RESEARCH

ARTICLE

A Longitudinal Study of Magnetic Resonance Spectroscopy Huntington’s Disease Biomarkers Aaron Sturrock, MRCP,1,2 Corree Laule, PhD,3 Katy Wyper, MSc,3 Ruth A. Milner, MSc,4 Joji Decolongon, MSc,2 Rachelle Dar Santos, BSc,2 Allison J. Coleman, BSc,2 Kimberley Carter, BSc,2 Susan Creighton, MSc,2 Natalie Bechtel, MD,5 Stefan Bohlen, MD,5 Ralf Reilmann, MD,5,6 Hans J. Johnson, PhD,7 Michael R. Hayden, PhD,1,2 Sarah J. Tabrizi, PhD,8 Alex L. Mackay, DPhil,3 and Blair R. Leavitt, MDCM, FRCP(C)1,2* 1

Centre for Molecular Medicine & Therapeutics, Vancouver, Canada Centre for Huntington Disease, University of British Columbia (UBC) Hospital, Vancouver, Canada 3 UBC MRI Research Centre, UBC Hospital, Vancouver, British Columbia, Canada 4 Child & Family Research Institute (CFRI), Vancouver, British Columbia, Canada 5 €nster, Mu €nster, Germany Department of Neurology, University of Mu 6 €nster, Germany George-Huntington-Institute, Mu 7 University of Iowa Carver College of Medicine, Iowa City, Iowa, USA 8 UCL Institute of Neurology, University College London, Queen Square, London, UK

2

ABSTRACT:

Putaminal metabolites examined using cross-sectional magnetic resonance spectroscopy (MRS) can distinguish pre-manifest and early Huntington’s Disease (HD) individuals from controls. An ideal biomarker, however, will demonstrate longitudinal change over short durations. The objective here was to evaluate longitudinal in vivo brain metabolite profiles in HD over 24 months. Eighty-four participants (30 controls, 25 pre-manifest HD, 29 early HD) recruited as part of TRACK-HD were imaged at baseline, 12 months, and 24 months using 3T MRS of left putamen. Automated putaminal volume measurement was performed simulta-

neously. To quantify partial volume effects, spectroscopy was performed in a second, white matter voxel adjacent to putamen in six subjects. Subjects underwent TRACK-HD motor assessment. Statistical analyses included linear regression and one-way analysis of variance (ANOVA). At all time-points N-acetyl aspartate and total N-acetyl aspartate (NAA), neuronal integrity markers, were lower in early HD than in controls. Total NAA was lower in pre-manifest HD than in controls, whereas the gliosis marker myo-inositol (MI) was robustly elevated in early HD. Metabolites were stable over 24 months with no longitudinal change. Total NAA

------------------------------------------------------------------------------------------------------------------------------

*Correspondence to: Blair R. Leavitt, Vancouver, BC, V5Z 4H4, Canada, E-mail: [email protected]

Funding agencies: This study was privately sponsored by CHDI, a notfor-profit organization dedicated to finding treatments for Huntington’s disease. NIH Grant Number; 20R-03189.

Medivation, Wyeth, ISIS Pharma, Link Medicine, Prana Biotechnology, the Cure Huntington’s Disease Initiative Foundation, MEDA Pharma, Temmler Pharma, AOP Orphan Pharmaceuticals AG. He received grant support from the Cure Huntington’s Disease Initiative Foundation (CHDI), the Deutsche Forschungsgemeinschaft (DFG), the European Union (EUFP7 program) and the European Huntington’s Disease Network (EHDN).

Full financial disclosures and author roles may be found in the online version of this article.

Dr. Johnson reports no disclosures.

Relevant conflicts of interest/financial disclosures: Nothing to report.

Dr. Hayden reports no financial disclosures. He receives funding from the Canadian Institutes of Health Research, Genome BC and CHDI.

Dr Sturrock was formerly a full-time Clinical Research Fellow on the TRACK-HD study. He received funding from CHDI. Dr. Laule, Miss Wyper and Ms. Milner report no disclosures.

Professor Sarah Tabrizi reports no financial disclosures. She receives funding from the Medical Research Council (MRC), The Wellcome Trust and CHDI.

Ms. Decolongon, Miss Dar Santos and Miss Coleman are Clinical Research Co-ordinators. Miss Carter was a research assistant. Each receive(d) a portion of their funding from CHDI.

Dr. MacKay’s research program is supported by operating grants from the Natural Sciences and Engineering Research Council of Canada and the Multiple Sclerosis Society of Canada.

Ms Creighton reports no disclosures.

Dr. Leavitt reports no financial disclosures. He receives research funding from CHDI, the Canadian Institutes of Health Research, and the Michael Smith Foundation.

Dr. Bechtel has received travel grants from the “Deutsche Gesellschaft €r Neurologie” and the European Huntington’s Disease Network (EHDN). fu Dr. Bohlen reports no disclosures.

Received: 5 May 2014; Revised: 6 October 2014; Accepted: 23 October 2014

Dr. Reilmann provided consulting services, advisory board functions, clinical trial services, quantitative motor analyses, and/or lectures for Novartis, Pfizer, Siena Bitoech, Neurosearch Inc., Ipsen, Teva, Lundbeck,

Published online 18 February 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/mds.26118

Movement Disorders, Vol. 30, No. 3, 2015

393

S T U R R O C K

E T

A L

was not markedly different in adjacent white matter than putamen, arguing against partial volume confounding effects in cross-sectional group differences. Total NAA correlations with disease burden score suggest that this metabolite may be useful in identifying neurochemical responses to therapeutic agents. We demonstrate almost consistent group differences in putaminal metabolites in HD-affected individuals compared with

The eventual development of Huntington’s disease (HD) symptoms can be predicted through detection of CAG trinucleotide repeat expansion in the HTT gene1 in ‘premanifest’ gene carriers. However, definitive HD biomarkers, able to objectively identify disease onset and progression, are lacking. Evaluation of potential biomarkers are the focus of longitudinal studies such as TRACK-HD2,3 and PREDICT-HD.4 One putative biomarker modality is magnetic resonance spectroscopy (MRS). We recently reported baseline results from the largest MRS study in HD to date. This study was performed at 3T, a higher field strength than most MRS studies in HD, in a homogeneous cohort comprising controls and individuals with premanifest and early symptomatic HD.5 Importantly, this study clarified discrepancies regarding the utility of MRS as an HD biomarker modality.6-13 A measure of neuronal integrity, N-acetyl aspartate (NAA), was reduced in premanifest HD (Pre-HD) and early HD, whereas myo-inositol (MI), a gliosis marker, was elevated in early HD. Furthermore, these metabolites correlated differentially with motor performance and disease burden. These evaluations were performed in the University of British Columbia, Vancouver cohort of the multinational TRACK-HD observational study. We report here longitudinal MRS findings from the same cohort over 24 months.

Methods Subjects All subjects were recruited and evaluated at University of British Columbia, Vancouver, as part of TRACK-HD.2,3 Participants were initially enrolled between January and August 2008. Participants were early HD, pre-manifest (Pre-HD) or controls (partners/ unaffected relatives of HD subjects). Early HD individuals had received genetic diagnoses and been diagnosed with clinical disease onset by an HD specialist, but maintained independent function compatible with early disease based on Unified HD Rating Scale–Total Functional Capacity (UHDRS-TFC) (early stage 1 TFC  11-13 or stage 2 TFC  7-10).14 Pre-HD recruitment was based on (1) CAG repeat expansion within HTT gene 40 CAG or greater; (2) UHDRS

394

Movement Disorders, Vol. 30, No. 3, 2015

controls over 24 months. Future work establishing spectroscopy as an HD biomarker should include multisite assessments in large, pathologically diverse C 2015 International Parkinson and Movement cohorts. V Disorder Society

K e y W o r d s : Huntington’s disease; MRS; MRI; Nacetyl aspartate (NAA); myo-inositol (MI)

motor component score 5 or less, indicating absence of marked motor deficits; (3) Disease Burden Score (DBS) 220 or greater at baseline. Disease Burden Score, an index of mutant huntingtin protein exposure, calculated as (age * [CAG – 35.5]),15 was used to select pre-HD individuals closer to predicted disease onset to maximize comparisons with controls. Of 25 pre-HD individuals at baseline, 22 had DBS of greater than 250 with scores 220 to 240 in the remaining three. Pre-HD subjects were divided into pre-HD A/B (further/nearer to clinical diagnosis) subgroups based on the TRACK-HD median (10.8 years) to predicted onset.2

Standard Protocol Approvals, Registrations, and Patient Consents The Study was approved by UBC Human Research Ethics Committee. Written informed consent was obtained from each participant.

Procedures Participants were assessed using UHDRS-99 motor component, performed by a trained rater (A.S.). The DBS was calculated for genetically diagnosed individuals. TRACK-HD employs novel quantitative motor assessments, including tongue force measurement (glossomotography) and force transducer–based selfpaced tapping tasks (digitomotography). Glossomotography assesses isometric forces during sustained tongue protrusion using a force transducer. Subjects matched target force (0.5N) presented by a monitor for 20 seconds (four trials). Tongue pressure variability (coefficient of variation) and contact time (% time tongue was contacting transducer) were calculated. During digitomotography, timing precision for selfpaced tapping was recorded. Subjects matched auditory cues (1.8Hz rate) and continued tapping at the same frequency after cues stopped (five trials). Variability of deviations of (1) tap initiation from predefined tapping frequency; (2) interval between midpoints of consecutive taps were calculated for selfpaced periods. These tasks are sensitive to pre-HD motor impairment.2 Because we were interested in left hemisphere MRS changes, right-handed self-paced tapping was used.

M R S

H D

B I O M A R K E R S

FIG. 1. Sample WM voxel placement. Please note that this was adjacent to the putaminal voxel, placement of which is demonstrated as outlined in Sturrock et al.5 [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Subjects underwent 3T magnetic resonance imaging (MRI) examination (Philips Achieva MR scanner). Spectra were obtained using single-voxel PRESS sequence in left putamen. Scan parameters included 3.5 cm 3 1 cm 3 1.5 cm (5.25 cm3) voxel size. A voxel of this size inevitably incorporated adjacent white matter (WM); however, no other gray matter or cerebrospinal fluid (CSF) was included. Repetition time (TR) 5 2000 ms, Echo time (TE) 5 35 ms, 1,024 samples performed/spectrum with 128 signal averages. Spectral bandwidth 5 2,000 Hz. Second-order shimming was performed. Water concentration 5 43,300 mM. The LCModel basis set for Philips was used for analysis. T1-weighted images were collected for volumetric analyses using 3DT1 Fast Field Echo (TR 5 7.7 ms, TE 5 3.5 ms, voxel size 5 1.1 3 1.1 3 1.1 mm3; 164 slices). Time constraints prevented repeat scanning in which spectra were inadequate/affected by movement. If TRACK-HD study images were also poor, that is, they did not meet quality control criteria, subjects were recalled for re-examination and repeat spectra used (see Supplemental Data for details). Spectra were fit using LCModel16 and metabolites normalized to unsuppressed water spectrum. LCModel estimates “concentration” measurement reliability, returning standard deviations (%SD) for each metabolite. Standard deviations below 20% are considered reliable,17 and higher %SD were excluded. Three metabolites of interest were examined: separate tNAA (summed NAA 1 N-acetyl aspartyl glutamate (NAAG)), NAA measurements (neuronal integrity marker), and MI (an astrocyte marker). As an ad-hoc analysis to qualitatively assess WM contribution to putaminal voxel metabolites, in a subset of quasi-randomized patients (2 controls, 3 preHD, and 1 early HD) returning for 36-month evaluation, an identical second MRS voxel was placed within peri-putaminal (predominantly) WM directly superior to the original putaminal voxel. To assess relative biomarker sensitivities of putaminal MRS against putaminal volume, measurement compari-

sons were performed against 3T MRI data. These included T1- and T2-weighted volumetric sequences.2 IXICO (London, UK) performed rigorous quality control to ensure scan stability over time. Putaminal volume was derived with fully automated segmentation/ subtraction with Brain Research: Analysis of Images, Networks, and Systems software (V2) (BRAINS, Iowa City, IA, USA).18 Segmentations and registrations were visually inspected by image analysts with blinding to ensure accurate tissue delineation (Fig. 1).

Statistics At baseline, participants were sex-matched across groups, across TRACK-HD sites. Across sites age among controls was balanced against combined age distribution of generally younger pre-HD and older early HD subjects. Group comparisons were made using one-way analysis of variance (ANOVA) with post-hoc NewmanKeuls Multiple Comparison Tests. Longitudinal assessments were performed where a subject’s paired evaluations were available at all time-points, using paired Student’s t tests with Bonferroni correction. Scatter plots and linear regression were used to compare metabolites against DBS and motor outcomes. Linear correlations were investigated using the method of least squares.

Results Subject Demographics Subject demographics are displayed in Table 1.

Cross-Sectional Analyses at Baseline, 12, and 24 Months At all time points, NAA, tNAA, and MI demonstrated reproducible group differences (Fig. 2). Of the metabolites, tNAA demonstrated almost consistent differences between pre-HD and control subjects, ranging from 2.7% (NS) to 7.1% (P < 0.01) lower in pre-HD versus controls. All metabolites were consistently altered in early HD. Sample spectra are provided in Supplemental Data.

Movement Disorders, Vol. 30, No. 3, 2015

395

S T U R R O C K

E T

A L

TABLE 1. Demographics of successfully scanned subjects Controls

Pre-HD

Early HD

Baseline (V1) Successful scans (n 5 84/85) Women % Age (yrs) (Mean/SD) Subgroups (N 5)

30/30 63 46.60 (11.9) —

25/25 56 39.90 (10.8)* Pre-HD A 5 15 Pre-HD B 5 10

29/30 28 48.16 (10.9) Early HD 1 5 13 Early HD 2 5 16

12 Months (V2) Successful scans (n 5 76/77) Women % Age (yrs) (Mean/SD) V1-V2 interscan interval (yrs) (Mean/SD) Subgroups (N 5)

28/28 64 47.98 (11.2) 0.97 (0.1) —

25/25 60 41.10 (10.2)* 0.92 (0.1) Pre-HD A 5 16 Pre-HD B 5 9

23/24 35 49.40 (10.7) 0.99 (0.1) Early HD 1 5 10 Early HD 2 5 13

24 Months (V3) Successful scans (n 5 74/76) Women % Age (yrs) (Mean/SD) Mean V2-V3 interscan interval (yrs) (Mean/SD) Subgroups (N 5)

27/27 59 49.17 (11.3) 1.10 (0.2) —

22/24 59 40.59 (10.8)** 1.06 (0.1) Pre-HD A 5 15 Pre-HD B 5 7

24/25 33 51.60 (10.7) 1.08 (0.1) Early HD 1 5 12 Early HD 2 5 12

Note that all subjects were evaluated as per their original subgroups (Pre-HD A/B or Early HD 1/2) regardless of their progression to the alternative subgroup over the duration of the study. This was to optimize recognition of longitudinal effects. Asterisks refer to Pre-HD age differences vs. Control and Early HD groups; *P < 0.05; **P < 0.01 (same level of significance for both comparisons). Ages were calculated to the nearest month. HD, Huntington’s disease.

FIG. 2. (A-C) Cross-sectional MRS in control, pre-HD, and early HD subjects at baseline, 12, and 24 months. Data arranged by metabolite. (A) NAA; (B) tNAA; (C) MI. Error bars denote standard error of the mean for each group. Asterisks indicate level of significance; *P < 0.05, ** P < 0.01, *** P < 0.001. Group comparisons were made using one-way ANOVA with post-hoc Newman-Keuls Multiple Comparison Tests. MRS, magnetic resonance spectroscopy; HD, Huntington’s disease; NAA, N-acetyl aspartate; tNAA, summed NAA 1 NAAG; MI, myo-inositol; ANOVA, analysis of variance. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

396

Movement Disorders, Vol. 30, No. 3, 2015

M R S

H D

B I O M A R K E R S

TABLE 2. Correlations of metabolite concentration with DBS and motor performance in HD Early HD

Baseline DBS UHDRS-TMS Metronome tapping: Self-paced tap precision Metronome tapping: Deviation of tap initiation Tongue contact time* Tongue pressure variability Speeded tapping: Tap duration Speeded tapping: Repetition time 12 Months DBS UHDRS-TMS Metronome tapping: Self-paced tap precision Metronome tapping: Deviation of tap initiation Tongue contact time* Tongue pressure variability Speeded tapping: Tap duration Speeded tapping: Repetition time 24 Months DBS UHDRS-TMS Metronome tapping: Self-paced tap precision Metronome tapping: Deviation of tap initiation Tongue contact time Tongue pressure variability Speeded tapping: Tap duration Speeded tapping: Repetition time

Pre-HD & Early HD

tNAA

NAA

MI

tNAA

NAA

MI

— NS NS

— NS 0.71 (

A longitudinal study of magnetic resonance spectroscopy Huntington's disease biomarkers.

Putaminal metabolites examined using cross-sectional magnetic resonance spectroscopy (MRS) can distinguish pre-manifest and early Huntington's Disease...
466KB Sizes 2 Downloads 13 Views