Pharmaceutical Statistics最新文献

Pearson's asymptotic χ² test is often used to compare binary data between two groups. However, when the sample sizes or expected frequencies are small, the test is usually replaced by Fisher's exact test. Several alternative rules of thumb exist for defining "small" in this context. Replacing one test with another based on the obtained data is unusual in statistical practice. Moreover, this commonly-used switch is unnecessary because Pearson's χ² test can easily be carried out as an exact test for any sample sizes. Therefore, we recommend routinely using an exact test regardless of the obtained data. This change of approach allows prespecifying a particular test and a much less ambiguous and more reliable analysis.

Administration of therapeutic protein products might potentially elicit an immune response via production of Anti-Drug Antibodies (ADA). This immune response can cause some clinical consequences ranging from mild to harmful for the patient, affecting the safety and efficacy of the drug. Therefore, assessment of Immunogenicity and the ability to follow possible associations between ADA assay measurements and clinical events is one of the key parts of clinical safety evaluation in both clinical and preclinical areas. In order to assess the immunogenicity of biological drug molecules, it is important to develop and validate reliable laboratory methods and evaluate various performance characteristics during development and validation phases. Determination of the screening assay cut-point and establishment of the confirmatory assay cut-point are fundamental aspects of ADA assay validation. Existing regulatory guidance documents addressing immunogenicity topics (immunoassays) cover the development and validation of reliable laboratory methods, but there is a need for more comprehensive discussions on statistical evaluation methods. While there is literature available on statistical methods for cut-point estimation, this tutorial aims to provide additional statistical considerations specifically tailored for ADA assay development and validation cut-points. Furthermore, practical R code snippets are provided to facilitate the implementation of the key evaluation steps. This resource aims to enhance the rigor and reliability of ADA assay validation cut-point evaluation, ultimately contributing to more robust immunogenicity assessments in clinical studies.

Recently, model-assisted designs, including a Bayesian optimal interval (BOIN) design with optimal thresholds for determining the dose for the next cohort, have been proposed for Phase I cancer studies. Model-assisted designs are useful owing to their good performance in addition to their algorithm-based simplicity. In this era of precision medicine, drug combinations are widely used to enhance treatment efficacy and overcome resistance to monotherapies. However, identification of maximum tolerated dose (MTD) combinations is complicated because the joint toxicity order of paired doses is only partially known. BOIN and Keyboard combination designs are the only model-assisted designs developed to date. Further, both these combination designs show similar operational characteristics. Despite the simplicity and superior performance of model-assisted designs, they have not been sufficiently studied in Phase I drug combination trials. In this study, to provide a new design with simplicity and superior performance compared to model-assisted designs for dose-combination cancer Phase I studies, we extend the cumulative sum interval design (CUSUMIN) developed for single-agent dose-finding design based on statistical quality control methodology, which improves on BOIN and other representative model-assisted designs in terms of controlling overdosing rates while maintaining similar performance in determining the MTD. CUSUMIN can be expected to provide a safer assignment than that of BOIN in drug combination dose-finding studies while maintaining MTD selection performance, as shown in the single-agent dose-finding settings.

The term "minimal clinically important difference" (MCID), though defined as the smallest change in an outcome that is meaningful to the patient, is often used to interpret differences between treatment groups. It is in this context that the limitations of MCID are discussed, which include: the omission of the role of time in its definition for progressive diseases; the unsuitability of adopting MCID derived from open-label studies for randomized, placebo-controlled, blinded studies; the unreliability of MCID in rare disease trials; challenges in interpretation when placebo patients also achieve MCID; the failure to account for how differences in patient populations affect MCID (e.g., inclusion or exclusion of patients on prior treatment); not recognizing the connection between the true treatment effect, MCID and power; lack of consideration of differences in analysis methods (e.g., the extent of missing data and how it is handled); and the limitations of an MCID-based responder analysis. Therefore, the recommendation made is to prospectively define a customized MCID that addresses each deficit. If the deficits cannot be adequately resolved, then the recommendation is that trial results should be interpreted without reference to MCID.

It is critical to evaluate the sensitivity of conclusions from a clinical trial to potential violations in the missing data assumptions of the statistical analysis. Sensitivity analyses should not consist of a few methods that might have been reasonable alternatives to the chosen analysis method, nor should they explore only a limited space of violations in the assumptions of the analysis. Instead, sensitivity analyses should target the same estimand as that targeted in the main analysis, and they should systematically and comprehensively explore the space of possible assumptions to evaluate whether the key conclusions hold up under all plausible scenarios. In a randomized, controlled trial, this can be achieved by tipping point analyses that vary assumptions about missing outcomes on the experimental and control arms to identify and discuss the plausibility of scenarios under which there is no longer evidence of a treatment effect. We introduce a simple, novel tipping point approach in which, for a variable that is quantitative or can be analyzed as if it is quantitative, inference on the treatment effect is based on the observed data and two sensitivity parameters, with minimal assumptions and no need for imputation. The sensitivity parameters to be varied are the mean differences between outcomes in dropouts and outcomes in completers on each of the two treatment arms. We derive the asymptotic properties of the proposed statistic and illustrate the utility of such an approach with two examples of drug reviews in which the methodology was utilized to inform regulatory decision-making.

Patient enrollment can be a substantial burden in rare disease trials. One potential approach is to incorporate external control (EC) into concurrent randomized trials, or EC borrowing, to reduce such burden. Extensive research has been conducted to explore statistical methodologies. As in all designs, type I error control is essential. Conditional type I error rate has been used in the literature as the de facto metrics for type I error rate. However, research has shown that controlling the conditional type I error rate at the alpha level will disallow EC borrowing. Therefore, EC borrowing is practically at an impasse. Kopp-Schneider et al. concluded that a more appropriate metrics for type I error is necessary. We show that a trial with EC borrowing can be considered as a two-stage adaptive design. With this perspective, we propose to define type I error as the weighted averages of conditional type I error rate in trials with EC borrowing. Dynamic borrowing methods for controlling type I error are proposed.

Randomization is a statistical procedure used to allocate study subjects randomly into experimental groups while balancing continuous variables. This paper presents an alternative to random allocation for creating homogeneous groups by balancing experimental factors. The proposed algorithms, inspired by the Theory of Evolution, enhance the benefits of randomization through partitioning. The methodology employs a genetic algorithm that minimizes the Irini criterion to partition datasets into balanced subgroups. The algorithm's performance is evaluated through simulations and dataset examples, comparing it to random allocation via exhaustive search. Results indicate that the experimental groups created by Irini are more homogeneous than those generated by exhaustive search. Furthermore, the Irini algorithm is computationally more efficient, outperforming exhaustive search by more than three orders of magnitude.

One of the main challenges in drug development for rare diseases is selecting the appropriate primary endpoints for pivotal studies. Although many endpoints can effectively reflect clinical benefit, their sensitivity often varies, making it difficult to determine the required sample size for study design and to interpret final results, which may be underpowered for some or all endpoints. This complexity is further compounded when there is a desire to support regulatory claims for multiple clinical endpoints and dose regimens due to the issues of multiplicity and sample size constraints. Joint Primary Endpoints (JPEs) offer a compelling strategy to address these challenges; however, their analysis in conjunction with component endpoints presents additional complexities, particularly in managing multiplicity concerns for regulatory claims. To address these issues, this paper introduces a robust two-stage gatekeeping framework designed to test two hierarchically ordered families of hypotheses. A novel truncated closed testing procedure is employed in the first stage, enhancing flexibility and adaptability in the evaluation of primary endpoints. This approach strategically propagates a controlled fraction of the error rate to the second stage for assessing secondary endpoints, ensuring rigorous control of the global family-wise Type I error rate across both stages. Through extensive numerical simulations and real-world clinical trial applications, we demonstrate the efficiency, adaptability, and practical utility of this approach in advancing drug development for rare diseases while meeting stringent regulatory requirements.

The estimand framework, introduced in the ICH E9 (R1) Addendum, provides a structured approach for defining precise research questions in randomised clinical trials. It suggests five strategies for addressing intercurrent events (ICE). This case study examines the principal stratum strategy, highlighting its potential for estimating causal treatment effects in specific subpopulations and the challenges involved. The occurrence of anti-drug antibodies (ADAs) and their potential clinical impact are important factors in evaluating biosimilars. Typically, analyses focus on subgroups of patients who develop ADAs during the study. However, conducting subgroup analyses based on post-randomisation variables, such as immunogenicity, can introduce substantial bias into treatment effect estimates and is therefore methodologically not optimal. The principal stratum strategy provides a statistical pathway for estimating treatment effects in subpopulations that cannot be anticipated at baseline. By leveraging counterfactuals to assess treatment outcomes, with and without the incidence of intercurrent events (ICEs), this approach can be implemented through a missing data perspective. We demonstrate the implementation of the principal stratum strategy in a phase 3 equivalence trial of a biosimilar for the treatment of rheumatoid arthritis. Using a multiple imputation approach, we leverage longitudinal measurements to create analysis datasets for subpopulations who develop ADAs as ICE. Our results highlight the principal stratum strategy's potential and challenges, emphasising its reliance on unobserved ICE states and the need for complex and rigorous modelling. This study contributes to a nuanced understanding and practical implementation of the principal stratum strategy within the ICH E9 (R1) framework.

Analyzing and effectively communicating the efficacy and toxicity of treatment is the fundamental basis of risk-benefit analysis (RBA). There is a need for more efficient and objective tools. We apply Chauhan Weighted Trajectory Analysis (CWTA) to perform RBA with superior objectivity, power, and ease of communication. We used CWTA to perform 1000-fold simulations of RCTs using ordinal endpoints that captured both treatment efficacy and treatment toxicity. RCTs were stochastically generated with 1:1 allocation at defined sample sizes and hazard ratios. We first studied the simplest case simulation of 3 levels each of toxicity and efficacy (a 3 × 3 matrix). We then simulated the general case of the advanced cancer trial, with efficacy graded by five RECIST 1.1 health statuses and toxicity graded by the six-point CTCAE scale to create a 6 × 5 matrix. Finally, the 6 × 5 matrix model was applied to a real-world dose escalation phase I trial in advanced cancer. Simulations in both the 3 × 3 simplest case matrix and the 6 × 5 advanced cancer matrix confirmed our hypothesis that drugs with both superior efficacy and toxicity profiles synergize for greater statistical power with CWTA RBA than either signal alone. The CWTA RBA 6 × 5 matrix meaningfully reduced sample size requirements over CWTA efficacy-only analysis. Despite a small sample size, application of the matrix to each of the seven cohorts of the dose finding phase I clinical trial provided objective and statistically significant validation for the dose subjectively selected by the trialists. CWTA RBA, by incorporating both drug efficacy and the trajectory of drug toxicity, provides a single test statistic and summary plot that analyzes, visualizes, and effectively communicates the risk-benefit assessment of a clinical trial. CWTA RBA requires fewer patients than CWTA efficacy-only analysis when the experimental drug is both more effective and less toxic. Our results show CWTA RBA has the potential to aid the objective and efficient assessment of new therapies throughout the drug development pathway. Furthermore, its distinct advantages over competing tests in visualizing and communicating risk-benefit will assist regulatory review, clinical adoption, and understanding of therapeutic risks and benefits by clinicians and patients alike.