Kopans calls for withdrawal of NEJM study linking mammo, overdiagnosis

An analysis published Nov. 22, 2012, in New England Journal of Medicine, that estimated more than 30 percent of all breast cancers diagnosed in 2008 represented overdiagnosis, should be retracted, according to Daniel B. Kopans, MD, director of breast imaging at Massachusetts General Hospital in Boston, in a point-counterpoint exchange published in the May issue of the Journal of American College of Radiology.

David Seidenwurm, MD, of Radiological Associates of Sacramento in California, disputed Kopans’ assertion and argued that the article was correct and should not be withdrawn.

Screening mammography does not cause overdiagnosis or overtreatment, according to Kopans. “Pathologists cannot distinguish cancers that will be lethal from those that do not require treatment, and oncologists cannot determine who will be cured without systemic treatment and who will be ‘overtreated.’”

Breast cancer treatment, like medicine, is inexact, which is not cause to deny access to screening and its lifesaving potential, Kopans continued. He pointed to the near universal agreement among professional societies regarding the link between screening and the reduced breast cancer death rate.

The analysts who authored the NEJM paper, Archie Bleyer, MD, of Oregon Health and Science University in Portland, and H. Gilbert Welch, MD, from Geisel School of Medicine at Dartmouth in Hanover, N.H., have no experience treating women with breast cancer and based their argument on “assumptions and estimates,” Kopans wrote. In addition, they grouped together DCIS and small invasive cancers, which muddy the results for small invasive cancers.

Bleyer and Welch underestimated the background increase in invasive cancers, which swelled from 60 per 100,000 women in 1940 to 128 per 100,000 in 2008, according to Kopans. The miscalculation misrepresents the incidence of overdiagnosis, he continued.

Kopans also took issue with the authors’ assertion that screening is not effective unless it lowers the rate of advanced cancers. “Women die from breast cancers diagnosed at all stages. Finding cancers at smaller sizes within stages saves lives,” he wrote.

In contrast, Seidenwurm supported Bleyer and Welch. “The data show that early-stage breast cancer has increased markedly, while late stage diagnosis has decreased only slightly, and that decrease is in highly treatable nodal disease, not metastases.”

This small reduction in late-stage disease, according to Seidenwurm, suggests minimal screening benefits and greater harms of overdiagnosis and overtreatment. He also attributed improved breast cancer outcomes to advances in surgery, radiation, chemotherapy and hormonal treatment.

Seidenwurm took issue with developments in breast imaging, offering that advanced imaging detects lesions occult to conventional mammography. At the same time, the biopsy threshold has lowered, upping the number of biopsies; and advances in pathology have changed the definition of disease. Outcomes include a high recall rate and overdiagnosis.

The proportion of DCIS in the U.S., which has increased to 20 percent of breast cancers in the U.S., poses two problems, according to Seidenwurm. “DCIS has increased more in the United States than in other countries, and it is treated more aggressively here.” However, he acknowledged that physicians face poor choices with DCIS; underdiagnose and undertreat or overdiagnose and overtreat.

DCIS may be the common ground between Kopans and Seidenwurm. “It is well known that diagnosis and treatment of DCIS is highly controversial. Efforts have been under way for decades to try to find a reasonable approach to these lesions,” Kopans wrote.

Seidenwurm closed with a call to arms and recommended observational studies focused on procedures and therapies; application of performance metrics such as surgical complications and comparative effectiveness studies. In addition to improving screening programs, he urged physicians to accept uncertainty about the value of screening at different ages and development of pathways to help women make informed personal decisions.

Kopans reiterated that screening mammography saves lives and also called for accurate dissemination regarding the risks and benefits of screening. “Probably at its best, it [mammography] can reduce the death rate by 40% to 50%. Clearly, a universal cure or safe prevention is needed, but neither is on the horizon,” he admitted.