Multiparametric MRI to improve detection of prostate cancer compared with transrectal ultrasound-guided prostate biopsy alone: the PROMIS study

Louise C. Brown, Hashim U Ahmed, Rita Isabel Neves De Faria, Ahmed El-Shater Bosaily, Rhian Gabe, Richard Kaplan, M Parmar, Yolanda Collaco-Moraes, Katie Ward, Richard G Hindley, Alex Freeman, Alexander Kirkham, Robert Oldroyd, Chris Parker, Simon Bott, Nick Burns-Cox, Tim Dudderidge, Maneesh Ghei, Alastair Henderson, Rajendra PersadDerek J Rosario, Iqbal Shergill, Mathias Winkler, Marta Ferreira Oliveira Soares, David Eldon Spackman, Mark John Sculpher, Mark Emberton

Research output: Contribution to journalArticlepeer-review


Background: Men with suspected prostate cancer usually undergo transrectal ultrasound (TRUS)-guided prostate biopsy. TRUS-guided biopsy can cause side effects and has relatively poor diagnostic accuracy. Multiparametric magnetic resonance imaging (mpMRI) used as a triage test might allow men to avoid unnecessary TRUS-guided biopsy and improve diagnostic accuracy. Objectives: To (1) assess the ability of mpMRI to identify men who can safely avoid unnecessary biopsy, (2) assess the ability of the mpMRI-based pathway to improve the rate of detection of clinically significant (CS) cancer compared with TRUS-guided biopsy and (3) estimate the cost-effectiveness of a mpMRI-based diagnostic pathway. Design: A validating paired-cohort study and an economic evaluation using a decision-analytic model. Setting: Eleven NHS hospitals in England. Participants: Men at risk of prostate cancer undergoing a first prostate biopsy. Interventions: Participants underwent three tests: (1) mpMRI (the index test), (2) TRUS-guided biopsy (the current standard) and (3) template prostate mapping (TPM) biopsy (the reference test). Main outcome measures: Diagnostic accuracy of mpMRI, TRUS-guided biopsy and TPM-biopsy measured by sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) using primary and secondary definitions of CS cancer. The percentage of negative magnetic resonance imaging (MRI) scans was used to identify men who might be able to avoid biopsy. Results: Diagnostic study – a total of 740 men were registered and 576 underwent all three tests. According to TPM-biopsy, the prevalence of any cancer was 71% [95% confidence interval (CI) 67% to 75%]. The prevalence of CS cancer according to the primary definition (a Gleason score of ≥ 4 + 3 and/or cancer core length of ≥ 6 mm) was 40% (95% CI 36% to 44%). For CS cancer, TRUS-guided biopsy showed a sensitivity of 48% (95% CI 42% to 55%), specificity of 96% (95% CI 94% to 98%), PPV of 90% (95% CI 83% to 94%) and NPV of 74% (95% CI 69% to 78%). The sensitivity of mpMRI was 93% (95% CI 88% to 96%), specificity was 41% (95% CI 36% to 46%), PPV was 51% (95% CI 46% to 56%) and NPV was 89% (95% CI 83% to 94%). A negative mpMRI scan was recorded for 158 men (27%). Of these, 17 were found to have CS cancer on TPM-biopsy. Economic evaluation – the most cost-effective strategy involved testing all men with mpMRI, followed by MRI-guided TRUS-guided biopsy in those patients with suspected CS cancer, followed by rebiopsy if CS cancer was not detected. This strategy is cost-effective at the TRUS-guided biopsy definition 2 (any Gleason pattern of ≥ 4 and/or cancer core length of ≥ 4 mm), mpMRI definition 2 (lesion volume of ≥ 0.2 ml and/or Gleason score of ≥ 3 + 4) and cut-off point 2 (likely to be benign) and detects 95% (95% CI 92% to 98%) of CS cancers. The main drivers of cost-effectiveness were the unit costs of tests, the improvement in sensitivity of MRI-guided TRUS-guided biopsy compared with blind TRUS-guided biopsy and the longer-term costs and outcomes of men with cancer. Limitations: The PROstate Magnetic resonance Imaging Study (PROMIS) was carried out in a selected group and excluded men with a prostate volume of > 100 ml, who are less likely to have cancer. The limitations in the economic modelling arise from the limited evidence on the long-term outcomes of men with prostate cancer and on the sensitivity of MRI-targeted repeat biopsy. Conclusions: Incorporating mpMRI into the diagnostic pathway as an initial test prior to prostate biopsy may (1) reduce the proportion of men having unnecessary biopsies, (2) improve the detection of CS prostate cancer and (3) increase the cost-effectiveness of the prostate cancer diagnostic and therapeutic pathway. The PROMIS data set will be used for future research; this is likely to include modelling prognostic factors for CS cancer, optimising MRI scan sequencing and biomarker or translational research analyses using the blood and urine samples collected. Better-quality evidence on long-term outcomes in prostate cancer under the various management strategies is required to better assess cost-effectiveness. The value-of-information analysis should be developed further to assess new research to commission.

Original languageEnglish
Number of pages175
JournalHealth technology assessment
Issue number39
Publication statusPublished - Jul 2018

Bibliographical note

© Queen’s Printer and Controller of HMSO 2018. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details


  • Aged
  • Cohort Studies
  • Continental Population Groups
  • Cost-Benefit Analysis
  • Endoscopic Ultrasound-Guided Fine Needle Aspiration/economics
  • Humans
  • Magnetic Resonance Imaging/economics
  • Male
  • Middle Aged
  • Neoplasm Grading
  • Prostate/diagnostic imaging
  • Prostatic Neoplasms/diagnosis
  • Quality-Adjusted Life Years
  • Sensitivity and Specificity
  • State Medicine
  • United Kingdom

Cite this