To investigate the short-term and long-term clinical efficacy of devascularization in the treatment of esophageal and gastric varices in cirrhosis patients with portal hypertension.

76 patients with portal hypertension admitted to Beijing Youan Hospital Affiliated to Capital Medical University from April 2015 to July 2018 were enrolled in this prospective study. Among them, 43 patients were male and 33 female, aged 25-65 years with a median age of 51 years. The informed consents of all patients were obtained and the local ethical committee approval was received. According to the conditions of patients, different devascularization procedures were performed including splenectomy + selective portal-azygos devascularization, splenectomy+traditional devascularization and splenectomy, etc. Patients complicated with primary liver cancer simultaneously underwent partial hepatectomy. All patients underwent gastroscopy at postoperative 1, 6, 12 and 24 months to observe the changes of esophageal and gastric varices. The degree of esophageal varices and classification of gastric varices at different time points before and after operation were statistically compared by Wilcoxon rank-sum test. The positive rate of red sign was statistically compared by Chi-square test or Fisher's exact test.

All patients completed the devascularization successfully, including splenectomy + selective portal-azygos devascularization in 60 cases, splenectomy + traditional devascularization in 8, splenectomy alone in 5, splenectomy + partial hepatectomy in 2 and splenectomy + radiofrequency ablation in 1, respectively. At 1, 6, 12 and 24 months after operation, the percentage of patients with severe esophageal varices was 51%, 49%, 41% and 35%, respectively, with a gradually decreasing trend, which was significantly lower than 83% before operation (Z=-4.786,-4.025,-4.418,-3.314; P<0.05). The positive rate of red sign in patients with esophageal varices was 37%, 35%, 33% and 30%, respectively, which tended to decline and was significantly lower compared with 87% before operation (χ2=36.963, 34.036, 35.475, 27.284; P<0.05). However, the positive rate in those without gastric varices was 53%, 60%, 62% and 65%, respectively, with a gradually increasing trend, which was significantly higher than 36% before operation (Z=-1.084, -2.127, -2.608, -2.668; P<0.05).

Devascularization can effectively mitigate the severity of esophageal and gastric varices in cirrhosis patients with portal hypertension. Splenectomy + selective portal-azygos devascularization may bea better option compared with the other procedures.