The development and validation of a microneutralization assay for the detection and quantification of anti-yellow fever virus antibodies in human serum

Abstract

The plaque reduction neutralization test (PRNT) has been used widely for the detection and quantitation of yellow fever (YF) virus-neutralizing antibodies in human serum; however, it is labor-intensive and challenging to adapt to high-throughput clinical testing needed for vaccine licensure. Here, we describe the development and validation of a new Vero cell-based YF microneutralization (MN) assay, with immunostaining readout, for the detection and quantification of YF virus-neutralizing antibodies in human serum. Comparison of neutralizing antibody titers measured with the YF MN assay versus the historical YF PRNT, based on a 50% reduction in plaque counts (PRNT₅₀), demonstrated 100% serostatus agreement at a titer of 10 (1/dil) in participants with a history of YF vaccination. For validation, intra-assay precision (repeatability), intermediate precision, dilutional accuracy, linearity, specificity, upper limit of quantitation (ULOQ), and lower limit of quantitation (LLOQ) were assessed. The YF MN assay demonstrated suitable intra-assay precision (repeatability) and intermediate precision of 36% and 54%, respectively, with an ULOQ of 10,240. At the lower end of detection, repeatability and intermediate precision were 38% and 41%, respectively, with a LLOQ of 10 (1/dil). Suitable dilutional accuracy, linearity, and specificity across orthoflaviviruses (dengue virus, Japanese encephalitis virus, and Zika virus) and serum matrices (hemolytic, lipemic, and icteric) were also demonstrated. Overall, these promising results led the Center for Biologics Evaluation and Research to confirm the suitability of the validated YF MN assay for the detection and quantification of YF virus-neutralizing antibodies.

Importance: With increased globalization and shifting climate patterns, yellow fever (YF) is re-emerging as a global threat. At present, vaccination remains the most effective prevention strategy. This study describes the development and validation of a new YF microneutralization (MN) assay for the detection and quantification of YF virus-neutralizing antibodies in human serum that offers increased throughput compared with the current standard assay. Overall, the YF MN assay demonstrated acceptable intra-assay precision (repeatability), intermediate precision, dilutional accuracy, linearity, and specificity and is suitable for the detection of YF virus-neutralizing antibodies. Further, the Center for Biologics Evaluation and Research (CBER) supports the use of the YF MN assay in the licensure of candidate YF vaccines.

Keywords: assay development; assay validation; microneutralization assay; neutralizing antibodies; orthoflavivirus; yellow fever virus.

Fig 1

The YF MN assay combined precision profile graphically represents data from the precision panel (blue dots), generated by testing the precision samples at a 1:10 serum starting dilution, the LLOQ panel (orange dots), generated by testing the LLOQ samples at a 1:5 serum starting dilution, and from the ULOQ panel (green dots), generated by testing the ULOQ samples at a 1:20 serum starting dilution. The statistical analysis of the combined YF MN assay precision data demonstrated acceptable precision from the LLOQ titer of 10 to the ULOQ titer of 10,240. Abbreviations: Dil, dilution; GCV, geometric coefficient of variation; LLOQ, lower limit of quantitation; MN, microneutralization; ULOQ, upper limit of quantitation; and YF, yellow fever.

Fig 2

The YF MN assay linear regression analysis was performed to assess the dilutional linearity of the method. Each of the six human serum samples (undiluted and pre-diluted 1:5, 1:10, 1:20, and 1:40) was tested in triplicate in three individual assay runs. One hundred percent (6/6) of the samples had an R² > 0.95 and a slope between 0.67 and 1.50. The y-axis represents the log-transformed observed GMTs. The x-axis represents the log-transformed expected GMTs. The blue shaded area represents the confidence interval of the “expected” y at the given x. Abbreviations: GMT, geometric mean titer; MN, microneutralization; and YF, yellow fever.