Long-term autonomy, professional activities, cognition, and overall survival after awake functional-based surgery in patients with IDH-mutant grade 2 gliomas: a retrospective cohort study

Abstract

Background: In isocitrate dehydrogenase (IDH)-mutant low-grade gliomas (LGGs), awake functional-based resection (i.e., resection based on intraoperative functional responses rather than anatomical margins) has emerged as an efficient method to reduce tumour volume (TV) while minimizing postoperative deficits. Here, our goal was to assess the long-term onco-functional outcomes after awake functional-based resection in IDH-mutant LGGs, in conjunction with clinico-radiological and molecular factors.

Methods: We retrospectively studied a consecutive cohort (June 1997-January 2023) of 949 patients. Six hundred patients with IDH-mutant LGGs benefited from an awake functional-based resection with a median follow-up of 7.8 years (95% Confidence interval [CI]: 7.1-8.4 years). The main outcomes were the overall survival (OS), the OS with Karnofsky performance status ≥80% (OS_{KPS ≥ 80%}), cognition measures, and professional activities at 12 months post-surgery.

Findings: 600 patients were included in the cohort (274 female [46.0%], median age: 36 years [Interquartile range, IQR: 30-44 years]). The rate of return to work was 93.7%. The impact of surgery on cognition was of limited magnitude. The median postsurgical TV of 2.5 mL (IQR: 0-8.0 mL). The median OS was over 20 years (median: NA, 95% CI: 17.0-NA years). The median OS_{KPS ≥ 80%} was 14.7 years (95% CI: 13.2-17.2 years). Factors associated with longer OS and OS_{KPS ≥ P80%} were 1p19q codeletion (Hazard ratio [HR]_OS: 0.27, 95% CI: 0.16-0.43, HR_{KPS ≥ 80%}:0.25, 95% CI: 0.17-0.36), supratotal resection (HR_OS: 0.08, 95% CI: 0.005-0.40, HR_{KPS ≥ 80%}:0.12, 95% CI: 0.03-0.34) and total resection (HR_OS: 0.31, 95% CI: 0.16-0.59, HR_{KPS ≥ 80%}:0.21, 95% CI: 0.12-0.36). Recursive partitioning analyses established three OS and OS_{KPS ≥ 80%} prognostic groups, highlighting the contributions of histomolecular status, extent of resection, postsurgical and presurgical TV. Further propensity-matching analyses confirmed the oncological benefits of supratotal resections.

Interpretation: Awake functional-based resection surgery in newly diagnosed IDH-mutant grade 2 LGG, was an effective strategy associated with long survival (median OS over 20 years) and long-term preservation of autonomy. More complete tumor resections favored better onco-functional outcomes across all molecularly-defined subtypes. Short-term effects were of limited magnitude regarding postoperative cognitive and professional outcomes. Supratotal functional-based resections offered additional survival benefits.

Funding: None.

Keywords: Awake surgery; Functional outcomes; Survival; WHO grade 2 gliomas.

Fig. 1

Data flow diagram. IDH indicates isocitrate dehydrogenase gene 1 or 2.

Fig. 2

Kaplan–Meier curves and Hazard ratios for overall survival and overall survival with Karnofsky performance status ≥80%. A., from top to bottom, Kaplan–Meier curves for overall survival from initial surgery in all patients (n = 600), in IDH-mutant astrocytomas (n = 335) vs IDH-mutant 1p19q codeleted oligodendrogliomas (n = 265), and Hazard ratio results for overall survival (n = 600) using univariate Cox proportional hazard models and multivariate Cox proportional hazard models in variables eligible for analyses. B., from top to bottom, Kaplan–Meier curves for overall survival with Karnofsky performance status ≥80% from initial surgery in all patients (n = 600), in IDH-mutant astrocytomas (n = 335) vs IDH-mutant 1p19q codeleted oligodendrogliomas (n = 265), and Hazard ratio results for overall survival with Karnofsky performance status ≥80% (n = 600) using univariate Cox proportional hazard models and multivariate Cox proportional hazard models in variables eligible for analyses. Astro., IDH-mutant astrocytoma; Cox-PHM, Cox proportional hazard model; KPS, Karnofsky performance status; NA, not available; Oligo., IDH-mutant; 1p19q codeleted oligodendrogliomas; OS, overall survival.

Fig. 3

Recursive partitioning analysis for overall survival, overall survival with Karnofsky performance status ≥80% and resulting Kaplan–Meier curves and Hazard ratios stratified by risk groups. A., Three risk groups were determined by recursive partitioning analysis (n = 600) for overall survival, based on the following factors: age at surgery, sex, histomolecular status, the extent of resection, tumour location, presurgical and postsurgical tumour volumes, adjuvant chemotherapy, adjuvant radiotherapy, presurgical and 3-month postsurgical Karnofsky performance status, and epileptic status. B., Three risk groups were determined by recursive partitioning analysis (n = 600) for overall survival with Karnofsky performance status ≥80%, based on the following factors: age at surgery, sex, histomolecular status, the extent of resection, tumour location, presurgical and postsurgical tumour volumes, adjuvant chemotherapy, adjuvant radiotherapy, presurgical and 3-month postsurgical Karnofsky performance status, and epileptic status. C., Kaplan–Meier curves and hazard ratios for overall survival stratified by risk groups, as determined by recursive partitioning analyses. D., Kaplan–Meier curves and hazard ratios for overall survival with Karnofsky performance status ≥80%, stratified by risk groups, as determined by recursive partitioning analyses. Cox-PHM, Cox proportional hazard model; EOR, extent of resection; KPS, Karnofsky performance status; LGG, low-grade glioma; OS, overall survival.

Fig. 4

Propensity score analysis of overall survival and overall survival with Karnofsky performance status ≥80% in patients with supratotal resection (supraTR) vs patients with total resection (TR) and in patients with total resection or more (TR+) vs patients with less than total resection (TR−). A., Characteristics of patients selected for comparison with propensity score matching (supratotal resection vs total resection). Given the discrepancy between the number of control subjects (TR, n = 134) and the number of patients with supratotal resections (supraTR, n = 49), a 2:1 ratio was applied. Of note, 1:1 ratio analysis indicated non-significant difference between survival curves, although the same trend was observed (Log-rank test, p = 0.062, see Supplements). Kaplan–Meier curves for overall survival and overall survival with Karnofsky performance status ≥80% stratified by resection group are presented below. Log-rank tests were used for statistical comparisons. B., Characteristics of patients selected for comparison with propensity score matching (at least total resection or TR+ vs less than total resection or TR−). Kaplan–Meier curves for overall survival and overall survival with Karnofsky performance status ≥80% stratified by resection group are presented below. Log-rank tests were used for statistical comparisons.