Making sense of serotype-specific pneumococcal antibody measurements

In 2012, a specialist group of the American Academy of Allergy, Asthma & Immunology (AAAI) collaborated to produce guidelines on the use and interpretation of diagnostic vaccination in primary immunodeficiency. ¹ Within this document there is extensive evaluation of the use of serotype-specific pneumococcal serotype testing for diagnosis of immune deficiency. A titre of 1.3 µg/mL is reported to confer protection for each serotype. This concentration has generated much discussion: in the UK these tests are only available in a few centres and all use 0.35 µg/mL as a putative protective titre. In an attempt to help clarify this discrepancy, this article summarizes the literature describing the development of these assays.

Although widely quoted, a ‘protective titre’ of 1.3 µg/mL may not have a sound evidence base. The evolution of assays to measure antibodies to serotype specific pneumococcus, particularly the unit of measurement conversions, has played a significant part in the confusion. A sensitive radioimmunoassay was developed by Schiffman et al. ² in 1980 which measured IgG antibodies to C polysaccharide generating antibodies in units of ng antibody nitrogen/mL. Later enzyme-linked immunosorbent assays (ELISAs) were developed which were more sensitive but also generated results in ng antibody nitrogen/mL. ³ In 1995, the World Health Organisation (WHO) produced reference and calibration sera to standardize assays. ⁴ The first ELISAs showed poor correlation of antibodies with vaccine efficacy studies. This was due to the presence of C polysaccharide and 22F. Antibodies to C polysaccharide, although naturally occurring in humans, are not opsonic and therefore do not protect against pneumococcal infection; antibodies to serotype 22F are cross reactive to a number of other serotypes.^5,6 Reabsorbing sera with C polysaccharide and 22F led to the development of 2nd and 3rd generation assays respectively with increased specificity for measuring protective antibodies; these assays gave results in µg/mL. ⁷ From this point onwards there was a divergence in interpretation of assay results depending on special interests such as whether the assays were used to assess vaccination response from a public health perspective or to compare base line and postvaccination response to help diagnose T-cell independent antibody deficiency.

The WHO approved the ELISA for monitoring response to newly produced vaccines. ⁷ WHO published data showing that pneumococcal conjugated vaccines should induce titres of 0.35 µg/mL or above four weeks after vaccination to confer protection against invasive pneumococcal disease.^8,9 This putative protective level was based on a 2nd generation ELISA. When reassessed on a 3rd generation (more sensitive) ELISA only a small decline in putative protection level was found (0.32 µg/mL); this meta-analysis was based on three clinical efficacy trials with in excess of 60,000 control and vaccinated infants. ¹⁰ A decision to stick to the original value (for ease) of 0.35 µg/mL was made. ¹⁰ Although derived from data in children, this level of protection has been extrapolated for adult use.

Development and interpretation of ELISA results in the adult immune deficiency world took a different approach. Landesman and Schiffman ¹¹ used the radioimmunoassay to assess pneumococcal antibody concentration with systemic pneumococcal infection >250–300 ng antibody nitrogen/mL for adults. This figure was derived from vaccine failure data where antibody levels were not protective against pneumococcal infection in high risk populations. The recommendation of concentration was based on 23 cases where 22/23 had one of eight serotypes below 300 ng antibody nitrogen/mL. A similar study in 79 children concluded that ≥200 ng antibody nitrogen/mL correlated with decreased nasopharyngeal colonization postvaccination. ¹² These definitions were used for conversion of units to µg/mL for ELISA once this method was optimized. Sorensen et al.^13,14 in 1998 stated: ‘We arbitrarily defined an adequate IgG antibody response to an individual serotype as a post-immunization antibody concentration as ≥1.3 µg/mL. This is equivalent to 200 ng antibody nitrogen/mL based on a conversion factor of 160 ng of antibody nitrogen to 1 µg/mL'. This conversion factor was based on a personal communication to the authors as cited in Sorensen et al.^13,14 It is unclear how the conversion factor of 160 was derived.

A year later Sorensen’s group again used this antibody concentration in evaluating the role of polysaccharide vaccines in children but made the following comment: ‘Further studies are needed to better define what constitutes a normal response as well as a putative protective response to various serotypes. Until better data becomes available an adequate response to individual serotypes may be defined as concentrations ≥1.3 µg/mL.’ ¹⁵ Fifteen years later this value is still widely used worldwide for immune deficiency interpretation. The conflicting protective levels may be cause for concern when clinicians are interpreting results. Regional assays of serotype specific pneumococcal antibody responses in the UK report the putative protective levels either in line with the WHO or based on internal assay development studies. Clinicians are reliant on the expertise of referral laboratories for interpretation of these assays.

An additional caveat to correctly interpreting serotype-specific pneumococcal assays requires knowledge of what vaccines patients have received and which serotypes are measured in the laboratory. In 2010, Prevenar13™ (Pfizer, Sandwich, Kent, UK) was introduced as the standard childhood vaccine. As this vaccine is conjugated, antibodies to these 13 serotypes are produced by T-cell dependent means. The use of investigating pneumococcal antibody vaccine responses in an immunology setting was to assess T-cell independent antibody responses to non-conjugated vaccines (Pneumovax™, Sanofi Pasteur MSD, Maidenhead, Berkshire, UK). Therefore recognition of defects in this pathway, in particular specific antibody deficiency can be made. Ideally laboratories need to include a large number of serotypes not present in Prevenar (but that are included in Pneumovax™) in their serotype-specific assays so that the T-cell independent antibody pathway can still be assessed for diagnosis of primary immune deficiency.

In conclusion, the consensus cut-off concentration of 1.3 µg/mL for individual pneumococcal serotype antibodies has little substantial evidential backing; with data based on small cohorts and an unsubstantiated conversion factor used when units of measurement were changed. Use of 0.35 µg/mL as a minimum protective concentration although proven on extensive populations has also not undergone validation in the primary immune deficiency setting. Audit of data within different cohorts would be highly advantageous to justify current local or national practices. Until such publications arise, interpretation of serotype specific pneumococcal antibody responses requires specific scrutiny and caution.