S Nagral1, S Shah, M Gandhi, S K Mathur. 1. Department of Surgery, King Edward VII Memorial Hospital, Mumbai. nagral@bom3.vsnl.net.in
Abstract
OBJECTIVE: Isolated gastric varices (IGV) are rare and are believed to be associated with left-sided portal hypertension. We studied patients presenting with bleeding from IGV and compared them with those bleeding from both esophageal and gastric varices. METHODS: A retrospective analysis of 14 patients with bleeding from IGV was carried out. Portovenography findings (pattern of collateralization and natural shunts) in these patients were compared with a matched group of 69 patients with both esophageal and gastric varices. RESULTS: Of 14 patients with IGV, 2 had isolated splenic vein thrombosis and 12 had generalized portal hypertension. Portovenograms in 11 of the latter 12 revealed predominantly 'left-sided' collateralization in 8 patients as compared to 17 of 69 (25%) patients with esophageal and gastric varices (p = 0.004); natural shunts were seen in 6 of 11 cases and 15 of 69 (22%) patients in the two groups, respectively (p = 0.05). Abdominal devascularization operation gave good short- and long-term control of bleeding. CONCLUSIONS: Contrary to belief most patients with isolated gastric varices may have generalized portal hypertension rather than splenic vein obstruction as the cause and hence should be treated by a more extensive procedure than just splenectomy. The IGV could be a result of predominant collateralization to the retroperitoneal area (left-sided collateralization and natural shunts) rather than the usual pattern to the azygos system which results in esophageal varices.
OBJECTIVE: Isolated gastric varices (IGV) are rare and are believed to be associated with left-sided portal hypertension. We studied patients presenting with bleeding from IGV and compared them with those bleeding from both esophageal and gastric varices. METHODS: A retrospective analysis of 14 patients with bleeding from IGV was carried out. Portovenography findings (pattern of collateralization and natural shunts) in these patients were compared with a matched group of 69 patients with both esophageal and gastric varices. RESULTS: Of 14 patients with IGV, 2 had isolated splenic vein thrombosis and 12 had generalized portal hypertension. Portovenograms in 11 of the latter 12 revealed predominantly 'left-sided' collateralization in 8 patients as compared to 17 of 69 (25%) patients with esophageal and gastric varices (p = 0.004); natural shunts were seen in 6 of 11 cases and 15 of 69 (22%) patients in the two groups, respectively (p = 0.05). Abdominal devascularization operation gave good short- and long-term control of bleeding. CONCLUSIONS: Contrary to belief most patients with isolated gastric varices may have generalized portal hypertension rather than splenic vein obstruction as the cause and hence should be treated by a more extensive procedure than just splenectomy. The IGV could be a result of predominant collateralization to the retroperitoneal area (left-sided collateralization and natural shunts) rather than the usual pattern to the azygos system which results in esophageal varices.