A new study suggests a hybrid scanner that unites cutting-edge 3D X-ray tomosynthesis with 3D molecular imaging tomosynthesis might help radiologists better detect hard-to-find breast cancers.

In the April issue of Radiology, scientists from the University of Virginia discussed a pilot study using their new in-house scanner. Dubbed the dual modality tomographic breast scanner (DMT), it marries three-dimensional structural and functional imaging to get results possibly several times more accurate than traditional mammography.

In X-ray tomosynthesis, a detector travels in an arc around the breast taking a series of X-ray snapshots to then render a three-dimensional image of the breast. X-ray tomosynthesis devices are currently only commercially available in Europe, but they have been generating buzz in the United States, where they might start appearing as early as next year.

"Compared with 2D mammography, X-ray tomosynthesis will likely be an improvement, but it will still supply only anatomical information," says Dr. Mark B. Williams, an associate professor of radiology, biomedical engineering and physics at the University of Virginia, and lead author of the study. "We are testing the value of ganging up anatomical and functional tomosynthesis information together."

That's why in his device, Williams teams up X-ray tomosynthesis with a new technique, molecular breast imaging tomosynthesis. In this modality, a radioactive tracer, called technetium-99m sestamibi, is first injected into the patient. Cancerous tissue eats up this substance much more than healthy tissue does, so it pools together in malignancies, where it fires off gamma rays. A small gamma camera, which captures those rays, is then moved in an arc around the breast to create a series of images. Much like in X-ray tomosynthesis, the images are combined to produce a 3D molecular picture.

While each modality is effective alone, by uniting them, the DMT scanner can line up, or co-register, the structural X-ray image with the functional molecular one. The goal is to "permit the radiologist to easily identify regions of the breast in which tracer accumulation is high and know exactly where those regions are in the X-ray tomosynthesis image," Williams says.

The pilot study

To test the accuracy of the device, last year Williams and his colleagues used the DMT to scan 17 women already scheduled for breast biopsies. They then compared the results of the scans with the results of the biopsy, the so-called gold standard for diagnosing breast cancer.

The results were encouraging: the DMT detected six of seven biopsied cancers, as well as one cancer missed by an earlier clinical exam. Even better, it was extremely specific. Every lesion it identified as benign was actually benign.

"We got better results than we expected, but I have no doubt in my mind that when we do a bigger study, those numbers are going to change," Williams tells DOTmed News.

Still, it's a promising start. The positive predictive value - or the fraction of positive findings that turn out to be true - was 100 percent in the study, a figure Williams naturally expects to be reduced during a larger clinical trial that enrolls more people. But even if it ends up at 80 or 70 percent, it would still be a huge improvement, he says. That's because in traditional mammography, the positive predictive value is about 25 percent, meaning that three out of four biopsies triggered by positive mammo findings are negative, says Williams.

The findings also compare favorably with breast MRI, generally held to be one of the most sensitive modalities. An International Breast MR Consortium study, published in the Journal of the American Medical Society, found an MRI sensitivity of 88 percent and a specificity of 68 percent. In Williams' pilot study, the DMT's sensitivity and specificity were 86 percent and 100 percent, respectively.

Road to adoption

Nonetheless, don't expect DMT scanners to replace mammography as a screening tool overnight.

While the hybrid scanner has the potential to improve detection of cancer and diagnosis, it also creates much more information than 2D mammography, and requires more time to review. At first, this could slow down a radiologist's workflow as they have to flip through many slices.

"I think there's definitely going to be a learning curve associated with going from digital mammography to digital tomosynthesis," says Williams. "But radiologists who read MRI and read CT have a much larger volume of data to handle, and they do it well, so I have no doubt that radiologists will be able to adjust to tomosynthesis, and will do it well."

Another challenge will be the transition. Radiologists check for breast cancer by comparing a current mammogram of a breast with an older one, to look for differences. That could be tricky if the older one is two-dimensional, and the new one is in 3D.

This transition resembles the one faced by radiologists making the switch from analog to digital mammography, observes Williams. To make it work, some simply compare the older film images on a light box with the newer digital ones on a monitor in the reading room. Others actually print out the digital images on laser film and hang them side by side with the older analog images on the light box.

"Different centers will take different approaches to this transition, as they did [when] going from analog to digital," Williams says.

Construction

The DMT scanner was the product of a joint effort, involving scientists and engineers from academia, industry and the government.

The University of Virginia, Dexela, a London-based device company, and the Jefferson Lab, a federal lab funded by the Department of Energy and located in Newport News, Va., all contributed to the device, Williams says.

The origins of the DMT scanner stretch back to 2000, when Williams created the first version, which combined 2D mammography with 2D molecular imaging.

About four years ago, he and his group built the first 3D tomosynthesis version of the DMT scanner, and he has been tinkering with it ever since. In 2008, he switched it for the upgraded model used in the study. Recently, he has replaced the X-ray detector with one he says is faster, reducing the scan time.

While it's going well, an obstacle in the short-term could be the ongoing shortage of the radioisotope, technetium-99m. As reported in DOTmed News, the molecular imaging industry is undergoing a crisis because two of the main nuclear reactors that create the agent are down for repairs until at least the end of the summer, if not longer.

"We're entering one of the worst periods right now, so we'll see how we do over the next few months," says Williams.