Predict the Risk of Variceal Rehemorrhage in Cirrhotic Patients with Portal Vein Thrombosis: A Two-Center Retrospective Study

Background: and aims Although portal vein thrombosis (PVT) was confirmed worsening portal hypertension and leading to poor prognosis, the risk stratification of such patients still be unclear. This study was to identify the clinical predictors of variceal rehemorrhage (VRH) and accordingly develop a competitive risk model in cirrhotic patients with PVT. Methods: Cirrhotic patients with PVT admitted for variceal hemorrhage were retrospectively included from two tertiary centres and respectively served as the training and validation cohort for 1-year follow-up. A competitive risk model was developed based on the result of multivariate Cox regression analysis, of which the predictive performance and the optimal decision threshold were evaluated by C -index, competitive risk curves, calibration curves and decision curve analysis. Results: A total of 199 patients completed 1-year follow-up, of which 64 (31.8%) patients occurred VRH. Multivariate Cox regression identified that CTP score ( P =0.020), time course of PVT ( P =0.036), CRP ( P <0.001) and AST ( P =0.047) were independently associated with VRH, which were incorporated into the competitive risk model, with high C -index (0.804 and 0.742 of the training and validation cohort, respectively). Competitive risk curves and calibration curves demonstrated its high risk stratification ability and consistency. The optimal decision range of the threshold probability was 0.2-1.0. Conclusions: This study found that CTP score, time course of PVT, CRP and AST were independent predictors of VRH, based on which the competitive risk model could be conveniently used for clinical decision making in cirrhotic patients with PVT.