Design Considerations for a Phase II Platform Trial in Major Depressive Disorder

Abstract

Major depressive disorder (MDD) is one of the leading causes of disability globally. Despite its prevalence, approximately one-third of patients do not benefit sufficiently from available treatments, and few new drugs have been developed recently. Consequently, more efficient methods are needed to evaluate a broader range of treatment options quickly. Platform trials offer a promising solution, as they allow for the assessment of multiple investigational treatments simultaneously by sharing control groups and by reducing both trial activation and patient recruitment times. The objective of this simulation study was to support the design and optimisation of a phase II superiority platform trial for MDD, considering the disease-specific characteristics. In particular, we assessed the efficiency of platform trials compared to traditional two-arm trials by investigating key design elements, including allocation and randomisation strategies, as well as per-treatment arm sample sizes and interim futility analyses. Through extensive simulations, we refined these design components and evaluated their impact on trial performance. The results demonstrated that platform trials not only enhance efficiency but also achieve higher statistical power in evaluating individual treatments compared to conventional trials. The efficiency of platform trials is particularly prominent when interim futility analyses are performed to eliminate treatments that have either no or a negligible treatment effect early. Overall, this work provides valuable insights into the design of platform trials in the superiority setting and underscores their potential to accelerate therapy development in MDD and other therapeutic areas, providing a flexible and powerful alternative to traditional trial designs.

Keywords: allocation; clinical trial simulations; futility stopping; major depressive disorder; multiple treatment arms; platform trial.

FIGURE 1

TRD platform trial design proposal. If patients are eligible for the platform, they are randomised to either control or one of the treatment arms within the platform. Different drugs can enter the trial at different time points. The design might also allow for interim analyses to drop treatment arms early. The time point of the interim analyses is indicated by a dotted line.

FIGURE 2

Experimental treatment arms per 1000 patients (without control arm) during the runtime of the platform trial, and rejection rates for platforms running at expected capacity, meaning three treatment arms at the beginning of the platform trial and a probability of 20% per month for a new treatment arm to enter. The targeted sample size per experimental treatment arm was fixed at $n_j=80$, and all effect sizes were assumed to be equally likely, i.e., ${\theta }_0={\theta }_{0.2}={\theta }_{0.35}={\theta }_{0.5}=0.25$. The circles give the values without the implementation of an interim analysis, and the triangles the corresponding values when a futility boundary of 0.5 for the p‐value is applied. The MAPC for the ‘$\sqrt{k}$ allocation’ was set to 35%. (A) depicts the mean number of arms that can be evaluated per 1000 patients in a corresponding platform trial. In (B), the percentage of rejected null hypotheses is depicted stratified by the four different investigated effect sizes. It equals the type I error rate for $d=0$ and the power for the other values of $d$. The type I error rate is always controlled at 5%. This value is indicated by the lower dotted line. The higher dotted line highlights the 80% mark.

FIGURE 3

Experimental treatment arms per 1000 patients (without control arm) during the runtime of the platform trial, and rejection rates when implementing different futility rules. On the x‐axis the futility boundary on the p‐value scale is shown, i.e., we stop for futility if the corresponding p‐value is larger than the futility boundary. No stop means that no futility analysis is performed. In the scenario with the filled squares all effect sizes are assumed to be equally likely, i.e., ${\theta }_0={\theta }_{0.2}={\theta }_{0.35}={\theta }_{0.5}=0.25$. With the empty squares we present results for a more pessimistic scenario with probabilities of assignment ${\theta }_0=0.5$, ${\theta }_{0.2}=0.3$, ${\theta }_{0.35}=0.1$, and ${\theta }_{0.5}=0.1$. (A) gives the mean number of experimental treatment arms per 1000 patients in the platform trial. All experimental treatment arms are included in this number regardless of the underlying effect size. In (B), the percentage of rejected null hypotheses is depicted stratified by the four different investigated effect sizes. It equals the type I error rate for $d=0$ and the power for the other values of $d$. The type I error rate is always controlled at 5%. This value is indicated by the lower dotted line. The higher dotted line highlights the 80% mark.

FIGURE 4

Experimental treatment arms per 1000 patients (without control arm) during the runtime of the platform trial, and rejection rates for different per‐arm sample sizes. All effect sizes are assumed to be equally likely, i.e., ${\theta }_0={\theta }_{0.2}={\theta }_{0.35}={\theta }_{0.5}=0.25$. The circles give the values without the implementation of an interim analysis and the triangles the corresponding values when a futility boundary of 0.5 for the p‐value is applied. The sample size depicted on the x‐axis was examined in steps of 10. The small variation in the x‐direction is based on jittering for better readability. (A) shows the mean number of experimental treatment arms per 1000 patients. In (B), the percentage of rejected null hypotheses is depicted stratified by the four different investigated effect sizes. The rejection rate equals the type I error rate for $d=0$ and the power for the other values of $d$. The type I error rate is always controlled at 5%. This value is indicated by the lower dotted line. The higher dotted line highlights the 80% mark.

FIGURE 5

Comparison of operating characteristics in different trial types. All effect sizes are assumed to be equally likely, i.e., ${\theta }_0={\theta }_{0.2}={\theta }_{0.35}={\theta }_{0.5}=0.25$. The circles give the values without the implementation of an interim analysis and the triangles the corresponding values when a futility boundary of 0.5 for the p‐value is applied. The sample size depicted on the x‐axis was examined in steps of 10. The small variation in the x‐direction is based on jittering for better readability. (A) shows the mean number of experimental treatment arms per 1000 patients for the three different trial types: Platform trial with maximum capacity utilisation, platform trial with expected load in the MDD case, and the traditional approach with a series of individual two‐arm randomised controlled trials. (B) gives the power for the same type of trials. The dotted lines indicate the 80% and 90% marks.