Is a clinical biomarker predictive of outcomes with immune checkpoint inhibitor (ICI) therapy in a phase I patient population.

Abstract

e14136 Background: NLR is reported to be a prognostic marker in multiple malignancies and has reported potential predictive value with ICI. We evaluated NLR as a predictor of response and time to treatment failure (TTF) in a phase I patient population treated with ICI. Methods: Regulatory approval was obtained. A retrospective chart review of patients treated on phase I trials with ICI based therapy at Karmanos Cancer Institute, was conducted. Data were collected on demographics, performance status, tumor type, response rate, TTF and NLR pre and post therapy. Associations with clinical benefit (response rate and stable disease) and TTF were conducted by univariable logistic and Cox regression analyses. Results: 100 pts; 52 female and 48 male were evaluated. Median age was 62 years (range 22-84 years). 11 were African American (AA) and all patients were pretreated with standard therapy and had progressed prior to enrollment. Median pretherapy NLR was 3.31 (range 0.91-25.5). 26 patients showed clinical benefit with ICI based therapy; 6 partial responses and 20 with stable disease. Univariable Cox analysis for baseline continuous NLR associated with TTF showed a significant relation between NLR and TTF (p = 0.023). Univariable logistic and Cox analyses for dichotomized baseline NLR with optimal cut-off values, resulted in low NLR favoring higher chance of clinical benefit (Odds ratio = 0.169 95% CI 0.028,1.013, p = 0.043) and longer TTF (Hazard ratio = 1.841, 95% CI 1.188,2.854 p = 0.006). Conclusions: Pretherapy NLR was a statistically significant predictor of clinical benefit with ICI therapy in a phase I patient population. NLR is an easily applicable clinical predictive biomarker that may help guide clinical trial choice, especially in a patient population with limited life expectancy.