New Window on Breast Cancer

An imaging technique that may improve diagnosis

Mammograms aren't perfect, though experts agree that so far they're the best screening tool we have for breast cancer. But a new technique currently in clinical trials at Massachusetts General Hospital may improve breast imaging enormously. Called tomosynthesis, it's a three-dimensional mammogram that's the next step in digital mammography. "We read it off a computer monitor, rather than as film," explains radiologist Betty Rafferty, MD, principal clinical investigator of the Breast Tomosynthesis Project at Massachusetts General, which developed the process.

Conventional mammography has two major weaknesses: A tumor can be obscured by surrounding tissue, or criss-crossed tissue can look like a lesion and lead to false positives. "The fundamental benefit of tomosynthesis is that on a computer screen, you can remove all the surrounding structures and focus on a tiny area. You're looking at the breast in small sections, allowing you to concentrate specifically on that area." Tomosynthesis will increase both specificity and sensitivity "resulting in far fewer callbacks for another mammogram because of a false positive reading," predicts Rafferty, instructor of radiology at Harvard Medical School. In her pilot study of 60 women, Rafferty found radiologists were able to reduce false positive callbacks by 83 percent. This could affect millions of women, sparing them both the psychological stress and the added expense of more mammograms.

Results So Far

About 400 women have already taken part in clinical trials at Massachusetts General, the only site where tomosynthesis is available. Dr. Rafferty's team is exploring how clearly lesions are seen with the technology, compared to conventional mammography. In a blinded study, radiologists found they could see lesions more readily 70 percent of the time with tomosynthesis, because the 3-D images were easier to read.

In a separate study of 40 women, 18 percent of cancers were detected only by tomosynthesis. "We'll need to look at the general populace, in adequate numbers, to verify the screening potential. The pilot strongly shows that more cancers will be found," says Dr. Rafferty. General Electric has licensed the technology for commercial use and will apply for FDA approval. Tomosynthesis should be more widely available as soon as late 2004.

The noninvasive procedure uses lower radiation doses than traditional mammography, and "patients say it's more comfortable," Dr. Rafferty reports. Because each breast is compressed once (rather than twice as for a mammogram), more women can be screened each day by a particular machine. "This is the first new thing in breast imaging since magnetic resonance. We envision tomosynthesis as a screening test, replacing conventional mammography for women over 40. Won't insurance companies be happy to have a test that results in fewer callbacks and better accuracy?" she asks.

Dr. Rafferty's team will test other possibilities for tomosynthesis, such as coupling it with an injection of a contrasting agent, to get physiological data along with the three dimensions. They're also looking at the potential for computer analysis, comparing last year's images to the current ones to detect any subtle changes. Merging tomosynthesis with ultrasound would permit an instant check on whether a detected mass is merely a cyst. "Fusing the technologies would allow more specific, immediate diagnosis," says Rafferty.

Another view

"This is a very exciting development," concurs Ruth Rosenblatt, MD, Professor of Clinical Radiology at Weill Cornell Medical Center and Director of Women's Imaging at New York Presbyterian Hospital. "It's based on having the power of computer technology as it is applied to clinical studies. The concept for screening using tomosynthesis- type technology is exciting because an imaginary slice, rather than compressing tissue and looking at it in a pancake slice, allows better examination. It also makes it more comfortable for the many women who find mammography uncomfortable."

Dr. Rosenblatt's concern is affordability. "Screening is for everyone. If it remains expensive, it won't be practical. There's a big difference between technological development and everyday practice because of economics." However, she adds, "we look forward to any possibility for upgrading our equipment."