Manganese concentration mapping in the rat brain with MRI, PET, and autoradiography.

Purpose: Mn2+ is used as a contrast agent and marker for neuronal activity with magnetic resonance imaging (MRI) in rats and mice, but its accumulation is generally not assessed quantitatively. In this work, nonradioactive Mn and Mn-52 are injected simultaneously in rats, and imaged with MRI, positron emission tomography (PET) and autoradiography (AR). Mn distributions are compared between modalities, to assess the potential and limitations on quantification of Mn with MRI, and to investigate the potential of multimodal measurement of Mn accumulation. Methods: MRI (in vivo), PET (in vivo and post mortem), and AR (ex vivo) were acquired of rat brains, for which animals received simultaneous intraperitoneal (IP) or intracerebrovertricular (ICV)targeted injections containing the positron-emitting radionuclide Mn-52 and additional nonradioactive MnCl2, which acts as an MRI contrast agent. Pre and postinjection MR images were fit for the longitudinal relaxation rate (R1), coregistered, and subtracted to generate R1 difference maps, which are expected to be proportional to change in Mn concentration in tissue. AR and PET images were coregistered to smoothed R1 difference maps. Results: Similar spatial distributions were seen across modalities, with Mn accumulation in the colliculus, near the injection site, and in the 4th ventricle. There was no Mn-52 accumulation measurable with PET in the brain after IP injection. In areas of very highly localized and concentrated Mn-52 accumulation in PET or AR, consistent increases of R1 were not seen with MRI. Scatter plots of corresponding voxel R1 difference and PET or AR signal intensity were generated and fit with least squares linear models within anatomical regions. Linear correlations were observed, particularly in regions away from very highly localized and concentrated Mn accumulation at the injection site and the 4th ventricle. Accounting for radioactive decay of Mn-52, the MnCl2 longitudinal relaxivity was between 4.0 and 5.1 s(-1)/mM, which is within 22% of the in vitro relaxivity. Conclusions: This proof-of-concept study demonstrates that MR has strong potential for quantitative assessment of Mn accumulation in the brain, although local discrepancies from linear correlation suggest limitations to this use of MR in areas of inflammation or very high concentrations of Mn. These discrepancies also suggest that a combination of modalities may have additional utility for discriminating between different pools of Mn accumulation in tissue. (C) 2017 American Association of Physicists in Medicine