NICOTINAMIDE FOR SKIN CANCER PREVENTION: WHAT DO THE NUMBERS SAY?

A vitamin with a big preventive assignment

Nicotinamide, also known as niacinamide, is a form of vitamin B3 used to correct deficiency, including pellagra. Why did it become the focus of a debate about skin cancer prevention? First, it is a precursor in the formation of nicotinamide adenine dinucleotide (NAD+), a molecule tied to cellular energy metabolism. Second, a phase III randomized trial, ONTRAC (2015), showed a preventive effect in high-risk patients with a mean of eight previous skin cancers [1]. A later trial, ONTRANS (2023), did not confirm the result in solid organ transplant recipients [2], that is, in patients who have received an internal organ such as a kidney, liver, or heart and who take immunosuppressive drugs afterward to prevent rejection. The question of whether nicotinamide prevents skin cancer therefore remains open.

Biological plausibility and a single positive trial that was not reproduced in another population do not settle the question of efficacy. Tan and Williams (2026) address this contradiction in an article in the American Journal of Clinical Dermatology. It is an expert critical appraisal rather than a new clinical trial: the results of a large observational study are set against earlier trials and meta-analyses. The central question is whether the accumulated evidence is sufficient for widespread preventive use [3].

The discussion concerns oral use to prevent keratinocyte carcinoma, that is, basal cell and squamous cell carcinoma of the skin. These results do not apply to melanoma prevention or to cosmetic products containing niacinamide. Correcting a vitamin deficiency should not be confused with preventing tumors either: these are separate goals that require separate evidence.

Availability is not evidence of benefit

In the United States, nicotinamide is sold over the counter as a dietary supplement. That availability makes the question of preventive use a practical one: a patient can start taking it without waiting for a consultation. But being able to buy a product does not mean its ability to prevent skin cancer has been convincingly established.

It helps to separate a familiar substance from a new proposed use. Using nicotinamide to treat vitamin B3 deficiency does not prove that it works in people without a deficiency who hope to avoid another tumor. A discussion of prevention should therefore begin not with how easy the supplement is to buy, but with what results have been obtained, in which patients, and how much confidence they deserve.

Almost 34,000 patients: convincing at first glance

The latest round of debate was set off by a retrospective study by Breglio and colleagues (2025) that included 33,822 patients with a history of skin cancer in the US Department of Veterans Affairs health care system. The investigators compared patients who received nicotinamide 500 mg twice daily for at least 30 days with patients who had no recorded use, and assessed the time to a subsequent tumor. This describes the use that was studied; it is not a ready-made prescribing regimen [4].

Use was associated with a 14% lower overall risk of subsequent skin cancer and a 22% lower risk of squamous cell carcinoma. The association was strongest when treatment began after the first tumor, with a reduction of about 50%; with later initiation, it weakened. Among patients with a history of organ transplantation, no overall benefit was found.

That sounds impressive, but these are relative measures. They do not mean that a tumor was prevented in 14 or 22 out of every 100 people, and on their own they do not show the absolute difference in risk. More importantly, an observed association does not prove that nicotinamide itself caused the difference.

Who was protecting the skin: nicotinamide or the patient?

In a randomized trial, participants are assigned to groups by chance. In ordinary practice, patients decide whether to take a supplement based on their conditions, behavior, and physician advice. Groups can therefore differ before the first tablet is taken.

In this study, the groups were matched on several known characteristics. However, several factors were not accounted for: skin phototype, cumulative ultraviolet exposure, family history, the burden of actinic keratoses, the intensity of dermatologic surveillance, and the degree of immunosuppression. Each can influence both the development of tumors and the likelihood that they are detected. A patient taking nicotinamide, for instance, may have been more careful about sun protection, while more frequent examinations could have increased the number of tumors found. These are possible explanations, not established differences between the groups. Still, without the relevant data, it is hard to separate the substance's effect from the circumstances that accompany it.

The way use was recorded also creates uncertainty: filling a prescription does not confirm regular consumption, and the absence of a prescription in an electronic record does not rule out an over-the-counter purchase. Availability thus becomes an additional source of uncertainty, because some patients may have been placed in a group that did not reflect what they actually took.

Ninety days that can change the result

Tumors diagnosed or treated within the first 90 days after baseline were excluded from the analysis. The rule was applied to both groups, but applying it symmetrically does not by itself eliminate temporal bias. If disproportionately many early events were removed from the nicotinamide group, the protective effect could look stronger; if they were removed from the control group, it could look weaker. The study has not established which scenario occurred, and it remains unclear how much time-handling could have influenced the final estimates.

In addition, the authors identified tumor outcomes from diagnostic and procedure codes rather than histologic confirmation for each case. Questions also remain about whether the results hold up under alternative analytic approaches. Using the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) version 2 tool, the overall risk of bias was judged serious. A large sample makes statistical estimates more precise, but it does not remove systematic bias.

What the trials say when treatment is decided by chance

That is precisely why observational data need to be read alongside randomized trials. As noted, ONTRAC found a preventive effect in high-risk patients [1], whereas ONTRANS found no comparable benefit in patients with a history of organ transplantation [2]; recruitment difficulties and early termination also limited ONTRANS. This is not a case where one trial cancels another, since the populations differed and both trials' limitations must be considered. But there is also no basis for extending a favorable result to all high-risk patients.

Meta-analyses do not settle the question either. A 2022 review pooled 29 trials, yet data for the primary outcome came from only five studies with 552 participants, and three of those five had not been designed to assess skin cancer in the first place. The number of trials included therefore creates a more impressive impression than the volume of data that directly addresses the preventive question.

In a meta-analysis by Tosti and colleagues (2023) that combined four randomized controlled trials, the relative risk was 0.82 (95% confidence interval 0.61–1.12) [5]. The point estimate favors nicotinamide, but the interval includes 1, the value of no difference. A statistically significant effect has not been established: this is not proof that the supplement is useless, but neither is it reliable confirmation of preventive efficacy.

From study to clinic: no promises beyond the evidence

How closely do the participants in the largest study resemble patients in an aesthetic practice? In the veterans cohort, 98% were men, about 95% were white, and the mean age was about 77 years; all had a history of skin cancer. The results cannot simply be transferred to younger patients, to women, or to people who have never had a tumor.

For the practitioner, the practical value of this debate lies in the precision of the consultation. When a patient asks about nicotinamide, the answer should reflect the purpose of use: correcting a vitamin deficiency and trying to prevent a new tumor are not the same thing. Data on oral use also do not justify claims that a cosmetic cream prevents cancer.

Safety is a separate question. The article analyzed here does not provide a detailed quantitative assessment of adverse events that would allow the risks of long-term preventive use to be judged. Neither the vitamin nature of the substance nor its over-the-counter status is a substitute for such an assessment.

The necessary next step is a large, multicenter, double-masked, placebo-controlled randomized trial with histologic confirmation of tumors and objective measurement of adherence. Such a trial should help separate the effect of nicotinamide from patient characteristics and from limitations in how the data are recorded.

Biological rationale and encouraging results explain the interest in nicotinamide, but they do not resolve the uncertainty. A positive association is not the same as an established preventive effect, and criticism of a study does not prove that there is no benefit. The practitioner's task is to hold on to both parts of that conclusion, and to explain to patients not only the attractive percentages but also the degree of confidence behind them.