Persiani · Nutrients 2026 · systematic review · n=23 studies (306 patients)

Ketogenic Diet in the Treatment of Malignant Gliomas: A Systematic Review.

Level 1 - systematic review of randomized trials

Systematic review of intervention studies including randomized controlled trials and non-randomized designs

PubMed 42451166 · doi:10.3390/nu18132166 · record verified 2026-08-26

What was done

Authors systematically reviewed MEDLINE, Embase, and Cochrane Library records up to February 1, 2025, for studies evaluating ketogenic diet interventions (classic, modified, medium-chain triglyceride, or Modified Atkins diet, with or without fasting) in human patients with glioma compared to standard care or other diets. Evaluated outcomes included overall survival, progression-free survival, quality of life, biochemical markers (glucose, ketones, glucose-ketone index), anthropometrics, safety, and tolerability. Study quality was assessed with JBI tools, and results were synthesized narratively due to heterogeneity.

What was found

Twenty-three studies with 306 total patients met inclusion criteria: 2 randomized controlled trials (RCTs), 11 quasi-experimental studies, 6 case series, and 4 case reports. The overall intervention dropout rate was 20.3%. Included RCTs showed no statistically significant differences in overall survival or progression-free survival between ketogenic diet groups and controls, though exploratory analyses suggested better outcomes with lower glucose levels. Ketosis was consistently achieved, but blood glucose reductions were variable, and target glucose-ketone index values were rarely reached outside fasting. Weight and BMI generally decreased modestly or remained stable, and adverse events were mostly mild.

Why it matters

Despite popular interest and mechanistic plausibility regarding tumor glycolysis, current human evidence does not support survival or quality-of-life benefits for ketogenic diets in malignant glioma.

Limits

The evidence base is small and low quality, comprising only 2 RCTs alongside mostly uncontrolled designs (306 total patients across 23 studies). Protocol heterogeneity precluded quantitative meta-analysis, adherence was hindered by a 20.3% dropout rate, and target metabolic thresholds were rarely attained.

Cited by