Clinical and laboratory parameters associated with rapid progression in advanced NSCLC patients treated with second or third-line single agent immune checkpoint inhibitors (ICIs).

Abstract

e21087 Background: While laboratory parameters including PDL1 expression and NLR are associated with outcomes in patients with metastatic NSCLC treated with ICIs, these measures have not identified patients with rapid progression (RP) defined as progressive disease within 30 days. Clinical factors including patient age and weight loss have been associated with cancer outcomes in general. Identifying clinical correlates for NSCLC patients who experience RP on ICI therapy would be useful. Methods: The objective of this retrospective study was to evaluate relationships between pre-treatment cachexia and inflammation and RP in NSCLC patients who received 2nd- or 3rd- line single agent ICIs. Associations of age, race, gender, smoking status, and longitudinal changes in weight and NLR (from at least 6 weeks prior to treatment initiation) with RP were analyzed by univariate and multivariate statistical (Kaplan-Meier and related) methods. Results: 195 patients were included: 59% female, 18% Black, and 78% current or former smokers. 14% of patients had RP. Black race was associated with RP (HR = 2.32, p = 0.03). 191 patients had pretreatment weight available. 63% had weight loss prior to ICI, 25% with > 5% loss. Any weight loss and weight loss > 5% over the time period ≥6 weeks prior to treatment initiation were associated with RP (HR = 3.19 and 6.40, p = 0.03 and < 0.01). 188 patients had pretreatment NLR values, 63% and 42% had NLRs > 3.5 and > 5, respectively. Pre-treatment NLR > 5 and higher baseline NLR were associated with increased risk of RP (p = 0.03 and p < 0.01). In multivariate analysis adjusted by age, smoking status and weight change, higher pre-treatment NLR is found to differentially increase RP risk for black patients (interaction HR = 1.27, p = 0.03). A model with these 5 variables provided 84.5% AUC of ROC curve (80% sensitivity, 75% sensitivity) for prognosticating RP. Conclusions: This retrospective study identified clinical variables including pre-treatment NLR > 5, weight loss > 0% & > 5%, black race, smoking status, and age that were associated with RP in previously treated advanced NSCLC patients receiving single agent ICIs. Though this requires validation with racially diverse data sets, these clinical parameters may be useful in identifying patients at high risk of RP on 2nd or 3rd line ICI therapy. Future directions include evaluating clinical characteristics and laboratory parameters in NSCLC patients treated with ICIs combined with chemotherapy and novel immunotherapy regimens, as well as single agent ICIs in the first line setting. If these clinical characteristics are associated with frequent RP in the setting of first line ICI treatment, it would be reasonable to consider novel immune strategies in this patient subset.