The Value of Selective Angiography in the Diagnosis of Tumors in Pancreas and Liver

The diagnosis of pancreatic diseases presents many difficulties, and consequently it was hoped that good results would be obtained from angiographic examinations such as celiac and superior mesenteric angiography. Many authors have pointed out the value of these examinations and have been able to demonstrate angiographic changes in almost 90 per cent of patients with carcinoma of the pancreas (1–4). These studies were of a retrospective nature, however, being analyses of angiographic findings in patients in whom the diagnosis had already been verified. Few authors have given statistics regarding diagnostic accuracy based upon preoperative evaluation of angiograms (5, 6). These authors regarded the diagnosis of pancreatic diseases with some skepticism, since both false positive and false negative findings were frequent. It appears to be generally agreed that angiography is of value in the detection of primary liver tumors (7–9), whereas it may be difficult to detect liver metastases on the angiogram (7). Material and Methods The present study comprises 71 patients (32 male and 39 female) examined at the University Hospital, Copenhagen, from 1961 to 1966. The majority of these were referred to us for examination because pancreatic or hepatic tumors were suspected clinically. The Seldinger technic was utilized in all the patients, with one or two Ödman catheters inserted from a femoral or axillary artery (10). Twenty-five patients underwent selective celiac angiography only; in 44 patients both celiac and superior mesenteric angiograms were obtained, and in 2 cases celiac angiography was performed in conjunction with aortography. Superselective technic (11) or injections of vasopressor substances (12) were not employed. In almost all cases alternating anteroposterior and lateral projections were obtained. A total of 4 films per second were taken for the first six seconds, 2 films per second for the next four seconds, and one film per second for the last ten seconds. Urografin 60 per cent was injected in doses of 30 ml or 60 ml with the use of one or two catheters, respectively. Contrast material was injected with the aid of a Gidlund injector at a pressure of 3 kg/cm2. Apart from hematomas at the site of injection and an isolated shaking spell after the injection, no side-effects related to the examinations were observed. During the review of the angiograms, we knew the referred diagnosis and the original description of the films, but we did not compare the surgical findings and course of the disease until final analysis of the material. Table I shows the classification of the patients according to the principal angiographic diagnosis. The revised diagnoses made by the authors are given, together with the original roentgen diagnoses. We decided to revise the diagnoses because the original evaluation was made by different radiologists with varying experience.