Article Text

Download PDFPDF
F22 Robust biomarkers of huntington’s disease progression: observations from the track-hd, predict-hd and image-hd studies
  1. Peter A Wijeratne1,
  2. Eileanoir B Johnson2,
  3. Sarah Gregory2,
  4. Amrita Mohan3,
  5. Cristina Sampaio3,
  6. Rachael I Scahill2,
  7. Sarah J Tabrizi2,
  8. Daniel C Alexander1
  1. 1Centre for Medical Image Computing, Department of Computer Science, University College London, Gower Street, London, UK
  2. 2Huntington’s Disease Research Centre, University College London, 2nd Floor Russell Square House, London, UK
  3. 3CHDI Management/CHDI Foundation, 350 7th Avenue, New York, NY, USA

Abstract

Background The TRACK, PREDICT and IMAGE-HD studies provide rich and varied datasets with which to identify robust imaging and clinical biomarkers of Huntington’s disease (HD) progression. A comparative analysis of biomarkers between studies has potential use in observational study design. Estimating the sequence in which these biomarkers become abnormal can provide important insights into HD pathology and a mechanism for disease staging.

Aims We have, for the first time, analysed and statistically compared structural imaging and phenotypic clinical data from these three observational studies. We hence aim to identify a common set of robust biomarkers, and explain observational differences between studies. We also propose how to use these biomarkers to inform a model of HD progression.

Methods We analysed structural imaging, clinical and behavioural data from a total of 357 TRACK, 1091 PREDICT, and 96 IMAGE-HD participants at baseline. The imaging data were segmented and parcellated using a common framework. Groupwise comparisons were made between controls, pre-manifest and manifest groups, and effect sizes compared between studies. An event-based model1 was trained to infer the most likely sequence of biomarker abnormality, and to stage participants.

Results We identified a core set of significant imaging, clinical and behavioural biomarkers common to all studies, plus biomarkers that were significant within, but not between studies. Consequently, the disease progression model reveals a distinct, cross-validated pattern of imaging and phenotypic abnormality.

Conclusions We successfully identified a set of robust biomarkers common to all studies, explored observational differences, and demonstrated that these biomarkers can be used to model HD progression.

Reference

  1. . Wijeratne, et al. Ann Clin Trans Neurol2018. doi:10.1002/acn3.558

  • biomarkers
  • multi study
  • disease progression

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.