External validation of the updated Leibovich prognostic models for clear cell and papillary renal cell carcinoma in an Asian population

Purpose The Leibovich model was updated to prognosticate oncological outcomes in postnephrectomy nonmetastatic renal cell carcinoma (RCC) for each major histological subtype including clear cell (ccRCC), papillary (papRCC), and chromophobe RCC. We evaluated its performance in an independent population of predominantly Asian patients from Singapore. Materials and Methods Nine hundred and forty two binephric patients with nonmetastatic unilateral RCC treated with radical/partial nephrectomy from 1990 to 2015 from Singapore were retrospectively reviewed. Based on the Leibovich model, ccRCC patients were scored from 0 to 25 and papRCC patients divided into 3 risk groups. Primary outcomes of progression-free survival (PFS) and cancer-specific survival (CSS) were assessed with the Kaplan–Meier method. Receiver operating characteristic curves and calibration plots were obtained to determine discrimination and calibration respectively. Results Eight hundred and twenty nine patients (88%) had ccRCC where 16.2% experienced disease progression while 11.9% died of RCC over a median follow-up of 76 (42–117) months. There was good discrimination (c-index 0.81 for PFS, 0.83 for CSS) and calibration (PFS calibration-in-the-large 0.002 and calibration slope 0.99, CSS calibration-in-the-large 0.005 and calibration slope 0.96). One hundred and thirteen patients (12%) had papRCC, where 18.6% progressed while 14.2% died from RCC over a median follow-up of 69.5 (36.0–112.0) months. Discrimination was slightly weaker (c-index 0.72 for PFS, 0.74 for CSS), and the model was only calibrated for CSS (calibration-in-the-large 0.002, calibration slope 0.98), not for PFS (calibration-in-the-large 0.09, calibration slope 1.93). Conclusions The updated Leibovich score is applicable for prognostication of progression and death in both ccRCC and papRCC, even when applied to an independent population of Asian patients. Further validation is required to ensure accuracy in prognosticating PFS for papRCC.