- Email: [email protected]
Is brain lactate increased in Huntington's disease?
Journal of the Neurological Sciences 263 (2007) 70 – 74 www.elsevier.com/locate/jns
Is brain lactate increased in Huntington's disease? W.R. Wayne Martin a,⁎, Marguerite Wieler a , Christopher C. Hanstock b a
b
Division of Neurology, University of Alberta, Edmonton, Alberta, Canada Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada Received 22 April 2007; accepted 30 May 2007 Available online 25 July 2007
Abstract Impaired brain energy metabolism with increased regional brain lactate may play a role in the pathogenesis of Huntington’s disease (HD). Magnetic resonance spectroscopy (MRS) has provided conflicting evidence, however, regarding metabolic changes. Our objective was to evaluate the potential contribution of CSF lactate to the changes observed with MRS in HD. We performed single voxel MRS at 3 T in 23 patients with HD and 28 age-matched control subjects using a method to segment voxels into grey matter, white matter, and CSF, and to extrapolate regional lactate content to a hypothetical voxel containing 100% brain in order to control for differences in CSF lactate. Lactate/ creatine and lactate/N-acetyl aspartate (Lac/NAA) ratios were significantly increased in parieto-occipital ( p b 0.05) and cerebellar ( p b 0.01) voxels in HD patients. After extrapolating group Lac/NAA results to a theoretical voxel containing 100% brain, this ratio was greater in the HD group than the control group, suggesting possibly increased lactate in this predicted voxel, although the difference between groups did not reach statistical significance. These results suggest an increase in brain lactate content in manifest HD, in a regionally non-specific fashion, although the possibility of a CSF contribution to this increase cannot be ruled out. Regardless, this supports the possibility of impaired mitochondrial function resulting in abnormal brain energy metabolism in HD. © 2007 Elsevier B.V. All rights reserved. Keywords: Huntington’s disease; Magnetic resonance spectroscopy; Energy metabolism; Lactate; Mitochondrial function
1. Introduction Huntington's disease (HD) is a progressive, inherited neurodegenerative disorder characterized by abnormal involuntary movements, cognitive impairment, and behavioral disturbances. Typical neuropathological changes include marked neuronal loss from the caudate and putamen, particularly affecting medium-sized spiny neurons. Although most descriptions of pathology have emphasized the striatal changes, there is evidence of more widespread neuronal loss, particularly in advanced disease [1]. The genetic abnormality underlying the development of manifest disease has been identified [2], but the pathophysiological mechanisms that link this abnormality to neuronal loss remain uncertain. ⁎ Corresponding author. Movement Disorders Clinic, Glenrose Rehabilitation Hospital, 10230 - 111 Ave., Edmonton, AB, Canada T5G 0B7. Tel.: +1 780 735 8805; fax: +1 780 735 8804. E-mail address: [email protected] (W.R.W. Martin). 0022-510X/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2007.05.035
Impaired energy metabolism may play a role in the pathogenesis of HD. Measurements of mitochondrial electron transport enzymes have shown abnormalities in platelets and postmortem brain tissue [3–5]. Animal models have shown that inhibition of oxidative metabolism can produce a pattern of striatal neuronal loss that mimics that seen in HD [6–8]. Using magnetic resonance spectroscopy (MRS), Jenkins and co-workers reported an increase in lactate content in occipital cortex in HD, providing further support for a defect in brain energy metabolism [9]. These observations have been challenged, however, by the suggestion that the reported increase in tissue lactate may reflect contamination from ventricular CSF [10]. The objective of the present study was to replicate and expand on the studies of Jenkins et al. [9] evaluating the potential contribution of CSF lactate to the changes observed with single voxel MRS in HD. We determined brain and CSF lactate levels in two groups of patients and age-matched controls. In the first group, metabolite concentration in a
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
parieto-occipital voxel was evaluated, using methods similar to those published previously [9]. To evaluate the potential contribution of CSF lactate content in more detail, we subsequently studied a second group, using a method to segment voxels into grey matter, white matter, and CSF, and to extrapolate regional lactate content to a hypothetical voxel containing 100% brain (no CSF). 2. Methods
71
subjects, a similar parieto-occipital VOI was placed, but without restricting contribution from adjacent CSF. The same PRESS sequence was used for segmentation, to allow measurement of the relative percentage of grey matter (GM), white matter (WM) and CSF from within the selected voxel. For this, the PRESS sequence used a double inversion recovery 1D-projection method to allow selection of each compartment based on differences in T1 relaxation [15]. This allows simultaneous nulling of the signal from 2 compartments, leaving only that from the third.
2.1. Subjects 2.3. Data analysis Ambulatory, functionally independent patients with manifest HD were recruited from the Movement Disorders Clinic at the Glenrose Rehabilitation Hospital. All patients had a positive family history of HD, chorea as their predominant disease manifestation, and DNA analysis demonstrating more than 39 CAG repeats. Clinical disease severity was rated with the motor score from the Unified Huntington's Disease Rating Scale (UHDRS) [11] and the Total Functional Capacity (TFC) score [12]. Control subjects were drawn from a population that was not at risk for HD, and were healthy and free of neurological or psychiatric disease. The study was approved by the Human Research Ethics Board of the University of Alberta and all subjects gave informed consent.
Following filtering with a 2 Hz exponential multiplication, spectroscopic data were Fourier transformed, and phase and baseline corrected. Lactate (Lac), N-acetyl aspartate (NAA), creatine (Cr) and choline (Cho) peaks were fitted to Laurentz–Gaussian lineshapes, minimizing the residual. Results were expressed as the ratio between metabolites. In the second group of patients, the segmentation data were utilized to allow extrapolation to a theoretical voxel containing 100% brain (by summing the contribution from GM and WM), thereby eliminating the potential contribution from lactate-containing CSF. All data manipulation was performed in the MATLAB (The MathWorks, Inc., Natick, MA) program environment.
2.2. Data acquisition
3. Results
MRS data were acquired using a Magnex 3 tesla magnet (Magnex Scientific Ltd., Abingdon, U.K.) with actively shielded gradients, controlled with a Surrey Medical Imaging Systems spectrometer console (Surrey Medical Imaging Systems, Ltd., Guildford, U.K). Signal transmission and reception were achieved using a quadrature birdcage resonator. In all subjects, multi-slice gradient echo images were initially acquired in transverse, sagittal and coronal planes (slice thickness 5 mm, 1 mm gap, echo time (TE) 20 ms, repetition time (TR) 500 ms). In the first group, a volume of interest (VOI) measuring 2 × 2 × 3 cm was then registered to the medial parieto-occipital lobe of the left hemisphere, avoiding CSF as much as possible. A second 2 × 2 × 3 cm VOI was registered to the cerebellum. Shimming of the selected volumes was accomplished with FASTMAP [13]; linear shims were optimized with an “in-house” automatic routine, resulting in a shimmed water linewidth typically b 4 Hz (∼ 0.03 ppm). Water suppressed point-resolved spectroscopy (PRESS) was performed from the selected volumes with TE 272 ms, TR 3 s, and 512 averages (total acquisition time = 27 min), each 4 averages phase-cycled through the CYCLOPS scheme [14]. A second group of subjects was subsequently studied with a tissue segmentation routine in addition to the MRS sequence in order to evaluate the potential contribution of the CSF compartment to brain lactate estimates. In these
Twenty-three patients with HD and 28 healthy controls were studied. Thirteen patients and 17 controls comprised the first group; 10 patients and 11 controls comprised the second. Data were acquired from two differently placed voxels in one patient from the second group, yielding a total of 11 tissue volumes for extrapolation to a theoretical voxel containing 100% brain. There was no statistically significant difference in age or sex distribution. Clinical data are summarized in Table 1. In the first group, suitable spectra were available for analysis from both VOI's in all 13 patients. These data were compared to spectra from parieto-occipital VOI's in 14 controls, and from cerebellar VOI's in 11 controls (since not all Table 1 Clinical characteristics of patients and healthy control subjects Group 1
N % male Age mean ± SD (yr) Symptom duration mean ± SD (yr) UHDRS motor score mean ± SD TFC mean ± SD
Group 2
HD
Control
HD
Control
13 46 52.3 ± 9.7 7.4 ± 5.6
17 41 45.8 ± 11.3 –
10 50 50.4 ± 7.6 6.0 ± 4.7
11 45 46.0 ± 9.4 –
32.3 ± 12.3
–
32.5 ± 12.8
–
10.4 ± 2.8
–
10.4 ± 1.8
–
UHDRS: Unified Huntington's Disease Rating Scale. TFC: Total Functional Capacity.
72
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
Table 2 MR spectroscopy results (group 1)
NAA/Cr Ch/Cr Lac/Cr Lac/NAA
Parieto-occipital VOI
Cerebellar VOI
HD n = 13
Control n = 14
HD n = 13
Control n = 11
2.66 ± 0.59 0.86 ± 0.16 0.28 ± 0.16 a 0.11 ± 0.07b
2.86 ± 0.65 0.89 ± 0.16 0.17 ± 0.09 0.06 ± 0.03
1.74 ± 0.23 1.16 ± 0.22 0.13 ± 0.04 b 0.08 ± 0.03 c
1.72 ± 0.31 1.23 ± 0.18 0.09 ± 0.05 0.05 ± 0.02
NAA: N-acetyl aspartate; Cr: creatine; Ch: choline; Lac: lactate. a p b 0.05 vs controls. b p b 0.02 vs controls. c p b 0.005 vs controls.
controls had data acquired from both volumes). Lac/Cr was significantly increased in HD patients in both the parietooccipital ( p b 0.05) and cerebellar ( p b 0.01) volumes. Lac/ NAA was also significantly increased in HD patients in both volumes (parieto-occipital p b 0.02; cerebellum p b 0.005). In contrast, neither NAA/Cr nor Ch/Cr differed significantly from controls. Cerebellar Lac/Cr was strongly correlated with measures of clinical disease severity as indicated by disease duration (r = 0.69), UHDRS motor score (r = 0.73), and TFC score (r = −0.66), whereas parieto-occipital Lac/Cr showed no significant correlation with any of these clinical measures. No other significant correlations were evident. These MR results are summarized in Table 2. In the second group, tissue segmentation yielded voxels containing a range of 6%–84% CSF in HD patients vs 9%– 76% in control subjects. There was a close correlation between Lac/NAA and the amount of CSF within these voxels in patients (r = 0.99) and controls (r = 0.97). By extrapolating group Lac/NAA data to a theoretical voxel containing 100% brain, we determined this ratio to be slightly greater in the HD group than the control group, suggesting possibly increased lactate content in this predicted voxel (Figs. 1 and 2). Calcu-
Fig. 1. Lac/NAA as a function of brain vs CSF content within the MRS volume of interest in HD and controls. Results are from Group 2 (see text). On the x-axis, 1/% brain = 0.01 corresponds to the value that would be obtained from a theoretical voxel containing 100% brain. Data from the lower left corner are enlarged in Fig. 2.
Fig. 2. Data from Fig. 1, limited to volumes of interest with low CSF content (1/% brain b0.016 corresponds to CSF content b37.5% in the volume of interest). In addition to the regression line, 95% confidence limits for each group are shown.
lation of 95% confidence intervals, however, showed significant overlap between groups, indicating that this trend to increased lactate was not statistically significant. 4. Discussion These results, obtained with high field (3 T) MRS, suggest that brain lactate content is increased in manifest HD in voxels placed to minimize CSF content in parieto-occipital and cerebellar regions. This is consistent with previous reports of increased occipital lactate in this disorder measured at 1.5 T [9]. The presence of elevated lactate in these two functionally separate regions of the brain, neither of which are thought to harbor major neuropathological changes in HD, suggests that the increase may be regionally nonspecific. This is consistent with the observation that the gene mutation itself is widely expressed without significant selectivity for brain regions targeted by the disease process [16,17]. Although it would be intuitively obvious to seek changes in brain lactate in the basal ganglia structures that bear the brunt of HD pathology, reproducible spectra from basal ganglia voxels are not easily acquired because their age-related high iron content leads to susceptibility changes and a marked decrease in acquired signal [18]. Unfortunately for the present study, these iron-induced changes are more prominent with data acquired on high field strength MR systems. Because our initial objective was to replicate and expand on the studies of Jenkins et al. [9] we included a parietooccipital voxel in our studies. Although we attempted to minimize the potential contribution from CSF lactate by careful voxel placement, the relatively large voxel size utilized to maximize signal to noise ratio likely included a contribution from sulcal (and perhaps ventricular) CSF that varied amongst subjects. Indeed, Hoang et al. have suggested that the changes in HD may be due to increased lactate in CSF
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
rather than in brain itself [10]. We therefore studied a second group of subjects to address the possibility of a potential CSF contribution to the elevated lactate. By varying CSF content in the parieto-occipital VOI in different subjects, we were able to correlate the contribution of CSF to the Lac/NAA ratio. There was a very close relationship between CSF content and Lac/NAA within the parieto-occipital voxel, not surprisingly since NAA is a neuronal marker, expected to be absent in CSF. As shown in Figs. 1 and 2, by extrapolating to a theoretical voxel approaching 100% brain (no CSF), we demonstrated a trend toward increased lactate in HD vs controls. The overlapping confidence intervals, however, suggest that the increase is not as great as implied by the simple ratio measurements utilized in our first group of subjects, and that the apparent increase in lactate may be present in both CSF and brain tissue. Conflicting data have been reported previously from direct measurements of the CSF concentration of lactate in HD. Concentration has been suggested to be either reduced in HD brain compared to controls [19], or to not differ significantly [20]. While lactate concentration in normal resting brain has been estimated to be ∼0.5mM [21], the intracellular concentration in HD has not yet been reported. To the best of our knowledge, there have been no reports of increased CSF lactate in HD based on direct measurements. This supports our contention that much of the increased lactate that we have observed derives from brain rather than CSF lactate. In contrast, in vivo evidence suggests decreased NAA, at least in frontal regions [22] and striatum in HD brain [9]. Neither our data nor data of Jenkins et al. [9] suggest reduced NAA/Cr ratios in parieto-occipital regions. The correlation between the degree of lactate increase and clinical estimates of disease severity, based on the duration of manifest disease and on the severity of symptoms as measured with clinical rating scales, implies a relationship between disease progression and the biochemical processes involved in lactate formation. Altered energy metabolism has been reported to be a feature of HD for many years, recently reviewed by Browne and Beal [23]. Several lines of evidence, both in vivo and in vitro, point to the presence of abnormalities in various aspects of oxidative phosphorylation and the tricarboxylic acid cycle in brain regions affected by HD. Increased lactate production, as demonstrated in the present study and by others [9,10], whether present in brain tissue, CSF, or both, is consistent with these observations. The lactate itself is not thought to be a neurotoxic metabolite, but may represent a marker for energetic changes such as reduced ATP production and excitotoxicity which may have a direct effect on neuronal function and survival in HD [23]. These observations provide in vivo support for the presence of bioenergetic abnormalities associated with manifest HD. Additional studies are required to further elucidate the extent and significance of these changes. Direct quantitation of lactate concentration with correction for T2 changes should provide information that is more robust than that
73
provided by ratios such as we have used. In addition, future studies should consider correlating the MRS changes with direct measurement of CSF lactate concentrations. Lastly, implementation of spectroscopic imaging in similar studies in the future may provide additional valuable insights regarding the regional specificity of the changes in lactate content. Acknowledgements This study was supported by the Huntington's Disease Society of America and the Canadian Institutes of Health Research. References [1] Vonsattel J-P, Myers RH, Stevens TJ, Ferrante RJ, Bird ED, Richardson EP. Neuropathological classification of Huntington's disease. J Neuropathol Exp Neurol 1985;44:559–77. [2] Huntington's Disease Collaborative Research Group. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 1993;72:971–83. [3] Brennan WA, Bird ED, Aprille JR. Regional mitochondrial respiratory activity in Huntington's disease brain. J Neurochem 1985;44:1948–50. [4] Mann VM, Cooper JM, Javoy-Agid F, Agid Y, Jenner P, Schapira AH. Mitochondrial function and parental sex effect in Huntington's disease. Lancet 1990;336:749. [5] Parker WD, Boyson SJ, Luder AS, Parks JK. Evidence for a defect in NADH: ubiquinone oxidoreductase (complex I) in Huntington's disease. Neurology 1990;40:1231–4. [6] Ludolph AC, Seelig MO, Ludolph A, et al. 3-nitropropionic acid decreases cellular energy levels and causes neuronal degeneration in cortical explants. Neurodegeneration 1992;1:21–8. [7] Beal MF, Brouillet E, Jenkins B, Henshaw R, Rosen B, Hyman BT. Age-dependent striatal excitotoxic lesions produced by the endogenous mitochondrial inhibitor malonate. J Neurochem 1993;61:1147–50. [8] Beal MF, Swartz KJ, Hyman BT, et al. Aminooxyacetic acid results in excitotoxin lesions by a novel indirect mechanism. J Neurochem 1991;57:1068–73. [9] Jenkins BG, Koroshetz WJ, Beal MF, Rosen BR. Evidence for impairment of energy metabolism in vivo in Huntington's disease using localized 1H NMR spectroscopy. Neurology 1993;43:2689–95. [10] Hoang TQ, Bluml S, Dubowitz DJ, et al. Quantitative protondecoupled 31P MRS and 1H MRS in the evaluation of Huntington's and Parkinson's diseases. Neurology 1998;50:1033–40. [11] Huntington Study Group. The Unified Huntington's Disease Rating Scale: reliability and consistency. Mov Disord 1996;11:136–42. [12] Shoulson I, Fahn S. Huntington's disease: clinical care and evaluation. Neurology 1979;29:1–3. [13] Gruetter R. Automatic, localized in vivo adjustment of all first- and second-order shim coils. Magn Reson Med 1993;29:804–11. [14] Hoult DI, Richards RE. Critical factors in design of sensitive highresolution nuclear magnetic-resonance spectrometers. Proc R Soc Lond A 1975;344:311–40. [15] Hanstock CC, Allen PS. Segmentation of brain from a PRESS localised single volume using double inversion recovery for simultaneous T1 nulling. Proceedings of the Society of Magnetic Resonance in Medicine; 2000. p. 1964. [16] Sharp NH, Love SJ, Schilling G, et al. Widespread expression of the Huntington's disease gene (IT-15) protein product. Neuron 1995;14: 1065–74. [17] Sapp E, Schwarz C, Chae K, et al. Huntington localization in brains of normal and Huntington's disease patients. Ann Neurol 1997;42: 604–12.
74
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
[18] Martin WRW, Ye FQ, Allen PS. Increasing striatal iron content associated with normal aging. Mov Disord 1998;13:281–6. [19] Garseth M, Sonnewald U, White LR, et al. Proton magnetic resonance spectroscopy of cerebrospinal fluid in neurodegenerative disease: indication of glial energy impairment in Huntington chorea, but not Parkinson disease. J Neurosci Res 2000;60:779–82. [20] Nicoli F, Vion-Dury J, Maloteaux JM, et al. CSF and serum metabolic profile of patients with Huntington's chorea: A study by high resolution proton NMR spectroscopy and HPLC. Neurosci Lett 1993;154: 47–51.
[21] Hanstock CC, Rothman DL, Prichard JW, Jue T, Shulman RG. Spatially localised 1H NMR of metabolites in the human brain. Proc Natl Acad Sci U S A 1988;85:1821–5. [22] Harms L, Meierkord H, Timm G, Pfeiffer L, Ludolph AC. Decreased N-acetyl-aspartate/choline ratio and increased lactate in the frontal lobe of patients with Huntington's disease: a proton magnetic resonance spectroscopy study. J Neurol Neurosurg Psychiatry 1997;62:27–30. [23] Browne SE, Beal MF. The energetics of Huntington's disease. Neurochem Res 2004;29:531–46.
Is brain lactate increased in Huntington's disease? W.R. Wayne Martin a,⁎, Marguerite Wieler a , Christopher C. Hanstock b a
b
Division of Neurology, University of Alberta, Edmonton, Alberta, Canada Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada Received 22 April 2007; accepted 30 May 2007 Available online 25 July 2007
Abstract Impaired brain energy metabolism with increased regional brain lactate may play a role in the pathogenesis of Huntington’s disease (HD). Magnetic resonance spectroscopy (MRS) has provided conflicting evidence, however, regarding metabolic changes. Our objective was to evaluate the potential contribution of CSF lactate to the changes observed with MRS in HD. We performed single voxel MRS at 3 T in 23 patients with HD and 28 age-matched control subjects using a method to segment voxels into grey matter, white matter, and CSF, and to extrapolate regional lactate content to a hypothetical voxel containing 100% brain in order to control for differences in CSF lactate. Lactate/ creatine and lactate/N-acetyl aspartate (Lac/NAA) ratios were significantly increased in parieto-occipital ( p b 0.05) and cerebellar ( p b 0.01) voxels in HD patients. After extrapolating group Lac/NAA results to a theoretical voxel containing 100% brain, this ratio was greater in the HD group than the control group, suggesting possibly increased lactate in this predicted voxel, although the difference between groups did not reach statistical significance. These results suggest an increase in brain lactate content in manifest HD, in a regionally non-specific fashion, although the possibility of a CSF contribution to this increase cannot be ruled out. Regardless, this supports the possibility of impaired mitochondrial function resulting in abnormal brain energy metabolism in HD. © 2007 Elsevier B.V. All rights reserved. Keywords: Huntington’s disease; Magnetic resonance spectroscopy; Energy metabolism; Lactate; Mitochondrial function
1. Introduction Huntington's disease (HD) is a progressive, inherited neurodegenerative disorder characterized by abnormal involuntary movements, cognitive impairment, and behavioral disturbances. Typical neuropathological changes include marked neuronal loss from the caudate and putamen, particularly affecting medium-sized spiny neurons. Although most descriptions of pathology have emphasized the striatal changes, there is evidence of more widespread neuronal loss, particularly in advanced disease [1]. The genetic abnormality underlying the development of manifest disease has been identified [2], but the pathophysiological mechanisms that link this abnormality to neuronal loss remain uncertain. ⁎ Corresponding author. Movement Disorders Clinic, Glenrose Rehabilitation Hospital, 10230 - 111 Ave., Edmonton, AB, Canada T5G 0B7. Tel.: +1 780 735 8805; fax: +1 780 735 8804. E-mail address: [email protected] (W.R.W. Martin). 0022-510X/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2007.05.035
Impaired energy metabolism may play a role in the pathogenesis of HD. Measurements of mitochondrial electron transport enzymes have shown abnormalities in platelets and postmortem brain tissue [3–5]. Animal models have shown that inhibition of oxidative metabolism can produce a pattern of striatal neuronal loss that mimics that seen in HD [6–8]. Using magnetic resonance spectroscopy (MRS), Jenkins and co-workers reported an increase in lactate content in occipital cortex in HD, providing further support for a defect in brain energy metabolism [9]. These observations have been challenged, however, by the suggestion that the reported increase in tissue lactate may reflect contamination from ventricular CSF [10]. The objective of the present study was to replicate and expand on the studies of Jenkins et al. [9] evaluating the potential contribution of CSF lactate to the changes observed with single voxel MRS in HD. We determined brain and CSF lactate levels in two groups of patients and age-matched controls. In the first group, metabolite concentration in a
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
parieto-occipital voxel was evaluated, using methods similar to those published previously [9]. To evaluate the potential contribution of CSF lactate content in more detail, we subsequently studied a second group, using a method to segment voxels into grey matter, white matter, and CSF, and to extrapolate regional lactate content to a hypothetical voxel containing 100% brain (no CSF). 2. Methods
71
subjects, a similar parieto-occipital VOI was placed, but without restricting contribution from adjacent CSF. The same PRESS sequence was used for segmentation, to allow measurement of the relative percentage of grey matter (GM), white matter (WM) and CSF from within the selected voxel. For this, the PRESS sequence used a double inversion recovery 1D-projection method to allow selection of each compartment based on differences in T1 relaxation [15]. This allows simultaneous nulling of the signal from 2 compartments, leaving only that from the third.
2.1. Subjects 2.3. Data analysis Ambulatory, functionally independent patients with manifest HD were recruited from the Movement Disorders Clinic at the Glenrose Rehabilitation Hospital. All patients had a positive family history of HD, chorea as their predominant disease manifestation, and DNA analysis demonstrating more than 39 CAG repeats. Clinical disease severity was rated with the motor score from the Unified Huntington's Disease Rating Scale (UHDRS) [11] and the Total Functional Capacity (TFC) score [12]. Control subjects were drawn from a population that was not at risk for HD, and were healthy and free of neurological or psychiatric disease. The study was approved by the Human Research Ethics Board of the University of Alberta and all subjects gave informed consent.
Following filtering with a 2 Hz exponential multiplication, spectroscopic data were Fourier transformed, and phase and baseline corrected. Lactate (Lac), N-acetyl aspartate (NAA), creatine (Cr) and choline (Cho) peaks were fitted to Laurentz–Gaussian lineshapes, minimizing the residual. Results were expressed as the ratio between metabolites. In the second group of patients, the segmentation data were utilized to allow extrapolation to a theoretical voxel containing 100% brain (by summing the contribution from GM and WM), thereby eliminating the potential contribution from lactate-containing CSF. All data manipulation was performed in the MATLAB (The MathWorks, Inc., Natick, MA) program environment.
2.2. Data acquisition
3. Results
MRS data were acquired using a Magnex 3 tesla magnet (Magnex Scientific Ltd., Abingdon, U.K.) with actively shielded gradients, controlled with a Surrey Medical Imaging Systems spectrometer console (Surrey Medical Imaging Systems, Ltd., Guildford, U.K). Signal transmission and reception were achieved using a quadrature birdcage resonator. In all subjects, multi-slice gradient echo images were initially acquired in transverse, sagittal and coronal planes (slice thickness 5 mm, 1 mm gap, echo time (TE) 20 ms, repetition time (TR) 500 ms). In the first group, a volume of interest (VOI) measuring 2 × 2 × 3 cm was then registered to the medial parieto-occipital lobe of the left hemisphere, avoiding CSF as much as possible. A second 2 × 2 × 3 cm VOI was registered to the cerebellum. Shimming of the selected volumes was accomplished with FASTMAP [13]; linear shims were optimized with an “in-house” automatic routine, resulting in a shimmed water linewidth typically b 4 Hz (∼ 0.03 ppm). Water suppressed point-resolved spectroscopy (PRESS) was performed from the selected volumes with TE 272 ms, TR 3 s, and 512 averages (total acquisition time = 27 min), each 4 averages phase-cycled through the CYCLOPS scheme [14]. A second group of subjects was subsequently studied with a tissue segmentation routine in addition to the MRS sequence in order to evaluate the potential contribution of the CSF compartment to brain lactate estimates. In these
Twenty-three patients with HD and 28 healthy controls were studied. Thirteen patients and 17 controls comprised the first group; 10 patients and 11 controls comprised the second. Data were acquired from two differently placed voxels in one patient from the second group, yielding a total of 11 tissue volumes for extrapolation to a theoretical voxel containing 100% brain. There was no statistically significant difference in age or sex distribution. Clinical data are summarized in Table 1. In the first group, suitable spectra were available for analysis from both VOI's in all 13 patients. These data were compared to spectra from parieto-occipital VOI's in 14 controls, and from cerebellar VOI's in 11 controls (since not all Table 1 Clinical characteristics of patients and healthy control subjects Group 1
N % male Age mean ± SD (yr) Symptom duration mean ± SD (yr) UHDRS motor score mean ± SD TFC mean ± SD
Group 2
HD
Control
HD
Control
13 46 52.3 ± 9.7 7.4 ± 5.6
17 41 45.8 ± 11.3 –
10 50 50.4 ± 7.6 6.0 ± 4.7
11 45 46.0 ± 9.4 –
32.3 ± 12.3
–
32.5 ± 12.8
–
10.4 ± 2.8
–
10.4 ± 1.8
–
UHDRS: Unified Huntington's Disease Rating Scale. TFC: Total Functional Capacity.
72
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
Table 2 MR spectroscopy results (group 1)
NAA/Cr Ch/Cr Lac/Cr Lac/NAA
Parieto-occipital VOI
Cerebellar VOI
HD n = 13
Control n = 14
HD n = 13
Control n = 11
2.66 ± 0.59 0.86 ± 0.16 0.28 ± 0.16 a 0.11 ± 0.07b
2.86 ± 0.65 0.89 ± 0.16 0.17 ± 0.09 0.06 ± 0.03
1.74 ± 0.23 1.16 ± 0.22 0.13 ± 0.04 b 0.08 ± 0.03 c
1.72 ± 0.31 1.23 ± 0.18 0.09 ± 0.05 0.05 ± 0.02
NAA: N-acetyl aspartate; Cr: creatine; Ch: choline; Lac: lactate. a p b 0.05 vs controls. b p b 0.02 vs controls. c p b 0.005 vs controls.
controls had data acquired from both volumes). Lac/Cr was significantly increased in HD patients in both the parietooccipital ( p b 0.05) and cerebellar ( p b 0.01) volumes. Lac/ NAA was also significantly increased in HD patients in both volumes (parieto-occipital p b 0.02; cerebellum p b 0.005). In contrast, neither NAA/Cr nor Ch/Cr differed significantly from controls. Cerebellar Lac/Cr was strongly correlated with measures of clinical disease severity as indicated by disease duration (r = 0.69), UHDRS motor score (r = 0.73), and TFC score (r = −0.66), whereas parieto-occipital Lac/Cr showed no significant correlation with any of these clinical measures. No other significant correlations were evident. These MR results are summarized in Table 2. In the second group, tissue segmentation yielded voxels containing a range of 6%–84% CSF in HD patients vs 9%– 76% in control subjects. There was a close correlation between Lac/NAA and the amount of CSF within these voxels in patients (r = 0.99) and controls (r = 0.97). By extrapolating group Lac/NAA data to a theoretical voxel containing 100% brain, we determined this ratio to be slightly greater in the HD group than the control group, suggesting possibly increased lactate content in this predicted voxel (Figs. 1 and 2). Calcu-
Fig. 1. Lac/NAA as a function of brain vs CSF content within the MRS volume of interest in HD and controls. Results are from Group 2 (see text). On the x-axis, 1/% brain = 0.01 corresponds to the value that would be obtained from a theoretical voxel containing 100% brain. Data from the lower left corner are enlarged in Fig. 2.
Fig. 2. Data from Fig. 1, limited to volumes of interest with low CSF content (1/% brain b0.016 corresponds to CSF content b37.5% in the volume of interest). In addition to the regression line, 95% confidence limits for each group are shown.
lation of 95% confidence intervals, however, showed significant overlap between groups, indicating that this trend to increased lactate was not statistically significant. 4. Discussion These results, obtained with high field (3 T) MRS, suggest that brain lactate content is increased in manifest HD in voxels placed to minimize CSF content in parieto-occipital and cerebellar regions. This is consistent with previous reports of increased occipital lactate in this disorder measured at 1.5 T [9]. The presence of elevated lactate in these two functionally separate regions of the brain, neither of which are thought to harbor major neuropathological changes in HD, suggests that the increase may be regionally nonspecific. This is consistent with the observation that the gene mutation itself is widely expressed without significant selectivity for brain regions targeted by the disease process [16,17]. Although it would be intuitively obvious to seek changes in brain lactate in the basal ganglia structures that bear the brunt of HD pathology, reproducible spectra from basal ganglia voxels are not easily acquired because their age-related high iron content leads to susceptibility changes and a marked decrease in acquired signal [18]. Unfortunately for the present study, these iron-induced changes are more prominent with data acquired on high field strength MR systems. Because our initial objective was to replicate and expand on the studies of Jenkins et al. [9] we included a parietooccipital voxel in our studies. Although we attempted to minimize the potential contribution from CSF lactate by careful voxel placement, the relatively large voxel size utilized to maximize signal to noise ratio likely included a contribution from sulcal (and perhaps ventricular) CSF that varied amongst subjects. Indeed, Hoang et al. have suggested that the changes in HD may be due to increased lactate in CSF
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
rather than in brain itself [10]. We therefore studied a second group of subjects to address the possibility of a potential CSF contribution to the elevated lactate. By varying CSF content in the parieto-occipital VOI in different subjects, we were able to correlate the contribution of CSF to the Lac/NAA ratio. There was a very close relationship between CSF content and Lac/NAA within the parieto-occipital voxel, not surprisingly since NAA is a neuronal marker, expected to be absent in CSF. As shown in Figs. 1 and 2, by extrapolating to a theoretical voxel approaching 100% brain (no CSF), we demonstrated a trend toward increased lactate in HD vs controls. The overlapping confidence intervals, however, suggest that the increase is not as great as implied by the simple ratio measurements utilized in our first group of subjects, and that the apparent increase in lactate may be present in both CSF and brain tissue. Conflicting data have been reported previously from direct measurements of the CSF concentration of lactate in HD. Concentration has been suggested to be either reduced in HD brain compared to controls [19], or to not differ significantly [20]. While lactate concentration in normal resting brain has been estimated to be ∼0.5mM [21], the intracellular concentration in HD has not yet been reported. To the best of our knowledge, there have been no reports of increased CSF lactate in HD based on direct measurements. This supports our contention that much of the increased lactate that we have observed derives from brain rather than CSF lactate. In contrast, in vivo evidence suggests decreased NAA, at least in frontal regions [22] and striatum in HD brain [9]. Neither our data nor data of Jenkins et al. [9] suggest reduced NAA/Cr ratios in parieto-occipital regions. The correlation between the degree of lactate increase and clinical estimates of disease severity, based on the duration of manifest disease and on the severity of symptoms as measured with clinical rating scales, implies a relationship between disease progression and the biochemical processes involved in lactate formation. Altered energy metabolism has been reported to be a feature of HD for many years, recently reviewed by Browne and Beal [23]. Several lines of evidence, both in vivo and in vitro, point to the presence of abnormalities in various aspects of oxidative phosphorylation and the tricarboxylic acid cycle in brain regions affected by HD. Increased lactate production, as demonstrated in the present study and by others [9,10], whether present in brain tissue, CSF, or both, is consistent with these observations. The lactate itself is not thought to be a neurotoxic metabolite, but may represent a marker for energetic changes such as reduced ATP production and excitotoxicity which may have a direct effect on neuronal function and survival in HD [23]. These observations provide in vivo support for the presence of bioenergetic abnormalities associated with manifest HD. Additional studies are required to further elucidate the extent and significance of these changes. Direct quantitation of lactate concentration with correction for T2 changes should provide information that is more robust than that
73
provided by ratios such as we have used. In addition, future studies should consider correlating the MRS changes with direct measurement of CSF lactate concentrations. Lastly, implementation of spectroscopic imaging in similar studies in the future may provide additional valuable insights regarding the regional specificity of the changes in lactate content. Acknowledgements This study was supported by the Huntington's Disease Society of America and the Canadian Institutes of Health Research. References [1] Vonsattel J-P, Myers RH, Stevens TJ, Ferrante RJ, Bird ED, Richardson EP. Neuropathological classification of Huntington's disease. J Neuropathol Exp Neurol 1985;44:559–77. [2] Huntington's Disease Collaborative Research Group. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 1993;72:971–83. [3] Brennan WA, Bird ED, Aprille JR. Regional mitochondrial respiratory activity in Huntington's disease brain. J Neurochem 1985;44:1948–50. [4] Mann VM, Cooper JM, Javoy-Agid F, Agid Y, Jenner P, Schapira AH. Mitochondrial function and parental sex effect in Huntington's disease. Lancet 1990;336:749. [5] Parker WD, Boyson SJ, Luder AS, Parks JK. Evidence for a defect in NADH: ubiquinone oxidoreductase (complex I) in Huntington's disease. Neurology 1990;40:1231–4. [6] Ludolph AC, Seelig MO, Ludolph A, et al. 3-nitropropionic acid decreases cellular energy levels and causes neuronal degeneration in cortical explants. Neurodegeneration 1992;1:21–8. [7] Beal MF, Brouillet E, Jenkins B, Henshaw R, Rosen B, Hyman BT. Age-dependent striatal excitotoxic lesions produced by the endogenous mitochondrial inhibitor malonate. J Neurochem 1993;61:1147–50. [8] Beal MF, Swartz KJ, Hyman BT, et al. Aminooxyacetic acid results in excitotoxin lesions by a novel indirect mechanism. J Neurochem 1991;57:1068–73. [9] Jenkins BG, Koroshetz WJ, Beal MF, Rosen BR. Evidence for impairment of energy metabolism in vivo in Huntington's disease using localized 1H NMR spectroscopy. Neurology 1993;43:2689–95. [10] Hoang TQ, Bluml S, Dubowitz DJ, et al. Quantitative protondecoupled 31P MRS and 1H MRS in the evaluation of Huntington's and Parkinson's diseases. Neurology 1998;50:1033–40. [11] Huntington Study Group. The Unified Huntington's Disease Rating Scale: reliability and consistency. Mov Disord 1996;11:136–42. [12] Shoulson I, Fahn S. Huntington's disease: clinical care and evaluation. Neurology 1979;29:1–3. [13] Gruetter R. Automatic, localized in vivo adjustment of all first- and second-order shim coils. Magn Reson Med 1993;29:804–11. [14] Hoult DI, Richards RE. Critical factors in design of sensitive highresolution nuclear magnetic-resonance spectrometers. Proc R Soc Lond A 1975;344:311–40. [15] Hanstock CC, Allen PS. Segmentation of brain from a PRESS localised single volume using double inversion recovery for simultaneous T1 nulling. Proceedings of the Society of Magnetic Resonance in Medicine; 2000. p. 1964. [16] Sharp NH, Love SJ, Schilling G, et al. Widespread expression of the Huntington's disease gene (IT-15) protein product. Neuron 1995;14: 1065–74. [17] Sapp E, Schwarz C, Chae K, et al. Huntington localization in brains of normal and Huntington's disease patients. Ann Neurol 1997;42: 604–12.
74
W.R.W. Martin et al. / Journal of the Neurological Sciences 263 (2007) 70–74
[18] Martin WRW, Ye FQ, Allen PS. Increasing striatal iron content associated with normal aging. Mov Disord 1998;13:281–6. [19] Garseth M, Sonnewald U, White LR, et al. Proton magnetic resonance spectroscopy of cerebrospinal fluid in neurodegenerative disease: indication of glial energy impairment in Huntington chorea, but not Parkinson disease. J Neurosci Res 2000;60:779–82. [20] Nicoli F, Vion-Dury J, Maloteaux JM, et al. CSF and serum metabolic profile of patients with Huntington's chorea: A study by high resolution proton NMR spectroscopy and HPLC. Neurosci Lett 1993;154: 47–51.
[21] Hanstock CC, Rothman DL, Prichard JW, Jue T, Shulman RG. Spatially localised 1H NMR of metabolites in the human brain. Proc Natl Acad Sci U S A 1988;85:1821–5. [22] Harms L, Meierkord H, Timm G, Pfeiffer L, Ludolph AC. Decreased N-acetyl-aspartate/choline ratio and increased lactate in the frontal lobe of patients with Huntington's disease: a proton magnetic resonance spectroscopy study. J Neurol Neurosurg Psychiatry 1997;62:27–30. [23] Browne SE, Beal MF. The energetics of Huntington's disease. Neurochem Res 2004;29:531–46.