Predicting the Need for Biopsy to Detect Clinically Significant Prostate Cancer in Patients with a Magnetic Resonance Imaging–detected Prostate Imaging Reporting and Data System/Likert ≥3 Lesion: Development and Multinational External Validation of the Imperial Rapid Access to Prostate Imaging and Diagnosis Risk Score
Max Peters, David Eldred‐Evans, P.H.J. Kurver, Ugo Giovanni Falagario, Martin J. Connor, Taimur T. Shah, Joost J.C. Verhoeff, Pekka Taimen, Hannu J. Aronen, Juha Knaapila, Ileana Montoya Perez, Otto Ettala, Armando Stabile, Giorgio Gandaglia, Nicola Fossati, Alberto Martini, Vito Cucchiara, Alberto Briganti, Anna Lantz, Wolfgang Picker, Erik Skaaheim Haug, Tobias Nordström, Mariana Bertoncelli Tanaka, Deepika Reddy, Edward Bass, Peter S.N. van Rossum, Kathie Wong, Henry Tam, Mathias Winkler, Stephen Gordon, Hasan Qazi, Peter J. Boström, Ivan Jambor, Hashim U. Ahmed
Abstract
BACKGROUND: Although multiparametric magnetic resonance imaging (MRI) has high sensitivity, its lower specificity leads to a high prevalence of false-positive lesions requiring biopsy. OBJECTIVE: To develop and externally validate a scoring system for MRI-detected Prostate Imaging Reporting and Data System (PIRADS)/Likert ≥3 lesions containing clinically significant prostate cancer (csPCa). DESIGN, SETTING, AND PARTICIPANTS: The multicentre Rapid Access to Prostate Imaging and Diagnosis (RAPID) pathway included 1189 patients referred to urology due to elevated age-specific prostate-specific antigen (PSA) and/or abnormal digital rectal examination (DRE); April 27, 2017 to October 25, 2019. INTERVENTION: Visual-registration or image-fusion targeted and systematic transperineal biopsies for an MRI score of ≥4 or 3 + PSA density ≥0.12 ng/ml/ml. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Fourteen variables were used in multivariable logistic regression for Gleason ≥3 + 4 (primary) and Gleason ≥4 + 3, and PROMIS definition 1 (any ≥4 + 3 or ≥6 mm any grade; secondary). Nomograms were created and a decision curve analysis (DCA) was performed. Models with varying complexity were externally validated in 2374 patients from six international cohorts. RESULTS AND LIMITATIONS: The five-item Imperial RAPID risk score used age, PSA density, prior negative biopsy, prostate volume, and highest MRI score (corrected c-index for Gleason ≥3 + 4 of 0.82 and 0.80-0.86 externally). Incorporating family history, DRE, and Black ethnicity within the eight-item Imperial RAPID risk score provided similar outcomes. The DCA showed similar superiority of all models, with net benefit differences increasing in higher threshold probabilities. At 20%, 30%, and 40% of predicted Gleason ≥3 + 4 prostate cancer, the RAPID risk score was able to reduce, respectively, 11%, 21%, and 31% of biopsies against 1.8%, 6.2%, and 14% of missed csPCa (or 9.6%, 17%, and 26% of foregone biopsies, respectively). CONCLUSIONS: The Imperial RAPID risk score provides a standardised tool for the prediction of csPCa in patients with an MRI-detected PIRADS/Likert ≥3 lesion and can support the decision for prostate biopsy. PATIENT SUMMARY: In this multinational study, we developed a scoring system incorporating clinical and magnetic resonance imaging characteristics to predict which patients have prostate cancer requiring treatment and which patients can safely forego an invasive prostate biopsy. This model was validated in several other countries.