Optimizing biopsy strategy for prostate cancer: Bayesian framework of network meta-analysis and hierarchical summary receiver operating characteristic model for diagnostic accuracy.

Overdiagnosis and overtreatment are well known problems in prostate cancer (PCa). The transrectal ultrasound (TRUS) Guided biopsy (GB) as a current gold standard investigation has a low positive detection rate resulting in unnecessary biopsies. The choice of optimal biopsy strategy needs to be defined. Therefore, we undertook a Bayesian network meta analysis (NMA) and Bayesian prediction in the hierarchical summary receiver operating characteristic (HSROC) model to present a method for optimizing biopsy strategy in PCa. Twenty eight relevant studies were retrieved through online databases of EMBASE, MEDLINE, and CENTRAL up to February 2020. Markov chain Monte Carlo simulation and Surface Under the Cumulative RAnking curve were used to calculate the rank probability using odds ratio with 95% credible interval. HSROC model was used to formulate the predicted true sensitivity and specificity of each biopsy strategy. Six different PCa biopsy strategies including transrectal ultrasound GB (TRUS GB), fusion GB (FUS GB), fusion + transrectal ultrasound GB (FUS + TRUS GB), magnetic resonance imaging GB (MRI GB), transperineal ultrasound GB (TPUS GB), and contrast enhanced ultrasound GB were analyzed in this study with a total of 7584 patients. These strategies were analyzed on five outcomes including detection rate of overall PCa, clinically significant PCa, insignificant PCa, complication rate, and HSROC. The rank probability showed that the overall PCa detection rate was higher in FUS + TRUS GB, MRI GB, and FUS GB. In terms of clinically significant PCa detection, FUS + TRUS GB and FUS GB had a relatively higher clinically significant PCa detection rate, whereas TRUS GB had a relatively lower rate for clinically significant PCa detection rate. MRI GB (91% and 81%) and FUS GB (82% and 83%) had the highest predicted true sensitivity and specificity, respectively, whereas TRUS GB (62% and 83%) had a lower predicted true sensitivity and specificity. MRI GB, FUS GB, and FUS + TRUS GB were associated with lower complication rate, whereas TPUS GB and TRUS GB were more associated with higher complication rate. This NMA and HSROC model highlight the important finding that FUS + TRUS GB, FUS GB, and MRI GB were superior compared with other strategies to avoid the overdiagnosis and overtreatment of PCa. FUS GB, MRI GB, and FUS + TRUS GB had lower complication rates. These results may assist in shared decision making between patients, carers, and their surgeons.