Benefits of cetalkonium chloride cationic oil-in-water nanoemulsions for topical ophthalmic drug delivery

Benefits of cetalkonium chloride cationic oil-in-waternanoemulsions for topical ophthalmic drug deliveryPhilippe Daull, Frédéric Lallemand and Jean-Sébastien Garrigue

Novagali Pharma, Evry, France

Keywordscationic; cetalkonium chloride; drug delivery;oil-in-water emulsion; ocular surface

CorrespondenceJean-Sébastien Garrigue, Novagali Pharma,1 rue Pierre Fontaine, 91058 Evry Cedex,France.E-mail: [email protected]

Received January 28, 2013Accepted April 7, 2013

doi: 10.1111/jphp.12075

Abstract

Objectives Topical ocular administration is the most convenient route of admin-istration of drugs for the treatment of eye diseases. However, the bioavailability ofdrugs following eye instillations of eye drops is very low. Over the past 20 years,extensive efforts have been put into research to improve drug bioavailabilitywithout compromising treatment compliance and patients’ quality of life.Key findings One of the most efficient ways to improve drug bioavailability is toincrease the precorneal residence time of the eye drop formulations. As a result,new eye drops, with bioadhesive properties, have been developed based on thecationic oil-in-water (o/w) nanoemulsion technology. These low viscosity eyedrop nanoemulsions have improved precorneal residence time through the elec-trostatic interactions between the positively charged oil nanodroplets and thenegatively charged ocular surface epithelium.Summary This review is the first to present the benefits of this new strategy usedto improve ocular drug bioavailability. The roles of the cationic agent in the stabi-lization of a safe cationic o/w nanoemulsion have been discussed, as well as theunexpected benefits of the cationic o/w nanoemulsion for the protection and res-toration of a healthy tear film and corneal epithelium.

Introduction

Aqueous eye-drop solutions are still the most common for-mulations for topical ocular drug delivery, since they are thesimplest, easiest and cheapest ocular dosage forms toproduce. The main drawbacks of these conventional oculardosage forms are that they are rapidly eliminated from theocular surface following instillation, resulting in a lowocular bioavailability (less than 1%) of the drugs, and arelimited to water soluble compounds.[1] Variants with vis-cosifying agents, penetration enhancers or spreading sur-factants have not fundamentally changed the paradigm.Suspensions, gels and negatively charged oil-in-water (o/w)nanoemulsions were developed to improve ocular bioavail-ability of lipophilic or poorly water soluble drugs.[2–4]

Among them, o/w nanoemulsions were demonstrated to beeffective ocular drug delivery vehicles.[5,6]

Anionic o/w nanoemulsions with ciclosporin (cyclospor-ine A; Restasis, Allergan) were developed to increase tearproduction in patients whose tear production was pre-sumed to be suppressed due to ocular inflammation (i.e. for

patients with dry eye disease), or with difluprednate(Durezol, Alcon) for the treatment of inflammation andpain associated with eye surgery.[7] Oil-in-water nanoemul-sions were demonstrated to be excellent vehicles forlipophilic drugs, such as ciclosporin or prostaglandin ana-logues like latanoprost or tafluprost, but also for deliveringwater unstable drugs.[8,9]

Cationic o/w nanoemulsions extended one step furtherthe benefits of the o/w nanoemulsions for drug delivery byimproving their residence time over that observed with theanionic o/w nanoemulsions. These cationic o/w nanoemul-sions take advantage of the negatively charged ocularsurface to increase through electrostatic interactions theirprecorneal residence time, and thus the ocular drug bio-availability.[10,11] As a consequence, the first generation ofcationic o/w nanoemulsions were developed to optimizepenetration of drugs (among them ciclosporin) in oculartissues.[9,11–14] This first generation of cationic o/w nanoe-mulsions used noncompendial cationic surfactants and

bs_bs_banner

And PharmacologyJournal of Pharmacy

Review

© 2013 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–•• 1

mailto:[email protected]

were not devoid of ocular toxicity side effects.[15–17] Hence,the challenges for the development of cationic o/w nanoe-mulsions are in the choice of the most appropriate cationicagent used to bring the positive charge to the oil nanodrop-lets, and in the improvement of the ocular tolerance ofthese positively charged nanoemulsions.

This review is the first to compile the informationpresent in the literature that describes this new strategyused to improve ocular drug delivery: the use of cationico/w nanoemulsion vehicles in ocular drug delivery. Themain steps involved in the pharmaceutical development willbe discussed, particularly the ones that led to the choice ofcetalkonium chloride (CKC) as a cationic agent compatiblewith the ocular surface.

Definition of a cationic oil-in-waternanoemulsion

By definition a cationic o/w nanoemulsion is a biphasic for-mulation that comprises positively charged oil nanodroplets(the oil phase) dispersed in water (the continuous phase).[18]

Table 1 summarizes the physicochemical properties of acationic o/w nanoemulsion. The positive charge of the oilnanodroplets is brought by a cationic surfactant that local-izes itself at the oil interface. Ideally, this cationic agentshould be sufficiently lipophilic to be almost exclusivelyentrapped in the oil with only very low amounts of the cati-onic agent present in the aqueous phase of the formulation.In addition to the biological effects of the cationic o/wnanoemulsion (discussed below), the positive charge of theoil nanodroplets helps improve the long-term stability ofthe nanoemulsion by generating a repulsive electrostaticforce (measured by the zeta potential) between the posi-tively charged oil nanodroplets, thus preventing theirmerging and avoiding the coalescence process of the nanoe-mulsion during shelf life.[18–20] Figure 1 presents the sketchof one of the oil nanodroplets present in the cationic o/wnanoemulsion.

While the oil phase of the nanoemulsion is generallymade of inert and stable oils, such as medium chain triglyc-

erides or mineral oil (i.e. non-vegetable liquid paraffin), thechoice of the right cationic agent needed to produce a safeand well tolerated cationic o/w nanoemulsion necessitated athorough examination of the cationic agents at hand.[21]

Choice of the cationic agent

The positive charge of a cationic o/w nanoemulsion is esti-mated by measuring the zeta potential. The zeta potential(z) is the electrical potential difference (DV) between thedispersion medium (i.e. water) and the stationary layer offluid attached to the dispersed oil nanodroplets.[22,23] Thezeta potential is a measure of the magnitude of the electro-static or charge repulsion between the oil nanodroplets, andis one of the fundamental parameters known to affect thestability of dispersed systems (i.e. o/w nanoemulsion); thusthe higher the zeta potential, the better (z � +40 mV).[22] Asa consequence, to obtain a high zeta potential for the cati-onic o/w nanoemulsion, all, to almost all the cationic agenthas to be entrapped in the oil nanodroplet, with the positivecharge located at the oil–water interface, and no to very lowamounts of freely soluble molecules of the cationic agentpresent in the aqueous phase (i.e. the dispersion medium),where they can contribute to ‘shield’ and reduce the zetapotential of the nanoemulsion. Thus, the cationic agentneeds to be lipophilic, i.e. amphiphilic; and the higher thelipophilicity the better.

A large number of cationic agents were describedin the literature that could have been potential cationicagents for the cationic o/w nanoemulsion, such as: stear-ylamine, oleylamine, poly(ethylenimine), poly(l-lysine), 1,2-di-(9Z-octadecenoyl)-sn-glycero-3-phosphoethanolamine(DOPE), 1,2-di-(9Z-octadecenoyl)-3-trimethylammonium-propane (DOTAP), and alkyl benzyldimethylammoniumcompounds of various alkyl chain lengths (IUPAC name:benzyl(dimethyl)azanium; but better known under theircommon name: benzalkonium chloride (BAK) derivatives).However, they were all hampered either by toxicity, stabilityor regulatory issues which avoided or limited their use inophthalmologic products.[16,23,24] Consequently, the search

Table 1 Summary of the physicochemical characteristics of a cationicoil-in-water nanoemulsion

Parameter Description

Aspect White opaque to slightly translucentpH 5.0–7.0Osmolality (mOsmol/kg) 270Droplet size (nm)a < 200Zeta potential (z, mV)a Positive (+40)Sterility Sterile

aDroplet size was determined by dynamic light scattering (HPPS,Malvern Instruments), and zeta potential by electrophoretic mobilitymeasurement (Zetasizer 2000, Malvern Instruments).

Oily CoreSolubilizes the drugs

SurfactantStabilizes the interface

Cationic agentBrings the Positive Charges

Figure 1 Schematic representation of one of the oil nanodropletspresent in the cationic oil-in-water nanoemulsion.

Philippe Daull et al.Cationic emulsion and ocular drug delivery

© 2013 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••2

for the appropriate cationic agent was limited to the onesalready registered, used in ophthalmic products or compli-ant with either the United States (US) or European (EU)pharmacopoeias.[11]

The most common cationic agents found in ophthalmicproducts belong to the family of quaternary ammoniums,such as BAK or polycationic polymers also known as poly-quaternium (e.g. Polyquaternium-1 (PQ-1) found in Poly-quad). However, these quaternary ammonium derivativesare used as preservative agents in conventional ocular drugproducts for their bactericidal and microbicidal properties(through a detergent action (see Furrer et al.[25])) at rela-tively low concentrations in aqueous solution: 0.001% PQ-1in Travatan (Alcon) or 0.02% BAK in Xalatan (Pfizer). Overthe past twenty years, a very large collection of evidence hasbeen published demonstrating the deleterious effects ofquaternary ammonium preserved eye drop solutions for theocular surface, especially for BAK-preserved solutions.[26–33]

The actual trend for conventional ocular drug products– i.e.for eye drop aqueous solutions– is to reduce the concentra-tion, or even remove completely quaternary ammoniums,and especially BAK, from their compositions.[34] As a conse-quence many soft-preserved and preservative-free ophthal-mic drug products have reached the market in the last fewyears.[35] Hence, can quaternary ammonium chlorides, andamong them BAK derivatives still make good cationicagents for cationic o/w nanoemulsions?

BAK derivatives as cationic agent forcationic oil-in-water nanoemulsion

As mentioned previously, the ideal cationic agent should belipophilic enough to localize itself exclusively within the oilnanodroplets with no freely soluble cationic agent moleculewithin the aqueous phase for better zeta potential and shelflife stability, and improved safety profile.[22,36]

According to the latest (2012) US and EU pharmaco-poeias BAK is a mixture of alkyl benzyldimethyl quaternaryammonium chlorides of various alkyl chain lengths(Figure 2). The alkyl chains are ranging from 8 to 18carbons, with the C12, C14 and C16 alkyl derivatives beingthe most common in the BAK mixture (Figure 2). Indeed,the pharmacopoeias specify that the BAK mixture contentof the C12 homologue should not be less than 40%, and thecontent of the C14 homologue not less than 20% of thetotal alkyl benzyldimethylammonium chloride content. Inaddition the sum of the C12 and C14 alkyl derivatives hasto represent at least 70% of the BAK composition. Note thateven for pharmacopoeia compliant BAK mixtures, the com-position and distribution of the alkyl chain derivatives canvary from one manufacturer to another.[37] Thus, is thereamong these alkyl benzyldimethylammonium of variousalkyl chain lengths an alkyl derivative that possesses physi-

cochemical properties that would make it compatible withthe ideal cationic agent for an o/w cationic nanoemulsion?

Physicochemical properties of theC12, C14 and C16 BAK derivatives

The C12, C14, and C16 alkyl derivatives of BAK are alsoknown as benzododecinium chloride, myristalkonium chlo-ride, and cetalkonium chloride (CKC), respectively. Table 2summarizes the physicochemical properties of theses alkylderivatives. It appears that CKC is the most lipophilic andless water soluble among the three major BAK derivativespresent in the BAK mixture. Thus, the C16 BAK derivative(i.e. CKC) with a calculated logP of 9.5 is a particularlyattractive cationic agent for cationic o/w nanoemulsions.Figure 3 illustrates the phase distribution for the differentalkyl derivatives of BAK in o/w nanoemulsions or aqueoussolution. Due to their lower lipophilicity, low concentra-tions of the C12 and C14 BAK derivatives can be found inthe aqueous phase of the cationic o/w nanoemulsion, whileno C16 BAK derivatives (i.e. CKC) are present in theaqueous phase of the o/w nanoemulsion. This is confirmedby the measure of the zeta potential. For cationic o/wnanoemulsions with BAK (0.02%) as cationic agent the zetapotential is approximately +20 mV, while for a cationic o/wnanoemulsion with CKC (0.005%) as the cationic agent thezeta potential is +40 mV.[38,39]

However, since these cationic BAK derivatives are sur-factants with detergent properties and cellular membranetoxicity, as such, the BAK mixture is very often added to eyedrop aqueous solutions for their preservation.[40] It wasdescribed by Kurup et al.[41] that only the free forms of thepreservative (i.e. BAK derivatives) present in the aqueousphase of an o/w nanoemulsion were available for antibacte-rial activity, and therefore exerted ocular surface cell mem-brane toxicity.[42,43] Thus, it is very important to have thelowest concentration of free/micelle BAK derivative mol-ecules in the aqueous phase to avoid or limit as much aspossible the side effects induced by these free molecules.

n = 8, 10, 12, 14, 16, 18

CnH2n+1

N+

Cl-

Figure 2 Benzalkonium chloride is a mixture of alkyl benzyldimethyl-ammonium chloride compounds of various chain lengths.

Philippe Daull et al. Cationic emulsion and ocular drug delivery


Hence, for a better ocular tolerance and safety profile of thecationic o/w nanoemulsion, the higher the lipophilicity ofthe cationic agent, the better. Consequently, CKC wasselected as the cationic agent of choice for the developmentof unpreserved, well tolerated cationic o/w nanoemulsionsas it is the most lipophilic BAK derivative present in theBAK mixture. The following sections will discuss the oculartolerance and safety profile of the unpreserved CKC-containing cationic o/w nanoemulsions, and the variousadvantages brought by CKC and its positive charge for theimprovement of drug bioavailability and the protection andhealing of the ocular surface.

Biological properties of cationicoil-in-water nanoemulsion eye drops

The rationale for developing cationic o/w nanoemulsion eyedrops arose from the observation that the ocular mucosa isnegatively charged. Both the corneal and conjunctivalhuman cells harbour O-glycosylated transmembranemucins with only 6% of their glycans not terminated by thenegatively charged sialic acid.[10] When a cationic o/wnanoemulsion eye drop is instilled onto the ocular surface,the resultant electrostatic attraction between the positivelycharged oil nanodroplets and the ocular surface manifestsitself macroscopically by an improved spreading of the eyedrop preparation onto the eye.[11] This was evidenced bydynamic contact angle measurements, which were rapidlyvery low with cationic o/w nanoemulsion (below 3° withinthe first second) upon instillation, while it remainedelevated (above 42°) with either the anionic nanoemulsionor the hyaluronate hydrogel.[17] The electrostatic interac-tions help to increase the residence time of the oil nanodro-plets on the ocular surface. The ocular residence time playsa major role for drug absorption, and various strategieswere developed to improve drug residence time as a meansto improve drug absorption, such as ophthalmic inserts, vis-cosity enhancers/mucoadhesives, anionic o/w nanoemul-sions, and cationic o/w nanoemulsions.[4] Classic eye drop

solutions are eliminated within minutes following adminis-tration, thus greatly reducing drug ocular bioavailability,which seldom exceed 1% of the delivered dose.[44] The ben-eficial role of the cationic o/w nanoemulsion on the ocularbioavailability was demonstrated with ciclosporin.[45] Cati-onic and anionic nanoemulsions of 0.05% ciclosporinabsorptions in the conjunctiva and cornea following asingle ocular application in the rabbit eye were measuredover time (up to 72 h). The pharmacokinetic data demon-strated that ciclosporin’s area under the curve (AUC) withthe cationic o/w nanoemulsion was approximately twice theone observed with the anionic o/w nanoemulsion (AUC:26477 vs 14210 ng/g.h, for the cationic vs anionic nano-emulsion, respectively). The maximum ciclosporin con-centration (Cmax) in the cornea for the cationic o/wnanoemulsion following instillation was 1371.8 ng/g, whileit was only 747.8 ng/g for the anionic o/w nanoemulsion.[46]

Interestingly, a second peak of ciclosporin absorption wasobserved in the cornea with the 0.05% ciclosporin cationico/w nanoemulsion two hours post instillation (and even11 h after instillation with the 0.1% ciclosporin cationic o/wnanoemulsion), while no such peak was observed with theanionic o/w nanoemulsion.[17,46] The presence of this secondpeak is a strong argument in favour of an extended ocularresidence time with cationic o/w nanoemulsions. Theextended ocular residence time of the cationic o/w nanoe-mulsion was suggested following the instillation of alatanoprost-loaded cationic o/w nanoemulsion.[47]

Safety profile of the cationicoil-in-water nanoemulsions

As indicated previously, the positive charge of the cationico/w nanoemulsions is brought by CKC, a quaternaryammonium that is structurally closely related to BAK.CKC’s alkyl chain contains 16 carbons. This seeminglyslight difference in alkyl chain length has a major impact onthe physicochemical properties of CKC (Table 2) and as aconsequence on the safety profile of the CKC-containing

Table 2 Physicochemical properties of C12, C14 and C16 alkyl derivatives of benzalkonium chloride

BAK derivatives C12 C14 C16 (CKC)

Molecular weight (g/mol) 340 368 396XlogP3 (PubMed compound) 7.4 8.4 9.5logP (calculated) 3.44 4.45 5.46logP (measured)a -0.17 < logP<-0.07 0.24 < logP < 0.44 2.4 < logP < 2.6Critical micellar concentration (CMC)b 4.5 mM 0.75 mM 0.55 mM

1.53 g/l 0.29 g/l 0.022 g/l0.153% 0.029% 0.0022%

Superficial tension at CMC (mN/m) 38 38 40Water solubility (g/l; 25°C) 1230 100 8.5

BAK, benzalkonium chloride; CKC, cetalkonium chloride. aMaximum solubility ratios in both organic and aqueous phases. bMeasured with a Wil-helmy blades tensiometer.



cationic o/w nanoemulsions when compared with BAK-containing cationic o/w nanoemulsion. Liang et al.[36] usedan in-vivo rabbit model to demonstrate that both BAK andCKC cationic o/w nanoemulsions were much better toler-ated by the rabbit ocular surface than their solution coun-terparts. When BAK is in solution the C12 and C14 alkylderivatives form micelles that have the possibility to interactwith the corneal and conjunctival cell membranes onceapplied on the ocular surface. Through their detergent

properties the C12 and C14 BAK alkyl derivatives alter theepithelial cells integrity, leading to the well-known deleteri-ous effects of BAK solutions. However, when formulated inan o/w nanoemulsion, a significant part of these C12 andC14 alkyl derivatives of BAK is entrapped in the oil (as aconsequence of the lipophilicity of the C12 and C14 alkylchains of BAK). Only a small proportion of the C12 andC14 alkyl derivatives of BAK remain in the aqueous phaseof the solution, thus greatly reducing the ocular toxicity of

Cetalkonium chloride in emulsion

(a)

(b)

(c)

Cetalkonium chloride

C16C16 C14 C12

Benzalkonium chloride in emulsion

Benzalkonium chloride in solution Benzalkonium chloride

Figure 3 Illustration of the phase distribution for the different alkyl derivatives of benzalkonium chloride. (a) Cetalkonium chloride (blue) in emul-sion; (b) benzalkonium chloride (BAK) mixture in emulsion; and (c) BAK mixture in aqueous solution.



the BAK o/w nanoemulsion (Figure 4). This is emphasizedwith CKC, the C16 alkyl derivative of BAK. Due to itsextreme lipophilicity, when formulated in emulsion almostall of the CKC is entrapped in the oil with no free CKC mol-ecules present in the aqueous phase of the emulsion. As aresult, an even better safety profile of the cationic o/wnanoemulsion can be expected with CKC as the cationicagent rather than with BAK (Table 3)[48–51]. This is con-firmed in vivo by the improved ocular tolerance of the CKCcationic nanoemulsion over the BAK cationic nanoemul-sion.[36] In a rabbit model of acute toxicity, 15 instillations ofCKC-containing cationic o/w nanoemulsion had a betterDraize test score and a lower in-vivo confocal microscopy(IVCM) score than the BAK-containing cationic o/wnanoemulsion (Figure 4), and were equivalent to the BAK-free saline control. The ciclosporin-containing CKC cati-onic o/w nanoemulsion was as well tolerated as –if notbetter than– the BAK-free Restasis.[52]

The CKC cationic o/w nanoemulsion was used as avehicle for lipophilic drugs such as ciclosporin or latano-prost. Nonclinical studies performed in the rabbit or in therat demonstrated that these ophthalmic drug products weresafe and well tolerated following single and repeated appli-

cations, with neither inflammatory cell infiltration norapoptosis.[47,52–55] This good safety profile was confirmed byclinical trials for both drug products, and by the commer-cialization (since 2008) of the CKC cationic o/w nanoemul-sion vehicle (Cationorm) as an improved artificial tearsubstitute for the relief of mild to moderate dry eyesymptoms.[17,56–59]

Protective properties of the cationicoil-in-water nanoemulsion vehicle

The increased residence time and better spreading proper-ties of the cationic o/w nanoemulsion designed to improvethe ocular bioavailability of lipophilic drugs was accompa-nied by unexpected beneficial effects for the ocular surface.

Applications of the cationic o/w nanoemulsion helprestore the integrity of the lacrimal film through the con-comitant action of the oil and the slightly hypoosmolaraqueous phase. The oil phase of the cationic o/w nanoemul-sion by mixing with the tear film lipid layer contributes toits stability, thus reducing the evaporation of water from theaqueous phase. This is of particular interest for meibomiangland dysfunction (MGD) patients with short tear filmbreak-up times (TFBUTs) due to the lipid deficiency oftheir tears. The cationic o/w nanoemulsion vehicle Cati-onorm was able to improve keratitis (corneal fluoresceinstaining), TFBUT, and significantly reduced the symptomsof dry eye disease (DED) (Figure 5).[60] The TFBUT was sig-nificantly greater with Cationorm than with Refresh in eyeswith MGD. Cationorm was even better in MGD conditionthan in non-MGD, suggesting a positive correlation, whileRefresh showed no better efficacy in eyes with MGD.

Thus, by mechanically stabilizing the tear film the cati-onic o/w nanoemulsion confirmed its benefits for the reliefof mild to moderate dry eye.[17] Hyperosmolarity of the tearis known to be pro-inflammatory, thus the hypoosmolarityof the aqueous phase may contribute to the management ofDED signs and symptoms by transiently (upon instillation)normalizing tear osmolarity post instillation.[61,62]

More surprising were the beneficial effects of the cationico/w nanoemulsion on the wound healing process. Repeatedinstillations of the cationic o/w nanoemulsion Cationormwere demonstrated to help the wounded corneal epitheliumrecover faster than following treatments with conventionalartificial tears in a rabbit model of corneal abrasion.[17] Bothin-vitro and in-vivo data demonstrated that the CKC cati-onic o/w nanoemulsion promoted wound healing.[55] Onscraped human corneal epithelial (HCE) cells a 30 minapplication of the cationic o/w nanoemulsion was able toincrease the pace of healing, as measured by the reductionof the size of the scraped area (Figure 6). These in-vitrodata were confirmed in-vivo in a rat model of cornealscraping. Following corneal mechanical abrasion, a twice

6

5

4

3

2

1

0

20181614121086420

IVCM test score

Draize test score(a)

(b)

H4

H4

D1

D1

D4

PBS

PBS

Sco

reSc

ore

0.02%BAK Sol

0.02%BAK Sol

0.02%BAK EM

0.02%BAK EM

0.002%CKC EM

0.002%CKC EM

0.002%CKC Sol

0.002%CKC Sol

Figure 4 (a) Draize test score and (b) in-vivo confocal microscopy(IVCM) score of cationic oil-in-water nanoemulsion, at four hours (H4),day 1 (D1) and day 4 (D4) after the instillations. **P < 0.01 againstphosphate buffered saline (PBS; nonparametric comparisons (Mann-Whitney)). BAK, benzalkonium chloride; CKC, cetalkonium chloride;Sol, solution; EM, nanoemulsion. Adapted from Liang et al.[36].



Tab

le3

Sum

mar

yof

the

phys

ical

and

biol

ogic

alpr

oper

ties

ofbe

nzal

koni

umch

lorid

eso

lutio

nsan

dca

tioni

coi

l-in-

wat

erna

noem

ulsi

ons

Aqu

eous

solu

tions

ofBA

K(c

onve

ntio

nalo

cula

rdo

sage

form

s)

Cat

ioni

coi

l-in-

wat

erna

noem

ulsi

ons

with

BAK

(C12

+C14

)C

KC

(C16

)

Solu

bilit

yin

wat

erSo

lubl

eSo

lubl

ePo

orly

solu

ble

toin

solu

ble

Solu

bilit

yin

oil

Not

appl

icab

le(a

queo

usso

lutio

n)So

lubl

eSo

lubl

eZe

tapo

tent

ial

/~+

20m

V~+

40m

VSt

ruct

ural

orga

niza

tion

Mic

elle

s(1

0–20

nm)

Emu

lsio

n(o

ilna

nodr

ople

t:15

0–20

0nm

)St

abili

tyU

nsta

ble

(dyn

amic

equi

libriu

m)

Stab

le(+

+)St

able

(+++

)Lo

caliz

atio

nin

form

ulat

ion

Wat

erfr

ee-fl

owin

gm

olec

ules

ineq

uilib

rium

with

mic

ella

rst

ruct

ures

Inth

eoi

lnan

odro

plet

san

da

smal

lpr

opor

tion

inth

eaq

ueou

sph

ase

Boun

din

the

oiln

anod

ropl

ets

Func

tion

info

rmul

atio

nPr

eser

vati

vero

le(r

esul

ting

from

the

free

-flow

ing

BAK

mol

ecul

esin

the

aque

ous

phas

e[41]

Hel

pso

lubi

lize

lipop

hilic

drug

s

Cat

ion

icsu

rfac

tan

tro

le–

Stab

ilizi

ngth

ena

noem

ulsi

on–

Brin

ging

posi

tive

char

geN

op

rese

rvat

ive

role

[41]

Effe

cts

Dos

e-de

pend

ent

det

erg

ent

effe

ctw

ithde

stab

iliza

tion

ofbi

olog

ical

mem

bran

es:

–M

icro

bici

dala

gent

–Irr

itanc

yof

tissu

es

No

det

erg

ent

effe

ct(m

ost

ofth

eBA

Kis

inth

eoi

ldr

ople

ts)[4

1,43

]

No

det

erg

ent

effe

ct(C

KC

boun

dto

the

oil

drop

lets

)[41,

43]

Pres

erva

tive

actio

nPr

eser

ved

eye

drop

from

0.00

4%to

0.02

5%de

pend

ing

onfo

rmul

atio

n(0

.005

%in

Lum

igan

;0.

02%

inX

alat

an)

Un

pre

serv

edca

tioni

coi

l-in-

wat

erna

noem

ulsi

oney

edr

op

Non

clin

ical

resu

lts[2

6,36

]To

xic

for

the

ocu

lar

surf

ace

with

ocul

arirr

itatio

n,in

flam

mat

ion

and

apop

tosi

s

No

tto

xic

for

the

ocu

lar

surf

ace

No

ocul

arirr

itatio

n,no

infla

mm

atio

nan

dno

apop

tosi

sA

ssa

feas

salin

eso

lutio

nC

linic

alef

fect

[17]

Tear

film

inst

abili

ty–

Low

erte

arbr

eak-

uptim

ew

ithBA

Kco

ntai

ning

eye

drop

s[48,

49]

Ocu

lar

surf

ace

alte

rati

on

s–

Con

junc

tival

epith

elia

lcha

nges

[50]

–C

orne

alal

tera

tion

Imp

rove

dte

arfi

lmst

abili

ty–

Impr

ovem

ent

ofte

arbr

eak-

uptim

ew

ithbo

thve

hicl

ean

dC

yclo

kat

afte

rth

ree

mon

ths

oftr

eatm

ent

Imp

rove

do

cula

rsu

rfac

e–

Impr

ovem

ent

ofco

rnea

lsta

inin

gw

ithbo

thve

hicl

ean

dC

yclo

kat

afte

rsi

xm

onth

sof

trea

tmen

t[51]

Imp

rove

dte

arfi

lmst

abili

ty–

Impr

ovem

ent

ofte

arbr

eak-

uptim

ew

ithC

atio

norm

and

Cyc

loka

tIm

pro

ved

ocu

lar

surf

ace

–Im

prov

emen

tof

corn

eals

tain

ing

with

Cat

iono

rman

dC

yclo

kat

BAK

,be

nzal

koni

umch

lorid

e;C

KC

,ce

talk

oniu

mch

lorid

e.



daily treatment with the CKC cationic o/w nanoemul-sion allowed for a complete and almost scar-freere-epithelization of the cornea (Figure 7a).[55] By opposi-tion, treatment with a 0.02% BAK aqueous solutionresulted in the formation of an opaque scar underneath thehealed epithelium. These data suggested that the CKC cati-onic o/w nanoemulsion was able to promote a safe healingprocess, without the formation of opaque scar tissue withinthe cornea, as if the CKC cationic o/w nanoemulsion wasable to manage the inflammation that resulted fromthe initial mechanical corneal abrasion. The number ofinflammatory cells was then determined on fixed andhaematoxylin-eosin stained rat corneas (Figure 7b).Corneas from the CKC cationic o/w nanoemulsion-treatedgroup presented a reduced number of inflammatorycells, while in the other groups (phosphate buffered saline-or 0.02% BAK aqueous solution-treated groups) the

100

80

60

40

20

0

–20

–40 Control (PBS) 0.005% CKC EM 0.02% BAK Sol

H2

% of closure

% o

f cl

osu

re

H24

Figure 6 In-vitro scraping assay (human corneal epithelial cells) meas-uring the pace of the healing process following a 30-min treatmentwith 1/10 dilutions of 0.005% cetalkonium chloride cationic oil-in-water nanoemulsion (0.005% CKC EM) or 0.02% benzalkonium chlo-ride solution (0.02% BAK Sol). *P < 0.05 against phosphate bufferedsaline (PBS); ***P < 0.01 against PBS (two-way analysis of variance fol-lowed by Fisher adjustment). H2, two hours; H24, 24 hours. Adaptedfrom Liang et al.[55].

(a)

(b)

No MGD0

–0.5

–1

–1.5

–2

–2.5

–3

6

5

4

3

2

1

0No MGD MGD (grade 1&2)

MGD (grade 1&2)

Cationorm

Refresh

Cationorm

Refresh

Lissamine green change from baseline

Gra

de

chan

ge

TBU

T in

crea

se (

s)

TBUT improvement from baseline

Figure 5 Tear film break-up time and lissamine green staining in dryeye patients with or without meibomian gland dysfunction (MGD)treated with Cationorm or Refresh. (a) Tear break-up time (TBUT)improvement over time, and (b) lissamine green change from baselinefollowing Cationorm or Refresh treatment. *P < 0.05 (generalized esti-mating equation models, analysis of variance).

Cornea opacity surface mm2 on D5

Surf

ace

mm

2

Inflammatory cells/mm2

Cel

ls n

um

ber

PBS

PBS

(b)

(a)

0.005%CKC EM

0.005%CKC EM

0.02%BAK Sol

0.02%BAK Sol

Catioprost

Catioprost

Xalatan

Xalatan

181614121086420

3000

2500

2000

1500

1000

500

0

Figure 7 (a) Scar size and (b) inflammatory cell count in the ratcornea at day 5 (D5), after corneal scraping and five days treatmentwith cationic oil-in-water nanoemulsion. 0.02% Benzalkonium chloridesolution (0.02% BAK Sol); 0.005% cetalkonium chloride cationic oil-in-water nanoemulsion (0.005% CKC EM). *P < 0.05 against phosphatebuffered saline (PBS) (two-way analysis of variance followed by Fisheradjustment). Adapted from Liang et al.[55].



inflammatory cell count remained elevated. This clearlysuggested that the CKC cationic o/w nanoemulsion mayhave harboured anti-inflammatory properties.[55] The sameresults were obtained with Cationorm, which contains0.002% CKC as the cationic agent, and was confirmed in therabbit following repeated instillations of a 0.05%ciclosporin cationic o/w nanoemulsion.[52,63] The mecha-nism underlying these observations is currently underevaluation.[64]

Conclusions

Cationic o/w nanoemulsions represent a new developmentstrategy to improve ocular drug delivery of lipophilic com-pounds. The main issue in the development of these prod-ucts was the choice of the cationic agent. CKC was found tobe the best cationic agent to produce stable unpreservedcationic o/w nanoemulsions with unexpected beneficial bio-logical activity for the ocular surface. The use of CKC overBAK as cationic agent appears obvious (Table 3) when com-paring the physicochemical properties of these compounds.In the continuous effort to improve ocular drug delivery aCKC cationic o/w nanoemulsion was developed to improvethe precorneal residence time and the spreading propertieson negatively charged ocular surface cells of the nanoemul-sions. This better spreading and improved residence time of

the CKC cationic o/w nanoemulsion translated into atwofold increase in ciclosporin ocular bioavailability overanionic ciclosporin nanoemulsions. This new type ofvehicle was demonstrated to be perfectly safe and well toler-ated by the ocular surface. While BAK in conventionalaqueous eye drops has a preservative role, thanks to itsdetergent action, CKC in the cationic o/w nanoemulsionexhibits neither detergent effect nor preservative role. Con-sequently, CKC cationic o/w nanoemulsion does not exhibitany of the observed ocular side effects related to BAK inaqueous eye drops (tear film instability, ocular surfacedamage, mucus removal). In addition to the improved bio-availability of the loaded drug, these CKC cationic o/wnanoemulsions have ocular surface protective propertiesthrough the restoration of a healthy tear film and by favour-ing the corneal wound healing process through the promo-tion of re-epithelization and inflammation management.

Declarations

Conflict of interest

The authors are employees of Novagali Pharma SAS.

Funding

This review received no specific grant from any fundingagency in the public, commercial, or not-for-profit sectors.

References

1. Urtti A. Challenges and obstacles ofocular pharmacokinetics and drugdelivery. Adv Drug Deliv Rev 2006; 58:1131–1135.

2. Kaur IP et al. Vesicular systems inocular drug delivery: an overview. IntJ Pharm 2004; 269: 1–14.

3. du Toit LC et al. Ocular drug delivery– a look towards nanobioadhesives.Expert Opin Drug Deliv 2011; 8:71–94.

4. Abdelkader H, Alany RG. Controlledand continuous release ocular drugdelivery systems: pros and cons. CurrDrug Deliv 2012; 9: 421–430.

5. Muchtar S et al. A submicron emul-sion as ocular vehicle for delta-8-tetrahydrocannabinol: effect onintraocular pressure in rabbits. Oph-thalmic Res 1992; 24: 142–149.

6. Naveh N et al. Pilocarpine incorpo-rated into a submicron emulsionvehicle causes an unexpectedly pro-

longed ocular hypotensive effect inrabbits. J Ocul Pharmacol 1994; 10:509–520.

7. Ding S et al. Nonirritating Emulsionsfor Sensitive Tissues, I. Allergan.Patent No.: 5474979, 1995: US.

8. Vandamme TF. Microemulsions asocular drug delivery systems: recentdevelopments and future challenges.Prog Retin Eye Res 2002; 21: 15–34.

9. Tamilvanan S, Benita S. The potentialof lipid emulsion for ocular delivery oflipophilic drugs. Eur J Pharm Biop-harm 2004; 58: 357–368.

10. Royle L et al. Glycan structures ofocular surface mucins in man, rabbitand dog display species differences.Glycoconj J 2008; 25: 763–773.

11. Rabinovich-Guilatt L et al. Cationicvectors in ocular drug delivery. J DrugTarget 2004; 12: 623–633.

12. Klang S et al. The stability of piroxi-cam incorporated in a positively-charged submicron emulsion for

ocular administration. Int J Pharm1996; 132: 33–44.

13. Klang S et al. Influence of emul-sion droplet surface charge onindomethacin ocular tissue distribu-tion. Pharm Dev Technol 2000; 5: 521–532.

14. Tamilvanan S et al. Ocular delivery ofcyclosporin A. I. Design and charac-terization of cyclosporin A-loadedpositively-charged submicron emul-sion. S.T.P. Pharma Sci 2001; 11: 421–426.

15. Benita S, Elbaz E. Oil- in-Water Emul-sions of Positively Charged Particles.Yisum Research Development Com-pany of the Hebrew University ofJerusalem. Patent No.: US 6007826 A,1993.

16. Campbell PI. Toxicity of some chargedlipids used in liposome preparations.Cytobios 1983; 37: 21–26.

17. Lallemand F et al. Successfullyimproving ocular drug deliveryusing the cationic nanoemulsion



novasorb. J Drug Deliv 2012; 2012:604204.

18. Becher P. Emulsions: Theory and Prac-tice, 3rd edn. New York: Oxford Uni-versity Press, 2001.

19. Rubino JT. The influence of chargedlipids on the flocculation and coales-cence of oil-in-water emulsions. I:kinetic assessment of emulsion stabil-ity. J Parenter Sci Technol 1990; 44:210–215.

20. Opawale FO, Burgess DJ. Influence ofinterfacial properties of lipophilic sur-factants on water-in-oil emulsion sta-bility. J Colloid Interface Sci 1998; 197:142–150.

21. Constantinides PP, Scalart J-P. For-mulation and physical characteriza-tion of water-in-oil microemulsionscontaining long- versus medium-chain glycerides. Int J Pharm 1997;158: 57–68.

22. Hunter RJ. Zeta Potential in ColloidScience: Principles and Applications.London: Academic Press, 1988.

23. Rabinovich-Guilatt L et al. Extensivesurface studies help to analyse zetapotential data: the case of cationicemulsions. Chem Phys Lipids 2004;131: 1–13.

24. Klang SH et al. Physicochemical char-acterization and acute toxicity evalua-tion of a positively-chargedsubmicron emulsion vehicle. J PharmPharmacol 1994; 46: 986–993.

25. Furrer P et al. Ocular tolerance of pre-servatives and alternatives. Eur JPharm Biopharm 2002; 53: 263–280.

26. Baudouin C et al. Preservatives in eye-drops: the good, the bad and the ugly.Prog Retin Eye Res 2010; 29: 312–334.

27. Baudouin C. Detrimental effect ofpreservatives in eyedrops: implicationsfor the treatment of glaucoma. ActaOphthalmol 2008; 86: 716–726.

28. Debbasch C et al. Quaternary ammo-niums and other preservatives’ contri-bution in oxidative stress andapoptosis on Chang conjunctival cells.Invest Ophthalmol Vis Sci 2001; 42:642–652.

29. Pisella PJ et al. Conjunctival proin-flammatory and proapoptotic effectsof latanoprost and preserved andunpreserved timolol: an ex vivo and in

vitro study. Invest Ophthalmol Vis Sci2004; 45: 1360–1368.

30. Jaenen N et al. Ocular symptoms andsigns with preserved and preservative-free glaucoma medications. Eur J Oph-thalmol 2007; 17: 341–349.

31. Ammar DA et al. Effects of benzalko-nium chloride-preserved, polyquad-preserved, and sofzia-preservedtopical glaucoma medications onhuman ocular epithelial cells. AdvTher 2010; 27: 837–845.

32. Kahook MY, Ammar DA. In vitro tox-icity of topical ocular prostaglandinanalogs and preservatives on cornealepithelial cells. J Ocul Pharmacol Ther2010; 26: 259–263.

33. Ammar DA, Kahook MY. Effects ofglaucoma medications and preserva-tives on cultured human trabecularmeshwork and non-pigmented ciliaryepithelial cell lines. Br J Ophthalmol2011; 95: 1466–1469.

34. EMEA. Emea Public Statement onAntimicrobial Preservatives in Oph-thalmic Preparations for Human Use– Emea/622721/2009. 2009.

35. Freeman PD, Kahook M. Preservativesin topical ophthalmic medications:historical and clinical perspectives.Expert Rev Ophthalmol 2009; 4: 59–64.

36. Liang H et al. Reduction of quaternaryammonium-induced ocular surfacetoxicity by emulsions: an in vivostudy in rabbits. Mol Vis 2008; 14:204–216.

37. Smith MJ et al. Method for the meas-urement of the diffusion coefficientof benzalkonium chloride. Water Res2002; 36: 1423–1428.

38. Bague S et al. Oil-in-Water TypeEmulsion with Low Concentration ofCationic Agent and Positive ZetaPotential, Novagali Pharma. PatentNo.: US 8298568, 2006.

39. Rabinovich L et al. Emulsion Com-positions Containing QuaternaryAmmonium Compounds, NovagaliPharma. Patent No.: US 7973081,2008.

40. Vlachy N et al. Determining the cyto-toxicity of catanionic surfactant mix-tures on HeLa cells. Colloids Surf BBiointerfaces 2009; 70: 278–280.

41. Kurup TRR et al. Preservative require-ments in emulsions. Pharm Acta Helv1992; 67: 204–208.

42. Kazmi SJ, Mitchell AG. Preservation ofsolubilized and emulsified systems I:correlation of mathematically pre-dicted preservative availability withantimicrobial activity. J Pharm Sci1978; 67: 1260–1266.

43. Sznitowska M et al. Physicochemicalscreening of antimicrobial agents aspotential preservatives for submicronemulsions. Eur J Pharm Sci 2002; 15:489–495.

44. Macha S, Mitra AK. Ocular disposi-tion of ganciclovir and its monoesterprodrugs following intravitrealadministration using microdialysis.Drug Metab Dispos 2002; 30: 670–675.

45. Abdulrazik M et al. Ocular delivery ofcyclosporin A. II. Effect of submicronemulsion’s surface charge on oculardistribution of topical cyclosporin A.STP Pharma Sci 2001; 11: 427–432.

46. Daull P et al. Distribution ofcyclosporine A in ocular tissues aftertopical administration of cyclosporinea cationic emulsions to pigmentedrabbits. Cornea 2013; 32: 345–354.

47. Daull P et al. A comparative study of apreservative-free latanoprost cationicemulsion (Catioprost) and a BAK-preserved latanoprost solution inanimal models. J Ocul Pharmacol Ther2012; 28: 515–523.

48.Baudouin C, de Lunardo C. Short-term comparative study of topical 2%carteolol with and without benzalko-nium chloride in healthy volunteers.Br J Ophthalmol 1998; 82: 39–42.

49. Ishibashi T et al. Comparison of theshort-term effects on the humancorneal surface of topical timololmaleate with and without benzalko-nium chloride. J Glaucoma 2003; 12:486–490.

50.Ciancaglini M et al. Conjunctivalmodifications in ocular hypertensionand primary open angle glaucoma: anin vivo confocal microscopy study.Invest Ophthalmol Vis Sci 2008; 49:3042–3048.

51. Amrane M et al. The Influence onSeasonality and Subtype in Vernal



Keratoconjunctivitis (Vkc) Patients ina Randomized Clinical Trial Investi-gating Nova22007, a Preservative-FreeCyclosporine Cationic Emulsion. FortLauderdale: ARVO, 2012.

52. Liang H et al. Ocular safety of cationicemulsion of cyclosporine in an invitro corneal wound healing modeland an acute in vivo rabbit model. MolVis 2012; 18: 2195–2204.

53. Liang H et al. Comparison of theocular tolerability of a latanoprostcationic emulsion versus conventionalformulations of prostaglandins: an invivo toxicity assay. Mol Vis 2009; 15:1690–1699.

54. Daull P, Garrigue JS. Preservative-freecationic nanoemulsion of latanoprost.Ophthalmol Times Eur 2013; 9: 8–10.

55. Liang H et al. In vitro and in vivoevaluation of a preservative-free cati-

onic emulsion of latanoprost incorneal wound healing models.Cornea 2012; 31: 1319–1329.

56. Mundorf TK. Treat glaucoma patientsglobally. Adv Ocul Care 2011; 71–72.

57. Godfrey DA. Glaucoma: what’s next?Ophthalmol Times 2011; 42–43.

58. Abelson MB, Lafond A. Glaucoma anddry eye: a tough combo. Rev Ophthal-mol 2011; 108–111.

59. Amrane M et al. The Effect of Veka-cia® (Unpreserved Cyclosporine Cati-onic Emulsion) on Severe CornealInvolvement in Patients with VernalKeratoconjunctivitis Participating ina Randomized, Parallel Group Con-trolled, Clinical Trial. Genève: SOE,2011.

60. Amrane M et al. Efficacy of Cati-onorm Preservative-Free CationicEmulsion Versus Refresh in Dry Eye

Disease (Ded) Patients with/withoutMeibomian Gland Dysfunction(Mgd). Kenchoji, Kamakura, Japan:TFOS Asia, 2012.

61. Luo L et al. Hyperosmolar saline is aproinflammatory stress on the mouseocular surface. Eye Contact Lens 2005;31: 186–193.

62. Li DQ et al. Stimulation of matrixmetalloproteinases by hyperosmolar-ity via a jnk pathway in humancorneal epithelial cells. Invest Ophthal-mol Vis Sci 2004; 45: 4302–4311.

63. Daull P et al. Comparison of the anti-inflammatory effects of artificial tearsin a rat model of corneal scraping.Nice: EVER, 2012.

64. Filion MC, Phillips NC. Anti-inflammatory activity of cationiclipids. Br J Pharmacol 1997; 122: 551–557.



Documents

Benefits of cetalkonium chloride cationic oil-in-water nanoemulsions for topical ophthalmic drug delivery