Breast tests under microscope

Does breast cancer screening really work?
ANDREW LAXON reports on the return of a medical controversy.

Campaigners won't appreciate it but Dr Peter Gotzsche and Ole Olsen have just done their bit for breast cancer awareness month.

The Danish researchers made headlines around the world at the weekend with an article in the Lancet medical journal attacking the value of breast cancer screening.

To the frustration of health authorities in most Western countries, Gotzsche and Olsen argued there was no reliable evidence that mammograms reduced death rates.

They also said screening had led to a 20 per cent increase in mastectomies (breast removal by surgery) and a 30 per cent rise in both mastectomies and removal of tumours, many of which might be unnecessary.

Their claims - which are based on an analysis of other scientists' research - have been rejected by health services and cancer specialists in many countries, including the United States, Britain and New Zealand.

Many critics have pointed out that the Lancet article revives arguments first put forward by the two researchers from the Nordic Cochrane Centre last year.

And to outside observers the debate sometimes seems to be more about the academic freedom to publish controversial findings than the implications for women's health.

Yet the research does challenge assumptions that have prompted most Western countries to urge women in their 50s and 60s to have regular mammograms and to spend hundreds of millions of dollars on screening in the past decade.

Is breast cancer a major health problem in New Zealand?

It is the leading cause of death in women aged 55 and under. One in 10 New Zealand women will get breast cancer - one of the highest rates in the world. Each year about 1800 develop the disease and 600 die from it. In rare cases men can get it.

It usually occurs in women aged over 40. Three-quarters of the women diagnosed with breast cancer and 84 per cent of the women who die from it are 50 and over.

What can women do to avoid it?

The causes of breast cancer are not known, so women can do little to prevent it. The main risk factors are increasing age, family history - although most women who get breast cancer are not in this category - late menopause and having your first child after the age of 30.

The most common sign is a new lump in a woman's breast, which is usually painless. Before a woman reaches menopause, most lumps turn out not to be cancer. After menopause, 50 per cent are cancerous.

Can breast cancer be treated?

Most women with breast cancer end up having surgery to remove all or part of the affected breast. But other possible treatments if the cancer is detected in the early stages include radiation, chemotherapy, drugs and hormones.

To ensure breast cancer is detected as early as possible, women aged 35 to 50 are advised to check their breasts regularly and keep in touch with their doctor.

Women aged 50 and over are advised to have mammograms (breast x-rays) every two years, which are free for women aged 50 to 64.

Does the free breast screening work?

It is too early to tell in New Zealand's case. Our $22-million-a-year screening programme has been going for less than three years but there have already been problems.

In January a monitoring group identified grave concerns about missing or inadequate data, missed coverage targets - some units had not screened 70 per cent of eligible women in the first two years of operation - and delays in following up women for assessment.

The Ministry of Health replied that it was working towards the 70 per cent coverage target, which it described as "ambitious".

Screening unit manager Julia Peters said that in the first 18 months to June last year, more than 120,000 women were screened and 8000 (6.6 per cent) referred for assessment because of an abnormality. Of these, 814 were diagnosed with cancer.

But none of these statistics reveals whether the programme is cutting breast cancer deaths.

Is there any evidence from overseas?

The chairman of the monitoring group for New Zealand's breast screening programme, cancer researcher Dr Brian Cox, cites the success of Finland's programme, which achieved a 23 per cent reduction in breast cancer deaths among women between 50 and 59 after five years.

Finland appears to have had more success than Britain, which began screening in 1990 and is falling well short of its target of 25 per cent fewer deaths from breast cancer by 2000.

A study in the British Medical Journal last year showed that while death rates had dropped by 21.3 per cent between 1990 and 1998, only 6.4 per cent of this reduction could be attributed to screening.

Better treatment - especially the drug tamoxifen - and women's greater awareness of breast irregularities, which led them to seek treatment earlier, accounted for the other 14.9 per cent.

Last week, in response to the Lancet article, the National Health Service said screening saved 1250 women's lives in Britain each year.

Cox says New Zealand will not be able to make similar predictions for at least another three years.

He believes Britain's results may be disappointing because women are screened only every three years, which is sometimes not often enough to catch new cancers.

Britain's decision to introduce breast screening was based on the success of trials in Sweden, which predicted a 25 to 30 per cent fall over 10 years in breast cancer deaths among women over 50.

These results have now been attacked by Olsen and Gotzsche.

What are the criticisms of breast screening?

Olsen and Gotzsche reviewed seven of the biggest mammography trials in the world, involving 500,000 women. They found two were of medium scientific quality, three were poor and two were so flawed their results should be discounted.

The two best studies found no reduction in breast cancer deaths among women who had regular mammograms, yet the three "poor quality" studies found an average 32 per cent reduction.

Olsen and Gotzsche argued that the poor studies failed to balance the number of women who had mammograms with the numbers who did not. There were also differences between the two groups, such as socioeconomic status, which could have affected breast cancer death rates.

More seriously, the Danish pair accuse researchers of bias in deciding whether the patients in their trials died of breast cancer or some other cause.

They note the trials that claim a reduction in breast cancer death rates show no overall drop in death rates, which would be expected.

Olsen told the Washington Post: "If we look at the data we can get access to, it seems as if there is a tendency for doctors to more rarely write 'breast cancer' on the death certificates of women who have been screened but more often on those of women who have not been screened. But the total number of deaths is the same."

Are women losing their breasts when there is no risk of cancer?

The Lancet report says mastectomies have increased by 20 per cent and mastectomies and tumourectomies by 30 per cent since screening programmes began.

The authors say screening identifies some slow-growing tumours "that would never have developed into cancer in the women's remaining lifetimes".

Cox, who believes the two researchers were unfairly harsh on some studies, disagrees strongly on this claim.

He says it is rare for a breast cancer tumour not to spread and there is no way doctors could let it remain to find out. "You can't do that. If you think it's breast cancer, you're committed to treatment."

One other factor is that 1 to 2 per cent of cases spotted by mammograms are confined to ducts drawing to the nipple and will not spread to the breast.

However, these are picked up at the biopsy stage, he says, and would not lead to an unnecessary mastectomy.

So what's the argument about?

In many ways, it is about the integrity of scientific research and allegations that the health establishment suppresses news it does not want to hear.

Olsen and Gotzsche are locked in a bitter dispute with the influential Cochrane Collaboration, which has a highly regarded system of reviewing medical treatments.

The two researchers deliberately put their earlier results through the Cochrane review method to gain greater credibility, but Cochrane editors refused to accept their results without amendments.

Olsen, Gotzsche and the Lancet now accuse Cochrane editors of censorship and an attack on academic freedom.

In a furious editorial, Lancet editor Richard Horton writes " ... women should expect doctors to secure the best evidence about the value of screening mammography. At present there is no evidence from large randomised trials to support screening mammography programmes".

With the caution of a professional researcher, Cox replies that by far the majority view is that breast screening works.

"There's a possibility that the minority view is, in fact, correct. But at the moment there's sufficient evidence to persuade Governments - and I would say they're not the most easily persuaded - that it's worthwhile and they've put their money into it."

nzherald.co.nz/health