Kirsten Grønbæk’s first suspicion was that her own randomization had gone wrong. Her team in Copenhagen had assigned 109 people with pre-leukemic blood disorders to a daily vitamin C capsule or a placebo, kept patients and clinicians blind for a year, and measured what the trial was built to measure, which was whether the mutant blood clones grew more slowly. They didn’t, not by any margin that counted. Yet by the end of follow-up, 24 patients on placebo had died and 11 on vitamin C. “This has been very puzzling to us,” she told CancerNetwork. “We thought for a long time there must be differences between the groups.”
| Group | Value (deaths) |
|---|---|
| Placebo | 24 |
| Vitamin C | 11 |
The trial is called EVITA, and it was published this fall in the American Cancer Society’s journal Cancer. The investigators were the first to point out that the survival result is an exploratory, post hoc analysis of a phase 2 study that missed its primary endpoint. They also put the numbers on the table: at a median follow-up of 33.6 months, the unadjusted hazard ratio for death on vitamin C was 0.35, with a 95 percent confidence interval of 0.17 to 0.71, and after adjusting for age, sex, baseline hemoglobin and diagnosis it was 0.42, with an interval of 0.20 to 0.88, according to Oncology Nursing News. The adjustment trimmed the signal without erasing it. The dose was 1,000 mg a day, two 500 mg capsules, the kind of amount anyone can buy without a prescription.
| Result | Value (HR) | Lower bound | Upper bound |
|---|---|---|---|
| Unadjusted | 0.35 | 0.17 | 0.71 |
| Adjusted | 0.42 | 0.20 | 0.88 |
Nothing to do but watch
The patients had clonal cytopenia of undetermined significance (CCUS), lower-risk myelodysplastic syndrome (MDS), or a related overlap disorder. In each, a mutated line of blood stem cells has claimed part of the marrow and blood counts have started to sag, but the disease has not become leukemia. Casey O’Connell, who ran the trial’s site at USC’s Keck Hospital, described what that diagnosis usually earns a patient: “The standard approach for patients with pre-cancer blood disorders is to monitor their blood cell counts over time.” He was blunter about what it feels like from the other side of the desk, saying that for these patients “there’s nothing that can be done other than watch and wait.”
EVITA enrolled from 2017 to 2022 at Rigshospitalet, Herlev and Aalborg in Denmark and at Keck in Los Angeles, taking patients with a measurable mutant clone and no anticancer therapy in the prior six months. The rationale traces to an enzyme called TET2, which restrains blood stem cells from self-renewing out of control, is among the genes most often mutated in these disorders, and depends on vitamin C to function. In 2017 a group led by Iannis Aifantis reported in Cell that vitamin C mimics TET2 restoration in Tet2-deficient mouse cells and suppresses leukemia progression in mice carrying human leukemia samples. Peter Jones of the Van Andel Institute, a co-corresponding author on EVITA, told Healio that “the first evidence that people with leukemia often are vitamin C deficient was discovered about 70 years ago,” knowledge he said never made it into clinical care.
The deficiency showed up on schedule. At baseline, 62 of the 109 patients, 57 percent, had deficient or inadequate vitamin C levels. Over the treatment year, plasma levels in the vitamin arm climbed from 45.85 to 81.90 µmol/L, while the placebo arm drifted from 43.75 to 48.73.
| Event | Placebo | Vitamin C |
|---|---|---|
| Serious adverse events | 57% | 33% |
| Grade 3 or worse | 53% | 33% |
| Pneumonia | 15% | 2% |
Medicine has a long memory on this vitamin, and not a fond one. In 1985 Charles Moertel and colleagues at the Mayo Clinic reported in the New England Journal of Medicine that 10 grams a day did nothing for 100 patients with advanced colorectal cancer: no delay in progression, no gain in survival. It was Mayo’s second test of claims championed by Linus Pauling, and the verdict hardened into consensus. But it concerned terminal solid tumors. EVITA asked about early blood disease, and the Van Andel Institute’s announcement notes that in cancer models the strongest effects appeared when vitamin C was given before cancer had become established. The institute also describes EVITA as the first randomized, double-blind, placebo-controlled trial of oral vitamin C in these conditions. Between Moertel’s paper and this one, by that account, nobody ran the experiment at all.
The clones kept growing
On the question it was designed to answer, EVITA came up empty. The between-arm difference in annual clonal growth was a negligible 0.016, with P = .70, and progression to higher-risk disease ran 9 cases on placebo against 5 on vitamin C, a gap that was not statistically significant. John Riches of Barts Cancer Institute, who was not involved, summarized it without ceremony: “Vitamin C did not significantly slow the growth of the abnormal blood-cell clones.” The vitamin did not move the mutation. The placebo arm, meanwhile, kept landing in the hospital.
Serious adverse events struck 57 percent of the placebo safety population (30 of 53; one randomized placebo patient fell outside the safety analysis) and 33 percent of the vitamin C arm (18 of 55). Grade 3 or worse events split 53 percent to 33. Pneumonia hit 15 percent of placebo patients and 2 percent of those on vitamin C. Bacterial infection was the leading cause of death in both arms, accounting for 58 percent of deaths on placebo and 45 percent on vitamin C, and median overall survival was 42.2 months on placebo while it had not yet been reached on the vitamin.
What connects those numbers is not settled. Six of 19 inflammatory markers diverged between the arms after correction for multiple testing, and one plausible reading is that repleting frail, often deficient patients helped them weather infections. This trial cannot prove that, and its authors do not claim it has. Grønbæk’s next move suggests where she thinks the answer lies: her proposed phase 3 will use all-cause mortality as its primary endpoint, because what kills these patients, she told CancerNetwork, is “also infections and other things.”
The capsule had costs of its own. Diarrhea, esophagitis and gastrointestinal infections were somewhat more common on vitamin C, three vitamin C patients stopped for adverse events against one on placebo, and one patient on the vitamin needed intervention for kidney stones. That patient’s previous stone episode was more than a year before enrollment, outside the trial’s exclusion window for stones within the past year.
The loudest caution comes from inside
Nobody is warning harder against overreading EVITA than the woman who led it. “We don’t have evidence that this will help you. We don’t have evidence that this will be better for your disease,” Grønbæk told Oncology Nursing News, adding that the danger comes when doctors are not aware what their patients are taking. She also drew a hard line around intravenous megadoses: a separate study giving 1 g/kg IV three times a week to 10 patients with TET2-mutated CCUS produced no hematologic responses at 20 weeks. Her peer reviewers, she said, worried that the paper “could bring the wrong message, too early with a phase 2 trial.” Jones, who calls himself “cautious but excited,” put the other half on record: “I don’t think ever in our wildest dreams we expected such a strong signal in terms of overall survival.”
The money behind the result is worth reading too. EVITA was paid for by the Van Andel Institute through its Stand Up To Cancer epigenetics team, the Danish Cancer Society and Greater Copenhagen Health Science Partners, and no drug company appears among the funders. The one commercial tie in the coverage is Jones’s paid consulting and advisory role with Zymo Research, disclosed to Healio. None of the published accounts names a sponsor or a timeline for the phase 3. It is hard to picture a patented molecule posting an exploratory hazard ratio of 0.35 in a population whose standard care is a blood draw and a follow-up appointment, and then waiting on charities to find out whether it holds. Ascorbic acid has no patent and no shareholder, so the confirmatory trial will be funded by whoever decides the question is worth answering without a payday attached.
Grønbæk’s group has been on this since a 2016 pilot that showed oral dosing could raise blood levels. Ten years on, the sentence she offers patients is still “We don’t have evidence that this will help you,” and the trial she has proposed is the one that could make her stop saying it. No one has yet announced who will pay for it.
Sources
- Mikkelsen SU et al., EVITA phase 2 randomized, placebo-controlled trial, Cancer (2026)
- Oncology Nursing News: Oral vitamin C misses primary end point in CCUS but shows survival signal
- Healio: Vitamin C may benefit people with preleukemia blood disorders
- CancerNetwork: What does phase 2 data reveal about oral vitamin C in lower-risk MDS?
- ScienceDaily / Van Andel Institute: Scientists discover a surprising link between vitamin C and cancer survival
- News-Medical: Daily vitamin C may benefit pre-cancerous blood disorder patients
- Nutrition Insight: Daily vitamin C and pre-leukemia trial outcomes
- Cimmino et al., Restoration of TET2 function blocks aberrant self-renewal and leukemia progression, Cell (2017)
- Moertel CG et al., High-dose vitamin C versus placebo in advanced cancer, New England Journal of Medicine (1985)