PSI Pre Clinical SIG Workshop 2022 (TRNG285)

Date: Wednesday 21st - Thursday 22nd September 2022

Time:

21st = 12:30-17:00 | 22nd = 09:00-17:00 BST

21st = 13:30-18:00 | 22nd = 10:00-18:00 CEST

Location: Online

Who is this event intended for? Pre-Clinical Statisticians in the Pharmaceutical Industry with interest and/or some basic knowledge of R/STAN/BUGS and Bayesian statistics.

What is the benefit of attending? This workshop offers the chance to meet with colleagues across industry and learn more about Bayesian Methodology and its applications in pre-clinical.

Overview

This is the pre-clinical SIG's 10th workshop and will be the first one run virtually. Our theme for this workshop is Bayesian; the workshop will run over a day and a half and will include a training course on Bayesian methods (see more below), two presentations on applications of Bayesian methodology in a pre-clinical setting and a breakout session.

Course

Bayesian Statistics for Preclinical Research: New Opportunities

The course will start by introducing the key concepts of Bayesian statistics, emphasizing the context and key objectives of preclinical research in pharmaceutical and medical device development. We then move on to show how Bayesian thinking and practices are a fit-for-purpose paradigm. Over the last decade, preclinical research has been identified in the literature as an area of research suffering from a lack of reproducibility. Causes for this are many, but in this course, we'll show how to frame a Bayesian strategy to address reproducibility concerns by proposing new study designs, modelling, and decision-making. Preclinical research is a learning process, making Bayesian statistical learning a very natural partnership.

Talks

Bayesian Tumor volume analysis with BRMS R package

Marie Miossec

In cancer drug development, demonstrated efficacy in tumor xenograft experiments on severe combined immunodeficient mice who are grafted with human tumor tissues or cells is an important step to bring a promising compound to human. A key outcome variable is tumor volumes measured over a period of time, while mice are treated with certain treatment regimens. The tumor growth inhibition delta T/delta C ratio is commonly used to quantify treatment effects in such drug screening tumor xenograft experiments In this presentation, we propose a Bayesian approach to make a statistical inference of the T/C ratio, including both hypothesis testing and a credibility interval estimate. Through a practical case, implementation, diagnosis, model selection and results with the BRMS R package will be discussed.

A Bayesian, Generalized Frailty Model for Comet Assays

Helena Geys

This paper proposes a flexible modelling approach for so-called comet assay data regularly encountered in pre-clinical research. While such data consist of non-Gaussian outcomes in a multi-level hierarchical structure, traditional analyses typically completely or partly ignore this hierarchical nature by summarizing measurements within a cluster. Non-Gaussian outcomes are often modelled using exponential family models. This is true not only for binary and count data, but also for, e.g., time-to-event outcomes. Two important reasons for extending this family are: (1) the possible occurrence of overdispersion, meaning that the variability in the data may not be adequately described by the models which often exhibit a prescribed mean-variance link, and (2) the accommodation of a hierarchical structure in the data, owing to clustering in the data. The first issue is dealt with through so-called overdispersion models. Clustering is often accommodated through the inclusion of random subject-specific effects. Though not always, one conventionally assumes such random effects to be normally distributed. In the case of time-to-event data, one encounters, for example, the gamma frailty model (Duchateau and Janssen 2007). While both of these issues may occur simultaneously, models combining both are uncommon. Molenberghs et al (2010) proposed a broad class of generalized linear models accommodating overdispersion and clustering through two separate sets of random effects. In Ghebretinsae et al, we used this method to model data from a comet assay with a three-level hierarchical structure. Whereas a conjugate gamma random effect is used for the overdispersion random effect, both gamma and Normal random effects are considered for the hierarchical random effect. Apart from model formulation, we place emphasis on Bayesian estimation.