IDENTIFICATION AND VERIFICATION OF NOVEL VARIABLES IN QUANTITATIVE MOTOR TESTS (Q-MOTOR) IN HUNTINGTON'S DISEASE, USING THE TRACK-HD DATA SET

Background Clinical rating scales in clinical trials for Huntington’s disease (HD) have shown limited sensitivity in the premanifest stage of HD and often suffer from inter- and intra-rater variability and bias. Thus, there is an unmet need for objective and reliable assessments to serve as outcome measures in clinical trials. The Q-Motor, a sensor-based assessment battery, was developed to provide accurate and precise quantification of motor performance. During the analysis of the TRACK-HD study, a couple of variables turned out to be particularly sensitive, but further approaches on testing alternative variables have not been conducted, yet. Objective To identify novel conceptual variables for the Q-Motor digitomotography (speeded tapping) assessment, that have the potential to be more sensitive and robust particularly in the premanifest stage. The applicability of these variables should be investigated for cross-sectional and longitudinal analyses, using the TRACK-HD data. Methods Q-Motor raw-data from the TRACK-HD study was used to extract the novel variables. The data sample included 4 year follow up data from 288 participants (age: 48 ± 10 years, female: 124, number of unaffected controls: 94). Statistical analyses were conducted using R. Generalized linear mixed models and ANOVA was used for group comparisons cross-sectional and longitudinally. Correlation with clinical rating scales and imaging parameters was performed. Results Most novel identified variables allow discrimination between controls pre-HD and manifest HD groups. Some even show significance in the more subtle distinction between pre-HD subgroups. Most variables also show good correlations with the clinical Total Motor Score (TMS) and with several magnet resonance (MR) imaging variables. Conclusion The sensitivity observed in the novel variables is comparable to that of previously used variables. However, the additional information may be useful for the creation of a combined measure, which will be explored in a next step.