Video capsule endoscopy as the initial examination for overt obscure gastrointestinal bleeding can efficiently identify patients who require double-balloon enteroscopy

Obscure gastrointestinal bleeding (OGIB) is defined by the American Gastroenterological Association as bleeding of unknown cause after upper or lower endoscopy [1]. OGIB can be further classified into occult and overt bleeding. Occult OGIB is defined as iron deficiency anemia, with or without a positive fecal occult blood test, while overt OGIB indicates clinically perceptible bleeding that recurs or persists despite negative initial endoscopic and radiologic evaluations. In overt OGIB, therefore, immediate identification of the bleeding lesion is needed for appropriate therapy, including hemostatic treatment to rescue unstable patients. However, the diagnostic performance and accuracy of conventional diagnostic strategies, including small intestinal radiography, abdominal computed tomography, magnetic resonance imaging, digital subtraction angiography, radionuclide scanning, and intraoperative endoscopy, have not been satisfactory for detecting small intestine diseases [2, 3].

Since the introduction of video capsule endoscopy (VCE) [4] and double-balloon enteroscopy (DBE) [5] in 2000 and 2001, respectively, these modalities have enabled direct observation of the lumen of the small intestine. VCE is a new type of endoscopy, which allows painless endoscopic imaging of the entire small bowel, and is well tolerated by patients [6]. However, lack of therapeutic capability, including hemostasis, is a critical drawback of the device. On the other hand, DBE is a new insertion method for enteroscopy, which enables endoscopic scrutiny of the entire small intestine, and endoscopic intervention, such as polypectomy, hemostasis, and sampling of tissue. Total enteroscopy may be achieved through a combination of antegrade and retrograde approaches [7]. However, it is an invasive procedure, and it typically requires patient sedation, two endoscopists, and additional medical staff.

Both VCE and DBE reportedly demonstrate similarly high diagnostic yields of small intestinal disorders [8, 9]. However, in the clinical situation of overt OGIB, the role of these new modalities has not been fully established. In particular, the choice of initial modality is still controversial. The policy of some endoscopists is that DBE by the antegrade route should be the initial procedure for overt OGIB, because the intestinal lesions causing overt OGIB are most likely localized in the proximal part of the small bowel [10]. Others advocate that VCE should be the initial choice because it is noninvasive and is capable of visualizing the entire small intestine. Recently published guidelines in Europe recommended VCE as the initial examination for overt OGIB, mainly due to its safety [11]. However, the validation of this strategy is incomplete.

Our recent policy for overt OGIB is also initial VCE, followed by DBE via the antegrade or retrograde route, as appropriate. Hence, consecutive patients with overt OGIB underwent VCE as the initial examination for small bowel lesions in our hospital. In this study, the clinical outcome of these patients is shown, and the strategy of initial VCE for overt OGIB is validated.

Recent meta-analysis of examinations for small bowel lesions reported that both VCE and DBE have similarly high diagnostic yields [9]. However, the choice of modality for use in overt OGIB has not yet been clearly established. In particular, the choice of the initial approach is controversial. Recommendation of initial VCE by the European Society of Gastrointestinal Endoscopy is based only on the excellent safety profile of VCE, [11] and the validation of the superiority of the strategy is incomplete.

On this background, we demonstrated the results of our clinical experience for overt OGIB. Our policy for overt OGIB was to perform VCE initially, followed by DBE as necessary. Our results have the following clinically relevant implications. First, an initial VCE procedure could precisely determine the necessity of subsequent DBE, except one case with perforation with obscure causal relationship with VCE. In fact, the sensitivity and NPV of VCE were 100 %, both for the presence of small bowel lesions and the requirement of hemostasis in the small bowel. These perfect results indicate that VCE never misses relevant findings in the small bowel, and that negative VCE findings correspond to the lack of necessity for further examination. Because missed overt OGIB may lead to critical clinical outcomes, VCE as the initial examination for OGIB appears to be optimal.

In this context, it is surprising that approximately 60 % of patients who had undergone VCE could dispense with DBE, suggesting that the strategy of initial VCE would greatly reduce the effort, burden, and cost of further examinations, including DBE. In particular, VCE for overt OGIB revealed that 25 % had bleeding sites outside the small bowel, but within the reach of upper endoscopy or colonoscopy, and that 10 % had no bleeding site throughout the intestine. The rate of presence of bleeding sites outside of the small intestine appears to be large. In fact, previous reports demonstrated that the rates ranged 12.5–28 % [12]. However, because the detection rate of small bowel lesions by VCE in our study was equivalent to that of previous reports (65 % vs. 59.4–72.5 %), [13‐15] our population does not appear to be skewed by inappropriate upper GI endoscopy or colonoscopy [16]. Further investigation should be performed on this issue.

As the second relevant finding of our study, VCE could determine the appropriate insertion route for DBE. Although DBE enables endoscopic interventions including hemostasis, observation of the entire small intestine usually requires both antegrade and retrograde insertions. Because the procedure is cumbersome and requires much time and medical staffing, a predetermined appropriate insertion route for bleeding lesions would reduce the burdens on patients, endoscopists, and medical staff. Hence, VCE as the initial study, followed-by DBE, appears to be the best strategy for overt OGIB.

Li et al. suggested the ability of VCE to guide the choice of insertion route of DBE [17]. Similar to our strategy, the authors also determined the insertion route by the time point at which VCE reached the target lesion. While our point of divergence for antegrade vs. retrograde insertion was 0.5, their criterion was 0.6, (i.e., if VCE reached the target at a time 60 % or less than the total small bowel transit time, DBE was inserted by the antegrade route). The result for reaching the target lesion with the initial DBE procedure was 100 % in their study, as well as in ours, although the indication for DBE in their study was not confined to overt OGIB. If the divergence criterion of 0.6 had been adopted in our study, 2 patients who underwent retrograde DBE would have qualified for the antegrade DBE route (the indices were 0.51 and 0.58, respectively). Naturally, there may be subjects with lesions in the mid-portion of the small bowel, which DBE by either route could reach. The optimal method for choice of route for these patients should be investigated further.

Our study has limitations. First, the study design is retrospective, and the results of the single-arm strategy (initial VCE) alone were shown. Our strategy for overt OGIB, i.e., initial VCE followed by DBE, seems to be fully validated by our clinical experience, despite the drawback of the small number of the analyzed patients. For more precise evaluation of our strategy, however, studies with larger number of patients from multiple institutes and/or a prospective controlled study comparing DBE with and without prior VCE should be performed. However, such a study appears to be too risky because the effectiveness of prior VCE was obvious, and overt OGIB is sometimes fatal. Second, not all patients underwent prior upper or lower endoscopy in our hospital. Therefore, there may have been misdiagnoses of overt OGIB in our patients.

In conclusion, a considerable proportion of overt OGIB patients do not require intervention including hemostasis, and can therefore avoid further invasive and costly procedures, such as DBE. The strategy of initial VCE for overt OGIB patients can efficiently identify patients who require DBE as the subsequent examination. Hence, the strategy is recommended, not only because of the safety of VCE, but also for better clinical outcomes for these patients.