A first WHO reference reagent for the detection of anti‐human platelet antigen‐15b

Abstract Background and Objectives Alloantibodies to human platelet antigen‐15b (anti‐HPA‐15b) have been detected in mothers with foetal–neonatal alloimmune thrombocytopenia and in multiply transfused patients. Assays used to detect this antibody, which aids in disease diagnosis, can be unreliable and vary in sensitivity. The objective was to generate a stable, lyophilized anti‐HPA‐15b preparation and evaluate its suitability as a World Health Organization (WHO) reference reagent for use in the quality control of platelet alloantibody detection assays. Results from an international collaborative study to evaluate the preparation were used to assign a minimum potency at which laboratories can be expected to detect the antibody. Materials and Methods Recalcified plasma containing anti‐HPA‐15b was aliquotted, lyophilized and coded 18/220. Twenty‐five laboratories in 16 countries tested doubling dilutions of the reconstituted material in glycoprotein‐specific assays such as the monoclonal antibody–specific immobilization of platelet antigen assay and reported the last positive (or endpoint) dilution. Results Twenty‐four laboratories (96%) detected antibodies with HPA‐15b specificity in preparation 18/220. Reported endpoint dilutions were normally distributed with a modal dilution of 1 in 16 and ranged from 1 in 2 to 1 in 128. Only two laboratories (8%) failed to detect anti‐HPA‐15b at 1 in 8 dilution. Conclusions When diluted 1 in 8, most laboratories detected anti‐HPA‐15b in preparation 18/220 using HPA‐15bb platelets but not with HPA‐15aa platelets. The participants agreed this to be an appropriate dilution for assignment as the minimum potency. In October 2020, the WHO Expert Committee on Biological Standardization approved 18/220 as an International Reference Reagent.


INTRODUCTION
To date, a total of 41 platelet alloantigens have been defined serologically, of which 12 are grouped in six bi-allelic systems (human platelet antigens; HPA -1, -2, -3, -4, -5, -15) [1,2]. The molecular basis of the 41 antigens has been resolved, and in all but one (HPA-14b), the difference between self and non-self is defined by a single nucleotide polymorphism in the gene encoding the relevant membrane glycoprotein [2]. Alloantibodies against HPAs are involved in foetalneonatal alloimmune thrombocytopenia (FNAIT), a disease in which foetal/neonatal platelets are destroyed by IgG alloantibodies from the incompatible, HPA-sensitized mother. HPA antibodies are also involved in platelet transfusion refractoriness (PTR), the repeated failure to achieve the desired level of blood platelets following transfusion and in post-transfusion purpura (PTP), the delayed reaction to a transfusion caused by recipient antibodies to platelet antigens. Identification of the HPA antibody specificity is essential to the diagnosis and treatment of the patient. In severe cases of thrombocytopenia requiring treatment with a platelet transfusion, it is important that the transfused platelets are negative for the target alloantibody specificity [3][4][5][6]. reference reagents established (anti-HPA-5b, anti-HPA-3a and anti-HPA-1a) [7][8][9], which are used for assay quality control. These reagents are used to validate the minimum sensitivity of tests for the respective HPA antibodies.
In addition to HPA-1a, -3a and -5b, HPA-15 is of clinical relevance. Studies have shown this antigen to be as immunogenic as HPA-5, and alloantibodies against HPA-15b can be detected in patients receiving multiple transfusions and mothers with FNAIT [10][11][12]. The platelet-based methods used for the detection of anti-HPA-15 antibodies can be unreliable and vary in their sensitivity because CD109, the glycoprotein on which HPA-15b is located, is expressed in low numbers by platelets and is labile [13,14]. An anti-HPA-15b minimum potency reference reagent would allow clinical laboratories to validate their methods. The need for this important reference material was emphasized in the proficiency testing scheme organized by National Institute for Biological Standards and Control (NIBSC) in 2017 showing that anti-HPA-15b detection in the samples distributed was generally poor. Furthermore, it was concluded in a report by the Platelet Immunobiology Working Party of the International Society of Blood Transfusion (ISBT) in 2018 [15] that a more standardized approach to the CD109 MAIPA is required.
The purpose of this study was to produce and evaluate a candidate WHO anti-HPA-15b reference reagent (minimum potency) for use as a quality control reagent in glycoprotein-specific assays used to detect/identify alloantibody specificity. The authors describe the evaluation of an anti-HPA-15b lyophilized preparation coded 18/220 in an international collaborative study, the results of which are used to assign a minimum potency to this preparation. Clinical laboratories testing the established reference reagent at the assigned minimum potency should expect a positive result if the sensitivity of their method is acceptable.

Candidate material
The material was provided to NIBSC by the National Blood Service, Oxford, United Kingdom. All donations of recalcified plasma came from just one consenting donor, and each was screened using the MAIPA to identify the anti-HPA-15b titre. Those with the highest titres were pooled to make a bulk (coded 18/220) prior to filling 0.5 ml/ampoule into 1896 glass ampoules (size 2.5 ml). Filled material was freeze-dried under conditions established in house for the routine lyophilization of WHO serum standards/reference reagents; a summary of the product information is shown in Table 1. The individual donations from which the candidate reference material was prepared and the pooled bulk were found to be negative for HBsAg, anti-HIV1 + 2 and anti-hepatitis C virus (HCV). HCV PCR testing of the pooled bulk was also negative, and no microbial contaminants were detected.

Participants
The invitation was distributed to participants of the ISBT Platelet Immunology Workshop and to participants of the HPA antibody detection quality assessment scheme organized by the National External Quality Assessment Site UK (NEQAS). In total, 27 clinical laboratories located across the globe accepted the invitation to participate (see Table 2).

Study design
Four ampoules of the definitive freeze-dried material (coded 18/220) were sent to each laboratory. Participants were asked to reconstitute the material immediately before testing and to titrate at doubling dilutions in one or more assay method(s) routinely used in clinical practice.
Reconstituted material from two ampoules at each temperature was tested in duplicate at a range of twofold serial dilutions in the MAIPA assay with HPA-15bb platelets. Absorbance readings were used to calculate the relative potencies of the accelerated thermal degradation samples by parallel-line analysis using the À70 sample as the reference. Relative potency calculations were performed using the European Directorate for the Quality of Medicines software Com-biStats, version 6.0, using a sigmoid curve model and logit transformation of responses. The Arrhenius equation, relating degradation rate to absolute temperature assuming first-order decay [16], was used to predict the degradation rates for each storage temperature.

RESULTS
A total of 27 clinical laboratories accepted the invitation to participate in the study, however, only 25 laboratories returned results.
Each laboratory was assigned a code number, which does not reflect the order of listing shown in   Participants were asked to provide technical information on their protocols; this is summarized in Appendix S1. Indeed, there was considerable variation among the MAIPA methods performed, the most notable differences (further to those

Stability
Estimates of the potency of 18/220 stored at elevated temperatures for a period of 13 months relative to 18/220 stored at À70 C for 13 months are summarized in Table 4. Tests for nonparallelism and non-linearity in the parallel-line analysis to estimate relative potencies were not statistically significant (p = 0.14 and 0.59, respectively). There was insufficient degradation at the elevated temperatures, even after 13 months of storage, to fit the Arrhenius model, with only a clear relative potency loss at +45 C.
This indicates that 18/220 will be stable for long-term storage at À20 C and sufficiently stable to allow for shipment of ampoules at ambient temperature.

DISCUSSION
The aim of the study was to prepare and evaluate a stable, reference reagent for anti-HPA-15b detection that clinical laboratories can use to assess and validate the sensitivity of their routine assays. The collaborative study has shown that candidate preparation 18/220 contains anti-HPA-15b antibody that could be detected at a dilution of 1 in 8 by A report of the 19th ISBT Platelet Immunology Workshop gave a comprehensive summary of the variations in the MAIPA procedure used by workshop participants and concluded that the MAIPA is far from harmonized which may contribute to variations in results [15].
While our study also showed a lack of inter-laboratory harmonization of the CD109 MAIPA procedure, we were also able to demonstrate notable intra-assay variability in reported endpoint dilutions that are likely due to the differential levels in CD109 expression by platelets from different donors, as previously reported [14]. Therefore, the harmonization of test methods using optimized assay conditions may well improve the sensitivity of testing across clinical laboratories to some degree but cannot overcome test variability as a result of donor-donor differences. The use of the anti-HPA-15b reference reagent will at least provide some guarantee as to the sensitivity of the platelets used.
All participants of the international collaborative study were invited to comment on the final study report, and their approval for use of 18/220 as a reference reagent for human IgG antibodies against HPA-15b with a minimum potency of 1 in 8 was sought. No objections were received from the participants, and, in addition, this proposal was endorsed by the ISBT Working Party on Platelet Immunology. In October 2020, all data were reviewed by the WHO Expert Committee on Biological Standardization, and 18/220 was approved for use as a first WHO anti-HPA-15b reference reagent. This material should be used at a dilution of 1 in 8 for assay validation (i.e., the minimum potency which should test positive) and may be used to qualify 'in-house' controls. It is hoped that wide use of this reference reagent will improve the sensitivity of test methods giving more confidence to the results generated.