why to/who should ask for digital mammography

[Lani]

Digital Mammography Better Than Film for Detecting Breast Cancer in Women Under 50

Richard Hyer
Information from Industry
Antibiotic Treatment for Many Outpatient and Inpatient Bacterial InfectionsLearn more about high-dose, short-course therapy, including PK/PD parameters, clinical benefits, and doctor and patient benefits.

November 27, 2006 (Chicago) — Digital mammography is significantly better than film mammography for detecting breast cancer in premenopausal or perimenopausal women with dense breasts who are younger than 50 years, according to a new subgroup analysis of the American College of Radiology Imaging Network (ACRIN) Digital Mammographic Imaging Screening Trial (DMIST).

The analysis was presented here at the 92nd scientific assembly and annual meeting of the Radiological Society of North America (RSNA).

The finding confirms and expands on results of the original DMIST trial presented last year, according to principal investigator Etta D. Pisano, MD, vice dean for academic affairs at the University of North Carolina School of Medicine at Chapel Hill. The original DMIST analysis of 42,760 women found digital mammography to be better than film mammography for this subset, in 15 data analyses (P < .0033).

"The question we hoped to answer with this new analysis is why did we find the results we did?" Dr. Pisano told the conference. "People have been asking since the first minute that we announced results, how come there is one group where there was better performance, and yet no better performance overall?"

A total of 10 additional analyses were performed to get the current data, according to Dr. Pisano, examining patients according to breast density, age (<50, ?50 but <65, and ?65 years), and menopausal category, as defined by the original DMIST trial.

The trial, conducted at 33 sites in the United States, initially enrolled 49,528 women who did not have signs of breast cancer. The women provided background health information and underwent both digital and film mammography, each with at least 2 views of each breast. Results were interpreted by 2 different radiologists. An analysis of 42,760 women from the original study population determined breast cancer status by biopsy or follow-up examination at 1 year.

[Lani]

The radiologists' interpretation of data was noted on a scale of 7 points for receiver-operating characteristic (ROC) analysis and Breast Imaging Recording and Data System (BIRADS). The study compared areas under the curve (AUC) for each of several modalities and controlled for multiple comparisons.

The only subgroup in which digital mammography was statistically better than film was pre- or perimenopausal women younger than 50 years (ROC areas, 0.79 vs 0.54; 95% confidence interval, 0.09 - 0.4; P = .0015). BIRADS-based sensitivity in this subgroup was 0.61 for digital and 0.35 for film, according to the study abstract. None of the remaining subsets showed significant difference in ROC areas.

"Because we've now done 25 analyses — 15 in the primary paper and 10 in this paper — the P value must be <.002 to be significant. The AUC favored digital in this population by .25," Dr. Pisano told the conference. No other combinations reached statistical significance, she said.

However, one group, women older than 65 with fatty breasts, approached statistical significance in favor of film over digital mammography. "These were close to significantly better for film than digital, but not statistically significantly so," said Dr. Pisano. The reasons for this are still under analysis.

As for the under-50 group, Dr. Pisano offered an opinion. "I can tell you that a preliminary look at the data suggests that contrast had something to do with it. My own hypothesis is that the contrast for digital mammography has been optimized for the dense breast and has not been optimized for the fatty breast."

"In terms of integrating new data with [existing] data, and in terms of retrievability of old images, and the improved speed of reading — all of those will be advantages of digital mammography," Dr. Rao said.

"With film mammography, one of the problems is trying to get old studies for comparison. Often, many days are spent trying to get old studies. With digital mammography, just like with any digital technique, those processes will be faster and faster, and improved," Dr. Rao pointed out.

[Lani]

"So overall, whether it's better statistically in detecting cancers in dense breasts vs fatty breasts, the word is not out yet. I don't think digital will be significantly worse than film mammography, and if everything else is the same, there will definitely be advantages to having digital," he concluded.


RSNA 92nd Scientific Assembly and Annual Meeting: Abstract SSA15-02. Presented November 26, 2006.