Restless legs syndrome (RLS) is a neurological disorder with a prevalence of 5-10% of the population. It is characterized by unpleasant sensations mainly in the legs and an urge to move during sitting or resting in the later part of the day¹. Previous studies have associated RLS to the abnormal dopaminergic function in the central nervous system (CNS) and dysfunction in iron metabolism^2-5. In vivo measures of tissue iron using relaxometry or MR phase have suggested a general trend of iron decrease in the brain especially in the substantia nigra (SN), yet many inconsistencies still exist in the reported iron change of different brain regions in different RLS phenotypes^5,6. The present study thus aims to assess possible brain iron deficiency in RLS using quantitative susceptibility mapping (QSM)^7-10 and test possible correlations between tissue magnetic susceptibility and RLS clinical features.

Age matched subjects (30 controls with Mean ± SD ages of 57.9±8.5 years, 40 primary RLS patients off all RLS medications for at least 12 days with ages of 58.2±9.4 years and IRLS severity scores Mean ± SD 25.3 ± 7.1) were scanned at 7T (Philips Healthcare) using a 32-channel Nova medical receiver head coil using 3D GRE sequence, with either 1mm isotropic resolution, TR/TE1/ΔTE=45/2/2 ms, 9 echoes, flip angle of 9°, bandwidth 1530 Hz/px (n=2), or 0.8 mm isotropic resolution, with TR/TE= 20/12 ms, flip angle 10°, bandwidth 169 Hz/px (n=68). For all the scans, MR phase data at TE=12 ms were used to generate quantitative susceptibility maps. Phase data were unwrapped with Laplacian-based phase unwrapping and the background field was removed using the V-SHARP method^11-12. QSM images were then generated using the LSQR method¹². The QSM image was referenced with respect to the CSF in the lateral ventricle. Regions of interest (ROI) were selected automatically using a QSM based atlas¹⁰ and corrected by a neuroradiologist if necessary (HL). In addition, a fixed shape ROI (3x3 square) was selected in the center of the SN in order to test the influence of different ways of ROI selection⁶ (Fig. 1). Group differences between control and RLS groups were tested using ANOVA controlling either the ROI volume or subject age. Pearson partial correlations were obtained between tissue magnetic susceptibility and RLS clinical measurements including IRLS severity score, serum iron, serum ferritin and periodic limb movement during sleep (PLMS) on the 2nd of two-consecutive-nights sleep studies controlling for ROI volume (for full ROI) and age.

RLS compared to controls showed significantly lower susceptibility in the dentate nucleus (DN) (p<0.01) and in the thalamus (p<0.05). There were no significant differences in the serum ferritin or serum iron in control versus RLS group. In most brain regions, no significant correlations were found between brain susceptibility and serum iron or serum ferritin except some weak correlations in caudate nucleus (CN) and putamen (PUT) (Table 1). Brain tissue magnetic susceptibility does not correlate with IRLS severity score, but a significant negative correlation was found between susceptibility in SN and PLMS measures. This finding persists with both automatically selected ROI and fixed-shape ROI (Fig. 2). In addition, for a subgroup of RLS patients (n=11) with severe clinical feature of PLMS/hr ≥ 100, a significant decrease of SN susceptibility was noticed as compared to control group (p<0.05 with automated ROI and p<0.01 with fixed shape ROI).As the iron homeostasis inside the CNS is regulated by multiple mechanisms, the lack of strong correlations between the measured tissue susceptibility and serum iron or serum ferritin is not fully unexpected. As compared to previous studies^4-6, our current results suggest possible iron deficiency also in the DN and thalamus in RLS. While previous RLS studies have indicated that SN is the main region affected by iron deficiency in RLS⁴, the current study seems to suggest that the iron deficiency in SN happens mostly in those RLS patients with severe PLMS symptoms. The strong negative correlation between tissue susceptibility and PLMS may also indicate the important role of iron in the pathophysiology driving the PLMS motor sign of RLS.The current study suggests lower brain iron in DN and thalamus in primary RLS and in SN in RLS patients with severe PLMS, though the serum iron and serum ferritin levels do not differ significantly. The iron status in SN as measured by magnetic susceptibility has been found to significantly correlate with PLMS, which is the motor sign of RLS.