Article Text

Download PDFPDF
Clinical characteristics
H04 Prognostic factors for the speed of progression of HD results of a re-evaluation of the riluzole study
  1. W Gaus,
  2. B Einsiedler
  1. Universität Ulm, Institut für Epidemiologie und Medizinische Biometrie, Ulm, Germany


Background One of the precursors of EHDN was the Riluzole-study performed from 1999 to 2004 in 44 centres in eight European countries. Results on Riluzole vs placebo were published in 2007.

Methods Now we did an extended re-evaluation of the data of in order to find prognostic factors influencing the speed of HD-progression. As statistical methodology a multiple linear regression model with backward elimination of variables was developed. Included were the 368 patients valid for efficacy and without data missing for regression. Outcome variable was a combined score derived with equal weights from motor function and total functional capacity subscores of UHDRS. As statistical software we used SAS Proc Mixed.

Results The following variables were removed from the model because they had no significant influence on the speed of progression:

▶ Study treatment Riluzole or placebo had no effect at all on progression (p=0.98).

▶ Patients with a higher Body Mass Index (BMI) had a slight tendency to faster progression (p=0.48).

▶ Third variable removed from the model was gender, although the speed of progression was 1.9 times faster for males compared to females (p=0.20).

▶ If parents of the patient were affected then progression was slower (p=0.19).

The final model had the following variables:

1st The combined score at admission was by far most important. The worse the patient was at admission, the faster was progression (p<0.0001).

2nd The longer the CAG-repeat of allele 2, the faster was progression (p=0.0073).

3rd Progression was faster for older patients than for younger patients (p=0.013).

4th Possibly the age at onset of HD has influence on the speed of progression (p=0.068). The tendency is, the older the patient is at HD onset, the slower is progression.

The correlation between the speed of progression predicted by the model and the actually observed speed was r=0.78. Thus, r2=0.61, that is, 61% of the speed of progression of HD is determined by the mentioned four variables.

Discussion The goodness of fit with r2=0.61 is neither really good, nor really bad. Obviously, speed of progression of HD is not constant over time, it accelerates during the course of the disease.

  • Progression
  • prognostic factors
  • methodology
  • clinical trials

Statistics from

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.