NIMG-18EVALUATING TUMOR PROGRESSION AND IMMUNOLOGIC REACTIONS DURING IMMUNOTHERAPY: NEUROIMAGING OBSERVATIONS FROM A MULTICENTER STUDY OF NIVOLUMAB MONOTHERAPY AND NIVOLUMAB IN COMBINATION WITH IPILIMUMAB IN RECURRENT GLIOBLASTOMA (CHECKMATE-143)

Abstract

BACKGROUND: Determining disease progression and response in glioblastoma (GBM) is challenging as standard therapies may alter tumor enhancement on MRI and technical factors complicate interpretation of imaging results in multicenter trials. The emergence of investigational immunomodulatory agents for GBM has further challenged investigators to distinguish possible treatment-related immunologic reactions at the tumor site from disease progression using non-invasive imaging techniques. METHODS: Eligible patients had first recurrence of primary GBM, no prior bevacizumab treatment, and KPS ≥70. Patients were randomized to nivolumab 3 mg/kg Q2W (NIVO) or nivolumab 1 mg/kg + ipilimumab 3 mg/kg Q3W followed by nivolumab 3 mg/kg Q2W (NIVO + IPI). Participating institutions underwent imaging standardization; sites were qualified and images were read locally then transmitted to an independent imaging core lab. MRI was obtained at baseline, week 7, week 13, then Q8W. RANO criteria were applied. Select patients (n = 7) underwent advanced and exploratory neuroimaging, including dynamic contrast-enhanced MRI (DCE-MRI) and [(18)F]-fluorodeoxyglucose positron emission tomography (FDG-PET). RESULTS: Twenty patients were enrolled (NIVO, n = 10; NIVO + IPI, n = 10). Median age: 57 (range 37–73); MGMT: methylated (n = 4), unmethylated (n = 10), and undetermined (n = 6). Investigator-assessed objective responses: partial response (n = 1), stable disease (n = 8), and progression (n = 10). A subset of patients demonstrated biopsy-confirmed immune-related treatment effects, with pathology showing reactive changes and minimal to no tumor. Additional neuroimaging in one of these patients showed hypermetabolism on the FDG-PET scan, while the DCE-MRI results were equivocal, with only mildly increased Ktrans and plasma volume. Standard and advanced neuroimaging results from this preliminary cohort will be reported. CONCLUSIONS: Use of checkpoint inhibitors in GBM may be associated with treatment-related imaging changes that may be difficult to distinguish from disease progression using standard MRI imaging. Additional multi-modality imaging studies, particularly perfusion MRI, may be helpful to address such discordant findings.