3250 - Patterns and Risk Factors of Locoregional Recurrence after Radical Nephrectomy for Localized Renal Cell Carcinoma: Implications for Patient Selection in Adjuvant Radiotherapy

Purpose/Objective(s): This study aims to analyze locoregional recurrence (LRR) patterns and identify risk factors in localized renal cell carcinoma (RCC) patients following radical nephrectomy (RN), with the goal of refining patient selection for adjuvant radiotherapy and improving survival outcomes.

Materials/Methods: We retrospectively analyzed 320 patients who underwent RN for non-metastatic RCC from January 2013 to December 2022. Univariate and multivariate Cox proportional hazards regressions were performed to measure the risk of LRR. LRR was stratified into two distinct components: local recurrence (LR) and regional lymph nodes metastatic (RLNM) for more detailed evaluation. We also mapped the position of LRR sites on reference CT which was reconstructed to be a 3D anatomy model including the blood vessels and bilateral kidneys using commercially available software.

Results: A total of 44 patients (13.8%) had LRR with a median follow-up duration 24.1 months (2.0-103.7months), while the median recurrence time from RN to LRR was 11.3 months (0.4-54.1months). T3-4, multifocality, and maximum diameter of tumor = 9 cm were identified as independent risk factors of LRR (P < 0.050). Patients with LRR had poorer overall survival (OS) (5-year OS rate 49.2 ± 8.2% vs 81.6 ± 2.6%, P = 0.000). We evaluated the RLNM patterns stratified by tumor locations. Left-sided RCC had more than 65.0% RLNMs in the para-aortic (PA) region. For right-sided RCC patients, RLNMs distributed in the PA (29.4%), interaortocaval (AC) (52.9%), and paracaval (PC) (17.7%) regions. No RLNM was observed above the diaphragmatic aortic hiatus or under the inferior mesenteric artery (IMA). We proposed a refined staging system for T3 stage: T3q (not invading perirenal fat), T3w (invading perirenal fat), and T3e (invading the vena cava or its wall), which enhances the LRR prediction. For LR, the optimal stratification model is “T1 + T2 + T3q”, and “T3w + T3e + T4” (5-year LR rate 5.3% vs 17.5%, P = 0.000). For RLNM, the optimal stratification model comprises “T1 + T2”, “T3q + T3w”, and “T3e + T4” (5-year RLNM rate 2.6% vs 8.9% vs 18.8%, P = 0.000).

Conclusion: T3-4, multifocality, and maximum diameter of tumor = 9 cm were the independent risk factors associated with LRR after RN for localized RCC. RLNM patterns were different according to the primary tumor sides. An enhanced T3 staging system had been proposed to improve the predictive accuracy for LR and RLNM. This study provides critical insights for identifying high-risk patients who may benefit from adjuvant radiotherapy, thereby guiding personalized treatment strategies and optimizing therapeutic outcomes.