There is considerable interest in methods that use accumulated data to modify trial sample size. However, sample size re-estimation in group sequential designs has been controversial. We describe a method for sample size re-estimation at the penultimate stage of a group sequential design that achieves specified power against an alternative hypothesis corresponding to the current point estimate of the treatment effect.
Background
New therapies are urgently needed for Alzheimer’s disease (AD). Sodium oligomannate (GV-971) is a marine-derived oligosaccharide with a novel proposed mechanism of action. The first phase 3 clinical trial of GV-971 has been completed in China.
Methods
We conducted a phase 3, double-blind, placebo-controlled trial in participants with mild-to-moderate AD to assess GV-971 efficacy and safety. Participants were randomized to placebo or GV-971 (900 mg) for 36 weeks. The primary outcome was the drug-placebo difference in change from baseline on the 12-item cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS-cog12). Secondary endpoints were drug-placebo differences on the Clinician’s Interview-Based Impression of Change with caregiver input (CIBIC+), Alzheimer’s Disease Cooperative Study-Activities of Daily Living (ADCS-ADL) scale, and Neuropsychiatric Inventory (NPI). Safety and tolerability were monitored.
Results
A total of 818 participants were randomized: 408 to GV-971 and 410 to placebo. A significant drug-placebo difference on the ADAS-Cog12 favoring GV-971 was present at each measurement time point, measurable at the week 4 visit and continuing throughout the trial. The difference between the groups in change from baseline was − 2.15 points (95% confidence interval, − 3.07 to − 1.23; p < 0.0001; effect size 0.531) after 36 weeks of treatment. Treatment-emergent adverse event incidence was comparable between active treatment and placebo (73.9%, 75.4%). Two deaths determined to be unrelated to drug effects occurred in the GV-971 group.
Conclusions
GV-971 demonstrated significant efficacy in improving cognition with sustained improvement across all observation periods of a 36-week trial. GV-971 was safe and well-tolerated.
Trial registration
ClinicalTrials.gov, NCT02293915. Registered on November 19, 2014
Summary. Two-arm group sequential designs have been widely used for over 40 years, especially for studies with mortality endpoints. The natural generalization of such designs to trials with multiple treatment arms and a common control (MAMS designs) has, however, been implemented rarely. While the statistical methodology for this extension is clear, the main limitation has been an efficient way to perform the computations. Past efforts were hampered by algorithms that were computationally explosive. With the increasing interest in adaptive designs, platform designs, and other innovative designs that involve multiple comparisons over multiple stages, the importance of MAMS designs is growing rapidly. This article provides break-through algorithms that can compute MAMS boundaries rapidly thereby making such designs practical. For designs with efficacy-only boundaries the computational effort increases linearly with number of arms and number of stages. For designs with both efficacy and futility boundaries the computational effort doubles with successive increases in number of stages.
Methods for controlling the type-1 error of an adaptive group sequential trial were developed in seminal papers by Cui, Hung, and Wang (Biometrics, 1999), Lehmacher and Wassmer (Biometrics, 1999), and Müller and Schäfer (Biometrics, 2001). However, corresponding solutions for the equally important and related problem of parameter estimation at the end of the adaptive trial have not been completely satisfactory. In this paper, a method is provided for computing a two-sided confidence interval having exact coverage, along with a point estimate that is median unbiased for the primary efficacy parameter in a two-arm adaptive group sequential design. The possible adaptations are not only confined to sample size alterations but also include data-dependent changes in the number and spacing of interim looks and changes in the error spending function. The procedure is based on mapping the final test statistic obtained in the modified trial into a corresponding backward image in the original trial. This is an advance on previously available methods, which either produced conservative coverage and no point estimates or provided exact coverage for one-sided intervals only.
A method is proposed for modifying a group-sequential clinical trial by restricting future enrollment to a subgroup and possibly altering the sample size of the subgroup, based on an interim analysis of the data already obtained. The method provides strong control of type 1 error without requiring prespecification of the list of possible subgroups or of the decision rule for selecting among them. Nevertheless, for regulatory submissions it is recommended that the subgroups and decision rule be prespecified. The method is applied to a large cardiology trial in which the subgroups are prespecified and the decision rules for subgroup selection and sample size alteration are based on conditional power. It is shown by simulation that substantial gains in power can be attained if there is a subgroup by treatment interaction.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.