Reproducibility of index lesion size and mean apparent diffusion coefficient values measured by prostate multiparametric MRI: Correlation with whole-mount sectioning of specimens

OBJECTIVE. The purpose of this study is to determine the intra- and interreader agreement for index lesion size and mean apparent diffusion coefficient (ADC) value measurements performed by five readers using whole-mount histopathologic specimens processed with a patient-specific, MRI-based, 3D-printed mold as the standard of reference. MATERIALS AND METHODS. All men who underwent multiparametric MRI of the prostate performed using a 3-T scanner with endorectal and phased-array surface coils, followed by prostatectomy conducted between November 2015 and July 2016 at our institution, were identified. MRI examinations were independently reviewed by five readers with varying degrees of experience, two of whom had essentially no experience in prostate MRI interpretation before the study, to assess index lesion size and ADC values. A linear mixed model–based intraclass correlation was used to assess intra- and interreader reader agreement for lesion size and ADC measurements and agreement for size measurements between pathologic analysis and readers. RESULTS. A total of 80 men met the study eligibility criteria. Overall inter- and intrareader agreement for ADC measurements was excellent, with interclass correlation coefficient (ICC) values of 0.84 and 0.90, respectively; both inter- and intrareader agreement between experienced readers (0.82 and 0.92, respectively) and inexperienced readers (0.86 and 0.87, respectively) were excellent as well. The agreement between mean lesion size on imaging and histopathologic analysis ranged from poor (0.32) to good (0.66), with overall agreement considered fair (0.49). CONCLUSION. Readers with varying degrees of experience achieved good-to-excellent agreement for index lesion size and ADC values on multiparametric MRI of men with prostate cancer. This degree of reproducibility may improve preoperative risk stratification, informed decision making, and treatment planning for men with known or suspected prostate cancer.