• Martin Lowry, Bashar Zelhof, Gary P Liney, Peter Gibbs, Martin D Pickles, and Lindsay W Turnbull.
• Centre for Magnetic Resonance Investigations, University of Hull, Hull, United Kingdom. m.lowry@hull.ac.uk
• Invest Radiol. 2009 Sep 1; 44 (9): 577-84.

ObjectivesThe ability to detect and identify malignant lesions within the prostate with conventional T2-weighted imaging is still limited. Although lesion conspicuity is improved with dynamic contrast-enhanced imaging there still remains some ambiguity as all tissues within the prostate may enhance. The aim of the current study was to take advantage of the improved signal-to-noise ratio at 3 T and assess the ability of 2 alternative pharmacokinetic models to clearly identify malignant areas within the prostate. We also aspire to assess the impact of tissue heterogeneity on variation in estimated pharmacokinetic parameters.Materials And MethodsQuantitative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of the prostate was implemented using multiple flip angles for T1 determination, and a rapid dynamic 3D T1-weighted acquisition with parallel imaging and a temporal resolution of 6.7 s. Pharmacokinetic analysis was performed for regions of tumor, normal-appearing peripheral zone (PZ), and central gland (CG) using fast exchange limit (FXL) or fast exchange regimen (FXR) models. Cell density was obtained from hematoxylin and eosin stained whole mount radical prostatectomy specimens.ResultsNative tissue T1 was significantly lower in tumor and PZ tissue than in CG. The FXL model revealed increased mean K(trans), k(ep), and v(e) in tumor and CG compared with PZ. With the FXR model, fitting was improved and all parameters were significantly increased, however, there were no longer significant differences between regions for v(e). The additional parameter of the FXR model, tau(i), nominally representing mean lifetime of intracellular water, was significantly decreased in tumor compared with both PZ and CG. Rate constants for CG were significantly lower than those of tumor for both models. In addition, for all tissues, K(trans) and v(e) were positively correlated with cell density.ConclusionsAccounting for a finite water exchange rate between cells and their environment improves the discrimination of malignant from benign tissues within the prostate and may aid staging accuracy and ability to monitor response to treatment.

31 keywords...

hide…