Detection of Unusual Increases in MRI Lesion Counts in Individual Multiple Sclerosis Patients

• Activity Number: 547
• Type: Topic Contributed
• Date/Time: Wednesday, August 6, 2014 : 2:00 PM to 3:50 PM
• Sponsor: SSC
• Author(s): Yinshan Zhao*+ and Albert John Petkau and David Li and Andrew Riddehough and Anthony Traboulsee
• Companies: University of British Columbia and University of British Columbia and University of British Columbia and University of British Columbia and University of British Columbia
• Keywords: Longitudinal count data ; negative binomial ; random effects model ; semi-parametric estimation

Abstract

Data Safety and Monitoring Boards for multiple sclerosis clinical trials consider an increase of contrast enhancing lesions on repeated magnetic resonance imaging an indicator for potential adverse events. To identify "unexpected increases" of lesion activity for individual patients, we consider as an index the likelihood of observing lesion counts as large as those observed on the recent scans of a patient conditional on the patient's lesion counts on previous scans. To estimate this index, we rely on random effects models. Given the patient-specific random effect, we assume that the repeated lesion counts from the same patient follow a negative binomial distribution and may be correlated over time. As the value of the conditional probability index can be sensitive to the choice of distribution for the random effects, we explore the utility of two simple semi-parametric estimation procedures for a model with an unspecified distribution for the random effects. We examine the performance of these methods using simulations and illustrate the approach using data from a clinical trial.