Ketogenic Diet for Cancer: What Clinical Trials Actually Show

Ketogenic diets have a plausible metabolic rationale in cancer research, but they have not earned a place as a proven cancer treatment. The central gap is surprisingly basic: the available randomized cancer trials have not reported blood ketone measurements alongside clinical outcomes, so readers cannot tell whether participants consistently achieved the metabolic state the intervention was meant to test.

That does not prove ketogenic diets fail. It means claims that keto slows tumors, improves survival, or reliably enhances standard treatment run ahead of the human trial evidence. The newer oncology studies focus on whether patients can follow the diet safely during demanding treatment—not whether the diet extends life.

What the trial evidence can—and cannot—answer

Two published meta-analyses sit at the top of this evidence base. Yang et al. (2021) reviewed randomized controlled trials of low-carbohydrate ketogenic diets as adjuncts to cancer therapy. It examined metabolic measures including cholesterol fractions, triglycerides, fasting glucose, insulin, body weight, and tumor markers such as CEA, CA19-9, PSA, and TNF-alpha. The authors used standardized mean differences with 95% confidence intervals and required studies to document adherence and dropout.

That adherence requirement was sensible. A trial of dietary advice is not necessarily a trial of sustained ketosis. Yet the publication does not provide the quantitative results needed here to report effect sizes, confidence intervals, adherence rates, or adverse-event rates. Its search ended in December 2019, so it also predates several newer feasibility studies.

Amanollahi et al. (2022) examined cardiometabolic outcomes in cancer patients using ketogenic diets. Its review included controlled clinical trials as well as randomized trials. That broader approach may increase the number of included studies, but it makes the pooled findings less useful for answering a narrow question: what do randomized trials alone show? The available report does not provide recoverable numerical outcomes for this article.

Neither review establishes an effect on progression-free survival or overall survival. Metabolic markers and tumor markers can be useful measurements, but they are not substitutes for evidence that patients live longer, feel better, or respond better to treatment.

The most relevant cancer trials are feasibility studies

The Keto-CARE trial, published in 2024, studied a well-formulated ketogenic diet providing 20 to 50 grams of carbohydrate per day in women with stage IV metastatic breast cancer. The trial was designed around adherence, safety signals, and metabolic feasibility rather than progression-free survival or overall survival. That is an appropriate first question in a population already at risk of nutritional compromise, but it cannot establish anticancer efficacy.

The published information available here does not report a numerical adherence rate, blood beta-hydroxybutyrate concentration, or clinical outcome estimate. Therefore, it cannot support a claim that the diet improved cancer control—or that it did not.

A newer trial, KOMPARC, is testing ketogenic-diet tolerability and safety in people with cervical or endometrial cancer receiving pelvic radiotherapy. It is prospective, randomized, and open-label, with preliminary results from an ongoing study. This setting deserves caution: pelvic radiotherapy can cause gastrointestinal toxicity and treatment-related weight loss, while dietary restriction can complicate efforts to preserve intake and body weight. The trial's safety findings may not transfer to people with other cancers or treatment regimens.

A feasibility trial can show whether a protocol is practical enough to test. It cannot answer whether the protocol treats cancer unless it was designed and powered to answer that question.

The MCT question has no oncology dose answer

Medium-chain triglycerides, or MCTs, are often promoted as a shortcut to higher ketones. The relevant controlled study does show that researchers can test this question rigorously—but it was not conducted in cancer patients.

In Nakamura et al. (2022), healthy adult men spent four days in an inpatient, double-blind, parallel-group trial. Researchers compared a ketogenic diet plus a 3:1 ketogenic-ratio MCT formula with a ketogenic diet plus a long-chain-triglyceride placebo formula and with a normal-diet control. The MCT formula provided about 1,112 kcal per day, and the target macronutrient distribution was 22% protein, 70% fat, and 8% carbohydrate.

This is the only controlled human study described here that was specifically designed to assess MCT-formula dosing for raising blood ketones. But its numerical beta-hydroxybutyrate and acetoacetate results are unavailable here, and its short, controlled setting matters. Healthy men do not have the metabolic demands of advanced cancer, cachexia risk, systemic treatment, nausea, altered appetite, or radiation-related bowel symptoms. The study does not establish an MCT dose that reliably produces therapeutic ketosis in oncology.

A blood beta-hydroxybutyrate level of at least 0.5 mmol/L is commonly used in broader ketogenic-diet literature to indicate nutritional ketosis. No study described here establishes that threshold as therapeutic for cancer, and no cancer-specific MCT dose-response trial has shown how to achieve it safely and consistently.

Indirect evidence does not solve the cancer question

Ketogenic diets and MCT supplementation can produce measurable metabolic changes in other settings. A 2025 meta-analysis of randomized trials of carbohydrate-restricted diets in adults assessed glucose, HbA1c, insulin, HOMA-IR, lipid measures, beta-hydroxybutyrate, and other metabolic outcomes. Feng et al. provides useful context for why ketogenic diets can alter metabolic biomarkers, but it does not test cancer outcomes or patients undergoing active cancer treatment.

Similarly, a 2024 randomized feasibility trial in Parkinson's disease evaluated an MCT-supplemented ketogenic diet for feasibility, acceptability, blood ketones, and exploratory clinical measures. A retrospective pediatric epilepsy study also examined an MCT-enhanced ketogenic formula, where formulated products may help with adherence and adverse effects. Neither population can answer whether MCTs improve cancer outcomes or identify an oncology dose.

Why the current evidence remains uncertain

The studies differ in cancer type, treatment context, dietary protocol, and intended endpoint. Metastatic breast cancer and gynecologic cancers treated with pelvic radiotherapy pose different nutritional and clinical problems. Combining them into one broad claim about "keto and cancer" would hide the question that matters most: which patients, under which treatment conditions, using which monitored diet?

The larger limitation is intervention verification. If a cancer trial does not pair blood ketone data with outcomes, a neutral result could mean the metabolic strategy had no useful effect. It could also mean participants could not sustain the diet well enough to test the strategy. Those explanations require different next studies.

Safety also deserves equal billing with mechanism. Advanced cancer can reduce appetite and lean mass. Treatments can worsen nausea, bowel symptoms, and weight loss. A diet that restricts carbohydrate may be workable for some patients, but it can be a poor fit for someone struggling to maintain calories, protein, or body weight. More dietary discipline is not automatically more therapeutic.

Verdict

Randomized evidence does not currently show that a ketogenic diet improves cancer outcomes or survival. The best available oncology trials are early-phase or feasibility-focused, and the available reports do not provide the numerical outcome data needed to quantify benefits or harms. No cancer-specific trial described here establishes an MCT dose that reliably achieves a clinically meaningful ketone target.

Anyone considering keto during cancer treatment should make that decision with their oncology team and an oncology dietitian, with active monitoring of intake, weight, symptoms, and nutritional status. Ketosis may be measurable; a cancer benefit remains unproven.