Speaker

Biography

Abstract

Tuesday 25^th October

Ondrej Slama

Ondrej joined Roche in the spring of 2019 as he started in SPA-DA team, contributing to the NEST core and other molecule-related projects. He was part of MS Floodlight, Admiral, and now works on Ophthalmology related projects, part of which is VIStA, which he is going to present. Ondrej comes from the Czech Republic where he studied the Faculty of Nuclear Sciences and Physical Engineering at CTU, mostly focusing on mathematics, which eventually brought him to machine learning and data science.

VIStA - Visualizing Images with Statistical Analyses.
In ophthalmology, but also in other areas, there is a growing need to promptly analyze non-standard clinical tabular data. This often includes imaging data with different imaging modalities. Visualizing Images with Statistical Analyses, or VIStA, is a framework that includes guidelines and template code to create interactive exploratory tools and analysis figures that relate clinical and/or imaging features with the underlying images. Our current use cases consist of analytical apps based mostly in R shiny, but special use cases also include JS, CSS, and Python. Our vision is to create a unified approach to use statistical analyses and analytical tools for data scientists or TA teams requiring interactive image visualization functionality.

Yilong Zhang

Dr. Yilong Zhang is a biostatistician at Meta and enjoys clinical development. Before joining Meta, he worked as a statistician at Merck and earned a PhD degree in biostatistics from New York University. He co-authored a book in using R to fill process, technical and training gaps for clinical trial analysis, reporting, and submission. His interests in statistical methods include study design, missing data, and survival analysis. 

R-based test submission to FDA.
On 22 November 2021, the R Consortium R Submissions Working Group successfully submitted an R-based test submission package through the FDA eCTD gateway. FDA staff were able to reproduce the numerical results.

This submission, an example package following eCTD specifications, included a proprietary R package, R scripts for analysis, R-based analysis data reviewer guide, and other required eCTD components.

To our knowledge, this is the first publicly available R-based or open-source-language-based FDA submission package. We hope that our materials and what we learned can serve as a good reference for future R-based regulatory submissions from different sponsors.

To bring an experimental clinical product to market, electronic submission of data, computer programs, and relevant documentation is required by health authority agencies from different countries. In the past, submissions have been mainly based on the SAS language. In recent years, the use of open-source languages, especially the R language, has become very popular in the pharmaceutical industry and research institutions. Although the health authorities accept submissions based on open-source programming languages, sponsors may be hesitant to conduct submissions using open-source languages due to a lack of working examples. Therefore, the R Consortium R Submissions Working Group aims to provide such examples as part of its focus on improving practices for R-based clinical trial regulatory submissions.

https://github.com/RConsortium/submissions-pilot1-to-fda

Monika Huhn

Monika Huhn has a background in mathematical statistics and has been with AstraZeneca for nearly

10 years. She has worked in a number of different roles, starting as a study statistician in asthma clinical trials and most recently working as a data scientist within the AZ Data Science & AI department. Her work has mainly focused on designing clinical trials and learning from clinical data. She has focused some of her time on writing software packages and creating web applications in order to enable others to easily connect with data.

Monika is very enthusiastic about utilizing data and finding new methods to answer scientific questions. She is also very interested in data visualization and different means of making data and analysis results accessible and understandable for all. 

OneView – A Shiny app to unlock the full potential drug repositioning investigations. 
Monika Huhn & Shameer Khader, Center for Artificial Intelligence, Data Science & Artificial Intelligence, R&D, AstraZeneca

In many clinical and pre-clinical projects, there is a need to connect biologists and clinicians with the data in a meaningful way. Data is often stored in repositories that are hard to access without programming knowledge or buried in a collection of unstructured and unconnected spreadsheets. R Shiny apps can often help us to bridge this gap and empower scientists to work with their data directly.

One of the Shiny apps we have developed in recent years is OneView, which brings together different data sources to accelerate drug repositioning investigations.
Drug repositioning is an area of growing interest in drug development that can accelerate the discovery of new treatment options to benefit patients worldwide. Briefly, drug repositioning refers to the systematic investigation of a novel disease indication for a drug molecule. Drug repositioning can be accelerated using various tools and technologies, including intelligent dashboards, data integration and human-in-the-loop machine learning.

The core data behind the OneView Shiny app are from an analysis comparing transcriptomic signatures of drug molecules with hundreds of disease transcriptomic signatures, creating connections between a compound and diseases based on an inverse correlation between the transcriptomic signatures. To fully understand the significance of the relationships, and find further evidence to substantiate them, OneView provides a dynamic dashboard enabling scientists to filter/search within the data, follow connections through multiple datasets, and provide meaningful interactive visualizations.

Tuesday 8^th November

Fiona Ehrich

Fiona Ehrich graduated from Columbia University in May with an MS in Biostatistics and currently works as an Assistant Research Biostatistician at Memorial Sloan Kettering Cancer Center. During graduate school, she interned at Cytel, where she developed R Shiny applications enabling users to conduct customized clinical trial simulations. Prior to graduate school, she worked for four years at Alkermes, a biopharmaceutical company developing medicines in neuroscience and oncology. Fiona is interested in clinical trial design and multi-omics biomarker development in oncology.

Adaptive Group Sequential Enrichment Designs App.
Enrichment strategies in group sequential clinical trials may be considered when there is evidence that a treatment may provide greater benefit to a particular subgroup of the patient population (eg, biomarker-positive). Enrichment designs begin by recruiting the full patient population but build in the option to selectively recruit the prospectively-defined subgroup of interest for the remainder of the study based on the results of an interim analysis. An R Shiny app was developed to enable users to conduct customized simulations of adaptive enrichment clinical trials, using a promising zone approach with sample size re-estimation and early stopping for efficacy, facilitating the selection of an optimized study design.

Guiyuan Lei

Guiyuan is a Senior Principal Statistical Scientist with Roche (UK). Currently she is a Project Lead Statistician and Data Sciences Team Lead for one oncology molecule. She had worked at Universities as a post-doc and research associate (statistician) in Sweden and UK for five years before joined Roche in 2008. At Roche, Guiyuan has developed extensive experience in drug development within multiple therapeutic areas including I2O, Hematology and Solid Tumors. Outside of the molecule work, Guiyuan dedicates her time to many initiatives, most prominently, Advancing Inclusive Research. She has long experience in programming and developed several R Shiny apps. Guiyuan is passionate to help patients both as a professional and as a breast cancer survivor.

R Shiny App for Trial Diversity Dashboard.
Diversity in clinical trials means enrolling trial participants to reflect the intended patient population. FDA released draft guidance for industry on diversity plan in April 2022 which recommends that sponsors should define enrollment goals for underrepresented racial and ethnic participants as early as practicable in clinical development for a given indication. The R Shiny app for trial diversity dashboard is for benchmarking and real-time monitoring of the percentage of each race, ethnicity, or gender category from clinical trials, overall, by indication, by study, by country, etc. With a live demo using dummy data, the author will demonstrate key features of the Shiny app: real-time data source, interactive visualization, and automatic reporting.

Katrin Roth

Katrin joined Bayer as a PhD Scholar in 2006 and continued as a clinical statistician in 2009 after receiving her PhD in Mathematical Statistics from the University of Magdeburg, Germany. Katrin has significant experience in designing and supporting clinical studies and projects in Women’s Health, Pulmonology, Anti-Infectives, Oncology, Radiology, and Pain therapeutic areas. Katrin is an active member of Bayer's Biostatistics Innovation Center (BIC) Dose-Finding group and a newly elected member and the speaker of the BIC Steering Committee. Katrin is also a member of the Conference Advisory Committee supporting the International Biometric Society.

The dosedesignR – an interactive tool for planning dose finding studies
Zhenglei Gao, Franco Mendolia, Christoph Neumann, Katrin Roth, Thomas Schmelter, Katrin Walkamp (Bayer AG)

We are presenting an interactive tool that applies the theory of optimal experimental design to facilitate the planning of dose finding studies. In drug development, phase 2 dose finding studies aim at estimating the dose-response relationship and thereby finding a therapeutic dose for further development in phase 3. The quality of the study design usually depends on the (unknown) shape of the dose-response curve. The planning of dose finding studies is therefore typically an effort that requires substantial input from non-statistical experts from, e.g., pharmacology or pharmacokinetics.

Statisticians from Bayer‘s Biostatistics Innovation Center (BIC) group on dose finding have developed an R shiny [1] app called dosedesignR. The tool is based on the dosefinding R package [2] and aims at facilitating the interactive process during the planning of a dose finding study.

[1] Chang W, Cheng J, Allaire JJ, Xie Y and McPherson J (2018). shiny: Web Application Framework for R. https://CRAN.R-project.org/package=shiny

[2] Bornkamp B., Pinheiro J. and Bretz F. (2016). DoseFinding: Planning and Analyzing Dose Finding Experiments. https://CRAN.R-project.org/package=DoseFinding.

Tuesday 15^th November

Vincent Shen

Vincent Shen is from Roche PD Data Sciences, and is currently the Chief Product Owner for NEST, which is a set of open-sourced R packages that generate, deliver and catalog insights for clinical studies, in both static and interactive formats. During his 8 years at Roche, Vincent has taken various data science-related roles in areas such as medical affairs, real-world evidence, and PD. Prior to that, he worked at Princess Margaret Cancer Centre as a biostatistician focusing on supporting research in lung and head & neck cancer. Vincent completed his Bachelor of Science degree from University of Hong Kong and Master of Mathematics from University of Waterloo. In his spare time, Vincent enjoys building LEGOs, playing piano and cheering for Manchester United.

NEST - productionize a comprehensive open-source toolkit in R. 
In this talk, we will share how we productionize a comprehensive open-source toolkit like NEST and transform how we generate and deliver insights for clinical trials. By applying product thinking mindset and introducing product owners to the various components of the project, we drive a roadmap that allows us to consistently make improvements on the toolkit and closely engage with users. The open-source nature of the product also leads to a multi-layer collaboration model for both internal users and external companies. We will also discuss how NEST delivers unique values to clinical trial analysis, how users are engaged in the development of the toolkit, and our vision on the influence this toolkit brings to clinical trial insights generation.

Ardalan Mirshani

Ardalan Mirshani works as a Data Scientist at Novartis. He holds a phd in statistics from Penn State University. He specializes on scientific tool creation, reproducible workflows, production-grade shiny applications, and automating repetitive operations associated with data exploration, modeling, reporting, and productization processes in the Innovation and Technology group.

Democratizing Shiny App Development.
As clinical data exploration continues to grow, there is an increasing need for interactive graphical displays created with Shiny apps. To date, the development and deployment of study apps have required specialized knowledge and considerable effort. However, the similarity across domains in clinical studies motivated us to build a comprehensive framework that scales shiny app creation across the portfolio. The Datapipeline harmonized framework democratizes the shiny app creation. It enables non-technical associates to create and deploy professional shiny apps quickly. It also empowers shiny developers to build reusable shiny modules that may be easily shared in a plug-and-play manner, ultimately accelerating future application development.