Pretherapeutic Tumor Growth Velocity Refines Survival Stratification Beyond Clinical, Molecular, and Surgical Factors in IDH-Mutant Lower-Grade Gliomas

BACKGROUND: Contemporary risk stratification of isocitrate dehydrogenase (IDH)-mutant lower-grade gliomas relies primarily on molecular characteristics and surgical tumor burden but does not capture dynamic tumor behavior. Our primary objective was to determine whether pretherapeutic tumor growth velocity (TGV) provides independent prognostic information for overall survival (OS) beyond established factors. METHODS: This multicenter retrospective study included 569 adults with grade 2-3 IDH-mutant gliomas from three French centers, comprising a development cohort (n = 440) and two external validation cohorts (n = 129). TGV was derived from automated, outcome-blinded volumetric analysis of serial preoperative MRI using mean diameter expansion and log-volume growth. Prognostic value was assessed using multivariable Cox regression, external validation, decision curve analysis, and exploratory random survival forests. RESULTS: During a median follow-up of 6.8 years, 73 deaths occurred in the development cohort. Higher TGV was independently associated with shorter overall survival after adjustment for age, sex, grade, 1p/19q status, and postoperative residual tumor volume (HR per 1-mm/year increase, 1.046; 95%CI, 1.011-1.083; P=.011), and was independently replicated externally (HR, 1.077; 95% CI, 1.005-1.153; P=.035). Findings were consistent using log-volume growth and after accounting for CDKN2A/B status. Adding TGV improved prognostic model fit (likelihood-ratio χ²=5.70; P=.017), with greatest incremental value in astrocytomas. Decision curve analysis demonstrated additional clinical utility, while survival forests identified TGV among the strongest survival predictors alongside 1p/19q status and postoperative residual tumor volume. CONCLUSIONS: Pretherapeutic TGV independently predicts survival beyond contemporary molecular and surgical factors in IDH-mutant lower-grade gliomas. This accessible MRI-derived biomarker may refine individualized risk stratification, particularly in astrocytoma.