StatisticAlps 2017, 7^thEdition

Biomarkers & Classifiers for Diagnostic and Therapeutic Research:

Discovery, Study Design and Analysis

Ponte di Legno (BS), 20^th – 23^rd March 2017

Lisa McShane & Richard Simon
Biometric Research Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute, Bethesda, USA 

DESCRIPTION

StatisticAlps is a residential training course on advanced statistical issues of interest in clinical research. It is traditionally held in Ponte di Legno, a place in the Alps with beautiful surroundings. The course has reached the 7th edition and this year will be in a winter format.
The focus of this edition will be on biomarkers and classifiers for diagnostic and therapeutic research. Biomarkers are becoming increasingly important for streamlining drug discovery and development. They are widely used as a tool for disease diagnosis and for personalized medicine, and they are also used as surrogate endpoints in clinical research.
This course is designed to give a comprehensive introduction to clinical research on biomarkers and multivariate classifiers, addressing the entire spectrum from discovery to evaluation. The goals of the course are:

• to understand how to conduct translational research on biomarkers and classifier
• to be able to analyze data on biomarkers and classifiers and to interpret results.

The methods covered in the course have applications in diagnostic and therapeutic research.
The course will consist of lectures and tutorials

CONTENTS COVERED

• Uses for biomarkers in drug development and clinical care

• Kinds of validation (Analytical, Clinical, Medical utility)

• Analytical validation (Single analyte, Sequencing panels, Gene expression signatures

• Introduction to case studies of prognostic single analyte biomarkers and biomarker scores

• Statistical methods for development of prognostic scores (Discriminant analysis, Shrunken centroids, Nearest neighbour methods, Support vector machines, Penalized logistic regression, Boosting, Random forests, Penalized proportional hazards model)

• Statistical methods for evaluation of prognostic scores (Intended use of prognostic biomarker scores; Performance assessment; Internal validation methods; Hands on workshop on methods for development and evaluation of prognostic biomarker classifiers)

• Retrospective study designs (Prospective-retrospective design with archived specimens; Efficient retrospective sampling (e.g., case-control, weighted); Sample size/power; Optimally splitting dataset)

• Prospective designs (Considerations to establish medical utility; Trial examples: TAILORx, MINDACT, RxPONDER (also some predictive aspects), Sample size/power)

• Introduction to case studies of predictive biomarkers

• Statistical methods for development of predictive scores/classifiers (Score development methods)

• Statistical methods for evaluation of predictive scores (Performance measures; Cross-validation for evaluation of predictive performance; Considerations in use of archived specimens; Special case sampling approaches in retrospective study designs)

• Hands on workshop on methods for development and evaluation of predictive biomarker classifiers

• Evaluation of intermediate endpoint biomarkers (Surrogate endpoints; Patient management biomarkers; Imaging biomarkers; Liquid biopsy based biomarkers)

• Biomarker driven clinical trial designs (Basket design, Enrichment design, Umbrella design, Stratified all-comers design, Strategy design, Adaptive signature design, Adaptive threshold design, Adaptive enrichment designs)