Objective: To analyse the outcomes of adjuvant chemotherapy vs observation in a multicentre cohort of patients with upper tract urothelial carcinoma (UTUC) in order to clarify whether such patients benefit from adjuvant chemotherapy after radical nephroureterectomy (RNU). Patients and Methods: Data from 15 centres were collected for a total of 1544 patients, treated between 2000 and 2015. Criteria for patient selection included pT2–4N0/x stage, or lymph node-positive disease, and prior RNU. The standardized difference approach was used to compare subgroup characteristics. Overall survival (OS) was the primary endpoint. The primary analysis used 1:1 propensity score matching, with inverse probability of treatment weighting in addition to this in the secondary analysis. The latter was also performed with the inclusion of covariates, i.e. with ‘doubly robust’ estimation. A 6-month landmark analysis was performed to exclude early events. Results: A total of 312 patients received adjuvant chemotherapy and 1232 underwent observation. Despite differences between the two groups, the standardized difference was generally <10% after matching. In the matched analysis no difference was observed in OS between adjuvant chemotherapy and observation (hazard ratio [HR] 1.14, 95% confidence inverval [CI] 0.91–1.43; P = 0.268). In the doubly robust estimate-adjusted comparison, adjuvant chemotherapy was significantly associated with shorter OS (HR 1.26, 95% CI 1.02–1.54; P = 0.032). Similar findings were confirmed in subgroup analyses stratified by pathological stage, and after landmark analysis. Results should be interpreted with consideration given to the inherent limitations of retrospective studies. Conclusion: Adjuvant chemotherapy did not improve OS compared with observation in the present study. These results contribute to the uncertainties regarding postoperative chemotherapy in UTUC, and suggest dedicated prospective trials, new more potent therapies, and the identification of enhanced patient selection criteria are required.
All Science Journal Classification (ASJC) codes