A prognostic classification system for extent of resection in IDH-mutant grade 2 glioma: an international, multicentre, retrospective cohort study with external validation by the RANO resect group

Abstract

Background: The efficacy of resection in IDH-mutant grade 2 gliomas remain controversial since terminology for the extent of resection has been inconsistently applied across studies. We aimed to establish a standardised classification for the extent of resection and assess the association between supramaximal resection and survival across molecular subtypes.

Methods: In this international, multicentre, retrospective study, patients aged 18 years and older with newly diagnosed grade 2 IDH-mutant glioma were identified from institutional databases across 16 centres in the USA, Europe, and Asia between between Sept 1, 1993, and May 10, 2024. We used Cox proportional hazard regressions to analyse the associations between residual tumour and progression-free survival and overall survival. Patients were stratified according to a previously postulated classification system based on residual tumour volume. A cohort of patients from UCSF diagnosed between Feb 16, 1998, and Nov 14, 2017, was used for geographically and institutionally independent external validation.

Findings: We identified 1391 patients with newly diagnosed IDH-mutant grade 2 gliomas, with a median follow-up of 81 months (95% CI 78-85). 728 patients (379 with astrocytoma and 349 with oligodendroglioma) received no first-line treatment beyond surgery, allowing us to study the isolated effects of resection. Patients with maximal T2-fluid attenuated inversion recovery (T2-FLAIR) resection (class 2; 0-5 cm3 remnant) had superior progression-free and overall survival compared with submaximal T2-FLAIR resection (class 3; 5-25 cm3 remnant) or minimal T2-FLAIR resection (class 4; >25 cm3 remnant), with 10-year survival rates of 82% (95% CI 76-87) versus 75% (62-84) versus 48% (29-65; p<0·0001) and 5-year progression-free survival rates of 44% (38-50) versus 25% (16-34) versus 12% (4-24; p<0·0001), respectively. Resection beyond T2-FLAIR borders (class 1) provided survival benefits, with a 10-year survival rate of 98% (95% CI 92-99) and a 5-year progression-free survival rate of 83% (76-88) for supramaximal T2-FLAIR resection (class 1). Associations between survival and extensive resection were evident after 3 years in astrocytomas, whereas survival curves separated after 6-8 years in oligodendrogliomas. The prognostic relevance of the four-tier classification was conserved in multivariable analyses, in 625 patients receiving first-line chemotherapy or radiotherapy (with or without chemotherapy), and in the external UCSF cohort of 381 patients with IDH-mutant grade 2 gliomas.

Interpretation: The proposed RANO classification for extent of resection could serve as a tool for prognostic stratification. Although associations between survival and extensive surgery are evident sooner in patients with astrocytoma, supramaximal resection also translates into survival benefits for patients with oligodendrogliomas.

Funding: None.