Steroids Diminish Survival Benefits From ICIs in Lung Cancer

Corticosteroid medications are associated with poorer outcomes of therapy with immune checkpoint inhibitors for patients with lung cancer.

Baseline steroid use at initiation of programmed death 1/programmed death ligand 1 (PD-1/PD-L1) immune checkpoint blockade is associated with shorter survival among patients with non–small-cell lung cancer (NSCLC), even after statistically controlling for the underlying conditions treated with corticosteroid therapy, like brain metastases. That is the conclusion of investigators who conducted a retrospective review of medical records from Memorial Sloan Kettering Cancer Center (MSKCC) in New York City and Institut Gustave Roussy, Villejuif, France. The study findings (abstract 9003) were presented at the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting, held June 1–5 in Chicago.

“Prudent use of steroids at the time of initiating PD-1/PD-L1 blockade is recommended,” said lead study author Kathryn Cecilia Arbour, MD, of MSKCC.

Nonsteroid alternatives for managing cancer symptoms should be considered if a patient might be administered PD-1/PD-L1–targeting immunotherapy, Arbour advised. However, she was quick to emphasize that medically necessary steroid therapy, such as for brain metastases, should not be avoided.

The study did not determine the efficacy implications of steroid therapy for patients receiving both immunotherapy and chemotherapy, she noted.

Little is known about the efficacy of immunotherapy in patients taking steroid medications, in part because patients taking prednisone and other steroids have been ineligible for clinical studies of anti–PD-1/PD-L1 agents. Steroids are known immunosuppressants and are used to control the inflammatory immune-related adverse events (irAEs) sometimes triggered by immunotherapy-but treatment of irAEs with steroids does not appear to diminish the antitumor efficacy of immune checkpoint blockade, Arbour noted.

To clarify the impact of steroids on immunotherapy for lung cancer, the researchers evaluated the efficacy of PD-1/PD-L1 blockade in patients with NSCLC who received baseline oral or IV steroids before single-agent immune checkpoint blockade for NSCLC at MSKCC (n = 455) and Gustave Roussy (n = 185). (Fewer than 20 patients who were taking prednisone at a dose lower than 10-mg daily were grouped with the no-steroids cohort.)

Baseline steroids were associated with a lower immunotherapy-related overall response rate (ORR) in both cohorts. At MSKCC, 6% of patients on steroids at baseline experienced complete or partial tumor response, vs 19% of patients not taking steroids. At Gustave Roussy, the outcomes were comparable: 8% of patients on steroids at baseline had complete or partial tumor response, compared with 18% of patients not taking steroids.

Progression-free survival (PFS) was also impacted by steroid use (MSKCC cohort: PFS hazard ratio [HR], 1.7; P < .0001; Gustave Roussy cohort: PFS HR, 1.5; P < .0001), as was overall survival (OS HR, 2.1 and 2.0, respectively; P < .0001 and P < .001).

Pooled subgroup analyses showed that the impact of steroids on immunotherapy outcomes held true regardless of age, sex, performance status, NSCLC histology (squamous or nonsquamous), and the presence or absence of brain metastases.

In a multivariate analysis, steroid use remained a statistically significant correlate of poorer immunotherapy ORR, PFS, and OS, even after controlling for smoking status, performance status, and history of brain metastases.