Statistics in Biopharmaceutical Research Best Papers Award 2023

We are pleased to announce the recipients of the 2023 Best Paper Award for the articles published in Statistics in Biopharmaceutical Research (SBR). The following five articles were selected from those published in the 2021 and 2022 issues. These articles exhibit excellent examples of current statistical advancements in biopharmaceutical research. In selecting the winners, the editors reflected SBRs goal of publishing articles that focus on the development of novel statistical methods, advanced applications of existing methods, or innovative applications of statistical principles that can be used by statistical practitioners in these disciplines. They also considered factors such as innovation, importance to the research community, impact, and clarity of presentation.

• Guizzaro, L., Pétavy, F., Ristl, R., and Gallo, C. (2021), “The Use of a Variable Representing Compliance Improves Accuracy of Estimation of the Effect of Treatment Allocation Regardless of Discontinuation in Trials with Incomplete Follow-up,” Statistics in Biopharmaceutical Research, 13,119–127.

• Hua, E., Janocha, R., Severin, T., Wei, J., and Vandemeulebroecke, M. (2022), “A Phase 3 Trial Analysis Proposal for Mitigating the Impact of the COVID-19 Pandemic,” Statistics in Biopharmaceutical Research, 14, 80–86.

• Zhan, T., Hartford, A., Kang, J., and Offen, W. (2022), “Optimizing Graphical Procedures for Multiplicity Control in a Confirmatory Clinical Trial via Deep Learning,” Statistics in Biopharmaceutical Research, 14, 92–102.

• Lui, G.F., Liu, J., Chen, F., Gutman, R., and, Lu, K. (2022), “Bayesian Approaches for Handling Hypothetical Estimands in Longitudinal Clinical Trials with Gaussian Outcomes,” Statistics in Biopharmaceutical Research, 14, 626–635.

• Xu, Y., Wu, M., He, W., Liao, Q., and Mai, Y. (2022), “Teasing Out the Overall Survival Benefit With Adjustment for Treatment Switching to Multiple Treatments,” Statistics in Biopharmaceutical Research, 14, 592–601.

In clinical trials, not all randomized patients adhere to the treatment to which they are assigned. The potential impact of such nonadherence has been increasingly recognized, especially since the ICH E9 Addendum (ICH 2019) redefined estimation objectives (“estimands”). Guizzaro et al. (2021) evaluated the performance of different estimation techniques in trials with incomplete follow-up after discontinuation when a treatment-policy strategy is implemented. Through simulated trials in major depressive disorder with varying real treatment effects, proportions of patients dropping out, and incomplete follow-up, they modeled and visualized as a directed acyclic diagram a reasonable data-generating model, and investigated which set of variables allows identification and estimation of such an effect. They provided several guidance on the estimation of an estimand that employs a treatment-policy strategy for treatment discontinuation. On the other hand, Lui et al. (2022) discussed Bayesian methods to implement the analysis for the estimands under the hypothetical strategy and missing data handling in longitudinal clinical trials. They provided analyses of three clinical trials to illustrate the applications of the methods with discussion of their advantages and disadvantages.

The COVID-19 pandemic posed unique challenges in medical research. In particular, the disruptions caused by the pandemic wreaked havoc on ongoing clinical trials. Hua et al. (2022) described some of these challenges and mitigation strategies observed in two identical, ongoing chronic dermatology disease clinical trials in during the pandemic. They shared their ideas and experience in devising a risk mitigation proposal for the statistical methodology of the trials during the COVID-19 pandemic, where the originally-predefined graph-based multiple testing procedures (graphical procedures) for individual trials was modified to pool the data from the two trials, but to maintain the power of key endpoints and comparisons while still controlling the relevant Type I error probability.

Graphical procedures (Bretz et al. 2009; Maurer and Bretz 2013) provides a very convenient tool for designing, visualizing, and performing multiplicity adjustments to control the Type I error probability associated with multiple hypotheses in confirmatory clinical trials. The graphical procedures need to reflect the relative importance, contextual relationships, and logical restrictions of individual hypotheses. Zhan et al. (2022) discussed a feedforward neural network-based optimization framework for the graphical procedures. Their framework takes advantage of the strong functional representation of deep neural networks and further uses constraint optimization techniques to locate the solution. They provided a case study to illustrate how to optimize a graphical procedure with respect to a specific study objective.

In cancer clinical trials, characterizing long-term overall survival (OS) benefit of an experimental intervention group is often unobservable if some patients in the control intervention group switch to the experimental intervention and/or other cancer treatments after disease progression. A key question is how to estimate the true OS benefit of the experimental treatment that would have been estimated in the absence of treatment switching. To address the question, Xu et al. (2022) discussed an extension of the rank-preserving structural failure time model to a multi-level model, and a random-forest-based prediction method. They conducted a simulation study to evaluate the utility of their methods compared to existing methods under different scenarios.

The authors of these five articles will be recognized at the SBR Invited Session “Recent Advances in Statistical Methodology for Medical Product Development: SBR Best Papers 2023,” which will be held at the 2023 Joint Statistical Meetings (JSM2023) in Toronto. Three of the five articles will be presented by the authors at that session. Open access for all five articles is available until JSM2023. We hope that the readers will enjoy reading them. We congratulate the authors for their excellent work and thank all SBR authors whose important, high-quality and far-reaching biopharmaceutical research becomes published in this journal.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.