Angiography in Diagnosis of Chronic Unexplained Melena

Abstract

After all clinical methods, including ordinary roentgen diagnostic procedures, have been attempted, the source of gastrointestinal bleeding remains unexplained in many patients. As many as 20 per cent of massive hemorrhages are undiagnosed (8). The cause of chronic or intermittent bleeding is even more difficult to detect. In this latter group of patients, the source, though undetected at barium examination, may be revealed by selective angiography. Thus, cases of unexpected tumor (3), angioma (10), arteriovenous fistula (18), and varices (1) have been diagnosed. At the Roentgendiagnostic Department, University Hospital of Lund, 24 patients with previously undiagnosed melena have been examined with selective celiac and superior mesenteric angiography. These examinations have been reviewed to assess the usefulness of the method in these patients, who present difficult diagnostic problems. Materials and Methods The patients in this series were chosen according to the following criteria: (a) two or more positive Weber tests for occult bleeding were performed during at least two different periods of bleeding; (b) the gastrointestinal barium examination was normal. Patients in whom barium examination of the gastrointestinal tract revealed a possible source of bleeding, such as a small hiatus hernia or coarse gastric mucosa, were excluded even though there was doubt these were the sites of bleeding; (c) both selective celiac and superior mesenteric angiography had been carried out. Thus, the material was highly selected to include only those patients with undiagnosed, chronic gastrointestinal bleeding, in whom complete angiographic examinations had been made. Catheterization was performed by the percutaneous technic via the femoral arteries, usually with a thin-walled KIFA red catheter (I.D./O.D. = 1.4/2.2 mm). Filming was done with a rapid film-changer, usually two exposures per second for four seconds, one per second for five seconds, and then one every other second for ten seconds. Injections of approximately 30 cc of 76 per cent Urografin (Schering, Germany) were made into the celiac and superior mesenteric arteries in the anteroposterior and right posterior oblique projections. A pressure injector was used, and the contrast medium was injected at approximately 11 cc/second. Findings The abnormal angiographic findings in these patients have been grouped into five general classifications, although some patients had abnormal findings in more than one classification (TABLE I).