Time-dependent efficacy of checkpoint inhibitor nivolumab in metastatic lung cancer patients.

Abstract

9096 Background: Nivolumab (NIV) is a Programmed-cell-Death-1 inhibitor approved as 2nd line treatment for metastatic Non-Small Cell Lung Cancer (NSCLC). NIV mainly targets T(CD8) cells, whose functions and trafficking are regulated by circadian clocks (Nobis et al. PNAS 2019), hence suggesting possible dosing time-dependent changes in NIV efficacy. Methods: Consecutive metastatic NSCLC patients (pts) received single agent NIV (240 mg iv q 2 weeks) at a single institution. NIV timing slots were randomly allocated for each course by the day hospital coordinator on a logistics basis and recorded. The median NIV timing and its intra-pt coefficient of variation (CVar) were computed over the whole treatment span. The study population was split into two NIV timing groups based upon the median value of the median treatment times of all the pts. CTCAE-toxicity rates were compared between groups with c2 or Fisher exact. Progression free survival (PFS) and overall survival (OS) were compared between both NIV timing groups with Log Rank. Results: From 9/2015 to 11/2020, the study accrued 95 stage 4 NSCLC pts (males, 83%; PS 0-1, 96%), aged 41-83 years (median, 67). Primary histological types were adenocarcinoma (55 pts, 58%), squamous cell carcinoma (37 pts, 39%) or unspecified (3 pts, 3%). The pts had a median of 4 metastatic sites, including bone (52% of the pts), pleura (41%), liver (25%), brain (24%) and adrenal gland (20%). A total of 1818 NIV courses were given as 2nd line for 72 pts (76%), or as 3rd or later line for 22 pts (23%). Median PFS and OS (months, mo.) were 3.9 mo. [95% CL, 2.1 – 5.8], and 14.0 mo. [9.5 – 18.4] respectively, for the 95 pts. The majority of NIV administrations occurred between 9:27 and 12:54 for 48 pts (‘morning’ group) and between 12:55 and 17:14 for 47 pts (‘afternoon’ group), with intra-pt NIV timing CVar ranging from 2% to 21% (median, 10%). Main pts characteristics were similar for both groups, except for fewer females (8% vs 26%) and younger age (median, 66 vs 69 years) in the ‘morning’ group compared to the ‘afternoon’ one. Grade 3-4 fatigue, anorexia or myalgias were less in the ‘morning’ group compared with the “afternoon’ one (6% vs 15%; 2% vs 6%; 0% vs 4%, respectively). Strikingly, median PFS [95% CI] were 11.3 mo. [5.5 - 17.1] for the ‘morning’ group as compared to 3.1 mo. [1.5 - 4.6] for the ‘afternoon’ one (p<0.001). Median OS were 34.2 mo. [15.1 - 53.3] for the ‘morning’ group vs 9.6 mo. [4.9 - 14.4] for the ‘afternoon’ group (p<0.001). Multivariate analyses identified NIV ‘morning’ timing and 2nd line administration, as significant independent predictors of longer PFS and OS. Conclusions: NIV was both less toxic and four times as effective following ‘morning’ as compared to ‘afternoon’ dosing in this study in Stage 4 NSCLC pts, possibly as a result of dosing time-dependent pharmacology. Translational and clinical nivolumab timing validation studies are needed, in order to optimize pts benefits from cancer immunotherapy.