Randomised controlled trialindependent cross-checkfull text checked★ key study I-INTA-08 in-depth review
Can short fasting periods help chemotherapy work?
Fasting mimicking diet as an adjunct to neoadjuvant chemotherapy for breast cancer in the multicentre randomized phase 2 DIRECT trial.
Overview
de Groot S et al.. Nature communications 2020 · PMID 32576828 · DOI 10.1038/s41467-020-16138-3
In 131 women with breast cancer, a fasting-mimicking diet around chemotherapy did not reduce severe side effects or increase complete tumour disappearance (the main outcomes); tumour shrinkage on imaging was more frequent, but this was a secondary result.
Does the study support the specific claim? Partly · Methodological quality: moderate
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Fasting-mimicking diet with chemotherapy
- How was the study done?
Randomised phase 2 trial
- How many people, animals or samples were studied?
131 women, HER2-negative breast cancer
- What was actually found?
More tumour responses with FMD; no toxicity difference
- How large was the observed effect?
pCR 10.8% vs 12.7% (no difference); radiological response OR 3.17 (1.06-9.45); pathological response OR 4.11 (per-protocol)
- What do the results mean in everyday life?
Fasting may become a supportive tool during chemotherapy, under supervision
- What are the study's limitations?
Phase 2; main outcomes not improved; benefit only in secondary and per-protocol analyses; poor adherence; lead author holds equity in the diet's manufacturer
- What does it say about Elawady's specific claim?
Supports research on fasting as an add-on, not fasting instead of treatment
- Do the results apply to Germany and Europe?
High
- What can explicitly not be concluded from this study?
That fasting alone treats cancer
Scientific in-depth review
1. Full reference
- Authors
- de Groot S et al.
- Journal
- Nature communications
- Year
- 2020
- Identifier
- PMID 32576828 DOI 10.1038/s41467-020-16138-3
- Verification
- verified – PubMed efetch 2026-10-25
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
Funded by Pink Ribbon (2012.WO31.C155) and Amgen (20139098). Competing interests: V.D. Longo has equity interest in L-Nutra (producer of the FMD); H. Pijl has shares in a company that invested in L-Nutra; other authors none.
2. Research question
Does a fasting-mimicking diet (FMD) around neoadjuvant chemotherapy reduce toxicity and improve response in breast cancer?
3. Study design and population
- Study design
- Randomised controlled trial – Multicentre, open-label (observer-blind pathology), randomised phase 2 trial (planned phase II/III; phase III abandoned on DSMB advice), DIRECT, NCT02126449
- Population
Women >=18 years with HER2-negative stage II/III breast cancer receiving neoadjuvant AC-T or FEC-T chemotherapy at 11 Dutch centres; BMI >19 kg/m2, no diabetes, WHO 0-2, adequate organ function; median age 49-51 years, median BMI about 26
- Country
Netherlands (11 centres)
- Period
Randomisation February 2014 to January 2018
- Sample size
131 randomised; 129 analysed (65 FMD, 64 regular diet)
- Inclusion and exclusion criteria
Histologically confirmed HER2-negative stage II/III breast cancer; adequate bone marrow, liver, renal and cardiac function; BMI >19; no diabetes; no allergy to FMD components. 1 withdrawal, 1 ineligible (liver metastases).
4. Methods and endpoints
- Exposure / intervention
4-day plant-based low-protein fasting mimicking diet (Xentigen; soups, broths, tea; ~1200 kcal day 1, ~200 kcal days 2-4) for 3 days before and on the day of each chemotherapy cycle; dexamethasone omitted during AC/FEC cycles in the FMD arm
- Comparator
Regular diet with standard dexamethasone premedication
- Endpoints
Primary (phase II): grade III/IV toxicity (CTCAE 4.03); planned primary phase III: pathological complete response (pCR). Secondary: radiological response (RECIST 1.1), Miller and Payne pathological response, QoL (EORTC QLQ-C30), distress, metabolic markers, gamma-H2AX DNA damage in PBMCs (side study)
- Statistical methods
Power: 128 patients to detect 50% reduction in grade III/IV toxicity (80% power, alpha 3.06%). Logistic regression (univariate and adjusted for stratification factors) for response; ITT and per-protocol (FMD compliant >=half of cycles vs compliant controls); Armitage trend test; mixed models for QoL.
5. Actual results
- Results
Grade III/IV toxicity not different (ITT text: 75.4% FMD vs 65.6% control; Table 2: 47.7% vs 56.3%, P=0.331); chemotherapy discontinuation 27.7% vs 23.8% (P=0.580). pCR 10.8% vs 12.7% (OR 0.830, 0.282-2.442, P=0.735). Radiological complete/partial response ITT: OR 2.886 (1.012-8.227, P=0.047) univariate; OR 3.168 (1.062-9.446, P=0.039) adjusted; stable/progressive disease 11.3% vs 26.9%. Per-protocol Miller-Payne 4/5: OR 3.194 (1.115-9.152) univariate, 4.109 (1.297-13.02, P=0.016) adjusted. QoL not different (P=0.841). Less chemotherapy-induced gamma-H2AX increase in T-lymphocytes with FMD (P=0.045). Compliance: 81.5% completed first FMD cycle, 33.8% >=4 cycles, 20.0% all cycles; 7.8% of controls fasted on their own.
- Absolute risks
pCR 10.8% (FMD) vs 12.7% (control); radiological stable or progressive disease 11.3% vs 26.9%; grade III/IV toxicity reported inconsistently (75.4% vs 65.6% in text; 47.7% vs 56.3% in Table 2).
- Authors' conclusion
FMD cycles appear safe and effective as an adjunct to neoadjuvant chemotherapy in women with early breast cancer; findings encourage further exploration of fasting/FMD in cancer therapy.
6. Strengths
- Randomised, centrally allocated with stratified block randomisation
- Blinded central pathology review
- ITT and PP analyses reported; metabolic markers document diet effect
- Clinically relevant population and standard chemotherapy
7. Weaknesses
- Primary endpoint (grade III/IV toxicity) not met; pCR (planned phase III primary) not different
- Positive findings are secondary endpoints with P values near 0.05 and wide CIs; pathological benefit only in per-protocol analysis (selection bias acknowledged by authors)
- Open label; co-intervention differs by design (dexamethasone omitted in FMD arm)
- Poor adherence (20% all cycles); 7.8% controls fasted
- Phase III abandoned because of low pCR rate and poor compliance
- Imbalance: triple-negative tumours twice as frequent in FMD arm
- Inconsistent toxicity figures between text and table
- Financial conflict: senior author holds equity in the FMD manufacturer; industry co-funding (Amgen)
- Survival not yet reported; excluded underweight patients and diabetes
8. Risk of bias
- Orienting tool
- RoB 2
- Type of assessment
- structured assessment based on the full text by a single reviewer (no formal double review)
- Full formal assessment
- no
- Overall risk of bias
- high
- Rationale
Raised from some_concerns to high after full-text review (null primary endpoints, emphasis on secondary/per-protocol results, design-level co-intervention difference). Financial conflict of the senior author with the FMD manufacturer noted (not a RoB 2 domain). Simplified single-reviewer judgement; a formal RoB 2 assessment needs two independent reviewers and the trial protocol/SAP.
| Domain | Judgement | Rationale |
|---|---|---|
| D1 randomisation process | low | Central web-based block randomisation with varying block sizes, stratified by stage, ER, BMI, regimen; baseline mostly balanced (triple-negative imbalance by chance). |
| D2 deviations from intended interventions | some concerns | Open label; poor adherence to FMD (20% all cycles) and some control fasting; protocol-mandated dexamethasone omission in FMD arm means the comparison is FMD-without-dexamethasone vs diet-with-dexamethasone. ITT used for effect of assignment. |
| D3 missing outcome data | low | 129/131 analysed for main outcomes; QoL data incomplete (57 and 54). |
| D4 measurement of the outcome | some concerns | Pathology centrally reviewed blind to allocation; blinding of radiological assessment and of toxicity grading by treating physicians not stated (open label). |
| D5 selection of the reported result | high | Primary toxicity endpoint and pCR null; emphasis on secondary radiological response and per-protocol Miller-Payne results, multiple analyses (univariate/multivariate, ITT/PP, trend); internal inconsistency in toxicity numbers. |
Full-text review (25 October 2026): Added period, eligibility, FMD composition, primary endpoint (toxicity, not met) and pCR (null), CIs, adherence, early termination of phase III, dexamethasone co-intervention difference, toxicity inconsistency, funding (Pink Ribbon, Amgen) and COI (equity in L-Nutra). RoB 2 overall raised from some_concerns to high; lay summaries corrected to state null main outcomes.
9. Assessment of the statistical interpretation
No effect on primary toxicity or pCR. Secondary radiological response OR 3.17 with CI 1.06-9.45 (imprecise, P=0.039); per-protocol pathological response OR 4.11 susceptible to selection bias. Results are hypothesis-generating.
10. Relevance to the specific claim
The best randomised evidence on fasting-type diets in cancer; tested only together with chemotherapy in well-nourished patients.
11. Independent contrary findings
Systematic review (I-INTA-10) found no demonstrated impact on treatment outcomes overall.
12. Transferability to Germany and Europe
High.
13. Overall methodological judgement
Phase 2 trial with null primary endpoints and imprecise secondary signals; supports further research on supervised FMD as an add-on to chemotherapy, not fasting as a cancer treatment or as a replacement for therapy.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | moderate |
| Risk of bias | high |
| Precision of results | low |
| Directness | indirect |
| Reproducibility | not replicated |
| Consistency with the overall evidence | unknown |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Phase 3 confirmation; 5-year survival (announced); whether benefits persist without dexamethasone difference; safety in underweight or diabetic patients.
15. Linked fact checks
- In his practice he treats cancer with fasting (fasting every other day) and by reducing solid foods; he himself can go for days without food, on sugar-cane juice only. – Studies contradicting or limiting the claim