Diagnostic Accuracy of Multiparametric MRI for Prostate Cancer Detection: A Meta-Analysis
摘要
Prostate cancer (PCa) is a well-known cancer that is frequently diagnosed in men and a significant cause of cancer morbidity and mortality on a global scale. Clinical significant prostate cancer (csPCa) is a concern that requires early diagnosis in order to maximize treatment and minimize overtreatment. Traditional diagnostic tools, which are prostate-specific antigen (PSA) and systematic transrectal ultrasound (TRUS)-guided biopsy, are limited in nature in the form of low specificity and sampling errors. Multiparametric MRI (mpMRI) has become a potentially effective non-invasive method to enhance detection and targeted biopsies.
ObjectiveTo determine the diagnostic accuracy of mpMRI in detecting csPCa within a meta-analysis of both prospective and retrospective analyses.
MethodsThe literature search was performed in the databases of PubMed, Embase, Web of Science, Scopus, and Cochrane Library until March 2026. Articles that provided diagnostic results of mpMRI or biparametric MRI (bpMRI) that were confirmed by histopathology were incorporated. Bivariate random-effects model was used to conduct data extraction, quality assessment (QUADAS-2) and statistical pooling.
ResultsThe number of studies involved was 23 with 8,000 or more participants. mpMRI sensitivity was pooled and was 93.3% (95% CIs: 87.6–96.4%), and specificity of 57.5% (95% CIs: 45.1–69.0%). The odds ratio of the diagnostic was 18.75 and the likelihood ratio was negative 0.12. The level of heterogeneity was high (I2 = 95%). The results indicate high sensitivity in the elimination of csPCa but moderate specificity reveals the need to interpret the results with care.
ConclusionmpMRI is very sensitive in the detection of csPCa and it is effective in eliminating unnecessary biopsies. The further improvement of diagnostic performance and patient management is possible with the help of standardized imaging protocols, radiologist experience, and incorporation of advanced techniques.