Abstract
BACKGROUND: Baseline hepatic venous pressure gradient (HVPG) has been applied for prediction of variceal rebleeding in patients after acute variceal bleeding. However, for patients receiving secondary prevention, there still lacks evidence about the predictive performance of baseline-HVPG for rebleeding. This study aims to investigate the predictive value of baseline-HVPG for variceal rebleeding in cirrhotic patients receiving secondary prevention. METHODS: This retrospective study included 122 patients with cirrhosis accepting secondary prevention of variceal rebleeding in a university hospital. All the included patients had HVPG measurements before rebleeding and had at least 1-year follow-up after HVPG measurement unless the rebleeding occurred. The rebleeding rate in patients with different HVPG levels and time-dependent predictive performance of baseline-HVPG were analysed. A Cox regression model and P for trend were used to assess the rebleeding risk. RESULTS: Variceal rebleeding occurred in 22 (18.0%) patients during 1-year follow-up. No significant difference was observed in rebleeding rate between patients with HVPG <16 mmHg and HVPG ≥16 mmHg (17.91% vs. 26.41%, P=0.200). A decreasing trend was observed in area under the curve of HVPG for predicting rebleeding by time. The multivariate Cox model showed an overall decreasing trend in hazard ratio of rebleeding (vs. patients with HVPG <12 mmHg) for patients with 12≤ HVPG <16 mmHg, 16≤ HVPG <20 mmHg and HVPG ≥20 mmHg; besides, an increasing P for trend was observed. CONCLUSIONS: A single baseline-HVPG measurement was insufficient for predicting rebleeding in patients with cirrhosis who received secondary prevention.