Comparison of Time-to-First-Event and Recurrent Event Methods in Multiple Sclerosis Trials

TL;DR

This paper compares traditional time-to-first-event methods with recurrent event analyses in MS trials, highlighting how recurrent methods can enhance power and clinical insights by accounting for multiple disability progression events.

Contribution

It introduces approaches for defining recurrent disability progression and evaluates various recurrent event models for MS trial analysis.

Findings

Recurrent event analyses can improve statistical power in MS trials.

Different models have varying suitability for analyzing recurrent disability events.

Defining disability progression accurately is crucial for effective recurrent event analysis.

Abstract

Suppression of disability progression is an important goal in the treatment of multiple sclerosis (MS). Randomized clinical trials in MS frequently use the time to the first confirmed disability progression (CDP) on the ordinal Expanded Disability Status Scale (EDSS) as an endpoint. However, especially in progressive forms of MS, patients may experience repeated CDP events over time. We investigate first how recurrent disability progression events can be defined, and consider then recurrent event analyses that could increase power and improve clinical interpretation of results. Data on disease activity and results from two simulation studies which compare analyses of the time to the first event with recurrent event analyses (including negative binomial, Andersen-Gill and Lin-Wei-Ying-Yang models) are presented to demonstrate challenges in defining disability progression and to identify analysis methods suitable for treatment comparisons in MS trials with disability progression endpoints. This article is based on the Master Thesis by Alexandra B\=uhler at the University of Ulm which was written under the supervision of Jan Beyersmann, Marcel Wolbers, Fabian Model and Qing Wang.