Forecasting individual progression trajectories in Huntington disease enables more powered clinical trials

Variability in neurodegenerative disease progression poses great challenges for the evaluation of potential treatments. Identifying the persons who will experience significant progression in the short term is key for the implementation of trials with smaller sample sizes. We apply here disease course mapping to forecast biomarker progression for individual carriers of the pathological CAG repeat expansions responsible for Huntington disease. We used data from two longitudinal studies (TRACK-HD and TRACK-ON) to synchronize temporal progression of 15 clinical and imaging biomarkers from 290 participants with Huntington disease. We used then the resulting HD COURSE MAP to forecast clinical endpoints from the baseline data of 11,510 participants from ENROLL-HD, an external validation cohort. We used such forecasts to select participants at risk for progression and compute the power of trials for such an enriched population. HD COURSE MAP forecasts biomarkers 5 years after the baseline measures with a maximum mean absolute error of 10 points for the total motor score and 2.15 for the total functional capacity. This allowed reducing sample sizes in trial up to 50% including participants with a higher risk for progression ensuring a more homogeneous group of participants.


Supplementary Figure 1: Association of individual parameters with genetics and socio-demographics.
Multivariate regression of each model parameter (right) against five covariates: sex, number of CAG repeats, level of education, number of visits and handedness. To estimate confidence intervals on the p-values, we bootstrapped the model 100 times and ran 100 multivariate analyses. The reported p-values correspond to the median over the runs. The two black vertical bars correspond to the 0.05 and 0.01 thresholds. Neither education level nor sex influences the speed of progression or the age at onset. Increased CAG repeats advance Stroop abnormality (p=5.5 10 -3 ) and delays grey-matter changes (p = 5.4 10 -8 ), brain atrophy (p = 4.9 10 -4 ) and ventricles volumes (p = 5.1 10 -4 ) compared to the average progression. Sex influences on circle tracing and grey matter needs further investigation. The number of visits influenced every clinical variable but the Stroop test, as well as all the volumes of basal ganglia structures, but it had no influence on grey and white matter volume or on ventricular volume. The number of visits as a covariate reflects a standard bias in cohorts: subjects with more visits are those that can be followed during years, which is usually related to a slow disease progression -fast progressors are less likely to be evaluated for long periods of time because of death, physical or mental disabilities. ROC curves are shown for the detection of participants with an annual change of SDMT of at least 3, 4, 5 or 6 points (in columns) in 1, 2, 3 or 4 years (in rows). Selection was made for ENROLL participants whose baseline SDMT value has already diverged from controls. Three methods are compared: selection based on the PIN, Burden and the SDMT change from baseline that HD COURSE MAP predicts. AUCs for the three methods are reported in the legend. ROC curves are shown for the detection of participants with an annual percentage change of at least 5%, 10%, 15% or 20% points (in columns) in 1, 2, 3 or 4 years (in rows). Selection was made for ENROLL participants whose baseline PBA-Apathy assessment has already diverged from controls. Three methods are compared: selection based on the PIN, Burden and the SDMT change from baseline that HD COURSE MAP predicts. AUCs for the three methods are reported in the legend. ROC curves are shown for the detection of participants with an annual change of TFC of at least 1, 2 or 3 points (in columns) in 1, 2, 3 or 4 years (in rows). Selection was made for ENROLL participants whose baseline TFC value has already diverged from controls. Three methods are compared: selection based on the PIN, Burden and the SDMT change from baseline that HD COURSE MAP predicts. AUCs for the three methods are reported in the legend.

Time to prediction
1 year 2 years 3 years

Targeted annual change of TFC (lower bound)
1 point 2 points 3 points Supplementary Figure 6: HD COURSE MAP identifies participants experiencing changes of TMS from a baseline value ranging from 0 to 5. ROC curves are shown for the detection of participants with an annual change of TMS of at least 1, 2, 3 or 4 points (in columns) in 1, 2, 3 or 4 years (in rows). Selection was made for ENROLL participants whose baseline TMS value was smaller than 5. Three methods are compared: selection based on the PIN, Burden and the SDMT change from baseline that HD COURSE MAP predicts. AUCs for the three methods are reported in the legend.

Time to prediction
1 year 2 years 4 years

Targeted annual change of TMS (lower bound)
1 point 3 points 4 points