How accurate are transrectal ultrasound guided biopsy and magnetic resonance imaging in staging prostate cancer and will we advise the wrong patient for active surveillance? A retrospective review in Hong Kong

Abstract

Objectives: To evaluate the accuracy oftransrectal ultrasound guided (TRUS)biopsy and magnetic resonance imaging(MRI) on staging prostate cancer (PCa)and to investigate any predictors ofupstaging of PCa in low risk patients.Patient and methods: 158 consecutiveseries of robotic assisted laparoscopic radi-cal prostatectomy during September 2011to October 2015 was retrospectivelyreviewed.Results: Total 134 patients with availablereport were analysed. The accuracy of later-ality for 12 core systematic TRUS biopsy,MRI and combination were 46.6%, 58.6%and 64.1%. The concordance of TRUSbiopsy, MRI and combination as indicatedby the Cohen Kappa coefficient were 0.21± 0.45, 0.14 ± 0.46 and 0.31 ± 0.68. Inaddition, the PPV, NPV, sensitivity, speci-ficity and accuracy were 25%, 80.2%, 7.4%,94.4% and 76.9% for prediction of extra-capsular extension by MRI scan. PSA density was a predictor of clinicalupstage of disease in low risks patient withodd-ratio of 2.337 (p = 0.024).Conclusion: TRUS biopsy, MRI and incombination had limited predictive valuein evaluating the laterality of tumour andMRI was a specific but not a sensitivetools in assessment of extra-capsularextension. In Chinese patients diagnosedwith low risks PCa and high PSA density,avoid active surveillance as the choice oftreatment.