Inflammatory signals hidden in the CBC are detectable prior to durvalumab-induced pneumonitis.

e21116 Background: For patients with unresectable stage III non-small cell lung cancer (NSCLC), chemoradiation followed by 1 year of durvalumab is standard of care. Treatment-induced pneumonitis can be life-threatening. Neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR) and systemic immune inflammation index (SII, defined as platelet x neutrophil / lymphocyte counts) are calculated from the CBC and have been studied as inflammatory biomarkers in a variety of settings. This study assessed trends in these indices among patients treated for stage III NSCLC. Methods: We reviewed medical records of adults with stage III NSCLC who completed chemoradiation and durvalumab from 1/2017 to 9/2021 as part of routine care. Patients who developed non-pneumonitis immune toxicity were excluded. NLR, PLR and SII were calculated at dates of first and last radiation dose and first durvalumab dose. For patients who developed pneumonitis, indices were also calculated at each of 8 weeks leading up to pneumonitis onset, defined as the first diagnostic CT scan and graded as 1-5 based on CTCAE criteria. Indices were compared between the pneumonitis and no pneumonitis group using rank-sum test. P values < 0.05 were considered statistically significant without adjustment for multiple testing. The slope of linear trend for each index was examined via log-gamma generalized linear regression model using a random intercept for each patient and robust standard error. Results: Of 65 patients who received chemoradiation followed by durvalumab, 2 were excluded because they were on clinical trial and 28 were excluded for non-pneumonitis immune toxicity. Among the 35 patients eligible for analysis, median age was 67 years; 20 (57%) were female; 6 (17%) were current smokers; 15 (43%) had COPD; and 17 (49%) developed pneumonitis (2 G1; 11 G2; 3 G3; 0 G4; 1 G5). In the pneumonitis group, NLR (p = 0.022), PLR (p = 0.007) and SII (p = 0.001) showed a statistically significant linear increase in the 5 weeks preceding pneumonitis onset. There were no differences in clinical characteristics, SII, NLR or PLR between the pneumonitis and no pneumonitis groups at the 3 time points examined. Conclusions: In this exploratory analysis, patients who developed pneumonitis showed a linear increase in inflammatory indices for 5 weeks before pneumonitis onset, suggesting information hidden in routine blood tests can anticipate development of a life-threatening complication. Our hypothesis-generating study will need validation in a larger prospective analysis with a control group. Future directions may also include examination of selected acute phase reactants or cytokines.