Abstract
Background: Huntington disease (HD) is characterized by striatal atrophy that begins long before the onset of motor symptoms.
Objective: To determine when striatal atrophy begins, the extent and rate of atrophy before diagnosis of motor symptoms, and whether striatal atrophy can predict when symptom onset will occur.
Methods: Caudate and putamen volumes were measured on MRI scans of 19 preclinical subjects with the HD gene expansion who were very far (9 to 20 years) from estimated onset, and on serial scans from 17 preclinical subjects, six of whom were diagnosed with HD within 5 years after the initial scan.
Results: Striatal volumes were significantly smaller for the subjects who were very far from estimated onset than for age-matched control subjects. Statistical models fit to the longitudinal data suggest that rate of caudate atrophy becomes significant when subjects are approximately 11 years from estimated onset and rate of putamen atrophy becomes significant approximately 9 years prior to onset. In the six incident cases, caudate and putamen were approximately one-third to one-half of normal volume at diagnosis, and caudate volume alone was able to predict with 100% accuracy those subjects who would be diagnosed within 2 years of imaging.
Conclusions: Striatal atrophy begins many years prior to diagnosable HD, and assessment of atrophy on MRI may be very useful in both predicting HD onset and in tracking progression in future therapeutic trials in preclinical subjects.
Huntington disease (HD) is inherited in an autosomal dominant fashion through a CAG-repeat mutation on chromosome 4 in the huntingtin gene.1 Since the discovery of the HD gene mutation in 1993, offspring of parents with HD have been able to choose to be tested to determine whether they have the gene expansion and will, therefore, eventually manifest the symptoms of HD themselves. Clinical diagnosis of HD, which generally occurs in the patient’s 40s or 50s, is based on the unequivocal presence of otherwise unexplained extrapyramidal movement disorder (e.g., chorea, dystonia, bradykinesia, rigidity). Efforts have been made to determine the best way to predict when clinical onset of symptoms will begin. Our group devised a formula for estimating the age at clinical onset, based on the CAG repeat of the gene-positive individual and the parent’s age at onset.2 We sought to determine whether MRI measurements of the basal ganglia during the preclinical stage might assist in predicting onset of diagnosable symptoms.
Abnormalities in structure and function of the basal ganglia have been previously demonstrated in individuals with the HD gene expansion prior to the diagnosis of HD (hereon referred to as “preclinical”).3–16⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓ Structural MRI studies by our group have demonstrated that preclinical subjects have smaller striatal volumes, as measured on MRI scans, than individuals without the gene expansion,14 and that volumes of caudate and putamen correlate positively with estimated years to HD onset.15 We also sought to provide a closer approximation of the time that striatal atrophy begins relative to onset of manifest HD and to determine rate of change in basal ganglia volumes as individuals approach their estimated age at onset.
Better definition of the changes occurring around onset could greatly facilitate future treatment trials, as it would allow selection of subjects who are most likely to convert to the symptomatic stage of HD during the clinical trial (in turn allowing delay of onset to be used as an outcome measure). Understanding the rate of striatal atrophy around the time of onset will assist in determining the usefulness of striatal volume measurements as outcome measures in clinical trials with preclinical subjects, where measurement of motor symptoms would not be a practical outcome measure. Finally, understanding when onset of striatal atrophy begins may assist in determining when treatments, as they become available, should begin in the preclinical stage.
Methods.
Subjects.
For the cross-sectional study, caudate and putamen measurements were obtained from MRI scans of 19 preclinical subjects with the HD gene expansion and 19 control subjects who were matched for sex and age. There were 7 men and 12 women in each group. Mean age was 32.1 years (SD = 6.6) for the preclinical subjects and mean CAG repeat length was 42.8 (SD = 2.2). All of the preclinical subjects were offspring of patients with documented HD but were themselves free of diagnosable HD symptoms. This group was restricted to individuals who were more than 8 years from onset, as estimated by a formula described below. Four of these subjects were included in previously reported studies. Ten of the preclinical subjects were recruited through the Huntington’s Disease Presymptomatic Testing Program at The Johns Hopkins University School of Medicine and scanned at Johns Hopkins. Among the remaining nine subjects, five were recruited through the Neurogenetics Clinic at the University of Washington, three through the Huntington Disease Medical Clinic at the University of British Columbia, and one through the Inland Northwest Genetics Clinic in Spokane, WA. These nine subjects were all scanned at the University of Washington.
Control subjects for the cross-sectional study were 7 men and 12 women with a mean age of 32.5 (SD = 6.7). Of the 19 control subjects, 10 were recruited through the Johns Hopkins Huntington’s Disease Presymptomatic Testing Program, were offspring of a parent with HD, and had tested negative for the HD gene expansion (i.e., CAG repeat length < 35). The remaining nine, selected from a sample of normal control subjects to match the preclinical subjects on age and sex, were recruited through the Division of Psychiatric Neuroimaging at Johns Hopkins. No attempt was made to match subjects and controls on education, as there is a possibility that individuals who test positive for the HD gene mutation may make decisions regarding educational attainment that would be different from those of gene-negative individuals. Thus matching on education as a way to control for potential brain differences is probably not appropriate. All control subjects were scanned at Johns Hopkins. Procedures were approved by The Johns Hopkins University Joint Committee on Clinical Investigation, the University of British Columbia Clinical Ethics Research Board, and the University of Washington Institutional Review Board. Written informed consent was obtained from all subjects after procedures were fully explained.
Subjects for the longitudinal study included five subjects from the cross-sectional sample (the others had only one scan, either because they had recently enrolled or had dropped out of the study) and 12 additional preclinical subjects (10 men and 7 women total). All were offspring of a parent with HD and had tested positive for the HD gene expansion. Subjects were included in this study if they had two or more MRI scans. Mean number of scans per subject was 4.1 (range = 2 to 6) and mean interval between first and last scan was 4.2 years (SD = 1.8; range = 1 to 6 years). Mean age at first scan was 37.1 (SD = 6.7) years and mean CAG repeat length was 44.0 (SD = 2.5). Six subjects were judged to be symptomatic sometime after the initial scan. Mean age at onset for these incident cases was 37.7 (SD = 4.13). Mean interval between first scan and diagnosed onset of symptoms was 2.26 years (SD = 1.2; range = 1.0 to 5.2 years). All of these subjects were recruited through the Huntington’s Disease Presymptomatic Testing Program at The Johns Hopkins University School of Medicine and scanned at Johns Hopkins. Ten of these subjects were included in previously reported studies, but with only two data points per subject. A control group of 20 individuals matched on age (mean = 38. 3, SD = 8.8) and sex (12 men, 8 women) was included. These subjects were scanned only once. Most were offspring of a parent with HD, but some were control subjects who were not related to individuals with HD. All had neurologic examinations and were found to have no CNS disorder. All controls were scanned at Johns Hopkins. Procedures were approved by The Johns Hopkins University Joint Committee on Clinical Investigation. Written informed consent was obtained from all subjects after procedures were fully explained.
Estimating years to onset.
For each of the preclinical subjects, expected age at onset was estimated using a formula based on the length of the subject’s trinucleotide repeat and the parental age at onset. The prediction equation (age at onset = [−0.81 × repeat length] + [0.51 × parental onset age] + 54.87) was derived from stepwise multiple regression analysis of data from 50 affected parent-child pairs in the Johns Hopkins HD clinic.2 The equation yielded a multiple R of 0.74 (p = 0.001). The addition of other variables (sex of affected parent or parent’s repeat length) did not add significantly to the equation’s ability to predict age at onset. Estimated years to onset (YTO) for the preclinical subjects in this study was calculated by subtracting the subject’s age at the time of the initial scan from his or her estimated onset age. Only subjects with YTO > 8 were selected for the cross-sectional study. (Our previous research suggested that subjects who are 12 or more years from onset might have normal striatal volumes. However, because these studies included very few subjects who were 8 to 12 years to onset, we chose to include these subjects as well in order to determine whether atrophy could be detected during this stage of the preclinical period.) Mean YTO for these subjects was 13.7 years (SD = 3.8; range = 9.0 to 20.0), and mean YTO for the subjects in the longitudinal study was 5.4 (SD = 5.5) at the initial scan. Because of the lack of sufficient subjects with the HD gene who have been studied prospectively from the preclinical stage through symptom onset, it has not yet been possible to validate the equation for predicting age at onset. However, the mean estimated onset age for the six incident cases in this study was 38.7 (SD = 3.2), compared to a mean actual age at onset of 37.7 (SD = 4.13), suggesting that the formula for estimating onset was fairly accurate for this small sample.
Neurologic examination.
All preclinical subjects were evaluated with either the Unified Huntington’s Disease Rating Scale (UHDRS)17 or Quantified Neurologic Exam (QNE).18 The Motor Section of the Unified Huntington’s Disease Rating Scale assesses motor abnormalities, with scores ranging from 0 to 124. For the nine preclinical subjects in the cross-sectional study who were evaluated with the UHDRS, mean score on the Motor Section was 0.33 (SD = 0.71, range = 0 to 2). The QNE is a standardized test of motor symptoms, with scores ranging from 0 to 129. Mean score for the 10 preclinical subjects in the cross-sectional study who were evaluated with the QNE was 2.3 (SD = 1.4). Diagnosis of HD was based on observation of new onset motor abnormalities, including chorea or incoordination, by experienced evaluators. All cases were reviewed by consensus conference. Evidence of cognitive or affective change without motor abnormalities was not considered sufficient for a diagnosis.
MRI.
At both University of Washington and Johns Hopkins, MRI scans were obtained on a General Electric 1.5 Tesla Signa scanner, using a protocol identical to that described in previous studies by our group.15,16⇓ This protocol includes a 1.5-mm spoiled gradient recalled (SPGR) echo in steady state (repetition time = 35, echo time = 5, number of excitations = 1, flip angle = 45°, voxel size = 0.9735 mm in x and y direction) coronal series, which was used for all measurements reported in this study. Measurements were performed using Measure, custom graphics software developed by Patrick Barta at The Johns Hopkins University School of Medicine.19
As in previous studies by our group,16,20⇓ caudate and putamen measurements were performed on axial images reconstructed from the 1.5 mm SPGR coronal series, using tri-linear interpolation. Because these reconstructions could be resliced at any angle, it was possible to ensure that the scans from all subjects were aligned identically along the axial plane passing through the anterior and posterior commissures (AC-PC line) and perpendicular to the inter-hemispheric fissure. After the brain was aligned with the AC-PC line, the images were resliced in the axial plane at a thickness of 0.9375 mm. As in our past studies,16,20⇓ measurement of caudate and putamen began in the most inferior slice in which the two structures are clearly separated by the internal capsule (see figure E-1, available online at www.neurology.org, for an illustration of measurement methods). The caudate is bordered laterally by the internal capsule and medially by the lateral ventricle. Measurement progressed in the superior direction until the body of the caudate was no longer visible. The borders of the putamen are defined laterally by the external capsule. At more inferior levels, the medial borders of the putamen are defined by the globus pallidus; at more superior levels, the medial borders are defined by the internal capsule. Structure outlines were rechecked in the coronal plane to make certain that no tissue was omitted. For each structure, areas within each slice were automatically calculated, summed across slices, and multiplied by slice thickness, yielding structure volumes.
After obtaining inter-rater reliability (intraclass correlation = 0.98 for caudate and 0.99 for putamen on 10 scans), one rater, blinded to diagnostic group, completed all of the measurements for the cross-sectional study. The other rater completed all of the measures for the longitudinal study by measuring all scans from one subject at the same time in order to maximize equality of the measurement technique. She was blind to the order of the scans being measured (i.e., she did not know whether the scan was an initial or follow-up scan) and blind to any clinical data. This rater also measured the scans for the control subjects who were matched with the preclinical subjects on age at the initial scan, but she was not blind to diagnostic group, as control subjects had only one scan.
Statistics.
t-Tests were used to assess group differences in caudate and putamen volume. Pearson correlations were performed to assess the association between striatal volumes and estimated YTO. Longitudinal data were analyzed using a generalized estimating equation (GEE) approach21 and random effects models approach.22 GEE is a method of estimating regression model parameters when dealing with correlated data. Data collected longitudinally on the same subjects are repeated measures that are generally correlated over time. If this correlation is not taken into account, the standard errors of the parameter estimates will not be valid and hypothesis testing results will not be replicable. To test the hypothesis that the rate of decline in putamen and caudate changes at a certain number of years prior to onset, we used GEE to fit a statistical model allowing different slopes prior to and after specific YTO time points. The criterion of maximized likelihood function was used to determine if and where the slope changes. A discriminant function analysis was also performed to determine whether subjects who would convert within 2 years could be identified based on striatal volumes. All reported p values are for two-tailed tests.
Results.
For the cross-sectional study, significant group differences were observed for both caudate and putamen (table), with far-from-onset subjects having significantly smaller structure volumes than the control subjects. Figure 1 shows cross-sectional measures of caudate volume from the 19 far-from-onset subjects as well as from the initial scans of those subjects in the longitudinal study who were < 8 years from estimated onset at the time of their first scan. Results for the putamen were similar (see figure E-2, available online at www.neurology.org). If the sample is restricted to those with YTO > 12 years, in order to compare with a previous study,15 the preclinical group still showed volume reduction in comparison to the age-matched controls (t = 2.08, df = 29, p = 0.05 for caudate; t = 2.08, df = 29, p = 0.05 for putamen), with values being between those of control subjects and those of preclinical subjects with 8 < YTO < 12.
Table Mean (SD) caudate and putamen volumes for 19 preclinical subjects who were > 8 years from estimated onset and for 19 controls
Figure 1. Caudate volume and estimated years to onset (YTO) in preclinical subjects. Solid squares represent subjects with YTO > 8; open squares represent subjects with YTO < 8. Solid line represents mean volume for control subjects; dashed lines represent ± 1 SD for control volumes.
Figures 2 and 3⇓ summarize the findings for longitudinal progression of caudate and putamen atrophy as subjects near estimated, or in some cases, documented onset. The regression lines in figures 2 and 3⇓ were fitted to the data using GEE, and show that for caudate, the rate of decline is negligible 11 years prior to the onset of HD (0.001 cm3 per year), but becomes significant within 11 years of onset, at the rate of 0.24 cm3 per year. A similar pattern holds for putamen (see figure 3), for which the rate of decline is estimated at 0.035 cm3 per year 9 years prior to the diagnosis of HD and increases to 0.23 cm3 per year afterward. For both measures, the difference in rates of decline is significant at the 0.05 level. (The results from random effects modeling were basically the same as for GEE.)
Figure 2. Longitudinal change in caudate volume for preclinical subjects. (Each color represents a single subject. Stars indicate visits at which the subject was considered symptomatic; i.e., the six incident cases. Heavy dashed line shows the model best fitting the data, according to generalized estimating equation analysis. Solid line represents mean volume for control subjects; lighter dashed lines represent ± 1 SD for control volumes.)
Figure 3. Longitudinal change in putamen volume for preclinical subjects. (Each color represents a single subject. Stars indicate visits at which the subject was considered symptomatic; i.e., the six incident cases. Heavy dashed line shows the model best fitting the data, according to generalized estimating equation analysis. Solid line represents mean volume for control subjects; lighter dashed lines represent ± 1 SD for control volumes.)
To determine rate of change in caudate during the preclinical period, we restricted our analysis to data from scans that were within 11 years prior to onset. Similarly, for rate of change in putamen, we restricted analysis to scans within 9 years of onset. Volumes from scan visits at which subjects were symptomatic were also not included. Percent per year rates of change were calculated by subtracting volume at the final scan from volume at the first scan that met the restrictions described above, dividing this result by volume at the first scan that met the restrictions described above, and dividing again by interval between scans. Rates of structural change were 4.3% per year for caudate and 3.1% per year for putamen. These rates reflect the change that occurs after atrophy actively begins (after 9 YTO for putamen and after 11 YTO for caudate) but before symptoms are observed.
For the far-from-onset subjects, caudate and putamen volumes were not correlated with YTO (r = 0.41, p = 0.08 for caudate; r = 0.33, p = 0.17 for putamen). As would be expected, correlations were higher for the subjects in the longitudinal study, as these subjects were, on average, closer to onset than the far-from-onset subjects. Correlations between YTO and volume at initial scan were 0.76 (p < 0.001) for putamen and 0.69 (p = 0.002) for caudate.
For all of the six incident cases (those who were diagnosed with HD sometime after the initial scan), caudate volume was < 4.6 cm3 at the time of diagnosis. In contrast, all subjects with caudate volume > 5.3 cm3 were preclinical. For putamen, all subjects with putamen volume < 3.3 cm3 were symptomatic and all with putamen > 5.1 cm3 were preclinical. For the 20 control subjects matched to the preclinical subjects with longitudinal data, mean caudate volume was 9.5 cm3 (SD = 2.2) and mean putamen volume was 9.8 cm3 (SD = 2.1). For the six incident cases, caudate was approximately 52% of control volume (5.0 cm3, SD = 0.2; range = 48 to 55%) and putamen was approximately 43% of control volume (4.2 cm3, SD = 0.7; range = 31 to 51%) at the time of diagnosis. High correlations were found for the six incident cases between actual years to onset and caudate and putamen volumes at first scan, controlling for age at first scan, but the correlations were not significant due to the small sample size (r = −0.64, p = 0.24 for caudate; r = −0.72, p = 0.17 for putamen; r = −0.85, p = 0.07 for total basal ganglia). A discriminant function analysis significantly predicted with 100% accuracy the subjects who would (n = 6) and would not (n = 11) become diagnosable with HD within 2 years, based on caudate volume (F = 12.35, df = 1.15, p = 0.003). Addition of putamen volume, age, or CAG repeat length did not improve the predictive value of the equation. Figure 4 presents caudate volumes for those subjects who did and did not become symptomatic within 2 years of the scan date. In figure 4, the largest caudate volume for a subject who would show symptoms within 2 years was 5.71 cm3 and the smallest caudate volume for subjects who would not convert within 2 years was 5.73 cm3. There was, therefore, no overlap, although this cannot be observed clearly in the figure.
Figure 4. Caudate volumes of subjects who were diagnosed with Huntington disease (motor symptoms) within 2 years of the scan and those who were not.
Discussion.
Using longitudinal and cross-sectional studies, we sought to determine 1) when basal ganglia atrophy begins in the preclinical stage, 2) the extent and rate of atrophy prior to onset, and 3) whether basal ganglia volumes can predict the onset of diagnosable HD. Although results of the study cannot be considered definitive because of relatively small sample size, we conclude that caudate and putamen volumes are reduced in individuals with the HD CAG repeat expansion many (9 to 20) years prior to onset of diagnosable HD, but that rate of atrophy is not significant until around 10 years from onset of symptoms. After the rate of atrophy becomes significant (at approximately 11 years prior to onset for caudate and 9 years prior to onset for putamen), rate of decline is approximately 4.3% per year for caudate and 3.1% per year for putamen. In our sample of six incident cases, caudate volume could predict onset of symptoms within 2 years with 100% accuracy. These findings are important, as they may guide selection of samples for detecting treatment efficacy in future clinical trials.
In a previous study,15 we speculated that striatal volumes would be normal in preclinical individuals who were 12 or more years from estimated onset. This supposition was based on a very small sample of four individuals whose caudate volumes (head of caudate only) and putamen volumes were very similar to the mean volume for the 27 control subjects. If our current sample is restricted to those with YTO > 12 years, as in the previous study, the preclinical group still shows significant volume reduction in comparison to the age-matched controls.
Thus, we were unable to detect a period during which putamen and caudate volumes were of normal volume in this sample of subjects who were very far from onset, and the data allow no definitive conclusion regarding the question of when striatal volume reduction begins. There are several possibilities. First, it has been hypothesized that basal ganglia are normal at birth, but begin to decline linearly from that point forward.23 Our results clearly refute this possibility, as this hypothesis would predict striatal volumes much smaller than we observe in the far-from-onset group (or that subjects with the HD CAG repeat expansion have striata several times larger than normal at birth). A second possibility is that striatal volumes in expansion positive individuals are developmentally smaller than normal. This hypothesis could be consistent with the idea that a normal function of huntingtin is pro-neurotrophic, and that glutamine repeat expansion causes a partial loss of function.24 The hypothesis of abnormal basal ganglia development is also consistent with the observations of a neuropathologic study25 that found increased oligodendroglial density in three of six preclinical subjects and concluded that this may reflect a developmental effect of the HD gene.
Although our data cannot rule out the hypothesis of abnormal development of basal ganglia, we favor a third possibility, that striatal volumes are essentially normal at birth and begin to atrophy at a point earlier than we were able to capture in this sample. Results from the cross-sectional study strongly suggest that progression of striatal atrophy proceeds in a nonlinear fashion, with extremely slow atrophy (beginning before 10 to 12 years prior to onset) followed by significantly increased rate of atrophy around 10 years prior to onset. Regardless of when striatal atrophy and dysfunction begins, our results suggest that when an effective treatment is discovered for symptomatic HD patients, it should be administered to preclinical subjects long before symptoms begin. Our estimates of the onset of significant rate of striatal atrophy may be useful in determining when treatment should begin.
When the data from the cross-sectional study are combined with the data from the longitudinal study (initial scan for each subject only), strong correlations are observed between YTO and caudate and putamen volumes. Careful observation of figure 1 suggests, however, that the function is not completely linear (with very little decline observed in the subjects with the greatest YTO). Analyses of longitudinal data with GEE also support the conclusion that atrophy proceeds in a nonlinear fashion (i.e., little or no decline for subjects very far from onset and more rapid, but steady decline that begins around 10 years from estimated onset). Results suggest that caudate volume, although significantly reduced in subjects who are very far from estimated onset, shows non-significant rate of decline between 12 and 20 years of estimated onset; at around 11 years prior to estimated onset the rate increases significantly. Similarly, although putamen is significantly smaller than normal in preclinical subjects who are very far from estimated onset, the rate of decline is not significant until around 9 years prior to estimated onset, at which time the rate of atrophy increases to a significant level. Whether the decline follows two linear rates, with a sharp transition, or a more complex curve, cannot be determined for certain with the sample currently available.
This study is the first to present data on striatal volumes from a prospective study that includes incident cases of HD. These data demonstrate that volume of caudate and putamen may be very useful for two purposes. First, striatal volumes may be informative in predicting onset of symptoms in subjects who have the HD gene expansion. Data from the six incident cases suggest that caudate volume is approximately one-half of normal volume at the time of diagnosis and putamen volume is one-third to one-half of normal volume at the time of diagnosis. The range of volumes for the incident cases at the time of diagnosis is remarkably small, especially for caudate (with < 1 cm3 separating all incident cases from all non-incident cases). Thus, the volume of caudate may be an excellent predictor of which subjects will convert within the next few years. This information could be useful in selecting cases for clinical trials involving preclinical subjects, where the outcome measure is delay of onset, as it would be possible to select only cases whose onset was expected within the duration of the trial.
Second, about 10 years prior to onset, rate of striatal atrophy increases significantly. A precise determination of the rate of change during this period will require study of a larger number of cases, but based on the current data, atrophy rate appears fairly steady throughout the 10 years prior to HD onset. Thus striatal volume loss may be an excellent neurobiological marker for progression of the underlying disease process of HD, and a highly useful outcome measure for therapeutic trials—especially in the preclinical period. Based on the GEE analysis of the longitudinal data, it appears that treatment trials using caudate volume as an outcome measure would maximize the probability of detecting an effect if the sample were restricted to those who are within 11 years of onset. For the 14 subjects in the longitudinal study with YTO ≤ 11, the change per year in caudate volume was 0.26 cm3 (SD = 0.20). Power analysis indicates that if a treatment were effective in reducing the rate of atrophy by one-half, approximately 48 preclinical subjects per group would be needed to demonstrate a difference in caudate volume change between treatment and placebo groups, with alpha = 0.05 and 0.90 power (using two-tailed tests). These calculations assume that the treatment trial would include only those subjects whose estimated onset was less than 11 years from the initiation of the trial, that the trial would be of 1 year duration, and that rate of change would be equal to that of the preclinical subjects in the current study. Fewer patients would be needed if the trials were longer and more patients would be needed if the treatment effect were expected to be smaller.
One limitation of our longitudinal data is the lack of serial data for the control subjects. However, in this age range, little striatal atrophy would be expected. Correlations between age and striatal volumes were not significant for the control subjects in this sample (r = −0.16, p = 0.49 for caudate and r = −0.30, p = 0.20 for putamen). In a cross-sectional study of healthy control subjects between 20 and 80 years of age, estimated rate of caudate shrinkage was approximately 3.3% per decade.26 Average caudate atrophy in preclinical subjects was approximately 3.5% per year, clearly far greater than the normal rate.
Interpretation of the longitudinal results of the GEE analysis must be limited by our assumption that two slopes best describe the data. As our sample includes no subjects who are more than 20 years from estimated onset, it is impossible to determine whether our conclusions regarding onset of significant atrophy can be extended to even earlier time periods. Interpretation is also limited by the reliance on estimated onset age rather than actual onset age for most subjects. Because the subjects in this study are still far from estimated onset, it will probably be many years before we can validate our predictions regarding their onset ages. In the meantime, however, we are following a large cohort of preclinical subjects whose data will assist us in validating or refining our formula for estimating onset.
Interpretation of the longitudinal results might also be limited by the fact that the rater was not blinded according to diagnosis (but was blinded to years to onset within the preclinical sample). The rater was not blinded to diagnosis because the control subjects had only one scan and the preclinical subjects all had more than one scan. The striatal volumes measured for the longitudinal and cross-sectional control samples are, however, quite similar, lending assurance that the volumes of control subjects in the longitudinal study were accurate. When the ages of the samples were equated by omitting the oldest subjects in the longitudinal sample, the mean caudate volumes were 9.21 (SD = 2.04) for the cross-sectional controls and 9.26 (2.48) for the longitudinal controls. Mean putamen volumes were 9.78 (SD = 1.83) for the cross-sectional controls and 9.90 (SD = 2.38) for the longitudinal controls.
Our studies are limited by small samples because of the limited number of preclinical HD expansion positive subjects available, especially those who are very far from onset. The results suggest the value of a larger multi-site study, such as the PREDICT-HD (Neurobiological Predictors of Huntington’s Disease) study that is now under way. Despite the small sample, our data strongly suggest that atrophy begins long before onset (approximately 10 years), that therapy should begin at least by this point, and that striatal volumes may be an excellent predictor of HD onset and an excellent outcome measure in clinical trials in preclinical as well as affected individuals.
Acknowledgments
Supported by grants from the National Institute of Neurologic Disorders and Stroke (NS16375), the Johns Hopkins School of Medicine General Clinical Research Center (NIH/NCRR grant M01 RR00052), and the Huntington’s Disease Society of America.
The authors thank Lynn Raymond, Elisabeth Almqvist, and Joji Decolongon at the University of British Columbia Huntington Disease Medical Clinic, Cynthia Dolan at the Inland Northwest Genetics Clinic in Spokane, WA, and Robin Bennett, Hillary Lipe, and Thomas Bird at the Neurogenetics Clinic at the University of Washington for helping recruit subjects. They also thank Godfrey Pearlson, previous director of Psychiatric Neuroimaging at The Johns Hopkins University School of Medicine, for providing control scans.
Footnotes
-
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the July 13 issue to find the title link for this article.
- Received October 24, 2003.
- Accepted March 3, 2004.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Topics Discussed
Alert Me
Recommended articles
Longitudinal change in basal ganglia volume in patients with Huntington's disease
Regional cortical thinning in preclinical Huntington disease and its relationship to cognition
Change in PDE10 across early Huntington disease assessed by [18F]MNI-659 and PET imaging
Hippocampal atrophy rates in Alzheimer disease