Radiologists make moves to reclaim prostate imaging

Prostate imaging can be a lonely, thankless line of work for radiologists. Specialists are scarce, and urologists have the upper hand. Cancer screening is controversial, and imaging research has yielded a mixed bag of results. Nevertheless, prostate guru Dr. Ethan Halpern is bullish about the future.

Halpern and others in the relatively small circle of radiologists with prostate imaging expertise have reason for optimism. Though still in the early stages, new techniques look promising for prostate cancer detection, notably contrast-enhanced ultrasound (CEUS), elastography, and MR spectroscopy imaging (MRSI). Advances have also occurred in treatment options such as image-guided tumor ablation.

Most recently, Halpern has been experimenting with contrast enhancement and microflow imaging, a type of harmonic ultrasound technique that significantly improved the accuracy of image-guided biopsy in a small preliminary study. Researchers will apply this new approach in a major three-year National Institutes of Health trial kicking off this summer.

"Contrast-enhanced ultrasound is extremely promising and will likely revolutionize the way we do prostate biopsies," said Halpern, a professor of radiology and urology at Thomas Jefferson University in Philadelphia.

Prostate cancer biopsy has plenty of room for improvement. Typically, the prostate-specific antigen test is a tip-off for cancer. Using gray-scale transrectal ultrasound (TRUS), urologists perform a systematic sextant biopsy, which involves harvesting randomly selected cores. This technique misses up to 35% of clinically relevant cancers.

"The most critical pieces of information that we need are the precise location and extent of cancer within the prostate. I can't think of anything more important," said Dr. Patrick Walsh, urologist-in-chief at Johns Hopkins Hospital in Baltimore.

"Right now, there is no proven method I know of for making that assessment, so we need that desperately. We don't want to treat people based on unreliable information," he said.

Screening and treatment of prostate cancer are both controversial. The disease progresses slowly, yet the PSA test and ultrasound-guided biopsy enable early detection, which can lead to treatment for cancer regardless of the nature of the tumor. Complications of treatment include impotence. Urologists stress the need to diagnose patients who stand to benefit the most from early detection-those with the most aggressive cancers.

Enter advanced ultrasound. Radiologists have been performing biopsies targeted at suspicious areas with high-end color and power Doppler ultrasound systems to enable visualization of increased blood flow, a possible sign of cancer. Doppler imaging can improve the accuracy of image-guided biopsy.

Despite these advances, however, targeted biopsy with image guidance has not yet proven good enough to supplant the gold standard random biopsy. That picture could change with the use of microbubble contrast agents. Research shows that enhancement with these agents allows better visualization of the vascular pattern of tumors on gray-scale and color power Doppler ultrasound.

"If you identify microvessels in the vasculature, you do a better job finding aggressive tumors. Generally, cancers with more blood flow are more likely to be aggressive and require treatment," said Halpern, who is also codirector of the Jefferson Prostate Diagnostic Center.

Starting in 2002, Halpern led a large three-year trial partly funded by the Department of Defense to examine the effectiveness of CEUS with intermittent gray-scale harmonic imaging in prostate cancer biopsy (Figure 1). Results were published in the December 2005 issue of Cancer.

The trial showed that CEUS-guided targeted biopsies were two to three times more likely to find cancer compared with systematic sextant biopsy. But CEUS still was not sensitive enough to detect all cancers and would therefore not eliminate the need for the random approach.

Ultrasound equipment and techniques are more advanced than even a few years ago, when the DOD trial began, Halpern said. In a small study due to be presented this year, he put microflow imaging and the microbubble contrast agent Definity to the test. With this method, high-power flash pulses destroy bubbles, then low-power pulses demonstrate contrast replenishment. A composite image depicting vascular architecture is constructed through maximum intensity capture of temporal data in consecutive low-power images.

Halpern and colleagues compared the microflow technique with continuous harmonic imaging in 11 patients referred for biopsy with TRUS. Imaging was followed by targeted biopsy in areas with abnormal vasculature and then by conventional systematic biopsy. The microflow technique enabled better visualization of individual vessels and was more sensitive for cancer detection, according to the researchers.

One drawback of CEUS is that benign prostatic hyperplasia (BPH), like prostate cancer, also demonstrates greater blood flow. To counter this effect in the upcoming NIH trial, researchers will administer a 5-a reductase inhibitor, which will reduce blood flow in areas of the benign condition. The combination of contrast-enhanced imaging with 5-a reductase inhibitor is expected to further improve prostate cancer detection.

ELASTOGRAPHY OPTIONS

Elastography is another new technique that has been drawing attention. The term refers to measurement of the elastic properties of tissues, based on the well-established principle that malignant tissue is harder than benign tissue (Figure 2). A color classification system registers tissue as benign (green) or malignant (blue).

A raw ultrasound is obtained before and after a slight compression of tissue, typically achieved with an ultrasound transducer. Elastography measures and displays strain, or the change in the dimension of tissue elements at various locations in the region of interest.

Halpern performed a trial of 137 patients, in which targeted biopsy with Doppler and elastography improved prostate cancer detection but missed a substantial minority of cancers found on systematic sextant biopsy.

"Success depends on applying pressure evenly to prostate, and that is difficult to achieve with current systems. We have a ways to go toward perfecting the technique," Halpern said.

Researchers in Austria have been working to overcome the technique's limitations. To make compression easier, for example, they use a narrower region of interest, examining the right, middle, and left parts of the prostate. This technique significantly improves overall accuracy but is much more time consuming, said Dr. Ferdinand Frauscher, director of uroradiology at the Medical University Innsbruck. Frauscher shared results using elastography at the European Congress of Radiology in March.

A software technique that assigns elasticity coefficient values for the stiffness of tissues, in addition to the color values, is also proving useful. The coefficient values offer a more objective measure of malignancy than do the colors, Frauscher said.

Results of a new unpublished study of elastography in 100 patients are encouraging. Radiologists obtained elastography images and performed a targeted biopsy if an abnormality was detected. They found cancer in 36 of 100 (36%) patients with a mean PSA of 4.1. Results were compared with systematic biopsy. Elastography detected cancer in 29 subjects, and systematic biopsy found cancer in 26 subjects. Elastography had a sensitivity of 81% and specificity of 84%.

"The results are better than expected. This technique has great potential," Frauscher said.

Some cancers are still missed, however, such as those located in the anterior part of the prostate and small masses. Frauscher noted that his results would not necessarily apply across the board, because he is working with a screening population. Patients in prostate screening programs tend to be younger with smaller glands, which are easier to examine with elastography. The technique is performed only in first biopsies, where it may be more effective than in glands that have been previously biopsied.

PROSTATE SKILLS GAP

Frauscher and Halpern agree that in coming years radiologists will play a greater role in prostate imaging, because techniques like CEUS and elastography require a higher skill level. Ultrasound continues to be the workhorse for prostate imaging, although interesting work is under way with high-field MR and MR spectroscopy imaging.

For the most part, prostate ultrasound work is done by urologists. About 290,000 TRUS studies were done in 2003 in the Medicare population, and radiologists performed roughly 9% of them, according to a study from Thomas Jefferson University presented at the 2005 RSNA meeting. Radiologists performed just 1% of prostate biopsies, most of which were image-guided.

Radiologists didn't fight for prostate work and lost control of the field, said Dr. Gary Onik, director of Florida Hospital/Celebration Health's prostate cancer research program. In general, they have been apathetic about advances. Image-guided prostate tumor ablation is proving itself and becoming more established, for example, but radiologists are more active in liver ablation, even though the need is greater in the prostate.

"Radiologists need to make a commitment to start training in prostate cancer," Onik said.

The skills gap doesn't surprise some observers.

"Radiologists don't get [many] referrals. Since they do not do the procedure, they cannot build up expertise. Only a few of us somehow cracked the barrier," said Dr. Duke Bahn, chair of radiology and medical director of the Prostate Institute of America, based at Community Memorial Hospital in Ventura, CA.

There is a perception in the urology community that image-guided biopsies do not require a high level of expertise, yet the high-end color Doppler systems used by radiologists produce better results, Bahn said. In a 1998 study, Bahn and colleagues compared targeted biopsy with high-end ultrasound in 110 patients with known cancer who had asked for a second opinion. They upstaged 26% of cancers from T1/T2 (tumors confined to the prostate) to T3/T4 (tumors that are nonconfined).

"We can not only confirm the prostate cancer diagnosis, but we determine the stage of cancer and whether it is confined to the prostate," he said. "We can't pick up all the cancers, but at least we get a significant percentage with relatively accurate staging."

Most cancer patients are seen at the time of diagnosis by a number of different specialists, including radiation therapists and radiologists. But management differs significantly when it comes to prostate cancer. Typically, the urologist manages care unless or until a patient develops very bad metastases. The prostate cancer patient might get a medical oncology consultation or visit the imaging suite only in the last moments of life.

"The field of prostate cancer is not nicely shared with other medical specialists," Bahn said.

Urologists point out that the scientific literature does not support a greater role for prostate imaging. In line with National Comprehensive Cancer Network guidelines, prostate cancer patients do receive imaging in cases of extensive disease.

"There is no conspiracy here," Walsh said.

Radiologists can expect greater support for referrals and reimbursement from prominent urologists like Walsh as more evidence of the effectiveness of imaging emerges.

"If someone has a good idea, I will go knock on the doors of Congress myself," Walsh said.

Ms. Hayes is feature editor of Diagnostic Imaging.