BACKGROUND: Magnetic resonance imaging (MRI) and MR spectroscopic imaging (MRSI) have been gaining acceptance as tools in the evaluation of prostate cancer. We compared the accuracy of transrectal ultrasound (TRUS)-guided biopsy and dynamic contrast-enhanced MRI combined with three-dimensional (3D) MRSI in locating prostate tumours and determined the influence of prostate weight on MRI accuracy. PATIENTS AND METHODS: Between March 1999 and October 2006, 507 patients with localised prostate cancer underwent radical prostatectomy (RP) at the Jules Bordet Institute. Of these, 220 had undergone endorectal MRI (1.5 T Siemens Quantum Symphony) and 3D-MRSI prior to RP. We retrospectively reviewed data on tumour location and compared the results obtained by MRI and by TRUS-guided biopsy with those obtained on histopathology of the RP specimen. RESULTS: Patient data were as follows: median age 62.4 years (45-74); median PSA 6.36 ng/ml (0.5-22.6); 73.6% of patients had non-palpable disease (T1c); median biopsy Gleason score 6 (3-9); median RP specimen weight 50 g (12-172); median pathological Gleason score 7 (4-10); 68.64% of patients had organ-confined (pT2) disease. Tumour localisation was correlated with RP data in a significantly higher percentage of patients when using MRI rather than TRUS-guided biopsy (47.4 vs. 36.6%, p < 0.0001). MRI was marginally superior to TRUS-guided biopsy in detecting malignancy at the prostate apex (48.3 vs. 41.9%, p = 0.0687) and somewhat better at the prostate base (46 vs. 39.1%, p = 0.0413). It was highly significantly better at mid-gland (52 vs. 41.1%, p = 0.0015) and in the transition zone (40.1 vs. 24.3%, p < 0.0001). MRI had higher sensitivity in larger (≥50 g) than smaller (<50 g) prostates (50.3 vs. 42.2%, p = 0.0017). CONCLUSIONS: MRI was superior to TRUS-guided biopsy in locating prostate tumours except at the gland apex. MRI was more accurate in larger (≥50 g) than smaller prostates.
BACKGROUND: Magnetic resonance imaging (MRI) and MR spectroscopic imaging (MRSI) have been gaining acceptance as tools in the evaluation of prostate cancer. We compared the accuracy of transrectal ultrasound (TRUS)-guided biopsy and dynamic contrast-enhanced MRI combined with three-dimensional (3D) MRSI in locating prostate tumours and determined the influence of prostate weight on MRI accuracy. PATIENTS AND METHODS: Between March 1999 and October 2006, 507 patients with localised prostate cancer underwent radical prostatectomy (RP) at the Jules Bordet Institute. Of these, 220 had undergone endorectal MRI (1.5 T Siemens Quantum Symphony) and 3D-MRSI prior to RP. We retrospectively reviewed data on tumour location and compared the results obtained by MRI and by TRUS-guided biopsy with those obtained on histopathology of the RP specimen. RESULTS:Patient data were as follows: median age 62.4 years (45-74); median PSA 6.36 ng/ml (0.5-22.6); 73.6% of patients had non-palpable disease (T1c); median biopsy Gleason score 6 (3-9); median RP specimen weight 50 g (12-172); median pathological Gleason score 7 (4-10); 68.64% of patients had organ-confined (pT2) disease. Tumour localisation was correlated with RP data in a significantly higher percentage of patients when using MRI rather than TRUS-guided biopsy (47.4 vs. 36.6%, p < 0.0001). MRI was marginally superior to TRUS-guided biopsy in detecting malignancy at the prostate apex (48.3 vs. 41.9%, p = 0.0687) and somewhat better at the prostate base (46 vs. 39.1%, p = 0.0413). It was highly significantly better at mid-gland (52 vs. 41.1%, p = 0.0015) and in the transition zone (40.1 vs. 24.3%, p < 0.0001). MRI had higher sensitivity in larger (≥50 g) than smaller (<50 g) prostates (50.3 vs. 42.2%, p = 0.0017). CONCLUSIONS: MRI was superior to TRUS-guided biopsy in locating prostate tumours except at the gland apex. MRI was more accurate in larger (≥50 g) than smaller prostates.