National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection - recommendations from a UK consensus meeting

Abstract

Objectives: To identify areas of agreement and disagreement in the implementation of multi-parametric magnetic resonance imaging (mpMRI) of the prostate in the diagnostic pathway.

Materials and methods: Fifteen UK experts in prostate mpMRI and/or prostate cancer management across the UK (involving nine NHS centres to provide for geographical spread) participated in a consensus meeting following the Research and Development Corporation and University of California-Los Angeles (UCLA-RAND) Appropriateness Method, and were moderated by an independent chair. The experts considered 354 items pertaining to who can request an mpMRI, prostate mpMRI protocol, reporting guidelines, training, quality assurance (QA) and patient management based on mpMRI levels of suspicion for cancer. Each item was rated for agreement on a 9-point scale. A panel median score of ≥7 constituted 'agreement' for an item; for an item to reach 'consensus', a panel majority scoring was required.

Results: Consensus was reached on 59% of items (208/354); these were used to provide recommendations for the implementation of prostate mpMRI in the UK. Key findings include prostate mpMRI requests should be made in consultation with the urological team; mpMRI scanners should undergo QA checks to guarantee consistently high diagnostic quality scans; scans should only be reported by trained and experienced radiologists to ensure that men with unsuspicious prostate mpMRI might consider avoiding an immediate biopsy.

Conclusions: Our consensus statements demonstrate a set of criteria that are required for the practical dissemination of consistently high-quality prostate mpMRI as a diagnostic test before biopsy in men at risk.

Keywords: consensus methods; multi-parametric MRI; prostate cancer; recommendations.

Figure 1

(A) Shows the mpMRI of a 62‐year‐old man, with a PSA level of 4.4 ng/mL and a gland volume of 25 mL at the level of the mid‐gland to apex region. On T2W imaging, there is diffuse and patchy low T2 signal and a lower T2 signal at the right lateral gland, with an equivocal high signal focus on diffusion high b value at 9 o'clock and corresponding equivocal low ADC signal with bilateral enhancement on DCE. The focal lesion (represented by number 1 in 1. (B) was reported with a Likert‐assessment of 3/5. Besides, the remainder of the gland was also assessed with the whole prostate divided into quarters for Likert assessment (C). Each quarter was reported as a ‘Likert‐assessment’ 3/5. The background changes scored 3 are represented by the shaded area in B. Upon transperineal template mapping biopsy, the prostate was found to harbour adenocarcinoma Gleason 3+4, (40% biopsy core involvement) at the right posterior apex, focal high‐grade prostatic intraepithelial neoplasia at the left posterior apex and Gleason 3+3, at eight different sites within the prostate (10–40% biopsy core involvement).

Figure 2

Summarises the key recommendations across the early prostate cancer diagnosis pathway to deliver consistently high‐quality mpMRI studies.