'Overdiagnosis Exists in Breast Cancer'

Danish Study Has Unique Data

Nick Mulcahy

January 09, 2017

Another major study has found that breast cancer screening does not substantially reduce the incidence of advanced cancer over time but does significantly increase the detection of nonadvanced cancers, resulting in the overdiagnosis of disease.

The new cohort study from Denmark is "unique" because that country had an extended period ― 17 years ― during which only 20% of the age-eligible population were invited to participate in screening, according to the authors, led by Karsten Juhl Jorgensen, MD, of the Nordic Cochrane Center in Copenhagen.

This allowed the Danish investigators, in estimating overdiagnosis, to include contemporaneous women who had not undergone screening as controls in their multipart analysis and not strictly rely on controls from an era with no screening.

The 17 years represent the longest "differential mammography access" period in any country, say the authors.

The new study also has the longest follow-up period after the start of screening of any such breast cancer study, they add.

"We need long follow-up to assess if the detection of early-stage cancers leads to the expected drop in late-stage cancers later on. We could not see a decline, even with almost 2 decades of screening," Dr Jorgensen told Medscape Medical News.

Early-stage or nonadvanced cancers were defined as cancers measuring ≤20 mm; late-stage or advanced cancers were those >20 mm.

Effective breast cancer screening should reduce the incidence of more dangerous, advanced cancers over time, explain the authors. But that was not the case in Denmark, which started offering organized biennial screening to women aged 50 to 69 years in 1991 in one region; other regions were added over time. Instead, the Danish investigators found that only the incidence of nonadvanced tumors increased "markedly" in the screening period vs the prescreening period (incidence rate ratio, 1.49; 95% confidence interval, 1.43 - 1.54).

In their new study, published online January 9 in the Annals of Internal Medicine, the authors estimate that the rate of overdiagnosis, when expressed as a relative risk, was 24.4% for invasive breast cancers and ductal carcinoma in situ and 14.7% for invasive disease alone.

The authors also observe that the average tumor diameter in Denmark decreased by 1 cm from the late 1970s to the early 1980s, "long before any screening." They acknowledge that factors other than screening can be important with respect to the size of breast cancers at detection.

Earlier studies that have reported the overdiagnosis of breast cancer have been criticized because they compared higher incidence rates during the screening era to lower rates in the years before screening began, observes Otis Brawley, MD, of the American Cancer Society, in an accompanying editorial.

"The argument is that the true incidence might be rising independent of screening," he writes.

Dr Brawley references one such recent criticism from Daniel Kopans, MD, a radiologist from Harvard University in Boston, Massachusetts (J Am Coll Radiol. 2016;13:R50-R52) who protested a new estimate of the overdiagnosis of breast cancer in the United States that was published in the New England Journal of Medicine in October 2016.

Dr Brawley says that the Danish researchers "sidestep" this concern with their study design, in which screened and unscreened populations were compared during the same period.

Notably, the results are in concert with other research, he says.

This is not to say that screening as currently practiced is without value. Dr Otis Brawley

"This is the latest of several studies using various methods and data sets to show that the phenomenon of overdiagnosis exists in breast cancer," writes Dr Brawley.

The fact of overdiagnosis "challenges the value of screening" and means that "the benefits of breast screening have been overstated," he continues, echoing a comment he made in 2009 that resulted in a huge controversy. It is unlikely that the new study will produce a controversy now, as the potential harms of screening have since been widely discussed in medical journals and at meetings.

Dr Brawley still endorses mammography screening. "This is not to say that screening as currently practiced is without value and should be abandoned," he writes.

Dr Jorgensen struck a slightly different note.

"Whether you should attend screening is a matter of balancing benefits and harms against each other, taking into consideration your own values and preferences.... The 'correct' answer for one patient may not be for the next," he said.

Dr Jorgensen also shared that he personally thinks it is time for large-scale screening programs to end. He agrees with an independent panel of experts in Switzerland, who published that opinion a few years ago, as reported by Medscape Medical News.

"Focusing our scarce resources on interventions with clear benefits for those with breast cancer, and [doing] no serious harm to healthy women, seem a much better proposition," he said.

The United States could benefit from such a resource allocation, Dr Jorgensen suggested. "Sadly, many women in the US do not get the breast cancer treatment they need," he observed.

Struggling to Accept Overdiagnosis? Look to History

In his essay, Dr Brawley defines cancer overdiagnosis as "the concept that some tissues fulfilling histologic criteria for cancer are of no threat and do not need to be cured." He acknowledges that this "is very difficult for some to accept, especially among the most ardent supporters of breast cancer screening."

So, Dr Brawley turns to history to explain how overdiagnosis works.

The biopsy definition of breast cancer was first provided by German pathologists in the mid-19th century, he writes.

Although the legacy of that definition lingers, modern technology has since radically altered how breast cancer is detected. "Because of advances in imaging and diagnostics, it is currently possible to find and biopsy lesions (some as small as 5 mm) that were not detectable 30 years ago and fulfill the 19th-century definition of cancer," he writes.

"If a lesion fits the profile of something that killed people in the past, the natural inclination today is to assume that the lesion will grow, spread, and eventually kill. However, some of these lesions may be genomically predetermined to grow no further and may even regress," Dr Brawley relates.

Oncologists are familiar with this heterogeneity, he says: "Experienced clinicians have long observed that tumors, including those within the same pathologic grade, have various biological behaviors."

Genomic tests will continue to help clinicians to evaluate the varied aggressiveness of breast tumors (and the subsequent need for some treatments, such as aggressive chemotherapy), he says. Thus, because technology is changing the definition of cancer, it is now time to accept the fact that some tumors are overdiagnosed, as they are of "no clinical threat" he says.

The study authors and Dr Brawley have disclosed no relevant financial relationships.

Ann Intern Med. Published online January 9, 2017. Abstract, Editorial

Follow Medscape senior journalist Nick Mulcahy on Twitter: @MulcahyNick

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.