We identified several prognostic factors for patients with ICC treated with resection. Our data support the use of adjuvant CT in patients with ICC. The results of prospective randomized controlled studies will clarify the role of adjuvant CT in the future.

Background & Aims The prognosis of biliary tract cancer (BTC) is poor. Standard treatment for advanced BTC is a chemotherapy (CT) with gemcitabine and cisplatin. Phase III evidence for a second‐line (2L) CT is lacking. We aimed to investigate the feasibility of a 2L CT, to estimate the outcome and to identify prognostic markers. Methods Patients of our institution with advanced BTC between 2000 and 2015 receiving CT were included. Data were analysed in univariate and multivariate analysis. Results Three‐hundred and fifteen and 144 patients (45.7%) received first‐line (1L) and 2L CT respectively. The OS of patients receiving 2L CT was 16.67 and 9.9 months from the beginning of 1L and 2L CT respectively. The overall response rate and the disease control rate after 3 months were 9.7% and 33.6% respectively. Adverse events of grade 3 or more were observed in 26.1%. One patient died of gemcitabine‐related haemolytic uraemic syndrome. Age of more than 70 years was not associated with a poor outcome. In multivariate analysis, CEA levels of >3 µg/L (P = 0.004, hazard ratio [HR] 1.89, 95% CI 1.22, 2.91), cholinesterase (CHE) levels of <5 kU/L (P = 0.001, HR 2.11, 95% CI 1.34, 3.31) and leukocytosis (P = 0.001, HR 2.90, 95% CI 1.51, 5.56) were associated with poor survival. Conclusions Despite a relevant toxicity, our data suggest that 2L CT may be feasible in fit BTC patients. CEA elevation, leukocytosis and low CHE levels are unfavourable prognostic markers. Results from prospective randomized trials are urgently awaited.