Background/aims: The pathophysiology of gastric varices may be due to generalized or segmental portal hypertension. A considerable debate has arisen regarding the role of injection sclerotherapy in the pathogenesis of gastric varices.
Methodology: During the period from 1987 to 1997, a total of 1686 cases with bleeding varices were presented to our center and 225 cases (13.3%) with bleeding gastric varices were diagnosed. There were 198 males and 27 females with a total mean age of 45.7 years (+/- 7.6). Primary fundal varices (FV) were found in 121 (54%) cases and secondary FV were found in 104 (46%) cases. All patients with isolated FV presented with repeated attacks of upper gastrointestinal bleeding.
Results: The pathological diagnosis was studied in 120 cases; it was schistosomal in 8.3% of cases, non-schistosomal in 33.3% of cases, and mixed (Schistosomal with post viral cirrhosis) in 58.3% of cases. Seventy-five cases were subjected to splenectomy and gastroesophageal decongestion (SGED), 64 cases were subjected to distal splenorenal shunt (DSRS), and 86 cases were subjected to sclerotherapy. Mortality after DSRS was 7.8%, after SGED it was 12%, and after sclerotherapy it was 21%. Rebleeding was the major complication and occurred in 3% after DSRS, in 13% after SGED, and in 18% of cases after sclerotherapy.
Conclusions: Gastric varices are not an uncommon condition as a cause of upper gastrointestinal bleeding. Our findings support the hypothesis that gastric varices may be considered a late sequel of injection sclerotherapy, though they may also be considered as one of the pathophysiologies of generalized portal hypertension. Finally, DSRS was found to be the treatment of choice in the management of fundal varices.