Importance: The association of adjuvant chemotherapy (AC) with survival in the general population of patients with resected biliary tract cancer (BTC) remains controversial. As such, the role of this treatment in the treatment of older adult patients (aged ≥70 years) needs to be evaluated.
Objective: To describe the patterns of use of AC and compare survival outcomes of AC and observation in older adult patients following resection of BTC.
Design, setting, and participants: This retrospective cohort study included 8091 older adult patients with resected BTC with data available in the National Cancer Database from January 1, 2004, to December 31, 2019. Patients were divided into 2 cohorts: AC and observation. The AC cohort was subdivided into single-agent and multiagent AC treatment.
Exposures: Adjuvant chemotherapy vs observation following BTC resection.
Main outcomes and measures: The primary outcome was overall survival (OS) of patients who received AC compared with observation following resection of BTC as evaluated using Kaplan-Meier estimates and multivariable Cox proportional hazards regression models. Inverse probability of treatment weighting and propensity score matching were performed to address indication bias.
Results: Between 2004 and 2019, of 8091 older adult patients with resected BTC identified (median [range] age, 77 [70-90] years; 5136 women [63.5%]; 2955 men [36.5%]), only one-third (2632 [32.5%]) received AC. There was an increase in the use of AC across the study period from 20.7% (n = 495) in 2004 to 2009 to 41.2% (n = 856) in 2016 to 2019. Age 80 years or older (odds ratio, 0.29; 95% CI, 0.25-0.33; P < .001) and gallbladder primary site (odds ratio, 0.71; 95% CI, 0.61-0.83; P < .001) were associated with a lower odds of AC. Following inverse probability of treatment weighting, as a composite, AC was not associated with improved survival (median OS, 20.5 months; 95% CI, 19.2-21.7 months) compared with observation (median OS, 19.0 months; 95% CI, 18.1-20.3 months). A longer median OS was associated with single-agent AC (21.5 months; 95% CI, 19.9-24.0 months) but not multiagent AC (19.1 months; 95% CI, 17.5-21.1 months) compared with observation (median OS, 17.3 months; 95% CI, 16.1-18.4 months). This improvement in OS with single-agent AC was not apparent on multivariable analysis (hazard ratio [HR], 0.97; 95% CI, 0.89-1.05; P = .44). However, age at diagnosis of 80 years or older (HR, 1.35; 95% CI, 1.28-1.42; P < .001) and treatment at nonacademic centers (HR, 1.14; 95% CI, 1.07-1.20, P < .001) were associated with worse OS.
Conclusions and relevance: In this cohort study of older adult patients, AC was not associated with an improvement in survival compared with observation following BTC resection. These findings suggest the need for further study of AC for older adult patients who may benefit after curative intent surgery for BTC.