Germline platelet polygenic risk score to predict renal cell carcinoma survival via sustained thrombocytosis.

Abstract

543 Background: Polygenic risk scores (PGS) reflect inherited germline variants fixed from conception and is measured early in life. While renal cell carcinoma (RCC) PGS has been developed for incident risk, its relevance to prognosis and treatment outcomes falls short. Moreover, several routine hematologic measures, partly under germline control, are prognostic in RCC, yet the extent to which their PGS influences survival remains unexplored. Methods: We assembled 495 RCC patients who underwent tumor sequencing and analyzed OS using left-truncated Cox proportional hazards regression with hematologic PGS and technical covariates. We performed survival analyses with different time origins: (1) from diagnosis (left-truncated at sequencing), (2) from sequencing, and (3) from ICI initiation. 321 PGS of hematological measures were drawn from the PGS Catalog. Results: Across 321 CBC-related scores, two PGS met false discovery rate (FDR) significance for OS in RCC: a PRS for higher platelet counts (PGS002660) associated with worse survival (Hazard Ratio [HR] 1.37, Confidence Interval [CI]: 1.17 - 1.62 , q=0.027), whereas a PRS for higher RDW (PGS004826) was protective (HR 0.74, CI: 0.63 - 0.87, q=0.027). PLT PGS showed clear fidelity capturing its intended biological trait (R² = 0.06, p = 0.001 in RCC; 0.03, p = 3x10 -51 pan-cancer). The PRS survival association was attenuated by 51% after adjusting for time-varying platelet counts measured longitudinally, while PLT remained strongly adverse (HR 1.40, CI: 1.29 - 1.58, p = 1.1x10 -11 ), suggesting effect operates through PLT levels. Tumor mutational burden did not modify PGS effects, consistent with prior observations of limited prognostic utility in RCC. Pan-cancer analyses demonstrated that PLT PGS effects during immune checkpoint inhibitor therapy were significant exclusively in RCC among 17 cancer types. Conclusions: Germline determinants of PLT are associated with RCC survival and operate predominantly through sustained elevation of circulating PLT rather than independent molecular pathways. These findings substantiate thrombocytosis’s adverse prognosis in RCC and establish PLT PGS as a heritable instrument that mechanistically links constitutional biology to longitudinal hematologic parameters and clinical outcomes. The integration of PLT PGS with serial CBC measurements offers potential to refine risk stratification and patient selection. Association of polygenic risk scores with overall survival in RCC. Variables 1 HR 95% CI P-value / q-value PLT (PGS002660) 1.37 1.17 - 1.62 p = 1.3 x 10 -4 q = 0.027 RDW (PGS004826) 0.74 0.63 - 0.87 p = 2.7 x 10 -4 q = 0.027 RDW (PGS001908) 0.76 0.65 - 0.89 p = 5.8 x 10 -4 q = 0.044 PLT (time-dependent) 1.43 1.29 - 1.58 p = 1.1x10 -11 PLT (PGS002660) adjusted for PLT 1.12 0.95 - 1.32 p = 0.20 PLT (PGS002660) adjusted for TMB 1.25 1.06 - 1.46 p = 6.9x10 -3 Interaction PLT (PGS002660) * PLT 0.88 0.80 - 0.97 p = 8.0x10 -3