Poster Abstract

Abstract

Anthony Calvert¹, Anthony Poles¹, Matthew Hopkins¹, Tim Hayes²

¹NHS Blood and Transplant, Filton, UK; ²Manchester University NHS Foundation Trust, UK

Heparin induced thrombocytopenia (HIT) is a rare complication of heparin therapy with mild thrombocytopenia but potentially fatal thrombosis. HIT antibodies target the epitope of platelet factor 4 (PF4) and heparin. Laboratory investigations commonly detect antibodies by ELISA. The British Society for Haematology guidelines suggest clinical significant IgG antibodies equate to an OD > 1.0. Studies show an OD ≥ 1.4 corresponds with ≥50% chance of positive serotonin release assay (SRA) (Warkentin et al. J Thromb Haemost 2008; 6(8):1304-12).

Common practice for a HIT positive patient is repeat testing until a negative result indicates the safe use of Heparin, which is then administered pre-procedure and ceased immediately afterwards. Conditioning includes antibody titre reduction by plasmapheresis, multiple sessions are costly and can delay surgery. PF4/heparin antibodies are detected in 5-22% cardiac surgery patients but only 1%–2% have HIT (Pishko et al. Semin Thromb Hemost 2017; 43(7): 691-698).

A HIT positive pre-transplant patient was tested after each plasmapheresis using ELISA (Immucor HAT45G) and a platelet activation assay (IQ Products HITAlert) to compare HIT antibody levels with ability to activate platelets. Results showed an OD > 1.4 associated with positive activation, comparable to Warkentin's observations. The advantage of HITAlert is it detects functional antibodies and measures the expression of activation markers by flow cytometry, not radiolabelled serotonin release. The assay is then suitable for use in the routine laboratory environment, but requires freshly isolated (<2 h) ABO O platelets.

The patient received a successful heart transplant after the first negative activation result, sooner than waiting for negative ELISA results.

Carla Rosser¹, Deborah Sage¹

¹NHS Blood and Transplant, Tooting, UK

The use of antibody-epitope analysis tools can improve the interpretation of in vitro HLA antibody identification assay data, particularly for identification of false positive reactions. HLA antibody-epitope analysis using HLAMatchmaker was performed on 22 crossmatched sera where donor-specific antibody (DSA) had previously been identified by One Lambda LABScreen™ Single Antigen (LABScreen). These sera were retrospectively tested by Immucor LIFECODES LSA (LIFECODES) and BAG Healthcare HISTO SPOT® (HISTO SPOT) assays.

All three HLA antibody identification assays identified DSA in serum that did not cause a positive flow-cytometric crossmatch (FCXM). Antibody-epitope analysis supports the presence of DSA in 12/22 of these sera, indicating a higher sensitivity of the solid phase assays compared to the FCXM. However, antibody-epitope analysis was unable to explain 13% (5/39) of LABScreen DSA, compared to 5% (1/19) by LIFECODES and 0% (0/9) by HISTO SPOT. In this selected sera cohort (i.e. LABScreen DSA positive), 50% of HISTO SPOT, 33% of LIFECODES assays and 8% by LABScreen correctly predicted a negative FCXM result. All three assays demonstrated 100% concordance for sera with a positive FCXM (10/22).

Disparities between the antibody identification assays, particularly LABScreen and LIFECODES occurred in several samples, where different interpretations of antibody-epitope interactions could explain the antibody reaction pattern. This data illustrates the benefits of verifying antibody assay results where they are being used for clinical risk interpretation. Whilst antibody-epitope analysis is complex and challenging to introduce routinely in a clinical laboratory, it offers the opportunity to develop a better understanding of complex antibody profiles.

Rachel Smith¹, Fiona Powell¹, Betia Nouri¹, Mazen Mabrok¹, Ambika Camille¹, Renuka Palanicawander¹, Eduardo Olavarria¹, Arthi Anand¹

¹Imperial College Healthcare NHS Trust, London, UK

Post transplant HLA typing at the H&I laboratory, Imperial College Healthcare NHS Trust is a new concept driven by the increased numbers of haploidentical transplants being performed and the implementation of Next Generation Sequencing (NGS) HLA typing.

Thirty percent of patients transplanted with a haploidentical donor go on to relapse. Treatment choices differ depending on whether the patient is relapsing with or without HLA loss. NGS HLA typing can help determine if relapse is occurring with HLA loss due to its increased sensitivity. Relapse with HLA loss is defined as the genomic loss or downregulation of the patient's mismatched haplotype in the re-emerging leukaemic cells. This is an attempt by the cancer to evade the graft vs. leukaemic effect (GvL) by donor T cells. Donor lymphocyte infusions (DLI), a common relapse treatment post HPCT is not recommended for those patients experiencing HLA loss due to the increased risk of Graft vs Host Disease without the benefit of the GvL response.

Here we describe two cases where post transplant HLA typing made a significant impact on treatment choice. Both patients experienced relapse following a haploidentical sibling transplant. NGS HLA typing was performed at the point of relapse diagnosis and these results combined with the post transplant chimerism and cytogeneic testing confirmed that one patient was experiencing HLA loss while the other was not. As a result, the patient experiencing relapse with HLA loss went on to have a second haploidentical transplant while DLI's were considered for the patient relapsing without HLA loss.

Anna Barker¹, Laura Ford¹, Stephine Whiteside¹, Poppy Greenaway¹, Rebecca Dench¹, Helena Lee¹

¹Manchester University NHS Foundation Trust, Manchester, UK

Lineage specific chimaerism analysis for engraftment monitoring following stem cell transplantation has greatly increased; by 51% in the past 2 years.

Sample processing is required within 24 h of the blood samples being taken, hence those received on a Friday must be processed the same day. Extending the time to cell separation beyond 24 h was investigated as part of a review of laboratory workflow, with the aim of reducing pressure on staff time and instrumentation.

CD3+ and CD15+ cells were isolated from blood using RoboSep™-S (STEMCELL™ Technologies) automated cell separators. Samples from four patients were each processed at three time points (samples 1–3 at 1, 3 and 5 days, sample 4 at 3, 6, and 7 days). Extracted DNA was tested as for routine chimerism using the GenePrint^® 24 System (Promega). The purity of each isolated cell population was determined using flow cytometry.

There was no detrimental effect of sample age up to 5 days (in one case 7 days) on purity of the cell population, percentage donor chimaerism (Table 1), and electropherogram peak height (all > 3000 RFU).

Although it is best to process samples immediately, workload can often exceed instrument capacity and the time available to process the samples. Quality of the results is not affected up to 5 days, which means that samples received late on a Friday can be processed on Monday, reducing pressure on staff time.

Eva Santos-nunez¹, Katrina Spensley², Corinna Freeman³, Michelle Willicombe², Arthi Anand¹

¹H&I Laboratory Hammersmith Hospital, North West London Pathology hosted by Imperial College NHS Trust, London, UK; ²West London Renal and Transplant Centre, Imperial College NHS Trust, London, UK; ³Clinical Transplant Laboratory (Viapath), Guy's Hospital, London, UK

Introduction: Using algorithms to calculate molecular mismatches requires HLA typing at high resolution or second field (2F). However, this level of resolution is not always available. To overcome this, computer algorithms to infer second field genotypes can be used.

Aim: To determine the accuracy of using a 2F prediction algorithm, Easy-HLA (https://hla.univ-nantes.fr/recherche/recherche.php) and whether discrepancies in results lead to differences in the molecular mismatch scores of two epitope analysis tools.

Methods: HLA types of 28 renal transplant recipients generated by PCR-SSO (n = 23) and PCR-SSP (n = 5) were uploaded to the “HLA-Upgrade tool” of Easy-HLA. Samples were re-typed by NGS and results compared. HLA eplet mismatch load and PIRCHE II scores were calculated using HLAMatchmaker (www.epitopes.net) and PIRCHE-II algorithm (www.pirche.com).

Results: The accuracy of the estimated 2F alleles per locus was HLA-A (93.5%), HLA-B (95.7%), HLA-C (93.5%), HLA-DRB1 (91.3%), HLA-DQB1 (97.8%) for SSO upgrades (E-SSO), and 80.0%, 80.0%, 80.0%, 60.0% and 40.0% for SSP upgrades (E-SSP). Ethnic accuracy differences were also noted; White (97.5%), Black (94.4%), Asian (91.2%) and Other (90.0%).

E-SSO concordance using HLA Matchmaker was 96% and 100% for Class I and II respectively, 60% and 40% for E-SSP. Deviations in PIRCHE II scores were seen in 7/23 (30%) of E-SSO samples (range 0–62) and 3/5 (60%) of E-SSP samples (range 0–95).

Conclusion: In summary, it is possible to successfully infer 2F data from intermediate level typing. The accuracy may be improved by using intermediate (SSO) typing results and statistical and population genetics data provided by algorithms such Easy-HLA.

Emma Holmes¹, Jennifer Gauss¹, Sue Jordan¹, Deborah Sage¹

¹NHS Blood and Transplant, Tooting, UK

Virtual crossmatching is widely used to assess compatibility between a renal transplant recipient and deceased donor to reduce cold ischaemia time. At NHS Blood and Transplant, Tooting, only patients that are HLA antibody negative and have not been previously transplanted are eligible for a virtual crossmatch. Seven hundred and thirty-five deceased donor crossmatches carried out from 2016 to 2020, from which 494 patients were transplanted, were analysed to determine whether the eligibility criteria could be extended to include HLA antibody positive patients that did not have a donor specific antibody (DSA) directed against the prospective donor. Five hundred and two (68%) HLA antibody negative patients were crossmatched, of which 383 were completed by virtual crossmatch. One hundred and twenty-seven (17%) laboratory crossmatches were performed for HLA antibody positive patients in the absence of DSA. All of these were reported as having a standard immunological risk, in line with the BSHI/BTS guidelines, which indicates that laboratory crossmatching did not provide additional information, beyond HLA antibody testing, that could have altered clinical patient management. One hundred and six (14%) laboratory crossmatches were performed for HLA antibody positive patients in the presence of DSA. All three groups include patients that have previously been transplanted. Analysis of clinical rejection episodes reported to NHSBT-OTDT showed that there were similar rates of acute rejection between the three patient groups. In light of this data, the virtual crossmatch policy has been extended to include sensitised patients in the absence of DSA. HLA antibody testing is used to confirm the patient's antibody profile remains unchanged at the time of transplant.

Jennifer Gauss¹, Susan Jordan¹, Deborah Sage¹

¹NHSBT, London, UK

Characterisation of a renal transplant recipient's HLA antibodies is necessary to assess their compatibility with a potential deceased or living donor. Single Antigen Bead (SAB) solid phase assays have simplified antibody analysis, however defining these antibodies by their epitope/eplet targets, rather than by their antigen targets, is considered more clinically relevant. In this case study, we present a 46-year-old, previously transplanted, male renal patient whose serum we tested using the One Lambda Labscreen SAB assay and was found to have antibodies directed at HLA-DP that could not be defined by a DPB1 or DPA1 target alone. Using a combination of the epitope registry and publicly available crystal structures it was determined that DPA1 position 50 (Glutamine,Q or Arginine,R) and DPB1 postion 85–87 (Glutamic Acid-Alanine-Valine (EAV) or Glycine-Proline-Methionine (GPM)) could feasibly constitute a single antibody target. Re-analysis of this patient's SAB profile showed that only the DP antigens that were Q and EAV at the aforementioned positions were positive. HLA-DP typing by NGS showed the patient was Q-GPM. Further HLA antibody testing using Immucor Lifecodes Single Antigen kits and Labscreen Supplementary Class II SAB kits also resulted in DP antibody profiles that could be explained by a Q-EAV epitope target. Absorption of this patient's antibodies only occurred when incubated with a cell carrying Q-EAV DP molecules; no absorption occurred with Q-GPM. Incorporating possible combined alpha-beta chain epitopes into epitope analysis software and listing epitopes, rather than antigens, as unacceptable, will provide a more targeted approach to compatibility assessment in the future.

Sejal Morjaria¹, Arun Gupta¹, Delordson M Kallon¹

¹Clinical Transplantation Laboratory, Barts Health NHS Trust, London, UK

Thymus transplantation is an emerging clinical option for babies born with no thymus, a symptom associated with DiGeorge syndrome (DGS).

Patient AN (3 months, female), diagnosed with DGS in early 2019, received an unconditioned thymus transplant in July 2019 from an unrelated donor at Great Ormond Street Hospital. No prior donor or recipient HLA typing was performed. Three-month biopsies did not detect thymic epithelium and there was no immune reconstitution after 1 year. Pre-transplant serum tested by Luminex single antigen assay showed no pre-existing donor specific antibodies (DSAs). However, by October 2019 the patient had developed de novo DQ7 DSA > 5000 MFI with signs of rejection.

Subsequently, the patient's DSA levels decreased and she was considered for a second transplant. Due to the presence of HLA antibodies, the H&I laboratory advised the clinical team to HLA type potential donors. Donor samples were typed by qPCR and a second thymus transplant was undertaken in January 2022 with cyclosporine induction. The patient was monitored weekly for DSA, developing de novo A24 DSA < 2000 MFI after 2 weeks, increasing to >2000 MFI at 3 weeks. Additional de novo DQA1*05 DSAs were detected at month post-transplant (peak DSA > 10,000 MFI).

AN was treated with rituximab followed by three cycles of plasma exchange with HLA antibody monitoring after each cycle. After three cycles her DSA levels had stabilised around 3000 MFI and the patient continues to be monitored.

This case highlights the role of the H&I laboratory in facilitating thymus transplants and helping to avoid antibody mediated rejection.

Paul Wright¹, Marcus Lowe¹, Robin Kippax¹, Anna Barker¹, Alison Logan¹, Helena Lee¹, Natalia Diaz Burlinson¹, Stephen Sheldon¹, Kay Poulton¹

¹Transplantation Laboratory, Manchester University NHS Foundation Trust, Manchester, UK

HLA allele frequency data is beneficial for identifying patients for which it may be challenging to find a suitable unrelated donor, and in routine analysis of genotyping data when scrutinising results. Despite the many advancements in genotyping the HLA gene complex, with next generation sequencing providing higher resolution typing than previously possible, there are few tools and reference materials available with high resolution data. Coding using Python was developed for extracting HLA allele frequency data from the local next generation sequencing services. Data from 11 HLA loci (HLA-A, -B, -C, -DRB1/3/4/5, -DQA1, -DQB1, -DPA1, -DPB1) were collected from all samples tested using AlloSeq HLA Tx17 next generation sequencing (CareDx, USA) with the MiSeq platform (Illumina, USA) (using the IMGT/HLA v3.43 reference data set) between April and December 2021 (n = 1636). Data for all loci were presented as allele frequency, with the exception of HLA-DRB3/4/5, which were presented as carrier frequency. This data set provides the department with new insights into high resolution HLA allele frequencies, such as the allele frequencies of HLA-DQA1 and -DPA1, and the allele frequencies at third field resolution for various alleles, such as HLA-A*68:01:01 (freq = 0.0122) versus A*68:01:02 (freq = 0.0251) and HLA-B*44:03:01 (freq = 0.0504) versus B*44:03:02 (freq = 0.0067). The next aim of the project is to advance the current data set to allow analysis by patient type (for example, deceased donor or renal transplantation patients) and to develop a data set for assessing HLA linkage disequilibrium frequencies at third field resolution.

Kim McShane¹, Emma Burrows¹, Deborah Pritchard¹, Tracey Rees¹

¹Welsh Transplantation and Immunogenetics Laboratory, Talbot Green, UK

Recent case reports have revealed HNA-3a antibodies are associated with increased antibody-mediated rejection and graft loss in kidney allograft recipients. Ninety-five percent of the UK population is HNA-3a antigen positive.

Two multi-parous female transfused patients presented with unexplained strong positive deceased donor T and B-cell flow cytometry crossmatch (XM). Further investigation confirmed them as HNA-3b/3b with HNA-3a antibodies. Both partners typed as HNA-3a antigen positive, indicating antibodies were likely to be pregnancy stimulated. Patient 1 was deemed unsuitable for transplantation due to co-morbidities. HNA compatible family donors were investigated for patient two with no success. This patient remains on the deceased donor list, but is expected to be FCXM positive with 95% of offers. Live donor transplantation with antibody removal is under consideration.

Due to the risk of HNA antibodies being missed by virtual XM, and reported poorer clinical outcomes, we subsequently screened patients awaiting kidney transplant (n = 298) for HNA antibodies using the LABScreen Multi kit, as this kit is currently used in our laboratory for TRALI risk reduction. Samples with a positive HNA-3 result (n = 22, 7.4%) were referred to a reference laboratory for confirmatory testing. Two (0.67%) were confirmed to be positive for HNA-3a antibodies. False positive rate 9%.

To mitigate the risk to virtual crossmatching, we now HNA antibody screen (once) all kidney transplant patients. Confirmed positives are unsuitable for virtual XM. HNA genotyping as a first line test is being considered to identify the 5% of patients requiring HNA-3 antibody testing.

Blanka Zamostna¹, Cem Mak¹, Jasper Taal¹, Pavel Jiroutek¹, Doug Bost¹

¹JETA Molecular, Utrecht, the Netherlands

Allogeneic hematopoietic stem cell transplantation (HSCT) is an established cure for haematological disorders. Early, accurate, and sensitive determination of hematopoietic chimerism is crucial for the adjustment of a transplant recipient's treatment. Onventional STR and qPCR methods to detect graft failure after HSCT are limited by their dynamic range of analysis (STR-PCR: 5%−100% and qPCR:0,1%−30%). Digital PCR (dPCR) is a 3rd generation technology and has proved in many studies to combine end point quantification over a wide measurement range (0.01%−90%) with excellent sensitivity, accuracy and reproducibility.

We have developed and validated a system of 42 dPCR assays and software enabling highly sensitive stem cell engraftment monitoring. Our approach allows for facile quantification of minute amounts of minor components in a mixture, while reducing workflow and analysis burdens. Each assay was validated using at least three independent control DNA mixtures with concentrations ranging from 90% down to 0.05% minor component in the presence of 50ngs input DNA. The results obtained using dPCR were concordant with the values generated using qPCR.

The use of dPCR allows for sensitive and precise quantification of target DNA and due to its unparalleled sensitivity, this technique represents a particularly useful diagnostic tool in chimerism monitoring. In comparison with the wide used qPCR chimerism analysis, dPCR has been shown to anticipate a relapse significantly earlier and therefore, it may guide timely post-transplantation therapeutic interventions. Compared to NGS methods, this system can deliver 72 answers within one working day, with highly reduced cost per sample and hands on time.

Olivia Shaw¹, Chloe Martin¹, Deeya Balgobin¹, Sarah Blears¹, Corinna Freeman¹

¹Viapath, London, UK

HNA-3 specific antibody may be implicated in rejection episodes following renal transplantation (Key et.al. J Renal Transplant Sci, 2(2):81–84). HNA-3 is bi-allelic, −3a/−3b, and homozygote individuals may develop antibody to the non-self variant.

HNA-3 is expressed on a variety of cell types including lymphocytes, leading to positive T and B cell flow crossmatches, and renal endothelial cells, thus a target for rejection.

Following two local cases of AMR in the face of HNA-3a specific antibody, and cases of positive crossmatches attributable to HNA-3a or -b specific antibody, we have introduced a screening programme for HNA-3 specific antibody using LABScreen Multi (OneLambda).

Between Jan 2020 and May 2022 7360 tests were performed. One hundred and twenty tests in 80 patients were positive for HNA-3a and 102 tests in 77 patients were positive for HNA-3b. ‘True’ positivity was assigned following HNA typing of patients testing positive on multiple samples. This gave confirmed antibody specificity in 20 patients for HNA-3a and 16 patients for HNA-3b, reflecting 7% of the active waiting list.

Patients testing ‘true’ positive require a prospective crossmatch prior to transplant regardless of HLA specific antibody status. Since this introduction an additional six prospective crossmatches have been performed, with 50% giving a positive result. To date no transplant has been refused due to HNA specific antibody.

We report that HNA-3 specific antibody can impact on crossmatch results and routine screening should form part of the immunological risk assessment and guidance given to Clinicians and patients prior to transplant.

Jennifer Lord¹, Rebecca Dench¹, Nicola Martin¹, Paul Wright¹, Anna Barker¹, Alison Logan¹, Helena Lee¹

¹Transplantation Laboratory, Manchester Royal Infirmary, Manchester, UK

Next generation sequencing (NGS) has become the routine methodology to obtain a high resolution HLA type for optimal donor selection in Haematopoietic Progenitor Cell Transplantation (HPCT). NGS provides sequencing beyond the antigen recognition domain (ARD), making it possible to detect HLA mismatches between the recipient and potential donor that are not detected by other HLA typing techniques.

A patient with T cell Acute Lymphoblastic Leukaemia was referred in 2021 for HPCT work-up and HLA typed by NGS (AlloSeq Tx17, CareDx). Two siblings were screened as potential donors by typing at HLA-B using LABType™ SSO (One Lambda). Sibling A appeared to be HLA matched with the patient at intermediate level resolution, whereas sibling B did not. Sibling A was HLA typed by NGS to ensure they were matched at HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1, which showed sibling A had a novel allele at HLA-B and hence a mismatch with the patient that had not been identified by LABType™ SSO. A single nucleotide polymorphism was observed at position 2798 in exon 5 of HLA-B, which excluded HLA-B*51:01:01, 52:01:01 detected by NGS in the patient. Further typing of both parents demonstrated the novel HLA-B allele was not inherited from either parent and may be a spontaneous point mutation in sibling A.

This case demonstrates the importance of high-resolution verses intermediate level HLA typing to optimise HPCT donor selection even in the presence of a family study, as HLA matching outside the ARD may improve transplant outcome.

Fiona Powell¹, Betia Nouri¹, Mazen Mabrok¹, Thet Myint¹, Sarah Blow², Delordson Kallon², Rachel Smith¹, Arthi Anand¹

¹Histocompatibility and Immunogenetics, North West London Pathology, Imperial College Healthcare NHS Trust, Hammersmith Hospital, London, UK; ²Clinical Transplantation Laboratory, Barts Health NHS Trust, London, UK

HLA specific mutations in leukemic cells lines may occur as part of leukaemia immune evasion mechanisms. These may be particularly evident in peripheral blood samples taken during blast crisis and may result in altered expression of HLA proteins. The sensitivity of HLA typing by NGS allows for the detection of these mutations and it is therefore recommended that when detected, they are confirmed in non-disease tissue such as buccal or skin plugs or through the HLA typing of relatives.

Here we describe the case study of an AML patient in which the HLA-A locus exhibited a mutation in exon 1 resulting in a novel null allele (HLA-A*31:01 novel), the other allele being HLA-A*24:02. At the time of detection, the patient was undergoing an unrelated donor search with the initial search being carried out as if the patient were HLA-A*24:02 homozygous.

Due to lack of related donors available to confirm the mutation an alternative source of DNA was sought. A buccal sample was taken but showed leukemic cell infiltration following NGS typing. An attempt was therefore made to use a skin plug sample as a source of non-diseased cells. The skin plug sample also showed contamination with leukaemia cells, though to a lesser degree. This case study highlights the potential risk of donor searches being initiated with the incorrect HLA type even following confirmation of the type using buccal or skin biopsy samples. It demonstrates that these DNA sources may not always be a reliable source of disease-free tissue.

Carla Rosser¹, Deborah Sage¹

¹NHS Blood and Transplant, Tooting, UK

HLA antibody identification is routinely performed in our laboratory using One Lambda LABScreen™ Single Antigen (LABScreen) kits. We identified 54 sera, crossmatched against donor lymphocytes, where donor-specific antibody (DSA) has been detected by LABScreen. Of this cohort, DSA was confirmed by Immucor LIFECODES LSA (LIFECODES) in 40/54 sera. These 40 sera were subsequently tested using Immucor LIFECODES® C3d Detection assay (LIFECODES-C3d) to determine whether assessment of LIFECODES-C3d DSA positivity could improve prediction of flow-cytometric crossmatch (FCXM) outcome in this cohort.

LIFECODES-C3d positive DSA was detected in 60% of sera (24/40). A trend towards C3d-binding DSA against HLA Class II proteins, particularly HLA-DQ, was identified alongside a higher proportion of non-C3d-binding DSA directed against HLA Class I proteins. When assessing C3d-binding capability of individual DSA in conjunction with FCXM outcome, LIFECODES-C3d positivity was seen in 68% of positive crossmatches and only 17% of negative crossmatches, with negative FCXMs attributed to HLA-DQ and HLA-DP DSA only. It has been suggested that complement-fixing capability is associated only with antibodies with higher mean fluorescence intensity (MFI) values but these results indicate that LABScreen MFI value alone could not be used to accurately predict antibody LIFECODES-C3d positivity.

Our results suggest a role for LIFECODES-C3d assay alongside other antibody identification assays, to enable enhanced prediction and risk stratification with regards to the clinical relevance of DSA, compared to the use of HLA antibody identification assays alone and may be a better indicator of HLA antibody clinical relevance than MFI.

Luke Foster¹, David Briggs¹

¹NHS Blood and Transplant, Birmingham, UK

In 2019, the UK National Kidney Allocation Scheme changed to allow unacceptable HLA-DPB1 antigens (UA-DPB1) listed with NHSBT-OTDT to be factored into allocation. The impact of transplanting across HLA-DPB1 donor specific antibodies (DSA) remains unclear. There is a risk of denying access to transplantation due to listed UA-DPB1 that may not produce a positive crossmatch and which may not be clinically significant. However, not listing UA-DPB1 may result in an increased risk of reallocation and thus onward shipping of organs if pre-transplant assessments conclude that the donor is incompatible. This in turn can increase cold ischemia time that is associated with worse transplant outcomes. We aimed to assess the impact of our approach not to list UA-DPB1 by reviewing all lymphocyte crossmatches that were performed between 2018 and 2019.

A total of 537 lymphocyte crossmatches were re-evaluated, considering the DPB1 typing information, DPB antibody history and DSA which impacted upon the decision to proceed to transplant.

Fifteen crossmatches were positive (2.8%), nine (1.7%) indicating an increased but acceptable risk for transplantation and six (1.1%) constituting a contraindication to transplant. All 15 were performed pre-transplant and there were no positive retrospective crossmatches. Five (0.9%) of the positive crossmatches involved DPB1 sensitisation, one of which proceeded to transplant and four which did not, with reasoning being multifactorial. Our findings support our approach of not listing UA-DPB1, but recommend that listing may be appropriate on a case-by-case basis, particularly for patients with consistently high DPB antibody levels that are likely to produce a positive crossmatch.

Selda Goktas¹, Franco Tavarozzi¹, Adam King¹, Molly Green¹, Margaret Walker¹, Momin Shah¹, Reetinder Grewal¹, Michael Hoddinott¹, Sandra Frater¹, Sharon Vivers¹, Lisa Walsh¹

¹Anthony Nolan, Hampstead, UK

The Anthony Nolan Histocompatibility Laboratory utilises GenDx NGSgo® as its primary method for high resolution HLA typing of 11 loci: HLA-A, -B, -C, -DRB1, -DRB3,4,5, -DQB1, -DQA1, -DPA1, -DPB1, including all exons for HLA-DRB1 and whole gene primers for HLA-DQB1.

To manage sample throughput, an automated Next Generation Sequencing (NGS) set up was investigated. This involved a collaboration with Hamilton Robotics and the development of customised methods to process samples utilising the Hamilton Microlab STAR liquid handling robot. The aim of this work was to facilitate larger runs and also to increase the flexibility of the current NGS set up, enabling more combinations of HLA loci to be amplified.

Depending on the number of samples, two automated protocols were developed with Hamilton Robotics (low and high throughput). Low throughput mimics the manual process whereas for higher sample numbers, the high throughput option is more efficient. These options were validated on 16 and 96 samples, respectively.

Data generated using both the Illumina MiSeq and iSeq instruments was analysed using NGSengine (GenDx). All runs met the predefined minimum acceptance criteria, which included specific metrics such as cluster density, Q30 score and lowest read depth.

This collaboration demonstrated that running low and high throughput workflows using the Hamilton Microlab STAR, in conjunction with Illumina MiSeq and iSeq sequencers, resulted in acceptable metrics and 100% HLA concordance. Future work will focus on increased run complexity including varying combinations of HLA loci on the same run.

Sarah Peacock¹, Miriam Manook¹, Vasilis Kosmoliaptsis¹

¹Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

There is a subset of kidney patients difficult to transplant who may be eligible for consideration of modifying their criteria for listing unacceptable mismatches (UMM). A systematic review of records was undertaken to identify patients who had modified criteria applied. This identified 21 patients where listing criteria had been modified with nine transplanted across specificities that had been removed. The average standard cRF was 99.5% (97–100), applying modified criteria reduced the average cRF to 95.4% (67–100). In patients transplanted across modified criteria, in six of these there was one donor specific antibody (DSA) present and in three there were multiple DSA present. In total 14 DSA were crossed with the majority directed against HLA-C. The mean peak MFI crossed was 9157 (2686–21,004) and the mean current MFI crossed was 3800 (614–11,690). The mean follow up time was 562 days (7–1431) with no adverse clinical events for seven of the nine patients transplanted across DSA and at last follow up DSA levels for these patients were MFI < 2000 or absent. One patient, transplanted across multiple DSA, lost their graft after 1 year due to antibody mediated rejection (AMR). Another patient 8 months post-transplant had increased serum creatinine and a biopsy showing Banff 1B acute rejection without features of AMR. Both of these patients had high level (MFI > 5000) DSA present at last follow up. This review demonstrates that for selected patients modified listing criteria can result in successful transplantation but risks of doing so remain.

Sarah Dyer¹, Emma Burrows¹, Deborah Pritchard¹

¹Welsh Blood Service, Ely Valley Road, Talbot Green, UK

We use LABScreen® (One Lambda, Inc) Mixed (LSM) kits to screen patients on the kidney transplant waiting list for HLA antibodies; positive samples are then tested using LABScreen® Single Antigen (LSA) kits. All sera are treated with EDTA.

We identified a male patient, cRF 92% monitored in the laboratory since 2018, that had tested LSM negative with current serum, but had previously been LSA positive for HLA-DQ2 (6 months previously, all five DQ2 beads positive with MFI 4219-17197). All other CII beads are consistently negative. DQ2 is a previous transplant mismatch listed as an unacceptable antigen.

There are four beads in the CII LSM test that carry DQ2 (Table 2). The ratio of these beads ranged from 0.3 to 1.34 (positive > 1.9). Previous LSM testing gave highest ratios of 2.15–5.

LSA testing on the current sera confirmed the DQ2 with MFI 3655-20275. Crossmatching against two DQ2 cells yielded positive results with both historic and current serum (auto xm negative). Lifecodes® Lifescreen beads were positive, and Lifecodes Single Antigen assay confirmed DQ2. These results add weight to the fact this is a LSM ‘missed’ antibody.

To investigate the unexpected negative CII LSM result, further tests were performed on the current serum; dilution testing, DTT treatment, full bead protocol (half-bead utilised routinely), and testing using LSM with IgM conjugate. All tests gave negative results (Table 2).

This case highlights that LSM screening kits may not detect all clinically relevant antibodies, and the importance of analysing samples alongside historic results.

Nikita Sinha¹, Stephen Weston¹, Paul Dunn¹

¹University Hospitals of Leicester, Leicester, UK

It has been well established that anti-HLA donor-specific antibodies (DSAs) are the main cause of antibody-mediated rejection (AMR). There is anecdotal evidence, however, that non-HLA antibodies might be involved in AMR episodes after solid organ transplantation with an increasing number of patients presenting with pathological features of AMR in the absence of anti-HLA DSA. The aim of this study is to establish the role of non-HLA antibodies in patients with suspected AMR, with no anti-HLA DSA, following kidney transplantation. To achieve this, serum samples from the study group (n = 23) were screened using LIFECODES® non-HLA antibody kits, which is a multiplex assay that allows the detection of antibodies against 60 non-HLA markers. Pre- and post-transplant serum samples were chosen to establish any changes in non-HLA antibody profiles following transplantation and whether non-HLA antibodies had a role in graft rejection in these patients. Of the 60 non-HLA markers, a significant difference (p < .05) in pre- and post-transplant mean fluorescence intensity (MFI) values was seen in 15 markers. Lower MFI values were seen post-transplant for all 15 statistically significant markers. Of these, five markers met the suggested cut-off value to be classed as positive – Actin (9/23), Collagen II (13/23), Collagen III (8/23), HARS (6/23) and LGALS3 (6/23). Lower MFI values post-transplant could be due to the effects of immunosuppressive drugs administered to patients following transplantation. Further research is still required. In particular, larger, collaborative studies could aid in establishing the clinical relevance of non-HLAantibodies in solid organ transplantation.

Sarah Peacock¹

¹Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

This patient is a 33 year old female who had previously received a combined liver, pancreas and jejunum transplant 2013; at that time she was highly sensitised (cRF 100%) due to blood transfusions. This transplant crossed multiple HLA class I and II donor specific antibodies (HLA-DSA) (MFI range 2946–21,999) corresponding to a positive T & B cell lymphocytotoxic crossmatch. Her pancreas failed in 2014 and there were episodes of T cell and antibody mediated rejection of the liver in 2020. Her native kidney failed in 2019 leading to dialysis dependence. The plan was to proceed with direct live kidney donation from her father as it would have been challenging to identify a compatible graft given her sensitisation which is predominately directed against HLA-A (she herself is homozygous HLA-A32). Her live kidney donor (LKD) shared repeat HLA-A3, -DQ6 and -DP2 mismatches with her previous transplant with current HLA antibody screening results show she remained highly sensitised including HLA-DSA against HLA-A3 (MFI:523) that had been seen at this low level from approximately 6 weeks post-transplant. Direct transplantation from her LKD proceeded with a current negative T & B cell lymphocytotoxic crossmatch (historical B cell positive). The HLA mismatch grade was 1.1.1 and overall this transplant conferred intermediate immunological risk with close post-transplant monitoring of both the kidney & liver allograft function recommended. There was delayed graft function which recovered by day 9 and 6 months post-transplant the patient remains well at home with no HLA-DSA and no episodes of rejection.

Rhea McArdle^1,2, Rebecca Cope^1,2, Afzal Chaudhry^3,4, Lisa Sharkey⁵, Sarah Peacock¹

¹Tissue Typing Laboratory, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK; ²Faculty of Biology, Medicine and Health, Division of Medical Education, School of Medical Sciences, University of Manchester, Manchester, UK; ³Department of Nephrology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK; ⁴Department of Medicine, University of Cambridge, Addenbrookes Hospital, Cambridge, UK; ⁵Department of Gastroenterology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

Despite recent advances that have improved patient outcomes following intestinal transplantation (ITx), achieving long-term survival and rejection-free survival is still challenging. Understanding the relevance of pre-transplant human leukocyte antigen (HLA) donor specific antibody (DSA) in ITx and the immunomodulatory potential of the liver within the allograft is crucial to providing an accurate assessment of pre-transplant immunological risk, which could influence and improve post-transplant outcomes further. Thus, this was the primary objective of this retrospective study in 95 adult ITx patients transplanted at Cambridge University Hospitals NHS Foundation Trust between 2007 and 2019.

Two novel programs were developed and validated to aid in DSA identification in this dataset. Fifty-four (57%) ITx cases contained a liver, and 28 (29%) harboured pre-transplant DSA. Using the Kaplan-Meier survival method, pre-transplant DSA greater than 500 mean fluorescent intensity (MFI) as identified by Luminex single antigen beads, seemed to negatively affect post-ITx survival and rejection outcomes. Furthermore, liver-inclusive allografts seemed to show resistance to HLA class I DSA. These findings could be clinically important but statistical significance was not achieved via log-rank tests and cox proportional hazard models, indicating a need for future analysis in a larger patient cohort.

Nevertheless, our data hints towards consistency with other ITx studies where deleterious DSA effects have been demonstrated, and where liver inclusion is protective from HLA class I DSA. This is in line with United Kingdom guidelines for immunological risk. We hope that our publicly available research programs developed will support ease of gaining statistically relevant data in the future.

Charlotte A. Cambridge¹, Thomas R. Turner^1,2, Jonathan A.M. Lucas¹, Gabriel J. Benitez¹, Neema P. Mayor^1,2, Steven G.E. Marsh^1,2

¹Anthony Nolan Research Institute, Royal Free Hospital, UK; ²UCL Cancer Institute, Royal Free Campus, UK

PacBio Single Molecule Real-Time (SMRT) technology facilitates full-gene sequencing of complex HLA genes. Several barcoding methods are possible, enabling sample and gene multiplexing: barcoded primer (BCP) or barcoded adapter (BCA), including blunt-ended (BCAv1) or overhang (BCAv2) adapters. Here, we compare data quality of the three methods. Twelve BCP, ten BCAv1 and seven BCAv2 libraries were sequenced on the PacBio Sequel, containing ≤96 samples for ≥2 full-length class II genes: HLA-DRB1, -DQB1 and -DPB1. Lima was used to demultiplex and filter subreads by barcode quality score (pre-filtering = barcode quality scores for all potential reads assigned to a barcode; post-filtering = reads meeting selection criteria). We observed higher mean quality scores for BCAv1 versus BCPs at the pre- (62.6 (n = 1092) vs. 51.7 (n = 1229); P = < .0001) and post-filtering (68.6 (n = 848) vs. 65.9 (n = 666), P = < .0001) stages. BCAv2 had higher mean scores versus BCAv1 at the pre-filtering stage (64.2 (n = 706) vs. 61.6 (n = 1249); P = < .0001), implying that raw subreads generated by BCAv2 are of higher quality. BCAs negate the need for barcoded target-specific PCR primers, resulting in improved amplification success for high quality DNA samples (≤98% BCAs vs. ≤88% BCPs), and can reduce cost per sample versus BCPs as multiplexing increases. Overall, BCAs generate high quality sequence data, with BCAv2 outperforming BCAv1 at the pre-filtering stage. This data will inform optimisation of library preparation and analysis workflows, improving efficiency of ultra-high resolution HLA typing on the SMRT platform.

Dominic Barker¹, Xenia Georgiou¹, Michael A Cooper¹, Thomas R Turner², James Robinson², Steven GE Marsh²

¹Anthony Nolan Research Institute, Hampstead, UK; ²Anthony Nolan Research Institute and UCL Cancer Institute, Hampstead, UK

The IPD-IMGT/HLA Database is the official repository for sequences named by the WHO Nomenclature Committee for Factors of the HLA system. It provides a highly curated dataset of sequences and metadata as a resource to the HLA community. In recent years the advent of Next Generation Sequencing (NGS) has increased the volume and complexity of data being submitted to the database. This technology has also been used to further characterise the International HLA and Immunogenetics Workshop (IHIW) cell lines, increasing their HLA typing to ultra-high resolution. These highly characterised cells are widely used as reference material by the HLA community for the development of new HLA typing methods. In response to the growing size and complexity of the database, IPD recently developed a REST API allowing developers to build tools to query the database directly. Using this API, we have developed a new IHIW Cell Query Tool aimed at improving the accessibility of data on these cells. This tool also allows this data to be downloaded in a variety of formats for offline use. The IHIW Cell Query Tool supports the continued study of the IHIW cells and their importance as a resource to the HLA community.

Amy De'Ath¹, Deborah Pritchard¹, Tracey Rees¹

¹UK NEQAS for H&I, Cardiff, UK

Scheme 3 assesses participants’ ability to correctly determine the specificity of HLA antibodies in 10 samples annually. Reporting of antibodies to HLA-DQA and -DPA is optional and not assessed. An analysis of the data submitted for DQA and DPA in 2020–21 and 2021–22 was performed.

In 2020–21, 45/64 (70%) participants submitted data for DQA antibodies and 24/64 (38%) for DPA antibodies. Three (2%) DQA specificities reached consensus present, 24 (18%) consensus absent, 79 (61%) were reported as negative by all labs and 24 (18%) did not reach consensus. For DPA 0 specificities reached consensus present, eight (11%) consensus absent, 30 (43%) were reported a negative by all labs and 32 (46%) did not reached consensus.

In 2021-22, 25/65 (38%) participants submitted data for DQA antibodies (9/24 37.5% UK&I) and 19/65 (29%) for DPA antibodies (8/24 33.3% UK&I). Eight (6%) DQA specificities reached consensus present, 23 (18%) consensus absent, 56 (43%) were reported as negative by all labs and 43 (33%) did not reach consensus. For DPA 3 (4%) specificities reached consensus present, 0 consensus absent, 63 (90%) were reported a negative by all labs and 4 (6%) did not reached consensus.

The DQA/DPA specificities that reached consensus present were DQA1*04:01, 05:01, 05:03, 05:05, 06:01 and DPA1*02:01, 02:02, 04:01.

In the UK&I 63% laboratories reported considering antibodies to DQA in clinical transplant compatibility assessment whilst 44% consider DPA antibodies. NEQAS encourage all labs to report results for DQA/DPA antibodies with the intention that these antibodies will be assessed in the future.

Thomas R. Turner¹, Matilda C. Tierney², Michael A. Cooper², Victor-Randolph N. Boatey², Neema P. Mayor¹, Steven G.E. Marsh¹

¹Anthony Nolan Research Institute and UCL Cancer Institute, London, UK; ²Anthony Nolan Research Institute, London, UK

During routine, ultra-high resolution HLA typing in our laboratory, we identified three novel intronic variants in a healthy individual, AN300903, who has Efik (Nigeria) and Ashanti (Ghana) ancestry. Full-gene HLA amplicons were generated using PCR with VeriFi polymerase (PCR Biosystems, UK). A sequencing library was made using PacBio ETPK 2.0 and barcoded adaptors, according to manufacturer's instructions, then ran on a Sequel instrument. Data was analysed using an in-house bioinformatics workflow. The extended HLA typing of AN300903 is: A*36:01:01:01, 02:02:01:01, B*53:01:01:01, C*04:01:01:75, E*01:01:01:01, 01:03:02:01, F*01:01:02:11, 01:03:01:03, G*01:04:04, 01:03:01:02, DRA*01:03, 01:02:02var, DRB3*01:62:01:01, 02:02:01var, DRB1*03:02:01, 11:01:02:03, DQA1*01:02:01:03, 04:01:01:04, DQB1*04:02:01:08, 06:02:01:04, DPA1*02:02:02:04, 03:01:01:05, DPB1*01:01:01:02, 105:01:01var. The novel DPB1 allele is likely a recombinant of a DPB1*105:01:01:03 and a DPB1*16:01:01:02/652:01/653:01/1286:01. All novel alleles will be submitted to the ENA and the IPD-IMGT/HLA Database. No family samples were available, so we used haplostats.org to predict the likely haplotypes of AN300903. With 93.6% certainty it predicted A*36:01∼C*04:01∼B*53:01∼DRB1*11:01∼DQB1*06:02 and A*02:02∼C*04:01∼B*53:01∼DRB1*03:02∼DQB1*04:02. These are the second (0.8% frequency) and 176th (0.07% frequency) most common haplotypes in black Africans, respectively. Based on our observations of allele associations in other samples, it's likely the novel DRA and DRB3 alleles are on the same haplotype. We also confirmed the recently identified DRA*01:03 and DRB3*01:62, observed with DRB1*03:02. Ultra-high resolution HLA typing of AN300903 highlights the under-appreciated non-coding diversity in a young, black, British male on the stem cell register. This potentially reflects further diversity in donors of African ancestry, that is only identifiable with extended, full-length HLA sequencing.

Laura Ford¹, Alison Logan¹, Anna Barker¹, Julie Johnson¹, Stephine Whiteside¹, Helena Lee¹

¹Manchester Transplantation Laboratory, Manchester, UK

The Transplantation Laboratory provides a service to clinicians which determines percentage donor chimaerism in a patient post bone marrow transplantation using the GenePrint24 STR methodology (Promega). The Laboratory receives ∼4000 of these samples per year and the current method of analysis and reporting was time-consuming and involved numerous manual transcriptions, with a potential risk of error.

As part of the NHS Long-term Plan, seven Genomic Laboratory Hubs have been generated which will be responsible for coordinating Genetic based work in a particular region of the country. The Laboratory's STR team has been assigned to the North West Genomics Laboratory Hub and as a consequence the turnaround time for reporting the STR chimaerism data may need to be reduced. A new technique to analyse and report STRs that was both robust and quicker to perform was required to meet any shorter turnaround times. As a substitute to expensive commercial software and to improve laboratory efficiency, an inhouse method was developed. A system of Excel spreadsheets was designed, validated and introduced which will automatically calculate percentage donor chimaerism when peak area values from the analysis software Genemapper (Thermofisher) are entered.

This new method saves 23 min per patient (42 min–19 min) for a set of three samples to be analysed, reported and authorised. This time equates to ∼68 working days a year, which frees up staff time for alternative work. Thus, not only is this new method quicker but also improves patient safety by reducing manual transcriptions and potential errors.

Julie Johnson¹, Kay Poulton¹, Lesley Lappin², Paula Ormandy²

¹Transplantation Laboratory, Mancheseter Royal Infirmary, Manchester, UK; ²University of Salford, Salford, UK

Accreditation is a procedure by which an authoritative body gives formal recognition that a laboratory is competent to carry out procedures according to specified standards. The rationale being that if these standards are met there is a level of assurance that the service provided is of an acceptable level of quality. Throughout the NHS, laboratories have been encouraged to implement accreditation as a tool with which to demonstrate an acceptable level of service quality. However, evidence to substantiate there is any true benefit of accreditation is lacking.

The current process for a laboratory to obtain accreditation is time consuming, bureaucratic, and costly, with demands on the laboratory which can delay innovation and improvements to the patient service. In European medical laboratories, an alternative approach has been adopted successfully. This alternative approach uses a Flexible rather than a Fixed scope of Practice which may benefit experienced laboratories, allowing autonomy within their accredited scope to remain patient focused and to adapt to innovation and science in a timely cost-effective manner. This approach has not yet been fully explored within NHS based hospital laboratories.

This single centre study aims to analyse the implementation of the Flexible scope accreditation using a quasi-experimental design with a mixed methods approach for data collection. Longitudinal data will be collected to evaluate the implementation of the Flexible Scope using a retrospective-prospective study design to validate and verify the overall impact of implementing a Flexible scope on an NHS laboratory service.