A single-institution retrospective study of multicentric gliomas stratified by <i>IDH</i> mutational status.

Abstract

2078 Background: Multicentric glioma (MCG) is a subset of diffuse glioma that can be synchronous or metachronous and is defined as the occurrence of two or more tumor foci, with separation of FLAIR (Fluid-attenuated Inversion Recovery) hyperintensity on MRI. MCG has not been extensively studied in studies stratifying IDH wild-type and mutant gliomas. This large single-institution study investigates the prevalence of MCG, examines the prognostic implications of MCG, and characterizes metachronous MCG (mMCG) in a cohort that has been stratified by IDH mutational status. Methods: In this IRB approved UCLA study, we identified diffuse glioma patients with known IDH mutational status and adequate MRI studies. Patients with multiple lesions on the MRI study pre-surgery or up to 3 months post-surgery were considered synchronous MCG (sMCG). Patients who developed a new independent lesion at least 6 months after initial surgery were considered mMCG. To qualify as MCG, we identified additional tumors on MRI that had no overlapping FLAIR borders and met one or more of the following: pathologically confirmed with biopsy, exhibited growth and thickening over time, and developed or increased in enhancement. Difference in prevalence was compared using Student’s t-test. Kaplan-Meier and Cox-multivariate analyses were used to analyze OS and time to metachronous (TtM) appearance. Results: We identified 911 consecutive IDH wild-type, high-grade diffuse glioma patients from 2013-2023 and 515 consecutive IDH mutant patients from 2007-2024 with pre-surgical MRI or MRI within three months of initial surgery. From the examined cohort, we found 39 IDH mutants with 21 sMCG and 18 mMCG and 153 IDH wild types with 95 sMCG and 63 mMCG. In eight IDH wild-type cases but no IDH mutant cases, mMCG arose from sMCG patients. We found that MCG had higher prevalence in IDH wild-types than in mutants (WT = 16%, Mut = 7%, p &lt; 0.0001), and IDH mutant MCG showed more male predominance than IDH wild-type MCG (Mut = 73%, WT = 58%, p &lt; 0.0001). In IDH mutant patients, mMCG, but not sMCG, was associated with lower OS (mMCG: HR = 2.476, p = 0.0115; sMCG: HR = 0.6437, p = 0.5027). However, in IDH wild types both sMCG and mMCG and were associated with lower OS (mMCG: HR = 1.589, p = 0.0025; sMCG: HR = 1.347, p = 0.0332). There was no difference in TtM between the two groups (HR = 0.5738, p = 0.5318). Amongst patients with multiple biopsied lesions, IDH wild types had consistent pathologies between lesions in all examined patients (29/29), but 71% of IDH mutants exhibited different pathologies between lesions (5/7). Conclusions: Our study examined a cohort of adult diffuse gliomas stratified by IDH mutational status and shows MCG is less common in IDH mutant gliomas and sMCG is not associated with worse prognosis. Further studies to identify molecular features underlying MCG will be valuable. Notes: For mMCG, FLAIR overlap might have occurred had the new lesion been observed synchronously.