Harmonization study between three laboratories for expression of malaria vaccine clinical trial IgG antibody ELISA data in µg/mL

The ability to report vaccine-induced IgG responses in terms of µg/mL, as opposed arbitrary units (AU), enables a more informed interpretation of the magnitude of the immune response, and better comparison between vaccines targeting different antigens. However, these interpretations rely on the accuracy of the methodology, which is used to generate ELISA data in µg/mL. In a previous clinical trial of a vaccine targeting the apical membrane antigen 1 (AMA1) from Plasmodium falciparum, three laboratories (Oxford, NIH and WRAIR) reported ELISA data in µg/mL that were correlated but not concordant. This current study sought to harmonize the methodology used to generate a conversion factor (CF) for ELISA analysis of human anti-AMA1 IgG responses across the three laboratories.

This study enabled two out of the three laboratories to harmonize their µg/mL readouts for the human anti-AMA1 IgG ELISA, but results reported from WRAIR are ~ twofold higher. Given the need to validate such information for each species and antigen of interest, it is important to bear in mind these likely differences when interpreting µg/mL ELISA data in the future.

Purified IgG was distributed to the three laboratories and, following a set protocol provided by NIH, AMA1-specific human IgG was affinity purified. A new "harmonized CF" was generated by each laboratory using their in-house ELISA, and the original clinical trial ELISA data were re-analysed accordingly.

Statistical analysis showed that the data remained highly correlated across all three laboratories, although only Oxford and NIH were able to harmonize their CF for ELISA and generate concordant data. Conclusions: This study enabled two out of the three laboratories to harmonize their µg/mL readouts for the human anti-AMA1 IgG ELISA, but results reported from WRAIR are ~ twofold higher. Given the need to validate such information for each species and antigen of interest, it is important to bear in mind these likely differences when interpreting µg/mL ELISA data in the future.