Disease-free survival as a predictor of overall survival in localized renal cell carcinoma (RCC) following first nephrectomy.

4581 Background: Intermediate endpoints (e.g., disease free survival [DFS]) have gained traction lately as potential surrogates for OS in oncology as they require shorter follow up to show clinical benefit. Given the high post-nephrectomy survival in patients (pts) with localized RCC, evidence on if DFS can be used as a predictor of OS in the disease is warranted. We assessed the association between DFS and OS in pts with newly diagnosed, completely resected, intermediate-high (pT2N0 high grade, pT3N0) or high-risk (pT4N0, pTanyN1) RCC post-nephrectomy. Methods: This retrospective observational study used the SEER-Medicare database (2007–2016). DFS was defined as time from initial nephrectomy date to first recurrence (diagnosis of metastatic disease, additional surgery, starting systemic treatment for advanced RCC) or death, whichever occurred first. OS from time of recurrence in pts with recurrence were compared with OS from comparable time point in pts without, using Kaplan-Meier analyses and adjusted Cox models. OS was also compared between pts with and without recurrence by landmark time points at 1, 2, 3, 4 and 5 years (yrs) post-nephrectomy; hazard ratios (HRs) between the two cohorts were estimated using adjusted Cox models. Correlation between DFS and OS was assessed using the Kendall’s τ rank correlation. Monthly healthcare costs were compared between the two cohorts using generalized linear model. Results: 643 post-nephrectomy RCC pts (269 with recurrence vs 374 without) met the inclusion criteria (Median follow-up: 23 months). The mean age was 75.5 yrs, 61% male, and 86% white. The median post-nephrectomy OS and DFS was 8.61 and 4.44 yrs, respectively. Pts with and without recurrence had comparable baseline characteristics. Pts with recurrence had significantly shorter OS than those without [median: 2.53 yrs vs not reached; adjusted HR (95% confidence interval [CI]): 6.00 (4.24–8.48)]. Pts with recurrence by each landmark time point had significantly shorter OS than those without [1 yr post-nephrectomy median OS: 2.35 vs 9.66 yrs, and the OS 1, 3, and 5 after the 1 yr landmark was 69.9 vs 96.5%, 41.8 vs 83.8%, and 37.0 vs 70.1%, respectively; all Ps (log-rank test) < 0.001]. Cox models indicated that pts with recurrence by each landmark time point had 2.6–3.5 times increased risk of death compared with those without. Kendall’s τ rank correlation model demonstrated a statistically significant correlation between DFS and OS (Kendall’s τ = 0.70; 95% CI: 0.65–0.74; P < 0.001). Pts with recurrence had $4,924 and $1,387 higher adjusted all-cause medical costs and pharmacy costs per month (P < 0.001). Conclusions: Post-nephrectomy recurrence is associated with significantly shorter OS among pts with intermediate-high or high-risk RCC, resulting in a strong positive association between DFS and OS in the population. Higher healthcare cost was also seen among pts with recurrence.