Disease progression modeling of the North Star Ambulatory Assessment for Duchenne Muscular Dystrophy

Abstract

Duchenne muscular dystrophy (DMD) is a rare genetic disorder caused by decreased or absent dystrophin gene leading to progressive muscle degeneration and weakness in young boys. Disease progression models for the North Star Ambulatory Assessment (NSAA), a functional measurement widely used to assess outcomes in clinical trials, were developed using a longitudinal population modeling approach. The relationship between NSAA total score over time, loss of ambulation, and potential covariates that may influence disease progression were evaluated. Data included individual participant observations from an internal placebo-controlled phase II clinical trial and from the external natural history database for male patients with DMD obtained through the Cooperative International Neuromuscular Research Group (CINRG). A modified indirect response model for NSAA joined to a loss of ambulation (LOA) time-to-event model described the data well. Age was used as the independent variable because ambulatory function is known to vary with age. The NSAA and LOA models were linked using the dissipation rate constant parameter from the NSAA model by including the parameter as a covariate on the hazard equation for LOA. No covariates were identified. The model was then used as a simulation tool to explore various clinical trial design scenarios. This model contributes to the quantitative understanding of disease progression in DMD and may guide model-informed drug development decisions for ongoing and future DMD clinical trials.

FIGURE 1

Modeling analysis schematic: DMD disease progression, data sources, and model structure. The phenotypic trajectory of disease progression (upper left panel) in boys with DMD begins with lower extremity muscle weakness that results in decreased motor function in the lower limbs and eventually leads to a complete LOA and wheelchair dependency by adulthood. At this time, diminished upper limb motor function is also seen, followed by cardiac and respiratory complications that can ultimately lead to death. The data sources (lower left panel) for NSAA total scores were studies B5161002 and B5161011. In B5161002, participants undergo within subject dose escalation, receiving three dose levels of domagrozumab (5 mg/kg, 20 mg/kg, and 40 mg/kg) each administered every 4 weeks for 4 months (16 weeks total per dose level) in period 1‐sequence 1, period 1‐sequence 2, and period 2‐sequence 3. Participants in period 1‐sequence 3 and period 2‐sequence 2 received placebo. Participants in period 2‐sequence 1 received the highest domagrozumab tolerated dose level from period 1. In B5161011, participants completed visits at baseline and in 3‐month intervals over the course of 1 year; assessments were completed for ambulatory participants at 3, 6, 9, and 12 months and for nonambulatory participants at 6 and 12 months. Additionally, long‐term follow‐up visits were conducted at 18 months (year 2; 6 months) and 24 months (year 2 and onward; 12 months). The joint NSAA‐LOA disease progression model structure (right panel) is a modified indirect response model with stimulation of k _out (dissipation rate constant) reflecting the progressive decline in NSAA scores as age increases. The model characterizes the full NSAA total score trajectory from birth to death and links the NSAA total score to the probability of LOA via k _out (parameter is a covariate on the hazard of LOA). DMD, Duchenne muscular dystrophy; k _in, production rate constant; k _out, dissipation rate constant; LOA, loss of ambulation; NSAA, North Star Ambulatory Assessment.

FIGURE 2

Participant NSAA total score time profiles. Observed NSAA total scores as a function of age for all participants (left panel; top and bottom) and participants who lost ambulation (right panel; top and bottom). Profiles stratified by study and colored with respect to lower limit of quantification (blue [above]; red [below]). BLQ, below the limit of quantification; NHS, natural history studies; NSAA, North Star Ambulatory Assessment; OBS, observed.

FIGURE 3

Final joint NSAA‐LOA model TTE‐VPC stratified by study. VPC stratified by study (B5161002, domagrozumab clinical trial; B5161011 CINRG NHS). Kaplan–Meier curve derived from observed data (solid black line) and 95% confidence interval obtained from model simulations (blue shaded region) or observed data (gray shaded region). Vertical black lines represent censored data. LOA, loss of ambulation; NHS, natural history studies; NSAA, North Star Ambulatory Assessment; TTE, time‐to‐event; VPC, visual predictive check.

FIGURE 4

Final joint NSAA‐LOA model NSAA‐VPC stratified by study. VPC for NSAA portion of final model stratified by study (B5161002, domagrozumab clinical trial; B5161011 CINRG NHS). Blue circles represent the observed data and the red lines represent the median (solid red), 2.5th percentile (dashed red), and 97.5th percentile (dashed red) of the observed data. For 1000 simulations, the black lines represent the median (solid black), 2.5th percentile (dashed black), and 97.5th percentile (dashed black) of the simulated data. The shaded regions represent the 95% confidence intervals for the simulated median (shaded pink) and each of the simulated 2.5th and 97.5th percentiles (shaded blue). LOA, loss of ambulation; NHS, natural history studies; NSAA, North Star Ambulatory Assessment; VPC, visual predictive check.

FIGURE 5

Simulated and observed NSAA total score time profile over 1 year. Observed NSAA total scores overlaid onto simulated 95% confidence interval with median trajectory. Profiles stratified by baseline NSAA score ranges (16–30) and colored by baseline age (orange, 4 years old; green, 5 years old; blue, 6 years old; and purple 7 years old). Colored arrows represent 1‐year trajectory for each baseline age. NSAA, North Star Ambulatory Assessment.

FIGURE 6

Power curves for clinical trial simulation scenarios. Percent power for various clinical trial sizes assuming treatment effect (1) disease progression is halted at the start of the trial (red) and (2) disease progression is halted, and some function is restored (blue).