MRI in patients with inflammatory bowel disease

Review

MRI in Patients With Inflammatory Bowel Disease

Michael S. Gee, MD, PhD1,2 and Mukesh G. Harisinghani, MD1*

Inflammatory bowel disease (IBD) affects �1.4 millionpeople in North America and, because of its typical earlyage of onset and episodic disease course, IBD patients of-ten undergo numerous imaging studies over the course oftheir lifetimes. Computed tomography (CT) has becomethe standard imaging modality for assessment of IBDpatients because of its widespread availability, rapidimage acquisition, and ability to evaluate intraluminaland extraluminal disease. However, repetitive CT imaginghas been associated with a significant ionizing radiationrisk to patients, making MRI an appealing alternative IBDimaging modality. Pelvic MRI is currently the imaginggold standard for detecting perianal disease, while recentstudies indicate that MRI bowel-directed techniques(enteroclysis, enterography, colonography) can accuratelyevaluate bowel inflammation in IBD. With recent technicalinnovations leading to faster and higher resolution bodyMRI, the role of MRI in IBD evaluation is likely to con-tinue to expand. Future applications include surveillanceimaging, detection of mural fibrosis, and early assessmentof therapy response.

Key Words: MRI; MR enterography; Crohn’s disease;ulcerative colitis; imagingJ. Magn. Reson. Imaging 2011;33:527–534.VC 2011 Wiley-Liss, Inc.

INFLAMMATORY BOWEL DISEASE OVERVIEW

INFLAMMATORY BOWEL DISEASE (IBD) is a boweldisorder affecting �1.4 million people in North Amer-ica, with an associated incidence of 20,000–100,000new cases developing per year (1–3). Although the pre-cise mechanisms underlying the development of IBDhave not been completely elucidated, it is widely

accepted that IBD results from an inappropriateresponse of the gastrointestinal mucosal immune sys-tem to gut flora and/or ingested food antigens (1).IBD patients can be divided into two disease forms,Crohn’s disease (CD) and ulcerative colitis (UC), whichdiffer primarily with respect to histological featuresand bowel disease distribution (4). UC is typicallyconfined to the colon and, following surgical resection,intestinal symptoms in UC patients usually resolve. Incontrast, CD patients can have disease anywhere inthe gastrointestinal tract, making medical therapy thetreatment of choice.

Both subtypes of IBD historically have demon-strated a bimodal demographic distribution with apeak incidence occurring during late adolescence andearlier adulthood, followed by a second smaller peaklater in adulthood (2,3). Both forms of IBD are charac-terized by episodic symptom recurrence over thecourse of years. UC patients are much more likely tobe cured of their abdominal symptoms because intes-tinal disease involvement is limited to the colon,providing an option for complete surgical excision ofdiseased bowel in patients refractory to medical ther-apy. In contrast, CD patients can develop diseaseanywhere in the gastrointestinal tract and often haveinvolvement of multiple discontinuous segments ofbowel. As a result, CD is more typically characterizedby long-term frequent symptomatic recurrence overthe lifetime of the patient.

ROLE OF IMAGING IN INFLAMMATORYBOWEL DISEASE

Imaging plays an important role in the evaluation ofIBD patients (5,6). Imaging plays a critical role in theinitial diagnosis of disease by providing evidence ofthe presence, as well as distribution, of abnormalbowel in patients with suspected IBD. The informa-tion provided by imaging is combined with clinical,endoscopic, and histological data to provide the diag-nosis. Imaging is of particular importance in the eval-uation of CD to assess the small bowel between theligament of Treitz and the ileocecal valve, which is notwell evaluated by endoscopy. Cross-sectional imagingmodalities (CT, MRI) can also provide informationregarding extraluminal disease complications (ab-scess, fistula, bowel perforation) likely to require moreacute intervention, as well as extraintestinal manifes-tations of IBD that can be symptomatic (primary

1Division of Abdominal Imaging and Interventional Radiology,Massachusetts General Hospital, Harvard Medical School, Boston,Massachusetts, USA.2Division of Pediatric Imaging, Department of Radiology,Massachusetts General Hospital, Harvard Medical School, Boston,Massachusetts, USA.

Contract grant sponsor: National Institutes of Health and HarvardMedical School; Contract grant number: Clinical and translationalscience award UL1 RR025758-02 (to M.S.G. and M.G.H.).

*Address reprint requests to: M.G.H., Massachusetts General Hospi-tal, Division of Abdominal Imaging and Interventional Radiology, 55Fruit St., White 270, Boston, MA 02114. E-mail: [email protected]

Received October 29, 2009; Accepted December 14, 2010.

DOI 10.1002/jmri.22504View this article online at wileyonlinelibrary.com.

JOURNAL OF MAGNETIC RESONANCE IMAGING 33:527–534 (2011)

CME

VC 2011 Wiley-Liss, Inc. 527

sclerosing cholangitis, pancreatitis, nephrolithiasis,sacroileitis) (4). Imaging also provides useful informa-tion guiding the treatment of patients with establishedIBD. Imaging features of the bowel and adjacent mes-entery correlate well with clinical indices of diseaseactivity (5) and provide noninvasive determination ofthe need for therapeutic modulation as well as an in-dicator of therapy response.

CURRENT IMAGING EVALUATION OF IBD

For years the imaging reference standard for IBD eval-uation was barium fluoroscopy. Enteroclysis andsmall bowel series were used to evaluate the smallbowel in CD, while barium enema was used to evalu-ate the colon in CD and UC. Barium fluoroscopic eval-uation of IBD relied primarily on visualization ofabnormalities in bowel mucosal pattern and intestinalcaliber (7). However, fluoroscopic techniques areinsensitive for depicting transmural inflammation orthe extraluminal complications of IBD. Over the lastdecade CT has become the primary imaging modalityfor evaluating gastrointestinal tract pathology due toits widespread availability, fast scanning time, andability to produce high-resolution 3D images (8). CTenterography (CT-E) in particular is tailored to detectbowel wall abnormalities, through the use of large vol-ume neutral enteric contrast and thin slice technique,and has become the preferred CT technique for evalu-ating IBD (9). Currently, routine CT evaluation of CDincludes assessment of bowel wall thickening, perien-teric and pericolonic mesenteric inflammation; lymphnode size and number; extraluminal collections (fistu-lae, abscesses, sinuses); and extraintestinal complica-tions (10). Although CT has proved to be an effectiveimaging modality for CD, one significant limitation is itsassociated patient exposure to ionizing radiation. Theissue of ionizing radiation risk to patients associatedwith diagnostic radiology examinations has receivedmuch attention in recent years, especially for children,in whom the relative cancer mortality risk per unit radi-ation dose is significantly higher compared with adults(11,12). This issue is particularly relevant to the IBDpopulation, which is often diagnosed at a young ageand likely to require frequent imaging over the course oftheir lifetimes. Epidemiological studies suggest a non-zero radiation-induced cancer risk at exposure levels aslow as 75 mSv, which are often exceeded in patientsdiagnosed with CD during childhood (13–15).

DEVELOPMENT OF MRI FOR INFLAMMATORYBOWEL DISEASE EVALUATION

MRI has several intrinsic advantages over other imag-ing modalities that make it desirable for evaluatingIBD. Foremost of these is the lack of ionizing radiationexposure to patients, which is of particular impor-tance to the IBD population, which is likely to requirenumerous imaging studies over the course of theirlifetimes. Because of the lack of radiation exposure,cinematic MR images can be obtained to examinebowel peristalsis over time. Additional, serial dynamic

postcontrast images can be obtained to assess thetime course of bowel enhancement. In addition, MRIprovides superior soft tissue contrast to CT in the ab-sence of intravenous contrast, which would be helpfulfor detecting mesenteric inflammatory changes andbowel wall edema. Historically, MRI of the abdomenhas been limited by long acquisition times and exten-sive motion artifact from respiration. As a result, theinitial application of MRI in IBD was pelvic MRI forevaluation of perianal disease in CD. The soft tissuecontrast of MRI is superior to CT for delineation of theanal sphincter complex and the pelvis can be imagedin multiple planes with little image degradation fromrespiratory motion. The last 5–10 years have seen thedevelopment of MRI pulse sequences that providemotion-free, high-resolution images of the body,which has made MRI of the bowel possible (16). MRIevaluation of the bowel relies predominantly on threesequences (17). The first is a single shot fast spinecho T2 sequence with half-Fourier reconstruction (eg,HASTE, SSFSE) that provides motion-free T2 weightedimages for evaluating bowel wall edema and extralumi-nal fluid collections. The second is balanced steady-state free precession (eg, True FISP, FIESTA), which isexquisitely sensitive to mesenteric changes such ashypervascularity, fibrofatty proliferation, and fistulae.The third is a dynamic fast 3D spoiled gradient echoT1 fat-suppressed postcontrast sequence (eg, VIBE,LAVA) to evaluate the pattern of bowel enhancement.Other technological innovations leading to effective MRbowel imaging include higher magnetic field strengths(1.5 T and more recently 3 T), multichannel phasedarray coils, and parallel image processing (reviewed(18)). These technological advances have led to the de-velopment and clinical implementation of MRI proto-cols designed to evaluate the small bowel (MR enterog-raphy, enteroclysis) and colon (MR colonography).

CURRENT ROLES OF MRI IN INFLAMMATORYBOWEL DISEASE EVALUATION

Pelvic MRI for Perianal Disease Evaluation

Pelvic MRI has become part of the standard imagingworkup of patients with CD and suspected perianalinvolvement (19,20). The lifetime risk of perianal fis-tula formation in CD ranges from 30%–50% (20), withthe presence of a fistula leading to significant morbid-ity due to cutaneous drainage or perianal abscess for-mation. The superior soft tissue contrast of MRI pro-vides detailed anatomic delineation of the analsphincter complex. It is important to determine theanatomic relationship of perianal fistulae with theinternal and external anal sphincters, as well as thelevator ani complex (Fig. 1), as these can affect the sur-gical approach and closure technique (21). Additionally,MRI is sensitive to detection of perianal abscessesrequiring urgent intervention.

MRI Evaluation of Bowel and EntericContrast Agents

MRI evaluation of the bowel typically combines large-volume enteric contrast distention of the bowel with

528 Gee and Harisinghani

dynamic imaging following intravenous contrastadministration to increase sensitivity for detectingbowel wall abnormalities (22). Enteric contrast can beadministered either orally (MR enterography) or vianasojejunal tube (MR enteroclysis). Several entericcontrast agents have been used for MRI in an attemptto achieve uniform luminal distension with minimalintestinal absorption. Other important considerationsfor the use of orally administered contrast includepatient acceptability and cost (23). MRI enteric con-trast agents are generally classified according to theirsignal intensity on T1- and T2-weighted images(23,24) and are categorized as positive agents thatdemonstrate high signal intensity on both T1 and T2images (eg, blueberry juice, pineapple juice), negativeagents (eg, oral superparamagnetic iron oxide par-ticles) that demonstrate low signal intensity on bothT1 and T2 images, and biphasic agents (eg, water,polyethylene glycol, dilute barium with sorbitol) thatdemonstrate low signal intensity on T1 and high sig-nal intensity on T2 images. Enteric contrast agentsare routinely used for bowel MRI studies in order todisplace intraluminal bowel gas that would otherwisecause significant image distortion on gradient echosequences typically obtained after intravenous con-trast administration. Theoretical benefits of low lumi-nal signal intensity on T2-weighted images includebetter visualization of bowel wall edema and mucosalenhancement as well as discrimination of intralumi-nal and extraluminal fluid, while positive luminal sig-nal intensity is often favored for the detection of bowelwall thickening. One study examined positive versusnegative enteric contrast agents in MR enteroclysis(25). Positive contrast agents on T2-weighted images

were found to be superior for detecting areas of bowelwall thickening. Negative contrast agents were foundto be superior for abscess detection, as low intensityintraluminal bowel fluid could be easily distinguishedfrom abscess fluid that remained T2 hyperintense.Biphasic agents have become the predominant oralcontrast agents used for bowel MRI, and typicallyinclude nonabsorbable high osmolarity additives suchas mannitol, polyethylene glycol, and sorbitol to mini-mize water absorption by bowel.

MR features of signs of active bowel inflammation inIBD include bowel wall thickening, bowel wall hyper-intensity, and hyperenhancement (26). A layered pat-tern of bowel wall enhancement consisting of brightlyenhancing mucosa (from inflammation) and hypoen-hancing submucosa (from edema) has also beenshown to be a specific sign of active bowel inflamma-tion (Fig. 2) (27,28). An advantage of MRI over bariumfluoroscopic evaluation of IBD is its ability to detectassociated mesenteric inflammatory changes sugges-tive of active disease. Mesenteric features associatedwith IBD include lymphadenopathy, vasa recta hyper-vascularity, and fibrofatty infiltration.

MR Enteroclysis

MR enteroclysis was the first dedicated MRI methodfor evaluating small bowel in CD and was based onfluoroscopic enteroclysis technique (reviewed (29)).The enteroclysis technique involves placement of anasojejunal balloon-tipped catheter under fluoro-scopic guidance followed by instillation of a large vol-ume of enteral contrast (1.5–2.0 L) through the cathe-ter, typically using a motorized pump to ensure

Figure 1. Pelvic MRI of IBD perianal disease. Representative axial T2 fat-suppressed (a,b) and postcontrast T1 fat-sup-pressed (c,d) images demonstrating an intersphincteric perianal fistula (a,c) and presacral abscess (b,d) in two patients withknown CD. Arrows indicate sites of disease.

Inflammatory Bowel Disease MRI 529

uniform distention. Balloon inflation minimizes con-trast reflux back to the stomach. Large-volume entericcontrast distention of the bowel increases sensitivityfor detecting areas of abnormal wall thickening orenhancement, which can be missed if the bowel isunderdistended (30). The instillation of contrast origi-nally was performed in the fluoroscopy suite prior topatient transfer to MRI. However, with the advent ofdynamic thick slab MRI techniques, contrast is nowroutinely instilled under real-time MR guidance untiladequate small bowel distention is achieved. Onceadequate distention is achieved, multiplanar MRbowel imaging is performed. In some institutions thepatient is placed in the scanner in a prone position tominimize abdominal cavity distention and reduce thenumber of slices needed to scan the bowel. However,supine positioning may be preferred in some patientswho are unable to tolerate scanning in a prone posi-tion. In some institutions, intravenous antiperistalticagents (eg, glucagon, butylscopolamine) are adminis-tered once maximum bowel distention is achieved inorder to minimize image degradation from motion aswell as to delay enteric contrast clearance. The MRenteroclysis protocol varies by institution and gener-ally involves the single shot T2, balanced steady state,and dynamic T1 fat-suppressed postcontrast sequen-ces described above. The coronal plane is used mostoften for imaging because of the quicker acquisitiontimes relative to the axial plane. The major advantageof MR enteroclysis over conventional fluoroscopicenteroclysis is the ability to demonstrate extraluminalmanifestations of IBD. This includes detection of mes-enteric inflammatory changes such as fibrofatty prolif-eration and mesenteric hypervascularity, as well asfistulae and abscesses.

MR Enterography

The MR enterography technique developed as a nonin-vasive alternative to enteroclysis for small bowel eval-uation, analogous to the development of CT enterogra-phy, due to the fact that a significant proportion ofpatients refuse nasojejunal catheter placement. Addi-tionally, MR enteroclysis can be logistically challeng-

ing due to the variation in time needed for nasojejunalcatheter placement and bowel distention. At high-vol-ume MRI centers, this variation can lead to issueswith MRI throughput. MR enterography relies on oralingestion of contrast by the patient to distend thesmall bowel; otherwise, the protocol is very similar toMR enteroclysis. Enteric contrast agents for MR enter-ography tend to be those that are either pleasant tast-ing or easily flavored (eg, water, blueberry/pineapplejuice, dilute barium with sorbitol [VoLumen, BraccoDiagnostics, Princeton, NJ]).

At our institution, patients receive an oral entericcontrast preparation consisting of 900 mL of dilutebarium with sorbitol (VoLumen) mixed with 300 mL offerumoxsil iron oxide suspension (GastroMARK,Tyco). The addition of the iron oxide suspension dark-ens the lumen on both T2 and T1 postcontrastimages, which we feel aids in the detection of bowelwall edema and abnormal mucosal enhancement.Oral contrast is ingested continuously over a period of45 minutes prior to the exam. Patients are scanned inthe supine position using a multichannel torsophased array coil. Imaging sequences include coronaland axial single-shot fast spin-echo T2 (SSFSE/HASTE), coronal balanced steady-state free precession(FIESTA/True-FISP), axial fast relaxation fast spin-echo T2 (FRFSE/RESTORE) with fat suppression, coro-nal 3D T1 with fat suppression before and followinggadolinium administration (0.2 mmol/kg of gadopente-tate, Bayer injected at 3 mL/sec). Dynamic postcontrastimages are acquired at 1, 3, and 5 minutes postcon-trast. This time period of postcontrast imaging has beenshown to be sensitive for visualizing the progressivetransmural bowel wall enhancement pattern indicativeof active inflammation (28). Afterward, postcontrastaxial 2D fast spoiled-gradient-recalled echo T1-weightedimages with fat suppression are acquired.

A recent prospective study compared MR enterogra-phy and MR enteroclysis in CD evaluation (31). Thisstudy involved 40 patients with histologically provenIBD assigned to undergo either MR enteroclysis or MRenterography. This study demonstrated that MRenteroclysis was superior to MR enterography forbowel distention and detection of mucosal bowel

Figure 2. MR enterographic detection of active bowel inflammation. Coronal T2 fat-suppressed (a) and postcontrast T1 fat-suppressed (b) images from an MR enterography study on a CD patient demonstrate an area of thickened terminal ileum(arrows) exhibiting T2 hyperintensity and layered enhancement consistent with active disease. Corresponding surgical bowelexcision specimen from the same patient (c) demonstrates neutrophilic invasion of mucosal crypts consistent with activeinflammatory changes.


abnormality. The two techniques were comparable fordetection of luminal narrowing, mesenteric abnormal-ity, and fistulae. The authors suggest that MR enter-oclysis would be the preferred imaging modality forCD evaluation but that MR enterography is an accept-able alternative in patients who are unable to toleratenasojejunal intubation.

Cine MRI Evaluation of Bowel

Historically, one radiographic sign of IBD by fluoro-scopic examination has been decreased peristalsis ofa segment of bowel. Traditional MRI evaluation of thebody has relied on analysis of static images; however,with the advent of rapid MRI techniques, real-timecine MRI evaluation of the bowel is now possible.Most often, the cine sequence used is a coronal thickslab balanced steady-state sequence (eg, FIESTA,True FISP) in which a single volume of the abdomen10 mm or thicker is continuously imaged over a pe-riod of seconds to evaluate peristaltic motion. Onestudy compared cine MRI using a 17-second balancedsteady-state dynamic acquisition with conventionalMR enterography in 40 patients with established CD(32). The cine sequence was used to identify segmentsof abnormal bowel motility, defined as zones of abnor-mally increased or decreased peristaltic motion com-pared with adjacent bowel. The addition of the cinesequence led to detection of an increased number ofabnormal bowel segments compared with static MRenterography images alone. The authors believe thatalteration in bowel motility is an early imaging sign ofCD involvement of bowel and helps to identify abnor-mal bowel segments with subtle signs of inflammationon static images. A significant limitation of this studyis the lack of endoscopic or histological verification ofimaging findings.

MR Colonography

MR colonography is a specific MRI technique for eval-uating the colon that combines retrograde instillationof water with intravenous contrast and thin sectionimage acquisition (reviewed (33)). MR colonographywas initially applied to imaging surveillance of colo-rectal cancer and its use in IBD was initially reservedfor cases where optic colonoscopy was incomplete dueto technical difficulty or patient intolerance. However,MR colonography does offer some advantages overendoscopy including evaluation of submucosal andmesenteric involvement, and its noninvasive naturemeans there is no risk of colonic perforation in patientspresenting with acute symptoms. Prior to MR colonog-raphy, patients typically undergo a bowel cleansing reg-imen similar to that used for colonoscopy. Unlike CTcolonography, which uses colonic insufflation with air,MR colonography typically involves colonic distensionvia warm water enema to avoid susceptibility artifactsthat could degrade image quality. MR colonographicfeatures of IBD are similar to those seen on MR enter-ography, including bowel wall thickening and edema,mural hyperenhancement, lymphadenopathy, andmesenteric inflammatory changes.

An additional potential benefit of MR colonographyis in helping to distinguish between CD and UC incases of indeterminate colitis. In these cases, MRI isadvantageous for detecting extraluminal disease man-ifestations such as fistulae and abscesses that aremore suggestive of CD. Additionally, because MR colo-nography also images the small bowel, detection ofsmall bowel disease also indicates CD rather thanUC. A 2005 study of 22 patients with known or sus-pected IBD who underwent MR colonography withpositive rectal contrast (gadolinium/water mixture)followed immediately by optical colonoscopy (34).Compared with endoscopic reference, MR colono-graphic sensitivity for detecting colonic involvement ofIBD on a per bowel segment basis was 31.6% for CDand 58.8% for UC. The authors concluded from thisstudy that MR colonography was not suitable forassessing the extent of colonic inflammation in IBD. Asecond study from the same year evaluated 15 normalsubjects and 23 subjects with suspected IBD by MRcolonography (35) to detect colonic inflammation,using endoscopic biopsy as reference. MR colono-graphic assessment of colonic inflammation (based onabnormal enhancement, wall thickening, lymphade-nopathy, and loss of haustral folds) was shown to be87% sensitive and 100% specific compared with histo-logic standard. Based on these data, MR colonographyis considered a promising noninvasive method formonitoring IBD activity and therapeutic efficacy.

COMPARISON OF MRI WITH OTHER IMAGINGMODALITIES

The MRI protocol most frequently compared withother modalities has been MR enterography. A num-ber of recent prospective studies have demonstratedMR enterography to be at least comparable to otherimaging modalities for detection of small bowel dis-ease in CD patients (Figure 3). One study comparingcontrast-enhanced MRI and CT performed on differentdays in adult CD patients demonstrated MRI to besuperior for detection of subtle bowel inflammatorychanges (36). Another study in 2005 compared MRenterography with fluoroscopic barium small bowelseries (37). In this study, 30 subjects with establishedCD referred for barium small bowel series also under-went MR enterography using dilute barium within 28days to assess for concordance of the two imagingmodalities. In 18 subjects the two modalities demon-strated similar CD imaging features (10 of the sub-jects were normal). Among the 12 subjects with dis-cordant results, small bowel follow-through (SBFT)demonstrated additional strictures or fistulae in four,while MRI demonstrated additional information ineight by identifying active inflammation in strictureareas (based on abnormal wall enhancement or mes-enteric inflammatory changes). No endoscopic or his-tologic validation was included in this study. A 2009study examined 30 patients with established CD whounderwent MR enterography and CT enterography thesame day, followed by barium SBFT and ileocolono-scopy within 1 week (38). A 1% sorbitol oral contrast


solution was used for both CT and MRI. Receiver op-erator curve analysis demonstrated all three imagingmodalities to be similar for detection of terminal ileumactive inflammation using endoscopic evaluation asthe reference standard (area under curve values rang-ing from 0.88–0.95). CT and MRI both demonstratedsuperior accuracy to SBFT for detection of extraen-teric CD complications including fistulae, sinustracts, and abscesses, when compared with physicalexam or surgical reference. Both CT and MRI alsodemonstrated more segments of active small bowelinflammation proximal to the terminal ileum thanSBFT, although the results were not statistically sig-nificant. A second study from 2009 prospectively com-pared MR enterography and CT enterography per-formed the same day on 30 patients with suspectedCD using ileocolonoscopic findings performed within30 days as reference standard. This study demon-strated the two modalities to have comparable sensi-tivity (MRI 90.5%, CT 95.2%) for detecting active smallbowel inflammation in adults with CD (39).

Wireless capsule endoscopy (WCE) is another mini-mally invasive technique that has been developedrecently for small bowel evaluation (40). This tech-nique involves a video capsule endoscope that is swal-lowed and propelled through the gastrointestinal tractby peristalsis. The endoscope captures images of thesmall bowel that are transmitted to aerials taped onthe body and then stored on a portable recorder. Thistechnique provides endoluminal evaluation of theentire small bowel. Advantages of this technique overMRI include the ability to detect subtle mucosalabnormalities. Disadvantages include potential inabil-

ity to pass the capsule endoscope in patients withsmall bowel strictures, as well as inability of WCE tovisualize extraluminal complications of CD. Very fewstudies have directly compared WCE to MRI for smallbowel evaluation. One study (41) compared WCE withMR enteroclysis in patients with suspected smallbowel disease. A total of 17 patients with known orsuspected CD were imaged. WCE depicted a highernumber of inflammatory lesions in the jejunum andproximal ileum compared with MR enteroclysis, whilethe two modalities demonstrated a similar number ofinflammatory lesions in the terminal ileum. Thisstudy did not include any histologic or endoscopicvalidation of findings.

NEW AND EMERGING ROLES OF MRI IN IBD

MRI has been shown to be sensitive for detecting cer-tain aspects of CD such as small bowel inflammation,perianal fistulae, and abscesses. Recent technicaladvances in body MRI, including higher magnet fieldstrength, parallel image processing, and motion arti-fact reduction techniques, should lead to shorter scantimes and increased spatial resolution for detectingsubtle inflammatory changes (18). Such advancesshould make it possible for MRI to replace CT as theprimary imaging modality for CD patients in the nearfuture. Such surveillance imaging would extend therole of MRI beyond its current indications to includedetection of colonic inflammatory changes, as well asextraluminal imaging features of disease such asmesenteric lymphadenopathy, ascites, and fibrofattyproliferative changes. The incorporation of MRI into

Figure 3. MRI detection of IBD imaging features. Corresponding image pairs from contrast enhanced CT (top row) and MRenterography (bottom row) studies on the same patients with known IBD demonstrate bowel wall thickening (a,d), intramus-cular abscesses (b,e), and mesenteric lymphadenopathy (c,f). Arrows indicate the abnormalities.


routine CD surveillance would lead to a significantreduction in patient lifetime radiation exposure, themajority of which currently derives from CT (14).

Another future role for MRI in CD is the detection ofmural fibrosis. Most studies examining the accuracyof MRI for assessing CD activity focus on the distinc-tion between active and inactive inflammation (26,39),with the presence of active disease considered to bean indication to initiate or modify medical therapy. Acomplementary approach would be to take advantageof the soft tissue contrast of MRI to detect mural fi-brosis, which would helpful for selecting patientslikely to require surgical bowel resection. Fibrosis isconsidered a late-phase irreversible result of chronicbowel wall inflammation leading to collagen fiber dep-osition in the submucosal and serosal layers of thebowel wall. Mural fibrosis frequently leads to luminalnarrowing associated with proximal bowel obstructionand, unlike acute inflammatory strictures, fibroticstrictures usually require surgical resection to allevi-ate the associated obstruction. Reported MRI findingsassociated with mural fibrosis include bowel wall T2hypointensity (42) and lack of enhancement (27).Early detection of mural fibrosis in CD patients poten-tially would be useful to facilitate surgical resection ofirrevocably diseased bowel segments, thereby reduc-ing the number of symptomatic recurrences.

MRI protocols have been developed successfully forevaluation of the small bowel (MR enterography) or thecolon (MR colonography). A future challenge would bethe development of a single combined protocol for eval-uation of both small and large bowel. Such a protocolwould be particularly useful in the CD population,which can develop areas of bowel inflammation any-where in the gastrointestinal tract. A 2005 study (35)combined MR enterography with 1.5 L of oral contrastand rectal distention via 500–1000 mL water enema in20 patients with known CD and compared it with 20patients who underwent MR enterography without rec-tal distention. Comparison of MRI findings was madewith colonoscopy performed within 7 days of the MRI.The addition of rectal water was associated withimproved distention of both the terminal ileum and therectum compared with oral contrast alone. Diagnosticaccuracy for inflammation of both the terminal ileumand the colon was improved with rectal enema adminis-tration. One issue with such combined enteric contrastadministration is patient compliance. Large-volume oralcontrast administration alone makes some IBD patientsdistended and uncomfortable, and the addition of rectalcontrast is likely to be less well accepted. An entericcontrast regimen leading to distention of small andlarge bowel by oral administration alone would be ideal.

A final emerging application of MRI in IBD would beits use as a biomarker of therapeutic response. Treat-ment of IBD has been revolutionized due to the recentintroduction of biologic therapies targeting molecularpathways thought to contribute to bowel inflammation,such as the proinflammatory cytokine tumor necrosisfactor alpha (TNF-a), lymphocyte signaling moleculesCTLA-4 and CD20, and the a4 integrin adhesion mole-cule mediating leukocyte migration (43,44). Theseagents are generally considered more specific for IBD

compared with traditional corticosteroids or immuno-modulatory agents. Indications for biologic agentsinclude patients with IBD refractory to standard thera-pies, patients who are steroid-dependent, and patientswith draining fistulae or systemic extraintestinal dis-ease manifestations. Most of the biologic agents cur-rently in clinical practice are either recombinant pep-tides or chimeric antibodies, which are more expensiveto produce compared with traditional compound-baseddrugs (44), meaning that therapy with these agentscan be associated with high financial costs to thepatients and the healthcare system as a whole. Anearly noninvasive assessment of treatment responsepotentially would be of great financial benefit topatients undergoing biologic therapy by ensuring thatpatients who are refractory to treatment do not remainon medication for longer than necessary. Additionally,biologic agents have their own unique side effect profileincluding increased risk of serious infections, as well asrarer side effects such as neurologic disorders, conges-tive heart failure, and hematologic malignancies. Muchrecent attention has been focused on cases of hepatos-plenic T-cell lymphoma observed in young adult IBDpatients treated with a combination of biologic andimmunomodulatory agents (45). An early imagingassessment of treatment response or failure would alsobe beneficial to minimize potential side effects associ-ated with unnecessary prolonged treatment. MRI isparticularly well suited to serve as an imaging bio-marker of therapeutic efficacy because of its lack ofionizing radiation, which makes it an ideal modality forrepeated assessment before and during treatment.

CONCLUSION

The role of MRI in the assessment of IBD continues toexpand due to its lack of ionizing radiation exposureand superior soft tissue contrast. MRI currently is themodality of choice for detecting perianal inflammationand fistulae, as well as extraintestinal disease mani-festations. Recent evidence suggests a role for MRI inthe detection of active small bowel inflammation inpatients with known IBD. Other potential roles forMRI in IBD evaluation include detection of mural fi-brosis and early assessment of treatment response.As the spatial resolution and scanning time of MRIcontinue to improve as a result of technical innova-tion, MRI will likely also prove to be suitable as theprimary modality for IBD imaging surveillance.

ACKNOWLEDGMENTS

The authors thank Dr. Katherine Nimkin for assis-tance with creation of MRI protocols for evaluatinginflammatory bowel disease.

REFERENCES

1. Baumgart DC, Carding SR. Inflammatory bowel disease: causeand immunobiology. Lancet 2007;369:1627–1640.

2. Friedman S, Blumberg RS. Inflammatory bowel disease. In:Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL,Jameson JL, eds. Harrison’s principles of internal medicine,15th ed. New York: McGraw-Hill, 2001. pp 1679–1692.


3. Loftus EV Jr. Clinical epidemiology of inflammatory bowel dis-ease: incidence, prevalence, and environmental influences. Gas-troenterology 2004;126:1504–1517.

4. Baumgart DC, Sandborn WJ. Inflammatory bowel disease: clini-cal aspects and established and evolving therapies. Lancet 2007;369:1641–1657.

5. Patak MA, Mortele KJ, Ros PR. Multidetector row CT of the smallbowel. Radiol Clin North Am 2005;43:1063–1077, viii.

6. Mackalski BA, Bernstein CN. New diagnostic imaging tools forinflammatory bowel disease. Gut 2006;55:733–741.

7. Carucci LR, Levine MS. Radiographic imaging of inflammatorybowel disease. Gastroenterol Clin North Am 2002;31:93–117, ix.

8. Horton KM, Fishman EK. The current status of multidetector rowCT and three-dimensional imaging of the small bowel. Radiol ClinNorth Am 2003;41:199–212.

9. Paulsen SR, Huprich JE, Fletcher JG, et al. CT enterography as adiagnostic tool in evaluating small bowel disorders: review of clin-ical experience with over 700 cases. Radiographics 2006;26:641–657; discussion 657–662.

10. Gore RM, Balthazar EJ, Ghahremani GG, Miller FH. CT featuresof ulcerative colitis and Crohn’s disease. AJR Am J Roentgenol1996;167:3–15.

11. Brenner DJ, Elliston CD, Hall EJ, Berdon WE. Estimates ofthe cancer risks from pediatric CT radiation are not merely the-oretical: comment on ‘‘point/counterpoint: in x-ray computedtomography, technique factors should be selected appropriateto patient size. against the proposition.’’ Med Phys 2001;28:2387–2388.

12. Rice HE, Frush DP, Farmer D, Waldhausen JH. Review of radia-tion risks from computed tomography: essentials for the pediatricsurgeon. J Pediatr Surg 2007;42:603–607.

13. Brenner DJ. Should computed tomography be the modality ofchoice for imaging Crohn’s disease in children? The radiation riskperspective. Gut 2008;57:1489–1490.

14. Desmond AN, O’Regan K, Curran C, et al. Crohn’s disease: fac-tors associated with exposure to high levels of diagnostic radia-tion. Gut 2008;57:1524–1529.

15. Pierce DA, Preston DL. Radiation-related cancer risks at low dosesamong atomic bomb survivors. Radiat Res 2000;154:178–186.

16. Furukawa A, Saotome T, Yamasaki M, et al. Cross-sectionalimaging in Crohn disease. Radiographics 2004;24:689–702.

17. Gourtsoyiannis NC, Papanikolaou N, Karantanas A. Magneticresonance imaging evaluation of small intestinal Crohn’s disease.Best Pract Res Clin Gastroenterol 2006;20:137–156.

18. MacKenzie JD, Vasanawala SS. Advances in pediatric MR imag-ing. Magn Reson Imaging Clin N Am 2008;16:385–402, v.

19. Koelbel G, Schmiedl U, Majer MC, et al. Diagnosis of fistulae andsinus tracts in patients with Crohn disease: value of MR imaging.AJR Am J Roentgenol 1989;152:999–1003.

20. Szurowska E, Wypych J, Izycka-Swieszewska E. Perianal fistulasin Crohn’s disease: MRI diagnosis and surgical planning: MRI infistulazing perianal Crohn’s disease. Abdom Imaging 2007 [Epubahead of print].

21. Halligan S, Buchanan G. MR imaging of fistula-in-ano. Eur JRadiol 2003;47:98–107.

22. Fidler J. MR imaging of the small bowel. Radiol Clin North Am2007;45:317–331.

23. Laghi A, Paolantonio P, Iafrate F, Altomari F, Miglio C, PassarielloR. Oral contrast agents for magnetic resonance imaging of thebowel. Top Magn Reson Imaging 2002;13:389–396.

24. Fidler JL, Guimaraes L, Einstein DM. MR imaging of the smallbowel. Radiographics 2009;29:1811–1825.

25. Rieber A, Aschoff A, Nussle K, et al. MRI in the diagnosis of smallbowel disease: use of positive and negative oral contrast media incombination with enteroclysis. Eur Radiol 2000;10:1377–1382.

26. Maccioni F, Viscido A, Broglia L, et al. Evaluation of Crohn dis-ease activity with magnetic resonance imaging. Abdom Imaging2000;25:219–228.

27. Koh DM, Miao Y, Chinn RJ, et al. MR imaging evaluation of the activ-ity of Crohn’s disease. AJR Am J Roentgenol 2001;177:1325–1332.

28. Del Vescovo R, Sansoni I, Caviglia R, et al. Dynamic contrastenhanced magnetic resonance imaging of the terminal ileum: dif-ferentiation of activity of Crohn’s disease. Abdom Imaging 2008;33:417–424.

29. Prassopoulos P, Papanikolaou N, Grammatikakis J, Rousomous-takaki M, Maris T, Gourtsoyiannis N. MR enteroclysis imaging ofCrohn disease. Radiographics 2001;21(Spec No):S161–172.

30. Low RN, Francis IR. MR imaging of the gastrointestinal tract withi.v., gadolinium and diluted barium oral contrast media com-pared with unenhanced MR imaging and CT. AJR Am J Roent-genol 1997;169:1051–1059.

31. Masselli G, Casciani E, Polettini E, Gualdi G. Comparison of MRenteroclysis with MR enterography and conventional enteroclysisin patients with Crohn’s disease. Eur Radiol 2008;18:438–447.

32. Froehlich JM, Waldherr C, Stoupis C, Erturk SM, Patak MA. MRmotility imaging in Crohn’s disease improves lesion detection com-pared with standard MR imaging. Eur Radiol 2010;20:1945–1951.

33. Rimola J, Rodriguez S, Garcia-Bosch O, et al. Role of 3.0-T MRcolonography in the evaluation of inflammatory bowel disease.Radiographics 2009;29:701–719.

34. Schreyer AG, Rath HC, Kikinis R, et al. Comparison of magnetic res-onance imaging colonography with conventional colonoscopy for theassessment of intestinal inflammation in patients with inflamma-tory bowel disease: a feasibility study. Gut 2005;54:250–256.

35. Ajaj W, Lauenstein TC, Langhorst J, et al. Small bowel hydro-MRimaging for optimized ileocecal distension in Crohn’s disease:should an additional rectal enema filling be performed? J MagnReson Imaging 2005;22:92–100.

36. Low RN, Francis IR, Politoske D, Bennett M. Crohn’s disease evalua-tion: comparison of contrast-enhanced MR imaging and single-phasehelical CT scanning. J Magn Reson Imaging 2000;11:127–135.

37. Bernstein CN, Greenberg H, Boult I, Chubey S, Leblanc C, RynerL. A prospective comparison study of MRI versus small bowel fol-low-through in recurrent Crohn’s disease. Am J Gastroenterol2005;100:2493–2502.

38. Lee SS, Kim AY, Yang SK, et al. Crohn disease of the small bowel:comparison of CT enterography, MR enterography, and small-bowel follow-through as diagnostic techniques. Radiology 2009;251:751–761.

39. Siddiki HA, Fidler JL, Fletcher JG, et al. Prospective comparison ofstate-of-the-art MR enterography and CT enterography in small-bowel Crohn’s disease. AJR Am J Roentgenol 2009;193:113–121.

40. Iddan G, Meron G, Glukhovsky A, Swain P. Wireless capsule en-doscopy. Nature 2000;405:417.

41. Golder SK, Schreyer AG, Endlicher E, et al. Comparison of cap-sule endoscopy and magnetic resonance (MR) enteroclysis in sus-pected small bowel disease. Int J Colorectal Dis 2006;21:97–104.

42. Maglinte DD, Gourtsoyiannis N, Rex D, Howard TJ, Kelvin FM.Classification of small bowel Crohn’s subtypes based on multimo-dality imaging. Radiol Clin North Am 2003;41:285–303.

43. Clark M, Colombel JF, Feagan BC, et al. American gastroentero-logical association consensus development conference on the useof biologics in the treatment of inflammatory bowel disease, June21–23, 2006. Gastroenterology 2007;133:312–339.

44. Bosani M, Ardizzone S, Porro GB. Biologic targeting in the treat-ment of inflammatory bowel diseases. Biologics 2009;3:77–97.

45. Cucchiara S, Escher JC, Hildebrand H, Amil-Dias J, Stronati L,Ruemmele FM. Pediatric inflammatory bowel diseases and therisk of lymphoma: should we revise our treatment strategies?J Pediatr Gastroenterol Nutr 2009;48:257–267.


Documents

MRI in patients with inflammatory bowel disease