Health & wellness

Mammograms may overdiagnose up to 1 in 4 breast cancers, Harvard study finds

Getting regular mammograms to screen for breast cancer can reduce a woman’s likelihood of dying from the disease, but 15 to 25 percent of invasive cancers detected on the X-rays never would have turned life-threatening, according to a study published Monday in the Annals of Internal Medicine.

Harvard School of Public Health researchers teamed up with scientists from Norway and Sweden to examine medical records from nearly 40,000 patients in Norway from 1996 to 2005 and compare breast cancer incidence and death rates in communities where mammography screening was offered every two years for women ages 50 to 69 and in those places where it wasn’t offered.

After taking into account a two- to five-year delay in the diagnosis of breast cancer in women who weren’t screened, the researchers found that women who had regular mammograms were 52 to 59 percent more likely to be diagnosed with breast cancer over the course of a decade compared with those who weren’t screened.


But screening afforded women only a slightly smaller risk of dying from breast cancer and didn’t appear to reduce the total number of deaths from any cause. That led the researchers to conclude that many of the breast cancers detected via mammograms were slow-growing and non-aggressive, or perhaps even capable of vanishing on their own without any treatment.

Get The Weekender in your inbox:
The Globe's top picks for what to see and do each weekend, in Boston and beyond.
Thank you for signing up! Sign up for more newsletters here

“Once a woman is diagnosed with breast cancer, she has to be treated since we don’t know which one of these cancers has the potential to kill her and which ones don’t,” said study leader Mette Kalager, a visiting scientist at Harvard and researcher at the Telemark Hospital in Norway. “But I think women should be informed about the possibility of overdiagnosis beforehand, so they can decide for themselves whether or not to be screened.”

For every 2,500 women screened in the study, 26 to 30 were diagnosed with breast cancer over 10 years (after adjustments were made to account for earlier detection) compared with 20 who weren’t screened. One woman per 2,500 in the screening group died of breast cancer compared with two women in the group that wasn’t screened, but there was a total of 15 deaths per 2,500 from any cause, including breast cancer, in each group.

In recent years, mammograms have come under fire for being a crude screening tool. At least half of women who get screened have a false finding that turns out not to be cancer, necessitating unnecessary followup testing that might include a biopsy. The X-rays also miss some cancers until after they’ve spread throughout the body. “Now we’re moving on to a new under-appreciated harm, which is overdiagnosis,” said Dr. Joann G. Elmore, a professor of medicine at the University of Washington School of Medicine who wrote an editorial that accompanied the study.

While experts agree that some breast cancers are overdiagnosed and overtreated, the notion that women should consider skipping mammograms out of concerns about finding harmless cancers is bound to be controversial.


“I think most women would say they’re willing to risk overdiagnosis and overtreatment as a tradeoff for the potential lifesaving benefits,” said Dr. Daniel Kopans, a breast imaging radiologist and mammography researcher at Massachusetts General Hospital.

But studies don’t agree on just how much mammograms reduce breast cancer deaths. Some suggest the screening’s early detection cuts the death rate by about half, while others find it reduces deaths by about one-third or -- in women screened and diagnosed in their 40s -- by as little as one-tenth.

Regardless, Kopans said he believes the latest study finding “is really questionable in terms of its science” and runs counter to older randomized trials that measured an overdiagnosis rate of about 1 to 10 percent.

The research does, though, point to the need for further studies to help oncologists determine which breast cancers can be watched for a while and which require immediate treatment -- an approach now being taken with prostate cancer.

Researchers are now conducting studies using a watchful waiting approach with tiny “in situ” breast cancers that haven’t spread into the surrounding tissue to see whether they can learn to detect those that could turn deadly.


“We didn’t include these cancers in our study,” said Kalager. “If we did, we would have found an even higher estimate of over-diagnosis since some studies suggest fewer than 50 percent of these in situ growths go on to become invasive cancer.”

Deborah Kotz can be reached at Follow her on Twitter @debkotz2.