Predictive value of baseline clinicopathological factors in non-small cell lung cancer (NSCLC) patients on checkpoint inhibitors (CPI).

Abstract

Listen

Translate article

e14119 Background: Tumor PD-L1 expression and tumor mutational burden currently serve as primary predictive markers for CPI efficacy in NSCLC. However, these markers are imperfect due to tumor heterogeneity, changes over time, and lab variations. These challenges highlight importance of developing dynamic, readily available predictive biomarkers. We evaluated predictive value of pre-treatment Neutrophil-to-Lymphocyte Ratio (NLR), smoking history (SH), body mass index (BMI) and smoking intensity (SI) in NSCLC patients (pts) on CPIs. Methods: Retrospective analysis of NSCLC pts treated with CPIs July 2015 to November 2017. Pt demographics, tumor PD-L1 status positive (pos) or negative (neg) (PD-L1 > 0% or 0% respectively), SH, SI (heavy smokers (HS) or non-heavy smokers (NHS) [ = / > 20 pack-year (PY) and < 20 PY respectively], NLR and BMI high or low based on cutoffs of 5 and 25, respectively were captured. Disease Control Rate (DCR) was defined as objective response or stable disease per RECIST 1.1 at 3 months (m). Median overall survival (OS) and time to progression (TTP) were calculated. Fisher’s exact test and chi-square test were used to compare DCR for each group. Kaplan-Meier curves were used to estimate OS and TTP for each factor. Results: 140 pts were included. NLR was associated with DCR, OS, and TTP. 62/90 (68.9%) pts in the low NLR group and 18/47 (38.3%) in the high NLR group had DCR (p < 0.0006). Median OS for low NLR was 15 m (95% CI: 11.75, 22.25) vs. 5.25 m for high NLR (95% CI: 2.75, 9.75) (p < 0.0005). Median TTP for low NLR was 8 m (95% CI: 6.00, 11.25) vs 3 m for the high (95% CI, 2.00, 4.00) (p < 0.0001). SH was not associated with DCR or survival, but was predictive of TTP. SI was predictive of DCR, but not OS or TTP. 65/100 (65.0%) HS and 15/34 (44.1%) NHS had DCR (p < 0.0320). BMI and PD-L1 (n = 55) were not associated with any outcome. There was association between high BMI and low NLR (P < 0.0381). There were no other associations between factors. Conclusions: NLR < 5 was associated with improvement in all measures of outcome while smoking was associated with some. Neither BMI nor PD-L1 were predictive of any outcome. A score incorporating NLR and smoking status may be beneficial in choosing patients for CPIs.