Benzoyl Peroxide and Benzene: Latest findings from John Barbieri, MD, MBA, FAAD

John Barbieri, MD, MBA, FAAD, recently spoke with Dermatology Times on his research into benzoyl peroxide containing products and benzene. Barbieri’s dual studies found no increased cancer risk for patients utilizing benzoyl peroxide products.

Transcript:

John Barbieri: Hi, I'm Dr John Barbieri. I'm a board-certified dermatologist at the Brigham and Women's Hospital and Harvard Medical School in Boston, Massachusetts. I'm also the co-chair of the American Academy of Dermatology acne guidelines workgroup.

Dermatology Times: What factors of the initial benzene research inspired your team to conduct these studies?

Barbieri: In March of 2024, Valisure released the citizen’s petition to the FDA raising concerns about the potential for benzoyl peroxide containing products to thermally decompose into benzene, a known human carcinogen, and this isn't an issue of something like contamination. This is really a fundamental aspect of the benzoyl peroxide molecule, this propensity for it to degrade and become benzene, especially when exposed to high temperatures.

Now, this Valisure report has some limitations. It described temperatures at 50 degrees Celsius, 122 degrees Fahrenheit, a little bit of data at 37 degrees Celsius, and very little data at room temperature, which is typically how most people are storing benzoyl peroxide. In addition, this exposure to benzene doesn't necessarily mean risk of outcomes like cancer.

Certainly, we want to minimize benzene to make this close to zero as possible. But unfortunately, there are benzene exposures in our everyday lives too, and the attributable clinical significance of any additional benzene exposure from benzoyl peroxide products, especially if very minuscule, might not be relevant because of these limitations.We wanted to try to look in our research whether or not use of benzoyl peroxide in routine practice, not under these extreme conditions, actually seems to confer meaningful risk to patients of either increase benzene levels in their blood or increased risk of cancer down the road, which is ultimately what we're concerned about.

DT: Can you give us a brief overview of each study?

Barbieri: The first study involved data from the National Health and Nutrition Examination Survey (NHANES). NHANES is a national survey that's done of individuals in the United States, they have information on medications, and they also happen to collect blood benzene levels from some participants in the study. We used this data set to look whether those who are using benzoyl peroxide containing products have an increased likelihood of having any benzene detected in their blood, or higher levels of blood benzene than a set of matched individuals who are not. We tried to find people who are as similar as possible in every way, age, sex, other demographics, income, things like that, and try to make sure that they're not likely to be using benzoyl peroxide to compare them.

What we found is that those who are using benzoyl peroxide containing products compared to those who are not, were not any more likely to have elevated levels of blood benzene, and they were not any more likely, if they did have detectable blood benzene, to have a higher level of benzene. There was no increased rate of blood benzene being detected at all. Among those who did have some benzene detected, the levels are not higher among those who are using benzoyl peroxide products compared to those who are not. This study, although small in scale, provides some context that it seems like those who are using benzoyl peroxide products are not ending up with evidence of increased levels of benzene their blood, and that's how we'd expect that benzene exposure from benzoyl peroxide would be dangerous to people. It's not absorbed through the skin. It's really minimal to no percutaneous absorption.The concern would be that it could become volatile, could go into the air, and that we'd inhale it and get it into our body through our lungs.

The cancer risks associated with benzene are generally leukemias or lymphoma with blood cancer, so blood levels of benzene would be really the relevant kind of benzene exposure that we'd want to be looking for. So again, to summarize this first study: It'srelatively reassuring. It's small in scale. It certainly has that limitation, but we don'tidentify any risks with typical benzoyl peroxide use of increased levels of blood benzene in this population.

In the second study, we used a cohort design. We found a population of individuals who were exposed to benzoyl peroxide, and we compared them to a map set of individuals who are not using the TriNetX database. Again, we tried to make sure these populations were as similar as possible to everything that we could control for with the exception of that benzoyl peroxide exposure. In this study, which had over 50,000 individualswho are included in it, we looked over the next 10 years at whether those who are exposed to benzoyl peroxide were any more likely to develop cancer than those who are not. What we found, and again, this matched cohort design study is that those who are benzoyl peroxide usersdidn't have any increased risk of leukemia, they had no increased risk of lymphoma, and they had no evidence for an increased risk of any solid organ cancers. We didn't detect any evidence for increased risk of any kind of cancer that we looked at. Again, these data are reoccurring that with routine use of benzoyl peroxide by our patients, it doesn't seem to be putting them at risk of cancer, which is the thing we'd worry about with benzene being a known human carcinogen.

DT: What are the limitations of this study?

Barbieri: While reassuring, these studies do have some limitations. The study, using NHANES data, as I mentioned, is limited by small size. Is the study in TriNetX is limited by the fact that it's a database study. There may not be perfect capture of exposures and outcomes, so we tried to use validated measures as much as we could. In addition, both of these studies are limited by something we call misclassification bias, which is that those in the control group who were trying to have not be exposed to benzoyl peroxide might be using benzoyl peroxide.It'sin many over the counter products. Sometimes it's not always clearly part of something. It might be in some kind of an “acne system.”Some individuals who are using benzoyl peroxide might not know it. It might not get captured in these databases.

If there is this type of misclassification, where those in the control group are using benzoyl peroxide, that could bias the study toward finding a reassuring finding. We did take as many steps as we could to try to choose control groups that we thought would be less likely to be using benzoyl peroxide. We kind of exclude anyone with a condition where they'd be using benzoyl peroxide, like acne, hidradenitis, or similar types of things. But nevertheless, there still is the possibility for this kind of misclassification bias, that is a limitation of both studies, and why we need ongoing prospective studies to characterize whether those who are using benzoyl peroxide are at risk for any of these outcomes we'd be concerned about.

DT: What do you think should happen next or continue to happen as dermatologists await action from the FDA?

Barbieri: I look forward to hearing from the FDA about their thoughts on this matter to help guide our practices. But in the setting of current uncertainty, I think there's a few things that we can think about to inform our clinical decision making. First, again, as I shared, there are these 2 studies with reassuring findings that we can use to help support if we do want to use benzoyl peroxide, the likely safety of that practice. In addition, if we just think about this from a kind of a theoretical level, let's say we take the data that Valisurepresented at baseline testing. Again, this data is a little bit incomplete, so we don't totally know the full story, but 85% of their products have less than 10 parts per million of benzene in them detected at baseline. Let's say that we took 1 of those bad products that had 10 parts per million or more, and let's just do a thought experiment. Let's imagine we used a gram of that product containing 10 parts per million of benzene. Well, that would be expected to be about a 10 microgram exposure benzene. Let's assume that it all aerosolized, which may not be the case, and that we could inhale it and be exposed to that meaningful risk.

Just to put that into context, during our daily lives, turning on a gas stove makes about 5 micrograms of benzene each minute. A lot of foods and other exposures that we have in our daily lives have 10s to hundreds of micrograms of benzene exposures in them. While I absolutely love to have zero benzene in our lives, the likely additional risk contribution of benzoyl peroxide products, even if we assume everything's right about the Valisure report and that we take some of the worst products. Even then, it doesn't seem to be a likely, very meaningful clinical risk.The dose makes the poison, and the total absolute levels of benzene being detected are likely low enough that even if they are there, they're unlikely to have a very meaningful clinical effect compared to other exposures in our daily lives.

We want to try and minimize benzene as much as possible. Doesn't mean we don't want to try to try to figure out how to transport, how to store, how to manufacture benzoyl peroxide products that minimize this effect. But just thinking about, should I be worried about it in my life? I think the answer is, is probably not just like when we fly on an airplane, we get exposed to X-ray radiation, but that level of exposure is not that clinically meaningful. Would I like it to be zero? Absolutely. But am I worried about flying on planes that I'm going to get cancer from the radiation? Not really, compared to other things in my daily life.I think that's important thing: We have to keep in context too.

We also have to think about, there's no substitute for benzoyl peroxide. There's no, “Oh, I've just used something else instead.”It's the only product we have that really has those characteristics. It's really the only thing we've rigorously studied that can prevent antibiotic resistance to topical and oral antibiotics. So there's not an easy,“Let's just flip it out for something else” to do. If we can't use it to help care for people with acne, well we're probably going to have to use something else to control their skin to the way they want it to be, and that might be now oral medications like spironolactone or oral antibiotics or isotretinoin.Those have their own risks, so we have to consider that too. If we're not going to use benzoyl peroxide, what are we going to use instead? Might that have its own risks that have to be considered in this discussion as well? 

I think there's certainly some uncertainty here. We have to be thoughtful about it.We have to have an informed discussion with their patients about the safety of benzoyl peroxide containing products, how to store them appropriately, and whether or not we want to use them. But I do think now we have some emerging data from our research group that helps provide some reassurance for those who want to continue using benzoyl peroxide containing products. I look forward to hearing more from the FDA and others who are conducting testing in the space. I think if we just think about this in a logical manner, the likely risk versus benefit of using benzoyl peroxide is favorable from a benefit standpoint, that it is something that we can continue to use safely with our patients.

DT: How do you address questions from patients regarding benzoyl peroxide and benzene?

Barbieri: Sometimes patients will bring up this question about benzene and benzoyl peroxide containing products. I've actually been surprised how infrequently it's brought spontaneously. I often will discuss it as just part of counseling and talking about how to store things appropriately. I do think that in a setting of uncertainty, there should be shared decision making. We should talk about risks and benefits and alternatives, while we don't have a clear answer to this question, but I do think in general, the benefits seem to be favorable compared to these risks, based on the clinical data we have.

[Transcript has been edited for clarity.]