External validation of previously reported prognostic classifications for patients treated with second-line axitinib after immuno-oncology combination therapy for advanced renal cell carcinoma: JUOG retrospective study.

482 Background: To validate the prognostic classifications previously reported for patients with advanced renal cell carcinoma who received second-line axitinib after treatment with nivolumab and ipilimumab. Methods: This study included 86 patients from 27 JUOG participating institutions. Treatment outcomes for axitinib as a second-line therapy were collected retrospectively. The prognostic classifications evaluated included the MSKCC risk model (5 factors), IMDC risk model (6 factors), JMRC risk model (4 factors [Cancer Sci. 2012]), ACL risk model (3 factors: albumin, LDH, CRP [Urol Oncol. 2019]), and ATP risk model (4 factors: metastatic disease at diagnosis, number of metastatic sites, albumin level, neutrophil lymphocyte ratio [Cancer Sci. 2020]). An external validation was conducted to evaluate the predictive accuracy for overall survival. Furthermore, risk factors associated with overall survival (OS) were examined using the Cox proportional hazards model. Variables with a p-value less than 0.05 in the univariate analysis were included in the multivariate analysis. Results: The median observation period was 17 months. The median OS and progression-free survival (PFS) for second-line axitinib treatment following immuno-oncology combination therapy were not reached (95% CI: 20-NR) months and 13 (95% CI: 8-14) months, respectively. The best overall response rates were CR 4.7% and PR 34.9%. The integrated Time-Dependent AUC for the MSKCC, IMDC, JMRC, ACL, and ATP classifications were 0.76, 0.76, 0.70, 0.78, and 0.69, respectively. In analyzing treatment prognostic factors, univariate analysis identified significant prognostic factors: less than one year from initial diagnosis, Karnofsky Performance Status <80%, hemoglobin < lower limit of normal, platelet count > upper limit of normal, neutrophil-lymphocyte ratio, albumin level, and CRP level. In the multivariate analysis, only the albumin level was identified as a significant prognostic factor (HR 1.85, 95% CI: 1.18-75.59, p=0.02). Conclusions: Pre-treatment albumin level was an independent prognostic factor for OS of the patients with second line axitinib treatment after immune-oncology combination therapy. Five prognostic models demonstrated comparable accuracy for predicting OS. It is noteworthy that the ACL risk model exhibited a good performance despite its simplicity, comprising only three factors. Comparison of trends in the area under the curve (AUC) values of time-dependent receiver operating characteristic (ROC) curves (prognostic performance) among five prognostic models for OS. Risk model 6 month 12 month 18 month 23 month Integrated ACL 0.77 0.80 0.77 0.74 0.78 IMDC 0.81 0.75 0.68 0.76 0.76 MSKCC 0.80 0.73 0.70 0.72 0.76 JMRC 0.71 0.65 0.67 0.67 0.70 ATP 0.73 0.65 0.68 0.68 0.69