Abstract
BACKGROUND: The role of second-line chemotherapy in advanced biliary cancers (ABCs) has only recently been established in phase III randomized trial and the optimal selection of patients most likely to benefit from it remains challenging. METHODS: A cohort of 98 ABC treated second-line chemotherapy was used as a developmental dataset to identify covariates independently associated with overall survival (OS). Kaplan-Meier analysis was used to investigate the association between variables and OS and those retaining statistically significance were combined in a multiplexed score. RESULTS: The following pretreatment variables were independently associated with OS: ECOG PS > 0, peritoneal disease, LDH > 430 UI/L, albumin <3.5 gr/dL, gamma-GT >100 UI/L, sodium <140 mEq/L, absolute lymphocyte count <1000/mmc, and PFS to first-line <6 months. Based on these results, a scoring system was developed that identified three subgroups with statistically different OS: low-risk (mOS 18 months), intermediate-risk (mOS 9.4 months) and high-risk (mOS 2.9 months) (p < 0.001). The prognostic model was both internally and externally validated in a multicentre cohort of 120 ABCs. CONCLUSION: The Modena score is a multiplexed scoring system capable of accurately risk-stratified ABCs treated with second-line chemotherapy. Based on its reproducibility, usability and generalizability, it has the potential for assisting therapeutic decision-making in the clinic and risk-stratification in future trials.