Investigators have found that multiparametric MRI (mpMRI) for prostate cancer becomes more accurate when lesion-level apparent diffusion coefficient (ADC) values and patient prostate-specific antigen density (PSAD) thresholds are incorporated. Stanford researchers reported their findings on August 12 in Radiology, showing that these additional metrics can improve the positive predictive value (PPV) of mpMRI for clinically significant prostate cancer (csPCa).
Traditional mpMRI can identify suspicious lesions, but only about 35% of these turn out to be cancer. In a study of 944 men with 1,388 PI-RADS 3 or higher lesions evaluated via MRI/ultrasound fusion biopsy, low ADC values and high PSAD were found to be independent predictors of true-positive lesions. When combined, these factors increased PPV from 12% for high-ADC/low-PSAD lesions to 60% for low-ADC/high-PSAD lesions, substantially improving accuracy across PI-RADS categories.
The researchers note, however, that incorporating ADC and PSAD thresholds may affect sensitivity and negative predictive value, which requires further study to refine prostate imaging guidelines.