Accuracy of multiparametric magnetic resonance imaging in detecting recurrent prostate cancer after radiotherapy

Abstract

To assess the role of multiparametric magnetic resonance imaging (mp-MRI) of the prostate in evaluating local recurrence of prostate cancer, using transperineal template-guided 5 mm-spaced biopsies as a reference standard, in men treated with external beam radiotherapy (EBRT) for prostate cancer. The study included 13 patients with evidence of biochemical recurrence after EBRT who had undergone mp-MRI and prostate mapping. Each MRI scan (consisting of T1/T2 weighting, dynamic contrast enhancement and diffusion weighting) was reported by two expert uro-radiologists. Each prostate was divided into four regions of interest (ROI), generating 52 paired datasets for analysis. The mean (range) age of the men was 65.5 (55-70) years, the mean prostate-specific antigen (PSA) level before EBRT was 36.6 (4.5-150) ng/mL, the mean time from EBRT to biochemical recurrence was 5.7 (3-10) years and the mean PSA level at the time of recurrence was 7.1 (0.83-27.9) ng/mL. Eleven men had histological evidence of recurrence, with 23 of 52 ROIs involved with cancer. Overall accuracy, as expressed by the area under a receiver-operator curve, was 0.77 and 0.89 for all cancer, with accuracies of 0.86 and 0.93 for those cancers with ≥3 mm biopsy core length. Inter-observer variability was measured by calculating κ coefficients, which showed fair and moderate agreement between radiologists. Interpretation of mpMRI of the prostate after previous EBRT is challenging. Our results show that the accuracy is good using an accurate reference standard. These results need verification in more patients, but have implications for determining presence or absence of local recurrence and subsequent local salvage therapy.