Digital mammography detects more cancers than screen film mammography

Sep 21 2005

A landmark breast cancer screening trial shows that digital mammography detected more cancers - up to 28 percent more - than screen film mammography in women 50 and younger, premenopausal and perimenopausal women, and women with dense breasts.

However, the trial, conducted by the American College of Radiology Imaging Network (ACRIN) in conjunction with the Center for Statistical Sciences at Brown Medical School, showed no difference between digital and film mammography in detecting breast cancer for the general population of women. The results are reported today in a special online publication of the New England Journal of Medicine and presented at ACRIN's fall meeting in Arlington, Va.

With 49,528 women enrolled at 33 clinical sites, including Rhode Island Hospital, the trial is one of the largest breast cancer screening studies ever performed. Brown's Center for Statistical Sciences developed the study's statistical design and analyzed the results.

"The data show that digital mammography is, on average, as good at detecting breast cancer as film mammography - and in some important subgroups of women, digital performs even better," said Constantine Gatsonis, network statistician for ACRIN, professor of community health and applied mathematics at Brown, and an author of the New England Journal article.

Because of its size and rigor, Gatsonis said, the study provides some of the best data gathered on the diagnostic accuracy of mammograms. "Neither film nor digital mammography is able to catch every cancer," Gatsonis said. "So this study data can be used to develop and improve mammography in the coming years. And that is good news for women's health."

The primary aim of the trial, dubbed the Digital Mammographic Imaging Screening Trial (DMIST), was to compare the diagnostic performance of digital and film mammography.

Women enrolled in the trial had no signs of breast cancer. Volunteers of all ages - the median age was 54 - took part. Each woman was given both digital and film screenings. Then they were randomized so that half had the digital exam first, the other half had the film exam first. Two different radiologists independently interpreted the mammograms.

Doctors determined participants' breast cancer status through follow-up mammography or available biopsy results. Biopsy information was available within 15 months of study entry. Mammograms were taken 10 months or later after study entry.

Barbara Schepps, M.D., oversaw DMIST at Rhode Island Hospital (RIH), a major trauma center for southeastern New England and the largest teaching hospital of Brown Medical School. The site enrolled 642 participants.

A radiologist and director of the Anne C. Pappas Center for Breast Imaging at RIH, Schepps said, "With this new information, we are going to strive to convert all our units to digital as soon as possible. While the study doesn't show that digital mammography benefits all women, it clearly makes a difference for younger women, a population that is known to be difficult to examine because they tend to have denser breasts.

"The important message is that all women get screened annually - whether with digital or conventional mammography," said Schepps, a clinical professor at Brown Medical School.

The study's principal investigator and senior author of the article, is Etta Pisano, M.D., the Kenan Professor of Radiology and Biomedical Engineering at the University of North Carolina-Chapel Hill. Pisano directs the Biomedical Research Imaging Center at the UNC and is a member of the UNC-Lineberger Comprehensive Cancer Center.

The National Cancer Institute funded DMIST. According to institute statistics, breast cancer is the most common non-skin cancer and the second leading cause of cancer-related death among women in the United States. An estimated 211,240 women will be diagnosed with breast cancer and an estimated 40,410 women will die of the disease in the United States in 2005.

At Brown's Center for Statistical Sciences, faculty, students, and staff develop bio-statistical methodology and conduct interdisciplinary research in medicine, public health and the biosciences. The Center is known for its expertise in diagnostic test evaluation and is home to the Biostatistics Center of ACRIN, a National Cancer Institute-sponsored clinical trials cooperative group made up of investigators from more than 100 academic and community-based facilities in the United States, as well as several abroad.