Phenotype Characterization of HD Intermediate Alleles in PREDICT-HD

Abstract

Background: Huntington disease (HD) is a neurodegenerative disease caused by a CAG repeat expansion on chromosome 4. Pathology is associated with CAG repeat length. Prior studies examining people in the intermediate allele (IA) range found subtle differences in motor, cognitive, and behavioral domains compared to controls.

Objective: The purpose of this study was to examine baseline and longitudinal differences in motor, cognitive, behavioral, functional, and imaging outcomes between persons with CAG repeats in three ranges: normal (≤26), intermediate (27–35), and reduced penetrance (36–39).

Methods: We examined longitudinal data from 389 participants in three allele groups: 280 normal controls (NC), 21 intermediate allele [IA], and 88 reduced penetrance [RP]. We used linear mixed models to identify differences in baseline and longitudinal outcomes between groups. Three models were tested: 1) no baseline or longitudinal differences; 2) baseline differences but no longitudinal differences; and 3) baseline and longitudinal differences.

Results: Model 1 was the best fitting model for most outcome variables. Models 2 and 3 were best fitting for some of the variables. We found baseline and longitudinal trends of declining performance across increasing CAG repeat length groups, but no significant differences between the NC and IA groups.

Conclusion: We did not find evidence to support differences in the IA group compared to the NC group. These findings are limited by a small IA sample size.

Keywords

Huntington disease intermediate alleles

INTRODUCTION

Huntington disease (HD) is an inherited, autosomal dominant neurodegenerative disease caused by a CAG repeat expansion on chromosome 4 [1]. For people in the affected range of 36 or more CAG repeats, age of disease onset is related to length of CAG repeat, with longer CAG repeats associated with earlier age of onset [2]. HD symptoms include motor, cognitive, behavioral, and functional changes, with a formal diagnosis based on the presence of characteristic motor signs [3].

Current genetic testing guidelines define ranges for disease manifestation based on CAG repeat length: ≤26 = normal controls (NC); 27–35 = intermediate allele (IA); 36–39 = reduced penetrance (RP); ≥40 = full penetrance (FP) [4]. Persons in the RP range may not develop a formal diagnosis in their lifetimes [5]. Individuals in the IA range are highly unlikely to develop a formal diagnosis, although there are several notable case reports [6 –10].

More recently, evidence from large observational studies suggests that persons in the IA range display subtle abnormalities in motor, cognitive, and behavioral domains compared to controls. In an analysis of the Cooperative Huntington’s Disease Observational Research Trial (COHORT), 50 of the 1985 participants were in the IA range and demonstrated worse saccade velocity, dystonia, and performance on the Stroop Color and Word test compared to controls [11]. In an analysis of the Prospective Huntington At Risk Observational Study (PHAROS) [12], 50 of the 983 participants were in the IA range and had significantly worse apathy and suicidal ideation than controls. Authors of the PHAROS study suggest the IA range might represent prodromal HD or a behavioral subphenotype. The purpose of the current analysis was to examine baseline and longitudinal differences in motor, cognitive, behavioral, functional and imaging outcomes between persons in the RP range, the IA range and at-risk individuals who underwent predictive testing and found to have normal allele lengths (normal controls, NC). All participants were enrolled in the Neurobiological Predictors of Huntington’s Disease (PREDICT-HD) study.

MATERIALS AND METHODS

Participants and data

Participants included in this analysis came from the PREDICT-HD study, a prospective, international, 32-site study that followed persons who previously underwent testing for the HD gene expansion. Those who tested with their longest allele length ≥36 participated as gene-expanded cases and those with longest allele length ≤35 participated as control participants. All participants provided written informed consent and were treated in accordance with the ethical standards of each site’s institutional review board. Inclusion criteria required independent HD genetic testing prior to entering the study, and required all individuals be age 18 and above at the time of study entry. Exclusion criteria mandated that cases not have sufficient motor signs for a clinical HD diagnosis at study entry, no history of traumatic brain injury or other central nervous system injury or diseases, no pacemakers or metallic implants, no prescribed use of antipsychotic or phenothiazine-derivative antiemetic medication in the past six months, and no clinical evidence of unstable medical or psychiatric illness. This dataset is ideal for exploring disease progression in HD prior to motor diagnosis due to the large sample of premanifest individuals and longitudinal data. These data may be sensitive to subtle changes that potentially begin several years before motor diagnosis.

Measures

We selected a sample of cognitive, motor, behavioral, functional and imaging measures from the PREDICT-HD battery that have shown sensitivity to disease progression [13 –18]. Measures from the Unified Huntington’s Disease Rating Scale [3] include total motor score (TMS), Stroop Color and Word Test [19], and Symbol Digit Modalities Test (SDMT) [20]. Behavioral variables include the total and subscale scores from the Frontal Systems Behavioral Scale (FrSBe) [21] and the Symptom Check List 90- revised (SCL-90-R) [22]. Functional variables included the total scores from the Everyday Cognition (ECog) scale [23], and the World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) [24]. Both participant-rated and companion-rated versions of the WHODAS and ECog are included to account for the possibility of decreased reliability of self-reported functioning resulting from disease progression [14 –18]. We also included MRI measures for striatal volume processed using BRAINS image processing software [25].

Participant stratification and analysis aims

Progression groups were defined by CAG repeat length according to American College of Medical Genetics (ACMG) and American Society of Human Genetics (ASHG) [4] guidelines as follows: normal controls (NC)≤26, intermediate alleles (IA) 27–35, and reduced penetrance (RP) 36–39. The primary aim of this analysis was to examine differences between IA individuals and the NC group, with particular attention paid to cognitive and behavioral manifestations. We included the RP group because this group might or might not phenoconvert to manifest HD; we were interested to examine whether the IA group would look more similar to the RP group or to the NC group. We excluded full penetrance participants because the phenotype differences between premanifest HD and NC is well documented and we did not want our results to be influenced by the full penetrance phenotype. Based on previous studies that support a behavioral subphenotype for the IA range, our hypothesis was that IA individuals would demonstrate worse average performances compared to the NC group with respect to behavioral measures. In PREDICT-HD analyses, IA individuals are usually grouped with controls [26].

Statistical analyses

All analyses were performed using the statistical software program R (version 3.1.2). First, sample sizes, measures of centrality, and measures of variability were obtained for the demographic variables age (at baseline) and years of education. Analysis of variance (ANOVA) F-tests were used to determine whether an overall statistically significant difference in means existed between groups. Pearson’s chi-squared test was used to assess differences in sex proportion by group. Second, linear mixed models (LMMs) [27] were used for the longitudinal analysis. Each outcome of interest was analyzed separately, using the following predictors: time, group, and interaction between time and group. We included the covariates age (at baseline), years of education, and sex to control for these variables. Time was measured as duration of follow-up for all longitudinal analyses. The intercept corresponds to the outcome measures at baseline, and the slope corresponds to the annual rate of longitudinal change in the outcome.

Three models were assessed for each outcome: Model 1 = no baseline or longitudinal group differences; Model 2 = baseline group differences but no longitudinal group differences over time; and Model 3 = baseline group differences and longitudinal group differences over time. The Akaike information criterion (AIC) [28] was used to select the optimal model from among the three. The AIC is known for its ability to select a model that balances the two competing goals of model building: adequacy of the model fit to the observed data and model parsimony (simplicity). LMMs yield unbiased parameter estimates under the assumption that the missing data are ignorable [29]. After the optimal model was selected, t-tests were carried out to assess differences at baseline and over time between the IA and NC groups and the RP and NC groups. The False Discovery Rate (FDR) was used to adjust for multiple tests within comparison type, namely: IA vs NC baseline, RP vs NC baseline, IA vs NC longitudinal change, and RP vs NC longitudinal change [30].

Finally, due to the limited sample size (n = 21) in the IA group, post-hoc power analysis was conducted for outcomes where a gradient effect of symptoms was observed across the groups. Gradient effects are defined by evidence of increasing impairment or dysfunction from the NC group to the RP group. For instance, if considering a cognitive measure for which higher values are indicative of cognitive impairment, a gradient effect would be said to exist if the NC group had the lowest baseline mean followed by the IA group, and then the RP group. This post-hoc power analysis was conducted via simulation with 500 trials, wherein resampling was conducted via bootstrapping by group. Thus, 280 NC individuals were sampled with replacement from the NC group, 21 IA individuals were resampled with replacement, 88 RP individuals were sampled with replacement. The model selection process described above was then followed, conducting the test of interest when applicable (for instance, if model 2 is selected, then the test for a difference in longitudinal change will fail to reject). The power for each comparison (baseline and longitudinal change for relevant outcomes that exhibited gradient effects) is then the proportion of trials in the simulation where the test was rejected divided by the total number of trials in the simulation.

RESULTS

Demographics

Our dataset consisted of 389 participants in three ranges according to their longest CAG repeat allele: 280 were in the NC range, 21 were in the IA range, and 88 were in the RP range. Demographic data, including group, sex, years of education, and age are presented in Table 1. Statistical evidence at the p = 0.05 level concluded that there were differences in mean age at baseline (p = 0.0017). Although years of education was not found to differ statistically by group, education level can also play a role in cognitive performance. Therefore, age (at baseline) and years of education, along with sex, were controlled for in the LMM analyses. Data on years of education were not available for one IA participant. Consistent with the female/male ratio in both the COHORT and PHAROS studies, our sample was approximately two-thirds female. The PREDICT-HD sample consisted of individuals who had already been tested and thus our female/male ratio is representative of the population who underwent testing [31].

Longitudinal analysis via LMMs

Table 2 presents results from the LMMs (estimates and model fit) via the process described above. Model 1 (no baseline or longitudinal differences between groups) was the best fitting model for most of the outcome variables. Model 2 (baseline differences) was the best fitting model for WHODAS (companion), BDI, SCL90 Obsession, SCL90 Hostility, SCL90 Psychotism, ECog executive functioning (companion), and ECog language (companion), and Model 3 (baseline differences and longitudinal differences) was the best fitting model for SDMT, WHODAS (participant), and TMS. However, there were no statistically significant differences in baseline or longitudinal change differences between the IA and NC groups. All significant findings were due to differences between the NC and RP groups, indicated by bolded estimates in Table 2. Specically, baseline differences between RP and NC were found for WHODAS (companion, p = 0.001), striatal volume (p = 0.005), SCL90 hostility (p = 0.005), SCL90 psychoticism (p = 0.016), and ECog executive functioning (companion, p = 0.011), and ECog language (companion, p = 0.004). Also, longitudinal change differences between RP and NC were found for SDMT (p = 0.002), TMS (p = 0.0004), and striatal volume (p = 0.0004). All p-values were adjusted for multiple comparisons using the FDR, as noted in the methods section.

While there were no significant baseline or longitudinal differences between the IA and NC group, our data do indicate evidence of gradient effects on several measures, including behavioral measures. Evidence of baseline gradient effects were found for SDMT, TMS, Striatal volume, BDI, SCL90 hostility, and SCL90 psychotism. Evidence of longitudinal gradient effects were found for striatal volume, but not the other two outcomes for which Model 3 was selected. Post-hoc power analyses (described in the methods) revealed that the power for detecting differences between IA and NC groups for the above effects was extremely low. Specifically, power for detecting a difference in slopes between the IA and NC groups for striatal volume was 0.2%, while power for detecting a difference in baseline means between the IA and NC groups ranged from 1% for SCL90 psychoticism to 12.8% for SDMT. These rates all fall well below the usually accepted power rate of 80% for design of clinical studies.

Figures 1, 2, and 3 provide visual representation of the longitudinal changes in three outcomes for which Model 3 was selected: the SDMT, TMS, and striatal volume. In these visual representations, those in the IA group show patterns of change similar to those in the NC group, and the RP group shows more degeneration.

DISCUSSION

Findings from this study showed no significant differences between the IA and NC groups whereas multiple significant differences existed between the RP and NC groups. The COHORT [11] and PHAROS [12] studies examined baseline and longitudinal differences in controls, IA, and full penetrance groups but did not examine RP groups separately, as we did in our analyses. We did not include the full penetrance group in order to retain the focus of our comparisons between the NC and IA groups and because differences between the <35 group and the >35 CAG group in the PREDICT-HD study have been well documented previously. We did, however, include the RP group to examine evidence of cumulative pathology with expanded CAG length. While we found evidence of baseline and longitudinal differences between the three groups, we did not find evidence of differences between the IA group and the NC group. All of the differences were between the RP and NC groups, but the lack of differences between IA and NC may have been due to extremely low power. Given the large number of outcome measures examined, we expected to find some significant differences between IA and NC groups, even if just by chance. These negative findings are consistent with current genetic testing guidelines that indicate persons in the IA range are unaffected by the HD gene expansion [4]. However, it is possible that with a larger sample size of IA participants some of these differences would be significant. Examination of the figures and statistical findings show that trends were observed between NC and IA and illustrations suggest that symptom/sign severity for IA was in-between that observed in the NC and the RP groups.

Some researchers have postulated that environmental or genetic modifiers might cause some people within IA ranges to express a behavioral subphenotype, including increased depression, apathy, suicidal thoughts, suicide attempts, and history of psychiatric disease [7, 12]. We did not find evidence to support a behavioral subphenotype for the IA group in our sample, although we found baseline gradient effects for hostility, and psychoticism on the SCL-90. At the same time, significant baseline differences in the RP group compared to the NC group (shown in Table 2) included two behavioral measures, depression (BDI) and hostility (SCL-90). The RP group also showed significant changes over time in longitudinal SDMT, companion-rated WHODAS, and striatal volume compared to the NC group. This suggests that even those in the RP range who do not display motor signs required for a definitive diagnosis of symptomatic HD exhibit some of the characteristics associated with manifest HD. Gradient effects are suggestive of a toxic gain-of-function pathology in HD (i.e. pathology increases with increased CAG repeat length even if it does not meet diagnostic criteria for manifest HD). This is consistent with the findings for the RP group described above.

Longitudinal gradient effects were also present for frontal behaviors affecting executive function and total frontal behaviors score. However, the only longitudinal gradient effect for a behavioral measure that was observed was for SCL-90 obsessive compulsive scale. We found increased baseline depression and hostility behavioral symptoms in the RP group compared to the IA and NC groups, which suggests that increased CAG length might be associated with behavioral changes even though they are not apparent in our IA sample. However, we did not find longitudinal differences between RP and NC groups on any behavioral measures.

Our data also showed longitudinal gradient effects for some cognitive outcomes (Table 2), including Stroop Color and Word Test, and participant-rated ECog memory and companion-rated ECog language. The RP group showed a decline in performance compared to the NC group over time on SDMT. There were baseline gradient effects for the participant-rated WHODAS and longitudinal gradient effects for the companion-rated WHODAS. There was a baseline gradient effect for the TMS and longitudinal gradient effect for the striatal volume. Therefore, visualization of data suggest gradient effects across groups in all domains: behavioral, cognitive, functional, motor, and imaging. These findings could change with larger sample sizes for the IA and RP groups since phenotype expression is likely to be heterogeneous, even in the RP range [32].

The precise mechanism of neurological damage in HD is unclear, although it likely involves multiple processes [1]. Two proposed pathways include a cumulative damage model and a one-hit model [33]. The cumulative damage model is supported by the negative association between CAG length and age of onset. However, the one-hit model supports the phenomenon of a threshold for manifest HD at 36 CAG repeats. Although our data demonstrate gradient effects for several measures across the CAG length ranges, we do not find evidence of pathology or decline over time in the IA group that differs significantly from control participants.

Figures 1, 2, and 3 reinforce that participants in the IA group show patterns of change similar to those in the NC group, while those in the RP group show patterns similar to those published in the literature in fully-penetrant HD. Genetic counselors and clinicians encounter the issue of explaining the significance of CAG repeat lengths in the IA and RP ranges to individuals undergoing HD genetic testing with frequency. Indeed, subtle differences between the IA and the NC group are evident in the figures. In Fig. 1, the upward slope for performance by controls on the SDMT is indicative of practice effects [26]. The RP group shows downward slopes indicative of cognitive impairment that overrides practice effects. While the IA group slope is slightly positive, it is flatter than the control group slope, which might indicate subtle cognitive changes that result in reduced ability to benefit from practice effects. In Fig. 2, the slope for TMS is slightly negative, unlike the slopes for the RP group. Thus, the motor phenotype is not displayed by the IA group in our sample. Figure 3 shows that striatal volume decreases over time in all groups, which might be correlated with aging. While the slope in the IA is not significantly different from the control group, the striatal volume is slightly lower and a longitudinal gradient effect was evident in the data.

More data are needed before definitely stating that the IA range displays a phenotype consistent with HD, including whether it might involve a behavioral subphenotype. There is evidence that there are likely environmental and genetic factors that impact whether persons in the RP range develop manifest HD. Once we have more specific information regarding the factors that impact phenotype expression in the IA and RP ranges, it might become increasingly important to more accurately report the length of a person’s longest CAG repeat allele. Little attention is given in the literature to the issue of inconsistent reporting of CAG repeat lengths, which occurs in up to 51% of tests [34]. The ACMG/ASHG guidelines state that acceptable error rates for CAG repeat lengths is±2 repeats for alleles with less than 50 repeats [4]. This error rate might not be acceptable for individuals at the edge of one of the repeat ranges. In the future, HD genetic testing might require a two-tier approach using an additional long-read sequencing platform such as PacBio [35, 36] for persons with CAG lengths in the equivocal ranges (i.e., within 2 CAG repeats of another range).

The major limitation of our findings is the small sample size of participants with CAG alleles in the IA and RP ranges compared to NC group in the PREDICT-HD study. The 21 participants in our IA group represent only 1.5% of our sample. This is not overly surprising considering that prevalence of IA in the general population has not been thoroughly examined, with estimates ranging from 1.9% –6% [37 –39]. The RP group of 90 represents 6.4% of our sample. Previous studies indicate that the CAG repeat length in the general population is bimodal, with an average of 17 for those with longest alleles in the NC range and an average of 41 in the fully-penetrant range. CAG repeat lengths 28–38 may be less common [12]. Most recently, Kay and colleagues reported that RP prevalence might be as high as one in 400, suggesting lower penetrance that previously considered for this sample. Despite this new finding, our RP group showed impairment consistent with our other premanifest HD publications.

A caveat to our analysis when comparing behavioral results from IA analyses in the COHORT and PHAROS studies with PREDICT-HD is that these three observational studies recruited and defined case and control groups differently. In PREDICT-HD, all participants know their HD gene status prior to enrollment; cases are persons who tested positive for the HD gene expansion and controls are persons who tested negative. In COHORT, participants do not have to know their gene status to enroll and controls consist of spouses or caregivers. In PHAROS, all participants are at risk for HD but do not know their gene status; cases are positive for the gene expansion and controls are negative for the gene expansion. Therefore, it is reasonable to expect that behavioral outcomes might differ between at-risk individuals who know their gene status versus those who do not; thus, readers should use caution when comparing our findings with those from the COHORT and PHAROS studies. There is evidence that persons who feel as though they will cope poorly with positive results self-select to not complete HD genetic testing [40]. Furthermore, genetic testing protocols and pretest counselling might screen out individuals who are more psychologically vulnerable [41]. Therefore, our sample, which only includes persons who have chosen to undergo testing for the HD gene expansion, might not be representative of all individuals at risk for HD in terms of psychological functioning. This could explain why we found less evidence of behavioral differences between those in the IA and the NC range than the studies that included participants who are blinded to their HD gene expansion status.

CONCLUSION

Our data compared baseline and longitudinal differences in cognitive, behavioral, functional, motor, and imaging outcomes across NC, IA, and RP CAG range groups. We found evidence of baseline and longitudinal differences in the RP group compared to the NC group. We also found gradient effects on a number of measures across domains, supporting a cumulative damage effect for the CAG repeat expansion. On the other hand, only persons in the RP range had outcome measure results significantly different from NC participants, supporting a threshold phenomenon of HD pathology at 36 CAG repeats. More data are needed to accurately characterize the IA subphenotype.

PREDICT-HD INVESTIGATORS, COORDINATORS, MOTOR RATERS, COGNITIVE RATERS

Isabella De Soriano, Courtney Shadrick, and Amanda Miller (University of Iowa, Iowa City, Iowa, USA);

Edmond Chiu, Joy Preston, Anita Goh, Stephanie Antonopoulos, and Samantha Loi (St. Vincent’s Hospital, The University of Melbourne, Kew, Victoria, Australia);

Phyllis Chua and Angela Komiti (The University of Melbourne, Royal Melbourne Hospital, Melbourne, Victoria, Australia);

Lynn Raymond, Joji Decolongon, Mannie Fan, and Allison Coleman (University of British Columbia, Vancouver, British Columbia, Canada);

Christopher A. Ross, Mark Varvaris, Maryjane Ong, and Nadine Yoritomo (Johns Hopkins University, Baltimore, Maryland, USA);

William M. Mallonee and Greg Suter (Hereditary Neurological Disease Centre, Wichita, Kansas, USA);

Ali Samii, Emily P. Freney, and Alma Macaraeg (University of Washington and VA Puget Sound Health Care System, Seattle, Washington, USA);

Randi Jones, Cathy Wood-Siverio, and Stewart A. Factor (Emory University School of Medicine, Atlanta, Georgia, USA);

Roger A. Barker, Sarah Mason, and Natalie Valle Guzman (John van Geest Centre for Brain Repair, Cambridge, UK);

Elizabeth McCusker, Jane Griffith, Clement Loy, Jillian McMillan, and David Gunn (Westmead Hospital, Sydney, New South Wales, Australia);

Michael Orth, Sigurd Süβmuth, Katrin Barth, Sonja Trautmann, Daniela Schwenk, and Carolin Eschenbach (University of Ulm, Ulm, Germany);

Kimberly Quaid, Melissa Wesson, and Joanne Wojcieszek (Indiana University School of Medicine, Indianapolis, Indiana, USA);

Mark Guttman, Alanna Sheinberg, Albie Law, and Irita Karmalkar (Centre for Addiction and Mental Health, University of Toronto, Markham, Ontario, Canada);

Susan Perlman and Brian Clemente (UCLA Medical Center, Los Angeles, California, USA); Michael D. Geschwind, Sharon Sha, Joseph Winer, and Gabriela Satris (University of California, San Francisco, San Francisco, California, USA); Tom Warner and Maggie Burrows (National Hospital for Neurology and Neurosurgery, London, UK);

Anne Rosser, Kathy Price, and Sarah Hunt (Cardiff University, Cardiff, Wales, UK); Frederick Marshall, Amy Chesire, Mary Wodarski, and Charlyne Hickey (University of Rochester, Rochester, New York, USA);

Peter Panegyres, Joseph Lee, Maria Tedesco, and Brenton Maxwell (Neurosciences Unit, Graylands, Selby-Lemnos & Special Care Health Services, Perth, Western Australia, Australia);

Joel Perlmutter, Stacey Barton, and Shineeka Smith (Washington University, St. Louis, Missouri, USA);

Zosia Miedzybrodzka, Daniela Rae, Vivien Vaughan, and Mariella D’Alessandro (Clinical Genetics Centre, Aberdeen, Scotland, UK);

David Craufurd, Judith Bek, and Elizabeth Howard (University of Manchester, Manchester, UK); Pietro Mazzoni, Karen Marder, and Paula Wasserman (Columbia University Medical Center, New York, New York, USA);

Rajeev Kumar, Diane Erickson, Christina Reeves, and Breanna Nickels (Colorado Neurological Institute, Englewood, Colorado, USA);

Vicki Wheelock, Lisa Kjer, Amanda Martin, and Sarah Farias (University of California, Davis, Sacramento, California, USA);

Wayne Martin, Oksana Suchowersky, Pamela King, Marguerite Wieler, and Satwinder Sran (University of Alberta, Edmonton, Alberta, Canada);

Anwar Ahmed, Stephen Rao, Christine Reece, Alex Bura, and Lyla Mourany (Cleveland Clinic Foundation, Cleveland, Ohio, USA);

Executive committee

Principal Investigator Jane S. Paulsen, Jeffrey D. Long, Hans J. Johnson, Thomas Brashers-Krug, Phil Danzer, Amanda Miller, H. Jeremy Bockholt, and Kelsey Montross.

Scientific consultants

Deborah Harrington (University of California, San Diego); Holly Westervelt (Rhode Island Hospital/Alpert Medical School of Brown University); Elizabeth Aylward (Seattle Children’s Research Institute); Stephen Rao (Cleveland Clinic); David J. Moser, Janet Williams, Nancy Downing, Vincent A. Magnotta, Hans J. Johnson, Thomas Brashers-Krug, Jatin Vaidya, Daniel O’Leary, and Eun Young Kim (University of Iowa).

Core sections

Biostatistics: Jeffrey D. Long, Ji-In Kim, Spencer Lourens (University of Iowa); Ying Zhang and Wenjing Lu (University of Indiana).

Ethics: Cheryl Erwin (Texas Tech University Health Sciences Center); Thomas Brashers-Krug, Janet Williams (University of Iowa); and Martha Nance (University of Minnesota).

Biomedical Informatics: H. Jeremy Bockholt, Jason Evans, and Roland Zschiegner (University of Iowa).

CONFLICT OF INTEREST

Jeffrey D. Long has a consulting agreement with NeuroPhage, LLC, and is a paid consultant for Roche Pharma (F. Hoffman La-Roche Ltd.) and Azevan Pharmaceuticals, Inc.

Footnotes

ACKNOWLEDGMENTS

We thank the PREDICT-HD sites, the study participants, the National Research Roster for Huntington Disease Patients and Families, the Huntington’s Disease Society of America and the Huntington Study Group. This research included collaboration with the Institute for Clinical and Translational Science at the University of Iowa, which is supported by the National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) program, grant U54TR001356. The CTSA program is led by the NIH’s National Center for Advancing Translational Sciences (NCATS). This publication’s contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

This research and the PREDICT-HD study are funded by Neurobiological Predictors of Huntington’s Disease grant 5R01NS040068 from the NIH, National Institute of Neurological Disorders and Stroke (NINDS), awarded to JSP; grants A3917 and 6266 from CHDI Foundation, Inc., awarded to JSP; and Cognitive and Functional Brain Changes in Preclinical Huntington’s Disease (HD) grant 5R01NS054893 from NIH/NINDS awarded to JSP.

Appendix

References

Ross

, Tabrizi

. Huntington’s disease: From molecular pathogenesis to clinical treatment. Lancet Neurol. 2011;10(1):83–98. doi: 10.1016/S1474-4422(10)70245-3

Lee

, Ramos

, Lee

, Gillis

, Mysore

, Hayden

, et al. CAG repeat expansion in Huntington disease determines age at onset in a fully dominant fashion. Neurology. 2012;78(10):690–5. doi: 10.1212/WNL.0b013e318249f683

Unified Huntington’s Disease Rating Scale: Reliability and consistency. Huntington Study Group. Mov Disord. 1996;11(2):136–42. doi: 10.1002/mds.870110204

ACMG/ASHG statement. Laboratory guidelines for Huntington disease genetic testing. The American College of Medical Genetics/American Society of Human Genetics Huntington Disease Genetic Testing Working Group. Am J Hum Genet. 1998;62(5):1243–7.

Quarrell

, Rigby

, Barron

, Crow

, Dalton

, Dennis

, et al. Reduced penetrance alleles for Huntington’s disease: A multi-centre direct observational study. J Med Genet. 2007;44(3):e68. doi: 10.1136/jmg.2006.045120

Andrich

, Arning

, Wieczorek

, Kraus

, Gold

, Saft

. Huntington’s disease as caused by 34 CAG repeats. Mov Disord. 2008;23(6):879–81. doi: 10.1002/mds.21958

, Jankovic

. Exploring the correlates of intermediate CAG repeats in Huntington disease. Postgrad Med. 2011;123(5):116–21. doi: 10.3810/pgm.2011.09.2466

Groen

, de Bie

, Foncke

, Roos

, Leenders

, Tijssen

. Late-onset Huntington disease with intermediate CAG repeats: True or false?J Neurol Neurosurg Psychiatry. 2010;81(2):228–30. doi: 10.1136/jnn2008.170902

Kenney

, Powell

, Jankovic

. Autopsy-proven Huntington’s disease with 29 trinucleotide repeats. Mov Disord. 2007;22(1):127–30. doi: 10.1002/mds.21195

10.

Squitieri

, Esmaeilzadeh

, Ciarmiello

, Jankovic

. Caudate glucose hypometabolism in a subject carrying an unstable allele of intermediate CAG(33) repeat length in the Huntington’s disease gene. Mov Disord. 2011;26(5):925–7. doi: 10.1002/mds.23623

11.

, Beck

, Jankovic

. Intermediate CAG repeats in Huntington’s disease: Analysis of COHORT. Tremor Other Hyperkinet Mov (N Y). 2012;2. pii: tre-02-64-287-4. doi: 10.7916/D8FF3R2P

12.

Killoran

, Biglan

, Jankovic

, Eberly

, Kayson

, Oakes

, et al. Characterization of the Huntington intermediate CAG repeat expansion phenotype in PHAROS. Neurology. 2013;80(22):2022–7. doi: 10.1212/WNL.0b013e318294b304

13.

Aylward

, Harrington

, Mills

, Nopoulos

, Ross

, Long

, et al. Regional atrophy associated with cognitive and motor function in prodromal Huntington disease. J Huntingtons Dis. 2013;2(4):477–89. doi: 10.3233/JHD-130076

14.

Downing

, Kim

, Williams

, Long

, Mills

, Paulsen

, et al. WHODAS 2.0 in prodromal Huntington disease: Measures of functioning in neuropsychiatric disease. Eur J Hum Genet. 2014;22(8):958–63. doi: 10.1038/ejhg.2013.275

15.

Duff

, Paulsen

, Beglinger

, Langbehn

, Wang

, Stout

, et al. “Frontal” behaviors before the diagnosis of Huntington’s disease and their relationship to markers of disease progression: Evidence of early lack of awareness. J Neuropsychiatry Clin Neurosci. 2010;22(2):196–207. doi: 10.1176/appi.neuropsych.22.2.196

16.

Paulsen

, Long

, Johnson

, Aylward

, Ross

, Williams

, et al. Clinical and biomarker changes in premanifest Huntington disease show trial feasibility: A decade of the PREDICT-HD Study. Front Aging Neurosci. 2014;6:78. doi: 10.3389/fnagi.2014.00078

17.

Paulsen

, Long

, Ross

, Harrington

, Erwin

, Williams

, et al. Prediction of manifest Huntington’s disease with clinical and imaging measures: A prospective observational study. Lancet Neurol. 2014;13(12):1193–201. doi: 10.1016/S1474-4422(14)70238-8

18.

Williams

, Kim

, Downing

, Farias

, Harrington

, Long

, et al. Everyday cognition in prodromal Huntington disease. Neuropsychology. 2015;29(2):255–67. doi: 10.1037/neu0000102

19.

Stroop

. Studies of interference in serial verbal reactions. J Exp Psychol. 1935;18:643–62. doi: 10.1037/0096-3445.121.1.15

20.

Smith

. Symbol Digit Modalities Test: Manual (revised). Los Angeles: Western Psychological Services; 1991.

21.

Grace

, Malloy

, Psychological Assessment Resources I. FrSBe, Frontal Systems Behavior Scale; Professional Manual: Psychological Assessment Resources; 2001.

22.

Derogatis

. SCL-90-R: Symptom Checklist-90-R: Administration, Scoring, and Procedures Manual: NCS Pearson; 1996.

23.

Farias

, Mungas

, Reed

, Cahn-Weiner

, Jagust

, Baynes

, et al. The measurement of everyday cognition (ECog): Scale development and psychometric properties. Neuropsychology. 2008;22(4):531–44. doi: 10.1037/0894-4105.22.4.531

24.

Ustun

, Chatterji

, Kostanjsek

, Rehm

, Kennedy

, Epping-Jordan

, et al. Developing the world health organization disability assessment schedule 2.0. Bull World Health Organ. 2010;88(11):815–23. doi: 10.2471/BLT.09.067231

25.

Magnotta

, Harris

, Andreasen

, O’Leary

, Yuh

, Heckel

. Structural MR image processing using the BRAINS2 toolbox. Comput Med Imaging Graph. 2002;26(4):251–64.

26.

Paulsen

, Long

. Onset of Huntington’s disease: Can it be purely cognitive?Mov Disord. 2014;29(11):1342–50. doi: 10.1002/mds.25997

27.

Laird

, Ware

. Random-effects models for longitudinal data. Biometrics. 1982;38(4):963–74.

28.

Akaike

. A new look at the statistical model identification. IEEE Trans Automat Contr. 1974;19:716–723.

29.

Little

RJA

, Rubin

. Statistical analysis with missing data. 2nd ed. New York: Wiley; 2014. 278p.

30.

Benjamini

, Hochberg

. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 1995;57(1):289–300.

31.

Scuffham

, MacMillan

. Huntington disease: Who seeks presymptomatic genetic testing, why and what are the outcomes?J Genet Couns. 2014;23(5):754–61. doi: 10.1007/s10897-013-9678-z

32.

Panegyres

, Shu

, Chen

, Paulsen

. Factors influencing the clinical expression of intermediate CAG repeat length mutations of the Huntington’s disease gene. J Neurol. 2015;262(2):277–84. doi: 10.1007/s00415-014-7559-5

33.

Sugaya

. Modeling the polyglutamine aggregation pathway in Huntington’s disease: From basic studies to clinical applications. Subcell Biochem. 2012;65:353–88. doi: 10.1007/978-94- 007-5416-4_15

34.

Quarrell

, Handley

, O’Donovan

, Dumoulin

, Ramos-Arroyo

, Biunno

, et al. Discrepancies in reporting the CAG repeat lengths for Huntington’s disease. Eur J Hum Genet. 2012;20(1):20–6. doi: 10.1038/ejhg.2011.136

35.

Doi

, Monjo

, Hoang

, Yoshimura

, Yurino

, Mitsui

, et al. Rapid detection of expanded short tandem repeats in personal genomics using hybrid sequencing. Bioinformatics. 2014;30(6):815–22. doi: 10.1093/bioinformatics/btt647

36.

Huddleston

, Ranade

, Malig

, Antonacci

, Chaisson

, Hon

, et al. Reconstructing complex regions of genomes using long-read sequencing technology. Genome Res. 2014;24(4):688–96. doi: 10.1101/gr.168450.113

37.

Costa Mdo

, Magalhaes

, Guimaraes

, Maciel

, Sequeiros

, Sousa

. The CAG repeat at the Huntington disease gene in the Portuguese population: Insights into its dynamics and to the origin of the mutation. J Hum Genet. 2006;51(3):189–95. doi: 10.1007/s10038-005-0343-8

38.

Goldberg

, McMurray

, Zeisler

, Almqvist

, Sillence

, Richards

, et al. Increased instability of intermediate alleles in families with sporadic Huntington disease compared to similar sized intermediate alleles in the general population. Hum Mol Genet. 1995;4(10):1911–8.

39.

Sequeiros

, Ramos

, Cerqueira

, Costa

, Sousa

, Pinto-Basto

, et al. Large normal and reduced penetrance alleles in Huntington disease: Instability in families and frequency at the laboratory, at the clinic and in the population. Clin Genet. 2010;78(4):381–7. doi: 10.1111/j.1399-0004.2010.01388.x

40.

Rivera-Navarro

, Cubo

, Mariscal

Analysis of the reasons for non-uptake of predictive testing for Huntington’s disease in Spain: A qualitative study. J Genet Couns. 2015;24(6):1011–21. doi: 10.1007/s10897-015-9840-x

41.

Meiser

, Dunn

. Psychological effect of genetic testing for Huntington’s disease: An update of the literature. West J Med. 2001;174(5):336–40.