Purchase this article with an account.
Paul Andrew Yates, Kristina Holbrook, Elias Reichel, Nadia K Waheed, James Patrie, TOGA Investigators; Designing a Clinical Study to Evaluate Potential Therapeutics for Geographic Atrophy Secondary to Non-Exudative Age-Related Macular Degeneration. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):2835. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
TOGA is a phase IIb/III trial (NCT01792989) randomized, double-masked, placebo-controlled study evaluating the efficacy of daily oral administration of the tetracycline derivative ORACEA® compared with placebo on the rate of change in area of geographic atrophy (GA) as measured by change in area of GA for patients with dry AMD. A rational statistical design, using data from prior natural history and interventional studies, was needed to ensure appropriate enrollment to meet study end points and optimize detection of a difference between Oracea and placebo for primary and secondary GA measures.
Study design was determined using demographic and outcome data from the Geographic Atrophy Natural History Study and the Age-Related Eye Disease Study, specifically considering initial lesion size, rate of GA enlargement based on lesion size, and rates of development of exudative AMD. Monte-Carlo simulations were used to generate a range of expected outcomes to determine study power calculations.
Monte Carlo simulation was used to generate a range of expected outcomes between placebo and Oracea treatment groups, powered to detect a 30% difference in mean yearly rate of change in area of GA between the ORACEA® and placebo groups, as a function of sample size and the bivariate correlation between the yearly rate of change in the area of GA and baseline area of GA. Under these assumptions simulations found a 1.13 mm2 mean difference in area of GA at 24 months (slope of 1.89mm2/year placebo, 1.32 mm2/year Oracea). Primary analysis will be determined by random coefficient regression and comparison of marginal slopes adjusted to a common observation phase rate in the change in area of GA. A permuted-block randomization scheme will be used in assigning patients to Oracea versus placebo.
Monte Carlo simulations demonstrate 286 subjects (143 assigned to placebo and 143 to Oracea) are required to have an 80% likelihood of detecting a 30% difference in the mean yearly rate of change in the area of GA between the placebo and Oracea treatment group. Finding a non-zero bivariate correlation coefficient requires fewer subjects and could be generated by analysis of initial lesion size during treatment period compared to rate of growth during treatment, or by comparison of growth during the observation versus treatment phase.
This PDF is available to Subscribers Only