Histopathologic review of suspected disease progression in patients with recurrent glioblastoma (GBM) receiving nivolumab ± ipilimumab: CheckMate 143.

Abstract

2001 Background: Patients treated with immunotherapies can have a transient increase in lesion size due to immune cell infiltration that is suggestive of disease progression, but undergo subsequent regression with continued treatment. For patients with GBM, distinguishing progression from immune-related treatment effects using neuroimaging is challenging. Histopathologic examination of tumors could show the activity of immunotherapies in GBM and potentially minimize premature discontinuation of therapy and increase clinical benefit. Neuropathologic data from patients with recurrent GBM treated in CheckMate 143 who underwent biopsy/resection after suspected progression are presented. Methods: Patients received nivolumab 3 mg/kg (nivo 3) Q2W or nivo 3 + ipilimumab 1 mg/kg (ipi 1) Q3W × 4 doses then nivo 3 Q2W. Tissue collected from patients with suspected radiologic progression was submitted for blinded central review by 2 neuropathologists. Potentially significant treatment effect was defined as ≥30% necrosis/reactive changes and ≤50% viable solid tumor by area in posttreatment samples, or significant morphological changes from a prior biopsy if available. Results were compared with those from automated morphometry, pretreatment biopsy, and control pairs from unrelated patients with GBM treated with standard of care. Results: Of patients treated with nivo 3 (n = 20) or nivo 3 + ipi 1 (n = 1), 13 had potential treatment-related effects and 8 had no evidence of treatment effects (consistent with disease progression). Results were sent in real time to the treating physician to help inform treatment decisions; 4/13 patients with potential treatment effects continued therapy. Treatment effects were unrelated to age, MGMT methylation status, or PD-L1 expression (clone 28-8; archival tissue). Results were similar for automated morphometry analyses and showed differences between patients receiving nivo ± ipi and unrelated controls. Conclusions: These results suggest that histopathologic analyses may help better inform decisions on continuing nivo ± ipi in challenging cases, and show that immunotherapies may have intracerebral biological activity. Clinical trial information: NCT02017717.