Approaches for the characterization of clinically relevant pre‐transplant human leucocyte antigen (HLA) antibodies in solid organ transplant patients

Introduced as commercial assays in the late 1990s, the Luminex® assays have become ubiquitous across UK H&I laboratories and are used by 97% of all laboratories participating in the UK External Quality Assurance scheme for H&I HLA Antibody Detection (Singleton & Darke, 2015). These assays allow for detection and identification of IgG HLA antibodies using the Luminex® xMAP® technology. The Luminex 100/200™ platform has the capacity to distinguish between up to 100 sets of polystyrene microbeads in each test, with different HLA molecules bound to each set, with the newer FLEXMAP 3D® allowing even greater flexibility with up to 500 sets of microbeads per test. Antibody binding is quantified by fluorescence from both the microbead and a secondary antibody (Immucor, 2020; One Lambda, 2015c; Tait, 2016). Compared to the ELISA and FlowPRA™ technology, these assays allow for greater flexibility and demonstrate a greater capacity for high-throughput testing (Luminex xMAP®, 2014).

Luminex® HLA antibody assays are available for variable levels of antibody detection from two commercial suppliers (Immucor, 2020; One Lambda, 2015c). The screening/identification assays contain native HLA antigens on the surface of their microbeads. These kits can be used to screen for HLA antibodies, to monitor established antibody profiles and to investigate spurious reactivity identified by higher resolution single antigen bead kits. The single antigen bead (SAB) assays have a single recombinant antigen (or dimer in the case of HLA Class II molecules) bound to each microbead set (Picascia et al., 2014) and provide a high level of resolution for HLA antibody identification.

The development of these Luminex® SAB assays revolutionized HLA antibody testing and improved both sensitivity and accuracy of identification, particularly pre-transplant, when compared to cellular techniques. These assays enable laboratories to accurately assign negative specificities for highly sensitized patients and their implementation into routine clinical practice is associated with improvements to transplant outcomes and a reduction in unexpected positive cellular crossmatches (Campbell, 2013; Tait, 2016).

3.3.1 Limitations of Luminex® SAB assays

Despite their advantages for antibody identification, there are many challenges and limitations associated with the Luminex® SAB assays. Some of the main areas of contention are described below.

Peptide repertoire

HLA proteins must be associated with a peptide in their peptide binding groove to maintain their correct conformation on the cell surface (Wieczorek et al., 2017). Details of the peptide repertoires used by commercial manufacturers are not generally available, but the specificity of the bound peptide will affect the overall structure and hydrophobicity of the HLA molecule (Wieczorek et al., 2017), which is likely to alter the affinity and avidity of antibody binding (Anderson et al., 2016; Duquesnoy, Marrari, Jelenik, et al., 2013; Duquesnoy, Marrari, da Sousa, et al., 2013; Mallon et al., 2015). Conversely, for a cellular assay, bound peptides will be sourced from the proteins available within the lymphocyte (Duquesnoy, Marrari, Jelenik, et al., 2013; Duquesnoy, Marrari, da Sousa, et al., 2013).

Relationship between titre and fluorescence intensity

From a biological standpoint, the strength of an antibody is dependent upon its affinity and avidity with its target molecule (Tambur et al., 2015). Luminex® SAB assays are marketed and licenced as qualitative assays (McCaughan et al., 2019; Tait et al., 2013; Tambur et al., 2015) but are often referred to as semi-quantitative in clinical practice and the literature, where the mean fluorescence intensity (MFI) value generated from the target microbead(s) is used as a proxy for antibody titre/strength (British Society for Histocompatibility and Immunogenetics (BSHI) and British Transplant Society (BTS), 2014; Courant et al., 2018; Filippone & Farber, 2016; Schinstock et al., 2016; Sullivan et al., 2017; Süsal, Roelen, et al., 2013; Tait et al., 2013; Tambur & Wiebe, 2018). However, the MFI value is a measure of the amount of antibody bound to a microbead and is dependent upon the intrinsic performance of the individual test, and therefore, it cannot be used as an accurate representation of antibody titre (British Society for Histocompatibility and Immunogenetics (BSHI) and British Transplant Society (BTS), 2014; Middleton et al., 2014; Tait et al., 2013). Illustrating this, Tambur and Wiebe (2018) reported that there can be a 10-fold difference in titre for antibodies detected within a range of 1000 to 5000 MFI (Tambur & Wiebe, 2018). Conversely, with cellular crossmatching techniques, it is possible to titrate serum until a negative crossmatch reaction is achieved with cells carrying the target antigen to give an indication of a clinical antibody titre (Filippone & Farber, 2016; Karahan et al., 2017; Tambur et al., 2015).

A further consideration is whether an identified antibody is directed against a structural conformation present on a few microbeads in the assay or on many microbeads. If there are a high number of potential assay targets for the antibody then individual microbead MFI values may be proportionally lower as the antibody is able to bind to multiple targets within the test (Ellis, 2013; Filippone & Farber, 2016; McCaughan et al., 2019; Middleton et al., 2014; Tambur & Wiebe, 2018), potentially leading to unexpected positivity in a cellular crossmatch.

The “positive” cut-off level applied to MFI value differs dramatically between laboratories and some laboratories also apply different numerical cut-offs based on HLA protein group (Sullivan et al., 2017). Inter-laboratory and inter-assay variation for MFI values for the same serum can also be considerable, making comparisons between studies challenging (Filippone & Farber, 2016; Gebel & Bray, 2019; Liwski et al., 2017; Sullivan et al., 2017). Ellis (2013) suggests that Luminex® SAB analysis should use MFI as a starting point rather than a cut-off and that laboratories need to consider any individual microbead performance issues. As such, while MFI may provide a method of trending antibodies over time, it is not recommended that MFI level alone be used as a clinical threshold (Campbell, 2013; Tait et al., 2013).

Interference factors

The Luminex® SAB assays are reported to be particularly susceptible to microbead antibody binding site saturation and to a “prozone” effect attributed to interference from non-HLA factors (e.g., complement proteins and IgM) (Schnaidt et al., 2011; Sullivan et al., 2017; Visentin et al., 2016)). Both phenomena can result in an under-estimation of true antibody “strength” (Guidicelli et al., 2013; McCaughan et al., 2019; Middleton et al., 2014; Tait et al., 2013).

Saturation of SAB microbeads has been reported to occur at ~1:1024 titre, with a linear correlation between IgG-MFI and titre only observed up to ~10,000 MFI (reported as a titre of ~1:32), beyond with correlation between MFI and titre is poor (Tambur & Wiebe, 2018). Additionally, there is a maximum detection limit ~20,000 MFI using the currently ubiquitous Luminex® 100/200™, although the newer Luminex® FLEXMAP 3D platform has been reported to have a greater detection range (up to 400,000 MFI), which may reduce saturation errors in the future (Gebel & Bray, 2019). The relative impact of saturation can only be proven by retesting following dilution of sera (McCaughan et al., 2019; Navas et al., 2019; Tambur & Wiebe, 2018). Similarly, the effects of prozone can be reduced by dilution or pre-treatment of sera using heat inactivation, ethylenediaminetetraacetic acid (EDTA) or dithiothreitol (DTT) to remove blocking factors (i.e., complement and IgM antibody respectively) (Anani et al., 2016; Battle et al., 2017; Courant et al., 2018; Gebel & Bray, 2019; Guidicelli et al., 2013; Middleton et al., 2014).

Tambur et al. (2015) report that in a cohort of 55 sensitized patients, 71% were affected by a prozone effect, while Battle et al. (2017), Battle et al. (2020) demonstrated that differences in the serum dilution ratio recommended by Luminex® SAB assay protocols may alter the relative impact of prozone on different commercial assays. As not all microbeads in an assay will be affected to the same extent by these confounding factors, duplicate testing with and without treatment is required to confirm the magnitude of the effect (Tambur & Wiebe, 2018), but it should be noted that sera dilution and/or pre-treatment can itself be associated with lowering assay sensitivity such that “weak” antibodies may become negative (Clerkin et al., 2017).

Denatured antigen

An individual can develop antibodies that cross-react with HLA molecules without being exposed to non-self-HLA, for example via infection (Gombos et al., 2013; Grenzi et al., 2013). This is thought to be a rare occurrence, but clinical laboratories have reported the presence of high MFI value antibodies directed against self-antigens or unexplained positive specificities in up to 30% of unsensitized patients since the introduction of the Luminex® SAB assays (Wehmeier et al., 2017). Where reactivity cannot be explained by sensitization history or epitope analysis, it is possible that it could be attributed to denatured antigen bound to the microbeads (Gombos et al., 2013; Lachmann et al., 2013; McCaughan et al., 2019; Schinstock et al., 2016; Sullivan et al., 2017).

Denatured HLA molecules are an artefact of the manufacturing process and refer to HLA molecules which are not in their native conformational state (Buttigieg et al., 2019; Grenzi et al., 2013; McCaughan et al., 2019; Ravindranath et al., 2018). It has been estimated that up to 39% of patients will demonstrate antibody binding to at least one denatured antigen (Schinstock et al., 2016). However, the presence of these antibodies is often only suspected in patients with limited antibody profiles due to the challenge of differentiating between reactions against native or denatured molecules in sera with high-panel reactivity. Furthermore, some antibodies can recognize both denatured and native protein structures (Gombos et al., 2013; Middleton et al., 2014; Otten et al., 2013).

The now discontinued “iBead” assay was previously used to demonstrate that antibodies directed solely against denatured antigen are not associated with graft loss/function (Otten et al., 2013). Several protocols have been developed to identify denatured molecules in the clinical laboratory, but as these must be performed independent of routine testing and reevaluated for each assay microbead lot, there is an additional cost implication (El-Awar et al., 2009; Schinstock et al., 2016; Taylor et al., 2017).

Non-HLA-specific binding

In addition to antibodies against denatured antigens, up to 40% of patient sera may also contain antibodies capable of binding to the Luminex® SAB microbeads themselves, exhibiting strong non-HLA-specific binding which can confound antibody analysis (Waterboer et al., 2006). Commercial products are available from each Luminex® SAB supplier for pre-treatment of sera to reduce background reactivity for their assay where this can be identified; Adsorb Out™ (One Lambda) and Serum Cleaner (Immucor).

Antigen density

In 2013, Lachmann et al. reported that HLA-C, HLA-DQ and HLA-DP have a higher relative antigen density on the Luminex®-based SAB microbeads in comparison with cells, with further disparities identified between suppliers and microbead lot numbers (Lachmann et al., 2013; Reed et al., 2013). Some laboratories apply alternative positive cut-off MFI values for different loci (Liu et al., 2012) to account for this, and however, there is an implicit danger of applying this pan-locus approach as extreme variation in antigen density have also been reported to occur within the same locus (Lachmann et al., 2013; Reed et al., 2013).

Summary of Luminex® SAB assay limitations

The two commercial Luminex® SAB assays are reported to be broadly comparable in their sensitivity and specificity for detection of HLA antibodies, although differences exist with regards to their antigen panel composition, relative antigen density, serum to bead mix ratios and analysis algorithms (Clerkin et al., 2017; Singleton & Darke, 2015).

Not unexpectedly, the numerical MFI output value does not consistently correlate between assays and should not be used for quantitative comparison (Karahan et al., 2015; Ravindranath et al., 2018). An increased level of denatured antigen has been reported in the LABScreen assay compared to the LIFECODES (Ravindranath et al., 2018) assay, although conversely LIFECODES has been reported to demonstrate lower sensitivity than LABScreen (Gautier Vargas et al., 2019; Ravindranath et al., 2018).

Despite study methodologies indicating use of the same commercial assay for antibody identification, there are huge variations in protocols reported in these studies that affect the bead volumes, wash volumes, number of washes, pre-treatment of sera, incubation temperatures and the positive analysis MFI cut-off, all of which will ultimately impact assay outcome and interpretation (Liwski et al., 2017).

3.3.2 Clinical relevance of Luminex® SAB assays

Clinical laboratories using SAB assays are now highly adept at characterizing HLA antibodies directed against the proteins produced by all 11 polymorphic HLA loci (HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DRB3, HLA-DRB4, HLA-DRB5, HLA-DQA1, HLA-DQB1, HLA-DPA1 and HLA-DPB1) using these assays. The development of Luminex® SAB assays allowed for the identification of HLA antibodies that could not be detected by cellular assays, and therefore, they have helped to improve antibody definition and reduce the number of unexpected positive crossmatches, with associated improvements in patient and graft outcome measures (South & Grimm, 2016). Historically, solid organ recipients with high-panel reactivity were considered to have a higher immunological risk than other recipients, and however, the introduction of these sensitive SAB assays demonstrated that, in the absence of DSA, there is no increased immunological risk to transplant (Wehmeier et al., 2017).

HLA antigens identified by CDC antibody screening could be listed as unacceptable antigen mismatches with confidence that they would cause a positive cellular crossmatch (Zecher et al., 2018). The improved sensitivity of the Luminex® SAB assays has increased the number of “unacceptable” antigens registered for a patient, with an associated increase in waiting time for transplant (Filippone & Farber, 2016; Süsal, Roelen, et al., 2013; Zecher et al., 2018). However, not all antibodies detected by SAB assays are capable of damaging a donor organ (Filippone & Farber, 2016; Gombos et al., 2013; Salvadé et al., 2016). Although MFI values have been shown to correlate generally with crossmatch outcome (Ellis et al., 2012; Tait et al., 2013), it has also been reported that a high MFI value does not always correlate with a positive cellular crossmatch (Sullivan et al., 2017).

While the comparative clinical significance of HLA antibodies directed against different HLA protein groups remains contentious, there is evidence that antibody directed against any HLA protein can contribute to a positive cellular crossmatch and/or rejection of a transplanted organ (Bosch et al., 2014; Couzi et al., 2011; Filippone & Farber, 2015; Leeaphorn et al., 2018; Picascia et al., 2014). For example, HLA-C is generally recognized as a lower-expressed HLA antigen and Ling et al. (2012) reported that only one of eight patients with an HLA-C DSA demonstrated a positive FCXM (Ling et al., 2012). However, Bosch et al. (2014) reported a case study of AMR attributed to a DSA directed against the 11AVR epitope, which was present on several HLA-C previous transplant mismatches (Bosch et al., 2014). As such, the clinical relevance of Luminex® SAB identified DSA remains unclear, regardless of the HLA protein target.

An additional confounding issue for Luminex® SAB assay interpretation is that the presence of a denatured molecule on a microbead does not automatically preclude the presence of genuine antibodies against the native structure also being present (Otten et al., 2013), but “false” assignment of positivity can lead to an increased registration of unacceptable antigen mismatches and prolonged waiting time (Ravindranath et al., 2018). For patients with a limited antibody profile, concurrent testing by screening/identification assays with native HLA conformations can provide evidence to discount an antibody identified solely by SAB assays, although this is often not possible for highly sensitized patients. To avoid reporting false reactions, it has been recommended that antibodies should be confirmed using an alternative kit or assay (Ellis, 2013; Gombos et al., 2013), and however, this has potential implications for cost and time in a clinical laboratory.

Given the high sensitivity of these SAB assays, we can be confident that a negative DSA result by a Luminex® SAB assay indicates an antibody is not present in the patients’ blood. This has allowed for the effective application of virtual crossmatching (vXM) approaches where the potential donor's HLA type is compared with the patient's HLA antibody profile (British Society for Histocompatibility and Immunogenetics (BSHI) and British Transplant Society (BTS), 2014; Gombos et al., 2013; Zecher et al., 2018). Where DSA is absent, the vXM is negative and the transplant may be able to proceed in the absence of a cellular crossmatch. This approach is highly desirable for transplant teams as it allows for faster assessment of compatibility (donor lymphocytes are not required) and minimizes cold ischaemia time (Roll et al., 2020). However, although this approach has many benefits, studies indicate that 13%–40% of DSA positive patients may be unnecessarily denied a compatible transplant due to a positive vXM assessment where a cellular crossmatch assessment is not performed (Ho et al., 2014; Moszkowska et al., 2014). As such, current approaches for registration of unacceptable antigen mismatches may be too conservative for a significant proportion of highly sensitized patients.

留言 (0)

沒有登入
gif