TRIM63-associated signal in sunitinib-resistant xenografts and an eight- gene prognostic score in renal cell carcinoma

Abstract Background Sunitinib resistance limits the durable activity of anti-angiogenic therapy in renal cell carcinoma (RCC). This study screened TRIM-family expression in the GSE64052 786-O xenograft model and evaluated a TRIM63-anchored eight-gene risk score (TRIM-8RS) comprising OTX1, LIMCH1, ANK3, TRIM63, CAVIN2, CLTRN, WFDC10B and BBOX1 for prognostic association. Methods The score was evaluated in The Cancer Genome Atlas kidney renal clear cell carcinoma (TCGA-KIRC) development cohort (508 patients; 171 deaths) and an independent Clinical Proteomic Tumor Analysis Consortium operational primary-RCC cohort (183 patients; 31 deaths). Model development was repeated with all outcome-dependent candidate screening confined to outer training folds, and discrimination was assessed with Uno's concordance index and inverse-probability-of-censoring-weighted time-dependent receiver operating characteristic analysis. Results TRIM63 expression was directionally higher in resistant xenografts, but the association did not survive genome-wide false-discovery-rate adjustment (log2 fold change, 1.84; nominal P = 0.038; adjusted P = 0.291). The apparent Uno concordance index was 0.738 in the development cohort and 0.716 in the external cohort. The 1- and 3-year areas under the curve were 0.804 and 0.745 in TCGA-KIRC and 0.793 and 0.770 in the external cohort, and overall survival differed between risk groups defined by the frozen TCGA median score in both cohorts (hazard ratios, 4.56 and 4.51, respectively). The leakage-controlled reconstruction yielded a mean outer-fold concordance index of 0.699, and multivariable models further adjusted for age, sex and, in TCGA-KIRC, stage and grade. In public single-cell data, TRIM63 was detected in 49 of 17,986 cells, each with one raw unique molecular identifier, and spatial data showed low and variable fractions of TRIM63-positive spots across 11 slides. Conclusion These findings support a preliminary prognostic association and further biological validation. They do not establish a causal resistance mechanism or predict clinical benefit.