Mouse, Rat & Cell Studies Don’t Make E-Cigarettes Carcinogenic

In determining what causes cancer in humans, epidemiologic and public health research is far superior to lab studies based on cells, mice or rats.  While the latter can provide important information about the biology of cancer, the vast majority of carcinogens have been discovered in studies of human exposures. 

Although numerous epidemiologic studies prove that smokers are more likely to contract a variety of cancers, decades of research on cells and animals have failed to establish which of the thousands of toxins in cigarette smoke cause human lung, bladder or esophagus cancer. 

Of the two major components of e-cigarette juice, we know this: Nicotine, the subject of thousands of studies, has never been shown to be a cancer-causing agent, and propylene glycol is generally recognized as safe for human use by the FDA. 

Regrettably, these facts haven’t stopped some researchers from scaremongering about e-cigarettes.  A study published in January (here) has led to a media frenzy suggesting that e-cigarette liquid may be as dangerous as smoke (here).  This is nonsense. 

Normal cells do not live forever.  But cancer cells are “immortalized” and are able to proliferate indefinitely.  The experiments reported in this study were conducted in immortalized cell cultures, which also included mutations of two important genes: p53, an anticancer gene that is active in normal cells, was “silenced”; and k-ras, a well-characterized oncogene, was “activated.” 

The researchers were essentially using a cancer cell line.  They measured the effect of two (unquantified) concentrations of nicotine e-cig solution and some sort of smoke extract on assays of growth and invasiveness after 10 days of exposure. 

Exposure of the cells to the low-nicotine e-cig solution and to the smoke extract had no effect on the invasiveness of the cells (a cancer trait).  They reported, “We will next examine the effects of high nicotine conditioned media on cell invasion,” indicating a future experiment.

The researchers noted that after 96 hours of exposure to e-cig solution, the cells showed changes in gene expression.  This is not particularly newsworthy.  Genes are the bits of DNA that tell cells what to do.  At any given time cells have many thousands of active genes.  Any environmental change can produce changes in the expression of large numbers of genes. 

In their effort to implicate nicotine, the researchers omitted information as to whether they had established appropriate experimental controls, such as exposure of the cells to other common agents such as caffeine or coffee extracts.

Cellular and molecular research explores the incredibly complicated biology of cancer, but it is of limited value in identifying carcinogens.  There are well established tests to determine if an agent is a possible mutagen, which is an indication that it might be cancer-causing.  A 2007 study of American smokeless products was essentially negative (here), which is completely consistent with epidemiologic studies.  It is likely that tests of e-cigarette liquids would produce similar results.   

Undistinguished research on smokeless tobacco products routinely generates headlines and soundbites best suited for the tabloids.  From a public health standpoint, it is shameful that researchers and media conflate vague, exaggerated and highly theoretical claims about e-cigarette juice to the very real risks of cigarettes.

Is “Never Recommend A Carcinogen” An Appropriate Policy? Apparently Only For Tobacco

The Wall Street Journal’s “Numbers Guy” Carl Bialik blogged in April on the scientific debate about promoting smokeless tobacco as a substitute for cigarettes (read it here). Bialik’s well-balanced article included expert comments on the science of tobacco harm reduction, in addition to quotes from the usual array of individuals and organizations opposed to offering smokers safer options.

On re-reading Bialik’s article, I was struck by one passage in particular:

“‘We wouldn’t recommend anybody using a product that causes cancer,’ said Cathy Backinger, chief of the tobacco control research branch at the National Cancer Institute.”

At first glance, the comment sounds perfectly reasonable. Surely no one at the NCI would recommend that ANYBODY use a product that causes cancer. Well, it turns out that this is completely false. In fact, the NCI recommends routinely that women take a medication that is known to cause two forms of cancer.

The medication is tamoxifen, which, according to an NCI website (available here), “interferes with the activity of estrogen, a female hormone that can promote the development of cancer in the breast.” The NCI states that “Tamoxifen is approved by the U.S. Food and Drug Administration (FDA) for the prevention of breast cancer and for the treatment of breast cancer, as well as other types of cancer.”

In other words, the NCI endorses the use of tamoxifen by women who don’t have breast cancer but who are at risk of developing it.

But there’s a problem. The NCI acknowledges that “Tamoxifen increases the risk of two types of cancer that can develop in the uterus: endometrial cancer, which arises in the lining of the uterus, and uterine sarcoma, which arises in the muscular wall of the uterus.” The NCI says tamoxifen doubles the risk for endometrial cancer; the magnitude of the sarcoma risk is not specified. The NCI also cautions that “women who took tamoxifen had an increased chance of developing blood clots and an increased chance of stroke.”

The NCI justifies the use of tamoxifen with a straightforward rationale: “The benefits of tamoxifen as a treatment for breast cancer are firmly established and far outweigh the potential risks. Patients who are concerned about the risks and benefits of tamoxifen or any other medications are encouraged to discuss these concerns with their doctor.”

There are legitimate reasons to recommend tamoxifen for women who are at risk for developing breast cancer. Similarly, there is a rock-solid scientific and medical foundation for recommending smokeless tobacco – with cancer risks that are barely measureable by modern epidemiologic methods – to smokers who otherwise face far higher risks of developing a plethora of medical illnesses.

If the Backinger principle was adopted by physicians, they would withhold tamoxifen from women at risk for breast cancer on the grounds that it may cause cancer of the uterus. That application would violate principles of public health and condemn thousands of women to a life with – and perhaps death from – breast cancer. Dr. Backinger’s refusal to consider tobacco harm reduction similarly violates public health principles and condemns millions of inveterate smokers to a disease and death.

Public health policies should be consistent in using scientific evidence of relative risk to promote legitimate life-saving strategies.

Observations on a New Tobacco Harm Reduction Commentary

In August I discussed a comprehensive meta-analysis of smokeless tobacco use and cancer, which was published by Peter Lee and Jan Hamling in the prestigious journal BMC Medicine. Recently that journal published a commentary by David Timberlake and Jason Zell, professors at the University of California at Irvine. The article, available here, will be widely read and will likely promote lots of debate, so I will make some critical comments.

Timberlake and Zell attribute the origin of the “controversy” over the use of smokeless tobacco as a “potential substitute for cigarettes” to the “steep decline” in smoking among Swedish men “over the past two decades.” They conclude that “it is premature to state that the increased use of ‘snus’ is causally associated with tobacco substitution and the decline in morbidity.” Unfortunately, Timberlake and Zell have repeated a common misunderstanding of the Swedish data.

Most commentators describe the “Swedish experience” as a recent phenomenon. This is incorrect. Philip Cole and I published a study earlier this year in the Scandinavian Journal of Public Health and discussed in this blog, which examined lung cancer rates among men and women in all European Union countries for the past 50 years. We concluded “that snus use is inversely correlated with cigarette consumption among men in Sweden, resulting in the lowest [lung cancer mortality rates] in Europe for most of the past 50 years.” Thus, Timberlake and Zell are mistaken that it is “premature to state that increased use of ‘snus’ is causally associated with tobacco substitution and the decline in morbidity.” In fact, nothing else adequately explains the profound and sustained differences between very low lung cancer mortality in Swedish men compared with other EU countries over the past half-century.

Timberlake and Zell correctly observe that clinical trials are starting to address the efficacy of tobacco harm reduction. They describe a Danish quit-smoking trial that used smokeless tobacco as successful at 7 weeks and unsuccessful at 6 months, but they could have explained that the investigators urged subjects to be tobacco-free at 3 months. It is perfectly clear that smokeless tobacco will not be any better for smoking cessation than pharmaceutical nicotine (that is, terrible) IF the ultimate objective is complete nicotine and tobacco abstinence. This objective is the Achilles heel of the conventional quit-smoking mindset, and it is entirely unnecessary. The appropriate objective is for smokers to lead longer and healthier lives, and this can be accomplished by not forcing smokers to abstain completely from nicotine and tobacco.

Timberlake and Zell repeat anti-tobacco extremists’ claims that “[c]arcinogens in smokeless tobacco include high levels of nitrosamines, polycyclic aromatic hydrocarbons and other agents,” and they cite a monograph from the zealous International Agency for Research on Cancer. These contaminants are present in smokeless tobacco, but in miniscule concentrations that aren’t relevant to human health.

Trace levels of contaminants are in virtually every product that humans consume. For example, in 2000 Bruce Ames, a well known expert in carcinogenesis (cancer causation), published an article noting that roasted coffee contains thousands of chemical agents. At that time about 30 of them had been tested as carcinogens, and 21 were positive. Thus, even though coffee contains cancer-causing chemicals, there is no epidemiologic evidence that coffee is a significant risk factor for any cancer in humans who consume it. Similarly, although smokeless tobacco contains trace levels of carcinogens, the cancer risks among smokeless users are so low that they are barely measurable with modern epidemiologic methods.

Timerlake and Zell represent that Lee and Hamling observed “excess risk” for oropharyngeal cancer among smokeless tobacco users, but this is perplexing. As I noted in August, Lee and Hamling found no risk related to smokeless tobacco after adjustment for smoking and alcohol, two risk factors acknowledged by Timberlake and Zell.

In some sections Timberlake and Zell get on the anti-industry train, which may be attractive to anti-tobacco extremists but may not be entirely accurate. For example, they suggest that Philip Morris had a crystal ball as early as 1984 regarding the benefits of smokeless tobacco. This is just short of nonsense, because until 2005, none of the major cigarette manufacturers had ANY stake in smokeless tobacco. Prior to that, cigarette and smokeless manufacturers were absolute competitors, representing a lost opportunity for tobacco control advocates who failed to focus on the most harmful products (cigarettes and other combustible tobacco products).

Timberlake and Zell also appear to agree with the anti-industry complaint that it is “more interested in the dual use of tobacco products, rather than tobacco substitution.” It is true that some ads promote smokeless use during “times when you can’t smoke.” But the industry has been forced into a marketing corner so tight that it doesn’t have any other choice. With the idiotic warning present on smokeless products since 1986 (This product is not a safe alternative to cigarettes), a marketing campaign focusing on substitution would be a ticket to self-destruction. It is depressing to think that FDA regulation is likely to make a science-based switch-to-smokeless pitch even less likely.

I have been critical, but there is much to recommend Timberlake and Zell’s commentary. For the most part, they provide a balanced discussion of key issues, which is absent in many other published articles. With respect to the Lee-Hamling study, they correctly conclude that “the overwhelmingly null (i.e., zero) associations with cancer in this high quality analysis are provocative, if not compelling.” Regarding the gateway issue (the idea that smokeless tobacco use leads to smoking, especially among children), their interpretation is on target: “even if a gateway effect to smoking exists, which is doubtful, only a minority of smokeless-tobacco users would be affected.” They address the role of human papillomaviruses, another commonly forgotten risk factor in oral and pharynx cancer. Finally, they recognize that FDA regulation “may not be the sole determinant of harm reduction’s fate in the USA,” and they “anticipate that much of the debate will shift from the discussion of epidemiologic data to the discussion of the marketing, health communication and economics of smokeless tobacco.” When that happens, it will be long overdue.

The FDA Crusade Against E-Cigarettes

On July 22, 2009, the FDA released the results of laboratory tests of e-cigarettes, which were conducted by the Division of Pharmaceutical Analysis at the FDA’s Center for Drug Evaluation and Research. In a press release, the FDA said: “These tests indicate that these products contained detectable levels of known carcinogens…” The FDA report can be downloaded here.

For many years, I have investigated the cancer risks of cigarette smoking and smokeless tobacco use. As I wrote in a recent post, the FDA has never regulated nicotine effectively, and the agency had previously signaled its intention to ban e-cigarettes. So while the agency’s new analysis of e-cigarettes comes as no surprise, it does undermine the assumption that the FDA bases it oversight activities purely on scientific principles.

The FDA analyzed 18 cartridges from two e-cigarette manufacturers, Smoking Everywhere and Njoy (there are many other manufacturers). With respect to “carcinogens,” the agency looked at four tobacco-specific nitrosamines (TSNAs) with very long chemical names; I’ll abbreviate the agents here as NNN, NNK, NAT and NAB.

I have some experience with TSNAs, since I participated in a project with a scientist at the Swedish National Food Administration to measure the levels of these agents in smokeless tobacco products. Our research showed that TSNAs are present in most American tobacco products at extremely low levels, about 0.1 to 12 parts per million by weight. At this level of TSNAs, someone who puts 1 gram (about 1/28th of an ounce) of smokeless tobacco in his mouth is exposed to, at most, about 10 one-millionths of a gram of TSNAs. There is abundant scientific evidence that exposure at this minuscule level is not associated with ANY cancer in smokeless tobacco users.

The FDA analyzed 14 products from Smoking Everywhere, but the agency only reported the TSNA levels for 7 of those products. Why did the FDA test only half of the company’s products for carcinogens? And how did they choose those products? There are some clues in the report. First, the products that weren’t tested simply had blank boxes in the results chart. A footnote says, “Open boxes indicate the sample was not available for testing.” Another note in the methods section admitted that “…not all sample lots were available for analysis…as they were consumed in other testing.” In other words, the FDA didn’t purchase enough of the products to conduct the testing in a systematic and scientific manner. Maybe it’s a budget problem. On the Smoking Everywhere website cartridges are $9.99 each.

The FDA tested 3 out of 4 Njoy products for TSNAs.

What the FDA didn’t test is even more important than what the agency tested. The report noted that the “Nicotrol Inhaler, 10mg cartridge was used as a control for some test methods.” That inhaler is a pharmaceutical nicotine product that is regulated by the FDA, but the agency didn’t test the product for TSNAs. This is a critical omission, because in 2006 a published research study revealed that pharmaceutical nicotine products contain TSNAs. In fact, it’s been known for almost 20 years that nicotine medications contain TSNAs.

Why did the FDA analyze e-cigarettes for carcinogens, when there is no evidence the agency ever conducted carcinogen studies of products that they have regulated for over 20 years? Is it possible that the FDA approved medicines that contained TSNAs, but the agency is now disapproving e-cigarettes because they contain the same contaminants? To answer this important question, we have to know how high – or how low – the TSNA levels are in these products.

Unfortunately, the agency did not report TSNA levels. Instead, it reported that TSNAs were either “Detected” or “Not Detected,” which is entirely inadequate. For hundreds of years, one of the basic tenets of medicine has been “the dose makes the poison.” Mere detection of a contaminant is meaningless; the critical question is: At what concentration is it present?

So what does “Detected” mean in the FDA analysis? In other words, what was the lowest TSNA concentration that the test detected?

As I noted earlier, many tobacco products have TSNA levels in the single-digit parts per million range, a level at which there is no scientific evidence that TSNAs are harmful. According to the report, the FDA used an analytic method published in 2008. The report notes that “the published method is quite sensitive for the TSNAs…” and it goes on to explain that the level of detection is 40 parts per TRILLION.

The implications of this are astounding. Apparently, the FDA tested e-cigarette samples using a method that detects TSNAs at about 1 million times lower concentrations than are even possibly related to human health.

In summary, the FDA tested e-cigarettes for TSNAs using a questionable sampling regimen, and methods that were so sensitive that the results may have no possible significance to users. The agency failed to report specific levels of these contaminants, and it has failed to conduct similar testing of nicotine medicines that have been sold in the U.S. for over 20 years.

These are not the actions of an agency that is science-based and consumer-focused. These pseudo-scientific actions are clearly intended to form the justification for banning a category of products that are probably 99.9% safer than cigarettes. According to Dr. Murray Laugesen, a respected New Zealand researcher, “Simply banning e-cigarettes will simply consign thousands of e-smokers back to smoking tobacco and an early death.”

The FDA and anti-tobacco extremists who support it should be held accountable for their prohibitionist actions.

The FDA has a legitimate interest in two matters involving e-cigarettes: assuring that cartridges contain the advertised quantity of nicotine, and that they do not contain contaminants.

I welcome the FDA to correct any errors in this critique. I attempted but was unable to reach the scientist who conducted the analyses.