Mammograms

Q:I am 47. My doctor is pressing me to have mammograms every other year. What evidence is there that it can help prevent cancer? M S, Hitchin.

A:Absolutely none as medicine embarassingly admits in its own literature while stepping up screening for all ages sharply. As the number two killer of women in the West after smoking, breast cancer has become a political football, with breast cancer activists on both sides of the Atlantic demanding government action. In America, Congress responded to pressure by breast cancer activists by ordering the National Institutes of Health to increase spending on breast cancer by nearly 50 per cent to some $132.7 million. Lately, the American Cancer Society has begun employing tactics like designating October 19 as National Mammography Day ("Be sure she makes the appointment that could save her life," says the ad in the medical literature, sponsored by the ZENECA HealthCare Foundation); or offering individually tailored letters and telephone calls.

In the UK, the government launched its National Breast Screening Programme in 1990, offering mammography to women aged 50-64, and in its first year exceeded its target of screening 70 per cent of the million women invited to participate every three years.

During a recent international conference on breast screening Dr Laszlo Tabar, a Swedish radiologist, said this isn't good enough. He took issue with the British health department policy of recommending screening for women over 50 once every three years. In his view, women should be screened every 18 months and women between 40 and 49, every year (Doctor, April 1993). The American College of Obstetricians and Gynecologists also has called for more mammograms among women over 50.

All this activity may comfort those who wish to see a government seen to be doing something. However, nobody is quite sure exactly who should be screened and how often. Wide variations exist between countries (and even between different governmental bodies) as to which groups of women would most benefit from screening.

Following the results of a Swedish overview, which pooled results from five studies conducted from five to 13 years on some 300,000 women, most members of establishment medicine have adopted as gospel its results that for women 50 and over,regular screening can reduce breast cancer mortality by 30 per cent (The Lancet, 12 June 1993). However, from the results of this test and others it is also generally agreed that no studies have shown a benefit for women younger than 50.

That apparently hasn't prevented the American Cancer Society and the American College of Radiology from urging all women over 40 which of course includes this limbo group between 40 and 49 to have annual mammograms. In early 1994, the National Cancer Institute broke ranks by reversing its recommendations dating back to 1987 advising all women over 40 to have routine mammograms. The reason for this volte face, it said, was that randomized clinical trials of routine screening mammography "have not shown a statistically significant reduction in mortality in women under the age of 50." (JAMA, 1994:271;2: 96).

This decision was followed by a similar move by the government of New Zealand. Top breast cancer specialists Ismail Jatoi and Michael Baum from London's Royal Marsden Hospital wrote a special feature labelling American doctors giving mammograms to the under 50s "negligent" because it can often do more harm than good (BMJ, 4 December 1993).

But even among the over 50s, there is no conclusive evidence that mammographic screening is doing any good. In the much quoted Swedish overview, which medicine has used as proof that screening works, the researchers came up with their figure by pooling all the results of three bands of age groups the 40-49 year olds, 50-69 year olds and 70-74 year olds into an overview. Although the study showed a positive benefit (29 per cent reduction in mortality) among the 50s, there was no significant benefit among those in their40s or 70s.

However, when you actually examine the statistics and science behind them, this is the only study to show clear benefit, even among the 50s. The 30 per cent improved survival figure being bandied about by medicine may also derive from several articles which have looked at all the studies of screening; although most studies did not show a clear benefit, the articles concluded that those that were most scientific, or "randomized" (that is, women assigned randomly to either screening groups or controls) all proved benefit (BMJ, 4 May 1991).

However, University of Dublin epidemiologists Petr Skrabenek and James McCormick, both scourges of unproven medical practice, point out that three of the four randomized controlled (that is, women randomly assigned to be screened or not) trials "failed to reach statistically significant benefit for women aged 50 and over." (BMJ, 8 June 1991). These include two studies of an aggregate of 30,000 women, which were dismissed as too small by one set of screening apologists (BMJ, 6 April 1991).

To reach their glowing statistics, charges the Dublin epidemiologists, academics have combined various types of scientific studies randomized controlled with case-controlled, for instance in an attempt to make the non-significant advantages of screening appear significant. Another Bristol doctor pointed out that two of the "finest centres" in the United Kingdom failed to produce a statistically sound reduction in mortality using annual clinical examination and every-other-year mammography.

Furthermore, for the over-50s, it's wise to keep in mind what this 30 per cent figure actually translates into. As Dublin's Petr Skrabanek points out, it means that one cancer death for every 12,000 women invited for screening every year would be prevented or postponed.

However, in correspondence with the authors of the big Swedish study, he points out that more women actually died in the screened group than in the controls from all causes. "In this sense, not one life was saved," he says. "A detailed analysis of overall mortality is necessary to exclude the possibility that a minor reduction in mortality from breast cancer (less than one per cent of all deaths) in the invited group was not offset by a similar increase in deaths from causes other than breast cancer (The Lancet, 1993: 341: 1531). He also points to the fact that this represents data for as many as 2.5 million woman years which fails to show a net benefit in survival.

As for those over 60, a column in the British Medical Journal recently provided a neat demonstration of medicine's usual assumption of benefit before scientific proof. Reporting on US experts attempting to convince the over 65s to get screening the columnist cheerily admitted that there is "little hard evidence but plenty of sound reasons for believing that screening is just as effective in the elderly as those in their 50s". In other words, as age is the most important risk factor for the disease, and 65 year olds may go on to live another 12 years, it must be good for them (BMJ, 8 May 1993).

Because we know that survival from breast cancer correlates with the size of the tumour, the rationale for screening is that the earlier you catch it, the smaller the tumour will be, and hence the greater your chances of beating the disease. However, Johannes Schmidt, an epidemiologist in Switzerland and long-standing critic of mammography, says that this rationale doesn't take into account that cancer doesn't always metastasize at the same rate.

"The (small) effect of early detection on mortality might be due to the fact that many breast cancers are systemic from inception (and outcome cannot be altered by early local treatment), and that many are clinically benign (and early local treatment does not change the good prognosis)," says Schmidt. (The Lancet, 1992; 339: 810). In other words, breast cancer isn't a tidy disease that progresses in the same way for every woman; sometimes it spreads throughout the body, othertimes it advances in the breast alone.

Medicine in general has downplayed the importance of regular physical examination of breasts as a diagnostic tool. Dr Joan Austoker, an advisor to Britain's chief medical officer, admitted that "more than 90 per cent of breast tumours are found by the women themselves." A seven year study of 33,000 women by the Pennine Breast Screening Assessment Clinic in Huddersfield, showed that self-examination could reduce breast cancer deaths by up to one-fifth. Although some lumps detected by mammogram aren't palpable, the reverse is true as well.

If you don't want a mammogram, make sure to opt for a regular programme of self-examination (your doctor can teach you how to do it) and breast examination by your doctor. But for all women, particularly those in the limbo period between 40 and 50, the most important measure is prevention to have plenty of anti-oxidants in your diet and to avoid estrogen domination, which increasingly has been linked with breast cancer (see Second Opinion, p 12).