We read the article by Fireman et al (Gut 2003;52:390–2) with great interest. We agree that full visualisation and imaging of the entire length of the small bowel is unsatisfactory at present and that capsule endoscopy (CE) is a novel technique and can be considered as a promising new approach for the diagnosis of obscure disease located in the small bowel.
The authors diagnosed Crohn's disease (CD) in 12 of 17 patients with clinically suspected CD according to the findings of CE. The authors state that the majority of diagnostic lesions were located in the distal ileum. At least one colonoscopy had been performed prior to CE in 15 of 17 patients. Unfortunately, the investigators do not report whether or not they were able to explore the terminal ileum in all of these patients. Hence the important question arises of which endoscopic and histological findings had been observed in the terminal ileum of these 15 study patients prior to CE, and whether this clinical information may have affected the interpretation of the CE findings in this investigational setting. Furthermore, the authors did not compare their non-diagnostic x ray findings with the CE results.
To date, we have performed a total of 130 capsule endoscopy procedures. In 50 patients with obscure gastrointestinal bleeding, we were able to disclose CD as the most probable underlying cause of bleeding in four patients. In addition, one patient suffering from Peutz-Jegher’s syndrome was diagnosed as also having CD of the small bowel. We also performed CE in eight patients in whom the diagnosis of CD had been established prior to CE to “stage” the small bowel for additional lesions that could influence treatment decisions. In the majority of our patients we found that the main pathological lesions were located in the terminal ileum. We were however able to confirm most CD lesions histologically by applying a second ileocolonoscopy with special emphasis on the small bowel biopsies in most of these patients, which allows for a greater diagnosis validity as small bowel ulcerations obtained with CE may also be caused by non-steroidal anti-inflammatory drug abuse, ulcerative ileitis, or coeliac disease. Hence from our experience we strongly recommend that patients with suspected CD should initially undergo careful ileocolonoscopy with close inspection of as much as the ileum as possible, and acquisition of multiple ileal biopsies to histologically establish CD prior to therapy.
We believe that at present CE is only clinically indicated in patients with signs and symptoms suggestive of small bowel CD in whom:
a stenosis/stricture has clearly been excluded,
the terminal ileum looks unremarkable on endoscopy, or
the ileum cannot be intubated for technical reasons.
The present study does not elucidate whether CE is really superior to conventional endoscopy plus histological assessment, which must still be considered the gold standard for the diagnosis of CD. As there is a substantial risk of capsule retention in the gastrointestinal tract in patients with stenosing CD, it should be determined if the benefits of CE findings outweigh the risks of this otherwise remarkable novel technique in individual patients.