Using Proton MR Spectroscopy and Quantitative Susceptibility Mapping (QSM) at 7 Tesla to Decipher Mitochondrial Membrane Protein–Associated Neurodegeneration (MPAN)
Ralf Mekle1, Florian Schubert1, Thoralf Niendorf2, Till Huelnhagen2, Antje Else2, Simon Daniel Robinson3, Bernd Ittermann1, Vince Madai4, Marta Skowronska5, Petr Dusek6,7, Jens Wuerfel8, and Susanne A. Schneider9

1Physikalisch-Technische Bundesanstalt (PTB), Braunschweig and Berlin, Germany, 2Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 3High Field MR Centre, Department of Biomedical Imaging and Image guided Therapy, Medical University of Vienna, Vienna, Austria, 4Department of Neurology and Center for Stroke Research Berlin, Charité Universitaetsmedizin, Berlin, Germany, 52nd Department of Neurology, Institute of Psychiatry and Neurology, Warsaw, Poland, 6Department of Neurology and Centre of Clinical Neuroscience, First Faculty of Medicine and General University Hospital in Prague, Charles University in Prague, Prague, Czech Republic, 7Institute of Neuroradiology, University Medicine Goettingen, Goettingen, Germany, 8Medical Image Analysis Center, Basel, Switzerland, 9Neurology Department, University of Kiel, Kiel, Germany

Synopsis

1H MR spectroscopy (MRS) in the white matter (WM) of the precentral region and quantitative susceptibility mapping (QSM) of basal ganglia acquired at 7 Tesla in patients (homozygotes), non-manifesting heterozygotes, and control subjects were used to investigate mitochondrial membrane protein-associated neurodegeneration (MPAN). MPAN is a rare, but severe disorder from the neurodegeneration with brain iron accumulation (NBIA) group. Comparison of results for the different subject groups showed no abnormalities in heterozygotes. In contrast, metabolic changes in patients detected by MRS, in particular an increase in glutamate, suggests an underlying mechanism for MPAN related to neurotransmission in corticospinal pathway.

Purpose

Mitochondrial membrane protein-associated neurodegeneration (MPAN) is a recently identified rare disorder from the neurodegeneration with brain iron accumulation (NBIA) group1. Symptoms include progressive spasticity, motor neuropathy, cognitive dysfunction, movement disorders, and optic atrophy. MPAN is putatively an autosomal recessive disorder caused by mutation in the c19orf12, a gene with yet unknown function. Information about possible subclinical changes in heterozygotes is lacking. Brain MRI in patients typically shows iron deposits in globus pallidus (GP) and substantia nigra (SN) without other abnormalities2. Yet, spastic paresis heralding the clinical picture suggests impairment of corticospinal pathway. The aim of this study was to apply 1H MR spectroscopy (MRS) in the white matter (WM) of the precentral region and quantitative susceptibility mapping (QSM) of basal ganglia at ultra-high (7T) field in patients (homozygotes) and non-manifesting heterozygotes to elucidate underlying metabolic and structural changes of MPAN.

Methods

All scans were performed on a 7T whole-body system (Siemens Healthcare, Erlangen, Germany) using a 24 channel RF head coil (Nova Medical Inc., Wilmington, MA). Three patients (18, 19, 23 yrs), 9 heterozygotes (50.4±9.8 yrs), and 9 matched control subjects (42.3±17.8 yrs) were scanned (one control subject was excluded from MRS analysis due to poor data quality). For MRS, localized RF calibration was performed, and first- and second-order shims were adjusted using a vendor-supplied algorithm. Single volume data from the left precentral WM were acquired using the spin echo full intensity acquired localized (SPECIAL) MRS technique3,4 with the following scan parameters: VOI = 20x20x20 mm3, TR/TE = 9000/11 ms, number of averages = 64, and Tacq = 512 ms. Metabolite quantification was performed using LCModel5. Resulting concentrations were corrected for the CSF content of each VOI. QSM maps were reconstructed from 3D gradient echo (GRE) acquisitions with TR/TE = 30/15.3 ms, flip angle = 30º, in-plane resolution = 0.3x0.3 mm2, slice thickness = 1 mm, BW = 180 Hz/pixel. Phase images were reconstructed offline using a virtual reference coil approach6. The QSM reconstruction was performed using the STI Suite 2.2 and consisted of Laplacian unwrapping, variable SHARP filtering (starting radius=10 mm) and iLSQR inversion7. ROIs were manually drawn in GP and SN. Results from MRS and QSM were compared for control subjects and heterozygotes and patients, respectively using a non-parametric Mann-Whitney test.

Results

Localized shimming resulted in water linewidths of 11.4 ± 0.9 Hz, 11.4 ± 0.9 Hz, and 12.2 ± 0.4 Hz for controls, heterozygotes, and patients, respectively. The high signal-to-noise ratio (SNR) of the spectra (Fig.1) allowed the quantification of 13 individual and 5 combined metabolites with Cramér-Rao lower bounds (CRLBs) < 20% for all three groups, including GABA, glutamine (Gln), glutamate (Glu), and lactate (Lac). No significant differences between metabolite concentrations for heterozygotes and controls were detected, where only for total choline (tCho) the Mann-Whitney test result exactly yielded the limits of critical values. Performing a two-tailed Student’s t-test for tCho in addition yielded p>0.14 though, rather indicating no significant difference. Comparing patients versus controls resulted in significant increases for Glu, taurine (Tau), and Glu+Gln. Mean concentrations from heterozygotes and controls are shown together with the individual results for patients in Table 1.

Discussion

In this study, differences in metabolite quantification for patients versus controls were observed. These have to be considered with caution due to the small number of patient cases. The increase in Tau might be partially attributed to the relatively young age of the patient group, since Tau decreases with age8. However, for the first time, using MRS increased Glu levels in MPAN patients were detected. This suggests a dysfunctional regulation of excitatory neurotransmission. No differences between heterozygotes and controls with respect to results from 7T MRS and QSM were found indicating that one functional copy of gene c19orf12 is sufficient to prevent iron accumulation and metabolic changes in the brain of heterozygotes. In contrast, quantitative measurements suggest that iron concentration is 3-4 times increased in GP and SN of MPAN patients compared to controls.

Conclusion

Using MRS and QSM at 7T allowed a more detailed characterization of MPAN heterozygotes and patients. While no abnormalities were detected in heterozygotes, metabolic changes in homozygotes detected by MRS indicated an underlying mechanism related to neurotransmission in corticospinal pathway.

Acknowledgements

No acknowledgement found.

References

1. P.Dusek et al., NeurolClin, 33(1): 175-204, 2015; 2. M. Hartig et al., Int Rev Neurobiol, 110: 73-84, 2013; 3. V. Mlynarik et al., MRM, 56(5): 965-70, 2006; 4. R. Mekle et al., MRM, 61(6): 1279-85, 2009; 5. S.W. Provencher et al., MRM, 30(6): 672-79, 1993; 6. D.L. Parker et al., MRM, 72(2): 563-9, 2014; ; 7. W.Li et al., Neuroimage, 108:111-122, 2015; 8. V. Govindaraju et al., NNR in Biomed, 13(3): 129-153, 2000.

Figures

Fig. 1. 1H MR spectra from a VOI in the precentral WM (inset) acquired at 7T with the SPECIAL sequence (TR/TE = 9000/11 ms) for a MPAN patient (left) and an age-matched control (right). Note the high SNR and comparable data quality of the two data sets.

Table. 1. Metabolite quantification. The unsuppressed water signal was used for referencing. Concentrations are given as mean values±standard deviations for 8 control subjects and 9 heterozygotes together with the individual results for three patients. All concentrations were determined with CRLBs < 20%. Differences between concentrations from controls and patients are in bold.

Fig. 2. QSM results: sample axial susceptibility maps for A) control and B) patient showing higher susceptibility in GP of the patient (blue arrows); insets show images of mesencephalon including SN with higher susceptibility in the patient (blue arrows). C) Comparison of susceptibility values for all groups in GP and SN; *** p<0.0001.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4016