■ site map
June 2010 - December 2013
2 good choices to prevent breast cancer
No more, no less. Breast cancer prevention got to that final point, and beyond. Now you have not one, but two good choices that will safeguard you from ever getting the dreaded disease. Fear breast cancer no more; the rest is history...
This is what the headline, widely distributed through the media, conveys. And, it isn't sloppy journalism. It is exactly what the June article by two medical doctors from M.D. Anderson Cancer Center, Huston, throws at the readers of the Cancer Prevention Research journal (Two Good Choices to Prevent Breast Cancer: Great Taste, Less Filling, Hortobagyi and Brown, 2010).
Is this for real?
Of course, not. The article is actually commenting on another article in the same journal issue, which reports the results of a follow-up on a 2006 trial concerning use of tamoxifen for breast cancer prevention (Update of the National Surgical Adjuvant Breast and Bowel Project Study of Tamoxifen and Raloxifene (STAR) P-2 Trial: Preventing Breast Cancer, Vogel et al.).
The commentary article tells very concisely what this is all about: for 12 years - i.e. since FDA in 1998 approved tamoxifen (estrogen-blocking drug manufactured by Zeneca Pharmaceuticals, sold under brand name Nolvadex) for breast cancer preventive treatment - the public has been avoiding it, as well as the alternative drug, raloxifene. It is time, say Hortobagyi and Brown, to reemphasize the "great preventive benefit" of these two drugs to the public.
What is it, specifically, that should be reemphasized about these two drugs and breast cancer prevention?
Report by Vogel et al. is a follow-up on 2006 study by the same lead author, STAR (Study of Tamoxifen and Raloxifene, within the National Cancer Institute supported - but taxpayer funded - National Surgical Adjuvant Breast and Bowel Project, or NSABP for short). First thing that we notice here is similar misrepresentation of the effect of these two drugs: the report puts in the headline "Preventing Breast Cancer", a potentially misleading twist of the proper expression that should read something like "preventive treatments for breast cancer".
Just to make it crystal clear:
neither drug prevents breast cancer.
What they do offer, according to study authors, is a significant reduction in the risk of developing the disease. The authors claim that the results of their follow-up on the 2006 study confirms this benefit, also supplying more specific information about each drug's effects - both, beneficial and adverse.
So, let's look at their results, starting with the original 2006 study (Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial, Vogel et al, JAMA).
The goal of this trial was to assess efficacy of the newly developed drug raloxifene with respect to existing invasive breast cancer preventive treatments, and its other beneficial and adverse effects vs. already FDA approved tamoxifen. A total of 19,747 women was divided in two groups, and subjected to 5-year treatment with either tamoxifen (9,726) or raloxifene (9,745). The results are shown in the table below.
Study conclusion was that raloxifene is "an effective alternative to tamoxifen for reducing the incidence of invasive breast cancer in postmenopausal women at increased risk of developing the disease", also causing fewer cases of endometrial cancer, pulmonary embolism and deep vain thrombosis (DVT). It also had lower incidence of less serious adverse effects (gynecologic problems, vasomotor symptoms, leg cramps and bladder control problems).
Study conclusion, however, fails to point out that raloxifene was associated with significantly higher incidence of non-invasive breast cancer. Also, it had significantly higher symptom severity scores for musculoskeletal problems, dyspareunia (painful intercourse) and weight gain.
But how did authors figure out that these drugs reduce breast cancer incidence, with
no control (placebo) group at all?
Who needs control group, anyway, if they can make their own little calculation about what the risk should be without treatment, and then simply compare that figure with the incidence in the treatments groups?
You may be thinking I'm joking around, but that is exactly what
study authors did: they used so called "modified" Gail model to calculate the
So, with the actual total of invasive breast cancer events in both groups of 331, they proudly credited the two drugs with 58% invasive cancer reduction rate. Since for both, invasive and non-invasive combined, the rate came out somewhat lower (40%), they settled with ~50% reduction rate, which is the figure usually mentioned in media articles.
What is, exactly, the Gail model? In the late 1980s, using data from 2,800 white women, a group of statisticians at the National Cancer Institute (NCI) developed a statistical scheme for estimating theoretical risk of breast cancer. The scheme is based on four criteria: age of first menstruation, number of breast biopsies, age at first live birth and number of immediate relatives with breast cancer.
Obviously, any risk figure extracted from this type of very limited, statistical data is inevitably crude, with possibly - and probably - large margin of error. Its own creators characterized the model as having three major sources of uncertainty, one of which is that no one really knows what are the risk factors for breast cancer. In other words,
Gail model's accuracy cannot be determined.
They also said that better model could be created from larger database, but that was never done.
What was the purpose of developing such unreliable breast cancer risk assessment tool?
During the 1980s, drug companies were actively researching opportunities in breast cancer treatment; drugs were also treatment of choice for the NCI which, mainly through the NSABP, conducted its own studies as well. A tool was needed to select study populations according to desired risk criteria, and the Gail model happened to fit the bill.
Zeneca, the manufacturer of tamoxifen, developed an updated (modified) version of the Gale, biased to overestimating the risk (rather expected, considering that tamoxifen was intended for high-risk women). This is the Gale version used in the STAR study; it came out pretty handy, since by projecting higher than actual "no-treatment" risk, it almost certainly helped both, tamoxifen and raloxifene,
appear to be more efficient preventive drugs
Why should study authors worry about tamoxifen's effectiveness, which was demonstrated in a 1998 study that led to its FDA approval? Well, the problem is that it is still sort of uncertain whether, or how much tamoxifen reduces the risk of cancer. It is a long story, but here are the highlights:
∎ FDA approved tamoxifen based on a single study that was ended one year before its planned 5-year period (Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study, Fisher et al. 1998), and whose lead author was, after congressional hearings, stripped of his position for complicity in fraudulent activities within this trial as well as 14 trials preceding it, then re-instated despite admitting the knowledge of these activities
∎ nevertheless, the FDA quickly approved the drug after reviewing less than 10% of study's case reports, ignoring clear indications of drug's toxicity, which made mortality rates in the treatment and placebo group even (3 from breast cancer, 3 from pulmonary embolism in the former, and 6 from breast cancer in the latter), as well as doubled rate of endometrial cancer in women under 50, quadrupled in women over 50, and doubled rate of breast cancer among a relatively few black women in the study
∎ closer look at the adverse effects reveals that the trial recorded 86 fewer breast cancer cases in the tamoxifen group, but it also had 21 more endometrial cancers, 14 more strokes, 12 more pulmonary emboli and 13 more DVT cases; adding 26 fewer cases of osteoporotic fractures and 67 more cases of cataract, there is no significant benefit-to-risk ratio to claim for tamoxifen treatment
∎ in addition, all 86 extra cancers in the placebo group (175 vs. 89, or 2.65% vs. 1.35% incidence) were ER positive, mostly small (less than 2 cm in diameter) and axillary node negative, with the overall 80-90% curability; on the other side, cancers in the tamoxifen group tended to be ER negative, larger and node-positive, with lower curability, indicating lower breast cancer mortality benefit than what the number of cases alone implies (seems as if tamoxifan does help some women avoid developing breast cancer, but those that do fall pray to it are likely to be worse off with the drug)
∎ FDA approved tamoxifen brushing aside the results of two European studies published that same year, which did not find any benefits of the drug, but did find the evidence of its significant toxicity (Interim analysis of the incidence of breast cancer in the Royal Marsden, Powles et al. 1998, and Prevention of breast cancer with tamoxifen: preliminary findings from the Italian randomised trial among hysterectomized women, Veronesi et al. 1998)
∎ FDA's own advisory committee - which, contrary to agency's definition of this body being made of people with "diverse professional education, training and experience", had 8 out of its 11 members closely involved with testing and use of chemotherapy drugs - stated unanimously that Fisher et al. did not demonstrate that tamoxifen has favorable benefit-to-risk ratio for the prevention of breast cancer for women at increased risk, as defined by study population; that, however, did not prevent the committee to issue recommendation to approve tamoxifen for "risk reduction of the short-term incidence of breast cancer in women at increased risk as defined by the study population"; obviously, the committee's two statements directly contradict each other, and the drug was never actually approved for extended preventive use, although it was readily promoted as such by the media and the American Cancer Society
∎ the committee further demonstrated its bias by voting it unnecessary to advise women on tamoxifen to have endometrial testing, or yearly exam for cataract - another drug's side effect - while passing the issue of tamoxifen-caused blood clots to someone else to look into
∎ initial results from IBIS-1, a placebo-controlled clinical intervention trial involving a total of 7152 women aged 35-70 that took tamoxifen for 5 years, with 50-month follow-up, had 32% lower breast cancer incidence in the tamoxifen group (First results from the International Breast Cancer Intervention Study (IBIS-I): a randomised prevention trial, Cuzick et al. 2002). But the overall numbers did not look good, again:
In short, the total major events score was even for both, while mortality was more than doubled in the tamoxifen group. The authors tried to downplay the deaths, but the best conclusion they could come up with was that the overall benefit-to-risk ratio for tamoxifen is "still unclear".
Note that the above table lists only major adverse effects of tamoxifen; those less serious - some of them mentioned earlier - are common, and it is a good part of the reason why the drug hasn't become more popular.
The above answers pretty well why the STAR study authors opted to replace placebo group with statistical risk estimate for their participants (e.g. modified Gail model). That helped control the study outcome and obtain the "right numbers" by eliminating drug's adverse effects - including excess of all-cause deaths - from the picture. All that's left was the breast cancer risk reduction efficacy of tamoxifen - its only declared, yet still uncertain benefit.
There is no doubt that the latest two articles on tamoxifen in the Cancer Prevention Research journal are a part of the concentrated effort to accelerate sales of this drug as an effective breast cancer preventive. Big money is at stake: we are talking about
annual market of $40 billion,
or so, potential.
It is worth taking the chance, isn't it, and mislead people a bit (well, all right, more than "a bit")? It is also worth suppressing alternative, safe prevention measures based on the number of factors actually linked to the increased risk of breast cancer. Unlike drugs, these measures do address probable causes of breast cancer, and do not cause adverse health effects.
Not known to me at the time, this article was only the beginning of a long journey into the world of breast cancer (BC), what is known about it, and what is not, from its causes to the preventive measures that could reduce the risk of getting it. It took three years of rambling through all kind of papers and articles, all of them to some extent incomplete, most of them biased, often contradicting, trying to put together any kind of a meaningful conclusion.
One thing that clearly showed up is a consistent campaign of the distortion of facts by the industry marketing to BC prevention, screening and treatment. In general, as the above text well illustrates, it is creating and propagating unrealistic picture of their products' efficacy, while downplaying their negatives, as well as the beneficial potential of the alternatives.
Since this text arrived at its final form only after all that needed to be considered was reviewed, processed and organized, it wasn't possible to publish monthly. Instead, it is one big news article covering a 3-year period. While still not complete, and not highly competent, it is to my knowledge the most complete coverage of the facts on breast cancer available online. Most importantly, it is fully independent: no interest of any kind is at stake on my side.
The journey begins with the very next page. The summary, and the most important part, is the last article Minimizing breast cancer risk.