EFSPI Meetings
Biomarkers & Subgroups
Date: Friday 24th June 2016
Venue:
Astellas Pharma Europe B.V.,
Sylviusweg 62,
Leiden,
Netherlands
Summary
On June 24th, one-day meeting on Biomarkers and Subgroups was held in Astellas B.V. in Leiden. This meeting brought together the experts from industry, academia and regulatory agencies to discuss about the challenges on biomarkers and subgroups.
The event started with a welcome drink; and afterwards the first speaker, Hans-Ulrich Burger (Roche), set the scene for the day providing an overview of subgroup selection for biomarkers. He touched upon the challenges with biomarkers and subgroups, leaving the floor to the other speakers to discuss about these challenges and to provide some recommendations. Afterwards, Norbert Benda (BfArM) presented regulatory view on the Biomarkers and Subgroups. He talked about stratified medicine and a research project on biomarker defined populations. The third speaker, Andrew Grieve (ICON plc), presented adaptive designs for subgroup selection. He focused on “enrichment” designs talking about that “One size fits all” approach may not work well in different situations, giving the example of Type I and II diabetes.
The last speaker before the lunch break was Tim Friede (University Medical Center Gottingen). His talk revolved around the methods for exploration and confirmation of subgroups. In his talk, he mainly focused on meta-analytic techniques, explaining the differences between different estimators for estimating the overall effect in a meta-analysis especially in the case of small number of publications and heterogeneity.
After the lunch break, Andrew Stone (Stone Biostatistics) gave a presentation about Including the Biomarkers in Clinical Development Plans.
Afterwards, Sebastien JobJörnsson, a PhD student from Chalmers University presented optimal trial design for targeted therapies; different example when focusing on full population, enrichment design or a stratified approach The last speaker of the day, Xavier Benain (Sanofi), talked about the Statistical Challenges in Diagnostic Biomarker Qualification. How to confirm diagnostic biomarker best threshold using different techniques (including resampling techniques) in simulations studies.
At the end of the day all speakers were invited to the stage for a panel discussion. There were many questions from the attendees to the speakers during the panel discussion. The speakers agreed on the take home message from the panel discussion: “Pre-specify subgroups”, “Multiplicity is an issue” and “Biological plausibility play an important role”.
The event was very successful with 7 speakers approaching the challenges with biomarkers and subgroups from different perspectives and more than 50 attendees.
Presentations