Chat with us in Facebook Messenger. Find out what's happening in the world as it unfolds.

'Overdiagnosis' of breast cancer may be higher than thought

Dr. Otis Brawley

Story highlights

  • Researchers suggest overdiagnosis rate of breast cancer is 18%-25%
  • Overdiagnosis refers to a tumor that's found but that never would have caused harm
  • 21st-century definition of cancer will probably involve genetics and genomics, expert says
Q: The journal Annals of Internal Medicine has an article in it this week that talks about the "overdiagnosis" of breast cancer. What is that?
A: I believe many will consider this one of the most interesting medical articles of the year.
In the study, Norwegian epidemiologists describe how they compared breast cancer incidence rates and mortality rates over a period of years -- once for a large group of women who had routine mammographic screening and another time for a group who did not.
The researchers concluded that the overdiagnosis rate of breast cancer in Norway is between 18% (one in six) and 25% (one in four). (The accompanying editorial suggests that since women in the United States start getting mammograms at age 40, not age 50 as in this Scandinavian country, the overdiagnosis rate may be higher in this country.)
"Overdiagnosis" is a term many are unfamiliar with. In cancer medicine, it refers to a tumor that fulfills all laboratory criteria to be called cancer but, if left alone, would never cause harm. This is a tumor that will not continue to grow, spread and kill. It is a tumor that can be cured with treatment but does not need to be treated and/or cured.
Many clinicians reject the idea that there is such a thing as a cancer that does not kill and does not need treatment. Indeed, it is impossible for a physician to tell whether an individual patient with a small localized cancer has an overdiagnosis cancer. On the other hand, most who study epidemiology and cancer treatment outcomes have accepted that overdiagnosis is a problem in breast cancer.
The difference between the view of the clinician and the view of the epidemiologist are understandable. The clinician sees the patient, and the epidemiologist sees the population as a whole. The old "tree vs. forest" metaphor applies well. Even so, most epidemiologists believed overdiagnosis represented less than 10% of all breast cancers diagnosed.
Overdiagnosis clearly exists in oncology. It is estimated to represent up to 60% of diagnosed prostate cancers. Indeed, it is a major reason that prostate cancer screening is so controversial. It also occurs in thyroid and lung cancer in rates that are unknown.
The problem is related to advances in technology. German pathologists developed our definitions of cancer in the 1840s. They used a light microscope to look at biopsies from people who had died of cancer and drew pictures that, to this day, define breast cancer, colon cancer, lung cancer and other malignancies.
Now, 170 years later, a small tumor can be found using radiologic imaging, technology that was developed in the past few decades. That tumor is then biopsied with a fairly new needle technology. A pathologist looking at it under a microscope then identifies it as fitting the profile of what the Germans called cancer.
What we do in medicine is tumor profiling. The tumor that the German pathologists described in the 1840s was clearly a cancer that killed. But the small tumors found in today's screenings look like bad tumors but may not be genetically programmed to behave badly.
I believe what we need is a 21st-century definition of cancer.
The 21st-century definition of cancer will probably involve genetics and genomics rather than a profile of what cancer looks like under the microscope. We will want to assess a tumor and know which genes are mutated, which genes are turned on and which are turned off. Genomic testing is in its infancy.
So what's a woman to do now? I, and most physicians, still recommend that women get a high quality mammogram and clinical breast examination by a health care provider every year starting at age 40.
Yet women do need to understand that mammography screening is imperfect and has significant limitations. Also, those who think they have a breast mass at any age should get it evaluated by a physician.
A woman who is diagnosed with cancer should get it treated appropriately. Unlike the case of prostate cancer, observation of an early breast cancer mass is not appropriate therapy at this time; we have definite proof that, while we may treat some women who do not need to be treated, we definitely cure many women who need to be cured.
There has been a greater than 30% decrease in breast cancer mortality or decreased risk of dying from breast cancer since 1991. This translates into a 30% decreased risk of death from breast cancer since 1991. Another way of looking at our success: This year, approximately 39,500 women will die of breast cancer, but we have good evidence that the screening and treatment of the past decade will prevent at least 16,000 breast cancer deaths.