MRI prostate analysis could indicate cancer sans biopsy, reduce unnecessary procedures

An interdisciplinary team used an MRI quantification technique called "shape atlasing" to detect subtle deformations in cancerous prostate glands, finding a consistent anterior bulge among the patient cohort.

Led by Anant Madabhushi, PhD, professor at the Case Western Reserve School of Medicine, the team hopes to eventually integrate this technology into the typical prostate cancer workup, possibly avoiding thousands of unnecessary biopsies a year. Their research was published in Nature subsidiary Scientific Reports.

“People have looked at volume differences before,” said Madabhushi. “That hasn’t resulted in any usable data; there’s plenty of conditions that can cause an enlarged prostate.”

Instead, Madabhushi and his team looked at the shape of the prostate, in hopes of finding a resilient biomarker that would hold up across a variety of patients, sites and scanner manufacturers.

“We found that in specimens that had cancer, the glands tend to be arranged in a very chaotic way. But if you looked at regions that were benign, the glands tend to be systematically ordered,” said Madabhushi. “If you see chaotic architecture at the cellular level, does that potentially manifest then at a macro level, on an MRI scan?”

The researchers gathered MRI scans of 70 cancerous and healthy prostates, and aligned them into a single frame, mitigating any variation stemming from different scanners or different sites. After controlling for size, the team identified a trend of anterior deformation linked to adenocarcinoma, the most common form of prostate cancer.

This statically significant trend held up when comparing differently sized prostates and prostates imaged at different sites. They may have found the biomarker they were looking for. Madabhushi believes this technology could eventually reduce unnecessary biopsies by 10 to 20 percent.

Most patients who are getting a biopsy were flagged for high amounts of prostate-specific antigen in their blood, found in a common screening for prostate cancer. According to Madabhushi, these biopsies are shots in the dark, with no image guidance and a small chance of finding cancer.

“Our hope with this imaging is that you’d be able to say—with a high amount of confidence—that this high-PSA patient doesn’t have cancer,” he said. “You don’t have to subject the patient to a biopsy, you can just keep monitoring the PSA and have them come back in six months.”

With about a million prostate biopsies preformed a year, the economic impact could be huge. Imaging the prostate is vastly cheaper than a biopsy, which often costs several thousand dollars and only finds cancer about a fifth of the time. A fully-fledged decision support tool could save millions of dollars a year, in addition to reducing pain and anxiety that biopsy patients sometimes experience.

However, before the researchers take shape atlas process into the clinic, they need to demonstrate the technique’s value as a predictor. This paper showed marked differences in the shape of cancerous versus healthy prostates, but they need to establish its efficacy at predicting cancer in a fresh patient cohort.

Madabhushi is optimistic about the technique, with another study linking different types of deformation to less common prostate cancers currently under review.

“My sense is that we’re a couple years away from going into clinical trials and providing this tool to radiologists to use in clinical decision support," said Madabhushi.