Very-high (0.05 cm3) spatial resolution three-dimensional 1h mr spectroscopic imaging of prostate pre-spost-brachytherapy

Abstract

To assess the feasibility of using active-acquisition (i.e. not post-process zero-fill interpolated) very-high spatial resolution (0.05 cm3) magnetic resonance spectroscopic imaging (MRSI) to assess changes in prostate metabolic profile pre- and post- 125I seed brachytherapy. Five patients, from a cohort of twenty-five already taking part in an ongoing MRSI study of prostate cancer pre- and post- permanent 125I seed implant brachytherapy, were asked to take part in the current investigation. Practically, this only required their staying in the magnet for the additional time needed to obtain the very-high resolution data following their regular MRI/MRSI scans. The larger study consisted of a baseline MRI/MRSI scan performed prior to the trans-rectal ultrasound (TRUS) prostate mapping to plan the implant, followed by subsequent MRSI scans in 6-month intervals post-implant up to and including the two year end-point. MRSI scans were performed on a GE 1.5T TwinSpeed MRI scanner in conjunction with a torso-phased-array/endorectal coil combination. 3D-MRSI data were acquired using the GE-PROSE pulse sequence with TE/TR of 130/1500 ms and 1 NEX in order to most clearly resolve the choline/creatine and citrate resonances, while keeping the overall scan time manageable. Magnetic field homogeneity at the site of the prostate was optimized through the use of higher-order gradient shimming. The whole volume of the prostate was covered by the spectroscopic grid consisting of 8 (S/I) x 16 (R/L) x 8 (A/P) voxels, with a 12 cm FOV (resolution of 0.4 cm3/voxel). Based on the results of the initial MRSI scan (viewable on the scanner screen immediately following the end of the scan) a second MRSI acquisition was prescribed wherein a very-high resolution spectroscopic grid was placed so as to include those regions of prostate indicating greatest cancer activity. All MRSI acquisition parameters were left unchanged except the FOV, which was halved, yielding a resolution of 0.05 cm3/voxel. Quantitative analysis was subsequently done off-line, voxel-by-voxel to determine the [choline+creatine]/citrate ratio. Despite the decreased signal-to-noise, all endogenous metabolites present could be readily identified and quantified both pre- and post-implant. Moreover, the smaller volume of each voxel resulted in an effective reduction in the number of 125I seeds included in each voxel post-implant, thereby reducing susceptibility-induced spectral distortion and thus improving overall spectral information by increasing the number of usable voxels. The results obtained demonstrate that very-high resolution 3D-MRS metabolic imaging of prostate pre- and post-brachytherapy is not only feasible, but can be used to non-invasively probe regions of tumor activity and assess response to therapy in greater detail than previously thought possible. Such information is sure to improve our understanding of tumor function and the mechanisms of therapeutic response.