Purpose There is limited evidence to support the use of adjuvant chemotherapy (AC) after radical nephroureterectomy (RNU) for upper tract urothelial carcinoma (UTUC). Against this backdrop, we hypothesized that such treatment is associated with overall survival (OS) benefit in patients with locally advanced and/or positive regional lymph node disease. Patients and Methods Within the National Cancer Database (2004 to 2012), we identified 3,253 individuals who received AC or observation after RNU for pT3/T4 and/or pN+ UTUC. Inverse probability of treatment weighting (IPTW) -adjusted Kaplan-Meier curves and Cox proportional hazards regression analyses were used to compare OS of patients in the two treatment groups. In addition, we performed exploratory analyses of treatment effect according to age, gender, Charlson comorbidity index, pathologic stage (pT3/T4N0, pT3/T4Nx and pTanyN+), and surgical margin status. Results Overall, 762 (23.42%) and 2,491 (76.58%) patients with pT3/T4 and/or pN+ UTUC received AC and observation, respectively, after RNU. IPTW-adjusted Kaplan-Meier curves showed that median OS was significantly longer for AC versus observation (47.41 [interquartile range,19.88 to 112.39] v 35.78 [interquartile range, 14.09 to 99.22] months; P < .001). The 5-year IPTW-adjusted rates of OS for AC versus observation were 43.90% and 35.85%, respectively. In IPTW-adjusted Cox proportional hazards regression analysis, AC was associated with a significant OS benefit (hazard ratio, 0.77 [95% CI, 0.68 to 0.88]; P < .001). This benefit was consistent across all subgroups examined (all P < .05), and no significant heterogeneity of treatment effect was observed (all Pinteraction > .05). Conclusion We report an OS benefit in patients who received AC versus observation after RNU for pT3/T4 and/or pN+ UTUC. Although our results are limited by the usual biases related to the observational study design, we believe that the present findings should be considered when advising post-RNU management of advanced UTUC, pending level I evidence.
