PLoS ONE (Jan 2021)

A multicenter analytical performance evaluation of a multiplexed immunoarray for the simultaneous measurement of biomarkers of micronutrient deficiency, inflammation and malarial antigenemia

  • Eleanor Brindle,
  • Lorraine Lillis,
  • Rebecca Barney,
  • Pooja Bansil,
  • Sonja Y. Hess,
  • K. Ryan Wessells,
  • Césaire T. Ouédraogo,
  • Francisco Arredondo,
  • Mikaela K. Barker,
  • Neal E. Craft,
  • Christina Fischer,
  • James L. Graham,
  • Peter J. Havel,
  • Crystal D. Karakochuk,
  • Mindy Zhang,
  • Ei-Xia Mussai,
  • Carine Mapango,
  • Jody M. Randolph,
  • Katherine Wander,
  • Christine M. Pfeiffer,
  • Eileen Murphy,
  • David S. Boyle

Journal volume & issue
Vol. 16, no. 11

Abstract

Read online

A lack of comparative data across laboratories is often a barrier to the uptake and adoption of new technologies. Furthermore, data generated by different immunoassay methods may be incomparable due to a lack of harmonization. In this multicenter study, we describe validation experiments conducted in a single lab and cross-lab comparisons of assay results to assess the performance characteristics of the Q-plex™ 7-plex Human Micronutrient Array (7-plex), an immunoassay that simultaneously quantifies seven biomarkers associated with micronutrient (MN) deficiencies, inflammation and malarial antigenemia using plasma or serum; alpha-1-acid glycoprotein, C-reactive protein, ferritin, histidine-rich protein 2, retinol binding protein 4, soluble transferrin receptor, and thyroglobulin. Validations included repeated testing (n = 20 separately prepared experiments on 10 assay plates) in a single lab to assess precision and linearity. Seven independent laboratories tested 76 identical heparin plasma samples collected from a cohort of pregnant women in Niger using the same 7-plex assay to assess differences in results across laboratories. In the analytical validation experiments, intra- and inter-assay coefficients of variation were acceptable at <6% and <15% respectively and assay linearity was 96% to 99% with the exception of ferritin, which had marginal performance in some tests. Cross-laboratory comparisons showed generally good agreement between laboratories in all analyte results for the panel of 76 plasma specimens, with Lin’s concordance correlation coefficient values averaging ≥0.8 for all analytes. Excluding plates that would fail routine quality control (QC) standards, the inter-assay variation was acceptable for all analytes except sTfR, which had an average inter-assay coefficient of variation of ≥20%. This initial cross-laboratory study demonstrates that the 7-plex test protocol can be implemented by users with some experience in immunoassay methods, but familiarity with the multiplexed protocol was not essential.