Vitamin C in Blood Disorders: The EVITA Trial

Table of Contents

1. What EVITA Set Out to Test
2. CCUS and Lower-Risk Myeloid Disease in Plain Language
3. Why Vitamin C? TET2 and the Laboratory Background
4. How the Trial Was Run
5. Vitamin C Blood Levels Before and After
6. The Primary Result: Clone Growth
7. Inflammation Markers, DNA Methylation and Progression
8. Side Effects and Serious Events
9. The Survival Finding and Its Limits
10. Earlier Vitamin C Research and What Remains Unknown
11. Key Research Papers
12. Connections

What EVITA Set Out to Test

EVITA is short for a trial registered on ClinicalTrials.gov as NCT03682029. It was investigator-initiated, meaning it was designed and run by hospital researchers rather than by a manufacturer. The paper describes it as "exploratory": a phase 2 study meant to see whether a larger, definitive trial would be worth running.

The question was narrow and biological. People with CCUS and lower-risk myeloid disease carry a population of blood stem cells with an acquired gene mutation. That population, called a clone, can expand slowly over years. The researchers measured the clone by its variant allele frequency (VAF) — the share of DNA reads in a blood sample that carry the mutation. A rising VAF means the clone is taking up more of the blood system. The registered primary outcome was the change in VAF over the 12 months of treatment, expressed in the paper as a clonal growth rate.

Three ideas made vitamin C a candidate. First, laboratory work showed that vitamin C acts as a helper for TET enzymes, which are among the genes most often mutated in these disorders. Second, the same Copenhagen group had earlier found that many patients with myeloid cancers have low blood vitamin C. Third, oral vitamin C is cheap, widely available and has a long safety record at the dose chosen. The trial asked whether restoring normal vitamin C levels would change the clone's course.

Back to Table of Contents

CCUS and Lower-Risk Myeloid Disease in Plain Language

Blood cells are made in the bone marrow from a small pool of stem cells. With age, some of those stem cells pick up mutations in genes such as TET2, DNMT3A or ASXL1. A stem cell carrying such a mutation can gain a slight growth advantage and leave behind many descendants, all carrying the same mutation. This is clonal hematopoiesis.

When clonal hematopoiesis comes with a persistently low blood count — too few red cells, white cells or platelets, called a cytopenia — but the marrow does not meet the criteria for a cancer, the condition is called clonal cytopenia of undetermined significance, or CCUS. When the marrow shows the abnormal shapes and features of a cancer, the diagnosis becomes a myelodysplastic syndrome (MDS), or one of the overlap disorders called MDS/myeloproliferative neoplasms (MDS/MPN), which include chronic myelomonocytic leukemia (CMML).

EVITA enrolled people at the milder end of this range:

• CCUS;
• low-risk or very-low-risk MDS, scoring 3 or less on the revised International Prognostic Scoring System (IPSS-R);
• MDS/MPN, including CMML-0 and CMML-1.

Every participant had a somatic (acquired) mutation present at a VAF of at least 5%, fewer than 5% immature "blast" cells in the bone marrow (fewer than 10% in CMML), and no current anticancer treatment. In this group the standard approach is often monitoring rather than active treatment, so a low-burden intervention that could slow the disease would matter to many people. Some of these conditions can progress over time to higher-risk MDS or to acute myeloid leukemia (AML), and infections are a frequent problem because of low white-cell counts.

Back to Table of Contents

Why Vitamin C? TET2 and the Laboratory Background

TET2 is an enzyme that edits chemical tags on DNA. DNA in blood stem cells carries methyl groups (5-methylcytosine, or 5-mC) that help switch genes off. TET enzymes convert 5-mC to 5-hydroxymethylcytosine (5-hmC), the first step in removing the tag. That process is part of how a stem cell decides when to stop copying itself and start maturing into a working blood cell. When TET2 is mutated and works at half strength, the methyl tags pile up, and stem cells tend to keep renewing themselves instead of maturing.

TET enzymes need several helpers to work. One of them is vitamin C (ascorbate), which keeps the iron atom at the enzyme's active site in the form it needs. In plain terms, vitamin C keeps the TET machinery running at full speed. That is why researchers asked whether more vitamin C could make up for the weaker output of a single working copy of TET2.

Two mouse studies published in 2017 gave the idea its strongest support:

• Cimmino and colleagues, writing in Cell, showed that restoring TET2 function blocked the abnormal self-renewal of TET2-deficient blood stem cells and slowed leukemia progression in mice, and that high-dose vitamin C could mimic that restoration in the laboratory.
• Agathocleous and colleagues, writing in Nature, found that blood stem cells hold unusually high levels of ascorbate. When mice were made ascorbate-deficient, stem-cell function changed and leukemia developed more readily. EVITA cites this work as background on how low ascorbate could favor leukemia.

Human evidence came next. In 2019 the same Copenhagen group (Gillberg and colleagues) reported that patients with myeloid cancers being treated with the drug azacitidine often had low plasma vitamin C, and that oral supplementation brought levels back to normal and was followed by measurable epigenetic changes in their blood cells. In 2023 a Finnish team (Taira and colleagues) reported that vitamin C increased DNA demethylation in people who carry an inherited TET2 mutation. These studies showed that the biological pathway could be moved in people. They did not show any clinical benefit, which is the question EVITA was designed to start answering.

Back to Table of Contents

How the Trial Was Run

EVITA was a multicenter, double-blind, placebo-controlled, parallel-group phase 2 trial. Neither the participants nor the investigators knew who was taking vitamin C.

Sites and timing

• Three hospitals in Denmark: Rigshospitalet (81 participants), Herlev (15) and Aalborg (5).
• One cancer center in the United States, in Los Angeles (8 participants).
• So 101 of the 109 participants were treated in Denmark.
• Enrollment ran from 1 November 2017 to 28 September 2022. The data cutoff was 27 September 2023.

What participants took

The vitamin C group took oral ascorbic acid, 1,000 mg once a day as two 500-mg capsules, for 12 months. The placebo group took identical-looking capsules. Anyone already taking vitamin C supplements stopped them before joining. After the 12 months of treatment, participants entered long-term follow-up. The dose is reported here as the trial's dose only.

Randomization

Fifty-five people were assigned to vitamin C and 54 to placebo. Allocation began at 2 to 1 in favor of vitamin C and was later changed so that the overall split came out close to 1 to 1.

End points

• Primary: the median clonal growth rate, calculated as the natural logarithm of the VAF ratio per year, from baseline to the end of treatment.
• Secondary and exploratory: plasma vitamin C levels, 19 blood cytokines (signalling proteins of the immune system), global DNA methylation, progression to higher-risk disease, and safety.
• Overall survival: a post hoc exploratory end point. The paper says it was "planned before the end of the study but not included in the protocol".

Size

The paper states that "a formal sample size calculation was not performed". The pragmatic target was at least 50 people per arm.

The two groups were not evenly matched

Randomization in a small trial can leave the groups different by chance, and EVITA's were:

• The placebo group was older (median age 75.0 vs 72.3 years) and had more men (78% vs 65%).
• The placebo group had more people with low counts in more than one blood-cell line, and lower blood counts overall.
• The vitamin C group had a higher mutational burden and worse molecular prognostic scores.
• Median follow-up was 28.0 months in the placebo group and 36.3 months in the vitamin C group.

Three people in each arm were later found not to meet the eligibility criteria. In the placebo arm this included one person with AML; in the vitamin C arm it included one person with CMML-2.

Funding

According to the paper's acknowledgments, the trial was funded by the Van Andel Institute through the VAI–Stand Up To Cancer Epigenetics Dream Team, the Danish Cancer Society and Greater Copenhagen Health Science Partners. The paper states that the funders had "no direct involvement in the trial", while noting that Dream Team members contributed to the trial's concept, enrollment, analysis and revision. Several authors report consulting fees or grants from pharmaceutical and laboratory-supply companies; none of those companies makes vitamin C.

Back to Table of Contents

Vitamin C Blood Levels Before and After

One of the clearest findings concerns how many participants started out low. At baseline, 57% (62 of 109) had plasma vitamin C in the deficient range (below 23 µmol/L) or the inadequate range (23–49 µmol/L). That matches the Copenhagen group's earlier finding of frequent low vitamin C in people with myeloid cancers. More on what low levels mean in general is on the vitamin C deficiency page.

Plasma levels after 12 months (medians, with the interquartile range for the vitamin C group):

• Vitamin C group: 45.85 µmol/L at baseline (29.80–69.12) rose to 81.90 µmol/L at 12 months (67.08–99.57); p < .0001.
• Placebo group: 43.75 µmol/L at baseline and 48.73 µmol/L at 12 months; p = .88, no meaningful change.

So the intervention did what it was meant to do biochemically. A daily 1,000 mg dose moved the typical participant from the borderline range into the normal-to-high range and kept them there. Any lack of effect on the primary outcome therefore cannot be explained by the capsules failing to raise blood levels.

For comparison, oral vitamin C reaches a ceiling. As intake rises, the gut absorbs a smaller fraction and the kidneys excrete more, so oral doses keep plasma levels within a fairly narrow range. That is very different from the much higher concentrations reached by intravenous infusion, which are covered on the IV high-dose vitamin C and cancer page. EVITA tested ordinary oral repletion, not pharmacological dosing.

Back to Table of Contents

The Primary Result: Clone Growth

The primary end point showed no difference between the groups.

• Median clonal growth rate on placebo: 0.046 per year.
• Median clonal growth rate on vitamin C: 0.030 per year.
• Difference: −0.016 (95% confidence interval −0.096 to 0.064); p = .70.
• This analysis included 98 of the 109 participants.

In everyday terms, the clone's VAF rose by a median of about 4.6% a year on placebo and about 3.0% a year on vitamin C. The confidence interval runs from a clearly slower rate to a clearly faster one, so the data cannot tell the two apart. The clones in both groups also grew slowly over a single year, which leaves little room to detect a difference.

The authors drew a broader conclusion from this. They wrote that clonal growth rate "is not a suitable surrogate end point for treatment efficacy in this patient population". In other words, measuring the clone over 12 months may not be a useful way to judge whether a treatment helps these patients. That is itself a finding for future trial designers.

The press release accompanying the paper also reported that vitamin C did not slow abnormal cell growth, which matches the paper.

Back to Table of Contents

Inflammation Markers, DNA Methylation and Progression

Cytokines

The researchers measured 19 cytokines in blood. After correcting for the number of comparisons (a false-discovery-rate, or FDR, correction), 6 of the 19 changed differently between the groups:

• IL-6, IL-10, CXCL10, M-CSF and G-CSF rose more in the placebo group. For IL-6 the FDR-corrected p was .0002.
• RANTES (also called CCL5) rose more in the vitamin C group.

The authors describe the overall direction as consistent with a lower-risk inflammatory profile on vitamin C. IL-6 in particular is a marker of chronic inflammation that is often raised in these disorders. This is a biological signal measured in blood. The trial did not show that the cytokine differences led to a clinical benefit. The press release described them as "favorable inflammatory changes"; the paper's own wording is more cautious.

Global DNA methylation

Given the TET2 background, the most direct test of the theory was DNA methylation. The trial found no difference between groups in total 5-mC (p = .36) or in the ratio of 5-hmC to 5-mC (p = .77). At the level of the whole genome, the epigenetic change that the laboratory work predicted was not seen. Global measures can miss changes at specific genes, so this result does not rule out effects at particular sites.

Progression to higher-risk disease

Nine people on placebo and five on vitamin C progressed to higher-risk disease. In a competing-risks analysis the hazard ratio was 0.51 (95% CI 0.17–1.51), p = .22. That difference is not statistically significant: with so few events, the result is compatible with anything from a large reduction to a moderate increase in risk.

Back to Table of Contents

Side Effects and Serious Events

The safety analysis covered 53 people on placebo and 55 on vitamin C.

• Adverse events: 60% on placebo (32 people, 82 events) vs 44% on vitamin C (24 people, 54 events).
• Serious adverse events: 57% on placebo (30 people, 66 events) vs 33% on vitamin C (18 people, 39 events).
• Stopped treatment because of side effects: 1 on placebo (2%) vs 3 on vitamin C (5%).
• Deaths during the treatment year: 2 on placebo, both related to the disease, vs none on vitamin C.

Individual events

Events that were more common on placebo:

• anemia: 15% vs 7%;
• pneumonia: 15% vs 2% (8 people vs 1);
• acute aseptic arthritis: 6% vs 0;
• internal bleeding: 6% vs 0 (3 people vs none).

Events that were more common on vitamin C:

• diarrhea: 5% vs 2%;
• esophagitis (irritation of the food pipe): 5% vs 0;
• gastrointestinal infections: 4% vs 0.

These are descriptive counts. The paper does not give significance tests for them, and with numbers this small a difference of a few people can arise by chance. The press release listed fewer cases of anemia, pneumonia, aseptic arthritis and internal bleeding; the counts it gave are correct, and the context above is what the paper adds. The placebo group was also older and had lower blood counts at the start, which on its own would be expected to bring more infections and bleeding.

One participant on vitamin C had a kidney stone that needed treatment. That person had had kidney stones more than a year before the trial. Kidney stones are a known concern with long-term vitamin C supplementation because part of the vitamin is broken down to oxalate; the research on that is collected on the vitamin C and kidney stones page. Digestive effects at higher intakes are covered on the digestive upset and diarrhea page.

Back to Table of Contents

The Survival Finding and Its Limits

The result that drew the most attention was overall survival. Over a median follow-up of 33.6 months there were 35 deaths: 24 in the placebo group and 11 in the vitamin C group.

• Median overall survival was 42.2 months on placebo and was not reached on vitamin C (fewer than half had died by the end of follow-up).
• Hazard ratio 0.35 (95% CI 0.17–0.71); log-rank p = .0025.
• After adjusting for age, sex, hemoglobin and whether the diagnosis was CCUS, the hazard ratio was 0.42 (95% CI 0.20–0.88); p = .021.
• The leading cause of death was bacterial infection: 14 of 24 deaths on placebo and 5 of 11 on vitamin C.
• Older age weakened the association.

Why the number needs its context

The death counts are correct, but several features of the trial limit what they can show:

• It was post hoc. Overall survival was added as an exploratory end point; it was not in the trial protocol. Findings from analyses that were not pre-specified are more likely to arise by chance.
• The trial was not built for it. The authors write that EVITA was "neither powered nor designed to assess survival outcomes".
• The groups differed at the start. The placebo group was older, had more men and had lower blood counts — all factors linked to shorter survival. The adjusted analysis takes some of these into account, and the hazard ratio moved from 0.35 to 0.42, but adjustment in a small trial cannot fully remove such imbalances.
• Follow-up was unequal. Median follow-up was 28.0 months on placebo and 36.3 months on vitamin C.
• Treatment stopped at 12 months. What participants did afterwards, including whether they took vitamin C on their own, was not recorded. The authors call this possible "informal crossover".

The authors describe the survival result as "hypothesis-generating" and conclude that "a phase 3 trial is warranted". On the evidence in the paper, the survival difference is a signal worth testing, not a demonstrated benefit. The fact that most of the gap came from deaths due to bacterial infection fits with the cytokine findings and with the known role of vitamin C in white-cell function (see vitamin C and immune function), but the trial was not designed to test that link.

Back to Table of Contents

Earlier Vitamin C Research and What Remains Unknown

The long history of vitamin C and cancer trials

Vitamin C has been tested in cancer before. Two randomized trials from the Mayo Clinic, published in the New England Journal of Medicine in 1979 (Creagan and colleagues) and 1985 (Moertel and colleagues), gave 10 g a day of oral vitamin C to people with advanced cancer and found no benefit over placebo. EVITA cites both. Those trials differ from EVITA in almost every way: advanced solid cancers rather than early blood-cell disorders, a tenfold higher dose, and survival or tumor response rather than clone growth as the measure. The story of the earlier claims that prompted them is told on the Linus Pauling and the Cameron collaboration page.

Intravenous vitamin C in TET2-mutated CCUS

A small US study by Xie and colleagues, published in Blood in 2024, took the opposite approach to dosing. As EVITA's authors describe it, it was a single-arm trial in 10 people with TET2-mutated CCUS who received 1 g per kilogram of body weight intravenously three times a week for 12 weeks. It reported no blood-count responses at 20 weeks and largely stable TET2 VAFs. Taken together with EVITA, neither ordinary oral repletion nor very high intravenous doses has so far been shown to shrink these clones over a few months.

What EVITA adds

• Low vitamin C is common in this group: more than half started deficient or inadequate.
• A 1,000 mg daily oral dose reliably restored normal plasma levels.
• Clone growth over 12 months did not differ, and global DNA methylation did not change.
• Several inflammatory cytokines rose less on vitamin C.
• Serious adverse events were fewer on vitamin C, and an exploratory survival analysis showed fewer deaths, mostly from bacterial infection.

What remains unknown

• Whether the survival difference holds up in a trial designed and powered for survival, with balanced groups. The authors state that a phase 3 trial is warranted.
• Whether any benefit is limited to people who start out deficient, or applies to everyone.
• Whether any effect works through infection risk and immune function, through inflammation, through epigenetics at specific genes, or through something else.
• Whether particular mutations, such as TET2, respond differently. EVITA was too small to answer this.
• How results would apply outside Denmark, since only 8 of the 109 participants were treated in the United States.
• What happens with treatment longer than 12 months, and what participants took after the trial ended.

Vitamin C occurs naturally in whole foods such as bell peppers, guava and oranges; food sources are listed on the vitamin C sources page. EVITA studied capsules at a fixed dose and did not look at diet.

Back to Table of Contents

Key Research Papers

1. Mikkelsen SU, Al-Mousawi A, Puglisi AB, Vallentin AP, Mortensen AØ, Madaj Z, et al. (2026). Oral vitamin C supplementation in patients with clonal cytopenia of undetermined significance or lower-risk myeloid malignancies: results from EVITA, a phase 2 randomized, placebo-controlled trial. Cancer 132(19):e70549 — PubMed PMID: 42764731
2. Cimmino L, Dolgalev I, Wang Y, Yoshimi A, Martin GH, Wang J, et al. (2017). Restoration of TET2 function blocks aberrant self-renewal and leukemia progression. Cell 170(6):1079–1095.e20 — PubMed PMID: 28823558
3. Agathocleous M, Meacham CE, Burgess RJ, Piskounova E, Zhao Z, Crane GM, et al. (2017). Ascorbate regulates haematopoietic stem cell function and leukaemogenesis. Nature 549(7673):476–481 — PubMed PMID: 28825709
4. Gillberg L, Ørskov AD, Nasif A, Ohtani H, Madaj Z, Hansen JW, et al. (2019). Oral vitamin C supplementation to patients with myeloid cancer on azacitidine treatment: normalization of plasma vitamin C induces epigenetic changes. Clinical Epigenetics 11(1):143 — PubMed PMID: 31623675
5. Taira A, Palin K, Kuosmanen A, Välimäki N, Kuittinen O, Kuismin O, et al. (2023). Vitamin C boosts DNA demethylation in TET2 germline mutation carriers. Clinical Epigenetics 15(1):7 — PubMed PMID: 36639817
6. Xie Z, Fernandez J, Lasho T, Finke C, Amundson M, McCullough KB, et al. (2024). High-dose IV ascorbic acid therapy for patients with CCUS with TET2 mutations. Blood 144(23):2456–2461 — PubMed PMID: 39352751
7. Creagan ET, Moertel CG, O'Fallon JR, et al. (1979). Failure of high-dose vitamin C (ascorbic acid) therapy to benefit patients with advanced cancer: a controlled trial. New England Journal of Medicine 301(13):687–690 — PubMed PMID: 384241
8. Moertel CG, Fleming TR, Creagan ET, et al. (1985). High-dose vitamin C versus placebo in the treatment of patients with advanced cancer who have had no prior chemotherapy: a randomized double-blind comparison. New England Journal of Medicine 312(3):137–141 — PubMed PMID: 3880867

PubMed Topic Searches

1. PubMed: vitamin C clonal cytopenia of undetermined significance
2. PubMed: ascorbate TET2 clonal hematopoiesis
3. PubMed: vitamin C myelodysplastic syndrome
4. PubMed: plasma vitamin C deficiency myeloid cancer
5. PubMed: ascorbate DNA demethylation 5-hydroxymethylcytosine

Back to Table of Contents

Connections

• Vitamin C Benefits Hub
• Vitamin C Overview
• IV High-Dose Vitamin C and Cancer — pharmacological dosing, a different question from EVITA's oral repletion
• Vitamin C and Immune Function
• Vitamin C, Iron Absorption and Anemia
• News, 11 October 2026 — the short report on EVITA
• Myelodysplastic Syndrome
• Acute Myeloid Leukemia
• Leukopenia
• Hematology
• Vitamin C Deficiency
• Vitamin C and Kidney Stones
• Vitamin C and Immune Defense
• Vitamin C Sources
• Linus Pauling and the Cameron Collaboration

Back to Table of Contents