Huang X, et al. Endoscopy 2020.
BACKGROUND : Oxaliplatin, used as first-choice treatment for colorectal cancer (CRC), induces sinusoidal endothelial injury and portal hypertension. This study investigated the characteristics of oxaliplatin-induced portal hypertension and evaluated the efficacy of endoscopic management of gastroesophageal variceal bleeding. METHODS : We performed a retrospective, multicenter, case-control study between January 2010 and December 2018. Patients who received oxaliplatin-based chemotherapy after
CRC surgery and presented with portal hypertension and gastroesophageal varices were compared with consecutive patients with hepatitis B-related cirrhotic portal hypertension receiving endoscopic treatment for variceal bleeding. RESULTS : 39 patients with oxaliplatin-induced portal hypertension were identified, 35 of whom had a history of variceal bleeding. The median period between start of oxaliplatin-based chemotherapy and the occurrence of varices was 50.4 months (n = 39). A total of 26 patients with oxaliplatin-related portal hypertension and 230 patients with hepatitis B-related portal hypertension underwent endoscopic treatment. Kaplan-Meier analysis revealed that the 1-year rebleeding rate was significantly higher in the oxaliplatin group than in the hepatitis B group (43.3 % vs. 19.0 %, P = 0.001). Multivariable Cox regression analysis showed that oxaliplatin-based chemotherapy was an independent factor for 3-year rebleeding (hazard ratio [HR] 2.46, 95 % confidence interval [CI] 1.24-4.87; P = 0.01) and 3-year overall mortality (HR 9.43, 95 %CI 2.32-38.31; P = 0.002). CONCLUSIONS : Oxaliplatin-related portal hypertension was characterized by massive ascites, splenomegaly, gastric varices, concomitant arterioportal fistula, and relatively normal liver function. Endoscopic treatment to prevent variceal rebleeding in these patients was unsatisfactory compared with endoscopic treatment for hepatitis B-related portal hypertension.