‘Fit-for-purpose’ validation of SearchLight multiplex ELISAs of angiogenesis for clinical trial use

• Published in: Journal of immunological methods Vol. 342; no. 1-2; pp. 106 - 114
• Main Authors: Backen, A.C., Cummings, J., Mitchell, C., Jayson, G., Ward, T.H., Dive, C.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 15.03.2009; Elsevier
• Subjects: Angiogenesis; Biological and medical sciences; Biomarkers; Biomarkers - blood; Clinical Trials; Cross Reactions; ELISA; Enzyme-Linked Immunosorbent Assay - instrumentation; Enzyme-Linked Immunosorbent Assay - methods; Female; Fundamental and applied biological sciences. Psychology; Fundamental immunology; Humans; Molecular immunology; Multiplex; Neovascularization, Pathologic - diagnosis; Ovarian Neoplasms - blood; Sensitivity and Specificity; Software Validation; Techniques; Validation; Validation; Biomarkers; Angiogenesis; Clinical Trials; Multiplex; ELISA; Biological marker; Clinical trial; ELISA assay; EUSA; Immunological method
• ISSN: 0022-1759; 1872-7905
• DOI: 10.1016/j.jim.2009.01.003

Validated assays of circulating biomarkers of angiogenesis to predict and determine the efficacy of vascular-targeted anticancer drugs would facilitate successful drug development. Multiple biomarker candidates exist and a multiplex approach was sought to minimise the requisite patient blood volume and to aid selection of those biomarkers with greatest potential clinical utility. Validation of the SearchLight™ multiplex ELISA platform comprising two multiplex assays of nine potential angiogenesis biomarkers was conducted (plex 1; VEGF R1 and R2, IL-8, KGF, PlGF; plex 2; PDGFbb, HGF, FGFb and VEGF). The study focused on instrument qualification, analyte specificity within the multiplex format, assay precision and reproducibility. No evidence was found within the multiplex that signals output from one analyte impinged on another or that antibody cross-reactivity occurred. Spike recovery for 5 between-experiment repeats was within ±15% of input values for 7 of the 9 multiplexed analytes, with a coefficient of variation (CV) of <20% for 6 of the 9 analytes. Plasma samples from 8 ovarian cancer patients (who were not receiving therapy) were assessed using the two multiplexes on this platform to explore the likely baseline variability in this disease context. This study suggests that the platform and the multiplex approach will be useful to evaluate pharmacodynamic responses to vascular targeted therapy in early clinical trials.