Evidence Explorer · MS-N290 · Nutrition and dietary patterns · Appraised November 10, 2026
A small randomized phase II screening trial found numerical progression-free and overall-survival differences favoring a medically supervised ketogenic diet, but the confidence intervals were wide and neither result met conventional two-sided statistical significance.
Evidence at a glance
- Population: 36 randomized and 32 evaluable people with previously untreated metastatic pancreatic ductal adenocarcinoma; 16 evaluable participants per arm.
- Intervention: gemcitabine, nab-paclitaxel and cisplatin with either a remotely supervised ketogenic diet or usual diet.
- Regimen: chemotherapy on days 1 and 8 of 21-day cycles; daily ketone goal 0.5–3.0 mM in the diet arm.
- Primary result: median progression-free survival was 8.5 versus 6.2 months (HR 0.53, 95% CI 0.21–1.37; one-sided P=.096). Median overall survival was 13.7 versus 10.2 months (HR 0.58, 95% CI 0.25–1.37).
- Evidence label: hypothesis-generating randomized signal; larger confirmatory trials are required.
- Evidence label: Small randomized signal; not definitive.
What was studied
This was explicitly a screening study using a relaxed one-sided alpha of .20. Fifteen of 16 diet participants reached nutritional ketosis at least once, but the median proportion of monitored days in ketosis was 39.4%; ketone measurements were submitted on a mean 58.5% of treatment days.
The study therefore shows that a monitored program can be attempted alongside this specific triplet regimen. It does not establish that an unsupervised ketogenic diet improves pancreatic-cancer survival, or that the result applies to FOLFIRINOX or gemcitabine/nab-paclitaxel without cisplatin.
What this evidence cannot establish
The sample was too small for a definitive survival conclusion, radiology review was not blinded, baseline imbalances were possible, adherence and ketone monitoring were incomplete, and the relaxed statistical threshold raises the chance of a false-positive screening signal. Microbiome changes are exploratory and cannot establish tumor control.
Safety and practical context
Diet-related events were grade 1–2 and included fatigue, constipation, weight loss, decreased appetite, dehydration, dizziness, nausea and body ache. In pancreatic cancer, where malnutrition, pancreatic insufficiency and cachexia are common, restriction can be harmful without oncology-dietitian oversight. Weight, intake, hydration, electrolytes and enzyme needs matter more than achieving ketosis.
Questions for the care team
- Is my weight and muscle trajectory stable enough for any restrictive diet?
- Would the diet be supervised by an oncology dietitian familiar with pancreatic insufficiency and chemotherapy?
- How would we interpret this small screening signal without assuming survival benefit?
Sources and identifiers
- Randomized phase II trial — PMID 41817106; NCT04631445
- Open full text — PMCID PMC12980701
- ClinicalTrials.gov record — NCT04631445
- ESPEN practical guideline
Funding and conflicts: The trial used a remote-care platform developed by Virta Health. Full sponsor and investigator financial disclosures are available in the paper; they were not independently re-adjudicated for this record.
Source access and dates: Primary reports, registries and official guidance available through the links above were checked November 10, 2026. Last literature search November 10, 2026. Fields unavailable from accessible full text are labeled not reported or not verified rather than assumed absent.
Educational evidence summary, not individualized medical advice. Restrictive diets or supplements should not replace established cancer treatment or adequate nutrition. Symptom, weight, biomarker, laboratory, imaging, pathology or animal findings do not by themselves establish tumor control, recurrence prevention or longer survival.
