Meta-analysis of diffusion-weighted magnetic resonance imaging in detecting prostate cancer

The objective of this study was to determine the diagnostic performance of quantitative diffusion-weighted magnetic resonance imaging in detection of prostate cancer.

A comprehensive search was performed for English articles published before May 2012 that fulfilled the following criteria: patients had histopathologically proved prostate cancer; diffusion-weighted imaging (DWI) was performed for the detection of prostate cancer, and data for calculating sensitivity and specificity were included. Methodological quality was assessed by using the quality assessment of diagnostic studies instrument. Publication bias analysis, homogeneity, inconsistency index, and threshold effect were performed by STATA version 12.

Of 119 eligible studies, 12 with 1637 malignant and 4803 benign lesions were included. There was notable heterogeneity beyond threshold effect and publication bias. The sensitivity and specificity with 95% confidence interval (CI) estimates of DWI on a per-lesion basis were 77% (CI, 0.76-0.84) and 84% (CI, 0.78-0.89), respectively, and the area under the curve of summary receiver operating characteristic curve was 0.88 (CI, 0.85-0.90). The overall positive and negative likelihood ratios with 95% CI were 4.93 (3.39-7.17) and 0.278 (0.19-0.39), respectively.

Quantitative DWI has a relative sensitivity and specificity to distinguish malignant from benign in prostate lesions. However, large-scale randomized control trials are necessary to assess its clinical value because of nonuniformed diffusion gradient b factor, diagnosis threshold, and small number of studies.