MRI signatures in the brain of patients with PD and iRBD
Silvia Mangia1, Philip Burton1, Alena Svatkova2,3, Igor Nestrasil4, Alejandra Sierra Lopez5, Karin Shmueli6, Lynn Eberly7, Michael Howell4, Paul Tuite4, and Shalom Michaeli1

1CMRR, Department of Radiology, University of Minnesota, Minneapolis, MN, United States, 2Department of Pediatrics, University of Minnesota, Minneapolis, MN, United States, 3Central European Institute of Technology (CEITEC), Masaryk University, Brno, Czech Republic, 4Department of Neurology, University of Minnesota, Minneapolis, MN, United States, 5A.I.Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland, 6Department of Medical Physics and Biomedical Engineering, University College London, London, United Kingdom, 7Division of Biostatistics, University of Minnesota, Minneapolis, MN, United States

Synopsis

The idiopathic rapid eye movement sleep behavior disorder (iRBD) is a condition that often evolves into Parkinson’s disease (PD), therefore by monitoring iRBD one can track the neurodegeneration of individuals that may progress to PD. Here we used a battery of MRI contrasts to characterize brain tissue properties such as microstructural integrity, iron loads, and functional connectivity in 10 iRBD, 10 PD and 10 age-matched healthy subjects. Rotating frame relaxation methods adiabatic T1,2ρ and RAFFn, along with DTI and rsfMRI detected heterogeneous abnormalities in several subcortical structures of PD and iRBD subjects.

Purpose

Parkinson’s disease (PD) is associated with abnormal synuclein protein deposits, iron deposition, gliosis and neuronal loss. Idiopathic rapid eye movement sleep behavior disorder (iRBD) is another condition that arises from underlying pontomedullary synuclein pathology which then evolves into PD or another synucleinopathy in up to 90% of cases [1-6]. Therefore by monitoring iRBD one may track neurodegeneration in individuals as they progress to PD. It is thought that iRBD pathology spreads through transmission of pathogenic forms of α-synuclein protein in a caudal-rostral manner [4], similarly to what Braak et al. proposed for PD [7]. Thus, it is crucial to have sensitive imaging methods that can detect the heterogeneous changes that ultimately extend throughout different brain regions and affect different cell populations. In this work we employed a battery of MRI modalities that are sensitive to various brain tissue properties, such as microstructural integrity, iron loads, and functional connectivity. Namely, we used diffusion tensor imaging (DTI), our novel rotating frame relaxation mapping methods including adiabatic T1ρ [8], T2ρ [9] and RAFF4 (Relaxation Along a Fictitious Field in the rotating frame of rank 4) [10], and resting state fMRI (rsfMRI). Here we present the results of a cross-sectional study conducted at 3T on 10 control, 10 iRBD and 10 PD subjects.

Methods

Ten mild-moderate advanced (Hoehn & Yahr Stage I-II; on medications) individuals with PD (67±6 years, 5F/5M, UPDRS=39), ten iRBD subjects (64±7 years, 5F/5M, UPDRS=15) and ten age-matched healthy controls (58±6 years, 4F/6M) underwent MRI using a 3 T Siemens Prisma system. Adiabatic T1ρ, T2ρ and RAFF4 measurements were collected from 30 AC-PC aligned slices covering brainstem and basal ganglia using segmented GRE readout (4 segments), voxel size: 1.6x1.6x3.6 mm3, GRAPPA= 3, TE=3.18 ms; TR=2s. T1-weighted, T2-weighted images, DTI, T2* and rsfMRI were collected in a whole-brain fashion. For adiabatic relaxation measurements, hyperbolic secant pulses were used with BW=1.6 kHz, pulse duration Tp=6 ms, ω1max /(2π)=800 Hz, 5 acquisitions with number of pulses = 0, 4, 8, 12, 16, MLEV4 phase cycling; for RAFF4, Tp was 4.56 ms for one P-packet, number of P-packets 0, 4, 8, 12, 16, ω1max /(2π)=323 Hz. Parameters for T2*: voxel size: 1.6x1.6x1.6 mm3, 96 slices, TR=51 ms, TE=10, 19, 27, 36, 45 ms, GRAPPA 3; Flip Angle 15o; for DTI: 128 directions, with 5 additional non-diffusion weighed (b0) images, b-value=1500s/mm2, voxel size 1.8x1.8x1.8mm3, TR=2820 ms, TE=72.6 ms; multi band (MB)=4; for rsfMRI: EPI, TR=900 ms, multi band (MB)=4; TE=30 ms; voxel size=3x3x3 mm3, matrix size=64x64, 48 AC-PC aligned slices with interleaved slice acquisition, 502 volumes. The MRI relaxation and DTI parameters were calculated from 12 regions of interest (ROIs) defined in subject space, and compared among the subject groups. Functional connectivity measures were also extracted using seed analyses from 3 out of 12 ROIs, namely the caudate, substantia nigra (SN) and midbrain (regions of primary analysis). Statistical significance of group differences was inferred by unpaired two-tailed t-test.

Results and discussion

Statistically significant differences of MRI relaxation and DTI parameters among groups were observed in regions of primary analysis including SN, caudate, and midbrain (Fig. 1). Interesting differences of MRI parameters were simultaneously observed also in other brain regions including amygdala and parahippocampal gyrus (PG). Observed differences of MRI relaxation and DTI parameters were consistent with neuronal degeneration and iron accumulation in multiple brain regions [11-13]. In particular, gradual trends across controls vs iRBD vs PD of T2*, T1ρ and T2ρ in PG and caudate might suggest progression of iRBD into PD. In addition, we observed significant differences between controls and PD of functional connectivity measures in the networks associated with SN and midbrain (Fig. 2), in agreement with previous observations of disrupted nigrostriatal connectivity induced by iRBD and PD [14, 15].

Conclusion

The chosen MRI modalities detect group differences in multiple subcortical brain regions of iRBD and PD subjects vs control individuals. Overall, having MRI techniques which are sensitive to iron, such as T2ρ and T2*, or sensitive to neuronal loss such as T1ρ, or sensitive to myelination such as RAFF4, as well as having the ability to image extended brain coverage, is critical for evaluating PD and iRBD. The rotating frame adiabatic T1ρ and T2ρ techniques are expected to provide greater sensitivity to detect tissue changes rapidly as compared to conventional MRI methods.

Acknowledgements

NIH grants: P41 EB015894, P30 NS076408, UL1TR000114. Academy of Finland, Sigrid Juselius Foundation.

References

[1] Bunzeck et al. Parkinsonism Relat Disord 2013;19:1136. [2] Schenck et al. Sleep 1986;9:293. [3] Schenck et al. Neurology 1996;46:388. [4] Boeve BF. Lancet Neurol 2013;12:469-82. [5] Boeve et al Mov Disord 2001;16:622. [6] Luk et al. Science 2012;338:949. [7] Braak et al. Neurobiol Aging 2003;24:197. [8] Michaeli et al. J Magn Reson 2006;181:135. [9] Michaeli et al. JMR 2004;169:293. [10] Liimatainen et al. MRM 2014:doi: 10.1002/mrm.25129. [11] Lach et al. Acta neuropathologica 1992;83:352. [12] Jia et al. AJNR 2015. [13] He et al. Hum Brain Mapp 2015:doi: 10.1002/hbm.22928. [14] Ellmore et al. Sleep 2013;36:1885. [15] Rieckmann et al. NeuroImage Clinical 2015;8:554.

Figures

Figure 1. MRI relaxation and DTI parameters measured in control, iRBD and PD subjects (mean±SEM, n=10/group); left and right ROIs were pooled together where applicable. SN: substantia nigra, PG: parahippocampal gyrus, VDC: ventral diencephalon. Even if significance was not reached, a clear trend in SN of longer T1ρ and a trend in PG of shorter T1ρ and T2ρ (iRBD and PD vs. controls) were observed. * p<0.05, uncorrected for multiple comparisons.

Figure 2. Functional connectivity parameters measured in control, iRBD and PD subjects (mean ± SEM, n=10/group). Mean correlation coefficients are calculated from all voxels within the network associated with the seed placed in each ROI of primary analysis, providing a global measure of functional connectivity of the network associated with that ROI. * p<0.05,uncorrected for multiple comparisons.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3401