Introduction: We sought to investigate the association of preprostatectomy magnetic resonance imaging (MRI) and surgical pathologic findings in patients with prostate cancer.
Materials and methods: All patients with prostate cancer and preprostatectomy MRI available between 2002 and 2015 were included. Age, prostate-specific antigen at diagnosis, Gleason score at biopsy, MRI technique, radiology report suggestive of prostate cancer, extraprostatic invasion and seminal vesicle involvement, lymphadenopathy and final pathology report were retrospectively reviewed. Data was analyzed for sensitivity, specificity, positive and negative predictive values of MRI findings for predicting T3 disease. Consistency of MRI findings with pathology report was compared between MRIs with or without endorectal coil (ERC).
Results: A cohort of 83 patients was identified. Eighty-seven percent of the patients had MRI findings suggestive of prostate cancer. MRI was performed with and without ERC in 21 (25.3%) and 62 (74.3%) patients respectively. Eighty-five percent of patients with ERC and 88.7% of those without ERC had MRI findings suggestive of prostate cancer (p = 0.659). MRI correlated with final surgical pathology stage T3 in 53 patients (64%). MRI findings were consistent with final pathology report in 70% of ERC group and 61.3% of non ERC group (p = 0.482). In terms of extra prostatic invasion or seminal vesicle involvement, MRI had specificity, sensitivity, positive and negative predictive values of 84.44%, 37.84%, 66.67% and 62.3% respectively.
Conclusions: MRI was specific but not sensitive in determining extraprostatic or seminal vesicle invasion. MRI was not accurate for lymph node involvement. In addition, using an ERC did not increase the accuracy of prostate MRI in this small cohort.