Dr Gershman discusses how prostate imaging has improved over time, and how and when clinicians implement mpMRI in their practice.
Dr. Boris Gershman: The question is how multi-parametric MRI has evolved over time, and the short answer is it has evolved tremendously, both in terms of image quality, acquisition techniques, and interpretation. Certainly, as scanners have gotten better, image quality has improved. And, point of fact, the early prostate MRI's, university used endorectal coil as part of the imaging protocol. Nowadays a lot of facilities omit endorectal coils and we can get high-quality images even with a 1.5 Tesla machine sometimes. There have been advances in terms of the standardized grading of prostate MRI's with the introduction and different iterations of the PI-RADS classification system, most recently version 2.1, which have allowed a more standardized way of assessing the images and communicating the perceived risk or probability of having a clinically significant prostate cancer, which actually is an important component of interaction between radiologists and clinicians, urologists, and other specialists treating men with prostate cancer, to really ensure that we are not missing a more aggressive cancer than what might be seen on a biopsy, and also assessing risk for prostate cancer in men just being screened and considering prostate biopsy.
The question is what barriers are there to the application of multi-parametric prostate MRI, and there are some barriers. One historical barrier that does not really exist today as much is issues with insurance coverage. In the past, about ten years ago, there were problems with insurers covering multi-parametric prostate MRI for men in the biopsy-naïve setting. Some insurers would decline coverage unless a man had a prior prostate biopsy, and so this was a barrier. Nowadays, it is very uncommon to see insurers decline prostate MRI, even in the initial biopsy setting, and this is a testament to the expanded data supporting its utility both in biopsy-naïve settings, in the repeat-biopsy setting, as well as in men on active surveillance, or in planning definitive local therapy for men with clinically significant prostate cancer. One existing barrier that remains, though, is the heterogeneity, or variation, in cancer detection rates, according to the PI-RADS classification system. Several seminal trials provide very high positive predictive values for the PI-RADS 3, 4, 5 classification for detecting prostate cancer on biopsy. And unfortunately, in real-world settings, we see lower rates of cancer detection and substantial variability, or heterogeneity, in these cancer detection rates. And this has been described in multiple studies. This is in part due to interpreter variability, but also other factors that we may or may not yet know for sure, including differences between study populations in those trials, and real-world patients being referred to urology practices. But in the ideal setting, we would have a more accurate assessment, based on the MRI, of what the probability of finding cancer is, to better determine who should undergo a biopsy and who should not. And I think this is one of the opportunities for ongoing work in the future; to improve the utility and performance of prostate MRI, whether from different acquisition techniques or different interpretation techniques.