Breast Cancer Overdiagnosis: A New Estimate in NEJM

Nick Mulcahy

October 12, 2016

The overdiagnosis of breast cancer via mammography screening is "larger than is generally recognized," conclude the authors of a new analysis published online October 12 in the New England Journal of Medicine.

An overdiagnosis of breast cancer refers to a tumor detected on screening that "never would have led to clinical symptoms," explain the investigators, led by H. Gilbert Welch, MD, professor of medicine at the Dartmouth Institute for Health Policy and Clinical Practice in Lebanon, New Hampshire.

"I think the main message [of the study] is that screening has both benefits and harms. There is no single 'right answer' ― values matter. Screening is a choice, not a public health imperative," Dr Welch told Medscape Medical News.

Overdiagnosis is considered one of the harms of screening, but it is not easily evaluated using clinical trial data because of the need for long-term patient follow-up, the study authors say.

So Dr Welch and colleagues used population-based data from the Surveillance, Epidemiology, and End Results (SEER) program for 1975 through 2012 to calculate the tumor-size distribution and size-specific incidence of breast cancer among women aged 40 years or older.

The authors acknowledge that tumor size is now not as important as a tumor's biological characteristics in breast cancer prognostics.

But they explain that screening mammography "is not an assessment of functional gene expression; rather, it is an anatomy-based search for small structural abnormalities."

The goal of mammography screening, say the authors, is to detect small malignancies before "they grow large enough to cause symptoms.

"Effective screening should therefore lead to the detection of a greater number of small tumors, followed by fewer large tumors over time," they state.

And that is what the new study found.

After the advent of widespread screening mammography, the proportion of detected breast tumors that were small (invasive tumors measuring <2 cm or in situ carcinomas) increased from 36% to 68%; the proportion of detected tumors that were large (invasive tumors measuring ≥2 cm) decreased from 64% to 32%.

But there was a problem.

The authors say that this trend was "less the result of a substantial decrease in the incidence of large tumors" and "more the result of a substantial increase in the detection of small tumors."

This was seen when data from women who received a diagnosis of breast cancer in the late 1970s were compared with data from those who received a diagnosis in the early 2000s: the incidence of large tumors decreased by 30 cases of cancer per 100,000 women, and the incidence of small tumors increased by 162 cases of cancer per 100,000 women.

The study authors explain that overdiagnosis was in evidence because only 30 of the 162 additional small tumors per 100,000 women that were diagnosed would be expected to become large.

Joann Elmore, MD, MPH, a professor of medicine and an epidemiologist at the University of Washington in Seattle, agrees.

"These data suggest extensive overdiagnosis of small tumors," she writes in an editorial that accompanies the study.

These data suggest extensive overdiagnosis of small tumors. Dr Joann Elmore

The new study is important because it adds to the growing literature on overdiagnosis, said Dr Elmore.

"If this were the only study on this topic, we might say, 'What an interesting hypothesis,' " she told Medscape Medical News. "However, now there are multiple publications using different study designs and statistical approaches describing possible overdiagnosis."

Dr Elmore says that Dr Welch and colleagues present "powerful data" but that "they also rely on data with extensive missing values, make assumptions about underlying disease burden that cannot be verified, and acknowledge that their estimates are imprecise."

A radiology expert believes that these issues should have disqualified the study for publication in the NEJM.

"This has been a pattern at the NEJM for years, where they publish papers regarding breast cancer screening that have major scientific flaws," said Daniel Kopans, MD, a professor of radiology at Harvard Medical School in Boston, who was asked for comment.

The Underlying Incidence of Breast Cancer

Dr Kopans focused on one of the study's assumptions ― that the underlying burden of clinically meaningful breast cancer was unchanged over time. This assumption is important because it supports the conclusion of overdiagnosis.

However, Dr Kopans explained that the ongoing rise in detected breast cancers during the study period is explained by a growing incidence of invasive breast cancer ― not overdiagnosis.

"In fact, the incidence of invasive breast cancer had been increasing by 1% each year from 1940 to 1980 prior to any national screening," he told Medscape Medical News, referring to data from the Connecticut Tumor Registry and SEER adapted from previously published research.

Furthermore, the authors' assumption that the underlying burden of clinically meaningful breast cancer was unchanged during the study period means that the incidence of cancer would have been a "flat line" — the same in 2012 as in 1977, he explained.

Dr Kopans stated that the "actual increase" was 1% per year — which is similar to what is seen in the new study's prescreening period, from 1979 to 1982, before widespread mammography was employed.

"If the correct extrapolation is used, there is no overdiagnosis," he concluded.

If the correct extrapolation is used, there is no overdiagnosis. Dr Daniel Kopans

He also said that the rate of metastatic disease "should have increased at the same rate as the other invasive cancers."

In their discussion of the study, Dr Welch and his coauthors address "proponents of screening mammography," such as Dr Kopans, and their contention, expressed in other publications, that the increase in breast cancer instead "reflects genuine disease and that overdiagnosis has been greatly exaggerated.

"Those who postulate such substantial increases in underlying incidence, however, must explain why the increase coincides temporally with the introduction of screening, why the incidence of the most aggressive form of the disease — metastatic breast cancer — remains essentially unchanged," write Dr Welch and colleagues.

Although acknowledging the new study's limitations, editorialist Dr Elmore argues that overdiagnosis of breast cancer is real and, furthermore, needs to be addressed.

Dr Elmore provides a long list of would-be action items in her editorial, including the need to "move toward targeted screening of persons who are at a higher risk for developing cancer, rather than broadly screening the general population, in which most people have a lower risk for developing harmful cancer."

Risk-based, personalized screening has been championed by other experts, including the authors of a 2014 meta-analysis of 50 years of breast cancer screening data, as reported by Medscape Medical News.

The new study in the NEJM also determined the size-specific cancer case fatality rate for two periods: a baseline period (1975 through 1979), and a period encompassing the most recent years for which 10 years of follow-up data were available (2000 through 2002).

Dr Welch can be seen and heard discussing these and other study findings here:

Dr Welch and Dr Elmore have disclosed no relevant financial relationships. Dr Kopans has received research support from GE Healthcare.

N Engl J Med. 2016;375:1438-47, 1484-86. 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.