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.
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.
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).
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.
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.
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.
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 .
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).
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. , 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.  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.  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 . 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.
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%) , 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 . Early detection of malignant small bowel tumors by VCE has been shown to have a significant impact on outcome in these patients .
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 . VCE has demonstrated a significant clinical impact on the management of polyposis patients and being advocated as the surveillance tool of choice .
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 .
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 . 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.
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:© 2009 Lippincott Williams & Wilkins, Inc.
magnetic resonance enterography; magnetic resonance imaging; small bowel disease; video capsule endoscopy