Better decision making in drug development through adoption of formal prior elicitation.

With the continued increase in the use of Bayesian methods in drug development, there is a need for statisticians to have tools to develop robust and defensible informative prior distributions. Whilst relevant empirical data should, where possible, provide the basis for such priors, it is often the case that limitations in data and/or our understanding may preclude direct construction of a data-based prior. Formal expert elicitation methods are a key technique that can be used to determine priors in these situations. Within GlaxoSmithKline, we have adopted a structured approach to prior elicitation on the basis of the SHELF elicitation framework and routinely use this in conjunction with calculation of probability of success (assurance) of the next study(s) to inform internal decision making at key project milestones. The aim of this paper is to share our experiences of embedding the use of prior elicitation within a large pharmaceutical company, highlighting both the benefits and challenges of prior elicitation through a series of case studies. We have found that putting team beliefs into the shape of a quantitative probability distribution provides a firm anchor for all internal decision making, enabling teams to provide investment boards with formally appropriate estimates of the probability of trial success as well as robust plans for interim decision rules where appropriate. As an added benefit, the elicitation process provides transparency about the beliefs and risks of the potential medicine, ultimately enabling better portfolio and company-wide decision making.

Epithelial desquamation observed in a phase I study of an oral cathepsin C inhibitor (GSK2793660).

AIMS: Cathepsin C (CTSC) is necessary for the activation of several serine proteases including neutrophil elastase (NE), cathepsin G and proteinase 3. GSK2793660 is an oral, irreversible inhibitor of CTSC that is hypothesized to provide an alternative route to achieve NE inhibition and was tested in a Phase I study. METHODS: Single escalating oral doses of GSK2793660 from 0.5 to 20 mg or placebo were administered in a randomized crossover design to healthy male subjects; a separate cohort received once daily doses of 12 mg or placebo for 21 days. Data were collected on safety, pharmacokinetics, CTSC enzyme inhibition and blood biomarkers. RESULTS: Single, oral doses of GSK2793660 were able to dose-dependently inhibit whole blood CTSC activity. Once daily dosing of 12 mg GSK2793660 for 21 days achieved ≥90% inhibition (95% CI: 56, 130) of CTSC within 3 h on day 1. Only modest reductions of whole blood enzyme activity of approximately 20% were observed for NE, cathepsin G and proteinase 3. Seven of 10 subjects receiving repeat doses of GSK2793660 manifested epidermal desquamation on palmar and plantar surfaces beginning 7-10 days after dosing commencement. There were no other clinically important safety findings. CONCLUSIONS: GSK2793660 inhibited CTSC activity but not the activity of downstream neutrophil serine proteases. The palmar-plantar epidermal desquamation suggests a previously unidentified role for CTSC or one of its target proteins in the maintenance and integrity of the epidermis at these sites, with some similarities to the phenotype of CTSC-deficient humans.

The perils with the misuse of predictive power.

In early drug development, especially when studying new mechanisms of action or in new disease areas, little is known about the targeted or anticipated treatment effect or variability estimates. Adaptive designs that allow for early stopping but also use interim data to adapt the sample size have been proposed as a practical way of dealing with these uncertainties. Predictive power and conditional power are two commonly mentioned techniques that allow predictions of what will happen at the end of the trial based on the interim data. Decisions about stopping or continuing the trial can then be based on these predictions. However, unless the user of these statistics has a deep understanding of their characteristics important pitfalls may be encountered, especially with the use of predictive power. The aim of this paper is to highlight these potential pitfalls. It is critical that statisticians understand the fundamental differences between predictive power and conditional power as they can have dramatic effects on decision making at the interim stage, especially if used to re-evaluate the sample size. The use of predictive power can lead to much larger sample sizes than either conditional power or standard sample size calculations. One crucial difference is that predictive power takes account of all uncertainty, parts of which are ignored by standard sample size calculations and conditional power. By comparing the characteristics of each of these statistics we highlight important characteristics of predictive power that experimenters need to be aware of when using this approach.

Practical usage of O'Brien's OLS and GLS statistics in clinical trials.

Multivariate techniques of O'Brien's OLS and GLS statistics are discussed in the context of their application in clinical trials. We introduce the concept of an operational effect size and illustrate its use to evaluate power. An extension describing how to handle covariates and missing data is developed in the context of Mixed models. This extension allowing adjustment for covariates is easily programmed in any statistical package including SAS. Monte Carlo simulation is used for a number of different sample sizes to compare the actual size and power of the tests based on O'Brien's OLS and GLS statistics.