Immunotherapy with PD-1 and PD-L1 immune checkpoint inhibitors has become the standard of care in patients with non-small cell lung cancer (NSCLC) who do not express targetable alterations such as epidermal growth factor receptor (EGFR) mutations, anaplastic lymphoma kinase (ALK), or ROS1 translocation/rearrangements. Because these agents function by stimulating the immune response, use of immunosuppressive agents such as corticosteroids has the potential to decrease the efficacy of PD-(L)1 blockade. Corticosteroids are often used in patients with advanced NSCLC to treat cancer-related symptoms such as fatigue, dyspnea, decreased appetite, or symptoms due to brain metastases. Steroids are also the key treatment for immunotherapy-related adverse events (irAEs). While use of corticosteroids to control irAEs does not appear to diminish treatment efficacy, the impact of their use at the time of immunotherapy initiation is not known, as patients on steroids are typically ineligible for immunotherapy clinical trials. A recent study evaluated real-world data from Memorial Sloan Kettering Cancer Center (MSKCC) and Gustave Roussy Cancer Center (GRCC) to determine the impact of baseline corticosteroid use on the efficacy of PD-(L)1 inhibitors in patients with advanced NSCLC.
Data from 640 patients treated with PD-(L)1 inhibitor monotherapy were included in this analysis, 14% of whom were receiving corticosteroids (≥10 mg prednisone) at the start of treatment. The most common reasons for corticosteroid use included dyspnea (33%), fatigue (21%), and brain metastases (19%). Among the cohort of patients treated at MSKCC, baseline corticosteroid use was associated with significant decreases in overall response rate (ORR; 6% vs 19%; P = .02), median progression-free survival (PFS; 1.9 months vs 2.6 months; HR 1.7, P = .001), and median overall survival (OS; 5.4 months vs 12.1 months; HR 2.1, P<.001). In the GRCC cohort, baseline corticosteroids were associated with significant decreases in PFS (1.7 months vs 1.8 months; HR 1.5, P<.001) and OS (3.3 months vs 9.4 months; HR 2.0, P<.001). While there was a decrease in ORR with baseline corticosteroids in the GRCC cohort, the difference was not significant (8% vs 18%; P = .2).
In the pooled cohort of patients from both centers, a multivariate analysis adjusted for smoking history, performance status, and brain metastases found a significant correlation between baseline corticosteroid use and decreased PFS (HR 1.31, P = .03), and OS (HR 1.66, P<.001). Prednisone doses of 10 mg to 19 mg had similar detrimental effects as doses >20 mg. Patients who discontinued corticosteroids 1 day to 30 days prior to PD-(L)1 inhibitor treatment had intermediate PFS and OS compared to those who remained on corticosteroids at the start of treatment and those who received no steroids for 30 days prior to treatment.
The investigators concluded that baseline corticosteroid use diminished the efficacy of immune checkpoint inhibitor therapy, indicating that the impact of corticosteroid use should be carefully considered in patients with NSCLC for whom PD-(L)1 inhibitor therapy is planned. They highlighted that these data apply only to patients receiving single agent PD-(L)1 inhibitor therapy, and that implications for patients receiving a combination of chemotherapy and PD-(L)1 inhibitor is uncertain.