Glioblastoma (GBM) is the most aggressive malignant primary brain tumor. Despite advances in multimodality treatment, including surgery, radiotherapy, chemotherapy, targeted therapy, and supportive care, the overall prognosis remains poor. Throughout the course of the disease, many patients with brain cancer also receive corticosteroids, such as dexamethasone, to treat symptomatic cerebral edema that develops due to tumor progression, as well as treatment (e.g. external beam radiotherapy). While immunotherapy with immune checkpoint inhibitors (ICI) is transforming the therapeutic landscape for many advanced cancers, results from phase III clinical trials of ICIs in recurrent (CheckMate 143) and newly-diagnosed GBM (Checkmate 498) have been disappointing. This suggests that the immune environment in the brain may be different to other organs, and GBM may exploit various mechanisms of immunosuppression. There is also growing evidence that corticosteroid use may negatively affect ICI effectiveness in cancer treatment. Indeed, subgroup analyses from the CheckMate-143 trial indicated that overall survival (OS) was significantly worse in patients who were receiving corticosteroids at baseline versus no corticosteroids.

An interesting, recently reported study has evaluated the effect of dexamethasone treatment in GBM mouse models, and in an observational cohort of 181 patients with GBM treated with programmed cell death protein 1 (PD-1) or programmed death ligand 1 (PD-L1) inhibitors. In the preclinical part of the study, the dose-dependent effect of dexamethasone on anti-PD-1 efficacy was investigated. In an immune-sensitive mouse model, anti-PD-1 treated mice that received concurrent dexamethasone experienced a dose dependent worsening of OS: at 100 days, OS was 47.1% with a dexamethasone dose of 1mg/kg (P=0.04), 31.3% at 2.5mg/kg (P=0.008), and 26.5% at 10mg/kg (P<0.001). Of note, when mice were given dexamethasone prior to anti-PD-1 antibody treatment, there was no impact on OS (P=0.64). For mice that received anti-PD-1 and radiotherapy, the addition of dexamethasone demonstrated a trend towards decreased median and long-term survival, but was not statistically significant (P=0.15). Similar results were seen in an immune-resistant mouse model. The complex mechanism by which dexamethasone exerts its negative effects is thought to involve a reduction of T-lymphocyte counts, partly through inducing apoptosis, in addition to a decrease in lymphocyte functional capacity, and a reduction in the number of myeloid and natural killer cells. Thereby, both the adaptive and innate immune responses are diminished.

In the cohort of patients with GBM who had received anti-PD-(L)1 therapy, 75.7% had recurrent disease and 24.3% were newly diagnosed. The median follow-up from diagnosis was 22.2 months, and 84.5% of patients have died. Median OS (not adjusted) was significantly decreased from 13.1 months in patients with no baseline dexamethasone (64.6%) to 8.1 months in patients receiving <2mg daily dexamethasone (16.0%; P<0.001) and 6.3 months in patients receiving ≥2mg (19.3%; P=0.001). This negative effect persisted in multivariate analyses adjusted for disease setting (newly diagnosed versus recurrent GBM), age, MGMT promoter methylation status, Karnofsky Performance Status, tumor volume at anti-PD-(L)1 initiation and extent of resection. Baseline dexamethasone was the strongest identified negative risk factor for OS.

The authors note that while the clinical data require prospective validation in a randomized, controlled trial, they are consistent with previous observations. Hence, these preclinical and clinical data indicate that concurrent administration of dexamethasone, even at low doses, may limit the therapeutic benefit of ICIs in patients with GBM. Therefore, the authors conclude that dexamethasone use in patients with malignant brain tumors undergoing immunotherapy clinical trials should be carefully monitored and that alternative approaches to treating symptomatic cerebral edema, to avoid exposure to dexamethasone in these patients, merits further study.

Reference

Iorgulescu JB, et al. Clin Cancer Research 2020; Nov 25. [Online ahead of print].