Statistical Interpretation and Comparison of Waterfall Plots.

Waterfall plots have gained popularity as a visualization tool to present antitumor activity of treatments in oncology, especially for phase I and II trials. The typical waterfall plot in oncology is a bar plot with each bar representing the best percent tumor size reduction from baseline for a patient sorted in descending order along the x-axis. As new therapies are routinely developed in combination with standard of care or other investigational treatments, waterfall plot comparison between combination therapy and monotherapy may facilitate development decisions in addition to overall response rate or duration of response. However, waterfall plots are often assessed heuristically in practice with lack of statistical rigor. In this work, we examine the correspondence between the waterfall plot and the empirical cumulative distribution function. We demonstrate how to derive key summary statistics directly from the waterfall plot. Using real examples from published waterfall plots, we show how comparisons of waterfall plots can elucidate clinically meaningful information, such as treatment effect patterns in progression-free survival and overall survival.

A unified framework for weighted parametric group sequential design.

Group sequential design (GSD) is widely used in clinical trials in which correlated tests of multiple hypotheses are used. Multiple primary objectives resulting in tests with known correlations include evaluating (1) multiple experimental treatment arms, (2) multiple populations, (3) the combination of multiple arms and multiple populations, or (4) any asymptotically multivariate normal tests. In this paper, we focus on the first three of these and extend the framework of the weighted parametric multiple test procedure from fixed designs with a single analysis per objective to a GSD setting where different objectives may be assessed at the same or different times, each in a group sequential fashion. Pragmatic methods for design and analysis of weighted parametric group sequential design under closed testing procedures are proposed to maintain the strong control of the family-wise Type I error rate when correlations between tests are incorporated. This results in the ability to relax testing bounds compared to designs not fully adjusting for known correlations, increasing power, or allowing decreased sample size. We illustrate the proposed methods using clinical trial examples and conduct a simulation study to evaluate the operating characteristics.

Independent action models and prediction of combination treatment effects for response rate, duration of response and tumor size change in oncology drug development.

An unprecedented number of new cancer targets are in development, and most are being developed in combination therapies. Early oncology development is strategically challenged in choosing the best combinations to move forward to late stage development. The most common early endpoints to be assessed in such decision-making include objective response rate, duration of response and tumor size change. In this paper, using independent-drug-action and Bliss-drug-independence concepts as a foundation, we introduce simple models to predict combination therapy efficacy for duration of response and tumor size change. These models complement previous publications using the independent action models (Palmer 2017, Schmidt 2020) to predict progression-free survival and objective response rate and serve as new predictive models to understand drug combinations for early endpoints. The models can be applied to predict the combination treatment effect for early endpoints given monotherapy data, or to estimate the possible effect of one monotherapy in the combination if data are available from the combination therapy and the other monotherapy. Such quantitative work facilitates strategic planning and decision making in early stage oncology drug development.

Testing monotherapy and combination therapy in one trial with biomarker consideration.

It is a common scenario that an experimental oncology therapy, as a monotherapy, may be more effective than standard of care (SOC) in a biomarker positive population but less so or even inferior to SOC in biomarker negative population. At the same time, due to synergistic or additive effect, the combination of the two may be more effective than SOC alone in the all-comer population. The conventional development paradigm is to conduct two separate Phase III trials, one with the monotherapy versus SOC in the biomarker positive population, and the other with the combination therapy versus SOC in the all-comer population. In this manuscript, we propose a one-trial design that stratifies by biomarker status and randomizes biomarker positive patients into three arms (combination therapy, monotherapy, and SOC) and biomarker negative patients into two arms (combination therapy and SOC). There are two hypotheses in the proposed design and each addresses a different question. The family-wise type-I error rate (FWER) is smaller, due to shared control, than that of two separate trials. Therefore, no FWER adjustment is necessary in the proposed design and each hypothesis can be tested at the conventional 2.5% (one-sided) alpha level. The population for comparison between the combination therapy and SOC is skewed in the proposed design. A two-step log-rank statistic is proposed to account for the skewness. Power and sample size of the proposed design are evaluated in comparison with the two-trial paradigm. The proposed design is more efficient.

Optimal two-stage randomized multinomial designs for Phase II oncology trials.

A new two-stage design is proposed that is suitable for early detection of the anticancer activity of experimental therapies in Phase II oncology trials. The endpoints of interest are response rate and early progression rate. The anticancer activity is defined by a positive signal in one endpoint and a non-negative signal in the other endpoint. The two endpoints are modeled by the multinomial distribution. The design is optimal in that it minimizes the patient exposure when the experimental therapies are inactive. The design parameters are found by a grid searching algorithm under type I and type II error rate constraints. Examples of the design are also presented in this paper.