Back in 2019, data from two noteworthy trials on the impact of taking antioxidant and vitamin supplements during cancer treatment were published. Their results are important to us. They provide additional data that may help inform our thinking. Unfortunately, they contradict our long-held belief that such supplements are safe to use. Our reaction is to immediately criticize the research looking to find fault in the methodology.

The question of whether antioxidant supplements will aid or hinder cancer treatment has become highly divisive over the years with sharply contrasting opinions and little room for a middle ground. The standard medical wisdom adhered to by a majority of oncologists is that because many medical interventions, especially radiotherapy and traditional forms of chemotherapy, work by increasing reactive oxygen levels in tumor cells and eventually triggering apoptosis, that taking antioxidants will quench these free radicals neutralizing the expected benefits of therapy. Thus, their general opinion has been that cancer patients should avoid antioxidant supplements like the plague.

On the other hand, most practitioners of a naturopathic persuasion have not taken this view, pointing out the lack of incriminating evidence that taking antioxidants is harmful and suggesting that antioxidants may offer systemic protection against harm secondary to treatment. They cite specific examples of individual supplements that appear to be associated with improved outcome.

This debate has gone on for years without adequate published data to help anyone reach a definitive resolution. It is with this backdrop in mind that two studies crossed our virtual desktops in 2019 that both bear examination.

The first, written by AY Jung et al was published January 1, 2019, in the American Journal of Clinical Nutrition. Acknowledging the paucity of information both on supplement use by breast cancer patients and that few studies had looked at the impact dietary supplements, particularly antioxidants, have on breast cancer prognosis, the authors undertook an investigation of supplement use in postmenopausal breast cancer survivors in Germany—specifically looking at postdiagnosis use of antioxidants and other supplements and associations with breast cancer mortality and recurrence free survival in women who were treated with chemotherapy and radiation therapy. Data from 2,223 postmenopausal women diagnosed with nonmetastatic breast cancer from the Mamma Carcinoma Risk Factor Investigation (MARIE) study were used. Women were interviewed at recruitment in 2002-2005 and again in 2009 and followed through June 2015. Multivariate Cox regression analysis was used to estimate Hazard Ratios (HRs) and corresponding 95% Confidence Intervals (Cis).

Before diagnosis 36% of the women reported they took supplements. This increased to 45% after diagnosis. By 2015, 240 of the initial 2,223 women had died, (134 from breast cancer) and 200 had their breast cancer recur. After adjusting for relevant confounders, using antioxidants during chemotherapy or radiation therapy was associated with increased risk of total mortality (HR: 1.64; 95% CI: 1.01, 2.66) and worsened recurrence-free survival (HR: 1.84; 95% CI: 1.26, 2.68). Supplement use in general was not associated with breast cancer prognosis. Only the antioxidants were.¹

The second paper of interest was written by Christine Ambrosone, from Roswell Park, and colleagues from across the US and Canada and published in the Journal of Clinical Oncology almost 12 months later, in December 2019. They too hoped to determine if an association exists between dietary supplement use during chemotherapy and outcomes for patients with breast cancer.

This was a prospective observational study, which used questionnaires to assess supplement use in patients with breast cancer enrolled in a phase III chemotherapy clinical trial. The questionnaires were administered twice, initially prior to beginning chemotherapy and then after completion of chemotherapy. A total of 1,134 patients completed both questionnaires. Among these patients, there were 251 recurrences and 181 deaths. Far fewer patients in this study reported taking supplements than we would have expected based on earlier studies, especially in comparison to the study by Jung et al just mentioned. Whereas in the Jung study supplement use increased dramatically during treatment in Ambrosone’s patients use dropped during treatment. This difference may reflect differing attitudes of German vs. American oncologists.

… prevalence of supplement use, particularly antioxidants, was low compared with reports in the literature of use by patients with cancer and tended to decrease during treatment. For example, vitamin C was used by 20.5% of patients before treatment but only 12.2% during therapy. Vitamins E and A were taken during treatment by 10% of patients. Use of any antioxidant during treatment (vitamins C, A, and E; carotenoids; or coenzyme Q10) was observed among 17.5% of patients, whereas 44% of patients took multivitamins during chemotherapy. [Ambrosone]

Because so few patients reported supplement use, many of the statistical tests lacked power and did not reach statistical significance. However, many of the trends suggested by the data did suggest unwanted effects.

Use of any antioxidants (vitamin A, vitamin C, vitamin E, carotenoids, and CoQ10) both before and during chemotherapy was non-significantly associated with an increased risk of cancer recurrence (adjusted hazard ratio [adjHR], 1.41; 95% CI, 0.98 to 2.04; P = .06) and death (adjHR, 1.40; 95% CI, 0.90 to 2.18; P = .14. The key word here is “non-significant” and that distinction was lost in the news stories that reported these results where in these non-significant associations were reported as fact. For example, Forbes headlined: “Taking Supplements During Breast Cancer Treatment Increases Risk Of Death And Relapse”²

As my dear friend Ian Beirs, ND, enjoys reminding me, non-significant numbers are non-significant and should not be suggested to mean anything. The fact that they might be ‘almost significant’ does not count.

But let’s be honest. Over the years some of us may have, on occasion, cited a paper or two to patients in which the findings did not reach statistical significance. One in particular comes to mind, a paper published in 2012 on metformin use by 63 diabetic patients with triple negative breast cancer against 67 who weren’t taking metformin. After a bit more than five years, the diabetic patients not taking metformin had a 63% greater risk of having had a distant metastasis than those who took metformin (hazard ratio [HR], 1.63; 95% confidence interval [95% CI], 0.87-3.06 [P = .13]).³ Obviously the confidence intervals reveal that this number wasn’t statistically significant. Thus, the study accurately concluded that taking metformin didn’t make any difference, but that didn’t stop me from asking patients, “Imagine there are two teams here, and the one not taking metformin is maybe 63% more likely to be diagnosed with cancer recurrence. Which team do you want to be on?”

We should never do this. We should stick with one set of rules; we can’t have it both ways. We should not ignore significance or lack of it to support something we want our patients to do and then ignore a study whose outcome says something we don’t like. We should ignore much of what Ambrosone reported because it wasn’t significant, but we should pay attention to Jung.

Ambrosone did detect several associations with supplement use that were significant and that we should take note of. First, vitamin B-12 use both before and during chemotherapy was significantly associated with poorer disease-free survival. Vitamin B-12 takers were 83% more likely to have the cancer return and twice as likely to die during the study period (disease free survival: adjHR, 1.83; 95% CI, 1.15 to 2.92; P < .01 and overall survival adjHR, 2.04; 95% CI, 1.22 to 3.40; P < .01).

Likewise, the use of iron during chemotherapy was significantly associated with recurrence (adjHR, 1.79; 95% CI, 1.20 to 2.67; P < .01). Results were similar for overall survival.

I’ve recently written in the past about my concern regarding vitamin B-12. Recall that high levels are significantly associated with cancer occurrence and also associated with lower short-term survival. Johan Arendt reported in 2013 that cancer incidence in people with elevated B-12 levels (>1084 ng/ml) was more than six-fold higher than for people with normal levels.⁴ Arendt went on to report in 2016 that for patients diagnosed with cancer, one-year survival for those with similarly elevated B-12 levels was about half of what it was for patients with moderate levels.⁵

Many practitioners have used vitamin B-12 to prevent and treat neuropathy secondary to cancer chemotherapy. This practice may need to be reconsidered. The relationship between vitamin B-12 status and development of chemotherapy-induced neuropathy may be better predicted by methylmalonic acid levels (MMA) than by serum B-12 status. There is much that we don’t understand about B-12.⁶

The association with iron supplementation is also a concern. It seems to be common practice by many of our local oncology offices to suggest patients take iron “because they are anemic.” Our response to this suggestion has always been to encourage testing ferritin levels in the hope of differentiating anemia of malignancy from frank iron deficiency. The negative association between iron and prognosis seen in Ambrosone’s data should lend further weight to our hesitation to supplement with iron.

In the course of conversations with multiple colleagues, I have heard a multitude of reasons why Ambrosone’s study results should be discounted for methodological reasons. This issue makes me want to rephrase that line from Hamlet to, “My colleague doth protest too much, methinks.” We want to continue believing that cancer patients should take high doses of antioxidant vitamins before during and after treatment. We believe that doing so will be beneficial. Any information that may contradict our belief must be wrong.

Jung’s study showed statistically significantly poorer outcomes associated with taking antioxidants. Ambrosone’s results, probably because fewer patients took vitamins than in the German study, did not have the statistical power to reach significance. We can quickly get lost arguing over whether this is really a bad idea or not, whether we are harming patients or not.

The main thing though is that neither study showed benefit from taking antioxidants; this argues against our long-held assumptions that taking antioxidants is a good idea. If doing so doesn’t help, then why do it at all? We have other interventions that we might invest patient energy and time into adopting that perhaps may have a positive influence on outcome. What comes to mind as I write this? Exercise, weight loss, dietary shifts in macronutrients, meal timing, caloric restriction, fasting and so on. Of course, there has always been a financial incentive to sell vitamin supplements to our patients. If we profited from these other interventions perhaps, we would be more eager to encourage such behavioral changes with the same zeal as we’ve sold antioxidants?

This topic brings to mind a detail from Kurt Vonnegut’s 1963 book, Cat’s Cradle. In his story, Vonnegut describes a fictional religion called Bokononism. Part of this religion’s beliefs entails the concept of a granfalloon, defined as a “false karass.” A granfalloon in Vonnegut’s thinking was a group of people who affected a shared identity or purpose but whose mutual association is meaningless. I recall, even after all these years, that Vonnegut used “Hoosiers” as an example to illustrate his idea. This idea of a granfalloon comes to mind when this debate about antioxidants comes up. We may have grouped too large a group of chemicals together into a shared identity where in reality their mutual association has become meaningless. We are trying to predict the behavior of a diverse group of materials based on a single trait and contrasting that action with the equally broad group of things lumped together as oxidants. Put simply, a blueberry may have more going on inside itself than the vitamin C it happens to contain. Biology is not this black and white. [reading through this many months after my initial draft, this last line wants me to take this a step further than ‘black and white.’ Our segregating the world into oxidants vs, antioxidants may have as much predictive value as trying to judge a person based on skin color. Of course, as absurd a practice as this is, there are no shortage of people who think this is a valid practice.] Our minds are attracted to binary segregation, but the world itself isn’t. This habit of putting value on anything labeled antioxidant may no longer be appropriate.

When deciding which supplements may be beneficial during cancer treatment, we should not slip into the granfalloon of calling them antioxidants to justify their use. Rather we must do the hard work and ask what happens when a specific nutrient, phytochemical, or I suppose food is used in conjunction with the specific chemo drugs against a specific cancer type. Does the total combination increase reactive oxygen species in the cancer, does it increase apoptosis? Even being this persnickety may not be adequate to predict long-term outcomes, in particular overall survival (that was for you, Ian); but at least it will be a step in the right direction. Our past strategy of assuming all antioxidants fight cancer is fast becoming outdated.