Ultra-small paramagnetic iron oxide (USPIO) compounds are often used to enhance fMRI sensitivity in preclinical studies measuring cerebral blood volume (CBV) [1]. They have also been proposed for MRA and contrast-enhanced imaging of patients with end-stage renal disease [2]. However, we have observed reduced utility for enhancing primate fMRI with repeated administrations. In this work, we test the hypothesis of iron accumulation in brain tissue under a daily dosing regimen in rats.4 male Sprague Dawley rats (Crl:SD, 8-10 wks) were each administered 8mg/kg of one of the USPIO compounds Feraheme®, Molday IonTM or Molday CLIOHTM, or sterile saline via a jugular catheter once daily for 8 weeks, and imaged once weekly 17-25 hours post-injection for 9 weeks on a 7T Bruker scanner using a 2D multi-echo gradient echo sequence (ΔTE= 3.92 ms, TR= 22.5ms, FA = $$$15^0$$$, voxel size = 0.2, 0.2, 1 mm). Quantitative susceptibility maps (QSM) were reconstructed from the complex gradient-echo data using the MEDIN algorithm [3]. For comparison, R2* maps were reconstructed from the same data set using the ARLO algorithm [4]. Rats were sacrificed after 9 weeks for serum and CSF iron level measurements and histology. Kinetic analysis of USPIO penetration into the cerebrospinal fluid (CSF) from plasma was performed by fitting region of interest (ROI) or voxel values from the time-series QSM images to a two-compartmental model with bolus dosing and linear clearance of the form: $$$dC_{CSF}/dt= k_{in}C_{pl}-k_{out}C_{CSF}$$$, where $$$C_{CSF}$$$ and $$$C_{pl}$$$ are the concentration in brain CSF and plasma respectively. Rats were sacrificed after 9 weeks for serum and CSF iron level measurements and histology.There was little contrast on R2* images because of the rapid decay of the gradient-echo signal, presumably from high USPIO concentration. These high concentrations did, however, significantly affect the field in the nearby voxels, which QSM exploited to obtain susceptibility of those and other areas in the brain. Accumulation of USPIO compounds was observed in brain interstitial fluid, ventricles and subarachnoid spaces. One week after dosing was stopped, USPIO concentration was reduced in most areas of the brain, but still remained relatively high in the brain ventricles (Figure 1, Week 9). CSF serum iron measurements (data not shown) confirmed an average increase in iron in the CSF and histopathology results indicated small quantities of iron in the choroid plexus, but not in other areas of the brain. A two compartmental model appeared to explain iron accumulation in CSF (Figure 2), although the influx transfer constant, $$$k_{in}$$$, was negative indicating that the transport flux was directed opposite the concentration gradient.We observed increased and persistent elevation in the susceptibility values in brain ventricles of rats receiving daily administration of SPIO indicating accumulation of iron oxide. Histopathology analysis revealed higher levels of iron in the choroid plexus, compared to a rat receiving saline injections. This suggests transport of iron from the blood into the CSF through the choroid plexus. A negative influx transfer constant obtained by modeling the time-series data with a two-compartment pharmacokinetic model suggested that iron is transported into the CSF against the concentration gradient, probably by a carrier protein. Our finding is reminiscent of the results obtained for manganese retention in the adult rat brain [4], suggesting a similar process of carrier-mediated cation influx, but a diffusion mediated efflux [5].A recent report of a “blooming” artifact in the lateral ventricles of patients receiving iron oxide therapy, is consistent with the results from our animal studies [6]. Taken together, these findings imply that it may be prudent to monitor subjects receiving frequent SPIO infusions with a susceptibility imaging technique such as QSM. It is also necessary to develop an accurate carrier protein-medicated pharmacokinetic model for the assessment of the effects of frequent SPIO dosing in animals and humans.