This article considers some recent news items about breast cancer screening. Screening for cancer seems to be a very reasonable thing to do. Test people before they get cancer and if they test positive then give them early treatment and avoid all the terrible effects of advanced cancer- if only it were this simple.
The main problem with screening is that there are as yet no perfect screening tests. A perfect test would detect all people who have cancer (it would have 100% ) and it would always be negative in people who do not have cancer (it would have 100% ). No test yet is both 100% sensitive and 100% specific. However it has become apparent that tests need to be even better than this to be completely successful. We now know that cancer comes in a variety of forms. There is not just one kind of breast cancer or one kind of prostate cancer for example. A significant proportion of early cancers either disappear, do not progress, or progress so slowly that they never cause any harm to the person, who may go on to have a normal and healthy life before finally dying of something else. Thus the perfect test will not only be 100% sensitive and specific, it will also be able to discriminate between dangerous and non-dangerous cancers. This is almost certainly too much to ask of just a single test. It just might be possible in the future to use a strategy that uses sequential tests. We could use a 100% sensitive test to find all the cancers then use a different test to sort the bad ones from the harmless ones. Unfortunately we can’t do this for any of the common cancers at present.
Let’s look at breast cancer screening to see what happens in real life. The Australian BreastScreen program saves hundreds of lives every year. The BreastScreen Australia program has been associated with a 21 to 28 per cent decline in the number of deaths among women aged 50 to 69 with breast cancer. This wonderful result is due to both the screening program and advances in treatment of breast cancer. However epidemiologist, Alexandra Barratt, and academic and general practitioner, Paul Glasziou, writing in the Medical Journal of Australia and reported in The Sydney Morning Herald, point to a recent Cochrane Review which found that for every 2000 women screened in a decade, one life would be saved and 10 healthy women would have unnecessary treatment. Most of the women who test positive on the mammogram screen and who have a biopsy, will have benign lumps, not breast cancer.
Some good news on this front is that the rate of false-positive mammograms in Australia is among the lowest in the world and two-thirds lower than in the US, Australian research shows. The study was presented last month at the Sydney International Breast Cancer Congress. In 50-year-old women there were 30 false-positives for 1000 screened, compared to 87 per 1000 in the US Preventive Services Task Force systematic review of breast screening said researcher Professor John Boyages. The cumulative risk of a woman being recalled at least once over 20 years was 18.6% for a woman aged 50 not using HRT and having a mammogram every couple of years. In the US it was 50–60% in private settings and in European public settings the rate ranged from 20–30%. Professor Boyages, chair of the conference and director of the Macquarie University Cancer Institute, said over 20 years the chance of a woman having a complicated biopsy causing significant stress was about 4.5%.
One more issue to consider is that researchers have suggested that about a third of cancers detected by the BreastScreen program would not cause harm if left untreated.
This means that, in order for mammography screening to help those women who are found to actually have dangerous breast cancers, many other women are potentially being diagnosed and treated for breast cancer unnecessarily in order to benefit the other two-thirds. Thus as well as the financial cost of screening and treatment, there is cost in human inconvenience and suffering due to unnecessary treatment. This is not due to the design of the screening program, but is simply the result of the natural limitations of using current tests for screening purposes. Drs Barratt and Glasziou feel that women need to be made aware of this problem and give truly informed consent before they undergo breast cancer screening.
This same problem occurs in prostate cancer screening in men and has been the subject of a previous news item on this site.
Can we develop better tests? There are a growing number of exciting new developments that suggest we may well have better tests that may be able to be applied to cancer screening. Researchers at Kansas State University have recently announced the development of a test that preliminary testing shows can detect early stage as well as advanced breast cancer. The test works by detecting increased enzyme activity in the body. Iron nanoparticles coated with amino acids and a dye is introduced to small amounts of blood or urine from a patient. The amino acids and dye interact with enzymes in the patient's urine or blood sample. Each type of cancer produces a specific enzyme pattern, or signature, that can be identified by the laboratory.
Elsewhere a team from the University of Leicester and Imperial College London is about to start a clinical study in the UK’s largest breast screening clinic at Charing Cross Hospital, London. Investigator Professor Charles Coombes, from Imperial College London, says that “When a woman has breast cancer we can tell by the DNA in their blood.” The researchers hope that this simple blood test could in future replace the mammogram to screen for breast cancer.
Some perspective on these new tests is given by Dr. Judy Garber quoted in the Harvard Health Blog. She was referring to the findings by the Kansas State University team but her comments are applicable to both studies. “These findings are very exciting, but extremely preliminary,” says Dr. Judy Garber, director of the Center for Cancer Genetics and Prevention at the Dana Farber Cancer Institute, and professor of medicine at Harvard Medical School.
For example, what happens if the test does identify a very tiny, very early cancer? Will a follow-up CT or MRI scan be able to detect it to pinpoint its location? If not, will doctors still need to wait until the cancer grows before they can remove it? Also, will the test lead to real increases in survival, or will patients who test positive be subjected to unnecessary tests and treatments for an early-stage cancer that might never have spread?
“The bottom line is that screening for early cancer in healthy people must be done with great care and tests must show not only that they can detect cancer early, but also that it makes a difference when they do,” Dr. Garber says.