Abstract
CONCLUSIONS
Patients progressing to frontline ICIs are at risk of early death, excluding them from experiencing potential benefit from chemotherapy PATIENT SUMMARY: Our analysis suggests that outcomes after failing immunotherapy are poor, particularly in UC patients who received no prior chemotherapy.
RESULTS AND LIMITATIONS
A total of 270 UC patients with PD to ICIs (69 frontline, 201 later line) were analyzed. Of the patients, 57% of frontline-ICI-PD and 34% of later-line-ICI-PD patients received SST, and SST had an impact on OS in MV (frontline: hazard ratio [HR] 0.22, 95% confidence interval [CI] 0.10-0.51, p < 0.001; later line: HR 0.22, 95% CI 0.13-0.36, p < 0.001). In the frontline-ICI-PD group, median OS with and without SST was 6.8 mo (95% CI 5.0-8.6) and 1.9 mo (95% CI 0.9-3.0), respectively. High disease burden (three or more metastatic sites: HR 2.49, p = 0.03; simultaneous liver/bone metastases: HR 3.93, p = 0.03) predicted worse survival. In later-line-ICI-PD group, response to ICIs (HR 0.37, p = 0.03), longer exposure to ICIs (HR 0.89, p = 0.002), and bone metastasis (HR 2.42, p < 0.001) predicted survival. The retrospective nature of this study and a lack of certain parameters limit the interpretation of our analysis.
BACKGROUND
Immune checkpoint inhibitors (ICIs) are approved for first-line (cisplatin unfit, PD-L1+) and platinum-refractory urothelial carcinoma (UC). Still, most patients experience progressive disease (PD) as the best response. Although higher response rates to subsequent systemic treatment (SST) have been described, post-PD outcome data are scarce.
OBJECTIVE
To examine the outcome of UC patients who received SST and no SST after progressing to ICIs.
INTERVENTION
Post-PD management as per standard practice.
DESIGN, SETTING, AND PARTICIPANTS
A retrospective analysis of UC patients progressing to frontline or later-line anti-PD-1/PD-L1 therapy in 10 European institutions was conducted between March 2013 and September 2017.
OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS
Overall survival (OS) was analyzed with a Kaplan-Meier model. Cox regression was used for multivariate analysis (MV). Impact of SST on OS was examined with a time-varying covariate model.