Articles

Cancer prevention: The forgotten stepchild of cancer research?

The New York Times has been periodically running a series about the “40 years’ war” on cancer, with most articles by Gina Kolata. I’ve touched on this series before, liking some parts of it, while others not so much. In particular, I criticized an article one article that I thought to be so misguided about how the NIH grant system leads researchers to “play it safe” and how we could cure cancer if we could just fund “riskier” research that I had to write an extended screed about the misconceptions in the article. The latest installment, Medicines to Deter Some Cancers Are Not Taken, also by Kolata, is much better in that it discusses a problem at the heart of cancer, namely that we have developed drugs that can decrease the risk of specific cancers but they are not as widely used as they could be.

The first part of the article contrasts a seeming incongruity:

Many Americans do not think twice about taking medicines to prevent heart disease and stroke. But cancer is different. Much of what Americans do in the name of warding off cancer has not been shown to matter, and some things are actually harmful. Yet the few medicines proved to deter cancer are widely ignored.

Take prostate cancer, the second-most commonly diagnosed cancer in the United States, surpassed only by easily treated skin cancers. More than 192,000 cases of it will be diagnosed this year, and more than 27,000 men will die from it.

And, it turns out, there is a way to prevent many cases of prostate cancer. A large and rigorous study found that a generic drug, finasteride, costing about $2 a day, could prevent as many as 50,000 cases each year. Another study found that finasteride’s close cousin, dutasteride, about $3.50 a day, has the same effect.

This is indeed a contrast. Think about it. Millions of Americans take statins, for instance, to lower their cholesterol and thereby try to prevent the complications of elevated cholesterol, such as heart disease, vascular disease, and strokes. Yet, for at least two common cancers, there are proven effective drugs that will lower the risk of cancer considerably with a side effect profile at least as favorable as that of statins.

Of course, preventing cancer is not like preventing heart disease. “Cancer” is not a disease, but many diseases arising from many different organs with many different biological behaviors. There is no drug that can prevent “cancer.” There are, however, specific drugs that can lower the risk of specific cancers. So right there you have a big difference that could partially explain the reticence. Which cancers should be prevented? How many different drugs are you willing to take to prevent how many different cancers?

One thing that the article points out that is quite true and has been quite disappointing to cancer researchers, and that’s that diet does not have nearly as large an effect as we had hoped. Kolata correctly points out that if we could eliminate smoking cancer deaths would decrease by about a third. Lung cancer is quite rare in nonsmokers. Indeed, back in the early part of the 20th century, lung cancer was so rare that there are reports describing how, when a patient died of it, medical students would be told that they had to go to see the autopsy because they might not ever see another case in their careers. Now lung cancer is the single largest cause of cancer deaths in both men and women, causing 30% of cancer deaths in men and 26% of cancer deaths in women, and it’s all due to smoking:

Cancer

No other preventative intervention could conceivably even come close in terms of effect as cutting smoking to zero. Obviously, such a thing will not come to pass in my lifetime, but as a cancer surgeon I can always hope, can’t I? Also as a human being, too. After all, I’ve already lost an aunt to smoking-induced lung cancer in 2008, and I have several relatives who smoke who, I fear, might well join her.

After that, Kolata suggests that avoiding hormone replacement therapy containing estrogen and progestin after menopause would be a good preventative, which may be true, but since the Women’s Health Initiative study reported in 2002 that showed an increased risk of breast cancer from such HRT, few women take them any more anyway.

What’s more interesting is how Kolata correctly points out just how little diet and other lifestyle factors seem to matter compared to our hopes. If anyone wants to counter the claims of woo-meisters that we evil “allopathic” doctors don’t pay attention to diet and lifestyle, one only needs to look at just how much money has been spent on studies of diet and exercise and how they impact cancer and heart disease. There have been many studies. Of course, it’s incredibly difficult to do a randomized study, particularly a double-blinded one, of diet and exercise; so most of them end up being observational. The problem with such studies is, as with all such cohort studies, controlling for confounders. It’s hard enough to control for confounders that the investigators know about, but there are often many confounders that they don’t know about and may not identify until after the study is over. In any case, suffice it to say that studies of diet as a means of preventing cancer have been disappointing of late, as Kolata points out:

For example, public health experts for years recommended eating five servings of fruits and vegetables a day to prevent cancer, but the evidence is conflicting, at best suggestive, and far from definitive.

Low-fat diets were long thought to prevent breast cancer. But a large federal study randomizing women to a low-fat or normal diet and looking for an effect in breast cancer found nothing, said its director, Ross L. Prentice of the Fred Hutchinson Cancer Research Center in Seattle.

Fiber, found in fruits, vegetables and grains, is often thought to prevent colon cancer, even though two large studies found no effect.

“We thought we would show relationships that were strong and true,” said Dr. Tim Byers, professor of epidemiology at the Colorado School of Public Health, “particularly for dietary choices and food and vegetable intake. Now we have settled into thinking they are important but it’s not like saying you can cut your risk in half or three-quarters.” Others wonder whether even such qualified support is misplaced.

There has to be a reason the research disappointed, said Colin B. Begg, chairman of the department of epidemiology and biostatistics at Memorial Sloan-Kettering Cancer Center. Perhaps the crucial time to intervene is early in life.

“That’s one possibility,” Dr. Begg said. “The other is that it’s all sort of nonsense to begin with.”

One potential reason for these results is that the effect of diet or other environmental exposures may have a “window of vulnerability” that occurs in youth or childhood. Certainly, in the breast cancer field, we are starting to think that and actively research the possibility. The problem is, if that’s true, then dietary interventions during adulthood may have little or no effect on ultimate cancer risk.

Which brings us to drugs and supplements.

Kolata mentions the cautionary tales of beta-carotene and selenium, the former of which was thought to prevent cancer and the latter to prevent prostate cancer. However, when a large study of beta carotene was done, not only did the supplement not prevent cancer but it appeared to increase the risk of lung cancer in smokers. Selenium and vitamin E similarly were found to have no protective impact on prostate cancer risk and, in fact, the trial studying the question was halted because there was a hint of actually increased risk. Even so, as Kolata points out, ads for such supplements continue to imply, with the Quack Miranda warning, that such supplements “improve prostate health,” whatever that means, although the implication is that they decrease the risk of prostate cancer.

Now here’s where the drugs come in. Because I’m a breast cancer surgeon, I’m going to focus mainly on strategies to prevent breast cancer. It turns out that tamoxifen can decrease the risk of breast cancer by 50% in high risk women. Tamoxifen is a selective estrogen receptor modulator (SERM) that antagonizes estrogen action in breast tissues but has partial agonist activity in bone and endometrium. Indeed, raloxifene, another SERM, is used primarily to prevent and treat bone loss in women with osteoporosis. A $110 million clinical trial showed that its ability to prevent breast cancer is not statistically significantly different than that of tamoxifen, with the advantage that, unlike tamoxifen, raloxifene does not increase the risk of uterine cancer and does not increase the risk of blood clots as much as tamoxifen does.

So, with such overwhelming evidence that in women at high risk for breast cancer, why is it that so few are offered the option of tamoxifen or raloxifene? Kolata explains:

“Those were your tax dollars and mine,” he [Dr. Victor Vogel] added. “You can’t do too many $110 million studies.”

He cannot understand why no one cares, but some doctors say they see a number of problems. It is usually not the cost; tamoxifen is about 30 cents a day and raloxifene $3.30 a day. It is doctors’ practices and women’s concerns.

Most doctors, said Dr. Therese B. Bevers, medical director of the Cancer Prevention Center at M. D. Anderson, do not take the first step — calculating a woman’s lifetime risk of getting breast cancer — in part because that can lead to the next step, spending an hour or so discussing cancer risk and drug risks and benefits.

Dr. Bevers suggests the drugs for women whose lifetime odds exceeds 20 percent. That could include, for example, a 55-year-old woman who began menstruating early (increasing the risk), had her first child late (again increasing the risk), and whose mother and sister got breast cancer. About half the time, though, women with that kind of risk turn down the drugs, Dr. Bevers said. “The No. 1 reason I hear is, ‘Oh, I just don’t like to take medications,’ ” she added.

Yet the same women will happily down supplements that woo-meisters like Dr. Mercola or Mike Adams will suggest to them as cancer preventing. Some will down many supplements a day without thinking twice about it, even though there is no good evidence that they do anything other than enrich the supplement manufacturers who make them and they often cost far more than the paltry 30 cents a day that tamoxifen costs. Moreover, in general for prevention the usual course of tamoxifen or raloxifene is five years, in marked contrast to supplements, where the woo-meisters recommend them for the rest of one’s life.

Of course, never having worked anywhere else other than NCI-designated comprehensive cancer centers, I am probably spoiled in this respect. Such centers always have genetic counseling and prevention clinics. I can simply refer suitable patients there, where our trusty and excellent genetic counselors will calculate patients’ lifetime risk of breast cancer, construct family trees, decide if testing for cancer susceptibility genes is appropriate, and then make science-based recommendations for risk prevention strategies and/or screening strategies tailored to the individual patient. (Who says “allopathic medicine” doesn’t individualize therapy?) In any case, I can see why primary care doctors don’t bring these issues up more often, but with the advent of genetic counseling clinics I am puzzled why more patients aren’t referred to a specialist for these conversations.

One problem mentioned in the article is that there aren’t any particularly good biomarkers for increased risk for cancer, especially ones that fall in response to preventative interventions, the way, for example, cholesterol does for heart disease. Another is what I mentioned earlier, namely that cancer is not one disease, meaning that it is likely that many drugs would be needed to prevent various cancers. True, for the really common cancers, like prostate, breast, and colon, it might be worth it to develop such drugs, but neither the risk-benefit profile nor the number of patients who would benefit would be likely to lead to such drugs for less common cancers. And, as has been pointed out, our risk assessment tools are currently crude at best. In fact, even with better biomarkers I’m not so sure that interest would improve because biomarkers have their own problems and many people have a tendency to fear pharmaceuticals more than they fear a hypothetical increased risk of cancer. In the case of drugs that prevent breast cancer, nearly all of them have anti-estrogen activities and thus produce side effects of estrogen deprivation, such as hot flashes and other menopausal symptoms. It’s a very high bar to meet to produce effective drugs that people will actually take, given that it’s hard enough to persuade patients to comply with drugs designed to treat actual diseases and conditions.

The end result of all of this is that drug companies, once interested in developing cancer preventing drugs, are now floundering. As Kolata notes:

But risk assessment is not easy, and biomarkers are still more of a dream than a reality. There are other problems, too. If each cancer requires a different drug for prevention, how many drugs can a person take? For now, Dr. Curt said, the very idea of cancer prevention is daunting. And since cancer can take decades to develop, by the time a study concludes, a drug’s patent life may be over.

It is not a pretty picture, Dr. Vogel said.

“You have to think that in boardrooms they are saying, Man, did we learn a lesson,” he said. “We will stay as far away as possible from cancer prevention.”

Sadly, supplement manufacturers are not faced with such problems because they don’t actually have to demonstrate that their supplements have the health effects that they claim they have. The war against cancer goes on, but one weapon is not being developed.

Posted in: Cancer, Clinical Trials, Herbs & Supplements, Nutrition, Politics and Regulation

Leave a Comment (8) ↓

8 thoughts on “Cancer prevention: The forgotten stepchild of cancer research?

  1. James Fox says:

    I was glad to see you discuss this article which I found quite interesting when I read it last week. I continue to be saddened and amazed when my friends who are cancer patients have been given a list of recommended supplements to take by their physicians, and more recently a friend who had been diagnosed with bone degeneration is a wrist and had his orthopedic surgeon recommend acupuncture. What gives a patient a sense of control and participation in their treatment must be the misguided motivation some doctors have when making these senseless recommendations.

  2. mgoozner says:

    While I agree with you about the woo-meisters, Kolata’s article was way off base when it comes to prostate cancer prevention. See my post here: http://www.gooznews.com/node/3169.

    Also, while you have nice things to say about her comments on smoking cessation and the impact of reducing hormone replacement therapy as cancer prevention, those issues received all of one paragraph each in her story.

    And as for your endorsement of tamoxifen and raloxifene for breast cancer prevention, while there’s no doubt it is beneficial for women at high-risk, there’s no effort to in her story or your post to compare the benefits of those drugs to their risks (endometrial canc r, for one) and their side effect profiles, which may be a major reason that many women at high-risk don’t want to take them.

    The Association of Health Care Journalists suggests in its best-practice guidelines that relative risk reduction ALWAYS be accompanied by absolute risk numbers, and that some effort should be made to compare absolute benefits to absolute risks. Neither her story nor your blog post met that best-practice test.

  3. wales says:

    Interesting article. Thanks to mgoozner for bringing up the highly important subject of absolute v. relative risk. Two books I’ve read recently discuss the statistics of health risks and cancer testing in depth. I highly recommend them for those trying to sort through all the information and make practical decisions in their own lives.

    Know Your Chances: Understanding Health Statistics by Steven Woloshin, MD, MS, Lisa M. Schwartz, MD, MS, and H. Gilbert Welch, MD, MPH

    http://www.ucpress.edu/books/pages/10893.php

    Should I Be Tested for Cancer? Maybe Not and Here’s Why by H. Gilbert Welch, M.D., M.P.H.

    http://www.ucpress.edu/books/pages/10079.php#tocTab

  4. David Gorski says:

    mgoozer,

    Your point about prostate cancer prevention is well taken. I’m not a prostate cancer doc and am not familiar with the background studies. Neither am I a health care journalist.

    That being said, my purpose was not to get into the full nitty-gritty of the statistics of risk reduction due to Tamoxifen, but mainly to point out that the benefits of tamoxifen prevention therapy in high risk women have been well-established. In case you’re interested, the absolute numbers for risk reduction due to tamoxifen reported from the NSAPB P-1 trial in 2005 showed a risk reduction for invasive breast cancer from around 44 per 1000 women in the placebo group to 25 per 1000 women in the tamoxifen group and for DCIS from 16 per 1000 women in the placebo group to 10. per 1000 women in the tamoxifen group. This was over 7 years. In contrast, perhaps I should have pointed out that supplements in general have nowhere near that level of evidence of doing anything. Mea culpa if I wasn’t precise enough, but I wasn’t going for the sort of discussion I’ve had, for instance, about mammography lately.

    That being said, the post to which you link reminds me that one area that I actually should have criticized Kolata for was her simple assertion that breast cancer incidence fell 15% after the publication of the WHI study showing the risks from HRT in 2002. It was simplistic to link that fall directly to only HRT and to label it as 15% with the implication that the entire 15% was due to reductions in HRT use, which is what Kolata did. The main problem with Kolata’s simplistic treatment of the issue is that breast cancer incidence had started to decline, at least in the U.S., before 2002, although the decline was sharper between 2002 and 2003. Another problem with interpretation is that the decline was very rapid, given how long breast cancer takes to develop, and that not all countries that experienced similar declines in HRT use saw a major decrease in breast cancer incidence. Indeed, at first consideration, one would expect it to take several years for the decline in HRT to lead to a decrease in breast cancer incidence. The probable explanation (as best as we can tell–more science needs to be done) is that perhaps HRT potentiated the growth of incipient DCIS or breast cancers already, cancers that might have regressed. (The possibility that as many as 1 in 5 breast cancers resolve spontaneously is one that I’ve discussed before.) Another issue is that screening mammography also decreased in that time period. True, an argument against the decreased use of mammography as a factor in a decreased single year incidence of breast cancer has traditionally been that the decrease in cancers was primarily in ER(+) tumors, as would be expected if HRT cessation were the main cause, but also remember that ER(+) tumors tend to be more indolent and thus more likely to be picked up on screening mammography–a difficult issue to tease out.

    Although I have no doubt that the decrease in HRT use has at least something to do with the decline and is very likely the dominant factor, the decline was also probably not 15%–at least not 15% directly attributable to HRT decline. Indeed, more recent data suggests that it was more on the order of 8.6%. Still impressive but less than 15% by nearly half (Ravdin PM, Cronin KA, Howlader N, et al. The decrease in breast-cancer incidence in 2003 in the United States. N Engl J Med 2007;356:1670-1674). Actually the article to which Kolata linked concluded that the reduction was 11%.

  5. TimMills says:

    Thanks for bringing this up. I hadn’t known that the diet evidence was so weak, or that there were such effective preventative drugs.

  6. Ali771 says:

    Another thanks for posting on this article. Very interesting.
    While the vitamins that have been studied so far have proven to be without use in terms of cancer prevention, I must say I will be interested to see more research regarding the role of vitamin D.

  7. Amy Alkon says:

    I second that on D. I take 5,000 iu because I live like a bat, and plan to get my levels tested next week. But, I worry about what the long-term effects might be.

  8. Pingback: gmarris

Comments are closed.