ABSTRACT
Objective To compare the diagnostic performance of PI-RADS v2.1 and PI-RADS v2 in the detection of clinically significant prostate cancer(csPCa) by Meta-analysis. Methods The major biomedical databases were searched (CNKI, CBM, Medline, and Embase) with the keywords "PIRADS v2.1" or "PI-RADS v2.1". The Quality Assessment of Diagnostic Accuracy Studies Tool v2 (QUADAS-2) was used to evaluate literature quality. Meta-analysis was performed using STATA17.0 and ReMan5.4 software. Forest plots were used to represent the sensitivity and specificity of PI-RADS v2.1 and PI-RADS v2 for each study. Sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio were combined, and diagnostic performance was evaluated using asummary receiver operating characteristic curve (SROC). Subgroup analysis was performed on three covariables: tumor location, threshold, and the nationality of authors. Results A total of 12 studies were included, involving 3 158 patients and 3 243 lesions. Forall zones and the whole gland, PI-RADS v2.1 had a larger area under the SROC curve (AUC) for csPCa performance, compared with PI-RADS v2. Subgroup analysis: PI-RADS v2.1 also had a larger area under the SROC (AUC) to detect transitional zone csPCa. Different diagnostic thresholds: when a score of 4 was used for the threshold, PI-RADS v2.1 had the maximum area under SROC (AUC) for csPCa performance detection. Author nationality: Researches of PI-RADS v2.1 in Chinese authors had the largest area under the SROC (AUC) in detecting csPCa performance. Conclusion Compared with PI-RADS v2, the diagnostic performance of PI-RADS v2.1 in detecting csPCa is not obviously improved and overall specificity is still low.