Autofluorescence by noemi

O he fi~ldo~retina ha~?een revolutionized in recent years with the advent ofnew Imagmg modalities that allow clinicians to perform a better and moredetailed evaluation of their patients. Among these, fundus autofluorescence is

increasingly used and has proven ro be a very helpful rool assisting the clinician in thediagnosis of many retinal disorders. It has been used also to determine appropriatetreatment options and to evaluate outcomes following these therapies in selectedcases. Fundus autofluorescence provides different and complementary information tothat gathered by other imaging modalities such as fluorescein angiography and opti-cal coherence tomography and it has been a pleasant surprise to those in the field howmuch information can be gleaned from this conceptually simple modaliry. Indeedthere is much more to discover.

We conceived this book with the goal of providing the reader with a detailed up-to-date resume of current knowledge on fundus autofluorescence, including 0) basicaspects of the synrhesis and degradation of lipofuscin, the main fluorophore in theretina generating the fundus autofluorescence signal; (ii) information on techniquesavailable to image and quantify fundus autofluorescence and their basis; (iii) theanaromo-pathologic correlations of autofluorescence findings established in animalmodels; (iv) the normal disrribution of autofluorescence across the fundus; and (v)that associated with disease. The value of fundus autofluorescence as a diagnostic andprognostic tool is underlined in each of the clinical chapters, covering a variety of reti-nal diseases in which krowledge of rhe distribution of aurofluorescence has beengathered since this imaging technique became available. The value of fundus autoflu-orescence in the evaluation of patients with posterior segment disorders is discussedin each chapter in the context of other available imaging techniques, such as fluores-cein and indocyanine green angiography and optical coherence romography.Furthermore, fundus autofluorescence findings are reviewed in light of their value forunderstanding the parhogenesis of rhe conditions imaged. Lastly, individual clinicalchapters provide a comprehensive update on all aspects of each condition discussed.

Although this book has been written for ophthalmologisrs with special interest inretina, we believe it will also be of interest and value to other readers such as generalophthalmologists, ophthalmologists in rraining, basic researchers in the field of retina,and ophthalmic photographers and oprometrisrs wirh special interest in imagingtechniques, particularly as autofluorescence imaging as a diagnostic tool becomesmore widespread.It is our hope that this book will help clinicians to undersrand and inrerpret aut-

ofluorescence images of the retina and to recognize the different autofluorescence pat-terns observed in the retinal diseases discussed, helping subsequently, with rhe diag-nosis and management of patients with retinal diseases.

Noemi Lois and John V Forrester

CHAPTER

Lipofuscin: The "Wear andT "p.rear tgment

m any substances fluoresce spontaneously (aurofluoresce, i.e., emit light of au;LI particular wavelength) when illuminated by light of a different wavelength.

Pathologists ate well used to observing autofluorescence (AF) in certaincells, such as macrophages, when viewing specimens with a cobalt blue light filterin the microscope. More recently, ophthalmologists have become accustomed to

visualizing AF in images of the fundus. This book reviews the basic science andclinical knowledge regarding fundus AF in the human eye, a phenomenon that wasfirst noted when fluorescein angiographic imaging of the eye was introduced andhas become mote clearly evident with the development of scanning laser ophthal-moscopy. AF is thought to be due to lipofuscin present in cells of the retina, espe-cially retinal pigment epithelial cells. This first chapter serves as an introduction tothe biochemistry and mechanisms behind the accumulation of lipofuscin and otherautofluorescent storage material in tissues, with a particular focus on the centralnervous system (CNS).

Lipofuscin, commonly referred as the "wear and tear" pigment, is an aucofluo-rescent storage material that accumulates as a result of cell senescence. Lipofuscinhas also been termed lipopigment (LP), aurofluorescent storage material, yellow-brown material, and aging pigment. Although all cells accumulate lipofuscin, it isseen in the highest quantity in tissues or cells that are posrrnitoric, such as neurons,retina, and muscle. However, aging is not the only phenomenon associated with ac-cumulation of autofluorescent storage material. Aurofluorescent LPs have also beenshown to accumulate as a result of pathological conditions, in which case the auto-fluorescent storage material is known as ceroid. Such conditions include the pedi-atric neurodegenetative disorders called neuronal ceroid lipofuscinoses (NCLs).This distinction between ceroid (AF material that accumulates in disease) and lipo-fuscin (AF material that accumulates as a result of aging), however, is not generallyused in ophthalmology. Borh ceroid and lipofuscin have been shown to primarilyaccumulate in the lysosome; however, they have also been shown to accumulate invesicles, in the cytoplasm, and in the perikaryon of neurons. Table Ll Iisrs differentdisease states and aging pathologies reported to involve an accumulation of lipofus-cin/ceroid. The mechanisms and biochemistry of lipofuscin will be ptimarily dis-cussed in the context ofLP accumulation in all tissue types; in the case of ceroid, thefocus will be on the CNS.

All rypes of aurofluorescenr LPs were originally described as aurofluorescenrmaterial in postmortem tissue. All autofluorescent storage materials are not identi-cal. The LPs are often defined based on their fluorescent spectral properties, i.e.,the wavelength of light use to excite rhe intrinsic fluorophote (excitation) and thelight emitted as a result of this initial excitation (emission). Lipofuscin has a yellow-brown appearance and a wide range of spectral properties, with excitation wave-lengths of 320-460 nm and emission wavelengths of 460-630 nm (I). A derailed

3

Aging "Age-relatedmaculardegenerationNeuronal ceroid lipofusclnosesMucolipodosisIVMPS III SanfilippodiseaseAlzheimer diseaseRetinitis pigmentosaConeand cone-roddystrophyX-linked retinoschisisLeber congenital amaurosis:p~at~te~r~n~d:YS~t~ro~p~hY~ _

Best diseaseStargardt diseaseMaternal inherited diabetes anddeafChoroideremiaOsteopetrosis with neuronalstorag.diS!<:Adult-onset glycogen storagediseaseI\1l!Macular ABCA4diseaseWilson diseaseCrohn diseaseChoroidal tumors

4 SECTION I BASIC SCIENCE

TABLE 1.1f LPs (Lipofuscin/Ceroid)Occurrences 0

d "" f h ectral characreristics of retinal lipofuscin can be rouescnpnon 0 t e sp .Chapter 2" The most marked signal for rhe auroAuorescenr sr?ra~e malenal

d f I "let excitation Ceroid from NCLs has excrtauon and emun er ar-u rravro .wavelengths similar to those of lipofuscin, with an excitation ~ax.imum of4tand an emission maximum of 539 nm (2). The range of excrtanon and emwavelengths for both lipofuscin and ceroid reflects (he different methods ofurement used, different types of tissue studied, and different COrteCUOll

spectrum. It is important (0 use age-matched controls in studies of LPbiola,distinguish between the accumulation of lipofuscin (the result of aging)ana(the result of different diseases).

Other properties examined in characterizing LPs are histochemical srainin~·niques, such as differential dyes, lectin binding, and ulcrasrrucrure anal),". Tsections for both NCLs and aging brains have been shown to stain with penoi,Schiff, a carbohydrate stain, and Sudan black, a lipid stain (3-5). Lectin his,istry has been shown to distinguish between ceroid in Ls and lipofuscin:caging brain, with both LPs binding concanavalin A, but only ceroid in Cl.brasue binding to agglutinin (6).

Electron microscopy (EM) has also been carried Out on LPs to derermine dertrastructure. It has been shown by EM that lipofuscin-loaded ri ue has gran"",rnophilic deposits (GRODs) that appear as very densely packed vesicles.ilh

" granules filling the entire vesicle (3,5). For ceroid from Ls, EM h..GRODS identical to lipofuscin (5). However, two unique ultrasrrucrures at<

found only m the ceroid: fingerprint and curvilinear profiles. Curvilinear profi",vesicles that have an amorphous arrangement of lamellar structures formingCshaped forms. Fingetprim profiles also have lamellar structures in me vesidelever, the arrangement is in swirling circles, similar [0 me skin on a fingertip.,wca more dense arrangement of the lamellar structures." Although considerable work has been done to characterize lipofuscin/ceroid"

mICroscopIC level, the basic componems of lipofuscin and its pathological counrcremain to be determined V . di h19°1< . " . anous Stu res ave shown that lipofuscin is comp'". 0 t

fOl51~o lIpIds and 30% to 58% proteins (reviewed in Ref 7). Further",,"

non a IpOLUScmhas show th th li id . cIter I h h I" id n at e Ipl component consists of rriglyceride- "o , p asp a Ipl s, and free f t id Th .mixt f '. a ty act s. e protein component is a hererogtGure a proteins, with ani a id if

(A"PP) (8) Th. Y ne I enn led component: amyloid l3-pr=,pr"' . e calbohydrate co f ,.ture (9) I . mponent a Ipofuscin is also a heterogeneoLlS. ron, copper, alummum' al' "

, ZinC, C Clum, and magnesium account forapp·

CHAPTER 1 LIPOFUSCIN: THE "WEAR AND TEAR" PIGMENT

mately 2% of the lipofuscin components (10). Retinal specific lipofuscin has beenshown to have a very particular composition (11,12) (see Chapter 2). One study oflipofuscin isolated from retinal pigment epithelium (RPE) demonstrated that the com-ponents were highly damaged by peroxidation and glucoxidation (13). These compo-nents were specifically damaged at lysine and cysteine adducts, such as malondialdehyde(MDAs) and 4-hydroxynonenal (HNE). Moreover, the same study identified ad-vanced glycation end products (AGEs). This and other studies suggest that the com-ponents oflipofuscin are highly modified by oxidative stress.

For ceroid in the NCLs, except for the infantile variant, it has been shown thatthe primary protein component (50%) is the subunit c of mitochondrial ATPase. Inthe infantile variant ofNCL, the primary component is composed of sphingolipid ac-tivating proteins (saposins/SAPs) A and D (14-18). In other NCL variants, ceroidcontains SAPs A and D, but not to the extent of subunit c accumulation. Otheridentified components include AJ3PP, dolichol pyrophosphate-linked oligosac-charides, lipid-linked oligosaccharides, and metals (primarily iron) (16,19-21).Although all LP components vary in terms of the types of autofluorescent storagematerial and tissue, rhey all appear to be composed of undegraded or partially de-gradedproteins.

To date, very little is known about the fluorescent components (fluorophores) inmost LPs that generate the spectral properties of the autofluorescent storage material.Some have hypothesized that the fluorophore is a single compound. However, thefluorescent signal may also be generated after interactions between several differentnonfluorescent molecules. It has also been hypothesized that the fluorescence comesfrom lipid oxidation; however, some favor the hypothesis that modifications to thestoredproteins result in the fluorescence. It is certain that the ranges of spectral prop-erties reponed for aurofluorescenr material make identification of a single fluo-rophore challenging. Isolation oflipofuscin/ceroid has also proven problematic, withspectral properties decreasing or attenuating during the isolation process. In vitrostudieshave shown that reactions between carbonyls and amino compounds that pro-duce Schiff bases such as 1,4 dihydropyridine and 2-hydroxy-I,2-dihydropyrrol-3-ones demonstrate natural lipofuscin-like spectral properties (reviewed in Ref. 22).In retinal lipofuscin, it has been shown that the major blue absorbing fluorophoreis pyridinium bisretinoid (AlE) (23,24) (see Chapter 2). Ceroid is similar to lipo-fuscin outside of the retina and currently has no identified fluorophore.Furthermore, it cannot be excluded that each type of ceroid or lipofuscin mighthave a specific fluorophore, Ot multiple fluorophores with overlapping spectralproperties that result in the overall auto fluorescent signal. Numerous studies haveexamined the biochemical properties of LPs (Table 1.2), bur the larger question is,'Why does this aurofluorescenr storage material accumulate?

MECHANISMS OF LIPOFUSCIN ACCUMULATIONThree different mechanisms have been proposed for accumulation of lipofuscin/ceroid: lysosomal dysfunction, autophagy, and cellular stress. The accumulation ofLP at the lysosome implies an underlying lysosomal dysfunction that results in abuildup of lipofuscin/ceroid. Autophagy, the major degradationlreeycling pathway,could be altered, leading to LP / ceroid deposition. Cellular stresses in the form of ox-idative stress or starvation could have an impact on the cell physiology and result inlipofuscin/ ceroid accumulation.

5

SECTION I BASIC SCIENCE

TABLE 1.2. h . al Properties of LPsBlDc emic

Ceroid_________ ~L~iP~O~fU~s~c~in~~;__-----p;;;;;riiY~;;;;;;;--Primarilylysosome -.:.Primarilylysosome

lysosomal Dysfunction As a Cause forlipofuscin AccumulationA large percentage oflipofuscin/ceroid has been shown ro accumulate in asp'"organelle, the lysosome. There are two possible Fundamemal mechanisms th"conllresult in accumulation of LP in the lysosomes: substrate accumulation due to mu-tation/dysfunction in enzymes, or an imbalance in lysosomal homeostasis resultint;an altered lysosomal environment changing multiple enzyme activities/funcriorn.the case of the NCLs, three variant diseases are caused by mutations in l)"osomJenzymes: congenital, infantile, and late infantile N L. These diseases are cau.sed~mutations in the lysosomal enzymes cathepsin D, palmiroyl protein thioestens ,and tripeptyl protease, respectively (reviewed in Ref 25). However, nor all,fm,aurofluorescent storage material can be accounted for by such specific enzyman:defects, Undefined ways to affect lysosomal enzymes, such as alterations in lyseS!>mal homeostasis, are also likely. Other NCLs have defects in proteins that haven"yet been assigned a definitive function. Juvenile NCL QNCL) has a defecrinrltCLN3 pro rein, which resides in rhe Iysosomalllare endosomal membrane (reviewedin Ref. 26). In fibroblasts from patients with JNCL, a decrease in lysosomalpHwas observed. Most recently, the pH of lysosomes was shown to be regularedbi"'membrane channel protein (TRP·MLl). Defecrs in this protein lead to lysosonWaccumulatJon of lipid deposits (27), so there may be a general mechanism ,,~ullIn. the accumulation of lysosomal material related to the intralysosomal milieu.Thi>shifr III Intralysosomal conditions may not be optimal For enzymatic acri,i"Suboptimal lysosomal enzyme activity could potentially underlie the accumul"ionoilipofuscin/ ceroid (28) In f hi . . dmi .",.. .' support 0 t IS are studies demonstrating mat a (IliJuo

non of leupepnn, a general lysosome inhibitor, or chloroquine, an amine rhar raiS6lysosomal pH to an alkaline e '. . ff fus. lik f1 nVIrOnn1ent, In fats resulted in accumulation 0 lpoem- I e auto uorescenr storag ial i b . 32)1st<

al Ch 4) e rnaren III ram tissue and hepatocytes (29-so apter .

LocationSpectraiproperties(nm)ExcitationEmission

Storage componentsProteins

Lipids

CarbohydratesMetalsStaining characteristicsSudanblackBPeriodicSchiffbaseLectinUItrastructures

320-460460-630

320-460460-630

heterogeneous mixArJPPtriglycerides,cholesterol .phospholipids,tree fatty acidsheterogeneous mixFe, Cu,AI, Zn.Mn. Ca

subunit CmitochondriaATPasesapos insA & 0, AI3PPphosphorylateddolichols,p/loSjjtneutral lipidsdolichol-linkedoligosacchandespredominantlyFe

YesYesconcanavalin AGROOS

YesYesconcanavalin A. agglutininGROOs,fingerprint,culVilinear

Autophagy is the process of transporting macromolecules and organelles to the lyso-some for degradation. During this process, macromolecules and organelles areengulfed in double-membrane vesicles that are trafficked to the lysosome for the break-down of proteins and organelles as basic ptecutsors for recycling in the cell. Threeforms of autophagy have been described in the mammalian cell (22,38-40). The firstis macroautophagy, which is the process of engulfing large organelles such as mi-tochondria andlor large portions of rhe cytosol into a double-membrane vesicle calledthe autophagosome, followed by trafficking to the lysosome for degradation. This isthe most widely studied form of autophagy. In contrast, microautophagy describes amechanism in which small organelles or proteins are brought into the lysosome fordegradation by invagination of rhe lysosome membrane. In the third form of au-tophagy, a chaperone complex aids in identifying cyrosolic proteins for degradation bythe lysosome (chaperone-mediated autophagy [CMA]). For a protein to enter theCMA pathway, it must be targeted by a KFERQ sequence (41). There is a degree ofinterplay between each type of autophagy, such that when there are defects in CMA,macroautophagy will compensate for those defects (42).

It has been shown in numerous studies that as cells age, all forms of autophagyslow down (43-47). Thus, as the rate of autophagy decreases, there may be increasedformation of lipofuscin. To determine whether alterations in autophagy can result inlipofuscin accumulation, Stroikin et al. (48) treated dividing fibroblasts with 3-methyladenine (3-MA), which inhibits the first step of autophagy. In this process,calledsequestration, the material is engulfed in double-membrane vesicles to form theaurophagosome. Their study showed an accumulation of lipofuscin-like material inthe fibroblasts as a result of the 3-MA treatment. Autophagy has been shown to beupregulated in all cells when cells are starved to the point where they break down allnonessential or slightly damaged proteins into amino acids for cell survival (49). Inanother study (50), fibroblasts that were exposed to prolonged hyperoxia and nutri-ent starvation and had accumulation of lipofuscin showed decreased survival com-pared to cells without lipofuscin. It was hypothesized that lipofuscin in the lysosomeblocks the efficient breakdown of material to allow for further survival. However,there is as yet no consensus as to wh~ther all lysosomal enzymes have attenuated ac-tivity in aging cells. What seems to be clear is that lysosomes loaded wirh lipofuscinare nor able to degrade proteins, organelles, and macromolecules for recycling as effi-ciently, This could resulr in the further accumularion oflipofuscin.

Experimental models are available in which there appears to be an accumulationof lipofuscinl ceroid. In rhe Cathepsin 0-1- mouse model for congeniral N CL,there is a buildup of vesicles in the brain that are auto fluorescent, contain subunit cof mitochondrial ATPase, and are posirive for rhe autophagosome marker, LC3(51). This study suggesrs that rhe ceroid is accumulating in autophagosome, andthis accumulation of ceroid outside the lysosome itself is indicative of an increase inautophagy most likely caused by autophagic stress. A mouse model for juvenileNCL, Cln3",,7-8 mouse, has also shown defects in autophagy. Brains from

CHAPTER 1 LIPOFUSCIN: THE "WEAR ANO TEAR" PIGMENT 7

Srudies on the aging process have shown that as lipofuscin accumulares, lyso-somal cysteine proteinases undergo a decrease in enzymatic activity (33,34). Otherlaboratories have shown an increased activity of lysosomal enzymes such as 13-galactosidase and cathepsin Band 0 (35-37). In any case, there is enough evidenceto suggest that changes occurring at the lysosomal level could be a possible cause ofLPaccumulation.

Autophagy


Cln36ex7-8 mice were shown to have an increase in LC3-positive v~sicles, speci?-cally LC3-II, suggesting an upregulation in autophagy bur possible delays In

hazi . I t r ticn (52) Additional work earned out on cerebellaraurop aglC vestc e rna u a· .'precursor cells from this mouse showed that upon stImulatIon. of autophagy, the

I I b th lysosomal and autophagosome vesicles was decreasednorma over ap etween ecompared to wild-type mice. Taken together, these results illustrate a dystegulation

of autophagy in JNCL.Each field of research has supporced alterations in autophagy as a cellular event

that can occur both in disease and in the aging process. Studies using inhibitors spe-cific for autophagy have shown that autophagy is a possible mechanism for lipofus-cin/ceroid accumulation. However, alterations in autophagy, which are linked to thelysosome and other key organelles such as mitochondria, could be a result of a pri-mary insult to the lysosome, protein trafficking, mitochondna functIon: or amophagyitself. The interplay between each organelle and the overall cellular environment maybe synergistic in the formation of aurofluorcscent storage matenal.

Cellular StressAlterations in cell physiology could also contribute to or be a possible mechanism un-derlying lipofuscin! ceroid accumulation. Certain types of stress can cause cells to altertheir intracellular microenvironment, for instance, via reactive oxygen species (ROS).

The major organelle responsible for the production ofROS (45) is the mitochon-dria (reviewed in Refs. 53-55). ROS are formed by the normal functioning of theelectron transport chain (ECT) during ATP generation. Mitochondrial and cyroso-lie enzymes, including catalase, glutathione peroxidase, and superoxide disrnurase(SOD), help rid the cell of harmful free radicals. However, this system is nor per-fect, and some ROS escape the mitochondria with the potential to oxidize proteins,lipids and other macromolecules. In a normal functioning cell, oxidized or damagedproteins are transported to the lysosome for degradation. Senescent mitochondria,also known as giant mitochondria, have been shown to swell, lose their cristae and,occasionally, the inner membrane, resulting in little or no ATP production(56-58). All of these characteristics of aging mitochondria would lead to the releaseof increasing amounts of free radicals. In younger cells, as mitochondria age, theyare targeted by autophagy and degradation at the lysosome. Once the mitochondriareach the lysosome, they are degraded. As mitochondria are broken down, H202 isreleased into the lysosomal compartment. The lysosome itself is the perfect environ-ment for Fenton chemistry to take place to propagate the free radical damage due tothe available iron and cysteine pools, and the acidic pH. Lysosomes are responsiblefor recycling ferritin for the cell's continued use of the iron (59). Fenton chemistryrefers to the reaction in which ferrous iron converts the less reactive free radical ofH202 into the most reactive oxidative species of OH·. This basic conversion allowsfor an increasing amplification of oxidative damage because of an increase in the arrayof reactive reactions resulting from the increase in reactivity.

LP has components such as oxidized lipids and proteins; this, along with otherex?er~ments showing increased oxidative stress and slow degradation of mitochon-dna, IS known as the mitochondrial-lysosomal axis theory of aging (7). This theorysupports the hypothesis that oxidative damage may lead to the accumulation of auto-fluorescent storage material. However, it has yet to be determined whether this oxida-tion occurs inside the lysosome itself or elsewhere in the cell, as ROS are membrane-permeable and can diffuse throughout the cell. With slowing autophagy, there couldbe a buildup of gIant mitochondria III the eytosolleading to an increased ROS in thecytosol, as well as the hydroxyl radical emanating from the lysosome from the reac-

CHAPTER 1 LIPOFUSCIN: THE "WEAR AND TEAR" PIGMENT 9

rions discussed above. This increase in ROS could lead to damage to nucleic acids,intracellular proteins, extracellular proteins, organelles, lipids, and membranes.Additionally, if the ROS interacts with reactive metabolites, such as nitrogen species,this could lead to the formation of a further class of reactive species, namely, reactivenitrogen species (RNS) or peroxynitirites, which would further propagate damageacrossthe cell. Regardless of whether the ROS or RNS arise in the lysosome or in thecytosol,the effect would be across the whole cell.

ln the case of ceroid in NCLs, it has been shown that subunit c of mitochondrialATPaseis the most prevalent protein srored. A study in the English Setter model ofNCL showed partial uncoupling between electron transport and ATP production inlivermitochondria (60). In NCL fibroblasts, mitochondria were shown to have dam-agedfatty acid oxidation (61). Recently, our laboratory showed that in the brain ofthe JNCL knockout mouse model, progressive oxidarive damage occurs throughouttheCNS (62). Thar study demonstrated an increased amounr of protein oxidation, adecreasein glutathione, and an increase in one form of SOD in the CNS. A similarstudyexamining two orher variants ofNCLs found that the protein and gene expres-sionsof SOD were also upregulated (63). Therefore, ceroid accumulation in NCLsmaybe the result of oxidative damage.

Cell starvation alters a cell's ability to use nonessential processes. Under suchharshstress, cells will induce autophagy of nonessenrial proteins and organelles in thelysosome,with the subsequent breakdown of these macromolecules into small basicprecursorsthat can then be used by the cell to maintain its life. This additional stresscouldpropagate lipofuscin/ceroid formation in a cell wirh already present autophagydefectsor oxidative stress. Specifically, during starvation the buffering enzymes to re-duce ROS exposure are down-regulated, whereas more macromolecules are beingtransported to the very reactive oxidative environment of the lysosome, as discussedabove. Additionally, deficiencies in various nutrients, such as vitamin E, have beenshownto cause accumulation oflipofuscin and ceroid (64). This accumulation ofLPishyporhesized to occur due to the loss of the antioxidant property normally providedby vitamin E. All of these types of stress may playa role in the same oxidative parh-way that leads to lipofuscin/ceroid accumulation.

CONCLUDING REMARKS ONL1POPIGMENT ACCUMULATIONOverall, there is evidence to support all three basic mechanisms of LP accumulationin both aging and NCLs. Figure 1.1 summarizes each mechanism oflipofuscin/ceroidaccumulation. The lysosome plays a role in the deposition ofLP in each of these basicmechanisms.The mitochondria are also affected in all. It is possible that in some caseslipofuscin/ceroid accumulation occurs mainly due to a defect in a single pathway;however, it is likely that in most instances this accumulation results from deficienciesor subtle alterations in all three pathways. The varying contributions of these path-waysmay result in differences in the makeup of the autofluorescenr storage material,challenging the characterization of this material.

Implications of Lipofuscin/Ceroid in the CentralNervous SystemDoes lipofuscin accumulation res~lt in cell deat~, or is the .\P ~nternal.ized into thelysosornes as a protective mechanism to help aVOIdapoptoS1S. Lipofuscin accumula-

c - _pH- ~

.-2IUO< ZHaO • 0.

FIGURE 1.1. Mechanism of lipofuscin/ceroid accumulation. (A) lysosomal dysfunction. The top panel illustrates a normally functioning Iyso.some with vesicles delivering organelles, such as mitochondria, and proteins/macromolecules for degradation The lysosomal enzymesIbluehallcircles) are properly degrading the material and maintaining an acidic pH. This organelle is also transporting the basic precursors that havealreadybroken down to other locations in the cell to be recycled back into proteins. However, in aging or NCls. there are alterations to the lysosome.asseen in the bottom part of the panel lysosomal pH is raised to a more alkaline environment and/or lysosomal enzymes are not properly function.ing This results in the accumulation of lPs in the lysosome because the lysosome is incapable of transporting them out of the lysosome.181Autophagy defects. During the process of autophagy in a normal cell, an autophagosome surrounds the mitochondria and protein/macromoleculesand transports them to the lysosome for degradation by direct fusion to the lysosome or fusion to the late endosome (top pane II. Transport vesi-cles move lysosomal enzymes to the late endosome for maturation into a lysosome or additionally to the autophagosome itself. In aging cells orin cells with lP pathologies. the trafficking of these vesicles slows. Although there may not be a permanent block. if transport is delayed. proteinsand organelles targeted for degradation can become trapped in an environment that is not conducive to degradation, and alterations to the com-ponents may then occur. Proteins/organelles may undergo modifications before they finally reach the lysosome. rendering them nondegradable.Additionally. the lysosomal enzymesinvolved in the trafficking would also experience delays in trafficking. which would affect the cell. Therecouldbe a shortage of enzymes in Iysosomes, or the enzymes could be held in environments that lead to their modification, both resulting in loweredenzymatic activity. Any slowing down or blocking of autophagy can result in accumulation of lP. (e) Cellular stress (oxidation). The mitochondriaare the power plant of all mammalians. The ETCIshown in orange) produces ATP, but as a result of this process. harmful ROS(441such as OH.and H,O, form. However, the cell has enzymesto control the effects of ROS.In the mitochondria, the enzyme SOD converts OH. to H,O,. a memobrane·permeable form of ROSthat goes across the mitochondrial membranes. Once it reaches the cytosol, another set of enzymes detoxify H,O,.Catalaseand glutathione peroxidase convert 2H,O, into 2 H20. This system is not perfect, and some ROSreach other organelles; however, the ef-fect is minimal. Top panel: As cells age, they start to accumulate mitochondria that are less functional and have an increased production of ROS.This could be a result of slowed trafficking for degradation or of degradation itself. The end result is that the enzymes Ipurplel are not able to keepup with the increased concentration of RDS. With more H202 in the cellular environment, the lysosome is an attractive location for accumulationbecauseas a result of the high quantity of iron (Fe),the lysosome can use Fenton chemistry to convert H,O, to OH., a more reactive ROS.Oxidationalters lysosomal enzymes, proteins, lipids, and organelles, resulting in slowed degradation because of the crossnnks and modifications (pink zig'zaqs)and possible lowered enzymatic activity resulting in the accumulation of lP (a substrate for oxidative damage I, propagating the cycle. largegray ovals: Iysosomes; green squares and rectangles: proteins/macromolecules; blue half circles: lysosomal enzymes.

10


rion has been implicated in the pathology of neurodegenerative disease and in aging,particularly in association with neuronal loss and increasing reactivity and number ofglialcells (64). Storage of neuronal lipofuscin has also been associated with increasedoxidarivestress and decreased antioxidant defense (reviewed in Refs. 7 and 66). It hasalso been shown that neuronal lipofuscin destroys neuron structure and is correlatedwith the development of cytoskeletal defects (67-69). Overall, it has been assumedthat lipofuscin in the brain is detrimental, impacting neuronal and glial homeostasis,which affects the brain physiology (reviewed in Refs. 65 and 70). For the NeLs,ceroid, although clearly a hallmark of disease, has not been demonstrated to be theprimary disease defect. Accumulated ceroid, while damaging cells, may occur as theconsequence of a yet-to-be-determined initial disease insult. It has not been deter-mined whether neurons are trying to remove lipofuscin/ceroid as a harmful compo-nent. It has also not been demonstrated that LP accumulation arises from the needto rid the cell of certain components. However, it has been shown that LP triggersapoptotic cascades to initiate cell death (reviewed in Refs. 71-73).

Continued research needs to pursue the different mechanisms that underlie accu-mulation oflipofuscin/ceroid. Depending on the tissue type and autofluorescent stor-age material pathology involved, different mechanisms may be the culprits in thecomplex biology of this accumulation.

l. Dowson JH. The evaluation of autofluorescence emission spectra derived from neuronallipopigment.J Microsc 1982;128(Pt 3):261-270.

2. Dowson JH, Armstrong 0, Koppang N, et a1. Autofluorescence emission spectra of neuronallipopigmentin animal and human ceroidoses (ceroid-lipofuscinoses). Acta Neuroparhol (Bed) 1982;58:152-156.

3. Elleder MA. Histochemical and ultrastructural study of stored material in neuronal ceroid lipofuscinosis.Virchows Arch B Cell PathoI1978;28:167-178.

4. ElIeder M. Deposition of lipopigmem-a new feature of human splenic sinus endothelium (SSE).Ulcasrruccural and histochemical study. Virchows Arch A Parhol Anar Histoparhol 1990;416:423-428.

5. Elleder M. Primary extracellular ceroid type lipopigment. A histochemical and ultrastructural study.Histochem J 1991;23:247-258.

6. Wisniewski KE, Maslinska D. Lectin histochemistry in brains with juvenile form of neuronal ceroid-lipo-fuscinosis (Batten disease). Acta Neuropathol (Bed) 1990;80:274-279.

7. Brunk UT, Terman A. The mitochondrial-lysosomal axis theory of aging: accumulation of damaged mi-tochondria as a result of imperfect autOphagocycosis. Eur J Biochem 2002;269:1996-2002.

S. Bancher C, Grundke-Iqbal I, Iqbal K, er aI. Immunoreactivity of neuronal lipofuscin with monoclonal an-tibodies to the amyloid beta-protein. Neurobiol Aging 1989;10:125-132.

9. Brunk UT, Terman A. Lipofuscin: mechanisms of age-related accumulation and influence on cell func-rico. Free Radic BioI Med 2002;33:611-619~

10. Jolly RD. Batten disease (ceroid-lipofuscinosis): me enigma of subunit c of mitochondrial ATP synthase ac-cumulation. Neurochem Res 1995;20:1301-1304.

11. Katz ML, Drea CM, Eldred GE, et al. Influence of early photoreceptor degeneration on lipofuscin in theretinal pigment epithelium. Exp Eye Res 1986;43:561-573. . . .. .

12. Katz ML, Drea CM, Robison Jr WG. Relationship between dietary retinol and lipofuscin III the retinalpigment epithelium. Mech Ageing Dev 1986;35: 291-305. .

13. Schutt F, Bergmann M, Holz FG, er al. Proteins modified by malondialdehyde, 4-hydroxynonenal, or a~-vanced glycation end products in lipofuscin of human retinal pigment epithelium. Invest Ophrhalmol VIS

Sci 2003;44;3663-3668.14. Palmer ON, Martinus RD, Cooper SM, cr aI. Ovine ceroid lipofuscinosis. The major lipo~igmem protein

and the lipid-binding subunit of mirochondrial ATP synthase have the same Nl-lz-terrninal sequence. JBioi Chern 1989;264;5736-5740.

15. Palmer ON Barns G Husbands DR, cr al. Ceroid lipofuscinosis in sheep. II. The major component ofthe lipopigmenr in liver, kidney, pancreas, and brain is low molecular weight protein. J BiD] Chern

1986;26];1773-1777. . .. ..16. Palmer ON, Maninus RD, Barns G, et al. Ovine ceroid-lipofuscinosis. I: Lipopigment composltlon IS Ill-

dicative of a lysosomal proteinosis. Am J Med Genet Suppl 1988;5: 141-158. . . .17. Palmer ON. Fearnley IM, Walker JE, et al. Mitochondrial ATP synthase subunit c storage m the ceroid-

lipofuscinoses (Batten disease). Am J Med Genet 1992;42:561-567 .

REFERENCES

...

12 SECTiON I BASIC SCiENCE

M 1 Sf age of saposins A and 0 in infantile neuronal ceroid-lipo-18. TyynelaJ, Palmer DN, Baumann ,et a. orfuscinosis FEBS Lett 1993;330:8-12. . .. b - f - _. .. .. . KE M I- ki S et al. Beta-protein imrnuncreacuvny m rains 0 patients With

19 Kiraguchi T, Wlsmewski , as lOS > • aJ d - N - Le- - i- f . '. I structural and biochellllc emonsrranon. euroscr ttneuronal ceroid Ipa L1SCInOSlS. u rra1990-112-155-160- - aJ C.', . . d K' W Kida E Abnormal processing of carboxy-termin rragmenr of ben

20 WlsOiewsla, KE, Gor 00- raJcer, . .' ) J I h - M b D- 1993- - (b PPJ in neuronal ceroid-JipofusCIOOS1S (NCL cases. n ern eta IS ;16precursor protem eta

312-316_ - h - f I id B --W- - ki KE M I- k D Kitaguchi T et al. TopographIc ererogenerry 0 amy OJ -protein epJ.21. ISJ11eWS I , as U1S a ,I'\..J . , . de- _

- b - - h . f1 f neuronal ceroid lipofuscinoses suggestmg erective processmg of[Opes 10 rains WJt vanous orms 0 '

amyloid precursor protein. Acta Neuropath~l (Berl) 1990;~0:26-34: .22. Terman A, Brunk UT. Lipofuscin: mechanisms of formation and mcrease WIth age. APMIS 1998;106:

265-276_ _ "E"- d d idi - bi23. Sakai N, Decatur], Nakanishi K, et al. Ocular age pigment A2- . an unprece erne pyn nuurn IS-

retinoid. J Am Chern Soc 1996;118:1559-1560. . ..24. Reinboth J], Gautschi K, Munz K, er al. Lipofuscin in the retma: quantjltanve assa.y for an uExnprecEedemed

autofluorescent compound (pyridinium bis-rerinoid, A2-E) of ocu ar age pigment. p ye Res1997;65;639-643_ _

25. Ramirez-Montealegre D, Rothberg PG, Pearce DA. Another disorder finds IrS gene. Brain 2006;129(P,6);1353-1356_ _ _ _ _

26. Phillips SN, Benedict JW, Weimer JM, er a]. CLN3, the protein associated WIth Barten disease: structure,funcrion and localization.] Neurosci Res 2005;79:573-583.

27. Miedel MT, Rbaibi Y, Guerriero CJ, et al. Membrane traffic and turnover in TRP-MLl-deficienr cells:arevised model for mucoiipidosis rype IV pathogenesis. J Exp Med 2008;205: 1477~1490.

28. Holopainen ]M, Saarikoski], Kinnunen PK, er al. Elevated lysosomal pH in neuronal ceroid lipofusci-noses (NCLs). Eur] Biochem 2001;268:5851-5856.

29. Ivy GO, Schonler F, Wenzel], et a]. Inhibitors of lysosomal enzymes: accumulation of lipofuscin-likedense bodies in the brain. Science 1984;226:985-987.

30. Ivy GO, Kanai S, Ohm M, er al. Lipofuscin-like substances accumulate rapidly in brain, retina and inter-nal organs with cysteine protease inhibition. Adv Exp Med Bioi 1989;266:3 J--45; discussion 45--47.

31. Ivy GO, Roopsingh R, Kanai S, er al. Leupeptin causes an accumulation of lipofuscin-like substances andother signs of aging in kidneys of young rats: further evidence for the protease inhibitor model of aging.Ann NY Acad Sci 1996;786:12-23.

32. Ivy GO, Kanai S, Ohta M, et al. Leupeptin causes an accumulation of lipofuscin-like substances in livercells of young rats. Mech Ageing Dev 1991;57:213-231.

33. Amana T, Nakanishi H, Kondo T, et al. Age-related changes in cellular localization and enzymatic activ-ities of carhepsins B, Land D in the rat trigeminal ganglion neuron. Mech Ageing Dev 1995;83:133-14l.

34. Gracy RW, Chapman ML, Cini ]K, er aJ. Molecular basis of the accumulation of abnormal proteins inprogeria and aging fibroblasts. Basic Life Sci 1985;35:427--442.

35. Gerland LM, Peyrol 5, Lallemand C, er a1. Association of increased autophagic inclusions labeled for beta-galactosidase with fibroblastic aging. Exp Gerontol 2003;38:887-895.

36. Keppler D, Walter R, Perez C, et al. Increased expression of mamre cathepsin B in aging rat liver. CeJlTissue Res 2000;302: 181-188.

37. Nakanishi H, T ominaga K, Amano T, et aI. Age-related changes in activities and localizations of cathep-sins D, E, B, and L in the rat brain tissues. Exp Neurol 1994; 126: 119-128.

38. Sundelin S, Wihlmark U, Nilsson SE, et a1. Lipofuscin accumulation in cultured retinal pigrnem epithe-lial cells reduces their phagocytic capacity. Curl' Eye Res 1998; 17:851-857.

39. Yin D, Brunk U. Aurofluorescent ceroid/lipofuscin. Methods Mol Bioi 1998;108:217-227.40. Terman A, Brunk UT. Ceroid/Lipofuscin formation in cultured human fibroblasts: the role of oxidative

stress and lysosomal proteolysis. Mech Ageing Dev 1998;104:277-291.41. Dice]F. Peptide sequences that target cytosolic proteins for lysosomal proteolysis. Trends Biochem Sci

1990; 15;305-309_

42. Massey AC, Kaushik S, Sovak G, et al. Consequences of the selective blockage of chaperone-mediated au.tophagy. Proc Nat! Acad Sci USA 2006; I 03:5805-581 o.

43. Cuervo A.M, Dice JF. Age-related decline in chaperone-mediated autophagy. ] Bioi Chern 2000;275:31505-31513_

44. Dice JF. Altered degradation of proteins microinjected into senescent human fibroblasts. J Bioi Chern1982;257; 14624-14627_

45. Bergamini E, Del Rosa A, Fierabracci V, et al. A new method for the investigation of endocrine-regulatedautophagy and protein degradation in rat liver. Exp Mol Pathol 1993;59:13-26.

46. I?onati A, CavaIlini G, .Paradiso C, et aI. Age-related changes in the autophagic proteolysis of rar isolatedbver cells: effectS of antlaging dietary restrictions. J Geromol A Bioi Sci Med Sci 2001;56:B375-B383.

47. ~ona~i A, CavaIlini G, Paradiso C, et aI. Age-related changes in the regulation of autophagic proteolysisIII rat lsolated hepatocytes.] Geromol A Bioi Sci Med Sci 2001;56:B288-B293.

48. Stroikin Y, Dalen H, Loof S, et al. Inhibition of autophagy with 3-methyladenine results in impairedt~rnover oflysosomes and accumulation of lipofuscin-like material. Eur J Cell Bioi 2004;83: 583-590.

49. Fllln PF, Dice JF. Proteolytic and lipolytic responses to starvation. N urrition 2006;22:830-844.


50. Terman A, Dalen H, Brunk UT. Ceroidllipofuscin-loaded human fibroblasts show decreased survivaltime and diminished aurophagocyrosis during amino acid starvation. Exp Ceronro! 1999;34:943-957.

51. Shacka JJ, Klocke B], Young C, et al. Cathepsin D deficiency induces persistent neurodegenerarion in theabsence of Bax-dependenr apoprosis. J Neurosci 2007;27:2081-2090.

52. Cao Y, Espinola ]A, Fossale E, er aL Autophagy is disrupted in a knock-in mouse model of juvenile neu-ronal ceroid lipofuscinosis. J Bioi Chern 2006;281 :20483-20493.

53. Balaban RS, Nemoto S, Finkel T. Mitochondria, oxidants, and aging. Cell 2005;120:483--495.54. Lenaz G, Bovina C, D' Aurelio M, er al. Role of mitochondria in oxidative stress and aging. Ann NY Acad

Sci 2002;959;199-213.55. Richter C, Gogvadze V, Laffranchi R, et al. Oxidants in mitochondria: from physiology to diseases.

Biochim Biophys Acta 1995;1271:67-74.56. Beregi E, Regius 0, Hurd T, er at. Age-related changes in the skeletal muscle cells. Z Gerontol 1988;

21;83-86.57. Ermini M. Ageing changes in mammalian skeletal muscle: biochemical studies. Gerontology 1976;22:

301-316.58. Terman A, Dalen H, Eaton JW, er al. Mitochondrial recycling and aging of cardiac rnvocyres: the role of

aucophagocyrosis. Exp Cerontol 2003;38:863-876.59. Radisky DC, Kaplan J. Iron in cyrosolic ferritin can be recycled through lysosomal degradation in human

fibroblasts. Biochem J 1998;336(Pt 1):201-205.60. Siakoros AN, Blair PS, Savill )D, et al. Altered mitochondrial function in canine ceroid-Iipofuscinosis.

Neurochem Res 1998;23:983-989.61. Dawson G, Kilkus J, Siakotos AN, et aJ. Mitochondrial abnormalities in CLN2 and CLN3 forms of

Batten disease. Mol Chern NeuropathoI1996;29:227-235.62. Benedict lW, Sommers CA, Pearce DA. Progressive oxidative damage in the central nervous system of a

murine model for juvenile Batten disease. J Neurosci Res 2007;85:2882-2891.63. Heine C, Tyynela J, Cooper JD, et al. Enhanced expression of manganese-dependent superoxide dismu-

rase in human and sheep CLN6 tissues. Biochem J 2003;376(Pt 2):369-376.64. Pattoreni P, Berroni-Preddan C, Casoli T, er al. Morphometry of age pigment (lipofuscin) and of ceroid

pigment deposits associated with vitamin E deficiency. Arch Ceroncol Geriatr 2002;34:263-268.65. Riga 5, Riga 0, Schneider F, er al. Processing, lysis, and elimination of brain lipopigments in rejuvenation

therapies. Ann NY Acad Sci 2006;1067:383-387.66. Terman A, Brunk UT Aging as a catabolic malfunction. Tnt J Biochem Cell Bioi 2004;36:2365-2375.67. Fifkova E, Morales M. Aging and the neurocyroskeleron. Exp GerontoI1992;27:125-136.68. Vickers Je, Delacourre A, Morrison JH. Progressive transformation of the cytoskeleton associated with

normal aging and Alzheimer's disease. Brain Res 1992;594:273-278.69. Grune T, Jung T, Merker K, et al. Decreased proteolysis caused by protein aggregates, inclusion bodies,

plaques, lipofuscin, ceroid, and 'aggresomes' during oxidative stress, aging, and disease. ImJ Biochem CellBioi 2004;36;2519-2530.

70. Riga D, Riga 5, Halalau F, er al. Brain lipopigment accumulation in normal and pathological aging. AnnNYAcad Sci 2006;1067;158-163.

71. Kurz T, Terman A, Custafsson B, et aI. Lysosomes in iron metabolism, ageing and apoptosis. HisrochemCdl Bioi 2008;129;389-406.

72. Kurz T, Terman A, Custafsson B, et al. Lysosomes and oxidative stress in aging and apoptosis. BiochimBiophys Acta 2008; 1780: 1291-1303.

73. Kurz T, Terman A, Brunk UT Autophagy, ageing and apoptosis: the role of oxidative stress and lysoso-mal iron. Arch Biochem Biophys 2007;462:220-230.

. funcri aI . f the RPE with increasing age. In old age, the eyto-reducing the nctron capacuy 0 . . id b. 1 . d b h quatorial and perIpheral RPE IS cons: era ly lessplasmic vo ume occuple y tee

than that in the central retina. .' .I dditi th egl·onal distribution of lipofuscin across the fundus, there ISn a I(1OJ1 to e r . . . .

considerahle cell-to-cell heterogeneity in lipofuscin density rn any particular region of

h . (F· 2 1) (12) It is evident thar two adjacent cells (presumably WIth thet e renna Ig.. . . di . . allC • aI d belie load) can have hizh, mterme late, or rrurum evelssame runcnon an meta b

of lipofuscin. This would support the proposal by Burke and Hjelmeland (12) that. . f h RPE· natural result of normal mechanisms regulating gene ex-mosaicism 0 ( e IS a

pression during development and postnatal aging. This heterogeneity may help ex-plain the highly focal narure of the lesions associated WIth pathologies such as AMD.

Among the different tissue lipofuscins, RPE lipofuscin has a number of uniqueproperties that determine its photophysical properties an~ Its pO(en~lal for co~trIbu[-ing to RPE dysfunction, as will be discussed in more detail below: First, RPE lIpofus-cin is derived at least in part from pas and thus con tams significant quannnes ofretinoids (4,13), including the diretinal conjugate A2E (4). Second, RPE lipofuscin isa broadband absorber thar is constantly exposed ro visible light and is located in a high-oxygen environment, which makes it an ideal substrate for retinal phorodamage (I).


LIPOFUSCIN GENESIS AND COMPOSITIONAlthough RPE lipofuscin is believed to result from the incomplete degradation ofPOSand autophagy of spent or damaged intracellular organelles, the relative contributions ofthese pathways remain unknown. It is thought that prior oxidative damage to these sub-strates renders them undegradable by the lysosomal system, the universal intracellulardegradation machinery that contains over 40 hydrolytic enzymes. However, prior oxi-dation does not account for the fluorescent properties of lipofuscin, since Eldred andKatz (14,15) showed that the aurofluorescenr products of lipid peroxidation differ fromthose of lipofuscin. Although we have gained an understanding of the initial substratesinvolved in lipofuscin genesis, we still have a limited understanding of the compositionof the mature lipofuscin granules. Probably the first atrempt at analysis was made byEldred and Katz (13), who used Polch's extraction to separate lipofuscin into three frac-tions: chloroform-soluble, methanol-soluble, and insoluble (Fig. 2.3A). Thin-layerchromotography analysis of the chloroform fraction revealed 10 discrete fluorophores:two green-emitting, which comigrated with retinol and rerinyl palmitate; three emittingyellow/green; and one golden-yellow and four orange-red fluorophores, the mostprominent of which was subsequently identified as A2E (Fig. 2.3) (16). Further studiesby other researchers identified the fluorophores as retinoid derivatives (4,17). This wasfollowed by many studies on the role of AlE, the only readily synthetic component oflipofuscin, and demonstrations of its photo toxic and lysosomorrophic properties.However, the content of the bulk (>70%) oflipofuscin material, which is located at theinsoluble interface on Felch's extraction, has largely been ignored (18).

It has long been considered that lipofuscin consists of cross-linked oxidatively mod-ified lipids, proteins, and sugats, with the protein content tanging from 10% to 70% ofthe lipofuscin granule. The first proteorne for RPE lipofuscin was reported by Schuttand colleagues (19) in 2002. They identified over 65 abundant cellular proteins, in-cluding cytoskeleton proteins, proteins of transduction, enzymes of metabolism, pro-tei~s of the mitochondrial respiratory chain, ion channel proteins, and chaperones,which supported the origin of lipofuscin as a combination of photoreceptor phage-cytosrs and autophagy. Furthermore, analysis of the samples revealed that many of

• 60~:lQ.:>c 50•Q,.•0 40~•U,

30~C"T

~20=....

AJ!l 10•a:

_ .~,,_._._._.~._.~._.---1*

CHAPTER2 LIPOFUSCINOFTHE RETINAL PIGMENT EPITHELIUM

70~-- ~

+

/380 420 460 500 540

Wavelength (nm)

C

FIGURE2.3. Lipofuscin composition and photoreactivity.IAl Photograph of lipofuscin after Felch's ex-traction.Thetop layer is the methanol/water phase. the bottom layer is the chloroform-soluble phase,andinsolublematerial is located at the interface. (B) Comparison of the action spectra of initial rates ofphoto-inducedoxygen uptake in suspensions of OMPC liposomes containing both the chloroform-soluble(LI) and insoluble ['V) interfacial material plus a combination ("reconstituted lipofuscin". 01. The arith-meticaddition (+ I is shown of the rates measured separately for the chloroform-soluble and chloroform-insolublefractions. The inset shows the change in content of dry mass of the chloroform-soluble (01 andchloroform-insoluble I_I material extracted from lipofuscin granules as a function of donor age.Horizontalbars: SOof the donors of the pooled samples; vertical bars SOof the dry mass measurements(reproducedcourtesy of Investigative Ophthalmology and Visual Science from Ref 181.(C) Comparisonofthe thin-layer chromatography profile of the chloroform-soluble fraction of pooled lipofuscin granulesfromdonors of different ages: 11 50-59 years; 2) 60-69 years; 3) 70-79 years. The fractions show anumberof different fluorophores when excited by UV irradiation.

theseproteins were damaged by the aberrant covalent modifications of MDA, 4-HNEandAGEs (20). In a subsequent proteome analysis of RPE lipofuscin, Watburton er aI.(21) identified 41 cellular proteins; surprisingly, only 11% these proteins had beenidentifiedin the proteomic analysis of Schutt and colleagues (19). Although there arenwnerousexplanarions for this discrepancy, the likely problem lies in the purity of thestartingmaterial. Lipofuscin is generally isolated by repeated sucrose density gradientcentrifugarion (19,21,22); however, there is usually some residual contamination by

18 BASIC SCIENCESECTION I

cellut~r debris (22). While investigating this further, we demonstrated that highly pu-rified debris-free RPE lipofuscin granules contain little or no protein, bur do conrainsignificant amounts of modified material, such as carboxyethylpytrole adducrs (23). Inconclusion, although our understanding of the composmon of lipofuscin 10 the RPEremains limited, it does appear that protein makes only a very small contribution, jfany, and that oxidized and modified lipids from outer-segment and mitOchondrialmembranes may be major contributors.

Spectral Characteristics of RPE LipofuscinRPE lipofuscin granules exhibit a broad and wavelength-dependent absorption spec-trum with a decrease in absorption toward increasing wavelengths (22). The excitationand emission spectra of lipofuscin granules are shown in Figure 2.4, and a summary ofexcitation and emission peaks is provided in Table 2.1. RPE lipofuscin granules typi-cally exhibit four main regions of interest in the emission spectra when excited at 364nm: the main peak located at 600-610 nrn, a blue-green shoulder located at 470 nm, agreen-yellow shoulder at 550 nm, and, in the case oflipofuscin from inclividuals over 50years, a fat red shoulder at 680 nm. Similar emission peaks are observed at 500, 610,and 680 nm when lipofuscin is excited at 476 nrn, which is similar [Q the excitation em-ployed by in vivo autofluorescence (AF) systems (2:488 nm) (24,25). Excitation spec-

Excitation Emission10080

A

~ 100=ii BO

... 60-'iii 40l:.. 20-.!: 0

100806040200

Wavelength (nm)

8

FIGURE 2.4. EXCitation and emission spectra of intact lipofuscin granules isolated from donors of dif-ferent ages. EXCitation spectra were monitored with emission performed at 570 d erniwere monitored with ' . nm an emiSSIon spectra

excitauon performed at 364 nm. Lipofuscin granules were prepared from differentage groups: (AI 5-29 years' (B) 30-49 years' (e) >50 .

. , .' . years. EXCitation and emission spectra are ex-pressed In arbitrary units (AU.). (Modified from Ref. 22)

CHAPTER 2 19LIPOFUSCIN OF THE RETINAL PIGMENT EPITHELIUM

_1l:s;:um=m:':a:ry:::o:if';ith"e:-Mo.:a:;:in:-;:F;-:lu~o~r~es~c~e~n:-:c-e'":P;:-e-a"':'k-s'":f"'o-r:"'Li'"p-o"'fu-s-c':"in----.Gra?ul~s From Different Age Groups Observed in Figure 2.4.Excitation spectra were monitored with emission performed at570 nm, and emission spectra were monitored with excitationerformed at 364 nm.

Sampleage Iy,)Total PeakAmplitude'Excitation Inm) Emission Inm)

5 2930-4950--80

370,405,470370,405,470

370,470470,550,600470, 550, 600570, 600, 680

1001.71195

&Normalized to the 5-29 yr group

teawith emissions monitored at 610 nm or 570 om show a main excitation peak ar 470nrn, with two subsidiary peaks at 370 and 405 nm (22). Time-resolved fluorescence mi-croscopy has identified four decay components in lipofuscin with decay times of ap-proximately 0.21, 0,65, 1.75, and 6.6 ns (26). Wavelength-resolved decay analysisdemonstrated that the three main fluorescing components of lipofuscin do not stronglyinteract and thus appear to be excited directly. The spectral characteristics of melano-lipofuscin are intermediate between those of lipofuscin and melanin.

The fluorescence properties of lipofuscin granules exhibit age-related changes.The fluorescence intensity of lipofuscin granules increases by up to 40% with increas-ing age (22). Whether this is due ro increased deposition of material in the lipofuscinor a change in composition is unclear. It should also be emphasized that there is con-siderable heterogeneity in the fluorescence properties between individual lipofuscingranules, with variations in both emission maxima and spectral shape (27,28). Thus,lipofuscin spectra generally represent an average of all these spectral differences.

Aging, Lipofuscin, and the Potential forChronic Light DamageThe spectral characteristics of lipofuscin together with irs localization in a high-oxygenenvironment and diurnal exposure to visible light make it an ideal substrate for photo-chemical reactions. Our laboratory was the first to demonstrate that lipofuscin is able tophotogenerate reactive oxygen species (ROS) when we demonstrated that exposure ofisolated human RPE lipofuscin granules to light results in the generation of the super-oxide anion (29). That study also showed that the phorogeneration of ROS was de-pendent on both light intensity and wavelength, with the highest levels of superoxideanions being generated at higher visible light intensities and shorter (blue) wavelengthsof visible light. Subsequent studies demonstrated rhat isolated human lipofuscin gran-ulesare able ro phorogenerate significant quantities of singlet oxygen, hydroxyl radical,hydrogen peroxide, and the longer-lived lipid hydroperoxides (30-32). There is evi-dence that the overall phororeactiviry oflipofuscin increases with increasing age (18). Itappears that this is due to an increase in one or more of the components in the chloro-form-insoluble fraction obrained from lipofuscin granules (Fig. 2.3). Analyses of blue-light photoreactiviry in isolated RPE cells from donors of different ages demonstrated asignificant age-related increase in oxygen photo-uptake (an indicator of ROS forma-tion), primarily due ro the presence of lipofuscin within the cells (30). Not surprisingly,the lipofuscin-phorogenerated ROS are capable of eliciting oxidative modification tolipids, proteins, and nucleic acids. Lipofuscin granules exposed to light induce oxidation

Notwithstanding these observations, it is known that synthetic AlE applied diE II' I" hororoxic when exposed to blue light, It can causeOx'reedy to RP ce smell ture IS p .' I .

idati f 11 I constituents rhrough rhe phorogenerarion of SIng er oxygen (40); It1 anon 0 ce u ar VEGF (47) . . I

ul h ·011 of angiogenic factors such as ; It IS ysoso.upreg ares t e expreSSl . . d dhi d bl destabilize lysosomal membranes, resulting In re uce lYsosomalmanop 1C an a e to . . I di .

.. (48)' nd it can reduce mitochondrial capacity, ea 111gto apoprosl'enzyme acnvrry , a I .(37) In addition, photo-oxidation products of A2E can lead to comp ernenr actlva·tion '(49). Thus it is clear that A2E has the potential. to cause RPE cell damage, buth h . . hi s pre or posrincorporation Into lipofuscin granules remalOsw et er 10 VIVO t 15 occur -

ro be shown.


Association Between Lipofuscin andRetinal DegenerationElevated lipofuscin levels, as demonstrated by fundus"AF imaging, are associated ~vithavariety of retinal degenerations (24,43,50,51) (see Fundus Autofluorescence In meDiseased Eye" section). Although ail of these conditions demonstrate Increased AF andelevated numbers of lipofuscin granules within the RPE, rhe composition of the gran-ules may well vaty significantly. Furtherrnore, it is difficult to determine whether lipo-fuscin is a cause or consequence of these conditions. However, in the case of AMD,there is considerable circumstantial evidence linking lipofuscin with the etiology of thedisease. First, lipofuscin can be observed in early drusen, and these may serve as an im-munogenic stimulus for the localized activation of dendritic cells and deposition of in-flammarory factors (52). Confocal microscopy dernonsrrares two lipofuscin profiles inthe overlying RPE. One presents as a ring of hypofluorescence surrounding drusen, andthe other presents as a central area with decreased lipofuscin (Njoh and Boulton, un-published results). Furthermore, subretinal exocytosis oflipofuscin granules appears tobe a feature of eyes >60 years of age and results in degradation of rhese granules in meextracellular sub-RPE space (Hageman, Boulton, and Njoh, unpublished results).Second, the highesr density oflipofuscin is located in rhe central retina, where the den-sity of rod outer segments is at its highest and where there is a preferential loss of rodcells in AMD (53). Third, the accumulation of lipofuscin is reponed ro be greater inwhites than in blacks (54,55), and AMD is more prevalent among whires than amongblack persons (56). Fourth, chronic phototoxiciry, Iysosomotrophic damage, and occu-parion of cytoplasmic volume in RPE cells by lipofuscin and irs consriruenr AlErhroughout life will contribute to RPE dysfunction and pur added stress on a highlymerabolically active cell type (1,4). Fifth, longitudinal in vivo monitoring of fundus AFhas demonstrated that the presence of focaJ increased AF is a risk factor for progressionof AMD (57) (Fig. 2.6). Funhermore, zones of RPE exhibiring increased fundus AFsignal are prone to atrophy (24,43,58), and these areas exhibit a variable loss of retinalsensitivity that suggests an association between RPE dysfunction and excessive accWTIU-lation of lipofuscin (59). Sixth, synthetic AlE when phoroacrivared can induce the ex-pression of angiogenic factors in RPE cells (47). Seventh, phoro-oxidation producrs ofA2E can activate complement, and this may contribute to the chronic inflammation as-sociared with AMD (49). Eighth, lipofuscin levels have been observed ro be elevated inrodent models for AMD (60-62). Furthermore, in Srargardr disease there is an abnor-mality in the abcr gene that leads to a buildup of the lipofuscin fluorophore AlE (seealso Chapter 11G) (63). The predisposition for excessive lipofuscin accumulation inretinal degenerations such as A1vfD and, in particular, its elevation during me earlystages of disease progression suggest that lipofuscin makes a sisnificanr contribution to

the pathogenesis of AMD rather than simply being a consegu;nce of the disease.

CHAPTER 2 LIPOFUSCIN OFTHE RETINAL PIGMENT EPITHELIUM

FIGURE2.6. IA) Fundus AF image of a 74-year-old patient with visual acuity 20/40 taken in 2005 with a Heidelberg HRA IIscanninglaser ophthalmoscope. Several areas of RPEatrophy are noted, with a few small areas of Increased AFat the bordersof the larger lesions At this time the AF in the fovea was regular and the patient was told that progression of his condition wasImminent.(B) Two years later the patient was reexamined and visual acuity had decreased to 20/200; AF taken with the samenstrument showed extensive geographic atrophy, including the foveal area. (Image and caption provided by Dr. Erik Van Kuijk,Ophthalmologyand Visual Sciences, University of Texas Medical Branch, Galveston, Texas.)

Reducing the Impact of lipofuscin on Aging andRetinal DegenerationIf, as the evidence above suggests, lipofuscin is cytotoxic and contributes to the patho-genesis of retinal degenerations such as AMD and Srargardt disease, then a strategy [Q

remove or prevent further formation of lipofuscin would be efficacious, There are anumber of paradigms to support this: (a) lipofuscin can be voided into sub-Rl'E space,where it appears to be degraded; (b) lipofuscin deposition can be decreased wirh appro-priate modificarion of the lysosomal compartment (64); and (c) knockout or modifica-tion of the RPE specific protein RPE65, a crirical component of the retinoid cycle, pre-vents accumularion of the retinoid component of lipofuscin (65,66). The most popularapproach to date is to inhibit AlE formation by reducing retinoid turnover in the visualcycle(see also Chapters 1DC and 11 G).

AlE deposition is an endpoint that can be easily measured both in vitro in culturedRPE cells and in vivo using the abcr-/- mouse model for Stargardt disease (67).Isotretinoin (Accutane) and the small-molecule isoprenoid RPE65 antagonists reducethe formation and accumulation of A2E and other retinoid pigments in abcr- /- mice(68). An alternative approach is [Q reduce AlE accumuJarion by reducing serum vita-min A using N-(4-bydroxypbenyJ)retinamide (69). Although all of these approaches arepromising, [hey have two major limitations: (a) targeting the retinoid cycle is likely toaffectvisual functions, and there is clear evidence that many of these agents slow darkadaptation and/or reduce impaired electroretinogram (ERG) responses (67,70); and (b)although AlE fluorophores are reduced, there has been little morphometric analysis [Q

confirm a significant reduction in me lipofuscin granules within the RPE. Fenrerinide,currently in clinical erial (see also Chapter 10C), is the most recenrly developed agentaimed at reducing the retinoid cycle, but evidence suggests that this may actually pro-mote choroidal neovascularization (CNV) in compromised eyes (71).

23

B

AI· h rhar are aimed primarily at attacking the nonretinoid corn.rernauve approac es . . . .. . f li c . . lude reducing the accurnujatjon of lipofuscin by elmer a) pre-posItIon 0 IPOIUSClO me . .. h f . of inrralysosomal ROS which crosslink intralysosornal materialventmg t e ormatron '.. .

. . hIt (64). or b) preventing the formation of cross-linking adducn inuSlllg Iron c e a Drs , . .. .lipofuscin by increasing levels ofintralysosomal glutarhlOne-s-transferase and facilitatingthe degradarion of lipofuscin substrates. There IS no doubt that the next decade willseethe development of a variety of therapeutic approaches for prevenong or removmg lipo-

fuscin accumulation.


L1POFUSCIN- THE UNANSWERED QUESTIONSIn the last two decades there has been a major emphasis on RPE lipofuscin research,and studies have provided invaluable informacion on the composition, photobiologi-cal properties, and association of lipofuscin with retinal degenerations such as AMDand Stargardt disease. However, despite our increased knowledge, there are still manyunanswered questions. First, to what extent does autophagy contribute to the bio-genesis of RPE lipofuscin? Research would suggesr rhar the mitochondria, which areabundant in RPE cells, are compromised in aging and AMD (72) and rhus theirturnover may be much greater. Second, if prorein is only a minor componenr in lipo-fuscin, whar are the non retinoid constituents? Of interest, aurofluorescenr lipofuscin-like granules accumulate in long-term RPE cultures without exposure to POS orrerinoids (73,74). These granules are highly phororeacrive, Third, whar is rhe impor-tance of AlE within the lipofuscin granule? Is it inactive when sequestered into thelipofuscin granule, and can it be released? Fourth, what are the fluorophores in lipo-fuscin? Clearly AlE is one, but there must be many more since AlE does nor accountfor a large part of the excitation or emission spectra of lipofuscin granules either invivo or in vitro. Fifth, is lipofuscin immunogenic and does its release into me sub-RPE space lead to a localized inflammatory response? Sixth, whar is melanolipofus-cin? Given the heterogeneous composition of these granules, they cannot be simplyderived from the fusion of melanosomes and lipofuscin granules as suggested in sometextbooks. Seventh, can we degrade existing lipofuscin or prevenr irs accumulation bypharmacological intervention? Finally, is lipofuscin a cause or a consequence of reti-nal degenerations? It is hoped that research over the next decade will clarify thesei~sues and both i~prove our understanding of retinal pathogenesis and identify effec-nve pharmacological treatments for diseases in which lipofuscin is prevalent.

REFERENCESJ. Boulton M, Rozanowska M, Rozanowski B, et al. The phororeacriviry of ocular lipofuscin. PhorocbemPhorobiol Sci 2004;3:759-764.

2. Sohal RS. Age Pigments. Amsterdam: Elsevier/North-Holland Biomedical Press, 198 J.3. Seehafer S, Pearce P You say lipofu . id DC' .. .. scm, we say ceroro. ennIng aurofluorescenr storage material.Neurobiol Agmg 2006;27:576-588.

4. Sparrow JR, Boulton ME. RPE lipofuscin and its role in retinal pathology. Exp Eye Res 2005.80:595--606.5. Terman A, Cust~son B, Bru~lk UT. Aurophagy, organelles and ageing. J PathoI2007'211:134--143.6. Boulton M. Ageing of the retinal pigment epithelium. In: Osborne N, Chader G, eds. Pro ress in RetinalResearch. Oxford: Pergamon Press, 1991: 125-151. g

7. ~eeney L. Lipofuscin and m~lanin of human retinal pigment epithelium. Fluorescence, enz me ochem-ical, and ultrastruct~ral studies. Invest Ophchalmol Vis Sci 1978; 17:583-600. y cyr

8. Feeney-Burns L, HIlderbrand ES, Eldridge S. Aging human RPE· morph . I' f I. I d . ' . ometnc ana YSIS 0 macu ar,

equat~na, an peripheral cells. Invest Ophrhalmol Vis Sci 1984;25:195-2009. Delon FC Coger DC Dore CK A I d I . .f I' bi ' y . ge-re ate accurnu anon and spatial distribution of lipofuscin in RPEo .norma Sll jeers. Invest Ophrhalmol Vis Sci 2001 ;42: 1855-1866.

10. Wmg GL, Blanchard Gc, Weiter JJ The topogra h d lari .. . .the retinal pigment epith 1" I . 0 h hal PI ~ an . age re atio rtship of lipofuscin conceutranon In

e lum. nvest p t mo VIS SCI 1978; 17:601-607.

CHAPTER 2 25LIPOFUSCIN OF THE RETINAL PIGMENT EPITHELIUM

11. Marsball]. The ageing retina: physiology or pathology? Eye 1987; 1:282-295.12. Burke ]M, Hjelmeland LM. Mosaicism of the retinal pigment epithelium; seeing the small picture. Mol

Inrerv 2005;5:241-249.

13. Eldred GE, Katz ML. Fluorophores of the human retinal pigment epithelium: separation and spectralcharacterization. Exp Eye Res 1988;47:71-86.

14. Eldred GE, Katz. ML. The autofluorescent products of lipid peroxidation may nor be lipofuscin-like. FreeRadic Bioi Med 1989;n57-163.

15. Eldred CE, Katz ML. The lipid peroxidarion theory of lipofuscin ogene sis cannot yet be confirmed. FreeRadicBioi Med 1991;10:445-447.

16. Eldred CE, Lasky MR. Retinal age pigments generated by self-assembling lysosornorropic detergents.Nature 1993;361:724-726.

17. Lamb LE, Simon JD. AlE: a component of ocular lipofuscin. Photochem PhotobioI2004;79:127-136.18. Rozanowska M, Pawlak A, Rozanowski B, er a1.Age-related changes in the photoreacdviry of retinallipofus-

cin granules: role of chloroform-insoluble components. Invest Ophthalmol Vis Sci 2004;45:1052-1060.19. Schutt F, Ueberle B, Schnolzer M, et al. Proteome analysis of lipofuscin in human retinal pigment epithe-

lial cells. FEBS Lett 2002;528:217-221.20. Schutt F, Bergmann M, Holz FG, er a1. Proteins modified by malondialdehyde, -l-hydroxynonenal, or ad-

vanced glycation end products in lipofuscin of human retinal pigment epithelium. Invest Ophthalmol Vis Sci2003;44,3663-3668.

21. Warburton S, Southwick K, Hardman RM, ec al. Examining the proteins offunctional retinal lipofuscinusing proreomic analysis as a guide for understanding its origin. Mol Vis 2005; 11: 1122-1134.

22. Boulton M, Docchio F, Dayhaw-Barker P, et a]. Age-related changes in the morphology, absorption andfluorescence of melanosomes and lipofuscin granules of the retinal pigment epithelium. Vision Res1990;30,1291-1303.

23. Ng KP, Gugiu B, Renganathan K, et al. Retinal pigment epithelium lipofuscin proteomics. Mol CellProreomics 2008;7: 1397-1405.

24. Holz FG, Bindewald- Wittich A, Fleckenstein M, er a1. Progression of geographic atrophy and impact of fun-dus autofluorescence patterns in age-related macular degeneration. Am J OphthalmoI2007;143:463-472.

25. von Ruckmann A, Fitzke FW, Gregor Z]. Fundus autofluorescence in patients with macular holes im-aged with a laser scanning ophthalmoscope. Br J OphthalmolI998;82:346-351.

26. Docchio F. Boulton M, Cubeddu R, et a1.Age-related changes in the fluorescence of melanin and lipofuscingranules of the retinal pigment epithelium: a time-resolved fluorescence spectroscopy study. PhorochemPhorobioI1991;54;247-253.

27. Clancy KMR, Krogmeier JR, Pawlak A, et al. Aromic force microscopy and near-field scanning optical mi-croscopy measurements of single human retinal lipofuscin granules. J Phys Chern B 2000; 104: 12098-121 0 1.

28. Haralampus-Crynaviski NM, er al. Probing rhe spatial dependence of the emission spectrum of singlehuman retinal lipofuscin granules using near-field scanning optical microscopy. Photochem Photobiol2001;7064--368.

29. Bouhon M, Lamb LE, Simon JD, et al. Lipofuscin is a photoinducible free radical generator. J PhorochemPhcrobiol B 1993; 19:201-204.

30. Rozancwska M, jarvis-Evans J. Koryrowski W, er al. Blue light-induced reactivity of retinal age pigment.In vitro generation of oxygen-reactive species. J Bioi Chern 1995;270: 18825-18830.

31. Rozanowska M, Wessels J, Boulton M, et a]. Blue light-induced singlet oxygen generation by retinallipofus-cin in non-polar media. Free Radic Bioi Med 1998;24:1107-1112.

32. Sparrow JR, Nakanishi K, Parish CA. The lipofuscin fluorophore AlE mediates blue light-induced dam-age to rerinal pigmented epithelial cells. Invest Ophthalrnol Vis Sci 2000;41: 1981-1989.

33. Wassell J, Davies S, Bardsley W, cr a]. The photoreacdviry of the retinal age pigment lipofuscin. J BiolChem 1999;274,23828-23832.

34. Davies S, Elliott MH, Floor E, er a]. Photocytoroxiciry oflipofuscin in human retinal pigment epithelialcells. Free Redic Bioi Med 2001 ;31 :256-265.

35. Godley BF, Shamsi FA, Liang FQ er al. Blue light induces mitochondrial DNA damage and free radicalproduction in epithelial cells. J Bioi Chern 2005;280:21061-2\ 066.

36. Shamsi FA, Boulton M. Inhibition ofRPE lysosomal and antioxidant activity by the age pigment lipo-fuscin. Invest Ophrhalmol Vis Sci 2001;42:3041-3046.

37. Sparrow JR, Cai B. Blue light-induced apoptosis of A2E-containing RPE: involvement of caspase-B andprotection by BcI-2. Invest Ophrhalmol Vis Sci 2001;42:1356-1362.

38. Haralampus-Crynaviski NM, Lamb LE, Clancy CM, et al. Spectroscopic and morphological studies ofhuman retinallipofuscin granules. Proc Nat! Acad Sci USA 2003; 100:3! 79-3 .184. . .

39. Gaillard ER, Atherton SJ, Eldred G, er al. PhotOphysical studies on human retinal lipofuscin. Photochem

Pborobiol 1995;6L448-453.40. Ben-Shabat S, hagaki Y, Jockusch S, et a1. Formation of a nonaoxirane ftom AlE, a lipofuscin fluorophore

related to macular degeneration, and evidence of singlet oxygen involvement. Angew Chern Int Ed Engl

2002;4 L814--817.41. Margrain TH, Boulton M, Marshall J, et aI. Do blue light filters confer protection against age-related macular

degeneration? Prog Retin Eye Res 2004;23:523~531. .42. Bindewald A, Bird AC, Dandekar 5S, et al. Classification of fundus autofluorescence patterns 111 early age-

related macular disease. Invest Ophrhalmol Vis Sci 2005;46:3309-3314.

43,


fl in patients with age-related macular degeneratiDliCRt al Fundus auto uorescence ILoisN,OwensSL, oeo ,e . I hal 12002'133:341-349.and high risk of visual loss. Am ] Op 1[ rno al I 'J . 1 and one-step preparation of AlE and iso-AlE

h i M Nakanishi K, et . so anor . USA 19989- 1460 '44 Parish CA, Has rrn ora , . . h I" Proc Nat! Acad SCI ;); 9-14613.• C J ri a] pigment eprr e Him. . h hfluorophores n-om inman r~ In . d rna e to the retina: role of rhodopsin c romop ore revisited.

4 R k M Sarna T Llght-mduced a g5. ozanows a ". 330Phorochem Phor.obiol 2005;8]: 1305-1 a" The role of AlE in prevention or enbancernenr of lightclam.

46. Roberts JE, KuJo~ICla~ BM, Hu .DN~ eCeel"Is.Phocochem Phorobiol 2002;75: 184-190.age in human retinal plgmem epithelia! c j induction of HNE- MDA- and AGE-adduct5, RAGE' J yP a] Mechanisms ror r ie In •

47. Zhouj, Cal B, ang .: er ru. . h r 1 Us Exp Eye Res 2005;80:567-580.and VEGF in retina] plgmenr ep" e hib ~~ . fl mal degradarive funcrions in RPE cells by a retinoidHolz FG, Schutt F, Kopirz ], er a1.l11 11 lit/ani 0VYSsO~Oj999.40'737_743.

f1' fi . I Ophrha rno IS CI ,.component a IpO ISClO. nvesr I ti arion by phorooxidation products of AlE, a lipofUScin

49. Zhou,], Jang YP,.Ki~ SR,.et al. COI~PI 'I~lenr ;c~~<Narl Acad Sci USA 2006; 103:16182-16187.. t of the rednal plgmenr eprt re rum. r C []

consntuen . W B· d AC Distribution of pigment epithelium aura uorescence in retin~50. von Ruckmann A, Firzke F , If. . G f. A ch C1in Exp OphthaJmoI1999'237'1_9

disease state recorded in vivo and its change over time. rae es .r. '. . ., k FW B' d AC In vivo fundus autofluorescence 10 macular dystrophies. Arch51. von Ruckmann A, Pita e ,Ir .

OphthaJmoI1997;115:609-615. b ~, h 'GS M II' RF MIl composition of drusen as related to su structural p enorype.1\10152. Hageman , U InS . 0 ecu ar

Vi, 1999;5,28, , I did 'I53. Curcio CA, Medeiros NE, Millican CL. Photoreceptor loss In age-re are macu ar egeneratJon. nV~1

OphthalmoJ Vis Sci 1996;37:1236-1249. . ... . .' JJ D 1 'FC W' GL·· 1 Retinal pigment epIthelIal iJpofuscln and melanm and chorOIdal54. Welter , eon , Ing , .... a.

melanin in human eyes. Invest Ophthalmol Vis Sci 198.6;27:1.45-152.. . .. ,D CK W G Eh 'D t a1 CeU loss in the agmg retina. RelationshIp to hpofuscln accwnulamlnorey , u , enstelll ,e .and macular degeneration. Invest OphthaIrnol Vis Sci 1989;30: 1691-1699. . .

56. Friedman OS, O'Colmain BJ, Munoz B, et a1. Prevalence of age-related macular degenerauon In theUnited States. Arch OphthalmoI2004;122:564-572. . . .

57. Sol bach U, Keilhauer C, KJlabben H, et aI. Imaging of retinal autofluorescence m patIents wuh age-related macular deoeneration. Retina] 997; 17:385-389.

58. Holz FG, Bellman C, Sraudt 5, et al. Fundus autoRlIorescence and development of geographic auophy inage-related macular degeneration. Invest Ophthalmol Vis Sci 2001;42: 1051-1 056.

59. Schmitz- Valckenberg S, Biilrmann 5, Dreyhaupt], et aJ. Fundus autofluorescence and fundu~ perimetryin the junctional zone of geographic atrophy in patients with age~relared macular degeneratlon. InVe51Ophthalmol Vis Sci 2004;45:4470--4476. . .

60. Ambati J, Anand A, Fernandez 5, et al. An animal model of age-related macular degeneratIon In s(:nescemCcl-2- or Ccr-2-deficienr mice. Nat Med 2003;9:1390-1397.

61. Malek G, Johnson LV, Mace BE, et al. Apolipoprotein E allele-dependent pathogenesis: a mode! forage-related retinal degeneration. Proc Narl Acad Sci USA 2005; 102: 1 1900-1 1905.

62. justilien V, Pang Jj, Renganathan K, et al. SOD2 knockdown mouse mode! of early AMD. InvestOphmalmoI Vis Sci 2007;48:4407-4420.

63. Mata NL, Weng j, Travis GH. Biosynthesis of a major lipofuscin f1uorophore in mice and humans withABCR-mediated retinal and macular degeneration. Proc Nad Acad Sci USA 2000;97:7154-7159.

64. Kurz T. Can lipofuscin accumulation be prevented? Rejuvenation Res 2008;11:441-443.65. Katz ML, Redmond TM. Effect of Rpe65 knockout on accumu.lation of lipofuscin f1uorophores in the

retinal pigment epithelium. Invest Ophthalmol Vis Sci 2001;42:3023-3030.66. Kim SR, Fishkin N, Kong J, et al. Rpe65 Leu450Met variant is associated with reduced levels of (he reno

nal pigment epithelium lipofuscin fluorophores A2E and iso-AlE. Proc Nat! Acad Sci USA 2004;101:1] 668-11672.

67. Radu RA, Mara NL, Nusinowitz S, et aI. Treatment with isotretinoin inhibits lipofuscin accumulation inamouse model of recessive Stargardr's macular degeneration. Proc Nat! Acad Sci USA 2003;100:4742-4747.

68. Maiti P, Kong j, Kim SR, et aI. Small molecule RPE65 amagonists limit [he visual cycle and preventlipofuscin formation. Biochemistry 2006;45:852-860.

69. Radu RA, Han Y, Bui TV, et al. Reductions in serum vitamin A arrest accumulation of toxic retinal flu-orophores: a potential therapy for treatment of lipofuscin-based retinal diseases. rnvest Ophrhalmol VisSci 2005;46,4393-4401.

70. Travis GH, Golczak M, Moise AR, et al. Diseases caused by defects in the visual cycle: retinoids as poten.tia! therapeutic agents. Annu Rev Pharmacol ToxicoI2007;47:469_512.

71. Sreekumar :G,. Zh~lI j,. Sohn j, et al. N-(4-hydroxyphenyl) retinamide augments laser-induced choroidaJneovasculaflzatlon lJl mIce. Invest Ophthalmol Vis Sci 2008;49: 1210-1220.

72. Xu H, Zhong X, Lin H, et a1. Endogenous ROS production increases and mitochondrial redox functiondecreases in me RPE as a function of age and stage of AMD. Invest Ophthalmol Vis Sci 2008;49:ARVOE-absrracr.

73. Wassell j, E}lis 5, Burke j, et aJ. Fluorescence properties of aurofluorescenr granules geneldted by culturedhuman RPE cells. Invest OphthaImol Vis Sci 1998;39:1487-1492.

74. Burke JM, SkumaLi. CM. Autofluorescent inclusions in long-term postconfluenr culuues of retinal pio-ment epithelium. Invest OphthaJmol Vis Sci 1998;39: 1478-1486. :0

48.

55,

CHAPTER

Lipofuscin: The Origin of theAutofluorescence Signal

O he retinal pigment epirhelium (RPE) digests the tips of the outer segment ofrhe photoreceptors, which are phagocytosed on a dally basis. The digestionprocess of rhese materials, which contain polysaturated fatty acids and byprod-

tiersof the visual cycle, is overall very efficient, but a tiny fraction is chemically incom-patible for degradarion and accumulares in lysosomes of the RPE. This undigestedfraction is called lipofuscin (LF). Above rhe age of 70, as much as 20% to 33% of thefree cytoplasmic space of rhe RPE cell may be occupied by LF granules and melano-lipofuscin granules, a compound producr ofLF and melanin (1).

LF is a pigment that exhibits a characteristic autofluorescence (AF) when excited inultraviolet (UV) or blue light. Fluorescence m..icroscopy of the RPE using UV or blueexcitation light reveals disrinctly brighr orange-red or golden-yellow granules. This flu-orescencealso makes it possible to visualize and measure LF noninvasively, since the ab-sorptionspectrum of LF is monotonic (2) w.ithout any distinct spectral signature, and itwould be very challenging to measure it by fundus reflectometry.

SPECTROFLUOROMETRY AND IMAGING OFFUNDUS AUTOFLUORESCENCEFundusAF was initially demonstrated in vivo by vitreous fluorophotometry, whichrevealed in preinjection scans a distinct "retinal" peak whose magnitude increasedwith age, as would be expected from LF fluorescence (3,4). These observarions led tothe development of a fundus spectrofluorometer (5,6) designed to measure the exci-rationand emission spectra of the AF from small retinal areas (2 degrees diameter) ofthe fundus and to allow absolute measurements of the fluorescence. The device incor-porated an image-intensifier-diode-array as detector, beam separation in the pupil,and confocal detection to minimize the contribution of AF from the crystalline lens.This device was used initially ro demonstrate that fundus AF essentially exhibits thecharacteristics of RPE LF (7).

Fundus AF is about 2 orders of magnitude less intense rhan the backgroundfluorescence of a sodium fluorescein angiogram at the brightest part of the dye tran-sit. As a result, AF imaging systems with high sensitivity and/or image averagingcapabiliries (8) are requ..ired ro record rhe fundus AF with acceprable signal/noise ra-tios and safe retinal exposures. Confocal scanning laser ophthalmoscopy (SLO) (9)optimally addressed these requirements, and the firsr clinical AF imaging system wasintroduced ar Moorfields in London with an excitation wavelength of 488 nm(l0,11). Subsequent developments using confocal SLO sysrems, nonconfocal cam-etas, and differenr excitarion wavelengrhs have further broadened the field of AFimaging (l2~16).


TRIBUTION OF AUTOflUORESCENCERETINAL DISal dark area caused by rhe lighr .h"l/pif h fu d s show a cenrr

AF images are nUl . r rhar srrongly ab orbs \va\'e!engtluE 1 . nd maw ar plgmen .borh RP me arun a . al i b lieved ro originare po tenor ro rhe phore"(17) The AP sIgn IS e I

rhan 540 nm. bsornri b the visual pigmencs when rhe1aIl"~,. . ffi r d by a sorprron y UI

tors because ir IS a ec e h fi I d'fference specrrum, log (dark-ad.pred IffuIly bleached. Indeed), r1 e PIara ovalesarh~ specrrum of rhodopsin (7, J 8). SimiJam1 (1' h d redAF c ear y reve . "og 19 r-a ap .. ' .: rhe choriocapillaris because 1(5 emlSS,on!p<n:AF is believed ro ongmare anrerror ~o hemo lobin (in contrast, rhe reflemnctj,does nor reveal absorption bands a xy b gds 540 nd 575 nm) (19). The:«trum of the fundus shows these ahsorprion an at . a

. '. h the norian rhar rhe bulk of the AF emanares from',servauons are consrsrenr WIt '. I' Th AF.. al . . fAP images rem force this cone usron. e /ItRPE. Clinic mterpreranons 0 . . " .

. . f 11 thickness macular hole IS similar m magmrude and '/"p<.510n spectrum In a u - . al .h d b r 7 degrees eccentricity because me neurosensory renn IJ!sl..[ at measure at a ou

b . h h 1 d the RPE is roughly intacr (7,10). onversely, RPEarroptare a sent 111 teo e an . dh AF " ecrrum from geographic arrophy m age-relare maculard<.causes t e emlSSlOn sp .

generacion (AMD) ro be significandy reduced and d.srorred (20).

SPECTRAL PROPERTIES OF FUNDUSAUTOFLUORESCENCE

The emission specrra of fundus AF are broad (Fig. 3.1) and maximal in rhe 600-6401lIDregion of rhe speen'um, shirring slightly roward longer wavelengths wirh increasingexcirarion wave1engrh (21). This "red shirr" OCCurs for excirarions ar rhe long ",,,.lengrh end of rhe absorprion specrrum for some fluorophores in viscous or polarenvi.ronments (22). The red shirr in large parr explains me variabiliry of emission 'permmaxima reporred in rhe lirerarure; wirh excirarion wavelengms ranging from 364nm(lamp used in microscopy) ro 580 nm (16), rhe emission peak will vary berween 590and 660 nm.

The excirarion speerra have maxima berween 490 and 530 nm. In COntrasr to thespecrra measured 7 degrees remporal ra rhe fovea, me foveal specrra are arrenuared anddisrarred. This is due in parr ro rhe lower amounr ofLF in rhe foveal area (23), and torhe absorprion of rhe excirarion lighr by macular pigment. The difference spectrum,log(foveal AF) - log(parafoveal AF), has me shape of rhe absorprion speerrum of mac-ular pigment (Fig. 3.1, insee), and rhis serves as me basis for me rwo-wavelength Afmerhod ro measure macular pigment (24). Nore mar rhe difference spectrum still de-creases wirh increasing wavelength ar 550-570 nm (arrow), where macular pigmemdoes nor absorb subseantially. This is because of rhe absorprion by RPE melanin, whichdecreases wirh wavelengrh and is larger ar rhe fovea rhan ar rhe parafo

vea(RPE melanin

is locared on rhe apical side of rhe RPE cell) (23,25).

INFLUENCE OF CRYSTALLINELENS ABSORPTION

Excieaeion and emission spectra of fundus AF are affected by absorprion and scareeringin the lens and rhe orher ocular media. For example, an exciraeion ar 488 run is atrenu-ared, on average, by factors of 2.2 and 3.3 for ages 20 and 70 years, respecrively (26).The AF emlSSlon 15 also attenuared, but ro a lesser exrenr, because ir occurs ar longer

CHAPTER 3 LIPOFUSCIN THE ORIGIN OF THE AUTOFLUORESCENCE SIGNAL 29

200

JOT'50

-,100Ec

::'50<:

---~------:430 '

5 0l-t-~--+------r~----r __ -"-I-

~200 ~: 0 , f'f~ i .20u,

~1~· o~Fovea

~

100

50

o400 500 600 700

Wavelength (nm)600

FIGURE3,1. Fluorescence spectra measured in a 52-year-old normal subject at 7 degrees temporal to thefovea(7degrees TIand at the fovea (sampling area: 2 degrees in diameter) (21). Continuous lines: emissionspectra[excitation wavelength as indicated); line with filled triangle: excitation spectrum for emission at620 rrn The excitation spectrum is not measured. but is constructed from the fluorescence at 620 nm andplaned against the excitation wavelength [dotted linel, The foveal spectra are lower than at 7 degreestemporal.particularly for excitation wavelengths shorter than 490 nrn, where absorption of the excitationlightby macular pigment is revealed, Inset Difference spectrum, [loglfoveal AF) - loglparafoveal AFIJ.convertedto macular pigment density using the absorption coefficients of the macular pigment

wavelengths (factors of 1.4 and 1.5, respectively, when averaged over emissions from500 to 700 rim). If absolure AF levels are needed, rhen corrections to account for theexcitation and emission losses are required. These losses can be estimated for each sub-ject by psychophysical (27) or retlectometric (28,29) methods. Alternatively, algo-rithms that predict the average light loss at a given wavelength and age can be used inpopulation studies (26,30). Corrections obtained with such algorithms (26) result ina marked increase in AF. The corrected excitation spectrum is shifted toward shorterwavelengths (where lens absorption is highest), but the shape of the emission spec-trum does not change substantially (Fig. 3.2, top, specna A and B).

AGEAND FUNDUS AUTOFLUORESCENCEFundus AF has been shown to increase significantly with age (4,7,11,21,31), confirm-ing data obtained in ex vivo studies (32,33). The rate and time course of this increasevaries among different studies, probably as a result of differences in population selec-tion, techniques, correction merhod for media absorption, and other factors (31). Theincrease of fundus AF with age for a population of normal subjects (Fig. 3.3) shows alarge but constant variability (the coefficient of variation per decade of age is about

I '


T'Temporal Fovea

345678 345678Age Decade

FIGURE 3.4. Age-related changes In the shape of the emission spectrum (Exc.: 470 nml as character·ized by the maximum wavelength (top) and the ratio of the AF at 560 nm to that at 660 nm Ibottom). Opencircles normal subjects; filled circles: patients with ARM. The spectra were corrected to account forab·sorption of the ocular media. A decrease in the peak wavelength and an increase in the ratio resultsintlle"blue shift" of the emission, spectra.

more substantial in older subjects. Furthermore. the blue shift is also more pro-nounced for spectra measured in age-related maculopathy (ARM) patients (Fig. 3.4).

Although an increase in the amount of melanolipofuscin could contribute ro thisshift (see above), the magnitude of the changes and the additional blue shift ar me siteof drusen suggest that BMD are responsible for the shift. This confirms ex vivo meas-urements (48) indicating that the emission from Bruch's membrane and BMD is simi-larly blue-shifted compared to the emission spectrum ofLF. Spectral analyses predictedthe BMD emission spectrum to be somewhat narrower than the LF emission and to

have a maximum around 540 nm (49). Spectral deconvolution of speerra from ARMpatients showed that the AF (Exc.: 488 nm) from BMD contributed on average 20%(range 5%-45%) of the total AF detected between 500 and 700 nm. This AF couldplaya role in the varying appearance of drusen in AF images in ARM patients.

The unique AF of drusen is also demonstrated by ratio imaging using pairs of AFimages obtained with different excitation and emission bands, as shown in Pig. 3.5.One image (Fig. 3.5A) was obtained using bands similar to those used in routine AFimaging (Exc.: 470 nm), whereas the other (Fig. 3.5B) was obtained with excitationat 550 nm. The ratio image (Fig. 3.5D) thus reflects the difference in the local exci-tation spectra for both the macular pigment and the drusen. The latter are seen in theratio image as uniform areas corresponding to their location in the monochromaticimage (Fig. 3.5C). Differences in the excitation spectra of drusen and their neighbor-ing area are thought to be tesponsible for this effecr.

OTHER SECONDARY FLUOROPHORESThe blue shift at rhe fovea in the youngest subjects (Fig. 3.4; decades 3 and 4 com-pared with decade 5) must be related to other secondary fluorophores. The AF from

LIPOFUSCIN, THE ORIGIN OF THE AUTOFLUORESCENCE SIGNALCHAPTER 3

FIGURE3.5. LF images (A,B) and a 550-nm monochromatic image Ie) from a 55-year-old ARM patient.(AI Obtainedwith excitation at 470 nm and a detection >510 nrn, and IB) obtained with an excitation at550nmand a detection >500 nm. (0) The ratio (after alignmentl of A to B. All images were recordedwith a nonconfocal camera (15). Note that the attenuation by the macular pigment seen in A is not seenin B. allowing examination of the foveal RPE.

LFat rhe fovea in young subjects is very low. Therefore, secondary fluorophores emit-ting in the 500-540 nm spectral range may become relatively more important. Thisfluorescence is believed to originate anterior to the macular pigment and could befrom the macular pigment itself (50), from flavins in the inner retina, from collagen(51) and hyaluronic acid (52) in the vitreous, and/or from lens LF reflected by thefundus.

AF from microglia in the inner retina of normal mice has been observed by AFimaging in some (53) but not all (54,55) studies. When seen (53), the AF emission ofmicroglia appears to have LF-like characteristics (see also Chapter 4), although deter-mination of the excitation spectrum would be required for comparison with RPE LF.However, AF of microglia has not been observed in normal human eyes by AF imag-ing. If the density and AF characteristics of microglia in normal human and micewere similar, then their observation in mice but not in humans might be explainedby a lower AF emanating from the mice RPE than from the human RPE, causingtheir contrast to be high enough to be observed. This could result from a loweramount of RPE LF in mice and from a higher density of RPE melanin (larger atten-uarion of the AF of RPE LF in AF images).

Finally, AF contributions from the choroid and sclera have been estimated in AMDpatients (69-81 years old) with geographic atrophy to be 6% to 15% of the total fun-dus AF for dark- and light-pigmented subjects, respectively (45). This suggesrs thatchoroidal melanin absorption modulates the AF from other choroidal fluorophores.The contribution of choroidal AF to total fundus AF will be less in younger subjects be-causeof the higher concentrations and absorption by melanin in the RPE and choroid(23,25,56). Emission spectra measured in geographic atrophy give little information

...

34 BASIC SCIENCESECTION I

about the origin of the weak AF, because of distortion byblood absorption andnoise~the spectral data. Endogenous fJuorophores m the choroid may include stromalelastinand collagen, porphyrins, melanocytic melanofuscrn, and macrophages (COntainingblue-emitting LF) (51,57,58).

TOPOGRAPHICAL DISTRIBUTIONEx vivo studies have shown that the amount of RPE LF is low at the fovea.increasesgradually to a maximum about 10 degrees from the fovea, and then decreasestOlVardthe periphery (23). Topographical distriburion of fundus AF, measured wirhanexci-tation at 550 nm (and thus not affected by macular pigment absorption), confirmsthese findings; the AF is maximal at an eccentricity of7 to 13 degrees. where theMis abour 1.7 times higher than at the fovea (31,59). The central minimum isarusedprincipally by LF in the RPE, since the higher concentration of RPE melaninat thefovea has been estimated ro be only 1.15 higher than at the parafovea (31). Fundu,AF is not symmetrically distributed around the fovea (Fig. 3.6); the AF is maximalatabour 12 degrees temporally and superiorly, lower inferiorly and nasally whereitismaximal at about 7-8 degrees from the fovea.

The AF distribution toughly matches the distribution of rod phororecepron (60),which is not surprising given that LF derives from precursors within phagocytosedpho-toreceptor outer segments. However, the distribution does not reflect the narrowdistri-bution of cones ar the fovea (60). The rare of LF formation from cones may be slower,as suggested by the observation in rhesus monkeys that the number of fovealcone-derived-phagosomes in the RPE was only one-third that of extrafoveal rod-derived-phagosomes (61). Other effects, such as high absorprion by RPE melanin ar rhefovea,cone/rod differences in the fractional content of indigestible materials in rhe phago-somes, and sparial1ydependent protection by the macular pigment conld alsoplayarolein reducing foveal LF formarion.

FIGU~E 3.6. Topographical distribution of fundus AF based on measurements along the horizontal(nasal temporal. N-T) and vertical (superror-Inferror: S-I) meridians (data from about 40 subjects)(31).TheAF IS expressed relative to that at the fovea.

CHAPTER 3 LIPOFUSCIN: THE ORIGIN OF THE AUTOFLUORESCENCE SIGNAL 35

SUMMARYThe fundus LF signal derives principally from RPE LF, as evidenced by its spatial dis-rribution, its spectral characteristics, and its relation to age. This signal is affected by ab-sorption by the ocular media, macular pigment, retinal blood vessels and capillaries, andRPEmelanin. Secondary sources ofLF in humans are BMD and vitreous fluorescence.TheAF of drusen may provide specific information about the type or content of drusen.

A case can be made for using excitation wavelengths longer than the 488 nrn cur-rently used in most fundus AF imaging systems. Indeed, there would be less ocularmedia absorption and scattering, and errors associated with correction for media ab-sorption would be decreased. Macular pigment absorption would be reduced, if noteliminared, allowing for clear visualization of the foveal RPE (Fig. 3.5B). One wouldalso expect an overall increase in the signal intensity and signal/noise ratio. Finally,longer wavelengths are less susceptible to photochemical damage to the retina.

REFERENCESI. Feeney-Burns L, Hilderbrand ES, Eldridge S. Aging human RPE: morphometric analysis of macular,equatorial, a.nd peripheral cells. Invest Ophchalmol Vis Sci 1984;25:195-200.

2. Boulton MD, Docchio F, Dayhaw-Barker P, er al. Age-related changes in the morphology, absorption andfluorescence of melanosomes and lipofuscin granules of the retinal pigment epithelium. Vision Res1990;30,1291-1303.

3. Delori FC, Bursell S-E, Yoshida A, et a]. Vitreous Iluorophotomerry in diabetics: study of artifacrual con-tributions. Graefe's Arch Clin Exp Ophthalmol 1985;222:215-218.

4. Kitagawa K, Nishida S, Ogura Y. In vivo quantification of autofluorescence in human retinal pigment ep-ithelium. Ophehalmologica 1989; 199: 116-121.

5. Delori Fe. Spectrophotometer for noninvasive measurement of intrinsic fluorescence and reflectance ofthe ocular fundus. Appl Optics 1994;33:7439-7452.

6. Delori Fe. Pluorophoromerer for noninvasive measurement ofRPE lipofuscin. Noninvasive assessmentof the visual system. OSA Technical Digest 1992; I: 164-167.

7. Delori FC, Dorey CK, Staurenghi G, et al. In vivo fluorescence of the ocular fundus exhibits retinal pig-ment epithelium lipofuscin characteristics. Invest Ophrhalmol Vis Sci 1995;36:718-729.

8. Wade AR, Fitzke FW. A fast, robust pattern recognition system for low light level image registration andits application [Q retinal imaging. Opt Express 1998;3:190-197.

9. Webb RH, Hughes GW, Deiori Fe. Confocal scanning laser ophrhalrnoscope. Appl Opt 1987;26:1492-1449.

10. von Ruckmann A, Pirzke FW, Bird Ae. Distribution of fundus autofluorescence with a scanning laser oph-thalmoscope. Br J Ophthalmol1995;1 ]9:543-562.

II. von Rnckrnann A, Pirzke FW, Bird Ae. Fundus aurotluorescence in age-related macular disease imagedwith a laser scanning ophthalmoscope. Invest Ophchalmol Vis Sci 1997;38:478--486.

12. Solbach U, Keilhauer C, Knabben H, er al. Imaging of retinal autofluorescence in patients with age-relatedmacular degeneration. Retina 1997; 17:385-389.

13. Holz FG, Bellmann C, Margaritidis M, et al. Patterns of increased in vivo fundus autofluorescence in thejunctional zone of geographic atrophy of the retinal pigment epithelium associated with age-related mac-ular degeneration. Graefes Arch Clin Exp Ophrhalrnol 1999;237: 145-152.

14. Gray DC, Merigan W, WolfingJI, ec al.In vivo fluorescence imaging of primate retinal ganglion cells andretinal pigment epithelial cells. Opt Express 2006;14:7144-7158... .. .

15. Delori FC, Fleckner MR, Goger DG, et al. Autofluorescence disrriburion associated with drusen In age-related macular degeneration. Invest Ophthalmol Vis Sci 2000;41 :496-504.

16. Spaide RF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology 2003;110:

392-399.17. Bone RA, Landrum jT, Cains A. Optical density spectra of the rnacular pigment in vivo and in vitro.

Vision Res 1992;32:105-110.lB. Prieto PM, McLellan jS, Burns SA. Investigating the light absorption in a single pass through the phorore-

cepror layer by means of the lipofuscin fluorescence. Vision Res 2005;45: 1957-1965.19. Delori FC, Pflibsen KP. Spectral reflectance of the human ocular fundus. Appl Op~ 19~9;28: 1O~1-1077.20. Arend GA, Weiter JJ, Coger DG, et al. In-vivo fundus-fluoreszenz-messungen bel patlenten nut alterab-

hangiger makulardegeneration. Ophrhalmologie 1995;92:647-653. .21. Delori FC Keilhauer C Sraurenghi G, et al. Origin of fundus autofluorescence. In: Holz FG, Schmitz-

Vall b 's S id RF' Bird AC eds Atlas of Fundus Autofluorescence Imaging. Berlin, Heidelberg,aucen erg , pal e ,I ,.

New York: Springer, 2007:17-29.

--

CHAPTER

•Lipofuscin and Autofluorescence InExperimental Animal Models

O undus autofluorescence (AF) imaging is increasingly being used in oph-thalmic practice to aid in the diagnosis and monitoring of pa[Jents with a va-riety of retinal disorders. Thus, it is particularly important to underst~d the

molecular and cellular sources of fundus AF. Only with such an understanding candoctors interpret and judge the significance of fundus AF changes in various patho-logical conditions. However, it is very difficult to obtain such information, ~romhuman studies; in this context experimental animal studies may have great additionalvalue. Improvements in in vivo fundus AF imaging techniques and the developmentof various animal models of disease enable us to carry out direct clinicopathologicalcorrelations as well as dissect the detailed mechanisms of lipofuscin formation.

THE VALUE OF EXPERIMENTAL ANIMAL MODELSExperimental animal studies are valuable because they allow us to investigate the clin-icopathological correlations of fundus AF and the pathological changes related to var-ious AF patterns. Such information is critical for clinicians to correctly interpret AFimages. Although a variety of animal models representing different types of humanretinal diseases have been developed over the years, only a few AF clinicopathologicalstudies have been published (1,2), including a study from the author's Iaborarory innormal aged mice. Another important use of experimental animal studies is to investi-gate the mechanisms oflipofuscin formation. Studies in this field have identified a fewimportant factors involved in in vivo lipofuscin formation, including the RPE65 geneand vitamin A. In addition, animal studies are of great value for testing the effective-ness of various treatments, particularly those targeting the formation of lipofuscin.

IMAGING TECHNIQUES

In Vivo ImagingCurrently, confocal scanning laser ophthalmoscopy (cSLO) is the most widely used~echmque to Image fundus AF 1ll VIVO. Because all commercial cSLOs are designed to

Image the ~umanfundus, slight modifications, such as the use of special lenses, areneeded to Image rat or mouse retina. Various groups have used this system to imagethe rodent fundus. In the author's laboratory, a custom-built cSLO is used in experi-mental animal studies. To be able to image the mouse fundus, we use a +25D lensplaced 1 em in front of the ~ouse cornea to further focus the laser beam. In addition,~ mouse hard contact lens IS also used to prevent tear evaporation and keep the eye-lids open. We are able to image a 43 X 32 degrees field of the mouse fundus (3). This

38

CHAPTER 4 LIPOFUSCIN ANO AUTOFLUORESCENCEIN EXPERIMENTAL ANIMAL MODELS 39

in vivo imaging technique is noninvasive. and therefore changes in the fundus can befollowed up at different stages of disease in the same animal.

ExVivo ImagingA5with human samples, lipofuscin in animal tissues can be extracted and examined invirro by various rechniques. Traditionally, tissue AF has been imaged in situ with theuseof various fluorescence microscopes (4,5). Fluorescence microscopy is still a simple,straightforward, and useful technique to detect lipofuscin. Currently, however, moreadvanced and sophisticated technologies are used to quantify lipofuscin content and ex-amine the detailed fluorescence emission specrra of lipofuscin. Bui and colleagues (6)used a fluorescence microplare reader (Tecan Satire II Fluorescence Microplate Reader,Tecan US, Research Triangle Park, NC) to examine the fluorescence emission spectra ofRPE-choroid and retina explants. In their system, fresh samples were placed side up in amodified 384-well microplate. Emission spectra were measured following excitation at420 nm using the top read mode (6). This technique allows one to measure the emis-sion spectra as well as the relative fluorescent units of RPE-choroid and retina explants.In the author's laboratory, the lambda mode of a confocal scanning laser microscope(LSM510 META, Carl Zeiss) is used to image fluorescence emission spectra of mouseRPE-choroid and retina explants (1). This technique has several advantages: (i) samplescanbe excited with different lasers and the characteristics of the emission spectra excitedby the different lasers can be studied; (ii) it detects the emission spectra of individualcellsor regions but not the whole tissue, allowing comparisons of the emission spectra ofdifferentcells or different regions; and (iii) the fluorescent intensity of different tissues ordifferent regions can be quanritatively recorded and analyzed. The lambda mode of theLSM510 META confocal microscope is designed to separate the fluorochromes withwidely overlapping emission spectra or fluorochromes that are excited by the same laser.It iswidely used to eliminate sample background AF. With the META scanning system,the emission fingerprints of each fluorochrome can be collected and a spectral databasecan be built (Fig. 4.1). It is therefore an ideal means of studying the fluorescence spectra

A

FIGURE 4.1. AF emission spectra of a mouse RPEcell The RPEflatmount was prepared from a 2-year-oldC57Bl/6 mouse and examined with the use of a LSM510 META confocal microscope (lambda model).Thesamplewas excited with a 458 nm HeNe laser and the emission signals were collected from 460 to697nm.(AI Emission spectra database of the RPEcell. (8) AF intensity at different emission wavelengths

extractedfrom the database in A.

...


oflipofuscin. With the META scanning mode, emission signals are detected by a poly-chromatic 32-channel detector, which allows the fast acqursmon of lambda stacks.Samples can be excited with lasers at 448 nm, 477 nm, 488 nm, 514 nm, 543 nrn, and633 nm separately. For each excitation, the emission spectral range for lambda stackacquisition can be controlled as required, and the upper limit IS 790 nm. For ~Ilambda acquisitions, the pinhole needs to be set ro I aIry urut and the scannIng speedset to 6. To be able to compare the emission fingerprints acquired from differentcells/tissues, the configuration, including the detector gain, amplifier offset, and ampu-fier gain, should not be changed during the process of acquisition. An example ofmouse RPE lipofuscin emission spectra excited by 458 nm is shown in Figure 4.1.

CLINICOPATHOLOGICAL CORRELATION OF FUNDUSAUTOFLUORESCENCE IN EXPERIMENTAL ANIMALSAlthough various pathological srudies in retinalfRPE lipofuscin have been carriedout in postmortem human eyes over the past decade, direct clinicopathological Cor-

relation is lacking. To date, the clinicopathological correlation of fundus AF find-ings in the majoriry of retinal diseases is still largely unknown.

Fundus AF in Normal MouseTo be able to understand fundus AlF changes in various animal models of diseases of theposterior segment, the characteristics of fundus AF in healthy control animals need to

be established. We have investigated the fundus AlF in normal C57BLl6 mice of differ-em ages. The overall fundus AlF increased with age in normal C57BLl6 mice by cSLO(I). Increasingly strong AlF signals were observed with age in the neuroretina and sub-retinalfRPE layer (Fig. 4.2A) . Unlike fundus AlF detected in normal human subjects,fundus AF appeared as discrete foci distributed throughout the retina (Fig. 4.2A).Most of the AF signals in the neuroretina were distributed around retinal vessels (I),Immunohistological studies of retinal or RPE-choroidal flarmounrs indicated thatmost of the AlF' signals derived from Iba-I + perivascular and subretinal microglia(FIg. 4.2B,C), and some derived from RPE cells (Fig. 4.3) (I).

FIGURE 4.2. FundusAFin a normal aged mouse. (AI FundusAF in an 18-month-oldC57Bl/6mouse.181 Confocalmicroscopyof RPE flatmount of the same mouse reveals manyautofluorescentcells locatedon the surface of the RPE celis. (e) The autofluorescent cells on the RPE cell surface stained positiveforthe microgliamarkerIba-1.

CHAPTER 4 LIPOFUSCIN ANO AUTOFLUORESCENCE IN EXPERIMENTAL ANIMAL MODELS 41

FundusAF in Mouse Models of DiseaseCCR21CCL2 Knockout MiceMice deficient in either monocyte chemoattractant prorein-I (CeL2, also known asMCP-I) or its cognate C-C chemokine reccpror-Z (CCR2) develop the cardinal fun-dus features observed in patients with age-related maculopathy (ARM) or age-relatedmacular degenerarion (AMD) (7). These include drusen, geographic atrophy, andchoroidal neovascularization. Both strains of mice develop drusen-like changes at theageof9 months. By 16 months, geographic atrophy can be observed. Choroidal neo-vascularization (CNV) can be detected in one-fourth of CCL2-1- and CCR2-I-mice ar rhe age of 18 months. Such mice are therefore considered good models forARM/AMD. Clinically, we observed drusen-like and geographic atrophy-like changesin our CCR2 and CCL2 knockout mice (purchased from the Jackson Laboratory),but no subretinal neovascularization. Fundus AF imaging using cSLO indicated manymore aurofluorescenr foci present in CCL2- or CCR2-deficient mice compared to age-matched wild-type control mice (Fig. 4.4).

Ex vivo confocal imaging of retinal flatmounts from this animal model revealedmany perivascular and sub retinal Iba-I + microglia with aurofluorescent lipofuscin,similar to those seen in normal aged mice, but in significantly higher numbers.

FIGURE 4.3. lipofuscin in subretinal microglia and RPEcells. (AI Subretinal microglia with lipofuscin.Inset Emissionspectra of subretina! microglial lipofuscin excited with 488 nm (BI RPEcell with lipofuscinInset: Emissionspectra of RPEcell lipofuscin excited with 488 nm.

The number of sub retinal microglia and the amount of intracellular lipofuscinincreased with age (I). Transmission electron microscopy (TEM) revealed many lipu-fuscin granules in the cytoplasm of subretinal microglia (I). Emission spectra de-tected by the META scanning mode of the confocal microscope indicated that thespectra of the lipofuscin in sub retinal microglia and the RPE cells were the same (Fig.4.3), suggesting that they may have the same chemical components.

CFH Knockout MiceGenetic studies have revealed a strong association between a broad range of AMDphenorypes and variants in the gene encoding complement factor H (CFH). CFH isthe major plasma prorein that exclusively regulates the alternative pathway of comple-ment activation. CFH protein has been detected in drusen, the pathogenic hallmarkof AMD. CFH-deficient mice exhibit significantly reduced rod responses on elec-troretinography compared to age-matched controls (2). Fundus AF of these mice by

...


FIGURE 4.4. Mouse fundus AF. (A) Fundus AF of a 1B-monlh-old C57Bl/6 mouse.IB) FundusAF01anlS-month-old CCL2knockout mouse.

SLO revealed an increase in aurofluorescenr subrerinal deposits, similar to those ob-served in CCR2- and CCL2-deficienr mice. More recently, using ex vivo confocal mi-croscopy of retinal and RPE/ choroidal flarmounts, we showed that the majority ofthe aurofluorescenr signals were from subretinal microglia (Fig. 4.5).

Fundus Autofluorescence in ExperimentalAutoimmune UveoretinitisClinical studies have shown altered AF in various choriorerinal inflammatory diseases(see Chapter 12). Inflammation may alter RPE and photoreceptor cell function andsubsequently affect the production of aurofluorescenr molecules. Furthermore, inflarn-matory cells, in particular macrophages, may contain aurofluorescenr materials as a re-sult of active phagocytosis of oxidized substances at the site of inflammation. It istherefore important to understand the correlation between AF patterns and parholog-ieal changes in chorioretinal inflammation.

We have carried. out clinicopathological studies in a mouse model of experimentalautoimmune uveoretiniris (£AU). £AU was induced by immunizing C57BLl6 mice

FIGURE 4.5. AF cells in the subretinal space. The retinal flalmounl was prepared from a 12~month-oldCFHknockout mouse and imaged by confocal microscope (LSM51 0 META). The image was taken fromthephotoreceptor side of Ihe retinal flatmount. Many AF cells (arrows) are seen within the outer segmentofthe photoreceptor cells.

CHAPTER 4 LIPOFUSCIN AND AUTOFLUORESCENCEIN EXPERIMENTAL ANIMAL MODELS

with human interphotorecepror retinoid-binding protein (IRBP) peptide 1-20. Inthis EAU model, retinal inflammation starts at day 10-12 postimmunization (p.i.),peaks ar day 20-25 p.i., and resolves at about day 30 (8). Fundus AF was imaged atdisease initiation (day 12 p.i.), peak (day 25 p.i.), and resolution (day 35 p.i.) stages,respecrively. At day 12 p.i., when inflammation was very mild and localized in theoptic disc or venule fragments, fundus AF was largely normal; neither an increasednor a decreased AF signal was observed in inflamed retina areas compared to nonin-flamed ones. As inflammation progressed, increased AF signals began to appear inareas of inflammation at peak stage EAU. AF signals appeared as discrete small dotsof increased AF surrounding inflamed vessels (Fig. 4.6).

At the late stages of inflammation (i.e., day 35 p.i., disease-resolution stage), a mas-sivedestruction of photoreceptor cells was observed. Highly increased AF signals wereseen throughout the entire retina (Fig. 4.7).

Ex vivo confocal microscopy of retinal RPE-choroidal tlarmounrs revealed manyautofluorescent cells at sites where inflammation was present, particularly in the areaaround the inflamed vessel segments. Immunofluorescent studies indicated that theseautofluorescent cells were F4/80+ macrophages.

FIGURE 4.6. Fundus AF in peak stage EAU. (AI Fundus image taken from a day 22 p.i. EAU mouse,shOWingmassive inflammation in the retinal venules (arrow) (81 FundusAF image showing increased AFsignalsin the optic disc and inflamed venules [arrowl.

FIGURE 4.7, FundusAF in late-stage EAU. Images were taken from a day 35 p.i. EAUmouse. (A) Fundusimageshowing atrophic retina and inflamed venule segment (arrow). (81 Many small foci of increased AFdistribute throughout the retina, with more in the perivascular area larrowl. The AF background is lowerthanthat of the peak-stage EAU retina (Fig 4.6).


EXPERIMENTAL ANIMAL STUDIES TO DISSECT THEMECHANISM OF LIPOFUSCIN FORMATION

. 1 . Hpofuscin formation in the eye have exclusivelyfocUJedStudies on the mec rarusm 0 I . fi . . RPE 115' I. b I' d h the formation of Iipo uscrn m ce mvovestheon RPE cells It IS e ieve t at I al fu .

. 1 b RPE d photoreceptors and the reduced ysosom n(t1oninvisual cyc e etween an h 2)RPE cells, in addition to other mechanisms (see also C apter .' . .

The visual cycle is a tightly regulated cycle in which retinoid (vitamin A)mov,b I de th c RPE ells to photoreceptor outer segments through theInter.ac< an lor lrom c . .. . ..h . (IPM) and is strongly retained in this loop. Retinoid '" reoP otorecepror matrix . " "I d f h bl d t m to RPE cells With light illurninarion, retinoid mov~ease rom t e 00 s rea . . .f h . RPE cells and with dark adaptation the process IS reVetsea.rom t e r'ctrn.a to, .. . .Substantial evidence indicates that vitamin A derivatives (rerinoids) are require] forRPE lipofuscin formation. In normal rat retinas, age-related lipofuscin accumula~ionoccurs in RPE cells, and vitamin A deficiency has been shown to reduce RPEIipo.fuscin (9). In the Royal College of Surgeons (RCS) rat, a strain. with an inheritedretinal dystrophy, lipofuscin-like AF develops in rhe degenerating phororeceprorcells, and vitamin A deficiency substantially reduces the AF associared with degener.ating photoreceptor cells. Furthermore a vitamin A-dependent fluorophore wasiso-lated from these retinas with the use of thin-layer chromarography (TLC) (10).However, this fluorophore differs (in fluorescence intensity and mobility in TIC)from that of aged RPE cells of normal rats. It appears thar the f1uorophore generatedin the photoreceptor cells must undergo chemical modification once it has beentaken up by the RPE (10).

The requirement of vitamin A for lipofuscin formation is further supported brastudy in which lipofuscin was induced in RPE cells by inrravitreal injection of thelysosomal protease inhibitor leu pep tin. Intravitreal injection of leupeprin causedrapid accumulation of inclusions with lipofuscin-like AF in rhe RPE of albino fa"(II). However, in vitamin A-deprived animals, similar inclusions formed in responseto leupeprin treatment, but they did nor become AF (II). It appears likelythatretinoids are directly incorporated into these inclusions. In another animal study, inwhich all-trans retinol was labeled with 3H and injecred into rats that were treatedwith an intravitreal injection of leupeptin, the radiolabel in the RPE was primarilyassociated with the leupeprin-induced inclusion bodies. Label was also present in thephotorecepror outer segments. The localization of vitamin A to the leupeptin-induced inclusions in the RPE strongly suggests that vitamin A is covalently boundtoouter segment proteins that have been phagocytosed by the RPE bur remain unde-graded as a resulr of protease inhibition. Vitamin A is nor likely ro be bound througha Schiff base linkage, since retinal-SchiFf base compounds do not exhibit lipofuscin.like fluorescence (12).

The importance of a normal visual cycle in the formarion of lipofuscin is alsosupported by studies in RPE65 knockout mice. RPE65 is essential in the operationof the visual cycle and functions as a chaperone for all-trans-retinyl esters, the sub-strates for isomerization in the visual cycle. During the visual cycle, the visual pig-ment chromopho~e, ll-c~s-retin~, is photoisomerized to the all-trans configurationa.nd then en~ymat1~all! retsomerized to the II-cis isomer. RPE65 proteins are essen-tial for the lsomenzatlOn of the all-trans- to II-cis-retinaL Mice with RPE65 defi-ciency (i.e., RPE65-

1- mice) accumulate almost no lipofuscin in RPE cells (4), and

RPE65+/~ mice also have significantly reduced accumulation of lipofuscin fluo-rophores 111 RPE cells. Further experiments in RPE65 knockout mice with proteaseinhibitor (leupeptin) treatment showed that although lipofuscin-like inclusions were

CHAPTER 4 LIPOFUSCIN AND AUTOFLUORESCENCE IN EXPERIMENTAL ANIMAL MOOELS 45

observed in both WT and RPE65 knockout mice, only the inclusions from WTRPE (and not RP£65-1~ RPE) autofluoresce (13). These observations suggest thatthe formation of RPE lipofuscin fluorophores is almost completely dependent on anormalvisual cycle.

In line with,these experimental studies in mice, a lack of AF in patients with early-onset severe rennal rod-cone dystrophy has been observed (14). Early-onset severerod-conedystrophy is believed to be associated with RP£65 mutations (15).

The fact that leupeptin treatment induces lipofuscin-like inclusions suggests thatlossof lysosomal function is also responsible for the formation of lipofuscin. The in-clusions are derived from phororeceptor outer segments1 which are normally phago-cyrosedand degraded by the RPE. A normal lysosomal function is essential for thedegradation of phagocytized materials. The tripeptide leu-gly-gly, which is similar toleupeptinexcept that it does not inhibit proteolysis, has no effect on RPE AF content.Likewise,nerilmicin, a purported inhibitor of lysosomal phospholipases, does not in-creaseAF in the RPE (11).

In summary, experimental studies on the formation of lipofuscin in RPE cellssuggestthat loss of lysosomal function is essential for the formation of lipofuscin in-clusions, whereas a normal visual cycle is required for lipofuscin formation.

TESTING ANTILIPOFUSCIN COMPOUNDS INANIMAL MODELSIncreasingknowledge regarding the mechanism of lipofuscin formation will enablethe development of antilipofuscin drugs, which could be of benefit in various retinaldegenerative diseases. As summarized above, inhibiting visual cycle function withsmallmolecules has been shown to be an effective approach to prevent lipofuscin for-mation.Various animal models have been used to test the effectiveness of compoundsdesignedto inhibit lipofuscin formation (16). The effect of isotretinoin (Accutane),whichhas been shown to slow the synthesis of l1-cis-retinaldehyde and tegenerationof rhodopsin, has been tested in ABCA4-1- knockout mice, a mouse model ofStargardt disease (see Chaptet 1IG). Fenretinide (N-[4-hydroxyphenyl]retinamide[HPR]) potently and reversibly reduces serum retinol. Adrninisrration of HPR toABG44-J~ mice caused immediate, dose-dependent reductions in serum retinol andretinol binding protein (RBP). Chronic administration can produce commensuratereductions in visual cycle rerinoids and arrest accumulation of A2E and lipofuscinAF in RPE cells. Physiologically, HPR treatment causes modest delays in dark adap-tation. Chromophore regeneration kinetics, light sensitivity of photo receptors, andphototransduction processes are normal. HRP could be a novel therapeutic approachwiththe potential to halt the accumulation oflipofuscin fluorophores in the eye (18).In this regard, a Phase II randomized, double-masked, placebo-controlled, multicen-tersrudy using fenretinide is currently under way (see also Chapter 10C).

Nonrerinoid isoprenoid compounds have been shown to serve as antagonists ofRPE65. These RP£65 antagonists block regeneration of 11-cis-retinal, the chro-mophore of rhodopsin. Chronic treatment of a mouse model of Srargardr diseasewithRP£65 antagonists abolished the formation of A2E. Thus, RP£65 is also on therate-limiting pathway of A2E formation. These nontoxic isoprenoid RP£65 antago-nistsarecandidates for treating forms of macular degeneration in which lipofuscin ac-cumulation is an important risk factor. These antagonists will also be used to probethemolecular function of RP£65 in vision (19).

SUMMARYd i the clinic for almosr a decade, ouruntbt,. . has been use In . Is fc

Although AF lmagmg .. d An' al models are Important tOO Orgaiing of fundus AF is still limite . d hei correlation to observed fundus AF en.. al h ges an t err .sight into parhologic can. h Cation of lipofuscin. They are alsohani derlymg t e ror'rnand the mec arusms lin . I' ofuscin formarion with the goalofpr

. IT' drugs targetmg Ip .for developing errecnve di d cred so far on the mechamsrns of Iipo. d . Stu les con uing retinal egeneranon. . r molecules and con equenrly new rrCltr. h aled many Imporran , . .formation ave reve d Th dy of clinicopathologIcal corrd.nornorh bei develope e stuapproac es are emg h b d onsrrared thar other cells besides RPEh . b Ir as een em .dus AF as just egun. F h k in particular using differenr animal mod,d r fU5CIn urt er war , . . .may pro uce IpO d d th bnormaliries in the dlsrnbunon of fund~di . d d to un erstan e atsease, IS nee e . al di as well as their significance. Ir should be

rell1lobserved in different reun rseases, . Id b

h b rions made in animal models shou su sequ,n",bered, however, r at any 0 servavalidated clinically in humans.


ACKNOWLEDGMENTSTh

h h k D A Manl'vannan (University of Aberdeen) for helpingwirhm,e aut or t an Sr..

SLO imaging, and Professor]. Greenwood (Insricure ofOphrhalmology, LondoolLproviding the CFH knockout mouse samples.

REFERENCES

1. Xu H, Chen M, Manivannan A, Lois N, cr aJ. Age-dependem accumulation oflipofu.scin in prnY2.Sl':llWand ,"beetin,) mitroglia in expeeimental mite. Aging Cell 200B;7;5B-6B. .

2. Coffey Pj, Gias C, McDermott C], er a1. Complement faeror H deficiency in aged mice causes retmal:ili.noem')iti" and vi,u,) dy,funttion. Peot Nacl Acad Sci USA 2007; J 04; 16651-16656. .

3. Xu H, Manivannan A, Goatman KA, et a/. Improved leukOCYtetradcing in mouse retinal andchorordaci'Culation. Exp Eye Rec 2002;74;403-4 J o. .

4. Katz ML, Redmond TM. Effect of Rpe65 knockout on accumulation ofl.ipofuscin fluorophoresIn dlt'etin,) pigm'nt epithdium. Invw Ophth,)mol Vi, Sci 2001;42;3023-3030. .

5. Marmorstein AD, Marmorstein LY, Sakaguchi H. et a.l. Spectra.! profiling of autofluorescencea.ssocwnlwith lipofuscin, Bruch's membrane, and sub-RPE deposits in normal and AMD eyes. InvestOphlhalmoJVis Sci 2002;43:2435-244l.

6. Bui TV, Han Y, Radu RA, et,). Chmcte,i"'ion of n"ive 'eeinal fluocophoe" involved in bio'y",h.,of AlE and lipofuscin-associated retinopathies. ] BioI Chem 2006;281: 18112-18119.

7. Amb"i], Anand A, Femande' S, et ,). An anim,) mode! of age-'e!ated moeul", degen''''ion in ''"'"''''Cc1-2- or Ccr-2-deficient mice. Nat Med 2003;9: 1390-1397.

8. BmdecickC, Hock RM, Fo"e'te, ]V, " ,). Co,mitutive eetinal CD200 expe"'ion eegular" ,,,id,,,, mi-ccoglia'nd activarionstate ofinflammarory coil, dming expeeimcnc,) autoimmune uv<o"cinici,.Iun]Mol2002; 161;1669- J 677.

9. Robison], WG. Kuwabata T, Bieci]G. Deficienciec of vitami", E and A in ,he tat. R,cin,) damag,,,dlipofuscin accumulation. Invest Ophrhalmol Vis Sci 1980;19:lD30-1D37.

10. KatzML, Eld'ed GE, Robison], WG. lipofUSCin autofluoeescence; evidence foe vitamin A invol,,"""in the retina. Mech Ageing Dev 1987;39:82_90.

11. Katz ML, Shank'e MJ. D,vdopment of lipofuscin-like fluo'escence in the min,) pigm'nt epithdilmlioresponse to protease inhibitor treatment. Mech Ageing Dev 1989;49:23-40.

12. K"z ML, Gao CL. Vitamin A inco'pomion inco lipofuscin_lik, inel"'ion' in the 'etin,) pigmem,pi<h,.Hum.Mech Ageing Dev 1995;84:29_38.

13. Ka.tzML, Wendt KD, Sande" DN. RPE65 gene mu'acion peev'n" deve!opmeot of aumfluoe"",,, 10reunal pIgment epJrhelJalphagosomes. Mech Ageing Dev 2005; 126:513-52l.

14. Lo"'" B, Wabbeh B, Wegscheidec E, et .1. Lack of fund", autofluotescence co 4BB nanommo fromchildhood on in p"ients with eady-on"t "v,,, "tinal dymophy "'oci"ed wim mutation' in RPE6;.Ophthalmology 2004;111:1585_]594.

CHAPTER 4 LIPOFUSCIN AND AUTOFLUORESCENCE IN EXPERIMENTAL ANIMAL MODELS

15. Lorenz B, Cyurus P, Preising M, er al. Early-onset severe rod-cone dystrophy in young children withRPE65 mutations. Invest Ophthalmol Vis Sci 2000;41 :2735-2742.

16, Radu RA, Mara NL, Nusinowitz 5, er a]. Treatment with isotretinoin inhibits lipofuscin accumulation in amouse model of recessive Srargardr's macular degeneration. Proc Nat! Acad Sci USA 2003; 100:4742--4747.

17. GollapaUi DR, Rando RR. The specific binding of rerinoic acid to RPE65 and approaches to the treat-rnenr of macular degeneration. Proc Nat! Acad Sci USA 2004; 101: 10030-1 0035.

18. Radu RA, Han Y, BUI TV, er al. Reductions in serum vitamin A arrest accumulation of toxic retinal flue-rophores: a potential therapy for treatment oflipofusein-based retinal diseases. Invest Ophthalmol Vis Sci2005;46;4393-4401.

19. Maiti P, Kong J, Kim SR, er al. Small molecule RPE65 antagonists limit the visual cycle and preventlipofuscin formation. Biochemistry 2006;45:852-860.

CHAPTER

Imaging Techniques of FundusAutofluorescence

Ohe imaging of fundus autofluorescence (AF) phenomena in vivo, .as dis~ussedin Chapter 1, was first demonstrated in the early days of f1uores~em. angIogra-phy, when prcinjection fluorescence of optic disc drusen and OptIC disc hamar-

tomas in tuberous sclerosis, and within lesions of Best virelliform macular dystrophywas detected (1---4). However, these observations were limited to a few patients withpathological accumulations of highly fluorescent material. By contrast, the naturallyoccurring intrinsic fluorescence of the ocular fundus is quite low-about 2 orders ofmagnitude lower than the background of a fluorescein angiogram at the most intensepart of the dye transir (5). Absorprion of excirarion and emission lighr wirh partly ad-ditional generation of fluorescence by anatomical structures anterior to the retina fur-ther interferes with the detection offundusAF. The main barrier is the crystalline lens,which has highly fluorescent characteristics in the short-wavelength range (excitationbetween 400 and 600 nm resulrs in peak emission at 520 nrn) (see also Chapter 3) (6).With increasing age and particularly the development of nuclear lens opacities, the flu-orescence of the lens becomes even more prominent. Therefore, fundus AF imagingwith a conventional fundus camera using the excitation and emission filters as appliedfor fluorescein angiography produces images with low conrrasr and high backgroundnoise in young persons (Fig. 5.1). In rhe elderly, the qualiry of rhe images drops evenfurther and ir becomes practically impossible ro evaluate the disrriburion of fundus AF.

RECORDING FUNDUS AUTOflUORESCENCETo better record fundus AF, adjustments and modifications of existing camera sys-tems or sophisticated new imaging devices are required. Such devices include the fun-dus specrrophoromerer, rhe confocal scanning laser ophthalmoscope (cSLO), and rhemodified fundus camera.

Fundus Spectrophotometer

The fundus specrrophoromerer developed by Delori and coworkers (5,7) was designedto measure and systematically analyze the excitation and emission spectra of the AFfrom small rerinal areas (2 degreesdiamerer) of the fundus. By incorporaring an image-intensifier-diode-array as detector, beam separation in the pupil, and confocal detectionto minimize contribution of AF from the crystalline lens, this device allows absolutemeasurements of AF. Groundbreaking work by Delori er al. (5) demonstrated that Iipo-fuscin IS the dominanr source of intrinsic fluorescence of the ocular fundus. However,the small field of view is nor practical for recording fundus AF in large patienr popula-nons or III the clinical setting.

48

CHAPTER 5 IMAGING TECHNIQUES OFFUNDUS AUTOFLUORESCENCE 49

FIGURES.1. Comparison of fundus AF imaging with a conventional fundus camera and cSLO. (A) Fundus photographofa 29-year-old subject with no eye disease and clear media shows good quality. with small retinal vessels and fovealreflexvisible. (8) Fundus AF imaging with the conventional fundus camera using the filters as applied for fluoresceinangiographyand maximum flash intensity [300 JI in the same subject produces a noisy image with low contrast. Theoptic nerve head appears to be artificially bright. which may be caused by scattered excitation light and thereforewouldbe attributable to pseudofluorescence. Ie) Fundus AF imaging with the cSLO. in contrast. results in a clear imagewith high image contrast and high sensitivity. (D) The quality of the fundus color photograph of this patient with geo-graphicatrophy secondary to age-related macular degeneration IAMO) IS slightly impaired as a result of concomitantnuclearsclerosis of the lens. IE) Corresponding fundus AF image obtained with a conventional fundus camera using theexcitationand emission filters used for fluorescein angiography. The image is hazy and no retinal details are visible.IF)FundusAF image obtained using a cSLO. The quality of the AF image is better than that obtained with the funduscamera,and macular abnormalities, including a central area of atrophy, are observed.

Confocal Scanning Laser OphthalmoscopyThe use of a cSLO optimally addresses the limitations of the low intensity of the AF sig-nal and the interference of the crystalline lens. It was nsed initially by von Rlickmannand coworkers (8) in a clinical imaging system.

The cSLO projects a [ow-power laser beam on the retina that is swept across thefundus in a taster pattern (9). The intensity of the reflected light at each point, after itpasses through the confocal pinhole, is registered by means of a detector and a two-dimensional image is subsequently generated. The use of confocal optics ensures thatout-of-focus light (i.e., light originating outside the adjusted focal plane but withinthe light beam) is suppressed and thus the image contrast is enhanced. This suppres-sion increases with distance from the focal plane, and signals from sources anterior tothe retina, i.e., the lens or the cornea, are effectively reduced.

In contrast to the 2-degree discrete retinal field of the fundus spectrophotome-ter, the cSLO allows imaging over larger retinal areas. The standard image encom-passes a retinal field of 30 degrees X 30 degrees. Addirionallenses allow for imaging

---


of a 55-degree field or, using the composite mod~, imaging over even ~argerretinalareas. To reduce background noise and enhance Image contrast, a s~rtes of severalsingle images is usually recorded (8,10-13). For rhe final fundus AF Image, a num-ber of these frames (usually 4 to 32) are averaged and pixel values are normalized.Given rhe higb sensitivity of the cSLO and the high frame rare of up to 16 framesper second, fundus AF imaging can be performed within seconds and at low excira-tion energies that are well below the maximum retinal irradiance hm1t~ of laserses-tablished by the American National Srandards Insrirure (14) and orher InternationalorganIzatIons. .'

Three different cSLOs have been widely used to obtain fundus AF images: meHeidelberg retina angiograph (HRA classic, HRA 2 and HRA Spectralis; HeidelbergEngineering, Dossenheim, Germany), the Rodenstock cSLO (RcSLO; Rodenstock,Weco, Dusseldorf, Germany), and the Zeiss prototype SM 30-4024 (ZcSLO; Zeiss,Oberkochen, Germany). For fundus AF imaging, all three use an excitation wavelengthof 488 nm. Emitted light is detected above 500 nm for the HRA, above 515 nm formeRcSLO, and above 521 nm for the ZcSLO. Clinically useful fundus AF imagingbasbeen reported with all three systems by several studies (8,10,12). A systematic com-parison among the three sysrems by Bellmann and coworkers (15) in 2003 showedthat both image contrast and image brightness were significanrly higher with theZcSLO and HRA classic compared to the RcSLO. Using a model eye, the highesrbackground noise was measured with the ZcSLO and the lowesr with the HRA clas-sic. Since image contrast and brightness, and background noise are important indica-tors of image quality, Bellmann et al. (15) concluded that the observed differencesmight be of great importance when comparing fundus AF findings obtained with dif-ferenr imaging devices. Subsequent to their study, further developments led to the in-troduction of the HRA 2, which is currenrly the only commercially availablecSLOforfundus AF imaging. This device allows teal-time imaging, i.e., recording of the finalmean and normalized image during acquisition. No time-consuming postacquisitionprocessing of a movie is required. Thanks to an improved algorithm for automatedimage alignment to correct for eye movements during acquisition, a higher number ofsingle frames can also be more easily captured, improving the signal-to-noise ratio andtherefore possibly leading to better visualization of details. Recently, the combinationof cSLO imaging with spectral-domain optical coherence tomography (OCT) in oneinstrument was made possible. This new imaging device enables simultaneous cSLOand OCT recordings, taking advantage of the registration of the fundus image andlateral eye movements by the cSLO system and synchronous topographic alignmenrto OCT scans. Several OCT scans can be averaged to reduce the background noise,and three-dimensional correlations of visible structural changes are possible.

MODIFIED FUNDUS CAMERAUntil recently, clinical applications of fundus AF imaging were limited to the cSLO,and in recent years the cSLO has gained increasing popularity and mote widespreadapplication. However, the major advantage of the cSLO-its sophisticated technicalsetup-also represents its major drawback: its considerable purchase and rnainre-nance costs .

.As ~entioned above, fundus AF imaging with a conventional fundus camera op-etanng ill the same wavelength range as the cSLO has limitations due to the relativelylow fundus AF signal, the absorption effects of the crystalline lens and the nonconfo-caliry, which makes the fundus camera prone to light scattering. Dcicri and cowork-

CHAPTER 5 IMAGING TECHNIQUES OF FUNDUS AUTOFLUORESCENCE 51

ers(16)described the possibility of obraining fundus AF images using a modified fun-dus camera. This included the insertion of an aperture in the illumination optics ofthecamerato minimize the loss of contrast caused by light scattering and fluorescencefromthe crystalline lens. However, this modification also resulted in the restriction oftheangleof view ro a 13-degree-diamerer circle; this, rogether with the complex de-sign,isthe likely reason why this setup has not been furrher pursued and used by othergroupsor in a clinical serting ro date. Spaide (17) reported his eleganr idea of modify-ing a commercially available fundus camera by simply moving the excitation andemissionwavelengths for fundus AF imaging toward the red end of the spectrum tobypassthe fluorescence of rhe lens. The theoretically inexpensive purchase of an addi-tionalfilter set, rogerher with the broad availability of the fundus camera, may repre-sentan attractive alternative. Figure 5.2 illustrates the excitation and emission filtersofthemodified fundus camera as introduced by Spaide in comparison with the cSLOforfundusAF imaging. Note that Spaide recently developed an additional modifica-tionof the filrers (here called "modified fundus camera 2" as opposed to "modifiedfundus camera 1" for the first modification). Currently, there is only one commer-ciallyavailablemodified fundus camera for fundus AF imaging (Topcon TRC-50IX;Topccn, Paramus, N]), based on the modification by Spaide (modified fundus cam-era2). Because only rhe filters are altered in the modified fundus camera, the othertechnicaldetails remain unchanged compared to the conventional fundus camera, in-cludingthe angle of coverage (up to 50 degrees) and possible range of flash light in-

System cSLO modified FC-l * modified FC~2*

Excitation 500-610 nm

615-715 nm

488 nm 535-580 nm

Barrier filter >500, 515or 521 nm

675-715 nm

cSLO

400 500 600 700Wavelength (nm)

modified Fe·1 *

400 sao 600 700Wavelength (nm)

modified FC-2*

800

800

400 500 600Wavelength (nm)

700 800

*according to R, specie

FIGURE 5.2. Overview 01 excitation and emission spectra of imaging systems for 1undus AF imaging ThecSLOsystem is based on excitation of a blue laser beam ().. ~ 488 nrn] Emission is detected with a cutofffilter (A'" 500 nrn. depending on the system; see text]. For the modified fundus camera, two bandwidth fil-ters are applied (17,271. operating in a longer-wavelength range compared to the cSLO. Subsequent to theoriginal modification [here called modified FC-l I, Spa ide recently introduced a second modification (here

calied modified FC-Z).

--

TABLE 5.1 Summary of Technical Differences Between the cSLO and theModified Fundus Camera for Fundus AF Imaging


.. C rl fundus AF imaging with the modified fundus camera is basedOntenSltles. unen y, . .excitation by one single flash and the immediate capture of a sJn?le Image. After ac-quisition, the image brightness an~ .contra~t v~ue~ of this single Im~ge ar~ .manuallyadjusted to better visualize the individual distribution of fundus AF mtensrues.

COMPARISON BETWEEN THE CSlO AND MODIFIEDFUNDUS CAMERA FOR FUNDUS AF IMAGINGBoth the cSLO and the modified fundus camera allow for fundus AF imaging in theclinical setting and both are commercially available. However, only one systematiccomparison of different retinal pathologies is available. Theoretical considerationsand preliminary observations suggest that fundus AF imaging with the modified fun-dus camera might not be entirely equivalent to the cSLO sysrem. Table 5.1 summa-rizes the technical differences between both imaging devices, which are discussed inmore detail below.

Excitation and Emission SpectraBecause of the broad excitation and emission spectra of ocular intrinsic fluorescence, itis conceivable that both systems, with their individual fluorescence settings, are able codetect fundus AF (5,7). However, fundus AF is made up of several f1uorophores withdifferent absorption and emission spectra as well as different amounrs of fluorescenceintensity (see also Chapters 1-3 and 8). For example, minor Iluorophores of the fun-dus include NAD-NADH+ (oxidized and reduced nicotinamidadenindinucleotide)and FAD-FADH' (oxidized and reduced f1avinadeninucleotide) in mitochondria, ad-vanced glycation end products (AGEs), collagen, and elastin (l8~21). Furthermore,lipofuscin as the dominant fluorophore is composed of several molecules that may notonly be disease-specific (and therefore may cause alteration of the spectrum), bur alsoknown to be characterized by a red shift of the emission with increasing excitationwavelength (22). These considerations would imply that the composition of the de-tected AF signal may vary between the cSLO and the modified fundus camera.

The different excitation and emission spectra may explain the different patterns ofdecreased intensity revealed by the two systems in the central macula in a normal sub-ject (Fig. 5.3). Generally, the decreased AF signal in the macula of healthy individuals

oSLO Modified fundus camera

Oneexcitationwavelength(lasersource)large emissionspectrum(cutofffilter)Continuous scanning at low light intensities in

a raster patternConfocalsystemLaser power fixed by manufacturer, detectorsensitivityadjustable

Imageprocessingwith averagingof singleframes and pixel normal ization

Bandwidth filters for excitation and emission

One single flash at maximum intensities

Entirecone of lightFlashlight intensity. gain and gammaof detectoradjustable

Manualcontrast and brightness

Schmitz-Valckenberg S. Fleckenstein M, Gobel AP. Evaluation of autofluorescence imaging with the scanning laser ophthal-moscope and the fundus camera in age-related geographic atrophy. Am J OphthalmoI2008;146(2j:183-192.

IMAGING TECHNIQUES OF FUNDUS AUTOFLUORESCENCE

I I

CHAPTER 5 53

FIGURE 5.3. Differences between the cSLO and modified fundus camera 1 in detecting fundus AF in-tensityin the central macular area. (A-D) High-magnification images of the same eye comparing the re-flectance(A) and AF image (8) obtained with a cSLO. and the color fundus photograph Ie) and the AFimageobtained using the modified fundus camera (DI. A decreased signal is detected in the center of themaculain both images obtained with the cSLO. suggesting absorption of the excitation light. In contrastto the cSLOimages, an irregular area of decreased intensity and overall a small area of decreased inten-sityis seenin the AF image obtained with the fundus camera. Note also that the contrast for retinal bloodvesselsis less on the image obtained with the fundus camera compared to the cSLO,which enables bet-tervisualization of the peri foveal vasculature.

hasbeen attributed to macular pigment absorption, increased melanin deposition, andlowerdensity of lipofuscin granules in central retinal pigment epithelium (RPE) cells(5,23).In particular, the absorption of macular pigment (peak absorbance at 460 nm,markedreduction> 510-540 nm [5,24]) should differ between the two systems.

Mode of Excitation and DetectionAmajor difference between both systems is embodied in the scanning mode and theuseof confocal optics with the cSLO as opposed to the excitation by a single flash andthe fluorescence exhibition within the entire cone of light at one single instance withthe fundus carnera. The latter mode is particularly prone to pseudofluorescence

, h e usually much more easily visualized'iilifl in anglOgrap y arthe leakage on uoresccr a {It has been speculated that rhey represent areaswiilithe modified fundus camera. , 'the result of gravity effects [28,29],) Anoili"

' al fl id d that their location IS hsubretin U1 an dvnarni c AF changes due to p orableaching areibl I ' is that short, ynarmPOSSI e exp ananon , nning mode as opposed ro rheshortU.h' I .h the conrrnuous seamore like y to occur wit de lying choroidal blood vessels are someeimer

f h f d Furthermore, un r , ,ate un us camera, d The ability to image OptiC dISCdrusen, which' h h d fi d fun us camera, ,seen Wit tne rna I ie b hi h AF ay be also be enhanced with the modifiolare usually chatactenzeld I Yflgh r~ :uried deeper in the optic nerve (30), Th",fu d era parnell ar y 1 t ey a

n us cam, b 11 lained by pseudofluorescence phenomena, Twoobservations cannot e tora y exp lik I rl d'f!i fil

' bi rion seem to be more ley: ie I erenr lerother reasons, alone or In com Ina 1 , .. I' d' , ( d ' Fdifferenr fluorophores, Including e asnn an coUagencombination e.g., erection 0 1 . d ' f

' , d I all) d the nonconfoca! setup (e.g, easier erecnon 0 OUt,within the bloo vesse w s anof-focus fluorescence).

5& SECTION I BASIC SCIENCE

Image AcquisitionTh SLO he reti onrinuously and the actual image is immediately digitizede c scans t e retlna c ..

d ' ali d puter screen The orientation and posiuon of the laserscan,an VlSU lze on a com. . .ning camera, detector sensitivity, and refractive correction can b.e easIly, adjusted d~r.ing the acquisition ptocess and in rea! time by rhe operator, ThIS pracrical and eastlyfeasible mode of imaging is only possible because of rhe high sensmvrry and relativelylow light levels (intensity per retinal area) of the cSLO, Nevertheless, It ISImportantto optimally adjust these three modifiable settings to achieve good Image quality andgain reliable information from these recordings. . .

The acquisition mode with the fundus camera allows for modification of moresettings. In addition to the orientation and position of the camera, gain of the detec-tor, and focusing, the intensity of the flash of light and the gamma of the derector canbe adjusted. These variations mean more choices for the user. However, it mightbedifficult or rime-consuming to find the correct settings for an individual patient,

Image ProcessingCompared to the cSLO, image processing with the modified fundus carrera is easy andstraightforward. To optimize the visualization of the distribution of inrensiries, imagebrighrness and contrast are usually manually adjusted. However, chis modification hasnot been standardized yet and it should be noted that, after manual adjusrment ofthese image parameters, an absolute comparison of intensities within twO images(even when taken with the same acquisition settings) is not possible.

Standard operating procedures for cSLO fundus AF imaging are available bur havenot found widespread use in Inany clinical centers (31). Postacquisition image process-ing with the cSLO allows for more modifications and is more time-consuming (unlessthe real-time mode is used) compared to the fundus camera, The key principle is rharcalculations of a mean and normalized image are required in order to overcome tWO

major general drawbacks of fundus AF: noise and low sensirivity of the signal, withsubsequent difficult visualization of the topographic distribution of intensities. As aresult, these two modifications introduce two minor limitations: slight distortion ofdetails (because of imperfect alignment of single frames) and inability to compare ab-solure intensities between images, In the following, the processing of cSLO fundusAF Images IS explamed m more detail and compared with imaging by means of themodified fundus camera.

IMAGING TECHNIQUES OF FUNOUS AUTOFLUORESCENCECHAPTER 5 51

The noise of the signal within a single cSLO frame is significant. Noise is gener-ally a random event and interferes with the actual signal. Image noise is reduced by afactorof ~ where n IS the number of frames to be averaged. In practice, averaging aseriesof single frames can greatly improve the signal-ro-noise ratio and therefore thevisualization of details. The number of images to be processed is usually limited. Formathematical reasons (there is no linear relationship between the number of framesand the reduction of noise» the required number of frames increases by the square toachievethe same amount of noise reduction. For example, using four single frames re-duces the noise by a factor of 2 (V4). To further decrease the noise by 2, one wouldneed 16 (v16 = 4) frames, not eight frames. Furthermore, other practical reasonslimit the number of images used to calculate the mean image, including the exposuretime to the laser beam needed to image the patient and the increasing difficulry ofcorrecting for eye movements with more and more frames. The latter in particularcan be troublesome and lead, when using a high number of frames (16-64, depend-ing on the patient's stability of eye movements), to an increase in image noise andblurriness (fuzzy border of retinal structures) due to poor alignment of single frames.Therefore, in most cases it is more practical to use a reasonable number of single imagesto calculate the mean image. Compared ro the cSLO, the modified fundus camera usesthe maximum or near-maximum flash light intensities (200-300 J) for a single frame.This means that the number of images that can be safely taken is limited because ofexposure levels. However, averaging even a small number of images has not yet beendemonstrated to reduce noise and improve the visualization of details with the funduscameraAF imaging. Furthermore, to date, the commercially available software hasnor implernenred averaging algorithms.

One additional factor affecting the number of frames required to calculate themean image is the increase in sensitivity of the AF signal. A standard digital grayscaleimageis based on pixel values ranging from 0 to 255 units. However, with current de-vices, one AF image encompasses a band of approximately 20-40 units in most sub-jects (because of the low intensity of fundus AF). Therefore, a modification of thepixel histogram using all available values (0-255 units) greatly improves the visualiza-tion of the intensity distribution and therefore the visualization of more details, at theexpense of losing a judgment on the absolute AF intensities. The easiest way toachieve this is to manually stretch the histogram and modify the contrast, as is donewith the modified fundus camera. The more sophisticated cSLO method uses an au-tomatic software algorithm with normalization of the histogram. Using the informa-tion of several frames, binary calculation allows for a smooth stretching of the pixelhistogram. No gaps in between values occur (such as with the manual method usingone image) and therefore the sensitivity of the signal is enhanced. Compared to themanual modification, such as with every computer software, the automatic normal-ization algorithm may not work properly. For example, incorrect settings during theacquisition process, particularly wrong detector settings, may not be seen on a nor-malized image at first glance and may lead to false interpretations. Therefore, it is pru-dent to review single images and always compare them with the mean normalizedimage. Furthermore, details within areas of very increased or very decreased signal in-tensities compared to most of the rest of the image may be also obscured on imagesby improper automatic normalization. In these cases, it is advisable not to use auto-matically generated normalized images and to adjust brightness and contrast valuesmanually in the nonnormalized mean image.

It is very important to understand that neither automatic normalization nor man-ualadjustment of image contrast and brightness allows for the quantitative assessmentof absolute AF intensities. Such image processing merely facilitates visualization of the

--


topographic differences of AF intensities within one image, ,and thus the inte~preta_tion of localized alterations in the distribution of fundus AF in one eye at the time ofacquisition. Because of the artificial change of the absolute pixel values and their rela-tion to each other, an absolute comparison of normalized Images among eyes of dif-ferent patients, as well as images of the same eye obtained at differe~t time points, isnot possible. The software of modern cSLO imaging systems permlrs users to easilydeactivate the normalization of mean images. However, ro date, the absolute quantifi-cation of intensities remains a challenge. No reliable method for quantifying AF insmall areas has been demonstrated. It is again the influence of the lens (as discussed atlength above) that appears ro be the major limiting facror for absolute quantificationof AF intensities using the current commercially available systems.

REFERENCESI. Schatz H, et al. Preinjecrion fluorescence. Disc leak. In: Schatz H, er al., eds. Interpretation of Fundus

Fluorescein Angiography. St. Louis, MO: C.V. Mosby, 1978;251-259, 514-537.2. Wessing A, Veranderungen der Papille M-SG. Papillenschwellung. In: Wessing A, Veranderungen der

Papille M-SG, eds. Pluoresaenzangiographie der Retina. Lehrbuch und Atlas. Srungarc Georg ThiemeVerlag, 1968;166-172.

3. Musrooen E, Nieminen H. Optic disc drusen-a photographic study. I. Autofluorescence pictures andfluorescein angiography. Acta Ophrhalrnol (Copenh) 1982;60:849-858.

4. Neerens A, Burvenich H. Autofluorescence of optic disc-drusen. Bull Soc Beige Opbralmol 1977;179;]03-110.

5. Delori FC, er al. In vivo fluorescence of the ocular fundus exhibits retinal pigment epithelium lipofuscincharacteristics. Invest Ophrhalmol Vis Sci 1995;36;718-729.

6. Bessems GJ, et al. Non-tryptophan fluorescence of crysrallins from normal and cataractous human lenses.Invest Ophrhalmol Vis Sci 1987;28; 1157-1163.

7. Delori Fe. Spectrophotometer for non-invasive measurement of intrinsic fluorescence and reflectance ofthe ocular fundus. Appl Optics 1994;33;7429-7452.

8. von Rdckmann A, Fitzke FW, Bird AC Distribution of fundus autofluorescence with a scanning laserophehalrnoscope. Br J OphmalmoI1995;79:407---412.

9. Webb RH, Hughes GW, Delori FC Confocal scanning laser ophrhalmoscope. Appl Optics 1987;26:1492-1499.

10. Bellmann C, et a]. [Topography of fundus autofluorescence with a new confocal scanning laser ophthal-moscope]. Ophrhalrnologe 1997;94;385-391.

11. jorzik j], er al. Digital simultaneous fluorescein and indocyanine green angiography, autofluorescence, andred-free imaging with a solid-stare laser-based confocal scanning laser ophthalmoscope. Retina 2005;25:405-416.

12. Solbach U, er a]. Imaging of retinal autofluorescence in patients with age-related macular degeneration.Retina 1997; 17:385-389.

13. Bindewald A, er aI. [cSLO digital fundus autofluorescence imaging]. Ophchalmologe 2005;102:259-264.14. American National Standard for the Safe Use of Lasers: ANSI Z136.1. Laser Institute of America.

Orlando, FL: A.N.S.lnstiwte, 2000.15. Bellmann C, et al. Fundus autofluorescence imaging compared with different confocal scanning laser oph-

thalmoscopes. Br J Ophrbalmol 2003;87; 1381-1386.16. I?elori Fe, er aI. Autofluorescence distribution associated with drusen in age-related macular degenera-

tion. Invest Ophthalmol Vis Sci 2000;41;496-504.17. Spaide RF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology 2003;110:

392-399.18. Schweitzer D, et aI. In vivo measurement of time-resolved autofluorescence at the human fundus. J Biomed

Opt 2004;9; 1214-1222.19. Schweitzer D, et al. Towards metabolic mapping of the human retina. Microsc Res Tech 2007-70:

4]0-419. '

20. Nelson DA, er a1. Special report: noninvasive multi-parameter functional optical imaging of the eye.Ophthalmic Surg Lasers Imaging 2005;36;57-66.

21. Win~er BS, ." at. Metabolic mapping in mammalian retina; a biochemical and 3H-2-deoxyglucose au-roradiographic study. Exp Eye Res 2003;77:327-337.

22. Sparrow J. Lipofuscin of the retinal pigment epithelium. In: Holz FG, er al., eds. Atlas of FundusAur.ofluorescence I~aging. Berlin, Heidelberg; Springer, 2007:1-16.

23. Welter J], er aI. Retinal pigment epithelial lipofuscin and melanin and choroidal melanin in human eyes.Invest Ophchalmol Vis Sci 1986;27:145-152.

CHAPTER 5 IMAGING TECHNIQUES OF FUNDUS AUTOFLUORESCENCE

24. Hammond Jr BR, Wooten BR, Snodderly DM. Density of the human crystalline lens is related to themacular pigment carorenoids, lutein a.nd zeaxanthin. Oprom Vis Sci 1997;74:499-504.

25. Machemer R, et a.l. Pseudofluorescence-..a problem in interpretation of fluorescein angiograms. Am JOphd"limo] 1970;70;1-10.

26. Lemke L, Tilgner S, June A. [Sources of error in fluorescence-photography]. Ophrhalmologica 1967;153;349-354.

27.Spaide RF. Autofluorescence imaging with the fundus camera. In: Holz FG, et aI., eds. Atlas ofAutofluorescence Imaging. Berlin, Heidelberg: Springer, 2007:49-53.

28. Dandekar 55, ct al. Autofluorescence imaging of choroidal neovascularizacion due to age-related maculardegeneration. Arch Ophrhalmol 2005;123: 1507-1513.

29. Spaide RF, Klancnik Jr JM. Fundus autofluorescence and central serous choriorerinopachy. Ophthalmology2005;112;825-833.

30. Holz FG, Schmirz-Valckenberg S, Spaide RF, et al, eds. Atlas offundus autofluorescence imaging. Berlin:Springer-Verlag, 2007:313-327.

31. Schmirz-Valckenberg S, et al. How to obrain the optimal fundus autofluorescence image with the cSLO.In: Holz FG, et al., eds. Atlas of Autofluorescence Imaging. Berlin, Heidelberg: Springer, 2007:37-48.

CHAPTER

Near-Infrared FundusAutofluorescence

F F) d . d f li ofuscin and its majorFundus autofluorescence (A errve rom Ipfluorophor, AlE, has developed into an i~por~an[ no~invasive imaging .tech-nique in the last decade. When interpreting Images In normal and dl~ea:e

states, ODemust keep in mind that several different fluorophores are present withinthe retina and the retinal pigment epithelium (RPE) (see also Chapters 1-3) (1,2). Aspecific selection of the wavelength for excitation of AF and the. wavelength of thecutoff filter for emitred light could provide additional possibilities for norunvasrveimaging of the RPE. . .

Near-infrared autofluorescence (NIA) imaging, which uses the same excrtanonlight and cutoff filters employed for indocyanine-green (rCG) angiography, is widelyavailable. NIA was first described as pseudofluorescence prior to rCG angiography(3). The intensity of the emitted AF is about 60-100 times lower compared to that ofAF (2), but imaging can be reliably repeated.

NIA imaging was first reported in 2006, and has been applied in several recentlypublished studies (2,4-15). This chaprer describes the imaging technique, rhe originof the NIA signal, the distribution ofNIA in the healthy eye, and the clinical findingsin selected retinal diseases and their implications.

IMAGING TECHNIQUEImaging of NIA can be done using a confocal scanning laser ophthalmoscope(cSLO, Spectralis HRA; Heidelberg Engineering, Heidelberg, Germany) (2).Different camera objectives provide either a 30-degree or wide-angle field-oF-viewmode. The image resolution is usually 768 X 768 pixels, but it can be varied.Focusing is achieved using the near-infrared reflectance mode at 815 nm. Diodelaser lighr (787 nrn) is used to excite AF. A band-pass filter with a cutoff at 800 nmincluded in the system is inserted in front of the detector. Six pictures per second arerecorded and about 15 single images are averaged depending on the fixation of thepatient (Fig. 6.1). Because of the lower-intensity signal, optimal imaging is per-formed with dilated pupils. Cideciyan er a1. (5) reported a technique termed near-infrared reduced-illuminance AF imaging (NIR-RAFI). Although they reduced theilluminance for FAF in their HRA2, they used the same manufacturer settings forNIR-RAFI as described above.

NIA images are evaluated qualitatively on a computer monitor. Abnormalities ofNIA distribution are documented and classified as either increased, reduced, or ab-sent NIA compared to other areas within the same image. Interindividual differencesin media transmission properties due to the presence or absence of cataract, aftercataract, or moderate vitreous opacities, and the associated difficulties with normal-ization in a clinical setting preclude an easy quantification of NLA intensity. Image

CHAPTER 6 61NEAR-INFRARED FUNDUS AUTOFLUORESCENCE

FIGURE 6.1. Normal distribution of FAFas recorded using conventional AF imaging and NIA imaging.At the posterior pole, AF shows the highest intensity in the perifoveal area with a decrease toward thefov8a.lncontrast, NIA has a peak intensity in the foveal area with a decrease toward the periphery.Moreperipherally,AF and NIA distributions are homogeneous.

postprocessingwith different programs and quantification of NIA images has beensuggestedin research serrings (2,5).

ORIGIN OF THE NIA SIGNALThe major fluorophore contributing to NIA is most likely melanin and its relatedcompounds (melanolipofuscin, melanolysosomes, and oxidized melanin) in the RPEcells(2).A low-intensity signal is obtained from melanin in rhe deep choroid, whichcanbedetecred in areas denuded of RPE cells (i.e., geographic atrophy).

The norion that the NIA signal derives from melanin is based on the followingevidence:The distribution ofNIA seems to correspond to rhat ofRPE melanin (16),although the contriburion of choroidal melanin is evident (2). The severely reducedNIAsignal in areas of RPE loss indicares that rhe major source for NlA is rhe RPE.The markedllyincreased NIA signal in choroidal nevi supports the view that NIA isderivedpredominantly from ocular melanin (see Chapter 15). Melanin absorptiondecreaseswirh increasing wavelength (17,18). AF emission spectra of melanin gran-ulesof human RPE have been mostly examined for shorter wavelengths (17,19). Atwavelengthsused for conventional AF imaging «500 nm), melanin AF is presentburwith much lower intensity compared to lipofuscin. For synthetic melanin andmelaninin the skin, AF with a maximum at 870-900 nrn was demonstrated after ex-citation at 785 nm (20). Melanin AF is increased with oxidarion (18,21,22)_Melanolipofuscin (melanin with a cortex of lipofuscin) and melanolysosomes(melanin with a cortex of enzyme-reactive material) accumulate with age and mayrepresentmelanin in the process of repair, modification, or degradation (23). The AFof melanolipofuscin is intermediate between melanin and lipofuscin (19). In sum-mary,AF of melanin, oxidized melanin, and compound granules containing melanincontribute to the NIA signal.

Keilhauer and Delori (2) thoroughly discussed the possible contributions ofother fluorophores to the NIA signal. Even for conventional AF imaging, only a fewfluorophores wirhin rhe lipofuscin complex (AlE, cis-isomers of AlE, and all-trans-retinaldimer conjugare [24-26]) have been clearly characterized (see also Chapter 3).Other fluorophores may be present, and rhe composition of fluorophores within the

....

84 SECTIONI BASIC SCIENCE

NIA d to conventional AF. When rhere is no detectable NIA fromRPEon compare . dII I . . NIA signal hom the deep choroid can be etecred betweendu'.ce s, a ow-mtensrry K

choroidal vessels.

CLINICAL FINDINGSGiven that NIA is a very new imaging technique, rhere are scarce data (to date)on~edistribution oENIA in various retinal diseases, its clinical usefulness, and its implica.tions for the understanding of pathophysiologic mechanisms of disease. Herein,wewill describe only those conditions in which sufficient data are currently available.

ABCA4-Gene Related Stargardt Disease andCone-Rod DystrophyIn Srargardt disease and cone-rod dystrophies associated with mutations in theABCA4gene, characteristic NIA findings have been reponed (5,8, J 1). Abnormalities in thedistribution oENIA, as it occurs with conventional AF abnormalities (see also ChapterII G) (50,51), may be limited to the posterior pole or progtess beyond the vascularn.cades. Whereas foci of increased AF are frequently seen when using conventionalAFimaging, foci of increased NIA signals are rare. NIA is frequently reduced in patehyorconfluent areas (Fig. 6.3); of interest, in these areas, foci of increased AF are frequendydetected in conventional AF imaging. Areas ofNIA abnormalities appear to belargerthan those observed on conventional AF images, which may indicate that NIA alter.ations precede conventional AF abnormalities. Atrophic areas have absent NIA. Therelative peripapillary sparing observed using conventional AF imaging (seealsoChapter 11G) is also present on NIA imaging.

FIGURE 6.3.. Excerpts of conventional FAF and NIA images demonstrating progression of Stargardldisease. The Images In the middle column were taken 6 months after the initial visit I/efti. and themages on the tight were taken 6 months later. The lesion in the center of the image shows a spotof In'creased NIA at the initial VISit(arrow) with progressive reduction of NIA' tens! b . '1

. . ' In ensrtv on su sequent VISI s.In contrast. In. the. same area, AF Intensity increases from the initial to the second visit but decreases atthe last examination. '

Follow-up on the above findings is limited. In a few documented cases, increasedNIA preceded increased conventional AF, with a further increase in conventional AFas NIA decreased; rhe reduction in the NIA also preceded rhe reduction of conven-tionalAF (Fig. 6.3). This sequence of events indicates that the initial increase in NIAsignalmay be the first sign of rhe parhological process rhar leads to RPE degenerationandloss.

The NIA findings in Srargardt disease are in accordance wirh hisroparhologyfindings(see Chapter I 1G).

CHAPTER6 NEAR-INFRAREDFUNDUSAUTOFLUORESCENCE

Retinitis PigmentosaNIAfindings in RP are characteristic (Fig. 6.4). Typically, conventional AF shows aring of increased AF in an area of relatively normal AF on borh sides (see alsoChapter IIA) (53,54). NIA is usually increased wirhin this ring, with a sharp declinebeyondthis ring toward rhe periphery (9,12).

Marked abnormaliries in the distribution of NIA and convenrional AF have beenobservedin infants preceding ophthalmoscopic abnormalities. Thus, NIA and con-ventionalAF imaging can be used as an important tool to establish rhe diagnosis ofRJ' in small infanrs in whom elecrrophysiology and visual fields may be difficult toobtain.

The NIA findings in RP correspond to histological findings, which in previousreportsshowed absence of melanin in areas with nonfunctioning photoreceptors butpreservedRPE cells (41,48), whereas melanin was present in areas with photorecep-torsthat kept rheir dendriric connections (41-44). As outlined above, rod outer seg-mentphagoeyrosis induces melanogenesis (36,37). Thus, ir could be speculated thatthe absent demand of phagocytosis leads to a decline of melanosomes afrer a periodof increased phagocytosis because of photoreceptor degeneration.It has been demonstrated that the ring of increased AF observed using conven-

tionalAF imaging corresponds ro the peripheral border of preserved cone function inRP (seealso Chapter IIA) (52,53). Combining tlhe information from imaging andhistology,one could speculate that the area of increased NIA encircled by a ring of in-creasedAF, as observed using conventional AF imaging, indicates the area of pre-servedcone function, whereas the more peripheral area witlh markedly reduced NIAand normal AF corresponds to preserved RPE cells undetlying defunct cones. Moreperipherally,the absence ofNIA and AF signals indicares loss ofRPE cells.

FIGURE 6.4. Retinitis pigmentosa. A ring of increased signal with conventional FAF is observed corresponding to an areawith increased NIA in a bull's-eye shape. AF and NIA alterations are not detectable in the color fundus photograph.

..

at R . allaminar architecture in human retinitis pigmem0139. Aleman TS, Cideciyan AV, Sum~rokat, er 0 ~:;:a1mol Vis Sci 2008;49: 1580-1590.

caused by rhodopsin gene mutatlOn.s. nve~tk dP Zapfendyscrophien. Ophrhalmologe 2009;106:99_108II S K1" ik nd Diagnose er .

10. Kellner U, Ke ner. 10l U al L' f . and melanin-related fundus 3urofluorescence innoK 11 U W be' BHF et Ipa USC1I1- f'-

11. Kellner 5, e ner , e 1 .' dvs hi Am J Ophchalmoi 2009:Epub ahead of prim.rienrs with ABCA4-associated reunalal YLs~ro~ le~. and melanin-related fundus autofluorescence Vi5ualiu

II 5 W b BH et . lporuscm- .12. Kellner U, Ke ner o, e e~ ial I . . acienrs with retinitis plgmemosa. Eye 2008:Epubaheaddifferent retinal pigment epitheli a reranons In P

of print. ideci AVer 301Retinal Disease in Usher Syndrome III Caused by Mutations13. Herrera W, Aleman T, CI eciyan , I V·, S . 2008.49'2651-2660.

. h Cl . 1 Gene Invest Ophthalmo IS CI ,. . .In t e arm- '. MIA fl orescence in aduir-onset foveomacular virelliform dystrophy.14. Parodi MB, Iacono P, Pedio , er a. uro u

Retina 2008;28:801-807. I h fl '. .,. 5 K T al N -infrared and shore-wave engt auro uorescence ImagIng In en-15. Ayara A, Tatlipinar , ar ,et . ear .. 9-82

tral serous chorioretinoparhy. Br J Ophrhalmol 2009,93.7. '.. . d I' d ', ' Fe W· GL J Retinal pigment epirheiial lipofuscin an me arun an choroidal16. Welter JJ, Delori , mg , et a. ..

melanin in human eyes. Invest Ophthalmol VIS SCI 1986;27:145-152: .M D hi F D h -B -ker P et al. Age-related changes rn the morphology, absorption and17 Boulton , occ 10 , ay aw ar , ···th I' V"

' f I d lipofuscin granules of the rerinal pIgment epl e rum. men Resfluorescence 0 me anosomes an1990;30,1291-1303, .

P Th G L h TI et al Oxidation causes melanin fluorescence. Invest OphthaJmoJVIS18. Kayatz P, umann , ut er , .5,i 2001;42,241-246. , .

19. Docchio F, Boulron M, Cubeddu R, er a1.Age-related changes in the fluorescence of melamn and ltpofus-cin granules of the retinal pigment epithelium: a rime-resolved fluorescence spectroscopy study.Photochem PhorobioI1991;54:247-253. . . .H Z Z H H m'avi I et al Cutaneous melanin exhibidng fluorescence emiSSIon under near-ln'20. uang , eng ,a , .frared light excitation. J Biomed Opt 2006;11 :340 IO. . . . .

2 I. Sarna T, Burke JM, Koryrowski W, et al. Loss of melallln from human RPE WIth agmg: poSSibleroleofmelanin photooxidation. Exp Eye Res 2003;76:89-98. '.

22. Korytowski W, Pilas B, Sarna T, et a1. Photoinduced generation of hydrogen peroXIde and hydroxyl radi-cals in melanins. Photochem Pborobioll987A5: 185-190.

23. Bou1roo M. Melanin and the tetinal pigment epithelium. In: Marmor MF, Wolfensberger TJ eds, Theretinal pigment epithelium. New York: Oxford Univerisry Press; 1998:68-85.

24. Sparrow JR, Fishkin N, Zhou J, et al. AlE, a byproduct of the visual cycle. Vision Res 2003;43:2983-2990,

25. Sparrow JR, Parish CA, Hashimoro M, et al. AlE, a lipofuscin fluorophore, in human retinal pigmemedepithelial cells in culture. Invest Ophthalmol Vis Sci 1999;40:2988-2995. . ,

26. Kim SR, Jang IT, Jockusch S, et al. The all-trans-retinal dimer series of lipofuscin pigmenrs In rel1nalpigment epithelial celis in a recessive Stargardt disease model. Proc Nat! Acad Sci USA 2007;104:19273-19278.

27. Wagnieres GA, Star \X7M, Wilson Be. In vivo fluorescence spectroscopy and imaging for oncological ap-plications. Phorochem PhorobioI1998;68:603-632.

28. Feeney-Burns L, Hilderbrand ES, Eldridge S. Aging human RPE: morphometric analysis of macular,equatorial, and peripheral cells. Invest Ophthalmol Vis Sci 1984;25:195-200.

29 . Smith-Thomas L, Richardson P, Thody A], et al. Human ocular melanocytes and retinal pigment epithe-lial cells differ in their melanogenic properties in vivo and in vitro. Curr Eye Re. 15(11):1079-1091.

30. Schraermeyer U. Does melanin turnover occur in the eyes of adult vertebrates? Pigmenr Cell Res1993;6,93-204,

31. Peters 5, Lamah T, Kokkinou D, et al. Melanin protects choroidal blood vessels against light toxicity.Z Naturforsch (C) 2006;61:427-433.

32. Wang Z, Dillon J, Gaillard ER. Antioxidant properties of melanin in retinal pigment epirnelial celli.Photochem PhotobioI2006;82:474-479.

33. Boulton M, Dayhaw-Barker P. The role of the retinal pigment epithelium: topographical variarion andageing changes. Eye 2001;15:384-389.

34. Feeney L. Lipofuscin and melanin of human retinal pigment epithelium. Fluorescence, enzyme cyrochem-ical, and ultrastructural studies. Invest Ophthalmol Vis Sci 1978; 17:583-600.

35. Burke JM, Henry MM, Zareba M, et al. Photobleaching of me1anosomes from retinal pigment epithe-lium: I. Effects on protein oxidation. Photochem Phorobiol 2007;83:920-924.

36. Schraetmeyet U, Peters S, Thumann G, et al. Melanin granules of retinal pigment epithelium are con-nected with the lysosomal degradation pathway. Exp Eye Res 1999;68:237-245.

37. Peters $, Kayatz P, Heimann K, et al. Subretinal injection of rod omer segments leads to an increase in wenumber of early-stage melanosomes in retinal pigment epithelial cells. OphthaLnic Res 2000;32:52-56.

38. Schraermeyer V, KopitzJ, Peters S, et al. Tyrosinase biosynthesis in adult mammalian retinal pigmenrep-ithelial cells. Exp Eye Res 2006;83:315-321.

39. Julien S,. Ko~iok N, Kreppel F, et aJ. Tyrosinase biosymhesis and trafficking in adult human retinal pig-ment eplthehal cells. Graefes Arch Clin Exp Ophthalmol 2007;245: 1495-1505.

40. Buch.anan T.A. Gardiner TA, Archer DB. An ultrastructural study of retinal photoreceptor degenerationassociated WIth bronchial carcinoma. Am J OphthalmoI1984;97:277_287.


CHAPTER 6 NEAR-INFRARED FUNDUS AUTOFLUORESCENCE 69

11

i

41. Szamier RB, Berson El.. Retinal ultrastructure in advanced retinitis pigmenrosa. Invest Ophrhalmol VisSci 1977;16;947-962.

42. Szamier RB, Berson EL, Klein R, er al. Sex-linked retinitis pigmenrosa: ultrastructure of phororecepeorsand pigment epithelium. Invest Ophrhalmol Vis Sci 1979; 18: 145-160.

43. Bum-Milam AH, Kalina RE, Pagan RA. Clinical-ultrastructural study of a retinal dystrophy. InvestOphthalmol Vis Sci 1983;24:458--469.

44. Sz.amierRB, Berson EL. Retinal hisroparhology of a carrier of X-chromo some-linked retinitis pigmemosa.Ophthalmology 1985192;271-278.

45. Steinmetz RL, Garner A, Maguire JI, et a]. Histopathology of incipient fundus fiavimacularus.Ophthalmology 1991198;953-956.

46. Lopez PF. Autosomal-dominant fundus flavirnacularus. Clinicopathologic correlation. Ophthalmology1990197]98-809.

47. Sarks JP, Sarks SH, Killingsworth MC Evolution of geographic atrophy of the retinal pigment epithe-lium. Eye 1988;2:552-577.

48. Milam AH, Jacobson 5G. Photoreceptor rosettes with blue cone opsin immunoreactivity in retinitis pig-menrosa. Ophthalmology 1990;97: 1620-1631.

49. Eagle Ir RC, Lucier AC, Bernardino Jr VB, et al. Retinal pigment epithelial abnormalities in fundus flav-imacularus: a light and electron microscopic study. Ophthalmology 1980;87:1189-1200.

50. Delcri FC, Scaurenghi G, Arend 0, er al. In vivo measurement oflipofuscin in Stargardr's disease-s-fundusflavimecularus.Jnvest Ophrhalmol Vis Sci 1995;36:2327-2331.

51. Lois N, Halfyard AS, Bird AC, er a]. Fundus autofluorescence in Srargardt macular dystrophy-fundus flav-imacularus. Am J Ophchalmol 2004;138:55-63.

52. Robson AG, Saihan Z, Jenkins SA, et al. Functional characterisation and serial imaging of abnormal fun-dus autofluorescence in patients with retinitis pigmentosa and normal visual acuity. Br J Ophrhalrnol2006;90,472--479.

53. Popovic P, jarc- Vidmar M, Hawlina M. Abnormal fundus autofluorescence in relation to retinal functionin patients with retinitis pigmenrosa. Graefes Arch Clin Exp Ophrhalmol 2005;243: 10 18-1 027.

54. Vaclavik V, Vujosevic S, Dandekar 5S, er al. Autofluorescence imaging in age-related macular degenerationcomplicated by choroidal neovascularization a prospective study. Ophthalmology 2008;115:342-346.

55. Kellner U, Renner AB, Tillack H. Fundus autofluorescence and mfERG for early detection of retinalalterations in patients using chloroquine/hydroxychloroquine. Invest Ophthalmcl Vis Sci 2006;47:3531-3538.

CHAPTER

Interpreting Fundus Autofluorescence

Interpretation of autofluorescence (AP) images with illu~i~ation at 4~8 n~(standard AF image) is based on the principle that the maJonty of the signal IS

derived from lipofuscin in the retinal pigment epithelium (RPE) (see Chapter3) 0-3). Although there are many fluorophores in the retina, the signal from RPElipofuscin is stronger than that from any other substance. The quantity of lipofuscin 1.0

the RPE is a balance between accumulation and clearance. Lipofuscin is formed of reo-nal and ethanolamine in the photoreceptor outer segment and is ingested by the RPEby phagocytosis of shed outer segment material (see Chapter 2) (4,5). Thus, accumu-lation of lipofuscin is driven by outer segment renewal. Clearance of Iipofsucin may bedue in part to discharge of long-term phagolysosomes by the RPE into the extracellu-lar space (see Chapter 2) and in part to photodegradation (6), although the halflife ofthe fluorophores is unknown. Thus, the "background" levels of AF in the normalhealthy eye reflect a normal photoreceptor outer segment turnover and retinoid cy_cling. A reduction in the number of phororeceptors causes loss of AF over time.Increased levels of AF are caused by RPE dysfunction either because of an intrinsic fail-ure of lipofuscin clearance or the presence of an abnormal metabolic load.

Imaging of AF depends on the clarity of the media. Extravascular blood internalto the RPE will prevent detection of the signal, and nudear sderosis will attenuate il-lumination. Under other circumstances, the variation reflects the lipofuscin content ofthe RPE. The image of a normal fundus with illumination at 488 am shows homoge-neous AF over most of the posterior pole (see Chapters 2 and 9) (2). The blood vesselsand fovea are dark because of light absorption by blood and luteal pigment, respec-tively, since both absorb short wavelength light. The optic disc is dark because of theabsence of a fluorophore. The AF level increases with age and tends to be highest atabout 10 degrees of eccentricity (see also Chapter 3). Deviation from this pattern indi-cates outer retinal disease. Current imaging techniques with commercially availablesystems are very valuable for detecting abnormalities in the distribution of AF, but arerelatively unreliable for measuring absolute levels of AF (see Chapter 5). Describedbelow are the circumstances in which evaluation of fundus AF is the most valuable.

CONTINUITY OF AUTOFLUORESCENCE (NORMALAUTOFLUORESCENCE PATIERN)In the presence of unexplained loss of vision, the existence of normal AF with illumi-nation of 488 urn usually implies that the Joss of vision is unlikely to be due to Outerreti~al disease. Howev~r, th~re are cer~alncircumstances in which normal AF is pres-ent In c.asesof outer. retinal disease. ThiS was observ~d in a study of patients with Lebercongenital amaurosis (7) (see also Chapter lID) (Fig. 7. I), in which it was concluded

CHAPTER7 INTERPRETINGFUNDUSAUTOFLUORESCENCE 71

FIGURE 7.1. AF images of the right and left eyes of a 16-year-old female with Leber congenital amaurosis whohadlight perception vision throughout life and a flat electroretinogram in the first year of life.

that photoreceptor cell dysfunction occurred wirhout population loss. The loss of vi-sion may be due to transduction failure or constant noise, i.e., the photoreceptor cellsmay behave as if they are in constant light, thus reducing signal-to-noise ratios. If itwerepossible to correct the metabolic abnormality, AF imaging suggests that such pa-tientsmight be able to recover vision. A similar situation of loss of function with nor-mal AF signal was found in acute zonal and occult outer rerinopathy (AZOOR), atleast in the first 5 years of disease, implying the possibility of spontaneous recovery (8).In some patients with retinitis pigmentosa (RP), there is a ring of increased AF at vari-able eccentricity around rhe fovea (see also Chapter llA) (Fig. 7.2) (9,10). On eitherside of the ring of increased AF, AF is normal but scotopic function is markedly re-duced; photopic function is also reduced in areas external to the ring ofincreasedAF.There is no explanation as to why the function is poor in the presence of a normal pop-

FIGURE 7.2. AF image of a 32-year-old male with RPshowing a ring of increased AF around the fovea

andcystoid macular edema.


ulation of photoreceptor cells unless the half life of rhe fluorophores is verylong.Curiously, near-infrared AF (NIA) coincides with funcnon (see also Chapter 6).

AF is also very important for determining the integrity of me outer retinaat[herime of treatment of choroidal neovascularizarion CCNY). Until recently, poor visualacuity was interpreted as indicating loss of central photoreceptor cells, but recovery ofvisual acuity following inrravitreal injection of antivascular endothelial growth faCtor

(VEGF) agents implies that this is not the case. Further suPPOrt for the conceprrhatthe outer retina may be physically intact comes from the observation of normal AFinsuch cases (see also Chapter lOB) (Fig. 7.3) (11,12). AF imaging would he of greatvalue for assessing the likely therapeutic benefit of such treatments.

INCREASED AUTOFLUORESCENCEIn many disorders, pale areas appear at the level of the ourer retina such as drusenandscarring. It is important to determine whether these areas represent lipofuscin inme

FIGURE 7.3. fluorescein angiogram (AI and Af image (81 f . .visual acuity of 6/36. showing CNV and absent Af t II fOI a patiant With an 8-month history of visual symptomsanda

. cen ra y. uOIescernanglog (e) d . .'It-month history of visual symptomsand a visual acuit f 6/60 . ram an Af Image (DJ of a patientWithan

yo, showing CNV and mtact Af centrally.

CHAPTER 7 INTERPRETING FUNDUS AUTOFLUORESCENCE

FIGURE 7.4. AF image of a 24-year-old female with Stargardt-fundus flavimaculatus and a visual acuityof6/60.showing focal increased AF and discontinuous atrophy around the fovea

RPEor in bloodborne macrophages. The distinction is easily made by determiningwhether these lesions auwfluoresce on AF images.

The pattern of increased AF may be characteristic of a disorder. Macular dystro-phiesmay present with focal lesions rather than continuous abnormalities. The for-merarecharacteristic of Stargardt disease (see also Chaptet II G) (Fig. 7.4), in whichthe initial change is discrete spots of increased AF that correspond with pale lesionsseenon ophthalmoscopy, with totally normal AF elsewhere (13,14). As the disorderprogresses)atrophy occurs at the sites of increased AF and new spots of increased AFappear.Focal increased AF with normal intervening levels is also seen in pattern dys-trophies(see also Chapter l l E) (Fig. 7.5).

Increased AF seen in an area of the fundus that appears normal occurs in avarieryof disorders and may be of diagnostic value. In bull's-eyc macular dystro-phies, a continuous ring of increased AF is seen early, followed by photoreceptorcellloss in a bull's-eyc pattern (see also Chapter lIB) (Fig_7.6) (13,15). There are

FIGURE 7.5. AF image of a 55-year-old female with 6/9 visual acuity and pattern dystrophy, showingfocalincreasesof AF and normal intervening AF Patches of reduced AFcorresponding to areas of atrophy

are also seen.


FIGURE 7.6. AF image of a 36-year-old male with bull's-eye maculopathy and a ring of atrophyalOundthe fovea, and increased AF at the perimeter of the area of atrophy.

no ophthalmoscopic abnormalities that correspond with the ring of increasedM.In retinal degenerative disease resulting from the 172 RDS mutation, the ~enrr~macula has increased AF at a time when there are no symptoms, the fundus IS nor-mal by ophthalmoscopy, and elecrrophysiological responses are normal (seealsoChapter lIB and E) (Fig. 7.7). .

In early age-related macular disease, focally increased AF occurs In some cas~andappears to be indicative of a risk of geographic atrophy, rather than eNV, asa hkelycause of visual loss (see also Chapter lOB and C) (16-18). In geographic atrophy,thepresence of increased AF implies likely progression of the atrophy. Drusen inage-related macular disease do not fluoresce brighrly (see also Chaprer lOA),whichIS Inmarked contrast to drusen seen in the young or as part of a monogenic disordersuchas Doyne honeycomb dysrrophy (Fig. 7.8).

REDUCED AUTOFLUORESCENCEReduced AF indicates loss of phororeceprors-c-or at least their outer segments, orRPE loss. Outer retinal atrophy is not always obvious on biomicroscopy. burisun-mistakably recognizable on AF imaging. It may have a distinctive distribution, suchas in disease associated with an A3243G mitochondrial mutation in which me lesionsare distributed and orientared in a circumferential fashion (see also Chapter IIH)(Fig. 7.9) (19). In Srargardt disease the arrophy is spotry (see also Chapter IIG)(Fig. 7.4), whereas it is continuous in bull's-eye lesions (Fig. 7.6).

IRREGULAR AUTOFLUORESCENCEDiffusely irregular AF indicates disease at the level of the outer retina; it is seen inava-riety of disease states and often has a distinctive distriburion. The A3243G mirochon-drial mutation (Fig. 7.9) and the 172 RDS murarion (Fig. 7.7) usually cause irregularAF associared with atrophy. Diffusely irregular AF also occurs in cenrral serousretinopathy after the first 6 months of detachment (see also Chapter 13) (20).lrsrhought that this IS due to outer segment shedding into rhe subrerinal space rharmaycollect injerjorly because of gravity and phagocytosis by the RPE or possiblyothermacrophages. Focal changes also occur that ofren correspond with rhe point ofleakageon fluorescein angiography.

CHAPTER7 75INTERPRETINGFUNDUS AUTOFLUORESCENCE

a

FIGURE 7.7. (A) Pedigree of a family with a dominantly inherited macular dystrophy resulting from mutation 172 in the RDSgene. (8) The proband had irregular AF occupying the whole of the posterior pole Her 16-year-old son had no symptoms,andnormal fundi and electrophysiological responses. (Cllncreased AF centrally showed that he was affected by the disorder.

FIGURE7.8. Reflectance (AI and AF images (8) of Doyne honeycomb dystrophy showing bright AF of drusen.

...


(AI d AF (8)· of disease associated with an A3243G mitochondrial mutation in a 50-yeal-01aFIGURE 7.9. Reflectance an Imagesfemale with deafness. Atrophy is associated with limited irregular AF.

SUMMARYFor the first time, it is possible to image changes at the level of the RPE that are inte-gral to outer retinal diseases. Although experience .with AF irnagi~g is short, it is~i-dent that the technique is clinically useful and at times can be an tmportanr facrormmanagement decisions.

REFERENCESL Delori FC, Dorey CK, Sraureuthi G, er al. In vivo fluorescence of the ocular fundus exhibits retinalpig-

ment epithelium lipofuscin characteristics. Invest OphrhaJmol Vis Sci 1995;36:718-719.2. von Ruckmann A, Finke FW, Bird AC. Distribution of fundus autofluorescence wirh a scanning laseroph-

thalmoscope. BrJ Ophthalmol1995;79:407--412.3. Eldred GE, Karz ML. Fluorophores of the human retinal pigmenr epithelium: separation and specrrsl

characterization. Exp Eye Res 1988;47:71-86.4. Jang Yl', Matsuda H, Itagaki Y, er al. Characterization of peroxy-A2E and furan-A2E phorcoxidaree

products and detection in human and mouse retinal pigment epithelial cell lipofuscin. J Bio]Chern2005;280;39732-39739.

5. Bui TV, Han Y, Radu RA, et al. Characterization of native retinal fluorophores involved in biosynthesisof A2E and lipofuscin-associated retinopathies. J Bioi Chem 2006;281: 181 12-181 19.

6. Fernandes AF, Zhou J, Zhang X, et al. Oxidative inactivation of the proteasome in retinal pigment epithe-lial cells: a potential link between oxidative stress and up-regulation of interleukin 2. J Biol Chern 2008;283;20745-20753.

7. Scholl HPN, Chong NHV, Robson AG, er al. Fundus autofluorescence with Leber congenital amaurosis.Invest Ophthalmol Vis Sci 2004;45:2747-2752.

8. Schmitz-Valckenberg S, Holz FG, Bird AC, er al. Fundus autofluorescence imaging: review and perspec-tives. Retina 2008;28:385--409.

9. Robson AG, EI-Amir A, Bailey C, et aI. Pattern ERG correlates of abnormal fundus autofluorescence in pa-tients with retinitis pigmentosa and normal visual acuity. Invest Ophthalmol Vis Sci 2003;44:3544-3550.

10. Ro~son AG, Eg~n CA, L.uon~ VA, ~t aI. Comparison of fundus autofluorescence with photopic and sco-topic fine-matnx mappmg 111 panenrs with retinitis pigmentosa and normal visual acuity. InvtS[Ophthalmol Vis Sci 2004;45:4119--4125.

11. Dandekar 55, Jenkins SA, Peto T, et aI. An analysis of autoflorescence of choroidal neovascularizarion dueto age-related macular disease. Arch Ophthalmol 2005; 123: 1507-] 513.

12. Vacl3~ik V, Vujosevic. S, Dandekar S~, e~ a]. Autofluorescence imaging in age-related macular degenemlc'complicated by choroidal neovascularizarion a prospective study. Ophthalmology 2008;]15:342-346.

CHAPTER 7 INTERPRETING FUNDUS AUTOFLUORESCENCE 77

13. von RUckmann A, Pirzke FW, Bird AC. In vivo fundus autofluorescence in macular dystrophies. ArchOphlh,lmoI1997;115,609~615.

14. Lois N, Holder GE, Bunce C, et al. Incrafamilial variation of phenotype in Srargardr macular dystrophy-fundus flavimaculacus. Invest Ophthalmol Vis Sci 1999;40:2668-2675.

15. Kurz-Levin MM, Halfyard AS, Bunce C, er al. Phenotypic assessment of bull's eye maculoparhy. ArchOphlh,moI2002;120,567~575.

16. von Riickmann A, Firzke FW, Bird AC. In vivo fundus autofluorescence in age related macular degenera-tion. Invest Ophrhalrnol Vis Sci 1997;38:478-486.

17. Holz FG, Bellmann C, Margariridis M, et al. Patterns of increased in vivo fundus autofluorescence in thejunctional zone of geographic atrophy of the retinal pigment epithelium associated with age-related mac-ular degeneration. Craefes Arch Clin Exp Ophrhalmol 1999;237: 145-152.

18. Hob: FG, Bellman C, Staudt S, et al. Fundus autofluorescence and development of geographic atrophy inage-related macular degeneration. Invest Ophrhalrnol Vis Sci 200 I ;42: 1051-1056.

19. Michaelides M, Jenkins SA, Barniou DE, ct ai. Macular dystrophy associated with the A3243G mirochon-drial DNA mutation: distinct retinal and associated features, disease variability, and characterization ofasymptomatic family members. Arch Ophrhalrnol 2008; 126:320-328.

20. von Ruckmann A, Firzke FW, Fan J, et a]. Fundus autofluorescence in central serous retinopathy imagedwith a laser scanning ophthalmoscope. Am J Ophthalmo12002;133:780-786.

CHAPTER

Quantifying Fundus Autofluorescence

INTRODUCTIONThe investigation of endogenous fluorophores has the potential to aid in evaluatingthe metabolic status of tissues and thus detecting the early stages of disease (1,2). Theredox pairs of fluorophores, the coenzymes NAD-NADH (oxidized and reducedforms of nicotinamide adenine dinucleotide) and FAD-FADHz (oxidized and re-duced forms of flavin adenine dinucleotide) are electron transporters in the basicprocesses of cell metabolism (Fig. 8.1). NAD and FAD participate in reactions thatoccur in J3-oxidarion of farry acids (acyl-CoA-dehydrogenase reaction), in glycolysis,in the citrate acid cycle (succinate dehydrogenase reaction), in the respiratory chain incomplex I (NADH+ +H+ ubichinone reducrase) reactions, in the succinate dehydro-genase reaction in complex II reactions, and in the connection between the citrateacid cycle and the respiratory chain (3).

It appears that rhe dominant f1uorophore of the fundus is rhe aging macular pig-ment lipofuscin, which appears ro play an important role in the pathophysiology ofseveral retinal diseases, including age-related macular degeneration (Al\.1D) (see alsoChapters 2 and 3). Other fluorophores include advanced glycation end-products(AGEs), which ate involved in the pathogenesis of diabetes mellitus; elastin and col-lagen, which change during sclerotic processes and in glaucoma; pyridoxal phosphate,the prosthetic group of all amino transferases; proroporphyrin IX, a tluorophore inhem synthesis; and the amino acids tryprophan, kynurenin, and phenylalanine, whichare strong fluorophores in connective tissues and the cornea, lens, and sclera (4) (seealso Chapter 3).

Since an evaluation of the metabolic status of tissues requires information on sin-gle fluorophores, however, the separation of individual signals corresponding to eachof these f1uorophores from within the sum of the fundus autofluorescence (AF) signalrepresents a challenge. This chapter reviews possible methods to achieve that goal, aswell as their limitations.

METHODS FOR DISCRIMINATIONOF FLUOROPHORES

According to the chemical structure, there are three characteristic properties-the ex-citation spectrum, the emission spectrum, and the fluorescence lifetime after excitationby short pulses~that allow discrimination offluorophores (5). In the eye, the spectralrange for exarmnanon IS limited to between 400 and 900 om by the transmission ofthe ocular me?ia in the short-wave ra.nge and the absorption of water in the long-waverange, respectively (6). The decrease 10 the ocular transmission, which occurs with in-creasing age) can be determined (7).

78

CHAPTER 8 QUANTIFYING FUNDUS AUTOFLUORESCENCE

RedoxpairsSufficientOxygen

Lack ofOxygen

NAD -NADH_ no fluorescence high fluorescence ofNADH

FAD _FADH2 _ high fluorescence ofFAD no fluorescence

FIGURE8.1. Change of fluorescence of coenzymes NAD and FAD depending on the availability of oxygen.

Since specific excitation maxima are shorter man 400 nrn, a differentiation of thefundus fluorophores is limired according to the excitation spectra. If the sample can bestrongly excited and the fluorescence signal can be detected with a high signal-ro-noiseratio, the contribution of single fluorophores can be recalculated in the sum emissionspectrum. In such a fitting procedure, the spectral course of the emission spectra ofeach fluorophore musr be known. Weighting facrors are optimized until the approxi-mation of the model function and the measured sum fluorescence spectrum are opti-mal. These weighting factors correspond to the contriburion of each fluorophore tothe sum spectrum. The signal-co-noise ratio is equal to the square root of the collectednumber of phorons. To obtain a high number of photons, the irradiance must be highor the measuring time will be long. Generally, spectral intensity measurements sufferfrom absorption and scattering in nonfluorescent substances. which weaken the exci-ration light and change rhe specrral shape of the emission spectrum.

The fluorescence lifetime. the third parameter for discrimination of fluorophores,hassome important advantages. The lifetime is independent of the absorption or scat-tering of neighboring substances. There is also no dependence on the concentration offluorophores. Weakly emirring fluorophores can be separated from strongly emittingfluorophores if the liferimes are sufficiently different. Only about one hundred photonsare necessary for approximation of a monoexponenrial decay (8). Because the lifetimedependson pH value and viscosity, information can be found related to the cellular em-beddingmarrix.

Limiting Conditions for MeasuringFundusFluorophoresThere are some limiting conditions that must be taken into account 10 rwo-dimensional (20) measurements of fundus fluorophores in vivo:

1. The fundus fluorescence is covered by the strong fluorescence of the anterior seg-ment of the eye. The techniques of aperrure division and confocal laser scanning canreduce the influence of fluorescence of, cornea and lens (see also Chapters 3 and 5);a combinarion of both methods is advisable for accurate measurements (9). Theemission spectra of the lens and cornea are minimal for wavelengths longer than570 nm (10), and thus their influence will be reduced when long waves are used.

2. Endogenous f1uorophores are present in all layers of rhe fundus.3. The strong fluorescence of lipofuscin covers the weak fluorescence of other

fluorophores. .4. Eye movement limits the time available for highly spatially resolved measurements.5. The maximal permissible exposure (MPE) is the most imporrant limitation for spec-tral measuremenrs in the eye (11).

Z(w) ~

" (Y,' W' 7.2~, ,"""1+w2'T2t=l i [51

[II


Fluorescence Lifetime Measurements" " best suited for detection and discrimination of fluLifetime measurements are d for derecri ,

.. f h c d s There are two merho s ror etectlng fluoresce,rophores 111Images 0 t e run U . . _L . ce. th c ency domain or measurement In me nme domain I)',decay: measurement In e lrequ

Fluorescence Lifetime Measurementsin the Frequency Domain .. .I h c. d ain the sample is excited by intenSIty-modulated Iighl T'n t e lfequency om , ..' Itt

emitted fluorescence light is also modulated. There rs a phase shift berween ex(i~tionand emission light because of ehe delay between absorption and errussron This ph",shift depends on the modulation frequency. The rnodularion In the fluorescenceli~1is also weaker than the excitation light and decreases With Increasmg modulationfrequency. .. . . .

From the phase shift, the fluorescence Iiferime "P IS determined according tothefollowing equation:

tan <I>7=---p W

where <P = phase shift and w = modulaeion frequency.The lifetime Te» as determined from demodulation measurements, is caJcularedbr

~ 1 -1m' [2)w

where m = demodulation.The demodulation m is determined by rhe rario of maximal and minimal inrensiry

in ehe emitted lighr, divided by the ratio of the maximal and minimal intensiry in theexcitation lighc. The same values ('fp = 'fml are deeermined in monoexponenrialdecay,but in multiexponential decay they are 'fp < 'f m- In rhe case of muleiexponencial decay,rhe decay times are calculated according CO the complicaeed sec of formulas (12):

tan <t>(w)= Z(w) [31N(w)

m(w) =~ N(w)' + Z(w)' [4J

[6)

i=1

!he 1.iferimesTj and the pre-exponential factors <Xi of the component i are calcularedrterativejy, not analytically.

. In ~rinciple.' one. can. determine the fluorescence lifetimes simuiraneousiy forall'mage pixels by illuminating the whole image and detecting phase shift or demodula-t1~n usmg a detector ma.tflx. Sequential excitation of all image pixels can be achievedWIth the .use of a scannmg system with only one detector. A combination of bothmethods IS also possible.

CHAPTER 8 QUANTIFYING FUNDUS AUTOFLUORESCENCE tI1

Fluorescence Lifetime Measurementsin the Time DomainFluorescence lifetime m~asurements are performed in the time domain by excitingthesampleby pulses havmg a shorr full width at half maximum (FWHM), e.g., in thepico- or femtosecond range. Because the decay of the fluorescence is too short for di-rect measures to be obtained, the fluorescence intensity is detected during multipleexcitations in a number of rime windows, in a single variable time window (boxcarprinciple), or by time-correlated single photon counting (TCSPC). In TCSPC, thesampleis excited by weak light pulses. The intensity is so low that only one photonwill be detected in a sequence of about 10 excitation pulses. Each fluorescencephoton will be accumulated in time channels according to the detection time afterthe excitation pulses. After a sufficient measuring time, the content of all time chan-nelsrepresents rhe probability density function of the decay ptocess. The principle ofTCSPC, technical details, and pracrical applications were previously explained byBecker(13).

The process is mostly assumed as an exponential decay, which can be approxi-matedby a sum of e-functions:

I(t) ~~ a .. e-t + bI£""o ;=1

[7]

where (Xi = preexponentia] factor of exponent i or amplitude, 't, = lifetime of expo-nent i, b = background, and p = degree of exponential function.

Because the excitation is not a Dirac pulse, the measured decay of fluorescence isthe convolution of excitation pulse with the decay process. Therefore, the criterionfor the fitting process is the minimization of Xr2:

1 " [N(t) - N(t)]'2 ~ J r )

X, = n- q . £... N(t.))=1 J

[8]

In thisequation, N(t) is the measured number of photons in the rime channel j; N,('!)isthe number of expected photons, which are calculated by rhe convolution of rhe in-srrumental response function and the model function; n is the number of time chan-nels;and q is the number of free parameters (o.; Ti' b).

If the detection of photons is a Poisson process, the mean square root error be-tweendetected photons and calculated photons is equal ro rhe square root of the de-tected events:

Noise ~J[N(t) ~ N(t)]' ~JN(t)) r ) )[9]

Thus, the rario in the sum of Eq. [8] is one for each time channel and rhe sum is n.Thismeans rhat the limiting value of xc' is one. The algorithm is independent of thedegree of exponenrial function, but the calculation time increases with the number ofexponents_ The time between two excitation pulses should be about five times thelongestexpected decay time.

The separation of fluorophores can be improved by global analysis (14). In thisanalysis, at least two data sets are considered that contain the same fluorophores, char-acterized by the decay time, but have different contributions a;

In addition to discrete e-functions, other models can describe the distributionof lifetime in complex sysrems, including the lifetime distribution (15), strerched e-funcrion(16), time-resolved area-normalized emission specrroscopy (17), and Laguerreexpansion techniques (18).

a.' 7"Q = -p;;-'-' ---"-

""''''7..... ' ,[II]


. f iifeti measurements, in addition to the single amplitud"an~For evaluatIon 0 rrenme ib ' nlif time 7 and relative conrn urron -o are helpfullifetimes, the parameters mean J e mean .

Here the mean lifetime is defined as:p

""''''7..... ' ,Tmean ---"=-'':-~---

L",~l [~

Th I· t ibution n,of the component i corresponds to the area undenhee re atrve con fl <ad dererrni d by the component i. This value is calculated according to:eeay curve, eterrrunc

;=1

Two-dimensional lifetime measuremen ts can be performed in scanning systems,with a pulse laser used as the light source. In contrast to applications in microscopy,the movement of the object eye must be detected and compensated for. The correcrelation between detected photons and the position in the sample can be determinedby the control signals, the frame and line clocks, and the (line) timer signal.Thetime between the excitation pulse (sync signal) and the detection time of the firstand unique photon is used for allocation in the time channel. If there are multipledetectors in different spectral channels, differences in the fluorescence spectra canbeused for further differentiation of f1uorophores, e.g., by global fitting. In addition,fluorescence spectra can be reconstructed from the photons detected in each spectralchannel.

Starting from the MPE of the eye, lifetime measurements in the time domainwere optimally suited for measuring the dynamic AF of the fundus (19). Only a weakexcitation power is required for detection of signal in TCSPC. This technique off,"the most sensitive detection of light. To ensure a high number of photons, whichisrequired for a good signal-to-noise ratio, the photons are added from a series of sin.gle image measurements after image registration.

STUDIES ON ISOLATED FLUOROPHORES

Excitation and Emission Spectra of ExpectedFundus FluorophoresKnowledge of the specific excitation and emission of expected fluorophores is re-quired for selected investigation and interpretation of time-resolved AF images. In ad.dition, the transmission of the ocular media must be taken into account, transform-ing this spectral information for studies at the fundus. In the study described below,ocular transmrssron data from a 20-year-old person were used. The excitation andemission spectra were measured with a Fluorolog (SPEX; Jobin Yvon LongjwneauCedex, France) after the absorbaney of the substances was adjusted to about 0.08.

In the excrrauon spectrum of NAD, the maximum is at 350 nrn and the emis-sion is maximal at 450 nm. Although NADH is a strong fluorophore in humantissue, the. ocular transrntssron blocks nearly all excitation at the fundus, and thusthe detection of NADH in living human fundus is unlik I I ' the ex-. . e y. n contrast, Incitation spectrum of FAD (Fig. 8.2) there are two maxima at 370 nm and 446 nrn.The excitation at 350 nm results in maxima of fluorescence at 441 nm and 520 nrn.For mvesnganon of the fundus, only the excitation around 446 nrn is relevant.

CHAPTER 8 OUANTIFYING FUNDUS AUTDFLUORESCENCE

1,2,~~~~--=~-:---~~~~~~~~~~~,Emission

)'ex= 350 nm

>- 0,8'"'"c:(])~ 0,6c:

E~0 0,4c

0,2

0300 350 400 450 500 550 600 650 700

EmissionIcex= 446 nmExcitation

ocular Transmission

fiGURE 8,2. Excitationand emission spectra of FAD.

Wavelength in nm

Excitation at both 446 om and 468 nrn results in the same maximal emission spec-trumat 526 nm. The area under excitation spectrum and ocular transmission is largeenough to excite FAD at the fundus in living eyes.

Lipofuscin is' the most intensive emitting fundus fluorophore in elderly humans(20,21) (see also Chapter 3). It consists of 10 fluorophores (22) (see also Chapter 2).The component VIII has been identified as N -rerinylidene- N -retinylethanolamine(A2E) (23), a by-product of the visual cycle that accumulates in the retinal pigmentepithelium (RPE) throughout life (24) (see also Chapter 2). Figure 8.3 shows the

1,2Emission

Excitation A ex = 446 nm

ocular Transmission

c-,.'" 0,8'"c(])~c 0,6E0 0,4c

0,2

a300 400 500 600 700 800

Wavelength In nm

fiGURE 8,3, Excitationand emissionspectra of AlE in relationto the ocular transmission.

excitation and emission spectra of AlE (12.8 mM, dissolved in ethanol), synrhesizeda;cording to Parish et al, (25). In the visible spectral range, the maximum excitationofAlE is at 441 nrn, An excitation at 446 nm results in maximal fluorescence at 600nm.Because the area is large under the curves of excitation and ocular transmission, A2Ecanbe excited at the fundus, which confirms practical experiences in AF measurements.

AGEs ate a mixture of different glycolyzed proteins. In a human sample (Fig.8.4), excitation at 446 om resulted in an emission maximum at 502 nm. Theex-tended area under the excitation curve and ocular transmission make the in vivoex-citation possible. A comparison of the excitation spectrum of melanin in relationrothe transmission of the ocular media shows that the in vivo excitation of detectablemelanin fluorescence at the fundus is unlikely (Fig. 8.5).

Ocular structures contain connective tissue composed of collagens and elastin,among other structures. Taking into accounr the strong increase of the ocular trans-mission starting from 400 nrn, the detection of fluorescence of connective tissuecan-not be excluded at the fundus, which is excited in the visible spectrum (Fig. 8.6).

Considering the excitation spectra of isolated substances in relation to the trans-mission of the ocular media, the effective excitation spectrum should be calculated asaproduct of the excitation spectrum and spectral ocular transm.ission. Ir is done forthefluorcphores with the highest excitation probability (AlE, FAD, and AGE). In th.way, a certain discrimination of fundus fluorophores can be achieved according to thespectral range of excitation (Fig. 8.7). Furthetmote, endogenous fundus fluorophorecan be discriminated accotding to the emission spectra (Fig. 8.8).

Fluorescence Lifetime of ExpectedFundus FluorophoresIn addition to the excitation and emission spectra, the fluorescence lifetime is a distin-guishing feature. The lifetimes of diffetent fluorophores, estimated using a published

1

Fluorescence Spectrum(Excitation at 360 nm)

0-P---r----r--~--~-~----""'";_--r_-___1300 350 400 450 500 550

Wavelength in nm

600 650 700


1,2 ~~=-:-;::-~~~~~~~~~~~~~~-Excitation Spectrum

(Fluorescence at 436 nm)

ocular Transmission~ 0,8 -r------/'---~---\------=-~-------__Ic:Q)-c: 0,6 i--+---+-------;A---------------jE~o 0,4-r-----I---\----I-------'I;-----------------1c:

0,2 -j-+--__ --l.-+- ~--------__I

FIGURE8.5. Excitation and emission spectra of melanin in relation to the ocular transmission

0,2 +\------------------------1

Collagen 2

Collagen 4

~ 0.15+---'t;-------------------------1c<Il.0~o~ 0,1<{

Collagen 3

Collagen 1

0.05~/ ~~~

ElastinO~-----,----,----~-----r----I300 400 500 600 700 BOD

Wavelength in nm

FIGURE8.6. Absorbance spectra of coliagens 1-4 and elastin.

...

SECTIONI BASICSCIENCE

0,6

0,5

>, 0,4..",<f)

C(J)~ 0,3c

E~0 0,2c

0,1

a400 450

FAD x ocular Transmission

AGE x ocular Transmission

A2E x ocular Transmission

500 550 600Wavelength in nm

650 700

FIGURE 8,7. Effective excitation spectra as a product of the excitation spectrum and ocular transmis-sion for A2E, FAD,and AGE.A2E, FAD, and AGE are excited by light in the short-wave range belween400 nmand 500 nm; predominantly A2E is excited by wavelengths longer than 500 nrn.

0,9 T-------r-----+'\-'rl---:f-------"\:c-----------JAGE

0,8T--~c::_I'-----+----T---W"------+----- __ ---J0,7I--T-t---v-t------.:b;o--""'~---_I0,6j--t----T-ntT----~1__----_I0,5 j-I'-------t----7---.:-f\;-\---------\----_~0,4 j---t--j---t--j--\--\;---------.J\-- _

0,3 t-~7""'----t-t-+-~""---~---_____.jFAD

0,2 1-------t-j"---.:---+-4~-------l,,---__i0,1

°4~00~--"'t"----5'00----L--6'0-0-----..:::::'=",70T-O-----=:::::~800

f~oc

Wavelength in nm

FIGURE 8,8. Selectedemission ranges for discrimination of A2E FAD a dAGE I h h .sian rangebetween 450 nm and 560 nm AGE FAD "n. n t e sort-wave erms-560 nm is dominatedby A2E. '" and A2E are detectable, but the emission range above


Lifetimes and Amplitudes of Isolated Fluorophores Expected atthe Fundus

Substance 71 in ps 0:'.1 in% 1'2 in ps cxzin%

Cullagen1 670 68 4040 32Cullagen2 470 64 3150 36Cullagen3 345 69 2800 31Collagen4 740 70 3670 30Elastin 380 72 3590 28A2E 170 98 1120 2Melanin 280 70 2400 30fAO 330 18 2810 82AGE 865 62 4170 28NAOH 387 73 3650 27

method (26), ate given in Table 8.1. This property (lifetime) is especially interestingfor investigationsof the macula behind the absorbing macular pigment. The absorp-tionofmacular pigment changes both the intensity and the spectrum of emitted lightoriginatingfrom the RPE. In contrast, the lifetime of the RPE fluorescence stays un-changed.Because the lifetime depends on viscosity and pH, information related topropertiesof the embedding matrix in the tissue can be obtained. In a less viscous en-vironment, the molecules display internal rotation and charge transfer, which resultsin radiationlessdecay. Ao a result, the quantum yield and the lifetime depend on vis-cosity.Fluorophores can have both a protonated and a deprotonated form. Bornformshave different lifetimes. The average lifetime is related to the equilibrium ofboth forms, which depends on the pH of the local environment.

STUDIES ON OCULAR TISSUES

Excitationand Emission Spectra of Ocular TissuesTo evaluate the excitation and emission spectra of ocular tissues, the cornea, aqueoushumor, lens, vitreous, neuronal retina, retinal pigment epithelium, choroid, andscleraobtained from porcine eyes were separated and the absorption spectra weremeasuredwith a Lambda 2 UVfVlS spectrometer (PerkinElmer, Waltham, MA).With excitation in the absorbance maxima, fluorescence spectra were measured withan LS 5 spectromerer (PerkinElmer). With illumination at 45 degrees, the fluores-cence was detected in reflection under 0 degrees. To measure the fluorescence inaqueoushumor and in vitreous, the absorbance was adjusted to 0.05. This adjust-ment was not possible in any other opaque or scattering ocular structures. The exci-tation spectra of ocular tissues, detected at 460 nrn, are shown in Figure 8.9.

To check FAD, the excitation spectra of ocular tissues were determined by meas-uring changes in the fluorescence intensity at 524 urn when excited at single wave-len~hs between 310 nrn and 490 nm (Fig. 8.10).

Fluorescence Lifetime of Ocular TissuesConsiderable differences among different ocular structures in histograms of meanlifetimeTmean (Fig. 8.11) are observed in biexponential approximation.


600

500Sclera

400~

Aqueous humour

'00c: 300

ChoroidOJ RPE-c:200

Retina

300 350Wavelength in nm

400

FIGURE 8.9. Excitation spectra of ocular tissue detected at 460 nm (mean of measurements obtained from10ocular samples) Excitation maxima around 350 nm were detectable from lens. sclera, and cornea aswellas from retina, and to a certain degree also from vitreous, In the RPEthis excitation maximum is only weaklydetectable, No clear excitation maximum at 350 nm was detected from choroid and aqueous humor. An ad-ditional shoulder in the excitation spectra was detectable at 380 nm from lens, sclera, and cornea,

800 ,-----.:~-------'------------__,Retina460

2001--r-Y-\--t-1I-f!~~Arf---:----,-----JAqueous humor

2520

Choroid1700

VitreousRPE-H---960 ~260

600

»,

g Sclera

Q) Cornea 17806- 400I-ttlt----tt----~;;;__--+--------~ l~OLL

1000 2000mean Lifetime "m in ps

3000

FIGURE,8.10. Comparisonof lifetime Tmean of ocular structures lexcitation 446 nrn. emission 50G-700 nm)The lifetime Tmoo, increasesnearly by a factor of 2 from the shortest value in the RPE1260 I .'in th it 1960 I' , , ps to a maximumIn e VI reous ps, with a value In between In the neural retina 1460psi Th d' t 'b ' ' ,'. . e IS n utlons of mean life-time Tmean cover each other qurts well for lens and cornea 11400psl, and for scle d h id Ithe aqueous humor, the distribution of mean lifetime Tm", is quite broad-arou~~ ;~20c p~rol 1750 psi In

CHAPTER B QUANTIFYING FUNDUS AUTOFLUORESCENCE

Excitation Imageregistration

Excitation pulse laser IR reflection75 psFWHM,BOMHz image

Detection

Images ot tirre - resolvedauto - fluorescence

TCSPC

Frarrn grabbe, 1500- ~60nml

Laser sc an nerophthalmoscope r---7"-----t--T---[~56~0~-~7~0~0~nmrJj

I CW 11 las., IfiGURE 8.11. Setup for fluorescence lifetime imaging of the fundus for excitationat differentwave-lengthsandfluorescence detection in different spectral ranges.

In histograms of rhe amplitude aj, the decay of AF is dominared by the shortcomponent in RPE (o , = 95%) and in neuronal retina (a, = 88%). In the choroid,lens, cornea, and sclera, the amplitude is much smaller and afe nearly the same (alabout 70%).

The liferimes and amplitudes of ocular structures are given in Table 8.2.Individual deviations from rhese values are on the order of 10%.

EXPERIMENTAL SETUP FOR FLUORESCENCELIFETIMEMEASUREMENTS IN THE HUMAN EYE

Technical DescriptionAt our institution, a scanning laset ophthalmoscope (SLO) was developed for 20measurements of time-resolved AF of the fundus (Fig. 8.11). It works in the timedomain. In this device, the parameters required for discrimination of fluorophoresarecombined; therefore, two wavelengths can be used for excitation, and the rime-resolved fluorescence can be measured in two spectral emission ranges ....Lifetimes and Amplitudes of Ocular Tissue of Porcine Eyes. .

in Biexponential Approximation (Excitation 446 nm,Emission 500-700 nm)

Tissue 71 in ps 0:1 in% 72 in ps azin %

RPE 200 95 1800 4Retina 240 88 2560 10Choroid 530 70 3400 30Lens 460 69 3200 31Cornea 470 70 3600 30Sclera 620 64 3640 36Vitreous 260 78 3200 22

The basic device is a commercially available SLO (HRA 2, Heidelberg Engineenn,. ) T Ise lasers can be used for excuanon, emirri,.Heidelberg, Germany. wo pu . kJ B r-'

446 nm or 468 nm (Lasos, Jena, Germany; Becker & H IC , er in, German~P· B l' G rrnany) These lasers deliver pulses of75 ps FWHM atar",lCoquant, er In, e . . I th -r. . f 80 MH The average radiation power IS ess an 100 IJ,Winuuon rate 0 z.

I Th f dus is irradiated by an infrared (IR) laser at 820 nmsimullcornea pane. e un . . . ., .I . h h it ti n light Since the acqursiuon rime ISshan fat S1n~'neous y Wit t e exci a 10· . ijI\;

., Iik I that eye movements will interfere with the measurementsnages, It 15 un 1 e y " ... hi h i e ch IR image but very weak In the AF Image. Subsequentun",contrast IS 19 In a . ._'"

are registered in relation to this IR reference Image. The calculated Imagetransfo,marion is also used for the registration of the lIfetime Images. Thus, each phoron~be added at the right position in the correct time channel for liferime measurementImages are automatically excluded if not enough strucrure IS found for reg~tr.lrionsuch as in the case of blinking and eye movement resulting In doubling of vesselstruc.tures or fixation outside of one-half of the reference image. The IR lighc issepararelfrom the fluorescence lighr by a dichroic filter (DM I). Additionally, the fluoro-cence light is detected in a short-wave spectral range (490-560 nm) and a long-waverange (560-700 nrn). A dichroic filter (DM 2) separates both beams. In bothspec-tral ranges, the fluorescence will be detected by a multichannel-photomultiplier(MCP-PMT, HAM-R 3809U-50; Hamamatsu, Herrsching, Germany) withjitter<50 ps. The fluorescence decay is detected in the TCSPC technique by a SPC liDboard (Becker & Hickl). This board works in the first-in/first-our mode andh"ili-recr memory access. A HRT41 router (Becker & HickJ) separates the photonsfromboth spectral channels. During measurement, online registration is performed inborh channels.

The measurements last until a certain number of photons are collected at eachpixel in the fluorescence image. Practical measuring times are realized with a spari~resolution of 40 X 40 /Lm2 Since rhe number of detected photons incresewith the pixel area, the measuring time can be reduced if the spacial resolution ~not so high.


Experimental ResultsImages and Histograms of Parameters of Dynamic Fluorescencein a Healthy SubjectIn the following examples, results are presented rhar can be calculated from measure-ments of rime-resolved fundus AF. The first example is the dynamic fundus AF ofa63-year-old healthy subject. The fundus was excired at 468 nm by a pulse laser.Thedynamic fluorescence ar each pixel was approximated by a biexponential modelfunonon using the program SPClmage 2.9.1 (Becker & HickJ). As result of the fir, imagoof the fluorescence intensity, lifetimes, and amplitudes, as well as histograms of theseparameters, can be demonstrated. By comparing the intensity images of channels Iand 2 (Fig. 8.12), one can see rhar the conrrasr of the vessel structure is higher inthelong-wave emission range than In the short wave one Thi h th . pan

- . 15 means t at e mainof long-wave fluorescenc~ light originates from behind the vessels. Also, almost nolong-wave fluorescence hght is detecred from rh d k . di h . the

. e ar OptIC 15C, W ereas Inshorr-wave range the Contrast IS weak and fluorescence I' h . al d d Ii the. di h 19 t 15 so etecte romopnc ISC.T erefore, rhe shorr-wave fluorescence light al " al ., at. . f f h so con tams SIgn ongmar-Ing In ront 0 t e vessels. In liferime images T d'fi"C 'd bJ b theal' ' mean 1 lers ConSI era y erweentwo specn channels (FIg. 8.13; in rhe color range, red means T

m= = 150 ps,and

500 nm- 560 nm


560 nm - 700 nmFIGURE 8.12. Images of fluorescence intensity of healthy fundus in different spectral channels (excite-tion468nrn]. Left: Short-wave channel Kl. Right: Long-wave channel KZ.

blue"mea" = 300 ps). The longest lifetime is detectable from the optic disc. This longdecay originates from connective tissue (collagen, elastiu, and cholesterol) (27).Especiallyin the short-wave range, the macula exhibits the shortest lifetime (RPE),and in the papillomacular range the lifetime Tmean is in between them. In contrast, inthe long-wave channel, this range exhibits nearly the same long lifetime as the opticdisk. Clear differences in lifetime distribution of'Tmean exist for the fluorescence inboth channels (Fig. 8.14).

500 nm - 560 nm 560 nm - 700 nm

150 ps 300 ps

FIGURE 8.13. Images of fluorescence lifetime "mea" in the short-wave channel [leftj and the long-wavechannellrightl.

o 50 100 150 200 250Lifetime Tm in ps

300 350 400


2000~-----------T--------

1500-TmKl-TmK2>.oc:

Q)::l 10000-Q)....u...

500

FIGURE 8.14. Histogramof mean lifetime "moe, in channels Kl and K2 of the fundus of a normal subject.

The amplitude IT, is generally highet in the short-wave channel Kl (94.4%)thanin K2 (85.6%). In the optic disc in KI, IT, is quite low, which means thatthemflu-ence of the component corresponding to TJ is considerably reduced in me opric dsc

Quasi-3D Images in Early Age-Related MaculopathyInvestigations of dynamic fluorescence are of special interest in early stages ofAMD.Here, the advantage of lifetime measurements is evident. The RPE fluorescencecsnbe studied with no interference from the macular pigment.

In addition to 2D images or histograms of lifetimes 'Tjl amplitudes OJ, andrelative contribution Q;, cluster diagrams of-r, VB. 'Tj, or aj VB. CXj' are well suitedfor detection and interpretation of pathological alterations. Of special interest arequasi-3D images, in which the value of these parameters is drawn in the thirdcoordinate.

Contrary alterations in lifetime in the macula ate detectable in both specrrslchannels. As demonstrated in nonexudative AMD (Fig. 8.15), in the shorr-wavechannel 73 is increased in an extended range temporal the optic disc, excludingthemacula, where low values of 73 are determined. In this macular range, the life{i~e73 is increased up to 4 ns in the long-wave channel. According to the emissionIIIthe long-wave range, a component of lipofuscin may be detected with a longdecaytime. For compatison, the lifetime of AlE (component VIII) was determinedwith 170 ps.


500-560 nm 560-700 nm

Infrared

FIGURE8.15. Complementary macular lifetime T3 in short- and tone-wave channels in nonexudativeAMD Left T3 in the short-wave channel. Right: T3 in the lonq-wave emission channel. Blue: Short life·timeT,; red long lifetime T3. Below IR fundus image for comparison. The encircled macular range of en-largedlifetime 'T3 only in the long-wave emission channel points to the accumulation of a component oflipofuscinwith a long fluorescence decay time.

SUMMARYQuantitative, independent evaluation of fundus fluorophores seems to be possible. Forthis purpose and in the clinical setting, fundus images demonstrating local distributionof fluorophores would be of great value. However, various problems can be encoun-tered, such as the variety of fluorophores contributing to the measurable fluorescence,and the weak fluorescence of interesting fluorophores, such as FAD, which is coveredby the strong fluorescence of lipofuscin, the predominanr fluorophore of the ocularfundus. Furthermore, the fluorescence of the crystalline lens should be eliminated. Agood separation between fluorescence from RPE and neuroretina on one hand and flu-orescence from the crystalline lens on the other hand can be achieved by fluorescencelifetime measurements. Lifetime measurements can effectively be performed in thetime domain with the use of pulse lasers in confocal SLOs and by derecting the dy-nannicfluorescence in the TCSPC technique. A nearly complete elimination oflens flu-orescence can be achieved in fluorescence measurements of the fundus by combiningthe confocal technique with division of the aperture diaphragm.

The combination of fluorescence lifetime measurements with selected excitationwavelengths and simultaneous detection of fluorescence in separate spectral ranges isan oprimal method for characterizing fluorophores in fundus images.

Because several fluorophores are excited at the fundus simultaneously, the inter-pretation of such measurements is challenging. For such interpretation, knowledge ofthe excitation and emission spectra as well as the fluorescence lifetimes of expectedfluorophores is required (28). Equivalent measurements on ocular tissue and com-parisons with the anatomy of the eye point to the origin of the measured fluorescence(0). The performance of different methods must be evaluared for analysis of dy-namic fluorescence measurements. Model studies on cell and tissue cultures (29-31)make it possible to compare spectrometric measurements with the results of other

REFERENCES


al . al h d such as high-performance liquid chromatography, mass 'fit.an ytlC met 0 s, C h d Ramans-s-c.:1 eric resonance, and surrace-en anee an scanenng I]trometry nuc ear magn J' c . •

1: . al . the interpretation of fluorescence rrenrne measurementsQ'In C InK practice, .. d b bolic provocation tests. Measurements on eyes wah In.be Improve y meta . J' . f h .

. al fu . al defects would enable a sunp e Interprerauon 0 [ ,1,[,,; •.anatomic or ncnon -"'IJl

information. diIn addition to investigation of endogenous fJuorophores, Stu es on exog,,,,,

fJuorophores are interesting because lifetimes may change between. th, &eelilI. b d nd such studies may facilirare further understanding ofc1Uu~proteln- on states, a

findings.

1. Niesner R, Peker H, Schluesche P, er al. Nonirerarive biexponennal fluorescence lifetime imagingintlitinvestigating of cellular metabolism by means ofNAD(P)H autofluorescence. Chern PhysChemloot.i1141-1149.

2. Wu Y, Zheng W, Qu ]Y. Sensing cell metabolism by time-resolved autofluorescence. OpDCSlentn2006;31;3122-3124.

3. Stryer L. Biochemie. Spekrrum Akad. Heidelberg, Berlin, New York:.Verl,ag, 1.~91.4. Berman ER. Biochemistry of the eye. In: Blakemore C, ed. Perspectives In VISionResearch. New-Yod:.

Plenum Press, 1991.5. Lakowicz JR. Principles of Fluorescence Spectroscopy. 2nd ed. New York: K1uwe:rAcademidPlenllJlJ,

1999.

6. Geeraets \Xl], Berry ER. Ocular characteristic as related to hazards from laser and other light sources AmJ Ophthalmol 1968;66; 15-20.

7. van de Kraars J, van Nerren D. Optical density of me aging human ocular media in rhe visibleandth.:UV. J Opt SocAm A 2007;24;1842-1857.

8. KoelJner M, Wolfrum J. How many photons are necessary for fluorescence lifetime measurements?QrmPhys Lerr 1992;200:2.

9. Schweitzer D, Hammer M, Schweitzer F. Grenzen der konfokalen Laser Scanning Technik beiMessungender zeitallfgekisten Aurofluoreszenz am Aligenhinrergrund. Biomedizinische T echnik 2005;50:263-267.

10. Schweitzer D, Jentsch S, Schenke S, er al. Spectral and time-resolved studies on ocular Structures.SP!E.OSA 2007;6628;662807-1-<'i62807_12.

11. American National Standard for the Safe Use of Lasers. ANSI Z 136.1-2000. Orlando. FL: LaserInscirurrof America, 2000.

12. Clegg RM, Schneider PC. Fluorescence lifetime-resolved imaging microscopy: a general descriptionoflifetime-resolved imaging measurements. In: Fluorescence Microscopy and Fluorescence Probes. NewYork: Plenum Press, 1996:15-25.

13. Becker W. Advanced time-correlated single photon COunting techniques. Springer Series in ChemiclPhysics 81. Berlin, Heidelberg, New York: Springer, 2005.

14. Knutson JR, Beechem ]M, Brand 1. Simultaneous analysis of multiple fluorescence decaycurves:a~ob~approach. Chern Phys Lett 1983;102:501_507.

15. AlcalaJR, Gratton E, Prendergast FG. Fluorescence lifetime distribution in proteins. Biophys] 1987;51:597-604.

16. Lee BKC, Siegel], Webb SED, er al. Application of stretched exponential function to fluorescenceliferim~imaging. Biophys] 2001;81:1265_1274.

17. Kori ASR, Krishna MMG, Periasaml N. Time-resolved area-normalized emission spectroscopy(TRANES). A novel method for confirming emission from two excited states. ] Phys Chern A 2001;J05:1767-1771.

18. ]0 ]A, Marcu L, Fang Q er ~. New m.ethods for time-resolved fluorescence spectroscopy data analysisbased on the Laguerre expansion techllJque-applicarions in tissue diagnosis Methods Inf Med 2007;46;206-211. .

19. SchweitzerD, Kolb A, Hammer M. er al. Tau-mapping of the autofluorescence of the human ocular fun-dus. Proc SPIE 2000;4164:79_89.

20. Feeney-Burns L, Berman ER, Rothman H Lipofuscin of hu . al· ith rAm]Ophthalmol 1980;90:783-791. . man retlll plgmenr epl e /Urn.

21. Delori F.C, !?otey.CK, S~aurenghi G, et al. In vivo fluorescence ofche ocular fundus exhibirs retinalpig·mem epIthelIUmhpofusclO characteristics. Invest Ophmalmol Vis Sci 1995;36718-36729.

22. Ehld<edGE, Km ML. Fluotophme< of the hum,n 'etina! pigment epi,heUum; ,epmtion ,nd ,p,arnlc aractenzatlOn. Exp Eye Res 1988;47:71_86.23. Eldred GE, Lasky MR Retinal age-pigme d b If. .

Nature 1993;361:724_726. nrs generate y se -assembll.llg lysosomotropic detergents.


24. Sparrow JR, Fishkin N, Zhau J, et a1. A2E, a by-product of the visual cycle. Vision Res 2003;43:2983-2990.

25. Parish CA, Hashimoto M, Nakanishi K, et al. Isolation and one step preparation of A2E and iso-AlE, flu-orophores from human retinal pigment epithelium. Prae Nat! Acad Sci USA 1998;95:14609-14613.

26. Schweitzer 0, Hammer M, Schweitzer F, er al. In vivo measurement of time-resolved autofluorescenceat the human fundus. J Biomed Optics 2004;9:1214-1222.

27. Marcil L, Grundfesr WS, Maarek JML Phorobleaching of arterial fluorescent compounds: characteriza-tion of elastin, collagen and cholesterol time-resolved spectra during prolonged ultraviolet irradiation.Phorochem Photo bioi 1999;69:713-712.

28. Schweitzer 0, Schenke S, Hammer M, er at Towards metabolic mapping of the human retina. MicroscResTech 2007;70:410--419.

29. Cubeddu R, Taroni P, Hu DN, er al. Phocophysical studies of AlE, putative precursor of lipofuscin, inhuman retinal pigment epithelial cells. Phocochem Photobiol 1999;70: 172-175.

30. Doccio F, Boulton M, Cubeddu R, et al. Age-related changes in the fluorescence of melanin and lipofus-cin granules of the retinal pigment epithelium: a rime-resolved fluorescence spectroscopy study.Phorochem PhorobioI199l;54:247-253.

31. Bui TV, Han Y, Roxana A, er al. Characterization of native retinal fluorophores involved in biosynthesis ofAlE and lipofuscin-associated retinopathies. J Bioi Chern 2006;281: 18112-18119.

32. Kneipp K, Kneipp H, Kneipp J. Surface-enhanced Raman scattering in local optical fields of silver and goldnanoaggregares-e-from single-molecule Raman spectroscopy to unrrasensirive probing in Eve cells. Ace Chern&,2006;390443-450.

CHAPTER

The Normal Distribution of FundusAutofluorescence

Ohis chapter concentrates on the image of the normal fundus itself and itsanalysis. In pa~ticular, it first discusses images obtained from the most widelyused and available systems, the Heidelberg Retinal Angiograph (HRA;

Heidelberg Engineering, Dossenheim, Germany) and its latest model, the HRA2.Theseconfocal scanning laser ophthalmoscopes (cSLOs) use an excitarion wavelengthof488 nm and a barrier filter of 500 nm to provide fundus autofluorescence (AF) im-agingin vivo (see also Chapter 5) (J ,2). A discussion on the variations in such imageswhen acquired with longer-wavelength excitation systems follows.

AF images acquired by the HRA consist of 3D-degree field-of-view laser scans,512 X 512 pixels in size, centered on the macula. The resolution of the HRA2 is768X 768 pixels. As explained in Chapter 5, several scans with a relarivelylow signal-to-noise ratio are registered and then averaged to improve the signal-to-noise ratio.The resulting image is then histogram-stretched by the HRA software [Q increase con-trastfor viewing. These details serve to remind us that the final images are not pixel-by-pixelrepresentations of absolute AF levels, such as would be obtained by spectropho-tometry;rather, they show the relative AF intensities of neighboring pixels. Further,just as in fundus photography, even the relative intensities of widely separated pixelsareaffected by intrinsic variability in the acquisition process. In other words, illumi-nation may vary gradually from one portion of a photograph or AF scan to another,with a resulting effect on measured intensities that is purely photographic, not phys-iologic.Nonetheless, a wealth of qualitative information may be gleaned from theseimages, properly interpreted, as reviewed throughout this book. Furthermore, it ispossible,as will be discussed in this chapter, to take these qualitative data back toquantitative interpretation by demonstrating that (i) a large portion of the variabilityobservedin the background AF of a normal AF image has a smooth and regular struc-ture, (ii) this strucrure can be mapped by an appropriate mathematical model, and(iii)this model of a normal AF image naturally provides a framework for the interpre-tationand quantification of AF abnormalities as variations from the normal model.

NORMAL DISTRIBUTION OF FUNDUSAUTOFLUORESCENCEInvivospectrophotometric recording of fundus AF was described by Delori et al. (3),whoshowed that AF arose predominantly from lipofuscin in the retinal pigment ep-ithelium(RPE; see Chapter 3). Thus, the intensity of fundus AF mostly parallels theamount and distribution of lipofuscin (subject to exceptions noted further below).Lipofuscin is derived, in large part, from phagocytosis of outer segment discs contain-ingbisrerinoid by-products of light absorption (see Chapters 1 and 2). The emissionoflipofuscin has a broad band ranging from 500 nm to 750 nm (3) (see Chapter 3).

100 SECTION II CLINICAL SCIENCE

FIGURE 9.1. NormalfundusAF pattern from the HRA. The optic disc and retinal bloodvesselsaredarlThe backgroundpattern is mildlyvariable. with lighter and darker areas throughout,andwithg,eale~"tensity roughlysurroundingthe fovea. The central macula is particularly dark due to absorptionofthebl",excitation lightby luteal pigments. There is a tiny "speck" of increased AF temporally,whichprobablyrepresents a small druse in this normal 54-year-old woman.

In the normal fundus AF pattern, diffuse AF is most intense berween 5 and15 degrees from the fovea. However, mild variability throughout the fundus is therule in a normal AF image, with lighter and darker areas due at least in part ro the imageacquisition process, much as can be seen in a normal fundus phorograph (Fig. 9. I). Theoptic disc and retinal blood vessels have a low (dark) aurofluorescent signal, and rheblood vessels mask the RPE beneath them. However, these fIndings are nor deer-mined by the AF oflipofuscin alone, a fact that complicates their interpretation. Oneof the complicating facrors is the absorption of the 488 nm blue light by macular pig-menrs, especially the carotenoids lutein and zeaxanthin (see also Chapter 3) (4,5).Onfundus photographs and under visual observation, these pigments are characterizedby strong yellow coloration (6). Their absorption is greatest in the center of the mac.ula. This absorption markedly diminishes the foveal AF signal (Fig. 9.1), whichwould otherwise approximate the remainder of the macular background. There is alsosome absorption of the 488 nm light by melanin granules located in the RPE (7).Further, cone phoropigments will be incompletely bleached after an approximately15- to 30-second exposure to the HRA in AF mode (8); hence, absorption by thesestructures will also diminish the AF signal centrally. On the other hand, an advantageof the HRA system is that the AF that is tecorded is dominated by lipofuscin. In par.ticuiar, AF from melanin, which can be recorded in the near infrared (IR) (9) (seeChapter 6), does not affect the HRA signal.

NORMAL DISTRIBUTION OF FOVEALAUTOFLUORESCENCE

Normal AF fundus images of the fovea are affected rhe rnosr by the anaromic disrribu-non of fluorescent lipofuscin and blue light attenuating pigments; hence, the structure

CHAPTER 9 101THE NORMAL DISTRIBUTION OFFUNDUS AUTOFLUORESCENCE

of such an. image deserves separate and detailed consideration. OUf approach todemonstratl~g the structure of a foveal image is as follows: Noise is first removed witha fine GaussIan H.lter, and then the image is further contrast-enhanced for visualiza-tion of the resulting contours of isoaurofluorescence (Fig 9 2) Th. .. ese contours arereferredto asisobars. The resulting normal foveal AF images exhibit finely resolvedconcentricelliptical isobars of AF, with AF increasing outward along any radius froma least-fluorescent center. Further, these elliptical foveal patterns are consistent withthe kn~wn a~lat~mic distribution of lipofuscin, luteal pigment, cone pigment, andRPEthICkenmgm the center of the fovea. A two-zone elliptic quadratic polynomialmodel provides an accurate fit for foveal data and can be used to reconstruct the en-tire fovealdata from small subsets of data (8). The fine structure of the foveal AF isillustratedin two examples in Figure 9.2.

Conversely, the imaging data are evidence for anatomic regularity in the normaleye. That is, the extraordinary geometric regularity and precise isobar resolutions

FIGURE 9.2. Normal foveal AF images and isobar patterns. The foveal regions of interest from two nor-malAFimages have been filtered IGaussian filter. radius 36 I'm) on a small scale for noise reduction. Thecentralfovea has on the average reduced AF. but no pattern is yet apparent (B,E). The fovea [1500-Wdiameterdisk) has been further filtered (Gaussian. radius 180 urnl to establish a very regular shading pat-ternof concentric elliptical isobars of isoautofluorescence Contrast enhancement has also been appliedto emphasize the geometry of the pattern (e). The individual isobars in B are still 100 fine to be dis-cernible.so three are highlighted with black (AF ~ 1121. gray (AF = 1201. and white [AF ~ 1231. respee-lively There are 30 distinct isobars in this pattern that fill the 750 I'm radius. yielding an average isobarresolutionof 750 I'm/3D = 25 I'm. The isobars illustrated are even finer. mostly one pixel [15 fLml inwidth. In this typical pattern. the central isobars are nearly circular. with the more peripheral isobars be-comingmore vertically elongated until the last ones are only partial annuli. (F) A similar demonstration ofthe isobar pattern in E. In Ewe see that the temporal fovea has a slightly more increased AF signal thanthat of other quadrants. and the reduced AF at the center appears somewhat elongated horizontally.Thesefeatures are dramatically more evident in the individual isobars in F. where isobars are highlightedinblack(AF = 621.gray (AF ~ 751. and white [AF = 971. respectively. The ellipses in this less-typical pat-tern become more elongated horizontally and an increased AF signal arc appears temporally. There are36 ISobarsin all in this pattern. yielding an average isobar resolution of 21 I'm. The isobars shown are1 pixel (15 urn) in width. with occasional single-pixel discontinuities.


b J AF patterns suggesc rhar che normal anatomic 'fdemonstrated by rhe dO;~"E. nt and Jipofuscin have similar eJJipcicregubvariations of luteal an plgmeand equally fine resolution.

MATHEMATICAL MODEL FOR AUTOFLUORESCENcIMAGES IN THE 6000-1J.REGION

. . f bnormal AF relative ro the image bacl.~.To make quantltatJve assessments 0 a ". "Slut[

d c h as rernenrs efficiencJy and uniformly In che seCClngof,;"an to penorm t ese me u ">J'. -e-. bId iabiliry ir is desirable ro level the AF Image [Q an Imag"icant ac <graun var , .a uniform background. Then areas of reduced or increased AF WIll appear agaiN'"level playing field" or basal level of fluorescence and can be calculared wirh amform threshold. To accomplish chis, Hwang er al. (10) excended rhe foveal model,AF background described above ro a I2-zone quadraric polynomial machemaciQmodel of the background in rhe 6000-fL region of a normal AF Image. The model,.tested and fit ro normal AF scans, as described below.

Four inner zones were defined: a 600-fL central disc and three annular ron.(600-1000, 1000-2000, and 2000-3000 fL diameter), and two Outer annularro,"(3000-4500 fL and 4500-6000 u). The two outer zones were each subdivided in.four quadrants (superior, nasal, inferior and temporal), giving eighc outer zones,whiclwith the 4 inner zones gave 12 zones in all. The two innermost zones were dlOseusoofor the foveal model described previously. The two-threshold Orsu method (I J) w.used throughout to define candidate regions in each zone with increased and decreasedAF, and local quadratic polynomials were fit ro the temaining normal, or background,pixel values, as described in a previously published study (12). Precisely, the candidaeregions were Co (nonbackground sources with decreased AF, e.g., vessels), C, (back.ground AF), and C2 (areas of increased AF). For each zone, there was an initial choicrof background, C" for input to the quadratic polynomial background model. There.sulting global model was formed from the 12 local models with appropriate radialandangular cubic spline interpolations at interfaces.

This model of macular background was fit ro 10 normal AF images from 10 subjeers with normal dilated retinal examinations. The average absolute errorswert3.8% ± 3.5% of net image range. The mean local standard deviations of the origi.nal images in each zone (exclusive of rhc pixels with increased and reduced AF)ranged from 3.0% to 4.1 % over the 10 images. Thus, if these mean local Standarddeviations were taken as representative of noise in the image, it foUows that the er-rors of the model were of the same magnitude as the noise in the original dara. Thisdemonstrated that the model was an excellent fir ro normal AF data. Finally, eachAF image was then leveled by subtracting its background model (with an offsetof125 gray levels to center the brightness of the resulting image within the usual [0,255J range), and the mean and standard deviation a of the leveled image (excludingvessels) was calculated. It was found chat the leveled image fell wirhin 2.0 (J of themean for 99.7% of pixels In each of the images (Fig. 9.3).

Because SUC? a consisten~ly small fraction of pixels in a normal leveled image fell2.0 a above the Image mean, It seemed reasonable to propose this as a working defini-non of focally Increased AF (FIAF) in a normal image I' e rh a 30' [pix-

, .. , not more an . ;000e1smay have a gtay level greater than 2.0 a above the image mean, after the image hasbeen leveled by the model. The same terminology can then be d rh kina def.. " fFlAF '. use as e war 0mmon a In an abnormal Image' all those pixels h I l' than

. . W ose gray eve 1S greater2.0 a above the Image mean, after the image has been leveled by che model.

CHAPTER 9 103THE NORMAL DISTRIBUTION OF FUNDUS AUTOFLUORESCENCE

FIGURE 9.3. Mathematical model and segmentation of a normal AF scan 16000-fJ, region) (AI Normal AF scan of the righteye of 54-year-old female. Note significant background variability and foveal decreased AF due largely to luteal pigment. Thel1-zone mathematical model of the AF background in A is presented in B as a contour graph Note how the model captures thebackground variability of the original scan. It is essentially smooth throughout, with the exception of a residual mild disconti-nuity superotemporally in this blend zone The contour lines are closer together in the fovea, where the background is morehighly variable. Ie) The image in A leveled by subtracting the model in B. The background of the leveled image is now quitehomogeneous, with mean gray level 126 and standard deviation 11.6 The global threshold of 2.0 standard deviations abovethe mean defining increased AF was therefore 149.2, which was applied to the entire leveled image and yielded the increasedAF shown in white 1028% of the 6000-fl- zone) Comparison of the increased AF with the original image (AI demonstrates avery reasonable selection. By contrast, the use of any single threshold in the unleveled image (AI to define increased AF would

cause major errors due to the image background variability

The leveling of an abnormal AF image is beyond rhe scope of this chapter. Fordetailsand applications, the interested reader is referred to published work on AF ingeographic atrophy (10) and drusen (13). Note that this also allows great flexibilityin the definition of abnormal FLAF: in a given circumstance, and because there is asyet no absolute consensus on the definition of FIAF, a definirion based on 1.5 a or2.5 tJ of deviation from image mean could equally well be used, Focally decreasedAF can be defined similarly, with the understanding that normal retinal vessels willalways fall in this category.

LongerWavelengthsSpaide et al. (14) demonstrated fundus AF using a fundus camera-based system with aband-pass filter for the excitation light of 580 nm and a barrier filter of 695 nm toavoidAF from the crystalline lens (lens AF is rejected by the confocal optics of the SLOsystems) (see Chapter 5). An advantage of the camera-based system is that the longerexcitation wavelength is minimally absorbed by the luteal and photopigments. Hence,the central macular AF signal is only slightly diminished (by the higher RPE melaninconcentration) relative to the remaining background signal, so central macular in-creasedor reduced AF may be easier to interpret. On the other hand, melanin itself,which fluoresces in the near IR (9), could contribute to rhe total detected AF above695 nrn. When there is attenuation of a normally uniform RPE melanin, the melaninin the choroid may be imaged in a banded pattern. This could complicate interprera-tion, as well as any attempt to model the normal AF background from such images.

SECTION II CLINICAL SCIENCE

SUMMARYA precise mathematical description of the normal AF image will be usefulin

. .. . fe all . creased or decreased AF InAF Images ob..:..'quanntative mterpretanon 0 roc yIn I4UlQI

from eyes with pathological fundus changes.

REFERENCESI. von Ruckmann A, Fitzke FW, Bird AC. In vivo fundus autofluorescence in macular dystrophies. Art:!Ophth,umoI1997:115:609-615. .. .

2. Holz FG, Bellmann C, Margariridis M, er ai. Parrerns of lilcreas,ed I~ VIVOfun~lls aut~f1uorescenceintli:junctional zone of geographic atrophy of the retinal pigment epithelium associated With age-relaroo/lll(.ular degeneration. Graefes Arch Clio Exp Ophthaimol 1999;237: 145-152. . .

3. Delori FC, Dorey CK, Sraurenghi G, er al. In vivo fluorescence of we ocular fundus exhibin retinal~~.merit epithelium lipofuscin characteristics. Invest Ophrhalmol Vis Sci 1995;36:718--729.

4. Bone RAj Landrum JT, Cains A. Optical density spectra of the macular pigment in vivo and ill vitro.V~Res 1992;32:105-110.

5. Handelman G), Snodderly DM, Adler A], er aL Measurement of carorenoids in human and monkeyrm.nas. Methods Enzyrnol 1992;220-230.

6. Gellerman W, Bernstein PS. Noninvasive detection of macular pigments in the human eye. J BiomdOptics 2004;9:75-85.

7. Bindewald A, jorzik )), Loesch A, er al. Visualization of retinal pigment epithelial cells in vivo usingdigi.tal high-resolution confocal scanning laser ophthalmoscopy. Am J Ophrhalmol 2004; 137:556-558.

8. Smith RT, KoniarekJP, Chan JK, er a]. Autofluorescence characteristics of normal foveas and reconsme,tion offoveal autofluorescence from limited data subsers. Invest Ophthalmol Vis Sci 2005;46:2940-2~,

9. Keilhauer CN, Delori FC. Near-infrared autofluorescence imaging of the fundus: visualization of ocularmelanin. Invest Ophrhalmol Vis Sci 2006;47:3556-3564.

10. Hwang ]C, Chan ], Chang S, er ai. Predictive value of fundus autofluorescence for development ofgeo.graphic atrophy in age-related macular degeneration. Invest Ophchalmo! Vis Sci 2006;47:2655-2661.

II. Ocsu N. A threshold selection method from gray-level histograms. IEEE Trans Sysr Man Cybemrt1979:9:62-66.

12. Smith RT, Chan JK, Nagasaki T, er a]. Automated detection of macular drusen using geometric back.ground leveling and threshold selection. Arch Ophrhalmol 2005; 123:200~206.

13. Smith RT, .Chan ]K, Busuoic M, er al. Autofluorescence characteristics of early, atrophic, and hig}Hiskfellow eyes 111 age-related macular degeneration. Invest Ophrhalmol Vis Sci 2006;47:5495-5504.

14. Spaide RF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology 2003;110:392-400.

SECTION

Clinical ScienceFundus Autofluorescence in the Diseased Eye

h b I 3% (24). In contrast, the risk fOt progressionto~advanced AMD was sown ro e . . do". . ith many intermediate or large rusen was IS Yo In theA.d AMD In patients WI . '~.

vance . 5 d (ARED5) (24). "Soft" drusen describes the ptesenceRelated Eye DIsease tu y f hi keni f

h d I d arcated lesions in the presence 0 t IC enlng 0 the inneramorp ous an poor y em .f B h' b "Confluent" drusen refers ro contiguous boundariesb,.peets 0 rue s mem rane. .. al ft d Eyes with soft confluent drusen are more likely to pr"""'-tweensever so fusen., -!T'"oJt

late-stageAMD (15.1%) (25). ..The so-called "reticular drusen" have been described as a variant of soft druocn

. h . c (21 26 27) The prevalence of reticular drusen seems to be~:.LWIt umque leatures " . .' . 'ug;crin AMD patients than in age-matched subjects without the disease (21,26Jn.Reticular drusen seem to represent an important risk factor for me deveIopmenrofneo.vascular AMD (28), although a subgroup analysis of ARED5 found a higher rnreofprogression to geographic atrophy than ro choroidal neovascularizarion in the pr""",of reticular drusen (29). Klein and colleagues (30) recently described a IS-year cum.lative incidence of reticular drusen (3.0%) in a population-based (n = 4926) prospec.tive study. Eyes with reticular drusen had a higher risk ro progress to geographit atr0-phy (cumulative incidence 21 %) or ro exudative AMD (cumulative incidence 20%1than those with soft indisrinct drusen. From histopathological findings in one eyewitlreticular drusen, it appeared that fundus changes did not correlate with the extratdJu_lar material deposited in the inner aspect of Bruch's membrane, but did correlarewithchoroidal alterations; therefore, the term "pseudodrusen" was proposed (27). However.the precise histopathological changes in reticular drusen is yet unknown.

In addition ro the above drusen types, a different type of material can be founddeposited between the RPE cell plasma membrane and its basement membrane. Thiscomplex composite of granular electron-dense material, coated membrane bodies,and fibrous collagen was initially termed "basaltinear deposit" (31,32). Based on lighrand electron microscopy, the deposit was later renamed "basal laminar deposit,'whereas the term "basal linear deposit" was introduced to describe vesicular materialunderneath the basement membrane of the RPE. Because of this confusing terminol-ogy, Loeffler and Lee (33) suggested the terms "basement membrane deposit" (BMD)for material located between the RPE cell and its basement membrane, and "bas~laminar deposit" (BLD) for vesicular material located within Bruch's membraneHowever, these deposits are detected only by light and electron microscopy and re-main invisible by slit-lamp biomicroscopy. Apart from these ophthalmoscopically in-visible deposits, Bonanomi et al. (34) and Gass er al. (35) described basal laminardrusen in association with pseudovitellifotm lesions that angiographically appearas"stars in the sky." Histologically, rhese basal laminar drusen correspond to nodular,hyaline thickening of the basement membrane of the RPE (35).

Over time, drusen may be subject to dynamic changes: (i) hard drusen mayen-large and turn into soft drusen (36), (ii) soft confluent drusen may lead to a drusenoidretinal pIgment epithelial detachment, (iii) drusen material may show signs of calcifi-canon, and (IV) drusen may regress with the occurrence of a corresponding areaofge-ographic atrophy (37).


IMAGING TECHNIQUES IN ARM

Fluorescein AngiographyDuring fluorescein angiography (FA), drusen may appear hyper-, iso-, or hypoiluo-rescent (FIg. lOA. I). Particularly soft drusen may be hy fl . I hases

po uorescent In ear y Pand hyperfluorescem due to staining in late phases of the angiogram. These angie-

CHAPTER lOA

A,B

FUNDUS AUTOFLUORESCENCE IN AGE-RELATED MACULOPATHY

D,E

FIGURE IDA.I. Left eve of a patient with soft and reticular drusen at the posterior pole at baseline (A-C) and 2 vears laterIO-F). Funduscopicallv visible drusen (A,D) demonstrate increased fundus AF (H,E) with enlargement and multiplication ofareasof increased AF over time. FA IC.FI shows hvperfluorescence in the presence of drusen and development of a smallchoroidalneovascularization over time (F)

graphic features are thought to depend on the chemical composition of drusen.Hyperfluorescentdrusen contain mainly polar phospholipids, whereas hypofluorescentdrusenare formed by neutral lipids (11). In contrast, areas of focal hyperpigmentationat the levelof the RPE are characterized by hypofluorescence due to blockage phenom-ena (38). FA is used only to evaluate patients with ARM when neovascular AMD issuspected.

Indocyanine Green AngiographyHard drusen are very difficult to distinguish on indocyanine green (leG) angiogra-phy. Larger soft drusen appear mosrly hypofluorescent throughout the leG an-giogram.leG angiography does nor usually reveal essential diagnostic/prognostic in-formation in eyes with ARM. As for FA, ICG angiography is only used in patientswithARM when certain forms of neovascular AMD, including polypoidal choroidalvasculoparhy and retinal angiomatous proliferation, are suspected.

Optical Coherence TomographyOptical coherence tomography (OCT) allows for cross-sectional imaging of theretina in vivo with a micron resolution, helping to precisely evaluate the anatomi-cal localization of pathological processes (39). OeT imaging of drusen shows

C

F


I I ions including focal elevations and irregularitl". . tructura a rerauo , ...,ftvanous mJcfrohs RPE hich normally appears as a continuous highly refl~tithe level 0 t e , wI' 11 di , .. I d to a focal irregu arrry as we as to ISruprlOno!t"straighr layer. Drusen may ea OCT all d . <. I h f soft conlluent drusen, ows erectlon.rRPE SIgnal. n t e case 0 t,

drusenoid RPE detachments. C..L d '1. I' I domain OCT has revealed rurrner erai S of anatom'"tHIgh-reso uuon, speccca. '. . ~

al . (40) USI'ng ultrahigh-resolunon (UHR) OCT with a 3 ILm axl~trn,~teracons . be i . th ....'1, h al limiting membrane appears to e mracr In e presen~.1reso litlan, t e extern .. . .

ARM, whereas in the presence of drusen, irregularities In me RPE and themneraodt f rhe photoreceprors can be seen In me OCT Images (41). Pieroni",outer segmen so. . ,

coworkers (42) described rhree patrerns of drusen In me presence of ARM: (,)dIJtinrRPE excrescences, (ii) a sawroorhed pattern of me RPE (multiple excrescen,,,.su:.gesring a wrinkling or bunching of the RPE), and (iii) nodular drusen. ocr andsp•.tral-domain OCT help to identify RPE changes WJm or without RPE cell rrugrntloninto me overlying neurosensory retina (Fig. 10A.2). Furrhermore, hlgh.reso!urionOCT may help to detect early exudative changes, i.e., extracellular fluid, rharmaynotbe visible on funduscopy or FA (42).

Fundus Autofluorescence ImagingAF imaging is nor only of interest to help understand me pamophysiological processof ARM; it can be also used to precisely diagnose and monitor phenotypic chango.Recently, variations in AF have been demonstrated in eyes with ARM (43-45).1rwasnoted mar alterations in the AF signal do nor necessarily correspond with funduscopi.cally or angiographically visible changes (43).

Hyperpigmenrarion and pigment mortling are usuaJly associated with an in-creased AF signal, which is thought to derive from melanolipofuscin, whereahypo. and depigmentation are characterized by loss of AF signal due to degeneraedRPE or absence of viable RPE. Of interest, hypetpigmentarion may be presentinthe vicinity of drusen associated with focal and Jinear increased AF (46). TheAFsignal from individual drusen may be slightly increased, normal, or decreasedcom-pared to normal background AF, Therefore, drusen mayor may nor be idenrifiedin AF images (47). The composirion of drusen (including possible aurofluorescentconstituents), drusen size, and alterations of the overlying RPE may be responsiblefor the variation in the AF pattern. In general, larger drusen are more fcequenclyas-sociated with extensive AF abnormalities compared to smaller drusen. Crystalline

A.8

FIGURE 10A.2. Fundusphotogtaph(A). fundus AFimage (8). and optical coherence tomography (C) ofa patientwithdrusenandgeographICatrophyWithcrystallinematenalln the atrophicarea, Inaddition to lackofAFdue to pigmentep'Ithellalatrophy.fundusAFIttegulantlesoutside the atrophic patch are seen (8) 0 h ildlv i dAf

. . rusen s ow very Ill! y Increase .Opuca: coherencetomographyreveals several "bumps" in the RPE due to drusen under the RPE.

CHAPTER lOA FUNDUS AUTOFLUORESCENCE IN AGE-RELATED MACULOPATHY

drusen rypically show a corresponding area of decreased AF signal. In contrast,large soft and confluent drusen with drusenoid RPE detachments are characterizedby a moderately increased and patchy AF signal (43,44,48). Such drusenoid detach-mentS have been found to be associated with increased risk of choroidal neovascu-larization (CNY) (49).

Delori and colleagues (50) demonstrated another distinct variation of the AF sig-nalassociated with drusen consisting of decreased AF in the center of the druse witha surrounding annulus of increased AF signal. Possible explanations for this patternare that (i) the RPE may be stretched over the druse wirh a subsequent reduction onthe density of LF granules; (ii) rhe druse makes rhe central overlying RPE releaselipofuscin granules, which are phagocytosed by RPE cells at the margin of rhe druse,creating reduced central AF surrounded by an annulus of increased AF; and (iii)drusenoccur in incipient RPE cell atrophy.

Reticular drusen (see above) or "pseudodrusen" are readily identified on AF im-ages.They show a unique rericular AF pattern with multiple small rounded or elon-gared areas of decreased AF surrounded by an interlacing network of normal AF. Thepreferential localization of reticular drusen is superior and superotemporal to the mac-ula, bur rhey may also spread toward rhe mid-periphery as well as nasally to the opricdisc.As explained above, it was assumed that "pseudodrusen" would represent an at-tenuatedchoroidal venous layer with regression of theechoriocapillaris and broadeningof intercapillary pillars due ro fibrous replacement. Based on fundus AF images to-gether with funduscopic findings, deposition of abnormal material underneath theRPE in the inner aspect of Bruch's membrane appears more likely to be the substrarefor this phenotypic change. The yellowish appearance similar to soft drusen couldhardly be explained if the alterations were in rhe choroidal layer. A decreased AF inten-sityasseen in correspondence with these yellowish lesions would be expected to resultfrom changes in the RPE cell layer or from a somehow strerched RPE. However, a de-creasein AF intensiry cannot be explained by broadened intercapillary pillars, since thelocation in an anatomical layer posterior to the RPE would not conceivably induce anauenuation of the AF signal derived from the RPE.

Classification of Fundus AF Patterns in ARMRecently, a classificarion system for AF changes associated wirh ARM was proposedby an inrernarional consensus group (43). AF findings in patients with ARM indicatemorewidespread abnormalities and diseased areas than assumed by fundus examina-tion or fundus phorography. The proposed ARM AF classification sysrem includeseight different AF patrenlS in ARM, as described below.

Normal PatternA normal AF parrern (see also Chaprer 9) with homogeneous background AF wirh agradualdecrease AF at the fovea can be seen in patients with ARM:. This pattern iscommonly seen in rhe presence of hard drusen (Fig. IOA.3).

Minimal Change PatternVery limited irregular increased or decreased background AF without an obviousropographic pattern in rhe presence of few hard drusen and pigment abnormalities

(Fig. lOA.4).

FocalIncreased PatternThe presence of ar leasr one well-defined spor «200 p.m diameter) of markedly in-creasedAF, which is much brighter than the surrounding backgroundAF. Some areas


FIGURE 10A.3. FundusAF image with a homogeneous background fluorescence and a gradualdecreasein the inner macula toward the foveola due to the masking effect of macular pigment (normalpaneml.Only small hard drusen were seen clinically. (Reprinted from Bindewald A. Bird AC. Dandekar.mal.Classification of fundus autofluorescence patterns in early aqe-related macular disease. InvestDphthalmolVis Sci 2005;46:3309-3314. with permission.)

of focal increased AF may be surrounded by a darker appearing halo. These areasoffocal increased AF, such as focal hyperpigrnenrarion or drusen, mayor may norcor-respond to visible alterations on color fundus photographs (Fig. lOA.5).

Patchy PatternThe presence of at least one larger area (>200 u.m diameter) of markedly increasedAF that is brighter than the surrounding background AF. The borders of theseareas

FIGURE 10A.4. Fundus AF image with only minimal variations from the normal background FAFImini.mal change pattern) due to multiple small hard drusen. clinically visible. (Reprinted from BindewaldABird AC. Dandekar. et al. Classification of fundus autofluorescence patterns in early age-related maculardisease. Invest Ophthalmol Vis Sci 2005;46:3309-3314, with permission.)

CHAPTER IDA FUNOUS AUTOFLUORESCENCEIN AGE-RELATEDMACULOPATHY

FIGURE IOA.5. Fundus AF image showing the focal increased pattern with several well-defined spotsof markedly increased AF. Multiple hard and soft drusen were detected on silt-lamp biomicroscopy.IReprintedfrom Bindewald A. Bird AC. Oandekar. et al. Classification of fundus autofluorescence patternsinearlyage-related macular disease. Invest Ophthalmol Vis Sci 2005;46:3309-3314. with permission]

are typically ill defined. Increased AF corresponds to large soft drusen that may beaccompanied by hyperpigmentations (Fig. IOA.6).

Linear PatternThe presence of at least one linear area with well-demarcated, markedly increasedAF.Theselinear structures of increased AF usually correspond to hyperpigmented lineson color fundus photographs (Fig. lOA. 7).

FIGURE 10A.5. Fundus AF image showing multiple large areas of increased AF [patchy pattern I corre-spondingto large soft drusen and/or hyperpigmentatlon clinically. (Reprinted from Bindewald A. Bird AC.Dandekar.et al. Classification of fundus autofluorescence patterns in eariy age-related macular disease.InvestOphthalmol Vis Sci 2005;46:33D9-3314. with permission. I

SECTION / BASIC SCIENCE

FIGURE IOA.7. The linear pattern is characterized by the presence of a/least one linearamawilarnarkedly increased fundus AF. A corresponding hyperpigmented line was visible clinically IRepnntrofrom Bindewald A. Bird AC. Oandekar. et al. Classification of fundus autofluorescence panemsin~flage-related macular disease. Invest Ophthalmol Vis Sci 2005;46:3309-3314. with permission.)

Lace-Like PatternMultiple branching linear structures of increased AF forming a lace-likepattern.Th,borders may be difficult to define because a gradual decrease of AF is occasionallyob-served from the center of the linear areas toward the surrounding background. Thislace-like pattern of increased AF may correspond to hyperpigmenracion or ronov~i.ble abnormality (Fig. lOA.B).

FIGURE IOA.B. FundusAF image demonstrating multiple branching linear structures of increasedAf inalace-like pattern This lace-like pattern corresponds to hyperpigrnentation on fundus examination.IReprmtedfrom Blndewald A Bird AC Dandekar et al CI if . f f .

.' , '. ass: Icatlon 0 undus autofluorescence patterns In earlyage-related macular disease. Invest Ophthalmol Vis Sci 2005A6·3309-3314 ith .. )

' . ,WI permIssion.

CHAPTER IDA 115FUNDUS AUTOFLUORESCENCE IN AGE-RELATED MACULOPATHY

FIGURE IOA.9. Multiplespecific small areas of decreased AFwith brighter lines in betweencharacter-izethereticularpattern. There may be visible reticular drusen on fundoscopy.(ReprintedfromBindewaldA. BirdAC.Oandekar.et al. Classificationof fundus autofluorescencepatterns in earlyage-relatedmac-ulardisease.InvestOphthalmolVisSci 2005;46:3309-3314, with permission.I

Reticular PatternThe presence of multiple small areas «200 u.m diameter) of decreased AF with mild in-creasedAF around them. The reticular pattern occurs not only in the macular areabutalsoin a superotemporal location. This pattern may be associated with funduscopicallyvisiblereticular drusen (Fig. 1OA.9).

Speckled PatternThe simultaneous presence of a variety of AF abnormalities in a larger area of the AFimage.The corresponding abnormalities visible on color fundus photographs includepigment abnormalities and multiple (sub-) confluent drusen (Fig. lOA.IO).This phenotypic classification may help to (i) identify prognostic determinants in

longitudinal studies, (ii) better evaluate the progression of the disease, and (iii) allowphenotype/genotype correlations ro be established. Similar longitudinal studies in at-rophic AMD have identified the prognostic relevance of different AF patterns in thejunctional zone of geogtaphic atrophy (51,52). In the case of ARM, more longitudinaldataare needed to evaluate the impact of distinct AF patterns on disease progression.

DIFFERENTIAL DIAGNOSISApart from normal aging, a variety of retinal diseases may mimic ARM. In the case ofidiopathic central serous chorioretinopathy (lCSC) (see also Chapter 13), age, ab-senceof drusen, and the occurrence of distinct small serous retinal detachments maybe helpful in differentiating this disease entity from ARM. However, chronic ICSCmaypresent with phenotypic alterations reminiscent of ARM and occult choroidalneovascularization. Pattern dystrophy can mimic ARM (see also Chapter lIE); how-ever,the high-intensity AF signal corresponding to the yellow deposits seen in patterndystrophy is not observed in ARM. Similarly, autosomal-dominant (AD) drusen,

SECTION" CLINICAL SCIENCE

42. Pieroni CG, Witkin AJ, Ko TH, et at Ultrahigh resolution optical coherence tomographyinno~rive age related macular degeneration. Be J Ophrhalmol 2006;90: 191-J 97.

43. BindewaJd A, Bird AC, Dandekar, er al. C1~sjfi~arion of fundus aurofluorescence patterns in e41JY1gl"rlared macular disease. Invest Ophrhalmol VIS SCI 2005;46:3309-3314.

44. Lois N, Owens 5L, Coco R, et al. Fundus autofluorescence in patients with age-relatedmacubr~~.arion and high risk of visual loss. Am J Ophchalmol 2002;133:341-349.

45. Spaide RF. Fundus autofluorescence in age-related macular degeneration. OphthalmologyZoo},II392-399.

46. von Rttckmann A, Fitzke FW, Bird AC. Fundus autofJuore~cen~e in age-related maculardisasei~with a laser scanning ophthalmoscope. Invest Ophthalmol VIS SCI 1997;38:478-486.

47. von Ruckmann A, Firzke FW, Bird AC. Autofluorescence imaging of the human fundus. In: M:l!llI<IrMFWolfensberger T], eds. The Retinal Pigment Epithelium. Oxford: Oxford University Press, 1998:224-234

48. Smith RT, Chan JK, Busuoic M, er al. Autofluorescence characteristic of early, atrophic, and high-risiftl.low eyes in age-related macular degenerarion. Invest Ophrhalmo! Vis Sci 2006;47:5495-5504.

49. Einbock W, Moessner A, Schnurrbusxch VE, er al. Changes in fundus autofluorescence in pali~nlSwithage-related maculopathy. Correlation to visual function: a prospective study. Graefes Arch air.Ophrhalmol 2005;24BOO-305. .

50. Delori FC, Fleckner MR, Goger DG, et a1. Autofluorescence distribution associated with drusenin .related macular degeneration. Invest Ophrhalmol Vis Sci 2000;41 :496--504. agt

51. Holz FG, Bindewald- Wirtich A, Fleckenstein M, et al. Progression of geographic atrophy and impaa:offtm.dus autofluorescence patterns in age-related macular degeneration. Am] OphrhalmoI2007;141463-472

52. Schmitz- Valckenberg S, Bindewald- Wittich A, Dolar-Szczasny J, er 31. Correlation between the areaofin.creased autofluorescence surrounding geographic atrophy and disease progression in parienrs with AMD.Invest Ophthalmcl Vis Sci 2006;47:2648-2654.

CHAPTER

I

IIFundus Autofluorescence in NeovascularAge-Related Macular Degeneration

m eovascular or exuda:ive age-r~la[ed macular degeneration (AM D) is the mostW common cause of VISual loss In patients wirh AMD (1,2). It is defined by thepresence of any of rhe following (3): (i) rerinal pigmenr epithelium (RPE) de-

rachmentfs), which may be associated with a neurosensory retinal detachment; (ii)subretinal or sub-RPE choroidal neovascularizarion (CNV); (iii) epiretinal (with ex-clusion of idiopathic macular puckers), intraretinal, subretinal, or sub-RPE scar/glialtissue or fibrin-like deposits; (iv) subrerinal hemorrhages not related to orher retinalvascular disease; and (v) hard exudates (lipids) within rhe macular area related to anyof the above and not related to other retinal vascular disease.

In addition to the typical neovascular AMD caused by the occurrence of CNV,two addirional phenotypes have been recognized: (i) so-called retinal angiomatousproliferarion (RAP) and (ii) idioparhic polypoidal choroidal vasculoparhy (lPCV).In RAP the neovascular process starts in the retina and extends into the subrecinaland sub-RPE space. RAP appears to represent abour 20% of exudarive AMD cases(4). [rCV affecrs rhe inner choroid and is characrerized by a dilared nerwork of ves-sels and mulriple rerminal aneurysmal dilarions in a polypoidal configuration (5).[rev represents 4%-23% of exudative AMD and is more frequent in Asian popu-lations (6,7).

The prevalence of neovascular AMD in people 50 years of age and older rangesbetween 0.10/0 and 7.4%; it is more common in Caucasians and its frequency in-creaseswirh increasing age (1,8-16).

Patients with neovascular AMD usually present with distortion or loss of centralvision. On slit-lamp biomicroscopy, there is usually a serous or hemorrhagic detach-ment of the neurosensory retina and/or RPE with or without hard exudation. Drusenand RPE changes are often present in rhe affecred eye or in rhe fellow eye.

Until recently, laser photocoagulation and photodynamic therapy (PDT) werethe only treatments shown through randomized controlled trials (RCTs) to be effec-tive in a small group of patients with exudarive AMD (17,18). However, new anrivas-cular endothelial growrh factor (VEGF) rherapies are now available and provide ben-efit to the majority of patients with active exudative AMD (19,20).

HISTOPATHOLOGY AND PATHOGENESISThe process of CNV formarion starts with rhe growrh of blood vessels from thechoroid, which then extends rhrough rhe Bruch membrane under the RPE (sub-Rl'ECNY) (21). The CNV may then extend further into the subretinal space (22). Thisabnormal neovascular process leads to serous or hemorrhagic detachment of the RPEand/or neurosensory retina. Eventually, if untreated, scarring wil~ ~nsue.and degener-arion of photoreceptors and RPE and subsequent loss of central VISIonwill occur (23).


. I rh invasion of choroidal vessels through Bruch',.Th . al th t sumu ate e .,~e SIgn sad d Under normal circumstances the RPE,Ierely un erxtoo . errbrane are not comp f: hi h helps [0 maintain the feneserations in the IVEGF at irs basolateral sur ace, w IC . .. h RP (

.' . h . accumularion of lIpofusCIn In teE and depositio,ncapijjaris (24). W" algIn1g"d) and subsequent chickening in Bruch's memb",'

ial ( dominant y Ipl ematen pre It in reduced oxygen transmission from thechoriucurs (25,26). These events resu .' . [

. .rh outer reti nal hypOXia and subsequent upregul'noillaris into the outer renna WI CNV fi '.. r

. b h RPE (27) which could lead [0 ormation. It" tho".VEGF secrencn y t e, .' d ~I lex in these cases provides nurnenrs an oxygentom,'that the neovascu ar corup . (). '

h . RPE/ . which is expressIng VEGF 28,29. Activated marrop""c erruc outer retma, . IIOf. f h horoid into Bruch's membrane With the goal of remo,;"may rmgrare rom tee . . . . -1)

ial d . d in this layer with the possible creation ofchannel'InBNwaste maten eposlte , . . .b h h hi h blood vessels could then Invade the rerina (30), Ith.,rnem cane t raug W rc . "

shown that the RPE can produce monocyte-chemoarrractant-proteln and !L8dulj . h ofCNV development which would stimulate the migrationofmo"t je actrve p ase ' . .. u-

cyres (macrophages) from the choriocapIllans. lI1to the outer surface of Bruchffillfbrane (31). Macrophages obtained from surgically removed CNVs and fromw

postmortem eyes with eNV showed increased expression of many inflammaro~r1'rokines by rhese cells (31). Increased expression of many growth factors, 'pecifidVEGF, in the RPE and photoreceptors, as well as recruited macrophages andprolfu·arion of vascular endothelium from choriocapiliaris, has been demonstraredjns~fmens obtained from patients with neovascular AMD (28,29,31,32).

Gass (33) suggested that CNV in AMD grows predominantly under theRPEI.called type 1 CNV), although in some patients it can extend mainly inca thesub,e.space (type 2 CNV). Several histopathology srudies have shown that in 00, "'P'graphically classified as occult, the growth of new vessels occurred predominandy•.neath the RPE (22,34-38). In contrast, lesions classified angiographically ~ ""'CNVs contained a subretinal neovascular component, with or without a sulrRPEOJITrponenr (22,34). A combination of both types has also been found (39). CNV'3lectusified as classic on fluorescein angiography (FA) when they demonstrate early,weI1<l.fined hypetfluorescencc and late leakage blurring the margins, and as oauI,,1oill-defined, late, stippled hyperfluorescence or late leakage of undetermined lO1lIT<!

present (2).Histopathology studies of surgically excised RAP lesions showed an inrmerirut

neovascular mass in all cases; in some specimens, an add.itional CNV was alsofouoo(40,41). Immunohistochemistry in these cases demonstrated expression ofhYl""1Oducedgrowth factors and macrophages, suggesting that ischemia and inA"m .. ri"may be 1Ovolved 10 the pathogenesis of RAP (41). Histopathology srudiesinen,dIated eyes of patients with IPeV showed large, thin-walled choroidal vesselsOld"neath the RPE WIth choroidal capillary proliferation (42,43).

CLINICAL FINDINGS AND IMAGING STUDIESSlit-Lamp BiomicroscopyOn slit-lamp biomicros . . , ,

. al . . copy, panenrs with exudative AMD may demonstrate (I) ,.,retm graYISh leSIOn (CNY) . . .JdJI' h ' an mtraretmal reddish lesion (RAP), Ot a round,IOJ"eSlOn at t e RPE level or deeper (IPCY) ( .. ) . d •

'd I d ; 11 subretinal or inrraretinal fluid,In U ~cYStDI macu ar e ema (CME) (... ) .serous or vas ul . d' ; or 111 llltraretinal, subretinal, or sub-RPE blood,81PED h d c danze plgmenr epithelial detachment (PED) including heInorrh~'

, ar exu anon, and, at the end f '. 'fa (JI'stage 0 the neovascular process, diso mt!

CHAPTER lOB FUNDUS AUTOFLUORESCENCE IN NEOVASCULAR AGE-RELATEO MACULAR OEGENERATION

ring.Drusen and RPE changes or disciform scarring are commonly present in the fel-loweye. In IPCV, the fellow eye may demonstrate no abnormalities. CME and in-traretinalhemorrhages are typically seen associated with RAP. Large serous or hemor-rhagicPEDs are commonly observed in IPCV, typically affecting the peripapillary areaandthe macula, but rhey can also be found in the peripheral fundus (44,45).

FluoresceinAngiographyFA remains the most valuable tool in the diagnosis of exudative AMD. In addition, itallowsphenotyping of the disease, i.e., by determining the rype of CNV (classic,mini-mallyclassic,or occult) or, in many cases, whether there is a RAP or possibly IPCV (al-thoughRAP and IPCV may be better imaged by indoeyanine green angiography [ICG][seebelow]), and the locarion of the neovascular process with respect to the fovea andits size.FA is also important to establish whether the neovascular process is actively leak-ing or mainly inactive, by demonstrating active leakage of dye in the former or onlystaining in the latter. Th.is is of particular importance when treatment is considered.

Precise localization of the neovascular process with respect to the fovea {extra-,juxta-, or subfoveal) is very important when deciding treatment options. Argon laserphotocoagulation may still be considered appropriate for patients with extrafovealneovascularization, whereas for those with subfoveallesions anti-VEGF therapies arenowthe trearment of choice. Similarly, rhe size of the CNV is important when choos-ing a suitable therapy. This is especially relevant when considering treatment withargonlaser photocoagularion or photodynamic therapy (PDT) (e.g., a small extra orjuxtafoveallesion could be treated wirh argon laser, but a large lesion on the same lo-cation might not be treated in this manner because of the resulting scotoma thatwouldoccur following treatment).

Knowing which rype of CNV is present is crucial when considering the effecrivetreatment oprions for each patient. Classic CNV can be treated by argon laser pho-tocoagulation, PDT, or intravitreal injections of anti-VEGF therapy. The treatmentof choice is based on the site and size of the lesion and the patient's preference. Onthe other hand, argon laser photocoagulation and PDT are not effective for occultCNV; to date, anti-VEGF therapy is the most successful form of treatment for suchcases (20). The classification of neovascular AMO (classic, minimally classic,predominantly classic, or occult), however, is subjective. Inrerobserver and in-traobserver agreement has been shown to be only moderate, in most studies >0.7(usingkappa statistics) (46-49). Ir should be taken into account that previous stud-ies on inter- and inrraobserver agreement of FA classification in exudative Al\1Dwere conducted without stereo angiography. Stereoscopic assessment of FA allowsthe observer to berrer appreciate the level of the lesion, whether deep or superficial,in relation to the RPE, as well as any elevation in the RPE, and is now commonpractice in most retinal clinics.

The main disadvantage of FA imaging is that it is invasive and occasionally canresultin reactions to the dye, ranging from itching (in 0.5%) and nausea (in 2.9%) toanaphylactic reaerions (in 0.2%) (50).

Indocyanine Green Angiographylndoeyanine green (ICG) angiography is an important adjunct in the evaluation ofpatients with exudative AMD, especially those wirh lPCV and RAP.

In cases of occult CNV, ICG angiography can sometimes be useful to delineatethe neovascular complex (51). Occulr CNV can appear on ICG as a plaque or an

122 SECTION /I CLINICAL SCIENCE

combinarion of both (52). fCG mac. h c I h perfluorescence, or a ..,area wit roca . Y CNV in redominantly hemorrhagic lesions (53).helpful for identifying rhe I' P ally classified on FA as occulr, ICGdeJno,f IPCV rhe esion ISusu

In the cases a , hi h pear as small areas of hyperfluorescenceinl1JJh of polyps, w IC apstrares t e presence . (54-56) The identification of polyps,andtheirto lare phases of the ICG angIOgram T ('55 57)d si h I uide argon laser or PO "

an Size, e ps to g . reas of hyperfluorescence that are usuallyt""a FA RAP lesions appear as an, I' It On ICG RAP lesions appear as a well-defined~tfied as minimally c assrc or occu . , . . fi h di .

ICG h fore allows physicians to can rrrn t e 'agnos~anjof hyperfluorescence. , r ere,. PDr' .ali he lesi hich is irnporranr If laser treatment Ot ISbe,~accurately loc rze t e csion, w. .)

id d I . combination with other therapies (58,59 .consi ere a one or III

Optical Coherence TomographyA . al I phy (OCT) appears to be useful for detecting CNY inptle co 1ererrce romogra . . " .

. . h the diagnosis of exudanve AMO IS suspected clinically,undopatIents 111 w omthese circumstances, a sensitivity of 96-97% and a specificity of 59-66% havebeen

d . . on with FA (60-62) OCT also appears to be useful fordiffer.repone In compans . .entiating classic (predominantly above the RPE) and occult (predorninanrjy under.neath the RPE) CNV (63,64). In our experience, however, rhis d,fferentlatlon15often difficult to achieve if only OCT is used. OCT may be helpful in identifYingRAP lesions by demonstrating inrraretinal areas of high reflectivity wirh imfarerin~fluid (65,66).

Recently, it was suggested that OCT could be used to monitor the activilJ'ofex.udative AMO specifically to monitor the response to anti- VEGF therapies andguideretrearmenr (62,67-72). However, the sensitivity and specificity of OCT to detecrz-rive exudative AMD in comparison with FA in this context remain [0 be elucidated.

Fundus AutofluorescenceFundus autofluorescence (AP) is emerging as a useful tool for the evaluationofP"tients with exudative AMD (48,73-82). This is highlighted by the fact that thisim-aging technique is now being included, together with FA and OCT, as part oftheroutine investigations in RCTs assessing the effectiveness of new treatments forthisdisease.

Several observarions seem to suggest rhar the integrity of the RPE layer ispreservedearly on during the course of the disease in patients with exudative AMD. Thus, inpa-tients with a CNV of recent onset, a normal or minimally abnormal distribution offundus AP at the macula is often observed (77). Furthermore, when small CNYsarepresent, the abnormalities on fundus AF seem to be restricted to the area of the lesion,with a preserved, homogeneous AF detected outside the area involved by the neovas-cular process in most cases (Fig. lORI) (74). Moreovet, a normal or near-normaldis-tribution of fundus AP has been recorded in fellow eyes of patients with exudativeAMO in whom age-related maculopathy changes were detected (74). In addition,noor minimal changes on the distribution of fundus AP, specifically a lack of increasedAP signal, have been detected before the development of CNV at the site of the lesion(Fig. IOB.2) (74). The above findings have important implications. First, and assug-gested by a recent study (75) in which a continuous and predominantly preservedAFpattern at the macula was correlated with better VA, fundus AP may be a usefulprog-nosnc tool In exudative AMO. Second, it is plausible that a relatively preservedRPEmay be needed for tmtratron of the neovascular process. In this regard. and as discussed

CHAPTER lOB FUNDUS AUTOFLUORESCENCE IN NEDVASCULAR AGE-RELATEDMACULAR DEGENERATION 123

fiGURE lOB. 1. (AI Optical coherence tomography (OCT)[axis 330 degrees), (B) fluorescein angiogram (FA),IC) conventionalfundusa.notluorescence IAF), and (DI near-infrared autofluorescence (NIA) obtained from a patient with a class.c subfovealchoroidalneovascular membrane (CNV). (AI OCT demonstrates a highly reflective lesion above the RPEsurrounded by subreti-nalfluid. IBI Early hyperfluorescence in observed on FA. (C) A decreased AF signal at the site of the CNV is seen on conven-tionalAF imaging; the neighboring RPEappears to be preserved. (D) On NIA there is reduced signal corresponding to the areaof the CNV; this reduced signal is surrounded by a ring on increased signal Note the lack of the normal pattern of increasedAFsignal at the center of the macula on NIA (see text for details].

above (see histopathology and pathogenesis section), the RPE is a major source ofVEGF, the key molecule in the development of neovascularization. This is also sup-ported by the reduced incidence of CNV in patients with geographic atrophy, and bythe fact that under the latter circumstances the CNV seems to develop in areas of rel-atively preserved fundus AF rather than in those showing abnormal and increased AF(personal observation) (Fig. 10B.3).

Although, as discussed above, the distribution of fundus AF in patients with recent-onset exudative AMD is minimally abnormal in many cases when studied by conven-tional AF imaging (wavelength of 488 run), this does not seem to be the case when near-infrared autofluorescence (NfA) is used (personal observation) (Figs. IOB.1 and IOB.2).The typical increased NlA signal at the center of the macula (see Chapter 6) is oftenlacking in patients with exudative AMD. Given that the NlA signal appears to emanatepredominantly from melanin in the RPE (83,84), conventional and NlA findings seemto indicate that a loss or depletion of melanin in a relatively intact RPE may be an

SECTIONII CLINICAL SCIENCE

FIGURE 10B.2. FA (A,CI and AF (B,D) images obtained from a patient before (A,B) and after (C,DI the developmentofaminimally classic CNV. (CI The CNV developed at a site where no changes on the distribution of AF had beenpreviouslyd.tected. The blocked fluorescence and the fovea on FA lCI and the increased AF signal at the fovea (D) appearto correspondclinically to an area of resolving blood.

initial step in the sequence of events that lead to neovascularization. In rhis regard,melanin is recognized as a protective substance by its photo-screening effect(85)and,more importantly, by its biochemical effect in reducing phorooxidative damage(8~.The melanin content of the RPE, as well as its protective effect, was found to decreasewith age (87-89). It was also found that aging melanosomes in the RPE containde·graded melanin that could be phororoxic ro the RPE and could aid in the parhogen~'of AMD (90). However, NIA is a very newly developed imaging technique andfurth"work is needed to support these preliminary observations.

It has been shown that the area of abnormal AF signal in most casesof ,xurhriv'AMD is larger than the total area seen on FA (77). This finding is also supponedbysome histopathology data (91) and may have implications for assessing the sizeofm'lesion to be treated and possibly the Outcomes following treatment,

CHAPTERlOB FUNDUS AUTDFLUDRESCENCE IN NEDVASCULAR AGE-RELATED MACULAR DEGENERATIDN

FIGURE 106.3. Early (A) and late (61 FA frames, an AF image (e), and NIA (D) of a patient with previously diagnosed at-rophic AMD and a newly developed classic CNV. Foci of increased AF signal are seen surrounding well-defined areas of re-ducedAF (RPEatrophy); these are not present at or around the site of the CNV. Note how the preservation of the foveal RPEismore readily detected by AF and NIA compared to FA

Differences in the distribution of fundus AF have been described in eyes withclassic, occult, and mixed CNV (74). A reduced AF signal at the site of classic CNVhas been observed in most cases (79-90%) (Fig. lOB.I) (73,74). In cases of occultCNV, multiple foci of low AF signal at the site of the CNV ate commonly seen(Fig. IDBA). Recent observations using combined AF and OCT imaging in pa-tients with newly developed classic CNV suggest that the reduced AF signal at thesite of classic CNV does not indicate loss/damage ofRPE, but rather, as previouslyhypothesized (74), masking of the RPE-AF signal by the CNV growing in the sub-retinal space (personal observation) (Fig. lOB,I). In contrast, it is likely thatthe foci of reduced AF signal observed in cases of occult CNV could represent smallareas of RPE loss or a more irregular pattern of growth followed by rhe CNV

,- __ ;,' H>-Q,rl .... __ ......

12li SECTION II CLINICAL SCIENCE

FIGURE 108.4. OCT(axis 90°) IA), FA 18), conventional fund~s AF (C), and NIA (DI obtained from a patient with anocculiCNV. (AI On OCTthe RPEappears to be preserved but elevated. It is difficult to establish whether there is a CNVunderneaththe RPE;there IS no obvious subretinal fluid. (81 On FA, however, elevated hyperfluorescence and late leakage correspondingto an occult CNVare observed.IC) On conventional AF imaging, a mottled AF Signal with small foci or reducedand increasedAF is seen at the site of the CNV.IDI NIA demonstrates multiple foci of reduced AF signal. There is relative preservationoftilenormally increased NIA Signal at the center of the macula despite the presence of the CNV.

AF imaging is useful for following the behavior of the CNV and the statusoftbeRPE following treatment. Framme et al. (73) found that 2-3 months afterPDTrbemajority of classic lesions became better defined, as evidenced by increasedAFsign~around the lesion compared to before treatment (73). This could be related to prolif.eration of RPE cells at the margin of the lesion, which may result from laseracriva-tion of the RPE or laser damage to the central RPE wirh more acriviryof me RPE"the perimeter of the eNV (73). This agrees with the observarions by Gass(92)cit"proliferation of RPE occurs at the edge of classic CNV induced by the separationofthe neurosensory retina from the RPE and in an attempt to engulf the new vessels.

It was recently shown that intact foveal A.F is associated with better visualout-comes following anti-VEGF therapy (93). It has also been suggested that poorerVI'sual outcomes after successful anti- VEGF therapy may be associated with arropb"RPE changes detected on AF (94).

CHAPTER lOB FUNDUS AUTOFLUORESCENCE IN NEOVASCULAR AGE-RELATED MACULAR DEGENERATION 121[ISUMMARY

The above imaging techniques can be used by the clinician, in combination, to ob-tain a more accurate diagnosis and perform an in-depth evaluation of the status of theretina.They can also guide the clinician in the treatment and counseling of patientswith exudative AMD.

FA remains, to date, the gold standard for establishing the diagnosis of exudativeAMD and is also required to determine the location, type, size, and degree of activity ofthe exudative lesion. ICG angiography is often needed to confirm the diagnosis of RAPand IPCV. ocr can be used to determine the presence of sub- or intraretinal fluid, andit maybe useful in the follow-up of patients mer treatment. AF imaging provides infor-mation on the status of the RPE and indirectly of the photo receptors; it may have aprognostic value for determining the natural history of exudative AMD and may helpto predict visual outcomes after treatment. Fundus AF may also assist in determiningthe type of CNV (classic vs. occult) and the extent of the lesion. Like conventional AF,NIA provides a means of assessing the status of the RPE; it is likely that this imagingtechnique will be of value in predicting visual outcomes following treatment. Both AFand NIA are providing insights into the pathogenesis of exudative AMD.

REFERENCESFriedman OS, Katz J, Bressler NM, er al. Racial differences in the prevalence of age-related macular degen-eration: the Baltimore Eye Survey. Ophthalmology 1999;106:1049-1055.

2. Subfoveal neovascular lesions in age-related macular degeneration. Guidelines for evaluation and treat-ment in rhe macular photocoagulation study. Macular Photocoagulation Srudy Group. Arch Ophrhalmol1991; 109;1242-1257.

3. Bird AC, Bressler NM, Bressler SB, er al. An international classification and grading system for age-relatedmaculopathy and age-related macular degeneration. The International ARM Epidemiological Study Group.Surv Opluhalmol 1995;39:367-374.

4. Massacesi Al., Sacchi L, Bergamini F, et al. The prevalence of retinal angiomatous proliferation in age-re-lated macular degeneration with occult choroidal neovascularizarion. Graefes Arch Clin Exp Ophchalmol

2008;246;89-92.5. Yannuzzi LA, Sorenson J, Spaide RF, er al. Idiopathic polypoidal choroidal vasculoparhy (Il'Cv). Retina

1990;10;1-8.6. Sho K, Takahashi K, Yamada H, et al. Polypoidal choroidal vasculoparhy: incidence, demographic fea-rures, and clinical characteristics. Arch OphthalmoI2003;12l:l392-1396.

7. Yannuzzi lA, Ciardella A, $paide RF, er a1. The expanding clinical spectrum of idiopathic polypoidalchoroidal vasculoparhy. Arch OphthalmoI1997;115:478-48S.

8. Kawasaki R, \'ifang JJ, Aung T, et al. Prevalence of age-related macular degeneration in a malay population:the Singapore Malay Eye Study. Ophthalmology 2008;115: 1735-1741.

9. Andersen MV, Rosenberg T, la Cour M, er al. Prevalence of age-related maculoparhy and age-related mac-ular degeneration among the Inuit in Greenland. The Greenland luuir Eye Study. Ophthalmology

2008;115:700-707.e1.10. Mitchell P, Smith W, Anebo K, et al. Prevalence of age-related maculoparhy in Australia. The Blue

Mountains Eye Study. Ophthalmology 1995; 102: 1450-1460.11. Vingerling JR, Dielemans I, Hofman A, er al. The prevalence of age-related maculoparhy in the

Rotterdam Study. Ophthalmology 1995; 102:205-210.12. Varma R, Fraser-BeU S, Tan S, et al. Prevalence of age-related macular degeneration in Latinos: the Los

Angeles Latino eye study. Ophthalmology 2004;111: 1288-1297.13. Oshima Y, Ishibashi T, Murata T, er al. Prevalence of age relared maculoparhy in a representative Japanese

population: the Hisayama Study. Br J OphthalmoI2001;85:1153-1157.14. Wong TY, Loon SC, Saw SM. The epidemiology of age related eye diseases in Asia. Br J Ophchalmol

2006;90;506-511.15. Li Y, Xu L, Jonas JB, et a]. Prevalence of age-related maculopathy in the adult population in China: the

Beijing Eye Study. Am J OphthahnoI2006;142:788-793. . .16. Klein R, Peto T, Bird A, et a1.The epidemiology of age-related macular degeneration. Am J Ophrhalmol

2004;137;486-495.17. Argon laser photOcoagulation for idiopathic neovascularization. Results of a randomized clinical trial. Arch

-- __ '""'fc--:~.I-.;,;-A~ 8-1361.


D d k SS J ki SAPerc T er at Autofluorescence imaging of choroidal neovascularizarjond

77. an e ar ,en ns oo, " ueroage-related macular degeneration. Arch Ophthalmo1 2005; 123:150~-1513.

R k A F·· k FW Bird AC Fundus autofluorescence 10 age-related macular diseasei d

78. von uc mann ,"Itz e,' . . mageicl I . hrl lmoscope Invest OphrhalmoJ VIS SCI 1997;38:478--486.Wl 1 a aser scanning op lill·. .

79. Lois N, Halfyard AS, Bird Ac' et a]. Quantitative evaluation 4°5ffundus autofluorescence Imaged"invivo'

in eyeswith retinal disease. Br J Ophthalmol 200,0;84:741-7 . ".80. Stanga PE, KychenrhaJ A, Pirzke FW, et al. F~nCtlonal assessment of the native retinal ~Ig~entepithelium

after the surgical excision of subfoveal choroidal neovascular membranes rype II: prehnuoal)'results,Inl

Ophrhalmol 2001 ;230309-316.81. Lois N, Owens SL, Coco R, et a]. Fundus autofluorescence in patiems with age-relatedmaculardegenera.

rion and high risk of visual loss. Am ] Opbthalmol 2002; 133:341-349. .82. Spaide RF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology2003;IlD:

392-399.83. Keilhauer CN, Delori Fe. Near-infrared autofluorescence imaging of the fundus: visualizationofocularmelanin. Invest Opbthalmol Vis Sci 2006;47:3556-3564.

84. Weinberger AW, LappasA, Kirschkamp T, er al. Fundus near infrared fluorescencecorrelateswithfuodwnear infrared reflectance. Invest Ophthalmol Vis Sci 2006;47:3098-3108.

85. Sarna T. Properties and function of the ocular melanin-a photobiophysical view.] PhorochemPhotobiol

B 1992;120215-258.86. Wang Z, Dillon ], Gaillard ER. Antioxidant properties of melanin in retinal pigment epithelial,elb.

Pborochem Phorobiol 2006;82:474-479.87. Sarna T, Burke ]M, Korytowski W, er al. Loss of melanin from human RPE with aging:possibleroleof

melanin phorooxidarion. Exp Eye Res 2003;76:89-98.88. Weiter JJ, Delori Fe, Wing GL. et a]. Retinal pigment epithelial lipofuscin and melaninandcboroidal

melanin in human eyes. Invest Ophrhalmol Vis Sci 1986;27: 145-152.89. Boulton M, Dayhaw-Barker P. The role of the retinal pigment epithelium: topographicalvariation and

ageing changes. Eye 2001;15:384-389.90. Rozanowski B, Cuenca ], Davies 5, er al. The phototoxiciry of aged human retinal melanOliomes.

Phorochem PhotobioI2008;84:650-657.91. Bynoe lA, Chang TS, Punara M, et al. Histopathologic examination of vascular patterns in subfovealoro-

vascular membranes. Ophthalmology 1994; 10 J : 1112-1117.92. Gass J. Scerioscopic arias of macular diseases: diagnosis and treatment, 4th ed. St. Louis:Mosby,

19960554.93. Kozak 1,Mojana F, Morrison VL, et al. Autofluorescence imaging in exudative age-relatedmacularde-

generation after anti-VEGF treatment. ARVO 2008;264.94. Perumal B, Lee )), Bearely S, er al. Correlation of RPE atrophy measured with fundus aurofluoresonc

with visual acuity outcomes in patients with NVAi\ID responsive to ami-VEGF therapy.ARVO2008;2238.

CHAPTER

Fundus Autofluorescence in AtrophicAge-Related Macular Degeneration

n :rop~ic age-related macular degeneration (AMD) represents the late stage of~ dry AMD. It IS characterized by the development of atrophic patches, which

may initially occur in the parafoveal area (1-4). These atrophic areas appearfunduscopically as sharply demarcated areas of depigmentation rhrough which deepchoroidal vessels can be seen. During the natural course of the disease, the areas of at-rophyslowly enlarge over rime. In atrophic AMD, characteristically, the fovea remainsuninvolved until the advanced stages of the disease, a phenomenon referred to as"fovealsparing." A widely established term used ro refer to this advanced form of dryAMD is "geographic atrophy" (GA). By contrast, the term "areolar choroidal atrophy"is usually used co refer to findings that are similar but caused by monogenetic retinalmacular dystrophies, which may manifest earlier in life.

Severe visual loss secondary to GA occurs in about 20% of all patients with AMD(5-8). Hence, GA is, after choroidal neovascularization (CNY), the second mostcommon cause of legal blindness due to AMD. Patients with primary GA tend to beolder than those with neovascular forms of AlvID at the time of initial presentation,and it has been speculated that GA occurs in eyes in which a neovascular angiogenicevent has not developed. As opposed to the recent breakthrough with anti-VEGF(vascular endothelial growth factor) therapy for active neovascular AMD, there is todate no treatment available for patients with GA, other than visual aids and visual re-habilitation. Therefore, a berter understanding of the pathogenesis ofGA appears tobe mandatory. Sensitive diagnostic tools and prognostic markers to evaluate diseasestage and future progression in the individual patient are needed.

It is clinically well established rhat GA atrophy can affect one or both eyes (9,10).The fellow eye can be affected by any other AMD manifestation and development,including CNV or disciform scarring. It has been reponed that eyes demonstratingtypical early features of GA can also develop CNV; in these patients, severe and sud-den visual loss occurs (11). When CNV develops in an eye with previously diagnosedpure GA, it usually has an evanescent appearance and it is often difficult to outline itsborders and differentiate between hyperfluorescence resulting from the CNV andthat caused by atrophic areas. In this context, careful slit-lamp biomicroscopy mayallow the identification of subretinal fluid or hemorrhage, which would indicate thepresence of a CNV. Optical coherence tomography (OCT) may also be helpful inidentifying a CNV in these cases.

PATHOPHYSIOLOGYAreas of GA in AMD occur at sites where macular changes at the level of the retinalpigment epithelium (RPE) and Bruch's membrane, such as pigmentalY alterations anddrusen, are present (9,10,12,13). Regression of confluent, large, soft drusen may lead

131


to atrophy (I). Similarly, calcified deposits seem to correlate well with the develop-ment of atrophy (14), In some cases, GA occurs following the collapse and flattening

of RPE detachments (I).The pathophysiological mechanisms underlying the atrophic process are not com-

pletely understood. It is believed that lipofuscin, a by-product of incompletely digestedphotoreceptor outer segments that accumulates in RPE cells m atrophic AMD, playsakey role in the disease process (see also Chapters 1 and 2) (15). Histopathological stud-ies have shown that clinical visible areas of atrophy are confined to areas of RPE andphotoreceptor cell loss, and choriocapillaris closure (13,16,17). Furthermore, lipofuscinand melanolipofuscin-engorged RPE cells have been observed in the junctional zone be-tween the atrophic and the relatively normal retina, whereas in areas of atrophy thereis

loss of RPE and thus oflipofuscin granules (17).These postmortem observations suggest that lipofuscin may playa direct patho-

genetic role in atrophic AMD by causing RPE cell death, with subsequent deleteriouseffects on phororeceptors and choriocapillaries. Alternatively, it is possible that the ex-cessive accumulation of lipofuscin is an expression of RPE cell dysfunction and is thus

the result, not the cause, of it.

IMAGING TECHNIQUES

Fundus PhotographyThe discrete areas of loss of RPE associated with loss of overlying phororeceprors inGA appear as areas of depigmentation on slit-lamp biomicroscopy in comparisonwith surrounding normal retina. The decreased retinal thickness at such sites is usu-ally not visible on nonstereo images. Deep, large choroidal vessels may be apparentand more distinctly visible through areas of atrophic retina. Due to the low contrast,particularly in the red spectrum, the distinction of GA on fundus photographs ischal-lenging and other imaging modalities ate required to accurately identify atrophic

patches,

Fluorescein AngiographyOn fluorescein angiography (FA), atrophic areas are much better delineable. Theseappear as areas of discrete hyperfluorescence, representing a transmission defectwithmild late staining. It may be difficult to differentiate between areas of atrophy andareas with fibrosis, regressed CNV, or hyperfluorescence from other causes.

Indocyanine Green AngiographyDue to atrophy of the choriocapillaris, GA appears as areas of discrete hypofluores-cence With loss of the normal background signal on indocyanine green (ICG) angiog-raphy. Latger,. deeper choroidal vessels are clearly visible. Compared to FA, it appear,to be more difficult to distinguish between atrophic and nonatrophic retinausingrcc.

Optical Coherence Tomography~etinal thinningwith loss of outer retina is observed over GA on OCT scans. SpeaIl1-

ornam OCT with SImultaneous confocal scanning laser ophthalmoscopy (cSLO

CHAPTER lOC FUNDUS AUTOFLUORESCENCE IN ATROPHIC AGE-RELATED MACULAR DEGENERATION 133

,I

IIII

lowsfor better 3D assessmenr of rerinal abnormaliries. Recenr dara suggesr that OCTcanrevealhighly variable morphological alterations in the atrophic area and in rhe sur-roundingrerina in eyes wirh funduscopically uniform-appearing GA (18).

Fundus AutofluorescenceFindingson Fundus AutofluorescenceinGeographic AtrophyFundusAF findings in patienrs wirh GA are in accordance wirh histoparhologic find-ings(19,20). Thus, because of rhe lack ofRPE lipofuscin, which contains rhe dominantfluorophoresinvolved in rhe AF signal (see Chapter 3), AF imaging shows a markedlyreducedAF signal at rhe sire of atrophic areas (Fig. lOCI). Compared with drusen,which may a!so exhibit a decreased AF signal, atrophic areas typica!ly show an evenstronger reduction of AF (21). The high-contrast difference between atrophic andnonatrophic retina allows for much better delineation of atrophic areas with AF

FIGURE lOC.l. Fundus photographs and autofluorescence (AF) images of two patients with geographicatrophy (GAl secondary to age-related macular degeneration (AMDI. In one of these patients an ill-definedarea of hypoprgmentation is observed. as well as calcified drusen (top lem A fundus AF image obtained fromthe same patient I top rigM better delineates the area of atrophy. Furthermore. surrounding areas of atrophy.at the Junctional zone between atrophic and relatively preserved retina. markedly increased AF signal is de-tected (top rlgM. this does not correlate clinically with any obvious changes. Similar findings are observed

in another patient (bottom left and righOwith atrophic AMO.


FIGURE 10C.2. Fundusphotograph of the left eye of a patient with atrophic AMD [leflj Singleat-rophic patches are clearly detected and quantified by customized image analysis software [detectedareas are markedby a white circlel on the fundus AF image (rightl. Note that drusenandpigmentabnor-malities seen on the fundus photograph can be easily differentiated from atrophic areaswith fund>sAf

imaging.

imaging compared to conventional fundus photographs (22,23). In contrast to FA, AFimaging is a noninvasive and less time-consuming method.

AF imaging has been applied to detect and precisely quantify atrophic areasincombinarion with customized image analysis software (Fig. lOC.2) (22,23). Thissoftware enables automated segmentation of atrophic areas by a region algorithmInterfering retinal blood vessels, which have similar AF intensities, are recognizedbythe software and can be excluded from the measurements. Further developments in-clude the alignment of AF images at different examinations using retinallandmarloto correct for different scan angles and magniftcations. This allows the ptogressionofthe atrophic process to be accurately assessed over time, and can be used in longitudi-nal observations, including interventional trials.

An even more striking finding of AF imaging in GA patients, which does not usu-ally correlate wirh obvious changes on fundus photography or FA, is rhe frequent visu·alization of high-intensity levels of AF surrounding the atrophic patches at the junotional zone of atrophy (Fig. lOC.l) (24). This observation is in accordance wilnhistoparhological data showing lipofuscin-engorged RPE cells surrounding areasofaI'rophy (see above).

Clinical Significance of Fundus Autofluorescence FindingsinGeographic AtrophyStudies of photoreceptor function have underscored the importance of abnormal AF in-rensines around atrophy and the pathophysiological role of increased RPE lipofuscinaI'cumularion in patients with GA due to AMD. Scholl and coworkers (25) demonstm,e<I[~at r~d photoreceptor function is more severely affected than cone function over arealwith increased AF using fine-matrix mapping. Wirh a combination of SLO-baseJ mi-croperimetty and AF imaging, impaired photopic sensitivity has been observed inare"of abnormal AF III rhe JunctlOnal. zone (26). Because normal photoreceptor funcrionis

dependent on normal RPE function, particularly wirh regard to rhe constant phagOCY'toSIS of shed distal outer segment discs for photoreceptor cell renewal, a negaovefe«!·back mechanism has been proposed whereby cells with lipofuscin-loaded secon@)lysosomes would phagocyto £; - h d h clse ewer s e p otoreceptor outer segments) subsequen }'

FUNDUS AUTOFLUORESCENCE IN ATROPHIC AGE-RELATED MACULAR DEGENERATION 135CHAPTER lOC

leadingto impaired retinal sensitivity in areas with increased AF intensity. This wouldalsobe in accordan~e w.ith experim~ntal data showing that compounds of lipofuscin,such as A2-E (N-retmybdene-N-retlnylethanolamin), possess toxic properties and mayinterferewith normal RPE cell function (see also Chapter 2) (27,28). The higher sus-ceptibiliryof the rod system relarive to the cone system may be explained by rhe relativererinoiddeficiency in cones.

The Fundus Autofluorescence in Age-related Macular Degeneration (FAM)Study Group introduced a classification system for distinct patterns of abnormal ele-vatedAF in the junctional zone ofGA (29). This morphological classification is basedon information that is solely detectable by AF imaging and consists of five main pat-terns: none, focal, banded, diffuse, and patchy. The "diffuse" pattern is further sub-divided inro five subtypes. These disrinct AF phenotypes may reflect heterogeneity ofrhe underlying disease process. The concomitant observation of a high degree of sym-metry of AF patterns in patients with bilateral GA in the presence of a high degree ofinrerindividual variability also suggests that genetic determinants rather than nonspe-cific aging changes may be involved (30).

Impact of Fundus Autofluorescence Findings on GeographicAtrophy ProgressionWhile the presence of atrophy represents a nonspecific end-stage manifestation of var-ious retinal degenerations, the identification of elevated levels of AF at the junctionalzone of atrophy is of particular interest. The distinct patterns of AF observed sur-rounding areas of GA, as described in a cross-sectional study undertaken by the FAMStudy Group (see above), seem to be of prognostic value for predicting the speed ofatrophy progression. This may be even more important when one considers the highvariability of atrophy progression over time among patients with GA, which has beenindependently demonstrated by several natural history studies (2,31,32). Currentdata on the spread of atrophy suggest that atrophy expands linearly over time, andthat the best predictor for this expansion appears to be the growth rate observed in theprevious year (33). Atrophy enlargement of very small areas (less than one disc area orca. 2.5 mm2) has been shown to be less rapid than that of larger areas; however, theoverall difference of atrophy progression between eyes (0-13.8 mm 'zper year) couldnot be explained by baseline atrophy or by any other tested demographic factors (34).

Early pilot studies with AF imaging for atrophy progression in GA patients demon-strated the occurrence of new atrophic patches and the spread of preexisting atrophysolely in areas with abnormally high levels of AF at baseline, suggesting that levels of in-creased AF precede cell death and, therefore, absolute scotoma (24). With larger groupsof patients and longer follow-up, it was shown that the extension of the total areaof in-creasedAF surrounding areas of atrophy at baseline had a strong positive correlation withthe rate of progression of atrophy over time (35). A recent analysis of atrophy progressionrates over time and AF patterns at baseline revealed that variation in GA growth rates isdependent on the specific phenotype of abnormal AF at baseline (31). The progressionrates in eyes with banded (median 1.81 mrnvycar, n = 24 eyes) and diffuse(1.77 mm2/year, n = 112) AF patrerns were significantly higher compared to those ineyes without AF abnormalities (0.38 rrurr'Zyear. n = 17) and focal FAF patterns(0.81 mm2/year, n = 14, P < 0.0001; Figs. 10C.3 and 10CA). Within the group of thediffuse pattern, eyes with the diffuse-trickling pattern exhibited an even higher spreadrate (median 3.02 mm2/year, n = 9) compared to the other diffuse types (1.67mm2/year, p = 0.001; Fig. 10CA). This study represents the largest longitudinal studyof AF fmdings in GA patients so far (median follow-up 1.80 years, range 0.52-7.14; n= 195 eyes of J 29 patients). Overall, phenotypic features of AF abnormalities had amuch stronger impact on atrophy progression than any other risk factor (including size


FIGURE 10C.3. Illustration of the relationship between specific fundus AF phenotypes andatrophyprogression in patients with GA due to AMD (Part I: slow progressorsl showing the baselineAFimage(left) and the follow-up AF image [(/917t)for each eye, respectively. Eyes with no abnormalAFchanges[top row: atrophy progression 0.02 mm'/year, follow-up 12 monthsl and with onlysmall areas offocallyincreased AFat the margin of the atrophic patch (bottom row: 0.36 mm'/year, follow-up15monthslus"·ally have very slow progression over time. (Adapted from Holz FG, Bindewald·Wittich A,FleckensteinM,et a! Progression of geographic atrophy and impact of fundus autofluorescence patterns In age·relatedmacular degeneration. AmJ Ophthalmol 2007;143:463-472.)

of baseline atrophy, history of smoking, hypertension, diabetes, age >80 years, hyperlipi-demia, and family history) that has been addressed in previous studies on progressionofGA due to AMD. These findings also suggest that AF phenotypes may be used to expjainthe great heterogeneity of atrophy progression rates among different patients (36).

Fundus Autofluorescence to Evaluate the Response to FutureTherapeutic StrategiesCurrently, there is no treatmenr available to halt at slow the progression of atrophicAMD. With the noninvasive visualization of prognostic determinants, the merehclicmapping of functional changes and the ability to accurately monitor the disease overtime,AF imaging may not only conttibute to the understanding of atrophic AMD, it mayahabe use~ to develop and assess new emerging therapeutic strategies for this disease. .

VISUal cycle modulators, which aim to target the detrimental accumulation ofto"clipofuscin in the RPE, are pharmaceutical agen ts that may slow down the progt"s;on

CHAPTER lOC FUNDUS AUTOFLUORESCENCE IN ATROPHIC AGE-RELATED MACULAR DEGENERATION 137

of atrophy. One agent is Fenretinide (N_[4_hydroxyphenyl]retinamide), an oral com-pound that has been shown to lower the production of toxic fluorophores in the RPEin a dose-dependent manner in albino ABCA4-1- mice (37). This vitamin A derivativeacts by competing with serum retinol for the binding sites of retinal-binding proteinand promotes renal clearance of retinol. The bioavailability of retinol for the RPE andphororeceptors is consequently reduced and less toxic retinoid by-products, such asAl-E, are generated. A Phase II randomized, double-masked, placebo-controlled multicen-ter srudy that included over 200 GA patients was initiated in 2006 (SirionTherapeutics, Tampa, FL; http://www.siriontherapeutics.com). The therapeutic con-cept of Fenretinide is underscored by growing evidence from experimental and clinicalstudies, including AF findings, on the pathophysiological role of deleterious accumula-tion of lipofuscin. To reduce the observational period in a slowly progressive disease,

FIGURE 10C.4. Illustration of the rela-tionship between specific fundus AF pheno-types and atrophy progression in patientswith GA due to AMD (Part I slow progres-sors) showing the baseline AF image (ietliand the follow-up AF image (righO for eacheye. respectively. Eyeswith no abnormal AFchanges (top row: atrophy progression 0.02mm2jyear, follow-up 12 months) and Withonly small areas of focally increased AF atthe margin of the atrophic patch [bottomrow: 036 mm'jyear, follow-up 15 monthslusually have very slow progression overtime. (Adapted from Holz FG, Bindewald-Wittich A, FleckensteinM, et al. Progressionof geographic atrophy and impact of fundusautofluorescence patterns in age-relatedmacular degeneration. Am J Ophthalmol2007;143463-472.1

. . . I' d better demonstrate possible treatment effects,p .rrumrmze the samp e Size, an .... k c. ahent. . hi dv i olves the idennficanon of hlgh-ns reatures accordinrecrtu tment Il1 t 15 stu y IIlV S d G _d gto

the AF pattern classification described by the PAM . tll Y roup. an only"rapid" . I d d i this large inrerventlOnal rrial in patients With GAseco dprogressors are me u e In - vlJ .

aryroAMD.


DIFFERENTIAL DIAGNOSISThe differential diagnosis ofGA includes scar tissue or fibrosis following tegressionofCNV, trauma, posriuflammarion, or other causes. Notably, GA IS a nonspecific diseasemanifestation and can be the result of retinal disorders other than AMD, such asreti.nal and macular dystrophies (e.g., Stargardt disease and Best disease), central areolarmacular dystrophy, and cone dystrophy. To establish the diagnosis ofCA, it mighrbehelpful ro evaluate retinal changes in the fellow eye and look for retinal abnormalitiosurrounding the area of atrophy. For example, hard and soft drusen might be veryin.dicarive of CA, while focal yellow flecks would suggest Srargardr disease, AF imagingmay be helpful for evaluating retinal abnormalities surrounding areas of atrophy.

SUMMARYFundus AF imaging allows for accurate delineation of areas of atrophy. Precisequan-tification of these areas is possible with AF, which is also useful for monitoring theprogression of the disease. Distinct phenotypic patterns of abnormal AF in the junc-tional zone of atrophy are associated with significant differences in rates of arrophyprogression over time, and appear to have the strongest predictive value compared toother factors, including smoking, age, and size of the area of atrophy at baseline.Ar",of increased AF in the junctional zone of atrophy are correlated with decreased reti-nal sensitivity, which may reflect the parhophysiologic role of increased lipofuscinac-cumulation. AF imaging has enhanced our understanding of the disease and mayprove to be essential in monitoring the response to therapeutic interventions.

REFERENCES1. Blair C]. Geographic atrophy of the retinal pigment epithelium. A manifestation of senile macular degea-

eraticn. Arch Ophrhalmol 1975;93: 19-25,

2. 5unness J5. The natural history of geographic atrophy, the advanced atrophic form of age-relared maculardegeneration. Mol Vis 1999;5:25.

3. Maguire P, Vine AK. Geographic atrophy of rhe retinal pigment epithelium. Am J Ophthalmol1986:102,621-625.

4. Schatz H, McDonald HR. Atrophic macular degeneration. Rate of spread of geographic atrophy and vi-sual loss. Ophthalmology 1989;96; 1541-1551.

5. Klei.n R, Klein BE, Knudtson MD, er al. Fifteen-year cumulative incidence of age-related macular degen-eration: the Beaver Dam Eye Study. Ophthalmology 2007;114:253_262.

6. Wang J], Roc.htchina E, Lee AJ, er al. Ten-year incidence and progression of age-related maculoparhy rheBlue Mountains Eye Study. Ophthalmology 2007;114;92-98.

7. Allgood CA, Vingerling JR, de ~ong PT, er a]. Prevalence of age-related maculopathy in older Europeans:the European Eye Study ~E~YE). Arch Ophthalmo!2006;124;529_535.

8. Kl~ver C~, Wolfs RC, Vmge.rhng JR, et al. Age-specific prevalence and causes of blindness and vislJ~im-panmenr In an older population: the Rotterdam Study. Arch Ophrhalmol 1998; 116:653--658.

9. Gass JD. Drusen and disciform .macular derachmenr and degeneration. Arch Ophthalmol 1973;90:2010. Green WR, Key 3rd SN. SenJle macular degeneration: a histoparhoiocj d T Am 0 hth

Soc 1977;75:180-254. Oglc Stu y. rans p

FUNDUS AUTOFLUORESCENCE IN ATROPHIC AGE-RELATED MACULAR DEGENERATION 139

11. Sunness jS, Gonzalez-Baron), Bressler NM, er al. The development of choroidal neovascularizarion ineyes with the geographic atrophy form of age-related macular degeneration. Ophthalmology 1999;106:910-919.

12. Sarks SH. Drusen patterns predisposing to geographic atrophy of the retinal pigment epithelium. ALLstJOphrhalmol1982; 10:91-97.

13. Sarks JP, Sarks SH, Killingsworth Me. Evolution of geographic atrophy of the retinal pigment epithe-lium. Eye 1988;2:552-577.

14. AREDS. A randomized, placebo-controlled, clinical uial of high-dose supplementation with vitamins Cand E, beta carorene, and zinc for age-related macular degeneration and vision loss: AREDS report no. B.Arch Opbchalmo) 200 1;119: 1417-1 436.

15. Sparrow JR, Boulton M. RPE lipofuscin and its role in retinal parhobiology. Exp Eye Res 2005;80:595-606.

16. Creeo WR, Engel C Age-related macular degeneration: hisroparhologic studies: the 1992 Lorenz E.Zimmermann Lecture. Ophthalmology 1992;100: 1519-1535.

17. Sarks SH. Ageing and degeneration in the macular region: a clinico-pathological study. Br J Ophrhalmol1976;60324-341.

18. Fleckenstein M, Charbel Jssa P, Helb HM, er a]. High resolution spectral domain-OCT imaging in geo-graphic atrophy associated with age-related macular degeneration. Invest OphthalmoI2008;49:4137-4144.

19. von Ruckmann A, Firzke FW, Bird AC. Disrriburion of fundus aurofluorescence with a scanning laserophthalmoscope. Br J Ophrbalmol 1995;79:407-412.

20. Holz FC, Bellmann C, Margaritidis M, et al. Patterns of increased in vivo fundus aurofluorescence in thejunctional zone of geographic atrophy of the retinal pigment epithelium associated with age-related mac-ular degeneration. Graefes Arch Clin Exp Ophchalmol 1999;237: 145-152.

21. Delori FC, Fleckoer MR, Coger DC, er a]. Autofluorescence distribution associated with drusen in age-related macular degeneration. Invest Ophchalmol Vis Sci 2000;41 :496-504.

22. Schmic-Velckenberg 5, Jorzik J, Unnebrink K, et a]. Analysis of digital scanning laser ophthalmoscopyfundus autofluorescence images of geographic atrophy in advanced age-related macular degeneration.Graefes Arch CEn Exp Ophrhalmol 2002;240:73-78.

23. Deckert A, Schmiu-Valckenberg S, jorzik J, er a1. Automated analysis of digital fundus autofluorescenceimages of geographic atrophy in advanced age-related macular degeneration using confocal scanning laserophthalmoscopy (cSLO). BMC Ophchalrnol 2005;5:8.

24. Holz FG, Bellman C, Staudt 5, ct a]. Fundus autofluorescence and development of geographic atrophy inage-related macular degeneration. Invest Ophrhalmol Vis Sci 2001;42:1051-1056.

25. SchoU HP, Bellmann C, Dandekar 55, et al. Photopic and scoropic fine matrix mapping of retinal areas ofincreased fundus autofluorescence in patients with age-related maculoparhy. Invest Ophchalmol Vis Sci2004;45574-583.

26. Schmitz- Valckenberg 5, Bulrmann S, Dreyhaupr .l- er a]. Fundus autofluorescence and fundus perimetryin the junctional zone of geographic atrophy in patients with age-related macular degenerario». InvestOphchalmol Vis Sci 2004;45:4470-4476.

27. Zhou J, jang YP, Kim SR, er al. Complement activation by phorooxidation produces of A2E, a lipofuscinconstituent of the retinal pigment epithelium. Proc Nat! Acad Sci USA 2006; 103: 16182-16187.

28. Bergmann M, Schutt F, Holz FG, er a]. Inhibition of the ATP-driven proton pump in RPE lysosomes bythe major lipofuscin fluorophore A2-E may contribure to the pathogenesis of age-related macular degen-eration. FA5EBJ 2004;18:562-564.

29. Bindewald A, Schmitz-Valckenberg S, Janik J), ct a1. Classification of abnormal fundus autofluorescencepatterns in the junctional zone of geographic atrophy in patients with age related macular degcneration. BrJ OphthalmoI2005;89:874-878.

30. Bellmann C, Jorzik J, Spital G, et al. Symmetry of bilateral lesions in geographic atrophy in patients withage-related macular degeneration. Arch Ophthalmol 2002; 120:579-584.

31. Holz FG, Bindewald-Wirrich A, Fleckenstein M, et al. Progression of geographic atrophy and impact offundus autofluorescence patterns in age-related macular degeneration. Am J Ophthalmol 2007;143:

463-472-32. Sun ness J, Margalir E, Srikurnaran 0, et al. The long-term natural history of geographic atrophy from age-

rdated macular degeneration. Ophthalmology 2007;114:271-277.33. Dreyhaupt j, Mansmann U, hitsch M, et al. Modelling the natural history of geographic atrophy in pa-

dents with age-related macular degeneration. Ophthalmic Epidemiol 2005; 12:353-362.34. Sunness jS, Gonzale:;:-Baron J, Applegate CA, et al. Enlargement of atrophy and visual acuity loss in the

geographic atrophy form of age-related macular clegenerarion. Ophthal mology 1999; 106: 1768-1779.35. Schmitl-Valckenberg S, Bindewald-\Vittich A, Dolar-Szczasny J, et al. Correlation between the area of in-

creased autofluorescence surrounding geographic atrophy and disease progression in patienrs with A.MD.Invest Ophthalmol Vis Sci 2006;47:2648-2654.

36. Schmitz-Valckenberg S, Fleckenstein M, Scholl HP, et al. Fundus autofluorescence and progression ofage-related macular degeneration. 5urv Ophthalmol 2009;54(1):96-1 17.

~ ,~adu RA Han Y Bui TV, et al. Reductions in serum vitamin A arrest accumulation of roxic rerinalherapy for treatment of lipofuscin-based retinal diseases. Invest Ophthalmol01.

CHAPTER

m etinitis pigmentosa (RP) is a heterogeneous group of disorders characterizedbyprogressive retinal dysfunction affecting mainly rods WIth secondaryconeIfl.volvernenr. Patients with RP classically presem with impaired nightVISIOnand

progressive visual field constriction with ultimate l~,ss of central ~,ision.Typically,"photoreceptors die, intraretinal pIgment mIgratIOn ( bone-spicnles") ISobservedalongthe midperipheral retina. Retinal blood vessel arrenuauon and disc pallor areseeninad.vanced stages. Fundus examination can be normal, especially early in me courseofthedisease. Full-field electroretinogram (ERG) demonstrates rod-cone dysfunctionandmay be essential in establishing the diagnosis and severity of the condition.

Autosomal dominant (AD), recessive (AR), and X-linked forms ofinheritancecanbe observed in families affected with RP, with tare cases of mitochondrial anddigenicforms. In general, AD RP has a berrer prognosis for retention of cenrral vision than re-cessive or X-linked disease, although there is wide mutation-specific variability. In ad.dition, RP can be isolated Ot be part of a syndrome. More than 180 genes implicatedin retinal dystrophies have been mapped and more 120 genes have been cloned,high.lighting the complexity of the disease (for a recent summary, see the Redna!Information Network, hrtp://www.sph.uth.tmc.edu/Rernetl). However, morethanhalf of all cases are due to unidentified genetic defects. Thus, precise phenoryping ucritical to better understand the disorder, identify candidate genes, and to developnovel therapies.

•Fundus Autofluorescence InRetinitis Pigmentosa

MOLECULAR BASIS AND PATHOLOGY

140

The majority of cases of RP are nonsyndromic (isolated); about 25% occur asparrofa syndrome, the most common being Usher syndrome (RP and neurosensory hearingloss), which represents 14% of a!l RP cases (I). Other syndromic forms includeBatdet-Biedl syndrome (RP, polydactyly, obesity, rena! abnormality, and mental reotardauon), Refsum disease, and other disorders associated with renal, metabolic, skele-tal, or neurological disease (2,3). The prevalence of isolated RP varies accordingrotheinheritance pattern; recent estimates suggest that AD RP forms account for approxi-mately 30%, AR RP for 20%, and X-linked for 15% of cases (4). A further 5%havebeen classified as early-onset forms of RP (Leber congenita! amaurosis). The other30% represenr sporadic cases, which are most likely to be AR, but X-linked or denovodominant forms might also be included in this group (4). To add to the complexity,different genes have been Implicated in similar patterns of inheritance. Since the iden-tification of the RHODOPSIN (RHO) gene as rhe first gene implicated in AD RP (5),more than 180 genes have been linked to the physiopathology of rerinal dystrophies.!he ~roducts of these genes are implicated in very diverse cellular functions,mcludmg the phototransduction cascade retinoid cycle h rures

' , p otoreceptor struc ,

transcription factors, outer segment renewal, splicing factors, and intracellular traf-ficking (3). For most of these genes, expression is restricted to the photoreceptors,especially rods, and/or the retinal pigment epithelium (RPE), but others, such assplicing factors, are more ubiquitously expressed. The mechanism by which a ubiq-uitously expressed gene is responsible for a restricted photoreceptor disease is notwell understood, but it may relate to the uniquely high metabolic demand of thephoto receptors.

In addition, in AD RP, nonpenetrance has been described associated with certaingenotypes (e.g., PRPF31 mutations [6]), which may complicate genetic counseling.Some genes are implicated in different phenotypes; for instance, RHO is the genemost commonly implicated in AD RP (20%-30% of AD RP [7]), bur it can also beinvolved (less commonly) in AR RP and AD congenital stationary night blindness(8,9). Similarly, NR2E3, which is responsible for Enhanced S Cone syndrome (10), arecessive disorder, can be found in 1%-3% of cases of AD RP (11,12).

Because of the genetic heterogeneity of RP, the precise molecular mechanisms lead-ing to photoreceptor cell death are still not fully understood. It is thought that photore-ceprors degenerate through a common final pathway by apoptosis (13,14), which mayinvolve calpains (15,16). Causes of apoptosis include ionic imbalance, protein aggre-gates, or default in photoreceptor strucrure (3). Explanations have also been given to thesecondary cone cell dearh in the case of mutation in genes expressed only in rods, andevidence suggests that cones rely on rods to survive (17,18).

CHAPTER llA FUNDUS AUTDFLUORESCENCE IN RETINITIS PIGMENTOSA 141

FINDINGS ON FUNDUS EXAMINATIONFundus examination can be normal (RP sine pigmento), especially in the early stages ofthe disease. In the course of RP, RPE changes will appear in the midperiphery withpigment migration into the inner retinal layers as photoreceptors die. Blood vessel at-tenuation and pallor of the optic disc ate hallmarks of advanced stages and are thoughtto result from the decreased metabolic demand with photoreceptor degeneration. Theposterior pole is usually preserved until late in the course of the disorder. However, inabout 63% of all cases of RP, foveal lesions are ptesent, including atrophic changes(43%) and cystic changes (20%), mainly oedernarous, that can be documented withfluorescein angiography (FA) or optical coherence tomography (OCT) (19-21) (seebelow). These macular changes will cause early loss of central vision. Premature poste-riorsubcapsular cataract may also occur, resulting in impaired visual acuity.

ELECTRODIAGNOSTIC FINDINGSERG is essential for establishing the diagnosis of RP when there are no or only subtleabnormalities on fundus examination. In RP, the dark-adapted bright flash ERG a-wave, which predominanrly reflects rod photoreceptor hyperpolarization, is abnormalwith milder photopic ERG abnormalities, indicating milder cone-system involve-memo The degree of macular cone system involvement varies among patients andmay be assessed by pattern ERG (PERG) (22) and multifocal ERG (mfERG) (23).The PERG P50 component is a response to an alternating checkerboard stimulus thatdepends on the integrity of macular cones and has been used extensively as an objec-tive index of macular function (24,25). Multifocal ERG is typically performed usinga stimulus array comprised of 61 or 103 hexagonal stimulus elements centered at thefovea and usually extending over the central 55-60 degrees. The mfERG responses


FIGURE l1A.2. FundusAF images obtained in two individuals with AD RP(mutation in PRPFJ.RP18)andgood visual acuity (A,CI compared with two individuals with X-linked cone-rod dystrophy(mutation'"exon ORF15of RPGR-RP3 [B,Oll. (AI from the maternal aunt (35 years oldl of (ClI17 yearsoldl;181 fromthe maternal nephew [40 years old I of (D) [74 years old I.

amaurosis (see also Chapter lID) (49), cone-rod dystrophy (Fig. IIA.2B,D) (50--54)and "cone dystrophy with supernormal rod ERG" (see also Chapter II B) (55,56), Be~disease (see also Chapter IIF) (57), Xclinked retinoschisis (see also Chapter l lC)(58),and other forms of maculopathy (59,60). The incidence of parafoveal ringsof increasedAF in non-Rf cases was recently reviewed (30). In cone-rod dystrophy, the annulsareas of increased AF may be indistinguishable from those seen in RP (Fig. IIAlA,B)(see also Chapter lIB), and full-field ERG may be essential in rhe differentialdiagno-sis. However, once the cone-rod dystrophy progresses, arrophic macular RPE changousually ensue, demonstrating a reduced AF signal within the ring (Fig. IIA2D). Animportant caveat is that dense macular pigment may resemble early atrophic changes"the fovea, and two-wavelength AF (61) utilizing wavelengths that are differendyabsorbed by luteal pigment may help to identify subtle abnormalities. Unlike in RP,the ring of increased AF in cone-rod dystrophy evolves differently as the maculopathyworsens (see later).

In sector RP, fundus AF can be decreased within the inferior vasculararcado,showing approximate correspondence with areas of superior visual field loss(62).Asemicircular parafoveal arc of increased AF may be present inferior ro the foveawi<hpatchy AF changes at the level of the vascular arcades (59). Figure IIAJA showsatypical example from a patient with sector RP who had full-field ERGs consisten'with restricted photoreceptor disease; multifocal ERG reducrion and superiorVi5U~field loss were concordant with the localized inferior retinal area of abnormalAf.

CHAPTER llA FUNDUS AUTOFLUORESCENCEIN RETINITIS PIGMENTOSA 145

B

FIGURE llA.3. (A) Fundus AF in a patient with sector RP,showing an arc of high density and an inferiorareaof abnormally reduced AF that corresponds with a superior visual field defect. (B) Radial pattern of in-creasedAF corresponding to a typical "tapetal reflex" in an obligate carrier for X-linked RP.

Functional Significance of Annular AutofluorescenceAbnormalities in Retinitis PigmentosaSeveralelectrophysiological and psychophysical srudies have sought to derermine thefunctional significance of the ring of increased AF. In a heterogeneous group of RP pa-tients with normal visual acuity, ring size was not directly related to the severity of gen-eralizedrerinal dysfunction, as determined by full-field ERG (28), This, however,wasexpectedsince full-field ERG does nor provide information regarding macular function.

In a heterogeneous group of patients with RP and normal visual acuity, a highpositive correlarion between rhe radius of the ring and the pattern ERG P50 compo-nent was found (28,30,31), suggesting that only areaswithin the ring contribute to thegeneration of rhe partern ERG. This hypothesis was first substantiated by recordingpattern ERGs to checkerboards of different diameters, Pattern ERGs to small stimu-lus fields were comparable to those recorded in normal subjects, but there was a cut-off in the expecred enlargement when the checkerboard srimulus diameter exceededthat of the ring (28,29), Multifocal ERGs corroborared the pattern ERG findings,showing widespread response attenuation over parafoveal areas with sparing or rela-tive sparing over the central macula (31,35), The lateral extent of mfERG responsepreservation varied between patients and was related to both the ring size and the ap-proximate area of central visual field preservation ascertained using automatedHumphrey (35) or Goldman perimetty (36). Figure 11A.4 illustrates typical AF ab-normalities (Fig_11A.4A), small-field pattern ERGs (Fig. 11AAB), visual field con-striction (Fig. 11AAD), and corresponding mfERGs (Fig. 11AAC) in a representa-tive patient with RP and normal visual acuity.

Fine matrix mapping involves psychophysical measurement of photopic and sco-topic thresholds, typically at more than 100 retinal locations within a 9 X 9 degreemacular area. The technique has been used to measure cone and rod system sensitivityin numerous rerinal disorders (29,63-65), Photopic fine matrix mapping in individu-als with RP and good visual acuity revealed preserved cone sensitivity within the ringof increased AF (Fig. 11A,5), There was a steep gradient of threshold elevation acrossthe ring and severe sensitivity loss over more eccentric areas (Fig. 11A.5C) (29,35,66),


B 6

s:2- 30~Q.

2'"u:WQ.

00

0

5 10 15

Field diameter {dew

FIGURE 11A.4. (AI Fundus AFin a patient with RP and normalvisual acuity, showing parafovealrings of increased AF. (BI Thepattern ERG P5D fails to show theexpected enlargement seen innormal subjects when responsesare recorded to increasingly largecheckerboard fields. Filled circlesand error bars in (B) show meanvalues and standard deviationsfor eight normal subjects; trian-qles and squares show patientdata from right and left eyes.Multifocal ERG shows marked re-duction over peripheral areas lei,consistent with eccentric visualfield loss (0). Maximum diameterof the mfERG stimulus array is 58degrees. (From Ref 35, repro-duced with permission from theBMJ Publishing Grcup.)

c

~~,.....,.-"""\.~~~

~rI"""'~V\r.yv..t-/W-~~

~"""""""'.r¥".../V'~.../Lr~~\I"Y'

~""'vw~V"Y~~'-V"""''IVM''

~"""""""~..,rvJ~........-v./............... '-..,.~~~.....;""'"

consistent with the PERG and mfERG data. The ring corresponded with the internaledge of photopic visual field constriction. Perhaps not surprisingly, given that patientshad predominantly rod photoreceptor disease, scotopic sensitivity losses weremoresevere than photopic sensitivity losses and encroached upon central macular areasen.circled by the ting (Fig. IIA.5D). These observations suggest that rod macular dys-function precedes significant lipofuscin accumulation and increased AF (see below).The significance of a similar ring in patients with impaired visual acuity is nor known.

The above-mentioned studies show that retinal function may be severely im-paired outside the ring of increased AF, despite normal or near-normal levelsofAF(Figs. 11A.5 and IIA.6). There is evidence that lipofuscin, and specifically AlE,rapidly fragments when exposed to light, and this could be a triggering factor forthe complement system that would predispose the macula to disease (67). If rhetime course of lipofuscin degradation were similar in vivo, then its presence in theRPE would suggest conrinuing outer segment turnover and metabolic demand(42). Preservation of AF may occur in the presence of dysfunctional but intact pho-toreceptors, and it has been suggested that these cells may be amenable to func-tional rescue. Identification of viable retinal areas has become increasingly impor-tant given recent advances in gene therapy (68,69). It is also possible that clinicianswiJI use AF as an outcome measure when assessing the efficacy of future treatments.However, recent OCT evidence has indicated disruption of the presumed phorore-ceptor layer in regions of preserved AF external to the ring (44). An implicarion isthat maintenance of normal lipofuscin levels may not be dependent on normal pho-toreceptor structure. Further studies are required to examine this possibiJiry.

Serial Imaging of Fundus Autofluorescence Abnormalitiesin Retinitis PigmentosaIn some RP pedigtees,. ring size may be related to the age of the patient or the dura-tion of the disease (FIg. IIA.2A,C) (28). It was anticipated that the ring would

FIGUREl1A.5. Photopic (A,Cj and scotopic (B,D) fine matrix mapping in one normal subject IA,BI andin one individual with a clinical diagnosis of RPand normal visual acuity (C,D). Contour plots (rows 1 and3) show sensitivity gradients over tested retinal areas; corresponding 3D plots (rows 2 and 4) show retinallocation (abscissa, degreesl and thresholds (ordinate, log units I. Labeling [x) shows correspondence be-tween the orientation of contour and threshold plots Pattern ERG,mfERG, and visual field from the sameindividual are shown in Fig. l1A.4. In the patient with RP,photopic thresholds at the fovea are normal, inkeepingwith preserved cone system function, but show marked elevation over areas external to the ring ofincreased AF. The ring aproximates to the internal edge of visual field constriction. Under dark-adaptedconditions, threshold elevation is seen internal and external to the ring (DI, in keeping with rod system dys-function over both the central macula and more eccentric areas.

A

c

CHAPTER llA FUNDUS AUTOFLUORESCENCEIN RETINITIS PIGMENTOSA

B

II

D

x

constrict as the severity of the disease and macular involvement progressed; however,to date, only one longitudinal study has documented serial changes in RP. In 9 of 12patients monitored over periods of up to 5 years, ring size and pattern ERG PSG am-plitude were stable. In the remaining three individuals there was progressive constric-tion and narrowing of rhe ring of increased AF (35). Progressive ring constriction re-suited in proportionate artenuation of the pattern ERGs to large but not smallstimulus fields (Fig. llA.6) and, as in previous studies, fine matrix mapping showedthat the internal edge of visual field constriction was spatially concordant with thering of increased AF (35). Itwas concluded that the visual field loss in RP mirrors theconstriction of the AF ring, and follows progressive rod photoreceptor dysfunctionover concentric macular areas within the ring.

The ring of increased AF may be of prognostic value in patients with normal vi-sual acuity, but longitudinal imaging studies may be necessary to establish the stabil-ity or rate of ring constriction. Small rings of increased AF are sometimes associated


• normal5

____right eye

--+-Ieft eye4

'r __ --:~ 1003

_~.?,p~l~=====:::::1005

o.l---.---~--~--.,o 5 15 "'0

field diam.er (deg)

2005

FIGURE llA.6. Comparisonof fundus AF images and pattern ERGswith different sizesof checkerboardsand Humphreyvisual fields In a patient with Usher syndrome type 2 (mutation of USH2Ajandnormalvisualacuity.Testingwas repeated after 24 months. The ring of increased AFshows progressiveconstrictionoverthe 2-year period, consistent with pattern ERGreduction and progressive visual fieldconstriction.Filleddr·cles and error bars in the graph show mean values and standard deviations for eight normalsubjects;tri-angles and squares show patient data from right and left eyes. (FromRobson AG,SaihanZ.JenkinsSA,etal. Functionalcharacterisation and serial imaging of abnormal fundus autofluorescence inpatientswilli,.tinitis pigmentosa and normalvisual acuity. BrJ OphthalmoI2006;90:472-479. reproducedwithpermissionfromthe BMJ PublishingGroup.)

with mild foveal photopic sensitivity loss (29). Snellen visual acuity may be preserved,but such cases may have a poorer prognosis for retention of central vision.

Unlike cases ofRP, some individuals with cone-rod dystrophy manifest rings thatshow progressive enlargement with time (54), suggesting an expanding centrifugalfront of macular photoreceptor dysfuncrion. This was also suggested by a cross-sec-tional analysis within families (54): older individuals tended to have larger AF rings(Fig. lIA.2B,D) that surrounded areas of central RPE atrophy (Fig. llA.2D). It ispossible that rings associated with other causes of maculoparhy (other than RP) mayalso expand with time as lesions become larger with age.

Fundus Autofluorescence in X-LinkedRetinitis PigmentosaX-linked RP accounts for approximately 5% to 20% ofRP cases (4,70-72); ir generallyhas a severe phenotype with early onset and often leads to complete blindness by mefourth or fifth decade of life. Most cases are consequenr upon mutation in RPGR orRP2. Typical fundus features of RP are presenr, which are usually severe in adul"(73,74). Perhaps not surprisingly, given the severity of the phenotype, there is a paucio/

CHAPTER llA FUNDUS AUTOFLUORESCENCE IN RETINITIS PIGMENTOSA

f'o--+lt-~al'ra'a specifically describing AF in male parienrs with XLRP. One report described aring of increased AF at an eccentricity of 3.5-15 degrees in a young affected male (47).

The detection of carrier status is important for genetic counseling purposes.Heterozygous carriers ofXLRP can manifest a wide range of fundus and/or full-fieldERG abnormalities that may be asymmetrical, resulting from random inactivation ofthe X-chromosome during embtyogenesis. Carriers may have a normal fundus and anormal ERG when young, but often manifest a typical radially orientared "rapetalsheen" that has been considered parhognomonic of rhe disorder (75). The prevalenceof eirher abnormal fundus or abnormal ERG increases with age (76). In one study offemale carriers ofXLRP (RPGR), abnormal AF was found in 9 of 11 individuals (47).Mosr of these showed a distinctive parchy, radially distributed pattern of increased AFin the parafovea that was not seen in other forms of RP (Fig. IIA.3B). The radialpattern of increased AF could be asymmetrical and occurred over parafoveal areas as-sociated with the most pronounced sensitivity losses. Such changes can occur in theabsence of severe fundus abnormalities and may prove useful in identifying carriers,and patient management.

DIFFERENTIAL DIAGNOSISRP must be distinguished from other causes of inherited or acquired pigmentatyretinopathies and from other causes of night blindness that may occur with or with-out fundus changes. The differential diagnoses include the following:

1. Genetically determined causes of inherited retinal degenerations other rhan typicalRP, including choroideremia, gyrate atrophy, enhanced S-cone syndrome/Goldman-Favre, inherited virreoretinopathies such as familial exudative vitreo retinopathy(FEVR) and Wagner-Sticklet syndrome, cone-rod dystrophies, Stargardt disease, andSorsby fundus dysrrophy.

2. Abnormal retinal pigmentation, including X-linked choroideremia carriers, X-linked ocular albinism carriers, and congenital hypertrophy of the RPE. Unilateralpigmentary retinopathy usually results from an acquired disease caused by rerinalinflammation (toxoplasmosis and diffuse subacute unilateral neuroretinitis) ortrauma (contusion, intraocular foreign body, and secondary siderosis).

3. RP sine pigmento must be differentiated from other causes of night vision impair-ment with normal fundus, including various forms of congenital stationary nightblindness, vitamin A deficiency, carcinoma associated retinopathy, or autoimmunererinoparhy.

4. Inflammatory condirions (syphilis, rubella, or other causes of posterior uveitis orchorioretinitis).

5. Drug toxicity (rhioridazine, cIofazimine, and chloroquine).6. Resolved rhegmatogenous or exudative retinal detachments.

Family history, clinical context, and ancillary studies such as electrophysiologyand fundus AF will be critical for accurate diagnosis.

SUMMARYFundus AF imaging is a useful tool in the diagnosis and evaluation of patients with RPand allows a precise characterization of the disease in conjunction with electrophysiol-ogy and OCT. An abnormal AF pattern may be observed in some patients wirh no

SECTiON Ii CLINiCAL SCIENCE

.. J ki SA et al Identification of novel RPGR53. Ebenezer ND, Michaelides M, en lOS , CORD) families. Invest Ophchalmol

linked progressive cone-rod dystrophy (XL1891-1898. fe d fl. id M L VA tal Functional correlates a run us auto uorescenceabnormal

54 Robson AG, Micbaeli es , liang , e.. d d h B J .. ... . irh RPGR RIMSI mutations causing cone or ccne-ro ysrrop y. r Ophthalmolrues III patients wt or2008;n95-102. h . d f" d h

55. Michaelides M, Holder GE, Webster AR, et a]. A detailed p enocypic stu yo cone ymop ywiths~.

pernormal rod ERG." Br J Ophthalmol 2005;89: 332-339. .. .6 R b AG M· h Iid M WHet a] Fundus autofluorescence and mrensiry-response funCtlonsin5.0500 , lCaeles , u, . . .

" d hv wi h al od elecrroretinogram" consequent upon muranon In KCNV2 Investcone ysrrop y Wit supernorm r .Ophrhalmol Vis Sci 2008;49:ARVQ Eiabsrracr 2167. ..., . .

57. jarc-Vidmar M, Kraut A, Hawlina M. Fundus autofluorescence Imagmg m Best S virelliforrn dysuopbJ·.

Klin Monatsbl Augenheilkd 2003;220:861-867.. .' . . .58. Tsang SH, Vaclavik V, Bird AC, et al. Novel phenotypic and genoryplC findings In X-hnked rennoscbisis.

Arch Ophth,]moI2007;125,259-267. .59. Fleckenstein M, Charbel Issa P, Fuchs HA, er al. Discrete arcs of increased fundus autofluorescence inreti·

nal dystrophies and functional correlate on microperimetry. Eye 2008; March 14 [Epub ahead of prim].60. Poloschek Cfvl, Hansen LL, Bach M. Annular fundus aurofluorescence abnormality in a case of macular

dystrophy. Doc OphthalmoI2008;1l6:91-9S. '"61. Wiistemeyer H, Jahn C, Nestler A, er aL A new instrument for the quantification of macular pigmentdell'

siry: first results in patients with AMD and healthy subjects. Graefes Arch Clin Exp Ophthalmol

2002;240,666-671.62. Meyerle Cb, Fisher YL, Spaide RF. Autofluorescence and visual field loss in sector retinitis pigmemosa.

Retina 2006;26:248-250.63. Chen JC, Firzke FW, Pauleikhoff D, er aL Functional loss in age-related Bruch's membrane changewirh

choroidal perfusion defect. Invest Ophrhalmol Vis Sci 1992;33:334-340.64. Westcott MC, Carway-Heach DF, Fitzke FW, et al. Use of high spatial resolution perimetry to idemifYsco.

romara not apparent with conventional perimetry in the nasal field of glaucomatous subjects. BrJOphth,]moI2002,86'761-766.

65. Scholl HP, Bellmann C, Dendekar 55, er 3.1. Photopic and scotopic fine matrix mapping of retinalareas ofincreased fundus autofluorescence in patients with age-related maculopathy. Invest Ophthalmol V~Sci2004;45,574--583.

66. Robson AG, Moreland JD, Pauleikhoff D, et 3.1. Macular pigment density and distribution: compariscnolfundus autofluorescence with minimum motion photometry. Vision Res 2003;43: 1765-1775.

67. Zhou J, Jang YP, Kim SR, et al. Complement activation by phocooxidarion products of AlE, a lipofuscinconstituent of the retinal pigment epithelium. Proc Narl Acad Sci USA 2006; 103: 16182-16187.

68. Bainbridge JW, Smith A], Barker 55, et al. Effect of gene rnerapy on visua.l function in Leber's congenit:llamaurosis. N Engl] Med 2008;358:2231-2239.

69. Maguire AM, Simonel1i F, Pierce EA, et al. Safety and efficacy of gene transfer for Leber's congenildlamaurosis. N Eng! J Med 2008;358:2240-2248.

70. Jay M. On the herediry of retinitis pigmemosa. Br J Ophchalmol 1982;66: 405-416.71. Bird AC, Heckenlive1y JR. X-linked recessive retinitis pigmemosa (X-linked pigmentary retinopathies).In:

Heckenlively ]R, ed. Retinitis Pigmenrosa. Philadelphia: JB Lippincott, 1988: 162-176.72. Prokisch H, Hartig M, Hellinger R, et 3.1. A population-based epidemiological and genetic studyofX·

linked retinitis pigmentosa. Invest Ophtha.lmol Vis Sci 2007;48:4012-4018.73. Flaxel C), Jay M, Thiselton DL, et 3.1. Difference between RP2 and RP3 phenotypes in X linked retjnitispig-

mentosa. Br J OphthalmoI1999;83:1144-1148.74. Lorenz B, Andrassi M, Kretschmann U. Phenotype in two families with RP3 associated with RPGRmu·

tations. Ophthalmic Genet 2003;24:89-101.75. ~eckenii:ely ].R, W~leber RG. X-linked recessive cone dystrophy with tapetal-like sheen. A newlyrecog·

filze~ entity w~th Mlzuo-Nakamura phenomenon. Arch Ophthalmol 1986;104: 1322-1328.76. Je~kms SA, Bnd AC, Moore AT, et a1. What is the reduced risk of carrier status in females of X-linked

retinal dystrophy families when the fundus and ERG are normal? ARVO 2006;E-abstracr 1021.

HAPTER

Fundus Autofluorescence in Coneand Cone-Rod Dystrophies

INTRODUCTIONThe cone dystrophies (COD) and cone-rod dystrophies (CORD) are a heterogeneousgroup of disorders in terms of both clinical features and underlying genetic basis.They are characterized by reduced central vision, color vision abnormalities, visualfield loss, and a variable degree of nystagmus and photophobia. There is absent or se-verelyimpaired cone function on electroretinography (ERG) and psychophysical test-ing. Patients with CORD develop additional rod system abnormalities that lead tonight-blindness later in the disease process.

Disorders of cone function can be usefully divided into stationary (cone dysfunc-tion syndromes) and progressive (COD and CORD) disorders (1,2). The stationarycone dysfunction syndromes often present shortly after birth or in infancy.Progressive cone dystrophies usually present in childhood or early adulthood, withmany patients developing rod photoreceptor involvement later in life, thereby leadingto considerable overlap between progressive COD and CORD. In this chapter rheterm CORD is used to describe those disorders in which subjects have significant sec-ondary involvement of the rod system at an early stage, in contrast to progressiveCOD, in which rod involvement, if present, occurs late in the disease process.

COD and CORD can be inherited as autosomal dominant (AD), autosomal reces-sive(AR), or X-linked (XL) recessivetraits. When an inheritance pattern can be reliablyestablished, it is mosr commonly AD (3,4). Mutations in 14 genes have been describedto date, with mutations inperipherin/RDS, ABC44, and RPGR being the most commoncausesof AD, AR, and XL COD and CORD, respectively (2,5).

This chapter will concentrare only on progressive disorders, including COD andCORD, since data on fundus autofluorescence CAP) imaging in the various stationarycone dysfunction syndromes are, to date, very limired (6).

MOLECULAR BASISThe molecular basis of COD and CORD is complex because of their generic hetero-geneiry. The molecular basis of some of the most common forms of COD andCORD is discussed below (see Fundus Autofluorescence subsection).

CLINICAL FEATURESPatients wirh COD are not usually symptomatic until late childhood or early adult-hood. The age of onset of visual loss and the rate and degree of progression isvariable;however, visual acuity usually deteriorates over time to the level of 20/200-counting


FIGURE 118.1. Color fundus photograph (leftl and AF image (right) in a patient with conedystropoyand bull's-eve maculopathy.

fingers. Photophobia is often a prominent early symptom. Individuals with CORDwill, in addition, complain of night-blindness. The rennal dystrophy IS usuallyiso-lated bur may be associated with systemic abnormalities (2).

In COD, fundus examination may show a typical hull's-eye appearance(Fig.IIB.l). However, in some cases there may only be minor macular retinal pigmemep-ithelium (RPE) disturbance. The optic discs show a variable degree of temporalpallor.The retinal periphery is usually normal. In CORD, fundus examination mayshow,bull's-eye appearance in the early stages, with macular atrophy developingovertime.Peripheral RPE atrophy, retinal pigmentation, arteriolar attenuation, and optic discpallor can be seen in the late stages of the disease, similarly to the rod-cone dysrrophie(seeChapter lIA). The sign of the "dark choroid" may be seen on fluoresceinangiog-raphy (7). A tapetal-like sheen, which may change in appearance on dark adaptation(the Mizuo-Nakamura phenomenon), has been described in association with XL-CORD (8).

DIAGNOSTIC TESTS

ElectrophysiologyIn COD or in the early stages of CORD, ERG shows normal rod responsesbutsig·nificantly abnormal cone responses. The 30Hz flicker ERG is usually of increasedim-plicit time, but rarely the implicit time is normal and amplitude reduction istheonlyabnormaliry (Table IIB.l) (9,10). In CORD, both rod and cone thresholdsare ele-vated on psychophysical testing and the ERG shows reduced rod and coneampli-tudes, with the cone ERGs being more abnormal than the rod ERGs. A negativeERG canbe also observed in some patients (Table 11B.l). An unusual formofCORD WIth abnormal cone function and supernormal rod responses has beenalsodescnbed (11,12). Obligate carriers of XL COD and CORD may show evidenceofcone dysfunction on elecrrophysiological or psychophysical testing (13-15).

Imaging StudiesFluorescein angiography, indocyanine green angiography, and optical coherenceto.mography do not, to date, have a significant role in the evaluation of patient5WIth COD and CORD.

CHAPTER 11B FUNDUS AUTOFLUORESCENCE IN CONE AND CONE-ROD DYSTROPHIES 155

TABLEllBl Genetics. Age of Onset. Fundus Features. Autofluorescence, and ElectrophysiologicalAbnormalities in Cone and Cone-Rod Dystrophies

Gene Deleet Age 01linheritanee) Onset Phenotype Fundus Changes AF PERG ERG

GUCA7A 3rd-5th COD, ranges lrom mild focal increased AF severely reduced severely reduced(AD! decade CORD macular RPE at the macula; or undetectable amplitude of single

disturbance to perifoveal rings of flash and flicker withRPEatrophy, with increased AF; minimal or no 30Hznormal peripheral reduced AF m flicker implicit timeretina areas of atrophy shift; in CORD,

additional reductionin rod responses

GUCY2D 1st. 2nd CORD macular and increased foveal severely reduced reduced cone and[AD) decade peripheral atrophy in areas of atrophy or undetectable rod responses;*

AF; reduced AF in "negative ERG"areas of atrophy

Peripherin 2nd, 3rd CORD macular RPE "speckled" AF reduced responses reduced cone and(AD) decade mottling, macular rod responses

atrophy, peripheralretinal atrophy andareas of peripheralRPEhyperpigmentation

CRX 1st decade CORD macular and later reduced AF in reduced responses reduced cone and(AD! peripheral retinal areas of atrophy, rod responses:

degeneration "speckled" AF in "negative ERG"the midperipheralretina

RIMS 7 2nd-5th CORD ranges from mild reduced AF at the absent or severely reduced cone and[AD) decade macular RPE center of the reduced rod responses. often

disturbance to macula surrounded normal 30Hz flickeratrophy and by a ring of implicit timepigmentation' increased AF

ABCA4 1st-3rd COD, macular atrophy, reduced AF at the absent or severely severely reduced cone(ARj decade CORD bull's eye center of the reduced responses; in CORD,

appearance, macula surrounded additional reduction inperipheral by a ring of rod responsespigmentary increased AF;changes in reduced AF onlyadvanced disease

KCNV2 1st. 2nd COO RPEdisturbance most commonly absent or severely reduced rod b-wave

IAR) decade at the macula reduced AF at the reduced amplitude with lowcenter of the intensity stimulus: oftenmacula surrounded higher than normal withby a ring of high flash energies, coneincreased AF responses severely

reduced

RPGR 5th decade COO macular atrophy perifoveal ring of absent or severely reduced cone responses

(X-linked! increased AF reducedRPGR 2nd-4th CORD range from mild reduced macular AF, absent or severely reduced cone and rod

(X-linked) decade macular RPE reduced AF reduced responsesdisturbance to surrounded by aextensive atrophy ring of increasedand AFhyperpigmentation

AF. autofluorescence; ERG, full-field electroretinogram; PERG, pattern electroretinogram; RPE, retinal pigment epithelium; ". ERG is usually nonrecordable by the fourthdecade of life; ", attenuation of retinal blood vessels and peripheral retinal atrophy can occasionally be observedSee text for detailed description of AF findings. In CORD. ERG reveals greater reductions in cone than in rod responses, AD, autosomal dominant AR. autosomal recessive


Fundus Autofluorescence .,.Fundus AF imaging has been used to assist in the diagnosis, to a~ddetaileddescri~tion of the phenotype, and to provide insights into the natural history andunderly.ing pathophysiology of COD and CORD. . '.

COD and CORD are classified below based on their mode of mhenranceandtheir causative genes (only those for which well-documented AF data ateavailablewillbe discussed in this section). Fundus AF features will be descnbed herein.Thedini.cal, electrophysiological, and AF findings are summarized in Table I 1B.1.

Fundus Autofluorescence Findings in Autosomal Dominant DiseaseTo date, seven genes have been associated with AD disease. AF data are availablefo!COD and CORD caused by mutations in GUCAIA, GUCY2D, Peripherin/RDs,CRX, and RIMS], as described below.

COD and CORD Associated with GUCAIAPhotophobia, reduced central vision, and generalized dyschromatopsia, withnoei-dence of nystagmus, are usually observed. In some subjects, RPE changesmaybesub.tle, especially in the early stages of the disease. In these patients, AF imagingishelpfulin confirming the macular abnormality by identifying localized areals) of incrOlSedmacular AF (Fig. 11B.2) (9,16). In some individuals, perifoveal rings ofincreaseriAFhave been described that are similar to those observed in retinitis pigmenrosa (seeChapter I I, Retinitis Pigmentosa subsection) (9); these rings are increasinglyrecog.nized as features of COD and CORD (17,18). In older subjects with rnaculararro.phy, corresponding areas of decreased AF are seen (Fig. II B.3).

The gene GUCAIA encodes rhe phorotransducrion protein guanylarecyclaseac-tivating protein- I (GCAPI). Mutant GCAPI protein activates retinal guanylarecy.clase-I (RetGCI) at low Ca2+ concentrations but fails to inactivate it at highCa1+concentrations, thereby leading to a constant activation of RetGel in photorecep-tors, even at the high Ca2+ concentrations of the dark-adapted state (Fig. 11B.4).The consequent dysregulation of intracellular Caz+ and cGMP levels is believedtolead to cell death.

CORD Associated with GUCY2D

Moderate .myopia is common, with photophobia and pendular nystagmus seeninsffected individuals, who experience the major visual reduction in the second or thirddecade oflife (19-21). Initially there is only absent tritan color discrimination, burthisprogresses to complete loss of color vision over time.

FIGURE 11B.2. AD·COD and CORD (GUCA 'AI. Fundus AF image showing localized increased macularAF

CHAPTER 118 FUNDUS AUTOFLUORESCENCEIN CONEAND CONE·ROD DYSTROPHIES '51

FIGURE118.3. AD-COD and CORD(GUCA7A). Fundus AF image showing decreased macular AF.

The earliest AF abnormality is increased AF at the fovea, suggesting that this isthe site of initial dysfunction (21). A markedly reduced AF signal is detected ar thesite of macular atrophy, indicating loss of photoreceptor cells, or at least their outersegments. In the later stages, subjects have an annulus of increased AF surroundingareas of central arrophy (see below). Increased AF ar the edge of atrophy is likely toindicate an area destined to become atrophic (17,18).

Families with AD-CORD associated with single GUCY2D missense mutationshave a much milder phenotype (only mild rod involvement) than subjects with com-plex mutations (moderate to severe rod loss) (19-23).

The gene GUCY2D encodes retinal guanylate cyclase 1 (RetCG 1) (Fig. 1IB.4).Mutant RetGCl has been shown to have a higher apparent affinity for GCAPI thanwild-type RetGCl, and an altered Ca2+ sensitivity of the GCAPI activation, withmarked residual activity at high Ca2+ concentrations (24). Therefore, as seen inGeAPl mutations, RetGCI mutations result in a failure to inactivate cyclase activityat high physiological Ca2+ concenrrations in the phororecepcors, with a subsequent

8m 1ft'

:ua.nyJaft cydan guanylate cydue ..adiv-atinl prottln )Ie; C,.u KGCAP E: ce

8.;pllOton rllodopsUI en'ilR tramdlldn (01 @) cGMP-pted~.tir cadon chumd

~ metall POE \

'bodOPrinldnY ~~E> IJlSi * aoted In ligJll

•POE* GM""+

FIGURE 118.4. Phototransduction cascade.


b I h d rion recovery phase. This dysregulation ofinrraceliularCa1-a norma p ororrans lieand cGMP levels may lead to cell death. . .

It is plausible that RetGCI, in addition to havlllg a role In photOlransduCiion,may have a function at the photoreceptot synaptic terminal, as suggested b~the"neg.arive ERG" often observed in patients carrymg mutauons In GUCY2D (2)).

CORD Associated with Peripherin/RDSTo date, four mutations in peripherin/RDS have been reported in Japanese familiesanda large British family (26-29). Affected individuals in the Japanese families had~'Pic'symptoms, elecrrophysiological findings associated With CORD, andllttle Imrafamili'variability. Ophthalmoscopy revealed macular RPE atrophy With peripheral retinalde.generation in the later stages. Although mutations in the Argl72Trp (RI72W) periph.en'nlRDSwere previously reported to cause a fully penetrant progressive maculardystro-phy with high intra- and inrcrfamilial consistency of phenotype (30,31), rhis mutationhas now also been described in a large Brirish family wirh marked intrafamilial pheno.typic variation, including a CORD phenotype (29). The majority of affecred individu.als had reduced central vision starting in the second or third decade, and several inm.viduals became aware of nyctalopia and slow dark adaptation at a later Stage.

AF imaging in rhe majority of patients reveals a highly characrerisric sperkledmacular appearance wirh areas of increased and decreased AF (Fig. 11B.5).

The peripherin/RDS prorein is found in rod and cone photoreceptor outerseg.menr discs in a complex with ROMI. Ir is believed to function as an adhesion mole.cule involved in the stabilization and maintenance of a compact arrangememofourersegment discs (32,33). To dare, four mutations in peripherin/RDS have been associ.ated with a CORD phenotype: Asn244His (26,27), Tyrl84Ser (27), Val200Glu (28),and Argl72Trp (29). The amino acids Argl72, Tyrl84, Val200, and Asn244 arealllocated in the second intradiscal loop (EC-2) of the peripherin/RDS protein, whichcontains seven conserved cysteine residues, six of which are imporrant for proteinfold-ing (33). The seventh cysteine (CI50) in the EC-2 domain forms an intermoleculardisulfide bond with the CI50 residue in another molecule of either peripherin/RDS orROMI to form higher-order oligomeric complexes that are necessary for outer seg.ment disc generation and stabilization (33). This loop, within which the CORD.asso-elated mutations are located, is therefore critical for the functioning of the protein.

FIGURE 11 B.5. AD-CORD(peripherin/RDS!. AF imaging showinq florid typical abnormal speckleda~pearance. with areas of Increased and decreased macular AF.

CHAPTER 11B FUNDUS AUTOFLUORESCENCE IN CONE AND CONE-ROD DYSTROPHIES

CORD Associated with CRXA severe early-onset AD-CORD phenotype associated with CRX mutations (cone-rodhomeobox-containing gene) has been reported in patients of diverse origin, with lit-tle or no visual function remaining after the age of 50 years (34-37).

AF imaging undertaken in a German CORD family with CRX mutations revealedrharAF was severely reduced in areas of atrophy bur otherwise well preserved at the pos-terior pole (37). A speckled appearance with areas of increased and decreased AF wasnoted in the mid-periphery in several bur not all subjects. This feature may prove help-ful in suggesting CRX as the underlying genetic cause (Fig. IIB.G) (37).

CRXis a photoreceptor-specific transcription factor and plays a crucial role in thedifferentiation and maintenance of photoreceptor cells (38,39). The electronegativeERG changes seen in patients with CRX mutations suggest that inner retinal functionis primarily impaired. Since retinal expression of CRX is limited to photoreceptors(38,39), this dysfunction may be the result of abnormal photoreceptor communica-tion with second-order retinal neurones. This is supported by the finding that pho-roreceptors in CRX knockout mice have severely abnormal synaptic endings in theouter plexiform layer (40).

In addition to directly regulating the expression of several photoreceptor-specificgenes (including the opsins and arrestin), CRX also interacts with the transcriptionfactors NRL (neural retina leucine zipper) and NR2E3 to affect expression of genescritical to photoreceptor morphogenesis and function (41,42).

CORD Associated with RIMS]A CORD phenotype associated with a point mutation in RIMS] (formerly knownas RIMl), a gene encoding a photoreceptor synaptic protein, was reported in afour-generation British family (43,44). Most of these individuals had experienced

FIGURE llB.6. AD-CORD (CRXI. AF image showing reduced AF in areas of atrophy, but otherwisegoodpreservationat the posterior pole.A speckled appearance with areas of increasedand decreasedAF is seen inthe mid-periphery.


FIGURE lIB.7. AD-CORD (RIMS7l. Fundus AF image showing decreased AF centrally with a surround·

ing ring of increased AF.

progressive deterioration in central vision, night vision, and peripheral visualfieldconstriction, especially during the third and fourth decades. They had mild pharo·phobia and no nystagmus. Mild to moderate genetalized dyschromatopsia was de-tected in the majority of the individuals.

AF imaging revealed decreased macular AF centrally surrounded by a ringofin-creasedAF in the majority of patients (Figs. II B.7 and IIB.8) (17,18,43). Aperifovealring of increased AF was also detected in the youngest individual (18 yearsold),whohad very mild RPE disturbance at the macula and was asymptomatic (Fig. IIB.8). Thepresence of a petifoveal ring of increased AF may be very helpful in establishingan earlydiagnosis in such cases.The size of the AF ring correlated with the ageof the patientandenlarged over rime (17). Photopic and scoropic fine matrix mapping and multifocalERG (mfERG) demonstrared that the rings of increased AF were associatedwithagra.dienr of scotopic and photopic sensitivity loss when comparing to retinal locationsin-ternal and external to the ring of increased AF (see also Chapter IIA). PatternERGPSOamplitude, when detectable, was inversely related to the size of the AF ring (see~soChapter IIA). Increased AF was therefore associated with reduced rod and conesersirivity, rather than photorecepror cell death. Bull's-eye lesions, present in two individu-als, consisted of a ring of decreased perifoveal AF bordered peripherally and centrallybyincreased AF (Fig. IIB.9). The normal ERG recorded in these patients sugges~thaithere is no widespread dysfunction of the RPE.

RIMS] is expressed in brain and retinal photoreceptors, where it is localizedtothe presynaptic ribbons in tibbon synapses. The protein product is believedto playm

FIGURE lIB.8. AD-CORD (RIMSlI AF' howi . .____________ .__ lm_ag:..e_s_owm.:g..:a:..:p:.:e~rt:.:fo:.:v:ea:l~r~m~9~o~f~re:la:ti~ve~i:nc:r:ea:se:d:.Af:..-

FUNDUS AUTOFLUORESCENCEIN CONEAND CONE-ROODYSTROPHIESCHAPTER 11B 161

FIGURE 118.9. AD-CORD (RIMS!). Fundus AF image showing concentric rings of increased and de-creasedAF in a bull's-eve maculopathy-like pattern.

important role in synaptic transmission and plasticity (45). Mutations in RIMS] mayalter the rate of synaptic vesicle docking and fusion in response to a Ca2+ signal (44).

Fundus Autofluorescence in Autosomal Recessive DiseaseTo date, six genes have been associated with AR disease. Data are available for CODand CORD caused by mutations in ABCA4 and KCNV2, as described below.

COD and CORD Associated with ABCA4Mutations inABCA4 are the most common cause of AR-CORD (46-48). All patientsexperience visual loss early in life, impaired color vision, and a central scotoma (47).

Subjects with COD may have a ring of increased AF surrounding decreasedfoveal AF (Fig. IIB.IO). Although some patients with CORD have the same patternof AF, the majority have decreased foveal AF only (Fig. IIB.II) (49,50). In a recentstudy involving patients with disease-causing variants in ABCA4 and either Stargardtdisease (STGD) or CORD, a peripapillary ring of normal-appearing AF was visibleat all stages of disease (51). This finding is highly suggestive of ABCA4-relatedrerinopathy.

ABCA4 encodes a transmembrane rim protein (an outwardly directed f1ippase ofall-trans retinal) that is located in the discs of rod and cone outer segments and is in-volved in ATP-dependent transport of retinoids from photoreceptor to RPE during

FIGURE 118.10. AR-COD (ABCA4I. AF image showing a ring of increased AF surrounding decreasedfoveal AF

SECTIONII CLINICAL SCIENCE

FIGUREllB.ll. AR-CORD(A8CA4) AF image showing decreased foveal AF.

the visual cycle. Failure of this transport results in deposition of AlE (N-tetinylidene_N-retinylethanolamine) in the RPE, with RPE cell dysfunction and subsequent lossofphototeceptots (see also Chapter II, Statgardt-Fundus flavimaculatus subsection).

COD Associated with KCNV2 (COD/SuperROD)An unusual AR-CORD has been described with abnormal photopic responsesasscd-ated with supernormal and delayed rod ERG b-waves (I1,12,52). Subjects presenwith reduced central vision and marked photophobia, are usually myopic, andhavereduced color discrimination predominantly along the red-green axes. Patients withmore advanced disease complain of nyctalopia.

In most patients, AF imaging reveals a perifoveal ring of increased AF. Insomepatients (generally older), however, an area of increased AF may be seen at the centralmacula (Fig. IIB.12), suggesting that accumulation of AF material at the centerolthe macula may occur over time. In the oldest subject studied, central atrophy waspresent surrounded by an annulus of relarively increased AF (Fig. IIB.l3), indicatingthat the accumulated autofluorescent material may lead to cell death (18,52). Thee1ectrophysiological data (see Table IIB.I) ate consistent with the site ofdysfuncrionbeing post-photorransducrion but pte-inner nuclear layer, most probably at the firstsynapse (52,53).

KCNV2 encodes a voltage-gated K+ channel subunit expressed in rods and con'(54). The effects of these mutations suggest that KCNV2 is involved in settingthetesting potential and voltage response of phororeceptors.

FIGUREllB.12. Cone dystrophy with supernormal rod ERG (KCNVZ). Fundus AF imageshowi~markedly Increased macular AF surrounded by a ring of relati dive ecreased AF.

CHAPTER 11B FUNOUS AUTOFLUORESCENCE IN CONE ANO CONE-ROO OYSTROPHIES 163:

FIGURE 118.13. Cone dystrophy with supernormal rod ERG{KeNV21. AF image reveals decreased mac-ular AF surrounded by a ring of relatively increased AF.

Fundus Autofluorescence in X-Linked DiseaseMutations in RPGR are a common cause of XL-CORD (14,55) and have also beenidentified in patients witb XL-COD (14,56).

COD and CORD Associated with RPGROnly two families with XL-COD have been reported, wirh patients having typicalCOD findings (Table llB.l) (56). However, they were unusual in having a late onsetof reduced vision in the fifth decade. In XL-CORD, affected males experience pro-gressive deterioration in central vision and subsequently night vision, mild photopho-bia, and moderate to high myopia (14).AF imaging in COD patients may reveal a perifoveal ring of increased AF (14).

Detailed AF imaging has been assessed in two unrelated XL-CORD families (14). Aring of increased AF around the fovea is often observed; older affecred subjects mayalso have decreased AF corresponding to areas of atrophy seen ophthalmoscopically(Fig. IIB.14). The ring of increased AF can be the sole abnormality detected inotherwise asymptomatic patients (Fig. IIB.15) (14), underlining the importance ofAF imaging in establishing an early diagnosis. Unlike female carriers of rod-cone

FIGURE 118.14. XL-CORD (RPGRi. Fundus AF image showing markedly reduced macular AF correspon-ding to atrophy seen on ophthalmoscopy. with a surrounding ring of increased AF.


AFIGURE 118.15. Xl-CORD (RPGR). Color fundus photograph (A) shows no abnormalities in an asymptomaticpatientin

whom AF imaging (8) revealed a ring of increased AF.

dystrophy associated with RPGR mutations (57), CORD carrier femaleshavenor-mal AF reported to date (14).

Mutations in the retinitis pigmentosa GTPase regulator (RPGR) gene arealsoImajor cause of XL RP. RPGR has been shown to interact with RPGRlPL Bothpro-teins colocalizc to the outer segments of rod and cone phororeceprors, principally In

connecting cilia of rods and cones, and th us are thought (0 have an important role inintracellular transport.

SUMMARYCOD and CORD represent an important cause of blindness in children andyoungadults for which there is currendy no treatment available. It is important, however, toestablish a correct diagnosis to provide patients with an accurate prognosis and in-formed genetic counseling.

Fundus AF imaging plays an important role in the evaluarion of patientswirhCOD and CORD. As shown in this chapter, in many instances, AF imagingmaj aicin establishing an early diagnosis before the onset of symptoms and before anyoph-rhalmoscopic abnormalities are detected. Furthermore, rhe pattern of AF maybesocharacteristic and recognizable (e.g., the speckled AF appearance associatedwithmu·rations in peripherin/RDS) that it may help the clinician to not identify the disorderon clinical grounds alone, but also to target genetic testing. The distribution ofAFhas been helpful also in shedding light on the natural history and pathogenesisofthese disorders. For example, in CORD associated with GUCY2D, the earliestabnor-mality of increased AF at the fovea implies that this is the site of initial dysfunction,in contrast to bull's-eye dystrophies, in which there is central sparing in the earlystages of the disease. Moreover, the high correlation found between AF and PERG,and mfERG and photopic and scotopic fine matrix mapping in patients withperi-foveal rings of increased AF demonstrates that AF abnormalities have function~sig·nificance. Therefore, AF imaging may help in identifYing suitable patients andviableareas of retina amenable to future therapies) and may also have a role in assessing theresponse to these treatments.

CHAPTER lIB FUNDUS AUTOFLUORESCENCE IN CONE AND CONE-ROO DYSTROPHIES

REFERENCESL Mjchaelides M, Hum DM, Moore AT. The cone dysfunction syndromes. BtJ Ophchalmol 2004;88;291-297.

2. Michaelides M, Hardcastle AJ, Hunt DM, et al. Progressive cone and cone-rod dystrophies: phenotypesand underlying molecular genetic basis. Surv Ophrhalmol 2006;51 :232-258.

3. Krill AE, Deutman AF, Fishman M. The cone degenerations. Doc OphthalmoI1973;35:1-80.4. Moore AT. Cone and cone-rod dystrophies. J Med Genet 1992;29:289-290.5. Maugeri A, Klevering BJ, Rohrschneider K, er al. Muracions in the ABCA4 (ABCR) gene are the major

cause of aurosomal recessive cone-rod dystrophy. Am J Hum Genet 2000;67:960-966.6. Michaelides M, Aligianis IA, Holder GE, er al. Cone dystrophy phenotype associated with a frameshift

mutation (M280fsX29l) in the a-subunit of cone-specific rransducin (GNAT2). Br J Ophthalmol2003;87; 1317-1320.

7. Uliss A, Moore AT, Bird AC. The dark choroid in posterior retinal dystrophies. Ophthalmology 1987;94:1423-1428.

8. Jacobson DM, Thompson S, Bartley JA. Xelinked progressive cone dystrophy. Clinical characteristics ofaffected males and female carriers. Ophthalmology 1989;96:885-895.

9. Downes SM, Holder GE, Firzke F\X7, er a1.Autosomal dominant cone and cone-rod dystrophy with mu-tations in the guanylate cyclase activator lA gene-encoding guanylate cyclase activating procein-L ArchOphthalmoI2001;119; 96-105.

10. Payne AM, Downes SM, Bessanr DA, er al. A mutation in guanylate cyclase activator IA (GUG41A) inaurosomal dominant cone dystrophy mapping to a new locus on chromosome Gp21.1. Hum Mol Genet1998;7;273-277.

11. Couras P, Eggers HM, MacKay C). Cone dystrophy, nyctalopia, and supernormal rod responses. A newretinal degeneration. Arch Ophrhalmol. 1983; 101 :718-724.

12. Alexander KR, Fishman GA. Supernormal scotopic ERG in cone dystrophy. Br J Ophrhalmol1984;68;69-78.

13. Reichel E, Bruce AM, Sandberg MA, er a]. An electroretinographic and molecular genetic study of X-linked cone degeneration. Am J Ophthalmol 1989; 108:540-547.

14. Ebenezer ND, Michaelides M, Jenkins SA, er al. Identification of novel RPGR ORF15 mutations in X-linked progressive cone-rod dystrophy (XLeORD) families. Invest Ophchalmol Vis Sci 2005;46:1891-1898.

15. Jacobson DM, Thompson S, Bartley JA. Xdinked progressive cone dystrophy. Clinical characteristics ofaffected males and female carriers. Ophthalmology 1989;96:885-895.

16. Michaelides M, Wilkie SE, Jenkins S, et al. Mutation in the gene GUG41A, encoding guanylate cyclaseactivating prorein-I (GCAPl) causes cone, cone-rod and macular dystrophy. Ophthalmology 2005;112;1442-1447.

17. Robson AG, Michaelides M, Luong VA, et al. Functional correlares of fundus autofluorescence abnormal-ities in patients with RPGR or RIMS J mutations causing cone or cone rod dystrophy. Br J Ophrhalmol2008;92;95-102.

18. Robson AG, Michaelides M, Saihan Z, er al. Functional characteristics of patients with retinal dystrophymat manifest abnormal parafoveal annuli of high density fundus autofluorescence; a review and update.Doc Ophrhalmol 2008; 116:79-89.

19. Gregory-Evans K, Kelsell RE, Gregory-Evans CY, et a1. Autosomal dominant cone-rod retinal dystro-phy (CORD6) from heterozygous mutation of GUCY2D, which encodes retinal guanylate cyclase.Ophthalmology 2000; 107:55-61.

20. Ito 5, Nakamura M, Nuno Y, et al. Novel complex GUCY2D mutation in Japanese family with cone-roddystrophy. Invest Ophcbalmol Vis Sci 2004;45:1480-1485.

21. Downes 5M, Payne AM, Kelsell RE, ec a]. Autosomal dominant cone-rod dystrophy with mutations in theguanylate cyclase 2D gene encoding retinal guanylate cyclase-I. Arch OphthalmoI2001;1l9:1667-1673.

22. Payne M1, Morris AG, Downes 5M, et al. Clustering and frequency of mutations in the retinal guanylatecyclase (GUCY2D) gene in patients wirh dominant cone-rod dystrophies. J Med Genet 2001;38:611-614.

23. Perrault I, Rozet JM, Gerber S, et aJ. A retGC-l mutation in autosomal dominant cone-rod dystrophy.Am) Hum Genet 1998;6%51-654.

24. Wilkie SE, Newbold RJ, Deery E, et al. Functional characterization of missense mutations at codon 838in retinal guanylate cyclase correlates with disease severity in patients with autosomal dominant cone-roddystrophy. Hum Mol Genet 2000;9:3065-3073.

25. Liu X, Seno K, Nisbizawa Y, et al. UltrasrfUcturallocalization of retinal guanylate cyclase in human andmonkey retinas. Exp Eye Res 1994;59:761-768.

26. Nakazawa M, Kikawa E, Chida Y, ec al. Asn244His mutation of the peripherin/RDS gene causing autoso-mal dominant cone-rod degeneration. Hum Mol Genet 1994;3: 1195-1196 .:

27. Nakazawa M, Kikawa E, Chida Y, et al. Autosomal dominant cone-rod dystrophy associated with mu-tations in codon 244 (Asn244His) and codon 184 (Tyr1845er) of the peripherin/RDS gene. ArchOphthalmoI1996;114: 72-78.

28. Nakazawa M, Naoi N, Wada Y, et al. Autosomal dominant cone-rod dystrophy associated with aVal200Glu mutation of me peripherin/RDS gene. Retina 1996;16:405-410.


.' G B d h K 1 Cone-rod dystrophy, intra~fa.milial variability and intorn_29 Michaelides M Holder E, fa s aw , et a . . h . (RDS 0. '. d . h h Rl72W mutation in the perlp am gene. phmalmology

plere penerrance associate WIt t e2005;112,1592-1598. f d 172 RDS I do SM F" k FW Hider GE er al. Clinical features 0 co on macu at ysuophr-simi_

30. ownes ,1[2 e : a ' I 99-117'1373-1383.Jar phenotype in 12 families. Arch Ophrhalmo 19,·. .

N->- M W d Y T . M Mocular dystrophy associated with monogenic Argl72Trp mmationof

31. akazawa , a a , amar ." . . ..- --2rhe peripherin/RDS gene in a Japanese family. Renna 1995, ].5.) 1~ ) 3. . .Arik K M Id LL M Id RS et al. Localization of penphenn/RDS m the disk membranes ofcone

32. "-' awOl , 0 ay , 0 ay , . d . a1 d .dd h I· hi t disk membrane morphogenesIs an renn egeneranon.J Ceil BioI

all ro p ororecepcors: re anons 'P 01992;116, 659-667. . h . IRDS d Rom-l iL C] M Id RS O· led ediatecl oligomerization of penp erm an om-I In photoretep.

33. oewen , 0 ay . ISUn e-m· '.tor disk membranes. Implications for phororecepror outer segment morphogenesIs and degeneranon.] Bicl

Chern 2000;275,5370-5378. .E K 0 II Y J

Eirzke FW er al Chromosome 19q cone~rod retinal dystrophy. ocularphem.34. vans , uva - oung , I "

rype.Arch Ophchalmol 1995; 113:195-201. . . . .35. lrabashi T, Wada Y, Sate H, et a]. Novel 615delC mutation In the CRX gene In a Japanesefamilywith

cone-rod dystrophy. Am J Ophthalmol2004;138: 876-877. . ,36. Papaioannou M, Bessanr D, Payne A, et a]. A new family of Greek ong1l1maps to the CRD locustorau.

rosomal dominant cone-rod dystrophy on 19q. J Med Genet 1998;35:42~--4~ I,37. Paunescu K, Preising MN, Janke B, et al. Genorype-phenotype correlation In a German familywitha

novel complex CRX mutation extending the open reading frame. Ophthalmology 2007;114:1348-1357.38. Freund Cl., Gregory-Evans CY, Furukawa T, et al. Cone~ro~ dystrophy due to murarionsin a novelpho-

toreceptor-specific horneobox gene (CRX) essential for malOrenance of the phororecepror.Cell1997;

91,543-553.39. Furukawa T, Morrow EM, Cepko CL, et al. Crx, a novel au-like homeobox gene, shows phororecepror.

specific expression and regulates phororecepror differentiation. Cell 1997;91 :531-541.40. Morrow EM, Furukawa T, Ravio]a E, et al. Synaprogenesis and outer segment formation areperturbedin

rhe neural retina ofCrx mutant mice. BMC Neurosci 2005;6:5.41. Peng GH, Ahmad 0,Ahmad F, et al. The phototeceptor-specific nuclear receptor Nr2e3 imeractswith

Crx and exerts opposing effects on the transcription of rod versus cone genes. Hum Mol Genet.

2005; 14,747-764.42. Pittler SJ, Zhang Y, Chen S, et aI. Functional analysis of me tod phororecepror cGMP phosphodiestel'.1Se

alpha-subwlit gene promoter: Nrl and Crx are required for full transcriptional activity.J BioIChern2004;279,19800-19807.

43. Michaelides M, Holder GE, Hunt DM, et al. A detailed study of the phenorype of an autosomaldomi-nant cone-rod dystrophy (CORD7) associated with mLitation in the gene for RIM1. Br J Ophthalmol2005;89,198-206.

44. Johnson S, Halford S, MatTis AG, et al. Genomic orga.nisation and alternative splicingof humanRIMl,agene implicated in autosomal dominant cone-rod dystrophy (CORD7). Genomics 2003;81:304-314.

45. Sun L, Bitmer tviA, Holz RW. RIM, a component of the presynaptic active zone and modulatorofexo·cytosis, binds 14-3-3 through its N-terminu$. J Bioi Chern 2003;278:38301-38309.

46. Fishman GA, Stone EM, Eliason DA, et aI. ABCA4 gene sequence variations in patients wirhautosomalrecessivecone-rod dystrophy. Arch Ophthalmol 2003; 121 :851-855.

47. KleveringBJ, BiankenagelA, Maugeri A, et al. Phenotypic spectrum of autosomal recessivecone-roddysrro-phies caused by mutations in me ABCA4 (ABCR) gene. Invest Ophthalmol Vis Sci2002;43:1980-1985,

48. Maugeri A, KJevering BJ, Rohrschneider K, et a1. Mutations in the ABCA4 (ABeR) genearethemajorcause of autosomal recessive cone-rod dystrophy. Am J Hum Genet 2000;67:960-966.

49. Michaelides M, Chen LL, Brantley Jr MA, et al. ABCA4 mutations and discordant ABCA4allelesinpa·dents and siblings with bull's-eye maculopathy. Br J Ophthalmol 2007;91: 1650-1655.

50. Kun-Levin MM, Halfyard AS, Bunce C, et al. Clinical variations in assessment ofbuH's-eyemaculopamy.Arch Ophth,lrnoI2002;120'567-575.

51. Cide~iyan AV, Swider M, Aleman TS, et al. ,ABCA4-associated retinal degenerations sparesrrucrureannfu~cnon. of the human parapapillalY retina. Invest Ophthalmol Vis Sci 2005;46:4739--4746.

52. MlchaeiJdes M, Holdet GE, Webster AR, et at. A detailed phenotypic study of "cone dysuophywithsu-pernormal rod ERG." Br J OphthalmoI2005;89:332-339.

53. Hood DC, Cideciyan AV, Halevy DA, Jacobson SG. Sites of disease action in a retinal dysrrophywifhsupernormal. and dela~ed ro.d electroretinogram b-waves. Vision Res 1996;36:889-901.

54. ",:",uH, Cowmg JA, ~Lchael~~es M, et al. Mutations in the gene KCNV2 encoding a voltage-garedporas"SlUmchannel subunit cause cone dystrophy with supernormal rod electroretinogram" in humans.AmJHum Genet 2006;79:574--579.

55, O~mirc.iFY, Rigatti BW, Wen G, et aI. X-linked cone-rod dystrophy (locus COOl): idenrificarionofmu-tatlons In RPGR exon ORF15. Am J Hum Genet 2002;70:1049-1053.

56. Yang Z, Peachey NS, Moshfeghi OM, et al. Mutations in the RPCR gene cause X-linked conedystrophy.Hum Mol Genet 2002;11:605-611.

57. Wegscheider E, Preising MN Loren B F d fl' . ' ... ..' z. un us auto uorescence m carriers of X-linked reces5[\'freunl[lS

plgmentosa associated with mutations in RPGR d I" . '_.1 d I"I h· I d G fi h ' an corre aoon With electrophyslOlogJC<Uan JllClOp ySlca ata. rae es Arc Clin Exp OphthaimoI2004;242:501_511.

CHAPTER

Fundus Autofluorescence inX-Linked Retinoschisis

a-linked retinoschisis (XLRS), also known as juvenile rerinoschisis, is the mostcommon cause of macular degeneration in young men (1,2), with an incidencebetween 1:5000 and 1:25,000 (3). Rarely, a homozygous female from a consan-

guineous marriage can be affected. The disease occurs in all races, with the highestprevalence in Finland (4).

Patients with XLRS have a highly variable clinical course. Patients present mostcommonly as school-aged boys who fail vision-screening examinations (60%) becauseof strabismus (30%) or vitreous hemorrhage (1,2,5). The classic "spoke wheel" appear-ance may be present in the fovea (3), but eventually nonspecific macular atrophy OCCLUS

in adulthood (6). Therefore, the diagnosis can be missed when patients present late oratypically.

RSl, the gene tesponsible fot XLRS, is located on chromosome Xp22 and en-codes the protein retinoschisin. It is not known to be expressed anywhere else in thebody besides the retina and there are no systemic associations with XLRS. To date,there are 132 different pathogenic mutations known to cause XLRS (7). RSI gene de-fects have complete penetrance, but the phenotypic expression of the disease is vari-able (8) and there is no known genotype-phenotype correlation (9,10).

In children and young adults with XLRS, the differential diagnosis includesuveitic macular edema, myopic foveal schisis, and other inherited retinal diseases suchas enhanced S-cone syndrome (ESCS)/Goldmann-Favre, congenital stationary nightblindness, Srargardr disease, and familial exudative virreorerinopathy. Adults withXLRS usually progtess to macular atrophy, and the major differenrial diagnosis is age-related macular degeneration (AMD).

MOLECULAR BASIS AND PATHOLOGYOFXLRSThe cellular location of retinoschisin in the retina is a controversial issue. It was ini-tially thought that retinoschisin is secreted by photo receptors and ganglion cells.Specifically, it was first found in the photoreceptor inner segment microsomal andsynaptic compartments (11-13).

In a more recent study using an epitope unmasking protocol, retinoschisin im-munoreactivity was also found in the plasma membrane of inner retinal cells in addi-tion to photoreceptors and ganglion cells (14). Since retinoschisin has not been con-clusively detected in Muller cells (14,15), the long-standing hypothesis that Mullercells contribute to the pathogenesis ofXLRS has yet to be verified experimentally.

In a mouse model of XLRS, loss of retinoschisin led to disrupted synaptic interac-tions between photoreceprors and bipolar neurons in the outer plexiform layer (OPL)(J 5,16). Dming development of XLRS mice, failure of centrifugal displacement of


IMAGING AND DIAGNOSTIC TECHNIQUES

dendrites and synapses, as well as other inner retinal neurons and synapses, wasthought to be responsible for the splitting of the putative fibers of Henle (14).

Improvement of the electronegative electroretinography (ERG) b-wave Was

found in adult RS1-deftcient mice treated with adeno-assoclated virus (MV) carryingthe wild-type gene (16). A single injection of AA V-RSI resulted in sustained RSI ex-pression and functional rescue. This demonstrated that even when loss ofretinoschisin was long-standing, function could be recovered (I6).

There is only one human pathology specimen of a patient with XLRS. In this 19-year-old man's enucleated eye, reduced levels of retinoschisin immunoactiviry wereob-served in both the macula and the peripheral retina compared to a normal eye of anage-matched control (17). Clinically, about 50% ofXLRS patients manifest periphernlretinoschisis, and some also have bridging vessels from the inner to the outer layer.Traction on these bridging vessels causes vitreous hemorrhage.

The classic foveal schisis seen in XLRS is best visualized at the slit lamp using red-freeillumination. Similarly, red-free photos demonstrate the cavities better than color

fundus photography (Fig. UClA).In XLRS, optical coherence ropography (OCT) shows cystic spaces in the mac-

ula at the level of the retinal nerve fiber layer and inner nuclear layer, correspondingto the schisis cavities (Fig. 11CIB). However, similar findings are observed in cystoidmacular edema. As in other diseases with macular thickening, there is no correlationbetween foveal thickness and visual acuity in XLRS. Macular atrophy, which is oftenobserved in late stages ofXLRS, is another nonspeciftc finding on OCT. In the pres-ence of vitreous hemorrhage, OCT is of limited value because the signal cannot pen·

etrate dense vitreous opacities.

FIGURE 11 C.l. A 34-year-old patient with XlRS IAIR~·free fundus photograph shows stellate cystoidstructurE'radiating from the fovea. (8) OCT with splittingofoul,rplexitorrn layer or Henle's fiber layer.---------

CHAPTER llC FUNDUS AUTOFLUORESCENCE IN X·L1NKEO RETINOSCHISIS

Rod Specific Maximum Photopic 30 Hz TransientScotopic Flicker Photopic

Normal

bLSr-t-- b-wave

~ ~

eFigure 1:

;\~ . wave

- ..a-wave

.. - ,. - -Rodfunction Mb: edrcd lind cone Cone function Cone function

function

XLRS

C6 ue ctrone e t~~.

_rr:. '# :i\L-._~ - .. a-wave-, . . - - -Undetectable Electronegative Simplified waveform Borderline norm al

FIGURE llC.2. Standardized full-field ERG in accordance with the International Society for ClinicalElectrophysiology of Vision (lSCEVI. Rod-specific, rnaxirnurn scotopic, photopic flicker, and transient pho-topic responses in norrnal subjects (above) and electronegative XLRS patients (below) are shown.

On fluorescein angiography (FA), rhe cystic changes observed in XLRS do nor fillor srain wirh fluorescein. Thus, FA is useful clinically to rule out leaking cystoid mac-ular edema when the diagnosis of foveal cysts is uncertain. As in XLRS, patients withESCS/Goldmann-Favre may also have nonleaking cystic spaces on FA. However,ESCS/Goldmann-Favre often can be distinguished from XLRS based on fundus ap-pearance and rhe ERG findings.

Elecrrodiagnostic testing is essential in the work-up ofXLRS and is important inesrablishing the diagnosis. FuJI-field ERG shows a reduced b-wave with a preserveda-wave (Fig. llC2). This finding, known as an electronegative ERG, is seen in alimited number of other inherited diseases, such as congenital stationary night blind-ness, Duchenne muscular dystrophy, Batten disease, and some forms of cone-roddysrrophy (CORD; see also Chapter liB). Late in the course of the disease, borh a-and b-waves may be reduced, bur fuJI-field ERG usually remains electronegative.Multifocal ERG may show reduced macular function, which is not specific to XLRS(8). There is no specific visual field defect associaredwith XLRS.

PHENOTYPES AND AUTOFLUORESCENCE FINDINGSDefects in RSI have great allelic heterogeneity, with different mutations causing a widespectrum of phenotypes. There are four distinct autofluorescence (AF) patterns foundin XLRS. Three of them--early cystoid changes, intermediate hyperfluorescenr rings,and eventual macular atrophy-are likely to represent different stages in the progressionof rhe disease. The fourth AF pattern ofXLRS is a rater presentation of white dots onfuundusexamination, as recently described in rhe literature (18).

In young patients with foveal schisis, AF discloses a stellate pattern of round-ovalareaswith signal similar to that of the background. At this site, and in normal circum-srances, a reduced AF signal is observed mainly related to blockage by luteal pigmenr.This AF pattern in XLRS seems to be the result of macular pigment displacement atthe sire of the cystic cavities (Fig. IIC.3); however, it is not specific for XLRS and can


FIGURE llC.3. A 29-year-old patient with XLRS. AF shows pigment displacement of the macularpi~·ment between cysts.

be seen also in ESCS/Goldmann-Favre, cystoid macular edema of any cause, andmy-opic foveal schisis (19).

In the inrermediate stare ofXLRS, a high-density ring is seen in the macula onAF,indicaring RPE involvement earlier than first realized (Fig. 11C.4). This ring is alsoseenin patienrs with retinitis pigmenrosa, CORD, and bull's-eye macular dystrophy(20-22). High-density AF lesions suggest increased RPE accumulation of lipofuscinfrom increased or abnormal phororeceptor outer segment shedding. This disruptionofRPE metabolism eventually leads to RPE atrophy.

A

FIGURE 11C.4. A 49-year-old patient with an intermediate phase 01XLRS (A) ell d hid Rpc,n. . . 0 or un us p otograph shows matte rthe macula.IB) AF reveals a hypertluorescent ring In the macula with early RPE t hv i ha rop y In t e center.

CHAPTERllC FUNDUSAUTOFLUORESCENCEINX-LINKEDRETINOSCHISIS 111

FIGUREllC.5. Patientwith advancedmacularchanges inXLRS.IAlColorfundusphotographshows central atrophicRPE.18) AFshows low-densitysignal oversevere macularatrophysurroundedbya ringofhighdensity.

In older patients, the schisis cavities often collapse, leaving nonspecific RPE atro-phy, which is observed as an area of low AF signal (Fig. IIC5). Ar rhis late stage, thediagnosis can be confused with AMD. XLRS can still be suspected, with clues such asdecreased vision in childhood and a family history of similar problems in males. Evenwhen the diagnosis is unclear in late stages of disease, the ERG remains electronega-tive. However, similar findings (atrophic macular changes and elecrronegative ERG)can be also found in some patients with CORD (see Chapter lIB).

Another unusual form of XLRS presents as atrophic changes associated withdrusen-like multiple white dots (Fig. 11C6). These dots differ from drusen associatedwith age-related maculoparhy by their puncrate appearance in individuals under age50. Orher conditions associated with foveal or parafoveal whire dors and flecks includeSrargardr-fundus flavimacularus, retinitis punctara albescens, ESCS/Goldmann-Favre,Wagner virreoretinal dystrophy, yellow dot dystrophy, and other forms of maculardystrophy (23). However, these other conditions have differenr funduscopic appear-ances and do not manifest an electronegative ERG (24,25).

Besides foveal cysts, another common presentation in childhood is vitreous hemor-rhage from vitreous veils rhar have rom retinal vessels (4). The differential diagnosis ofvitreous hemorrhage in childhood includes Norrie disease, familial exudative vitreo-retinopathy, retinopathy of prematurity, persistent hyperplastic primary vitreous, andCoats disease. Ultrasonography is useful in such cases to rule out retinal detachment. AFis not as useful in this situation, but it would show a blocking defect from vitreous veilsor hemorrhage (Fig. IIC?).

RECENT INSIGHTS INTO THE PATHOGENESISOF XLRSThe defective protein, retinoschisin, is expressed mainly in photoreceptors and thenquickly raken up by Muller cells and transported to the inner retina (14). This mayexplain rhe electronegariviry of rhe ERG b-wave rhar is typically seen in XLRS

B


FIGURE IIC.6. IA) Fundus photograph and (B) AF 01 a patient with line intraretinal white lesions along the ven"lesandart~rioles of the periloveolar vascular network. Note that the AF images reveal increased AF signal corresponding to the white dots.

FIGURE 1IC.7. A 12-year-old patient who presented with vitreous hemorrhage and vitreous veils at birth. (A) Red-freelundus photography. (B) AF shewing a blocking delect caused by vitreous veils.

A

A

CHAPTER l1C FUNDUS AUTOFLUORESCENCE IN X-LINKED RETINOSCHISIS 113

patients, which suggests that the disease causes post-phorotransducrion or inner-retinal pathology. The exact pathogenesis of the disease is still not completelyunderstood, as retinal splitting occurs along all layers of the retina, most oftensuperficially.

Fundus AF has provided new insighrs into the pathogenesis of XLRS. AlthoughXLRS was previously thought to be primarily a disease of the retina, high-intensitysignals on AF suggest that there is increased RPE metabolism early in the course ofthe disease. As seen in other conditions, such as retinitis pigmentosa, a ring of mal-functioning RPE may be due to improper photoreceptor outer segment shedding andsubsequent accumulation of lipofuscin. This accumulation of lipofuscin eventuallymay lead to RPE dysfunction and atrophy. The exact role of RPE dysfunction inXLRS has only been recently realized and is not well understood.

SUMMARYAF findings are useful in patients with suspected XLRS based on previous ocular orfamily history. Thus, if small round-oval areas of increased AF signal at the fovea areobserved in a young male, the diagnosis of XLRS should be suspected. FA can con-firm that these are truly schisis cavities if they are not hyperfluorescent, narrowing thediagnostic possibilities. In a patient with a central area of reduced AF signal in themacula, with or without a surrounding ring of increased AF, the diagnosis ofXLRSshould be added to the differential diagnosis.

Unlike FA, AF and OCT are noninvasive and take seconds to obtain. This is ofgreat advantage when evaluating children and the elderly. However, AF, like OCT, isof limited value in the presence of vitreous hemorrhage or media opacities.

Fundus AF, in isolation, is of limited value for establishing with certainty the diag-nosis of retinoschisis; it should be used with other imaging techniques and electrophys-iology testing (see above). RSJ gene testing is available to confirm a diagnosis ofXLRS(hnp.r/www.nei.nih. gov/ resources/ eyegene.asp, http://www. ngrl.org. ukiManchester/).Genetic testing also identifies female carriers of RSJ (26,27). Early and accurate diagno-sis may be important in genetic counseling of RS families. By age 60, the visual acuityof most XLRS individuals declines steadily to approximately 20/200.

Future RSJ gene therapy will be of benefit for individuals with molecularly diag-nosed XLRS (16). Imaging techniques such as AF, OCT, and electrophysiology willhelp in determining which patients are eligible for treatment.

ACKNOWLEDGMENTSWe thank the staff of the Medical Imaging Division of the Edward S. Harkness EyeInstitute for their excellent work, and Brian Song for contributing to Figure llCI.We thank the members of the Graham Holder and Andrew Webster laborarories forsharing their advice and ideas. S.H.T. is a Burroughs-Wellcome Program inBiomedical Sciences Fellow and is also supported by the Charles E. CulpeperScholarship, Foundation Fighting Blindness, Hirschi Trust, Schneeweiss Stem CellFund, Crowley Research Fund, Joel Hoffmann Scholarship, Barbara and DonaldJonas Family Fund, Hartford/American Geriatrics Society, Eye Surgery Fund,Bernard Becker-Association of University Professors in Ophthalmology-Research toPrevent Blindness, and National Institutes of Health (EY004081).

REFERENCES, A, d Kerk AL. Dominantly inherited cystoid macular edema.Am J

1. Deurman AF, Pickers AJL, an e

Ophrhalmol 1976;82:540-548. d 0 fi d: Oxford University Press 1998-8992 TraboulsiE,ed.GencticDiseaseoftheEye, l sr e .. x,or. . ' '.

, M AT X I" ked rerinoschlslS. Bf J Ophchalmol 1995,79:697-702.3 George NO, Yates JR, J oore .' In D LA X link d reti "

'. CU" A FrascA Annesley Jr \XfH, on050 . ~ 1 e retllloschisls:ac~tli.4. Tanrri A, Vrabec TR, "" ~Jleng , :UmoJ 2004;49:214-230.

cal and molecular genetic review. SUI v Ophrh .' . .. I h" G ND Y 'JR ec al. X-linked retinoschisis: clinica p enorype and RSl genon'llf'i~5. Pimenides D, eorge .Yates , . r-

86 UK patients. J Med Genet 2005;42:e35. ., ..IL B S F . MH et a.I Xi-linked congeoltal rerinoschisis. Graefes Arch Clinhp6. Ke rer U, rurnmer , oerster , .

Ophchalmol 1990;2280432-437, ,', ,C 'R F ' J ' licarions of the specrrum of m utattons found In 234 cases With X-hnkedju,7. onsomum . uncnona Imp I _

'I ' I" Th R ' schisis Consortiunl. Hum Mol Genet 1998;7:118)-1192.vem e rennosc lISIS. e etlno '- , . ...Ek dl L And S Ab hamson M Juvenile X-linked rerinoschisis WIth normal scotopic b-wavein8. san 1, reasson o, ranarucou r-o . .

the electroretinogram at an early stage of the disease. Ophd~a.lml~ Gen~t 2005;2~: 111.-117.9. Inoue Y, Yamamoto 5, Okada M, et al. X-linked rerinoschisis With pomr mutations In the XlRSl gene.

Arch Ophrhalmol 2000;118:93-96. . ... .10. Takerani R, Yokoyama T, Horta Y, ct al. A case of juvenile rerinoschisis diagnosed by analysis of theXLRS

1 gene. jpn J OphrhalmoI2000;44:319. ." . ..11. Grayson C, Reid SN, Ellis IA, er al. Retinoschisin, the Xi-linked recinoschisis prareln, IS a secretedphOTore-

cepror protein, and is expressed and released by Weri-Rb 1 cells. Hum Mol Genet 2.000;9:1~T~-1879.12. Reid SN, Yamashita C, Farber DB. Rerinoschisin, a photoreceptor-secreted protem, and HS Jnteraction

with bipolar and Muller cells. J Neurosci 2003;23:6030-6040.13. Molday LL, Hicks 0, Sauer CG, er a]. Expression of Xdinked recinoschisis protein RSl in photoreceptor

and bipolar cells. Invest Ophrhalmol Vis Sci 2001;42:816-825.14. Takeda Y, Fariss RN, Tanikawa A, er al. A retinal neuronal developmental wave of rerinoschisin expres-

sion begins in ganglion cells during layer formation. Invest Ophchalmol Vis Sci 2004;45:3302-3312.15. Weber BH, Schrewe H, Mol day LL, et al. Inactivation of the murine X-linked juvenile rerinoschisisgem,

Rs 1h, suggests a role of retinoschisin in retinal cell layer organization and synaptic structure. ProcNat!Acad Sci USA 2002;99:6222-6227.

16. Zeng Y, Takada Y, Kjellstrom 5, et al. RS-l gene delivery to an adult Rs 1h knockout mouse model restoresERG b-wave with reversal of the electronegative waveform of X-linked retinoschisis. Invest OphthalmolVis Sci 2004;45:3279-285.

17. Mooy CM, Van Den Born Ll, Baarsma S, ec al. Hereditary X-linked juvenile retinoschisis: a revie\\Ioftherole of MUller cells. Arch Ophrhalmol 2002; 120:979-984.

18. Tsang SH, Vaclavik V, Bird AC, et a1. Novel phenotypic and genotypic findings in X-linked rerinoschisis.fueh Oph'h,lmol 2007; 125,259-267,

19. Sayanagi K, lkuno Y, Tano Y. Different fundus autofluorescence patterns of rerinoschisis and macularhole retinal detachment in l-ugh myopia. Am J OphthaJmol 2007; 144:299-301.

20. Robson AG, Saihan Z, Jenkins SA, et al. Functional characterisation and serial imaging of abnormalnm-dus autofluorescence in patients with retinitis pigmentosa and normal visual acuity. Br J Ophlhalmol2006;900472-479,

2 L Robson AG, Moreland JD, Paulei1clloffD, et al. Macular pigment density and distribution: comparisonoffundus autofluorescence with minimum motion photometry. Vision Res 2003;43: 1765-1775.

22. Robson A, Saihan Z, Jenkins 5, et a1. Macular function and serial imaging of abnormal fundus allloRuo-rescence in patients with retiniris pigmentosa and good visual acuity. Invest Ophthalmol VisSci2006;ARVO E-AbsuacL

23. Audo I, Tsang SH, Fu AD, et a1. Autofluorescence imaging in a case of benign familial fleck retina.ArchOphthalmoI2007;125:714-715.

24. Tsui f, Song B,.Lin C-S, et al. A practical approach to retinal dystrophies. Retin Physician 2007;4:18-26,25. Robson AG, Richardson EC, Koh AH, et al. Unilateral electronegative ERG of non-vascular aetiology.Br

J Ophthalmol 2005;89,1620-1626,

26. Si.eving PA, Bi.ngham EL, Kemp J,. et al. Juvenile X-linked retinoschisis from XLRS1 Arg213Trp muraoonWith Freservatlo~ of the electroretinogram scotopic b-wave. Am J Ophthalmol 1999; 128: 179-184.

27. H~wltr ~~'. F'tzGer~ld L~, Scatter LW, et al. Genotypic and phenotypic spectrum of X-linke.:!retinOSclllSIS In Australia. Chn Exp Ophthalmol 2005;33:233-239.

174 SECTION II ClINICAL SCIENCE

I

I

Fundus Autofluorescencein Leber Congenital Amaurosis

INTRODUCTIONLeber congenital amaurosis (LCA) accounts for around 5% of all inherited retinal dys-trophies and is the earliest and most severe form of inherited retinal disease (1-4). LCAis clinically and genetically heterogeneous, although most forms show autosomal reces-sive inheritance. Patients with nonsyndromic LCA typically have an onset of poor vi-sion and nystagmus before 6 months of age, sluggish pupillary reactions, and unde-tectable electroretinogram (ERG) (5). Their vision, when they are old enough forformal assessment, is usually less than 20/400. Children with LCA are usually hyper-opic and may demonstrate the oculodigital sign (repetitive pushing of the knuckle orfinger into the eye). The appearance of the fundus is highly variable (6-8). A normal-appearing fundus may be encountered in infancy (6,8,9), although later in childhood avariety of fundus abnormaliries may be present (6,7). These include typical retinitispigmentosa (RP) (6,7,10), salt-and-pepper appearance of the fundus (10-15), in-creased granularity of the retinal pigment epirhelium (RPE) (10,13,15), white spots orfundus flecks (6,10,16), macular coloboma (6-8,10,17), marbled fundus(6,7,10,18,19), peripheral nummular pigmentation (6,10,20), attenuation of the reti-nal vessels (14,15), and optic atrophy (11). Macular changes, peripapillary hypopig-mentation, and lack of pigment migration into the retina (bone spicules) are commonfeatures observed in adult patients (15,21).

The clinical heterogeneity of the disease is reflected by the genetic heterogeneity.To date, 14 causative genes have been found to be mutated in patients with LCA andjuvenile retinal degeneration, and explain approximately 70% ofLCA cases (22). Thesegenes are expressed preferentially in the retina or the RPE. Their putative functions arediverse and include vitamin A metabolism (RPE65) (23,24), phororransduction(RetGClIGUCY2D) (25,26), retinal embryonic development (CRX) (27), protein traf-ficking (AIPLI and RPGRIPl) (28-30), photoreceptor cell structure (CREl) (31), andG protein trafficking (CEP290) (32,33).

Compound heterozygous or homozygous mutations in RPE65 result in a numberof different retinal degenerations, including LCA and early onset severe tetinal dystro-phy (EOSRD) (15,23,24,34-36). In various series, mutations in RPE65 accounted for3% to 16% of cases of LCA/EOSRD (34,37-42). Very recently, RPE65 has gainedparticular interest because of the initiation of small pilot studies in which patients withretinal dystrophies due to mutations in this gene received treatment with subrerinal in-jections of recombinant adena-associated virus vector expressing RPE65 complemen-tar), DNA (43-45). These studies represent the first attempt to use gene therapy to treatan eye disease. The results of similar treatments in animals have been very promising,and investigators were able to restore vision in a naturally occurring animal model(46-49). Patients with RPE65 mutations have better visual function than is typicallyseen in LCA, especially in childhood (15,21,35). Although severe visual impairment is

115


d i . f: ith visual responses elicited only in bright surroundings,childrennote In In ancy, W b fid visiwith LCA and RPE65 mutations generally have poor ur use VISionIn earlyl;~.. f fcen i proves during the first years of life, allowmgthechildre,Visual per ormance 0 ten un .' th "

d 1 h I b t then gtadually declines dunng e school-ageyearsAro atten regu ar sc 00 5, U ." alth . . .number of patients retain residual islands of penpheral VISIOn, ough It IS consider.bl . d in the third decade of life (39). In higher-age groups, progr~s;vea y compromIse. .visual field loss and severe visual loss ISthe norm (50). Nysragmus IS often presentihow.

he rovi ents commonly seen in LCA are rarely seen in LCAcaUJedever, t e rovmg eye movernby mutarions in RPE65 (5).

MOLECULAR BASIS AND PATHOLOGYClinical and genetic studies suggest that, although rhere is a relatively uniformlos;o1retinal function in LCA, the underlying pathophysiological mechanisms andrerinalmorphological changes may be extremely heterogeneous. As stated above,severn!genes have been found to be mutated in LeA; howe~e~,many others remain un-known (22). The complex disease mechanisms underlining LCA were recentlyreoviewed by den Hollander and colleagues (22)

CEP290, GUCY2D, and CRE] are the genes most frequently involvedinLeA(22). The protein codified by CEP290 appears to have a role in intracellularproreintrafficking in photoreceptor cells, specifically in ciliary transport processeslSI).GUCY2D encodes a protein, RetGC-I, that is involved in the resynthesisofcGMP(see also Chapter lIB). cGMP is needed for the recovery of rhe dark starein pho-toreceptor cells following light exposure. Loss of function of RetGC-l wouldmimic a situation in which the phororeceptors would be continuously exposed to

light, which would lead to photoreceptor cell degeneration. CRBI (RP12) appearsto be essential in the morphogenesis and orientation of the photorecepror ourerseg-merits (52).

Very few histopathology srudies of genoryped patients with LeA havebeenpub.lished. In one such study (53), eyes obtained from a 33-week-old feruswith a rnuta-rion in the RPE65 gene were examined. Compared with normal tissue, the RPE65mutated retina demonstrated cell loss and thinning in the photoreceptor celllayer(outer nuclear layer [ONL]), decreased immunoteacrivity of phototransduction pro.reins, and aberrant synaptic and inner retinal organization. There was also thickeningin Bruch's membrane and the choroid was abnormally vascularized.

Histopathology studies of the retinas obtained from an 11.5-year-old individualwith LCA and a mutation in GUCY2D have also been presented (54). The alfeCledpatient had vision of only light perception, no measurable kinetic visual fields,andaflat ERG before death. Lack of photoreceptor outer segments was observed:however,there. were reduced but present rods and cones in the macula and peripheral retina.The inner nuclear layer appeared normal in thickness and there were a reducednum-ber of ganglion cells.

IMAGING AND DIAGNOSTIC TECHNIQUESImaging patients with LeA is often challenging because of their common inabilitytofixate and th~presen~e of nystag~~s: However, imaging techniques often providevery valuable information to the cllll1Clan,and will likely be essential in the evaluationof children and adults wirh this condition once treatrn b ail blcnts ecome av a e.

CHAPTER 110 FUNDUS AUTOFLUORESCENCE IN LEBER CONGENITAL AMAUROSIS

Fluorescein AngiographyFluorescein angiography (FA) may be useful for determining RPE changes and atro-phy that are not detected clinically. However, fundus autofluorescence (AF; seebelow), being noninvasive, has now replaced FA for this purpose.

Optical Coherence TomographyOptical coherence tomography (OCT) studies in LCA are scarce, but they have pro-vided important information regarding the retinal structure in patients with this dis-order, including patients with mutations in RPE65, RPGRIPl, CEP290, RDHI2,and CRBI.

In patients with LCA and mutations in RPE65, OCT demonstrated reducedONL thickness at the fovea, even in the youngest patient studied (age 3 years) (55).Although reduced, measures of the foveal ONL thickness suggested preservation ofsome foveal cones even until later in life (55). RPE pigmentation (a measure of RPEintegrity), as demonstrated by the sub-RPE backscattering index, was normal in LCAcaused by mutations in RPE65 (55).

OCT images obtained from a patient with LCA due to a mutation in the RP-GRIP1 gene demonstrated retained central retinal architecture with normal ONLthickness, which decreased to immeasurable levels outside the fovea (56). Similarly,OCT srudies demonstrated that patients with mutations in CEP290 retain photore-ceptors and inner laminar architecture in the cone-rich central retina, independently ofthe severity of visual loss. Photoreceptor ceUloss and distorted retina, suggesting neu-ral-glial remodeling, were present elsewhere (57).

In contrast to patients with RPE65, RPGRIP1, and CEP290, patients with muta-tions in RDHl2 were found to have a lack of retinallarnination (distorted retinalstructure) with variable retinal thickness (thin or thick) on OCT (58). Similarly, inthose with mutations in CRBI, OCT demonstrated a very thick retina lacking on thenormal retinal architecture and resembling immature retina (59).

On the basis of the above findings, patients with mutations in RPE65, RPGRIP1,and CEP290 may be good potential candidates for gene replacement therapy. OCTappears to be an important tool to evaluate which patients might be eligible for po-tential treatments for this disease.

Fundus AutofluorescenceIn vivo recording of RPE AF provides indirect information on the level of metabolicacrivity of the RPE, which is largely determined by the rate of turnover of photore-ceptor outer segments. Progressive loss of lipofuscin occurs when there is reducedmetabolic demand due to photoreceptor cell death. This is consistent with studies onpatients with RP that showed a correspondence between areas of decreased AFand areas of photoreceptor cell Joss (60). It suggests that decreased AF may be a goodmarker for the integrity of the RPE/photoreceptor cell complex, and that AF imagingmay be useful to evaluate whether there is capacity to restore retinal function follow-ing treatment.

Patients with LCA and EOSRD may have a normal distribution of AF through-out the fundus (patient I, Table IID.I; Fig. IID.1) (61,62). In some patients, aparafoveal ring of mildly increased AF can be detected (patients 2 and 3, Table IID.I,Fig. IID.I). A moderately decreased AF signal along the arcades and in the rnidpe-riphery can be detected. In contrast, patients with typical RP, who also often demon-

· I d Funduscopic FindingsI ~s~um~m~a~ry~O~f~C~r~m~ic~a~I.~E~I:e:ct:ro~p~h~Y:S~iO~I:Og:':c=a:,_a..n-:-:------;::::;-----;.\ Patient 3 Patient 4 Patient5

EOSRD -1020/400 RE,10!5U~

lE lall mlPaleopticdi~,

macula,andperiphe~,witospottygiitte"n~rellexes


TABLE 110,1

Patient 1 Patient 2

DiagnosisAge at examination (yearslBCVA

LCA37Light perception

EOSRD720/200-20/100 DU

LCA24light perception

LCA15Light perception

fundus features Pale optic discs,a normal macularappearance,attenuated retinalvessels. and bothmild hypopigmen-tation at the levelof the RPE andintra retinalpigment in themidperiphery

Scotopic andphotopicnondstectable

Pale optic disc,central RPE defect,pale fundus, clearlyvisible choroidalvessels

Mild attenuationof the retinaivessels;otherwisenormal

Mild pallor of theoptic disc; mildattenuationof the retinalvessels; subtlesan-and-pepperappearance inthe midperiphery

ERGIISCEVstandard) Scotopic andphotopicnondetectable

Scotopic andphotopicnondetectable

Scotopic:non recordable,maximal responseamplitude 25% ofmean; 3D-Hz flicker:severely prolongedimplicit timas"

Yes after birth,none at present

Nystagmus Yes Yes Yes Yes

Scotopicandphotopicnondetec~ble

BCVA, best-corrected visual acuity; EOSRD,early onset severe retinal dystrophy; ERG, electroretinogram; ISCEV, International Society for Clinical EleGtrophysioJogyofVisi{h~lE, left eye; au, both eyes: RE. right eye: RPE. retinal pigment epithelium"Data from examination at the Department of Pathophysiology of Vision, University of TUbingen, Germany (Head Prof. Dr. E ZrennerJ.

strate a ring of increased AF at the macula, have very reduced AF signal in the rnidpe-ripheral retina (Fig 110.1; see also Chapter IIA),

Preservation of the AF signal in LCA indicates the presence of Structurally intactphotoreceptors and the inregriry of the photoreceptor/RPE complex (see below).Thedistinction between photoreceptor cell death and cell dysfunction is important andwill be essential for distinguishing those patients who may benefit from future rhera-peutic interventions. If the photoreceptor cells are viable but dysfunctional, genetherapy might allow recovery of function; under such circumstances, cell transplama-tion would be inappropriate. In contrast, if loss of vision is due to photoreceptor celldeath, gene therapy would serve to delay the progress of the disease by preventing ceUdeath of compromised bur surviving cells. Fundus AF and OCT findings (seeabove)suggest that the time course of progressive photoreceptor cell death may be slowinasubset ofLCA families/patients; viable photoreceptors may still be present evenuncilmidlife in these patients, Thus, there should be a window of opportunity to treatsome of the patients affected with this devastating disease.

As in patients with RP1 a ring of increased AF can also be observed in parienawith LCA (61,62). The significance of this ring of increased AF in patients with RPis explained in detail in Chapter IIA. However, the significance of this AF fearoreand its correlation to visual function in LeA is not known.

In contrast to the preserved AF signal in the above subser of LCA and EOSRD pa.tienrs, patients with LCAJEOSRD associated with RPE65 mutations were foundto

CHAPTER110

A,B

O,E

FUNDUSAUTOFLUORESCENCEIN LEBERCONGENITALAMAUROSIS

C

F

FIGURE 110.1. AF images of IA.B) a 32-year-old subject with no eye disease (A, image aligned lrom 16 single images;B, single frame of Af imagel.lCI a patient with RP,IO-FI three patients with LCA(patients 1-3. Table 1101). and IG,H) tWDpatients with EDSRD(patients 4 and 5, Table 1101). Comparedwith the normal AF distribution IA,B), the patient with RPIC)exhibits a paratoveal ring of moderatelv increased AFand severely decreasedAF eccentric to the macula, including the periph-ery.IO.E) In tWDLCApatients (patients 1 and 2, Table 110.11.a relatively normal distribution DIAF is shown. IF) In one patientwith LCA(patient 3, Table 110.1). the AD distribution is normal at the fovea but there is a paratoveal ring ot moneratelv in-creased Af. In the rnidperipherv. a moderately reduced AF signal with some topoqaphic correspondence tD areas of pigmentmigration into the retina is observed (G,H) Patients with EDSRD{who were not associated with RPE65 or LRAT mutationsl

demonstrated a clear Af signal.


A

have absent or very low AF signal beginning in the first decade of life (62). FigureIID.2 shows images from parienrs with LCAJEOSRD, homozygous or compoundheterozygous for mutations in RPE65. The fundus AF signal outside the maculaisverylow and very similar to that recorded ar the optic nerve head and large retinal vessd;At the macula, the signal is low at the fovea, as seen also in individuals with no retimldisease, resulting from blockage of rhe AF signal by the macular pigment (seealsoChaprers 3 and 9), but is surrounded by an area where AF is present (Fig. 110.2). Theimages are generally blurred and coarse-grained. With the detection sensitivity ofcon-mercially available instruments, such as the HRA2, it is difficult [Q detect anyAFsig-nal in most patients with LCNEOSRD and RPE65 mutations.

The finding of absent AF in patients with LCAlEOSRD and mutations inRPE65 was not due to media opacities or a high refractive error. Because the AFsig.nal was very low not only in the averaged images but also in the single-frame image,nystagmus can be ruled out as the underlying reason for the absent AF. In patiennwith nystagmus of other origins, although rhe averaged AF image may be sometimesdark due to eye movements, ir is generally much brighter that those observedinLCA/EOSRD and mutations in RPE65, and single-frame images have a dernonstra-ble AF signal (Fig IID.I) (61-63). However, OCT findings in patients withLCNEOSRD and mutarions in RPE65 indicated still-viable phororeceprors despitethe absence of AF (see above) (55-59,62), suggesting that the lack of AF signal isnotdue to atrophy of the RPE.

The lack of AF observed in patients with LCAIEOSRD and mutations inRPE65, even from early on in the course of the disease, is characteristic and appearsto be in accordance with the biochemical defect present in these patients andwithfindings in animal models of the disease. The RPE65 gene encodes an RPE-specific65 kD protein (RPE65) and is localized on chromosome Ip31 (24). RPE65 play"key role in the metabolism of vitamin A in the retina because it controls the finaliso-merization step from all-trans retinyl ester to II-cis retinal. RPE65-1- mice accumu-lace all-trans retinyl ester during illumination cycles (64). The absence of AF in

FIGURE 110.2. Af imagesobtainedfrompatients with EOSRD associated with mutations inRPE661AI NeariyabsenlAfina lO-year-oldpatient (B) VerylowAFsignal is observed in a 14-year-old patient; some preservationoftheAFsignalisobservedaroundthe fovea.

CHAPTER 110 fUNDUS AUTOfLUORESCENCE IN LEBER CONGENITAL AMAUROSIS

RPE65~/~ mice was shown by Katz and Redmond (65) to result from a failure oflipofuscin fluorophore formation. Lipofuscin accumulates from lysosomal degrada-tion end products of all-trans rerinal from shed photoreceptor disks phagocytosed byrhe RPE (66). The absence of AF in patients with compound heterozygous or ho-mozygous mutations in RPE65 indicates that in humans as well, the formation oflipofuscin fluorophores is dependent on the normal function of the visual cycle anda normal RPE65 gene product in the RPE. The absence of retinal, both all-trans andll-cis, prevents the formation of lipofuscin fluorophores in both RPE65-1- miceand patients with RPE65 mutations.

Recently, it was demonstrated that there may be some residual AF in some pa-tients with RPE65 mutations (42); to date, however, it remains unclear how this find-ing can be explained based on the biochemical data and findings in animal models(see above). Since these patients had the same mutation in RPE65 as those in whomno AF signal was detected, it is possible that the residual AF signal observed was dueto AF signal from the sclera. Further studies are needed to elucidate this discrepancy.

i. Perrault I, Rozet JM, Gerber 5, er al. Leber congenital amaurosis. Mol Genet Merab 1999;68:200-208.2. Kaplan J, Bonneau 0, Prezal J, er al. Clinical and genetic hercrogeneiry in retinitis pigrnenrosa. Hum

Genet 1990;85:635-642.3. Foxman SG, Heckenlively JR, Bateman JB, et al. Classification of congenital and early onset retinitis pig-mentosa. Arch Ophrhalmol 1985;103: 1502-1506.

4. Fazzi E, Signorini SG, Scelsa B, et aI. Leber's congenital amaurosis: an updare. Eur J Paediatr Neurol2003;n3~22.

5. Heckenlively JR. Retinitis Pigmentosa. Philadelphia: Lippincott, 1988.6. De Laey JJ. Leber's congenital amaurosis. Bull Soc Beige Ophtalmol 199[;241 :41-50.7. Harris EW. Leber's congenital amaurosis and RlJE65. Inr Ophrhalmol Clin 2001;41:73-82.8. Margolis 5, Scher 8M, Carr RE. Macular colobomas in Leber's congenital amaurosis. Am J Ophthalmol1977;83;27-31.

9. Lambert SR, Taylor D, Kriss A. The infant with nystagmus, normal appearing fundi, but an abnormal

ERG. Surv Ophth,lmoI1989;34;173-186.10. Heher KL, T raboulsi El, Maumenee IH. The natural history of Leber's congenital amaurosis. Age-related

findings in 35 patients. Ophthalmology] 992;99:24 J-245.11. Schapperr-Kimmijser I, Henkes HE, Bosch J. Amaurosis congenita (Leber). Arch Ophrhalmol

1959;61;211-218.12. Francois J. Leber's congenital tapetoretinal degeneration. Inr Ophrhalmol Clin 1968;8:929-947.13. Smith 0, Oestreicher], Musarella MA. Clinical spectrum of Leber's congenital amaurosis in the second

ro fourth decades of life. Ophthalmology 1990;97: 1156--1 J 6].14. Perrault 1, Rozet JM, Ghazi 1, er al. Different functional outcome of RetGC1 and RPE65 gene mutations

in Leber congenital amaurosis. Am J Hum Genet 1999;64:] 225-1228.

SUMMARYIt is important to evaluate the distribution of fundus AF in patients withLCA/EOSRD. Such an evaluation may serve as a guide for genetic testing (e.g., alack of fundus AF signal, especially in young patients, would indicate rhe likely pres-ence of a mutation in RPE65), and provide important information with regard tothe potential for visual recovery in patients with this disease, once genetic treatmentsbecome clinically available. This imaging technique could be particularly helpful inyoung patients who are not capable of cooperating with procedures to obtain de-tailed measurements of visual function, ERG, or other imaging techniques, such asOCT. It may also guide the location where therapeutic suhretinal injections shouldbe performed.

REFERENCES

CHAPTER

HISTOPATHOLOGY AND MOLECULAR GENETICS

Fundus Autofluorescencein Pattern Dystrophy

II attern dysttophy (PO), a term coined by Marmor and Byers (I), referstoa• gtoUp of inherited retinal dysttophies charactenzed by the deposirionof,

highly aurofluorescent material at the level of the retinal pigment eplrhelium(RPE), with an onset late in life. The incidence and prevalence are nor known;how·ever PO is considered to be rare. In most cases there is an autosomal dominant(AD)mode of inheritance, although penetrance is variable. However, some casesmay ~esporadic or inherited as an autosomal recessive (AR) trait (the latter in associationwith reticular dystrophy). Patients may be asymptomatic or ptesent with blurredvi·sion and/or metamorphopsia. In advanced stages of the disease, reduced centralvisionand reading difficulties may be noted as a result of the development of atrophy or,I~sfrequently, choroidal neovascularization (CNY).

The most common form of PO, adult vitelliform macular dystrophy (AVMDj,ischaracterized by a bilateral solitary yellow, round ro oval sub foveal lesion wirh orwithout a central pigmented spot (2). The next most common forms are mulrifocalPO simulating Stargardt disease-fundus flavimacularus (STGO-FFM) and butterfly.shaped PO (3,4). Rarer forms of PO include Sjogren reticular dystrophy, rnacrore-ticular dystrophy, and fundus pulverulenrus, These various phenotypes can beseenindifferent individuals in the same family (5,6). Also, depending on the stage of theds-ease, the phenotype can change in appearance in the same individual over rime(7,8).

PD may be present as an isolated condition or as pan of a syndrome associatedwith other systemic disorders. PO has been found in a proportion of patientswithmaternally inherited diabetes and deafness (MIOOM) (9) (see Chapter II H),my·atonic dystrophy (10), pseudoxanthoma elasticum (PXE) (ll), Friedreicb araxia(12),and Crohn disease (13).

Clinicopathologic studies of patients with PO have shown similar findings (14-16).A histopathological study of the posrmortem eyes of a 61-year-old female reve~edfocal atrophy of the RPE in the foveolar area bordered by hypertrophic RPE,withfusrform collagenous plaques of eosinophilic material located between the atrophicRPE and Bruch's membrane. The sensory retina over the abnormal RPE displayedsignificant atrophy of the outer nuclear layer with loss of photoreceptor inner layerand outer segments. PIgment-laden macrophages with periodic acid Schiff-po,i,ivematerial had migrated Into the atrophic outer retina. Ultraviolet fluorescent mi-croscopy d~m~nstrated massive accumulation of lipofuscin within the macularRPEas well as within the macrophages' th hi . . .In e atrap rc Outer retina. Scanning electron mI-croscopy revealed a confluent area of flattened hi RPE 11 d d LVatrop ic ce s surroun e u,

184

CHAPTER l1E FUNDUS AUTOFLUORESCENCE IN PATIERN DYSTROPHY 185

taller hypertrophic RPE cells. Transmission electron microscopy demonstrated thatthe RPE cells contained many lipofuscin granules (14,15). In three patients withAVMO, RPE and photoreceptor cell loss were observed in the central area. A mod-erate number of pigment-laden macrophages were present in the subretinal space andourer retina. The RPE was distended to both sides of the central lesion with abun-dant lipofuscin (16).

PO is genetically heterogeneous. To date, the most common genes known to har-bor mutations that can cause nonsyndromic PO include PRPH2 (Peripherin 2) (Stone,personal communication referred to in Grover et al. [17]) (18), ELOVL4 (19), andBESTI (20) (previously known as VMD2; see also Chapter 11F). In addition, as notedabove, gene mutations in OTM (myotonic dystrophy) and mtONA 3243 may be as-sociated with a "PO-like" phenotype.

PRPH2 encodes peripherin 2, a membrane-associated glycoprotein restricted to

photoreceptor outer segment discs (21). Normal levels of this protein are requiredfor the morphogenesis and maintenance of the phororeceptor outer segments (22).The rds mouse exhibiting the retinal degeneration slow phenotype was found tohave a single spontaneous mutation (23). The phenotype in the rds mouse is char-acterized by abnormal development of photoreceptor outer segments in the retinafollowed by a slow degeneration of the rods and cones, resembling the phenotypesdescribed in human retinal dystrophies associated with PRPH2 mutations (24). Afeature seen in both humans with PRPH2 mutations and mice with the rds muta-tion is a loss of photoreceptor function. Ali et al. (25) demonstrated that a subreri-nal injection of recombinant adeno-associated virus (rAAV) encoding an rds trans-gene resulted in the stable generation of outer segment structures and formation ofnew stacks of discs, which were morphologically similar to normal outer segments,and electtophysiological tests confirmed function. fAA Vi-mediated gene replace-ment of peripherin 2 restored retinal ultrastructure and function for as long as 14weeks in the rds mouse. In this model, however, AAV-mediated gene replacementwas not sustained in the long term.

Although studies have described a common consistent phenotype caused bycertain mutations in peripherin 2, such as the Arg172Tryp mutation associatedwith macular dystrophy (26), significant variation in phenotypic expressivity in thesame family has also been described in PO associated with PRPH2 mutations (5).For example, the same mutation can cause retinitis pigmentosa, PD, and fundusflavimacularus in a single family (5). Yang et al. (4) described three separate fami-lies, each of which had a distinct PO phenotype that differed from those of theother families, although they all had the same mutation in PRPH2 and an identicaldisease haplotype.

ELOVL4 is a photoreceptor-specific gene that is involved in the biosynthesis ofvery long chain fatty acids. It is expressed in retina and skin. Defective protein traf-ficking may underlie the molecular mechanism associated with degeneration of themacula (27). It is possible that some patients with multifocal PO simulating STGO-FFM have mutations in this gene, as mutations in ELOVL4 have been found in pa-tients with AD Stargardt-like disease (19,28).

CLINICAL FEATURESThe age of onset in PO is usually in late adulthood, but with a positive family histotyit may occur in early adulthood. The progression of visual loss is usually slow, andmost patients maintain reading vision until later in life. The condition can be quite


asymmetrical. Monitoring vision, especially in drivers, is impO~[anL Regular se~.monitoring with an Amsler grid should be encouraged, and patients shouldbead.vised to seek ophthalmic review if sudden onset of blurring or distortion ocwhich could indicate the development of CNV. CNV was originally thoughtI~trate in PD (29), but is probably not uncommon (30,31). Macular attophyhasbobserved both at presentation and over time in patients with PO (32). ~n

Adult Vitelliforrn Macular DystrophyAVMD is most commonly dominantly inherited, with incomplete penettanceandhighly variable expression, but there are a significant number of sporadic cases,Inone study, 91 % of patients with AVMD had no family history (8). Furthermore,I,.sions may be unilateral in a high proportion of patients (8). Typical findings includea yellow (Fig. 11E.l) or pigmented central deposit, around which a c1epigmentoohalo is often seen (Fig. l1E.2). There may be a central SpOt alone or there maybeafew deposits near the lesion or in the peripheral retina (Fig. llE.3). The electro.om.logram (EO G) light rise is usually normal or mildly affected. Color vision mavb,abnormal. .

A

C

FIGURE 11E.!. Color fundus photogrnph(A). AF image (BI. and OCT(CIobtalnedlroma 54-year-old lemale with AVMD. Slit·lampbiomicroscopy disclosed a smallyellowd.posit at the left fovea IAI, with a high,increased signal on AF imaging(BI,OCTdemonstrated a well-delineated, dam.shaped elevation 01 the anteriorreflectiveband and increased rellectlvltywithinthevitelillorm lesion ICI. PRPH2 testingdemon·strated a mutation Inthe gene,

CHAPTER11E FUNDUS AUTOFLUORESCENCEIN PATTERNDYSTROPHY 187

I

I

Butterfly-Shaped Pattern Dystrophy

FIGUREllE.2. Color fundus photograph obtained from a 56-year-old male with central visual distur-bance and AVMD. A central area of increased pigmentation surrounded by a halo of depigmentation isobserved

Butterfly-shaped PD is characterized by an accumulation of yellow or brown pig-ment at the level of the RPE in a butterfly configuration (Fig. lIE.4). A subnormalEOG and normal or slightly diminished visual acuity have been described, as wellas atrophic changes and an AD inheritance (3). Although Deutrnan proposed arelatively benign course for butterfly-shaped PD, it has now clear that butterfly-shaped PD, like other types ofPD, is usually a progressive disorder with varying de-gtees of visual deterioration (3). Older individuals may have atrophic, depigmentedlesions extending into the peripapillary region, with markedly reduced visual acu-ity (33).

FIGUREllE.3. Color fundus photograph of a patient with AVMD demonstrating accumulation of yellowmaterial at the fovea and additional deposits near this main lesion distributed throughout the macula andinto the midperipheral retina.

SECTION II ClINICAL SCIENCE

Fundus Pulverentulus

A

FIGURE llE.4. Color fundus photograph (AI. FFA image (BI. and AF (el obtained from an 82-year-old femalewithbuttlrl!.shaped POfrom a family with various POphenotypes. AD inheritance. and a mutation in PRPHZ Deposition ofyellowrnatelial A,Sin a "butterfly" distribution was observed on slit-lamp biomicroscopy (AI. which demonstrated a high AF signal onAf imagin~(el. FA demonstrated linear areas of hypofluorescence surrounded by ill-defined hyperfluorescence. (Courtesyof Prole"'rAlan Bird. Moorfields Eye Hospital. London. England.)

Reticular and Macroreticular Pattern DystrophyReticular dystrophy, first described by Sjogren (34) in 1950, is characterizedbyapar.tern of pigment c1wnping that has been likened to a fishnet with knots (Fig. lIE.5A,B).Patients may have normal vision, normal electroretinogram (ERG) and EOG, andnor-mal color vision and peripheral visual fields (34,35).

Macroreticular PD was described in 1970 (36). It is extremely rare. It ischar-acterized by a larger meshwork of rericular changes in the fundus (Fig. lIE.5B).The ERG and EGG may be abnormal (37).

The term "fundus pulverentulus" has been used to describe a granular fundusappear-ance, and has been reported in families manifesting other PD phenotypes (Fig.l1E.6) (38,39).

A

FIGURE llE.5. Color fundus photograph (AI d FA (B) b - ... an 0 tamed from a patient with macroretlCularPD.

Hypo/luorescence In areas of macroreticular pigment is observed on FA. (Courtesy of Dr. PaulusDeJong.l

CHAPTER 11E FUNDUS AUTOFLUORESCENCE IN PATIERN DYSTROPHY

Multifocal Pattern Dystrophy Simulating StargardtDisease-Fundus FlavimaculatusThe form of PO known as mulcifocal PO simulating STGO-FFM is characterized bythe presence of fleck-like deposits and AO inheritance (Fig. 11E.7) (40). Itmay be in-disringuishable from AD STGO-like fundus dystrophy associared with mutations inthe ELD VI4 gene, and molecular genetic testing may be necessary to differentiate be-tween these two conditions.

IMAGING AND OTHER DIAGNOSTIC STUDIES INPATTERN DYSTROPHY

Fluorescein and Indocyanine Green AngiographyWith the advent of fundus autofluorescence (AP) imaging, fluorescein angiography(FA) and indocyanine green (reG) angiography are probably better reserved forinvestigating complications ofPD, such as eNV, rather than for diagnostic purposes.

A

FIGURE l1E.6. FA obtained from a46-year-old patient with PO (a rela-tive of the probandwith buttertly dys-trophy shown in Fig. 11E.4I. In her lefteye, at presentation, retinal pigmentepithelial defects In a granular pulv-erentulus-like appearance at the cen-tral macula were seen IA). Twentyyears later, these changes had be-come more diffuse 18). (Courtesy ofProfessor Alan Bird. Moortlelds EyeHospital, London,England.)

8

FIGURE llE.7. Colorfundus photograph IA) and AF image (8) of a patient with Friedreich ataxia and multifocal POsimulating STGO-FFM(12) Visual acuity and retinal electrophysiology were normal until optic atrophy supervened aspart of Friedreich ataxia.


FIGURE llE.8. FAobtained from a patient with AVMO demonstrating the corona sign. inwhich~ereis hyperfluorescence surrounding a central area of hypofluorescence. ICourtesy of ProfessorAlan8irfjMoorfields EyeHospital. London. England.1

In the absence of AF imaging, FA is useful for highlighting changes in rheRPE(31,39,41). FA can also demonstrate a "dark choroid" in some individualswithSTGD-FFM (see Chapter 1IG), which may help in the differentiation betweenPOand STGD-FFM (see later).

Typical features seen on FA in AVMD include a central hypofluorescemsporcor-responding to the area of central increased pigmentation and a ring of hyper fluorescencearound it (referred to as the "corona sign" [Fig. I lE.8]), which relates to a windowde-fect in areas where atrophic changes have occurred (8,42). In a fewpatients,FAmaybenormal. On ICG angiography, a foveal nonfluorescent SpOthas been observedthrough-out the angiogram, with a hyperfluorescenr area surrounding the central sporthar isev-ident in early frames of the angiogram (43).

The ICG and FA findings from a 37-year-old female with reticular dystrophycomplicated by CNV showed that in the areas of reticular changes there weresignifi·cant abnormalities characterized by intense hyperfluorescence (31).

Da Pozzo et al. (44) suggesred caution in interpreting ICG angiography,as vitelli-form marerial present in PD may bind to the ICG molecule and cause hyperfluorecence simulating an occult CNV.

Optical Coherence Tomography (OCT)OCT alone will not permit accurate diagnosis of AVMD or other typesof PD.However, it can be a useful adjunct to fundus AF and color fundus imagingto ex·c1udeassociated edema, as seen in cases complicared by CNV. In the latter, FAwouldbe indicated.

Pierro er a1. (45) carried out a retrospective review of 43 patients (72 eyes)withAVMD. In all eyes, OCT showed a well-defined central region of thickeninginthereflective band representmg the RPE. Similar findings have been reported byHayamler el. (46). In patients WIth a thinner neurosensory retinal layer overlying theAVMDlesion, the VISUalacuIty was reduced (45).

Benhamou et al. (47) described the OCT features in evolving lesionsof AVMD.the diseaseprogressed. In the vitelliform stage a w II . ib d d h pedeb.-. , e -circumscn e omc-s anon of the anterior reflective band a moderate b I . b I th . reflee·, ac <. scattenng e ow e anrenor

CHAPTER l1E 191FUNDUS AUTOFLUORESCENCE IN PATTERN DYSTROPHY

rive band, well-delineated posterior boundaries in the plane of the RPE, and an increasein the reflectivity within the vitelliforrn lesion as the lesion became smaller (atrophicstage) were observed. These authors pointed out that it was not possible to be certainabout the exact location of the yellow material seen on histopathological studies becausethose observations were made in eyes with end-stage disease. According to Benhamouer al. (47), the posterior reflective band corresponds to the RPE, whereas the highly re-flective anterior band represents the pseudo.vitelliform lesion (47). They stated that thelesion is located between the photoreceptor layer and the RPE, but it sometimes appearsthat the lesion is within the RPE.

Therefore, the common feature seen on OCT imaging in patients with PO is athickened, well-circumscribed, central focal dome-shaped lesion that appears to lie inbetween the photoreceptors and RPE (Figs. IIE.IC and IIE.9). If a central lesion isobserved in OCT images of patients wirh PD, as in AVMD, there does not seem tobe any specific difference between rhe yellow deposit and the pigmented deposit.However, the discrete dome-shaped, smooth elevation on OCT seems to depend onthe type of PD (i.e., the presence of AVMD) rather than, for instance, multifocal POsimulating STGO-FFM (personal observation). Findings from OCT alone, however,may nor be specific enough to distinguish PD from a CNV. In cases of age-relaredmacular degeneration (AMO)-associated CNV, OCT tends to show much more ex-tensive changes throughout the retinal layers, in contrast to PO, in which changes aremore focal and smooth in contour. However, in idiopathic, myopic, or inflammatoryCNV, especially if it is inactive, OCT findings may be indistinguishable from rhoseobserved in PD. Under these circumstances, history and ancillary tests, such as AF, arevel)' helpful. A vety high A.F signal is seen in PD and, most commonly, a reduced ormottled A.F signal is seen at the site of a CNV.

C

FIGURE l1E.9. Color fundus photograph IA), Af Image(BI. and OCT(CI obtained from a 56-year-old fe-male with AD POwith a mutation in PRPHl Accumulation of yellow material at the fovea was observedon silt-lamp biomicroscopy (A). The material demonstrated a very increased AF signaIIB). OCTdiscloseda smooth, well-circumscribed, dome-shaped elevation in the area of the lesion, with a highly reflectiveband corresponding to the RPEIC).


FIGURE 11E.l0. Color fundus photograph (AI and AF Image (8) obtained from a 63-year-old male with PD.A relalivelylarge,ill-defined yellowish lesion at the macula was observed on slit-lamp biomicroscopy IA) ..Fundus AF disclosed a slgnlfic,,~Increased AF signal at the site of the lesion, suggesting accumulation of lipofuscin matenal as typically observedIn POIBI.

Fundus AutofluorescenceFundus AF is very helpful in the diagnosis and evaluation of patients with PD.!nAVMD a very characrerisric well-circumscribed, very high AF signal is seen in thevitelliform stage (Figs. IIE.IB and IIE.9B)_ In patients with a central clump ofpig-ment and a surrounding halo of depigmentation, AF images usually demonstrateacentral area of significantly increased AF surrounded by a halo of reduced AF. Theevo-lution of these appearances during the course of the disease has not been welldocu-mented, but with the disappearance of the vitelliform lesion a corresponding lossofAFsignal is observed, indicating the occurrence of atrophy. AF imaging may allowthevi-sualization of disease-specific distributions of lipofuscin in the RPE even whenmoechanges are not yet visible on fundoscopy (48). Fundus AF can also be useful in thedi-agnosis of PO in cases with an amorphous deposit, allowing its differentiation fromother lesions, such as a drusenoid pigment epithelial detachment (PED) (Fig. J IE.lOl.

In a study by Renner cr aI. (8) in which 13 patients (25 eyes) were imaged with fun-dus.AP, a yellow lesion was present in 22 eyes, and central pigmentation was presenr in6 eyes. In 19 of the 22 eyes (86%) an increased AF signal was seen, and in 8 of thesemerewas an additional small Spot of reduced AF signal at the center of the area of inrreeelAF (8). Parodi et al. (49) evaluated 15 patients with AVMD by AF imaging and com-pared the findings with those observed in 10 healthy volunteers. AF imaging wasper-formed with borh short-wave conventional AF and near-infrared AF (N1A) (seeillsoChapter 6). The former technique revealed three different patterns of AF (normilfocal, and patchy), whereas only two patterns (focal and patchy) were seen withNIA,When AF patterns were correlated with functional tests (vision and microperirnerrjl."was found that patchy AF was associated with the worst functional outcome. Abnorm~short-wave convenrional AF was seen in 86% of patients with AVMD, and abnormalNIA was observed in 100%.

Saito et aI. (50) evaluated six]apanese patients (12 eyes) with what they diagnosedas AVMD. Bilateral macular lesions were present in all patients and varied fromthetypical vitelliform (five eyes) to faded vitelliform changes with RPE atrophy (fiveeyes)or a normal fovea associated with small flecks around it (two eyes). AF imagingdemonstrated small Spots of increased AF throughout the posterior pole in allcases.

A

CHAPTER 11E FUNDUS AUTOFLUORESCENCE IN PATIERN DYSTROPHY

Multifocal ERGs (mFERGs) were significantly reduced not only in the macular area,but also in the outermost ring of the mFERGs (20-30 degrees). The authors sug-gested that morphological and functional abnormalities in AVMD may not be limitedto the macula, bur may be present throughout the posterior pole. This was supportedby their electrophysiological findings (50). It is possible that these observations re-flected previous repofts describing the coexistence of fundus pulverentulus andAVMD. The images of the Japanese patients were not typical of AVMD; in one case,the lesions were larger than those typical of AVMD, and it is not clear from rhe EOGfindings whether these parients mighr have had Best disease. Another case shown hada central lesion with multifocal vitelliform peripheral lesions, which could representSTGD-FFM or mulrifocal Best disease. The patients were not reported to be from thesame family and had no family history. Thus, their findings on AF imaging do not ap-pear to be typical of AVMD. However, ir is srill possible that rhese patients have anatypical phenotype of PO that has nor been previously described, especially consider-ing rhat most previously published cases of AVMD affected Caucasians.

AF imaging is expected to playa very important role in monitoring response to fu-ture therapies. It may also be useful for identifying individuals with no known geneticmutarion, and those in whom the early signs are subtle on fundoscopy. Alrhough AF im-aging cannot identify all presymptomatic cases of PO, in many cases it can highlight de-posits that are not clearly seen on slit-lamp biomicroscopy. AF imaging is very useful fordistinguishing PO from other conditions, especially AMD (see later). AF imaging in POhas replaced FA and is much preferred by patients because it is noninvasive and less time-consuming. FA, however, is still indicated if a CNV is suspected.

ElectrophysiologyIn general, PO is not associated with gross electrophysiological abnormalities. TheEOG may be normal or milclly abnormal; the ERG is most often normal (1,8,51).However, abnormal ERG values (b-wave amplitude of the maximum rod-cone re-sponse, single flash cone response, and 30-Hz flicker response) may be elicited insome patients (8). Previously normal ERG recordings may become mildly reducedwith long-term follow-up (1); only occasionally will a marked reduction in ERG val-ues be seen (8).

Weleber et al. (5) reported a family with clinically disparate phenotypes caused bya mutation in 153/154 codon of peripherin 2. A profoundly abnormal ERG wasrecorded in one member of the family with a retinitis pigmentosa phenotype, a mod-erately abnormal ERG was found in a member with a macular dystrophy phenotype,and a markedly abnormal ERG was found in another who had also a form of macu-lar dystrophy (5).

Reduced mfERGs representing the fovea have been recorded in patients withPO, and a generalized decrease in amplitude has also been seen in a small proportionof patients (14%) (8).

DIFFERENTIAL DIAGNOSESPO should be differentiated from Best disease, STGD-FFM, acute exudative polymor-phous vitelliform maculopathy, central serous chorioretinopathy (CSR), and AMD.

In contrast to PO, in Best disease the EOG light rise is usually extinguished orseverely reduced and onset occurs in childhood or early adulthood (see alsoChapter 11F). Furthermore, the size of the virelliform lesion tends to be larger.


SUMMARY

Multif I PD can easily be mistaken for STGD-FFM (52), particularlyiftheretll0Ca . . h ~c. '1 hi STGD FFM is associared with rnutanons in t e ABCA4gene(no rarru y istory. - .' see

al Ch IIG) and it usually has a worse visual prognosIs than PD. Thepres"so apter , ..' . ceof a dark choroid on FA and a relative penpapIllary sparIng 111 STGD-FFMwillhelpin rhe differentiation. AD Stargardt-like dysrrophy caused by mutations inELQVIAmay be parr of the PD spectrum.

Acure exudarive polymorphous vitelliform maculoparhy (53), characterizedbythe presence of rransient, mulrifocal, and numerous small yellowish lesions,mayalsobe misraken for PD. As in PD, affected panents may have reduced or normalamph.tudes of ERG and EOG (54). Abnormalities in dark adaptometry have beenfoundinparients with acure exudative polymorphous virelliform rnaculoparhy (54). UnlikeIe.sions in PD, those in acute exudative polymorphous virelliforrn macuJoparhyCanclli.appear over a relatively short period of rime, with gradual recovery of vision.However, this is not always the case. Also. the onset of lesions associated with anneexudacive polymorphous dystrophy may be accompanied by headache, and ifimagedwith ICG the choriocapillaris may be abnormal, which is not seen in PD.

Old RPE changes in CSR can simulare those observed in PD (seealsoChapter13). Increased levels of AF can be seen in both condicions. However, in CSRchangesare more diffuse and ill defined than in PD and, unlike PD, only very rarelywill,well-delineated increased AF signal at the fovea be observed.

AVMD is often misdiagnosed as AMD, and fundus AF is very helpful in differen.ciating these conditions. Alchough weakly increased AF can be dereccedin patientswithearlyAMD and drusen, the high-intensity AF signal present in PO is rarelyobservedinAMD. This is parcicularly useful in patients with large drusenoid PEDs, whichclinicallymay look like vicelliform lesions. Because AVMD generally has a better prognosisandmay be associated with an AD inheritance, it is very important (Q distinguish betweenAMDandPD.

Different forms of PD have been described, with AVMD and multifocal PD sirnc-laring STGD-FFM being the most common. PD can vaty in presentation evenwithin a family with AD disease, and combinations of phenotypes can be seeninin-dividuals In the same family. Furthermore, variable penetrance has been recognizedInPD. AF ImagIng can be helpful in making the diagnosis of PD and in differenti-atmg PD from other retinal diseases, such as AMD. It is likely that fundusAF willbecome a key rool for evaluating response to future therapies.

REFERENCES1. Marmor MF, By~r~ B. Pattern .dystrophy of the pigment epithelium. Am J Ophrhalmol 1977:84:32-44.2. ~9a;:;~~:~39~1~1;I.lIcopathologlC study of a peculiar foveornacular dystrophy. Trans Am (IphrhalSoc

3. DeutmanAF,vanBlommestein]DA Henke' HE I B ~Arch Ophthalmol 1970:83:558-569.' S, er a. urrerfly shaped pigment dystrophy of the O\'t1.

4. Yang Z, Li Y, Jiang L er al A novel RDS/ . h .,h 0 h h I '. G' penp enn gene mutation associated with diverse maCUlarp enotypes. pta mrc ener 2004;25:133-145.

5. Weleber RG, Carr RE, Murphey WH, et al Phen . .... . . .. .dystrophy and f d [lavi I . : oryprc vanabJllty including renruns plgmentosa, pa£lem

, un us avtmacu arus III a Single family . h d I' f f h . herin/RDS gene Arch Opl [h I 119931 Wit a e euon 0 codon 153 or 1540 t epenp'

. 1 armo ; 11:1531-15426. Felber U, Schilling H, Weber BHF Adultvirell'r. .rations in the peripherin/RDS . H M· I arm macular dystrophy is frequently associated wirhne

gene. um utar 1997;10:301-309.

CHAPTER 11E FUNDUS AUTOFLUORESCENCE IN PATIERN DYSTROPHY 195

7. van Lirh-Verhoeven jjC, Cremers FPM, van den Helm B, et a]. Genetic heterogeneity of burrerfly-shapedpigment dystrophy of the fovea. Mol Vision 2003;9: 138-143.

8. Renner AB, Tillack H, Kraus H, er al. Morphology and functional characteristics in adult vitelliform mac-ular dysrropby. Retina 2004;24:929-939.

9. Massin P, Virally Monad M, Vialenes B, er al. Prevalence of macular pattern dystrophy in maternally i11-herired diabetes and deafness. GEDIAM Group. Ophthalmology 1999;106: 1821-1827.

10. Kimizuka Y, Kiyosawa M, Tarnai M, er al. Retinal changes in myotonic dystrophy. Clinical and follow-upevaluation. Retina 1993;13:129-135.

11. Agarwal A, Patel P, Adkins T, et al. Spectrum of pattern dystrophy in pseudoxanchoma elasticum. ArchOphdtalmoI2005;123:923-928.

12. Porter N, Downes SM, Fratter C, er a]. Catastrophic visual loss in a patient with Friedreich ataxia. ArchOphth,lmoI2007;125;273-274.

13. De Franceschi P, Cosragliola C, Soreca E, ec aJ. Pattern dystrophy of the retinal pigment epithelium inCrohn's disease. Optbalrnologica 2000;214: 441-446.

14. Parrinely JR, Lewis RA, Fom RL. Foveomacular virelliform dystrophy, adult type: a clinicopathologicsrudy including electron microscopic observations. Ophthalmology 1985;92: 1712- l 718.

15. Jaffe G], Schatz H. Histopathologic features of adult-onset foveomacular pigment epithelial dystrophy.Arch Ophrhalrnol 1988; 106858-960.

16. Dubovy SR, Hairston R), Schatz H, er al. Adult-onset foveomacular pigment epithelial dystrophy.Clinicopathologic correlation of three cases. Retina 2000; 20:638-649.

L7. Grover S, Fishman G, Stone EM. Atypical presentation of pattern dystrophy in two families with peri ph-erin/RDS mutations. Ophthalmology 2002;109:1ILO-I 117.

18. Fossarello M, Berrini C, Calanruomo MS, er aJ. Deletion in the peripherin/RDS gene in rwo unrelatedSardinian families with aurosomal dominant butterfly-shaped macular dystrophy. Arch Ophchalrnol1996;114;448-456.

19. Bernstein PS, Tarnmur ], Singh N, et a!. Diverse macular dystrophy phenotype caused by a novel complexmutation in the ELOVL4 gene. Invest Ophrhalmol Vis Sci 2001;42:3331-3336.

20. Seddon ]M, Afshari MA, Sharma S, et a]. Assessment of mutations in the Best macular dystrophy VMD2gene in patients with adult-onset foveomacular vitelliform dystrophy, age related maculoparhy and bull's eyemaculopachy. Ophthalmology 200 1;108:2060-2067.

21. Travis GH, Sutcliffe ]G, Bok D. The retinal degeneration slow (rds) gene product is a photoreceptor discmembrane associated glycoprotein. Neuron 1991;6:61-70.

22. Lee ES, Burnside B, Flannery JG. Characterization of peripherin/rds and Rom I transport in rod phorore-cepcors of transgenic and knockout animals. Invest Ophrhalmol Vis Sci 2006;47:2150-2160.

23. Van Gulik P], Korrweg R. Suscepribdiry to follicular hormone and disposition to mammary cancer in fe-male mice. Am J Cancer 1940;38:506.

24. Van Nie RD, Ivanyi D, Demanr P. A new H-2linked mutation, rds, causing retinal degeneration in themouse. Tissue Antigens 1978;12:106-108.

25. Ali RR, Sarra G-M, Stephens C, et a]. Restoration of photoreceptor ultrastructure and function in retinal de-generation slow mice by gene therapy. Nat Genet 2000;25:306-310.

26. Downes SM, Firzke FW, Holder GE, er a]. Clinical features of codon 172 RDS macular dystrophy. ArchOphrhalmol J 999;117;1373-1383.

27. Ambasudhan R, Wang X, Jablonski MM, et al. Arrophic macular degeneration mutations in ELOVL4 re-sult in the intracellular misrouring of the protein. Genomics 2004;83:615-625.

28. Zhang K, Kniazeva M, Han M, er aL A 5-bp deletion in ELOVL4 is associated with two related forms ofautosomal dominant macular dystrophy. Nat Genet 2001;2789-2793.

29. Vine AK, Schatz H. Adult-onset foveomacular pigment epithelial dystrophy. Am ] Ophthalmol1980;8%80-691.

30. Battaglia Parodi M, Da Pozzo S, Ravalico G. Photodynamic therapy for choroidal neovascularisarion associ-ated with pattern dystrophy. Retina 2003;23: 171-176.

31. Zeldovich A, Beaumont P, Chang A, et a]. Indocyanine green angiographic interpretation of reticular dystro-phy of the retinal pigment epithelium complicated by choroidal neovascularisation. C1in Exp Ophthalmol2002;30;383-385.

32. Marmor MF, McNamara]A. Pattern dysrrophy of me retinal pigment epithelium and geographic atrophyof the macula. Am] Ophthalmol 1996; 122:382-392.

33. Prensky ]G, Bresnick GH. Butterfly macular dystrophy in four generations. Arch OphthalmolJ 983; 10];]198-1203.

34. Sjogren H. Dystrophia reticularis laminae pigmenrosae retinae: an earliet not described hereditary eye dis-ease. Acta Ophthalmol 1950:28:279-295.

35. Deutman AF, Rumke A1v1.Reticular dystrophy of the retinal pigment. Dystrophia reticulatis laminae pig-menrosa retinae ofH. Sjogren. Arch Ophthalmol 1969;82:4-9.

36. Mesker RP, Oosterhuis JA, Delleman )\Q. A retinal lesion resembling Sjogren's dystrophia reticularislamiiniae pigmenrosae retiniae. In: Perspectives in Ophthalmology, Vol. 2. Amsterdam: Excerpta MedicalFoundation, 1970:40---45.

37. Fishman GA, Woolf MB, Goldberg MF, et aL Reticular tapero-retinal dystrophy as a possible late stage ofSjogren's reticular dystrophy. Br] Ophthalmol 1976;60:35---40.


38. Slezak H, Hammer K. Fundus pulverentulus. Albrecht von Graefes Arch Klin Exp OphthaJmoI1969iI7E:

176-182. d . d h f h . al .39. O'Donnell FE, Schatz H, Reid P, er a]. Autosomal orrunanr yscrcp yo r e term p'gmem epithdilinL

Arch Ophthaimol 1979;97,680-683. ." .40. Boon C), van Schooneveld M], den Hollander Al, er a,!. Mutations In the pe:lphenn/RDSgeneareanim.

Pc'tw" cause of multifocai partern dystrophy simulating STC l/fundus flavimacuiann BrJ Ophtba1mt..2007;9U504-1511. . if d hv i

41. Hittner HM, Ferrell RE, Borda RP, er a]. Atypical virelli orm ystrop y In a 5 generation family. BrjOphthalmoI1984;6S,199-207. . . .

42. Epstein GA, Rabb MF. Adult virelliform macular degeneratlOn: diagnosis and natural history, BtlOphthalmoI1980;64,733-740. . .

43. Parodi MB, Iusruklin D, Russo D, er al. Adult-onset foveomacular virelli form dystrophy and indocyanimgreen videoangiography. Graefes Arch Clin Exp Ophrhalmol 1996;234:208-211.

44. Da Pozzo 5, Parodi MB, Toro L, er al. Occult choroidal neovascularisarion in adult-onset foveomaeuhIvitelliform dystrophy. Ophthalmologica 2001;215:412-414.

45, Pierro L, Tremolada G, Incroini U, er al. Optical coherence tomography findings in adult onset fOl'eomaarJar vitelliform dystrophy. Am J Ophthalmol2002; 134:675-680.

46. Hayami M, Decock CHR, Brabant P, er ai. Optical coherence tomography of adult onset vitelliformd}'S-trophy. Bull Soc Beige Ophtalmol 2003;289; 53-61.

47. Benhamou N, Souied EH, ZolfR, et a]. Adult-onset foveornacular virelliform dystrophy: astudybyopt~cal coherence tomography . .Am J Ophthalmol 2003; 135:362-367.

48. Wabbels B, Demmier A, Paunescu K, er al. Fundus autofluorescence in children and teenagerswithhe/t.ci.irary retinal diseases. Graefes Arch Clio Exp Ophrhalmol 2006;244:36-45.

49. Parodi MB, Iacono P, Pedio M, et 31.Autofluorescence in adult onset foveornacular vitelliform dysuoph)"Retina 2008;28:801-807.

50. Saito W, Yamamoto S, Hayashi M, er al. Morphological and functional analyses of adult onset vitelliformmacular dystrophy. Br J Ophthalmo12003;87:758-762.

51. Theischen M. Schilling H, Steinhorst UH. EOG in adult viteliiform macular degeneration, burrerflys.pattern dystrophy and Best disease. Ophrhalmologe 1997;94:230-233.

52. Aaberg TM, Han DP. Evaluation of phenotypic similarities between Srargardt flavimacularus and rerinslp~.rnenr epithelial dystrophies. Trans Am Ophthairnol Soc 1987;85: 101-119.

53. Gass ]0, Chuang EL, Cranek H. Acute exudative polymorphous virelliform maculopathy, TransAmOphthalmol SOt 1988;86,354-366.

54. Chan CK, Gass ]D, Lin SG. Acute exudative polymorphous vitelJiform mauculopathy syndrome.RetiD!2003;2H53--462.

PTER

Fundus Autofluorescence in Best Disease

a est disease was first repotted in 1905 as a disease affecting the macula (1). TheIE.I initial report was on eight members of one family from Giessen with mixed bi-lateral and unilateral disease presenting with a clearly demarcated macular le-

sion below the fovea of light red to yellow-white color, which Best described as com-pleted central "choroiditis." Various stages of the lesion were recognized as showingrectangular and crescent-shaped forms (1). These stages were later classified by Gass etal. (2). Currently, five stages are generally accepted (Fig. llF.l):

Previtelliform stage: No fundus abnormalities; mutation carriers are in this stageand will not be recognized if the family does not ptesent with a history of Best dis-ease (Fig. llF.lA,B).Vitelli form stage: Prominent, yellow or light red, well-demarcated central macular le-sion that, over time, increases in size to occupy the entire macular area; the lesion iscircular and the color is quite uniform (Fig. llF.lC,D). This stage was the primarystage described by Best (1).

• Pseudohypopyon stage: The yellowish material occupies predominantly the lowerhalf of the lesion (Fig. llF.lE,F) .

• Virelliruptive Ot "scrambled egg" stage: The lesion develops an elliptical shape inthe vertical axis that is described in association with a loss of visual acuity and dis-ruption of the uniform distribution of the yellowish material, which appears, at thisstage, to precipitate in the subretinal space and at the margins of the lesion (Fig.llF.lG,H).Fibrotic stage: Corresponds to the complete "choroiditis" described by Best (1) inwhich there is cicatrization of the macula and subrerinal fibrosis. A minimalamount of yellow material may still be visible within the scar (Fig. llF.lI,J).

The age of onset of the disease, the progression through the different stages, and thepenerrance of the symptOms is very variable among different families. Symptoms mayappear in the first decade of life or may never develop, as it occurs in carriers of thedisease. The disease may initially affect one eye only, with both eyes latet progressingindependently. Some patients may have lesions in the virelliforrn stage that remainunchanged for many yeats, whereas others may progress to the pseudohypopyon stagewithin a few months. This heterogeneity in the progression of the disease was origi-nally recognized by Best (1) and confirmed in several later reports, including our ownstudies (3).

MOLECULAR BASIS AND HISTOPATHOLOGYHuman Best l gene (hBESTI) codes for a protein (bestrophin) involved in transmem-brane transport in membranes of the retinal pigment epithelium (RPE). The gene prod-

197

FIGUREllF.l. Fundusphotographs (A,C,E,G,I) and fundus AF images (B,D,F,H,JI offivepati"tswiilivariousstages of Bestdisease. A clinical heterogeneity with regard to the age of onset and theprogressionof the disease is manifested in these photographs of a 39-year-old patient in the previtelliformstageIA,BI,a 6-year-oldpatient in the vitelliformstage (C,D). an S-year-old patient in the early pseudohypopionstage(E,F),a 16-year-oldpatient in the vitelliruptive stage (G,H). and a 17-year-old patient inthe fibroticandlinalstage of the disease (I,J).

uct is part of a Ca2+-activated cr channel (CACC). Bestrophin was considered robe

the channel itself (4), but its functional characteristics did not support the notion rh"bestrophin acts as a CJ- channel (5). Recent studies have solved the problem by show-_ing that bestrophin interacts wirh f3-subunits of the channel, influencing the acriviryo;the CACC, and thus explaining me functional characteristics of besrrophin (6),Histological studies locate bestrophin on the basolateral side of me RPE, and the co~.man notion was that me protein is part of me basolateral membrane (7). Expression'


1lg er in t e extramacu ar than in the macular RPE, as shown based on protein andRNA levels (8). The question of the functional impact and pathological effect remainsto be resolved. As a first hypothesis. Fischmeister and Hanzell (9) proposed thatbestrophin is involved in a cell volume-dependent current that decreases when the vol-ume of the RPE cell increases. Such an increase in volume may be caused by osmoticstress in the inrerphororecepror space, or follow phagocytotic activity when the RPEcleats the shaded outer segments at night. In this regard, phagocytosis may be hamperedby the imbalanced osmotic equilibration caused by improper besuophin function (9).

Best disease (VMD2) segregates in an autosomal dominant way. Reduced pene-trance has been associated with certain mutations, including c.969delTCA, a very fre-quent in-frame deletion (3).

Patients affected by Best disease present to the ophthalmologist with decreasedvisual acuity and color vision deficits, or ask for an appointment for genetic counsel-ing due to a positive family history. In carriers and in patients up to the vitelliformstage of the disease, the reduction of visual acuity may not be as prominent as thefundus changes would suggest. A rapid drop in visual acuity occurs when the patientprogresses through the vitelliruprive stage, although useful visual acuity may be re-tained into the fibrotic stage (3). Visual acuity depends on surviving photoreceptorcells and a preserved photoreceptor cell layer on which the image can be mapped.Any change in surface, receptor density, and receptor distribution will lead to faultyimage mapping. A faulty image mapping reduces visual acuity. The faulty mappingmay be tolerated to a certain extent by the adaptability of the visual system and maybe recognized by rhe patient in later stages of the disease only. Especially when pho-toreceptor loss becomes profound in the fibrotic stage or rearrangements of the pho-toreceptor layer take place in vitelliruptive srage, the visual acuity will drop (3,10).

The lesion described by Best was seen on fundus photographs to be located underthe fovea and within the retinal layers (1). Histological studies dating from the 1980sreponed on cases in rhe vitelliruptive (11) and fibroric (12) srages. These studies lo-calized the lesion to the level of rhe RPE. Weingeist et al. (12) described accumula-tion of vesicles in the RPE below the lesion that stained with lipophilic substances onlighr microscopy. These vesicles were distribured throughout rhe cells and identifiedas lipofuscin granules (11,13).

The histology of an eye from a patient in a lare stage of rhe disease and with aknown hBESTI mutation (T6R) supported the resulrs ofWeingeist et al. (12). The re-sults showed that accumulation of lipofuscin granules occurred peripheral to the scarregion and that, benearh rhe scar, the RPE was arrophic with only rare inclusions (8).Results from a 93-year-old patient with peripheral flecks carrying a Y227N murationcontradicted the findings in the T6R patient concerning lipofuscin accumulation. TheY227N mutation was associated with normal distribution of AF in the macular region.Mullins et al. (8) judged rhis as a variability of phenotypical expression. Lipofuscindensity was studied in purified intracellular granules in a further patient carrying a ho-mozygous missense mutation (W93C) in hBESTI compared with a heterozygous in-dividual carrying the T6R mutarion (8) and age-matched controls (13). The severityof the disease was no higher in the patient carrying the homozygous mutation. Bothpatients showed a reduction of rhe classical lipofuscin fraction oflight dense granules.A shift to denser lipofuscin granules was noted by fluorescence measurements of frac-rionated RPE granules (13). The denser granules were multilobed, indicating fusion oflipofuscin granules and thus impaired trafficking oflipofuscin granules within rhe RPEas a result of the dysfunction of besrrophin.

Some additional features were noted, such as a reduced number of melanosomesand an increased amount of secondary lysosomes. Mitochondria showed abnormal

___ -lPJp,edo-like shapes and small electron dense panicles were shown in rhe extracellular

CHAPTER 11F FUNDUS AUTOFLUORESCENCE IN BEST DISEASE

h d . fine fibrillar substance (12). Hypertrophy of rhe ER was"pon'"space enmes e In a . d f h ~. h II as loss of fenesrranon an occurrence 0 g ost vesselsin"III a furt er case, as we llIt

choriocapillaris (8).


IMAGING AND DIAGNOSTIC TECHNIQUES

B di . ular disorder wirh prominenr and characreristic featuresseenonest isease 15 a mac .fundus photography (see Introduction). The diagnosis of Best disease is usuallymad,by slit-lamp biomicroscopy, taking into. account visual acuiry, color vlSJon., and visualfield findings. Functional testing, especially elecrro-oculography (EOG), ISoftenob.rained to confirm the diagnosis.

Fluorescein and Indocyanine Green AngiographVAngiography is only rarely used to evaluate patients wirh Best disease. Since fluor""ioangiography (FA) displays the retinal vasculature, 1tS applicarion may be useful10 di>cern elder and late-stage Best disease patienrs from patients wirh adult vitelliformmac.ular dystrophy (AVMD) and age-related macular degeneration (AMD). The lanershow leakage of retinal vessels throughout the retina by FA, whereas in pariennwirhBest, the fluorescein leakage is restricted to the macular lesion. Few reportson theweof indocyanine green (ICG) angiography in patients with Best disease are available(14,15). Maruko et al. (14) found many hyperfluorescenr Spots in rhe peripheral mdmidperipheral retina on ICG angiography in patients with confirmed mutationsinhBESTI. The hyperfluorescenr spots in the periphery did not correspond to virelliforrnlesions resulting from lipofuscin accumulation. This was shown by AF of the peripb.eral vitelliform lesions. Maruko et al. (14) concluded rhat rhe diffuse hyperfluor"cemspots seen on ICG angiography were located on the RPE/Bruch's membrane levelanclwere associated with fibrillar and drusenoid material. Quaranta er al, (IS) focusedtheirfluorescein and ICG angiography on the macular lesion, showing the diffuse hyperllu.orescent spots also reporred by Maruko et al. (14). Quaranta er al, (15) did nor teporron any mutation in hBESTl in the patienrs presenred. Both groups showed that fluo.rescence in rhe macula did not occur on FA bur resulted from rhe AF inside the lesion(15). Finally, Pollack er aI. (16) reporred rhe FA findings in a case of Best diseasewithpositive mutation detection. In that report, hyperfluorescenr Spots were presemed atthe macula corresponding to light Spots of yeJlowish material in the macular lesion.FAis helpful in detecting the rare occurrence of choroidal neovasculanzation (CNY)inpatients wirh Best disease, especially in rhose in advanced stages of rhe disease.

Optical Coherence TomographyModern imaging techniques such as optic coherence tomography (OCT) imaginghave shown the lesion to be a hypo reflective structure underneath the neurorerina(Fig. IIF.2) (3,10). Studies reported to date have not provided OCT imaging witharesolution sufficienr to discern which retinal layer is actually involved. OCT imagesshow the lesion splitting a layer that contains photoreceptor outer segments RPE,and Bruch's membrane (outer retina-choroid complex [ORCC]). It is cu;renclythought that the split occurs between RPE and choroid in Bruch's membrane (10),This contradicts the reports by Mullins et al. (8), who located the split between RPEand neuroreuna, The hyporeflectlvlty of the lesion in OCT argues against a cellu1~nature o~ the content and poinrs toward a uniform refractive index, which is in accor-dance With a substance oflipophilic nature seen in histological sections (12).

CHAPTER11F FUNDUSAUTOFLUORESCENCEINBESTDISEASE

BA

FIGURE llF.2. IA) AFimagingofa controlprobandand IB) a patientinthe fibroticstage (see Fig.11F.1J).Thewhite arrowindicatesan area of lost RPEidentifiedbymissingAF;the blackarrowindicatesan areaof remainingRPEidentifiedby its AF.which is comparableto the AFin the macula;and the asteriskindi-catesan area of backgroundAF.which is comparableto normalAFintensity

Functional TestingMarked functional changes are seen on elecrro-oculography (EGG) in patients andcarriers. EGG allows evaluation of RPE function by testing the alternating potentialbetween the posterior pole and anterior segment of the eye during eye movements(I7,18). EGG derermines the maximal reduction of the standing potential of theRPE during a dark phase (dark trough) and the maximal increase of the standing po-tential of the RPE during a subsequent light phase (light peak). The ratio of the lightpeak vs. the dark rrough is called the Arden ratio and is considered abnormal when itis less than 2.0 (3). Since the standing potential is measured with electrodes posi-tioned at both canthus and a ground electrode at the forehead, the EGG measures thewhole retinal potential, and thus, if abnormal, implies functional disturbancesthroughout the retina.

Although a few patients and carriers with normal EGG in the early stages of thedisease have been reporred (3,16), an Arden ratio below 2.0 is the classical feature ofBest disease. Patients presenting with normal EGG were associated with hBESTl mu-tations showing reduced penerrance (3,16,19). Normal EGG recordings may be pres-ent up to the vitelliruptive stage (3,16).

Elecrrophysiological recordings localize the lesion to the RPE layer.Phororecepror-genetated signals and the subsequent responses from bipolar cells andother cells in the neuroretina as recorded by Ganzfeld-e1ectroretinography (ERG) andmultifocal (mf)ERG demonsrrare reduced responses and prolonged latencies starringin patients within the vitelliruptive stage when the lesion begins to disintegrate, butnot early on in the course of rhe disease (3). This suggests a secondary effect on neu-roretinal function that may result from besrrophin-mediated dysfunction of rhe RPE.Photoreceptor degenerarion as a result of RPE dysfunction is corroborated by the het-erogeneous results of visual acuity tests showing reduced visual acuity starting fromthe vitelliform stage in some patients but sustained normal visual acuity throughoutthe disease up to rhe vitelliruptive stage in other patients (16).

Fundus Autofluorescence ImagingThe distribution of fundus autofluorescence (AF) in the different stages of Besrdisease is shown in Figure IIF.!. In the previtelliform stage, no abnormality in thedisrribution of AF is detected (Fig. I1F.1B). In the vitelliform stage, however, thereis a marked and uniform increase in the AF signal at the site of the macular lesion


INTERPRETATION OF FUNDUSAUTOflUORESCENCE FINDINGS

hi h AF' 1 correlates well with the histoparhology findin" f(Fi IIF ID) The Ig sIgna ..,,0g. . . ".". . . I . hi the lesion, which was later Identified as lipofu",'increased lipophilic materia wtt In . '. In

(3 II 12). In the seudohypopyon stage, the lesion reaches Its maximal expansion,co~er;n the whol~ macula area (Fig. IIF.IF). Background AF ISunchangedou~id,

f h Ig d it seems to be lower than normal in the pseudohypopyon stage(Fig.ate eSlOn an d b h . " .IIF.IF). This decrease, however, is not real but cause y t e senSItivity adJustm'Di

f h dina i t ro limit rhe snong AF from the lesion, Smce thehpofu.o t e [eeor mg lilstrumen .' . .. I h . hisrological studies WIll be hardly resolved by rhe imaging""ern granu es s own In ". "J'"

b h . all' e a granular panern of AF IS less likely than a umformdis-tern ecause t elf SID SlZ,iburi f i d AF Of interest no increased AF can be derecred outsideth,rn unon 0 mcrease. ,

macular lesion (3). The AF signal remains very high ar the sire of the lesion, especiallyinferiorly, where the remains of the yellowish mater:,al are predommamly deposito!(Fig. IIF.IF). In the vitelliruptive or "scrambled egg srage, rernains of rhe yellowishmaterial are found at rhe margins of the lesion and inferiorly (FIg. IIF.IH).

Although AF changes should be expected from. the histological data rhrougho",the retina, since hBestl is expressed in all RPE cells, it IS Interesung thar fundusAFinpatients with Best disease is not increased or decreased outside the macularlesion(Fig. IIF.2).

It was assumed that Best disease is caused by a generalized defecr on the RPE williwidespread accumulation of lipofuscin material in RPE cells, and thar the hBESTJgene is expressed in all RPE cells. Therefore, ir was unclear why areas of increasedAfshould be restricred to the center of the macula. A recent repon by Mullins et al.(81solved this dilemma. The authors showed less intense immunolabeling ofbestrophinin the macula, indicating reduced expression of hBESTl compared to the regionsoutside of the macula. Given a function of besrrophin in phagocyrosis, as indicatedby Fischmeister and Hartzell (9), the accumularion oflipofuscin material in themac-ula area could be explained by the reduced levels of besrrophin expression and phago-cytosis at that site. In this regard, the yellowish material may be considered asshedbur nor phagocyrosed photoreceptor outer segments. AF outside of the lesionisin-conspicuous, as shown by several authors (20-22), and may be understood asa resultof haploinsufficieney restricted to the macula as the area presenting with the highesidensity of photo receptors and lowest concentration of bestrophin. In the periphery,minimal bur sufficient residual function of besnophin results in sufficient phagocy·totic activity in patients affected by mutations in hBESTl (8).

Further reports have supponed the notion that the peripheral RPE may norbeedependent on bestrophin function as the macular RPE. Bakall et al. (13) presented'patient homozygous for a missense mutation of uncertain functionality (W93C),andSchatz et al: (23) reported on a family with two patients carrying compound heterozy-gous rnuranons m hBESTl, an obvious null mutation (T29X), and another muranonthat may provide residual function (RI4IH). Family members carrying each of the;<rnurauons 111 the heterozygous state presented with Best disease that was lessseverethan in the compound heterozygous patients (23). Both reports provide data toevluate a maxlmum loss of functionality of bestrophin. From the histological darapro'vided by these reports, at least a normal AF can be expected (13), which arguesforagreat adaptabIlity In compensating for bestrophin dysfunction in the peripheral RPE.Unfortunately, none of these repons included AF imaging (13,23).

CHAPTER l1F FUNDUS AUTOFLUORESCENCEIN BESTDISEASE 2113

The disrriburion of fundus AF abnormalities in Best disease indicates that themutant besrrophin affects the fate of lipofuscin or its precursors within the RPEcells, as well as the fare of the RPE cells themselves. Both may well be due toimpaired phagocytosis or storage of lipofuscin within the RPE cells, as indicated byFischmeisrer and Hartzell (9). Degeneration will subsequently occur and the de-struction of the macula will ptogress until the final fibrotic stage is completed. Thequestion as ro why photoreceptor cells degenerate may be at least partially answeredby AF and OCT findings. The presence of spots of increased AF throughout themacular lesion and mostly along its margins from the vitelliruptive stage onward in-dicates residual AF material in regions of the lesion that did not reattach. This no-tion is in accordance with OCT findings that show disorganized RPE and neu-roretina and remaining hyporeflective areas that correlate with the areas of increasedAF (3,10). Thus the degeneration of the phororeceprors occurs by rhe missing reat-tachment of the split tissues, and therefore from a reduced support of oxygen andnutrients and reduced waste disposal from and to the choroid, respectively.

It remains unclear how mutations in hBESTl affect the RPE to produce the yel-lowish material in the macular lesion. The highly increased AF signal inside the mac-ular lesion observed in the vitelliform stage decreases centrally to levels similar tothose of the background in the vitelliruptive stage (Fig. IIF.IH). The fact that the AFsignal inside the lesion from the vitelliruptive stage onward returns to normal or near-normal background levels indicates that the yellowish material is not inside the RPEcell layer but is extracellular (Fig. IIF.3). If the fluorescent material present in thevitelli form stage were inside RPE cells, the AF signal should not decrease in the vitel-liruptive stage as long as RPE cells were present below the macular lesion. Also, theintensity of the AF signal supports the concept that the RPE does not accumulatemore fluorescent material in the macula (outside the central lesion) and peripherythan RPE in unaffected individuals (Fig. lIF.3). In the virelliruptive stage, the areaof the lesion that is not filled with yellowish material shows AF intensity comparableto that of the surrounding fundus (Fig. IIF.1 G,H). In the same area, reduced AF in-tensity occurs later in the fibrotic stage (Figs. IIF.lI,J and IIF.3). This notion is sup-ported by Mullins er al. (8), who argued that the yellowish material is located in thesubretinal space, which would support the notion that Best disease is caused by

FIGURE llF.3. OCTimaging at a 13-year-old patient with Best disease. The right eye demonstrates a lesion IAI. whereasthe left eye is in the previtelliform stage 181 NFL. nerve tiber layer; GCl. ganglion cell layer; ONl. outer nuclear layer; ORCC.outer retina-choroid complex.

, " f the RPE, From the histological data rer,d d phagocytotIC acnvity 0 "oned bre uce , () d M Ilins er al. (8), RPE degeneration rnusr be ex, yWelllgelSt er a]. 12 an u I 1 I'd ,eqed'

C d d AF inside the maeu ar eSIOn compare to back Inlare stages, Therefore, re uce 1 ' I (F' gtoUnd' f he lesi indicates loss ofRPE eel s 10 ater stages 19,lIP))levels outside 0 t e estort In 1 . . "II ' ,

C f h fl nt macerial inside the lesion ISstJ unresolved, A, lThe late 0 t e uoresce ". , SliOw

(3) d h rs (1024) the neurorenna overlying the leslOn"n ,nby our group an ot e " ,Ot~,d P' 1 (10) showed some OCT scans that allow the Interpret"io frupre. ianta et a. . d no a

di , f th basolateral membrane of the RPE, which woul aJlow leakingof Ilsrupnon 0 e f '. [nefl ial i to the choroid (24), However, final proof 0 this Interpret'uorescent maten III ". anon' 'II I ki D dation of the AF material filling the lesion has not yetbIS Sf I ac mg. egra eenshown in fundus phorographs and AF images,


DIFFERENTIAL DIAGNOSISAt the virelliforrn stage, the differentiaJ diagnosis includes AVMD (see also Ch'pterlIE) and vitelJiform detachments occurring in patients with basaJ laminar drllSen,Theformer can be distinguished from Best disease by the later age of onset and thelackofprogression through stages of the macular lesion; rhe latter is discerned by the lackofdrusen in Best disease.

SUMMARYFundus AF imaging in Best disease can be used to provide immediate suppOrt to theclinical diagnosis made by slit-lamp biomicroscopy, given that, in most cases,mechanges observed clinically, including color, size, form, and structure of the macularlesion, are characteristic of the disease. The presence of a positive family histol)' willbe helpful in confirming the diagnosis, EOG is usually used to Support the diagnosisof Best disease, which is then confirmed by molecular genetic testing.

AF can assist in differentiating late stages of Best disease from other maru-Ioparhies that may cause fibrosis of the macula, AF findings correlate well with hisro-logical data and OCT findings, The presence of areas of reduced AF in advancedstages of the disease is an important factor in estimating remaining RPE cellsbelowthe lesion, However, to date, fundus AF, as well as all orher ancillary studies, doesnotappear to provide information regarding the prognosis for progression of the diseasein an individual patient.

ACKNOWLEDGMENTS

Many thanks go to Prof, Birgir Lorenz, head of the Deparrmenr of Ophthalmology srthe Medical Faculty of the Justus-Liebig University, Giessen, who provided the ima,,"sshown in this chapter.

REFERENCES1. Best F. Ober eine hereditare .Ma~ulaaffektion. Z Augenheik 1905; 13: 199-212.2. Gass 1DM. Heredodystropinc disorders affecting the pigmenr 0 irh I' d reri I G JDM, ,d

. . '-- '--pI e turn an renna. n: assStereoscopic Atlas of Macular Disease-Diagnosis and Treatment. St. Louis: Mosby, 1997:303-43

CHAPTER 11F fUNDUS AUTOflUORESCENCE IN BEST DISEASE

3. Wabbels B, p-eeing MN, Krerscbmanc U, et al. Cenorype-phenorype correlation and longitudinalcourse in ren families with Best virelliform macular dystrophy. Graefes Arch Clin Exp Ophrbaunol2006;244; 1453-1466.

4. Sun H, Tsunenari T, Yau KW, et al. The virelliform macular dystrophy protein defines a new family ofchloride channels. Proc Nat! Acad Sci USA 2002;99:4008-4013.

5. Pusch M. Ca2+ -activated chloride channels go molecular. J Gen Physiol 2004; 123:323-325.6. Strauss 0, Milenkovic VM, Srriessnig J, et aI. Direct interaction of Bestrophin-l and beta-subunits of volt-

age-dependent calcium channels. ARVO Absu 2008;49:5182.7. Marmorstein AD, Marmorstein LY, Wang X, er al. Bescrcphin, the protein encoded by the best macular

dystrophy gene (VMD2), is a component of the plasma membrane. Invest Ophrhalmol Vis Sci2000;4;5398.

8. Mullins RF, Kuehn MH, Faidley EA, er al. Differential macular and peripheral expression of besrrophinin human eyes and its implication for best disease. Invest Ophthalmol Vis Sci 2007;48: 3372-3380.

9. Pischmeisrer R, Hartzell He. Volume sensitivity of the bestrophin family of chloride channels.] Physiol2005;562(P, 2);477-491.

10. Pianra M], Aleman TS, Cideciyan AV, et al. In vivo micro pathology of Best macular dystrophy withoptical coherence tomography. Exp Eye Res 2003;76:203-211.

II. O'Gorman $, Flaherty WA, Fishman GA, er aL Histopathologic findings in Best's vitelliform maculardystrophy. Arch Ophrhalmol 1988; 106: 1261-1268.

12. Weingeist TA, Kobrin JL, Wanke RC. Histopathology of Best's macular dystrophy. Arch Ophchalmol1982;100:1108-1114.

13. Bakall B, Radu RA, Stanton JB, et al. Enhanced accumulation of AlE in individuals homozygous or het-erozygous for mutations in BESTI (VMD2). Exp Eye Res 2007;85:1;34-43.

14. Maruko I, Iida T, Speide RE, et al. Indocyanine green angiography abnormality of the periphery in vitel-liform macular dystrophy. Am] Ophthalmol 2006; 141 :976-978.

15. Quaranta M, Buglione M, Lo Schiavo ER, er al. Angiographie au vert d'indocyanine des drusen de la mem-brane basale de I'epithelium pigmentaire retinien associes a du materiel pseudo-vicelli forme. J Fr Ophralmol1998;2U 85-190.

16. Pollack K, Kreuz FR, Pillunar LE. Morbus Best mit normalem EOG-Fallvorstellung einer [amiliarcnMakuladyscrophie. Der Ophrhalmologe 2005;102:891-894.

17. Marmor MF. Standardization notice: EOG standard reapproved. Electro-oculogram. Doc Ophrhalmol1998;95;91-92.

18. Marmor MF, Zrenner E. Standard for clinical electro-oculography. International Society for ClinicalElectrophysiology of Vision. Arch Ophrhalmol1993ill1: 601-604.

19. Kramer F, White K, Pauleikhoff D, er al. Mutations in the VMD2 gene are associated with juvenile-onsetvirelliform macular dystrophy (Best disease) and adult vitelliform macular dystrophy but not age-relatedmacular degenerarion. Eur J Hum Genet 2000;8:286-292.

20. Spaide R. Autofluorescence from the outer retina and subretinal space: hypothesis and review. Retina2008;28;5-35.

21. Spaide RF, Noble K, Morgan A, et al. Virelliform macular dystrophy. Ophthalmology 2006;113:1392-1400.

22. Wabbels B, Demmler A, Paunescu K, er al. Fundus autofluorescence in children and teenagers with hered-itary retinal diseases. Graefes Arch Clin Exp Ophrhalmol 2006;244:36-45.

23. Schatz P, Klar J, An:dreasson S, er al. Variant phenotype of Best vitelliform macular dystrophy associatedwith compound heterozygous mutations in VMD2. Ophthalmic Genet 2006;27:51-56.

24. Aleman T, Stone EM, Hernandez R, et al. Ophthalmologic findings in a retinitis pigmenrosa family withrhodopsin gene mutation Asp-190-Asn. Invest Ophrhalmol Vis Sci 1996;37;S667.

CHAPTER

O rargardr disease (STGD) (I) (also termed fundus flavimaculatus [2])isiliemost common recessively inherited macular dystrophy: affecting appro".mately one person in 10,000 (3). STGD can affect individuals of any gender

and race (1,4-8) and there is wide variability in age of onset, visual acuity, fundusap.pearance, and severity of the disease (1,5-10). Visual acuity may vary between20120to 20/400; rarely will it drop below 20/400 (7,11). Parienrs wid, STGD maybeasymptomatic or complain of visual acuity loss, photophobia, and, less commonly,nyctalopia (5).

Fundus examination may be normal in early stages of the disease or revealre[in~pigment epithelium (RPE) motding or a buil's-eye appearance at the macula, andactive (deposition of yellow material at the level of the RPE) and/or resorbed (RPEde.pigmentation/atrophy) flecks and atrophy at the macula and rnidperiphenl retim(1,5,6,9,12). Characteristically, the flecks have a pisciform ("fish-like") appearance,but they can also be round, like dots, and appear either as individual lesions orjoinedtogether (6,13). Different clinical classifications of STGD have been proposed basedon the presence or absence and distribution of the fundus lesions (6,9,12); however,none of these have been Widely accepted.

Electrophysiology testing in patients wid, STGD may demonstrate maculard}~function alone or macular and peripheral cone Ot cone and rod dysfunction (10,14).These patterns of functional loss cannot be predicted by the fundus appearance (10,14).Mutations in the ABCA4 gene, located in the shorr arm of chromosome l, arerespon-sible for all cases ofSTGD (15,16).

Currendy, there is no treatment available [or patients with STGD. However, lab-oratory studies suggest that progression of the disease may be slowed by protectingthe eyes from light exposure (17). Additionally, new treatment strategies ro reduceorprevent AlE accumulation in the RPE are also being investigated (see also Chapres2 and 4) (Fig. llG.1) (17-20).

•Fundus Autofluorescence InStargardt Disease

MOLECULAR BASIS AND PATHOLOGYThe mechanisms by which phototeceptors degenetate in STG D are not completelyunderstood. Recent laboratory studies investigating the function of the ABCA4pro.tein, as well as studies conducted in the ABC44 knockout mice, an animal modelofthe disease, have shed light on the molecular basis of STGD. The evidence sugges<sthat the ABCA4 protein facilitates the transport of retinoids, preferentially N-retinyli.dene-phosphatldylethanolarnme (N-retinylidene-PE) and all-trans-tetinal (21,22),from the cytoplasmic side of the photoreceptor disc membrane to the cytosolic side,making them accessible to all-trans-retino!-dehydrogenase and facilitating its conver-

CHAPTER11G FUNDUS AUTOFLUORESCENCEIN MATERNAL INHERITED DIABETES AND DEAFNESS

Disc cytoplasm

a Photoreceptor Outer segment

11..{;is.RAL..- cis-ROl ....- all.tfans_ROl

:[email protected] Pigment Epithelium

b Photoreceptor Outer Segment

Cytosol Disc cytoplasm

A2E

Retinal Pigment Epithelium

FIGURE llG.l. (A) AII-trans-retinal (all-trans-RAL) can react with phosphatidylethanolamine (PEl andform N-retinylidene-PE. Free all-trans-RAl and all-trans-RAl contained in N-retinylidene-PE are reduced toall-trans-retinol [all-trans-ROl) by the all-trans-retinol dehydrogenase (tRDH). Evidence suggests thatABCA4 transports N-retinylidene-PE and all-trans-RAl from the cytoplasmic side of the disc membrane tothe cytosolic side, where they are reduced to all-trans-ROl by the all-trans-retinol DHase (trDH). (8) In pa-tients with STGD, there is an impaired transport of all-trans-RAl and N-retinylidene-PE, with the subse-quent accumulation of both molecules in the photoreceptor outer segment disc membrane. Condensationof all-trans-RAl and PE gives rise to N-retinylidene-N-retinyl-ethanolamine [A2E). A2E, the major fluo-rophore of lipofuscin, then accumulates in the RPE after photoreceptor outer segment disc shedding(modified from Refs. 2B and 57). CRDH,Cis-retinol DHase.


. . . 1(2223) (Fig. llG.IA,B). This assists in the recoveryof~sion Into all-trans-renno , h ll noi eIJ c IJ . light exposure, reduces p ororecepror ce noise (thel~~

Photoreceptor ce 10 OWIng ich h bi d . ,. . all - ti al and opsin, whi w en com me can aCUvarethe'i.of an Increase In -trans re In ulati f alJ .

. d ) d diminishes the accum anon 0 -tram-reunal andN.sual transduction casca e , an . .. lid PE' h' th disc membranes (22). The latter In turn mcreasestheprodur.fenny J ene- WI t In e . th· fl

tion of N_retinylidene-N-retinyl-ethanolamIne (AlE), . e maJor. uoraphoreof

J. f . . h RPE ( also Chapter 2) which has potential cytotOJuceffec~onRPEIpa uscm, III t e see , . .celJs (24-27). RPE damage/loss is then followed by photoreceptor cell degeneraUonand

loss of vision (33). . .H· holozi al Iuarion of eyes from pauents with STGD have shownRPElstopat 0 ogle eva £1

lJ d 1 k d ith a substance with ultrastructural, autottuorescenf andhist~ce s ense y pac e WI . . .

chemical characteristics consistent with lipofuscin In both the macula and penphetal. (29 30) 0 Jy one histopathology study failed to detect increased lipofwcinin

reuna , . n bretinalthe RPE in a case of STGD without rnaculoparhy (31). Su renn desquamaredRPEcelJs macrophages engorged with melanolipofuscin in the outer retina, RPE andchoriocapillaris atrophy, and phororecepror-cell Ioss at the fovea have been alsoDb.served (29,30).

DIAGNOSTIC TECHNIQUESFundus flecks are the hallmark of STGD. Although "active" flecks are often seenbyslit-lamp biomicroscopy or indirect ophthalmoscopy (5,9,32), "resorbed" flecks aremore difficult to visualize and can be missed by the examining ophthalmologist. Inthe latter case, the diagnosis of STGD may be difficult.

Fluorescein and Indocyanine Green AngiographyFluorescein angiography (FA) is only rarely required for rhe diagnosis or evaluationofpatients with STGD. In both early and late frames of FA, "active" flecks appear hypD-fluorescent (5). "Resorbed" flecks may appear either hypo- (5) or hyperfluorescenc(9),Areas of overt macular atrophy are visualized as areas in which no choriocapillaris ispresent but large choroidal vessels are seen. Patients with STGD may have a "darkchoroid" or "choroidal silence" sign (33), characterized by a lack of early hyperfluore-cence coming from the choroid, such rhar rhe retinal blood vessels, even the smallcap-illaries, are easily seen over a very dark background where rhere is no choroidal fluo-rescence. Not all patients with STGD will demonstrare a dark choroid. In a recentstudy, only 62% of patients with STGD had this FA sign (7). Similarly, the darkchoroid is not a specific sign of STGD; it has also been observed in patients with coneand cone-rod dystrophy (see also Chapter lIB) (7,33,34). However, if ptesenc, thissign may be useful in the differential diagnosis ofSTGD (see below), especiallywhenthe diagnosis of multifocal pattern dystrophy simulating STGD is entertained (seealso Chapter I IE). Of interest, the overall increase in the fundus autofluorescence(AF) signal observed in patients with STGD (see below) seems ro be independent ofthe presence or absence of a dark choroid (35).

Indocyanine green (ICG) angiography allows the choroidal details to be seenevenin patients with a dark choroid (36). It is also possible to detect choroidal vascularclo-sure, such as mat present in patients with atrophic macular lesions. Active fundusflecksappear hypo fluorescent on ICG and are typically best detected in lare frames of thean-giograrn (36).

CHAPTER 11G FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS

Optical Coherence TomographyThe clinicalusefulnessof optical coherence tomography (OCT) in STGD remains tobe elucidated, Earlier studies using OCT concentrated on evaluating the location ofretinal flecks (37,38). In time-domain OCT imaging (Stratus OCT 300; Zeiss,Germany) fundus flecks are seen as hyperreflecrive deposits, either as dome-shapedlesionsat the levelof the RPE or jusr above me RPE, or as small, linear, highly reflectivelesions at the level of the photoreceptor inner segments or outer nuclear layer (37).Fourier-domain OCT imaging (University of California-Davis, prototype), however,revealsonly well-demarcated oval "bumps" within the RPE in all casesof STGD (38).

Recently, Querques et a1.(39) found that the two types of highly reflective dome-shaped deposits described above did nor correlate with foveal thinning or best cor-rected visual acuity. Ergun et a1. (40) used ultra-high-resolution OCT to assesstrans-verse photoreceptor cell loss and compared the results with visual acuity and changesdetected on AF and FA imaging. They found that a lower visual acuity correspondedto a greater transverse photoreceptor cell loss, which also correlated with the extent ofreduced AF (transverse diameter) and atrophy seen on FA. OCT rhus may providevaluable structural information in patients with STGD.

ElectrophysiologyElectrophysiology alone cannot be used to establish the diagnosis of STGO. However,it is essential for gathering information on the location and extent of retinal dysfunc-tion in patients with this disease.

The degree of functional loss can be assessed by using the pattern electroretino-gram (PERG) and the full-field electroretinogram (ERG) (5,6,9,12,13,41-43).Patients may demonstrate macular dysfunction alone (abnormal PERG with normalfull-field ERG), macular and peripheral cone dysfunction (abnormal PERG and pho-topic ERG responses), or macular and peripheral cone and rod dysfunction (abnormalPERG and scotopic and photopic ERG responses) (10). It is imporrant to note matthese patterns of functional loss cannot be predicted by the fundus appearance (10).There seems to be a high degree of intrafamilial homogeneity with respect to the pat-tern of functional loss present as determined by electrophysiology (9,42).

Electrophysiology is a valuable prognostic tool mat can help the clinician to iden-tify, early on in the course of the disease, those patients with peripheral cone and rod in-volvement who will likely have a higher chance of developing not only central but alsoperipheral visual loss and a more severe form of the disease.

Fundus AutofluorescenceTo date, fundus autofluorescence (AF) seems to be the most effective clinical adjunctfor the diagnosis and evaluation of patients with STGO (10,42,44). On fundus AFimaging, both "active" and "resorbed" flecks and areas of outer retinal atrophy can beeasily identified (Fig. 11G.2) (44). Active flecks appear as foci of high AF signal (Fig.IIG.2B), indicating an increased lipofuscin content at the site of the fleck. In con-trast, resorbed flecks are seen as foci oflow AF signal on AF imaging (Fig. IIG.2B)(42,44). Given that resorbed flecks seem to occur most commonly at sites previouslyoccupied by active flecks, the low AF signal observed on AF imaging could representdamaged/lost RPE, probably as a direct or indirect result of the previously increasedlipofuscin content in the RPE at the site of the fleck (44). Both resorbed and active


FIGURE llG.2. Color fundus photograph (A,D). fundus AF image (B,E), and OCT image (CI from two patients withSTGOshowing active (A-C. black arrows) and resorbed (A,B, circles) flecks. Active flecks appear on slit-lamp examina-tion as white-yellowish lesions formed by accumulation of material in the outer retina lA, black arrows). OnAF images(B. black arrows) the active flecks are seen as foci of increased AF signal. Resorbed flecks IB, circlesl. which appear assmall areas of depigmentation in the RPE.are difficult to detect on slit-lamp biomicroscopy but are easily visualized onAF images as foci of low AF signal lB. circlesl. FundusAF is helpful in demonstrating peripapillary sparing (B, asterisks;Fig. lIG.3C). Active flecks are seen on OCT imaging as dome-shaped bumps at the level of RPE IC. black arrow,90 degree section).Well-defined areas of low AF signal IE, white arrowl corresponding to areas of clinically detectable10. white arrowl or undetectable IA. black arrows) retinal atrophy can be seen at the macula and midperipheral retina.

CHAPTER nG FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS

flecks can be seen either confined to the macula or distributed throughout and be-youd the posterior pole (10,28,35,42,44,45) with a particular predominance nasally(45). A normal retinal sensitivity, as determined by fundus micro perimetry, has beendetected over areas of increased AF corresponding to active flecks. In contrast, re-duced retinal sensitivity has been found over areas of reduced AF signal that corre-sponded to patches of atrophy observed clinically (46).

When resorbed flecks are present in patients with advanced disease and atrophicmacular lesions but no active flecks, the diagnosis of STGD may represent a challengeto the clinician. In these circumstances, AF imaging can be extremely useful bydemonstrating multiple, small foci of decreased AF signal at the macula or midperiph-eral retina corresponding to resolved flecks. It has also been noted that in STGD thereis a typically relative peripapillary sparing (lack of flecks and atrophy around the opticnerve head), even in cases with diffuse RPE abnormalities and atrophy (Figs. II G.2Band IIG.E,F) (36,44,47,48). This pteservation of the retinal tissue around the opticnerve is easily appreciated on fundus AF imaging and is a very useful sign when estab-lishing the diagnosis in patients with advanced disease in whom fundus flecks are nolonger visible (44).

Well-defined areas of low AF signal corresponding to areas of clinically detectableor undetectable retinal atrophy are typically seen at the macula but may also be presentin some patients in the midperipheral retina (Fig. II G.2D,E) (42,44). The presence ofmultiple well-defined areas of low AF signal in the midperipheral retina has been ob-served only in patients with reduced macular and peripheral cone and rod function(10), and thus seems to indicate a poorer prognosis. Furthermore, a diffuse very highAF signal can be detected in some patients throughout the macula, where no funduschanges are evident on slit-lamp biomicroscopy. This finding seems to indicate a fasterspeed of progression of atrophy and a pooret prognosis (see below).

Quantitative evaluations of fundus AF have demonstrated high levels of AF inthe majority of patients with this disease (35,44,49,50), independently of whethera dark choroid was present on FA (35). More recently, fundus AF levels across themacula in patients with STGO were found to be high, normal, or even low com-pared to those in age-matched normal volunteers (44). Furthermore, there seemsto be a relation between levels of AF across the macula and the peripheral retinalfunction, as demonstrated by full-field ERG. Low levels of AF across the macula,including the fovea, were detected in all patients with peripheral cone and rod dys-function (44). Although quantitative evaluation of AF levels appears to be impor-tant, since in many cases clinicians cannot predict high or low levels of AF simplyby looking at the AF images (49); it appears that an accurate quantitative evalua-tion of AF levels cannot be achieved with current commercially available instru-ments (see also Chapters 5 and 8).

Fundus AF imaging is very useful for monitoring the progression of the disease.Serial imaging can demonstrate the development of new foci of increased AF signal(active flecks) over time, even in patients with long-standing disease (Fig.IIG.3A-D). The appearance of new active flecks tends ro follow a centrifugal pat-tern. In contrast, resorbed flecks tend to appear more centrally and with a smallerdispersion radius than active flecks (51). Enlargement of preexisting areas of atro-phy can be also documented (Fig. IIG.3A-H). Areas of atrophy seem to expanduniformly, with no quadran tic preference (51), and more often toward areas withpreviously increased AF signal, suggesting that, as in AMD (53-55) (see alsoChapter 10C), increased AF and thus lipofuscin may presage directly or indirectlyphororecepror-RPE cell demise. Long-term follow-up AF data from our group (un-published results) suggest that the development of new areas of low AF signal


FIGURE llG.3. Serial fundus AF im-ages obtained from four patients withSTGOat baseline (A,C,E,G) and at fol-lnw-up (B,D,F,H; A,B, C,D, U: 3-yearfollow-up; G,H: 2-year follow-up]. AF im-aging demonstrates development of newfoci of increased AF signal (active flecks,white arrowsl and the disappearance offoci of increased AF (black arrowsl overtime lA-D). The development of newareas of atrophy and the enlargement ofexisting ones can also be documentedlA-HI. The speed of enlargement of pre-existing areas of low AF signal may bedetermined by the pattern of fundus AFobserved. When homogeneous back-ground AF was detected surroundingareas of low AF signal (atrophvl a slowrate of enlargement was documented[A,B, 55-year-old male, age at onset ~19years, duration of disease ~ 34 years,3-year follow-up, rate of progression LE= 1.13 rnrn'/year; C,D, 43-year-old fe-male, age at onset = 11 years, durationof disease = 32 years, 3-year follow-up,rate of progression LE= 0.51 mm'/yearlin contrast when a widespread patternof increased backgroundAF signal inter-spersedwith foci of low and high AF sig-nals was identified, multiple new areasof low AF signal (atrophy) and/or a rapidenlargement of preexisting areas of lowAFsignal were detected (E,F, 41-year-oldfemale, age at onset ~ 37 years, dura-tion of disease ~ 4 years, 3-year follow-up, rate of progression 5.79 mm'/year;G,H, 40-year-old male, age at onset 31years, duration of disease = 10 years,AF 2-year follow-up, rate of progressionLE= 4.37 rnm'/year)

CHAPTER l1G FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS 213

(atrophy) and the speed of enlargement of preexisting areas of low AF signal may bedetermined by the pattern of fundus AF observed. Thus, when homogeneous back-ground AF was detected surrounding areas of low AF signal (atrophy) a slow rate ofenlargement was documented (Fig. 11G.3A-D). In contrast, when a widespreadpattern of increased background AF signal intersperse wi th foci of low and high AFsignals was identified, multiple new areas of low AF signal (atrophy) developedand/or a rapid enlargement of preexisting areas of low AF signal was detected (Fig.IIG.3E-H).

Although to date there is no available treatment for patients with STGD, phar-macological strategies aimed at reducing the synthesis and accumulation of A2E andother retinoids are under investigation 07-20). Isotretinoin (l3-cis-retinoic acid orAccutane, a drug commonly used to treat patients with acne) was found to biochem-ically suppress rhe accumulation of AlE in the RPE, and also inhibited the accumu-lation of lipofuscin granules in the RPE as detected by electron microscopy in the ro-dent model ofSTGD (abet-(- mice) (Fig. 11G.IA) (see also Chapter 4) (19). Futureclinical trials with these or other agents would need to identify objective outcomemeasures to evaluate the response to these treatments. Electrophysiology testing is un-likely to be useful for this purpose, since most patients with STGD have normal full-field ERG responses and a flat, unrecordable PERG (10). Furthermore, it is expectedthat long follow-up and a high number of patients will be required to detect statisti-cally significant variations in electrophysiology recordings. Under these circum-stances, fundus AF images would provide objecrive and measurable data on the devel-opment of new areas of increased or decreased AF at the macula and midperipheralretina and/or on the speed of enlargement of preexisting ones (see Fig. 11G.3).

It is possible that the new imaging technique of near-infrared autofluorescence(NIA) will also be very helpful in the evaluation of patients with STGD (seeChapter 6).

DIFFERENTIAL DIAGNOSIS OF STGDThe differential diagnosis of STGD includes autosomal dominant Stargardr-likemacular dystrophy, Best disease, cone and cone-rod dystrophy, central areolarchoroidal dystrophy, age-relared maculoparhy (ARM)/age-related macular degenera-tion (AMD), pattern dystrophy, and retinitis pigmenrosa (RP). A detailed family his-tory should be obtained and can be helpful in establishing this differentiation. Thedominant pattern of inheritance in Stargardr-like disease can help distinguish it fromthe recessive mode of inheritance in patients wirh STGD. As in STGD, in advancedBest disease (see also Chapter liP), cone dystrophy (see also Chapter liB), centralareolar choroidal dystrophy, and AMD (see also Chapter 10), a cenrral area of arro-phy at the macula is often seen. However, in contrast to STGD, active and resolvedflecks are not present in these retinal diseases. Drusen in patients with ARM may sim-ulate the round flecks observed in some patients with STGD; fundus AF can help dif-ferentiate between the two by demonstrating a very high AF signal at the site of theround flecks and no abnormaliry or a mildly increased Ot decreased AF signal in thecase of drusen. Like patients wirh STGD, patients with multifocal pattern dystrophysimulating fundus flavimaculatus (56) present with multiple pisciform lesions at themacula and midpetipheral retina (see also Chapter 11E). In both retinal diseases,these lesions will demonsrrate a high AF signal on AF imaging. However, the modeof inheritance (dominant in pattern dystrophy) and the absence of peripapillary spar-ing will assist the differential diagnosis. Furthermore, most patients with STGD willdemonsrrate a flat PERG, which only very rarely will be obtained in patients with

SECTIONII CLINICALSCIENCE

d h S corms of STGD can simulare cone-rod dystrophy (se, aIpattern ystrop y. everc 11 .' soChapter lIB) and RP (see also Chapter IIA): Multiple foci of decreasedAF signal"

h I d id . heral retina indlCatlllg the presence of resorbed flecks ",-llt e macu a an ill! penp' . )b b d . . irh STGD but not in those with cone-rod dystrophy Or RPe 0 serve In patIents WI ,

SUMMARYFundus AF imaging provides a rapid and noninvasive way [~ evaluate patientswithSTGD. AF imaging allows a clear visualization of both acnve and resorbed flecks,areas of macular and midperipheral atrophy, and the presence of the relanve peripap.illary sparing, helping to establish the diagnosis of the disease: AF imaging is excep-tionally helpful in the diagnosis of patients with advanced disease, In whom activefundus flecks are no longer visible and only resolved flecks are present. FundusAFimaging may have also a prognostic value by demonstrating areas of low AF signalinthe rnidperipheral retina in patients with the most severe form of the disease,andwidespread increased AF signal at the macula interspersed with foci of low andhighAF signals in patients with a fastet rare of atrophy progression. It is likely that fundmAF imaging will become the imaging technique of choice to evaluate the responserofuture treatments for patients with STGD.

REFERENCES1. Prancescherri A. Ueber rapero-rerinale Degenerationen im Kindesalrer (Kongenirale Form (Leber),amauro-rische Idiorie, rezessive-geschlechtsgebundene rapero-rerinale Degenerationen, Fundus albipuncrans ClllllHemeralopia, Fundus flavimacularus). Drirrer Fortbildungskurs der Deurschen OphrhalmologsdeaGesellschafr,Hamburg 1962, herausgegeben von Prof. Dr. H. Sautter. In: Enke F, ed. EnrwickluegucdPortschriu in der Augenheilkunde. Sturrgart: Verlag, 1963:107-120.

2. PranceschetciA, Francois J. Fundus flavimacularus. Arch Ophrhalrnol 1965;25:505-530.3. Blacharski PA Fundus Havimacularus. In: Newsome DA, ed. Retinal Dystrophies and DegenerationsNew York: Raven Press, 1988:135-159.

4. Armstrong JO, et al. Long-term follow-up of Srargardr's disease and fundus flavimsculmuOphthalmology 1998;1050448-458.

5. Hadden OE, Cass 10M. Fundus flavimacularus and Scargardt's disease. Am J Ophthalmol1976;82;527-539.

6. Klien BA, Krill AE. Fundus flavimaculatus. Clinical, functional, and histopathologic observations.Am)Ophthalmol 1967;64;3-23.

7. RorensrreichY, Fishman GA, Anderson RJ. Visual acuity loss and clinical observations in a largeseriesa[p.a-tienrswith Srargardrdisease. Ophthalmology 2003; 110: 1151-1158.

8. Srargardt K Uber familiare progressive degeneration in der makulagegend des auges.AlbrechtvonGraefnArch Klin Ophrhalmol 1909;71:534-550.

9. Aaber~ '!"M.St~gardt's disease and fundus flavirnaculann- evaluation of morphologic progressionand in-rrafamilial co-existence. Trans Am Ophthalrnol Soc 1986;84:453-487.

10. Lois N, et a]. Phenotypic subtypes of Stargardr macular dystrophy-fundus Ilavimacularus.ArchOphthalmol Zrjuj.I [9;359-369.

11. Fishman GA, ec ai. Visual acuity loss in patients with Stargardr's macular dystrophy. Ophthalmology1987;94;809-814.

12. Fishman GA: Fundus fl~vimaculatus. Arch Ophrhalmol 1976;94:2061-2067.13. Franceschetti A. A special form of tapetoretinal degeneration: fu d n· I Am '--'

L • n us f1avunacuatm. Trans !\l.<lUOphthalmol Otolaryngol 1965;69: 1048-1 053.

14. O~ KT, er al. Clinical phenotype as a prognostic factor in Srargardt disease. Retina 2004;24:254---262.is. ~ltkmets R, et al. A photoreceptor cell-specificATP-binding transporter gene (ABeR) is mutatedinreces-

siveStargardt macular dystrophy. Nat Genet 1997; 15:236-246.16. Kaplan J, et al. A gene for Stargardt's disease (fundus flavimacularus) maps to the shon armofchromo'

some 1. Nat Genet 1993;5:308-311.17. Radu RA, et al. Light exposure stimulates formation of AlE ox" d I fS dr'smac-

u1 d . p Iranes III a mouse mo e 0 targaear egeneratlon. roc Nat! Acad Sci USA 2004; 101:5928-5933.

CHAPTER llG FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS Z15

18. Main P, er a]. Small molecule RPE65 antagonists limit the visual cycle and prevent lipofuscin formation.

Biochemistry 2006;45:852~860.19. Radu RA, et al. Treatment with isotretinoin inhibits lipofuscin accumulation in a mouse model of reces-

sive Srargardr's macular degeneration. Proc Nat! Acad Sci USA 2003; 100:4742--4747.20. Radu RA, er al. Reductions in serum vitamin A arrest accumulation of roxie retinal fluorophores: a poten-

tial therapy for treatment of lipofuscin-based rerinal diseases. Invest Ophrhalmol Vis Sci 2005;

46A393--440 1.21. Behan)' 5, Zhong M, Maida)' RS. N-rerinylidene-phosphaddylerhanolamine is the preferred retinoid sub-

strate for the phororecepror-specific ABC transporter ABCA4 (ABCR). J Bioi Chern 2004;52:53972-53979.22. Sun H, Molday RS, Narhans ]. Retinal stimulates ATP hydrolysis by purified and reconstituted ABCR, the

photoreceptor-specific ATP-binding cassette transportet responsible for Srargardr disease. J Bioi Chern1999;27408269-8281.

23. Weng], er a!. Insights into the function of Rim protein in phororeceprors and etiology of Srargardc's dis-ease from the phenotype in aber knockout mice. Cell 1999;98: 13-23.

24. Eldred GE, Lasky MR. Retinal age pigments generated by self-assembling lysosomotropic decergenrs.Nature 1998;361:724-726.

25. Holz FG, et al. Inhibition of lysosomal degradative functions in RPE cells by a retinoid component oflipofuscin. Invest Ophthalmol Vis Sci 1999;40:737-743.

26. Sparrow JR, Nakanishi K, Parish CA. The lipofuscin fluorophore AlE mediates blue light-induced dam-age to retinal pigmented epithelial cells. Invest Ophrhalmol Vis Sci 2000;41:1981-1989.

27. Suter M, er al. Age-related macular degeneration. The lipofuscin component N-retinyl-N-retinylideneethanolamine detaches proapoprotic proteins from mitochondria and induces apoprosis in mammalianretinal pigment epithelial cells. J Biol Chern 2000;275:39625-39630.

28. Lois N. New perspectives in Srargardr's disease. In: Holz FG, Spaide RF, eds. Essentials inOphthalmology: Medical Reriua. Berlin: Springer-Verlag, 2007: 166-177.

29. Birnbach CD, er al. Histopathology and immunocytochemistry of the neurosensory retina in fundus flavi-macularus. Ophthalmology 1994;101:1211-1219.

30. Eagle RC], er al. Retinal pigmenr epithelial abnormalities in fundus flavimaculatus: a light and electronmicroscopic study. Ophthalmology 1980;87: 1189-1200.

31. McDonnell PJ, er al. Fundus flavimaculatus without maculoparhy. A clinicopathologic study.Ophthalmology 1986;93: 116-119.

32. Irvine AR, Wergeland FL]. Srargardr's hereditary progressive macular degeneration. Br J Ophrhalmol1972;56;817-826.

33. Bonnin P. Le signe du silence choroidien dans les degenerescences rapero-retiniennes cencrales examineessous fluoresceine. Bull Soc Ophchalmol Fr 1971;71:1423-1427.

34. Uliss AE, Moore AT, Bird AC. The dark choroid in posterior retinal dystrophies. Ophthalmology1987;94; 1423-1427.

35. Von Riickmann A, Firzke FW, Bird AC. In vivo fundus autofluorescence in macular dystrophies. ArchOpbchalmol 1997; 115;609-615.

36. Wroblewski JJ, et al. Indocyanine green angiography in Srargardr's flavimacularus. Am ] Ophrhalmol1995; 120;208-218.

37. Querques G, et al. Analysis of retinal flecks in fundus flavimacularus using optical coherence tomography.Be J Ophthalmo12006;90J 157-1162.

38. Gerth, C. Visualization of lipofuscin accumulation in Stargardr macular dystrophy by high-resolutionFourier-domain optical coherence tomography. Arch Ophrhalmol 2007;125:575.

39. Querques G, et al. Correlation of visual function impairment and OCT findings in patients with Scargardrdisease and fundus flavimaculams. Eur J Ophrhalmol 2008;18:239-247.

40. Ergun E, er al. Assessment of centra! visual Function in Stargardr's disease/fundus flavimacularus with ul-trahigh-resolution optical coherence tomography. Invest Ophchalmol Vis Sci 2005;46:310-316.

41. Carr RE. Fundus flavimaculatus. Arch OphthalmoI1965;74:163-168.42. Lois N, et a]. Intrafamilial variation of phenotype in Stargardt macular dystrophy-fundus flavimaculatus.

Invest Ophthalmol Vis Sci 1999;402668-2675.43. Stavrou P, et a!. Electrophysiologica..l findings in Stargardr's-fundus flavimaculatus. Eye 1998;] 2:953-958.44. Lois N, et al. Fundus autofluorescence in Stargardt macular dystrophy-fundus fJavimacularus. Am ]

OphthalmoL 2004; 138:55-63.45. Boon C), Jeroen KJevering B, Keunen JE, et a1. Fundus autofluorescence imaging in retinal dystrophies.

Vision Res 2008;48:2569-2577.46. Gomes NL, et al. Visual function, fundus autofluorescence and structure in Stargardt disease. ARVO

2008;2041.47. De Laey JJ, Verougsrraete C. Hyperlipofuscinosis and subretinal fibrosis in Srargardt's disease. Retina

1995; 15;399--406.48. Klein R, et al. Subretinal neovascularization associated with fundus flavimacularus. Arch Ophthalmol

1978;96;2054-2057.49. Delori FC, er al. In vivo measurement of lipofuscin in Stargardt's disease-fundus flavimaculatus. Invest

Ophdlall11ol Vis Sci 1995;36:2327-2331.

50. Von Ruckmann A, Firzke FW, Bird AC. Distribution of fundus autofluorescence with a scanninglasa

thalmosccpe. S, J OphthalmoI1995;79A07-412. op'51. Smith, T, ee a]. Autofluorescence memes in Srargardr disease. ARVO 2008;2203.52. Lois, N, er a]. Quantitative evaluation of fundus autofluorescence "in vivo" in eyes with retinal dse

J Ophth,jmoI2000,84,741-745. "<B,53. Holz FG, et al. Progressio.n of geographic atrophy and impact of fundus autofluorescence patternsin ~.

related macular degeneration. Am J Ophthalrnol 2007; 143:463-472. 't54. Hwang JC, er ai. Predictive value of fundus autofluorescence for development of geographic h

age-related macular degeneration. Invest Ophthaimol Vis Sci 2006;47:2655-2661. 3lrop rln55. Schmit~-Valckenberg S, er al. Correlation .berw~en the, area of ~ncreased autofluorescence 5urroundin ~

ographic atrophy and disease progressIOn In panents With AMD. Invest Ophehalmol Vi gt:2006;47,2648-2654. " 10

56. Gass JDM. Heredodysrrcphic disorders affecting the pigment epithelium and retina. In; Gass]DMStereoscopic Atlas ~f~acular Disease, Diagnosis and Treatment. St. Lou~s: Mosby, 1997:314-325. ,(~

57. Molday RS. A'I'Pcbinding cassette transporter ABCA4: molecular properties and role in vision dI d . J S· S· b all mac~·at egeneranon. JOeng rornem r 2007;39:507-517.


CHAPTER

Fundus Autofluorescence in MaternalInherited Diabetes and Deafness

mitochondria are ubiquitous in eukaryotes and are essential for survival. TheirIW primary function is ro support aerobic respiration and provide energy to thecell. Mitochondria also play an important role in cell signaling for apoptotic cell

death. Genes located in the mitochondrial DNA (mtDNA) encode subunits of the mi-tochondrial respiratory chain where ATP is generated. Because pathogenic mutations ofmtDNA usually do not affect all mtDNA within a cell, the clinical phenotype of the mi-tochondrial disease depends on the relative proportion of mutant and wild-typemtDNA in different tissues. Purrhermore, depending on the energy demand of a par-ticular cell, the level of mutated genomes required to produce a phenotypic expressionwill vaty. Disease is expressed when, at a parricular threshold, ATP production fallsbelow the energy demand. Until relatively recently, there was a general lack of awarenessof mitochondrial disease. Maternal inherired diabetes and deafness (MIDD) was first re-ported only in 1992 (1,2).

The prevalence of MIDD described in rhe general population is 0.06%; the dis-ease is found in about 1.5% of diabetic populations in different countries and peopleof differenr erhnic backgrounds (1--4). MIDD is considered to be a subtype ofdiabetes mellitus that cosegregates with the most common mitochondrial DNA pointmutation, an adenine-to-guanine transition at position 3243 of the mitochondrialDNA (A3243G) (I).

Clinically, mitochondrial diabetes is typically combined wirh neurosensory hearingloss and retinal dystrophy. Therefore, this disease is easily distinguishable from the idio-pathic forms of diabetes (4). Additional findings, such as a progressive defect in insulinsecretion and neuromuscular signs, may be helpful in identifying the disease. The finaldiagnosis is made by derecting rhe mitochondrial DNA point mutation A3243G in pe-ripheral blood leukocytes by molecular-generic procedures using DNA from oral mu-cosa cells, hair follicles, or muscle biopsy.

Diaberic rerinopathy changes are seldom found in parients with MIDD, whereasretinal dystrophy changes are vety common. The latter have been described with a preva-lence of up ro 85.7% (5,6), and they may lead to a diagnosis of the condirion. These reti-nal changes are concentrated on rhe posterior pole and range from mildly abnormal pig-mentation at the posterior pole to extensive atrophy of the retinal pigment epithelium(RPE) with a symmetric distribution berween both eyes (Fig. II H.I) (5-19). Differentpatterns of pigmentary retinopathy have been observed within the same family, whichsuggests that the particular pattern presenr in a patient may change over time (11).Furthermore, in a recent srudy of seven parients with MIDD, 10 our of 34 marernal rel-atives presented the typical rerinal changes ofMIDD with abnormal pigmentation at theposrerior pole. All of them tested positive for theA3243G mutation, suggesring that reti-nal abnormalities may be a reliable clinical indicaror for a positive mutation (20).Despite the retinal changes, visual prognosis is generally good in patients with MIDD.In a multicenter study, 80% of parients presented with visual acuity of 6/7.5 or better in

FIGURE llH.l. Fundus photograph (A) and AF image (B) obtained from the left eye of a patient with MIDD.PatchesolA~atrophy preserving the fovea were observed on fundus examination (A). AF images disclosed well-defined areasofreducedMsignal (8) AF images defined more clearly areas of RPE loss. In fact, at a location where clinically there seemedtobeRPEII·rophy (superior to the foveal, AF images disclosed a background AF signal. indicating preserved AF and thus RPE10 Ihisar~.Very small foci of increased and decreased AF signal are also visible outside the fovea and in between areas of atrophy.

both eyes (5). No correlation has been found among the severity of retinal changes,thepatient's age, and the percentage of mutant mtDNA (11). Moreover, there is no knowncorrelation between the proportion of mutant mrDNA and the clinical featuresoffam-ily members (14). However, the percentage of mutant mrDNA may vary from onea,·sue to another and may not correspond to the percentage of mutant mrDNA in min.tissue in a patient with th.is disease.

SECTION II ClINICAL SCIENCE

A

MOLECULAR BASIS AND PATHOLOGYThe parhogenesis of the retinal dystrophy in MIDD is nor clear. Retinal pigmenraryab-normalities commonly occur in other mitochondrial disorders, such as Kearns-Sayresyndrome, a neuromuscular disorder that is characterized by, in addition (Q retinalchanges, chronic progressive external ophthalmoplegia, and heart disease. In MEWsyndrome, rhe combination of mitochondrial myopathy, encephalornyoparhy, lactateacidosis, and stroke-like episodes leads to a high prevalence of retinal changes (2l). Bothsyndromes are caused by the same mitochondrial A3243G mutation as describedlOrMIDD. Morphological studies in patients with Kearns-Sayre and MELAS syndromerevealed ultrastrucrural changes in the RPE with enlarged mitochondria (22). Pathologicabnormalities were most marked posteriorly and included both hypo- and hyperpigmen-ration (23-26). These studies showed further degeneration of photoreceptor outer"g_menrs as well as complete atrophy of photoreceptor cells. It was suggested rhar thesechanges were most likely secondary to RPE degeneration. Chang and coworkers (24)reported ocular histopathologic and ultrastrucrural changes in two patients who sufferedfrom Kearns-Sayre and MELAS syndromes. In one of the patients srudied, Buores",inangiography (FA) had been obtained 3 years before the patient's death. The blockedflu-orescence described on FA was thought to be due to the accumularion of lipofuscininRPE cells. Parhologic evaluation of the same eye revealed a degenerared RPE withover-lying photoreceptor cell arrophy in the central rerina. In adjacent areas, there wasinmaRPE conrammg melanin granules, lipofuscin globules, and large melanolipofuscin con-

CHAPTERl1H FUNDUS AUTOFLUORESCENCEIN MATERNAL INHERITED DIABETES AND DEAFNESS

glomerates with overlying disorganized photoreceptor outer segments. However,McKechnie and coworkers (23) could not confirm an increased amount of lipofuscin inRPE cells in a case with Kearns-Sayre syndrome.

The mechanisms by which the RPE is affected in MIDD are not completely under-stood . .An age-dependent somatic selection favors the persistence of mitochondria carry-ing the mutation in mitochondrial disease, which explains why patients with MIDD typ-icallyare not identified before the fourth decade of life (5). Defective mitochondria maybe not properly autophagocyrosed. Theit components may undergo further oxidativemodification within the lysosomes, resulting in the formation of additional undegradablematerial, such as the lipofuscin in RPE cells, and progressively less mitochondrial recy-cling. Consequently, compensatoty mechanisms may fail with time, followed by dys-function and cell death, particularly in relation to postmirotic tissues with high energydemands, such as phororeceptors and RPE cells (27). In this regard, fundus autofluores-cence (AF) findings have shed some light on the pathogenesis of the retinal changes thatoccur in MIDD (see below).

Fluorescein AngiographyFluorescein angiography (FA) is rarely, if ever, required in the evaluation of patientswith MIDD and has been replaced by AF imaging (see below). FA may disclose mot-tled hyper- and hypofluorescence at the macula from RPE window defects and in-creased RPE pigmentation, respectively (5). In this regard, FA may be a useful tool toestablish the diagnosis of MIDD. However, the vety distinct phenotypic appearance ofpatients with MIDD on AF imaging, and the fact that AF is noninvasive have made AFthe preferred imaging technique for evaluaring patients with a possible diagnosisofMIDD.

IMAGING TECHNIQUES

Fundus AutofluorescenceAF is a useful tool for evaluating retinal changes in MIDD. AF imaging in MIDDcan clearly demonstrate RPE involvement (Figs. IIH.l-ll HA) (19). A decreased

AB

FIGURE llH.2. Fundusphotograph (A) and fundus AF 181 image obtained from the right eye of a patient with MIDD with ad-vanced macular atrophy involving the fovea. A characteristic speckled pattern of AF is observed outside the area of atrophy


FIGUREllH.3. Fundus photograph (A) and AF (B) image obtained from the right eye of a patient with MIDD with earlyreli·nal changes in the parafoveal region. Small areas of increased pigmentation clinically (A) corresponded to areason increaser!AF signallB)

AF signal is observed corresponding to areas of RPE atrophy detected clinically,byslit-lamp biomicroscopy and fundus photography. However, a reduced AF signalin·dicating RPE loss may also be observed in areas rhat are nor clearly identified diri-cally. An increased AF signal is found adjacent to areas of RPE atrophy, wherenochanges are detected clinically. Increased AF is rhought to be the result of a highmetabolic turnover of photoreceptor outer segments leading to lipofuscin accumula-

FIGUREllH.4 .. Multifocal ERGstimuli superimposed on fundus AF images obtained from fourdifferentpatients (A-D) With MIDD: Extensive RPEatrophy with decreased fundus AF is visible at the posterior~IeIn patients Band C. In adjacent areas the AF signal is irregularly increased. Patients A and 0 presentamore speckled AF appearance. In all subjects, mfERG trace array amplitude changes are presentThesechanges are not limited to the zones of RPEatrophy but are present in areas of increased AF uSingtheS!,hexagon stimulus. No distinction can be made between areas of RPE t h d . h'· ed". .. . a rap yan areas Wit mcreas fU

(Reprrnted from Ref. 19 With perrmssion from the Association for Research in Vision and Ophthalmology"the copyrrght holder.)

A

CHAPTER11H FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS 22\

FIGUREllH.5. AF image obtained from the right eye of a patient with geographic atrophy due to AMD.Inareas of atrophy the AF signal is decreased, and in adjacent areas at the posterior pole the AF signal isIncreased.

tion in RPE cells, with subsequent retinal impairment (19,28). Similarities betweenthe pattern of AF observed in MIOO and that described in patients with maculardystrophies and geographic atrophy due to age-related macular degeneration (AMO)(Figs. IIH.5 and IIH.6) (29,30) suggest a common pathogenic pathway in the de-velopment of RPE atrophy in retinal degenerations, which may be explained by thefact that mitochondrial abnormalities are present not only in cells of patients with

FIGURE llH,6. AF image obtained from the right eye of a patient with Stargardt disease. A central areaof atrophy with reduced AF signal is observed, surrounded by small foci of increased AF corresponding tofundus flecks

m SECTION II CLINICAL SCIENCE

. . d d' b t also in any aging posrrnitotic cells (31),Ret'",maternally inherite iseases, u. fl I ,I,,,,. d . h it chondrial disease thus may re ecr rne accelerationofchanges associate WIt IDl 0 . f . h . an

d Th ent finding that a vanant 0 a rrutoc ondnal protein'age-relate process. e rec . h . ( ". d . h h d lopmenr of AMD supports this hypot esis 32).assocrare WIt t e eve . h f ialTh I . fAF bnormalities in MIDD, WIt pre erenu paracenrralretin"e ocanon 0 a . ill

. I . -iki (19) The observation of a restricted area of damage;,mvo vement, IS stn ng . . UJ

MIDD correlates with the e1ectrophysiological results obtained from a smallmho,f oati hi h d onstrared restricred rather than generalized photorecepror,,"o patients, W tc em . . udysfunction. Ganzfeld full-field electroretinogram (ERG) abnormalltleswere n~to~.served in the majority of patients tested, whereas amplitude changes In multlfoolI . s (mfERGs) corresponded well With the area of abnotmali'" de.e ectroretmogram . 'J

tecred on AF imaging (Fig. II HA) (19). These findmgs demonstrate nonunifonnretinal damage (19) and suggest damage ro the cone phororeceptor curer segmentsioMIDD. Functional findings also support the hypothesis that both phororeceptorouter segments and RPE cells are involved in the pathogenesis of MIDD, consistentwith the histological data described in Kearns-Sayre and MELAS syndromes (s~above).

In 12 patients with different stages of MIDD, a typical patlern of AF wasfound with vety small foci of highly increased AF signal combined with smallfociand larger patches of reduced AF signal (Figs. II H.l and 11 H.2). The small fociofhighly increased AF signal corresponded, in some cases, ro small pale depositsvisi,ble on fundus examination. However, in many cases the diffuse speckled appear.ance of the macula on AF imaging did not correspond to any obvious changesonslit-lamp biomicroscopy (33). In fact, the area of abnormal AF at the posteriorpolein MIDD is significantly larger than that expected from the fundus appearance(33). These findings underline the usefulness of AF imaging in MIDD.

DIFFERENTIAL DIAGNOSISMIDD should be considered in the differential diagnosis of individuals presen"ngwith macular pigmentary abnormaliries, especially when combined with paracentralareas of atrophy (33). MIDD should be differentiated from atrophic AMD (seealsoChapter II C), central areolar choroidal sclerosis, Stargardt disease (see also ChapterIIG), and pattern dystrophy (see also Chapter lIE). In atrophic AMD, areascfdecreased AF signal corresponding to areas of atrophy may be surrounded by adjacentareas of increased AF (Fig. IIH,5), but may be similar to AF changes describedinMIDD (Figs, IIH.l and IIH.2). However, the speckledAF signal observed through.out the macula, which is characteristic of patients with the A3243G mtDNA muta-tion, is not usually present in patients with atrophic AMD. Unlike MIDD, centralareolar choroidal sclerosis affects the fovea and, in the latter, no abnormalities in [hedistribution of background AF, outside areas of reduced AF, are found, Similar/ylOMIDD, Stargardr disease may present with multiple foci of increased and reducedAFsignal at the macula (Fig. I1H.6). However, in most cases of Stargardt disease,rbevety small fOCIof increased and decreased AF do not coalesce, as occurs in MIDD.lnthe vety few cases in which this may occur, it usually happens in patients withad-vanced disease 111 whom, unlike MIDD, no foveal preservation will be presentF1I1d1l1gssimilar to those observed in MIDD can be detected in patients with panerndystrophy, speCIfically 111 the maculopathy caused by the dominant RI72W periph·erin mutation (33). The AF results from such patients appear to depend on thes"geof the disease (34). In the early symptomatic stages, patients with the RI72W

CHAPTER l1H FUNDUS AUTOFLUORESCENCE IN MATERNAL INHERITED DIABETES AND DEAFNESS

1. van den Ouweland )MW, Lemkes HHPj, Ruitenbeek W, et ai. Mutation in mitochondrial [RNA Leu(UUR) gene in a large pedigree with maternally transmitted type n diabetes mellitus and deafness. Nat Genet1992;1;368-377.

2. Ballinger SW, Shoffner )M, Hedaya EV, er al. Maternal transmitted diabetes and deafness associated witha IO.4kb mitochondrial DNA deletion. Nat Genet 1992; I: 11-15.

3. Cuillausseau PJ, Massie P, Dubois-La Forgue D, er al. Maternal inherited diabetes and deafness: a multi-center study. Ann Intern Med 2001;134:721-728.

4. Cerbirz KD, van den Ouweland JM, Maassen JA, er al. Mitochondrial diabetes mellitus: a review.Biochim Biophys Acta 1995;1271:253-260.

5. Massin P, Virally-Monod M, Vialetres B, et al. Prevalence of macular pattern dystrophy in maternally in-herited diabetes and deafness. GEDlAM Group. Ophthalmology 1999;106:1821-1827.

6. Fukui M, Nakano K, Obayashi H, et al. High prevalence of mitochondrial diabetes mellitus in Japanesepatients with major risk factors. Metabolism 1997;46:793-795.

7. Massin P, Guillausseau PJ, Vialetres B, et al. Macular parrern dystrophy associated with a mutation of mito-chondrial DNA. Am J OphilialmoI1995;120;247-248.

8. Bonte CA, Macrhijs GL, Cassiman J), er al. Macular pattern dystrophy in patients with deafness and dia-betes. Retina 1997;17:216-221.

9. Latkany P, Ciulla TA, Cacchillo PF, et a]. Mitochondrial maculopathy: geographic atrophy of the maculain the MELAS associated A to G 3243 mitochondrial DNA point mutation. Am J Ophthalmol1999;128;112-114.

10. Smirh PR, Bain Sc. Good PA, er al. Pigmentary rerinal dystrophy and the syndrome of maternally inher-ited diabetes and deafness caused by rhe mitochondrial DNA 3243 tRNA(Leu) A to G mutation.Ophthalmology 1999; 106: 1101-1108.

11. Harrison T), Boles RG, Johnson DR, et al. Macular pattern retinal dystrophy, adult-onset diabetes, anddeafness: a family study of A3243G mitochondrial heceroplasmy. Am J Ophrhalmol 1997;124:217-221.

12. Souied EH, Sales M], Soubrane G, er al. Macular dystrophy, diabetes, and deafness associated with a largemitochondrial DNA deletion. Am J OphthalmoI1998;125:100-103.

peripherin mutation present a diffuse macular abnormality on AF imaging, describedas speckled areas of increased and decreased AF within the macula. Later in the courseof the disease, areas of atrophy develop within the areas of abnormal AF, although thisdoes not occur perifoveally, as is observed in patients with MIDD. Unlike A3243GmacuJopathy, the changes seen in Rl72W patients appear to be confined to the mac-ula and peripapillary regions until very late in the disease, when atrophic changes canextend beyond the arcades (33,34).

SUMMARYThe prevalence of macular changes in MIDD is higher than was assumed until re-cently. Pigmentary or atrophic retinal changes may be an early sign of the disease. Thecombination of retinal changes typical of MIDD with diabetes and/or deafnessshould lead to screening for a mitochondrial DNA mutation to establish the diagno-sis of MIDD. Careful fundus examination combined with AF imaging is needed toguide genetic testing. In addition to helping in the diagnosis of MIDD, AF imagingcan provide valuable information regarding the degree of retinal involvement andthus the expected functional loss in these patients.

ACKNOWLEDGMENTSThe fundus photographs and fundus AF images presented in this chapter were obtainedduring the Medical Retina Fellowship of the authors at Moorfields Eye Hospital,London, UK.

REFERENCES


I BJR di P er al. Mitochondrial DNA disease masquerading as age- I

13. Andrews RM, McNee a , ea 109 , ~atedI d . Eye 1999;13;595-596.

macu at egeneratlOn. CC 1 Maternally inherited diabetes and deafness (MlDD)synd14. C~~n YN'dLiOUICWuI'Hua.n~ tUd'ytoaf"aTaiwanese family. Chang GungMedJ 2004;27:66-73.rome:a

clinical an rno ec ar genetic S d hv i . . hC M

hii GL C . an JJ er a1. Macular panern ystrop y In panents Wit deafnessand"15. Bonte A, art IJS ,aSSlffi , lW·

heres. Retina 1997;17:216-221. . I h hik P C· II TA C chillo PF et al. Mitochondnal macu apat y: geograp lC atrophyofthema~·l

16. Lar any , III a ,ac, "a! DNA' . Am ~,. ELAS . d A to G 3243 mirochondn pamt mutation. J Oph,l., ,In the M associate' llldlJllO!

1999; 128;1I2-114. .Smith PR, Bain SC, Good PA, et al. Pigmentary retinal ~ysuophy and the syndrome ofmaternallrinhrr.

17. ired diabetes and deafness caused by the mirochondnal DNA 3243 tRNA(Leu) A to G mutation.Ophthalmology 1999;106;1101-1108. ....

18. Latvala T, Musronen E, Uusiralc R, et al. Pigmentary retinopathy In patients With the MEW mU!llioD

3243A->G in mitOchondrial DNA. Graefes Arch Clin Exp Ophthalm~1 20?2;240:795-80L19. Bellmann C, Neveu MM, Scholl HPN, er aI. Localized retinal ele~rodphrlolo~lca1 and fundusaurofiuolt¥

cence imaging abnormalities in maternal inherited diaberes an ea ness. nvesr OphthalmolVisSci2004;452355-2360. . .

20. Michaelides M, Jenkins SA, Bamiou DE, er al. Macular dystrophy assoclat~d :V.IIDthe A3243Gmlroch(\![.drial DNA mutation. Discincc retinal and associated features, disease variability, and characterization~iasymptomatic family members. Arch Ophthalmol 2008; 126:?20-328. . .

2]. Sue eM, Mitchell P, Crimmins OS, er a]. Pigmentary retinopathy associared WIththe mitorhomlrialDNA 3243 point mutation. Neurology 1997;49:1013-1017.

22. Newell FW, Polascik MA. Mitochondrial disease and retinal pigmentary degeneration. In:Proceeding.oithe 3rd International Congress of Ophthalmology, Kyoco, Japan. New York: ElsevierSciencePubl~brn~1979;1613-1617.

23. McKechnie NM, King M, Lee WR Retinal pathology in the Kearns-Sayre syndrome. BrJ Ophlh~mGI1985;6%3-75.

24. Chang TS, Johns DR, Walker D, er al. Ocular clinicopathologic study of the mitochondrialencephalomr'apathy overlap syndrome. Arch Ophrhalmol 1993; Ill: 1254-1262.

25. Rummeh V, Polberg R, Ionescu Y, er al. Ocular pathology of MELAS syndrome with mitochondrialDNAnucleotide 3243 point mutation. Ophthalmology 1993; 100:1757-1766.

26. Eagle RC, Hedges TR, Yanoff M. The atypical pigmentary retinopathy of Kearns-Sayresyndrome:aIigh!and electron microscopic study. Ophthalmology 1982;89: 1433-1440.

27. Wallace DC. Diseases of the mitochondrial DNA. Annu Rev Biochem 1992;61:1175-1212.28. Rtlckmann Av, Pitzke FW, Bird AC Distribution of fundus autofluorescence with a scanninglaseroph-

rhalmoscope. Br J Ophthalmol1995;79:407-412.29. Holz FG, Bellmann C, Margaricidis M, et al. Patterns of increased in vivo fundus auwfluorescenceintht

junctional zone of geographic atrophy of the retinal pigment epimdiwn associatedwith age-relatedmKIrlardegeneration. GraefesArch Clin Exp OphmalmoI1999;237:145-152.

30. Bindewald A, Schmitz-Valckenberg S, Jonik JJ, et al. Classification of abnormal fundusauwfluOf&ellr.patterns in the junctional zone of geographic atrophy in patients with age related maculardegeneraliou.BrJ Ophth,]mol 2005;89;874-878.

31. Wallace DC. Mirochondrial diseases in man and mouse. Science 1999;283:1482-1488.32. Kanda A, Chen W, Othman M, et al. A variant of mitochondria.! protein LOC387715/ARMS2, DDI

HTRA1, is strongly associated with age-related macular degeneration. Proc Nacl Acad ScirIA2007; I04; 16227-16232.

33. Rath PP, Je~1c.insS, ~ichaelides M, et at. Characterisation of the macular dystrophy in patient.>widitiltA3243G ffiltochondnal DNA point mutation with fundus autofluorescence. Br J O?hth:llincl2008;92;623-629.

34. ~o,,:nes SM, Fitzke FW, Holder GE, et a1. Clinical features of codon 172 RDS maculardysrropby.SImIlar phenotype in 12 families. Arch Ophthalmol 1999; 117: 1373-1383.

Fundus Autofluorescencein Choroideremia

IIP3 horoideremia (CHM) is a progressive retinal dystrophy with an X-linked mode~ of inheritance (1-3) caused by mutations in the CHM gene (4). The name

"choroideremia" means an absence (eremia) of the choroid and points out thetypical findings in advanced stages of the disease: a complete atrophy of the choroid andvisible bare sclera. CHM is characterized by a progressive degeneration of the photore-ceptors and retinal pigment epithelium (RPE), followed by a degeneration of thechoroid (5). Full-field electroretinogram (ERG) is reduced early on in the disease, show-ing a rod-cone dysfunction, and with disease progression the ERG becomes nonrecord-able (6). Fundus changes are visible during the first decade of life, with mottled RPEalterations in the periphery being the first manifestations of the disease. Subsequently,areas of RPE and choroid atrophy develop in the far petiphery and midperiphery. Theatrophic lesions increase in size and become confluent, spreading toward the center overthe years; however, the macula remains spared for decades of life. Finally, bare sclera isseen throughout the fundus. Nyctalopia is one of the first symptoms in CHM patients.Visual field defects develop in the mid periphery, followed by progression to concenrricvisual field loss, color vision defects, photophobia, and loss of visual acuity.

Because of the X-linked inheritance, only males are affected. Female carriers whomanifest CHM are very rare (2,5,7-10). Normally, all female carriers show patchyfundus changes, including various grades of mottled RPE alterations, RPE stippling,or spOtty pigment atrophy in the periphery. Lyonization, i.e., a random Xvinactiva-rion, explains the various phenorypes observed in carriers of the disease (7). The full-field ERG is mostly normal in carriers of CHM.

In addition to a complete eye examination, important diagnostic tools includedetailed case and family histories, tests of color vision and visual field, recording offull-field ERG, and fundus autofluorescence (AF). Family members should be exam-ined and, in suspected CHM, a genetic analysis of the CHM gene should be provided.To date, no treatment for CHM is available.

MOLECULAR BASIS/PATHOLOGYThe gene underlying CHM was first described in 1990 (4) and encodes Rab escortprotein 1 (REP-I). More than 70 CHM gene mutations have been revealed so far (fordetails, see http.!lwww.retina-international.orglsci-newslrepmut.htm). All CHM mu-tations lead to complete loss of the gene product REP-I. Of interest, the gene is ubiq-uitously expressed; however, mutations affect only the eye (11). This phenomenon iscurrently explained by a specific gene substrate in the retina and choroid: Rab27, (12).REP-l is necessary for rransferring geranylgeranyl groups to Rab proteins, which aresmall GTP-binding proteins (4,13,14). Rab proteins are involved in functions such asprotein trafficking, endocytosis, intracellular vesicle transportation, and signal transduc-tion (15). For a detailed overview of REP-l functions, see Preising et a1. (16).


CHM gene expression has been shown in rods and RPE cells (5), indicatingrh,.. f th di . rods and/or RPE cells. Therefore, rhe degenerationof hongm 0 e isease In t e

cones and choroid seems to be secondary to rhe demise of rods and RPE.Onl1' f,whistopathologic studies of CHM have been published, and those dealrmainlywillithe eyes of female carriers (5,17-19); rhe data regarding affected malesarescarce(20). These studies showed patchy degeneration of rhe retrna m female carriers,willimixed areas of normal photoreceptors, photoreceptors that had lost theirOUrerseg.merits, and areas of loss of the entire photoreceptor cells. In addition, RPE changeswere found, including abnormal RPE cells wirh irregularities In thICkness,variousamounts of melanin and lipofuscin granules with some clumping of melaningran.ules, and areas with focal RPE hypertrophy or thinning (5). The chotiocapillarisw"normal except in areas with severe retinal degeneration (5).

IMAGING TECHNIQUESFluorescein AngiographyFluorescein angiography (FA) facilitates the visualization of areas ofRPE andchoroid,atrophy. In these areas, the remaining large choroidal vessels can be easilyidentified(21-23). Patchy fundus changes and RPE mottling in female carriers can bedemon.strated with FA (7) (Fig. 111.1). In the case of suspected subretinal neovascularizacion,a very rare complication of CHM, FA is essential (9,24,25). Similarly, FAisneeded10detect the very rare occurrence of intraretinaI neovascularization in CHM (26).

Indocyanine Green AngiographyIndocyanine gteen (ICG) angiography is not toutinely used in clinical practicefordi.agnosing CHM. Only limited data exist regarding the use of ICG in patientswithCHM. Forsius et a1. (27) reported that ICG shows the choroidal vesselsin are.where RPE and choriocapillaris are still present, whereas RPE atrophy and remaining

FIGURE 111.1. FA of the right eye of a female CHM carrier, 26 years of age. There are patchychangesof the RPEthroughout the fundus.

CHAPTER 111 FUNDUS AUTDFLUDRESCENCE IN CHDRDIDEREMIA

choriocapillaris are less well visualized by ICG. The choroidal blood circulation wasslow in advanced stages of CHM.

Optical Coherence TomographySeveral retinal changes can be detected by optical coherence romography (OCT) im-aging in patients with CHM. Retinal thickening, which may be due to Muller cell ac-tivation and hypertrophy, has been claimed to be a marker of the earliest stage of thedisease (28). As the disease progresses, loss of photoreceptor nuclei and shortening ofthe outer and inner segments of the photoreceptors, followed by disorganization andsubsequent slow thinning of the retina over decades, and abnormal laminar architec-rure have been reported (28~30).

Fundus AutofluorescenceOver the past few years, AF has become a very useful tool for the diagnosis and follow-up of patients with inherited retinal dysrrophies. Data on AF in CHM and femalecarriers demonstrate the usefulness of AF as a diagnostic tool in this disease (10,31,32).In young CHM patients, the peripheral retina shows only RPE mortling, which is oftendifficult to detect clinically. AF imaging demonstrates these RPE irregulatities well, inthe form of densely packed small areas with reduced AF signal sparing the fovea; at thefovea, AF remains homogeneous. Under these circumstances, the diagnosis of CHMmay be missed, but AF findings will point toward the diagnosis of CHM (Fig. 111.2).The extent of RPE alterations can be evaluated more precisely with AF than with anyother imaging technique. Furthermore, AF is noninvasive and is faster and easier to per-form than FA. These advantages of AF imaging are especially important when evaluat-ing children suspected to have CHM. In areas of RPE atrophy, a low AF signal is ob-

A

FIGURE 111.2. Fundus photography (AI and fundus AF (8) of the right eye of a male. 16 years of age. with CHM. VISualacu-ity was 1.0. Slit-lamp biomicroscopy disclosed what could be considered a normal. although very mildly pigmented. fundus (A)Only a very fine and subtle mottling in the RPE was present. AF imaging demonstrated marked abnormalities in the distributionof Af. with a characteristic diffuse pattern of multiple. coalescent areas of reduced and increased AF sparing the fovea. whereAf was stili homogeneous.

8


f I f, d . RPE j'rregularities and it reveals more widespread changesin·lu or erecnng , . . lUeRPE than can be observed funduscopically. The charactenstlc AF pattern infem.,

. f CHM . f I f r detecting the disease.earners 0 15very use u 0

DIFFERENTIAL DIAGNOSISThe diagnosis of CHM can be difficult to establish in early stages of the disease.Themost important differential diagnosis in these cases 15 Xvlinked retiruns pigmentlJla(RP). In cases of suspected CHM, female carriers in the affected familyhavero~,examined. This is important to exclude the diagnosis of X-linked RP, in whichfe.males carriers have no fundus changes but a reduced full-field ERG, in comr,,"oCHM, where female carriers show fundus changes bur mostly a normal full-fieldERG. A further important differential diagnosis is gyrate atrophy, whichcanb,treated with a special diet. Like CHM, gytate atrophy shows atrophic a1terarionso!RPE and choroid beginning in the periphery and spreading toward the center.However, the inheritance is autosomal recessive, and a thorough family hisrorycanhelp differentiate between the two diseases. In addition, gyrate atrophy is accompa-nied by an increased blood level of ornithine acid, which should be measuredto con-firm or exclude the diagnosis of gyrate atrophy. To dare, there are no publisheddataon AF in gyrate atrophy, likely because of the extreme rareness of the condition.Itispossible that AF may help in differentiating between CHM and gyrate atrophy.Tb,atrophic areas in gyrate atrophy are very sharply demarcated, and there arenomertied RPE changes between the lesions or parchy RPE changes on FA. Therefore,ili,speckled pattern in AF should be seen only in CHM and nor in gyrare arrophy.

SUMMARYAF imaging is an excellent method for obraining informarion about the RPEandinalterations in vivo. In contrast to FA, AF is fast and noninvasive. Because Rl'Eahe-ations are common in CHM and, in general, mosr of the hereditary retinaldysao·phies, AF imaging should be accepted as a standard diagnosric rool and canbeUJedinstead of FA in the majority of cases. The AF pattern in female CHM carriersseemsto be specific to female carriers of CHM and may provide an addirional phenotypiccriterion for diagnosis in carriers of this disease (10). In affected males, AF imagingisvery helpful in demonsttating the extent of RPE defects at first examination and,overtime, the increasing RPE loss, which is useful for evaluating the progressionofthedisease,

REFERENCES1. ~ill ~> Arche~D. Cl,assification of the choroidal atrophies. Am J Ophthalmol 1971'72'562-585.2. Kama J. Choroideremia A clinical ad' d f . ,. .

OphtbaJmoI1986;(Sup~I)176:1_68.n genetic Stu yo 84 Finnish pacienrs and 126 femaleearners,Aet1

3. Roberts MF, Fishman GA Roberts DK et I R . f'at '" .' . ' a. etrospecnve, longitudinal, and cross sectionalsrudyo\1'su acUity rrnparrmenr III chofOlderaemia. Br J Ophthalrno12002'86'658-662

4. Crerners PPM, van de Pol DJR, van Kerkhoff LPM . •. . . .with choroideraemia. Nature 1990;347:674-677. ,et al. Cloning of a gene rhar is rearrangedIn pallen:s

5. Syed N, Smith JE, John SK, et al Evaluation f . al . ' fImale carrier of choroiderem] a h h I I a renn photo receptors and pigment epitheliumIn a (.

ra. pta rna ogy 2001;108:711-720.

CHAPTER 111 FUNDUS AUTOFLUORESCENCE IN CHOROloEREMIA

Sieving PA, Niffenegger ]H, Berson E1. Electroretinographic findings in selected pedigrees with choroi-deremia. Am J Ophrhalmol 1986; 10 I :361-367.

7. Rudolph G, Preising M, Kalpadakis P, er al. Phenotypic variability in three carriers from a family withchoroideremia and a Irameshift mutation 1388delCCinsG in the REP-l gene. Ophthalmic Genet2003;24;203-214.

8. Cheung MC, Nune Ge, Wang M, et al. Detection of localized retinal dysfunction in a choroideremia car-rier. Am] Ophrhalmo12004j137:189-191.

9. Potter MJ, Wong E, Szabo SM, et at. Clinical findings in a carrier of a new mutation in the choroideremiagene. Ophthalmology 2004; Ill: 1905-1909.

10. Renner AB, Keltner U, Cropp E, er al. Choroideremia: variability of clinical and e1ectrophysiological char-acteristics and first report of a negative electroretinogram. Ophthalmology 2006; 113:2066-2073.

11. Shi W, van den Hurk )AJM, Alamo-Bethencourt V, et al. Choroideremia gene product affects trophoblastdevelopment and vascularization in mouse extra-embryonic tissues. Dev Bioi 2004;272:53-65.

12. Seabra MC, Ho YK, Anent )S. Deficient geranylgeranylation of RamJRab27 in choroideremia. J BioiChern 1995;270;24420-24427.

13. Seabra MC, Brown MS, Slaughter CA, er al. Purification of component A of Rab geranyigeranyl trans-ferase: possible identiry with the choroideremia gene producr. Cell 1992;70:1049-]057.

14. Seabra MC, Brown MS, Goldstein JL. Retinal degeneration in choroideremia: deficiency of rab geranyl-geranyl transferase. Science 1993;259:377-381.

15. Seabra Me. New insights into the pathogenesis of choroideremia: a tale of two REPs. Ophthalmic Genet1996;17043-46.

16. Preising M, Ayuso C. Rab escort protein 1 (REP!) in intracellular traffic: a functional and pathophysiologi-cal overview. Ophthalmic Genet 2004;25:] 01-1 ]0.

17. Ghosh M, McCulloch C, Parker JA. Pathological study in a female carrier of choroideremia. Can JOphthalrnoI1988;23;181-186.

18. Flannery )G, Bird AC, Farber DB, ec al. A histopathologic study of a choroideremia carrier. InvestOphthalmo! Vis Sci 1990;31:229-236.

19. MacDonald 1M, Chen MH, Addison OJ, et al. Histopathology of the retinal pigment epithelium of a fe-male carrier of choroideremia. Can J OphthalmoI1997;32:329-333.

20. Rodrigues MM, Ballinrine E), Wiggert BN, er al. Choroideremia- a clinical, electron microscopic, and bio-chemical report. Ophthalmology 1984;91:873-883.

21. Hayakawa M, Pujiki K, Horta Y, et a], Visual impairment and REP-I gene mutations in Japanese choroi-deremia patients. Ophthalmic Genet 1999;20:107-115.

22. Itabashi T, Wada Y, Kawamura M, et a1. Clinical features of Japanese families with a 402delT or a 555-556delAG mutation in choroideremia gene. Retina 2004;24:940-945.

23. Kellner U. Choroideremie. In: Heimann H, Kellner U, Foerster MH, eds. Atlas of fundus angiography.Stuttgart: Thieme, 2006.

24. Robinson D, Tiedeman). Choroideremia associated with a subrerinal neovascular membrane. Case re-porro Retina 1987;7:70-74.

25. Endo K, Yuzawa M, Ohba N. Choroideremia associated with subretinal neovascular membrane. ActaOphrhalmol Scand 2000;78:483-486.

26. Sawa M, T amaki Y, K1ancnik)R )M, et al. [ntraretinal foveal neovasculatization in choroideremia. Retina2006;26;585-588.

27. Forsius H, Hyvarinen L, Nieminen H, er al. Fluorescein and indocyanine green fluorescence angiographyin study of affected males and in female carriers with choroideremia. A preliminary report. ActaOphrhalmol (Copenh) 1977;55; 459-470.

28. Jacobson SG, Cideciyan AV, Sumaroka A, et al. Remodeling of the human retina in choroiderernia: rabescort protein I (REP- 1) mutations. Invest Ophthalrnol Vis Sci 2006;47:4113--4120.

29. Katz B), Yang Z, Payne M, er al. Fundus appearance of choroiderernia using optical coherence tomogra-phy. Adv Exp Med Bioi 2006572:57-61.

30. Mura M, Sereda C, Jablonski MM, et al. Clinical and functional findings in choroideremia due to com-plete deletion of the CHM gene. Arch Ophthalmo12007;125:1107-1113.

31. Wegscheider E, Poloschek eM, Preising M, er al. Fundus autofluorescence in carriers for choroideremia.Invest Ophthalmol Vis Sci 2005;46:e-abstract 4088.

32. Poloschek Clvl, Kloeckener-Gruissem B, Hansen LL, et al. Syndromic choroiderernia: sublocalization ofphenotypes associated with Martin-Probst deafness mental retatdation syndrome. Invest Ophthalmol Vis

Sci 2008;49;4096-4104.

231

CHAPTER

Fundus Autofluorescencein Posterior Uveitis

mosterior segment intraocular inflammation (PSII, uveitis affecting t~e posterior• segment of the eye) comes in many forms (1). These range from mild, chronic,

low-grade intermediate uveitis affecting the penphe~al retinal ves~elsWIthocca-sional inflammatory cells in the vitreous to acute, severe, sighr-threarening ocularin-flammation (STOI) wirh occlusive rerinal vasculiris and/or optic disc swelling.lnaddi.rion, STOI may occur because of rhe consequences of inflammarion, particularly edema(macular or retinal edema with or without neurosensory retinal detachment) andneo-vascularizarion (retinal and choroidal with hemorrhage at either sire).

Fundus imaging is of great value in the diagnosis and evaluation of pariemswithPSI!. In particular, it may help in the assessment of STOI. However, since clearmedia are required for fundus imaging, there are limi rations to the use of fundusim-aging in PSII. In particular, patients with pan uveitis and posterior synechiae wit]seclusio pupillae, or with dense vitreitis and vitreous hemorrhage are excluded fromthese investigations.

Despite these caveats, there are many cases of PSI! and STOI that are amenableto and can benefit from good fundus imaging. Cystoid macular edema (eMO) andchoroidal neovascularization (CNV) are two of the possible mechanisms by whichSTOI leads to registrable blindness. Moreover, many cases of chronic STOI escapedetection because the clinical signs are not very obvious and the cause of thepoorlevel of vision is not fully appreciated. As a result, patients lose sight unnecessarily.

There are many methods of fundus imaging: slit-lamp biomicroscopy, colorandred-free fundus photography, fluorescein angiography (FA), indocyanine greenan-giography (IeGA), optical coherence tomography (OCT), and, most recently, fun·dus autofluorescence (AF) and near-infrared autofluorescence (NIA). Each of rheseimaging modalities has advantages and deficiencies for evaluating STOI; these willbereviewed in this chapter.

HISTOPATHOLOGY

In ocular inflammatory disease, one of rhe first events is the breakdown of theblood-retinal barri,er (BRB). Experimentally, this is manifested by a reduction insheMflow in the large retinal vessels, occlusion of small capillaries, adhesion ofieukocyreslOthe lining cells of the BRB, and extravasation of fluid and cells into the exttavasculMspace. The extravascular space in the retina includes the retinal parenchyma andthesub retinal space. In addition, the lming cells of the BRE include not only the retinalendothelium but also the RPE cells. Consequently, cells and fluid can accumulateoneither SIde of the RPE and produce fe h h . al i .. erects t at ave consequences for renn Imagtng(FIg. 12.1).

CHAPTER 12 FUNDUS AUTOFLUORESCENCEIN POSTERIORUVEITIS 233

• •

=~ ---lBA .-FIGURE 12.1. Experimental autoimmune uveoretinitis (EAUI in the mouse.IA) Normal retina (B) EAU. Note the in-tense retinal vasculitis (RVI, granuloma (GI, and RPEswelling. Fluid and inflammatory celis accumulate on either side ofthe RPE.indicating breakdown of the BRB.

CLINICAL FINDINGS AND IMAGING TECHNIQUESRetinal edema (specifically maculat edema, including CMO), inflammatory cell infil-tration in choroid and retina, neovascularization, and subretinal fibrosis occur in pa-tients with PSII. The following imaging systems can be used to evaluate thesechanges:

Fluorescein Angiography and IndocyanineGreen AngiographyIn fluorescein angiography (FA) and indocyanine green angiography (ICGA), the mostobvious effect of inflammation is leakage of dye from the vessels into the extracellularspace. In the retina, any leakage of dye is a sign of abnormaliry because the retinal ves-sels are normally impermeable through tight junctions to protein-bound fluorescein.Choroidal vessels are normally leaky through physiological fenestrations; however, ininflammation, dye leakage can be increased, leading to intense fluorescence, whereascellular extravasation into me normally paucicellular choroidal stroma can lead to

blockage of the dye signal. Thus, on FA and ICGA, accumulation of inflammatory cellscan interfere with the signal and produce hypofluorescence in early frames of the an-giogram and hyperfluorescence in late frames, as dye accumulates around inflammatorylesions. Similarly, as the RPE layer becomes "leaky," fluorescent dye will accumulate atthis level in a "pericellular" distribution on FA (Fig. 12.2). Images of the macula andmid-peripheral retina can be obtained with FA and ICGA using conventional funduscameras or wide-angle systems. This is important because inflammatory lesions canoccur at any site in the posterior segment.

Maculat edema, particularly CMO, is the most immediate direct cause ofSTOI inPSII. Detection and, more importantly, evaluation of macular edema in PSII clinicallyis highly subjective. Stereoscopic fundus photography is used to assess clinically signif-icant macular edema in diabetic retinopathy, and similar methods are useful in PSII,but studies have shown a lack of correlation between observers, particularly regardingthe degree of CMO (2,3). More objective and reproducible measures of macularedema can be obtained with FA (4-6). Macular edema can be demonstrated readily on


A

FIGURE 12.3. FundusAF images obtained from a patient with birdshot choroid apathy in (AI earlyandIBI

advanced stages of the disease. (A) Marked CMO is detected by AF imaging. (B) Widespreadareas01r.duced AF signal due to diffuse RPE atrophy are also easily visualized. An ERG obtained from thepatientshown in Figure 12.3A disclosed reduced amplitude in rod and cone responses with increasedimplicitlimeinphotopic and 3D-Hzflicker; the a-wave of the mixed rod-cone response was also delayed.

RPE. Under these circumstances, CMO may be missed clinically and angiographically.Fundus AF allows evaluation of the entire area of edema, which can be captured nonin-vasively in one image in a few seconds (Fig. 12.3). Moreover, fundus AF allowsan eval·uarion of the RPE and, indirectly, photo receptors, thus providing essential informarionto balance the aggressiveness of treatment against the possibility for functional recovery(Fig. 12.3). AF imaging can also be used to noninvasively monitor the effectofimmunosuppressive therapies and to guide treatment in patients with PSI! (Fig. 12.6).

A

FIGURE 12.4. .Fundus AF images obtained from patients with serpiginous chorioretinopathyIAI anetoxoplasma chorioretinitis (B). (A) A large area of increased AF signal is observed extendinglromapalelof reduced AF. suggesting active disease. (B) A patch of reduced AF signal is observed superiortoiliefovea. With abnormal. predominantly low AF signal at the fovea. Note the presence of an areaofin'creased AF Signal occupymg most of the inferior aspect of the macula and extending toward thetemporal mid periphery, Indicating active disease.

CHAPTER 12 FUNDUS AUTOFLUORESCENCE IN POSTERIORUVEITIS Xl7

FIGURE 12.5. Composite of (A) color fundus photographs and (8) AF images obtained from a patient with inflammatory sub-retinal fibrosis syndrome. Note the extensive changes in the distribution of AFpresent in contrast fundus photographs demon-strate abnormalities only at the site where subretinal fibrosis is observed.

A

C

8

FIGURE 12.6. FundusAF images obtained froma patient with multifocal choroiditis who pre-sented with vitritis and CMO. (AI Despite thedense vitritis, AF images suggested the presenceof CMO. (8) The vitritis gradually settled after sys-temic treatment with corticosteroids, but the CMOpersisted and was easily detected on AF images(C) When systemic immunosuppressive treatment[FK506)was added, the inflammation becamecon-trolled and the CMO resolved.

CLINICAL SCIENCESECTION II

When one compares findings on FA, OCT, and AF in patients with maculf . f at

edema and posterior uveitis, interesting di ferences In areas 0 RPE "leakage"andpatches of choroidal nonperfusion can be seen (see legend of Fig. 12.7 to comparecorresponding areas). The RPE signal detected In OCT Images IS usually quitedis·tinct (Fig. 12.7) and thus it might be presumed that defects in the RPE as a resulof inflammation are readily detectable. Howevet, this is predicated in part byth:amount of retinal thickness and edema in the pre-RPE layers. Thetefore, OCTchanges in the RPE might not correlate directly with changes in the distributionof fundus AF. Thus, complementary infotmation obtained by combining theseimaging techniques is valuable when managing patients with posterior uveitis.Preliminary data using fundus AF images to evaluate chorioretinal infilttationin

PSI! suggest that this imaging technique may be of value for evaluating patientswiththis group of disorders. A recenr reporr on AF in birdshor choroidopathy suggestedthat this rechnique might reveal the true extent of RPE cell atrophy, which isnotev·idenced by slit-lamp biomicroscopy (Fig. 12.3) (28). In addition, similar findings

A

C D

FIGURE 12.1. (A) OCT 190 degree cutl, (8) AF, (C) NIA, and (0) late Ira . ..'choroiditis affecting predominantly the macular area (A) OCT d me 01 FA obtained Irom a patient with mulofoc,118) OnAF imaging an increased AF signal was obse' d h lemonstrated retinal thicksninc/edena and prominentRPEsigoo

l.

rve at t e ovea (white full ) d .temporal paraloveal area (dotted white arrow) with I AF . arrow an a reduced AF signal wasseenrnili.

, norma surrounding th! (bl "signal was also observed in the temporal paraloveal a (0) Th IS area ack arrowl,lC) On NIA, a reduced"an area demonstrating late leakage (white full arrowl ::~A'char e area 01 reduced AF and NIA signal appears to co~espondro

, aida I nonpertusinn was also observed temporally(blackarrowl

FUNDUS AUTOFLUORESCENCE IN POSTERIOR UVEITIS 239CHAPTER 12

were obtained in a series of cases of multi focal choroiditis and panuveitis (29). Of in-reresr, AF revealed small incipient lesions that were not always visible on fundusphotographs- Hence, it is possible that fundus AF can be used for early detection andscreening of active disease in patients with these disorders, without the need to resortto invasive FA or ICGA methods. We have observed similar findings, for instance,in setpiginous cborioretinoparhy. Fundus AF often reveals larger patches ofRPE at-rophy than can be observed on slit-lamp biomicroscopy. Additionally, an increasedAF signal can be detected at the edge of atrophic lesions, likely indicating the pres-ence of active disease (Fig. 12.4). Therefore, it is possible that fundus AF can be usedto evaluate the risk of lesion extension and encroachment onto the fovea, with result-ant visual loss.

Although FA remains the gold standard in the evaluation of parienrs with inflam-matory CNV, useful information can also be obtained in these cases by AF imaging.On AF images, a CNV most often appears as an area of reduced AF signal, which canbe surrounded by a complete or incomplete halo of increased AF (Fig. 12.8). Areas ofrecenr onset or long-standing subretinal fluid appear as areas of reduced (Fig. 12.8) orincreased AF signal, respectively. However, it is not always possible to determinewhether the CNV is active based on AF imaging. In CNV associated with AMD, arelatively normal distribution of AF at the fovea appears to have prognostic signifI-cance with regard to the potential for visual acuity improvement following treatment(see Chaprer 1DB). This may also be the case in inflammatory CNV, although to datethere are no data available to support this.

In areas of subretinal fibrosis, where presumably the RPE is atrophic, has lost itsmelanin, or is packed with lipofuscin-like material or other fluorophores, changes to

FIGURE 12.8. AF Image obtained from a patient with inflammatory CNV. A decreased AF signal is ob-served at the site of the CNV, surrounded by a halo of increased AF; the latter is likely the result of prolif-eration of RPEcells and/or macrophages around the neovascular process. III-defined reduced AF signal ISalso observed around the CNV, likely related to the presence of subretinal fluid. Note the increased AFsignal around the optic nerve; interestingly, in this area two patches of reducedAF signal compatible withRPEatrophy are also observed.


al AF' al r The extent of the subrerinal fibrosis can be c1earlYddinthe norm sIgn occur, fi . .d d h b re widespread than SlLSpecred on color undus Imaging(Figeate an s own to e rna . . diff . .

12.5). In addition, differences in the level of AF sl~nal. 111 ,: erenr regions of thefun.

d . th also relare to differences m acnvuy of the disease th,o""\'OUtus In ese cases may .' . ,"&,1

h f d (F' 12 5) Assuming that rhe AF signal represents hpofuscm aceumul,.t e un us Ig. . . . d AF' al .tion inside RPE cells or macrophages, areas with increase sIgn may Jndicate~.rivared RPE cells or macrophages in those regIOns. Clearly, longitudinal studies01h I· ild be required to confirm their changing nature.t ese esrons WOl

ELECTROPHYSIOLOGYThe functional abnormaliries ascertained by elecrrophysiological testing in patienuwith PSII may correlare wirh the fundal changes observed ~n clinical examination.Patients with chorioretinitis typically have normal or only mildly reduced amplitudein full-field electrorerinogram (ERG, both a- and b-wave), but normal implicittim.(30). Similarly, parients wirh localized inflammatory disease, such as toxoplasmosis,can also have a normal or only very mild reduction in full-field ERG with norm.implicit times (30,31). However, in diffuse and generally chronic inflammatory can.ditions of the retina and choroid, full-field ERG findings mote commonly showboth an amplitude reduction and an implicit time delay of the full-field ERGre-sponses (31). Inflammatory condi tions rhat lead to pigmemary degenerative chang~can mimic the fundus appearance of retinitis pigmemosa (RP) (see below); however,full-field ERG abnormalities tend ro be less marked (30). Furthermore, full-fieldERG abnormalities in patients with PSII may be unilateral or bilateral, in whichcasethey are often substantially asymmetric, in contrast to the typically symmetricalERG reduction and implicir time delay seen in RP.

The majority of pariems with birdshot choroidopathy have abnormal scotopicandphotopic ERGs and, in addition, the amplitude of the b-wave is frequently reduced[0a greater extent than that of the a-wave (electronegative ERG), suggesting dysfunctionof the inner retina (30,32). The presence of delay in the 3D-Hz flicker ERG t<.spons'in all patients wirh birdshot choroidopathy suggests that it is the most sensitive indio-tor of retina] dysfunction (32). Electrophysiological assessment of patients with birdshot choroidopathy is extremely useful for monitoring the response to treatment (e.g.,normalization of the b:a ratio and flicker implicit time). Subjective and clinicals~nsare, in contrast, poorer indicators of a therapeutic response (32,33). However, patternERG recovety after treatment does not always mirror the improvement in the full-fieldERG, due to the possible presence of persistent macular edema (32).

DIFFERENTIAL DIAGNOSIS

Although visualization of inflammatory cells in the anterior chamber and vitreouscavity on slit-lamp examination suggests rhe diagnosis of PSI!, this finding is noralways presem. Similarly, a typical "partem" of inflammatory infiltration such as thatobserved in POHS, PIC, birdshor choroidopathy, or other forms of PSI! may notU).ways be seen. Under these circumstances, the diagnosis of PSII may represent a ,hU)-lenge. TIllS IS particularly rrue when PERG and scotopic and photopic ERGs arereoduced (FIg. 12.3A). In the latter cases, there might be a misdiagnosis of inheritedretinal dystrophy, and no trearrnenr would then be considered for the patient. TheopposIte mighr also occur and patients could be treated unnecessarily. Fundus Af

CHAPTER 12

A

FUNDUS AUTOFLUORESCENCEIN POSTERIORUVEITIS 241

fiGURE 12.9. IA) Color fundus photographs and IBI fundus AF image of a patient with RPsine pigmento. Note the markedabnormalities In the distribution of (B) fundus AF compared with IAI the clinical findings Reduced amplitude in rod and coneresponseswith increased implicit time in the cone response and additional delay in the a-wave of the mixed rod-cone responsewere found on the full-field ERG.

may be helpful in establishing the correct diagnosis by demonstrating preservation ofthe mid-peripheral AF signal (Fig. 12.3A) or well-defined patches of reduced AF (Fig.12.3B) in cases of PSII and multiple smaller areas of reduced AF throughout the midperiphery in patients with inherited retinal dystrophies (Fig. 12.9).

SUMMARYFundus AF is a relatively new modality for fundus imaging based on a well-recognizedphenomenon from the early days of FA. Interpretation of the AF signal should be basedon a clear undemanding of the pathology of the condition under investigation. In thecase of uveitis, recent information about lipofuscin in microglial cells and macrophageshas added a new dimension to what was already understood abont RPE cell damageand loss in this condition. Fundus AF may be useful in monitoring disease progressionand establishing, noninvasively, the activity of disease in PSII. Furthermore, AF mayalso be helpful for evaluating patients who are allergic to FA or IeGA.

REFERENCES1. Forrester JVF, Okada A, BenEzra D, er al. Posterior Segment Intraocular Inflammation: Guidelines. TheHague: Kugler, 1998.

2. Arend 0, Remky A, Elsner AE, er aJ. Quantification of cystoid changes in diabetic maculopatby. InvestOphchalrno! Vis Sci 1995;36:608-613.

IIIIIII

d hi B al Uveiric macular oedema: correlation berween Optica!rnhT THdS MDBoagl ,er.. "Verence3. ran , e met '~L' 1 "rv and fluorescein angIOgraphy. Br J OphrhalmoJ 2008;92'~"2all

hy patterns WIUI vrsua acurry C f oarienrs w; h . ~ -J ..tomograp . . T M S er a.I. Reading perrormance 0 panents Wit uveitis-asso',4. Kiss CG, Barisani-Asenbauer , aca , 4 620---624 Clatf«cystoid macular edema. Am J Ophthalmol 2006; 1 2: . f ul d . .

. . N Halki daki 1 Pantelia E, er a1. Patterns 0 mac ar e ema In panenn wirh uvej"..5 Markomichelakis N, as, . h h 0 hthal I ~. .. d .. ssmenc using Optical co erence romograp y. p rna ogy2004'1l1~qualitative an quantitatIve asse I

946-953. fl . . hi d ical6 Kan SW, Park CY, Ham D1. The correlation ~erween uorescem anglograp rc an opt! coherencell>-. g hi C . 1" ically sionificant diabetic macular edema. Am J Ophthalmo! 2004;137:313-322.mograp IC rearures In C 1111 e 'c.' fh fl 1 r.,

. . PG T ka M t a.I Dereccion and quantHlcanon 0 yper uorescenr ea""!>ebyco!ll7. Phillips RPai"Rof"fu d'us ~:oresc'e~n al~giograms, Graefes Arch Clin Exp aphtha/mol 199Ii229:329_3Ji'

purer an ysrs 0 n . . C' 'fy 1 I c__ .Philli RP R PG Sh PF er ai Use of remporallOlormatlon to guano vascu ar ea''''lJeinfluores.-8 I IpS ,oss , arp ". . 1 ).

.. . h f rhe retina Clin Phys Physiol Meas 1990,11(Supp A .81-85,cern anglOgrap yo. b 'cal h

9. ~tc~~o~~c:~:f:~i~;:p~:a~ra1~~S'd::e~ri;no:rr~~~ider::;w~t1oed::r~~cep:~:~~ra~v~~anu~.:

Ophthalmology 2000;107;593-599. . . .M T AI b i M M han G er al Relationships between clImca1 measures of vISUalfunction flu(\.10. outfay , ar I ,a , . . " '

. . h' and optical coherence tomography features III paneors with subfoveal chol£lidalrescem anglOgrap ICneovascularisarion. BrJ Ophthalrnol 2008;92:361-364., . ..

11. Xu H, Chen M, Mayer E], et al. Turnover of resident reunal microglIa m the normal adult mouse,GI~2007;55;1189-1198.

12. Xu H, Chen M, Manivannan A, et al. Age-dependent accumularion of lipofuscin in perivascuJarandlUl>-retinal microglia in experimental mice. Aging CeU 2008;7:58-68.

13. Forrester JVF. Intermediate and posterior uveitis. Chern Immunol Allergy 2007;92:228-243.14. Jiang HR, Hwenda L, Maldnen K, et al. Sialoadhesin promores the inflammatory response in experimen_

tal autoimmune uveorerinitis. J ImmunoI2006;177:2258-2264.15. Jiang HR, Lumsden L, Forrester ]VF. Macrophages and dendritic cells in IRBP-induced experimemaJau.

toimmune uveoretinitis in BIORIII mice. Invest Ophthalrnol Vis Sci 1999;40:3177-3185.16. Lim W'K, Chee SP, Sng I, et al. Immunopathology of progressive subretinal fibrosis: a variam of sympa.

thetic ophthalmia. Am J Ophthalmol 2004; 138:475--477.17. ]akobiec FA, Marboe CC, Knowles II DM, et aI. Human sympathetic ophthalmia. An analysisOflh~

inflammatory infiltrate by hybridoma-monoclonal antibodies, immunochemistry, and correlariveelec-tron microscopy. Ophthalmology 1983;90:76-95.

18. Deeg CA, Raith AJ, Amann B, et at CRALBP is a highly prevalent auroanrigen for human auroirnmumuveitis. Clin Dev ImmunoI2007:39245.

19. Wang M, Bai F, Pries M, et aJ. Identification of MHC class I H-2 Kb/Db-restricted immunogenicpep'tides derived from retinal proteins. Invest Ophthalmol Vis Sci 2006;47:3939-3945.

20. Umeda S, Suzuki MT, Okamow H, et al. Molecular composition of drusen and possible invoJvememofanti-retinal autoimmuniry in f\¥O different forms of macular degeneration in cynomolgus monke)'(Macacafascicularis). FASEB] 2005;19:1683-1685.

21. Janssen J], Janssen BP, van Vugt AH. Characterization of monoclonal antibodies recognizing wina/ pig.ment epithelial antigens. Invest Ophthalmol Vis Sci 1994;35:189-198.

22. Wang D, Yu QC, Schroer ], et aJ. Human cytomegalovirus uses rwo distinct pathways to enrer retinaJpig.mented epithelial cells. Proc Natl Acad Sci USA 2007; I 04:20037-20042.

23. Cai S, Bra11dt CR. Induction ofinterleukin_6 in human retinal epithelial cells by an attenuated Herpesrim-plexvirus vector requires viral replication and NFkappaB activation. Exp Eye Res 2008;86:178-188.

24. Liu B, Li Z, Mahesh sr, et al. HTLV-l infection of human retinal pigment epithelial cells and inhjbici~nof viral infection by an antibody to ICAM-I. Invest Ophthalmol Vis Sci 2006;47:! 51D-1515.

25. Haamann P, Kessel L, Larsen M. Monofocal outer retiniris associated wirh hand, foot, and mourh diseas~caused by coxsackievirus. Am J Ophthalmol 2000; 129:552-553.

26. Kadrmas EF, Buzney SM. Coxsackievirus B4 as a cause of adult chorioretinitis. Am J Ophtha/mol1999;127;347-349.

27. McBain VA, Forrester]V, Lois N. Fundus autofluoresence in the diagnosis of cysroid macular oedema.BrJ OphthalmoI2008;92;946_949.

28. Koizumi H, Pozzoni MC, Spaide RF. Fundus autofluorescence in birdshot chorioretinopadl)',Ophthalmology 2008; I 15:elS-20.

29. Haen SP, Spaide RF. Fundus autofluorescence in multifocal choroiditis and panuveids. Am] Ophrhalmol2008;145;847-853.

30. Fishman GA: Birch DG, Holder GA, et at. Electrophysiologic Testing in Disorders of the Retina,Op~cNerve and VISUal Pathway. 2nd ed. San Francisco: American Academy of Ophthalmology, 2001.

31. Scholl HPN, Zrenner E. Electrophysiology in the investigation of acquired retinal disorders. SUNOphthalmoI2000;4529-47.

32. Holder GE, Robson AG, Pavesio C, et al. Electrophysiological characterisation and monitoring in theman~gement of birdshot chorioretinopathy. Br J Ophthalmol 2005;89:709-718.

33. Sobnn L, Lam BL, Liu M, et al. Elecrroretinographic monitoring in birdshot chorioretinographr. Am!Ophth,lmoI2005;140;52_64.

242 SECTION II CliNICAL SCIENCE

Fundus Autofluorescence in CentralSerous Chorioretinopathy

1':1~ntral sero~s ,chorio,retinopat.hy (esc) is a common retinal disease character-~ ized by an idiopathic flat retinal detachment within the macula (1,2). Ir ryp-

icallyaffects young and middle-aged adults between 20 ro 50 years of age andoften reveals a shallow, round, and serous detachment of the neurosensory retina;however, small detachments of the rerinal pigment epithelium (RPE) may also occur(2). Primarily male patients (male:female ratio about 10:1) are affected and rypicallya type-A behavior in these patients can be observed. Moreover, emotional stress fre-quently accompanies the visual disturbances. esc has also been associated with vaso-constrictive agents, endogenous hypercorrisolism, and systemic corticosteroid use (3).

"When the serous detachment involves the foveal region, patients become symp-rornatic and usually complain about blurred vision, scotoma, micropsia, or meta-morphopsia. This can easily be detecred by Amsler Grid resring. Decreased visualacuiry can be improved by rhe addition of small hyperopic correction, focusing rhelighr bundles ro the detached central region. Of interest, visual acuity remains largelypreserved despite the prolonged separation of the neurosensory layer from the RPE.The long-term visual prognosis for most patients is excellent and improvement canusually be achieved without specific treatment. However, about 20% to 30% of pa-tients will have one or more recurrences, and a small percentage (about 5%) will de-velop choroidal neovascularization or chronic detachment with cystoid macularedema from this condition (4,5). Chronic forms of CSC are characrerized by multi-ple sites of prolonged and recurrent serous retinal detachments in one or both eyes,and are particularly seen in Latin and Oriental people (5). Such parients may beasymptomatic for a prolonged time if the localized areas of retinal detachment areoutside the foveal area. On biomicroscopy, multiple areas of RPE atrophic tracts,particularly in the inferior site of the macula and in the peripapillary regions, are ob-served (6,7). Angiography then reveals multiple sites of staining rhar correspond tothe areas of RPE atrophy; however, no significant leakage is observable in these areas.Even though in these patients the macula is usually attached, the photoreceptors arechronically damaged due to previous long-term dysfunction. Thus, these patientsusually suffer from significant loss of vision and paracentral visual field defects. Inthe case of chronic detachment, lipid exudates and cystoid macular edema mayoccur, complicaring rhe disease (5). Whereas focal laser phorocoagulation is recom-mended in acute esc if no resolution of exudates appears after 4-6 weeks, in theatrophic stages of the disease no treatment can be offered. However, in chronicrecurrences with prolonged or repetitive serous retinal detachments, laser photocoag-ulation can improve the visual course. A faster resolution of the edema and a fasterrehabilitation of visual acuity are then observable; however, there is often no substan-tial benefir with regard to visual acuiry following laser photocoagulation. Usuallylaser treatment is directed to the site of leakage. Laser therapy induces damage of theRPE layer wirh migrarion and proliferation of neighboring RPE cells to cover the

243

defect resulting in a resroration of the outer blood-retina barrier (8-10). As ar~ul,of this biologic tissue reaction from the laser phorocoagularion, the neurOsenSilryretinal detachment disappears and the visual acurry recovers.


PATHOLOGY OF CSCIn contrast ro the well-defined clinical appearance of ese, a clear understandingo!the exact pathogenesis of accumulation of subrerinal fluid is lacking. It is widely".cepred that rhe origin of the subretinal fluid is rhe choroid. Because of a defectinmeRPE layer, choroidal fluid enters rhe subretmal space and leads ro the detachmemofthe neurosensory layer (I). ese was induced in monkeys by repeared injectionso!epinephrine (II). Hisrologic examinarion of rhe monkeys' eyes revealed foe:"RPEde.generation and endorhelial cell destruction in rhe underlyrng chonocapdlary layer(12). This supports rhe generally adopred opinion thar rhe RPE plays a crucialrolei,rhe development of eSc. Measurements of rhe merabolic activity ofRPE cellsahore.vealed significant changes in ese (13). The cause of rhe focal leak is unclear.It."inirially proposed that a simple breakdown of rhe RPE layer is responsible formeleak (I4). Larer, rhe theory of parhologically hypersecrering RPE was proposed(15);however, this did nor explain rhe observation of widespread hyperpermeabiliry in theareas of neurosensory detachment seen with indocyanine green (leG) angiograpoy(16,17). In facr, leG suggested impaired choroidal circulation as a cause of esc oyshowing delayed choroidal capillary filling in areas of hyperpermeability (18). It ""proposed that localized capillary and venous congesrion in disrincr areas leadsrois.chemia, increased choroidal exudation, and a focally hyperpermeable choroid.Because of rhis excessive choroidal fluid extravasation, the RPE deraches and afterfur.ther accumularion of fluid, breaks wirhin rhe RPE appear, allowing the fluid to crearea neurosensory retinal detachment (19). However, earlier bur limired hisropathologirexaminations in humans showed no abnormalities in the choriocapillaris underlyingthe RPE detachment (20). On the other hand, ir was nored that the gray-whiteexu-dates contained fibrin, which was taken as evidence that serum proteins escapedfromthe choriocapillaris. This supports the hypothesis that a focal increase in the permesbi-ity of rhe choriocapillaris is the primary cause of damage to the overlying RPE leadingto disrinct breaks and subsequent neurosensory derachrnenr (2,5).

IMAGING TECHNIQUES IN THE DIAGNOSISOF CENTRAL SEROUS CHORIORETINOPATHYBiomicroscopyThe diagnosis of ese is primarily clinical and usually confirmed by angiography.Biomicroscopicajlc, and best seen with a narrow lighr beam from the slit-lamp, awell-defined round or oval area of shallow elevation of the rerina, which usually presenuaslightly darker color than the surrounding normal retina (5), can be observed. Thederached retina is usually transparent and of normal rhickness and the subretinal Auidis also usually dear; however, sometimes gray-white serofibrinous exudates canbeseen. Because of the retinal detachment, rhe visibiliry of rhe xanthophyll pigmenrwithin rhe center of the fovea may be enhanced, presenting as a central yellowspoc.Also, through gravlry, the subretinal fluid often pools inferiorly within rhe areaofceti.nal detachment (5). Somerrmes small dot-like deposits on the inner side of the retina

CHAPTER 13 FUNDUS AUTOFLUORESCENCE IN CENTRAL SEROUS CHORIORETINOPATHY 245

or on the RPE surface within the detached area can be seen, most likely representingprotein ptecipitates(2).

FluoresceinAngiographyFluoresceinangiography (FA) plays an important role in the evaluation of CSC. It isused to detect the distinct site of one or more RPE breaks and to determine the amountof leakage,which can be very heterogeneous. Thus, some patients revealonly small de-tachments and Jessangiographic leakage, whereas others presenr with large derachmentsand heavyleakage. In FA, usually the dye from rhe choroid leaks rhrough the focal RPEdefect and pools in the subretinal space. In more than 95% of patients with CSC, atleastone leaking point can be seen. Typically, the dye spreads symmerrically in the sub-retinal space but does not extend outside the borders of the detachment. Sometimes theclassic"smokestackphenomenon" can be observed showing the dye percolating upwardin the subretinal space with subsequently pooling into rhe whole space. This pattern,firstdescribed in 1971 (21), is thought ro result from an osmotic pressure gradient gen-erared by differences between the protein concentration of the subretinal fluid underthe detachment and the fluorescein dye entering the derachment (13).

Indocyanine Green AngiographyTogether with FA, ICG angiography is an important imaging technique for the diag-nosis and follow-up of parienrs with CSc. For cases in which FA findings are not ryp-ical of CSC, ICG angiography may be helpful in establishing the diagnosis, often re-vealing multifocal choroidal hyperfluorescence in affected and unaffecred regions ofactive and fellow eyes (17,22). These hyperfluorescence areas have been hypothesizedro be causative factors in the pathogenesis of CSC and may not be observed by FAalone. The lCG is a larger molecule highly bound to proteins and therefore does notleak extensively through the fenestrations of the choriocapillatis (13). Thus, choroidalvasculature can be observed in detail, in contrast to fluorescein, which leaks rapidlyand easily through the fenestrations, immediately obscuring the choroidal vascula-ture. One study (13) reported that all patients with CSC examined had mulriplebright area,"of choroidal hyperfluorescence up ro three disc areas around the leakagepoints and elsewhere, and the boundaries were independent of the neuroretinal de-tachment. Of interest, these areas of hyperfluorescence were also noted in eyes withinactive disease (22) and after resolution of the disease (13). Frequently observed de-tachments of the RPE (PED) accompanying CSC showed distinct characreristics inICG angiography as a pooling of the dye in the late phase in the particular area ofPED forming a hyperfluorescenr ring (13).

Optical Coherence TomographyOptical coherence tomography (OCT) has been found to be a useful tool in the diag-nosis of CSC (23). Sometimes a slight elevation of the neurosensory retina is not de-tectable on slit-lamp biomicroscopy; however, such elevations can easily be detectedby OCT and usually overlie an optically dear fluid-filled caviry.The mostly attachedunderlying RPE can be observed as a highly reflective band at the base of the caviry.Furthermore, additional PEDs can be easily seen on OCT as localized elevations ofrhe highly reflecriveband over a dear caviry.The detached RPE then causes artenua-rion of the reflecredlighr, resulting in extensive shadowing of the underlying choroidalsignal (24). OCT may also be helpful in defining anatomic changes of the detached


neurosensory retina as thickening of the ~e(jna or intraretinal ~stic changes in CSc(23). Moreover, OCT is able to disringuish berween serous retinal detachmentandother pathologies, such as choroidal neovascula~:nembrane, seen as a disrinctthieken_ing of the outer rerinallayers in the latter condmon (24). However, a cleatdifferenti~diagnosis for this condition should be obtamed by angiography; OCT isbestused wmonitor rhe disease during the follow-up penod because It can reveal the resolutionorthe subretinal fluid by a reatrachmenr of the neurosensory layer.For studyPUtpos~,the retinal thickness can even be measured to obtal~ quanutanve follow-up valUtiafter interventions. When changes from acute ro chronic CSC appear, enhancedOcrsignals at the outer surface of the neurosensory retina can be observed, indicatingfi~rinous precipitates accumulating in this area (25).

Fundus AutofluorescenceIn patienrs with acute esc, a focal area of decreased autofluorescence (AP)istypi.cally seen at the site of the focal leakage on FA, and a reduced AF signalisalsoo~served within the enrire area of serous retinal detachment (Fig. 13.1) (25-27).

In chronic-recurrent eSc, the focal leakage point and/or the areaof an~o.graphic hyperfluorescence most commonly demonstrates a decreased APsignal(27). In some patients, these areas of reduced AF signal are surrounded byzonewith mottled, irregular AF, with foci of increased or decreased AF. In a smallgroupof patients, no abnormal AF, in comparison with that of the background,canbedetected at the leakage point. However, an increased.AF signal is most oftenseenin

A

C ---------

FIGURE 13.1. A 36-year-old male presentedwitha~Uvisual loss in his left eye for 3 weeks. Visual acuitywasOBFundus AF revealed a sharply demarcated area01d~creased AF (A) corresponding to the angiographically."·ble blockage of fluorescence IB).ICI A neurosenso~r'tin,detachment is confirmed by OCT(vertical scanl Anarea"further decreased AF is also observed correspondingloti:!RPE leaking point on FA.

CHAPTER 13 247FUNDUS AUTOFLUORESCENCE IN CENTRAL SEROUS CHORIORETINOPATHY

A

FIGURE 13.2. This 38-year-old woman complained of de-creased vision for several months. Visual acuity was 0.8. (AI AFrevealed significantly increased intensity. especially at the bot-tom area of the slight neurosensory retinal detachment. indi-cating long-standing detachment (8) Corresponding FA show-ing only minimal hyperfluorescence just superonasally to thecenter of the fovea. OCT (horizontal scan) confirms the pres-ence of a neurosensory retinal detachment (C).

C

the area of presumed former or residual neurosensory retinal detachment (Fig.13.2) (25,27). Fundus AF is extremely helpful in the noninvasive diagnosis ofchronic esc, demonstrating multiple areas of mottled, increased, or decreased AFsuggestive of previous episodes of active disease (25,27,28). In these cases, onlyminimal ophthalmoscopic changes may be present, and OCT imaging usuallyshows only slight abnormalities as little sub retinal fluid (Fig. 13.3).

INTERPRETATION OF AF FINDINGS IN escIn the acute phase of esc, the reduced AF signal observed at the leaking point couldbe rhe result of the ptesence of damaged RPE, which may, at least pattially, explainthe occurrence of leakage from the choroid into the subrecinal space. Alternatively, itcould be related to a blockage of the AF signal by the presence of subretinal fluid. Thereduced AF signal at the site of the subretinal fluid is most likely related to the block-age of the AF signal caused by the subrerinal fluid.

In chronic-recurrent esc, RPE atrophy may contribute to the decreased AF ob-served at the former leaking point. The increased AF observed in areas of presumed fat-mer neurosensory retinal detachment may be the result of a higher metabolic activity ofthe RPE leading to a higher phagocytosis rate of debris from the subretinal space(27,28) and shed photoreceptor outer segments (25). However, an accumulation of flu-otophores within the serous fluid may also contribute to this enhanced AF. The low AFsignal observed in cases of chronic-recurrent esc in some areas (Fig. 13.3) may be theresult of reduced metabolic activity, presumably due to photoreceptor ceUloss in areasof chronic neurosensory retinal detachment (25,28).


A

cFIGURE 13.3. (A) This 51-year-old male presented with decreased visual acuity of OJ in both eyes with onlyminimalchanges in ophthalmoscopy. The patient had experienced recurrent visual disturbances throughout the past years.IBI fA re-vealed hypedluorescence in the peripapillary region and centrally, but no active leakage. (Clln OCTsections, no signili~nledema was noted. (0) FundusAF disclosed reduced AF signal in the peripapillary area, compatible with RPEatrophy,atthesri'of angiographic hyperfluorescence. Perifoveolar increased AF was also observed, suggesting former areas 01neurosen,,~retinal detachments that led to RPEchanges. The diagnosis of chronic-recurrent CSR was made.

RELATIVE IMPORTANCE OF ANGIOGRAPHY, OCT,ANDAF FORTHE DIAGNOSIS OF CSCFrom a clinical point of view, in most cases, the correct diagnosis of acute esc can beestablished by history and slit-lamp biomicroscopy, showing the typical significanrround Or oval shaped neurosensory retinal detachment within the macular area. In theabsence of other techniques, fundus photography may provide some objectivebasisfor follow-up of the natural history of CSc. However, FA is an appropriateandnecessary investigation to confirm the diagnosis of CSC and to determine thelOG!-tion of the RPE leak(s) during the first episode of the disease, especially whenIf''''me.fit is considered, and also to obtain information on the presence of activedis~as Judged by the leakage of dye at the area(s) of RPE damage. lCG angiographyISless informative concerning the activity and prognosis of the disease, OCT ~Iows

CHAPTER 13 FUNDUS AUTOFLUORESCENCE IN CENTRAL SEROUS CHORIORETINOPATHY 249


evaluation of the amount and extent of subretinal fluid, and is a useful tool for follow-up in patients with esc, allowing monitoring of the resolution of the neurosensoryretinal detachment and PED(s) if ptesent. In patients with worsening symptoms, a re-peat FA may be useful to detect possible new areas of RPE leakage.

In cases of acute CSC, AF may be able to detect the distinct leakage poinrfs] bydemonstrating focal areas of reduced AF levels. In chronic-recurrent esc, AF im-aging is of significant value, especially considering that these cases often represent adiagnostic challenge. As shown in Figure 13.2, well-defined areas of increased AFcan be detected in chronic cases. Similarly, multiple areas of mottled, increased, ordecreased AF can also be observed, unilaterally Ot bilaterally, strongly suggesting thediagnosis of chronic-recurrent CSC (Fig. 13.3). These findings are important, par-ticularly when no subretinal fluid is present, since in these cases OCT might fail toestablish the diagnosis of csc. In the latter cases, AF is invaluable as noninvasivediagnostic tool to establish the diagnosis of esc. In the presence of subretinal fluidin chronic cases, which can be observed on OCT images, precipitates that demon-strate increased AF signal can often be seen at the level of the outer retinal layer(25). Furrher evidence of long-rerm chronic CSC can be provided by the presenceof areas of reduced AF signal in the peripapillary zone and exrending inferiorly, in-dicating damage and atrophy of the RPE from the previous presence of subretinalfluid in these areas.

In principle. serous elevations of the neurosensory retina in the macular areacan be pro-duced by diseases of the choroid, the RPE, and the retina itself The following diseasescan produce localized serous detachments similar to those observed in CSC: congenitalpit of the optic nerve, malignant hypertension, toxemia of pregnancy, Harada disease, id-iopathic uveal effusion, vitreomacular tractional syndrome, and age-related macular de-generation (AMD) (5). One of the most important diseases to rule out, particularly inolder patients, is neovascular AMD, especially now when more effective treatments forthis condition are available. The presence of a drusen supports the diagnosis of AMDrather than CSc. FA will allow the differentiation between CSC and exudative AMD.

SUMMARYDistincr Af patterns can be observed in patients with acute and chronic-recurrent esc.Wheteas acute CSC is usually suspected by slit-lamp biomicroscopy and diagnosedusing FA, the diagnosis of chronic-recurrent CSC is often challenging. Combined AFand OCT may have the potential to replace invasive angiography in the diagnosis ofacute eSc. AF may allow recognition, in some patients, of the point of RPE leakage,which cannot be distinguished in OCT. If that is not possible, FA will still be needed.AF seems to be particularly helpful in the diagnosis of chronic-recurrent stages of thedisease, which sometimes may be difficult to differentiate from other pathologic condi-tions because ofthe poorly defined and mottled, partly mulrifocal angiographic leakage.The AF patterns of increased AF levels and findings of RPE atrophy (mottled and de-creased AF), as described above, provide strong evidence to support the diagnosis ofchronic-recurrent CSc. Follow-up changes after treatment can also be easily detected byAF and usually correlate well with OCT findings (Fig. 13.4). Thus, fundus AF alsoserves as an appropriate tool to monitor the resolution of esc during follow-up.

CHAPTER

•Fundus Autofluorescence InFull-Thickness Macular Holes

n n idiopathic full-thicknessmacular hole (FTMH) is a defect in the neurosen-Ka sory retina at the fovea, from the internal limiting membrane to the outer seg-ment of the photoreceptors. The first reported case of a macular hole, de-

scribed by Knapp (1) in 1869, occurred in a patient who had sustained prior blunttrauma to the eye. Subsequent case reports and series pointed to an antecedentepisode of ocular trauma prior to FTMH formation, such that the rwo were linked toeach other. It later became clear that most cases occur spontaneously (idiopathicFTMH) and few are associatedwith trauma to the eye.

Patientswirh FTMH usuallycomplain of blurred vision and/or meramorphopsia.As the FTMH becomes larger, patients become aware of a central scotoma. Some,however,may be asympromaric, and rhe FTMH will be diagnosed during a routineeye examination.

The visual acuity of a patient with FTMH varies according to the size, location,and stage of the macular hole. Patients with small, eccentric holes may retain excellentvisualacuity in rhe range of20/25 to 20/40. However, in mosr cases rhe visual acuityvariesfrom 20/80 to 20/400, with an averagevision of 20/200.

A FTMH visualizedwith direct ophthalmoscopy appears as a well-defined round orovallesion in the cenrerof the macula, often with yellow-white deposits at the base. Onslit-lamp biomicroscopy, a round or oval excavation with well-defined borders inter-rupting the beam of the slit-lampcan be observed, surrounded in most cases by a cuffof subretinal fluid (neurosensoryretinal detachment). An ovetlying semitranslucent tis-sue representing a true operculum or a pseudo-operculum can be seen over the hole.Cysticchangesof the retinamay be evident at the margins of the hole. The retinal pig-ment epithelium (RPE) is usuallyintact at the site of the hole, although in long-stand-ing FTMH it may appear atrophic or hyperplastic.Fine wrinkling of the inner retinalsurfacecausedby the presenceof an epiretinalmembrane may also be detected.

The Watzke-Allentest (2) has been widely used as a diagnostic test to distinguishFTMH from other lesions, such as lamellar macular holes (LMHs) and macularpseudoholes (MPHs). In this test, a thin beam oflight is projected over the area of theholewhile the patient is askedwhether he sees the beam being broken ot intact. It wasassumed that most patients with FTMH would see a broken beam of light, becauseof the corresponding lack of tissue at the site of the macular hole. However, a recentstudy suggestedthat, in fact,most patients with FTMH did not repon a broken beamoflight but instead just thinning of the beam oflight (3).

In 1988, Johnson and Gass (4) described a classification for idiopathic FTMH.Gass (5) recently updated this biomicroscopic classification as follows: In the firststage (stage l a}a yellowspot 100-200 fLmin diameter, resulting from a foveolar de-tachment secondary to spontaneous tangential traction by the prefoveolar vitreouscortex, is observed on slit-lamp biomicroscopy.The yellow Spot is presumed to becaused by the ptesence of inttaretinal xanthophyll pigment, which becomes more

CHAPTER 14A FUNDUS AUTOFLUORESCENCE IN FULL-THICKNESS MACULAR HOLES 253

--""'"!J"~ e as a result of the foveolar detachment. In stage lb (occult hole), the yellow spotis transformed into a doughnut-shaped yellow ring of approximately 200-300 u.m insize centered on the foveola. The visual acuity in stage 1 lesions ranges typically from20/25 to 20/70 and there is often some degree of metamorphopsia. The first evidenceof the presence of a full-thickness retinal defect occurs in stage 2 holes, which are de-fined as holes :0;400 urn. Most stage 2 holes progress to stage 3 holes (>400 p.m). Inboth stages 2 and 3 there is an absence of complete posterior vitreous detachment.W'hen complete separation of the vitreous from the entire macula and optic disc oc-curs, the hole is classified as a stage 4 FTMH, independently of its size.

HISTOPATHOLOGYAND PATHOGENESISHistopathology studies frequently have demonstrated cystic spaces in the outer plex-iform and inner nuclear layers in patients with FTMH (6). There is also frequent glialproliferation from' the edges of the macular hole over the inner retinal surface aroundthe hole (5). Nodular proliferations of the RPE, at the RPE level, can be also found atthe site of the hole (5). In the majority of cases, associated epiretinal membranes areseen (6). A variable degree of phororeceptor cell degeneration at the margins of thehole has also been observed (6).

One immunocytochemisrry study (7) demonstrated that, in addition ro glialcells, photoreceptor cells (cones) were usually present in the operculum lying in frontof the macular hole (true operculum). In some cases, however, only glial tissue waspresent (pseudo-operculum). It was also shown that eyes with opercula containingmore than five phororeceptors were associated with higher anatomical failure (themacular hole remained open after surgery) compared to those in which less man fivephotoreceptor cells were found (7).

Anteroposterior and tangential vitreomacular traction has been suggested as apossible mechanism in idiopathic FTMH formation (8-1 I). Furthermore, the role ofthe internal limiting membrane facilitating the proliferation of cellular components,which could cause tangential traction around the fovea and FTMH formation, hasalso been postulated (for review see Abdelkader and Lois [12]).

IMAGING TECHNIQUES

Fluorescein AngiographyThe diagnosis ofFTMH is usually made by history and slit-lamp biomicroscopy with anoncontact or contact lens. However, occasionally it may be difficult to differentiate be-tween a FTMH and an MPH or LMH (see also Chapter 14B). In such cases fluoresceinangiography (FA) can be used to help in this differentiation. FA in FTMH stages 2 and3 typically discloses hyperfluorescence in early names, with no leak in late frames (win-dow defect). However, FA is invasive and carries potential risks for adverse reactions(13). Optical coherence tomography (OCT) and fundus autofluorescence (AF) imageshave now replaced FA in the evaluation of patients with FTMH (see below).

Optical Coherence TomographyOCT allows the physician to noninvasively detect the presence of an FTMH andchanges in the surrounding retina (Fig. 14A.I) (14). Cystic spaces in the retina and


FIGURE 14A.l. OCT(90 degree cut) obtained from an eye with an FTMH. Cystic spacesin theretioaanaan oparculum/pseudo-uaerculum overlying the area of the defect are seen.

a neurosensory rerinal detachment surrounding the hole are usually visualized(Fig.14A.2). In addition, the status of the vitreomacular interface can be evaluated(Fig.14A.2) (10,14).

OCT can be used to determine early macular hole closure followingsurgery(24hours postoperatively) (15). However, in many cases it may be difficult to obtainappro-priate images because of the presence of gas in the eye, especially in pseudophakicejes

Three different OCT patterns were described after what was considered to beasur-cessful surgical repair of an FTMH (macular hole no longer visible or macularholestillvisible but with disappearance of the neurosensory retinal detachment around it),whichcorrelated with postoperative vision (16). These were described as a U pattern (norm~foveal contour), a V pattern (steep foveal contour), and a W pattern (persistenceofaneurosensory retinal defect at the site of the hole but with lack of associatedneurosea-sory rerinal detachment). Postoperatively, rhe highest levels of vision were recordedinthe former, and the lowest were recorded in the latter (16). Similarly, it wasrecentlyshown that the presence of a normal inner segment and outer segment junction inOCTimages at the site of the hole postoperatively was associated with good visual recoveryfollowing surgery (17).

OCT can also provide information in addition to that gathered by slit-lampbie-microscopy, especially in early stages of macular hole formation. In earlystagesof

FIGURE 14A.2. OCT (330 degree cutl obtained from an eye with an eccentric FTMH. Thevitreousr"mains adherent to the retina at the edge of the hole. Small cystic changes around the holearepresent-

CHAPTER 14A fUNDUS AUTOfLUORESCENCE IN fULL-THICKNESS MACULAR HOLES 255

development (stage 1), OCT demonstrates in most cases a macular cyst, rather thana foveolar derachment, as proposed by Gass (18). Furthermore, rhe presence of a nor-mal foveal contour and normal retinal thickness but a preretinal, minimally reflective,rhin band inserring obliquely on ar least one side of rhe fovea (which has been rermeda srage 0 macular hole) was found ro be a significant risk facror in fellow eyes for rhedevelopment of an FTMH (19).

Alrhough OCT demonstrates very well the full-thickness neurosensory retinal de-fect in patients with FTMH, in some cases it may not be able to detect whether lossof inner retina has occurred, and thus whether an LMH is present (see Chapter 14B).

In the first study ever published on the distribution of fundus AF in patients withmacular holes (20), were evaluated AF images and rhe corresponding color fundusphoto graphs and FAs of the affected eye and AF images of the unaffected, ccntralat-eral eye. The AF intensity at the sire of rhe macular hole was compared with rhar arthe corresponding area iq. the unaffected eye. It was found that in some patientswirh srage 1 FTMH the distribution of AF and rhe corresponding FA was normal;in some, however, rhe foveal AF signal appeared slightly increased compared ro thecontralareral eye. Similarly, rhe FA in rhese latter cases showed mild central hyper-fluorescence in the mid phase of the angiogram. All stage 2 FTMH demonstraredan increased AF signal at the site of the hole (Fig. 14A.3); a window defecr that cor-responded exactly in location, size, and shape with that seen on AF imaging was de-rected on FA. This was also the case in srage 3 and 4 FTMHs. These findings canbe explained as follows: It has been demonstrated that the AF signal derives pre-dominantly from the lipofuscin in the RPE (see Chapters 3 and 9) (21-23). Thissignal is attenuared at the center of the macula by the presence of the luteal pigment(see also Chapter 14B) (23). In FTMH, there is no neurosensory retina at thesite of the hole, and thus there is no lureal pigment overlying the defect, An intenseAF signal is subsequently observed at rhe hole.

FIGURE 14A.3. fundus Af image of a patient with fTMH (A) Note the increased Af signal at the site of thehole and a mildly increased Af signal around the hole m a "petalloid" pattern. The latter is due to the presence

of cystoid spaces in the retina. as demonstrated by OCT (180 degree cut) (B).

Fundus Autofluorescence

A B

The cuff of neurosensory retinal detachment surrounding the macular holedemOIl_strared a reduced AF signal (Fig. 14A.4A). In contrast, on FA the cuff of subretinalfl .appeared hyperfluorescenr in the majority of cases; in about a third it disclosedhlP fl~d. d d 0 u-orescence. Color photographs and FA were exarrune ro etermine whether anycetin.elevation could be detected beyond the cuff of the subretinal fluid surroundin thmacular hole. Shallow retinal elevation extending beyond the cuff was seen inabou:h':of the cases; at this site, reduced AF signal was observed, although to a lesserextentthanat the site of the cuff of sub retinal fluid. The cuff of sub retinal fluid and the retinald.varion extending beyond rhis cuff demonstrated reduced AF, likely due to the effiecaused by the presence of subretinal fluid and/or thickening of the neurosensory .Oq. retmawhich would attenuate the AF signal coming from the RPE. '


A

o

FIGURE 14A.4. FundusAF images obtained from patients with tcreased AF signal at the site of the hole (A C) ad' f a sage 3 FTMH (A) and a stage 2 FTMH (B). Noleme 10'

• n a nng a reduced AF' Isubretinal fluid (AI. IC) A mildly increased AF sign I d signa around the hole that corresponds10aCuffDI

.. a aroun the hole with" II'" .spondsto cystic spaces In the retina. (B 0) After suc f I a peta Old appearance IS also seenthalcOml', cess u macular hole hsurgery, t e normal AF signal is restored.

C

CHAPTER 14A FUNDUS AUTOFLUORESCENCE IN FULL-THICKNESS MACULAR HOLES 257

The presence of a preretinal operculum or pseudo-operculum was demonstratedby the presenceof a mobile disc-like area of reduced AF signal overlying the area ofincreased AF signal corresponding to the macular hole or its surroundings. This wasinterpreted as the shadow casted by the operculum on the RPE. The presence of theoperculum or pseudo-operculum could not be documented by FA.

Afrer successful surgery, the high-intensity AF signal and rhe hyperfluorescenceon FAat the hole disappeared (Fig. 14A.4). This suggested that the RPE was coveredby retina and/or glial tissue, as previously demonstrated histologically (24), againblocking the AF signal from this layer. Similarly, the cuff of subretinal fluid was nolonger visible on AF imaging or FA after successful surgery.

Our findings on the distribution of fundus AF in FTMH, as described above,were confirmed more recently by other researchers (25,26).

There is one limitation of fundus AF in the diagnosis of FTMH: the differentia-tion between a LMH and a FTMH. As explained in detail in Chapter 14B, in bothLMH and FTMH, an increased AF signal would be detected because of the completeor partial loss of neurosensory retina and, subsequently, of luteal pigment at the siteof the FTMH and LMH, respectively. However, the presence of a cuff of subretinalfluid around the defect visualized on fundus AF as a well-defined area of reduced AFsignal would point roward the diagnosis of FTMH.

COMPARISON OF AVAILABLE IMAGING TECHNIQUESPatients with FTMH can be imaged before and after surgery. The most widely usedimaging techniques for FTMH include color fundus photography, FA, OCT, andfundus AF.The resolution of color fundus photographs may be insufficient ro allowfor a consistent and reliable recognition of small macular holes or to demonstrate thepresence of a preretinal operculum/pseudo-operculum. On FA, FTMH is usually im-aged appropriately; however, the disadvantage of this imaging technique is that it is aninvasive procedure that carries a small but significant risk of morbidity. In addition, itis time-consuming and requires the presence of an experience photographer and anurse or doctor. The images may not be instantly available unless a digitized imagingsystem is used. The combination of fundus AF and OCT appears to be the best avail-able method to evaluate patients for whom a diagnosis of FTMH, MPH, or LMH isentertained (see also Chapter 14B):

SUMMARYIn most patients with FTMH, the diagnosis can be reliably established with the useof slit-lamp biomicroscopy and a noncontact or contact lens; however, occasionallydifficulties may be encountered, even by experienced ophthalmologists, in diagnosingthis condition (see also Chapter 14B) (27). In such cases OCT and AF imaging willbe helpful in establishing the diagnosis of FTMH. Although FA is also an importanttool for evaluating eyes with FTMH (28,29), with the advent of OCT and AF it isnow only rarely needed.

ACKNOWLEDGMENTSI thank Dr. Vikki McBain and Dr. Noemi Lois for providing the photographs that il-lustrate this chapter. I also thank Prof. Alan C Bird, Prof. Fredrick W. Fitzke, and Dr.Zdenek J. Gregor for their cooperation in the initial srudy on AF of macular holes.

REFERENCES.. . . d Aderhaur in Polge von Traumen auf den Augenapfelk-L

L Knapp H. Uber isolierre Zerreissungen er . '-ll

Augeheilkd 1869;H~29, ill Am J 0 h hal 119696w, k RC All L Subjective slit-beam sign for macular sease. p t rna ; 8:449--4512. watz e , en. """ . I h I f db' a] h' ' TH WI k All lit beam test In macu ar 0 es con Hille y opnc co ereJl3. Tanner V, WtliJamson . wan e- en S 1 cetomography Arch OphthalmoI2000;118:1059-1063.. .

Jh RN G JDM Idi parhic senile macular holes: observanons, stages of formation,and implb.

4. 0 nson ,ass .. JO 88-95-917-924cions for surgical intervention. Ophthalmology 19 , . . '. . .

DM M I d f . used by vitreous and vlrreoreunal interface abnormalities. In' G:!l!5. Gass J . acu ar 'Is uncuon ca .' 4 h d "

d S ' A 1 f Macular Diseases' DiagnosIs and Treatment. t e. Sr Lows:MmDJ']DM, e. rereoscoprc t as o· >

1997, , f fidi chi I hi6. Guyer DR, Green W'R, de Busrros 5, er al. Histopathologic earures a I ropa C macu ar 0 esandC;'!l\Ophthalmology 1990;97, 1045~1 051. , 'E E F ' RN P , DE r al In1munocytochemical charactenzatlon of macular holeoperculaArch7. zra ,anss >. OSSlO , e . 'Ophthalmol 200 1;119,223~231. ,

8, Gass JD. Idiopathic senile macular hole. Its early stages and pathogenesis. Arch Ophthalmnl1988;106,629~639,

9. Gass JD. Reappraisal of biomicroscopic classification of stages of development of a macularhole.Am J

OphthalmoI1995;119,752~759, ,',10. Gaudric A, Haouchine B, Massi» P et al. Macular hole formation: new data provided by Optical coherence

tomography. Arch Ophrhalmoi 1999;117:744-751.. ,.11. Johnson MW, van Newkirk MR, Meyer KA. Perifoveal vitreous detachment IS the pnmarypathogenic

event in idiopathic macular hole formation. Arch Ophthalmol 2001;119:215-222.12. Abdelkader E, Lois N. Internal limiting membrane peeling in virreo-rerinal surgery. SurvOphthalmol

2008;530368~396,13. YannuzziLA, Rohrer KT, Tindel LJ, et a]. Fluorescence angiography complication survey.Ophthalmology

1986;93,6] 1~617,14. Hee MR, Puliafiro CA, Wong C, et al. Optical coherence tomography of macularhah

Ophthalmology, 1995; 102,748~756,15. Kasuga Y,Arai ], Akimoto M, et a]. Optical coherence tomography to confirm early closureofmacular

hole, Am J Ophrhalmol 2000;130,675~676,16. Imai M, Iijima H, Ooroh T, et a]. Optical coherence tomography of successfully repairedidiopathicmac·

ular holes. Am J Ophthalmol 1999;128:621-627.17. Baba T, Yamamoto S, Arai M, er al. Correlation of visual recovery and presence of phororecepor

inner/outer segment junction in optical coherence images after successful macular hole repair.Reim2008;280453-458,

18. Azzolini C, Parelli F,'Brancato R. Correlation between optical coherence tomography data andbicmicoscopic interpretation of idiopathic macular hole. Am J Ophrhalrnol 2001; 132:348-355.

19. Chan A, Duker ]5, Schuman JS, er al. Stage 0 macular holes: observations by optical coherencetomogra·phy, Ophthalmology 2004;111 ,2027~2032,

20. von Ruckmann A, Fitzke FW; Gregor ZJ. Fundus autofluorescence in parienrs with macularholesima"uedwith a laser scanning ophthalmoscope. Br J Ophchalmol 1998;82:346-351.

21. von Ruckmann A, Fizke FW';Bird AC. Distribution of fundus autofluorescence with a scanninglasereph-thalmoscope. Br J Ophthalmol 1995; 119:543-562.

22. von Ruckmann A, Fizke FW'; Bird AC. Fundus autofluorescence in age related maculat diseaseimagdwith a laser scanning ophthalmoscope. Invest Ophthalmol Vis Sci 1997;38:478-486.

23. Delori FC, Dorey K, Staurenghi G, et al. In vivo fluorescence of the ocular fundus exhibitsretinalpigmenrepithelial lipofuscin characteristics. Invest Ophthalmol Vis Sci 1995;36:718-729.

24. Funata M, Wedel RT, De La Cruz Z, et al. Clinicopathologic study of bilateral macular holesrrearedwirhpars plana vitrectomy and gas tamponade. Retina 1992;12:289-298.

25. Framme C, Raider ]. Fundus autofluorescence in macular hole surgery. Ophthalmic SurgLasers2001 ;32,383~390,

26. Wakabayashi T, Ikuno Y, Sayanagi K, et aI. Fundus autofluorescence telated to retinal morphologicalannfunctional changes in idiopathic macular holes. Acta OphthalmoI2008;86:897-901.

27. Martinez J, Smiddy WE, Kim J, et al. Differentiating macuJar holes from macular pseudoholes,AmJOphth,lmoI1994;117,762~767,

28. Thompson JR, Hiner CJ, Glaser BM, et aI. Fluorescein angiographic characteristics of macularholesIt-f~re and after viuectomy with transforming growth factor beta-2. Am J Ophthalmol 1994;117:291-30~,

29. Fish RH, Anand R, Izbrand DJ. Macular pseudoholes. Clinical features and accuracyof diagn()S1S.Ophthalmology 1992;99, 1665~1670,


CHAPTER

Fundus Autofluorescence in LamellarMacular Holes and Pseudoholes

DEFINITION OF PSEUDOHOLE AND LAMELLARMACULAR HOLEThe definitions of macular pseudohole (MPH) and lamellar macular hole (LMH)have been a matter of great debate in the past. Today there is a general consensus todefine an MPH as a macular lesion that has the appearance of a full-thickness mac-ular hole (FTMH) but does not have a loss of foveal tissue. By contrast, a loss offoveal tissue is mandatory for a diagnosis ofLMH (1,2). However, and in contrastto FTMH, in LMH only part of the foveal tissue is lost (there is no full-rhicknessdefect).

Before optical coherence tomography (OCT) was available, the definitions ofMPH and LMH were based on certain characteristic history and examination find-ings, which were shown to be not sensitive enough for a clear differential diagnosis.Clinically, both MPH and LMH have a similar appearance, with a round and well-circumscribed reddish lesion at the center of the fovea (1,3-10). Functional tests,such as the Watzke-Allen test (11) and microperimerry (12), in which no scoroma isdetected in either MPH and LMH are not useful for differentiating between theseconditions, and both clinical entities can lead to similarly impaired vision (median20/40) (1-3).

OCT evaluation has proved valuable in the diagnosis of macular holes, as it isable to visualize retinal anatomy with near microscopic resolution (= 10 j.LID forSTRATUS OCTTM [Carl Zeiss Meditec]; =5-7 urn for spectral domain OCT).However, although OCT studies have added valuable information regarding the def-inition, parhogenesis, and progression of macular holes (6-10), it is not always possi-ble ro differentiate between MPH and LMH. In particular, when there is residualtissue at the bottom of the foveal defect, as occurs with macular holes in stage 2 (ac-cording ro [he OCT classification of Azzolini et a1. [9]), OCT imaging may nor beable to indicate with certainty whether there is loss of retinal tissue. In such cases, thediagnosis of MPH or LMH is often a matter of speculation.

PATHOGENESISThe pathogenesis of MPH and LMH is not fully understood, but they have alwaysbeen considered ro be quite different processes. It has been hypothesized [hat MPH isattributable to [he centripetal contraction of an epiretinal membrane (3). In contrast,LMH is thought to be the result of an abortive process in the formation of an FTMH.Posterior vitreous detachment is the main initiating process of [he latter, but epireti-nal membrane contraction has been suggested as a likely secondary facror (2). Thismechanism is supported by two findings: 0) 62% to 89% ofLMHs may present wirh

259


. . al bane (12)' and (ii) pseudo-opercula, suggestiveofanaborredmaan epuetln mem r " .' c-ular hole, have been reporred in only 24% of patients WIth LMH (1), Tberefore,IIseems that the pathogenesis of LMH cannot simply be attribured to an ,bani"process in FTMH formation.

FUNDUS AUTOFLUORESCENCEIt is believed that fundus autofluorescence (AF) derives mainly from the Upofuscin.laden rerinal pigment epithelium (RPE; see Chapter 3) (13). It is generallyaccopledthat lipofuscin represents the product of degradation of photoreceptor outersegmen•.In a normal fundus, the distribution of AF IS diffuse, WIth decreased Intensityat Iheoptic nerve head, the retinal blood vessels, and the macula (13,14). MacularAF is".tenuated by the presence of luteal pigment, which has a higher concentration alonglheouter plexiform layer at the fovea (Fig. 14B.l) (15). Any foveal defect, includinganLMH that spares the phororeceptors (2), may alter the degree of fovealAF bydecreas-ing the amount of masking luteal pigment and thus increasing the fovealAF.

IMAGING TECHNIQUES

Slit-Lamp BiomicroscopySlit-lamp biomicroscopy in patients with MPH and LMH may simply showthecorn-man feature of a round, reddish lesion at the center of the macula, but it is notsersitive enough to detect a small loss of foveal tissue, which is present in LMH withpie·served visual acuity. Additionally, the presence of an epiretinal membrane is notdefinitive in the differential diagnosis of LMH and MPH; as mentioned above,62%to 89% of patients with LMH may have an associated epiretinal membrane(1,2),which will always be present in MPH (3).

In a study by Haouchine et al. (1), only 28% ofLMH cases diagnosedwithOCTwere diagnosed as LMH on fundus examination. Likewise, Witkin er al, (2)reportedthat only 37% ofLMH cases diagnosed using ultrahigh-resolution OCT weredereceiclinically on fundus examination. These data show the limitation of slit-lampbiom-croscopy in the diagnosis of LMH.

A~=

FIGURE148.1. IA) Macular pigment distribution (yel/owl in a macaque. (From SnodderlyOM, Auran JD, Delori FC.The macular pigment. II. Spatial distribution in primate retinas.Invest Ophthalmol Vis Sci 1984;25674-685.1181 Normal distribution of fundus AF.

CHAPTER 148 FUNDUS AUTOFLUORESCENCE IN LAMELLAR MACULAR HOLES AND PSEUDO HOLES 261

Optical Coherence TomographyRecently, Haouchine et al. (1) defined new criteria for the OCT diagnosis of MPHand LMH, and Witkin et al. (2) further expanded this topic using ultrahigh-resolu-tion OCT, a not yet commercially available technology capable of =3 p.rn axial imageresolution in the human eye. Haouchine et al.'s (1) OCT criteria were established byquantitative image analysis of six radial macular images obtained using a standardizedSTRATUS OCTTM imaging protocol. Briefly, retinal thickness was measured manu-ally with software calipers ar the foveal center and 750 urn temporally and nasallyfrom the center on the horizontal scan. In MPH, the OCT profile was a steepenedcontour with increased perifoveal retinal thickness and normal or slightly increasedcentrofoveal thickness. In LMH, the OCT profile was an irregular contour, withnear-normal perifoveal thickness and thinner than normal centrofoveal thickness.Using these OCT criteria, the study indeed showed two clearly different groups thatthe authors termed MPH and LMH (l).

Witkin et al. (2) used qualitative image analysis of ultrahigh-resolution OCT sec-tions in an attempt to define criteria for LMH diagnosis that do not necessitate meas-urement of retinal thickness. They presented four basic criteria for an OCT diagnosisof LMH: (i) an irregular foveal contour, (ii) a break in the inner fovea, (iii) a dehis-cence of the inner foveal retina from the outer retina) and (iv] an absence of a full-tbickness foveal defect with intact foveal phororeceprors. To avoid confusion betweenMPH and LMH, they proposed that the definition of MPH be expanded ro includeany macular lesion that has the appearance of a macular hole but lacks a full-thicknessfoveal defect. An LMH would then be a subcategory of MPH in which there is alamellar defect caused by separation of the innet from the outer retinal layers (2).It is clear that one of the major problems encountered in an OCT diagnosis of a

foveal defect is the difficulty of determining with certainty whether there is loss ofretinal tissue. Furthermore, if there is loss of foveal tissue, it is difficult to determinethe anatomic location of this loss.

Fundus AF ImagingIn vivo imaging of fundus AF can be performed with commercially available adaptedfundus cameras (Topcon USA, Paramus, NJ) or a confocal scanning laser ophthalmo-scope (see Chapter 5) (14). The usefulness of this technique has already been demon-strated for the diagnosis of FTMH (see Chapter 14A), and its accuracy has been re-ported to be comparable to that of fluorescein angiography (16,17).

Fundus AF has challenged the current OCT concepts regarding the differentia-tion between MPH and LMH (19). In a study of patients with stage 2 macular holesbased on OCT classification (9) and further classified as MPH and LMH accordingto the OCT profiles established by Haouchine et al. (I), we found rhat the foveal AFintensity did not differ between these two conditions (18). In both LMH and MPH,an increase in the foveal AF signal with respect to background was found (Figs. 14B.2and 14B.3). However, in patients with epiretinal membrane and macular pucker (Fig.14B.4) diagnosed by slit-lamp biomicroscopy, and in patients with impending macu-lar holes (foveal detachment) (Fig. 148.5), no focal increased AF at the fovea was de-tected. As noted above, fundus AF derives from lipofuscin-laden RPE (13) (seeChapter 3), which in the macula is attenuated by the luteal pigment. Most of this pig-ment in the fovea resides in the outer plexiform layer (15), interposed between thefoveal photoreceptors and the stimulating light. Thus, even very thin foveal defects,such as those affecting only the outer plexiform layer and sparing the photoreceptors,as in LMH (2), may increase the foveal AF. Thus, the AF findings (i.e., increased AF

(text continues on page 264)

A


c

c

FIGURE 14B.2. (A) Red-Iree image. (B) AF image.and(eh"cal OCT scan. On the basis 01 OCT (centroloveal thickness149,rnasal periloveal thickness 331 I'm. temporal perifovealthid~~282 I'm) the diagnosis 01 MPH was established. However.fundlAF demonstrated an increased in the foveal AF signal.indicativsjloss 01 loveal tissue and the presence 01an LMH.

A

FIGURE14B.3. (A) Red-Iree image. (BI AF image, and IClverticalOCTscan. On the basis 01 OCT (centroloveal thickness 93 um nas<periloveal thickness 356 I'm, temporal periloveal thickness390 ,m:the diagnosis 01 LMH was made. This diagnosis was confirmedb<lund us AF imaging demonstrating the presence of an increasedA'signal at the fovea.

CHAPTER 148 FUNDUS AUTOFLUORESCENCE IN LAMELLAR MACULAR HOLES AND PSEUOOHOLES 263

c

fiGURE 148.4. IA) Red-free image. 181AF image, and IC) hori-zontal OCT scan in a patient with macular pucker. No increasedAF at the fovea is observed, indicating a lack of tissue loss at thefovea.

A

c

fiGURE 148.5. IA) Red-free image, (81 AF image. and (CI verticalOCT scan of a patient with an impending macular hole (stage 1al.No increased foveal AF is present

264 SECTION III CLINICAL SCIENCE

. . . h MPH and LMH defined by OCT suggestthatinb"at the fovea) In patients WIt . d 14B 3) Th DUI

. . h . I ffoveal tissue (FIgs. 14B.2 an .. erewas~soalaclconditions t ere IS ass a hi len f'd a1 .. h aunt of AF and the t IC ess a resi u retirn]tissueof correlation between team . C hi h ". h MPH LMH. A likely explanation ror t ISISt at oncemeoU!the bases of err er or .. er

I .C I . the fovea is affected, the absence of masking pIgment allowstheAlp exirorm ayer In 1< ., f h h'., • C RPE II to be easily detected, mespeetJve 0 t e t icknes oftbongiuanng Hom ce S eI . h t r cell layer This is also supported by me lack of inc"".Jover ylllg p otorecep 0 -. .. 'IJ

C al AF d d i patients with macular pucker and impending macularholesinrove reeor e In . I

hi h h . d C t at the level of the outer plexiform layer or, therefore,ontbeW ic t ere IS no erccluteal pigment (Figs. 14B.4 and 14B.5).

Fluorescein AngiographyFluorescein angiography (FA) has been used in the past to diagnose FTMH(19),AIme site of the macular hole, the fluorescence from the underlying choroidbecomoclearly visible due to the lack of luteal pigment, which would normally attenuatetb,choroidal fluorescence. Therefore, foveal hyperfluorescent lesions are typically Db.served in stages 2, 3, and 4 FTMH (19); in stage 1 FTMH and in LMH, thehypel'fluorescence may vary depending on the attenuation caused by the foveallut~ pig.rnent; however, if present, it is usually mild.

Fundus AF imaging appears to be as accurate as FA in me diagnosisofFfMH(seeChapter 14A) (16,17). Likewise, AF imaging is sensitive enough to detectasmallloss ofluteal pigment in cases ofLMH, and the lack ofloss of luteal pigmentinMPH(Fig. 14B.6) (18). Together with the fact that, unlike FA, AF is noninvasive,thesefindings favor me use of AF instead of FA in rhe evaluation of patients withITMHLMH, and MPH.

CLINICAL IMPLICATIONS OF AF FINDINGSThe lack of a significant difference in foveal AF between LMH and MPH, asdiag·nosed by means of OCT imaging, raises questions concerning the validityof thisdif·ferentiation. Fundus AF may be a more valuable tool to evaluate the lossofinnerretina in such cases by demonstrating the presence or absence of me lutealpigment,which is located predominantly at the level of the outer plexiform layer in thehumanfovea, An absence or decrease of macular pigment (i.e., a loss of outer plexiformci,·sue) wonld increase foveal AF. In such cases, a diagnosis of LMH should beestab·lished. Fundus AF imaging can thus be used clinically to establish a differentialdiag·nosis between LMH and MPH, conditions that are usually difficnlt to rellapanevenby expett ophthalmologists (20).

Fundus AF imaging may replace FA in the evaluation of FTMH, LMH,andMPH. Fundus AF imaging has many advantages over FA: it is noninvasive, rapid,andshows an accutaey comparable to that of FA (16). Fundus AF imaging cannotrepllaOCT examination in all cases. For example, OCT may srill be needed whenthediag'nosis of FTMH is not clear on slit-lamp biomicroscopy. In such cases,OCT willdemonstrate a full-thickness defect, whereas AF, as in LMH, will show an area ofin·creased foveal AF. at the site of the hole. The similarly increased fovealAF inLMHand FTMH implies that fundus AF imaging is not suirable to differentiate a lamellarfrom a full thickiness macular hole. Therefore, OCT may be more appropriatethanAF ImagIng to moruror the progressive thinning of a LMH. OCT, however,doesnotappear to be sensitive enough to differentiate between an LMH (whereAf will

TER 14B FUNDUSAUTOFLUORESCENCEIN LAMELLAR MACULAR HOLESAND PSEUDOHOLES Z65

A

C

FIGURE 148.6. LMH. An OCTclassification stage 2 macular hole with residual retinal tissue at the bottom of the hole (A)OCThorizontal scan, (8) red-free image. and (C) AF image showing some faint foveal AF (01 FA shows subtle foveal alter-ations only during late phases of the angiogram.

demonstrate increased foveal AF) and an MPH (where AF will show a normal patternof foveal AF).

An accurate diagnosis of FTMH, LMH, and MPH is important to determine theproper surgical treatment of these lesions. Different options may be selected accord-ing to the OCT and fundus AF imaging findings. For instance, in the absence offoveal AF, the integrity of the foveal tissue is almost certainly confirmed. Therefore, itis likely that removal of the epirerinal membrane alone is all that is needed in suchcases. However, if foveal AF is present, a loss of foveal tissue has very likely occurredand the decision to operate will depend on many factors, such as the residual visualacuity and progression of signs and symptoms. Some recent reports have suggestedthat early intervention in LMH should probably be avoided because of the often dis-appointing results (2).

8

SUMMARYFundus AF imaging is becoming increasingly important in the examination ofmanymacular diseases, including macular holes. FTMHs may be properly. diagnosedby

f I· I bi . roscopyand OCT exammatlOn. In contrast, lesion, charaa_means 0 S 1[- amp iorruc w

ized by the presence of residual retinal tissue ar the bottom of rhe. foveal defect(Ocrclassification stage 2 according to Azzolini er al. [9]) may be more difficult to classifybe.cause OCT cannot indicate with certainty the loss of retinal tissue. Both LMHandMPH may produce a similar degree of functional loss. In such cases, fundus AFissen.sitive enough to differentiate an LMH (AF positive) from an MPH (AF neganve),andthus should be performed routinely for the clinical diagnosis and follow-up of pariennwith these macular diseases.


ACKNOWLEDGMENTSWe are greatly indebted to Francois Delori, PhD, for his help in designing thissrud)'and providing the picture of macular pigment distribution in a macaque (Fig. 14B.I).

REFERENCES1. Haouchine B, Massin p, Tadayoni R, et al. Diagnosis of macular pseudohoJes and lamellar macular holesbyoptical coherence tomography. Am ] Ophrhalmol 2004;138:732-739.

2. Witkin AJ, Ko TH, Fujimoto ]G, er al. Redefining lamellar holes and the vicreomacular interface:anIll·trahigh-resolurion optical coherence tomography study. Ophthalmology 2006; 113:388-397.

3. Allen jr AW, Gass JD. Contraction of a peri foveal epirecinal membrane simulating a macular hole.AmJOphch,lmolI976;82;684--691.

4. Gass JD. Idiopathic senile macular hole. Its early stages and pathogenesis. Arch OphrhalmoI1988jHi6:629-639.

5. Gass ]D. Reappraisal of biomicroscopic classification of stages of development of a macular hole.Am]Ophrhalmol 1995; 119;752-759.

6. Hee MR, Puliafito CA, Wong C, et al. Optical coherence tomography of macular holes. Ophrhalmclcg1995;102;748-756.

7. Gaudric A, Haouchine B, Massin P, et at Macular hole formation: new data provided by optical coherencetomography. Arch Ophthalmol 1999;117:744-751.

8. Tanner V, Chauhan OS, Jackson TL, er al. Optical coherence tomography of the vitrecretinal intemctinmacular hole formation. Br ] OphthalmoI2001;85:1092-1107.

9. Azzolini C, Patelli F, Brancato R Correlation between optical coherence tomography data and biomicescopic interpretation of idiopathic macular hole. Am J Ophrhalmol 2001; 132:348-355.

10. Haouchine B, Massin P, Gaudric A. Foveal pseudocyst as the first step in macular hole formation:aprospective study by optical coherence tomography. Ophthalmology 2001; 108: 15-22.

11. Watzke R, Allen L. Subjective slitbeam sign for macular disease. Am ] Ophrhalmol 1969;68:449--453.12. Tsujikawa M, Ohji M, Pujikado T, er al. Differentiating full thickness macular holes from impendingmac-

ular holes and macular pseudoholes. Br] Ophthalmol 1997;81: 117-122.13. Delori Fe, Dorey CK, Sraurenghi G, et a]. In vivo fluorescence of the ocular fundus exhibits reunalpig·

ment epithelial lipofuscin characteristics. Invest Ophthalmol Vis Sci 1995;36:718-729.14. von Ruckmann A, Firzke FW, Bird AC. Distribution of fundus autofluorescence with a scanning laserop~

thalmoscope. Br J Ophrhalmol 1995;79:407--412.

15. Snodderly DM, Auran ]D, Delori FC The macular pigment. II. Spatial distribution in prirnate rems-Invest Ophthalmol Vis Sci 1984;25:674--685.

16. v~n Ruckmann A: Fitzke FW; Gregor ZJ. Fundus autofluorescence in patients with macular holesimagedw~th a laser scannlllg ophthalmoscope. Br] Ophthalmol 1998;82:346-35 L

17. Ciardella Ap, Lee GC, Langton K, er al. Autofluorescence as a novel approach to diagnosing macularholes.Am) OphthoJmoI2004;137;956-959. e

18. Bot.toni F, Carmassi L, Cigada M, er a]. Diagnosis of macular pseudoholes and lamellar macular h(lles:~optical cohe~ence ~omog.raphy the gold standard? Br] Ophthalmol 2008;92:635--639.

19. Gass JD. Idiopathic senile macular hole. Its early stages and h . Ar h 0 hth 1m 11988'106:629-635. pat ogenesis. cpa 0 '

20. OM'hrtihn'l'J'ISmiddy WE, Kim J, et aI. Differentiating macular holes from macular pseudoholes.AmJpc, rna 1994;117;762-767.

Fundus Autofluorescence ofIntraocular Tumors: Choroidal Nevusand Melanoma

r:Ihoroidal melanoma is the most frequent primary intraocular tumor and the~ second most frequent malignant melanoma of rhe body (1). Although it is a

malignancy of rhe melanocytic cells of the choroid, it directly affects the reti-nal pigment epithelium (RPE). This secondary epithelioparhy appears as areas of arro-phy, drusen, lipofuscin accumulation, and localized detachment of the RPE (2).

Large rumors are nor difficult to diagnose as malignant, but the differential diagno-sis and management of small choroidal melanomas remains controversial. Documentedgrowth of a recently diagnosed small choroidal melanocyric lesion is considered a hall-mark for rhe diagnosis of choroidal melanomas, and quantirative (rumor size) and qual-itarive factors, such as symptoms, drusen, subretinal fluid, RPE changes, juxtapapillarylocation, and orange pigmentation, may be predictive of tumor growth in patients withthese melanocyric lesions (3).

Secondary epitheliopathy and particularly orange pigmentation overlying the le-sion have been described as a major risk factors for progression to melanoma (2,4-6).Orange pigment, indicative of lipofuscin accumulation within the RPE, overlyingsmall choroidal melanocytic lesions has been found to be significantly correlated withthe risk of subsequent growth (3,7).

Recently, several groups have investigated the role of fundus autofluorescence(AP) in the diagnosis and prognosis of choroidal lesions (8-12). It is accepted thatthe AP signal in the fundus comes predominantly from lipofuscin in the RPE (seealso Chapter 3). Thus, AF imaging may provide important information regardingpatients with choroidal tumors.

BASIC PRINCIPLESIn 1976, Shields er al. (13) demonstrated that lipofuscin accumulates in RPE cellsand macrophages of malignant choroidal rumors from incomplete degradation anddigestion of photoreceptor outer segments. It is usually seen as orange pigment overthe lesion; however, depending on whether the tumor is melanotic or amelanotic, thecolor may vary from orange to brown or red-brown, respectively.

Lipofuscin is a mixture of proteins, lipids, and small chromophores, and its accu-mulation in the RPE results from impaired or overwhelmed lysosomal digestive activ-iry (see Chapters 1 and 2). Evidence demonstrates that lipid peroxidation-derived pro-tein modifications are able to induce lysosomal dysfunction and lipofuscinogenesis inthe RPE, and suggests that such lipid peroxidation-induced lysosomal dysfunction


CLINICAL CHARACTERIZATION AND FUNDUSAUTOFLUORESCENCE FEATURESChoroidal Nevus

FIGURE 15.1. Autofluorescent deposits within the RPE and macrophages overlyingthe chorni,1melanoma seen by fluorescence microscopy.

considerably conrributes to cell damage and subsequent retinal degeneration(14).Theexact mechanism by which choroidal melanoma induces lipofuscinogenesis remains 10

be elucidated.In one histopathologic study (15), retina overlying choroidal melanoma

showed areas of degeneration of outer retinal layers with a reduction in the numberof nuclei of photo receptors. Sheets of proliferared RPE cells and clustersofpig.merit-laden macro phages were present under the degenerated neurosensory retiu,corresponding in position to the orange pigment observed in the macroscopicvie\\'of the tumor. On electron microscopy, most of the pigment granules appeaeiround ro oval and demonstrated moderate homogeneous density with ittegularlyindented outlines, similar to typical lipofuscin granules normally observedthrough·out the retina to a lesser extent, and different from melanin granules.

Histochemical analysis indicates that orange pigment in macrophagesandRPEcr&of tumors stains positive for PAS, Sudan black B, and Long Ziehl-Nielsen (8);reduassilversalrswith the Fontana-Masson method; bleaches partially with potassiumperman·ganate; and is acid-fast and oil red 0 positive in paraffin sections. It alsoexhibimgolden-yellow AF when examined with ultravioler light or a fluorescencemicrosopsuggesting that this pigmem is in faer lipofuscin (Fig. 15.1) (15). Lohmannandcoi-leagues (16) studied the endogenous fluorescence of ocular malignant melanomasandfound that lipofuscin granules were cleaved off and broken into small remnanrs inmemelanoma.

A choroidal nevus can be classified clinically according ro the CoilaborariveOculuMelanoma Study(COMS). Benign lesions are usually less than 1 mm in heighlan~5 mm m basal diameter according to ultrasound measurements, and haveno r~kfactors for growth such as specific s '. breti al u,,;d

J JJ ymptoms, orange pigrnentauon, su reno llUl'

FUNDUS AUTOFLUORESCENCE OF INTRAOCULAR TUMORS: CHOROIDAL NEVUS AND MELANOMA

FIGURE 15.2. Choroidal nevus with normal background AF.

juxtapapillary location, and "hot spots" on angiography. Nevi may be brown ro grayin color and their appearance may vary from homogeneous, small, flat, and well de-lineated to more heterogeneous with secondary changes overlying the nevus, such asdrusen and pigment mottling.

AF in choroidal nevus may show a normal pattern of background fundus AF, withno areas of increased or decreased AF signal, indicating a lack of RPE involvemenr (Fig.15.2). Most nevi will demonstrate this pattern (8). In a minority of cases, nevi may alsoreveal areas with mild decreased or increased AF signal (11). The size of these lesionsdoes not appear to influence the AF pattern, since large lesions may show normal back-ground AF and small lesions with some RPE chronic degeneration may have a faint in-creased or decreased AF signal.

Drusen, often overlying nevus, when large or coalescent, may appear as areas oflocalized increased AF. However, the increased AF signal under these circumstancesmay not be a sign oflipofuscin accumulation in the RPE, such as occurs in choroidalmelanoma; rather, it may result from the accumulation of different fluorophores in-side these large drusen of drusenoid detachments, or relate to the presence of de-tached RPE cells within the latter.

Pirondini et al. (II) observed drusen in 56% of nevi, the majority of whichdemonstrated no changes in the distribution of AF or a traced increased AF signal.Less commonly, a reduced AF signal was also detected. Another clinical feature ofnevus, although seen less commonly than drusen, is overlying RPE atrophy, which iseasily idenrified as areas of low AF signal on AF images.

Choroidal MelanomaChoroidal melanoma may appear as a pigmented lesion with orange pigment in areticular or confluent pattern throughout its surface. Other features present in this le-sion are subretinal fluid, superficial fibrosis with RPE metaplasia, and adjacent retinaldetachment.

In the majority of cases, fundus AF correlates with the presence of orangepigmentation within choroidal melanoma. Gunduz et al. (9) found a complete


FIGURE 15.3. Correlation between orange pigmentation and subretinal fluid on AF imagesandocr(horizontal scan) in a choroidal melanoma.

correlation between an increased A..F signal and orange pigmenr in 61.5% oftumos

a partial correlation in 23.1 %, and no correlation in only 15.4%.AF areas in small, medium, and large choroidal melanomas share commonchan;

teristics, such as a homogeneous increased AF signal and a confluent plaque-likeconfig-uration. An increased AF signal can be seen even before the appearance of orangepig·ment on ocular fundus photography (8)_

Materin et a1. (10) studied small choroidal melanoma and noted that allrunesthat presented with orange pigment had increased AF signal, although of differentintensities, and the majority of lesions showed mild increased AF (Fig.15.3).Subretinal fluid associated with choroidal melanoma disclosed increasedAf sigo~in 61% of the cases. Gunduz et a1. (9) found that nearly 90% of the choroidJimelanoma showed at least one focus of increased AF signal corresponding to thelocation of lipofuscin or hyperpigmentation over the lesion. In 75% of ameianoschoroidal melanomas, increased AF was observed in areas of hyperpigmenracion(9). Shields et a1. (12) demonstrated that the AF signal over small choroidJimelanomas was indistinguishable from that of the background, but that areasofsub retinal fluid and overlying orange pigment disclosed increased AF that seem~to be accentuated by the low AF signal present between each clump of orangepigment. A direct correlation between areas of increased AF and orange pigment w~identified clinically (Fig. 15.4).

Medium and large choroidal melanomas also appear to have increasedAf. Inlarger tumors, superficial fibrosis and subretinal fluid have been noted, andanin-creasedAF signal has been observed throughout the surface of the lesion correspondingto areasof increased pigmentation and fibrous metaplasia of the RPE. However, inhighly elevated tumors, artifacts related to the imaging technique may limittheine-pretation of fundus AF images (8).

TER 15 FUNDUS AUTOFLUORESCENCE OF INTRAOCULAR TUMORS: CHOROIDAL NEVUS AND MELANOMA

FIGURE 15.4. Choroidal melanoma with increased AF areas over the lesion (photograph obtainedusing a 55 degree lens] corresponding to areas of orange pigment

OTHER ANCILLARY TESTS

Fluorescein AngiographyFluorescein angiography (FA) has been used extensively for the diagnosis and differen-tiation of choroidal tumors, although its accuracy in evaluating suspected choroidalmelanomas is limited. There is no pathognomonic angiographic sign for a choroidalmelanoma; however, certain features can be identified, such as blockage of the back-ground fluorescence, punctuate hyperfluorescence due to changes in the pigment ep-ithelium, an independent vascular network, and late staining as the dye leaks from thetumor vessels. In larger, dome-shaped melanomas, a characteristic "double circulation"pattern is evident in the early phase of the angiogram (I7). Confocal FAmay revealves-selswithin the tumor, although these are only visualized in the very early arterial phaseof the angiogram.

Bernard and Dhetmy (18) performed a histological study in a choroidalmelanoma with orange pigment and pinpoint foci of hyperfluorescence on FA. Theformer relates to areas of increased lipofuscin, whereas the latter relate to exudative,hyaline, and gluco-lipido-proteic "blebs" in connection with Bruch's membrane or inthe subretinal space. The authors concluded that these findings are an expression ofchoriocapillaris and! or RPE dysfunction, and today we are able to identify these find-ings by AF. However, the complete correlation and value for diagnosis of FA and AFremain to be determined.

Indocyanine Green AngiographyIt has been reported that certain histologically identified microcirculation patterns arean independent risk factor for the metastatic behavior of choroidal melanomas(19,20). These patterns can be classified as "parallel with cross-linking" or "net-works," both of which indicate higher risk for metastatic disease, or as "silent" and"parallel without cross-linking," which suggest a better prognosis.

Confocal indocyanine gteen angiography (IGA) is capable of detecting microcir-culation patterns and appears to be superior to FA with a conventional fundus cam-era (21). Today, with videoangiography we are able to image very early phases of theangiogram, which can facilitate the identification of these patterns.


FI . d IGA are better suited fat large tumors with more promin"uorescein an \. t1 d although AF might add data on the status of the overl)'i'n

vesse structure, an. gretina and RPE ro the diffetential diagnosis, as the tumors get larger, the roleofAIprobably becomes more limited. However, fat smaller lesions, AF could behelpfulfor the early diagnosis of choroidal melanoma in combmatlon with other Imagingtechniques by demonstrating areas of increased AF SIgnal related to rhe presenceoiorange pigment or sub retinal fluid, which IS rarely observed over benign choroidal

nevus.

Optical Coherence TomographyOptical coherence romography (OCT) provides cross-sectional imaging of rheretinawith a resolution of 10-15 fLm (STRATUS OCTTM; Carl Zeiss Meditec) romorere·cent instruments such as spectral domain OCT, which reaches -3-fLm axialimagere-olutions. Several srudies used this technique ro evaluate choroidal melanocyrictu-mors, including choroidal nevi and small choroidal melanoma. OCT findingsofchoroidal nevi and melanomas are limited to the anterior portion of themass, sincethe laser is scattered by the RPE, causing a weak reflectivity within rhe tumor.Therefore, OCT is not useful for studying the choroidal tumor itself. However,ilmay help identify changes in the RPE and neurosensory retina (22,23).

Shields and colleagues (22) studied retinal findings at the site of choroidalnevibyOCT, which included overlying retinal edema, subretinal fluid, retinal thinningdrusen, and RPE detachment. They also dererrnined the thickness of the overl~ngretina and the status of the phororeceptors (i.e., whether there was a loss or attenuation of this layer). Findings specific ro the lesion are more difficult ro identity;how·ever, in one study (8) we were able ro show in the majoriry of nevi an increasedthick·ness of rhe RPE/choriocapillaris layer with an attenuation of the reflectivityheneaththat area. AF in these cases was useful for determining the functional statusoftheRPE, rather than the anatomical status, since most cases of choroidal nevusappearwith normal background AF, suggesting a normal or at least relatively healthyRPEand neurosensory retina overlying the tumor. Therefore, the combination ofocrand AF for choroidal nevus is probably tecommended for early detection ofsignsofmalignancy, and is certainly better than clinical examination for detecting related reti-nal edema, subretinal fluid, retinal thinning, and overlying RPE damage (8).

OCT can also be very useful for early diagnosis of small choroidal melanomasbe-cause localized subrerina] fluid is strongly associated with tumor growth (24,25).Espinoza and colleagues (26) identified an active OCT pattern of subrerinelfluidwith localized full-thickness retinal detachment overlying or adjacent to choroidalmelanoma, which may indicate an active tumor that is more prone to grow. Theycompared these features with a more chronic pattern featuring retinal thinningorin-traretinal cysts that could be confused with fluid and could suggest a lesionlesslikelyto grow. Moreovet, subretinal fluid demonstrates increased AF signal in casesofchoroidal melanoma, and this feature may be associated with findings ofOCT robet·rer differentiate and diagnose melanocytic tumors (8,12).


AF may also be useful for the differenriarion of choroidal and tetinallesions otherthan nevus and melanoma.

FUNDUS AUTOFLUORESCENCE OF INTRAOCULAR TUMORS: CHOROIDAL NEVUS AND MELANOMA

Congenital Hypertrophy of the RetinalPigment EpitheliumCongenital hypertrophy of the RPE (CHRPE) is normally identified as a flat, pig-mented lesion at the level of the RPE, and it can be usually differentiated fromchoroidal nevi and small melanoma by its clinical appearance. However, when thediagnosis is not straightforward, fundus AF may be beneficial because CHRPE ap-pears, in the majority of cases, as lesions with low AF signal. This correlates to thehistopathological finding of absence of lipofuscin in these lesions, probably due toextensive photoreceptor degeneration and a defect in the phagocytosis of photore-ceptor outer segments (Fig. 15.5) (27). There is often a pigmented halo surround-ing CHRPE, which generally appears with normal or slightly increased AF. Usingfluorescence microscopy, Lloyd et al. (28) demonstrated an increase in the amountof lipofuscin in the surrounding RPE of CHRPE, which may explain the halo ptes-ent clinically. Since choroidal nevus often does not interfere with the normal back-ground AF, these features ate helpful for the diagnosis and differentiation ofCHRPE (Fig. 15.5).

FIGURE 15.5. CHRPE demonstrating reduced AF (AI with a 488 nm laser [AFI. and reduced-normal AF(8) with a halo of IncreasedAFwith a 787 nm laser INIAl. in contrast to the choroidal nevus. which hasnormal AF (e) and increasedAFwith NIA (0).

11- SECTION II CLINICAL SCIENCE

FIGURE 15.6. Congenital simple pigment epithelium hamartoma demonstrating reducedAF.

Congenital Simple Pigment Epithelium HamartomaCongenital simple pigment epithelium hamartoma is generally found in routinea-aminations in asymptomatic children and young adults. lt is composed ofhyperpl.-tic RPE, with a variable vascular component. Shields and colleagues (29)describedsubtle feeder vessels in the clinical exam, although this finding was not confirmedan-giographically. This lesion has a superficial, full-thickness retinal involvementandconsequently will appear as a lesion with low AF signal due to blockageofnorm'background AF (Fig. 15.6) (30).

Retinal Astrocytic Hamartomasin Tuberous SclerosisTuberous sclerosis is a hereditary disease characterized by multiple hamartomainseveral organs, such as the central nervous system and [he eye. Retinal asuocyrichamartomas may arise from the inner surface of the retina or from the opticnervihead. They can be a circular or oval solitary, flat, gray mass wirhout calcificarion(~1"1), or they may contain several lesions with multiple calcified nodular areas(rype1i,A third type combines features of the other two and appears with a whirish-grayg~'tening central calcification and a peripheral irregular rim (type 3). AF maybeusefWfor differentiating these types of lesions from each other, as well as fromotherrecin'and choroidal tumors. Type 1 hamartomas can have either normal backgroundAFOIreduced AF signal due to blockage of normal backgtoundAF. Types 2 and3maybn'lesion-specificAF, but they may also have areas of reduced AF signal due to blockageorbackground AF (31).

Near-Infrared AutofluorescenceNear-infrared AF (NLA) uses a laser diode excitation at 787 nm and a detection@«'with transmitted light at 800 nm (see also Chapter 6). The NLAsignal appearstoorig'mare 111 the RPE and the choroid, and has been related to oxidized melaninorlOrn'

pounds associated with melanin (32).

FUNDUS AUTOFLUORESCENCE OF INTRAOCULAR TUMORS: CHOROIDAL NEVUS AND MELANOMA 275

fiGURE 15.7. Choroidal hemangioma demonstrating reduced AF with both a 488 nm laser and NIA witha 787 nm laser.

Weinberger and colleagues (33) studied NLA images of several patients, includingrwo cases with choroidal nevi that presented with increased NLA signal but lackedblue-light-excited AF, demonstrating the absence oflipofuscin accumulation in theselesions.

We studied choroidal melanocytic lesions using NLA, and both choroidal nevnsand choroidal melanoma showed an increased signal that correlated with near-in-frared reflectance and did not correlate with blue-light-excited AF (488 nrn).Although this techniqne may not be useful for the differentiarion of nevus andmelanoma, it could be useful for the differentiation of retinal and choroidal lesions,such as CHRPE, choroidal hemangiomas, and other nonmelanocytic tumors.CHRPE and choroidal hemangiomas show decreased NLA, in contrast to the in-creased signal of nevus and melanoma (Figs. 15.5 and 15.7).

SUMMARYChoroidal melanomas appear to have a pattern of confluent increased AF signal overthe tumor, secondary to accumulation of lipofuscin. Most nevi do not have areas ofincreased AF, although secondary changes such as large drusen may appear occasion-ally as areas with mildly increased AF signal. It is possible that transforming lesionsmay be identified by detection of subclinical lipofuscin (orange pigment) by AF im-aging or by a newly diagnosed area of increased AF. Therefore, serial AF images ofchoroidal nevus and melanoma are advisable in the follow-up of patients withchoroidal melanocytic lesions. However, prospective studies with a larger number ofpatients are necessary to confirm these findings.

REFERENCES1. Margo CEo The collaborative ocular melanoma study; an overview. Cancer Conrro12004;1 J :304-309.2. Damato BE, Foulds W$. Tumor-associated retinal pigment epithelioparhy. Eye 1990;4;382-387.3. Singh AD, Mokashi AA, Bcna ]F, cr al. Small choroidal melanocyric lesions: features predictive of growth.Ophthalmology 2006; 113: I 032-1039.

4. Augsburger JJ, Schroeder RP, Terriro C, ec al. Clinical parameters predictive of enlargement ofmelanocyric choroidal lesions. Br ] OphthalmoI1989;73:911-917.

Healthcare

Autofluorescence by noemi