Defining and Predicting Early Recurrence after Resection for Gallbladder Cancer

Abstract

Background: The optimal time interval to define early recurrence (ER) among patients who underwent resection of gallbladder cancer (GBC) is not well defined. We sought to develop and validate a novel GBC recurrence risk (GBRR) score to predict ER among patients undergoing resection for GBC.

Patients and methods: Patients who underwent curative-intent resection for GBC between 2000 and 2018 were identified from the US Extrahepatic Biliary Malignancy Consortium database. A minimum p value approach in the log-rank test was used to define the optimal cutoff for ER. A risk stratification model was developed to predict ER based on relevant clinicopathological factors and was externally validated.

Results: Among 309 patients, 103 patients (33.3%) had a recurrence at a median follow-up period of 15.1 months. The optimal cutoff for ER was defined at 12 months (p = 3.04 × 10^-18). On multivariable analysis, T3/T4 disease (HR: 2.80; 95% CI 1.58-5.11) and poor tumor differentiation (HR: 1.91; 95% CI 1.11-3.25) were associated with greater hazards of ER. The GBRR score was developed using β-coefficients of variables in the final model, and patients were classified into three distinct groups relative to the risk for ER (12-month RFS; low risk: 88.4%, intermediate risk: 77.9%, high risk: 37.0%, p < 0.001). The external validation demonstrated good model generalizability with good calibration (n = 102: 12-month RFS; low risk: 94.2%, intermediate risk: 59.8%, high risk: 42.0%, p < 0.001). The GBRR score is available online at https://ktsahara.shinyapps.io/GBC_earlyrec/ .

Conclusions: A novel online calculator was developed to help clinicians predict the probability of ER after curative-intent resection for GBC. The proposed web-based tool may help in the optimization of surveillance intervals and the counselling of patients about their prognosis.