Shuyen Ho

Design of Dose-Ranging Trials

Typically, Phase I clinical trials are designed to evaluate escalating doses in order to find the maximum tolerable dose (MTD) so as to establish the upper anchor of doses for future development of new medicinal products. After Phase I, one of the most important objectives for Phase II clinical development is to answer the Go/NoGo question—proof of concept (PoC). For new drugs treating non-life-threatening diseases, the next important objective is to establish the “dose range.” A range of doses can be identified by the high dose and the low dose, and the high dose is limited by MTD. As such, a key Phase II objective is to find a low dose of this dose range. Usually this low dose is considered as minimum effective dose (MinED). Simply put, Phase I is moving doses upward to find MTD; Phase II is moving doses downward to find MinED.

Data Analysis for Dose-Ranging Trials with Continuous Outcome

Traditionally, proof of concept (PoC) study and dose-ranging trials are conducted separately in a sequential fashion. A two-group PoC design involves a comparison between MTD of test drug and a placebo control, a comparison that is conducted first. Only after the concept is proven, then dose-ranging trials are designed to study the dose-response relationship.

Confirmation and Exploration

It should be noted that contributions of statistical science to modern medicine are not limited to statistical methods for estimating sample sizes, calculating p-values or confidence intervals.

Design a Proof of Concept Trial

In typical clinical development programs, Phase II is the first time for the candidate product (compound or biologic) to be tested in patients with the target disease under study. For medicinal products discovered and developed to treat chronic diseases, most of the Phase I clinical trials recruit normal healthy volunteers and, as a result, the disease improvement cannot be observed in Phase I. Under this situation, therefore, study results cannot be used to assess product efficacy. In Phase I oncology trials, although cancer patients are recruited to help study the maximal tolerable dose (MTD), these patients may or may not be with the tumor type for which the compound is developed for. Also, the sample sizes used for Phase I oncology trials are not large enough for study of product efficacy. Therefore, in most of product development programs, Phase II is the first time the product efficacy can be observed and evaluated. In other words, before the first Phase II clinical trial, the sponsor does not have an opportunity to study whether or not the test product, or compound, works in human patients.

Analysis of a Proof of Concept Study

From the data analysis perspective, clinical statisticians are typically faced with continuous data, discrete data, or time-to-event data. For early Phase II clinical trials, the most frequently seen data types are continuous data or binary data. In this chapter, and the next two chapters covering data analysis, the discussion will mainly be about analyzing continuous data or binary data.

Concept of Alpha

During a tea party in the 1920s in Cambridge, United Kingdom, the waiter presented cups, a pot of tea and milk in a tray to the party guests.

Data Analysis of Dose-Ranging Trials for Binary Outcomes

Continuing the data analysis for dose-ranging clinical trials from Chap. 9 for continuous outcomes, we illustrate the analysis for binary data in this chapter. We will use the same data from Chap. 9, but dichotomize the continuous response data into binary data simply for the pedagogical purpose of continuing from the previous chapter and illustrating the methods on analyzing dose-ranging clinical trial with binary data.

Combining Proof of Concept and Dose-Ranging Trials

Chapters 4 and 5 discuss proof of concept (PoC) and dose-ranging studies respectively and separately.

Overview of Phase III Clinical Trials

After the success of Phase II development, it comes to the Phase III development. Phase III has fundamental objectives in confirming clinical efficacy and investigating no unacceptable safety concerns. Clinical efficacy generally includes effects that are measurable as reflected by clinical outcome assessments such as patient-reported and clinician-reported outcomes and objective measures like survival or lab data.