August 25, 2008

Reducing Biopsies — Researchers Are Investigating Two New Potential Tools for Diagnosing Breast Cancer
By Beth W. Orenstein
Radiology Today
Vol. 9 No. 17 P. 26

Researchers have long been looking for ways to reduce unnecessary breast biopsies. Doing that would, in turn, relieve stress on patients and reduce healthcare costs.

Generally, when a suspicious lesion is found, doctors recommend that it be biopsied to determine whether it is malignant or benign. The sensitivity of the various imaging modalities is high, but the specificity is not. Up to 80% of breast lesions that are biopsied turn out to be benign, according to the American Cancer Society. Suspicious lesions found on screening mammography may be followed with additional mammographic views, MRI, or ultrasound.

Now, researchers at Jefferson Medical College of Thomas Jefferson University in Philadelphia are investigating contrast-enhanced subharmonic ultrasound as a noninvasive exam that could help physicians make a diagnosis. In subharmonic imaging, pulses are transmitted at one frequency, but only echoes at half that frequency are received.

Subharmonic imaging of the breast is not meant to find lesions but to characterize them once they are detected by other modalities, says Flemming Forsberg, PhD, FAIUM, FAIMBE, head of research for the Jefferson Ultrasound Research and Education Institute and an associate professor in the department of radiology. He has been working with the technique for the past eight years.

Forsberg explains that subharmonic imaging provides a higher contrast between biological tissues and blood flow because echo signals are generated only from blood containing the contrast agents. Cancer is known to have high vascularity. Using this technique, it is possible to tell whether the suspicious lesion has the vascular morphology that is often indicative of cancer, Forsberg says.

“Blood flow alone is not enough to tell you what is cancer and what is not,” Forsberg says. “We really need to look at vascular morphology, the connections between vessels, how many shunts there are, [and] what is the general behavior of these vessels. Are they nice and straight with a gentle curve, or are they completely chaotic and odd looking? The irregular pattern is what we believe indicates whether the lesion is malignant.”

In a study reported in the September 2007 issue of Radiology, researchers tested their technique on 14 women ranging in age from 37 to 66 who had 16 biopsy-proven lesions. The researchers used a GE Logiq 9 ultrasound machine that was modified to perform grayscale subharmonic imaging, transmitting at 4.4 megahertz and receiving at 2.2 megahertz. The women underwent precontrast imaging and imaging using contrast.

 The researchers used the same contrast agent that is used in echocardiography. “It is not that it is necessarily the best agent to choose for this imaging mode, but it is the safest in the eyes of the U.S. Food and Drug Administration,” Forsberg says.

 One of Forsberg’s graduate students has since developed an algorithm that enabled the researchers to account for motion. “He also figured out some criteria for what were good frames and bad frames that had to be discarded and then constructed cumulative maximum intensity images,” Forsberg says. “That’s where we are right now. We have these new processed images on top of our novel imaging modality, and we now have a new way of depicting the vascularity associated with breast masses.”

The researchers’ results using subharmonic imaging were better than conventional ultrasound and mammography. Of the 16 lesions, four were malignant. Mammography had 100% sensitivity and 20% specificity for these lesions. Subharmonic imaging had 75% sensitivity and 83% specificity for the same lesions. “Specificity was higher for all ultrasound modes than for mammography,” Forsberg says.

Forsberg believes the results were impressive. However, he says, it was a very limited study. “It’s too small a study to start talking about the test as a diagnostic tool, but at least I can say the data are very encouraging. It worked the way it was supposed to work. We could see the lesions, and we had a lot of interesting cases where we could see a lot of the vasculature in the tumors.”

One case was particularly interesting, he says. A woman had two lesions in the same area of her breast, both of which her physician thought were normal cysts. However, after looking at the subharmonic images, the physician changed her mind, designating one lesion as potentially malignant and the other as benign. Further testing showed that the subharmonic imaging was correct, and one lesion proved to be ductal carcinoma in situ and the other a cyst. “It was just one individual,” Forsberg says, “but we were very pleased with that result.”

Forsberg says it took approximately 30 to 45 minutes to scan each patient but only because of the study protocols that had to be followed. Normally, he says, subharmonic imaging would take five to 10 minutes. “The one disadvantage,” he says, “is that you do have to have an IV line inserted for the delivery of the contrast agent.”

The U.S. Department of Defense’s Breast Cancer Research Program funded Forsberg’s research in this pilot study. The group is looking for more funding from the National Institutes of Health (NIH) so it can continue to investigate subharmonic imaging’s possible role in improving the diagnosis of breast cancer.

Forsberg says that without larger studies, it is hard to say with any authority, but he believes that subharmonic imaging could provide a real improvement to breast imaging and patient care.

“Mammotomography”
Researchers at Duke University in North Carolina have developed a new scanner that they believe is better at finding early cancers in women than conventional mammography, and it can also be used for diagnosis and monitoring of therapeutic response(s). It is a hybrid between a SPECT and a CT scanner they are collectively calling mammotomography.

Martin P. Tornai, PhD, a biomedical physicist and associate professor of radiology at Duke, says the hybrid scanner he and his colleagues have built has advantages because it produces a high-resolution 3D image that doesn’t require the breast to be compressed, and it does so with lower doses of radiation—about one tenth of that delivered during mammography.

Tornai says his team has built two systems—a dedicated breast SPECT scanner and dedicated breast CT scanner—separately, but their intention all along was to combine the two so that the resulting hybrid images would provide anatomical and functional information, making it possible to not only find and characterize lesions but also determine where they are in the breast. The hybrid scanner contains both imaging modalities positioned perpendicularly to each other with respect to a shared field of view.

The researchers have done imaging observer studies using phantoms to compare x-ray digital mammography with CT. “We have been able to show a significant statistical improvement using CT compared to mammography,” Tornai says. “In mammography, you lose a lot of information because you have only a 2D image. With the 3D image that the SPECT/CT scanner produces, lesions become more conspicuous because overlapping tissues are removed. In contrast to x-ray tomosynthesis, a pseudo-3D x-ray imaging modality, the SPECT/CT system produces a uniform 3D image and does not require any breast compression.”

The hybrid scanner that the researchers have built from novel configurations of conventional equipment circles the breast as the patient lies on a specially built table.

No one has wanted to use clinical CT for breast screening because of the high radiation dose thought to be required, says Tornai. However, he says, “With our scanner, the dose is significantly lower because we use higher energy than mammography and highly modulated x-rays. They are highly monochromatic. Thus, we have been able to acquire a 3D mammotomogram at a dose equal to or less than a screening mammogram that results in a road map which is basically an anatomical framework for what is going on.”

Tornai says he also has been able to “take advantage of the physics of the modified monochromatic x-ray spectrum … that allows the CT images to be a bit more quantitative. So instead of just looking at the different colors on the screen that tell you white is skin, black is air, gray is fat, and you’ve got connective tissue on the inside, the really different shades and the numerical values associated with those shades can tell you whether or not the values are consistent with the known internal breast components, including cancer.”

 The scanner is also able to see areas, including the chest wall, that traditional mammograms may not, Tornai says. By registering and fusing the images from the nuclear medicine SPECT scan and CT, the radiologist can see exactly what’s going on and where, Tornai says.

 He says he initially built the scanner as two separate pieces of equipment because it was easier that way. “It was easier to ensure they work the way we envisioned and then to initiate patient studies and then to put the two together. That was the intent all along in my NIH-funded research,” he says.

 Tornai has tested the hybrid scanner extensively with phantoms and has begun successfully scanning subjects with known cancer. He says it could take anywhere from five to 10 years to commercialize the scanner because of what is necessary to get more funding and obtaining FDA approval.

One of Tornai’s goals in designing the scanner was that it be comfortable. Many women, he says, do not undergo screening mammography because the compression is painful. He thinks more women might be amenable to getting their annual breast exams if it were a more comfortable procedure.

 Because SPECT requires IV injection of imaging agents, the SPECT portion of the scanner would not likely be broadly used for routine screening mammography, Tornai says. However, he says, if it proves to be more effective, the hybrid SPECT/CT system might be especially helpful for women who are at high risk for developing breast cancer because of familial history or a genetic predisposition.

Breast SPECT/CT also could be used for women with dense breasts or implants because mammography is known to miss up to 25% of cancers in these women, Tornai says. The scanner and associated imaging procedure also should be less costly to employ than MRI, he adds.

— Beth W. Orenstein is a freelance medical writer and a regular contributor to Radiology Today. She writes from her home in Northampton, Pa.