Giving steroids for indications other than cancer palliation doesn’t compromise the effectiveness of immunotherapy for advanced non–small cell lung cancer (NSCLC), suggests a single-center retrospective cohort study.
The immunosuppressant activity of corticosteroids and recent reports linking them to poorer outcomes has raised concern about their use during immunotherapy, noted, of Dana-Farber Cancer Institute and Harvard Medical School, Boston, and coinvestigators. But mechanisms underpinning this association are unclear.
The investigators studied 650 patients with NSCLC treated with immunotherapy targeting programmed death 1 (PD-1) or programmed death-ligand 1 (PD-L1), either as monotherapy or with other immunotherapy. Overall, 14.3% were receiving 10 mg or more of prednisone daily when they started the immunotherapy, a cutoff selected for study because it has been used an exclusion criterion for clinical trials.
Results reported in theshowed that, compared with other patients, those who received 10 mg or more of steroids indeed had poorer median progression-free survival (2.0 vs. 3.4 months; P = .01) and overall survival (4.9 vs. 11.2 months; P less than .001).
However, when the indication for steroid therapy was considered, median progression-free survival was just 1.4 months among patients who received 10 mg or more prednisone for cancer-related palliation, compared with 4.6 months among patients who received 10 mg or more prednisone for cancer-unrelated reasons (for example, autoimmune disease, chronic obstructive pulmonary disease flare, hypersensitivity prophylaxis, or management of noncancer pain) and 3.4 months among patients who received 0-10 mg of prednisone (P less than .001 across groups).
Similarly, median overall survival was just 2.2 months among patients who received 10 mg or more prednisone for palliative indications, but 10.7 months among patients who received 10 mg or more prednisone for cancer-unrelated reasons and 11.2 months among patients who received less than 10 mg prednisone (P less than .001 across groups).
In a multivariate analysis that adjusted for performance status and PD-L1 positivity and that used patients receiving up to 10 mg prednisone as the comparator, patients receiving 10 mg or more for cancer palliation had a trend toward high risk of progression-free survival events and a higher risk of death (hazard ratio, 1.40; P less than .06 and HR, 1.60; P = .02, respectively). In contrast, patients receiving 10 mg or more for cancer-unrelated reasons did not have elevated risks (HR, 0.62; P = .14 and HR, 0.91; P = .79, respectively).
“These data suggest that the significantly worse outcomes among patients who receive corticosteroids before immunotherapy are driven by the group of patients treated with corticosteroids for palliative oncologic symptom management, rather than by patients receiving corticosteroids for other reasons,” Dr. Ricciuti and coinvestigators wrote. “Corticosteroid use for cancer symptom management might simply correlate with patients who have adverse prognostic factors (e.g., brain metastases and poor performance status) rather than cause a clinically significant blunting of the response to [immune checkpoint inhibitors].
“Our data suggest that corticosteroids should not necessarily be decreased or discontinued before the start of immunotherapy out of a theoretical concern that corticosteroids could impair a response to immunotherapy,” the investigators concluded. “Additional mechanistic studies are needed to identify whether the use of corticosteroids affects specific aspects of the immune system necessary for immunotherapy activity.”
Dr. Ricciuti reported that he has no relevant conflicts of interest. The study did not receive any funding.
SOURCE: Ricciuti B et al. J Clin Oncol. 2019 Jun 17. .