Objective: To prospectively evaluate the accuracy of transrectal ultrasonographic (TRUS)-guided biopsies by using combined magnetic resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI) in patients with persistently high prostate-specific antigen (PSA) levels and negative TRUS-guided biopsy results.
Material and Method: Twenty-one patients (age range 50-77 years, average 61.4 years) with negative TRUS biopsy were enrolled. Suspicious areas were identified by discrete low signal intensity in T2 on standard MRI. MRSI was interpreted by using spectral approach and given score of 1 (benign) to 5 (malignant). Suspicious voxels were localized for guided TRUS biopsy. All patients underwent sextant TRUS biopsies with up to four additional biopsies targeted at suspicious sites. Diagnostic accuracy of MRI/MRSI in patient-by-patient and voxel-by-core were analyzed.
Results: Prostate cancer was detected in 2 of 21 patients (9.5%). The sensitivity, specificity, PPV, NPV, and accuracy of combined MRI/ MRSI for detection of prostate cancer were 100%, 84%, 40%, 100%, and 86%, respectively. The site of positive biopsy correlated correctly with voxels were 80%, 85%, 21%, 99% and 85% on sensitivity, specificity, PPV, NPV and accuracy, respectively.
Conclusion: MRI/MRSI have the potential to guide biopsy to cancer foci in patients with persistently high PSA levels and prior negative TRUS biopsy results.

Keywords: Prostate cancer, Magnetic resonance imaging, Spectroscopic imaging, Transrectal ultrasonography