Digital Gait Biomarkers Allow to Capture 1‐Year Longitudinal Change in Spinocerebellar Ataxia Type 3

Abstract

Measures of step variability and body sway during gait have shown to correlate with clinical ataxia severity in several cross‐sectional studies. However, to serve as a valid progression biomarker, these gait measures have to prove their sensitivity to robustly capture longitudinal change, ideally within short time frames (eg, 1 year). We present the first multicenter longitudinal gait analysis study in spinocerebellar ataxias. We performed a combined cross‐sectional (n = 28) and longitudinal (1‐year interval, n = 17) analysis in Spinocerebellar Ataxia type 3 subjects (including seven preataxic mutation carriers). Longitudinal analysis showed significant change in gait measures between baseline and 1‐year follow‐up, with high effect sizes (stride length variability: P = 0.01, effect size rprb = 0.66; lateral sway: P = 0.007, rprb = 0.73). Sample size estimation for lateral sway indicates a required cohort size of n = 43 for detecting a 50% reduction of natural progression, compared with n = 240 for the clinical ataxia score Scale for the Assessment and Rating of Ataxia (SARA). These measures thus present promising motor biomarkers for upcoming interventional studies. © 2022 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society

Responsiveness of the Scale for the Assessment and Rating of Ataxia and Natural History in 884 Recessive and Early Onset Ataxia Patients

“…In advanced ataxia (SARA > 25), a relative 2‐ to 3‐fold increase in trial size—irrespective of genetic stratification—renders the use of the SARA as primary COA in rare genetic ataxias practically impossible. For these patients—almost 15% of our cohort—better COAs, or acceptance of other, nonclinical outcome measures, are thus needed, for example, fluid markers such as neurofilament light chain (with reductions of sample size in SCA3 by possibly up to a factor of 40) 3,44 or digital‐motor measures (with reduction by possibly up to a factor of 6) 45,46 …”

“…For these patients-almost 15% of our cohort-better COAs, or acceptance of other, nonclinical outcome measures, are thus needed, for example, fluid markers such as neurofilament light chain (with reductions of sample size in SCA3 by possibly up to a factor of 40) 3,44 or digitalmotor measures (with reduction by possibly up to a factor of 6). 45,46 Rank-Optimized SARA An improved version of the SARA might be established by a generic ranking and selection of only the most responsive SARA items, across ataxia types (roSARA). Our study demonstrates that finger-chase and nose-finger are not beneficial or "neutral" subitems, but detrimental to the performance of the SARA as COA.…”

Capturing clinical progression in multisystemic genetic ataxias: lessons from a prospective study of 884 patients with autosomal recessive or early-onset ataxia

“…For these patients -almost 15% of our cohort-better COAs, or acceptance of other, non-clinical outcome measures (e.g. fluid 39 or digital-motor measures 40, 41 ) are thus needed.…”

Digital gait measures capture 1-year progression in early-stage spinocerebellar ataxia type 2

“…Measures of spatiotemporal gait variability, such as step length/stride length and step time/stride time variability, have been shown in cross-sectional studies 15, 24-28 and a longitudinal study in SCA3 29 to be sensitive to ataxia-related gait changes and to be associated with an increased risk of falls 30 . Variability measures were calculated using the coefficient of variation CV=σ/μ, with the standard deviation normalized to the mean 31 .…”

“…From the rich source of possible gait measures, we adopted a hypothesis-driven approach here, selecting only those measures that were considered promising candidate features based on previous studies: Stride length and stride time variability: Measures of spatiotemporal gait variability, such as step length/stride length and step time/stride time variability, have been shown in cross-sectional studies 15,[24][25][26][27][28] and a longitudinal study in SCA3 29 to be sensitive to ataxia-related gait changes and to be associated with an increased risk of falls 30 . Variability measures were calculated using the coefficient of variation CV=σ/μ, with the standard deviation normalized to the mean 31 .…”