Small bowel push-type fiberoptic enteroscopy for patients with occult gastrointestinal bleeding or suspected small bowel pathology.

Abstract

Endoscopic investigation of the small intestine remains the last frontier for gastroendoscopists. We evaluated the diagnostic efficacy and safety of two different push-type, fiberoptic enteroscopes in the investigation of patients with occult gastrointestinal bleeding or suspected small bowel pathology. Sixty-six patients (28 men, 38 women; mean age, 68) underwent push-type enteroscopy with either the SIF 10.5L (n = 17; scope length, 2495 mm) or the SIF 3000 (n = 49; length, 2995 mm) enteroscopes (Olympus America, Inc., Lake Success, New York). All patients enrolled had an indication of either occult gastrointestinal bleeding (n = 55) or suspected small bowel pathology (n = 11). Push-type enteroscopy was performed with an overtube and fluoroscopic guidance to better estimate the degree of insertion. Assessments of mucosal visualization, tip deflection, procedure tolerance, and complications were noted. Visualization of the small bowel mucosa and tip deflection was rated excellent in all patients. The mean length of insertion past the ligament of Treitz for the SIF 10.5L was 90 cm and 113 cm for the SIF 3000. Only one complication, a pharyngeal tear, occurred with the SIF 3000. Lesions that may explain the occult GI bleeding were found in 35 out of 55 patients (64%). Twenty-one of these 35 patients (60%) had lesions proximal to the ligament of Treitz, and 14 patients (40%) had lesions distal to the ligament of Treitz. If only distal lesions were considered, the diagnostic yield was 25% (14/55) for occult GI bleeding and 73% (8/11) for suspected small bowel pathology. The most common lesions were arteriovenous malformations (34%), which were successfully cauterized. With a relatively high diagnostic yield and therapeutic capability, push-type enteroscopy should play an important role in the investigation and treatment of lesions causing occult GI bleeding and small bowel pathology.