Frontiers in Oncology (Mar 2023)
Proposal for a Two-Tier Re-classification of Stage IV/M1 domain of Renal Cell Carcinoma into M1 (“Oligometastatic”) and M2 (“Polymetastatic”) subdomains: Analysis of the Registry for Metastatic Renal Cell Carcinoma (REMARCC)
Abstract
PurposeWe hypothesized that two-tier re-classification of the “M” (metastasis) domain of the Tumor-Node-Metastasis (TNM) staging of Renal Cell Carcinoma (RCC) may improve staging accuracy than the current monolithic classification, as advancements in the understanding of tumor biology have led to increased recognition of the heterogeneous potential of metastatic RCC (mRCC).MethodsMulticenter retrospective analysis of patients from the REMARCC (REgistry of MetAstatic RCC) database. Patients were stratified by number of metastases into two groups, M1 (≤3, “Oligometastatic”) and M2 (>3, “Polymetastatic”). Primary outcome was overall survival (OS). Secondary outcomes were cancer-specific survival (CSS). Cox-regression and Kaplan-Meier (KMA) analysis were utilized for outcomes, and receiver operating characteristic analysis (ROC) was utilized to assess diagnostic accuracy compared to current “M” staging.Results429 patients were stratified into proposed M1 and M2 groups (M1 = 286/M2 = 143; median follow-up 19.2 months). Cox-regression revealed M2 classification as an independent risk factor for worsened all-cause mortality (HR=1.67, p=0.001) and cancer-specific mortality (HR=1.74, p<0.001). Comparing M1-oligometastatic vs. M2-polymetastatic groups, KMA revealed significantly higher 5-year OS (36% vs. 21%, p<0.001) and 5-year CSS (39% vs. 17%, p<0.001). ROC analyses comparing OS and CSS, for M1/M2 reclassification versus unitary M designation currently in use demonstrated improved c-index for OS (M1/M2 0.635 vs. unitary M 0.500) and CSS (M1/M2 0.627 vs. unitary M 0.500).ConclusionSubclassification of Stage “M” domain of mRCC into two clinical substage categories based on metastatic burden corresponds to distinctive tumor groups whose oncological potential varies significantly and result in improved predictive capability compared to current staging.
Keywords