Evaluating the Prognostic Variables for Overall Survival in Patients with Metastatic Renal Cell Carcinoma: A Meta-Analysis Of 29,366 Patients

Scoring systems are a method of risk assessment used to stratify patients with metastatic renal cell carcinoma (mRCC) and guide systemic therapy. The variables are weighed equally when calculating total score. However, the difference of even 1 positive predictor can change one's risk category and therapy. To compare the relative strength of association between predictive variables and overall survival (OS) in mRCC. A search of Medical Literature Analysis and Retrieval System Online (MEDLINE) and Embase was conducted. Clinical studies, retrospective and prospective, were included if the association of at least 1 predictor and OS in patients with mRCC receiving first-line systemic therapy was evaluated. Meta-analysis was performed to generate pooled hazard ratios (HRs) and 95% CIs for OS for predictors with ≥ 5 included studies. Sensitivity analysis identified outlier heterogeneity and publication bias. Sixty-six studies containing 29,366 patients were included. Meta-analysis indicated lung metastases, bone metastases, thrombocytosis, time to systemic therapy < 1 year, liver metastases, hypercalcemia, anemia, elevated neutrophil-lymphocyte ratio, multiple metastatic sites, neutrophilia, poor Eastern Cooperative Oncology Group (ECOG) status, no previous nephrectomy, elevated lactate dehydrogenase, Fuhrman grade 3 or 4, central nervous system metastases, elevated C-reactive protein, and Karnofsky Performance Status < 80% were associated with significantly worse OS. The HRs varied from 1.34 to 2.76, representing heterogeneity in predictive strength. The effects of study heterogeneity and publication bias were minimal to moderate across all predictors. Based on the differences in pooled HRs, prognostic strength between the variables is likely not equivalent. Restructuring scoring models, through inclusion of other variables and usage of relative weighting, should be considered to improve accuracy of risk stratification.