Secondary Logo

Share this article on:

Comparison of magnetic resonance enterography and video capsule endoscopy in evaluating small bowel disease

Crook, David W.a; Knuesel, Patrick R.e; Froehlich, Johannes M.b; Eigenmann, Franzc; Unterweger, Martind; Beer, Hans-Juergc; Kubik-Huch, Rahel A.d

European Journal of Gastroenterology & Hepatology: January 2009 - Volume 21 - Issue 1 - p 54-65
doi: 10.1097/MEG.0b013e32830ce7a7
Original Articles: Small Intestine

Purpose The goal of this study was to compare magnetic resonance enterography (MRE) and video capsule endoscopy (VCE) in suspected small bowel disease.

Materials and methods Nineteen patients with suspected small bowel disease participated in a prospective clinical comparison of MRE versus VCE. Both methods were evaluated separately and in conjunction with respect to a combined diagnostic endpoint based on clinical, laboratory, surgical, and histopathological findings. The Fisher's exact and κ tests were used in comparing MRE and VCE.

Results Small bowel pathologies were found in 15 out of 19 patients: Crohn's disease (n=5), lymphoma (n=4), lymphangioma (n=1), adenocarcinoma (n=1), postradiation enteropathy (n=1), NSAID-induced enteropathy (n=1), angiodysplasia (n=1), and small bowel adhesions (n=1). VCE and MRE separately and in conjunction showed sensitivities of 92.9, 71.4, and 100% and specificities of 80, 60, and 80% (κ=0.73 vs. κ=0.29; P=0.31/κ=0.85), respectively. In four patients, VCE depicted mucosal pathologies missed by MRE. MRE revealed 19 extraenteric findings in 11 patients as well as small bowel adhesions not detected on VCE (n=1).

Conclusion VCE can readily depict and characterize subtle mucosal lesions missed at MRE, whereas MRE yields additional mural, perienteric, and extraenteric information. Thus, VCE and MRE appear to be complementary methods which, when used in conjunction, may better characterize suspected small bowel disease.

aDepartment of Nuclear Medicine, University Hospital of Lausanne, Lausanne

bGuerbet AG, Zürich

Departments of cInternal Medicine

dRadiology, Kantonsspital Baden AG, Baden, Switzerland

eDepartment of Radiology, Sunnybrook Health Sciences Centre, Toronto, Canada

Correspondence to David W. Crook, MD, Service de Médicine Nucléaire, Centre Hospitalier Universitaire Vaudois, Rue du Bugnon 46, Lausanne 1011, Switzerland

Tel: +41 21 314 76 35; e-mail: dcrook@hotmail.com

or

Patrick R. Knuessel, Department of Medical Imaging, University of Toronto, University Health Network and Mt Sinai Hospital, Toronto, Canada

Received 3 February 2008 Accepted 13 June 2008

Back to Top | Article Outline

Introduction

The small bowel has been a challenging region to evaluate clinically and by imaging. As endoscopic techniques have been usually limited to the most proximal and distal regions, small bowel follow through and conventional enteroclysis have long been the mainstay of small bowel assessment. The diagnostic yield of these methods, however, is low, particularly in detecting sources of obscure gastrointestinal (GI) bleeding, early inflammatory changes as well as small mass lesions, and their diagnostic accuracy is to a relatively high degree dependent on examiner experience [1,2]. Moreover, conventional enteroclysis entails exposure to ionizing radiation, is semi-invasive, and may not be comfortable for a significant number of patients [3]. Multidetector computer tomography (CT) as well as magnetic resonance enterographic (MRE) techniques are emerging modalities, increasingly supplanting conventional radiological methods in the evaluation of small bowel pathology, whereas providing vast extraluminal information [4].

More recently, the development of wireless video capsule endoscopy (VCE) in 2000 has been widely heralded as a breakthrough for endoluminal visualization of the entire small bowel, the ‘last frontier’ in gastroenterology, with over 340 000 examinations performed to date (http://www.givenimaging.com/Cultures/enUS/Given/English/About/NewsAndEvents/PressReleases/). VCE has since demonstrated an unprecedented high sensitivity in detecting sources of obscure bleeding and is widely established in the work-up algorithm of these patients [5]. Additionally, VCE has shown promise in the evaluation of suspected Crohn's disease and small bowel tumors [6,7]. A number of studies have compared VCE with conventional enteroclysis, however, data comparing MRE with VCE are limited to relatively small study cohorts in the assessment of Crohn's disease and obscure GI bleeding and is largely lacking for other small bowel indications [8–12]. The aim of this study was to prospectively assess VCE and MRE in the diagnostic work-up of patients with a broad range of indications in suspected small bowel disease.

Back to Top | Article Outline

Materials and methods

Patients

Nineteen consecutive patients (mean age 56 years old; range 25–83 years old, 11 female, eight male) were recruited between March 2003 and April 2006 during work-up for suspected small bowel disease after referral to our hospital's gastroenterology service. Indications are listed in Table 1. VCE and MRE were performed within an average of 7.2 days (range 1–21 days) and the sequence of the examinations was variable based on capacity and scheduling limitations. No treatments or interventions were undertaken in any patients in the interval between MRE and VCE.

Table 1

Table 1

The study design was approved by our hospital's ethics committee. Informed written consent was obtained from all patients in accordance with our institutional review board guidelines. None declined participation or were excluded. VCE exclusion criteria included pregnancy, dysphagia, bowel obstruction, and the presence of an implanted electromedical device (i.e. pacemaker, defibrillator, spinal electrodes). MRE exclusion criteria included pregnancy, implanted electromedical devices (i.e. pacemaker, defibrillator, and spinal electrodes), ferromagnetic cardiac valves prostheses, ferromagnetic aneurysm clips as well as metal fragments located within the orbit or cranium (i.e. shrapnel).

Back to Top | Article Outline

Magnetic resonance enterographic procedure

All examinations were performed with a 1.5 Tesla scanner (Sonata; Siemens Medical Solutions, Erlangen, Germany) equipped with a phased array surface coil. Patients were asked to fast overnight and to drink 1 l of a 3% mannitol solution beginning 1 h before imaging. When possible, imaging was performed in the prone (n=13) position to facilitate separation of the bowel loops and to decrease the anterior–posterior diameter of the abdomen for faster coronal imaging. Bowel motility was assessed with several two-dimensional coronal true-fast imaging with steady-state precession (FISP) sequences during breath holding with the following parameters: TR/TE 4.3/2.1 ms, flip angle (FA) 77°, field of view (FOV) 250 mm, matrix 308×384, slice thickness 10 mm. Subsequently, 20 mg of N-butyl-scopolamine (Buscopan; Boehringer Pharma, Ingelheim, Germany) were administered intravenously (i.v.) in all patients (no contraindications found).

Axial T2-weighted half-fourier acquisition single shot turbo (SE; TR/TE 1110/96 ms, FA 150°, slice thickness 3 mm, gap 0.9 mm, FOV 400 mm, matrix 436×512) and two-dimensional fat suppressed true-FISP sequences were acquired coronally (TR/TE 4.3/2.1 ms, FA 63°, slice thickness 3 mm, gap 0.9 mm, FOV 350 mm, matrix 440×512), after an axial T1-weighted two-dimensional fat suppressed true-FISP sequence (TR/TE 4.3/2.2 ms, FA 63°, slice thickness 3 mm, gap 0.5 mm, FOV 350 mm, matrix 300×512).

After a second i.v. dose of 20 mg Buscopan, high resolution three-dimensional fat suppressed T1 weighted volumetric interpolated breath hold examination data sets (TR/TE 3.25/1.42 ms, FA 12°; 60 slices with a thickness of 1.5 mm, FOV 400, matrix 197×256) were acquired during subsequent single 20 s breath holds: precontrast, 20–40, 60–80, and 120–140 s after an i.v. bolus injection of gadolinum contrast agent (gadoterate, Gd-DOTA, dose of 0.2 mmol/kg body weight=0.4 ml/kg; Guerbet, Roissy, France). Finally, a T1-weighted two-dimensional fat-suppressed FLASH sequence was performed axially (TR/TE 159/2.64 ms, FA 80°; slice thickness 6 mm, gap 1.8 mm, FOV 370 mm, matrix 228×512) and coronally (TR/TE 113/2.5 ms, FA 80°; slice thickness 6 mm, gap 1.5 mm, FOV 370 mm, matrix 332×512).

Back to Top | Article Outline

Magnetic resonance enterographic image analysis

MR image interpretation was performed by two experienced radiologists in consensus (M.U. and P.K.). The readers were blinded to the results of VCE but had access to pertinent clinical information for each case as indicated on the imaging requisite forms from referring clinicians (i.e. history, abnormal clinical and laboratory findings, and rationale for imaging).

MRE were evaluated for segmental motility, luminal alterations, and wall thickening. Bowel wall thickness greater than 3 mm was considered pathologic. Dynamic contrast enhancement patterns of the small bowel were studied qualitatively and quantified in abnormal segments using three-dimensional volumetric interpolated breath hold examination data sets obtained precontrast and 20–40, 60–80, and 120–140 s after i.v. Gd-DOTA administration. Signal enhancement was measured on single slices at each time point in the small bowel (duodenum, proximal jejunum, middle ileum and distal ileum as well as in any suspicious segments) as well as in the aorta, psoas muscle, and background to calculate signal-to-noise and contrast-to-noise ratios. Relatively high signal-to-noise and contrast-to-noise values at 20–40 and 60–80 s were suggestive of active inflammation based on a recent unpublished study [13].

In addition, perienteric changes were sought (e.g. comb sign, fistula, abscess, lymph nodes, metastases, etc.). MRE findings considered highly suggestive for respective disease entities are listed as follows: tumor-mass lesion; Crohn's disease (some degree of clinical suspicion a precondition) – segmental wall thickening (>4 mm), wall enhancement, positive ‘comb sign’, ‘lead pipe’ type changes with loss of motility on two-dimensional coronal true-FISP dynamic sequences, stenoses; radiation enteritis (high degree of clinical suspicion and earlier history of radiation therapy a precondition) – stenoses, segmental wall thickening, wall enhancement. Findings considered consistent with Crohn's disease included mild wall thickening (3–4 mm), slight early and moderate late wall enhancement, and reduced segmental motility on two-dimensional coronal true-FISP dynamic sequences. All nonsmall bowel related MR findings were excluded from statistical analysis as this falls outside the scope of this study and would result in bias against VCE.

Factors limiting the quality of the exam, such as motion artifacts and suboptimal bowel distension, were assessed in each examination. Bowel distension was assessed in each of the main four small bowel landmarks (duodenum, proximal jejunum, middle ileum, and distal ileum) based on their respective locations and the appearance of their folds, whereas overall image quality was graded on a five-point scale as nondiagnostic, poor, fair, good, or excellent (Tables 2 and 3).

Table 2

Table 2

Table 3

Table 3

Back to Top | Article Outline

Video capsule endoscopy procedure

VCE was performed with the M2A capsule, which has been renamed the Pillcam SB during the study period (Given Imaging Ltd., Yoqneam, Israel). Manufactures software (RAPID) was used for imaging analysis.

In brief, this system consists of three major components: a miniature ingestible video capsule endoscope (Fig. 1), data recording hardware, and a workstation. After activation and ingestion, the capsule endoscope acquires video images during natural peristaltic propulsion through the GI tract at a rate of two pictures per second. The capsule endoscope transmits these images through radio waves to external sensor arrays and a data recorder worn as a belt by the patient. Patients were instructed to fast overnight before VCE. After activation and swallowing of the capsule, patients were allowed to drink liquids and eat solid food after 2–4 h, respectively. The data recording hardware was disconnected after 8 h and acquired data were downloaded. Patients were instructed to report natural excretion of the capsule in their stool.

Fig. 1

Fig. 1

Back to Top | Article Outline

Video capsule endoscopy image analysis

Studies were read by a single gastroenterologist (F.E.) experienced in both enteroscopy and VCE. The reader had access to pertinent clinical information for each case as indicated on the imaging requisite forms from referring clinicians (i.e. history, abnormal clinical and laboratory findings, rationale for imaging), but was blinded to the results of MRE. During data analysis, the location (i.e. proximal, middle or distal small bowel) and severity of abnormal findings as well the diagnostic quality of the exam (Table 2) were assessed on the basis of the following factors: completeness of small bowel coverage (in particular passage of the capsule into the cecum within 8 h), presence of obscuring bowel contents, and technical errors in data recording.

In VCE, neoplastic entities were suggested by polyps and submucosal or mucosal masses. Findings suggestive of postradiotherapy enteritis included teleangiectasis associated with ulcerations and strictures given an earlier history of radiotherapy. Crohn's disease was suggested by irregular or fissuring ulcers, cobblestone appearance, and luminal stenosis due to fibrosis or edema. The later criterium when viewed in isolation was considered suggestive of Crohn's disease given a high index of clinical suspicion, otherwise it was considered consistent with Crohn's disease. A single longitudinal ulceration was considered consistent with Crohn's disease. Villous denudation, edema, patchy mucosal erythema, and bleeding without an evident source lesion were considered nonspecific. Findings consistent with NSAID enteropathy included multiple erosions or ulcerations and diaphragm-like strictures, typically in a more proximal distribution of the small bowel.

Back to Top | Article Outline

Interpretation and statistical analysis

After the initial blinded readouts, MRE and VCE findings were assessed in conjunction (F.E., M.U., and P.K.) to establish a tentative, combined MRE–VCE diagnosis. In cases where the results of one study were suggestive of a diagnosis (e.g. Crohn's disease) and those of the other study were equivocal/compatible with one or more diagnoses (e.g. nonspecific inflammatory changes), the readout conclusions of the MRE–VCE correlation were weighted on the study with the most suggestive results (e.g. Crohn's disease).

Subsequently, the results of MRE, VCE, and combined MRE–VCE readouts were correlated with the final diagnosis based on histopathological findings (when available or conclusive, n=12) or based on a combined endpoint (F.E.) of clinical, laboratory, conventional endoscopic, surgical, and imaging findings.

Particularly, in the absence of confirmatory (i.e. specific) histology in suspected Crohn's disease, a positive diagnosis was based on symptoms and clinical findings [e.g. repeated episodes of GI symptoms (e.g. abdominal pain, diarrhea) associated with elevated C-reactive protein] which resolved in the long term after initiating medical therapy for Crohn's disease.

Statistical analysis of MRE and VCE findings for small bowel diseases (extraenteric findings excluded) was performed on a patient level: the sensitivity and specificity of these modalities were determined separately and then in conjunction with regard to the final diagnosis. Results were compared using the Fisher's exact test (two-tailed) assuming statistical significance at a P value of less than 0.05. This test was selected based on the relatively small sample size. To quantify agreement of both diagnostic methods with the final diagnosis, the Cohen κ-statistic (unweighted) was calculated.

Back to Top | Article Outline

Results

Final diagnoses with corresponding MRE and VCE findings are listed in Table 1. On the basis of the combined diagnostic endpoint of all imaging, clinical, laboratory, and histopathologic findings, small bowel pathologies were found in 15 of the 19 patients studied: Crohn's disease (n=5; Figs 1 and 2), nonHodgkin's lymphoma (NHL; n=4; Fig. 3), jejunal adenocarcinoma (n=1; Fig. 4), postradiotherapy enteritis (n=6; Fig. 5), jejunal lymphangioma (n=1; Fig. 6), jejunal angiodysplasia (n=1), NSAID-induced enteropathy (n=1), and small bowel obstruction due to adhesions (n=1). The final diagnoses were supported by histopathology in 12 cases. In five patients, the combined endpoints of all imaging, clinical, and laboratory findings were considered sufficient in making a definitive diagnosis. In two patients (patient numbers 1 and 7) with a high degree of clinical suspicion for Crohn's disease, biopsy results showed unspecific inflammatory changes with lymphoid hyperplasia, whereas clinical and imaging findings together were deemed sufficient in establishing a diagnosis of Crohn's disease. In one of these two patients, the endoscopist encountered technical difficulties, which prevented biopsy of the most prominent lesions.

Fig. 2

Fig. 2

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Fig. 5

Fig. 5

Fig. 6

Fig. 6

In two other patients (numbers 2 and 3), no abnormality was ultimately found. Conventional colonoscopy revealed a rectal carcinoma in patient number 12 and an ulcer in the ascending colon in patient number 4, who was also diagnosed with cholecystolithiasis at MRE. Only the small bowel findings were considered in determining the diagnostic accuracy of MRE and VCE in this study.

Back to Top | Article Outline

Magnetic resonance enterography

The overall diagnostic quality of MRE was judged to be good in eight (42.1%), fair in seven (36.8%), and poor in four (21.1%) patients (Table 2). Lack of small bowel distension was the most frequent factor resulting in poor image quality (n=3, 15.8%). Motion artefacts degraded image quality severely in only one patient (5.3%).

MRE findings were highly suggestive of (n=11) or consistent with (n=2, patient numbers 7 and 12) the final diagnosis in 13 patients. False negatives were found in four instances (patient numbers 11, 13, 15, and 17) and a false-positive diagnosis was made in two cases (patient numbers 8 and 12) (Table 4). Among patients presenting with obscure GI bleeding, MRE was diagnostic in two out of six patients (vs. five out of six in VCE). MRE demonstrated a sensitivity and specificity of 71.4 and 60%, respectively, in detecting small bowel disease.

Table 4

Table 4

Nineteen findings not related to the small bowel were observed at MRE in a total of 11 patients, however, these were not found to be clinically relevant. In none of the five patients diagnosed with Crohn's disease were extraenteric complications detected (e.g. abscess or fistula).

Back to Top | Article Outline

Video capsule endoscopy

The diagnostic quality of VCE was graded as good in 16 (84.2%) and fair in three (15.8%) patients. In 11 (57.9%) patients, the entire small bowel was covered (Table 2). Incomplete coverage of the small bowel was judged to be the most significant factor limiting diagnostic quality (n=8, 42.1%). Obscuring bowel contents were a minor factor in two patients (10.5%). Capsule retention led to surgical intervention in two patients (numbers 10 and 14) because of acute small bowel obstruction resulting from impaction of the VCE device at an ileocoecal stenosis in active Crohn's disease and at the site of a jejunal tumor, respectively. In these patients, however, obstructive signs and symptoms were absent before the examination and, as conventional small bowel studies were not routinely carried-out before VCE, these patients were not excluded from the examination.

VCE findings were highly suggestive of (n=15) or consistent with (n=2, patient numbers 7 and 11) the final diagnosis in 17 patients. VCE was falsely positive and falsely negative in patient numbers 8 and 16, respectively (Table 2). Consequently, VCE showed a sensitivity and specificity of 92.9 and 80% and positive and negative predictive values of 92.9 and 80%, respectively. VCE revealed gastric erosions in a patient (numbers 12) with a known hiatal hernia.

Back to Top | Article Outline

Video capsule endoscopy magnetic resonance enterography correlation

VCE and MRE findings were concordant and consistent with the final diagnosis in 13 cases. In four cases, MRE failed to depict limited mucosal disease detected by VCE (patient numbers 11, 23, 15, and 17). In one case of a duodenal lymphoma (patient number 8; Fig. 7), VCE and MRE findings were concordant and falsely positive in suggesting an inflammatory process (i.e. NSAID-induced enteropathy). In patient number 16, small bowel adhesions were diagnosed at MRE, whereas VCE showed no abnormality.

Fig. 7

Fig. 7

In interpreting the combined findings of MRE and VCE, sensitivity and specificity approached 100 and 80%, respectively, (Table 4). MRE yielded additional information compared with VCE not only for extraenteric processes, but also in defining the extent of mural disease. Although VCE alone showed in a higher diagnostic agreement (κ=0.73 for VCE; κ=0.29 for MRE with overlapping confidence-intervals) compared with MRE, the Fisher's exact test could not reveal a significant difference between the two modalities (P=0.31). When both diagnostic methods, however, are combined, their κ value approaches 0.85.

Back to Top | Article Outline

Discussion

VCE and MRE have both emerged relatively recently and are increasingly utilized for small bowel assessment. Technical advances have enhanced MRI's diagnostic capability in small bowel imaging and recent reports claim that MRI is becoming the method of choice for the evaluation of Crohn's disease in some reference centers, particularly in younger patients who may require repeated examinations [14].

VCE is a unique tool, which allows direct visualization of the entire small bowel mucosa. After its introduction to the clinical routine in 2001, VCE has been hailed as a revolution in the evaluation of obscure GI bleeding and has been widely established as in the work-up algorithm for these patients [4,15]. A growing body of literature shows that VCE holds promise in the evaluation of diverse small bowel processes, including Crohn's disease, NSAID-induced enteropathy, tumors, graft versus host disease, inherited polyposis syndromes, celiac disease, radiation enteritis, abdominal pain, and even unexplained diarrhea [4,15,16].

In this study, VCE was found to be more sensitive than MRE in diagnosing small bowel pathology (Table 2), detecting limited mucosal pathologies in four patients, which were missed at MRE (i.e. two NHLs, NSAID-induced enteropathy, and angiodysplasia). The differences in diagnostic accuracy between the two modalities, however, lacked statistical significance, which was precluded by a small study population and selection bias with a low number of true negative cases. When combining VCE and MRE findings, agreement of the imaging results with the final diagnosis was high (κ=0.85).

Back to Top | Article Outline

Crohn's disease

In all four patients with a suspected initial presentation of Crohn's disease, VCE and MRE findings were concordant with a diagnostic yield of 100%. In contrast, VCE revealed ulcera in the jejunum missed at MRE in one patient with established Crohn's disease and an ileocecal stenosis. No extraenteric complications were detected (e.g. abscess or fistula). The exclusion of such complications, however, are key advantages of MRE versus VCE alone.

Although our small patient population precludes any robust statistical analysis, our results are consistent with other comparisons of MRE and VCE in Crohn's disease. In a recent study by Albert et al. [10], VCE showed a slightly higher diagnostic yield than MRE in 27 patients with suspected Crohn's disease (92 vs. 77%; specificity 100 vs. 80%, P>0.05). VCE was slightly more sensitive than MRE (12 vs. 10 of 13 in suspected Crohn's disease and 13 vs. 11 of 14 in established Crohn's disease).

Gölder et al. [11] suggest that VCE is significantly more sensitive to Crohn's disease in the proximal small bowel, where VCE detected 12 lesions compared with a single lesion found at MRE (P=0.016) in 18 patients (n=16 with established and n=2 for suspected Crohn's disease). No significant difference in sensitivity was present between the two methods of disease in the terminal ileum or in regard to the overall diagnostic yield.

In a prospective comparison of VCE and CT enteroclysis, Voderholzer et al. [8] showed Crohn's disease of the small intestine to be much more frequent than considered earlier. In their study of 56 patients, VCE demonstrated small bowel involvement in approximately 60% of patients with previously nonestablished Crohn's disease. Jejunal and proximal ileal lesions were found in 25 patients by VCE compared to 12 by CT enteroclysis. Therapy was changed to appropriately address the VCE findings in 10 patients who subsequently showed clinical improvement. Thus, in a subset of patients, the lack of a therapeutic response to drugs released into the terminal ileum or colon may be related to yet undiagnosed more proximal small bowel disease.

In addressing the role of VCE compared with a wide range diagnostic modalities (barium radiography, CT enterogrpahy, MRE, and colonoscopy with ileoscopy), a meta-analysis of 250 patients in nine studies demonstrates that VCE is superior in diagnosing reoccurrences of nonstricturing Crohn's disease of the small bowel with a number needed to test three in order to yield one additional diagnosis over small bowel barium radiography and a number needed to test seven over conventional endoscopy [7]. In contrast, no significant difference was found between VCE and the other modalities in the initial diagnosis of Crohn's disease, although small sample sizes and lower pretest probabilities in this patient population may have led to a type II error. On the basis of this analysis and our own findings, larger study populations are needed to establish the role of VCE in evaluating patients with a suspected initial presentation of Crohn's disease.

Back to Top | Article Outline

Small bowel tumors

In this study population, six small bowel tumors were diagnosed: NHL (n=4), adenocarinoma (n=1), and lymphangioma (n=1). The diagnostic yield for VCE was 100% compared with 67% at MRE. In one patient with an early stage NHL of the terminal ileum, MRE missed limited mucosal disease. In a patient with histologically confirmed early stage NHL of the duodenum, VCE, and MRE findings were concordant in incorrectly suggesting an inflammatory process because of irregular wall and fold thickening (Fig. 7) and lack of lymphadenopathy. The incidence of small bowel tumors in this study was much higher than earlier reported (31.6 vs. 8.9%) [17], which is in part a reflection of selection bias here.

Tumors of the small bowel are difficult to diagnose because of nonspecific clinical symptoms and a high frequency of inconclusive diagnostic tests. Consequently, malignant lesions often present late with metastatic spread. In these cases, one would expect MRE to be advantageous in the detection of metastatic disease. In a retrospective analysis of 562 patients, capsule endoscopy revealed small bowel tumors in 9.7% of those presenting with obscure GI bleeding [17]. Early detection of malignant small bowel tumors by VCE has been shown to have a significant impact on outcome in these patients [18].

Studies addressing the utility of MRE in diagnosing small bowel tumors are limited to date and no formal comparisons with VCE have been reported thus far. MRE and VCE, however, having been evaluated in patients with polyposis syndromes, where VCE was superior in detecting polyps less than 15 mm and was the only method in which polyps less than 5 mm were seen [12]. VCE has demonstrated a significant clinical impact on the management of polyposis patients and being advocated as the surveillance tool of choice [19].

Back to Top | Article Outline

Other small bowel processes

The combination of MRE and VCE was helpful in evaluating a patient with suspected post radiation therapy enteritis. In view of the history, VCE findings of telangiectasias and stricture formation were highly suggestive of the diagnosis, whereas MRE accurately localized thickened wall and stenotic segments, which aided in planning surgical resection.

In a patient with obscure GI bleeding and a history of NSAID use, MRE showed no abnormality. In contrast, VCE detected a bleeding diaphragm-like stricture. Small bowel diaphragms are believed to be pathognomonic for NSAID-induced enteropathy, frequently manifest with GI bleeding and obstruction, and may be misinterpreted as normal plicae circulares at conventional small bowel imaging [20].

Back to Top | Article Outline

Limitations

A relatively small patient population and selection bias are the two major limitations of this study. These factors lead to an artificially increased disease prevalence here. Histological confirmation was lacking in nine out of 19 patients. Capsule coverage of the small bowel was incomplete in eight cases due to slow transit in six patients and capsule retention as a result of stricturing in two patients (patient numbers 10 and 14), requiring surgical removal. The 10.5% rate of retention seen here is considerably higher than the rate of 2% reported in a retrospective analysis of 733 consecutive cases [21]. Considering the underlying pathologies and ultimate necessity for small bowel surgery, capsule retention was, however, not considered a major complication in this study. Our relatively high rates of VCE retention compared with the literature may in part be because of the lack of a preceding conventional small bowel barium study in this study group to exclude stenoses. In VCE, lesion localization can usually only be roughly approximated based on landmarks such as the pylorus and cecum. Furthermore, VCE lacks sampling capabilities, as does MRE.

MRE was limited by suboptimal bowel distension and motion artifacts with a resultant diagnostic quality of good in eight, fair in seven, and poor in four patients. In the four false negative MRE findings, suboptimal bowel distension may have precluded detection of limited disease (poor in one and fair in two cases). The lack of key diagnostic extraenteric findings in this study population (e.g. suspicious lymph nodes, metastases, fistulae, abscesses, etc.) is a major limiting factor in the comparison of MRE and VCE, which seems to have biased diagnostic outcomes in favor of VCE. From the practical point of view, one must also consider that MRE is widely available, requires approximately 30 min machine time, and provides a rapid overview of the entire small bowel. In contrast, VCE examination time is 8 h and requires 1–2 h for image analysis. In general, costs must also be considered as significant limitation to the use of VCE and MRE (approximately 1000 Euros vs. 550 Euros in Switzerland, respectively).

In conclusion, VCE can readily depict and characterize subtle mucosal lesions, which may be missed with MRE, whereas MRE provides a rapid overview with additional mural, perienteric, and extraenteric information. Thus, VCE and MRE appear to be complementary methods, which in conjunction may increase diagnostic yields in the evaluation of suspected small bowel disease. Furthermore, in selected patients with a high index of suspicion and negative findings with one method, the alternate modality may be useful for further work-up.

Back to Top | Article Outline

References

1. Maglinte DD, Burney BT, Miller RE. Lesions missed on small-bowel follow through: analysis and recommendations. Radiology 1982; 144:737–739.
2. Miller RE, Sellink JL. Enterocylsis: the small bowel enema. How to succeed and how to fail. Gastrointest Radiol 1979; 4:269–283.
3. Schreyer AG, Geissler A, Albrich H, Schölmerich J, Feuerbach S, Rogler G, et al. Abdominal MRI after enteroclysis or with oral contrast in patients with suspected or proven Crohn's disease. Clin Gastroenterol Hepatol 2004; 2:491–497.
4. Mackalski BA, Bernstein CN. New diagnostic imaging tools for inflammatory bowel disease. Gut 2006; 55:733–741.
5. Carey EJ, Leighton JA, Heigh RI, Shiff AD, Sharma VK, Post JK, Fleischer DE. A single-center experience of 260 consecutive patients undergoing capsule endoscopy for obscure gastrointestinal bleeding. Am J Gastroenterol 2007; 102:89–95.
6. Schwartz GD, Barkin JS. Small-bowel tumors detected by wireless capsule endoscopy. Dig Dis Sci 2007; 52:1026–1030.
7. Triester SL, Leighton JA, Leontiadis GI, Gurudu SR, Fleischer DE, Hara AK, et al. Meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with non-stricturing small bowel Crohn's disease. Am J Gastroenterol 2006; 101:954–964.
8. Voderholzer WA, Beinhoelzl J, Rogalla P, Murrer S, Schachschal G, Lochs H, Ortner MA. Small bowel involvement in Crohn's disease: a prospective comparison of wireless capsule endoscopy and computed tomography enteroclysis. Gut 2005; 54:369–373.
9. Hara AK, Leighton JA, Heigh RI, Sharma VK, Silva AC, De Petris G, et al. Crohn disease of the small bowel: preliminary comparison among CT enterography, capsule endoscopy, small-bowel follow-through, and ileoscopy. Radiology 2006; 238:128–134.
10. Albert JG, Martiny F, Krummenerl A, Stock K, Lesske J, Göbel CM, et al. Diagnosis of small bowel Crohn's disease: a prospective comparison of capsule endoscopy with magnetic resonance imaging and fluoroscopic enteroclysis. Gut 2005; 54:1721–1727.
11. Gölder SK, Schreyer AG, Endlicher E, Feuerbach S, Schölmerich J, Kullmann F, et al. Comparison of capsule endoscopy and magnetic resonance (MR) enteroclysis in suspected small bowel disease. Int J Colorectal Dis 2006; 21:97–104.
12. Caspari R, von Falkenhausen M, Krautmacher C, Schild H, Heller J, Sauerbruch T. Comparison of capsule endoscopy and magnetic resonance imaging for the detection of polyps of the small intestine in patients with familial adenomatous polyposis or with Peutz-Jeghers' syndrome. Endoscopy 2004; 36:1054–1059.
13. Knuesel PR, Kubik-Huch RA, Crook DW, Eigenman F, Beer HJ, Froehlich JM. Dynamic contrast enhancement in active Crohn's disease of the small bowel with 3D MR enterography. Unpublished study currently under revision.
14. Wiarda BM, Kuipers EJ, Houdijk LP, Tuynman HA. MR enteroclysis: imaging technique of choice in diagnosis of small bowel diseases. Dig Dis Sci 2005; 50:1036–1040.
15. Eliakim AR. Video capsule endoscopy of the small bowel (PillCam SB). Curr Opin Gastroenterol 2006; 22:124–127.
16. Neumann S, Schoppmeyer K, Lange T, Wiedmann M, Golsong J, Tannapfel A, et al. Wireless capsule endoscopy for diagnosis of acute intestinal graft-versus-host disease. Gastrointest Endosc 2007; 65:403–409.
17. Cobrin GM, Pittman RH, Lewis BS. Increased diagnostic yield of small bowel tumors with capsule endoscopy. Cancer 2006; 107:22–27.
18. Bailey AA, Debinski HS, Appleyard MN, Remedios ML, Hooper JE, Walsh AJ, Selby WS. Diagnosis and outcome of small bowel tumors found by capsule endoscopy: a three-center Australian experience. Am J Gastroenterol 2006; 101:2237–2243.
19. Burke CA, Santisi J, Church J, Levinthal G. The utility of capsule endoscopy small bowel surveillance in patients with polyposis. Am J Gastroenterol 2005; 100:1498–1502.
20. Scholz FJ, Heiss FW, Roberts PL, Thomas C. Diaphragm-like strictures of the small bowel associated with use of nonsteroidal anti-inflammatory drugs. Am JRoentgenol 1994; 162:49–50.
21. Rondonotti E, Herrerias JM, Pennazio M, Caunedo A, Mascarenhas-Saraiva M, de Franchis R. Complications, limitations, and failures of capsule endoscopy: a review of 733 cases. Gastrointest Endosc 2005; 62:712–716.
Keywords:

magnetic resonance enterography; magnetic resonance imaging; small bowel disease; video capsule endoscopy

© 2009 Lippincott Williams & Wilkins, Inc.