How to improve the quality of your MRI

Multiparametric MRI (mpMRI) is now widespread across the UK as standard of care for establishing a man’s potential prostate cancer status. It has been shown to detect a greater percentage of clinically significant prostate cancer, whilst simultaneously lowering the rate of detection for clinically insignificant prostate cancer. mpMRI involves three sequences: dynamic contrast-enhanced (DCE), diffusion-weighted imaging (DWI) and T2-weighted imaging (T2W).
However there has been some debate in the clinical community as to the variation in quality of mpMRI images. This variation can lead to differing outcomes for men depending on which centre they have their imaging at.

Coupled with the idea that bpMRI, when performed at a high standard, could be as just as safe and effective as mpMRI, the PRIME trial, and its quality control phase called the GLIMPSE study (1), sought to understand these issues further.

The PRIME trial and Pre-Trial Quality Control Phase (the GLIMPSE study)

The PRIME trial (Prostate Imaging using MRI +/- contrast Enhancement) is an international, multicentre prospective clinical trial assessing biparametric MRI (bpMRI; DWI and T2W, without DCE) vs. mpMRI in diagnosing clinically significant prostate cancer. Quality control to prospectively evaluate the MRI scanners was undertaken before sites could participate in the trial. This pre-trial quality control phase was called the GLIMPSE study, which has now been published in a world-leading radiology journal and can be found here.

The first phase assessed the baseline quality of all MRI scanners, and showed substantial variation across the three different sequences (DCE, DWI and T2W) in many different centres (of the 71 scanners which were used, 48 received feedback to improve).

The Prostate Imaging Quality (PI-QUAL) score from the PRECISION trial is a standardised scoring system used to evaluate image quality of prostate MRI. It is a five-point scale that assesses image quality against a set of objective technical parameters (as per Prostate Imaging Reporting and Data System [PI-RADS] minimum technical requirements and standards for prostate mpMRI reporting) together with visual criteria obtained from the image which was assessed by an expert in prostate MRI.

The PI-QUAL SCALE:

● 1-2 – non diagnostic
● 3 – sufficient (it is possible to rule in but not to rule out all clinically significant lesions)
● 4 – adequate
● 5 – optimal

The centres with scanners not achieving a PI-QUAL score of 5 (48 scanners in total from 36 centres spanning 17 countries) were invited to resubmit new imaging data (second phase ) after using a modified version of the PI-RADS v2.1 protocol (2,3) resulting in a total of 62 (n = 64 [97%]) scanners with a final PI-QUAL score of 5.

In order to compare bpMRI and mpMRI within the trail, radiologists evaluated a two-part biparametric scan without the dye, and then separately evaluated a three-part multiparametric scan with the dye for each patient. If necessary, a targeted prostate biopsy was performed to confirm the accuracy of the diagnosis.

Results and Conclusions

Substantial variation in global prostate MRI quality has been shown, in particular with DCE sequences. Only 32% of scanners received an optimal score of 5 at baseline compared to 97% of the scans at the end of the quality control phase. Basic evaluation by a trained expert and modifications to MRI protocols using a modified version of the PI-RADS v2.1 protocol and PI-QUAL scoring systems can lead to substantial improvements in quality.

In addition, the newly published PRIME trial results show excitingly that bpMRI can be just as good as mpMRI in diagnosing prostate cancer. Removing the contrast step from a three-part MRI scan could make the process faster, cheaper, and more accessible without negatively impacting diagnostic accuracy.

What does this mean for men?

Good quality prostate MRIs have been shown to enable more accurate detection of prostate cancer and decision making. The GLIMPSE Study has demonstrated that there is a large variation in the quality of scans at different centres, but that very simple changes to MRI settings can easily improve the quality of scans.

Having the ability to make accurate diagnoses without using contrast could reduce scan time by around 33%, meaning hospitals can offer scans to more men using the same number of scanners and clinical staff.

Next steps

The PRIME Study has shown that a shorter, more streamlined prostate bpMRI is just as good as mpMRI in detecting prostate cancer. However, it is important that the quality of MRI scans (bpMRI or mpMRI) is of good quality to accurately detect prostate cancer, and this is more important with a shorter, two sequence scan.

However, for centres wishing to switch to bpMRI due to these results, it is essential that they first ensure that the MRI scan is of good quality to accurately detect prostate cancer. This is more important with a shorter, two sequence scan.

Find out more...

In March we hosted our second research webinar in the From Ideas to Impact series: How to improve MRI. Find out how Dr Veeru Kasivisvanathan, Dr Alexander Ng and Dr Aqua Asif from University College London are revolutionising the way we diagnose prostate cancer. Together they are aiming to make MRI scans faster, cheaper and less invasive, so that more men can access one.

You can watch the webinar here.