Brigham and Women's Division of Urological Surgery and Center for Surgery and Public Health
Introduction: Negative surgical margins predict favorable outcomes in surgery for renal cell carcinoma (RCC). The effects of the surgical approach on the rate of indeterminate surgical margins (Rx) and the effect of Rx on overall survival (OS) are unknown. Rx is defined as a margin that is unable to be adequately evaluated for involvement of cancer. Recent studies have associated Rx with worse overall survival in lung cancer. We aim to report the proportion of Rx by surgical approach and compare OS by surgical margin status among patients with RCC. Methods: We retrospectively queried the National Cancer Database for patients with RCC between 2004-2020 who underwent open, laparoscopic, or robot-assisted partial- or radical nephrectomy. Hospital characteristics and patient-level sociodemographic and clinical variables including grade, histology, stage, and major vein involvement were compared across surgical margin status (R0: negative; R1: microscopic positive; R2: macroscopic positive and Rx: indeterminate). We compared the proportion of Rx by surgical approach using chi-square tests. We utilized multivariable logistic regression to predict Rx status adjusting for covariates. Multivariable Cox proportional hazard regression analysis was used to evaluate the impact of Rx compared to R0 on OS. Results: The cohort included 504,571 patients of which 36.2% received partial nephrectomy. Surgical approaches included robot-assisted (37.1%), laparoscopic (30.5%) and open (32.4%). Among patients receiving surgery Rx was reported in 0.58% for robot assisted, 0.47% in laparoscopic, 0.46% in open surgical approach. Rx was reported in 0.8% of partial- and 0.49% in radical nephrectomy cases (all p<0.01). On multivariable logistic regression, laparoscopic approach, partial nephrectomy, and any major vein involvement had increased odds of Ri (aOR 1.27; 95% CI 1.07-1.51; aOR 2.95; 95% CI 2.52-3.47; aOR 1.48; 95% CI 1.13-1.94; all p<0.01). Median OS was 172 months for R0 and 134 months for Ri (Log Rank Test; p<0.01). On adjusted multivariable Cox regression analysis, Rx was associated with worse OS compared to R0 (aHR 1.37; 95% CI 1.2-1.55; p<0.01). Conclusions: We report differences in rates of indeterminate surgical margins between surgical approaches for RCC. There appears to be an association between Rx status with worse OS. Our findings warrant further investigation but may help guide risk stratification after surgery. SOURCE OF Funding: None
