Besasie BD, Sunnapwar AG, Gao F et al: Restriction spectrum imaging magnetic resonance imaging to improve prostate cancer imaging in men on active surveillance. J Urol 2021; 206: 44.

Accumulating data support the use of multiparametric magnetic resonance imaging (mpMRI) to reduce sampling errors of prostate biopsy targeting of suspected tumors.¹ Due to the significant variation in mpMRI techniques and reporting, in February 2012 the European Society of Urogenital Radiology (ESUR) published recommendations for the performance of mpMRI using a structured reporting scheme named Prostate Imaging Reporting and Data System™ (PI-RADS™).² The PI-RADS score is defined on a Likert scale based on suspicion of prostate cancer: highly unlikely (1), unlikely (2), equivocal (3), likely (4), highly likely (5). A randomized, multicenter clinical trial showed that prostate MRI for initial biopsy enhances prostate biopsy precision by significantly increasing the detection of aggressive cancers while reducing the number of prostate biopsies performed.³

However, men on active surveillance (AS) have already been diagnosed with prostate cancer, which makes the PI-RADS intended for the suspicion of prostate cancer no longer applicable. As such, the success of MRI in the AS population has been more tempered than in the diagnostic population. In a randomized clinical trial of MRI in AS, physicians did not improve the upgrading rate using MRI-targeted biopsies and systematic biopsies compared with systematic biopsy alone.⁴ Data in a large observational cohort from the Canary-PASS study also noted no additional benefit of MRI in predicting upgrading.⁵ These results demonstrate a need for substantial improvements in MRI characteristics in AS.

To overcome the shortcomings of standard mpMRI, we have developed a short-duration (<5 minutes), targeted magnetic resonance restriction spectrum imaging (RSI-MRI) scan that can be added to current mpMRI sequences to improve specificity.⁶ Diffusion-weighted imaging (DWI) is a key component of the standard mpMRI protocol used to generate quantitative maps of the Apparent Diffusion Coefficient (ADC). Hyperintense regions on DWI with corresponding low ADC is a combination that suggests reduced water diffusion mobility and an increased likelihood of aggressive cancer.⁷ The RSI-MRI scan provides a more sophisticated analysis of high b-value DWI data to disambiguate and quantify the restricted water signal from within cancer cells providing a more sensitive and specific imaging biomarker of aggressive prostate cancer compared with standard DWI/ADC.^6,8 The RSI-MRI output value is a quantifiable imaging biomarker and is a salient step in MRI that we have previously associated with tumor aggressiveness (Gleason score) in prostatectomy specimens.⁹

As an example using figure 1, part a, we show a PI-RADS 4 lesion with an RSI overlay. The ability to quantify the RSI output from this value is shown in histogram (fig. 1, b) noting RSI differences between normal tissue and the suspected lesion. The RSI output generates values that can be used as the imaging biomarker. The suspected lesion in figure 1 is a grade group 4 prostate cancer (Gleason 4+4=8) on targeted biopsy of this location. The implementation and validation of RSI-MRI in combination with current biomarkers is the last prerequisite to gaining widespread clinical acceptance of a shift from current AS practice (timed prostate specific antigen tests, prostate exams and biopsies) to a more individualized, risk-based approach.

Figure 1. a, T2 prostate MRI with RSI overlay. Blue is normal prostate tissue and red is the target lesion detected on RSI. b, histogram of RSI output values (Restricted Signal Map, x-axis) and number of voxels that contain those values (count, y-axis). Pathology reported a grade group 4 lesion (Gleason 4+4=8).

Figure 2. a, T2 prostate MRI with RSI overlay. Blue is normal prostate tissue and red is target lesion detected on RSI. b, histogram of RSI output values (Restricted Signal Map, x-axis) and number of voxels that contain those values (count, y-axis). Pathology showed no tumor in target and negative 12-core prostate biopsy in a man on active surveillance.

In our recent manuscript in The Journal of Urology®, we apply the RSI-MRI technique to men on AS prior to their next prostate biopsy.¹⁰ Our primary outcome was progression or reclassification in AS. We prospectively enrolled men prior to their upcoming AS prostate biopsy in order to predict biopsy outcome. The radiologist did not use RSI in decision making. All images were marked using the PI-RADS system and underwent targeted biopsy. In our analysis, we were able to use RSI-MRI output values to improve the specificity of prostate MRI over ADC value and PI-RADS, while achieving improved accuracy to an area under the curve of 90%. We acknowledge our sample size is small and will require a larger cohort for validation, yet the findings show that using a simple acquisition technique can substantially enhance MRI. Also important is that RSI-MRI acquisition does not need hardware or contrast material and can be deployed within the current MRI workflow.

In order to improve accuracy, one must improve the detection of true positive lesions but also reduce the detection of a false-positive. False-positive lesions on MRI are more common among PI-RADS 3 and 4 lesions, yet drive down the accuracy of MRI. For example, in figure 2, part a, we show a PI-RADS 4 lesion on targeted biopsy was shown to be negative for cancer. The histogram (fig. 2, b) shows an overlapping RSI-MRI signal with normal tissue; therefore, we may have anticipated a negative or low-grade biopsy result. Inflammation may appear as lesion on standard MRI. We describe that RSI-MRI is able to better distinguish inflammatory nodules from cancer nodules based on the RSI-MRI signal values by outperforming ADC values. RSI-MRI also has the potential to be normalized across institutions unlike ADC, which remains machine- and technique-dependent. RSI-MRI is now commercially available through Cortechs.ai with their latest U.S. Food and Drug Administration–cleared product, OnQ Prostate.

Dr. Liss has no financial conflicts of interest with Cortechs.ai or the OnQ Prostate platform.