Development and evaluation of a time to event endpoint for clinical trials in Duchenne muscular dystrophy (DMD)


Topic:

Clinical Trials

Poster Number: 93

Author(s):

James Signorovitch, PhD, Analysis Group, Inc., Craig McDonald, MD, University of California Davis Health, Francesco Muntoni, MD, UCL Institute of Child Health and Great Ormond Street Hospital for Children, Jessica R. Marden, ScD, Analysis Group, Inc., Nathalie Goemans, MD, PhD, University Hospitals Leuven, Andres Gomez-Lievano, Analysis Group, Inc., Adina Zhang, MSc, Analysis Group, Inc., Susan Ward, Ph,D, cTAP, James Signorovitch, Analysis Group, Inc.

Event-driven clinical trials, commonly used in oncology, offer several benefits over continuous outcome measures: i) each patient is assessed individually, ii) after the event has occurred, patients can receive therapy without breaking the blind. This study aimed to develop a time-to-event outcome measure for the North Star Ambulatory Assessment (NSAA) corresponding to clinically meaningful disease progression in DMD. Three candidate definitions based on changes in NSAA were investigated: A) Loss of 3 NSAA total score points; B) Loss of 2 distinct NSAA items; C) Loss of 3 NSAA points or loss of 2 NSAA items. To evaluate these definitions, 619 DMD patients with expected decline in motor function (baseline rise from supine ≥5 seconds) were studied from 3 clinical trial placebo arms and 4 natural history studies. At baseline, mean age was 9.3 years (range: 6-18yrs), mean NSAA total score was 22 (range:12-34); 63% were receiving prednisone, 37% deflazacort, and 0.4% no steroids. For each definition, median time to progression was 11.0, 14.0 and 8.3 months, respectively, based on Kaplan-Meier analyses. As a measure of prognostic validity, a progression event during the first year of follow-up was associated with significantly shorter time to loss of ambulation, with hazard ratios of 2.5, 6.5 and 3.2, respectively (all p<0.05), based on landmark analyses. Stability of the progression definitions varied, with approximately 10%, 19% and 12%, respectively, regaining their baseline level of function within a year. Definition A exhibited a favorable combination of event rate, prognostic validity, and stability. A hypothetical, well-powered, randomized, event-driven trial based on endpoint A and enrolling 160 DMD patients over a period of 9 months would require an expected duration of 10, 14, or 25 months to detect treatment effects delaying progression by 13.0, 8.4 or 5.4 months, respectively. Thus, event-driven trials may be viable for drug development in ambulatory DMD.