Breast tomosynthesis on verge of U.S. approval, but questions linger

November 24, 2010
by Brendon Nafziger, DOTmed News Associate Editor
By Brendon Nafziger and Olga Deshchenko, DOTmed News Reporters

Hologic Inc.'s breast tomosynthesis device received near-final U.S. Food and Drug Administration approval, and is poised to become the first such device to be sold in America.

The 3-D breast scanning technology aims to improve detection of cancer by revealing cancerous masses hidden by overlying breast tissue, but some experts say not enough published data exists to determine how useful the technology will be.

Plus, there are concerns about how it will fit into existing workflows.

"We do have a responsibility to be very careful when we're introducing new technologies, that we have the data to support that it is better for patients," said Dr. Wendie Berg, a breast imaging consultant with the American College of Radiology Imaging Network in Maryland.

Bedford, Mass.-based Hologic said Tuesday it received an "approvable" letter for the Selenia Dimensions 3-D digital mammography system from the FDA. This means final approval will come once the agency makes sure the company's manufacturing plant is up to snuff.

The technology is already available in Europe from Hologic and other manufacturers, including Siemens Healthcare. It works by spinning around the breast in an arc and taking multiple snapshots, then reconstructing all the slices into a 3-D image.

Proponents of the technology tout improvements over 2-D digital mammography in detecting cancerous masses in breasts that could be obscured by overlapping breast tissue.

"Because you can, in effect, see through tissues on top of other structures, that makes it easier to see the edges of masses," Berg said. She said based on data presented by Massachusetts General Hospital, but never published, the technology could improve cancer detection by about 10 percent.

Also, the hope is that the technology could reduce the recall rate - that is, the number of women called back for a follow-up exam based on ambiguous findings.

Berg said recall rates vary by doctors, centers and patient populations, but early research suggests with tomosynthesis the rate could fall from about 10 percent to around 6 percent. "It's an important decrease," she said, as the harms due to false positives were one of the issues raised by a controversial government task force on mammography last year.

Dr. Martine Backenstoss, section chief of mammography with Rochester General Hospital in New York, hopes 3-D technology enables radiologists to forego spot compression imaging because it displays the breasts in slices.

"Right now, we're looking at the forest and sometimes what we need to do, is look at the tree because that's where you have pathology," said Backenstoss.

"Certainly the intent of 3-D images is to reveal the inner architecture of the breast, which you can't always see satisfactorily on single plane images because you have so much superimposed fibroglandular tissue," she added.

The early promise of the technology has led to some speculation that the breast tomosynthesis device could generate $1 billion in sales, according to Reuters. Already, Hologic's shares have risen more than 4 percent since an FDA panel in September recommended clearing the device. But despite the promise, serious questions remain.

Berg said some evidence suggests tomosynthesis could actually lag behind standard digital mammography in demonstrating cancerous calcifications, as radiologists will have to look through multiple slices together to perceive the lesions and the longer acquisition time for the 3-D images could result in motion blur.

"Again, we don't have a lot of published information on this, but there certainly seems to be a concern that we won't see malignant calcifications as well," she said. "We may be missing some of the early cancers if we were to switch to tomosynthesis alone."

While protocols for screening needs to still be worked out, this could lead radiologists to take one tomosynthesis image and one standard 2-D mammogram, which "may actually make things more complicated," Berg said.

"We need more papers published," she said. "There's just very little published."

Workflow issues also remain to be worked out, as the images contain much more data than 2-D mammograms. Radiologists might have to spend more time reading exams.

"I'm told you can get relatively faster at it, but I can't imagine it's not going to take you longer than it takes now. But how much longer, I don't know," Berg said.

More data means the files are, relatively speaking, huge. Currently, mammograms average around 40 megabytes per image. But for tomosynthesis, it could be as high as 1 gigabyte per acquisition. And radiologists might want to take two sets of tomosyntheses of each breast. Vendors and providers will have to work out the best way to load up images before exams. Already, Berg said with the smaller-sized files patients can be forced to wait if there are problems calling up studies the night before.

But some factors might not be a problem. Despite early worries, Berg said that the published radiation doses only slightly surpassed the 0.4 mSv dose from a digital mammogram. "Now it's in the same ballpark," she said.

Still, many radiologists are excited. Rochester General Hospital acquired Hologic's digital mammography system about three months ago with hopes of upgrading it to 3-D technology once it's approved in the U.S.

"I have not seen a lot of data on the technique but we've purchased the machine because I thought that going forward it would certainly be a very interesting tool to use in the detection of breast cancer," Backenstoss said.

But as it will likely take radiologists longer to look through the tomosynthesis images, she said it's unclear whether it would be used selectively or across the board on all patients.

Plus, radiologists can only speculate what reimbursement policies are in the cards for the 3-D mammography system. For facilities like Rochester General Hospital, even the cost of an upgrade to the latest technology remains a mystery.

"We were trying to get the information from Hologic and they weren't able to tell us," said Backenstoss.

But before any of that, the FDA still has to give the device its final stamp of approval.

"This is only part one; we're waiting for part two," Backenstoss said.

Hologic did not respond to requests for comments by press time.