A Rational Optimization Approach for the Development of a Multiplexed Lateral Flow Immunoassay: Detection of Nonepithelial Ovarian Cancer Markers in Human Serum

With the rise of non-communicable diseases, lateral flow immunoassays (LFIAs) are well-positioned to address the demand for disease monitoring. We present the use of peroxidase-mimicking platinum nanozyme conjugates targeting three ovarian germ cell tumor markers (alpha-fetoprotein (AFP), human chorionic gonadotropin (HCG), and cancer antigen 125 (CA125)) in LFIA. A "design of experiments" (DoE) approach was used to optimize antibody-nanozyme conjugation, superseding conventional "one-factor-at-a-time" optimization strategies, which neglect factor-to-factor interactions obscuring identification of optimal conditions. Crucial to disease monitoring, assays must demonstrate limits of detection (LoD) within clinically defined ranges and produce quantitative readouts. We address limitations of typical LoD calculations, which assume homoscedastic variance across marker concentrations, presenting an alternative model and applying it to assess LFIA performance, subsequently demonstrating clinically relevant LoDs in human serum. To robustly derive marker concentrations from LFIA readouts, the model was coupled with the resolution molecular concentration and, using patient samples, validated against laboratory gold standard values. AFP and HCG assays align well with gold standard measurements, with sensitivities of 87.5% and 100%, and specificities of 98.3% and 100%, respectively. This work outlines a development pipeline of a patient sample validated semiquantitative LFIA, utilizing DoE for streamlined optimization and improving modeling approaches to quantify performance in a manner representative of assay function.