According to a study published online Feb. 22 in the Journal of American Medical Association, MRI-derived parameters applied to a risk model could reduce unnecessary biopsies and improve accuracy in diagnosing prostate cancer.  

MRI has improved the detection of prostate cancer when combined with MRI-transrectal ultrasound (TRUS) fusion biopsies. However, the clinical question remains of whether MRI and multivariable prediction models can reduce the number of biopsies for patients who may have prostate cancer. 

"This model improved risk stratification among patients with positive findings on prostate magnetic resonance imaging and can be applied to other independent patient populations," wrote lead author Sherif Mehralivand, MD, a postdoctoral fellow at the National Institutes of Health in Bethesda, Maryland, and colleagues.  

Study participants underwent MRI, MRI-TRUS fusion-guided biopsy and 12-core systematic biopsy in one session. A development cohort of 400 participants used to derive the prediction model from one institution were enrolled from May 2015 to August 2016. A validation cohort of 251 patients from two independent institutions had biopsies between April 2013 and October 2016.  

Overall, 48.3 percent of the 400 patients in the development cohort and 38.2 percent of the 251 patients in the validation cohort were found to have prostate cancer. Researchers also found 18 of 100 fewer biopsies could've been performed with no misdiagnosis.  

"By applying the model to the external validation cohort, the area under the curve increased from 64 percent to 84 percent compared with the baseline model," the researchers wrote. "At a risk threshold of 20 percent, the MRI model had a lower false-positive rate than the baseline model, with only a small reduction in the true-positive rate versus 99 percent."