Background Immune checkpoint inhibitors (ICIs) have become a cornerstone of treatment for patients with advanced non-small cell lung cancer (NSCLC). Many patients take corticosteroids, which can potentially decrease the efficacy of treatment; however, the potential impact of baseline corticosteroids at the time of treatment initiation and the underlying mechanisms remain elusive.

Methods We retrospectively evaluated 88 patients with NSCLC who received ICIs with or without peri-treatment steroids at medium or high doses as defined by a medium dose of >7.5 mg, but ≤30 mg prednisone equivalent a day and a high dose of >30 mg, prednisone equivalent daily. We assessed the progression-free survival (PFS), overall survival (OS), and overall response rate (ORR)

Results Eleven (13%) of 88 patients were taking corticosteroid at the start of ICI therapy. Patients who received steroids at any dose had a significantly lower PFS (mPFS: 324 days vs. 86.5 days and 96 days for no steroids (n=77) vs. medium (n=7) and high (n=4) steroids; p = 0.0152) while significantly lower OS was identified in patients with high doses of steroids (mOS: 639 days and 613 days vs. 207 days for no and medium steroids vs. high steroids; p = 0.0249). Patients who received steroids tend to have lower ORR (9.1% vs. 36.3% for steroids vs. no steroids, p = 0.0635). We analyzed the frequency of peripheral blood (PB) CD8+ T cells expressing a differentiation marker, CX3CR1, and discovered that the baseline frequency of CX3CR1+CD8+ T cells was significantly lower in patients who were on steroids (p = 0.0293). Using a bedside-to-bench approach, we confirmed that concurrent steroid use significantly decreased anti-tumor efficacy of anti-PD-1 therapy and PB CX3CR1+CD8+ T cells in mice bearing MC38 tumors while discontinuation of steroids at the start of treatment did not negatively impact survival and T-cell differentiation.

Conclusions Baseline corticosteroid use was associated with worse outcome in NSCLC patients treated with ICI therapy and decreased frequency of circulating CX3CR1+CD8+ differentiated T cells. Results from a pre-clinical model suggested that the discontinuation of steroid at the time of treatment initiation could maintain the efficacy of anti-PD-1 therapy and T-cell differentiation. Taken together, these findings support the recommendation to taper steroids to the lowest possible dose before ICI administration to alleviate symptoms without compromising the benefits of ICIs, if possible, and provide mechanistic insight into the impact of corticosteroid on T-cell differentiation.

Acknowledgements We would like to thank all our patients and their families who generously participated in our study. This work was supported by National Cancer Institute grants P30CA016056 and by the National Center for Advancing Translational Sciences of the National Institute of Health UL1TR001412. This work was further supported by the Department of Defense Lung Cancer Research Program (LC180245) and the National Cancer Institute grant K08CA197966, R01CA255240–01A1.

Ethics Approval All patients gave informed consent.

All patient information including blood samples, archived tumor tissue, and medical records was reviewed under the Institutional Review Boards of Roswell Park Comprehensive Cancer Center protocol approval (I 188310) and the University of Southern California protocol approval (HS-22–00134).