Evaluating inflammatory biomarkers as predictors of outcomes and immune-related adverse events (irAE) in the Big Ten Cancer Research Consortium (BTCRC) LUN16-081 phase II clinical trial.

e20087 Background: Immune checkpoint inhibition has improved patient outcomes across nearly every stage of NSCLC, and consolidation immunotherapy has become the standard of care following CCRT for unresectable stage III disease. However, these drugs can have significant adverse events, and improved biomarkers to improve patient selection are desperately needed. Numerous prior studies have evaluated the role of inflammatory biomarkers to predict outcomes in metastatic disease, but few have looked at their use in unresectable stage III NSCLC treated with immunotherapy. Methods: The BTCRC LUN 16-081 trial was a prospective, open-label phase II trial randomizing patients to either nivolumab or nivo/ipilimumab following CCRT for unresectable stage III NSCLC. From October 2017 to April 2021, 105 patients were enrolled and treated, and inflammatory biomarkers including neutrophil-lymphocyte ratio (NLR, <2.5 vs. ≥2.5), platelet-lymphocyte ratio (PLR, <200 vs. ≥200), von Willebrand Factor (vWF, <150% vs. ≥150%), erythrocyte sedimentation rate (ESR, <20 vs. ≥20), C-reactive protein (CRP, <median vs. >median), and ferritin (<median vs. >median) were collected at screening and each treatment timepoint. An analysis of samples from screening, C1D1, and C2D1 was performed and correlated with rates of irAEs as well as PFS and OS. An analysis comparing differences in inflammatory biomarkers between genders, histologies, and by PD-L1 status was also performed. Results: None of the biomarkers demonstrated an ability to predict for or against the development of irAEs at any of the timepoints. When evaluating changes from C1 to C2, there was a trend toward increased irAEs for increasing NLR (p=0.162) and CRP (p=0.168) but these were not statistically significant. In looking at PFS/OS across biomarkers, only vWF and CRP correlated with prognosis. Elevated vWF at C1 (median 24.8 months vs. NE, p=0.0194) and C2 (median 24.8 months vs. NE, p=0.0127) correlated with reduced PFS, and OS was numerically lower in this group though not statistically significant. At C2D1, high CRP correlated with reduced PFS (median 22.8 vs. 30.9 months, p=0.0427) and OS (median NE vs. NE, p= 0.0192). When comparing between groups, females demonstrated increased inflammation with higher median PLR, rates of elevated PLR, median ESR, and rates of elevated ESR at screening. At C1D1 and C2D1, only higher median ESR remained significant. By histology, median ferritin was higher in non-SqCC vs. SqCC patients at screening but not different at other timepoints. Median CRP and rate of elevated CRP were higher in SqCC patients at C2D1 but not other timepoints. None of the biomarkers correlated with PD-L1 score (pos vs neg) at any timepoints. Conclusions: Inflammatory biomarkers did not correlate with the incidence of irAEs. Most biomarkers did not appear predictive of PFS/OS. High vWF at C1 and C2 correlated with reduced PFS, and high CRP at C2 correlated with reduced PFS and OS. Novel biomarkers are needed to help personalize care in unresectable stage III NSCLC. Clinical trial information: NCT03285321 .