PHASTAR at PSI 2017

PSI 2017 - Conference Review

PHASTAR staff made multiple contributions to this year's PSI Conference, which was held in London this past May.

Members of our statistical staff presented on Overdispersed Count Data, describing various models available to account for overdispersion, and presented a simulation study comparing their performance when dispersion differs between treatment groups. They concluded that the standard Negative Binomial model is quite robust to violation of the assumption of a single common dispersion parameter. However, in extreme circumstances, fitting a Heterogeneous Negative Binomial model can provide improved standard errors.

PHASTAR also presented on “Statistical Considerations for Safety Assessment in Cancer Immunotherapy Trials”, with a focus on immune-mediated adverse events. Design and sample size considerations based upon the expected delayed effect of IO agents were discussed, including the analysis challenges with traditional oncology endpoints of ORR, PFS, and OS and emerging alternative endpoints and statistical analysis approaches. Special considerations for the immune-mediated adverse events around data capture, toxicity management, and reporting were also discussed.

PHASTAR’s Head of Statistical Consulting presented in a session of presentations from the EFSPI Integrated Data Analysis Special Interest Group. Her talk focused on the synthesis of safety data using individual patient level data, which provides an opportunity to allow for the impact of varying exposure. A motivating example of a Rheumatoid Arthritis development program was used, and was assessed using simulations of the performance of various methods, including grouped survival analysis.

As a representative of the PSI Scientific Committee, one of PHASTAR’s our Principal Statisticians chaired a session showcasing the work of young statisticians, whilst another presented a poster showing how SAS can be leveraged for large datasets in the pharmaceutical sector, focusing on the similarities/differences of two big data frameworks: Spark and SAS Visual Analytics.

The PHASTAR delegates also attended sessions on a range of interesting topics which they are feeding back to their colleagues, including combination trials, enrichment design for rare diseases, multi-regional studies, extrapolation of survival curves and statistical policy and guidance at the FDA. They also enjoyed networking with fellow statisticians and clients.

Next year’s conference is in Amsterdam in early June - perhaps we will see you there?