CT enteroclysis in the diagnosis of obscure gastrointestinal bleeding: initial results

To evaluate the usefulness of computed tomography (CT) enteroclysis in patients with obscure gastrointestinal (GI) bleeding.

In a prospective study, CT enteroclysis was performed in 21 patients (median age 50 years; range 13-71 years) with obscure GI bleeding in which the source of the bleeding could not be detected despite the patient having undergone both upper GI endoscopic and colonoscopic examinations. The entire abdomen and pelvis was examined in the arterial and venous phases using multisection CT after distending the small intestine with 2l of 0.5% methylcellulose as a neutral enteral contrast medium and the administration of 150ml intravenous contrast medium.

Adequate distension of the small intestine was achieved in 20 of the 21 (95.2%) patients. Potential causes of GI bleeding were identified in 10 of the 21 (47.6%) patients using CT enteroclysis. The cause of the bleeding could be detected nine of 14 (64.3%) patients with overt, obscure GI bleeding. However, for patients with occult, obscure GI bleeding, the cause of the bleeding was identified in only one of the seven (14.3%) patients. The lesions identified by CT enteroclysis included small bowel tumours (n=2), small bowel intussusceptions (n=2), intestinal tuberculosis (n=2), and vascular lesions (n=3). All vascular lesions were seen equally well in both the arterial and venous phases.

The success rate in detection of the cause of bleeding using CT enteroclysis was 47.6% in patients with obscure GI bleeding. The diagnostic yield was higher in patients with overt, obscure GI bleeding than in those with occult obscure GI bleeding.