Baseline neutrophil-to-lymphocyte ratio as a predictive and prognostic biomarker in patients with metastatic castration-resistant prostate cancer treated with cabazitaxel versus abiraterone or enzalutamide in the CARD study.

BACKGROUND:There is growing evidence that a high neutrophil-to-lymphocyte ratio (NLR) is associated with poor overall survival (OS) for patients with metastatic castration-resistant prostate cancer (mCRPC). In the CARD study (NCT02485691), cabazitaxel significantly improved radiographic progression-free survival (rPFS) and OS versus abiraterone or enzalutamide in patients with mCRPC previously treated with docetaxel and the alternative androgen-receptor-targeted agent (ARTA). Here, we investigated NLR as a biomarker. PATIENTS AND METHODS:CARD was a multicenter, open-label study that randomized patients with mCRPC to receive cabazitaxel (25 mg/m2 every 3 weeks) versus abiraterone (1000 mg/day) or enzalutamide (160 mg/day). The relationships between baseline NLR [< versus ≥ median (3.38)] and rPFS, OS, time to prostate-specific antigen progression, and prostate-specific antigen response to cabazitaxel versus ARTA were evaluated using Kaplan-Meier estimates. Multivariable Cox regression with stepwise selection of covariates was used to investigate the prognostic association between baseline NLR and OS. RESULTS:The rPFS benefit with cabazitaxel versus ARTA was particularly marked in patients with high NLR {8.5 versus 2.8 months, respectively; hazard ratio (HR) 0.43 [95% confidence interval (CI) 0.27-0.67]; P < 0.0001}, compared with low NLR [7.5 versus 5.1 months, respectively; HR 0.69 (95% CI 0.45-1.06); P = 0.0860]. Higher NLR (continuous covariate, per 1 unit increase) independently associated with poor OS [HR 1.05 (95% CI 1.02-1.08); P = 0.0003]. For cabazitaxel, there was no OS difference between patients with high versus low NLR (15.3 versus 12.9 months, respectively; P = 0.7465). Patients receiving an ARTA with high NLR, however, had a worse OS versus those with low NLR (9.5 versus 13.3 months, respectively; P = 0.0608). CONCLUSIONS:High baseline NLR predicts poor outcomes with an ARTA in patients with mCRPC previously treated with docetaxel and the alternative ARTA. Conversely, the activity of cabazitaxel is retained irrespective of NLR.