Machine learning in Huntington's disease: exploring the Enroll-HD dataset for prognosis and driving capability prediction.

BACKGROUND: In biomedicine, machine learning (ML) has proven beneficial for the prognosis and diagnosis of different diseases, including cancer and neurodegenerative disorders. For rare diseases, however, the requirement for large datasets often prevents this approach. Huntington's disease (HD) is a rare neurodegenerative disorder caused by a CAG repeat expansion in the coding region of the huntingtin gene. The world's largest observational study for HD, Enroll-HD, describes over 21,000 participants. As such, Enroll-HD is amenable to ML methods. In this study, we pre-processed and imputed Enroll-HD with ML methods to maximise the inclusion of participants and variables. With this dataset we developed models to improve the prediction of the age at onset (AAO) and compared it to the well-established Langbehn formula. In addition, we used recurrent neural networks (RNNs) to demonstrate the utility of ML methods for longitudinal datasets, assessing driving capabilities by learning from previous participant assessments. RESULTS: Simple pre-processing imputed around 42% of missing values in Enroll-HD. Also, 167 variables were retained as a result of imputing with ML. We found that multiple ML models were able to outperform the Langbehn formula. The best ML model (light gradient boosting machine) improved the prognosis of AAO compared to the Langbehn formula by 9.2%, based on root mean squared error in the test set. In addition, our ML model provides more accurate prognosis for a wider CAG repeat range compared to the Langbehn formula. Driving capability was predicted with an accuracy of 85.2%. The resulting pre-processing workflow and code to train the ML models are available to be used for related HD predictions at: https://github.com/JasperO98/hdml/tree/main . CONCLUSIONS: Our pre-processing workflow made it possible to resolve the missing values and include most participants and variables in Enroll-HD. We show the added value of a ML approach, which improved AAO predictions and allowed for the development of an advisory model that can assist clinicians and participants in estimating future driving capability.

Huntington's disease influences employment before and during clinical manifestation: A systematic review.

Huntington's disease (HD) is an inherited neurodegenerative disease. People at risk for HD can choose to get predictive testing years before the clinical onset. HD is characterized by motor, cognitive and psychiatric symptoms and has a mean age at onset between 30 and 50 years, an age at which people are usually still working. This systematic review focuses on summarizing which disease-specific characteristics influence employment and working capacity in HD. Twenty-three studies were identified and showed that while employment and working capacity in HD are negatively influenced by cognitive decline and motor impairments, apathy already plays a role in the prodromal stage. Moreover, the influence of HD transcends the clinical manifestation of the disease, as some people at risk are already experiencing the impact of HD on employment through fear of or actual genetic discrimination. Employment and working capacity are not influenced by predictive testing for HD in and of itself.

Predictors of Working Capacity Changes Related to Huntington's Disease: A Longitudinal Study.

BACKGROUND: Huntington's disease (HD) is an inherited neurodegenerative disorder that is characterized by motor, cognitive, and psychiatric symptoms. Although 65%of HD expanded gene carriers report changes in employment as the first functional loss, little is known about the predictors leading to changes of working capacity. Given the impact on quality of life, understanding of these factors is of great clinical value. OBJECTIVE: This study evaluates disease specific characteristics and their predictive value in loss of working capacity in HD. METHODS: Longitudinal data was collected through the worldwide observational study (Enroll-HD), with 15,301 participants in total and 2,791 HD and healthy control participants meeting the inclusion criteria. Changes in working capacity were analyzed by means of a survival analysis. Predictive values of demographic factors and clinical characteristics were assessed for premanifest and manifest HD through Cox regressions. RESULTS: HD expanded gene carriers, manifest and premanifest combined, had a 31%chance of experiencing changes in employment after three years, compared to 4%in healthy controls. Apathy was found to be the most crucial determinant of working capacity changes in premanifest HD, while executive and motor dysfunction play an important role in manifest HD. CONCLUSION: HD expanded gene carriers are more likely to lose working capacity compared to healthy controls. Disease progression, altered motor function, cognitive decline, and in an early stage of the disease apathetic symptoms are indicative of negative changes in working capacity. Clinicians should recognize that early disease related changes, especially apathy, can affect working capacity.

The prevalence and the burden of pain in patients with Huntington disease: a systematic review and meta-analysis.

It is remarkable that studies focusing on the prevalence and the burden of pain in patients with Huntington disease (HD) are scarce. This may lead to inadequate recognition of pain and hence lack of treatment, eventually affecting the quality of life. The aim of this review is to investigate the prevalence of pain and its burden in HD by performing a systematic literature search. In February 2018, a systematic search was performed in the electronic databases of Pubmed, Embase, Cinahl, Cochrane, and PsycINFO. Studies focusing on patients with juvenile HD were excluded. All other types of study were included without language restrictions. In total, 2234 articles were identified, 15 of which met the inclusion criteria and provided information on 2578 patients with HD. The sample-weighted prevalence of pain was 41.3% (95% confidence interval: 36%-46%). The pain burden, which was measured with the SF-36, is significantly less compared with that in the general population. The sample-weighted mean score on the SF-36 was 84 (95% confidence interval: 81-86), where a score of 100 represents the lowest symptom burden. The results demonstrate that pain could be an important nonmotor symptom in patients with HD, and there are indications that the pain burden could be diminished because of HD. Larger and high-quality prospective cohort and clinical studies are required to confirm these findings. In the meantime, awareness about pain and its burden in patients with HD is warranted in clinical practice.

Visual Object Perception in Premanifest and Early Manifest Huntington's Disease.

OBJECTIVE: In Huntington's disease (HD), a hereditary neurodegenerative disorder, cognitive impairment in early disease stages mainly involves executive dysfunction. However, visual cognitive deficits have additionally been reported and are of clinical relevance given their influence on daily life and overall cognitive performance. This study aimed to assess visual perceptual skills in HD gene carriers. METHODS: Subtasks of the Visual Object and Space Perception battery and Groningen Intelligence Test were administered in 62 participants (18 healthy controls, 22 participants with a genetic confirmation of HD without symptoms, i.e., premanifest HD, and 22 participants with a genetic confirmation of HD with symptoms, i.e., manifest HD). Group differences in task performance were measured using analysis of covariance with and without correction for age. Receiver Operating Characteristics (ROC) analysis was performed to examine which task best discriminated between groups and cut-off scores were provided. RESULTS: Manifest HD performed significantly worse compared to both controls and premanifest HD on all visual perceptional tasks. Premanifest HD did not differ in task performance from controls. Besides the Shape Detection, all tasks were robust in discriminating between groups. The Animal Silhouettes test was most accurate in discriminating manifest HD from premanifest HD (AUC = 0.90, SE = 0.048, p < .001). CONCLUSION: Visual perceptual deficits are present in early manifest HD, especially an impaired recognition of animals and objects from sketched silhouettes, and not in premanifest HD. This suggests that decline in visual processing only occurs in clinical disease stages. The visual cognitive battery, especially the Silhouettes tasks used in this study is sensitive in discriminating manifest HD from premanifest HD and controls.