Mammography - The Hidden Downside - Part I

Mammography is the term used to describe any imaging technique used for the screening and diagnosis of breast disease - and in particular, breast cancer. There are various ways of creating a mammographic image of the breast - ultrasound, thermography, MRI, etc., but by far the commonest form of mammography used for mass screening utilizes ionizing radiation (X-rays) to detect ‘lesions’ (i.e., areas of abnormal tissue) that are suspicious for breast cancer. The terms ‘mammography’ and ‘mammogram’ as used in this article therefore refer exclusively to the X-ray imaging technique.

There is a widespread belief that screening mammography unequivocally saves lives. The National Cancer Institute, the American Cancer Society, and the American College of Radiology recommend annual mammography for all women over the age of 40. The statistic that is most commonly quoted is that by detecting breast cancer early, before it has become large enough to be clinically apparent as an obvious lump in the breast, mammography reduces the mortality rate from breast cancer by 20 to 30 percent. So fixed has this statistic become in the minds of women, the medical profession and the media that by repetition alone it has now attained the status of unimpeachable fact.

How well-founded is this belief? A closer examination of the data yields a somewhat less optimistic picture.

First of all, how much benefit can one truly expect from regular mammography? Just how effective is it in terms of saving lives?

To come to grips with that question it helps to have an understanding of the concept of absolute risk. Absolute risk is a statistical concept that expresses the number of people who can be expected to succumb to a disease over a certain period of time. Women generally perceive that their risk of developing breast cancer is very high. But in reality the absolute risk of dying from breast cancer depends on your age.

For a 60 year old woman, the chance of dying from breast cancer in the next 10 years is 9 in 1,000. Mammography screening has been estimated to reduce the absolute risk of dying from breast cancer for this 60 year old woman by around one third; i.e., instead of having an absolute risk of 9 in 1,000 over the next 10 years, her chance might at best be reduced to around 6 in 1,000 by screening. For younger women, whose absolute risk of dying of breast cancer is commensurately lower to start with (around 6 in 1,000), the reduction in risk that might be conferred by mammography would also be smaller. For 50 year olds, 10 years of regular mammography might at best be expected to reduce the absolute risk of dying from 6 in 1,000 to around 4 in 1,000.

However, advocates of screening rarely ever talk in terms of absolute risk. Instead, they prefer to express the benefits of screening in terms of relative risk, a statistical concept that, because it is expressed as a percentage, makes the benefits of screening appear much more dramatic.

For example, they will say that mammography reduces your chances of dying by 30 percent (the relative risk) but will neglect to tell you that the chance of dying of the disease (the absolute risk) is very small to start with. The relative risk, expressed as a percentage, therefore makes screening look dramatically effective, whereas when expressed in terms of absolute risk the picture is considerably less persuasive. Which of these two statements sounds more impressive: mammography saves 2 lives out of 1,000 over 10 years, or mammography reduces breast cancer deaths by 30 percent? No wonder proponents of screening are so enamored of quoting relative risk rather than absolute risk.

As Prof. Samuel Epstein, MD, of the University of Illinois and colleagues have pointed out: “Even assuming that high quality screening of a population of women between the ages of 50 and 69 would reduce breast cancer mortality by up to 25 percent, yielding a reduced relative risk of 0.75, the chances of any individual woman benefiting are remote. For women in this age group, about 4 percent are likely to develop breast cancer annually, about one in four of whom, or 1 percent overall, will die from this disease. Thus, the 0.75 relative risk applies to this 1 percent, so 99.75 percent of the women screened are unlikely to benefit” (Epstein 2001).

Finding Indolent Tumors

There are other facets of mammography that are seldom discussed by the many enthusiastic advocates of mass screening. Cancers - even breast cancers - vary greatly in their malignancy. For any screening technique to be worthwhile, it should be capable of picking up the most dangerous kinds of cancer rather than the most indolent. It should also be highly sensitive, giving few false positives and false negatives. Sadly, x-ray mammography does not score well on either count.

Mammography is undoubtedly good at picking up slow-growing cancers. It is also good at detecting so-called ‘in situ’ lesions, that is, the latent, precancerous lesions that have not yet developed - and might never develop - into truly invasive cancers. But these are not the kinds of breast cancer that are most likely to kill. That distinction belongs to the faster-growing tumors, and it is precisely these faster growing malignancies that mammography typically fails to catch.

Thus, a woman can have a clear mammogram at one annual screening, and yet, less than a year later, can discover that she has a highly aggressive form of breast cancer. Women who develop such so-called ‘interval cancers’ (i.e., cancers that are discovered in the interval between two screenings) are more than twice as likely to die as are women whose cancers are detected through routine mammography. Like most screening tests, therefore, mammography suffers from the drawback that it misses many of the deadliest cancers entirely, while zealously identifying slow-growing or latent cancers, a significant proportion of which might never progress or pose a threat to life. This accelerates the trend towards finding and curing �cancers� of dubious malignancy, thus exaggerating the benefits of both diagnosis and treatment.

Another important aspect of breast tumor growth is the ‘doubling time’ of the tumor. This is the time taken for the tumor to double in size. It has been estimated that there are approximately 40 doublings between the development of a single malignant cell and the point at which a patient dies of widely metastatic breast cancer. For a tumor to be detectable by clinical breast examination (i.e., by the human hand, feeling for a lump) the tumor needs to be around 1 centimeter in diameter (i.e., around half an inch across). A mammogram can detect a tumor at half this size, i.e., 5 millimeters in diameter. This is just one doubling less than the size at which a small tumor becomes detectable manually, by self-examination. This single doubling is probably not a sufficiently wide difference to be able to affect the overall outcome of the disease very significantly.

- Ralph W. Moss, Ph.D.

www.CancerDecisions.com