Development and validation of diagnostic and pharmacodynamic cancer biomarker methods and their (pre)clinical application

Abstract

Biomarkers as biological signposts of cancer have become an essential component of oncology drug development. The importance of biomarkers in oncology has grown substantially due to the advent of targeted anti-cancer therapies, which are directed against a specific biological target that tumor cells rely on for their growth. In clinical cancer trials, biomarkers are used for the selection of patients most likely to benefit from a certain (personalized) treatment. Biomarkers can also facilitate the early detection of treatment outcome, which may provide essential guidance for treatment modification or discontinuation. For these reasons, the interest in the use of biomarkers in cancer trials has grown considerably over the past 60 years. Many challenges are encountered in the development of reliable and robust methods used for the measurement of biomarkers. Firstly, biomarkers need to be detected in patient material such as body fluids and tissue that contain a complex mixture (the matrix) of similar endogenous molecules. Secondly, endogenous compounds may affect the stability of a biomarker, or the biomarker is susceptible to post-translational modification. Thirdly, obtaining biomarker-free matrices for the method development is often impossible. Lastly, sometimes a pure and certified biomarker is not available to serve as reference standard. The level of purification of these reference standards may vary considerably between manufacturers, resulting in high lot-to-lot variation of 76% in a European Bioanalysis Forum case study. In this thesis, we focus on the development and validation of diagnostic and pharmacodynamic biomarker methods for the support of (pre)clinical cancer studies. Diagnostic biomarkers provide essential insight into the patient’s disease condition, for instance whether metastases are present. In contrast, pharmacodynamic biomarkers are used to establish whether a cancer drug is hitting its pharmacological target and thereby help to guide the clinical decision making about drug dose and optimal administration schedule. In regard to the importance of biomarkers, laboratory methods for their measurement need to be extensively validated to ensure that the used methods are specific, accurate, precise, and robust. We found, however, that only 13% of the pharmacodynamic biomarker methods used in clinical cancer trials for the past 7 years were fully validated. The majority of biomarker methods used in clinical trials have only partly been validated, which may explain why these methods not always perform as expected in the clinic. It may also be an important factor in the relatively low approval rate of 27% for new cancer drugs in 2018. This lack of validation for clinical cancer biomarker methods may be explained by the absence of international guidelines for their validation. However, since 2018 recommendations for the validation of biomarker methods have existed from the Food and Drug administration. Here, we demonstrate the development of seven cancer biomarker methods, and show that full validation of these biomarker methods is possible. These biomarker methods were designed for high sensitivity and precision to further enhance their clinical usefulness. We successfully applied these fully validated biomarker methods in (pre) clinical cancer studies. The clinical chemistry laboratory of our hospital, Antoni van Leeuwenhoek, has implemented our biomarker method for the detection of epithelial circulating tumor cells as a standard diagnostic tool for the diagnosis of leptomeningeal metastases.