Bogstedt A, Groves M, Tan K, Narwal R, McFarlane M, Höglund K
J. Alzheimers Dis. 46 (4) 1091-1101 [2015-09-25; online 2015-09-25]
Utilizing decision making biomarkers in drug development requires thorough assay validation. Special considerations need to be taken into account when monitoring biomarkers using immunoassays in the presence of therapeutic antibodies. We have developed robust and sensitive assays to assess target engagement and proof of mechanism to support the clinical progression of a human monoclonal antibody against the neurotoxic amyloid-β (Aβ)42 peptide. Here we present the introduction of novel pre-treatment steps to ensure drug-tolerant immunoassays and describe the validation of the complete experimental procedures to measure total Aβ42 concentration (bound and unbound) in cerebrospinal fluid (CSF) and plasma, free Aβ42 concentration (unbound) in CSF, and Aβ40 concentration in CSF. The difference in composition of the matrices (CSF and plasma) and antigen levels therein, in combination with the hydrophobic properties of Aβ protein, adds to the complexity of validation. Monitoring pharmacodynamics of an Aβ42 specific monoclonal antibody in a non-human primate toxicology study using these assays, we demonstrated a 1500-fold and a 3000-fold increase in total Aβ42 in plasma, a 4-fold and 8-fold increase in total Aβ42 in CSF together with a 95% and 96% reduction of free Aβ42 in CSF following weekly intravenous injections of 10 mg/kg and 100 mg/kg, respectively. Levels of Aβ40 were unchanged. The accuracy of these data is supported by previous pre-clinical studies as well as predictive pharmacokinetic/pharmacodynamics modeling. In contrast, when analyzing the same non-human primate samples excluding the pre-treatment steps, we were not able to distinguish between free and total Aβ42. Our data clearly demonstrate the importance of thorough evaluation of antibody interference and appropriate validation to monitor different types of biomarkers in the presence of a therapeutic antibody.