John Lawrence

Strategies for changing the test statistic during a clinical trial.

This article discusses the design of a clinical trial where a new treatment will be compared to a control. For a specific type of endpoint, there are a wide variety of test statistics that can be used. Also, the investigator must decide how many patients to accrue in each arm as well as the duration of the study. After an interim look at the data, the investigator may decide that a different test statistic would be more powerful or that more patients or longer follow-up is needed. In this article, we discuss a strategy for making these types of changes. This strategy controls the probability of making a type I error and can result in a procedure that has higher power than the test without adaptation. *The views expressed are those of the author and not necessarily those of the U.S. Food and Drug Administration.

Updating clinical endpoint definitions

The 6-Minute Walk Distance (6-MWD) has been the most utilized endpoint for judging the efficacy of pulmonary arterial hypertension (PAH) therapy in clinical trials conducted over the past two decades. Despite its simplicity, widespread use in recent trials and overall prognostic value, the 6-MWD has often been criticized over the past several years and pleas from several PAH experts have emerged from the literature to find alternative endpoints that would be more reliable in reflecting the pulmonary vascular resistance as well as cardiac status in PAH and their response to therapy. A meeting of PAH experts and representatives from regulatory agencies and pharmaceutical companies was convened in early 2012 to discuss the validity of current as well as emerging valuable endpoints. The current work represents the proceedings of the conference.

Testing non‐inferiority and superiority for two endpoints for several treatments with a control

Some multiple comparison procedures are described for multiple armed studies. The procedures are appropriate for testing all hypotheses for comparing two endpoints and multiple test arms to a single control group, for example three different fixed doses compared to a placebo. The procedure assumes that among the two endpoints, one is designated as a primary endpoint such that for a given treatment arm, no hypothesis for the secondary endpoint can be rejected unless the hypothesis for the primary endpoint was rejected. The procedures described control the family-wise error rate in the strong sense at a specified level α.

Is Randomization to Placebo Safe? Risk in Placebo-Controlled Angina Trials: Angina Risk Meta-Analysis

Objective: It was the aim of this study to document the risks of symptomatic patients with angina in placebo-controlled, anti-anginal drug development trials in which symptom-limited exercise testing was used as the primary endpoint. Patients and Methods: The original case report forms submitted to the United States Food and Drug Administration in support of approval of new or supplemental new drug applications between 1973 and 2001 were identified and subjected to a by-patient meta-analysis, utilizing both a maximum likelihood analysis and classical Mantel-Haenszel methods. Results: There were 63 placebo-controlled, clinical trials that randomized 10,865 patients, with 1,047 patient-years of observation time. The trials involved 21 different chemical entities from 4 different drug classes. The relative risk (RR) for withdrawal (placebo compared to drug-treated patients) was not increased [RR = 0.92, 95% confidence interval (CI) 0.78–1.08; p = 0.28]. Of interest, a RR of 0.54 (95% CI 0.26–1.04; p < 0.068) for irreversible harm (a combination of cerebrovascular accidents, myocardial infarction and death) and a RR of 0.89 (95% CI 0.61–1.30; p = 0.56) for serious cardiovascular events (myocardial infarction, congestive heart failure, cerebrovascular accidents) both non-statistically significantly favored being randomized to placebo. Conclusions: For the development of current or future drugs for the treatment of angina, there is no obvious contraindication to the use of placebo controls and exercise tolerance testing.

Is pulmonary vascular resistance index predictive of exercise tolerance in adult patients with idiopathic pulmonary arterial hypertension

BACKGROUND Clinical trials for adults with pulmonary arterial hypertension use exercise capacity, as measured by walking distance, as the primary endpoint to measure symptomatic improvement. In this article, we look at the relationship between walking distance and a hemodynamic variable, pulmonary vascular resistance index (PVRI), from the available trials. METHODS Patient-level data from 16 randomized controlled clinical trials were obtained. All idiopathic subjects with a baseline and study endpoint measurement of both hemodynamic and exercise endpoints were included. Changes from baseline in both endpoints and the relationship between the endpoints were summarized. Receiver operating characteristic curves were used to investigate the predictive ability. Measures of surrogacy were also calculated. RESULTS There is a weak correlation between changes in PVRI and exercise capacity. Receiver operating characteristic analysis shows a high false positive rate of using one variable to predict the other. Measures of surrogacy show the proportion of variability in exercise capacity explained by PVRI is approximately 5%. CONCLUSIONS PVRI should not be used as a surrogate marker to predict changes in exercise capacity.

Designing Group Sequential Trials with Survival Endpoints

This paper discusses an algorithm to calculate information about the trial that is needed at the design stage implemented as an S-plus function. The function allows the user to specify arbitrary length of accrual, control, and treatment survival distributions; number and timing of interim analyses; stopping boundaries or an alpha-spending function; and which member of the Harrington-Fleming Gρ class of statistics it is. Three examples from real treatment protocols are presented.

A Multiple Comparisons Procedure for Comparing All Arms in Three-Arm Clinical Trials

In three-arm studies, there are often multiple hypotheses to be tested among the arms. In this article, we describe a multiple comparisons procedure that controls the familywise error rate in the strong sense. The procedure is appropriate when one of the arms is a control, and both treatment arms will be compared to the control as well as to each other. The procedure is shown to be consistent and the asymptotic relative efficiency (compared to the uniformly most powerful test for the individual comparisons) is close to 1 in many scenarios. Furthermore, the special case where the arms are placebo, active control, and test drug with one goal of showing noninferiority is examined.

Analysis of Combination Trials with No Placebo Arm

When two drug products are combined together, the traditional design of a trial to investigate the efficacy of the combination product usually includes four arms: a placebo arm, both monotherapy arms, and the combination arm. For ethical reasons, it is sometimes not possible to include a placebo arm in the study when one drug has already been shown to be effective. In this article, we will specifically look at the case when drug A has already been shown to have a mortality effect relative to placebo from an historical study. A new drug is believed to have a similar effect to drug A, and the effect of the combination of the two drugs is believed to be more effective than drug A alone. A clinical trial is designed with three arms to show these conclusions, but no placebo arm is present.