Combined morphological, [1H]-MR spectroscopic and contrast-enhanced imaging of human prostate cancer with a 3-Tesla scanner: preliminary experience.

PURPOSE: The objective of this study was to explore the feasibility of combined morphological magnetic resonance imaging (MRI), [(1)H]magnetic resonance spectroscopic imaging (MRSI) and quantitative dynamic contrast-enhanced MRI (DCE-MRI) of human prostate cancer at 3 Tesla using a pelvic phased-array coil.
MATERIALS AND METHODS: MRI, MRSI and DCE-MRI with a 3-Tesla whole-body scanner were performed in 30 patients with biopsy-proven prostate cancer before radical prostatectomy. High-resolution T2-weighted turbo spin echo (TSE) images were evaluated for visualisation of the peripheral zone, central gland, visibility of the cancer lesion, prostatic capsule delineation and overall image quality according to a five-point scale. Relative levels of the prostate metabolites citrate, choline and creatine were determined in cancer and in the normal peripheral zone (PZ) and central gland (CG). Spectra were also evaluated for the separation of the signal of citrate, choline and creatine and suppression of lipid and water signals. Time-intensity curves were obtained for prostatic cancer and healthy PZ and CG from DCE-MRI. Finally, time of arrival, time to peak, maximum enhancement and wash-in rate in cancer, normal PZ and CG were calculated.
RESULTS: The high signal-to-noise ratio (SNR) at 3 Tesla provided T2-weighted TSE images with excellent anatomical detail (in-plane voxel size of 0.22 x 0.22 mm) and good T2 contrast. The increased spectral resolution was sufficient to separate the choline and creatine resonances and allow delineation of the four peaks of citrate resonance. The (choline + creatine)/citrate ratio was elevated in cancer in comparison with PZ and CG (p<0.001). Dynamic contrast-enhanced images showed good temporal resolution. All parameters obtained from DCE-MRI showed a statistically significant (P<0.05) difference between cancer tissue and normal PZ and CG. Wash-in rate and (choline+creatine)/citrate ratio were significantly correlated (r=0.713, P=0.001) in PZ cancer, whereas the correlation was not significant (r=0.617, P=0.06) in CG and in PZ (r=0.530, P=0.08).
CONCLUSIONS: It is possible to perform MRI of prostate cancer at 3 Tesla using a pelvic phased-array coil with high spatial, temporal and spectral resolution. The combination of vascular information from DCE-MRI and metabolic data from MRSI has excellent potential for improved accuracy in delineating and staging prostate carcinoma. These results suggest that high magnetic field strengths offer the possibility of studying prostate cancer without use of an endorectal coil.