Development of a Prognostic Model for Patients Aged ≤65 Years With Metastatic Renal Cell Carcinoma

Background/Aim: Existing prognostic models for metastatic renal cell carcinoma (mRCC) were developed using all-age cohorts, and their validity in non-elderly patients remains unclear. We aimed to develop and validate our prognostic model for patients aged ≤65 years with mRCC, including non–clear cell histologies, and to establish a practical risk stratification system. Patients and Methods: We retrospectively analyzed 210 patients with mRCC from multiple Japanese institutions. The primary endpoint was overall survival (OS). Independent prognostic factors were identified using a multivariable Cox regression analysis. A simplified scoring system was constructed to stratify patients into risk groups. Model discrimination was evaluated using the concordance index (C-index), with internal validation by bootstrap resampling. Clinical utility was assessed by a decision curve analysis (DCA). A nomogram was constructed based on the identified independent prognostic factors. Results: During a median follow-up of 57 months, 77 OS events occurred; median OS was 65.7 months. Histopathological type, liver metastasis, C-reactive protein, serum-corrected calcium, and time to systemic therapy were independently associated with worse OS. Our prognostic model stratified patients into favorable- (n=89), intermediate- (n=76), and poor- (n=45) risk groups, with median OS of 95.1, 37.7, and 9.6 months, respectively (p<0.001). The 5-year C-index was 0.74, exceeding that of the International Metastatic RCC Database Consortium (IMDC) model (0.70). In DCA, our prognostic model showed higher net benefit across threshold of 0.3-0.6, corresponding to an average reduction of 3.26 unnecessary treatment escalations per 100 patients. The nomogram showed time-dependent C-indices of 0.83, 0.84, and 0.87 at 1, 3, and 5 years. Conclusion: Our prognostic model provides superior discrimination to and higher clinical utility than the IMDC model for predicting OS in non-elderly patients with mRCC, supporting age-specific risk stratification in this population.