People with pre-leukemia blood disorders who took 1,000 mg of oral vitamin C a day for a year were significantly less likely to die in the years that followed than people on placebo, according to the phase 2 EVITA trial published September 21 in CANCER, the American Cancer Society journal. The same trial failed the goal it was built to test: vitamin C did not slow the growth of the abnormal blood cells that mark the road toward acute myeloid leukemia.
The survival difference came out of an exploratory analysis, not the question the trial set out to answer. If you have myelodysplastic syndrome, or a low blood count no one has explained, that makes this a trial to raise with your hematologist, not a reason to start a supplement on your own. It does not tell a healthy person to start 1,000 mg a day to hold off cancer.
Vitamin C has been aimed at cancer before, usually through an IV and at solid tumors. Whether those IV doses add anything to chemotherapy is a separate question, and one phase 1 trial tested it in pancreatic cancer.
Key takeaways
- EVITA randomized 109 adults in Denmark and the US to 1,000 mg of oral vitamin C daily or an identical placebo for 12 months.
- Vitamin C did not change the growth rate of abnormal blood cells, the endpoint the trial was designed to test.
- Survival favored vitamin C in an exploratory analysis, meaning that question was asked after the data came in, not before.
What the study found
The trial enrolled adults with clonal cytopenia of undetermined significance, known as CCUS, or with lower-risk myeloid cancers. Both can turn into acute myeloid leukemia, and neither has an approved therapy to stop that. Enrollment ran from November 2017 to September 2022, with 55 people on vitamin C and 54 on placebo.
The primary endpoint was the median clonal growth rate, a measure of how fast the abnormal cells expand. The groups did not separate (difference of -0.016, p = .70). A p-value that high means roughly a 7 in 10 chance of seeing a gap this size from luck alone, the statistical way of saying nothing happened.
Serious adverse events did separate. They hit 33% of the vitamin C group and 57% of the placebo group, about 24 fewer people per 100. Put another way, in this trial about 4 people had to take vitamin C for one of them to avoid a serious adverse event.
Then the exploratory finding. Overall survival was significantly longer with vitamin C, about a 65% lower risk of death (hazard ratio 0.35). The range around that estimate is wide but stays on the benefit side, very likely between 29% and 83% lower (95% CI 0.17 to 0.71), with about a 1 in 400 chance of a gap this large appearing by coincidence (p = .0025).
Dr. Kumar’s take
A survival result pulled from an exploratory analysis, in a 109-person trial that missed its primary endpoint, is exactly the setup that produces the headline “vitamin C fights cancer” and then evaporates in phase 3. The endpoint picked in advance came back flat. The survival number was found afterward. Those two facts belong in the same sentence every time this study gets described.
The less-reported detail is the more useful one. Over half these patients were outright deficient before they started. Correcting a real deficiency in sick older adults is a different story than a supplement with anticancer powers, and it points somewhere much narrower. Kirsten Grønbæk, who led the trial, put it in softer language: “It is too early to make definitive recommendations based on our results.”
The safety signal is the part I would not dismiss. Fewer serious adverse events is a real finding in a randomized comparison.
What it means for you
If you have CCUS, MDS, or a persistent unexplained low blood count, the move is a conversation, not a supplement run. Ask whether your vitamin C level has ever been checked. Deficiency is measurable, it is cheap to fix, and more than half of this trial’s participants had it without knowing.
This trial enrolled adults with CCUS or lower-risk myeloid cancers. It says nothing about vitamin C in anyone outside that group, and within it, vitamin C did not slow the growth of the abnormal cells.
How long can a promising early signal sit before anyone knows if it was real? Fifty years of trials on one cholesterol drug class answer that. A phase 3 trial is the next step, and until it reports, this is a hypothesis, not a treatment.