High neutrophil to lymphocyte ratio as a predictor for hyperprogressive disease in patients with metastatic non-small cell lung cancer treated with pembrolizumab as a second line.

Abstract

e21546 Background: In this study we evaluated the incidence of early non-responders (which are potential hyper-progressors (HP)) and its relation to Neutrophil-Lymphocyte Ratio (NLR), Platelet-Lymphocyte Ratio (PLR) as well as their dynamics in patients with Non-small Cell Lung Cancer (NSCLC) treated with pembrolizumab (P) as a second line. Methods: Patients with metastatic NSCLC (n=119) whose tumors expressed PD-L1≥1% were retrospectively analyzed between Apr 2017 and Oct 2019. All patients received platinum-containing chemotherapy (CT) as a first line treatment. Hematological parameters of interest were absolute neutrophil (ANC), absolute lymphocyte (ALC) and platelet (APC) counts, enabling calculation of NLR (ANC/ALC) and PLR (APC/ALC). NLR1 and PLR1 were calculated before CT, NLR2 and PLR 2 – before the first P infusion. ΔNLR (NLR2-NLR1) and ΔPLR (PLR2-PLR1) were calculated. The tumor response was assessed according to the RECIST (version 1.1) at every 3 months. Early progressors (EP) were defined as non-responders at the first computed tomography scan evaluation of the CT. As HP on P were considered patients with time to progression ≤ 3 months, clinical deterioration and appearance of ≥2 new lesions in organ already involved or spread to a new organ. Results: Twelve (10.1%) patients in the CT group were EP. Fourteen (11.8%) patients in the immunotherapy group were assessed as HP. Four patients (3.4%) did not respond either to CT or to P and had clinical aggravation. HP had significantly higher NLR2 (9.1±1.9 vs 4.8±3.2, p<0.001), PLR2 (400±173.8 vs 214.7±117.3, p<0.001), ΔNLR (2.3±1.9 vs 0.3±2.0, p=0.001), ΔPLR (107.9±167.3 vs 0.6±94.9, p=0.012) than the rest of the patients. ROC analysis revealed that at the optimal cutoff values of all markers, NLR2 achieved the greatest AUC=0.855 (95% CI, 0.77-0.93) and could distinguish between patients with or without hyperprogressive disease (HPD) with sensitivity of 85.7% and specificity of 72.4%. HP had significantly shorter mean OS - 12.5 months (95%, CI 10.9-14.1) than the rest – 32.2 months (95%, CI 28.4-36.1). Patients who did not respond either to CT or to P had significantly shorter mean OS – 9.2 months (95%, CI 6.6-11.8) than the rest – 29.8 months (95%, CI 26.4-33.4). NLR was found to be an independent predictive factor for HPD, HR=1.22 (95% CI,1.11-1.35; p<0.001). Conclusions: Our data suggest that the incidence of intrinsic aggressive disease phenotype, irrespective of treatment, is low. NLR is a novel predictive marker for HPD.