gms | German Medical Science

65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS)

06.09. - 09.09.2020, Berlin (online conference)

Assessing impact of immunogenicity on clinical outcome: Estimating the treatment effect in post-randomization categorical variables in an RCT

Meeting Abstract

Search Medline for

  • Dominik Heinzmann - Roche, Basel, Switzerland
  • Shengchun Kong - Genentech, South San Francisco, United States

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. 65th Annual Meeting of the German Association for Medical Informatics, Biometry and Epidemiology (GMDS), Meeting of the Central European Network (CEN: German Region, Austro-Swiss Region and Polish Region) of the International Biometric Society (IBS). Berlin, 06.-09.09.2020. Düsseldorf: German Medical Science GMS Publishing House; 2021. DocAbstr. 54

doi: 10.3205/20gmds266, urn:nbn:de:0183-20gmds2662

Published: February 26, 2021

© 2021 Heinzmann et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License. See license information at http://creativecommons.org/licenses/by/4.0/.


Outline

Text

Background: Immunogenicity refers to the immune response of patients against the therapeutic protein, and in certain instances may affect clinical efficacy or safety. Immunogenicity can be measured by anti-drug antibodies (ADAs) and neutralizing anti-bodies (Nabs) against the experimental drug.Immunogenicity can be measured by anti-drug antibodies (ADAs) and neutralizing anti-bodies (Nabs) against the experimental drug.

Methods: ADAs and Nabs are in general summarized as categorical variables measured post-randomization in an RCT. Determination of treatment effects for those ADA (Nab) subgroups are not straight forward since it is not evident which patients in the control arm would have developed ADAs (Nabs) had they been treated with the experimental drug. Hence a naÏve comparison of the ADA (Nab) groups with control does not estimate the treatment effect of interest and more advanced methods are required.

A solution is based on the principal stratum estimand as outlined in ICH E9 (R1). An application to the above problem of interest is presented, assumptions underlying the approach are shown and various estimators are compared and model performance assessed. Handling of missing data in the context of the estimand framework will also be discussed.

Results: It is shown how the principal stratum framework addresses the core clinical questions around potential immunogenicity impact on the clinical outcome of an experimental drug. Furthermore, re-weighting based the propensity score is shown to be superior to matching in this setting across a range of model performance assessment tools.

Conclusion: The application demonstrates the usefulness of the prinicapl statum estimand in the setting of immunogenicity analyses.

Employee of F. Hoffmann-La Roche

The authors declare that a positive ethics committee vote has been obtained.