The never-ending quest for antibody assays standardization and appropriate measurement units

In this issue of the Journal, Hansen et al. are reporting an interesting observation on the proposed WHO standard for anti-SARS-CoV-2 antibodies [1]. In their opinion, introducing separate units for results obtained using neutralising antibody (Nab) assays and for results from binding antibody (bAb) assays is not appropriate as it represents ‘a deviation from international nomenclature conventions used by WHO to assign International Units to CRM’. In addition, the authors bring on a proposal to use a common term for international units (IU) while maintaining a distinction according to the target antibodies of different assays, e.g., neutralizing antibodies targeting specific portions of SARS-CoV-2 spike proteins, binding antibodies to spike or binding antibodies to the nucleocapsid. We recognize this point to be valid, but we would like to add some comments. From the reference they used [2] it looks like they consider the units of that WHO standard as SI units, that by definition have recognized dimensions and are independent of measurement procedure. However, this is not the case for biological controls where the measurand is classified as a class B. The three elements of a class B analyte, together making up the measurand, are the system (sample matrix) the component such as Ig class and target specificity, and the kind of quantity (e.g., the biological activity) [3]. As we commented in a previously published Editorial [4], the signal generated by antibody assays is influenced by the Ig class(es) involved and by the relative affinity to the antigenic targets, and thus to time after infection as a low affinity antibody response is raised in the early stages of infection and a high affinity characterizes past as well as chronic infections. Both factors (Ig classes and affinity) will hamper the reliability of antibody standards that are usually prepared by pooling plasma specimens collected from many individuals whose infection stage is unknown. By a probabilistic estimate, majority of samples should come from people in late stages, with an overabundance of high affinity IgG, which will make the standardization of different methods detecting only IgG, only IgM or all Ig classes (‘total’ antibody assays) an almost impossible task. A living example of the difficulties in manufacturing and using reliable standards for infectious diseases serology is provided by the standardization of IgG antibodies to Rubella virus, that has been proposed since many years but is still failing to reach a sustained agreement across assays, both in general and at the supposed ‘immunity’ threshold of 10 IU/mL [3]. We may therefore conclude that even adopting this target-based distinction proposed by Hansen et al. [1] will not be sufficient to harmonize, let alone standardize, the results generated by different SARS-CoV-2 antibody assays. We may also comment on this issue from a wider perspective. More than 30 years ago Roger Ekins established the two categories for assays employed in the biological sciences: analytical (or “structurally-specific”) and comparative (or ‘functionally specific’) [5], where the latter compare the relative effects of substances, or mixtures of substances, not necessarily of identical chemical structure, on a biological system and whose results should be represented by units of effect and not by units of “amount” of the substance(s) measured and therefore cannot be ‘standardized’ by the use of a calibrant. While this general statement may look harsh, we shall acknowledge that view and accept the metrologic imperfection of antibody assays. In this specific field, focusing on absolute thresholds or reference values does not looks feasible and efforts shall be better aimed to establish assay-independent ranges of antibody response that shall characterize different infection stages or satisfy clinical needs, such as establishing an adequate response to vaccination or the antibody levels that may trigger a medical intervention e.g., start or stop of a specific treatment or switch to a different schedule or drug combination. While we totally agree with the concerns raised by Hansen and Coll. to avoid an overflow of measurement units in laboratory medicine, which should create confusion in both laboratory professionals, physicians and patients [1, 6], we have to recognize the role of the WHO International Standard for anti-SARS-CoV-2 immunoglobulin as a primary calibrant to harmonise the measurement of such antibodies. In fact, it facilitated comparison between clinical trials expressing the results of neutralization assays from different laboratories and allows the use of such measurements as a correlate of protection against infection, also in vaccinated individuals [7, 8]. A new SARS-CoV-2 antibody standard is currently under evaluation under the auspices of WHO and will possibly be available a few months from now. In our opinion, the scientific societies operating in the enlarged clinical laboratory environment that includes immunology, microbiology and infectious diseases shall be involved in that process to guarantee not only the validation process to be run according to the correct principles and verification, but also that units will be assigned, based on functional affinity and not on mass concentration and will consider the intended use of SARS-CoV-2 antibody assays.