F026 Refined changepoint model with an interval-censored anchoring event: an investigation of motor function using enroll-HD

<h3>Background</h3> There is an acceleration in rate of motor decline with progression of Huntington’s disease (HD). We identified this inflection point in a changepoint model using Enroll-HD data for various CAG lengths as the time preceding clinical motor diagnosis (CMD), i.e. when the diagnostic confidence level of 4 is assigned by the clinician. However, since the change may have occurred between two study visits, a refined model was constructed. <h3>Objective</h3> To study the inflection point of motor decline for individual CAG accounting for the interval-censoring nature of the CMD. <h3>Methods</h3> A refined changepoint model was applied to Enroll-HD study PDS5 for each CAG length 40–48. The results are compared to original model to draw a cohesive conclusion. <h3>Results</h3> On average, this inflection point occurred approximately two years before CMD with a decline rate of 0.35 points of UHDRS® total motor score prior to the inflection and 2.79 points thereafter; its timing and magnitude varied by CAG length. The decline rates before and after the acceleration are generally consistent between the two methods, with median [min, max] differences of 0.03 [0.00, 0.11] and -0.07 [-0.11, -0.04], respectively. The time window between the inflection point and CMD, accounting for interval censoring, is around a median of six months shorter (min=5.4, max=9.1) than the original method, aligning with half the period between scheduled follow-up study visits. <h3>Conclusions</h3> The alignment of results from both models confirms the inflection point and opens up the potential to generate transformative insights that can be used in clinical management and clinical trial design.