Using Item response theory to yield information from the MDS-UPDRS items in Parkinsonís disease clinical trials
S.Buatois (1), S.Retout (1), R. Gieschke (1), S.Ueckert (2,3), N.Frey (1)
(1) Pharma Research and Early Development, Clinical Pharmacology, Roche Innovation Center Basel, Switzerland; (2) IAME, UMR 1137, INSERM, F-75018 Paris, France; (3) Pharmetheus AB, Uppsala, Sweden
Objectives: Item response theory (IRT) has been recently introduced in pharmacometrics by Ueckert et al. to model the Alzheimer’s disease progression based on the ADAS-Cog scale . It permits a more precise analysis by integrating the whole available items information, it increases the probability to detect changes due to a drug effect and helps to determine the most informative items of the global score in function of the population of interest. The aim of this work was to apply IRT to the MDS-UPDRS scale to better assess the natural disease progression of Parkinson’s disease (PD) patients.
Methods: First, a baseline IRT model was built to analyze baseline data from the Parkinson’s Progression Marker Initiative (PPMI) database . The dataset includes 431 de novo idiopathic PD patients, 199 healthy controls and 65 PD patients with Scans without Evidence of Dopaminergic Deficit. This work focus on the MDS-UPDRS score , which consists of 65 items measuring the disturbance of non-motor experiences of daily living (i.e. sleep, cognition, mood) as well as motor experiences (i.e. Tremor, Rigidity, Bradykinesia) and motor complications. All item-specific-parameters were implemented as fixed effect via ordered categorical models and the baseline hidden neuronal disability as a random effect. In a second step, a longitudinal model was used to describe the neuronal disability time course [4,5]. Visual predictive checks both on the item as well as on the MDS-UPDRS score level were used to evaluate the model.
Results: We were able to describe the evolution of the neuronal disability over time using a longitudinal IRT modeling approach, for a better leverage of the whole MDS-UPDRS items information. The developed IRT model allowed also to rank the information provided by each item with respect to different severity of patient population and could be useful afterwards to derive simplified but informative MDS-UPDRS sub-scores according to the targeted patient population.
Conclusions: IRT is a powerful tool which enables to yield information from the MDS-UPDRS scale used in clinical trials, therefore maximizing the likelihood to bring new medicine to Parkinson’s disease patients.
 Sebastian Ueckert, Improved Utilization of ADAS-Cog Assessment Data Through Item Response Theory Based Pharmacometric Modeling, 2014 Aug, 31(8):2152-65
 PPMI: Parkinson’s Progression marker initiative www.ppmi-info.org/data
 C G Goetz et al, Movement Disorder Society-Sponsored Revision of the Unified Parkinson’s Disease Rating Scale (MDS-UPDRS): Scale, Movement Disorders, 2007 Jan, 22(1):41-7
 N H.G. Holford et al, Disease Progression and Pharmacodynamics in Parkinson Disease – Evidence for Functional Protection with Levodopa and Other Treatments, Journal of Pharmacokinetics and Pharmacodynamics, 2006 Jun, 33(3):281-311
 T C. Vu et al, Progression of motor and nonmotor features of Parkinson’s disease and their response to treatment, British Journal of Clinical Pharmacology, 2012 Aug, 74(2):267-83