An Adaptive Phase II Trial Design for Dose Selection and Addition in Microfilarial Infections

TL;DR

This paper presents a frequentist adaptive two-stage, basket trial framework for dose selection across four helminth infections using oxfendazole. It extends the partial conditional error approach to allow adding a high-dose arm at interim while borrowing safety information across diseases and testing efficacy separately within each disease. The design employs surrogate endpoints for interim decisions and handles a zero-inflated, mixture primary endpoint through a combination of elementary and intersection hypotheses with closed testing. Simulation studies demonstrate potential gains in dose identification efficiency and reductions in total sample size and trial duration, with recommendations for practical implementation and software support. Overall, the approach offers a robust, efficient path to adaptive dose selection in multi-disease infectious disease trials with complex outcomes.

Abstract

We propose a frequentist adaptive phase 2 trial design to evaluate the safety and efficacy of three treatment regimens (doses) compared to placebo for four types of helminth (worm) infections. This trial will be carried out in four Subsaharan African countries from spring 2025. Since the safety of the highest dose is not yet established, the study begins with the two lower doses and placebo. Based on safety and early efficacy results from an interim analysis, a decision will be made to either continue with the two lower doses or drop one or both and introduce the highest dose instead. This design borrows information across baskets for safety assessment, while efficacy is assessed separately for each basket. The proposed adaptive design addresses several key challenges: (1) The trial must begin with only the two lower doses because reassuring safety data from these doses is required before escalating to a higher dose. (2) Due to the expected speed of recruitment, adaptation decisions must rely on an earlier, surrogate endpoint. (3) The primary outcome is a count variable that follows a mixture distribution with an atom at 0. To control the familywise error rate in the strong sense when comparing multiple doses to the control in the adaptive design, we extend the partial conditional error approach to accommodate the inclusion of new hypotheses after the interim analysis. In a comprehensive simulation study we evaluate various design options and analysis strategies, assessing the robustness of the design under different design assumptions and parameter values. We identify scenarios where the adaptive design improves the trial's ability to identify an optimal dose. Adaptive dose selection enables resource allocation to the most promising treatment arms, increasing the likelihood of selecting the optimal dose while reducing the required overall sample size and trial duration.

Figures (51)

• Figure 1: Scheme of the trial design. The trial starts in stage 1 with placebo and medium and low doses, represented by colored bars. The trial continues in stage 2 with further investigation on the selected arms which are chosen based on interim results. The active doses that can be selected for stage 2 (low, medium and high doses) are depicted with bars with a striped pattern. The placebo group will always be continued to the second stage. Which of the active arms are selected for the second stage, depends on the efficacy and safety results in the first stage and maybe also informed by safety results obtained in the other diseases.
• Figure 2: Mansonellosis S1: Simulated power values for the low, medium, and high doses and the disjunctive power are depicted (in the columns) for simulation S1 (standard reduction rates) as a function of $\alpha_1=\{0.1,0.2,0.3,0.4,0.5\}$ for several scenarios of reduction rates (in the rows).
• Figure 3: Mansonellosis S3: Simulated power values for the low, medium, and high doses and the disjunctive power are depicted (in the columns) for simulation S3 (modified baseline parameters) as a function of $\alpha_1=\{0.1,0.2,0.3,0.4,0.5\}$ for several scenarios of reduction rates (in the rows).
• Figure 4: Mansonellosis S5: Simulated power values for the low, medium, and high doses and the disjunctive power are depicted (in the columns) for simulation S1 as a function of $N_1=\{80,100,120\}$ for several scenarios of reduction rates (in the rows).
• Figure B.5: Diagram of the hypotheses tested in each scenario depending on the selection at the interim. The diagram on the top-left shows the full closure of hypotheses. Hypotheses in grey nodes refer to those tested in the final analysis; dashed nodes refer to hypotheses which are accepted at the interim analysis. Arrows illustrate the intersection hypotheses contained in the subsequent hypotheses.