Can Multiparametric MRI Make Prostate Biopsies Obsolete?

No one looks forward to a prostate biopsy, but it is still the only way to obtain a definitive diagnosis of prostate cancer (PCa). This is because currently, a visual microscopic examination of tissue is the most reliable way to identify the presence of cancer cells. On the other hand, needle biopsies –especially those guided by transrectal ultrasound (TRUS)—are notoriously inaccurate because they can often overdetect insignificant PCa, underdetect significant PCa, or miss cancer altogether due to their hit-or-miss nature.

Looking for a better way

All of this may soon change, thanks to a new study being launched under the direction of Professor Mark Emberton (University College London). The study is under the auspices of Britain’s Medical Research Council (MRC) as part of their Stratified Medicine Initiative. This initiative is investing £16.8 million to create innovative approaches to treating PCa, kidney disease, alcoholic hepatitis and childhood arthritis. The portion allocated to PCa is £4.1m, enriched by another £1m from Cancer Research UK. Investigators plan to enroll 1,000 men with medium- to high-risk PCa, with a goal to ultimately develop alternatives to needle biopsy that would provide comprehensive diagnostic details in order to match the treatment to the patient—including no need for immediate treatment (Active Surveillance).

The study will also enroll 300 men (ages 6075) who have never had a PSA test, and administer mpMRI scans to determine how well it detects abnormal prostate conditions.

What’s unique about this study?

As the study is designed, it is ambitious because of its complex agenda which addresses 2 interrelated questions:

  1. Can mpMRI scans be used to screen men more accurately than the PSA blood test?
  2. Can the combination of mpMRI scans, genomics, and machine learning (e.g. training computers to take over detection/diagnostic tasks) replace needle biopsies?

According to an MRC bulletin, PCa patients who are acting as study advisors “…say they are particularly excited by the prospect of large reduction in biopsies, as they have serious side effects in the majority of patients, which include pain, bleeding, infections leading to sepsis, and urine retention…”[i]

In one sense, this study isn’t unique because it expands upon earlier work by Prof. Emberton and colleagues (see my blog about the PROMIS study). Published in Feb. 2017, their results show that for men with a suspicious PSA, multiparametric MRI before biopsy could avoid 27% of biopsies. What’s more, using mpMRI to target biopsies to the suspicious area, such as we do at the Sperling Prostate Center, found 18% more significant PCa than TRUS biopsy.

A vision of improving men’s health

Prof. Emberton described the shared vision of this collaborative research:

Our recent studies have begun to show how MRI technology will transform prostate cancer screening and diagnosis. Now we’re starting an ambitious new study, to combine MRI with the latest technologies – such a machine learning on MRI images and detecting DNA shed by cancers in blood – to see if we can find a way to make prostate cancer testing more reliable and maybe even do away with the need for biopsies altogether. We want to use MRI combined with new diagnostic tests to predict how the cancer will progress and to target the right treatment to the right person.[ii]

The University College London group will work in partnership with King’s College London, the Imperial College London, and over a dozen industry partners.

Given the scope of the study and the number of participants, it will be likely be a few years before the results are in and statistically analyzed. Hopefully, the time will fly by quickly, as we all envision and wait for the day when needle biopsies are a thing of the past.


[i] Medical Research Council Bulletin. “Researchers to investigate screening for prostate cancer using MRI—potentially replacing the PSA test. Apr. 4, 2018. https://medicalxpress.com/news/2018-04-screening-prostate-cancer-mripotentially-psa.html

[ii] Ibid.