"Do I think 3D is ready for prime time? I really do," said senior author Emily Conant, Penn's chief of breast imaging. "It's evolving technology, but this is a big step forward."
"I definitely believe tomosynthesis is going to become the standard of care," said lead author Sarah H. Friedewald, chief of breast imaging at Advocate Lutheran General Hospital in Park Ridge, Ill. "We hope this will convince insurers to cover it so it can be available for all women."
In an editorial, two radiology researchers were more cautious, calling the results "promising."
Tomosynthesis ( tomo is Greek for "slice") acquires a set of images from an X-ray tube that sweeps over the breast. A computer algorithm instantly creates a 3D reconstruction.
A machine equipped with tomosynthesis can perform 2D, or 3D, or both during a single breast compression, so the patient perceives no difference. Combining 2D and 3D doubles the radiation dose, but it remains below safety limits. Last year, the Food and Drug Administration approved new software from manufacturer Hologic Inc. that will cut the radiation by creating 2D views from 3D data.
Tomosynthesis addresses mammography's primary limitation: When the breast is squashed, overlapping tissue can hide tiny malignancies while making normal structures appear suspicious.
"With 2D, you mostly see the covers of the book," Conant analogized. "With 3D, you can leaf through and see the pages."
The study - funded by Hologic and the National Cancer Institute - compared the performance of breast screening before and after 3D was added. For every 1,000 patients screened, the cancer detection rate increased from 4.2 cases to 5.4. The number called back for more X-rays to clarify suspicious findings fell from 107 to 91 per 1,000 screened.
The biopsy rate increased slightly, but women who underwent biopsy because of suspicious 3D images were more likely to have cancer.
Adding 3D did not increase detection of cancer confined to the milk ducts, ductal carcinoma in situ. Ductal malignancies, which account for almost a quarter of new cancer diagnoses, present a dilemma because doctors can't predict which ones will never invade surrounding tissue. Thus, these preinvasive cancers are treated - critics would say overtreated - with surgery, radiation, and, often, hormone therapy.
"The important thing about 3D is that we weren't detecting more of the 'in situ' cancers," Conant said. "What we were really impacting is invasive cancer."
Marian Forman, 66, of Philadelphia, can attest to that. With a family history of breast cancer, she was happy two years ago to be among the first patients screened with 3D at Penn. Her 2D images showed no evidence of cancer. But on 3D, there it was: a thumbnail-size mass.
"It was very deep and very early" stage, said Forman, who had a lumpectomy and radiation. "My father was an oncologist at Penn. If he had been around to see this technology help his daughter, he would have been so thrilled."
The study does not shed light on the ultimate value of screening for a disease that kills 40,000 U.S. women annually. Previous research found screening reduces breast cancer-related deaths by up to 50 percent, but expert organizations continue to debate whether to start mammography at age 40 and how often to do it.
Adding to the angst, numerous states, including Pennsylvania and New Jersey, recently passed laws requiring that women who have dense breasts - and most do - be told that mammography may miss their cancer and they may want to consider MRI or ultrasound imaging.
"The study didn't set out to address the controversy over who should be screened or when or how often," Friedewald said. "It was just trying to show tomosynthesis is a better mammogram. I think that finding is definitive."
It remains to be seen how quickly insurers agree. Some centers (not Penn) charge patients $40 to $100 extra to add 3D.
Don Liss, vice president of medical management at Independence Blue Cross, e-mailed: "We applaud the researchers for their fine work. The publication of a single retrospective study doesn't automatically lead to a change in our coverage position. However, we will consider this new study as we review our coverage policy on digital mammography with tomosynthesis."