Approach to management of intussusception in adults: a new paradigm in the computed tomography era.

Intussusception has been considered an operative indication in adults as a result of the risk of ischemia and the possibility of a malignant lead point. Computed tomographic (CT) scans can reveal unsuspected intussusception. All CT reports from July 1999 to December 2005 were scanned electronically for letter strings to include the keyword intussusception. Identified CT scans were analyzed to characterize the intussusception and associated findings. Clinical, laboratory, pathological, and follow-up variables were gleaned from medical records. Findings were analyzed by treatment and findings at operation. Review of 380,999 CT reports yielded 170 (0.04%) adult patients (mean age, 41 years) with intussusceptions described as enteroenteric in 149 (87.6%), ileocecal in eight (4.7%), colocolonic in 10 (5.9%), and gastroenteric in three (1.8%). Radiological features included mean length of 4.4 cm (range, 0.8-20.5 cm) and diameter of 3.2 cm (range, 1.6-11.5 cm). Twenty-nine (17.1%) had a lead point, and 12 (7.1%) had bowel obstruction. Clinically, 88 (48.2%) patients reported abdominal pain, 52 (30.6%) had nausea and/or vomiting, and 74 (43.5%) had objective findings on abdominal examination. Thirty of 170 (17.6%) patients underwent operation, but only 15 (8.8%) patients had pathologic findings that correlated with CT findings. Seven had,enteroenteric intussusceptions from benign neoplasms (two), adhesions (one), local inflammation (one), previous anastomosis (one), Crohn's disease (one), and idiopathic (one). Three had ileocolic disease, including cecal cancer (one), metastatic melanoma (one) and idiopathic (one; whereas five patients had colocolonic intussusception from colon cancer (three), tubulovillous adenoma (one), and local inflammation (one). Of the 15 without intussusception at exploration, five had pathology related to trauma, four had nonincarcerated internal hernia after Roux-en-Y gastric bypass, four had negative explorations, one had adhesions, and one had appendicitis that did not correlate with CT findings. No patient in the observation group required subsequent operative exploration for intussusception at mean 14.1 months (range, 0.25-67.5 months) follow up. All operative patients demonstrated gastrointestinal symptoms versus 55.3 per cent of the observation group (P < 0.006). Analysis of CT features demonstrated differences among patients observed without operation, those without intussusception at exploration, and confirmed intussusception with regard to mean intussusception length 3.8 versus 3.8 versus 9.6 cm, diameter 3.0 versus 3.2 versus 4.8 cm, lead point 12.1 per cent versus 30 per cent versus 53.3 per cent, and proximal obstruction 3.8 per cent versus 0 per cent versus 46.7 per cent, respectively. Intussusceptions in adults discovered by CT scanning do not always mandate exploration. Most cases can be treated expectantly despite the presence of gastrointestinal symptoms. Close follow up is recommended with imaging and/or endoscopic surveillance. Length and diameter of the intussusception, presence of a lead point, or bowel obstruction on CT are predictive of findings that warrant exploration.