key: cord-0042185-sqi25kkz
authors: nan
title: Poster Presentations
date: 2018-04-18
journal: HLA
DOI: 10.1111/tan.13251
sha: fdb47e4747d71e18778d527bb91c3197a3a13777
doc_id: 42185
cord_uid: sqi25kkz

nan

In recent years, routine use of high-sensitive alloantibody detection assays clarified the post-transplant detection of serum donor-specific anti-HLA antibodies (sDSA) as the main cause leading to tissue damage and renal transplant failure. However, the actual graft-damaging potential of sDSAs is not yet completely clarified, as a proportion of patients developing sDSAs do not show relevant graft damage even in long-term follow-up. Furthermore, recent data obtained in adult kidney graft recipients indicated that DSA graft homing should be considered as a promising biomarker for antibody-mediated graft lesions and worse graft outcome. We analyzed graft biopsy specimens and posttransplant sera from 4 patients with a suspected graft rejection. Graft rejection was defined as a sudden and persistent serum creatinine increase, with exclusion of other major causes (infection or dehydration), significant proteinuria development and post-transplant detection of dn sDSA. Patient sera were studied with Luminex (standard and C1Q test). Three paediatric patients and one adult were considered; two were at first transplant, and two re-transplanted. One of the two patients enrolled for the first transplant received the kidney from a living donor. Pre-transplant serum antibodies screening were negative in these patients, positive in the other two (PRA 80%). One of the retransplanted patients received a kidney with 3 HLA A, B, DR mismatches and the other had one repeated mismatch from the first donor. The two re-transplanted patients did not show evidence of DSA in post-transplant sera, on the contrary we detected the presence of DSA in their biopsy specimens (A68, C16 and DP3, DQA1*02:01). The patient who received the kidney from a living donor did not show evidence of DSA in the sera nor in the biopsy specimens. The second patients at the first transplant developed de novo post-transplant sDSA but no antibodies were detected in the graft biopsy specimen. Kidney graft was successful for the two patients at the first transplant in one-year follow-up, on the contrary the two re-transplanted patients were explanted for a severe humoral rejection soon after the transplant. Our preliminary data show that the presence of intra-graft DSA is an unfavorable prognostic factor for early transplant outcome, even in the absence of sDSA.

2013 in a single-center. The variants of the FcGR3A gene were identified by the SNAP SHOT method. The level of CD16 Fc receptor engagement expression was assessed by flow cytometry analysis of alloantibody-induced CD16. Down regulation was observed within CD3-CD56 peripheral blood natural killer (NK) effector cells exposed to allogeneic target cells in the presence of donor specific antibodies (DSA). Thirty-nine recipients developed CLAD, eight with a restrictive allograft syndrome and 31 with a bronchiolitis obliterans syndrome; median survival was 35 months. DSA were detected in 31% of cases at M1 and in 18% of cases at M3. C1q bound 52% and 33% of DSA at M1 and M3, respectively. DSA at M3 were associated with a lower survival (p = 0.02) and a higher occurrence of CLAD (p = 0.01), independently of C1q binding. The [158F/V] FcGR3a polymorphism had no impact on functional outcome of the allograft. Interestingly, and in contrast to our previous finding in the kidney and heart transplantation setting, DSA-mediated engagement of CD16 was very low, whatever the MFI intensity, the C1q binding or functional outcome. Our data indicate that the [158F/V] FcGR3A polymorphism has no major impact on CLAD occurrence and overall survival of LTxR, suggesting that in the LTx setting, specific mechanisms of DSA dependent pathogenicity, which cannot be fully explained by the complement or CD16 ligation of the Fc fragment of DSA, may be at play in the development of CLAD. These results prompt further study on larger cohorts of LTxR Kidney transplantation represents the best care for patients with severe and irreversible renal failure and provides more treatment opportunities than dialysis. In Sicily, overall, there are 541 patients awaiting transplantation (September 2017), divided over three regional Transplant Centers (CT): n=209 at Policlinico Catania, n=173 at ISMETT and n=159 at ARNAS Civico. The number of patients on the waiting list with a high-risk immunological profile and with a positive PRA is well above the national average (57% vs 25%) and, of these, around 14% are to be considered hyperimmunized (PRA>80%); this causes the growth of the list and the extension of the waiting times. Most of these patients are not included in the National Hyper-immune Program (PNI) criteria; furthermore, the crossover program till now is poorly used. According to the national guidelines issued by the National Transplant Center (CNT), "Regional Transplant Centers (CRT) will have to assign kidneys taken in the region by selecting the recipients on a single regional list, this meaning the group of patients on the waiting list, regardless of the chosen CT. Sub-lists are not allowed for the selection of the recipients". In our region, before 1 October 2016, the effective date of the single regional list, the kidneys were assigned according to a specific shift and each CT was authorized to choose between a group of 5 patients with their own selection criteria, determining a different access to care and less patient protection. With the single regional list, however, the same criteria are applied for all centers and it has made it possible to guarantee a better preliminary assessment for hyperimmunized patients and, above all, it guarantees all patients a fairer access to organ allocation. In our opinion it represents the necessary basis for the activation of a "Hyper-Regional Regional Protocol" with less restrictive access criteria compared to the National Protocol, shared with all the regional CTs. exhaustion of the vascular bed and impossibility to perform peritoneal dialysis. Unfortunately, all the crossmatches performed with all cadaveric donor available in Sicily resulted positive and the patient was excluded. In 2014, because the worsening of her clinical conditions, the patient performed a series of immuno-adsorption treatments and obtained a favorable opinion for her inclusion onto the nationwide hyperimmune programme (NHP): all available kidneys in Italy are primarily proposed to highly sensitized patients with a panel reactive antibody above 80%. Finally, in November 2017, a negative CDC-XM was obtained despite the patient showing two DSA: anti-B39 and A3 with MFI respectively 4000 and 3000; the transplantation was performed. At the time of transplant, the FCXM was negative; four post-transplant sera showed a PRA of 50% for HLA class I and the DSA disappeared. Biopsy performed two months after transplantation, was identical to the one performed at the time of transplantation and no damage due to humoral rejection was detected, but only cytotoxic damage, probably due to pharmacological treatment. Every clinical evaluation is premature, but this case represents a positive outcome of the application of the single regional list. The omission of a prospective crossmatch (XM) for deceased donor kidney recipients negative for HLA antibodies (Abs) is standard practice in numerous centers. In our unit many sensitized patients also proceed to transplant (Tx) without a prospective XM when a day of Tx Ab sample has been tested by single antigen beads and a virtual XM review has been performed by a senior H&I scientist. Results on the current HLA Ab status are generated before donor XM material is available. This approach is supported by clinical staff as it reduces cold ischemia time (CIT). To assess XM omission risk we investigated the incidence of early AMR (histological evidence of microcirculation inflammation +/-C4d staining within 3 months of Tx) and 1 year graft loss attributed to rejection in three renal Tx cohorts, transplanted between January 2010 and November 2016; 155 patients negative for HLA Abs who proceeded without a XM (Group 1), 41 patients positive for HLA Abs who proceeded without a XM (Group 2) and 54 patients who had a prospective XM (Group 3). Overall 78% of patients received a Tx in the absence of a prospective XM; median CIT was 11.1hr compared to 14.2hr for XM patients (p=0.011). The incidence of AMR was 1/155 (0.6%) in Group 1, 2/41 (4.9%) in Group 2 and 3/54 (5.6%) in Group 3 (p>0.05). For the two AMR cases in Group 2 both had a retrospective negative Flow XM. One of the three AMR cases in Group 3 had known HLA donor specific Abs (DSA) at the time of Tx. There was no difference in 1-year graft loss between the three groups and no grafts were lost to rejection. A high rate of omission of the prospective XM can be achieved in well characterized sensitized patients with up to date Ab screening results without an increased risk of AMR. As donor HLA typing and HLA Ab definition continue to improve only those patients with an identified HLA DSA should require a prospective XM to assess risk. In 2017 in our center, 95% of deceased donor kidney Txs proceeded without a prospective XM.

Complement fixing anti-HLA-antibodies (Abs) in the serum of transplant (Tx) recipients have a negative impact on graft survival. Little is known about the influence of the noncomplement fixing IgA Ab isotype. Previously we reported a high prevalence of IgA Abs in serum of kidney re-Tx patients leading us to investigate the donor specificity and the effect of these IgA isotype on graft survival. 694 kidney re-Tx candidates from a retrospective multicenter cohort were selected. Importantly, the time to first dialysis after Tx (TTD) -measured in months (mo) -was used as a precise endpoint marker for loss of graft function. IgG-and IgA Abs were investigated in post-Tx sera, using Luminex microbead-based screening-and subsequently specificationassays. Of the 694 sera, 214 were tested positive for IgA and 518 for IgG Abs. Interestingly, the presence of IgA was highly linked to the presence of IgG (odds ratio 2.8, p<0.0001). As expected, stratification by IgA-and IgG status revealed the highest median graft survival in IgA/IgG double negative patients (TTD 118.5 mo) . Patients tested positive for either IgA or IgG Abs displayed intermediate graft survival (TTD 93 .7 mo and 101.5 mo, respectively), whereas double positive patients had a significantly reduced median graft survival compared to any other subgroup (TTD 77.1 mo, p<0.05) . Importantly, the presence of donor specific IgA (IgA-DSA) in serum (n=96) correlated with a significant worse outcome (TTD 76.0 mo) compared to the total cohort (TTD 102.3 mo, p<0 .0001), as well as compared to IgA positive patients without showing IgA-DSA (TTD 87.1 mo, p=0.006) . Presence of serum anti HLA-IgA Abs in conjunction with anti-HLA IgG Abs marks a highrisk subgroup of kidney re-Tx candidates with reduced graft survival. Notably, the detrimental impact of IgA-DSA among total anti-HLA IgA may indicate a novel functional contribution of these anti-HLA IgA Abs to graft failure. Rituximab is commonly prescribed in ABO or HLAincompatible kidney transplantation. When used before transplant, it interferes with B lymphocyte complementdependent crossmatch (CDCXM) or FACS crossmatch (FCXM) giving false positive results rendering their interpretation impossible. We present here an adaptation of the CDC or FC-XM in order to abolish this effect on the B lymphocytes by incubating the patient's sera with an anti-Rituximab mouse monoclonal antibody (10C5 clone, ABNOVA) (A-R Mab). This easy, efficient and relatively inexpensive method eliminates Rituximab interference on both CDC and FACS XM. We first determined the smallest quantity of A-R Mab needed. Five patients with and without class I and II HLA antibodies were chosen, all of them received Rituximab (10 to 42 days before the XM). For CDC-XM, four different A-R Mab concentrations (only 2 in FC-XM) were tested: 0.1 g/l, 0.17 g/l, 0.33 g/l, 0.5 g/l A-R Mab incubated with the patient's serum during 15 minutes at room temperature. Then the classical cross match was done with patient's sera containing Rituximab and with the same sera blocked by this A-R Mab. This concentration was then tested in CDC and FCXM on 17 sera from 11 patients desensitized by Rituximab for ABO or HLA incompatibility to confirm the results. A concentration of 0.17g/l of the A-R Mab abolished all false positive results in both CDC and FACS XM. All expected negative XM were found negative in patients without anti HLA antibodies. While this concentration of A-R Mab dilutes the serum it did not cancel expected positive results. CDC and FACS XM remained positive in sera with donor specific HLA antibodies above

The improvement in the definition of anti-HLA patterns after Luminex-assay implementation is clear. This success has permitted the development of highly-sensitized patients allocation and donor-pair exchange programs improving the access to transplantation. However, the incidence of non-HLA antibodies in highly-sensitized patients is unclear. The aim of the work was to assess the concomitant presence of serum anti-HLA and non-HLA antibodies in highly-sensitized patients. A total of 21 highly-sensitized patients awaiting for kidney transplantation were studied. The anti-HLA antibody profile was confirmed in serum by Single Antigen (One Lambda) and the reaction against non-HLA antigens by LABScreen Autoantibody class I and II (One Lambda) was performed. Upon anti-HLA analysis, all patients had a panel reactive of antigens >98%. Moreover, all of them presented concomitant anti-HLA and non-HLA reactions. Within the autoantigens tested, the most frequent reaction against non-HLA was anti-glutathione-s transferase theta-1 (GSTT-1) (61%) and anti-C-terminal fragment of perlecan (LG3) in 52% of the patients included in the study. The incidence of non-HLA antibodies in patients on waiting list was largely unexplored. The concomitant anti-HLA and non-HLA antibody reactions in highly-sensitized patients is extremely high, and several cases of vascular rejection after kidney transplantation have been described despite a negative virtual crossmatch test against HLA antigens (1). The screening of non-HLA antibodies in highly-sensitized patients could allow us to identify patients with increased risk of vascular rejection after kidney transplantation.

to living donation between a donor and recipient beside the ABO incompatibility are the presence of HLA antibodies and/or the presence of a positive crossmatch. Different methodologies were developed to overcome these immunological barriers and match the highly increasing number of patients with end-stage renal disease in Saudi Arabia, which currently ranges between 120 to 150 patients per million per annum. At our hospital, King Faisal Specialist Hospital Jeddah, we therefore successfully established the desensitization and ABOi renal transplantation program. Recently we have also started to use a kidney exchange program (KEP) . We hereby report about our experience to overcome the immunological hurdles by using KEP. We used an extensive computerized donor exchange program with high grade flexibility by performing different options including simple two-pair exchanges, more complicated domino exchanges and chain donations. All recipients with willing donors but with high titer HLA antibodies and/or strong positive cross match were included in the program. HLA typing was performed by using One Lambda sequence specific oligonucleotide reverse and sequence specific primers (SSOr/SSP). HLA Antibody Identification, Single Antigen Class I/II. T and B cell IgG XM is performed by FACSCanto II flow cytometer, where the cut-off for positive XM is determined based on normal human studies. Twenty patients were transplanted successfully by using the KEP within a period of 12 months only. Considering a total number of 220 transplanted kidneys at this time, which means we have created the largest center for renal transplantation in Saudi Arabia till now, we were able to increase the number of transplanted organs significantly by 9% and we continue in extending and optimizing our histocompatibility services at the hospital. Our experience demonstrates the high efficiency of the donor exchange program in overcoming immunological barriers in renal transplantation. This program provides in some cases also an alternative to the invasive and costly desensitization protocols. Further efforts and thoughts are going on, to extend this successful program to include diseased donors also. To match the increasing demand on renal transplantation (RTX) at our hospital, beside invasive technologies, like desensitization and ABOi RTX we also established noninvasive strategies, like the Kidney Pair Exchange (KPE) resulting in an increase of RTX and reduction in the waiting list time. However, to avoid immunological complications related to HLA incompatibilities all invasive and noninvasive procedures must be supported by outstanding HLA laboratory. The reduction of immunological obstacles and the evaluation of their residual risks require knowledge and special skills in the nature and immunological mechanisms of involved HLA-antibodies. To support these critical responsibilities of our staff, we still successfully use the CDC test beside other advanced technologies like flow cytometry for crossmatch and Luminex platform for detection of antibodies and C1q test. Here, we report about two KPE cases in which we have used CDC as an additional test along with flow crossmatch assay. The first case is of a 52-year-old female patient with multiple non-DSA and high titer DSA (DRB1*07:01 13964 MFI, DRB4*01 4718 MFI, DPB1*02:01 9770 MFI), negative T-Cell flow crossmatch and positive B-Cell flow crossmatch (FXM 216 MCS) . To evaluate the clinical relevance of these DSA and crossmatch results, we performed CDC assay for B cells which was clearly negative. With acquisition of this result along with all the above-mentioned laboratory findings and since the multi-sensitized patient has no other suitable donor, hence we took a cumulative decision to perform the RTX. The donated organ functioned immediately without any sign of rejection. Now three months after transplantation the patient is with normal renal retention parameters. The second KPE case was 48-year old male patient with multiple non-DSA and DSA (A*11 7524 MFI, A*30 7311 MFI, A*08 5974 MFI, B*52 10084 MFI, DQ5 16746 MFI), strong positive T-Cell flow crossmatch (FXM 492 MCS) and strong positive B-Cell flow crossmatch (FXM 506 MCS) . CDC assay in both T and B cells showed strong reactions even after 1:8 serum dilutions, so we were not able to initiate desensitization procedure and this case is still pending on the waiting list. These two cases demonstrated that CDC could support and enhance our decisions to evaluate the immunological risks for the patients in the KPE setting.

IL-17A and IL-17F are pro-inflammatory cytokines produced by activated CD4+ cells, for which receptors are present on different cell types, with specific expression on cells of the spleen and kidney. These proteins are mediators of the inflammatory responses, following the activation of T cells and are associated with certain inflammatory diseases and transplant rejection reactions. The present study aimed to assess the relationships between polymorphisms within genes coding for IL-17A and IL-17F, and their receptors as well as serum levels of both cytokines in dialyzed patients. For this purpose, patients (n=109) and controls (n=125) were genotyped for IL17A (rs2275913), IL17F (rs763780), IL17RA (rs4819554) and IL17RC (rs708567) alleles using LightSNiP assays. Levels of both cytokinesIL-17A and IL-17F in the patients' serum samples were assessed by ELISA. It was observed that patients more frequently presented with the G allele of the IL17F (rs763780; A7488G) polymorphism as compared to controls (OR=22.44, p<0.001). Furthermore, the presence of the G allele was associated (p=0.016) with higher IL-17F serum levels (> 6 pg/ml). No significant differences were observed while analyzing polymorphism and expression of the IL-17A encoding gene. There was no significant difference in the distribution of the IL17RA and IL17RC genotypes when dialyzed patients and controls were compared. These results showed that patients awaiting kidney transplantation and healthy controls differ with respect to the distribution of the IL17F genetic variants and that IL17F polymorphism correlates with IL-17F protein expression. TGF-β has been known to act as a potent immuneregulatory cytokine, which blocks T cell activation. It is considered as a potential target for more specific and less toxic immunosuppression and control of allo-immune responses over the long term in solid organ transplantation (SOT). Yet, recent studies have revealed that TGF-β also has proinflammatory functions playing a critical role developing an effector immune response. In this setting, TGF-β main function after SOT remains yet unclear and is important to further characterize its role under different clinical circumstances not only for operational tolerance, but also amongst other comorbidities such as opportunistic infection (OI) . The aim of this study was to assess the concentration of TGF-β in the supernatant of stimulated whole peripheral blood (WPB) in a cohort of SOT recipients (SOTr) and correlate it with the primary study outcome which was the occurrence of OI. Thirty liver (LTr) and 31 kidney (KTr) transplant recipients as well as 15 healthy volunteers (HCs) were recruited from 2010-2012 and prospective monitored for one year in our Immunology Service in the University Clinic Virgen de la Arrixaca Hospital in Murcia, in the southeast of Spain. Enzyme-Linked Immune Assay (ELISA) was carried out to calculate TGF-β concentration after WPB culture with Concanavalin A (ConA) for 72 hours. Prior transplantation, SOTr showed higher TGF-β concentration compare to HCs. The stratification analysis showed that TGF-β was significantly higher in patients with OI within the first six months following transplantation. A TGF-β >363.25 ng/ml in LTr and >808.51 ng/ml in KTr were shown to be the most accurate cut-off values to stratify SOTr at high risk of OI. The regression model confirmed this biomarker as the main recipient risk factor for developing OI. Our data show that the quantification of TGF-β could provide valuable information as to the occurrence of opportunistic infection in liver and kidney transplant recipients.

antigen bead assay which assigns antibodies at an allelic level, it has become necessary that HLA typing for potential solid organ recipients be done at high resolution. However, in a developing country such as ours easy accessibility and affordability to high resolution typing still remains a challenge. Against this background, we aimed to study how intermediate resolution typing results compared with high resolution typing results. Luminex based tests for HLA typing used in our institution offers an intermediate level resolution and summarizes the most likely allele based on a particular probe hit pattern. High resolution typing results are available for the group of samples tested in our laboratory which have been sent to registries overseas or external quality assurance samples received. Over a three-year period, 160 samples which had both intermediate and the high-resolution data were used for this study where the high-resolution typing was compared to the most probable allele assigned as per the intermediate resolution kit. Concordance was assessed for the HLA-A, -B, -DRB1 and -DQB1 loci. Of the 160 samples tested, there was an overall concordance for 159 samples. The concordance rate locus wise was 100% for HLA-A, -DRB1 and -DQB1 loci and 99.4% for B locus respectively. The disparity was seen in one sample for B locus and the disparity noted was HLA-B*15:78 which on the intermediate resolution kit had the first probability assigned as B*15:01:01. While the concordance was greater than 99%, larger numbers of samples need to be studied. However, in a developing country such as ours, until high resolution typing becomes more easily accessible and affordable, use of intermediate resolution typing might offer an interim feasible solution.

Anastasiya Mihaylova 1 , Petiya Yankova 1 , Anzhela Antonova 2 , Nevena Gesheva 2 , Valentina Atanasova 2 , Milena Ivanova-Shivarova 1 , Elissaveta Naumova 1 1 University Hospital Alexandrovska, Medical University, Sofia, Bulgaria, 2 University Hospital Alexandrovska, Sofia, Bulgaria Correspondence: mihaylova_ap@abv.bg

The importance of the pre-transplant crossmatch test for final confirmation of compatibility between the donor and the recipient and for proper decision making is undoubted. The regular assessment of the sensitization status of the recipients in the waiting list is an important prerequisite for assessing the probability of a positive/negative crossmatch (virtual XM). However, unexpected positive prospective crossmatches may be observed, the clinical significance of which can be questionable. This study aims to discuss unanticipated flow cytometry crossmatch (FCXM) results observed in our transplant practice. A total of 884 FCXM performed between January 2010 and November 2017 with 101 potential deceased and 148 living donors for kidney transplantation were evaluated. A standard three-color FCXM was performed. HLA antibodies were screened and identified with bead-based assays (Luminex and/or Flow-PRA). Nineteen (2.58%) of 737 tested patients were FCXM positive: three (0.41%) patients had a T-cell(+)/B-cell(+), five (0.68%) had a T-cell(+)/B-cell(-), and 11 (1.5%) had a T-cell(-)/B-cell(+) FCXM. Our analyses showed that the Tcell(+)/B-cell(+) FCXM were due to donor specific alloantibodies (DSA). Based on the auto-FCXM results and alloantibody testing of the T-cell(-)/B-cell(+) patients, autoimmune reactivity was detected in six of them. No positive auto-FCXM or DSA were found in the remaining five patients. HLA class I DSA were detected in the patients with T-cell(+)/B-cell(-) FCXM mainly against HLA-B donors MM antigens. Thus, less than half of the unexpected prospective positive FCXM may be attributed to HLA DSA. According to our policy a positive FCXM (T and/or B) is a contraindication for transplantation. Therefore, the reasons for an unanticipated positive crossmatch should be elucidated before transplant to ensure efficient organ allocation and optimize patient outcome. Immunogenicity of HLA antigens is related with molecular differences between immunizing and responder antigens but probably also with the individual ability to mount a humoral response against these differences that may be related with the generation of DR restricted T-cell epitopes. A computer algorithm (HLAMatchmaker) is used to determine HLA matching based on the number of exposed molecular mismatches in the form of eplets. The aim of this work was to analyze the incidence of HLA antibodies as a function of both the number of non-self-peptides that could be theoretically generated from the immunizing antigens, independently of their location in the molecule, and the theoretical affinity of them for the HLA-DRB1 of the responder (measured as a percentile rank). The study was performed in two contexts: pregnancy-induced immunization and kidney transplantation. The number and affinity of non-selfpeptides were determined by using an internet resource (www.IEDB.org) coupled with an in-house algorithm. In addition, the number of eplets was determined for each allogeneic HLA using HLAMatchmaker. Mismatched HLA antigens were classified as immunogenic or nonimmunogenic depending on whether antibodies could be detected or not. In the ever-pregnant group, comparisons between immunogenic and non-immunogenic alloantigens showed highly significant differences for both the number of non-self-peptides (p<0.0001) and the number of eplets (p = 0.001). In the immunization by transplantation group significant differences for the number of non-self-peptides (p = 0.016) but not significant for the number of eplets (p = 0.796) were shown. In both groups a lower percentile rank value is associated with the development of HLA antibodies. In conclusion, our data suggest that the likelihood of developing HLA antibodies increases with greater differences at the molecular level including exposed and nonexposed residues which could form T-cell epitopes. Virtual crossmatching (v-XM) refers to the comparison of the anti-HLA antibodies of the recipient, as defined by Luminex Single antigen (LSA) bead testing, with the HLA of the donor, providing reasonable prediction of the result of the actual cross-match (a-XM) performed with the use of complement-dependent cytotoxicity (CDC) method with and without addition of dithiothreitol (DTT). The aim of this study was to analyze the correlation between negative v-XM and a-XM performed in three years period (2014) (2015) (2016) in a cadaver donor kidney transplantation programme. In carrying out v-XM, the donor HLA molecules against which patients showed LSA-detected antibodies with normalized mean fluorescence intensity (MFI) ≥1500 for HLA-A, -B, -DRB1, -DQB1 and MFI≥5000 for HLA-C were considered to be unacceptable antigens. All cadaver donors and patients had been typed for HLA-A, -B, -DR, and -DQB by PCR sequence-specific primer methods. In Eurotransplant allocation system, XM lists are produced for selection of potential kidney transplant recipients, only including patients for which there is a prediction of the negative result of a-XM. On the basis of a negative v-XM, a-XMs were performed in kidney allocation procedure from 369 cadaver donors. In total we have performed 2309 CDC only and 166 CDC and CDC/DTT a-XMs. Positive a-XM has been obtained for 82 (2.8%) sera samples from 62 patients, 16 patients being always historically PRA < 5%. A total of 20 (2.3%) patients had a positive a-XM with more than one cadaveric donor. The analysis of positive a-XM revealed that the possible reasons for a-XMs positivity were mostly due to the presence of IgM antibodies and non-HLA antibodies, but also to unreported transfusions leading to non-detected unacceptable antigens prior to offer. Overall, the negative v-XM showed 96.7% correlation with CDC XM, showing high sensitivity in predicting donor-recipient immunologic compatibility. In solid organ transplantation the presence of donor specific antibodies (DSA) directed against HLA antigens of the recipient indicates shorter graft survival. However, the clinical significance varies according to the sensitivity: cytotoxic tests are less sensitive than bead array tests but have a higher clinical impact. Flow-cytometric crossmatches are in between. We studied 478 cases of solid organ transplantation using bead array techniques for the detection of HLA specific antibodies, mean fluorescence intensities (MFI) >1000 are considered as positive. We also performed a flow cytometric crossmatch using pronase treated B and T lymphocytes. Signal detection was performed in the linear mode, a fluorescence shift of six was deemed positive. In order to correlate an MFI of the bead array test with a positive FACS crossmatch Receiver Operating Characteristic (ROC) curve analyses were performed. An AUC (area under the curve) value of 0.875 was observed when we compared DSA positivity (defined as the sum of the MFI of all class I and class II antibodies) with FACS positivity (defined as positive T and/or B cell crossmatch). This value indicated a good accuracy of the tests. Using this curve, we found an optimal threshold of 5000 MFI. A contingency table was constructed with this threshold and analyzed by Fishers Exact Test (p<0.00001). Out of the 478 tests, 427 were true negative and 26 true positive; 15 tests were false negative (DSA positive, FACS negative), 10 false positive (DSA negative, FACS positive). The false positive cases all were weak positive in the FACS crossmatch. A higher cut-off could eliminate them. The false negative cases could be due to lower expression levels of alleles or due to limitations of the bead array test. We found evidence of possible low expressed alleles in some cases. In conclusion, we found a good correlation between MFI of the bead array test and outcome of a FACS crossmatch. The relevance of a positive complement-dependent lymphocytotoxicity crossmatch (CDC-XM) in the hyper acute rejection after kidney transplantation has been known for more than 40 years. The introduction of more sensitive technologies such as Luminex has provided alternative methods for HLA donor specific antibody (DSA) detection. It may be interesting to know what levels of fluorescence correspond with a positive CDC-XM especially in cases such as cardiac transplantation in which the decision of accepting an organ for an immunized recipient cannot be delayed. This work aimed to establish what anti-HLA DSA fluorescence intensity (MFI) levels correspond with a positive CDC-XM. We used a total of 33 sera with a positive virtual XM against HLA class I molecules (DSA-I) and 20 sera with a positive virtual XM only against HLA class II molecules (DSA-II). The criteria to consider a DSA-I and/or DSA-II as positive was that patient serum had a MFI value ≥1,500 against at least one of the class I (A, B or C) and/or class II (DRB1, DQA1 or DQB1) molecules of its donor. CDC-XM with T and B cells were performed for DSA-I and DSA-II sera respectively. The Mann-Whitney test was used to compare the median distributions. The MFI median values of the positive and negative CDC-XM were statistically significant different. In the case of the anti-HLA class I antibodies, the median values were 4,750 (2,000-14,000) and 11,000 (6,000-17,000) for the negative and positive CDC-XM-T respectively (p = 0.00005). Regarding the class II antibodies, the MFI median values were 6,750 (2,500-11,000) and 15,000 (9,500-20,000) for the negative and positive CDC-XM-B respectively (p = 0.0003). The likelihood of having a positive CDC-XM in case of DSA with MFI values ≥ 9,000 is about 85%. In conclusion, although the sample size is small, data suggest that MFI value greater than 9,000 predicts a positive CDC-XM. The presence of DSAs in kidney recipients, is strongly associated with hyper-acute and accelerated acute rejection. The production of de novo post-transplant DSA in kidney recipients is often indicative of antibody mediated rejection (AMR). Ample clinical data support the importance of HLA-DSA monitoring for risk assessment in recipients pre-and posttransplant. The aim of this study was to evaluate the development of alloantibodies in kidney transplant patients from the year 2001 to 2014 at the University Hospital of Parma. The study was conducted on 141 consecutive patients (mean age 46; 74% males) transplanted with kidneys from cadaveric donors (mean age 52). Inclusion was restricted topatients on the first kidney transplant with no HLA pre-transplant alloantibodies. Antibody development and monitoring was conducted using bead array technology and data was collected at baseline and one, three, six and 12 months post-transplant and annually until five years post-transplant. The percentage of patients with de novo DSA class II was related to the number of mismatches at the DR locus with a statistically significant difference between patients with 2 mismatches than those who had none (p=0.02) or one mismatch (p=0.04) at the DR locus. In three years, 70% of patients with two mismatches in HLA-DR developed DSA against the class II antigens while only 38% of patients with no or one mismatch developed DSA against the class II antigens. The correlation between class mismatch I and development of specific class I DSA is not statistically significant, perhaps owing to the limited number of patients. Despite this low sample size, our data confirm that mismatch at class II loci represents a major antigenic stimulation that is reflected in DSA production and then in a reduced long-term transplant survival. Highly sensitized patients can wait many years for an offer of a transplant due to the many unacceptable antigens listed with UK Transplant. Previous work at our center has shown that cumulative antibodies detected at <5000 MFI using Luminex Single Antigen beads correlated with a negative flow cytometric and CDC crossmatch. We therefore instigated a monthly multi-discipline team meeting to discuss highly sensitized, long waiting patients, deselecting antigens less than 5000 MFI. Whilst the median calculated Reaction Frequency (cRF) only fell from 100% to 95%, the median number of listed unacceptable antigens was halved from 41 to 21. Using the new UK transplant matching tool this resulted in a predicted increase in the median number of potential tier 1-3 donors from just 5 in 10,000 to 112 in 10,000 donors. Over the past two years, 10% of transplants at this hospital have been from this cohort of highly sensitized patients. The number of patients waiting longer than 1000 days for a transplant has reduced from 23 to just seven and some of these transplants have been from living donors who were previously excluded due to unacceptable antigens. The pre-transplant crossmatch was negative in all cases and outcomes are good with a median 12 month creatinine of 143 μmol/L. The role of anti-HLA-DP and -DQA antibodies in renal graft has been poorly understood for a long time. However, the introduction of the solid phase assay, allowing their identification and better understanding of their clinical importance in kidney transplantation. This study describes the incidence of anti-DP and -DQA antibodies in a cohort of 126 (80 male, 46 female) patients transplanted between January 2016 and December 2017 at the University Hospital of Parma. The mean age at transplantation was 53 years. Pre-and post-transplant sera were identified using the Luminex Single Antigen (LSA, Luminex Single Antigen class I and II, One Lambda). Anti-DP and -DQA antibodies were detected in 17 of 126 (13%) patient post-transplant sera. Of these, 71% (12/17) of the patients had antibodies against these loci before the transplantation (three of these were re-transplants), whereas 29% (5/17) developed them de novo within the follow-up period. Nine of these 17 (53%) patients had previous pregnancies or blood transfusions. Among post-transplant antibodies detected, anti-DP specificities showed a mean fluorescence intensity (MFI) level less than 3000 (positivity cut-off, AIBT Guide Lines 07/2016), while anti-DQA specificities also reached MFI values greater than 3000. In summary, we have shown that anti-DP and -DQA antibodies are common in primary renal transplants but they commonly accompany other antibodies: this makes their impact on graft survival hard to determine. Moreover, typing for HLA-DP and -DQA antigens is still not routinely performed, so it's difficult to classify if an anti-DP antibody as donor-specific, or not. Therefore, the currently available LSA panel for anti-DP and -DQA antibody detection, only covers a small number of the identified alleles. Future studies and extended follow-ups are then warranted to further characterize the role of these antibodies.

FLOW CYTOMETRY CROSSMATCH: FCXM VS. FLOWDSA-XM technique used to detect them (CDC vs. flow cytometry/ solid phase). The new FlowDSA-XM technique combines sensitivity of the classical FCXM with specificity of the single antigen bead assay, allowing a better correlation between crossmatch result and patients' HLA class I and II sensitization detected by solid phase assay. We performed a comparative analysis between FCXM and FlowDSA-XM results by testing 46 sera (9 PRA=0%, 14 HLA class I positive, seven HLA class II positive and 16 HLA class I&II positive) with five different lymphocytes population (typed for all HLA class I&II loci by high resolution PCR-SSP/ SSO techniques).The standard FCXM technique allowed detection of IgG-antibodies specific for T and B lymphocytes. The FlowDSA-XM consisted of an incubation of lymphocytes and serum was performed. Then, three different Capture Beads (HLA I A, B, C beads, HLA IIa DQ beads and HLA IIb DR, DQ, DP beads) were added. Lastly, a Lysis/Stain buffer was used to solubilize the HLA molecules and to label the antibodies bound to them. FlowDSA-XM and FCXM showed strong concordance in the nine PRA negative sera and in the 14 anti-HLA class I positive sera; FlowDSA-XM was partially discordant in one of seven anti-HLA class II positive sera because of FCXM B cells showed a very weak positivity while FlowDSA-XM HLA IIb showed a result near cut-off. Moreover, the FlowDSA-XM was more sensitive than the FCXM showing higher values of positivity, especially in the presence of anti-HLA class I DSA and gave "specific" results discriminating between HLA class I and/or II DSA. On the contrary the FCXM gave B positive results for the presence of both HLA class I and class II DSA. In conclusion the FlowDSA-XM, allowing a better interpretation of the crossmatch result, could represent the crossmatch of the new millennium. The clinical impact of anti-HLA antibodies in the liver transplantation (LT) setting has been unclear, although recent studies have described an increase in the diagnosis of antibody-mediated damage in this population. The aim of our study is to correlate the development of de novo donorspecific anti-HLA antibodies (dnDSA) after LT with the immunosuppression, as well as graft outcome. A total of 65 liver transplanted patients (50 males) of the 2010-2015 period (out of 102; 37 patients were excluded due to early graft loss <3 months) were included for analysis. All patients were screened by Luminex for anti-HLA antibodies before and 6 months, 1 year and 2 years after LT. HLA typing was performed for all donor and recipient pairs. De novo antibodies emerged in 26 patients (40%) who had 38.5% liver dysfunction or graft loss two years post-transplantation, versus 28.2% in de novo negative patients. Fifteen out of 26 positive patients (57.7%) had dnDSA (four class I and 13 class II) with 33.3% graft dysfunction (fibrosis), versus 28.2% in negative patients. Concerning the immunosuppression regimen, patients who received mTORi as maintenance therapy had more dnDSA (31%) and dnDSA class II (24.1%) against 18.8% in patients who did not receive mTORi. Moreover, less HLA compatible patients (0-3 out of 6 matches) had more dnDSA (27.7%) than 18.2% in more compatible patients (4-6 out of 6 matches). Less compatible patients with dnDSA had 35% graft dysfunction, while more compatible patients negative for dnDSA had 14.3% graft dysfunction. In conclusion, the preliminary results of this ongoing study show that de novo anti-HLA antibodies seem to affect the 2-year liver graft outcome. HLA compatibility as well as immunosuppression, seem to have an effect on dnDSA emergence. However, the differences are not yet statistically significant. alleles and of these 148 HLA-B null alleles in January 2018. We hereby report the identification of a novel HLA-B*08 null variant in a potential kidney transplant recipient from Denmark using Next Generation Sequencing (NGS The Luminex (LUM) platform permits detection of HLA antibodies not detectable by complement-dependent lymphocytotoxicity (CDC), but the clinical significance of these antibodies is not completely understood. The aim of this study was to evaluate whether the presence of pre-transplant LUM-detected HLA antibodies is associated with the development of allograft failure. In this single-center cohort study, 573 consecutive kidney recipients transplanted at our kidney transplantation center from deceased donors between January 1, 1996 and December 31, 2005 were enrolled. Pretransplant plasma samples were retrospectively analyzed for the presence of LUM-detected antibodies. Patients were followed for 10 years and censored at allograft failure, death or end of follow-up, whichever came first. Survival analyses were performed to assess the risk of allograft failures by LUM detection. LUM-detected antibodies were found in 171/573 (29.8%) subjects. Of those with LUM-positivity, 46 subsequently developed an allograft failure for a survival rate of 70.0% (95% confidence interval (CI95) 62.6-77.5%), whereas among those without antibodies 54/402 (13.4%) returned to dialysis, with a survival rate of 85.4% (CI95, Over the last decade we have observed remarkable progress in clinical outcome of renal transplant patients with the development of novel effective immunosuppressive agents. Despite these advances, graft dysfunction/rejection or reduced graft survival remains a major burden for clinicians as well as for patients. One of the major reasons for this problem is the formation of anti-HLA-DQ donor specific antibodies (DSA). The present study was done to evaluate the incidence of anti-DQ antibodies and their role in renal transplant rejections. HLA antibody (Ab) screen positive patients (n=40) were assessed by Single Antigen Bead (SAB) assay to determine specificity. HLA typing was performed in the donors to determine the DSA. Initially only HLA-A, -B, and -DR typing was performed in donors and patients. However, in donors with anti-HLA-DQ or anti-HLA-C antibodies, HLA-C and HLA-DQ typing was performed to determine the DSA, if any. Out of 40 patients, 26 (65%) patients developed anti-DQ DSA which was present in association with other HLA class II antibodies in seven (27%) patients and alone in eight (31%) patients. The most common anti-DQ antibodies were anti-HLA-DQB1*06 (42.5%), followed by anti-HLA-DQB1*03(30%) and anti-HLA-DQB1*02(17.5%). This study suggests a high incidence of anti-DQ DSA antibodies in renal transplant recipients. The Single Antigen Bead test and donor typing for HLA-DQ should be included as a routine investigation in the work-up of solid organ transplant.

Claudia Lehmann 1 , Ramona Landgraf 1 , Ilias Doxiadis 1 1 Transfusion medicine, HLA-laboratory, Leipzig, Germany Correspondence: c.lehmann@medizin.uni-leipzig.de Antibody screening and specification is regarded as a prerequisite for patients in organ and stem cell transplantation and platelet substitution. In our center we perform the testing using the generic and the single antigen assays (One Lambda, USA). In the present report we show data from the year 2017. In total 1716 sera were tested starting with the generic assay. 576 sera (33.6%) were class I positive, 525 (30.6%) class II, and 407 (23.7%) MICA positive respectively. Positive sera entered the single antigen assay where additional 253 (43.9%) MHC class I and 260 (49.5%) MHC class II were negative. This result lead us to reconsider the cut-off value for the generic assay for future testing. Using the remaining sera, we asked the question whether the different loci of the MHC, HLA-A, -B, -C, -DRB1, -DRB345, -DQA1, -DQB1, and -DP have a differential immunogenicity. Every defined specificity was calculated as a single hit, the sum of the hits shows how often antibodies towards a given locus were observed. We present below the analysis for the 258 solid organ transplantation patients and their 434 sera (against class I, class II, and class I and II) recognizing 1263 different specificities in total. From those 217 hits were against HLA-A (17.2%), 276 against HLA-B (21.9%), 126 against HLA-C (10.0%), 146 against HLA-DRB1 (11.6%), 75 for HLA-DRB3,4,5 (5.9%), 149 against HLA-DQB1 (11.8%); 176 for DQA1 (14.0%), 98 for HLA-DP (7.8%) respectively. For stem cell transplantation patients including platelet substitution patients a total of 317 hits were counted: 111 for HLA-A (30.8%), 132 for HLA-B (36.7%), and 74 for HLA-C (23.3%). The preliminary data presented here clearly suggest that all HLA loci induce antibody responses: HLA-B>HLA-A>HLA-DQA1>HLA-DQB1 and HLA-DRB1> HLA-C>HLA-DP when all here tested loci are included. These results although observed in a rather small number of patients suggest that for allocation procedures besides the known loci, HLA-A, -B, -DRB1 as well as HLA-DQA1, DQB1 and HLA-DP should be used. Serum samples from transplant recipients are regularly screened for IgG HLA antibodies due to their clinical relevance for transplant outcome. However, other isotypes of HLA antibodies, such as IgA may also contribute to alloimmune responses. Yet, properly validated methods for IgA HLA antibody detection are not available. Therefore, we aimed to modify and optimize the commonly used Luminex single antigen bead (SAB) assay for reliable IgA HLA antibody detection by validation with recombinant IgA HLAspecific human monoclonal antibodies (mAbs), as well as by applying IgG depletion on serum samples. We replaced the standard Luminex SAB IgG secondary antibody with optimal concentrations of IgA1-and IgA2-specific secondary antibodies. We verified the specificity of the assay by using IgG, IgA1 and IgA2 isotype variants of the HLAspecific human mAb MUS4H4 (recognizing HLA-Bw4/ A24/A32/A25). Specificities found for the IgA1 and IgA2 isotype mAbs were identical to those of IgG isotype. Furthermore, we found no cross-reactivity of the IgA1 secondary antibody with IgA2 or IgG isotype mAbs, or vice versa. Mixing experiments using IgA and IgG MUS4H4 mAbs showed that IgG dose-dependently interfered with IgA detection. Therefore, we developed an IgG depletion protocol by using protein G columns to eliminate IgG competition. Next, we validated the modified SAB on a small set of serum samples from which IgG was depleted (n=10). Serum samples from the 2 individuals without a history of alloantigen exposure showed no IgA isotype HLA antibodies, while, IgA isotype HLA-specific antibodies were detected in two out of eight serum samples from individuals with a history of alloantigen exposure. Our preliminary results show that the current modified SAB assay can be used to screen serum samples for HLA-specific IgA antibodies. This may facilitate studies aimed at deciphering the clinical relevance of IgA HLA antibodies in solid organ transplantation.

Gwendaline Guidicelli 1 , Mamy Ralazamahaleo 1 , Charlene Bouthemy 1 , Thoa Nong 2 , Jar-How Lee 2 , Jean-Luc Taupin 1 , Jonathan Visentin 1 1 Centre Hospitalier Universitaire de Bordeaux, Bordeaux, France, 2 One Lambda, Inc, Canoga Park, CA, United States of America Correspondence: jonathan.visentin@chu-bordeaux.fr

Most of the studies attempting to predict T-lymphocyte flow cytometry crossmatch (T-FCXM) results with the mean fluorescence intensity (MFI) of the single antigen flow beads (SAFB) assays were conducted before the description of the high frequency of class I anti-denatured antibodies (anti-dHLA). We previously described that anti-dHLA against HLA-C were particularly frequent and that antinative HLA-C antibodies provide random T-FCXM results, yet the number of T-FCXM performed was small. We reassessed T-FCXM prediction in the presence of anti-HLA-C only with MFI from classical SAFB and iBeads SAFB, largely devoid of dHLA at their surface. Using sera from 45 patients and cells from 44 organ donors, a total of 101 T-FCXM were performed. Sera were selected to react against only one Cw donor antigen with a classical SAFB MFI≥2000. Fourteen among the 15 (93.3%) existing HLA-C antigenic specificities were tested. Twenty-five among the 101 (24.8%) T-FCXM were positive. Fifteen among the 101 (14.9%) T-FCXM studied anti-dHLA, defined as an iBeads MFI<300, none of them induced a positive T-FCXM. The correlation between classical SAFB MFI and T-FCXM ratio was very low (rho=0.189, p=0.06) and higher with iBeads MFI (rho=0.336, p=0.0006). ROC AUC for T-FXCM prediction with SAFB MFI was 0.651 (p=0.01). The optimal threshold for classical SAFB MFI for T-FCXM prediction was 3370, with 92% sensitivity, 40% specificity, 32.9% positive predictive value (PPV) and 93.5% negative predictive value (NPV). ROC AUC was significantly higher with iBeads MFI (0.798, p<0.0001, ROC AUC comparison p=0.05). The optimal threshold for iBeads MFI was 2100, with 92% sensitivity, 55% specificity, 39.7% PPV and 95.3% NPV. In conclusion, iBeads MFI is more accurate to predict T-FCXM results for anti-HLA-C antibodies. This is in favor of the resurgence of iBeads production because this assay has proved to be efficient in eliminating anti-dHLA, which are highly prevalent among anti-HLA-C antibodies. Sensitive Luminex technology can be used for detection and identification of anti-HLA antibodies. While detection can be performed by using pooled antigens providing information on the presence of antibodies, identification of specificity is achieved by single antigen bead assay (SAB). In the Slovenian transplantation program, we perform antibody detection using CDC and Luminex assays twice a year. Additionally, identification is always performed by SAB (Labscreen Single Antigen, One Lambda). For determination of patient immunological status, we combine his/her clinical data and all test results. It is well known that Luminex based antibody detection assays are less sensitive than SAB assays. In order to compare assays and influence of results on patient immunological status we implemented LABScreen Mixed assay (LSM12, One Lambda) for anti-HLA antibody detection. A total of 110 sera were tested for anti-HLA antibodies with LSM12 and SAB. We determined antibody positivity above MFI 2000 in SAB and 2:2 ratio in LSM12 and compared results for each serum. We found 72% of results obtained with LSM12 and SAB to be concordant. Discrepancies were found mainly (25%) in cases where LSM12 was positive and SAB was negative. In those cases, immunological status of the patient would not be affected since SAB as a basis for final decision on the status would be performed. In few LSM12 negative sera that turned up positive in SAB (2%) immunological status would be affected, since the patients would not be tested with SAB and important sensitization could be overlooked. Based on these findings, we changed strategy of testing patients entering waiting list for kidney transplantation with addition of SAB assay. Axl, a TAM receptor tyrosine kinase is expressed in macrophages, monocytes, dendritic cells and natural killer cells. Axl modulates the immune response by prevention of activation of antigen-presenting cells, down-regulation of proinflammatory cytokines and up-regulation of suppressor of cytokine signaling (SOCS) 1 and 3. We previously reported that in mixed lymphocyte reactions (MLR) post-kidney transplantation (KTx) some display a down-regulation while others do not on the addition of aspiration biopsy cultures supernatants (s/n) seven days post-KTx. We report our observations of Axl, both on donor-recipient (R/D) or thirdparty recipient (R/T) combinations. MLR were done between six and 24 months post-transplant, stable, half of wells without s/n, the other with s/n. At the end of MLR cytospins were stained for Axl by immunoperoxidase using anti-human Axl antibody from R&D. The results are shown for R/D and R/T combinations and for MLR displaying a positive stimulation index (SI) and a negative SI. When cpm decreased by at least 30% as compared to wells not supplemented with s/n we called it down-regulation. Patients were first cadaver KTX under CsA-MMF-Pred treatment. There was 12 R/D and 10 R/T, eleven showed positive SI and 11 negative SI by addition of s/n. Axl expression (quartiles) was 88-619 and 454-1189 for R/D and R/T, respectively, (p=0.064) and 87-636 and 94-1189 for stimulated and inhibited MLR, respectively (P=0.12). These findings show a trend for a higher expression of Axl in R/T as compared to R/D and in inhibited MLR surmising a participation of Axl in the modulation of this immune response. Previously we showed significant differences in SOCS expression at the end of MLR in KTx and we speculate that Axl may be behind this observation. A further study encompassing rejecting cases and different treatments is of potential significance. IgA nephropathy (IgAN) is a glomerulary disease that causes end-stage renal disease (ESRD). COSMC is a molecular chaperone in IgA1 glycosylation, together with C1GALT1. Inadequate galactosylation of IgA1 leads to galactose-deficient IgA1 (GdIgA1). This study aimed to demonstrate the relationship between COSMC gene promoter methylation, COSMC mRNA and GdIgA1 level. Patients that didn't have treatment diagnosed as having IgAN after biopsy (n=30), first degree healthy relatives of the patients (n=11) and healthy controls (n=6) were included in the study. B-lymphocytes were cultured in three different groups with DNA methylation modifying agents [Group I: blank, Group II: with IL-4, Group III: with 5-Aza-2'-deoxycytidine (AZA)]. The COSMC gene DNA methylation and expression level were studied using RT-PCR and the GdIgA1 level in serum and cell culture supernatant was studied using ELISA. Serum GdlgA1 level in IgAN patients was detected significantly higher (p<0.0001) compared with healthy relatives and controls. COSMC promoter DNA methylation rate was not significantly (p>0.05) in IgAN patients compared with healthy relatives and controls. COSMC mRNA level in IgAN patients was found significantly lower compared with healthy relatives and controls (p<0.0001). COSMC methylation (p=0.019) and GdIgA1 level (p<0.0001) decreased and COSMC mRNA level (p<0.0001) increased significantly in group III. Increase of COSMC DNA methylation and GdIgA1 level were limited and COSMC mRNA level was significantly decreased (p=0.004) in group II. Our findings support that hypermethylation of COSMC gene promoter region was associated with low COSMC mRNA expression level and it is believed these findings are informative about pathogenesis of the disease. IgA nephropathy (IgAN) is the most frequent primary glomerular disease and is an important cause of end-stage renal disease (ESRD). In this study, we aimed to investigate the diagnostic value of markers for progression of IgA nephropathy. The long-term clinical course of IgAN patients were retrospectively evaluated with initial serum levels of Gd-IgA1 and Oxford classification of IgAN. All clinical data were subject to statistical analysis; estimated glomerular filtration rate (eGFR), sex, body mass, laboratory tests (i.e. proteinuria, serum creatinine) and treatment. Serum galactose deficient-IgA1 (Gd-IgA1) levels in patient sera were measured by using the KM55 ELISA assay (Immuno-Biological Laboratories (IBL) Co). Renal biopsies were scored according to the Oxford MEST scoring system by an experienced pathologist. Primary outcome was defined as ESRD or need for dialysis. eGFRs of patients were calculated with CKD-EPI formula. Forty-four patients (M/F:29/15,mean age 38 years) were examined. Eleven patients (25%) reached ESRD during the follow-up period. Initial serum Gd-IgA1 levels were significantly lower in patient with ESRD (p=0.032). Also, MEST scores were significantly higher in patient with ESRD (p=0.043). On survival analysis, IgA deposition, segmental sclerosis and tubular atrophy were found to be significant predictors of primary outcome (p<0.001, p=0.046, p<0.001 respectively). Serum Gd-IgA1 levels and MEST scores may be useful for predicting the progression of IgA nephropathy. Overt proteinuria may be causes of lower serum Gd-IgA1 concentrations in patient with end stage renal disease. Many kidney recipients will need more than one kidney transplant during a lifetime. Patients often develop anti-HLA antibodies after losing their graft. We investigated the effect of (donor-specific) HLA antibodies (DSA) in retransplanted patients (re-Tx) versus patients transplanted for the first time (first-Tx) in the Dutch PROCARE Consortium study. The 10-year graft survival of 3127 first-Tx and 142 re-Tx patients without HLA antibodies is comparable (80% vs 81%). If patients have HLA antibodies that are not donor-specific (NDSA) the 10-year graft survival was lower for the 311 re-Tx compared to 577 first-Tx (76% vs 71%). This was also the case for patients with DSA: graft survival of 69% for first-Tx (n=297) vs 57% for re-Tx (n=270). This effect might be HLA antibody independent but may also be caused by previous transplants inducing more different HLA antibodies. To investigate whether a higher level of immunization might be driving the graft survival differences in the NDSA and DSA groups we analyzed differences in the percentage panel reactive antibodies (%PRA) at time of transplant (current %PRA) and the historically highest level before transplant (high %PRA) between first-Tx and re-Tx. We found that both %PRA to be considerably higher in re-Tx if DSA were detected by single antigen bead (SAB) assays in the pre-transplant serum. The relative number of transplantations with a current %PRA equal to 0% (these are transplants for which all of the CDC crossmatches against a panel of donors were negative, yet for which SAB assays did pick up HLA antibodies) is considerably higher in first-Tx. After excluding all transplantations with a current % PRA of 0% we reduced the difference in immunization level between first-Tx and re-Tx and found no remaining difference between graft survival of first-Tx and re-Tx for NDSA positive transplantations, and a greatly reduced difference for DSA positive transplantations. In conclusion, in this cohort the relation between re-transplantation and graft survival seems to be predominantly dependent on the presence of pre-transplant HLA antibodies (NDSA and DSA). Therefore, one should consider not including the variable retransplantation in a multivariable model to study the effect of HLA antibodies on long term graft survival, as that will likely mask part of the studied effect.

Adrienne Seitz 1 , Katherine Mounsey 1 , Ruth Foster 1 , Pamela Hughes 1 , Richard Baker 1 , Brendan Clark 1 1 Leeds Teaching Hospitals NHS Trusts, Leeds, United Kingdom (Great Britain) Correspondence: umacls@leeds.ac.uk HLA donor specific antibodies (DSA) are associated with acute rejection and decreased graft survival. The UK national deceased donor allocation algorithm does not currently take HLA-DP sensitization into account when allocating kidneys. We present our center's recent experience of HLA-DP incompatible offers and outcomes. Patients were included in the study if HLA-DP DSA alone was identified in, or within six months of, the time of offer (TOO) sample. HLA-DP specificities with an MFI value >1000 were considered positive. Cross-matching and clinical outcomes data were compared to a control group, which included recipients with a TOO sample containing non-donor-relevant HLA-DP antibodies. Between 2013 and 2017, 31 recipients were identified and followed for a median of 654 days. Seventeen received a DP incompatible transplant (DP-DSA group), with a plan for augmented immunosuppression (n=10) and plasma exchange (n=4). The majority (94%) of patients received a deceased donor graft, and 65% were re-transplants. In addition to possessing HLA-DP antibodies, 59% were highly sensitized patients with a calculated reaction frequency ≥85%. Flow cytometric (FCXM) and CDC T-cell crossmatch results were negative in all cases. Five patients in the DP-DSA cohort generated B-cell positive FCXM results, but no association with HLA-DP MFI levels was observed (10381 B-FCXM negative, vs 12141 B-FCXM positive, p=0.672). The estimated 1-year rejection free survival was reduced in the DP-DSA cohort compared to control (34% vs 67%, p=0.07). The 1-year graft survival was 86% in both groups. Within the follow-up period, two grafts were lost in the DP-DSA cohort. Interestingly, this represented two of three patients who had received fully HLA matched grafts (HLA-A, -B, -C, -DR, -DQ). Neither MFI levels, nor crossmatch results were able to risk-stratify patients. This could reflect the variability of HLA-DP expression on donor tissue. Vigilance is required when transplanting across this barrier. We have developed a new FCXM assay that distinguishes between harmful complement-activating (IgG1 and IgG3) and non-complement activating (IgG2 and IgG4) antibodies using donor cells and recipient sera. Our new FCXM assay has led to successful heart and kidney transplants in adult and paediatric cases. PBMCs isolated from the donor samples were incubated with the pre-transplant sera from eight hearts, and ten kidney recipients with appropriate controls. The cells were then incubated in the lyophilized custom cocktail of antibodies that specifically recognize the various IgG subtypes bound to the cells, followed by FCXM analysis. C1q testing was carried out on all sera. Most of the heart transplant cases studied had a positive crossmatch due to IgG2 or IgG4 antibodies (non-complement activating), correlating with C1q results. Two cases were positive for C1q; probably due to prozone effect of HLA-specific IgM antibodies. All cases had positive 30-day and 90-day survival post-transplant with no PGD or >2R rejection. Two cases with documented AMR continued to have normal graft function. There was almost complete agreement between the IgG subtype assay and C1q assay results in most of the cases studied. A positive FCXM was due to presence of non-complement activating IgG 2 or IgG4 antibodies. Only one case showed the presence of IgG3 antibodies with a negative C1q; probably the result of denatured antibodies. Our new FCXM assay has also helped two paediatric kidney recipients who have no rejection episodes reported. Our preliminary results demonstrate that our new assay can facilitate safe transplants even in the presence of a contraindicating standard positive FCXM. The assay has shown itself to be highly accurate in detecting the IgG subtype/s causing a positive flow crossmatch. Clinical implementation of our IgG subtypes assay would have a great impact on increasing the number of successful transplants carried out, especially in sensitized recipients. In kidney transplantations, donors and recipients are matched on ABO compatibility and on HLA. Since January 2018 all laboratories within Scandiatransplant perform HLA typing on HLA-A, -B, -C, -DRB1, -DQB1, -DQA1 and -DPB1. In addition to HLA matching, donor-specific HLA antibodies (DSAs) are taken into account when matching donors and recipients, because DSAs of high intensity are associated with a high risk of antibody-mediated rejection (AMR) and impaired long-term allograft survival. The consequences of DSAs against DQA1 and DPB1 are not fully understood, but has been suggested to be of equal, or nearly equal, importance. The aim of this study was to quantify prevalence of DQA1 and DPB1 antibodies among patients on the waiting list for kidney transplantation in the western part of Denmark. All patients active on the waiting list for kidney transplantation in the western part of Denmark, by September 1st 2017, were included (n=196). The most recent luminex analysis was re-examined for DQA1 and DPB1 antibodies for each patient. DQA1 and DPB1 antibodies were frequently found among patients on the waiting list, with 11% (21/196) of all patients having DQA1 antibodies, and 19% (37/196) having DPB1 antibodies. Among patients waiting for their first transplantation, DPB1 antibodies were more frequent than DQA1 antibodies, with 10.5% (14/133) having DPB1 antibodies, whereas only 1.5% (2/133) had DQA1 antibodies. Among patients waiting for repeated transplantation the frequency of DQA1 and DPB1 antibodies were comparable with the presence of 30% (19/63) and 36% (23/63) respectively. We here show that DQA1 and DPB1 antibodies are frequently found in patients on the waiting list for kidney transplantation. We suggest that these antibodies should be included in the immunological risk assessment prior to transplantation. Calculated panel-reactive antibody (cPRA) is an accurate measure for the definition of candidates' immunization to a transplant. Based upon unacceptable HLA antigens to which the patient has been sensitized, cPRA is computed with HLA allelic and haplotypic frequencies from a pool of possible donors and represents the percentage of donors that express one or more of the antigens unacceptable for a given transplant candidate. The aim of this study is to compare cPRA values obtained from HLA frequencies of Portuguese donors with values obtained from cPRA calculators from international established sources. We randomly generate HLA-A, -B and -DRB1 unacceptable antigens to a simulated cohort of 100 transplant candidates. With the respective antigens for each patient we compute the Portuguese (PT) cPRA values using the formula developed by Zachary and Braun based on HLA allelic and haplotypic frequencies of 37,993 unrelated bone marrow donors. The results obtained were compared with cPRA values defined from Canadian Transplant Registry (CTR), United Network for Organ Sharing (UNOS) and Eurotransplant (ET) calculators. For each calculator we also classified cPRA between 0-19% as low sensitized, between 20-79% as sensitized and >79% as highly sensitized. Spearmans rho test was used to compare results from each calculator and to assess the correlation between them; Cohen's kappa coefficient was used to access agreement for categorical values. Values obtained for cPRA from PT calculator are highly correlated with cPRA obtained from each one of the three calculators. When we categorized cPRA values from each calculator, substantial agreement was obtained between PT and ET (kappa = 0.67) and between PT and UNOS (kappa = 0.62) while only a moderate agreement was obtained between PT and CTR values (kappa = 0.47). The cPRA gives us the probability of each transplant candidate to have a positive crossmatch with the next available donor for transplantation. Reliability of cPRA values depends on the donor pool from which it is calculated, and it is crucial for highly sensitized patients to be classified as so. The genotypic frequencies used for cPRA computation frequencies should be known and scrutinized to assure a correct measure of candidates' sensitization and must correspond, as far as possible, to those from future organ donors. The HLA-G gene stands out as a natural immunosuppressant on patients going through transplantation. Its low expression has been associated with graft rejection. Single nucleotide polymorphisms at the 3´UTR, affect the mRNA stability, promoting or inhibiting the molecule expression. Our study focused on the 3´UTR sequences of HLA-G regarding its polymorphisms in kidney transplanted patients (n=64, between 18 and 75 years) and their post-transplant rejection episodes. Haplotype, genotype and allele frequencies of the 3´UTR SNPs were compared to detect association with risk of kidney allograft rejection. Eleven out 40 haplotypes described in the literature were observed in the present study. No haplotype showed association with kidney graft outcome. However, the genotype +3010G/G (SNP: rs1710) was characterized as a protective factor (p=0.03; OR=0.25; IC=0.07-0.88) against the development of rejection, while the heterozygous genotype on this position was considered as a risk factor according to Fischer Exact test and multivariate analysis (p=0.012; OR=5.168; IC=1.43-18.56). Also, the high value of PRA (higher than 30%) was shown as a risk factor, which is in agreement with results from others clinical studies. The heterozygous genotype on position +3142 (SNP: rs1063320) is a risk factor as detected by Fischer Exact test (p=0.02; OR=4.17; IC=1.30-13.37), but not evidenced by multivariate analysis. Additionally, it was observed that strong linkage disequilibrium between +3010C/+3142G and +3010G/+3142C (r 2 =0.94; D = 1) exists. It is worth noting that in different studies +3010G was considered a risk factor for preeclampsia development, while +3142C was a risk factor for auto-immune disease. The present linkage disequilibrium and association observed with kidney transplantation outcome should be better investigated in order to clarify the suggested double effect of +3010 and +3142 polymorphisms in up/down regulation of HLA-G expression as a protector/risk factor. We report the case of a 54-year-old patient who experienced antibody mediated rejection (ABMR) two months after a second kidney transplantation (KT). The patient was transplanted in 1998 for a chronic glomerulopathy with kidney malformation. The first transplant was lost in 2012 due to chronic allograft rejection. The patient had developed HLA-C antibodies (Ab) anti-HLA-C*01, C*02, C*04, C*05, C*06, C*15, C*17, C*18 (epitope 80K, DSA C*01). Transplantectomy was performed in 2013. The patient received a second kidney transplant in 2015 from a deceased donor (HLA-A2,68; B7,44; C*08,10; DR1,15; DQ7 incompatible). Anti-thymocyte globulin was used for induction therapy and immunosuppressive regimen contained tacrolimus, mycophenolate mofetil, steroids and three cures of intravenous immunoglobulins (IVIg). A renal allograft biopsy made at two months for an acute renal dysfunction showed a C4d negative acute humoral rejection (C4d ABMR -) according to BANFF 2013 classification. Despite a treatment with Rituximab, IVIg and steroids, features of C4d-ABMR persisted on the two subsequent biopsies performed three and ten months later. We did not find anti-HLA DSA in serum (sDSA) neither anti-MICA and we decided to search intra-graft Abs. Elution was performed with ELU-Kit II elution. The specificities were detected using Single Antigen Luminex kit (One Lambda). The final wash supernatant was tested to control for efficiency of washing and no HLA Abs were detected. Intra-graft anti-HLA antibodies were detected in two biopsies performed 23 months and 33 months after KT. The result of elutions showed positivity only against HLA-C*05. Matchmaker analysis showed that HLA-C*05 shares the eplet 138K with the allele C*08:02, C*08:04, C*08:12 and HLA-C high resolution typing of the donor revealed the presence of C*08 :02, allele not tested for in the SA kit. This clinical observation illustrates the usefulness of searching HLA intra-graft Ab in case of ABMR without detection of sDSA. When risk assessing patients before re-transplantation knowledge of the HLA type of previous organ donors is crucial. However, these can be completely or partially unknown mainly due to the HLA typing techniques applied at the time of transplantation. The clinical significance of detected HLA antibodies, as well as the risk of repeated incompatibility with a new donor cannot be determined definitively without a complete type of the initial donor. When historic samples from deceased donors are not available it is clearly beneficial if the HLA type can be determined from graft biopsy material. This is, however, complicated by the fact that graft biopsies will contain both donor tissue and infiltrating lymphocytes from the patient. Thus, the HLA type deduced from biopsy material will represent both the donor and the recipient. Our aim was to investigate whether HLA typing performed on biopsy material would allow HLA typing of the previous donors and thus enable a better pre-transplant risk assessment. Initially we tested mixtures of DNA with known HLA types using Linkage qPCR kits and Suretyper software and were able to detect 40 out of 44 expected alleles. Next, we tested two allograft biopsies from cardiac transplanted patients and 3 biopsies from renal transplant patients. The transplantations were performed between 11 and 45 years ago, and we were able to detect between two and nine HLA alleles of donor origin. We conclude that it is possible to determine multiple donor HLA types that differ from the patient type. As expected, we found a correlation between graft age and number of detected alleles. Testing has enabled us to detect donor specific antibodies and furthermore to define unacceptable mismatches. We believe this is a feasible method to be used in routine, but it must be stressed that not all possible mismatches will be found. De novo donor-specific antibodies (dnDSA) detected posttransplant are an independent risk factors for late graft failure (GF), developed as a response to HLA mismatches (MM). However, individual MM can have different effects on GF in kidney transplantation (KT). The aim of this work was to determine the association between MM, using HLAMatchmaker (HLAMM) and Prediction of Indirectly ReCognized HLA Epitopes (PIRCHE II) with development of dnDSA in KT with living donor (LDKT). A total of 149 LDKT were performed between 2007 and 2014 in Hospital Santo António (HSA), 96 (64.4%) with living-related donor (LRD) and 53 (35.5%) with living unrelated donor (LURD), were included in the study. HLA-A, -B and -DR low resolution typing was performed for all donor-recipients pairs, and high resolution determined for HLA-A, -B, -DRB1/3/4/5, -DQA1/DQB1. PIRCHE II (version 2.4) score was determined and epitope MM load (EpMM) assigned with HLAMM (version 2). DSA were assessed longitudinally (at a maximum follow-up of 10 years) with single-antigen bead (SAB) assay. Patients that developed dnDSA (20, 13.4%) were more frequently re-transplants (P=0.043) and had a higher PIRCHE II (P=0.012). Re-transplantation (HR=4.019, P=0.009), younger recipient age (HR=0.963, P=0.047) and higher PIRCHE II (HR=1.013 per unit increase, P=0.007) were independent predictors of dnDSA in multivariable Cox analysis, while EpMM (HR=1.025 per unit increase, P=0.446) and antigen HLA MM (HR=0.844 per unit increase, P=0.405) were not. Considering EpMM and PIRCHE II as terciles, cumulative incidence of dnDSA was 13%, 25% and 29% for EpMM 0-10, 11-21 and ≥22 (log-rank P=0.052), and 7%, 22% and 32% for PIRCHE II 0-44, 45-77 and ≥78 (log-rank P=0.004), respectively. Highest tercile of EpMM (≥22, HR=11.742, P=0.030) and PIRCHE (≥78, HR=8.991, P=0.018) were both independent predictors of dnDSA. Our study showed that both eplet mismatch load and PIRCHE II score are predictors for dnDSA and may contribute to a significant improvement in HLA matching, allowing the identification of low and high risk HLA mismatches in kidney transplantation. The presence of donor specific anti-HLA antibodies (DSA) is a major risk factor for poor graft outcome in kidney transplant recipients. Desensitization protocols have been developed in order to overcome the immunological barrier. Between 2006 and until 2012 we, at the Rabin Medical Center in Israel, implemented a program for desensitization of highly-sensitized (positive DSA, negative NIH-CDC cross match) living donor recipients. The long-term outcome of 36 highly-sensitized kidney transplant recipients, treated with IVIG and plasmapheresis, with or without rituximab (added when >7500 MFI), was compared to all non-sensitized living-donor recipients who were transplanted in those years in our center. Thirty-six highlysensitized recipients and 252 non-sensitized living donor recipients were included. During a mean follow-up of 121.7 months, death-censored graft failure occurred in six (16.6%) recipients in the highly-sensitized group and 15 (6%) recipients in the non-sensitized group (p=0.021).

Ten-year death-censored graft survival rate was 81% versus 93%. There was no difference in patient survival. In the highly-sensitized group, graft failure or DSA persistence after transplantation was significantly associated with the presence of class II antigens (RR-4.7, 95% CI 1.06-21.4, p=0.041). No association was found with class I antigens or peak DSA level. Three out of five (60%) recipients who had DSA against HLA DR7 lost their graft, compared to 12/30 (40%) with DSA against other antigens. Twenty of the 28 highly-sensitized patients who remained alive and did not lose their graft did not have DSA at the end of followup period. Long term death-censored graft survival in highly-sensitized recipients is less favorable compared to non-sensitized recipients, although in this cohort they are better than previously reported. However, for those who did not have DSA rebound, the long-term graft outcome is excellent. DSA against class II (especially DR7) was associated with a significant risk for graft failure or DSA persistence. We conclude that adding rituximab to desensitization therapy for high-risk recipients may offer a benefit.

David Pinelli 1 , Muralidhar T. Gopalakrishnan 1 , Anat R. Tambur 1   1 Northwestern University, Chicago, IL, United States of America Correspondence: a-tambur@northwestern.edu

The temporal model of IgG subclass switching supports sequential IgG conversion. However, the diversity of previously stimulated B cells is known to shape responses in different directions. Recent studies suggest that different subclasses of IgG (1-4) possess varying capacity to cause graft injury. As IgG subclass switching is a dynamic process, we performed a preliminary exploratory study in 23 adult and pediatric recipients (14 kidney/3 kidney-pancreas/3 heart/2 kidney-liver/1 pancreas) to investigate the kinetics and relative magnitude of donor-specific antibodies (DSA) IgG subclasses and compare to total IgG over time. Serial post-transplant samples were tested for total IgG, IgG3 and IgG4 subclasses using class I and II Luminex SAB. All patients had total IgG DSA >1000 MFI at all time points tested and most were receiving treatment for active AMR. Total IgG antibodies (Ab) were compared with those detected by IgG3 or IgG4 reagents for each HLA Ab specificity. Interestingly, in 12 patients, some Ab directed at the same target were positive for both IgG3 and IgG4 in a single serum sample. In 21 patients, MFI values for the positive total IgG beads were higher or comparable to those seen with IgG3 or IgG4 reagents. However, two patients showed much higher MFI values for the IgG4 reagents than the total IgG. While most patients had noticeable IgG3 or IgG4 subclass binding in at least one serum sample, the MFI varied dramatically over time, and generally followed a similar pattern of increase/decrease as the corresponding specificity in the total IgG assay. Notably, we did not see class switching over time from IgG3 to IgG4, although this was not a comprehensive study. Among the ten patients that were tested at the time of active AMR with C4d+ biopsies, only two had IgG3+ DSA and 3 three had IgG4+ DSA. In summary, MFI of IgG subclasses fluctuate over time and do not seem to correspond to the clinical status of the patient. Some Ab may be positive for both IgG3 and IgG4, which can complicate analysis in clinical studies. We suspect that the reagents currently available are not ready for clinical use, and we advise significant caution when interpreting results, especially those obtained at a single time point.

Sevim Gönen 1 , Özant Helvacı 1 , Yaşar Kandur 1 , Hakan Sözen 1 , O guz Söylemezo glu 1 modulate immunity for BKV. Recent studies show that HLA antigens within major histocompatibility complex (MHC) may play a role in response to viral infections and may influence the course of BKV infection. Identification of HLAs, that alter the course of infection will facilitate risk stratification, and customization of pre-emptive intervention strategies. The aim of the present study was to determine HLA antigens associated with BK viremia after kidney transplantation. This was a retrospective single-center study. We looked for the association of recipient origin of BKV in renal transplantation and HLA antigens of the patients followed in an University Transplant Centre. We analyzed HLA antigen association in 25 kidney transplant patients with BKV nephropathy and in control group of 50 unaffected transplant recipients who were on tacrolimus-based immunosuppression. HLA was analyzed by a PCR-SSO method using Luminex technology. Our results showed that there was a significant difference between patients and control group in the mean frequency of HLA-DRB1*15 (p=0.050), HLA-DQB1*02 (p=0.029) and HLA DQB1*06 (p=0.014) locus. However, HLA-A*23 was significantly less frequent in patients than the control group (p=0.014). Although patient numbers are limited we can suggest that kidney transplant recipients who carry DRB1*15, DQB1*02, DQB1*06 HLA allele groups and who do not carry HLA-A*23, are at risk of BKV infections. Further randomized studies with larger sample sizes will increase our knowledge on the issue. In kidney transplantation, it is pivotal to prevent human leukocyte antigen (HLA) donor specific antibody (DSA) formation, which is triggered by immunogenic epitopes present on mismatched HLA. The optimal donor can theoretically be selected by avoiding these epitope mismatches. To implement epitope matching, it is essential to define the relevant epitope mismatches between recipient and donor. Therefore, we aimed at developing and validating a tool to define amino acid mismatches on donor HLA. This tool can be used to define the immunogenicity of HLA mismatches for a large number of recipient-donor pairs. We developed a computer algorithm containing all HLA amino acid (AA) sequences from the IPD-IMGT/HLA Database and all theoretical and antibody-verified eplet AA (position and type) from the HLA Epitope Registry. The HLA input is preferably second field typing, as the HLA sequences of the recipient are compared to each mismatched HLA allele of the donor to define the number of AA mismatches. In addition, the physiochemical properties of the AA mismatches are defined, which can be correlated to DSA formation as detected by Luminex single antigen after transplantation to define the immunogenic epitopes. This tool and other approaches (EpVix, Cambridge AA mm score) were used to analyze 233 first kidney transplant recipient-donor pairs as part of the 17th IHIW epitope component. Comparing the results of all approaches similar number of mismatches were observed. For both HLA class I and class II, we found that increasing numbers of AA mismatches resulted in a higher probability of DSA formation. This newly developed computer algorithm can be used to define HLA AA mismatches for a large number of recipient-donor pairs from all kind of populations and to select optimal donors in kidney transplantation. An advantage of this tool is that all HLA alleles, including rare alleles, can be analyzed.

Pregnancy is an excellent model to study antibody responses to HLA mismatches. Such antibodies are induced by paternal alleles of the unborn child during same time intervals and in the absence of immunosuppressive drugs. From the HLA types of the mother and child one can readily determine the repertoire of mismatched eplets. This study was done on 155 first pregnancy cases and immediately after delivery, each serum was tested for IgM-type and IgG-type HLA class I antibodies and evaluated for epletassociated patterns of antibody reactivity. An analysis of the effects of eplet loads and PIRCHE-II numbers on antibody responses showed generally correlations but often enough, with low levels of statistical significance. These findings have prompted new interpretations of the roles of eplet loads and PIRCHE-II numbers. This study also provided data about the immunogenicity of mismatched eplets as determined from considerable differences between the frequencies of specific antibody formation. This information about eplet immunogenicity could be applied to a new system of eplet loads of mismatched HLA alleles. Since the implementation of Luminex for anti-HLA testing, the improvement in the definition of anti-HLA patterns is clear. However, several caveats should be taken in account, false positive and negative results due to technical issues, denatured molecules, cross-reaction, lack of cut-off consensus, wide inter-and intra-assay variability, prozone effect, and interference with complement may appear. We sought to improve the definition of anti-HLA results after Luminex assay. Serum from 40 individuals newly listed for kidney transplantation in our institution and 50 solid organ transplant patients followed post-transplant within one year were tested for anti-HLA antibodies both Single Antigen (SAB) and screening (One Lambda). Acid treatment of the beads to identify reactions against denatured-HLA (dHLA) or native-HLA (nHLA) molecule and further complementdependent cytotoxic (CDC) and flow cytometry cross-match (FCXM) to confirm dHLA or nHLA reactions were performed. In 20 patients, (seven listed and 13 after transplantation) specific reactions with screening negative with SAB positive results were found. The MFI in the specific reaction was significantly increased in SAB vs screening (4949 +/-473.2 vs 154.5 +/-13.61, p<0.001). After acid treatment of the beads, 95% of the reactions against (dHLA) were identified. Moreover, the negative results after CDC and FCXM confirmed non-nHLA specific reactions. Performing SAB in non-sensitized patients in waiting list identify reactions against dHLA molecule without relevance in cross-match. Acid treatment of the beads allows the identification of dHLA and should be performed to improve the accuracy in anti-HLA antibody assignment. After solid organ transplantation, the presence of DSA should be confirmed for nHLA reaction in asymptomatic patients. The Luminex Platform has emerged as a favored technology for detection of HLA antibodies in addition to the CDC crossmatch (XM) which is still the gold standard. Luminex DSA crossmatch is being introduced in India as a prognostic marker for graft survival to detect antibodies that are missed by CDC XM as antibody mediated rejection (AMR) is still an issue of concern in transplant (Tx) patients with a CDC negative XM. The aim of the present study was to evaluate the impact of the pre-transplant DSAs detected by Luminex crossmatch on the clinical outcome of the renal graft over a period of one year in the recipients who underwent renal transplant at the Sir Ganga Ram Hospital. Pretransplant sera from 46 renal Tx recipients with a negative CDC crossmatch were assessed for DSA on the Luminex Platform using Lifecodes DSA kit (Immucor). The serum samples with DSA (HLA-class I or II or both) of MFI (mean fluorescent intensity) value more than 500 was considered to be positive. The results were then correlated with the clinical outcomes of the renal allograft. DSAs were found in 11 out of 46 recipients (23.9%). Of the eleven DSA positive patients, three patients (27.27%) developed acute graft rejection. All these three patients had positive C4d staining in their biopsies and the MFI value of the DSA on Luminex platform was found to be more than 1000. The remaining eight DSA positive patients showed no rejection and had stable graft function. The MFI value of the DSAs in these patients ranged from 500-1000. All the 35 DSA negative patients (76.1%) also had stable graft function in one year follow up. The AMR was more frequent in the DSA positive group with no or less AMR in DSA negative group (27.27% versus 76.1%). The present study evaluates the importance of Luminex DSA crossmatch testing in detecting the donor specific HLA antibodies over the CDC crossmatch. There was a higher incidence of AMR in patients with pre-transplant DSA despite a negative CDC crossmatch. The present study clearly establishes that the Luminex DSA crossmatch is helpful for predicting post-transplant graft outcome or rejection and DSA MFI value above 1000 should be considered for further evaluation by Single bead assay (SAB) by Luminex. The advancements in DNA based technologies has led to a switch from serologic typing to high resolution DNA typing methods, but these DNA sequencing techniques assign all B15 antigens (B62, B63, B70(B71 and B72), B75, B76, B77) to the HLA-B*15 allele group. The presence of antibodies defined as serological splits in the patient urges the identification of serological B15 splits of the donor HLA antigens as the presence of Donor Specific Antibodies is an important contra-indication for kidney transplantation. There are 289 B*15 alleles out of 400 without assigned serological type in the HLA Dictionary. Hence, we aimed to facilitate the assignment of B*15 alleles to specific serological subtypes by identifying specific nucleotide motifs for each B15 subtype based on single nucleotide polymorphisms (SNPs). For this purpose, we obtained the full-length sequences of 21 B*15 alleles from the IPD-IMGT/HLA database and extended the genomic sequences of another 5 alleles by group specific hemizygous Sanger sequencing. These 26 B*15 alleles with defined serologic specificities were aligned by Clustal Omega. This alignment lead to the identification of specific nucleotide motifs of each B15 split antigens at 38 different SNPs in introns 1 and 2 and exons 2 and 3 of B*15 alleles. The prediction power of these nucleotides was validated with a data set of 85 B*15 alleles with assigned serologic types (HLA dictionary) and 77 were concordant. Moreover, the predicted serologic subtypes of two new B*15 alleles are confirmed by serological typing. In the panel of 289 B*15 alleles with unknown/unidentified serologic specificities, we predicted the 257 B*15 serologic equivalents, only 32 alleles revealed inconclusive nucleotide patterns for splits at the identified positions. This fast and reliable method enables prediction of B15 serological equivalents, that is of importance in case Donor Specific Antibodies are present, to prevent kidney transplant rejection. Human leukocyte antigen (HLA) matching is not practically usable in heart or lung transplant allocation. The effect of donor-recipient HLA matching on outcomes remains relatively unexplored for thoracic transplants. The objective of this study was to investigate the effects of donor-recipient HLA matching on graft survival in heart and lung transplantation. A retrospective analysis was performed on data obtained from Centro Regionale Trapianti registry in Piemonte, Italy. An overall cohort of more than 600 heart and 300 lung transplants performed in the Transplant Center in Torino. The Kaplan-Meier model was used to assess the influence in graft loss among HLA match groups and other parameters, such as age, era of transplant. For the purpose of the multivariate Cox regression analysis continuous variables of transplant year and age were divided to create categorical variables: early transplant era (graft performed in the period 1990-2005) versus late transplant era (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) (2017) . Additional analysis of graft survival was performed between matching at each individual HLA-A, -B, -DRB1 locus. A greater degree of HLA donor-recipient match occurs randomly in thoracic transplants, but is associated with improved graft survival, especially in the early era and within one year from the date of transplant. For heart transplant of the early era (1990-2005, n=350 There are differences in the pharmacokinetics of mycophenolic acid (MPA) among individuals. MPA is primarily metabolized by uridine diphosphate-glucuronosyltransferase (UGT). UGT1A8 is an important UGTs for the glucuronidation of MPA. The aim of this study was to explain MPA pharmacokinetics in UGT1A8*2, UGT1A8*3 polymorphisms in Turkish renal transplant patients. 125 patients with renal transplants were included in this study. UGT1A8*2, UGT1A8*3 genotyping was performed using PCR-RFLP. Concentrations of MPA were determined with Cloned Enzyme Donor Immunoassay (CEDIA). Blood samples were taken from 125 renal transplant patients taking 2 g MPA daily with at least three months history of transplantation. All patients were monitored for acute rejection and graft function during the study period. The UGT1A8 *1*1, *1*2, *2*2, *1*3 and *3*3 genotype frequencies among patients were respectively, 68.0%, 15.2%, 9.6%, 4.8% and 1.6% (*3/*3 was not found). The frequency of heterozygous or homozygous UGT1A8*2 was observed in 31 of the 125 patients (24.8%). Of the 125 patients, eight (6.4%) were heterozygous for the UGT1A8*3 polymorphism. All single nucleotide polymorphisms were in Hardy Weinberg equilibrium. At one month, levels of MPA were significantly lower in UGT1A8*1*1 carriers than in *1*2 and *2*2 carriers (p<0.05). The ratio of blood concentration to dose of MPA in daytime were less in the UGT1A8*2*2 genotype group than in the *1/*1 and *1/*2 genotype groups, but the ratios were not significantly different (p>0.05). Also, there was no association found between UGT1A8*2, UGT1A8*3 polymorphisms and acute rejection (p>0.05). Our results demonstrated a correlation between the UGT1A8 polymorphism and MPA pharmacokinetics among renal transplant patients. UGT 1A8 polymorphism can be partly responsible for inter-individual differences among the renal transplant patients, although most of our patients had acceptable MPA plasma levels. The aim of the present study is to investigate the role of lymphocytotoxic T-cell and B-cell crossmatch (XM) results in liver transplantation outcomes. This retrospective cohort data was obtained from a single transplant center in Italy. We analyzed 3048 orthotropic liver transplantations consecutively performed from deceased donors between 1990 and 2017 in Città della Salute e della Scienza di Torino University Hospital. Whether a positive crossmatch has any relevance to liver graft outcome is controversial: some studies have reported that the presence of pre-transplant positive result is not a prognostic factor. We focus the impact of crossmatch results according to first or re-transplant outcomes, especially in the early follow-up (within three months). Crossmatch was prospectively performed by testing T and/or B cells, mainly with recipients' sera at the time of transplant, by complement dependent long incubation cytotoxicity test. The Kaplan-Meier product limit method and Cox regression analysis was used to assess graft survival and significance was determined by log rank test or Cox test as appropriate. Our results show a statistically significant impact in graft survival in first transplant recipients with T/B-cells positive crossmatch (at three months: 0.92 for T-cells negative XM, 0.89 for T/B-cells positive XM. At 1 year: 0.89 for T-cells negative XM, 0.79 for T/B-cells positive XM; p=0.01580). No effect evidence in 236 retransplant outcomes and no effect concerning B-cells only positive XM. Anti HLA class I pre-existing antibodies have a relevant impact on liver graft outcome, although with a less extent in respect other solid organ transplantations.

Aida Turganbekova 1 , Zhandos Burkitbaev 1 , Saniya Abdrakhmanova 1 , Indira Ramilyeva 1 , Dana Baimukasheva 1 , Zhazira Saduakas 1 , Perizat Zhunus 1 1 Republic scientific and production center of transfusiology, Astana, Kazakhstan Correspondence: dr.aida@mail.ru Detection of donor-specific antibodies (DSA) is a risk factor for transplant rejection. Therefore, early diagnosis of DSA may affect the result of transplant engraftment. Monitoring of pre-existing antibodies until 2017 was carried out by ELISA methods (Lat1240, One Lambda Inc) and from 2017 by flow cytometry using Luminex technology and One Lambda kits. Donors and patients typing were performed using the PCR-SSP method at a low resolution of the loci HLA-A, -B, -DRB1 (Prortans, mdp gmbh). The compatibility test before the transplantation was carried out on the basis of a lymphocytotoxic test (CDC-cross-match). All samples were negative. Post-transplantation monitoring was performed using the Single Antigen Bead Assay (Luminex, One Lambda Inc) technology. Positive results were those that showed MFI (Mean Florescent Intensity) above 1000. Monitoring of pre-existing and post-transplant leukocyte antibodies was performed for 34 patients who underwent heart transplantation. The results showed that 25 (74%) patients did not have pre-existing antibodies, seven (21%) patients had antibodies directed only to class I antigens, one patient had only class II antigens and one patient had a high level of sensitization on antigens of class I and II. Posttransplantation monitoring showed the following results: 24 (71%) patients had negative results, five (15%) patients had class I antibodies, two patients (6%) had class II and three (8%) patients had antibodies to class I and II antigens. One patient identified a donor-specific antibody (DSA) against HLA-B8. Complete detection of DSA for class II was not possible, due to the incomplete HLA genotype of the donor and the low typing resolution. Based on the results of this study, the government is considering the possibility of changing the protocols for the donor survey, including, if necessary, typing all the necessary loci at high resolution for the subsequent diagnostics of DSA. Heightened sensitivity of assays detecting anti-HLA antibodies has greatly improved renal allograft outcome. Though reduced, antibody mediated rejections continue to challenge. Non-HLA antibodies such as the anti-angiotensin 1 receptor, anti-MICA and the anti-endothelial cell antibody have been shown to impact. The latter has been known to be induced by viral infections. In India outbreaks of viral infections are known to occur seasonally. To investigate two clusters of acute antibody mediated rejections in ten patients last year this study was performed. We aimed to test pre-transplant serum for anti-endothelial cell and antiangiotensin 1 receptor antibodies. Of 98 renal transplants last year,15 developed acute antibody mediated rejection. Ten occurred in two clusters with four of ten patients transplanted in August and six of 20 transplanted in November & December. Controls were five patients transplanted in the same time period with no rejection. Both patients and controls were negative for donor specific anti-HLA antibodies, by complement dependent cytotoxicity and flow cytometric crossmatches and the Luminex based single antigen bead assay. Antibodies to angiotensin 1 receptor were tested using the EIA kit (One Lambda) and was negative in patients and controls. Anti-endothelial antibodies were tested using the Euroimmun kit and the titer plane technique. Nine of ten patients were strongly positive indicating the presence of anti-endothelial cell antibodies. The one patient who was negative showed a positive reaction in post-transplant serum. Of the five controls 2 were negative, 2 were borderline positive and one strongly positive. Our data, though limited by small patient numbers, suggests that in a tropical, developing country like India, anti-endothelial antibodies should be considered as possible culprits in clustered unexplained acute antibody mediated renal transplant rejections. Whether the etiology is viral needs further investigation Specific non-HLA antibodies directed to the angiotensin II type 1 receptor (anti-AT1R) and endothelin-1 type A receptor (anti-ETAR) of the vascular cells can be responsible for antibody mediated rejection in the absence of donor specific anti-HLA antibodies. We aimed to determine the presence of non-HLA antibodies in both the pre-transplant and posttransplant sera, its participation in acute humoral rejection, and its relation to donor specific anti-HLA antibodies. The presence of anti-ETAR and anti-AT1R antibodies was evaluated in stored serum samples of 125 consecutive recipients in kidney pre-transplant and during the first 12 months post-transplantation. The study of these antibodies was evaluated by ELISA. Concentrations larger than 10 U/ml of anti-ETAR and anti-AT1R were considered as positive. The determination of the presence of HLA antibodies, included in this study, was routinely performed by Luminex technology. Levels of non-HLA antibodies (anti-AT1R and/or anti-ETAR) were positive in 30 (24.0%) out of 125 of the studied patients. The performed tests were positive in 28 (22.4%) patients with anti-AT1R antibodies and 12 (9.8%) with anti-ETAR from the total of 125 patients, 10 of them had presence of both non-HLA antibodies. In the pre-transplant sera, the average level of positive non-HLA antibodies was found to be equal to 16.12 U/ml anti-ETAR and 15.39 U/ml anti-AT1R, while the latter was equal to 17.61 U/ml anti-ETAR and 15.10 U/ml anti-AT1R in post-transplant sera. Graft loss mediated by acute humoral rejection during the first year post-transplantation was detected in eight patients with positive result in non-HLA anti-bodies, although only four showed donor specific anti-HLA antibodies. Chronic humoral rejection was found in one patient, and acute cellular rejection in another patient. The presence of C4d was detected in two of the 30 mentioned positive patients. A high correlation between the levels of non-HLA antibodies and reduced kidney function during the first 12 months after transplantation was found, especially in presence of donor specific anti-HLA antibodies. Anti-AT1R and anti-ETAR antibodies can be useful to evaluate the risk of allograft loss. Uromodulin (UMOD) is a urinary protein secreted by epithelial cells of the thick ascending limb of the Henle loop. It plays a critical role in the prevention of urinary tract infections, in the maintenance of the osmotic pressure, and in electrolytic homeostasis. UMOD usually precipitates in the urinary tract due to the acidic pH, making its measurement in urine samples challenging. Although its concentration in plasma is lower, ELISA assays guarantee a precise estimate of UMOD concentrations that correlate well with renal function. In this work, we correlated UMOD serum levels to the rs12917707 single nucleotide polymorphism (SNP), located in the promoter region of the gene and leading to a G>T substitution, in a cohort of cadaveric kidney donors (n=357). In this cohort, allelic frequencies were 0.8 for the G and 0.2 for the T alleles, in line with what reported in the literature. Results indicate that GG carriers (n=234) have mean UMOD plasmatic levels of 121+/-51 ng/ml, while UMOD levels in GT carriers (n=108) are 81.5+/-41 ng/ml and in TT (n=15) are 61+/-35 ng/ml. Our results indicate a dose dependent relationship between the presence of the G allele and UMOD serum levels. Furthermore, recipients of GT kidneys were characterized by a shorter graft survival compared to GG, with a follow-up of >10 years. Specifically, at one year the survival of GG grafts was 96% compared to 95% of GT, while at 5 years survival of GG was 89% vs 84% of GT grafts (p=0,027). The recipients of TT homozygous kidneys were excluded from the analysis because of low numbers. These results show that rs12917707 influences plasmatic, and not only urinary, UMOD levels, potentially identifying in the heterozygous grafts a less performing subset of kidneys, leading in the long run to decreased graft survival. In conclusion, this work confirms a dose-dependent relationship between the presence of the T allele and lower UMOD serum levels, in turn associated with reduced kidney function. Further studies will tell whether adding serum UMOD to the tests performed during the evaluation of a kidney donor may add to the pre-transplant diagnostic power. Antibody-mediated rejection (ABMR) is the most common cause of kidney allograft failure. Renal transplant recipients with pre-existing donor-specific antibodies (DSA) or patients who develop de novo DSA after transplantation are at risk of ABMR. Therefore, HLA antibody monitoring before and after transplantation as well may help to identify patients at risk of graft loss. HLA antibody screening after kidney transplantation has been routinely tested on the day of transplantation and one month after. From December 2015, we started to perform post-transplantation monitoring of HLA antibodies at months 1, 3, 6 and 12. This retrospective analysis included HLA antibody screening results for 36 patients that have been transplanted from December 2015 until February 2017. All patient's sera have been tested by CDC and Luminex (Immucor) techniques. Before transplantation, 30 (83.3%) patients were negative and six (16.7%) CDC and/or LUM positive. Among negative patients, 26 (72.2%) remained negative, while four (11.1%) became positive after transplantation. In one patient, class II DSA were detected three months after transplantation and nine months later he had graft loss. In another patient, class II non-DSA were detected six months after transplantation without signs of graft dysfunction. Class I/II non-DSA were detected in two patients 12 months after transplantation associated with transitory graft dysfunction. Among pretransplant immunized patients, one has developed class II DSA three months after transplantation along with graft dysfunction. Adjustment of immunosuppressive therapy was successful. These results suggest that development of de novo DSA so as non-DSA may have detrimental effect on graft function after kidney transplantation, but for the reliable conclusions posttranslational monitoring needs to be continued including larger number of patients.

Angeliki G. Vittoraki 1 , Maria Darema 2 , Dimitra A. Skoumi 1 , Maria D. Apostolaki 1 , Ioannis Bokos 2 , Alexandra A. Siorenta 1 , Georgios Zavvos 2 , John N. Boletis 2 , Aliki G. Iniotaki 1 1 Immunology Department and National Tissue Typing Center, General Hospital of Athens, G. Gennimatas, Athens, Greece, 2 Transplantation Unit, Laikon Hospital, Athens, Greece Correspondence: avittor2@yahoo.com Anti-HLA donor specific antibodies (DSA) and ABO incompatibility (ABOi) are the major immunological barriers for successful kidney transplantation. The Paired Kidney Donation (PKD) program enables kidney transplant candidates with incompatible living donors to join a registry of incompatible pairs in order to find a suitable donor. The PKD program in Greece was started on 2016 at Laikon Transplantation Unit with the support of the National Tissue Typing Center. Here, we present the results from the first four transplantations. In total, 15 patients with either one (n=12) or two (n=3) living donors have been registered to the program. Patients and donors were typed for HLA-A, -B, -C, -DRB1, -DQA1, -DQB1 (PCR-SSO) and anti-HLA antibodies were defined by Luminex, Single Antigen Bead analysis (One Lambda). Nine patients had preformed DSA (HLAi) with positive actual CDC and/or T/B flow crossmatch and six patients were ABOi with their intended donors. The pairs recorded in an Excel-based database and were compared in a pairwise, all-versus-all fashion according to ABO compatibility and absence of DSAs. Four pairs were identified to match (2 HLAi and 2 ABOi) in a twoway exchange program. One of the two HLAi patients was highly sensitized (PRAs=98%). The CDC and T/B flow crossmatches between the selected pairs were negative. The two-way exchange transplantations were performed the same day consequently without undesirable events. One recipient, 10 days after transplantation, experienced an acute cellular rejection episode 4-2A according to Banff classification, which was successfully treated with corticosteroids. Six months post-transplantation the four recipients had serum creatinine 1.2, 1.3, 1.5 and 1.5 mg/dl respectively and all donors have normal renal function. These results confirm that KPD program help to overcome immunological barriers in renal transplantation, increase access to transplantation especially for HLA sensitized patients and should be incorporated to the National Transplantation Program. Nephrologists rely on the MFI for IgG-type Donor Specific Antibody (DSA) identification tests to assess the risk of graft loss by humoral rejection. However, there is heterogeneity in their pathogenicity. Following a first DSA research study in renal transplant patients 50 patients with humoral rejection at biopsy (Banff score g + cpt> 3; g> 0 and cpt> 0) and 7 patients without rejection with a DSA were included. A Genprobe test to measure the ability of DSA to fix C3d in biopsy day serum. In the humoral rejection group 27 DSA C3d + were identified, in 20 patients: seven in class I and 20 in class II, mostly anti-HLA-DQ. In univariate analysis, DFG at the time of renal biopsy was found inversely associated with graft loss (p = 0.0008). The presence of a DSA fixing the C3d (p = 0.025), the number of DSA fixing the C3d (p = 0.0206), the MFI of the strongest DSA (p = 0.0162) and the sum of the MFI of the DSA fixing the C3d (p = 0.0106) were found to be positively associated with graft loss. No significant associations were found between the presence of DSA, the strongest DSA MFI, the sum of DSA MFIs, and the number of DSA and graft loss with either Genprobe or Ingen IgG assays. The Banff "cg" score was found to be positively correlated with graft loss (p = 0.0052). The C4d score was not found to be significantly associated with graft loss, but close to the threshold (p = 0.0688). In multivariate analysis only GFR at the time of biopsy was identified as an independent risk factor for graft loss (p = 0.0004). In the subgroup of patients with Genprobe DSA and acute rejection (cg < or equal 1 score), in multivariate analysis, GFR at the time of biopsy and the presence of a C3d-binding DSA were found to be risks independent of graft loss (p = 0.0092 and p = 0.0198). There is a good correlation between Genprobe MFIs and C3d result (r = 0.75) and the agreement of the two tests is good (ICC = 0.74 with 95% CI 0.6391 to 0.8176). The C3d test would provide additional prognostic information for Single Antigen IgG type tests. It would, combined with DFG, at the time of diagnosis of acute A previous gene expression microarray study in preimplantation biopsies (PIB) from kidneys of 53 deceased donors (GeneChip Human Gene 1.0 ST Arrays, Affymetrix; GEO accession number GSE54888) showed higher expression of HLA class II genes in cases that presented poor (eGFR <45 ml/min/1.73 m2) graft function (GF) at 1-Y post-Tx. The present study aimed to: (a) investigate the relationship between DPB1 expression in these PIB and 5-Y GF; (b) validate, in a different cohort, by real-time PCR (RT-PCR), the association between high HLA-DPB1 expression in PIB and 1-Y poor GF; (c) investigate the relationship between SNP rs9277534 alleles and DPB1 expression in PIB; (d) investigate the relationship between the presence, in the donor, of the high expression rs9277534 allele (G) and the 1-Y GF in Tx with one or two DPB1 mismatches (MM). RT-PCR was performed with TaqMan Assays (Applied Biosystems); HLA-DPB1 typing was done with PCR-SSO (One Lambda); rs9277534 alleles were determined with TaqMan Assay C_29841700_20. High HLA-DPB1 expression in PIB from the microarray study was associated with poor 5-Y GF (p<0.05). The association between high DPB1 expression in PIB and poor 1-Y GF was confirmed in a different cohort, with 51 cases with poor and 58 cases with good 1-Y GF (p<0.05). PIB from donors homozygous for the high DPB1 expression SNP allele (n=19) presented higher DPB1 expression levels compared to PIB from donors homozygous for the low DPB1 expression allele (n=44) (p<0.01). No significant difference was observed in the proportion of cases with poor 1-Y GF between transplants with the high or the low expression allele in the MM DPB1 antigen. There is an association between increased expression of HLA-DPB1 in PIB and poor 1-and 5-Y graft function and despite the association of the rs9277534G allele with high DPB1 expression in PIB, the presence of this allele in the mismatched DPB1 allele was not associated with poor 1-Y GF. Mean fluorescence intensity (MFI) values from Luminexbased HLA antibody assays are used to determine the relative strength of HLA antibodies. However, MFI values can be inaccurate, especially when the patient's serum contains heterophile antibodies that are present in up to 40% of normal blood donors. Heterophile antibodies have been shown to cause false-positive results in Luminex-based immunoassays that use animal proteins (bovine serum albumin and casein) to block the reactive sites of the beads. However, these proteins provide a potential source of assay interference (false-positive results) as a result of heterophile antibody binding. The aim of this study was to determine whether fetal bovine serum (FBS) would be an effective and reliable way to reduce heterophile antibody interference in Luminex-based HLA antibody assays. Toward this, we tested selected sera (n=5) containing heterophile antibodies by means of the LABScreen Single Antigen assays (One Lambda, Inc.). The sera were pre-treated with FBS (5% and 10%) and normal human serum (NHS) (5% and 10%), and their MFI values were compared to the ones from untreated samples. In addition, samples (n=20) without heterophile antibodies and containing well-defined HLA antibody profiles were also pre-treated with FBS (5% and 10%), and their MFI values were compared to the ones from untreated samples. We show herein that pre-treatment with FBS (both 5% and 10%) eliminates heterophile antibody interference (>1000 MFI) in several HLA class I (A29, A68, B67, and B76) and HLA class II (DR4, DR16, DRB3*02:02, DQA1*02:01/DQB1*03:01, and DPB1*19:01) specificities. On the contrary, NHS did not show any effect on the heterophile antibody interference in any of the samples. Moreover, FBS did not have any effect on the HLA antibody profile of sera displaying well-defined HLA antibody specificities. The results of this study indicate that FBS-treatment is a costeffective and simple procedure to eliminate heterophile antibody interference in Luminex-based HLA antibody assays. This interference can lead to a significant disadvantage for selected patients displaying heterophile antibody interference. Serum pre-treatment with FBS would prevent these patients from being inappropriately excluded from receiving an organ due to false-positive virtual crossmatch results. We aimed to study the distribution of killer cell immunoglobulin-like receptors (KIRs) on NK cells of human leukocyte antigen (HLA)-matched unrelated donorrecipient pairs receiving hematopoietic stem cell transplantation (HSCT). Sequenced-based typing (SBT) was used to type the HLA genotypes of donor-recipient pairs and sequence-specific-primer polymerase chain reaction (SSP-PCR) was selected for KIR typing in 46 HLA-matched unrelated donor-recipient pairs. The KIR genotypes were found in all 46 HLA-matched unrelated donor-recipient pairs and the distribution frequency of the 16 genotypes showed no significant difference in donors and recipients. In KIR haplotypes, the distribution frequency of AA, AB and BB was 29.34%, 23.91% and 46.75%, respectively. Eleven cases of KIR-HLA mismatches were identified, accounting for 24%. No significant difference in the distribution frequency of KIR genes of donors and recipients was found. In HLA-matched unrelated donor-recipient pairs, the rate of KIR-HLA mismatch is low. Loss of heterozygosity (LOH) can cause erroneous homozygosity of human leukocyte antigen (HLA) typing in pretransplant patients with hematological malignancies. A 40year-old woman was diagnosed with stage IV lung cancer in April 2015, and thus received concurrent chemoradiotherapy and lobectomy surgery. Because of peripheral blood blastosis discovered in her regular follow-up care in May 2017, she was referred to the hematological department at National Taiwan University Hospital. After bone marrow (BM) study, the patient was diagnosed with acute myelomonocytic leukemia expressing CD34/HLA-DR/CD117/ CD7/ CD11b/ CD38/ cytoplasmic MPO. Cytogenetic study revealed a normal karyotype, while an NPM1 mutant gene was detected with the level of 60%. The peripheral blast cells were at 52% and HLA sequence-based typing (SBT) revealed homozygosity at three HLA class I loci and heterozygosity at HLA-DRB1 and -DQB1 loci. She then received chemotherapy. Her HLA types were reconfirmed when the blood leukocyte count was 3.6 K/uL with 86% neutrophils and <1% blasts and NPM1 mutant level was about 0.15% in BM. Homozygosity for HLA class I loci was still reported. However, HLA typing results of her parents indicated that she indeed lost her maternal alleles of HLA-A, -B, and -C loci. We further used the HLA SBT group specific primers for the maternal alleles, and finally demonstrated the patient was heterozygosity at HLA class I loci. In the case, the maternal alleles of HLA class I loci were not detectable in the blood samples with high (52 %) and really low (<1%) blast cells, suggesting that LOH affected the most of leukocytes, including blasts and neutrophils. Erroneous homozygosity results due to LOH would cause erroneous histocompatibility, and then could result in fatal consequences. HLA laboratories should be aware of any HLA homozygosity results of patients who will undergo hematopoietic stem cell transplantation. Homozygosity for HLA loci is reported only when proven by a family study.

Rayna Ivanova Derin 1 , Pascale Loiseau 1 , Kahina Amokrane 1 , Sophie Caillat Zucman 1 1 APHP Hôpital Saint Louis, Paris, France Correspondence: rayna.ivanova-derin@aphp.fr Because of our important and growing activity of HLA typing and DNA extractions respectively (about 50 samples/ day) we introduced in 2013 an automated DNA purification using the QIAsymphony system and the DNA MIDI Kit (QIAGEN). During our quality validation study, we found two minor problems with extracted DNA. The spectrophotometric analysis showed a deviation at 260/230 ratio measurement, suggesting organic, salts or solvents contamination. Accelerated ageing tests (according to the Biomatrica protocol) showed that DNA degradation was more important compared with other purification methods (FUJIfilm DNA purification system and LabTurbo 24 Compact System) This problem was later solved by QIAGEN. We did not observe any effect on the different PCR-based HLA typing techniques used in the laboratory (PCR-SSP and PCR-SSO). However, consequences of DNA degradation after long-term storage remained an open issue. Moreover, since we store DNA samples at +4C (and not at -20C as per manufacturer's instructions, a regular quality control was required. At the end of 2017, we tested nine DNA samples extracted in 2013 and stored at +4C. We checked their quality and integrity using several criteria: monitoring of DNA degradation by agarose gel electrophoresis, PCRbased HLA typing by PCR-SSO (Luminex technology), and accelerated ageing. The results were compared to those using DNA samples extracted in 2013 with the QuickGene 610L system (FUJIfilm). We did not observe DNA degradation on agarose gel electrophoresis, or impact on PCR-based amplification and HLA typing with both DNA purification methods. The only difference we observed was a lower accelerated ageing using FUJIfilm DNA extraction. In conclusion, the quality of DNA samples extracted with the fully automated QIAsymphony purification system remains correct after 5 years. Storage at +4C does not affect DNA integrity and stability. Repeated studies will be required in the next few years to monitor DNA quality over time. Acute graft-versus-host disease (aGVHD) is a leading cause of morbidity and transplant-related mortality post-allogeneic hematopoietic stem cell transplantation (AHSCT). It is a severe inflammatory complication involving T-cell activation that plays a crucial role in its pathogenesis. MicroRNA-146a (miR-146a), a modulator of the immune system, plays an important role in the regulation of inflammatory innate immune responses and is an important negative regulator of T cells. Nuclear factor Kappa Beta (NF-kB), a transcription factor in inflammatory processes, regulates the expression of miR-146a that in turn regulates NF-kB activation through a negative feedback loop involving adapter molecules. Dysregulation of mir-146 and NF-kB activity is associated with inflammatory disorders and autoimmune diseases. In the present study we analyzed the potential association between polymorphisms in miR-146a (rs2910164, G>C) and NF-kB1 (rs28362491, -94 ins/del ATTG) and risk of aGVHD. Genotyping of these polymorphisms was performed in 75 HLA-matched related donors using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). All patients received a myeloablative conditioning regimen with busulfan and cyclophosphamide, and cyclosporine and methotrexate as GvHD prophylaxis. Incidence of aGVHD grades IIIV in the whole cohort was 18.6%. We did not find a significant association between miR-146a and NF-kB polymorphisms and aGVHD (p>0.05). Combined genotype analysis of these genes did not show significant association with aGVHD (p>0.05). Our preliminary data do not support the association of these molecules with aGVHD. More research is needed to outline the possible mechanism of miR-146a and NF-kB in the underlying inflammatory reaction in aGVHD, including study of a larger cohort and inclusion of other molecules involved in this pathway. Combined immunodeficiency due to dedicator of cytokinesis 8 protein (DOCK8) deficiency is a form of T and B cell immunodeficiency characterized by recurrent cutaneous and sinopulmonary viral infections and susceptibility to cancer. Laboratory findings include elevated IgE level in serum and eosinophilia. T cell lymphopenia and reduced T cell function have been reported. Patients can also have NK and B cell lymphopenia. IgG levels may be elevated but specific antibody responses are impaired. IgM and IgA levels may be reduced in DOCK8 immunodeficiency syndrome, referred to as autosomal recessive hyper IgE syndrome (AR-HIES) and considered a rare form of HIES. Therefore, we present three cases with DOCK8 immunodeficiency syndrome who received an HLA-identical sibling donor bone marrow transplant (BMT). The first (age:12 years) and second (age:1 year) cases were siblings and their parents were first cousins. All three cases had recurrent pulmonary infection, erythematous and eczematous rash on the whole body and hepatosplenomegaly. Furthermore, the first case had bilateral chronic supurative otitis and a mass in the left auricular. A third female patient (age:10 years) also had a clitoral vegetative mass. Immunological investigations of three cases revealed that mean absolute neutrophil count was 4553.33/ mm 3 , mean absolute lymphocyte count was 4053.33/mm 3 , mean absolute eosinophil count was 7274.33/mm 3 , mean IgG level was 1091.32 mg/dl, IgA level was 92.33 mg/dl, IgM level was 133.2 mg/dl and IgE level was 7153.33 IU/ml. Lymphocyte subsets in all cases were analyzed using flow cytometry to determine the percentage of CD 3, CD4, CD8, CD16+56 and CD19 positive cells. No patients had abnormally low CD3, CD4, CD8, CD16+56 or CD19 counts (absolute counts below 200). Therapeutic measures included antiviral and antibacterial prophylaxis, immunoglobulin replacement and allogeneic BMT. The auricular and clitoral masses after ABMT regressed spontaneously and they are follow-up in our clinic. To conclude, early BMT should be strongly considered as a potential curative therapy in DOCK8 immunodeficiency syndrome. Here we describe a case of a male patient with a hematological malignancy (ALL) transplanted with peripheral blood stem cells from a female donor. In our laboratory posttransplant chimerism is monitored with short tandem repeat (STR) analysis using the PowerPlex Fusion kit, 23 loci + amelogenin (Promega) and occasionally the Identifilier kit, 15 loci + amelogenin (Life Thecnologies). The pretransplant patient study identified the absence of AMELY allele in three DNA patient samples collected in three different pre-transplant periods (two extracted from peripheral blood and one from buccal cells). AMELY loss and its absence in the female donor makes this locus not informative for the chimerism quantitation. We extended the study of Y chromosome to other methods: cytogenetics, Y chromosome microdeletion analysis (Yq11-AZF a, b, c regions) and STR PowerPlex Y23 System (Promega) for the co-amplification of 23 male-specific STR loci (DYS576,  DYS389I/II, DYS448, DYS19, DYS391, DYS481,  DYS549, DYS533, DYS438, DYS437, DYS570, DYS635,  DYS390 , DYS439, DYS392, DYS643, DYS393, DYS458, DYS385a/b, DYS456 and Y-GATA-H4). The cytogenetical analysis showed a normal 46, XY karyotype in 20/20 mitoses and fluorescent PCR (Devyser AZF_V2 kit) demonstrated the absence of AZFa, AZFb, AZFc microdeletions. The Powerplex Y23 analysis evidenced the deletion of the following markers: DYS576, DYS570 and DYS456. All these markers lie on the short arm of Y chromosome, p11.2. Analyzing the markers examined by the different methods, we observed a large Y p11.2 deletion (DYS456-AMELY-DYS570-DYS576), spanning at least a 2.9 Mb region. The nearest markers conserved are DY393 and DYS458, telomeric and centromeric respectively to the deletion. This study allowed us to obtain further information on the deletion localization on the Y chromosome. From this work we highlight the importance of using different available methods and collaboration with external laboratories in order to solve difficult cases. Many solid tumors are characterized by a loss or decrease in the expression of HLA antigens on the surface of tumor cells due to the complete or partial deletion of HLA genes from one of the chromosomes. In lymphoproliferative diseases, changes in HLA antigen expression are less common. A 46-year-old man diagnosed with PTCL-NOS, stage IV, with the involvement of lymph nodes, spleen, bone marrow and leukemization attended National Research Center for Hematology, Moscow, in 2017. After two courses of induction chemotherapy partial response was achieved, the size of the lymph nodes decreased, however persistence of the residual tumor clone in the peripheral blood and bone marrow was noted. Therefore, transplantation of allogeneic hematopoietic cells was suggested as first-line therapy regimen and search for matched unrelated donor was initiated. Haplotype-specific sequencing of the HLA genes from the DNA obtained from the patient's blood cells, revealed a heterogeneous C/T position in the codon 180 of the HLA-A*02 allele in the exon 3. The chromatogram of the second allele (HLA-A*03) was clearly homogeneous. Thus, the obtained data indicated the presence of 3 alleles in the HLA-A locus in the patient: A*02:01:01G; A*02:227N; A*03:01:01:G. When DNA was sequenced from another blood sample, two homogeneous chromatograms of the A*02:01:01G; A* 03:01:01G alleles were obtained. An additional analysis of the HLA-A genes in DNA obtained from a patient's lymph node biopsy was performed resulting in an unambiguous answer -A*02:227N; A*03:01:01G. In the presented case, a third additional HLA-A allele was identified, probably as a result of a mutation in tumor cells. Nucleotide replacement of C>T at the indicated position results in the formation of a stop codon and loss of expression of the HLA-A*02:01:01G gene in tumor cells. Due to the possible influence of the tumor clone on the results of HLA typing care should be taken to perform HLA-typing on tumor-free samples. The human Major Histocompatibility Complex (MHC) lies within the short arm of chromosome 6 and is responsible for the production of Human Leukocyte Antigens (HLA). The HLA genes are the most polymorphic genes in the human genome, encoding over 17000 allelic variants. Our laboratories are involved in HLA tissue typing of patient, their relatives and volunteer donors for the Italian Bone Marrow Register. DNA has been extracted from EDTA blood samples using automated system Maxwell 16 (Promega). PCR-SBT has been performed using SBTExcellerator GenDx and Protrans S4 HLA-B. During HLA-B high resolution typing of our relative potential donor, we found a sequence largely homologous to B*14:02:01 with a mutation at position 506 a C instead a G. This mutation results in a coding change from Arginine to Proline in exon 3 (codon 145b CGC!CCC). We repeat this sequence in isolation from the second allele and analyzed by SBTengine 3.14.0.2783GenDx. We couldn't study the parents due to unavailability. The described mutation is not unique but it has been observed in other HLA-A, -B and -C alleles but not in HLA-B*14. This allele was then submitted to WHO Nomenclature Committee via the IPD-IMGT/HLA Database for naming and was assigned the official name B*14:58. We consider sequence-based typing in both directions a good method for rapid new allele definition. High-resolution HLA matching reduces graft-versus-host disease and improves overall patient survival after hematopoietic stem cell transplantation. Sanger sequencing has been the gold standard for high-resolution HLA typing since 1996 until next-generation sequencing was developed. In our laboratory we use Luminex methodology for HLA typing of patients and donors and, when a sequencing study is required, samples are sent to laboratories abroad. Therefore, implementation of next generation sequencing HLA typing is urgently needed in our country. Our aim was to implement clinical HLA-NGS typing, and to validate the protocol clinically used at our lab. We performed HLA-NGS typing on 48 samples. Libraries were prepared with the Illumina Trusight HLA v2 Sequencing Panel and then sequenced on the MiSeq system. Results were analyzed with Assign 2.1 software. First, we ran four batches of six samples each. Three out of the 24 samples were from the UCLA exchange and the remaining ones were from patients previously typed by Luminex at our lab. To validate the procedure, we tested a 24-sample panel from the 17th International HLA and Immunogenetics Workshop. We identified two new alleles in two samples from our patients (i) a silent substitution in exon two of B locus and (ii) a non-synonymous substitution in exon two in DPB1 locus. In five samples, the alleles observed were consistent with the intermediate resolution typing, but they were among the less expected ones according to the frequency distribution of these alleles in Caucasian databases. Ambiguous results were most common at the DPB1 locus according to the limitations of the methodology. In this study we confirmed that NGS is a highly accurate and fast method to perform high-resolution HLA typing since very low ambiguity levels were found and a significant decrease of turnaround time could be achieved. PRICAI is the first lab in Argentina which introduced this method to confirm matches for bone marrow transplantation patients. Further work will be done soon to establish local HLA frequencies at allelic level. We report a 23 year old male patient with sickle disease, frequent pain attacks, multiple hospital admissions, no history of stroke, priapism or acute chest syndrome. The patient received multiple transfusions and had a hemoglobin baseline of 9 g/dl, with low MCV (62 fL) and high RDW (23.1 %). His blood film showed microcytic hypochromic picture with notable presence of sickle cells. His hemoglobin electrophoresis showed Low A1-Hb (17%), F-Hb (8.6%) and S-Hb (69%). The patient offered related allogeneic BMT as possible treatment. Patient HLA typing for class I and II was performed by using One Lambda sequence specific oligonucleotide probes and sequence specific primers (SSOr/SSP). Six family members had their HLA typing for class I and II performed, including his parents as possible donors. The parents are cousins. The results based on haplotype illustration are as follows: Patient (A*24, 24, B*48, 50, C*08, 06, DRB1*07:01, 07:01, DQB1*03:03, 02:02), Father (A*01, 24, B*50, 48, C*06, 08, DRB1*07:01, 07:01, DQB1*02:02, 03:03), Mother (A*24, 68, B*51, 50, C*15, 06, DRB1*04:03, 07:01, DQB1*03:02, 02:02), Brother (A*24, 68, B*48, 50, C*08, 06, DRB1*07:01, 07:01, DQB1*03:03, 02:02), Sister-1 (A*24, 24, B*48, 51, C*08, 15, DRB1*07:01, 04:03, DQB1*03:03, 03:02), Sister-2 (A*24, 68, B*48, 50, C*08, 06, DRB1*07:01, 07:01, DQB1*03:03, 02:02), Sister-3 (A*01, 24, B*50, 51, C*06, 15, DRB1*07:01, 04:03, DQB1*02:02, 03:02). All tests were confirmed by repeating the tests with new samples and possible sampling error was excluded. Further investigations by family study analysis showed a cross over in the region of paternal A locus, as the patient showed A*24 instead of expected A*68. Possibility of rare cross over must be considered in the interpretation of unexpected HLA results of families worked up for BMT. The illustration of HLA haplotypes and the significant similarity of HLA typing in class I and II in the parents explained the presence of single mismatch in class I between the patient and sister-2. Hematopoietic stem cell transplantation (HSCT) is the best therapy for hematologic malignancies such as acute leukemia. Although, graft versus-leukemia (GVL) is an important approach for prevention of relapse, it could be associated with severe graft versus host disease (GVHD). Similar to T cells, NK cells may facilitate engraftment, combat infection, and control malignant cells but do not cause GVHD. KIRs are a family of inhibitory/activating receptors that are expressed mainly by NK cells which bind HLA as ligands. Because of high variability in KIRs, careful selection of donors based on HLA and KIR is essential to optimize HSCT outcomes. Since donor KIRs are potentially encountered in non-self-recipient HLA class I epitopes, this study was designed to find out the impact of KIR/HLA combinations on the outcome of HSCT in patients with acute leukemia. From March 2015 to March 2015, a total of 31 patients with acute leukemia who received HSCs were followed for one year by chimerism analysis (30, 60, 90, 180 and 365 days after HSCT). During this period, 17 of these patients were excluded from further analysis due to death of GVHD or GVL. KIR genotyping of the donors and their corresponding ligands in recipients was also investigated. All alive patients received HLA-A, -B, and -DRB1-matched HSCs from their siblings. Three patients received HSCs from opposite sex, four patients from ABOincompatible and one patient from Rh-incompatible donor. Chimerism was not observed in none of the recipients at the studied time points after HSCT in 14 patients who were alive at least for the first year post-HSCT. All donors had KIR-B haplotype but just five of them carried KIR2DS1 while seven patients were C2/C2, three patients were C1/C2, four patients were C1/C1. Our results showed that HSCT from donors with KIR-B haplotype to C2-C2 or C1-C2 bearing recipients was associated with a better outcome during the first year after transplantation. Obscure points will be cleared by analysis of chimerism and KIR/HLA combinations in the samples of expired patients. Aplastic anemia (AA) is an autoimmune bone marrow failure syndrome closely linked to clonal hematopoiesis. This mechanism allows hematopoietic stem cells to escape recognition of the cytotoxic T lymphocytes, giving an advantage in disease progression. Identification of the specific clonal changes such as loss of human leukocyte antigen alleles helps diagnosis bone marrow failure and sometimes can be predictive of response to immunosuppressive therapy. A 10-year old male previously diagnosed with AA was referred to clinic with relapse after two previous courses of Atgam and an unrelated donor search was initiated. However, after the third block of immunosuppressive therapy he achieved remission and HSCT was postponed. Cytogenetic analysis was normal. HLA-typing by SBT indicated that the patient was homozygous for HLA class I alleles: HLA-A*24:02, B*07:02, Cw*07:02, DRB1*07:01, 15:01, DQB1*06:02, 02:02. A second re-typing by NGS for 11 HLA loci (TruSight HLA v2 Sequencing panel, Illumina) revealed no matches for alleles HLA-A and HLA-B and a heterozygous result for HLA-C with second rare allele. Detailed review of data showed the presence of secondary HLA alleles (HLA-A*02:01, B*13:02 and C*06:02) at about 12% depth of coverage. This case shows that loss of heterozygosity can lead to incorrect typing results. Homozygous results in patients with AA should be confirmed by typing of buccal swabs, family members or using other typing methods if possible.

Olga Shragina 1 , Viktoria Zakharova 1 , Elena Raykina 1 , Michael Maschan 1 , Alexei Maschan 1 1 Dmitry Rogachev National Research Center of Pediatric Hematology, Oncology and Immunology, Moscow, Russian Federation Correspondence: olga.shragina@mail.ru

The correct HLA-typing results of patient and donor are the first necessary step on the way to successful hematopoietic stem cell transplantation. Even the difference of one nucleotide can cause a locus mismatch in the donor-patient pair and induce worse outcomes in overall survival, treatmentrelated mortality and acute GVHD. The huge allelic multiplicity characterizing genes of HLA complex poses serious problems for the routine genotyping of these loci. Until recently, genotyping of HLA has been carried out using various methods based on PCR: SSP (sequence-specific primers), SSOP (sequence-specific oligonucleotide probes) and SBT (Sanger's sequence-based typing). These methods do not allow the determination of chromosomal-phase between assessed sequence features, which results in considerable ambiguity regarding an HLA genotype. HLA typing methods based on next generation sequencing (NGS) allow amplification of the full gene including exons and introns, minimizes the number of ambiguous results, increases amount of sequencing reads per sample and locus, allows the study of a large number of samples simultaneously and significantly reduce the per sample cost of typing. Since the implementation of the HLA-typing by NGS (HLA-A, -B, -C, -DRB1, -DQB1) 700 tests have been performed on patients and their potential related donor (Miseq Illumina platform, NGSgo (GenDx), NGSengine (GenDx) software). Even in such a small number of samples, 7 novel alleles (A*11:01:79, C*06:02:50, C*07:593N, C*17:38, DQB1*02:02:05, DQB1*03:37, DRB1*13:01:17) were detected. When novel alleles are detected, no additional studies are required. The sequence of nucleotides is easily unloaded from the analysis program. Rapid progress in NGS technology has resulted in significant changes in medical genomics. This method enables us to accelerate polymorphism discovery, expand the number of known alleles and improving our understanding of allelic diversity. Human leukocyte antigens (HLA) play an important role in the immune responses that determine the outcome of a transplant. Allele-level HLA matching between donors and recipients reduces the likelihood of rejection and graft-versus-host disease. The HLA genes are characterized by a diversified polymorphism that affects the specificity of antigen presentation and histocompatibility in transplantation. In recent years, next-generation sequencing (NGS) technology allows us to provide the highest possible resolution. Most of the NGS methods are based on traditional PCR of the target regions followed by massive parallel sequencing of the amplicons. However, amplicon-based methods are laborious and require extensive optimization steps. The aim of our study was to determine 11 HLA loci (HLA-A, -B, -C, -DRB1/3/4/5, -DQB1, -DPB1, -DPA1, -DQA1) using AllType NGS 11-loci Amplification Kit (One Lambda). DNA samples isolated from peripheral blood were obtained from 6000 bone morrow donors with the use of the chemagic Prepito instrument. Typing of 11 HLA loci was performed according to the HLA assay manual using the MiSeq Illumina platform. The raw sequencing data were assembled and analyzed by the TypeStream Visual NGS Analysis Software. In our study, this new assay allowed single tube amplification for all 11 HLA loci, completely removing the need for amplicon pooling. Besides, samples were transferred directly to library preparation and made ready for sequencing in a single workday. In addition, we also confirmed that the advantage of NGS with the use of new Alltype kit was a possibility to obtain the unambiguous, high-resolution genotyping results within two days. Anti-HLA immunization has been correlated with the risk of graft failure in allo-HSCT. Then, it is strongly recommended to screen recipients for anti-DSA, especially with HLA mismatched or haploidentical familial donors. The objective was to analyze the incidence of anti-HLA abs and DSA in the sera of recipient candidates to allo-HSCT between 2014 and 2016 in Hospital Donostias Transplantation Unit. 151 allo-HSCT have been performed; 39 of them from HLA identical donors; 61 out of 151 from 10/10 (8/8) matched unrelated donors (UD) and 25 HSCT from 9/10 UD. 25 Haplo HSCT were carried out. HLA-A, -B, -C, -DRB1, -DQB1 typing, screening of HLA specific abs on Luminex platform and CDC and cytometry crossmatches were performed in positive cases. The results of the study were: UD mismatched (9/10) HSCT: 2 out of 25 patients (8%) presented positive anti-HLA abs. (anti-B and anti-DPB1 respectively and MFI >1000), none had DSA. Haplo-identical HCST: 5 out of 26 recipients (19%) had positive results in HLA abs. One patient was DSA negative but four of them had positive DSA against the haploidentical donor candidate (HLA-C, -DRB1; -DRB3; -DRB4 and -DQB1 alleles, MFI between 1500-9000). Crossmatches could only be performed in two of the DSA positive cases. One cytometry crossmatch was positive in lymphocyte B according to the anti-class II DSA found. According to recommendations in allo-HSCT and haplo-HSCT protocols, all the donors with HLA specificities, targets of positive DSA (MFI>1000) recipients, were excluded and in all cases alternative donors were found. Although only a few positive DSA cases have been found, the incidence of anti-HLA and DSA found is similar to other reported studies. All the patients should be screened for anti-HLA abs when they enter the HSCT program in order to be informed of the sensitization status of the recipient if several donor options are proposed Sequencing based typing (SBT) is considered to be a gold standard for high resolution HLA genotyping. However, like other HLA typing technologies, it has certain limitations. Next generation sequencing (NGS) is a new method for both clinical and research laboratories that is being implemented into the routine practice worldwide. An introduction of NGS (GenDx, Omixon) in our laboratory led to the detection of a novel polymorphism in the HLA-C locus of a patient indicated for stem cell transplantation straight at the beginning of the testing. For the method validation we selected samples that included rare alleles, one of which was HLA-C*06:19 (combined with C*02:02). The original typing was performed by SBT and PCR-SSP.

Using NGS due to the separation of allele sequences we discovered that the correct typing was C*06:02 and a novel allele C*02. The whole HLA-C gene sequencing revealed mismatches in exon 1 and exon 2 compared to C*02:02. Patient had no HLA-matched sibling and no fully matched unrelated donor was found in registries worldwide. Therefore, two unrelated donors (both C*02:02, 06:02 + one mismatch in other locus) were chosen for confirmatory typing. The transplantation from donor 1 (MM in HLA-C+HLA-DQB1) was successful besides slight complications one month after the HSCT. NGS techniques have their place in HLA laboratories. As the number of different HLA alleles is continuously increasing the demand for non-ambiguous typing and distinguishing existing from novel alleles is more and more challenging. NGS delivers a way of unambiguous typing of a whole HLA gene of each allele separately with a high accuracy which can improve the proper HSCT donor selection.

Maria Paola Albergoni 1 , Daniela Bovo 2 , Aurora Strano 2 , Roberta Favero 1 , Genny Franceschetto 2 , Giustina De Silvestro 1 , Roberta Destro 2 1 UOC Immunotrasfusionale -DIMT Padova -Azienda Ospedaliera di Padova, Padova, Italy, 2 Oncoematologia Pediatrica -Padova Cord Blood Bank -Azienda Ospedaliera di Padova, Padova, Italy Correspondence: daniela.bovo@aopd.veneto.it Stem cell transplantation from unrelated cord blood (CB) units for the treatment of patients suffering from hematological diseases has developed considerably from 2000 for about a decade. Recently the preferred choice of haploidentical transplantation has determined a consistent decrease of CB use all over the world. The rate of CB units shipped for transplantation by the Italian Cord Blood Network has decreased dramatically from 2010 to 2017. Nevertheless, many Transplant Centers (TC) are still using CB as preferred source of stem cells for hematopoietic stem cell transplantation, in particular for patients lacking a matched sibling or unrelated donor and/or requiring an urgent HPC transplantation. In fact, one of the advantages of using CB is the rapid procurement. From 2011 to 2017 Padova Cord Blood Bank (PDCBB) forwarded to different TCs 624 CB unit reports, on average, in less than one day from the Transplant Center request, and in 224 cases the HLA laboratory provided high resolution (HR) HLA typing results (for one or more loci), on average, in 9 days. 44 of 624 were selected for procurement and issued in accordance with the TC shipment time requirements with a minimum time of three days from the TC request. The HLA confirmation typing took an average time of 2 days (range 0-7). Our data show that the urgent procurements (shipment in less than 8 days from the TC request), increased from 28.57% in 2011 to the 100% in 2017. More than 60% of the issued CB units were typed at HLA HR level at the time of the first report request; this supported our choice of performing HR HLA typing (HLA-A, -B, -C, -DR) of all our CB units since 2013, when it was not yet mandatory for cord blood banks of the Italian Cord Blood Network, in addition to the banking of CB with higher TNC, as result our CBB is always ranking in the first positions for the CB release rate (released units/banked units). A prompt reply to TCs and the availabilities of quick HLA typing results are crucial elements for a CB Bank and for its HLA laboratory which needs to be actively involved in the strategy and policy of the CBB activities. So PDCBB and the HLA laboratory are planning to perform HR HLA typing on the remaining CB units with higher TNC banked before 2013. The time needed for identifying an unrelated donor is critical for the treatment of patients with malignant disease by hematopoietic stem cell transplantation (HSCT). The aim of this study was to compare the times required for a Turkish patient to receive a transplant from an unrelated donor from Turkish versus international registries. To this end, we retrospectively analyzed data from 57 patients who received unrelated donor HSCT at the Istanbul Medical Faculty Bone Marrow Bank (TRIS) between 2016 and 2017. Delayed or deferred transplantations (n=28) due to bureaucratic procedures or worsening of the clinical condition of the patient after the start of the donor search were excluded from the analysis. For the remaining 29 transplants, donors were from international registries (n=12) or from the national Stem Cell Coordination Center of Turkey (TURKOK) (n=17). The mean time from the donor search start to transplantation was 124AE8 days (71-175 days) for international donors and 66AE4 days (48-109 days) for international and Turkish donors, respectively. This was at least in part due to a shorter time for confirmatory typing, which was mean 47AE6 days (23-97 days) for international donors and 19AE2days (10-35 days) for Turkish donors. TURKOK has been established 3 years ago and our data show that it is working efficiently to reduce donor search times for Turkish patients. Chimerism is the key analysis of monitoring post hematopoietic stem cell transplantation (HSCT). When chimerism is performed on sorted cells, the EFI standards recommend documenting the purity of sorted cell populations. The current reference method for analyzing purity of sorting is the flow cytometry (FCM) method, but it is relatively expensive and requires a specialized laboratory. In this context, new Non-T Genomic Kit (Accumol, Calgary, AB, Canada), based on the quantification by real-time PCR of genomic rearrangements of TCR genes, undergone by T cells during their differentiation seems very interesting. The Non-T Genomic Kit method has been assessed via a range of samples initially purified with the EasySep Human Whole Blood CD3 Positive Selection Kit (STEMCELL technologies), in comparison to FCM assays using the CYTO-STAT tetraCHROME kit (Beckman Coulter). The purity of CD3+ cell sorted is estimated by calculating the ratio of the height of the two peaks (Non-T Peak / ultra-conserved sequences peak), obtained by dissociation curve after amplification. Contamination rates have been determined in 30 CD3+ of recipients with lymphopenia or lymphocytosis. The detection of non-T cell contamination by the Non-T Genomic qPCR Kit is correlated with that of the FCM, with a y = 0.0142x-0.017 equation (r= 0.98). The method is repeatable and reproducible with CVs less than 20% over the entire range of cell contamination, except for purified samples at the detection limit of the dissociation curve. It is capable of accurately detecting non-T cell contamination ≥ 1%. The patient sample series shows that CD3+ purification is reliable when the CD3+ cells / total lymphocytes ratio is ≥ 10%. In our hands, the qPCR Non-T Genomic Kit is an efficient tool, allowing accurate assessment of contamination levels in a CD3+ sorted cell sample. Routine use will help meet EFI requirements and improve the follow-up of HSCT recipient at a cost equivalent to that of FCM.

Sally Elfishawi 1 , Raafat Abdel Fattah 1 1 National Cancer Institute, Cairo University, Cairo, Egypt Correspondence: sally.elfishawi@nci.cu.edu.eg HLA-G molecules have immunosuppressive functions where HLA-G inhibits the lysis of NK cells, the alloproliferation of CD4+ T cells and the antigen-presentation of dendritic cells. It aids in the production of regulatory T cells and the apoptosis of CD8+ cells. Importantly, the 14 bp del/ins polymorphism in the 3 0 untranslated region of HLA-G (rs66554220) controls mRNA stability and thus HLA-G expression. Expression of human leukocyte antigen G (HLA-G) has been associated with graft versus host disease and stem cell transplantation related morbidity and mortality. In order to investigate the effect of the HLA-G 14 bp deletion on transplant outcome, we performed HLA-G genotyping (14 bp del) on a cohort of 52 patient/donor pairs transplanted from matched sibling donors at the BMT lab unit of the national cancer institute. All patients and their donors showed identical HLA-G 14 bp deletion status. Seventeen patient/donor (32.6%) were homozygous for the 14 bp deletion, they showed a trend towards a higher risk of developing acute graft-versus-host disease (aGvHD) compared to patients heterozygous or homozygous for the absence of the deletion for the 14 bp insertion (p=0.05). No association was found between HLA-G and relapse in those patients. Therefore, the 14 bp polymorphism could be an important predictive factor for aGvHD following bone marrow transplantation. Given the endless expansion of the IPD-IMGT/HLA Database, it is essential to properly document each entry in order to establish which alleles should be considered in routine HLA genotyping and which alleles should be ignored. This documentation should evaluate allele frequency in the main world populations and linkage disequilibrium. A good rule of thumb is to consider as common alleles with a gene frequency of g > 0.001 and an allele count of n > 2 in any well studied major population. This means that a 'well studied' population requires a sample of at least 1500 subjects and 3000 haplotypes to evaluate which alleles are really common in that population. Only well documented rare alleles should be considered for inclusion in the HLA allele catalogue used for routine genotyping. By 'well documented' we mean 1) the identification of a given allele at least 3-5 times in independent samples and different labs, 2) consistency in geographical origin; and 3) consistency in linkage disequilibrium. In this study we evaluate a number of rare alleles in regard to consistency in geographical origin and consistency in linkage disequilibrium. Major improvements in the results after allogeneic HSCT were driven by deeper insights into the immunology of stem cell transplantation and more sophisticated donor selection. Still, the failure rate after allogeneic HSCT is relatively high due to relapse or non-relapse mortality. Additional immunogenetic donor selection criteria that would increase the success rates are desperately needed. Currently, matching for HLA-A, -B, -C, -DRB1 and -DQB1 at the allele level is still considered the gold standard for immunogenetic donor selection.The application of next generation sequencing (NGS) has considerably decreased costs for high-throughput genotyping. In addition, it opens up completely new opportunities mainly because, in contrast to Sanger, the addition of other genetic loci increases total costs only minimally. On the other hand, the retyping of large numbers of samples at a later time point would require significant investments. Here, we propose applying a forward-looking profile for genotyping potential stem cell donors. This profile includes several factors that have been reported to improve donor selection but are currently not considered in clinical practice. In particular, besides the six HLA genes (HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1), the complete KIR gene family at allelic level resolution, blood groups ABO and Rh, CCR5, MICA/B and HLA-E. We anticipate that in the next years independent studies will confirm at least some of the candidate genes as relevant parameters for donor selection. By incorporating these factors into the genotyping profile now, several million donors will be available in a few years and will then facilitate quick implementation of those additional criteria thereby improving the outcomes of patients after matched unrelated donor HSCT. Hematopoietic Stem Cell Transplantation (HSCT) with HLA mismatched or haploidentical donors is for many patients the only curative option. Pre-existing donor specific anti-HLA antibodies (DSA) have been reported to be implicated in graft failure but their frequency and clinical impact remains unclear. The aim of this study was to evaluate the need of a protocol to manage hematological patients undergoing mismatched or haploidentical HSCT and to prevent negative outcomes in the presence of HLA-immunization. We evaluated the presence of DSAs, using Luminex technology (Labscreen Mixed and Single Antigen class I and II, One Lambda), in 68 adult patients undergoing unrelated mismatched HSCT (USCT) and haploidentical SCT (HaploSCT). HPA antibodies were detected by using an ELISA kit (Immucor GTI-PAK Auto). For patients with unrelated mismatched donor the time of assay was 30 days before transplantation, for haploidentical SCT the same time of confirmatory test. DSA were detected in three patients: two patients underwent HaploSCT and 1 USCT, without receiving desensitization treatment before transplant. All of them failed to obtain allogeneic engraftment. One patient had DSA against HLA-Cw10 (MFI 22400); the second patient had DSA against HLA-A2 (MFI 10500) and also anti-HPA1a antibodies associated with platelet transfusion refractoriness. The third patient had DSA against HLA-B50 (MFI 900) and anti-HPA1a antibodies. In our study DSA patients were associated with graft failure despite MFI value and platelet transfusion refractoriness representing a hallmark of DSA positivity. As HLA-immunization remains a problem in the management of hematological patients undergoing to mismatched transplantation, despite using leucodeplete units, a laboratory protocol will be necessary to monitor outcome transplantation. The aim of the current study was to analyze characteristics of MUD search procedures performed in our hospital in the last five years and to compare them to previous findings from earlier years (2007) (2008) (2009) (2010) (2011) (2012) . The tissue typing laboratory at Rabin Medical Center facilitated 270 MUD search procedures between 2013 and 2017, 244 of them resulted in an allogeneic transplant (226 MUD, 12 unrelated CB, 6 haploidentical). 92% of patients found a 9/10 or 10/10 alleles MUD. 81% of the MUDs were found in the LDR. A median of two donors were requested for high resolution (HR) typing to be performed throughout the search procedures. The likelihood of finding a donor in the LDR's was higher for Jewish than for non-Jewish patients (84% versus 65%; p<0.01). A 10/10 alleles MUD was found for 69%, 50%, and 37% of the Jewish, non-Jewish and Arab minority patients, respectively (p<0.05). The likelihood of finding a fully matched donor was higher for adults than for pediatric patients (69% versus 55%; p=0.04). In the current study a larger portion of patients were transplanted from 10/10 or 9/10 allele matched donors in comparison to the previous study (92% versus 85%, respectively; p<0.04). This elevation is possibly due to the overall growth of the LDR in the corresponding years. Current donor searches required a significantly lower number of HR typing requests to be carried out by the registries in comparison to the previous study. This could be related to the higher resolution of HLA data provided by the registries in recent years. Our former study showed a significant gap between the likelihood of adult and pediatric patients to find a fully matched donor. This inter-generational gap is still currently evident but at a lower extent than in the past. This finding may be attributed to the efforts of the LDR's to recruit younger donors from more diverse ethnic backgrounds. Our findings show that donor search procedure characteristics for Israeli patients are dynamic and are influenced by long term transformations in the donor search environment.

Katarina Stingl Jankovic 1 , Marija Maskalan 1 , Zorana Grubic 1 , Marija Burek Kamenaric 1 , Renata Zunec 1 1 University Hospital Centre Zagreb, Zagreb, Croatia Correspondence: kstingl@kbc-zagreb.hr Chimerism analysis is a routine test performed in order to monitor engraftment or relapse after HSCT. A gold-standard method for this test was STR analysis. However, emergence of more sensitive qPCR-based methods has made them a more preferable method for chimerism analysis. The aim of the study was to evaluate applicability of a qPCR based chimerism monitoring method. For that purpose, 74 recipient/ donor pairs have been tested using KMRtype Core kit (GenDx, Utrecht, the Netherlands). qPCR was carried out using Applied Biosystems 7500 Real-Time PCR System. Analysis was performed using KMRengine Chimerism Analysis Software. Among tested subjects, 36 patients were transplanted from a related donor, while 38 patients had an unrelated donor. Median number of informative markers (M) was calculated for patients and donors in both groups and was as follows: for patients with related donors M=4 (range 1-7); for related donors M=3 (range 0-7); for patients with unrelated donors M=6 (range 2-10); for unrelated donors M=6 (range 3-12). Analysis of informativeness for individual markers revealed that out of 30 tested markers, eight markers were informative in 15 or more cases for either patient or donor. However, only three markers (KMR013, KMR053, and KMR054) were placed among those eight for both patients and donors. Conversely, several markers showed a low level of informativeness as well as a high percentage (≈20%) of atypical results (KMR011, KMR035, KMR041, KMR057). In conclusion, the qPCR method has been shown to be adequate for majority of our patients. Additional testing was needed for only four patients with a related donor for whom a single informative marker was identified in the first analysis. Improvement of qPCR method applicability in comparison to the initial qPCR-based methods is encouraging and, although the level of PCR-STR method applicability has still not been reached, results of our study justifies implementation of qPCR in routine chimerism monitoring procedure. We aimed to investigate the distribution of HLA antibodies in patients after multiple platelet transfusions and the relationship between the outcome of platelet transfusions and HLA antibodies or structural epitope (eplets) in patients serum. A group of 42 children and 7 adult patients with hematologic diseases and long-term platelet transfusions were analyzed. 24h CCI10 suggested platelet refractoriness. HLA-A and -B were typed in the patients and donors. Single antigen beads were used to identify the specificities of HLA antibodies in the patients' serum. For the patientdonor pairs, eplet based matching was done by HLAMatchmaker. A total of 11 platelet transfusions for children and 7 platelet transfusions for adult were refractory. The results of HLA antibody specificities showed that 31 of 49 patients serum were antibody positive (63.27%). 24 of 42 children patients' serum were antibody positive (57.14%), while 7 of 7 adult patients serum were antibody positive (100%). At the allele level, 91 antibody specificities including 31 HLA-A specificities, 45 HLA-B specificities and 15 HLA-C specificities were detected in 24 positive children patients' serum. All the 97 antibody specificities including 31 HLA-A specificities, 50 HLA-B specificities and 16 HLA-C specificities were detected in 7 positive adult patients serum. The specificities of HLA antibodies in the serum of patients with long-term platelet transfusions exhibit certain characteristics: The most frequent specificities for HLA-A and-B were A*68:02 and B*15:12, respectively. The most frequent eplets were 12SMR, 163LG, 245VA, 267PE, 35Q. The analysis of antibody specificities in patient serum and the compatibility between patients and their donors at eplet level illustrated that 12SMR, 163LG, 9H and 71NT might be clinical important. All the results may need to be confirmed a larger set of samples. Recently, Killer Immunoglobulin like receptors (KIR) present on Natural killer cells (NK) have been implicated in affecting transplant outcomes. A total of 41 patient donor pairs undergoing T-cell replete HLA haplo-matched transplant at Tata Memorial Centre, Mumbai were included in this study. HLA typing and KIR genotyping were carried out using PCR-SSP. The median overall survival (OS) was 13.9 months and median relapse free survival (RFS) was 9.2 months. The patients receiving transplants from male donors resulted in better outcomes with OS (p= 0.007) and RFS (p= 0.005). The effect of KIR ligands between the patient donor pairs suggested that a mismatch between the ligands resulted in significantly better OS (p=0.04). However, the impact of HLA B ligands on transplant outcomes suggested that patients who did not carry the Bw4 ligand (Bw6), had favorable OS (mean 34 months) and RFS (mean 26 months), as compared to patients carrying the Bw4 ligand [OS (mean 21 months, p=0.32) and RFS (mean 13 months, p=0.10)], although this data was not significant. Patients carrying the C1C1 ligand for HLA C epitope with Bw6 had significantly better OS (p=0.01) and RFS (p=0.007).We further assessed the impact of mismatches between the donor KIR receptors and the cognate HLA ligand in the patient on transplant outcomes. We observed that mismatches in the graft versus host direction led to better OS (p=0.02). It was also observed that donor inhibitory KIR receptor mismatch with patient HLA ligand led to higher incidence of acute graft versus host disease (p=0.02). Further, patients with Bw6 ligand for the receptor KIR3DL1 present in donor resulted in better OS (p=0.04) and RFS (p=0.01). We conclude that although HLA match is of primary importance, knowing the patient KIR ligands and donor KIR genotypes could help in selecting the best donor and minimize transplant related complications. Transplantation of hematopoietic cells (HSCT) from unrelated donors to patients suffering from hematological diseases has become a well-established treatment modality. However, transplanted patients still face significant complications, including graft-versus-host-disease (GVHD). To understand better the varying outcomes in patients receiving transplants matched for alleles of the HLA-A, -B, -C, -DRB1, -DRB3/4/5 and -DQB1 loci, the role of HLA-DPB1 has come increasingly into focus. Due to a recombination hot spot between the HLA-DP loci and the rest of the HLA class II region, 75 % or more of otherwise HLA matched donors have HLA-DPB1 mismatches. HLA-DP expression levels have been associated with a SNP (rs9277534A/G) located in the 3´untranslated region (UTR) of the HLA-DPB1 gene and can predict outcome. We typed 46 HLA-A, -B, -C, -DRB1, -DRB3/4/5 and -DQB1 matched patients-donor pairs in addition for their HLA-DP alleles. All patients were diagnosed with acute myeloid leukemia. To assess mismatches optimally we performed full length Next Generation Sequencing (NGS) typing; in addition, we evaluated the rs9277534A/G SNP by analyzing sequences from the 3´UTR. Analysis of the NGS data revealed a high association of SNPs and HLA-DPB1 alleles. The clinical outcome in our patient group indicated a possible association of the SNP in mismatched patients and donors. In our small patient cohort we found significant effects on transplant related mortality (rs9277534G in patients, p=0.015) and the occurrence of chronic graft versus host disease (rs9277534A donors, p=0.04) applying univariate analyses. Although these data are in accordance with published results, a larger number of transplant are required for definite proof. Methodically, full length NGS typing allows in one single test the allelic resolution of HLA alleles and, additionally, the assignment of functional SNPs. Therefore, this typing strategy provides a comprehensive risk assessment before HSCT. The simultaneous presence of host-and donor-derived cells, termed mixed chimerism (MC), is observed in a large proportion of beta-thalassemia (thal) patient (pts) after hematopoietic stem cell transplant (HSCT). When MC is detected early after transplantation it often evolves towards complete chimerism (CC), and when the proportion of donor cells is very low, MC leads to graft rejection or persistent MC (PMC). Pts with PMC do not require red blood cell support and are cured from the disease. T regulatory type 1 (Tr1) cells are characterized by the co-expression of CD49b and LAG-3 and the ability to secrete IL-10 and have been associated with PMC. We investigated MC in 152 beta-Thal pts at early stage (on day 30) after allo-HSCT by monitoring the amount of residual host cells (RHCs) in BM and PB cells. We observed that 60% of the pts showed CC and 40% MC. Among MC pts, 35% presented MC level 1 (RHCs <10%); 3% MC level 2 (RHCs 10-25%); and 13% MC level 3 (RHCs >25%). Interestingly, 70% of transplanted pts who showed MC level 3 shortly after allo-HSCT rejected the graft, whereas around 80% of transplanted pts who displayed MC level 1 and 2 progressed to CC. Interestingly, 10% of pts with MC level 1 and 10% of pts with CC on day 30 post-HSCT developed PMC. One year after allo-HSCT, 13% of the pts rejected the graft, 77% developed CC, and 10% developed PMC and were cure from the disease. T-cell cloning of CD4+ T cells from peripheral blood (PB) of two PMC pts confirmed that a high frequency of Tr1 cell clones is present in the PB of beta-Thal patients with PMC. Moreover, using CD49b and LAG-3 we showed that the frequency of Tr1 cells correlates with the maintenance of MC in a beta-Thal pt that underwent haplo-identical HSCT. Based on these data we can conclude that Tr1 cells can be easily detected in PB and are reproducibly observed in the PB of beta-Thal pt with PMC, sustaining the hypothesis that these cells have a role in promoting long-term MC and tolerance in beta-Thal pts after allo-HSCT. In silico prediction of high-risk non-permissive donorrecipient HLA mismatches after unrelated donor (UD) hematopoietic cell transplantation (HCT) is an attractive yet elusive objective. Non-permissive T-cell epitope (TCE) group mismatches have been defined by alloreactive T-cell cross-reactivity patterns for 72/942 (7.6%) HLA-DPB1 (DPB1) alleles identified to date (TCE-X), including 52/80 (65%) DPB1 alleles reported as common or well documented (CWD) in the EFI and ASHI catalogues. A numerical functional distance (FD) scoring system for in silico prediction of TCE groups based on the median impact of key amino acid polymorphism in exon 2 was developed for all DPB1 alleles (TCE-FD), including 28/80 (35%) CWD alleles not assignable by TCE-X. Here, we compared clinical outcome associations of non-permissive DPB1 mismatches defined by TCE-X or TCE-FD in 2730 patients who received an 8/8 HLA-matched UD-HCT for AML, ALL, MDS or CML between 1999 and 2011. Concordance between the two models was 92.3%, with most differences arising from DPB1*06:01 and DPB1*19:01 differently assigned by TCE-X and TCE-FD. Only 9/28 (32.1%) of CWD alleles not assignable by TCE-X were found in this cohort, occurring in 33/2730 (1.2%) of pairs. In both TCE-X and TCE-FD, nonpermissive mismatches were associated with reduced overall survival (HR 1.15, p<.006 and HR 1.12, p<.03), increased transplant-related mortality (HR 1.31, p<.001 and HR 1.26, p<.001), as well as acute (HR 1.16, p<.02 and HR 1.22, p<.001) and chronic graft-versus-host disease (HR 1.20, p<.003 and HR 1.22, p<.001). Our finding that in silico prediction of non-permissive DPB1 mismatches based on experimentally-elaborated FD scores is feasible for any DPB1 allele with known exon 2 sequence and improves prediction of major transplant outcomes including acute and chronic graft-versus-host disease, has important implications for UD searches. TCE-FD is likely to become increasingly useful as new alleles are detected by refined tissue typing techniques, and with the extension of unrelated HCT to new ethnic groups. Our proof-of-principle observation opens new potential avenues for developing risk prediction models across the HLA system. Matching for 10/10 HLA-A, -B, -C, -DRB1, -DQB1 alleles and DPB1 epitopes between unrelated donors (UD) and recipients significantly improves the outcome of hematopoietic stem cell transplantation (HSCT). In case DPB1 epitope matching is not performed over 80% of 10/10 donor/recipient pairs are DPB1 mismatched due to the weak linkage disequilibrium between HLA-DPB1 and other HLA loci. Selection algorithms based on DPB1 T cell epitope groups can identify mismatches that might be tolerated and those that would increase risk of GvHD after HSCT. The aim of phase one (A) of our study, regarded as validation, was to define how often is it in practice possible to improve donor selection by including HLA-DPB1 epitope matching. We implemented new search algorithm in the routine and analyzed its impact on matching during phase two (B). This study included 71 (A) and 44 (B) patients for whom the search for UD was initiated between 2014-2015 (A) and 2016-2017 (B) . All patients and their respective UD were typed for DPB1 alleles. In DBP1 allele mismatched donor/ recipient pairs T cell epitope matching was performed with DPB TCE Tool (IPD-IMGT/HLA) retrospectively in first and prospectively in second phase. Among all patients analyzed 55% had 10/10 matched UD, 62% of them had more than one UD. In phase one only 10% of donor/recipient pairs were allele matched for DPB1. Epitope matching for DPB1 allele mismatched donor/recipient pairs revealed 56% predicted permissive immunogenicity. In 75% of donors with GvHD risk DPB1 epitopes the matching could be improved by choosing another donor with permissive mismatch. Our results in phase one suggested that in 30% of 10/10 matched HSC recipients GvHD risk could be decreased with improved search algorithm extended to HLA-DPB1 T cell epitope matching. This finding was confirmed in the second phase where DPB1 epitope matching was implemented. As predicted by validation of algorithm, we were not able to find a DPB1 permissible donor for only 12% of patients.

HLA PHASED HAPLOTYPES IN SWITZERLAND, PREDICTIVE VALUE FOR MATCHED UNRELATED DONOR SEARCHES AND CLINICAL OUTCOME AFTER HSCT Stéphane Buhler 1 , Helen Baldomero 2 , Sylvie Ferrari-Lacraz 1 , José M. Nunes 3 , Alicia Sanchez-Mazas 3 , Stravoula Massouridi-Levrat 1 , Dominik Heim 2 , Joerg Halter 2 , Gayathri Nair 4 , Yves Chalandon 1 , Urs Schanz 4 , Tayfun Guengoer 5 , Grazia Nicoloso 6 , Jean-Marie Tiercy 1 , Jakob Passweg 2 , Jean Villard 1 A crucial criterion for successful allogeneic hematopoietic stem cell transplantation is HLA matching between the recipient and donor. In practice HLA matching is performed at the phenotypic level but recent studies have proposed that further matching at the haplotypic level could be beneficial (e.g. reduced graft versus host disease). The aims of this study were two-fold. First, we used the high resolution typings of 291 patients living in Switzerland and potential candidates for an unrelated donor search to determine their HLA haplotypes by descent using family members. The sum of ranks of the two haplotypes for each patient (as compared to the frequencies estimated for 6114 volunteer donors of the Swiss registry) was used as a surrogate predictor of a successful search (i.e. finding a 10/10 MUD) in the BMDW database. Second, the putative impact of phased HLA haplotypes (i.e. considered as frequent or rare as an indirect parameter of haplotype matching) on the clinical outcome of HSCT was analyzed in a cohort of 212 patients transplanted with 10/10 MUD. These 212 patients were mainly affected by malignancy diseases, a large majority being represented by acute leukemia, myelodysplastic and myeloproliferative syndromes. Logistic regression analysis showed a highly significant effect of the haplotype ranks in the outcome of a search (p=2.25e -14 ). However, we could not find any significant effect on overall survival, acute or chronic graft versus host disease and relapse following HSCT despite considering different frequency cut-offs and genotype subgroups for categorizing HLA haplotypes. This study provides useful data for the optimization of unrelated bone marrow donor searches in Switzerland and beyond but could not confirm previous findings that matching at the haplotypic level has a clinical impact following HSCT. Due to the extreme polymorphism of HLA genes further studies are warranted to better comprehend the many factors at play.

Alexandre Walencik 1 , Estelle Geffard 2 , Sophie Limou 3 , Anne Cesbron 1 , Nicolas Vince 2 , Pierre-Antoine Gourraud 2 1 EFS Centre-Pays de la Loire, Nantes, France, 2 INSERM, Université de Nantes, Nantes, France, 3 INSERM, Ecole Centrale de Nantes, Nantes, France Correspondence: alexandre.walencik@efs.sante.fr EasyMatch-R is a component of the Easy-HLA web applications suite. It calculates the probability of finding an HLA matched HSCT donor and recommends a parsimonious HLA typing strategy required before requesting blood sample. A major challenge for clinical laboratories is to trust and integrate such statistical decision support programs. We present a comparative analysis of searches and strategies performed on a 202 consecutive patient cohort with HSCT indication. We first compared the number of expected donors obtained with EasyMatch-R with the number obtained from the BMDW book interrogation. We found a good correlation between EasyMatch-R and the BMDW book (r=0.99, p<0.001 for all searches, but a weaker correlation for searches with less than 15 expected donors: r=0.39, p<0.001). When categorizing the BMDW book results in five categories: very favorable, favorable, unlikely (existence of potential donors but only with low resolution typing), unfavorable, no potential donor. We found the correlated number of expected donors with respective medians of 920, 95, 0, 4 and 0. We also retrospectively compared the impact of the algorithm recommendation and the number of additional typings requested by the lab at the time of the donor search. The recommendation of additional typings is very dependent of the threshold fixed by the user: EasyMatch-R recommends 614 different typings with no threshold, 184 with 5% threshold and 134 with a 10 % level of risk considering an input at 2D. Compared to the standard strategy confirmatory typing requests, we demonstrated that EasyMatch-R is cost-effective when considering threshold. EasyMatch-R facilitates the search of a donor by providing a statistically quantified argument supporting early adoption of alternative options when the BMDW book is not favorable. It improves the effectiveness and diminishes the cost related to additional HLA typings. Pre-transplant donor specific antibodies (DSA) have been associated with poor outcome after haploidentical hematopoietic stem cell transplantation (SCT). We report a case of appearance of specific Abs against recipient (RSA) after SCT. A 40 year old woman was diagnosed with AML on 05/2016. After two induction course failures she received 5AZA with no significant improvement, requiring extensive tranfusional support. As she has no matched sibling nor suitable UD she received peripheral blood stem cell (PBSC) from her haploidentical son on 12/2016 after sequential conditioning regimen and high-dose post-SCT cyclophosphamide. Allele mismatches between recipient/donor were B*35:08, C*07:02, DRB1*11:25, DQB1*03:01, DPB1*04:02 and donor/recipient A*32:01, B*53:01, C*04:01, DRB1*13:02, DQB1*06:04, DPB1*04:01. Screening for HLA Abs by SAB before SCT revealed presence of Abs non DSA (Abs anti-DR4, DR7, DR9, DR53 epitopes 96Y2, 4Q). Severe complications (cholecystitis, pulmonary aspergillosis) occurred during aplasia. Neutrophil recovery was sustained from D+15 but she remained thrombopenic, requiring transfusions: a CD34-selected PBSC boost on 07/2017 was only briefly effective. No GvHD was observed and immunosuppression (IS) was promptly tapered (stopped at D+90). Relapse occurred on 08/2017 resulting in death one month later. Five months after graft (before boost), the profile of the Abs changed with apparition of specificities anti-HLA DQ2, DQ4, DQ7, DQ8, DQ9 (epitope 84QL3). The RSA anti-HLA-DQ7 has high MFI value > 21000. Retrospective assessment of this this temporal profile could not assume that RSA appeared after stopping IS. We observed no clinical (e.g. any form of GvHD) nor biological significant change (e.g. blood counts, chimerism or relapse) after highlighting RSA. We retrospectively analyzed all 16 haploidentical SCT performed in 2016-17 in our center and did not find HLA-Abs in the post SCT period. This illustrates need for cautious prospective follow up of HLA-Abs (in particular RSA) in recipients post haploidentical SCT. Therefore clinical/biological significance of this apparently rare occurrence may be unmasked, potentially leading to medical intervention as previously described for pre SCT allo-immunization involving DSA. Bambino Gesù Children's Hospital, Rome, Italy Correspondence: paola.giustiniani@opbg.net Pediatric patients candidate to hematopoietic stem cell transplant (HSCT) may present circulating antibodies directed against HLA molecules, due to previous immunizations (e.g. blood transfusion). Several studies showed that, in haploidentical HSCT, the presence of anti-HLA antibodies against antigens encoded by the non-shared donor haplotype, is associated with a higher percentage of rejection in the recipients. This observation suggests performing routinely this test. The rejection and failure incidence may be related to the median fluorescence intensity (MFI) of the antibody and to the pattern type. In our study, carried out from May to December 2017, IgG anti-HLA antibody identification was performed in 90 paediatric patient candidates to HSCT, 38 females and 52 males, median age nine (1-18 years old), as a first determination within pre-transplant screening. The study included 50 patients affected by hematological malignancies: acute lymphoblastic leukemia (28), acute myeloid leukemia (19), Fanconi anemia (2), hemophagocytic lymphohistiocytosis (1). The other 40 patients were affected by non-malignant hematological diseases: thalassemia (24), aplastic anemia (6), immune deficiency (7), chronic granulomatous disease (1), Blackfan Diamond Anemia (2). The anti-HLA antibody identification was performed by Luminex Single Antigen technique for HLA classes I and II IgG antibodies, considering the MFI cut off value of 1000. In patients affected by non-malignant hematological diseases we found an increased incidence of anti-HLA IgG antibodies (72.5%), especially in patients with thalassemia, compared to patients with hematological malignancies (40%). Our study confirms the importance of routine anti-HLA antibody identification, particularly on patients who are candidate to haploidentical HSCT, in order to start desensitization procedures on time.

Gabriela Berardi 1 , Cintia Llull 1 , Nicolas Fernandez Escobar 2 , Gregorio Jaimovich 2 , Maria Beatriz Rodriguez 1 , Pablo Morikone 1 , Karin Padros 1 , Ulises Toscanini 1 1 PRICAI, Buenos Aires, Argentina, 2 CEHT, Buenos Aires, Argentina Correspondence: utoscanini@pricai.com.ar Engraftment monitoring after bone marrow transplantation is usually performed by genotyping polymorphic markers on DNA extracts obtained from whole white cell populations of post-transplant samples. However, in specific circumstances the analysis of different cell subsets may be the only approach to detect chimerism. We set up a method for manual separation of three white cell subsets (CD3+, CD15 + and CD19+) and the subsequent STR analyses on these cell populations. Briefly, a density gradient separation with Ficoll-Hypaque was performed starting from 10-20ml of ACD anti-coagulated whole blood. Then, CD19+ and CD3 + subsets were sequentially selected from the mononuclear layer harvested above the Ficoll-Hypaque, and CD15+ cells were additionally selected from the granulocyte layer below the Ficoll-Hypaque. Clonal selection was performed using monoclonal antibodies conjugated to magnetic beads with the EasySepTM HLA Chimerism Selection kits. After separation, the purity of each preparation was determined by flow cytometry. Following, DNA was extracted from each subset and a 22-STR panel was genotyped for each DNA extract. The analyses were first carried out on a mixture of blood from two healthy volunteers and then post-transplant samples from patients with several clinical conditions were studied. In every case, the status of the whole white cell population was also determined in order to compare them with the subsets results. The average purity of the subpopulations assessed by flow cytometry was above 90%. The STR analyses in the cell subsets showed different correlation with the results of the whole leukocyte analyses in patient samples, indicating that STR analyses of CD3+, CD15+ and CD19+ cells separated from post-transplant blood samples by the method described can provide additional information to the analyses of the whole leukocyte population itself in some specific cases. The aim of our prospective study was to assess the influence of donor KIR genes and genotypes on outcome in AML patients treated with allo-HSCT from HLA-identical sibling in first complete remission (CR1). 32 AML patients in CR1 were included in the study. All patients received T-cell replete bone marrow grafts from HLA-identical siblings. KIR genotyping was performed by PCR-SSP kit (Olerup, Sweden). Overall survival (OS) and event-free survival (EFS) were calculated by Kaplan-Meier method, and compared with log-rank test. OS was defined as survival after allo-HSCT without death from any cause, EFS was defined as survival after allo-HSCT in complete remission without death from any cause. A median followup was 28 months (range: 3-77). 5-year estimated OS after allo-HSCT was 78%, EFS 54%. 5-year EFS in patients with KIR B/x donor was significantly higher, than in patients with KIR A/A donor (74% vs 28%, p<0.05). Centromeric and telomeric KIR B motifs both improved survival. The survival was also higher in HLA-C1/Cx recipients of KIR2DS1-positive allografts. Our data suggest that KIR genes content influences on outcome of allo-HSCT from HLA-identical related donor. KIR B haplotypes, who express more activating KIR, are associated with higher survivals, probably due to graft-versus-leukemia effect. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is one of the curative approaches for hematological malignancies. The best donor for allo-HSCT is a human leukocyte antigen (HLA)-matched sibling donor. In situations when such a donor is not found, the alternative is an HLAhaploidentical donor who shares at least one HLA haplotype with the recipient. Biological parents and their children are potential haploidentical donors. The first successful haploidentical allo-HSCT (haplo allo-HSCT) was provided in the Institute of Hematology and Blood Transfusion in 2014 and the number of realized haplo allo-HSCT increases every year. To this day 56 haplo allo-HSCT have been performed in total, mostly for patients suffering from acute myeloid leukemia (AML) (59 % of all diagnosis) as in other types of allo-HSCT (non-haploidentical). Differences are in conditioning, with haplo allo-HSCT patients undergoing more often non-myeloablative, and other allo-HSCT patients undergoing myeloablative conditioning. Haplo allo-HSCT was the second transplantation for 11 patients. Chimerism status is an important factor for survival of patients after haplo allo-HSCT. Patients with complete chimerism survive in 76 % of cases, on the other hand patients with mixed chimerism survive in only in 45 % of cases. The results do not differ from non-haploidentical, related and unrelated allo-HSCT. Relapse occurred in 10 patients after haplo allo-HSCT, all of them with AML diagnosis. All these patients died due to relapse. The median of complete chimerism achievement was day 28 after haplo allo-HSCT. 13 patients never achieved complete chimerism and all of them died. Mixed chimerism is thus a high-risk factor in post-transplantation course. The techniques now used in haploidentical transplantation (using of cyclophosphamide) have had a positive impact on several outcome measures, namely, overall survival, disease-free progression, and survival free from acute or chronic graft versus host disease. The main objective of this prospective study is to evaluate the clinical significance of anti-HLA antibodies on early complications and outcome of unrelated donor hematopoietic stem cell transplantation (HSCT). From June 2014 to September 2016, 90 adult patients, who reached 6-month follow-up after their first unrelated donor HSCT or died within this period, were recruited into the prospective study. All donor-patient pairs were 10/10 or 9/10 matched at HLA-A, -B, -C, -DRB1, and -DQB1 loci. Additionally, HLA-DRB3,4,5 and -DPB1 loci were typed. Patient sera sample collection was scheduled before the start of conditioning treatment and at 1, 2, 3, and 6 months after HSCT. Anti-HLA antibodies were screened and specificities were identified by solid-phase immunoassays (Immucor) on the Luminex platform. HLAMatchmaker programs were used to perform structural HLA class I and class II matching at the eplet level. Patients median age was 56 years (range 19 -74). Of 90 patients, 47 (52.2 %) were men. Seventy-four donor-patient pairs (82.2 %) were 10/10 matched. All patients achieved neutrophil engraftment after HSCT and 4 patients (4.4 %) had secondary graft failure. The overall prevalence of preformed anti-HLA antibodies (no donorspecific antibodies were found) was 41.1 % and 32 of 37 patients retained anti-HLA antibodies after HSCT. The presence of preformed anti-HLA antibodies was found to be associated with the development of early grade III and IV acute graft-versus-host disease (aGVHD) post-transplant complications (p = 0.0446), higher transplant-related mortality (p = 0.011), and shorter overall survival (p = 0.021). The preformed anti-HLA antibodies did not affect engraftment, incidence of relapse, graft failure, event free survival or development of non-aGVHD. Anti-HLA antibodies detected after HSCT had no impact on development of early complications and outcome of HSCT. Nine patients developed de novo anti-HLA antibodies after HSCT and in the only case they were donor-specific and directed against mismatched DQA1/DQB1. Preformed anti-HLA antibodies have impact on development of severe aGVHD, higher transplant related mortality and reduced overall survival after unrelated donor HSCT. Therefore, routine testing for anti-HLA antibodies may be considered in to pre-transplant work-up of recipients. One of the many factors influencing hematopoietic stem cell transplantation (HSCT) is the presence of donor-derived alloreactive NK cells that are formed mainly as the result of KIR-HLA ligand mismatching. Three models for predicting alloreactive NK cell formation exist: the ligand-ligand model, KIR receptor-ligand model and KIR haplotype model. The purpose of this study was to analyze the association of patient and donor KIR ligands, KIR genes and genotypes on HSCT outcome and to see which of the three models the best algorithm to predict HSCT outcomes is. The investigated group consisted of 56 patient-donor pairs that had undergone unrelated HSCT in University Hospital Centre Zagreb. Analysis according to the ligandligand model in the studied group resulted in only four mismatches as opposed to the KIR-ligand model where 31 patient-donor pairs were KIR-ligand mismatched, leading to the conclusion that determining KIR ligand mismatches only following patient-donor HLA typing, without KIR genotyping of donors, is not good prognostic algorithm. The inhibitory KIR-HLA ligand matches/mismatches determined with the KIR-ligand model did not show any beneficial effect of KIR ligand mismatches on any investigated HSCT endpoint. Assigning the AA or Bx KIR genotype to each patient and donor (patient/donor combinations: AA/AA, AA/Bx, Bx/AA and Bx/Bx) enabled its analysis according to the third model and, although no statistically significant difference was found, a trend towards better survival was observed in AA patients receiving a Bx graft. The conclusion is that any of existing models to predict alloreactive NK cell formation and their effect on HSCT outcomes can't give the answer alone and to clarify the KIR influence on HSCT outcome, different approaches in the analyses which encompasses all properties of KIR genes/receptors, KIR-ligands and NK cells together needs to be taken into consideration.

Tsvetelin Lukanov 1 , Lyudmila Quin 1 , Antoaneta Nedyalkova 1 , Bushra Al Hadra 2 , Milena Ivanova 2 , Elissaveta Naumova 2 1 University Hospital Alexandrovska, Sofia, Bulgaria, 2 Medical University Sofia, University Hospital Alexandrovska, Sofia, Bulgaria Correspondence: ts_lukanov@yahoo.com

The Bulgarian bone marrow donor registry (BBMDR) has been a member of WMDA since 2005. Different HLA typing techniques have been used throughout the years: CDC, SSP, SSOP, SBT, so the HLA profiles of our donors is diversified. Our current policy for HLA typing of donors, states that young men and women should be typed for HLA-A, -B, -C, -DRB1 and -DQB1 loci at high resolution. All other donors should be typed for HLA-A, -B, and -DRB1 loci at low resolution. Recently, next-generation sequencing (NGS) for genotyping of young donors was implemented. One third (30.9%) of our donors have been typed at high resolution. Almost half of them (48%) have been typed at allele level within the past year since the introduction of NGS. HLA typing by next-generation sequencing generates phase-resolved unambiguous result, as opposed to the classic Sanger sequencing which helped us to precisely obtain the allele frequencies in BBMDR donors. The NGS technology allowed us to shorten the period for HLA typing of young volunteer donors, and significantly increased our throughput capabilities. Several rare alleles, as well as three new alleles were already found due to the better coverage of the genes. Moreover, NGS kits include lowexpression loci, such as HLA-DQA1, DPB1 and DRB3/4/5, which additionally increase the quality of the HLA data included in the BMDW and may be used as additional selection criteria allowing further clarification of the relevance of these loci in HSCT. In conclusion, NGS technology allows for high-quality HLA data and increased throughput of the registries, at a reduced cost per sample compared to Sanger sequencing. High resolution data will shorten donor search procedures, provide more accurate matching to improve patient survival, and substantially reduce the problem with donor non-identification. Additionally, the better characterization of HLA genetic diversity within patient populations will help to predict the probability of finding matched unrelated donor and timely change the HSCT strategy. CMV matching was achieved for 205 (91%) transplants and 198 (88%) were undertaken using a blood group matched or compatible donor. We looked at the time taken to identify the final donor for transplant, i.e. the time between initiation of search and receipt of verification typing sample from the donor selected for transplant. In 2012, the time taken had a median of 31 days (range 8-86) days. This figure had improved to 14 days (range 4-60 days) during 2016, the last year of the audit period. During the final two years of the audit (2015 and 2016), the search process involved identifying an alternative donor for transplant in 21/104 (20%) cases due to the final donor selected for transplant subsequently not being available for donation. This occurred due to a variety of reasons including donor withdrawal, failing medical or deferral due to medical reasons. These audit data reflect the strategy of utilizing UK donors where possible and improvements in HLA data on the UK registry. Loss of heterozygosity in the HLA region may provide tumor cells the mechanism to escape immune surveillance. Partial or complete deletion of the HLA locus on chromosome 6 has been shown for solid tumors and for hematological malignancies, including leukemias. This phenomenon poses a challenge for histocompatibility testing since stem cell transplant from an HLA homozygous donor to a mistyped patient may lead to fatal graft versus host disease. Moreover, patients are often drawn for transplant workup at the time of diagnosis, when they may have significant numbers of malignant cells in their circulation. The case of a 57 year old female AML patient is presented who was referred to bone marrow transplant. HLA typing from peripheral blood using a high resolution SSO method showed homozygous typing for all loci, with the exception of C locus. Next generation sequencing (NGS) typing however revealed the presence of previously undetected HLA B*40:01:02 allele present at an allelic imbalance of 14%. To investigate the sensitivity of HLA typing methods for detecting underrepresented haplotypes, homozygous samples were mixed at ratios of 5, 10, 15, 20, 30 and 50%, and the samples were tested by NGS using the Holotype kit (Omixon), and by high resolution SSO method using LAB-Type XR (One Lambda). Notably, the NGS method was able to detect and quantify the presence of underrepresented HLA alleles at levels as low as 10%, showing higher sensitivity than the SSO method. This study highlights the importance of confirmatory typing for patients with hematological malignancies where false homozygosity for HLA may increase risk for GVHD. Homozygous HLA results should be confirmed by a second HLA typing on remission sample or buccal sample. HLA typing by NGS is a sensitive technique for the identification of HLA alleles underrepresented due to malignancy induced loss of heterozygosity affecting the 6p region. Relapse of malignant disease remains one of the most important complications after allogeneic stem cell transplantation (alloHSCT). It is poorly controlled by the currently used methods and substantially contributes to posttransplant mortality. The success of transplantation depends on the development of a graft versus leukemia (GvL) reaction which in the case of HLA-matched transplantation depends on the response to mismatches in minor histocompatibility antigens (MiHAs). To prevent the relapse and facilitate GvL alloHSCT could be supplemented with the adoptive transfer of donor-derived CD8+ T cells, modified with MiHA specific T cell receptors (TCRs). Clinically relevant MiHAs should be presented on a frequent HLA allele, have optimal allelic frequency and their expression should ideally be limited to hematopoietic tissue. ACC-1Y is presented in a frequent HLA-A*24 allele and has optimal allelic distribution (around 27% of the European population has an immunogenic allele, therefore immunogenic mismatch occurs in 25% of HLA-A*24 matched transplantation cases). Also, this antigen is derived from gene BCL-2, specifically expressed in hematopoietic tissue and overexpressed in many leukemias. In our project, we aim to develop a method of adoptive T cell therapy targeted to ACC-1Y. From HLA-A*24+ donor, we obtained naive CD8+ T cells and performed antigen-specific expansion using peptide-loaded dendritic cells, isolating ACC-1Y specific cells by tetramers. Cytotoxicity was confirmed by specific lysis of HLA-A*24+ LCL cells, pulsed with exogenous peptide. From this expansion, we obtained a set of ACC-1Y specific TCRs, sequenced their alpha and beta chains and cloned one TCR into a lentiviral vector. We harvested lentiviral supernatant and performed lentiviral transduction of CD8+T cells. Next, we plan to confirm the cytotoxic activity of transduced cultures on HLA-A*24+ LCL cells presenting either endogenous or exogenous peptide. To further assess the clinical implication, cytotoxic activity should be confirmed on patient-derived leukemia samples. This method of immunotherapy is in the early stage of development and further research regarding its effectiveness is needed. Blood transfusion is currently a therapeutic support used in many pathologies (hemorrhagic contexts andcancers etc.). Under the combined effect of population growth and aging as well as advances in medicine, the need for red blood cells has increased by 29% in 10 years. The advent of generating red blood cells in vitro from stem cells could in the long run supplement supply needs or even replace them. The erythroid culture is accessible through the use of a cocktail of several cytokines, but it must be optimized. Many factors intervene in the processes of proliferation and differentiation. It has recently been demonstrated that the Human Leukocyte Antigen G molecule (HLA-G) may have an impact on these proliferative mechanisms. The expression or not of this molecule, proliferation and differentiation of CD34+ erythroid progenitors within the culture media could therefore have the potential to increase erythropoiesis and production yields. Three human CD34+ cultures from cord blood pools were induced to erythroid differentiation in vitro by the addition of erythropoietin from the 6th day (D6). The phenotypic expression of HLA-G was determined from D6 to D15 by flow cytometry, confocal microscopy and Western Blot. Its transcriptional expression was analyzed by Reverse Transcription-Quantitative Polymerase Chain Reaction (RT-qPCR).

More than 90% cells express erythroid markers at D13 with over 50% mature erythroblasts at D15. The expression of HLA-G was only demonstrated on few non-defined cells by confocal microscopy between D6 and D10. This expression has not been confirmed with the other techniques, including RT-qPCR. These data suggest that the HLA-G molecule is not expressed on either immature or mature erythroid cells in our culture model. So, as this molecule is of interest in the proliferative and differentiation phenomena of certain cell types, its expression and its impact must be more objectified in the erythroid line, with the use of more sensitive techniques and through more efficient red cell culture models. Around 14% and 10% of adults suffer from chronic kidney disease (CKD) or chronic respiratory disease (CRD), respectively. Among those, most patients with advanced illnesses evolve to end-stage disease and become candidates for kidney or lung transplantations. We developed a precision medicine application for kidney and lung transplantation patients in order to monitor individual trajectories of posttransplant outcomes using data from the DIVAT (Donnees Informatisees et VAlidees en Transplantation) and COLT (Cohort in Lung Transplantation) studies. Together, information from more than 3,000 patients with renal or lung transplantation, including clinical and immunological items, were collected since 2008. We present here two personalized contextualization algorithms: 1) a populational contextualization where we compare data trajectory of a given patient to a sub-population with similar characteristics selecting by filters or nearest neighbor approaches, and 2) a referential contextualization where we compare data trajectory of a given patient to extreme groups (defined by clinicians) such as acute graft rejection, humoral rejection, cellular rejection or tolerance. The comparative properties of these algorithms are individually determined at reference group level. An example is the normality assessment of the 50-, 100-, 500-, 1000-, or 1500-nearest neighbors for a given kidney transplantation patient with a 150 mol/L creatinine level at 1-year post-transplantation; we find him at the 65e, 57e, 67e, 67e, or 66e percentiles, respectively. This precision medicine application therefore facilitates access to large amount of data and allows their visualization and comparison in order to optimize medical care and guide clinical decision. Finally, we aim to extend our algorithms to various chronic medical conditions and settings to improve patients care.

APPLICATION OF EASY-HLA TO 2,260 SOLID ORGAN TRANSPLANT DONOR-RECIPIENT PAIRS FROM 2 COHORTS: STATISTICALLY UPGRADED HLA TYPING FOR RESEARCH USE Estelle Geffard 1 , Alexandre Walencik 2 , Sophie Limou 1 , Pauline Scherdel 1 , Adrien Tissot 1 , Divat investigators 1 , Colt investigators 1 , Gilles Blancho 1 , Magali Giral 1 , Antoine Magnan 1 , Anne Cesbron 2 , Nicolas Vince 1 , Pierre-Antoine Gourraud 1 1 INSERM, Université de Nantes, Nantes, France, 2 EFS Centre -Pays de la Loire, Nantes, France Correspondence: nicolas.vince@univ-nantes.fr HLA polymorphisms have a central role between clinic and research. However, large size datasets are usually difficult to analyze as HLA is typed with a wide array of techniques for partially overlapping loci at different resolutions. The Easy-HLA software suite delivers statistically updated HLA information from all HLA typing resolution. Easy-HLA is a userfriendly web application available at: http://hla.univ-nantes.fr. We present its uses of published haplotype frequencies to solve typing ambiguities, predict HLA-typing of loci (HLA-Upgrade) as well as haplotypes pairs (HLA-2-haplo). We applied HLA-Upgrade and HLA-2-Haplo to two solid organ transplantation cohorts including serological HLA-A~B~DRB1~DQB1 genotypes of donor-recipient pairs in kidney transplantations (n=1,501, DIVAT cohort) and in lung transplantations (n=759, COLT cohort). In COLT, HLA-Upgrade resolved genotypes to high-resolution in 63% of the subjects with confidence >0.5. It predicted un-typed HLA-C locus in 42% of the subjects with confidence >0.5. When using high resolution typing, HLA-2-Haplo imputed haplotype pairs with high confidence. The average value of prediction probability is 83% [82-84] ranging from 28% to 100%. Easy-HLA tools facilitate analysis of immunogenetic parameters in graft survival for both kidney and lung transplanted patients. In addition, it provides new functional immunogenomics parameters to analyze: HLA-expr imputes HLA-C expression, HLA-AA provides Amino Acid equivalence of HLA allele and HLA-KIRlig gives KIR Ligand classification of HLA allele. Easy-HLA facilitates large-scale analysis and promotes the multi-faceted HLA expertise beyond allelic associations. The HLA region is arguably the most important part of the human genome, but due to its peculiarities and complexities, most researchers are hesitant to work with this region. Recent developments in genome biology equally apply to the HLA region, but the tools may look daunting to nonspecialists. To facilitate research in genome biology, especially by HLA scientists, we have developed simple online tools. The first tool SNPsop facilitates examination of the functionality of any SNP in the genome. The user simply enters the SNP ID in the search box and selects from a number of tools listed that provide information about its functions (including eQTL status) and publications citing the SNP. The app then opens each of those web tools in separate tabs in the browser to display information specifically for the SNP queried. New tools will be added to the search menu as relevant ones are published. The second tool HLASNPs is an online SQL database (also available as a light-weight Microsoft Access database for offline use) which allows the user to query about any of 470,000 SNPs within the extended HLA region to obtain detailed information. This integrated database brings together information that is either extracted from other relevant genome-wide databases (including genomic location, eQTL/meQTL status, functionality scores, location within CpG islands/transcription factor or microRNA binding sites, chromatin features, chromatin interactions, GWAS/PheWAS disease associations) or from the literature (proxy status for HLA alleles) and from our own Immunochip data on IHWG cell lines (proxy status for HLA supertypes, ancestral lineages, and genotypes in each of the examined cell line). This manually curated database will also have constantly updated information from the current literature related to genomic features of the HLA region and disease associations. The tools are in trial and final verification phase of their development and will be fully functional and online by May 2018. We hope that these tools will introduce the life scientists interested in HLA region research to the latest advances in genome biology. The widespread use of next generation sequencing for HLA genotyping has led to a rapid increase in the discovery of novel alleles. Unfortunately, many of the allele sequences submitted to the IPD-IMGT/HLA and IPD-KIR databases comprise only coding sequence or, for HLA, only the exons encoding the antigen recognition domain. This impedes the establishment of long-read-based genotyping technologies. It is desirable that novel alleles are characterized and submitted in full-length, and that known alleles are extended to cover the complete gene sequence. The manual annotation and submission of full-length sequences to the IPD-IMGT/ HLA and ENA databases is a time-consuming and errorprone task. In 2016, we developed and published TypeLoader, a tool that takes the full-length sequence of a novel HLA allele in FASTA format as input, automatically annotates it, and creates all files necessary for submission. This reduced the manual effort for submission by over 95%. It was implemented as a web application to run on a Linux server and has been used at the DKMS Life Science Lab to successfully submit more than 1000 novel HLA alleles. To enable a more widespread use of the tool by other labs, we have adapted TypeLoader to make it available as a Windows standalone application that can easily be used on standard PCs without dependency on other software. TypeLoader now automatically detects null alleles generated by premature stop codons or frameshift mutations and features enhanced integration with ENAs automated submission API. TypeLoader also implements experimental support for annotation and submission of KIR alleles. We hope that the increased convenience and scope of TypeLoader will foster the submission of more full-length sequences to the IPD-IMGT/HLA and IPD-KIR databases, ultimately promoting the widespread use of full-length sequencing for genotyping of both HLA and KIR. HLA imputation from SNPs is a powerful statistical alternative from expensive NGS HLA typing. While extensive studies using SNP genotypes grew extremely fast, much work is still needed to increase capacities of studying HLA alleles and their association with different diseases. To fill this gap between genomic data availability and HLA alleles studies, we propose to facilitate the HLA imputation from SNP studies by developing the SNP-HLA reference consortium. The SNP-HLA reference consortium main objective is to bring together immunogeneticists and other scientists in order to (1) build several large reference panels for HLA imputation from SNPs genotypes, and (2) share these reference panels within a publicly available database. Indeed, imputation accuracy is dependent of the reference panel quality, as well as to the matching of the queried data (e.g. ancestry background or genomic array coverage). We already have access to large cohorts of over 4,000 ethnically diverse individuals with HLA types and SNPs genotypes. We offer to provide to the consortium our own reference panels and the logistics, including an access to a highcomputing infrastructure that is necessary for the reference panel building. As an example, only for HLA-A, it took 2,700 CPU-hrs for 917 individuals with 49 alleles and 5,780 SNPs to build a reference panel using HIBAG. As the reference panel bagging with the HIBAG R package uses bootstrap sampling, all the dataset will be anonymized and can therefore be released in a public database. To develop this very ambitious tool, we encourage willing participants with large HLA types + SNPs data to join the project and contribute to the development of a consortium that will empower the immunogenetic community to shift in the immunogenomic association era. Until recently, sequencing exons encoding Antigen Recognition Sites (ARS) and ruling out Common and Well Defined (CWD) null alleles was considered the gold standard for high resolution typing of HLA genes. This was mainly due to limitations in the older generation sequencing technologies. It is well known that variations in any site of a gene could modulate or knock out the expression of the HLA proteins. Advances in the 3rd generation sequencing technologies such as PacBio SMRT sequencing have made it possible to perform high volume, high throughput HLA class I whole gene typing. In our laboratory, we have been performing whole gene sequencing for class I genes as one of the routine typing methods since July 2016. We have typed 460,000 samples using whole gene class I sequencing on PacBio SMRT platform on PCR amplicons spanning from 5´to 3´Untranslated Regions (UTRs). 355 samples had novel variations in ARS exons and over 3251 samples had novel variations in non-ARS regions. Over 50,000 samples had a novel intronic sequence. We have found that 61 samples had variations in non-ARS exons that could result in premature stop codon or nonsense proteins. Three samples had variation at the intron boundary of the non-ARS exon that predicted a premature stop codon. 89 samples had variations in the spliceosome sites of the introns that could change the length of the translated protein and cause a premature stop codon or nonsense protein. Whole gene sequencing also enables us to identify known and new markers for haplotypes and population specific variation in coding and non-coding sequences. Those markers could be important predictors of the level of histocompatibility of the patient and donor cells. Our results indicate that there are enough convincing data to change our definition of Gold Standard of HLA class I typing to whole gene sequencing, 5´to 3´, including all exons and introns. Conventional high resolution HLA typing technologies produce ambiguous results, and it is often necessary to perform additional tests to resolve these ambiguities. Next generation sequencing (NGS) is a promising technology, which can overcome this problem, determining HLA compatibility between donor and recipient by phase-resolved HLA typing results. Here we report our preliminary experience in introducing NGS technology using AllType kits NGS on Ion Chef and S5 platform, followed by TypeStream software analysis (Thermo Fisher -One Lambda, Canoga Park, CA, USA) for HLA-A, -B, -C, -DRB1, -DRB3,4,5, -DQA1, -DQB1, -DPA1 and -DPB1 loci. All 11 loci were amplified in a single PCR reaction, avoiding the need for amplicon pooling. Fifty-three samples, for a total of 801 alleles, were included in this study, which were previously typed at highresolution using a combination of Sanger sequencing, sequence-specific primer (SSP) or sequence-specific oligonucleotide probe (SSOP) technologies and recorded at the two-field level. The mean read length was 243 bp, while average read depth for the key exons was 694, with the maximum of 1128 and the minimum of 327. An overall concordance of 99.8% with reference typing was obtained, with only one HLA-C allele not assigned by the software. All the alleles were typed at the minimum level of two fields; a three-field level was reached by all class I alleles and in 97.4%, 98.6% and 72.2% of DRB1, DQB1 and DPB1, respectively. Finally, allelic resolution was achieved in 100% HLA-A and -C, 85.7% of HLA-B, 96.1% of DRB1, 87.1% of DQB1 and 52.0% of DPB1 alleles. In conclusion NGS-based HLA typing has much potential to replace routine clinical SBT typing with its power to provide allelic level typing for the 11 clinically relevant HLA loci within 2-3 working days, with an excellent sequence coverage, streamlined workflow, and overall low cost. TypeStream analysis software correctly assigned all the alleles at high percent concordance. Among the many factors that influence the clinical success of HSCT, polymorphism of the classical HLA genes represents one of the major barriers. Next-generation sequencing (NGS) has recently become a useful technique for high resolution (HR) HLA typing. In the present study we describe the validation of NGS technology before its introduction in the routine activity in our laboratory for allelic resolution (AR) HLA typing. We analyzed 174 DNA samples using both NXType NGS or AllType kits NGS, first on an Ion Torrent PGM and subsequently on the S5 platform, followed by TypeStream software analysis (Thermo Fisher -One Lambda, Canoga Park, CA, USA) for the loci HLA-A, -B, -C, -DRB1, -DRB3,4,5, -DQA1, -DQB1, -DPA1 and -DPB1, for a total of 2385 alleles. We obtained 100% reproducibility of the method repeating the HLA typing on the same DNA samples (2-5 times) for a total of 582 tests. We subsequently compared the results previously obtained at 2-fields for all the 2385 alleles studied, with 100% concordance of the results. We further analyzed 57 out of the 174 DNA samples previously typed at HR level. We observed only two differences (0.12%) in which previous SBT analysis did not solve two cis/trans ambiguities: C*06:02, 12:02:08 vs C HLA alleles are observed in specific haplotypes, due to Linkage Disequilibrium (LD) between certain alleles. Recent improvements in Next-Generation Sequencing (NGS) technologies enable us to handle the peculiarities of the MHC, notably the tight LD between genes, as well as their high degree of polymorphism. In order to assign HLA haplotypes at allele level (4-fields) in a Greek population, in the context of 17th IHWS, 25 families including parents and siblings (95 subjects), all of Greek origin, were typed at 11 loci (HLA-A, -B, -C, -DRB1, DRB3/4/5, -DQA1, -DQB1, -DPA1 and -DPB1) by NGS technology. Instrument and chemistry: Illumina MiSeq, v2 chemistry, 300 cycles (2X150) and standard flow cell size, and Illumina TruSight HLA v2.0 and Omixon Holotype HLA kits were used. HLA haplotype count and frequency (%) were estimated according to the number of parental haplotypes identified in the whole cohort. Additionally, haplotype counts in families (n=25) and subjects (n=95) were calculated. All but one of the haplotypes were observed once (haplotype frequency 1%). HLA-A*01:01:01:01~C* 07:01:01:01~B*08:01:01:01~DRB3*01:01:02:01~HLA-DRB 1*03:01:01:01/02~DQA1*05:01:01:02~DQB1*02:01:01D PA1*01:03:01:02~DPB1*04:01:01:01/02 haplotype was found in three subject members of two families (haplotype frequency 2%). It is confirmed that the HLA-A*01-Cw*07-B*08-DRB1*03:01-DQB1*02 haplotype, defined by low-resolution molecular techniques, is the most common in Greeks, as in all Caucasians. These results are useful as reference data (controls) in disease association studies as well as in population genetics.

Alice P. Pedersen 1 , Menaka Andersen 1 , Tore Jensen 1 , Marte K. Viken 1 1 Department of Immunology, Oslo, Norway Correspondence: mavike@ous-hf.no It is well known that certain HLA alleles are associated with disease. Testing for HLA-B*27 is used as support for diagnosing ankylosing spondylitis and other HLA-B*27 associated diseases. In our laboratory the number of samples tested for HLA-B*27 per week is around 180. Methods like real time PCR are quick and can test multiple samples simultaneously. Our previous HLA-B*27 real time PCR typing was based on primers in exon 3, utilizing Sybr Green and melt curves to detect HLA-B*27 positive samples. However, the primers used, failed to detect several HLA-B*27 alleles which could lead to false negative results. Our aim was to make a triplex real-time assay with fewer false negative results and change from Sybr Green to Taqman probe detection. First, we designed a probe for the HLA-B*27 exon 3 assay. Secondly, we performed a literature search to identify other PCR-based HLA-B*27 assays. We selected primers and probe for an exon 2 assay. Our new method is a triplex PCR combining the assays for exon 2 and exon 3 with a commercial RNAseP assay. The instrument used in our laboratory is a QuantStudio 6 Flex from Life Technologies AS. By using a pipetting robot and 384well plates we now have the flexibility of running up to 381 samples per run. The triplex PCR was optimized with regards to the amount of the assays and DNA input in the reaction, and annealing/elongation temperature. We validated the method by running 275 samples typed with our previous HLA-B*27 assay, a panel of 16 samples selected for HLA-B*27 especially (B27-panel), and a set of 73 samples with various HLA-B alleles to cover the most common alleles. We do not consider the various disease associations found for the different HLA-B*27 alleles, but report results as HLA-B*27 positive or negative. Results are manually scored by registering the cycle of threshold, and thereby assigning the results as either positive or negative for the presence of HLA-B*27 allele(s). The validation gave a 100% concordance with earlier results. The B27-panel confirmed that our previous method probably has led to false negative, and false positive results. By combining HLA-B*27 assays for exon 2 and 3 we reduced the possibility of false negative and positive results but did not exclude this completely. This method is robust in our hands and gives us a flexible and cost effective method for HLA-B*27 typing.

Marine Cargou 1 , Charlene Bouthemy 1 , Mamy Ralazamahaleo 1 , Gwendaline Guidicelli 1 , Jonathan Visentin 1 1 Centre Hospitalier Universitaire de Bordeaux, Bordeaux, France Correspondence: jonathan.visentin@chu-bordeaux.fr Human Leukocyte Antigen (HLA) genotyping by Next Generation Sequencing (NGS) is now widespread. We aimed to evaluate the performances of the NXType kit from One Lambda performed with Thermo Fischer Scientific reagents (Ion 520 & 530 Ext Kit Chef ) and Ion 530 chips on a S5XL platform. The first 10 runs carried out in our laboratory allowed DNA from 289 patients, 15 external proficiency testing (EPT) and a control DNA included in each run to be typed at HLA-A, -B, -C, -DRB1/3/4/5/, -DQB1, -DPB1 loci. A median of 32 samples were loaded per chip. We observed only two drop-outs among the 2240 alleles typed with another method. Among the 662 new alleles detected by the software (TypeStream V1.1.0.11, IPD-IMGT/HLA Database release 3.25), two were confirmed by Sanger SBT because mismatches were located in key exons and 193 were considered as probably new but unconfirmed (mismatches in non-key exons or introns). The 467 others were rejected because of homopolymers (n=279), nucleotide doublet repeats (n=10), low read depth (< 200 reads, n=84), high background (>20%, n=92) or were additional DRB3 alleles (see below, n=18). The comparison of the most likely allele obtained with SSO-XR for 253 patients evidenced a 99.5% concordance for HLA-A, -B, -DRB1 and -DQB1 loci at second field resolution. For HLA-C and -DRB3/4/5 loci, 99% and 93.4% of results, respectively, were consistent with common linkage disequilibrium. Indeed, we observed 27 additional DRB3 alleles falsely detected by the software. For EPT samples, we obtained the consensual results for 100% of alleles at the first and second fields (206/206), 96% at the third field (190/197 ) and 100% at the fourth field (n=44/44) resolutions. In conclusion, NXType kits displayed satisfactory performances. An experienced user can easily manage the falsely new alleles caused by homopolymers which are frequently proposed. The DRB3 assignment should be improved in order to eliminate the supernumerary alleles, which are nevertheless easily detected by studying the linkage disequilibrium. The GenDx NGSgo workflow for HLA typing using nextgeneration sequencing (NGS) consists of four practical steps: amplification, library preparation, sequencing and data analysis. Out of these four steps the library preparation is the most labor intensive and most interspersed by incubations. To decrease hands-on time and reduce the risk of human error, NGSgo library preparation can be automated from start to finish using a liquid handler workstation. Here we describe the results of the automation of the NGSgo library preparation for Illumina on a Beckman Coulter Biomek 4000 workstation. The workflow was fully automated from amplicon pool to finalized library, making it completely hands-off to accommodate overnight runs and allow the operator to step away from the machine. The Biomek setup included a thermal cycler, 96-well plate magnet, two peltiers devices for cooling 96-well plates and 1.5ml tubes, a gripper tool, reagent reservoirs and liquid waste disposal. Pipetting was performed by two pipet tools. This setup allowed for flexible processing of four to 24 samples per run, with any number of loci. Per run only three boxes of pipetting tips were used for 24 samples. Panels with variable numbers of samples were processed with the Biomek. Subsequently, the pooled libraries were sequenced on a MiSeq system and the resulting data was analyzed with NGSengine software (GenDx).Read depth, mappability and noise levels of the sequence data was reviewed and showed that the automated workflow resulted in high quality reads and accurate typings. We conclude that automation of the NGSgo library preparation on the Biomek 4000 yields good and consistent libraries, with a significant reduction of hands-on time.

Luyanda Laura Kwofie 1 , Pieter W. Meyer 1 1 University of Pretoria, Pretoria, South Africa Correspondence: luyandakwofie@gmail.com Most countries' transplant guidelines advocate the screening of kidney transplant patients serum for donor-specific anti-HLA antibodies. These antibodies play a crucial role in graft rejection and eventually graft loss. Therefore, detection of cytotoxic antibodies is one of the most important investigations undertaken in potential organ transplant work-ups. Three procedures were compared for the detection of cytotoxic antibodies to identify which is best suited to optimize pretransplant donor-recipient matching. Serum samples from 15 renal transplant recipients awaiting transplantation were tested for presence of HLA antibodies by CDC, FCXM and Immucor Luminex based DSA crossmatch methods with using donor lymphocytes. Two (13%) of the 15 patients had positive HLA class I crossmatch for all methods tested, the two patients for CDC/FCXM are the same patients. DSA identified one patient that was CDC/FCXM positive, but the other was negative for CDC/FCXM. CDC showed a high percentage of positive HLA class II (40%) while DSA identified two (13%) patients and FCXM one (7%). When the four noncongruent CDC HLA class II-positive samples were analyzed with the lymphocyte single antigen (LSA) assay, none had any HLA class II antibodies present, thus indicating all four (67%) samples identified by CDC were false-positive. The results suggest that the CDC HLA class II method has no diagnostic application because of the high false-positivity rate. CDC HLA class I testing has limited application in cadaver donor transplant and may be beneficial in living related donor transplant when used in conjunction with another more sensitive crossmatch method. The DSA and FCXM methods showed equivalent performance. However, one patient with a positive FCXM reaction and negative DSA was identified. It was established that this patient was pre-sensitized (having HLA class II antibodies as per LSA) but did not have specific antibodies to the donor's HLA class II alleles. We therefore conclude that DSA could possibly add value in pre-sensitized kidney transplant patient screening to identify those presensitized patients that may not reject the donor graft due to the absence of donor-specific antibodies. HLA typing is always necessary for the execution of hematopoietic stem cell transplantation and scientific societies recommend the extended high-resolution typing of HLA alleles for transplants from unrelated, haploidentical and identical-related donors. Until 2016, in our laboratory, we performed high-resolution typing with the DNA sequencing Sanger method using a one-capillary instrument (ABI PRISM 310) and for this reason it was not possible to run many sequences in a short time. Our strategy was to execute the PCR-SBT of the exons 2, 3, and 4 of the HLA class I loci without the use of gSSP primers and additional exons. This led to results with many ambiguities due to the heterozygous positions and to the non-sequenced exons; therefore, in 2016, it was necessary to repeat 210/405 (52%) tests using a PCR-SSP method. We also performed highresolution class II typing using only the PCR-SSP method. This strategy demanded a lot of work on the part of the operators and a long time to obtain the final results. Since 2017 we have been using the sequencer 3500xL Dx with 24 capillaries by Life Technologies and this has allowed us to be able to run many sequences over a short time; thus we have also been able to extend the DNA sequencing to all the HLA-class II loci with SBTexcellerator Kits by GENDX and also to use the gSSP primers and additional exons to solve ambiguities. In 2017 this new strategy allowed us to obtain 367/372 (99%) high resolution results with no ambiguity, without having to resort to the use of PCR-SSP and therefore saving a lot of time and effort on the part of the operators. In our opinion, the PCR-SBT Sanger method using a 24-capillary sequencer is an excellent strategy for achieving unequivocal, high-resolution results, while in the future we would like to further improve our strategy by introducing the NGS; this is ideal for obtaining results at allelic resolution when having to test a very large number of samples.

Reinhard Kelsch 1 , Hanne Grawe 1 , Bram Luiken 2 , Daniel Mathow 2 , Sake van Wageningen 2 , Malte Heyer 3 , Benjamin Vogel 3 , Rafael Alvarez 3 , Sirak Kifle 3 1 Institute of Transfusion Medicine and Transplantation Immunology, Münster, Germany, 2 GenDx, Utrecht, Netherlands, 3 Tecan, Männedorf, Switzerland Correspondence: r.kelsch@uni-muenster.de HLA-typing by Next-Generation Sequencing (NGS) yields a higher resolution than other HLA-typing approaches. Amongst the different suppliers for NGS platforms predominantly Illumina and Ion Torrent systems are currently used for NGS-based HLA typing with the GenDx NGSgo workflow. Despite the advantages in terms of HLAtyping resolution, NGS sample processing is comparably laborious, an issue which can be circumvented using robotics. Here we show results of the NGSgo workflow for Illumina sequencing automated on the Freedom EVO NGS workstation. The DNA library preparation consists of DNA fragmentation, adapter ligation and indexing. Cleanup and pooling steps complete the workflow. The Freedom EVO NGS workstation was set up to perform the library preparation for four to 48 samples. The PCR indexing step is performed in an offline thermocycler to allow workflow optimization. Dual indices in a 96 well format are applied to facilitate automated indexing. To test the reliability of the automated workflow, we have run the same DNA sample on eight consecutive positions in one plate, resulting in 8 differently indexed libraries. These libraries were pooled and sequenced on an Illumina MiSeq system (300 cycles). The number of reads per library ranged from 0.85 to 1.17 times the average (451 reads). Read mappability ranged from 86% to 87%. Locus mappability was >90% in all cases. Further validation was performed by running independent sample panels. We conclude that the Freedom EVO NGS workstation can be used to successfully execute the NGSgo workflow for HLA typing by NGS. The automated workflow is a fast and efficient solution and can accurately HLA type gDNA samples. The application of NGS technology has provided rapid allelic HLA typing, enabling better matching and improved clinical outcomes for patients. NGS primers for full-gene HLA class I typing have been established but current commercial NGS primers for class II typing do not amplify the whole gene, resulting in some typing ambiguities. In this study, full-gene (5´-3´UTR) HLA-DPB1 amplification for NGS typing was developed using previously reported primers. Library preparation was carried out using the Illumina TruSight HLA library kit and sequenced on the MiSeq with GenDx NGSengine analysis. HLA-DPB1 NGS typing was validated with 24 International Histocompatibility and Immunogenetics Workshop proficiency testing samples and 374 stem cell donor samples previously typed by PCR-SSOP. A concordance of 99% was achieved between the two methods with the 1% (n=4) showing 3rd field discrepancies. Forty-eight samples were also amplified for HLA-DPB1 and library prepared using the One Lambda library kit with Ion Chef and sequenced on the Ion S5. Analysis of the HLA types using TypeStream showed 100% concordance with historical PCR-SSOP typing. Phase ambiguities were observed with both systems due to the long introns in the gene and the limitations of the technologies. In addition, a single PCR amplification by multiplexing the HLA-DPB1 primers with previously reported full-gene HLA-DQB1 primers was developed for NGS typing. Preliminary results typing 16 samples on the MiSeq and eight samples on the Ion S5 showed 100% concordance with previous NGS types. Furthermore HLA-DPB1*13:01:01/107:01 and HLA-DQB1*06:01:01/06:01:15 ambiguities, previously observed with primers excluding exon 1, were resolved. NGS typing for full-gene HLA-DPB1 has been developed and validated. HLA-DPB1 and HLA-DQB1 primers have also been multiplexed to produce a single amplification of full-gene sequences of both loci for NGS typing. Further work to reduce the amplifications required for HLA class I and II NGS typing will help streamline the workflow, reducing costs and improving test turnaround times. Next Generation Sequencing is increasingly considered to be the gold standard method for HLA typing. In this context, the French organization for blood transfusion Etablissement Français du Sang (EFS) developed a national research and development (R&D) program to provide a complete range of reagents and instrumentation leading to a fast, simple and robust NGS strategy. The developed tools cover all phases of the NGS process, apart from the sequencing reaction. The specially-designed NG-Mix amplifies all classical and non-classical class I and class II genes (HLA-A, -B, -C, -DRB1, -DRB3, -DRB4, -DRB5, -DQA1, -DQB1, -DPA1, -DPB1, -E, -F, -G, -DQA2 and -DQB2). Amplicon fragmentation is achieved with NG-Frag, a novel instrumentation recently developed by EFS which, in less than one minute, allows ultra-sonicationbased fragmentation of 96 samples in a unique single step. This highly standardized sonication step generates fragments in a uniform range of optimal size (average 500 bp), which eliminates the requirement to perform timeconsuming fragment size selection and purification steps. The library is subsequently generated in less than one hour by using NG-Lib, which are specific reagents that produce a normalized library without the need to perform any extra PCR reactions for tagging. Library molarity is consistently and reliably quantified using NG-Quant, a q-PCR method using a specific probe. The sequencing reactions are performed on the MiSeq system (Illumina) before analysis with NG-View, a data compilation and interpretation software designed by the EFS. This method was used to generate results on 41 homozygous reference cell lines and 190 random BMVD samples previously genotyped by another NGS commercial product (Omixon). These results will be presented along with other important aspects: the technical advantages of the EFS NGS approach at each step of the process, a description of DQA2/DQB2 allelic system, and the observed linkage disequilibrium between classical and non-classical HLA genes and description of new alleles. At the time of preparation of this abstract, 41 alleles discovered by this method are already listed on the IPD-IMGT/HLA Database.

Sandra G. Guerra 1 , Siobhan Hamilton-Jones 1 , Colin J. Brown 1 , Cristina V. Navarrete 1 , Winnie Chong 1 1 NHS Blood and Transplant, London, United Kingdom (Great Britain) Correspondence: sandra.guerra@nhsbt.nhs.uk High resolution HLA typing by NGS has reduced the need for extended typing of registry donors which has resulted in faster identification of matched unrelated stem cell donors. NHS cord blood bank (NHS-CBB) donations have been NGS typed since 2015 and are from a diverse cohort of donors in London that self-report ethnicity; this can lead to inaccuracies in ethnicity data. In this study, DNA samples from 450 NHS-CBB donors comprised of: 65% European Caucasoid, 12% mixed ancestry, 11% South Asian, 7% African and Caribbean, 2% other Asian (Central, East and South East Asia), 1% Middle Eastern, 0.6% South American and 0.5% Jewish, were NGS typed for HLA-A, -B, -C, -DRB1,-DRB345, -DQB1, DQA1, -DPB1 and -DPA1 using the One Lambda NXType (212 samples) or AllType (238 samples) kits on the Ion S5 and Ion Chef systems and analyzed using TypeStream visual and PyPop software. The analysis revealed 494 distinct alleles (k): HLA-A k= 64, -B k=114, -C k=58, -DRB1 k=72, -DRB3 k=10, -DRB4 k=8, -DRB5 k=5, -DQB1 k=41, -DQA1 k=54, -DPB1 k=48, -DPA1 k=20. The NXType and AllType kits showed resolution of non-common alleles including; A*31:14N, B*35:68:02, C*14:23, DRB1*16:09:01, DQB1*06:42, DQA1*05:10, DPB1*618:01 and DPA1*01:05. As expected, overall haplotype analysis showed the most frequent probable haplotypes to be A*01:01:01:01~B*08:01:01:01~C*07:01:01: 01~DRB1*03:01:01:01~DQB1*02:01:01 with DPB1*01: 01:01 or DPB1*04:01:01 (1.3%), of which the latter was the most frequent probable haplotype (2%) in European Caucasoids (n=293). Furthermore, the addition of DPA1 typing revealed an increased level of diversity within the overall cohort; DPB1*01:01:01 with DPA1*02:01:02 (1.8%) or DPA1*02:01:08 (4.9%) and DPB1*04:01:01 with DPA1*01:03:01:02 (11.7%) or DPA1*01:03:01:04 (6.5%). The application of 11 HLA loci resolution by NGS has revealed an increased level of diversity in the NHS-CBB. The incorporation of this data coupled with ethnicity data could lead to improved donor selection. The human killer-cell immunoglobulin-like receptors (KIR) have been reported to effect HLA-matched hematopoietic stem cell transplantation outcomes. To facilitate the selection of potential stem cell donors based on KIR genotypes we established an exon-based NGS workflow for KIR typing in 2015. Using this workflow, 2.6 million donors have been typed and thousands of novel KIR allele sequences have been discovered. Here we present a new long-range amplification approach developed for the full-length characterization and submission of novel KIR alleles. Based on data from haplotype KIR sequencing projects we designed a comprehensive long-range PCR-approach for the specific amplification of each of the 16 KIR genes (KIR2DL1-5, KIR2DP1, KIR2DS1-5, KIR3DL1-3, KIR3DP1, KIR3DS1). To amplify distinct KIR genes including the UTRs, a set of 19 forward and 22 reverse primers was used in various combinations resulting in amplicons up to 17.5 kb in length. To evaluate the specificity of this approach, we generated long sequencing reads using Pacific Biosciences Sequel SMRT technology for each primer combination. On average 87% and at least 70% of these reads mapped to the targeted KIR gene. These assays have been successfully adapted to different sample types including DNA extracted from buccal swabs which are known to be a challenging target for long-range PCR.

To generate reference quality sequences of novel KIR alleles, we applied a dual-sequencing approach by combining data from long-and short-read technologies. We successfully employed this method to amplify and sequence novel KIR alleles of more than 400 samples. Submitting these sequences to the IPD-KIR Database will contribute considerably to our knowledge of KIR diversity. HistoGenetics, Ossining, NY, United States of America Correspondence: nezih@histogenetics.com DNA sequencing of exon 2 encoding the Antigen Recognition Site (ARS) was considered the gold standard for high resolution typing of HLA class II. It is well known that variations in any site of a whole gene could modulate or knock out the expression of the HLA proteins. In our laboratory, we have been performing in-phase long range sequencing for class II genes as one of the routine typing methods since July 2016. We have typed 460,000 samples using this approach on PacBio SMRT sequencing platform on PCR amplicons spanning from intron 1 to intron 3 or intron 1 to 3´UTR of those genes. 211 samples had novel variations in class II ARS exons and over 1795 samples had novel variations in non-ARS regions. Close to 30,000 samples had novel intronic sequences.We have found that 77 samples had variations in non-ARS exons that could result in premature stop codon or nonsense protein.

119 samples had variations in the intron spliceosome sites that could change the length of the translated protein and cause premature stop codons or nonsense proteins. However, sequencing of long intronic regions of class II genes is still a challenge in certain regions, such as long homopolymer and short tandem repeat regions. This is could be PCR as well as sequencing artifacts. Until better methods developed currently we are masking these regions from analysis. In-phase long range sequencing of class II genes is an important step forward in raising the bar for high resolution or Gold Standard HLA typing. Improvements in DNA sample collection that can yield longer amplicon size will facilitate achieving the whole gene class II sequencing. Once this is accomplished, today's SMRT technology can sequence those amplicons. Recent studies indicate that in HSCT, a mismatch between donor and recipient in the Major Histocompatibility Complex (MHC) class I chain-related molecule MICA may have implications as severe as a mismatch in the classical HLA genes. Consequently, we added this marker to our donor recruitment genotyping profile and adapted our NGS algorithm neXtype to obtain genotyping results of MICA. So far neXtype supported the genes HLA-A, -B, -C, -E, -DRB1, -DQB1, -DPB1, ABO, RhD, CCR5 and KIR for highthroughput genotyping. Given indications that MICB matching might also be of relevance for donor selection, we decided to genotype both loci in a multiplexed setup. Our set-up comprises primers for individual amplicons for exons 2, 3 and a single amplicon spanning exons 4 and 5. Thereby, we are able to uniquely determine 136 out of 148 known MICA and MICB alleles. In particular MICA-129 sequence polymorphism can be resolved unambiguously. For sequencing, we employ our well-established concept involving Illumina MiSeq and HiSeq instruments. To cope with the lack of locus specificity of our primer design, we took advantage of a quantitative algorithm developed for KIR genotyping. Validation of our workflow was carried out using 118 and 94 samples with known genotypes for MICA and MICB, respectively. 115 MICA and 93 MICB genotypes were in full concordance with the pre-typing results. For the remaining three MICA and one MICB samples no result was obtained due to PCR failure. Since the start of the routine, we typed more than 380,000 potential donors for both newly added loci. For all typing results, we can identify the MICA-129 position unambiguously. Regarding first field resolution, for MICA (MICB) 95.5% (75.5%) of the typing results were unambiguous. We observed potentially new alleles in 0.4% (0.4%) of the results Thus, our MICA and MICB typing results are a valuable contribution to the standard typing profile for potential stem cell donors. Welsh Transplantation and Immunogenetics Laboratory, Pontyclun, United Kingdom (Great Britain) Correspondence: deborah.pritchard3@wales.nhs.uk UK NEQAS for H&I Scheme 4A2 assesses participants' ability to correctly HLA type samples at 2nd field resolution. However, the introduction of next generation sequencing technology has enabled laboratories to routinely HLA type samples at a higher resolution, for which no external quality assessment (EQA) scheme exists. In 2017 we introduced a pilot EQA scheme to monitor 3rd and 4th field HLA typing results. Participants were invited to HLA type 10 Scheme 4A2 blood samples to the 3rd or 4th field resolution. Participants could report any combination of 3rd/4th field results for any combination of HLA loci. Results were not assessed. 17 labs tested between 5-10 samples, resulting in 1763 alleles reported at the 3rd or 4th field resolution. The highest number of 3/4th field results were reported for HLA-B (n=306), -C (292), -A (287), -DRB1 (277) (27). 1017 alleles (57.7%) were reported as an unambiguous 3rd field result (e.g. B*07:02:01) and 639 (36.2%) were reported at as unambiguous 4th field result (e.g. B*07:02:01:01), while 102 (5.8%) contained reports of multiple alleles with ambiguities at the 3rd or 4th field (e.g. B*07:02:01:01/03). The highest proportion of unambiguous 4th field results were reported for HLA-A (65.2% of A results) and the least for DPBA (16.3%). The highest proportion of 3rd/4th field ambiguities were reported for HLA-DRB1 (11.2% of DRB1 results), while none were reported for DQA1, DPA1 or DPB1. There were six alleles where the 3rd or 4th results from participants were not all in agreement. For example, a HLA-C allele was reported as C*04:01:01 (n=3), C*04:01:79 (n=7), C*04:01:01:06 (n=1) and C*04:01:NEW (n=2). The other discrepant alleles involved DRB1*04:01, DPB1*04:01 (x2), DRB1*15:01, C*14:02. The finding of discrepant 3rd/4th field HLA types shows that an EQA scheme is required. UK NEQAS for H&I will formally assess results reported at the 3rd field from 2018 in Scheme 4A2. Killer-cell Immunoglobulin-like Receptors (KIRs) are encoded by 17 genes in the Leukocyte Receptor Complex (LRC). These genes are highly polymorphic, on the level of structural variants and copy-number variation. Furthermore, the KIR genes show a high degree of sequence homology. These characteristics make it challenging to design specific amplifications and to perform reliable analysis resulting in accurate identification of each KIR allele present in a sample. To enable high-resolution KIR genotyping, we developed an NGS-based strategy that involves whole gene amplification of KIR genes, using KIR-specific primers that amplify one or more KIR genes. The amplicons are processed by library preparation with NGSgo (GenDx) and sequenced on an Illumina MiSeq platform (2x151 bp). The resulting sequences are genotyped with NGSengine software (GenDx), using KIR database IPD-KIR 2.7.0. Multiple IHWG samples with known and unknown KIR genotypes were analyzed using this new strategy. A proof-of-concept study of the KIR framework genes (3DL2, 3DL3, 2DL4 and 3DP1) demonstrated that amplification, sequencing and subsequent typing was successful for the majority of samples. Many samples could be typed unambiguously, despite the fact that phasing was not fully accomplished due to the length of the genes (up to 17 kb). For the other KIR genes (3DL1, 3DS1, 2DS1/2/3/4/5, 2DL5A/B, 2DP1) the locusspecific amplification and subsequent analysis proved to be more challenging, largely due to the high degree of homology in sequence between the genes. Here we present the results, as a proof of concept, that show the feasibility of performing KIR genotyping of the framework genes at high resolution, possibly at allelic level, using the short read sequencing technology of Illumina. This new typing strategy may be an attractive alternative to existing KIR genotype assays that only determine KIR gene content at a limited resolution, or complex sequencing methods aiming to sequence the entire LRC. The simplicity of this assay would allow for integrating the assay with the already existing NGSgo workflow for HLA (GenDx), with the benefit of combining both KIR and HLA typing at a high-resolution level. The MHC class I in cattle is highly diverse, including large copy number variations of polymorphic MHC class I genes, but our knowledge of this diversity is mostly limited to allelic diversity and not haplotype structure, with only three full haplotypes characterized. Haplotypes contain at least one and up to four classical class I genes between genes NC1 and TRIM26, with each classical gene belonging to one of six different allele groups. To date 114 cattle class I alleles have been deposited on the IPD-MHC database of which most are derived from the Holstein breed. The high structural diversity has made it difficult to assemble and fully characterize this region with short sequencing reads produced by second generation sequencing. Through targeted genome enrichment with Roche Nimblegen SeqCapEZ probes and subsequent long-read Single-Molecule Real-Time (SMRT) sequencing we were able to de novo assemble five Holstein MHC haplotypes using individuals from an MHC homozygous herd. Approximately 80% of bases that mapped to the cattle chromosome 23 were located within the target region of~3.4Mb, which included the class II region, producing an average coverage of 20x. The allele content was correct in each assembly and for the first time we have confirmed the location of each classical gene on the haplotype and generated the full gene sequence. Currently we are adapting our analysis pipeline for heterozygous animals, to identify new MHC haplotypes and alleles with confidence to develop our understanding of cattle MHC evolution.

Vicky Van Sandt 1 , Aleksandar Senev 1 , Johan Kerkhofs 1 , Ming Li 2 , Tommy Liu 2 , Heylen Christine 3 , Marie-Paule Emonds 1 1 Belgian Red Cross-Flanders, Mechelen, Belgium, 2 Immucor, Mountain View, CA, United States of America, 3 Immucor Transplant Diagnostics, Nijlen, Belgium Correspondence: vicky.vansandt@rodekruis.be Next generation sequencing separates alleles, making the test more sensitive, but also more prone to cross contamination. Next to that NGS can increase HLA genotyping resolution. Allelic resolution is, however not clear cut, especially for alleles that differ in the number of short tandem repeats (STRs). To map the power versus pitfalls of our NGS workflow we evaluated the MIA FORA test sensitivity and the impact of cross contamination. We assessed the value of allelic resolution looking at the quality of STR calling and compared these results with third generation SMRT sequencing (TGS, PacBio) and Sanger data. The MIA FORA NGS test sensitivity was very high. Results with a minimum coverage of 40 were concordant with the reference genotype (allelic res.) down to 2 ng/l for HLA-A, 0.1 ng/l for -DRB1345 and down to 0.05 ng/l for HLA-B, -C, -DQ and -DP. To evaluate the impact of cross contamination we spiked homozygous samples (33 ng/l) with a DNA reference sample (0.001 to 10 ng/l). Class I data showed no impact of the cross contamination. In class II DBR1345, DPA1 and DPB1 data a second allele was added to the homozygous result starting from a spike concentration of 1 ng/l. For DQA1 and DQB1 the impact of the cross contamination was seen even down to 0.05ng/l. These data reflect the power and limitations of the NGS interpretation algorithms, where allelic imbalance is more tolerated in class II alleles and especially in DQ. MIA FORA allelic resolution was reproducible, with just one fourth field discordance out of 270 allele calls, demonstrating the robustness of the workflow and software. Analysis of homopolymer regions in MIA FORA NGS data showed confident calls, with only limited background of reads with indels (5%). PacBio CCS (circular consensus corrected reads) data of the same samples showed significantly higher background in the homopolymer regions (up to 20%). STRs of more than one nucleotide showed a more diverse distribution in the number of repeats in the NGS data, rendering allelic resolution in these cases questionable. The analysis of the TGS data of these samples is still on going.

Sue Davey 1 , Dimitra Niokou 1 , Zareen Deplano 1 , Monica Kyriacou 1 , Trevor Green 1 , Vasnti Jesani-Hethcoat 1 , Thusa Sathananthan 1 , Colin Brown 1 1 NHS Blood and Transplant, London, United Kingdom (Great Britain) Correspondence: suerdavey@aol.com Use of Next Generation Sequencing (NGS) for HLA typing is increasing throughout Europe. Although NGS has the advantage of HLA allele level definition, the complexity of the laboratory protocol presents practical challenges compared to previous HLA typing technology. At NHSBT Colindale we have developed an in-house protocol which consists of DNA purification, DNA quantification and normalization, whole gene PCR amplification for HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1, amplicon quantification and pooling, library preparation using TruSightHLA (Illumina) followed by sequencing on a MiSeq. For a single sample this requires 24 sample transfer steps and an additional 44 pipetting events. This increases to 6528 pipetting steps, including 2304 sample transfers, when processing a batch of 96 samples. As well as the physical demands presented by manual pipetting and associated risks of repetitive strain, each transfer step has the potential for sample mix up or contamination. To minimize these risks, we have implemented automation for each step of our NGS protocol using Biomek liquid handling robots. Programs for bead-based purification and DNA library preparation were made available from Beckman Coulter, with the remaining automated processes developed in-house, facilitated by sample tracking applications. NHSBT Colindale was the first UK laboratory to implement whole gene NGS for stem cell registry HLA typing using an in-house protocol. Our automated NGS pipeline has reduced the hands-on time by over 70% compared to the manual protocol, with manual pipetting limited to the preparation of bulk reagents. Sample transfer using liquid handling robots and associated sample tracking software has decreased the risk of human error, providing assurance of sample integrity. It is also less operator dependent and enables the high throughput HLA typing of more than 20,000 donors per year. HLA typing in stem cell transplantation diagnostics requires high-resolution typing of a combination of class I and class II HLA genes of donor and recipient. There is evidence that the genotype of the non-classical HLA-E gene may also be clinically significant and may impact transplantation outcome. To enable HLA-E genotyping at high to allelic level resolution, we developed an NGS-based strategy that involves PCR amplification of HLA-E gene, library preparation with NGSgo (GenDx), sequencing on an Illumina MiSeq platform and genotyping with NGSengine software (GenDx). Typing is based on the whole HLA-E gene, including exon 1-8 and all introns. We have typed multiple IHWG and UCLA gDNA samples (n= >100) to validate the performance of the assay. HLA-E amplification is highly robust and gives typing results that are concordant with the expected pre-types. Here we will present our results which show that for most samples an unambiguous HLA-E genotype at the 3rd or 4th field level was obtained. Most alleles identified belong to the HLA-E*01:01:01 and HLA-E*01:03:02 groups. However, other alleles with sequence variation in introns and exons have also been identified, albeit at a much lower frequency. These also include new alleles, which will be submitted to the IPD-IMGT/HLA Database to expand the list of known HLA-E alleles. This new strategy allows for extending the already existing NGSgo workflow for class I and class II with HLA-E, such that ultimately HLA typing can be based on the full spectrum of clinically relevant HLA loci in one single assay.

Amalia Dinou 1 , Maria Spyropoulou-Vlachou 2 , Vasiliki Vrani 3 , Efstathios Michalopoulos 1 , Catherine Stavropoulos-Giokas 1 1 Hellenic Cord Blood Bank-Biomedical Research Foundation of the Academy of Athens, Athens, Greece, 2 Alexandra General Hospital of Athens, Athens, Greece, 3 General Hospital of Athens "G.Gennimatas", Athens, Greece Correspondence: adinou@bioacademy.gr HLA typing requirements for unrelated donors and cord blood units have changed over the years reflecting the changes in the methodologies used. Whereas in the 1990s, when the first registries were established, donors were listed with only HLA-A and -B serology typings, today more than 50% of donors and 20% of CBUs are DNA typed for HLA-A, -B, -C,-DRB1, -DQB1, -DPB1 mostly by SBT. The typings listed in the registries can be available for years and will not be repeated unless requested. Discrepancies can then be revealed, that are not always due to sample mix-ups, or clerical errors, but on the accuracy of the methods used. At the Hellenic Cord Blood Bank (HCBB), the CBU typings were performed by PCR-SSO for HLA-A, -B, -C and-DRB1, while since 2015 for HLA-A, -B, -C, -DRB1, -DQB1, -DQA1 and -DPB1 loci by NGS. The corresponding mothers were also typed for NIMA-matching purposes, but also to help resolve ambiguities and spot sample mix-ups. Recently, 87 CB units that were initially typed by PCR-SSO (Lifecodes (Immucor), LabType (OneLambda)) in 2008 at the National Tissue Typing Center, were retyped by NGS (Holotype, Omixon) at the HCBB HLA Typing lab using the same DNA samples. All the obtained results were concordant at the first field, with the exception of one sample at B locus: for the CBU the typing in file was HLA-B*44, 50 while the NGS typing revealed HLA-B*45:01:01, 49:01:01. The mother was typed as HLA-B*50,51 and HLA-B*49:01:01,51:01:01 respectively. All other loci were consistent with the initial typing for both samples and sample mix-up could be dismissed. The initial typing reports were requested, but only one for the CBU was found and the proposed assignments were B*44:18, 50AD or B*45CKB, 49AE, which are consistent with unphased polymorphisms. A quick search showed that they can't be differentiated by PCR-SSP. Unfortunately, the database containing the maternal typing was corrupt and the initial typing could not be re-examined. In the Twin software the B*44:18, 50:01 pair was not a best match as it presents non-exon mismatches to the obtained phased sequence. It appears that the B*44, 50 pair was assigned in 2008, using the mother typing as reference. This example illustrates the power of NGS methodology, which can resolve most phasing ambiguities especially in class I genes. University Complutense,The Madrid Regional Blood Center (Centro de Transfusion), Madrid, Spain, 2 Immunogenetics, Universidad del Norte., Barranquilla, Colombia Correspondence: arnaizville@hotmail.com America's first inhabitant populations (Amerindians, Na Dene and Eskimos) underwent a drastic population reduction and gene exchange after European and Africans arrival after 1492 AD. The Barranquilla population may be a good model to study present day population admixture in South America. HLA-A, -B and -DRB1 DNA typing has been performed in 188 unrelated individuals originating from the area. They speak Spanish and showed apparent European / African and mixed characteristics. HLA genetic European / African features were found while only 1.85 % showed Amerindian characteristics. This contrasts with the neighboring Cuban population where 10 % Amerindian characters appear. The population of the Ossetians from Vladikavkaz, Russia, is a unique among the ethnic groups of the Caucasus region. Whereas their geographic neighbors all speak Caucasian languages, Ossetians are speak an Iranian language, belonging to the Indo-European language family and they could possibly be descendants of Scythian-Sarmatian tribe of Alans. Therefore, an assessment of Ossetians immunogenetic profile is of particular interest. One hundred and twenty-seven healthy potential bone marrow donors from the Registry of National Research Center for Hematology, Moscow were typed for HLA-A, -B, -C, -DRB1, -DQB1 (Luminex, Immucor, USA). All donors were from Vladikavkaz and Ossetians by nationality. The most frequent alleles of the HLA-A locus were A*02 (frequency 0.248), A*30 (0.133), A*03 (0.122) and A*01 (0.118). Three most common alleles in the HLA-B locus were B*49 (0.161), B*51 (0.129), B*07 (0.106), and in the HLA-C locus -C*07 (frequency 0.421), C*06 (0.153) and C*12 (0.083). Among the alleles of HLA class II genes, HLA-DRB1*13 (0.216) and *11 (0.185), HLA-DQB1*06 (0.338) and *03 (0.311) were prevalent. One hundred and thirty-eight fivelocus haplotype variants were identified among Ossetians (Arlequin 3.5). The most frequent haplotypes were A*30~B*49~C*07~DR*13~DQB*06 (frequency 9.1%), A*03~B*07~C*07~DR*15-DQB*06 (3.9%), A*02~B*51C *16~DR*11~DQB*03 (3.1%) and A*32~B*08~C*07D R*03~DQB*02 (2.8%). Haplotypes A*23~B*50~C* 06~DR*03-DQB*02, A*01~B*49~C*07~DR*11~DQB*03 and A*01~B*08~C*07~DR*03~DQB*02 were found in 2% cases. Thus, the population of the Ossetians of Vladikavkaz combines both features of European and Asian populations this indicated by high frequency of such allelic variants as A*02 and A*30, B*07 and B*51, B*49. A distinctive feature of Ossetians is the high frequency of HLA-B*49 and HLA-DRB1*13 alleles and A*30~B*49~C*07 DR*13~DQB*06 haplotype as compared with other Russian and world populations. The population of Ossetians is a unique material for inclusion in the registry of potential bone marrow donors, since the presence of representatives with such unique genotypes in the register increases the chance of finding an unrelated donor for patients with rare HLA genotypes. We determined the incidence of HLA haplotypes associated with CD in a group of pediatric patients in the Slovak population with CD, evaluated the risk of antigens HLA-DQ2 and HLA-DQ8 to the development of CD and importance of their examination in clinical practice, assessed whether the double dose of the associated genes affects the predisposition to CD. The group of patients included 66 patients with CD (40 women, 26 men) aged from 2 to 27 years with histologically confirmed diagnosis of CD. HLA alleles/haplotypes were determined by a PCR-SSP commercial kit (BAG Health Care GmbH, Germany). For statistical analysis we used chtest, Pcorr, Odds Ratio and validity tests. HLA haplotypes associated with CD had in the group of CD patients the following frequencies: haplotypes with the allele DQB1*02:01 (57.6%), haplotypes with the allele DQB1*02:02 (36.4%) and haplotypes with the allele DQB1*03:02 (19.7%). 89.4% of patients had in their genotype one of the associated HLA haplotypes. About 10% of patients had no haplotype with predisposition to CD. Immunogenetic analysis based on the molecular genetics confirmed the association of antigen HLA-DQ2 (allele DQB1*02:01+DQB1*02:02) with the CD in the Slovak population. The frequency of DQ2 in the patients group was 83.3% and in the control population 37.8%, OR=8.2. The antigen HLA-DQ8 was associated with CD only in DQ2 negative patients and represents an additional genetic risk with OR=7.4. The combined genetic risk for antigens DQ2+DQ8 for CD is characterized in the Slovak population by the following values the frequency in CD patients 89.4%, frequency in a population of 44.8%, OR=10.4, specificity 92%, predictive value of a positive finding of 89%. Patients with a double dose of the associated alleles DQ2 (composed homozygous), show a significantly higher risk of developing CD in comparison with patients which have only one dose of allele DQ2 (OR=2.7).

Turkan Patiroglu 1 , Murat Cansever 1 , Aykut Poyraz 1 1 Erciyes University, Kayseri, Turkey Correspondence: turkanp@erciyes.edu.tr Combined immunodeficiency (CID) is a primary immunodeficiency diseases with severe loss of T and B lymphocyte function. Lymphopenia is an important finding in diagnosis of CID. Therefore, we planned the screening lymphopenia in cord blood samples during labor. Complete blood count (CBC) was measured in cord blood samples of every baby born in our hospital from January to December 2018. If lymphocyte counts were below 3000/ mm3, it was classified as neonatal lymphopenia. Immunological investigations including serum immunoglobulin levels, lymphocyte subgroups and chest radiography for thymus shadow were evaluated in newborns who had lymphopenia. In this study, CBC was measured in 1500 cord blood samples during labor and lymphopenia was found in 39 of them. Mean lymphocyte count was calculated as 2250/mm3. Two of 39 newborns with lymphopenia had CID according to immunological investigations. One patient with CID had RAG 1 deficiency. Lymphopenia can occur and may include infectious, genetic, systemic and iatrogenic causes. Early diagnosis of CID can be lifesaving, so neonatal screening of lymphopenia is important for early diagnosis of CID. The Samara Region is located in the southeastern part of the European territory of Russia, in the middle of the Volga region. The Chelyabinsk Region is situated 867 km to the east from Samara Region at the slopes of South Urals and in Trans-Urals. The majority of the populations of these two regions are ethnic Russians. According to the last population census (2010) 89.47% of Samara Region and 86.53% of Chelyabinsk Region habitants are Russians. The present study was initiated to investigate the HLA allele and haplotype diversity in these two Russian populations. HLA typing was performed using SSO and SSP techniques at low resolution for HLA-A, -B and -DRB1. We analyzed the HLA data of 2795 cord blood units from the Samara regional cord blood bank and 591 bone marrow donors from the Chelyabinsk regional blood transfusion station. All the donors and CBUs parents were surveyed according to genealogic recommendations and live full time in Samara and Chelyabinsk Regions. Despite the fact that most frequent alleles were the same for both populations we found significant differences in presence of A*02 (27.3% in Samara population vs 30.7% in Chelyabinsk population), A*30 (3.3% vs 1.1%), A*68 (4.4% vs 2.7%), B*07 (10.8% vs 13.2%), B*14 (3.0% vs 1.9%), B*45 (0.1% vs 0.4%). In allele frequencies of DRB1 locus we found no significant difference. Although we saw some differences in allele frequencies, the 10 most frequent haplotypes in both populations were the same: A*01~B*08~DRB1*03, A*03~B* 07~DRB1*15, A*03~B*35~DRB1*01, A*02~B*13~DR B1*07, A*02~B*07~DRB1*15, A*02~B*41~DRB1*13, A*25~B*18~DRB1*15, A*24~B*07~DRB1*15, A*02~B* 27~DRB1*01 and A*26~B*38~DRB1*13. We found that Russians in different populations of Samara and Chelyabinsk Regions have the small difference in HLA allele frequencies. On the other side the top 10 frequent haplotypes were the same. The data of HLA allele and haplotype frequencies is primarily intended for population genetic purposes and disease association studies. Severe acute respiratory syndrome (SARS) was known as the first major epidemic at the beginning of 21st century, causing thousands of infections worldwide. China was the place of the epidemics origin and had the most severely affected population around the world. This highly infectious disease was caused by a novel coronavirus (SARS-CoV) with the immune reaction still remaining unclear. Human leukocyte antigen (HLA) variation is reported to be associated with a wide range of infections due to its essential role in immune response. Here, we aimed to describe the genetic patterns of HLA-A, -B, -C, -DRB1 and -DQB1 loci in 70 recovered SARS patients from Beijing and examine the association between HLA genes and susceptibility or resistance to SARS using the data of north Chinese bone marrow donors as control. A total of 70 recovered Chinese Han SARS patients were recruited to donate convalescent plasma in 2003. All participants were unrelated and diagnosed according to the World Health Organization case definition for SARS. HLA high-resolution typing was carried out using SBT. We report a novel allele HLA-DRB1*12:57, which was officially assigned by the World Health Organization (WHO) Nomenclature Committee in April 2015, identified in a volunteer Chinese cord blood donor. The donor's DNA was amplified and sequenced with AlleleSEQR HLA-DRB1 sequencing kit. The sequence reaction products were processed with an ABI 3500 DNA Analyzer and analyzed with Assign 3.5+ software. Nucleotide sequences of HLA-A, HLA-B, HLA-C exons 2C4, HLA-DRB1 exon 2 and HLA-DQB1 exons 2 and 3 were analyzed. The DNA sequencing result of HLA-DRB1 locus of the sample did not match any known HLA-DRB1 allele combination. Low resolution typing was performed for HLA-DRB1 using GT HLA SSP typing kits and the DRB1 locus was typed as DR12 and DR15. To further characterize the novel HLA-DRB1 allele in the sample, exon 2 sequences of HLA-DRB1*12 and HLA-DRB1*15 alleles were amplified separately using TBG group-specific sequencing kit. In the present study, we were able to describe, with coverage of full-length HLA sequences and phased genotypes, singular 4-field diversity and associations of both HLA alleles and haplotypes which were not evident in results obtained by conventional HLA typing methods. Therefore, this level of resolution obtained by NGS allows studying in detail the HLA diversity at a coding and non-coding sequence level. The HLA typing data of this Spanish cohort was also collected as part of the 17th International Histocompatibility and Immunogenetics Workshop (IHIW) under the study of unrelated subjects by NGS HLA. These results may contribute as a useful reference source for future population studies, disease association studies as a healthy control group and for future planning of the unrelated bone marrow donor registry. HLA-A*01~B*08~Cw07~DRB1*03~DQB1*02:01~DPB1*04:01 was the most frequent (6%) 6-loci HLA haplotype detected. The most common 3-loci HLA haplotypes in our studied group were: HLA-DRB1*11~DQB1*03:01~DPB1*04:01 (10%), HLA-DRB1*11~DQB1*03:01~DPB1*04:02 (10%) and HLA-DRB1*03~DQB1*02:01~DPB1*04:01 (8%). Our initial data regarding HLA-DPB1 haplotypes show similarities with frequencies of haplotypes observed in a previous study of our center with 816 HLA-A, -B, -C, -DRB1 haplotypes. The allele frequencies, as well as the preliminary haplotype frequencies correspond to the frequencies of respective HLA alleles and haplotypes in Greek and European population. These results will be useful in further population and disease association studies, genetic distance studies in the Balkan and Mediterranean area, paternity cases, as well as in HSCT and SOT taking into account the HLA-DP implementation in donor specific antibodies monitoring.

Maria Anagnostouli 1 , Serafeim Katsavos 1 , Artemios Artemiadis 1 1 Medical School of Athens National and Kapodistrian University, Aeginition Hospital, Athens, Greece Correspondence: managnost@med.uoa.gr Multiple Sclerosis (MS) is the most studied demyelinating disease of the Central Nervous System, affecting mainly young adults. Early-onset-MS (EOMS, pediatric-adolescent) is a rare MS manifestation, representing 3-5% of all cases. The role of HLA-DRB1*15:01 allele is well established in adult-onset-MS (AOMS) and confirmed in many Caucasian populations, the Hellenic included. HLA-DRB1 allelic frequencies in EOMS have been found to be different in various ethnic groups, with a preponderance of DRB1*03 in Hellenic EOMS patients. Recently, carriage of DPB1*03 allele has been identified as another genetic predisposing factor for developing AOMS, in Hellenic and other Caucasian populations. The total absence of studies regarding DPB1 allelic frequencies in EOMS though, prompted our team to examine this characteristic in a group of 77 Hellenic MS patients. We recruited 18 EOMS and 59 AOMS patients and genotyped them for DPB1 polymorphisms (154 alleles), after obtaining formal consent. HLA-genotyping was performed by a Sequence Specific Oligonucleotide (SSO) technique. We compared allelic frequencies between the EOMS and AOMS subgroups, as well as with a sample of Hellenic healthy controls, genotyped for DPB1 in a previous study (492 alleles). AOMS patients had significantly lower DPB1*02 (11.9% vs. 19.3%) and higher DPB1*03 (15.3% vs. 7.1%), DPB1*33 (1.7% vs. 0.2%) and DPB1*35 (2.5% vs. 0.4%) allelic frequencies than healthy controls. In patients with age at MS onset (AAO) younger than 15 years significantly lower DPB1*33 frequency (6.3% vs. 0.2%) was also found. EOMS patients had significantly higher frequency of DPB1*02 than AOMS patients (27.8% vs. 11.9%). The previously confirmed distinct DRB1* profiles between EOMS and AOMS, along with the discovered differences in DPB1* allelic frequencies in this study, highlights a possible fundamental role of genome in determining AAO in MS. Our results need replication in larger MS cohorts, of different ethnic groups, aiming at more individualized therapeutic decision-making in the future.

Antonio Arnaiz-Villena 1 , Jose Palacio-Gruber 1 , Ester Muñiz 1 , Cristina Campos 1 , Javier Alonso-Rubio 1 , Ignacio Juarez 1 , Jorge Nieto 1 , Shadallah Fareq Salih 2 , José Manuel Martín-Villa 1 , Rawand Al-Qadi 2 1 University Complutense, The Madrid Regional Blood Center (Centro de Transfusion), Madrid, Spain, 2 Dohuk Specialized Laboratory Center, Dohuk, Iraq Correspondence: arnaizville@hotmail.com Kurds from Iraq (Dohuk and Erbil Area, North Iraq) have been analyzed for HLA genes. Their HLA genetic profile has been compared with that of other Kurdish groups from Iran and Tbilisi (Georgia, Caucasus) and also worldwide populations. A total of 7,746 HLA chromosomes have been used. Genetic distances, neighbor joining dendrograms and correspondence analyses have been carried out. The haplotype HLA-B*52~DRB1*15 is present in all three analyzed Kurdish populations. HLA-A*02~B*51~DRB1*11 is present in Iraqi and Georgian Kurds. Haplotypes common to Iranian and Iraqi Kurds are HLA DRB1*11~DQB1*03, HLA DRB1*03~DQB1*02 and others in a lower frequency. Our HLA study conclusions are that Kurds most probably belong to an ancient Mediterranean / Middle East / Caucasian genetic substratum and that present results and those previously obtained by us in Kurds may be useful for Medicine in future Kurd transplantation programs, HLA epidemiology (HLA linked diseases) and pharmacogenomics (HLAassociated drug side effects) and also for anthropology. It is discussed that one of the most ancient Kurd ancestor groups is in Hurrians (2,000 years BC). Genetic distances, neighbor joining and correspondence analyses also showed that Azeris were close to Kurds, who have shown a closer Mediterranean/Caucasus HLA profile, and Gorgan (Turkmen) who have shown a closer Central Asia profile, as expected. It is shown that three different Iranian populations according to language, history and geography: Gorgans, Kurds and Azeris are genetically close. In fact, the old Azeri language (Adari) was an Iranian language and not a Turkic one, which they nowadays speak. Also, the present study does not support Aryan invasion from the East in accordance with many other previous studies. Finally, our results are useful for establishing preventive medicine programs in transplantation and HLA and pharmacogenomics/disease linkage. University Complutense,The Madrid Regional Blood Center (Centro de Transfusion), Madrid, Spain, 2 Dohuk Specialized Laboratory Center, Dohuk, Iraq Correspondence: arnaizville@hotmail.com HLA-G and HLA-A frequencies have been analyzed in Amerindians from Ecuador. HLA-G allele frequencies are found to be closer to those of other Amerindians (Mayas from Guatemala and Uros from Peru) and closer to European ones than to Far East Asians groups, particularly, regarding to HLA-G*01:04 allele. HLA-G~A haplotypes have been calculated for the first time in Amerindians. It is remarkable that the HLA-G*01:05N allele is found at a very low frequency (like in Amerindian Mayas and Uros) and is also found in haplotypes belonging to the HLA-A19 group of alleles (HLA-A*30, A*31, A*33). It was previously postulated that HLA-G*01:05N appeared in HLA-A*30~B*13 haplotypes in Middle Eastern Mediterraneans. It may be hypothesized that in evolution, HLA-G*01:05N existed primarily in one of the HLA extant or extinct A19 haplotypes, whether this haplotype was placed in Middle East or other world areas, including America. However, the highest present day HLA-G*01:05N frequencies are found in Middle Eastern Mediterraneans.

Cinzia Vecchiato 1 , Evi Lochmann 1 , Ivo Gentilini 1 1 Ospedale di Bolzano, Bolzano, Italy Correspondence: cinzia.vecchiato@sabes.it

The IBMDR Donor Centre in South Tyrol started its activity in 1992, from then 7704 donors subscribed and 72 of them have donated. From 2012 SBT typing was systematically used for registry donors at subscription. Between 2012 and 2017, 2306 new donor samples were typed at high resolution for HLA-A, -B, -C, -DRB1. According to our policy, no ambiguities are accepted if null alleles are involved. Moreover, we decided to solve all ambiguities for DRB1 and for HLA class I 99.5% of typing data entered in the registry present less than 3 ambiguities. In an ongoing study on HLA allele frequencies in our population, we observed that, thanks to introduction of SBT, twelve alleles, considered to be rare, were detected. In our samples we detected the null alleles B*08:30N and C*04:09N and six alleles, that meet the definition of Well Defined according to Mack SJ et al. (Tissue Antigens 2013; 81:194203) , not previously reported in Caucasian population. Nine new alleles were identified and described. Four new alleles are located in locus B, two new alleles at locus A and C and one at locus DPB1. In six of them the difference with already known alleles encompassed exon 2-3 and resulted in a different protein at antigen binding site. Since a rare allele could be dismissed in reporting, because it is considered not likely to be found, the retrospective study demonstrates that even in Caucasian populations, that are the most represented in bone marrow registries, rare and infrequently described alleles are not impossible to detect. The data confirm the relevance of a high-resolution typing at subscription for Bone marrow donors in order to select the most suitable donor for patients, waiting for bone marrow transplantation, in an acceptable time. It also put the attention on the importance of sharing typing data for a better knowledge of population genetics. The dissemination and confirmation of rare alleles HLA is of paramount importance to improve the field of immunogenetics as well as in the daily working of a histocompatibility laboratory. The HLA-C*06:76:02, allele not yet confirmed, has been found in a Spanish patient diagnosed with psoriasis. The complete HLA class I typing for this sample was the following: A*03, 26; B*14, 15; C*06:76:02, 08. The DNA-based HLA class I typing was performed by PCR-SSOP Luminex method using Lifecodes HLA typing kits (Gen-Probe) in accordance with the manufacturer's instructions. However, when it was studied by low resolution by PCR-SSP using Fluogene technology, the most plausible typing option given by the software was C*08, *12. This sample has also been studied by sequencing and the typing obtained has been the same as that found by luminex, confirming this result (C*06:76:02). It is important to note that low-resolution kits do not allow the detection of rare alleles and also, in some cases may lead to incorrect low-resolution typing, while with luminex technology we improve in resolution and thus in the detection of possible allelic variants. In summary, we have found a rare HLA-C allele in a Spanish patient and with the complete study of gene sequencing, this allele will be confirmed. The HLA loci are among the most polymorphic genes in the human genome. Mechanisms for generating polymorphism within MHC include point mutations, gene conversion and recombination events, that provide a means of generating novel haplotypes. During routine HLA typing we identified a suspected unequal crossing over on the short arm of chromosome 6 in a 7 year old patient of Bengali origin, affected by Adrenoleukodystrophy (ALD), a demyelinating disease inherited as a sex-linked recessive trait. Allogenic bone marrow transplantation is the only treatment that can stop the worsening effects of ALD. Peripheral blood samples were collected from the patient as well as from his parents and one brother. HLA-A, -B, -C, -DRB1, -DQB1, -DPA1 and -DPB1 alleles were typed by PCR-SSO (One Lambda Inc) and PCR-SSP (Olerup SSP). The DPB1 locus was further investigated by SBT analysis trough the sequence of exons 2, 3 and 4 in forward and reverse directions (One Lambda Inc), which provided a result without ambiguities. On the basis of family typing, haplotypes were assigned as follows: Father (haplotypes a,b) : a) A*11:01~B*15:02~C*08:01, DRB1*07:01~DQB1*02:02D PA1*01:03~DPB1*04:01/ b) A*24:07~B*35:05~C*04:01, DRB1*12:02~DQB1*03:01~DPA1*02:02~DPB1*04:01; Mother (haplotypes c,d): c) A*03:01~B*35:01~C*04:01D RB1*01:01~DQB1*05:01~DPA1*01:03~DPB1*04:02/ d) A*11:01~B*35:03~C*04:01~DRB1*12:01~DQB1*03:01D PA1*?~DPB1*--. The patient inherited the haplotypes a/d, the brother the b/c haplotypes. The patient's appeared DPB1*04:01 homozygous, thus apparently with no HLA-DPB1 specificity derived from his mother. A STR analysis was performed (PowerPlex 16 HS Promega, that coamplifies in a single PCR for 16 loci: D18S51, D21S11, TH01, D3S1358, FGA, TPOX, D8S1179, vWA, CSF1PO, D16S539, D7S820, D13S317, D5S818, Amelogenin, Penta E and Penta D), that confirmed sibling relationship among the 4 individuals. The generation of a such HLA haplotype lacking any DPB1 specificity was most likely to be due to an unequal crossover occurring during meiosis in a previous generation, i.e. a recombination event that happens when the breaks do not occur at precisely homologous points in two chromatid strands, and hence results in localized duplication of genetic material in one chromatid and complementary deletion in the other, resulting in the generation of new haplotypes.

Carlos H. Parga Lozano 1 , Franklin Torres 2 , Aracely Garcia 2 1 Universidad del Atlántico, Barranquilla, Colombia, 2 Universidad Libre, Barranquilla, Colombia Correspondence: pargacarlos@yahoo.com Barranquilla is a city in northern Colombia. The city has had high rates of immigration in the last 200 years, mainly from Central Asia, Africa and Europe. That caused a significant miscegenation of the population along with native Amerindians in the region. 43 individuals from Barranquilla were typed for the detection of the HLA-DRB1 and DQB1 alleles. The HLA allele frequencies of 738 chromosomes were analyzed and compared these with 9 Colombian Amerindian populations using genetic distances, neighbor joining dendrograms (NJ) and correspondence analysis. The most frequent alleles were DRB1*15:01 (0.088889), DRB1*01:01 (0.077778), DRB1*04:01 (0.077778) and DQB1*06:01 (0.16667), DQB1* 03:02 (0.15556), DQB1*05:01 (0.13333). Several new haplotypes were obtained. Of which the most frequent were DRB1* 01:01~DQB1*05:01 and DRB1* 04:01~DQB1*03:02. According to the NJ results and correspondence analysis, the population is not related to the Amerindians, due to its presence as a group in the NJ phylogenetic tree. Based on these analyses it was found that there is a close relationship between the indigenous people from Colombia, mainly with the indigenous from Sierra Nevada de Santa Marta. Several DRB1~DQB1 haplotypes were not reported previously and this population from Barranquilla is not related to any of the Amerindians directly, which suggests a miscegenation and therefore in future studies should be included for mestizo, Caucasian, Asian and Black populations to obtain a better phylogenetic relationship. During HLA typing of a patient with multiple myeloma and his brothers and sisters, we found a new DQA1*01 allele in three of his five siblings. OneLambda rSSO typing (Luminex) together with the analysis Software Fusion gave as DQA1 typing results: DQA1*01:07Q and *03. Because the first DQA1 allele in our family should be part of the DRB1*01:01~DQA1*01:01~DQB1*05:01 haplotype and DQA1*01:07Q (most similar to DQA1*01:04) is normally part of the DRB1*14:01P~DQA1*01:07Q~DQB1*05:03 haplotype, we were not sure if these were the right results. For that reason, we typed the DQA1 with next generation sequencing (NGS). We developed a workflow based on long range PCR (LR-PCR) and NGS. Therefore, we designed different HLA locus and/or allele specific LR-PCRs and sequenced the generated amplicons on a MiSeq platform (Illumina). The subsequent NGS data evaluation was performed with two different HLA software tools (Omixon Twin, Omixon and NGSengine, GenDx). The phased sequence alignment according to the individual single nucleotide variants (SNVs) pattern present ended up with allele-specific contigs. The final alignment of these contigs was done with AliView (Muscle) and BioEdit (ClustalW) software along with published IPD-IMGT/HLA Database DQA1 sequences. The full-length sequence analysis of the described allele unraveled a quite high similarity to the DQA1*01:01 allele. As in DQA1*01:07Q and DQA1*01:13 in the new allele we found a point mutation (C> T, CGC > TGC) resulting in Arginine (R) > Cysteine (C) in exon 2. This supplementary Cysteine seems to be the cause of the questionable (Q) expression of these three alleles. LR-PCR and NGS including phased sequence analysis revealed the unambiguous full-length sequence of the HLA-DQA1*01Q:new allele. The resulting HLA-B13 protein should have 8 additional amino acids. In HLA C*04:09N the deletion of one nucleotide in the penultimate exon (7), causes a frameshift and an elongation with 32 additional amino acids in the cytoplasmic tail. For a so far unknown mechanism, this elongation switches C*04:01:01:01 into the C*04:09N allele and therefore we expected this novel HLA-B*13:02 variant could be a null allele as well. To answer this question, we performed a serological typing for HLA-A, B and Cw with the help of CDC typing trays from OneLambda. To verify the mutation in exon 7 of B*13:02:new (found with NGS) we used Sanger sequence analysis with specific amplification and sequencing primers.

Serological typing revealed the following antigen pattern: A1, A31; B8, B13; Bw4, Bw6; Cw6, Cw7. Sanger sequence analysis confirmed the minor sequence differences between B*13:02:01:01 and B*13:02new. The positive recognition of B*13:02:new by the anti-B13-antibodies in CDC typing, showed that this allele is expressed on the cell surface and that the HLA-B13 protein is folded correctly. Its functionality in respect of signal transduction needs to be elucidated. Because the stated point mutation is outside from exon 2 and 3, this allele is a part of the B*13:02P and the 13:02:01G groups. With our approach, applying LR-PCR and NGS including phased sequence analysis, we were able to determine the complete gene sequence of HLA-B*13:02. The patients included 64 individuals, age range 17-66 years old with diagnosed HBC infection and the control group were consisted of 21 healthy people age range 23-50 years old. IL-28B gene polymorphism (rs8099917 G/T) were determined by a polymerase chain reaction restricted fragment length polymorphism. Among all participant, 56(87.5%) were HB Ag-positive compared to 29 (12.5%) were HB Ag-negative. IL-28 (rs80) TT and GT genotypes were more frequent in HBC patients in compared with the control group. There was no statistically significant association between the frequencies of IL-28B alleles in the patients and the normal population. This study showed that genotype TT in married group was higher than the single group and there was a significant difference in allele frequencies of the patient group. This study indicates that IL-28B TT and GT genotypes may influence increasing susceptibility to HBC infection. Next-generation sequencing is increasingly used in transplantation settings, but also as a method of choice for indepth analysis of population-specific HLA genetic architecture and its linkage to various diseases. With respect to ethnic admixture characteristic for the East Croatian population, we aimed to investigate the frequency and linkage between seven HLA loci in 112 healthy, unrelated, blood donor volunteers (female/male:32/80, 20-61 years, median age 34.5) originating from four eastern Croatia counties. Genomic DNA was extracted and HLA genotypes determined using high-resolution Omixon Holotype HLA 96/7 and 24/7 kits on Illumina MiSeq sequencing platform. HLA-A, -B, -C, and -DQA1 loci were sequenced for their entire length, -DQB1 and -DPB1 from intron 1 to 3´UTR, and -DRB1 from intron 1 to intron 4. Collected reads were analyzed using Omixon Twin software and IPD-IMGT/ HLA Database release 3.29. Allele frequencies and linkage disequilibrium statistics were obtained by direct counting and PyPop 0.7.0 software. Arlequin v3.5.2.2 was used for haplotype frequencies (ML-EM algorithm) and Hardy-Weinberg departures (Guo-Thompson test). In total, 215 alleles were determined at third/fourth field resolution, with 25, 40, 28, 39, 34, 20 To identify genetic markers for allopurinol-induced SCAR, we carried out a case-control association study. HLA class I (A and B) typing was performed by complement dependent cytotoxicity (CDC) on 11 patients with allopurinol-induced SCAR and 136 control individuals (group 1: 13 allopurinoltolerant subjects and group 2: 123 healthy subjects from the general population). We described for the first time a positive association of HLA-A10 and HLA-B8 antigens with allopurinol-induced SCAR in our Tunisian population. The comparison of the frequencies of HLA-A10 and HLA-B8 antigens in patients and control group 2 was significant (36.36% vs 9.76%; p= 0.03; OR=5.29 and 45.45% vs 13.01%; p= 0.02; OR=5.57 respectively). The HLA-B17 antigen, which has a frequency of 12 % in Tunisian population, was not found significantly associated with allopurinol-induced SCAR (p=0.08). It was detected in 36.36% of cases (4/11) and 15.38% of controls group 1 (2/13) and only 12.2% of controls group 2 (15/123). Unlike the findings of the other association studies, we described for the first time, that HLA-A10 and HLA-B8 antigens were associated with allopurinol induced SCAR in our South Tunisian population. Blood donors joining the BBMR are HLA-A, -B, -C, -DRB1, and -DQB1 typed using whole gene NGS methodology. In 2017 a whole gene HLA-DPB1 typing system, using published primers, was implemented. Analysis of the first 6997 donors has revealed a high level of diversity and here the HLA-DPB1 alleles that have been unambiguously identified in north Europeans, African/African-Caribbean and Chinese donors have been reviewed and compared. Eleven HLA-DPB1 alleles were identified in all three groups. The northern European donors comprised the largest group (n= 5370) with the highest diversity of HLA-DPB1 alleles. Twenty-five HLA-DPB1 alleles were only identified in this group and whilst 21 of these were identified in only one individual, the other four alleles were present in between 8 to 63 of the panel (phenotype frequency 0.15% to 1.2%). There were 289 donors of African ancestry; 28 different HLA-DPB1 alleles were identified in this group, ten of which did not occur in the northern European or Chinese cohorts; six of these occurred in only one individual, the other four alleles were present in 5 to 30 people (phenotype frequency 1.7% to 10.4%). Twenty different HLA-DPB1 alleles were identified in the group of 70 Chinese donors. Five of these alleles were not identified in the north European or African groups; two of these were found in six and eight of the Chinese donors (phenotype frequency 8.6 to 11.4%). Assignment of HLA-DPB1 alleles, without the use of the G group nomenclature is problematic with most NGS typing systems. However, where HLA-DPB1 types have been unambiguous defined, these data show that, similar to other HLA loci, some HLA-DPB1 alleles are characteristic of particular ethnic groups. Matching for HLA-DPB1 is now widely considered in the selection of unrelated stem cell donors and knowledge of the relative frequency and occurrence of HLA-DPB1 alleles in different ethnic groups can contribute to estimating the likelihood of identifying a matched donor. To complete the genomic sequence and to clarify the degree of recombination within this allele, we developed a workflow based on long range PCR (LR-PCR) and next generation sequencing (NGS). Therefore, we designed different HLA locus and/or allele specific LR-PCRs and sequenced the generated amplicons on a MiSeq platform (Illumina). The subsequent NGS data evaluation was performed with two different HLA software tools (Omixon Twin, Omixon and NGSengine, GenDx). The phased sequence alignment according to the individual single nucleotide variants (SNVs) pattern present ended up with allele-specific contigs. The final alignment of these contigs was done with AliView (Muscle) and BioEdit (ClustalW) software along with published IPD-IMGT/HLA Database sequences. The full-length sequence analysis of the DRB1*03:24 shows, the insertion of parts of DRB1*03:01 (donor) to DRB1*13:02 (acceptor) is restricted to exon 2, only. LR-PCR and NGS including phased sequence analysis revealed the unambiguous full-length sequence of the HLA-DRB1*03:24 allele. In mid-2017 we extended the upfront genotyping profile for registry samples to include the MICA and MICB genes. Because both genes show a high degree of sequence homology, we established a multiplexed PCR amplifying exon 2, exon 3, and an amplicon spanning exons 4 and 5. Using this NGS based workflow, we genotyped more than 400,000 registry donors and report on the observed allele frequencies in a predominantly European cohort. The unprecedented depth of the study allowed us to observe 59 of the 86 described MICA alleles distinguished at the protein coding level. The thirteen most abundant alleles account for a cumulative allele frequency of 99%, while the 37 least frequent alleles account for less than 0.1%. This newly developed NGS-based genotyping approach offers the opportunity to analyze the genetic diversity of MICA and MICB in large cohorts at high-resolution.

functions. The aim of the IHIW component Immunogenetics of Ageing is to identify new biomarkers for successful aging focusing on the immune response genes. Collaborative studies showed that KIR and functionally-relevant MBL2 haplotypes are important factors for control of CMV infection in the elderly and some cytokine genotypes related to increased anti-inflammatory profiles are positively associated with longevity. Longevity in all populations studied is associated with positive selection of HLA-DRB1*11 and DRB1*16 haplotypes, shown to be protective for diseases. Therefore, within the 17th IHIW we applied NGS approaches in order to precisely identify immunogenetic determinants of successful ageing. Statistically significant associations with longevity were observed for 18 HLA alleles and 20 haplotypes. Among them were alleles: A*01:01:01:01 and C*07:01:02, shown as a part of the ancestral haplotype to be related to one of the hallmarks of ageing -Th1 to Th2 shift of the immune response; DRB1*04:01:01:01, which although associated with autoimmunity confers better immune response against viral infections; DRB1*04:10:01, DRB1*14:04:01; DPB1* 20:01:01. Our data confirmed and clarified further positive association of ageing with DRB1*11:04:01 haplotypes. Additionally, the B*51:01:01:01~C*15:02:01:01 haplotype, shown to be protective for some hematological malignancies was significantly increased in elderly. These results allow elucidating further HLA associations with longevity and suggest that HLA alleles and haplotypes could be an informative immunogenetic marker for successful ageing. Discovery of potential new markers that influence the longevity and susceptibility to age related diseases will contribute to development of strategies for rejuvenating the immune system and preventing replicative senescence. In June 2017 high resolution HLA-typing in the setting of an unrelated hematopoietic stem cell transplantation (HSCT) donor search was carried out in our laboratory for a 74 year old acute myeloid leukemia patient. The initial blood samples were HLA-typed by sequence-based typing, using generic as well as antigen-specific primers. Due to an unexpected result, the samples were retested using a next generation sequencing (NGS) long-range PCR methodology. The initial typing revealed a normal HLA-B*51:01 allele as well as a chimeric B*07:02/*07:new HLA-B variant. The point mutation consisted of a C>T base change at position 4 of exon 3 resulting in an amino acid change from Serine to Phenylalanine at position 93. The identification of the dual HLA-B*07 allele was made possible through HLA-B*07 specific sequencing primers that allowed sequence phasing. NGS testing showed that the prevalence of the allelic variant carrying the point mutation represented about 20-25% of patient's overall HLA-B locus content. In order to exclude the possibility that this point mutation was leukemia-associated, an additional mouth swab specimen was requested. The results from this sample confirmed the initial typing. Subsequent high-resolution typing of the patients two sons (B*15:01, 51:01 and B*07:02, 13:02 respectively) disclosed that only the mutation-free B*07:02 variant passed on to the patient's descendant. Considering the similar results between blood and saliva samples in conjunction with the high prevalence rates of the new allelic variant, it is plausible to presume that the latter must have arisen at some early phase of patient's embryogenesis and hence is most likely expressed on non-blood cells as well.

The patient was finally transplanted with an unavoidably Ballele mismatched peripheral blood stem cell (PBSC) graft. This patient case is a perfect example of how modernization of HLA-typing techniques can unravel rare genetic alterations potentially significant for a patient's treatment and prognosis, that could go unnoticed by other still in use, albeit outdated methodologies. The extensive linkage disequilibrium (LD) in the extended HLA (xHLA) region is well-known, but there is also data suggesting no difference from the rest of the genome and haplotype-specificity of LD. Current algorithms do not consider LD beyond 500 kb. Given that xHLA covers more than 7 Mb, we used Haploview to analyze the Immunochip data from 95 HLA-typed IHWG reference cell lines covering the whole of xHLA, and 1000 Genomes (1KG) Project data for replication. More than 1000 pairs of SNPs mapping to the extreme ends of xHLA showed very longrange (>1.5Mb) LD with r 2 >0.80. The strongest longrange LD (r 2 =1.0; D=1.0; LOD=39.6) was observed between common SNPs rs7341211 (chr6: 29,459,360 in OR2H1) and rs589428 (chr6: 31,880,443 in SLC44A4) / rs535586 (chr6: 31,892,560 in EHMT2) with no HLA specificity. The longest-range (3.7Mb) absolute LD (r 2 =1.0; D=1.0; LOD=6.4) was noted between rs7761746 (chr6: 29,355,878 in OR5V1) exclusive to HLA-A*68:02:01:01) and rs6928954 (chr6: 33,093,045) exclusive to HLA-DPA1*03:01. The second longest-range (3.2Mb) LD (r 2 =0.83; D=1.0; LOD=6.4) was between rs1235162 (chr6: 29,569,447) and rs9276689 (chr6: 32,784,185), which were exclusive to the ancestral haplotype (AH) 8.1 (HLA-A*01~B*08~C*07~DRB1*03). In fact, there were hundreds of additional SNP pairs in strong LD (r 2 >0.80) within the same genomic range unique to AH8.1. Only those SNP pairs specific to AH8.1 showed strong LD in the 1KG project data at the population level in any ethnicity. Our effort confirmed that very long-range LD exists in the xHLA but is unique to one ancestral haplotype. These results suggest that any GWAS associations with these AH8.1-specific SNPs are particularly difficult to interpret as very strong LD among them may extend beyond 3 Mb since most AH8. Although antigen-related fluctuation in HLA-C (C) surface expression has been reported for African Americans (Apps et al., Science, 2013) , no coherent data exist for European Caucasians. Furthermore, it is still unclear to what extent the two known C-related polymorphisms can predict C expression levels. The aim of this study was to shed light on both questions, by quantifying C surface expression in 400 healthy German blood donors, who were also genotyped for both, rs9264942 (rs C/T) and rs67384697 (rs ins/ del). C expression on buffy-coat derived lymphocytes of 400 Caucasian individuals was determined by flow cytometry as previously described (Apps et al., Science, 2013) . All subjects were also genotyped for (rs C/T) and (rs ins/del) by NGS. Median Intensity Fluorescence coefficients were calculated for each C allotype through implementation of a linear regression model. Accordingly, rs genotypes for each C antigen were imputed. Flow cytometry measurements revealed highly significant overlapping between our results and those of Apps et al. in African Americans. When C allotypes were divided into high-and low-expressed, there was only one discrepancy observed concerning C*15. As to the impact of the two polymorphisms on C surface expression, our data confirmed the association of T and ins genotypes with lower expression levels. However, with regard to rs ins/del, this appeared to result from the strong linkage disequilibrium (LD) of this genotype (p<0.0001) with the very common low expressed C*07 and C*03 allotypes, since the high expressed, yet less common, C*01 and C*14 were also found in strong LD (p<0.0001) with the ins allele. Rs C/T on the other hand appeared to better predict C expression levels. Our results on C expression levels in European Caucasians highly match those of African Americans previously reported by Apps et al. Additionally, our data imply that rs9264942 (C/T) is probably more reliable as C expression surrogate marker than the rs67384697 (ins/del). Undeniably, before definitive conclusions can be drawn, further research is required on additional Caucasian ethnicities. The olive baboon often serves as a model species in immune response-related studies such as in renal and xenotransplantation research, and therefore knowledge on MHC-KIR interactions is needed for this species. In humans, Bw4 and Bw6 are mutually exclusive epitopes mapping on the alpha 1 domain of class I molecules at position 77-83, which may act as a substrate for antibodies and diverse KIR allotypes. These epitopes are also documented in macaques where a specific KIR3DL molecule has been shown to recognize a broad range of MHC class I molecules with Bw6/Bw4-like but also non-Bw4 or -Bw6 motifs. Therefore, we searched for presence/absence of these motifs on MHC class I molecules of the olive baboon. As recorded in macaques, no identical Bw4 motif (NLRIALR) is located on Paan-A or -B molecules. Conserved Bw6 (SLRNLRG) motifs, however, are present on two Paan-A and two Paan-B allotypes, as well as all Paan-B*18 lineage members. In addition, the canonical Bw6 motif (NLRNLRG), which is present in Mamu-A1*002:01 and responsible for binding to KIR3DL05 is also present in several Paan-A allotypes, including those of the Paan-A*01 and A*14 lineages. The nucleotide sequences of most of them form a separate branch in the phylogenetic tree. The clustering of those alleles encoding the canonical Bw6 motif may be indicative for a specialized function of these molecules.

Nowadays, one of the most important issues for successful human tissue transplantation is to find compatible donors at multiple HLA loci. Besides the availability of HLA genotype data for millions of stem cell donors throughout the world, a deep knowledge on the occurrence of HLA haplotypes in different populations is also of great potential in this context, as haplotype compatibility may decrease the risk of graft versus host disease, among other putative undesirable consequences of transplantation. In addition, haplotype frequencies and linkage disequilibrium are most informative in population genetics to analyze the effects of demography and/or natural selection on HLA molecular evolution. After the publication of the first EFI catalogue of Common and Well-Documented (CWD) HLA alleles in 2017, a further objective of the EFI Population Genetics Working Group was to create a record of common HLA haplotypes. Currently, haplotypes are estimated from genotypes by using appropriate expectation-maximization algorithms and are most often well-determined by linkage disequilibrium significance tests. In addition, family segregation analyses are essential to confirm the presence of such haplotypes in different populations. We have thus started to analyze both large databases of multi-locus genotypes and available family data with the aim to constitute a repository of Confirmed or Well-Determined (CWD) HLA haplotypes in Europe. In this pilot study, we propose specific criteria to define CWD haplotypes and we present a preliminary map of multi-locus haplotypes across this continent. We also show the implementation of a new haplotype viewer in the hla-net.eu Gene[rate] tools that will be most useful in both transplantation and population genetics studies.

The Human Leukocyte Antigen (HLA) loci are highly polymorphic and serve a key role in immune response and histocompatibility. However, allele phasing information is vital for the most detailed analysis. Therefore, constructing a high-resolution haplotype database can provide a powerful tool for downstream analyses and greatly support the study of population genetics and disease association. Significant work has been done on haplotype frequencies but is often restricted to resolution at 2-3 fields by traditional methods. Additionally, previous work usually relies on imputation or makes use of exclusively homozygous cell lines. Lastly, to our knowledge, all 11 major HLA loci are yet to be haplotyped together, prompting a need to overcome these aspects all in one analysis. We demonstrate a method to build high resolution HLA haplotypes for all 11 major HLA loci, without relying on imputation. Approximately 1700 family trios are sequenced using NGS technology. Each individual sample is genotyped at a high resolution on 11 HLA loci using NGS analysis software, and manually reviewed for confirmation. The four haplotypes for each family trio are then generated by phasing analysis with Mendelian constraints. Ambiguities resulting from all members sharing the same heterozygous genotype are resolved by using coalescencebased imputation. Results at 4-field resolution show that, of the possible haplotypes resulting from the 1685 family trios, 86.0% can be successfully resolved into complete 11-loci haplotypes, increasing to 93.1% when ambiguities are resolved. Linkage disequilibrium is measured for each gene, and a phasing expectation-maximization algorithm is implemented for comparison of pedigree analysis and imputation.

The results show that a large high-resolution HLA haplotype database can be successfully constructed from family trios, serving a valuable resource for downstream analysis for population and disease association studies.

Thomas Goeury 1 , Jose M. Nunes 1 , Alicia Sanchez-Mazas 1 1 University of Geneva, Geneva, Switzerland Correspondence: thomas.goeury@unige.ch

Due to complex and fluctuating selective pressures acting on HLA genes, unraveling the molecular evolutionary history of this region remains a tough task. We present an original approach to this question using the identification of patterns of similarity between sequences based on their transition probabilities. For this purpose, we computed Markov chains for all known DNA sequences (taken from the IPD-IMGT/HLA Database) of the different gene regions (introns, exons and UTRs) of 11 HLA loci. The transition probabilities between these sequences were extensively analyzed to identify emerging patterns, including a t-distributed Stochastic Neighbor Embedding (t-SNE) and a Principal Component Analysis (PCA). The resulting plots for class I sequences reveal a greater similarity within each gene region than among them, meaning that any gene region (including introns) is more similar to the equivalent gene region on the other genes than to the other regions on the same gene. This suggests full-gene duplication events of the three HLA-A, -B and -C loci from a common origin, consistent with current hypotheses on HLA class I evolution. The plots obtained for class II genes display a distinct pattern: exon 2 sequences coding for the beta chains of the HLA molecules form an outlier group compared to all other sequences, which can be the result of selective pressures acting on the peptide-binding region of the HLA molecules. Another, more unexpected, finding is the great similarity between different introns, as well as between different exons, within and among the HLA-DR, -DP (and to a lesser extent DQ) genes, whereas introns and exons are very dissimilar from each other. This suggests a distinct origin of class II introns, on one side, and class II exons, on the other side. Finally, DPB1 exon 2 sequences also display two differentiated clusters corresponding to two non-overlapping 1st-field allele groups, suggesting a peculiar evolutionary history of the DPB1 locus. To determine peripheral blood and cord blood HLA-C antigen expression level and mRNA transcription level in the Chinese population, high-resolution HLA-C sequencing typing performed using PCR-SBT of peripheral blood. Total HLA-C mRNA transcription level were analyzed by qPCR by using locus-specific primers. Peripheral blood was analyzed by qPCR for HLA-C*01:02, C*07;02, C*06:02 and C*03:04 mRNA transcription level using allele-specific primers. HLA-C antigen expression level were analyzed by flow cytometry. The rs2395471 genotypes were determined by sequencing. Among 105 healthy volunteers, seventeen HLA-C alleles were detected. Mean levels of HLA-C antigen expression level in peripheral blood was 25886(MFI). There was significant difference in each HLA-C antigen expression level. HLA-C antigen expression level were higher for C*01:02, 14:02 and lower for C*03:02, 07:01, 07:02, 08:01 and 12:02. The mean level of HLA-C mRNA transcription in peripheral blood was 0.77. Only HLA-C*06:02 and 15:02 showed higher mRNA transcription level, with statistical significance (P<0.05). Among 150 cord blood samples, eighteen HLA-C alleles were detected. The mean level of HLA-C mRNA transcription level was 0.42. There was no significant difference for HLA-C mRNA transcription level in cord blood. HLA-C antigen expression level was highest for rs2395471_AA genotype, lowest for rs2395471_GG genotype and moderate for rs2395471_AG genotype in Chinese Han population. rs2395471_A was in complete linkage disequilibrium with HLA-C*01:02, 04:01, 06:02, 12:02, 14:02, 15:02 and rs2395471_G was in complete linkage disequilibrium with HLA-C*03:02, 03:03,03:04, 07:02, 08:01. Peripheral blood HLA-C antigen expression level and mRNA transcription level were associated with HLA-C allotypes in Chinese Han population. HLA-C alleles were in complete linkage disequilibrium with rs2395471 genotypes and can be divided into higher expression level and lower expression level groups. The HLA system shows a combination of extensive polymorphism and conserved haplotypes. Allelic and haplotypic frequencies of HLA genes vary greatly across human populations, and many alleles and haplotypes are region-specific. Populations in Latin America have a complex history of migration resulting in a broad variety of admixture proportions involving mainly European, African, and Native American components. We hypothesized that admixture proportions in Latin American populations can be approximated based on the putative continental origin of extended HLA haplotypes. We determined high-resolution HLA-A, -B, -C, and -DRB1 alleles in a sample of 947 healthy subjects from 4 Mestizo populations (Costa Ricans from [1] the Central Valley and [2] the province of Guanacaste, [3] Mexicans from Mexico City, and [4] Nicaraguans), plus Costa Ricans of Afro-Caribbean descent, and Amerindians from Costa Rica, and estimated extended haplotypes forming the diplotypes in each subject. We then assigned a putative continental origin based on the alleles and linkages present in each haplotype. Our results show a great degree of allelic and haplotypic diversity within and across these populations, with most extended haplotypes being private. Mestizo populations show haplotypes from the three ancestral continental regions, albeit with significantly different proportions (p<.001): a large European component in Costa Rican Mestizos from the Central Valley (66% of the haplotypes), a strong African component in Costa Rican Mestizos from Guanacaste (28% of the haplotypes), and larger Native American components among Nicaraguan (43% of the haplotypes) and Mexican (62% of the haplotypes) Mestizos. Amerindians and Afro-descendants from Costa Rica show high proportions of Native American (77%) and African (89%) haplotypes, respectively, albeit with evidence of gene flow from other continental groups. Multidimensional scaling analyses comparing HLA profiles of these populations and those of a large set of indigenous populations from Iberia, Sub-Saharan Africa, and the Americas reflect their differential admixture processes. Overall, our novel approach reaffirms the complexity of Latin American population genetics and confirms the usefulness of detailed analysis of HLA alleles and haplotypes to understand these processes.

Cigdem Kekik 1 , Sonay Temurhan 1 , Sebahat U. Akgul 1 , Yasar K. Caliskan 1 , Yasemin Ozluk 1 , Serra Ertan 1 , Erol Demir 1 , Figen Abatay 1 , Halil Yazici 1 , Aydin Turkmen 1 , Kilicarslan Isik 1 , Mehmet S Sever 1 , Fatma Oguz Savran 1 1 Istanbul Faculty of Medicine, Istanbul, Turkey Correspondence: citcim@gmail.com IgA nephropathy (IgAN) is the most frequent primary glomerular disease and is an important cause of end-stage renal disease (ESRD). In this study, we aimed to investigate whether SNPs (CFH gene rs6677604, HORMAD2 gene rs2412971, TNFSF13 gene rs3803800, DEFA gene rs2738048) and HLA (-DQA1, -DQB1, -DRB1) are risk factors for ESRD in IgAN patients. All patient and control DNA samples were isolated from peripheral blood with EDTA. HLA typing was performed by luminex (OneLambda HLA typing kits). SNP detection was performed by sequence base typing (SBT). One hundred and eighteen patients [M/F: 67/51, mean age 39 (16-75)] with clinical follow-up data and 18 controls [M/F: 8/10, mean age 37 (23-52)] were included in this study. Twenty eight patients (23.7%) reached ESRD [M/F: 17/11, mean age 36 (20-69)] during the follow period. We found that rs3803800 A allele was significantly higher in IgAN than controls (p=0.009 OR=3.09 CI=1.3-7.3). Compared with IgAN patients, rs6677664 A allele, DRB1*04:07 were significantly higher in ESRD patients (p=0.006 OR=7.08 CI=1.7-29.3 and p=0.001 OR=23.6 CI=1.2-464). The present study's results indicate that the SNPs and HLA alleles are associated with the susceptibility or predictive parameters in IgAN and reached ESRD in IgAN. We detected that the relationship between the presence of rs3803800 A allele and IgAN susceptibility may shed light on the etiology of the disease. We showed that rs6677664 A allele, DRB1*04:07 alleles may predispose to ESRD in Turkish nephropathy patients. Although genetic factors are thought to play a role in the pathogenesis of the disease, the etiology of IgAN is unclear. The genus Carduelis (Fringillidae family) includes goldfinches, siskins, redpolls, greenfinches and crossbills. Many of the species classified within this genus and other related genera have been grouped by using molecular systematics and the mitochondrial cytochrome b (mt cyt b) gene. According to this, the Eurasian siskin (C. spinus) is the only extant direct ancestor of several North American finches; North American/South American radiations may have been originated by Eurasian siskin (or an extinct relative). In the present work, we aim to perform a study of transspecies and transcontinental analyses of MHC (Major Histocompatibility Complex) class I alleles in several genus Carduelis/Spinus species in order to draw evolutionary conclusions in several wild bird species belonging to the genera Carduelis/ Spinus. We found two matches between MHC-I DNA alleles from different South American siskins at DNA level. Also, it was observed that the Eurasian siskin shares a protein with pine siskin and another with three South American siskins. Eight South American siskins species also share the same MHC protein. In addition, studied songbirds' MHC class I intron 2 is longer than that of Gallus gallus. We have drawn the following conclusions: 1) We present the first direct evidence that Minimal Essential MHC does not exist for birds; one of its main definition characters, i.e. small intron size does not hold for songbirds. 2) We also report that MHC genes transspecies evolution exists in birds by showing also for the first time that worldwide bird species keep the same MHC protein and DNA alleles. 3) New evidence on MHC alleles conservation from Eurasian Carduelis spinus (most ancient) to South American siskins (most recent) over a million years support that Eurasian siskin is the parental species for American genus Carduelis (Spinus) species. It is uncertain whether Eurasian siskin (or extant relative) had initially an Holoartic distribution, including America.

Antonio Arnaiz-Villena 1 , Ana Carballo 1 , Ignacio Juarez 1 , Ester Muñiz 1 , Cristina Campos 1 , Rawand Al-Qadi 2 , Jorge Nieto 1 , José Manuel Martín-Villa 1 , Jose Palacio-Gruber 1 1 University Complutense, The Madrid Regional Blood Center (Centro de Transfusion), Madrid, Spain, 2 Dohuk Specialized Laboratory Center, Dohuk, Iraq Correspondence: arnaizville@hotmail.com Atlantic Europe populations were analyzed with HLA genes in order to establish their relationship among themselves and with other populations. Standard genetic and statistical software analyses were used. Celtic populations (British Isles and French Bretons) have genetically been found close together: Irish, Welsh, Orkney Islanders (Scottish), French Bretons, Galicians, Spanish Basques, Portuguese, cluster together in DA genetic distances, correspondence analysis and Neighbor Joining dendrograms. Genetics has been shown by itself not suffice in determining populations' migration/relatedness. Aristotle and Herodotus placed Celts in Iberia and R1b chromosome Y marker is high in Iberia and all Celtic European populations above mentioned (probably stemming from Iberian Ice refugee after the Last Glaciation) and Ancient Celt language (Gaelic) is being translated from Iberian-Tartesian language: these suggest that Celts and Iberians, so named by classic authors, constitute the same population. On the other hand, a) R1b gene analysis of the Canary Islands' ancient inhabitants (Guanches), b) abundant Iberian scripts are also found in Canary Islands, c) an established North Africa/Iberia ancient gene flow, and d) no evidence of demic diffusion from eastern to western Mediterranean according to human ancient skeleton studies is noticed in Mesolithic/Neolithic transition. These facts suggest that ancient Canary Islanders may be included within the Iberian/Celtic population. Our conclusions are that: 1) Celts are concentrated in Atlantic Europe, 2) Iberians and Celts mentioned by classic authors most probably refer to the same population living in Iberian Peninsula (Spain/Portugal); in addition, North African Berbers and ancient Canary Islanders also belong to this group 3) Postulated farmers demic diffusion in an East to West Mediterranean direction never occurred.

Marie Fechtenbaum 1 , Judith Desoutter 1 , Vincent Goeb 1 , Nicolas Guillaume 1 1 Amiens University Medical Center, Amiens, France Correspondence: guillaume.nicolas@chu-amiens.fr Spondyloarthritis are inflammatory chronic diseases for which HLA-B*27 alleles explain a part of the genetics. However, other genes could participate in physiopathology of this group of diseases. The major histocompatibility complex class I polypeptide-related sequence A (MICA) encodes a glycoprotein which mediates the activation of the NKG2D receptor expressed on NK and CD8+ T-cells. Methionine vs valine at position 129 in exon 3 results in strong (MICA-129 met) or weak (MICA-129 val) binding to NKG2D receptor. The MICA A5 allele causes a premature stop codon, leading to a truncated protein synthesis. NKG2D polymorphisms result in alleles associated with low (NKC3 C/C and NKC4 C/C) or high (NKC3 G/G and NKC4 T/T) cytotoxic activity. The primary objective of this study was to determine whether polymorphisms at MICA and NKG2D loci are associated with spondyloarthritis (ankylosing spondyloarthritis and psoriatic arthritis). One hundred and sixty-two patients with spondyloarthritis were included. Frequencies of MICA-129 met/val, MICA A5.1 mutation, NKC-4 C/T and NKC-3 C/G polymorphisms were compared to those of 90 healthy controls. Each polymorphism was reported in each cohort, before and after adjusting on HLA-B*27 status. MICA-129 methionine (met/met and met/val) was significantly associated with spondyloarthritis compared with controls (OR=4. 84, CI 95 [2.75, 8 .67]) whereas MICA-129 val/val, MICA A5.1 and NKC3 C/C were protective (OR= 0.20, CI 95[0.11, 0.37], OR=0.15, CI 95[0.06, 0.36], OR=0.24, CI 95[0.13, 0.44] ). After adjustment with HLA-B*27, only NKC3 C/C remains protective for spondyloarthritis (OR=0.14, CI 95 [0.06,0.33]). On the other hand, by studying each cohort, homozygosity MICA A5.1 was protective for ankylosing spondyloarthritis and NKC3 C/C, MICA-129 val/val protective for psoriatic arthritis. In conclusion, MICA and NKC3 polymorphisms, related to a low cytotoxic activity of NKcells, were protective against arthritis disease. With deep sequencing, the tailing of overlapped reads provides phase information and thus HLA allele sequences. Automatic data analysis also provides allele level genotyping. Heterogeneous combinations of genes are present in Mexicans, depending on the region. Oaxaca City was founded in 1486 by Aztec warriors, and the population results from the admixture of Spaniards and 18 native local groups, mainly Mixteco and Zapoteco. An African gene pool prevails. We selected 104 BM donors from Oaxaca City. DNA samples were HLA typed by amplification by long range PCR with full gene coverage for class I and class II including all exons. All amplicons for each sample were pooled and tagged; 24 samples at a time were sequenced with a Miseq instrument. Analyses of sequences was done with Mia Fora software. Genotypes were assigned unambiguously. Allele frequencies ( Calabria is the southernmost region of the Italian peninsula and it is surrounded by Tyrrhenian Sea to the west, Ionian Sea to the east and it is separated from Sicily by the Strait of Messina to the south. Calabria suffered many invasions over centuries. Around 700BC Greeks occupied the territory that became Magna Grecia. Bruzi (an Indo-European population) founded Cosenza in IV BC. Rome conquered Calabria during the IIIBC. From 536AD, Byzantines dominated the region. The first Saracen attack occurred in 812. The Arab presence ended when Normans constituted the Duchy of Calabria. Angevins and Aragonese were the last invaders. In a correspondence analysis (first axis 65.8%, second axis 18.8%), we compared the frequencies of HLA-A, -B, -C alleles from Reggio Calabria (206), Cosenza (144), Catanzaro (205) and Vibo Valentia (108) provinces with 11 populations representing modern descendants of the invaders: Greece, Turkey, Albania, Tunisia, Naples, Sicily, France, Catalan, Basque, South Spain, and Sweden (Jordan as an outliner). All HLA data are available at allelefrequencies. net. Calabrians were adult stem cell donors. We observed a tight cluster composed by Calabrians and Greeks, Sicily, Naples, Albania, and Tunisia. We also observed two small clusters: the first composed by France and Sweden, the second composed by Catalan and Basque. Spain was between France and Turkey, Turkey was between Calabria and Spain. Calabrians show an immunogenetic fingerprint derived from few colonizers: Greeks (Magna Grecia, Sicily), Byzantines (Albania), Angevins (Naples) and Arabians (Tunisia). Unexpectedly, Calabrians are nearer to Tunisians than Turkish. The influence of Byzantines is marked by the proximity to Albanians, but not to Turkish. Saracens made incursions without settling in Calabria and inbreeding with local population, conversely Greeks founded colonies in Magna Graecia and Tunisia, that's maybe why Calabrians are close to Tunisians. Human leukocyte antigens (HLA) are considered to be the most important genetic risk factor for rheumatoid arthritis (RA), a chronic systemic inflammatory autoimmune disease of unknown etiology. HLA-DRB1 alleles encoding the shared epitope -a five amino acid sequence motif in residues 7074 of the HLA-DRβ chain (HLA-DRB1*01, *04, *10 and *14) are known to be associated with the onset and the severity of the disease. The frequency of HLA-DRB1 genotypes associated with RA varies within different geographical areas. The aim of this study was to evaluate distribution of HLA-DRB1 alleles in patients with RA in Sinj region (SR) and in the rest of Split-Dalmatia County (SDC), as well as to determine the association of HLA-DRB1 alleles with the severity and progression of RA in these patients. We analyzed 74 patients and 80 healthy subjects from SR as well as 74 patients with RA and 80 healthy subjects from the rest of the SDC. Low-resolution HLA-DRB1 genotyping was performed using PCR-SSO (Luminex platform), while high-resolution typing of HLA-DRB1*04 alleles was performed using PCR-SSP. In patients with RA from the SR frequency of HLA-DRB1*04 allele was significantly higher than in patients from the rest of the SDC (18.2% vs. 9.5%; P=0.014), while HLA-DRB1*15 allele frequency was significantly lower (7.4% vs. 16.2%; P=0.010). There was no difference in the frequency of HLA-DRB1*01, *10 and *14 alleles between patients (P=0. 119, P=0.325, P=0.278, respectively) . The frequency of acute phase reactants and rheumatoid factor was higher in shared epitope positive patients from SR compared with the shared epitope positive patients from the rest of the SDC (P=0.014, and P=0.004, respectively). Shared epitope positive patients from SR also showed higher disease activity (P=0.043), more severe degree of functional disability (P<0.001) and more severe radiological damages (P=0.003), compared with the SE-positive patients from the rest of the SDC. These findings indicate that higher incidence of more severe forms of RA in Sinj region could be explained with higher frequency of certain shared epitope HLA-DRB1 alleles (HLA-DRB1*04) in patients from SR compared to patients from the rest of SDC. The duration and success rate of an unrelated donor search in hematopoietic stem cell transplantation greatly depends on patients' HLA typing. Studies about HLA haplotype frequency in Croatian Bone Marrow Registry revealed 50 most frequent HLA-A~B~DRB1 haplotypes (>0.1%). The rest of them were defined as less frequent (<0.1%) or rare if they included at least one HLA allele observed three times at maximum in our population. The aim of the study was to analyze 362 patients transplanted with an unrelated donor from 2012 to 2017 in UHC Zagreb. All patients had family derived haplotypes. We investigated distribution of HLA haplotypes in order to evaluate the probability of finding a donor in the national or international registries. In the group of 73 patients (20.17%) who received a transplant from national donor (group 1), 92 different haplotypes were found while in the group of 289 patients (79.83%) who found an international donor (group 2), the diversity of haplotypes was much higher, 432 different haplotypes. HLA-A*01:01~B*08:01~DRB1*03:01 haplotype showed the highest frequency in both groups; however, its presence was significantly higher in group 1 (18.49% and 2.94%, respectively; P<0.0001). HLA genotypes consisting of two frequent haplotypes, as well as genotypes consisting of one frequent and one less frequent haplotype were significantly more present in group 1 than in group 2 (23.29% vs. 4.84%, P<0.0001 and 60.27% vs. 44.29%, P=0.0179, respectively). Genotypes with both less frequent haplotypes were significantly more observed in group 2 (47.06% vs. 15.09%, P<0.0001). The same states for the genotypes consisting of one less frequent and one rare haplotype, which we found in six patients with international donors and just one patient with a national donor. These data suggest that the presence of at least one frequent haplotype serves as a positive predictive factor for finding a donor in our registry while patients bearing less frequent or rare haplotype will mostly benefit from finding a donor worldwide. Identification of HLA allele frequencies is an essential element for the donor-recipient matching process in kidney and bone marrow transplantation. At the same time, the analysis of genetic heterogeneity within particular geographical areas is a very important tool for the characterization of population groups and constitutes a basic element for the anthropological studies. In our study, the HLA typing tests were performed from peripheral blood lymphocytes by serological methods (CDC -complement dependent cytotoxicity) and by molecular biology methods (SSP -sequence specific primers and SSO -sequence-specific oligonucleotides). The samples were collected from individuals enrolled on the waiting lists of the transplant centers from Chisinau and Iasi, as well as from volunteer donors enrolled in the National Registry of Hematopoietic Stem Cell Voluntary Donors from Romania. The statistical analysis of this data allowed us to calculate the relative frequencies and to compare these results with the data available for the Caucasian populations. Among the population of the Republic of Moldova the hierarchy of the most common alleles is as follows: HLA-A*02, A*03, A*24; B*18, B*35, B*44; DRB1*11, DRB1*13 and DRB1*16. The analysis of the Iasi group led to the following hierarchy: A*02, A*01, A*03; B*18, B*35, B*44; DRB1*11, DRB1*13, and DRB1*03. This study represents the first comparative HLA genetic profile of the populations from the Republic of Moldova and from the Romanian region of Moldova. The presented data should be considered as preliminary, requiring a significant extension of the subgroup from Chisinau, as well as the inclusion of the entire population of the Republic of Moldova. Human killer-cell immunoglobulin-like receptor (KIR) genes play an important role in the immune system and have been reported to affect hematopoietic stem cell transplantation (HSCT) outcome. KIR specificity and affinity is influenced by multiple facets of allelic polymorphism. Therefore, high resolution KIR genotyping might contribute towards improving donor selection and success of HSCT. As previously reported we established an approach based on next generation sequencing (NGS) to generate high resolution KIR genotyping data. After successful validation we implemented this workflow in October 2016. Since then we have genotyped more than one million stem cell donor registry samples for KIR at high resolution. Based on these experiences we have further optimized our neXtype analysis software. For quality control, we are monitoring precision using a set of 280 samples. However, unlike for HLA genotyping, external proficiency testing services are not yet available for comprehensive high resolution KIR genotyping. Therefore, we exchanged 366 samples that had been previously analyzed comprehensively for KIR based on NGS capture sequencing data and the PING pipeline. 96.7 % of the genotype calls on individual KIR gene level were in concordance with the PING data. After detailed reanalysis of the discordant results we identified 0.6% potentially spurious genotyping results in our data. Re-evaluation of the PING data is still in progress to agree on a common comprehensive genotype set. Given the early days of comprehensive high resolution KIR genotyping in conjunction with the limited coverage of the KIR diversity in the IPD-KIR database, comparisons with results obtained using independent approaches are essential to spot systematic biases or shortcomings. This study demonstrates the maturity of the current approaches and underscores that comprehensive high resolution KIR genotyping is possible in a high throughput setting. Interactions between hepatic stellate and NK cells are important determinants of liver fibrosis progression and regression. Chronic alcohol consumption produces alterations in the innate immune system and in the development of liver cirrhosis but biological mechanisms are unknown. NK and T cells express KIR-receptors, which interact with HLA class I ligands regulating their activity. Their association would be very useful in clinical practice to diagnose patients and even adjust possible treatments. The aim of this study was to determine the role of telomeric inhibitor KIR genes in alcohol cirrhosis. Genomic DNA was extracted from peripheral blood by automatic extraction. KIR genes were determined in 281 alcoholic cirrhosis patients and 319 controls using PCR-SSP method. Patients were grouped according to whether or not viral infections (n=68 and n=213). Our data show a similar frequency of KIR2DS1/S3 and KIR2DS4 between both groups. KIR2DS5 and KIR3DS1 appeared increased in both groups, but these differences were only significant for KIR2DS5 (P=0.033). An increase of KIR2DS5 frequency in non-viral patients with respect to controls was also observed (39.9% vs 27%; P=0.002). A decrease of KIR2DS5 was observed in viral (20.6%) with respect to non-viral patients (39.9%; P=0.004). Telomeric aKIR genes (KIR2DS1, KIR2S5 and KIR3DS1) were increased in patients, but only KIR2DS5 was significantly increased (P=0.002). Similar trends were observed in patients without viral infections with KIR2DS5+/KIR3DS1+ (P=0.002). Finally, the influence of the three centromeric KIR genes (KIR2DS1, KIR2DS5, KIR3DS1) was jointly analyzed, with an increase observed in patients with alcohol cirrhosis without associated viral infection. In conclusion, the genes KIR2DS1/S5 and KIR3DS1 appear to be more represented in the population of patients with alcoholic cirrhosis, being able to participate in the biological mechanisms of the cirrhotic process. Previous studies have suggested that the rat harbors a repertoire of MHC class Ib-reactive Ly49 activating receptors that can induce adaptive-like alloresponses. We now provide molecular and functional evidence for this. The Ly49 stimulatory receptor 4 (Ly49s4) and its inhibitory counterparts, Ly49i4 and Ly49i3, recognize two distinct MHC class Ib molecules, RT1-U and RT1-Cl, while the RT1-Eu molecule is recognized by the Ly49s5 and Ly49i5 pair of receptors. Although these ligands are categorized as non-classical, they share structural features with classical class I molecules, and display gain and loss haplotypic variation. The ligand-binding of several of the stimulatory Ly49 receptors appear to be stronger than that of their inhibitory counterparts. We observed a dramatic expansion of Ly49 expressing and highly alloreactive NK cell subsets in the peritoneal cavity of rats after repeated MHC class Ib alloimmunizations. This demonstrated a direct involvement of activating Ly49 receptors in the induction of potent NK cell alloresponses in vivo. KIR and HLA genotypes influence clinical outcome in relapsed patients treated with lenalidomide-based regimens, but poor is known about their impact on high-risk young (HRY) MM patients. The general aim of this study is to evaluate how KIR and HLA genotypes influence clinical outcome of HRY-MM patients. KIR genes and their HLA ligands were analyzed in 30 healthy subjects (controls) and 69 MM patients using the genotyping SSP and SSO kits. In 50 HRY-MM patients (median age 54.1 years, 30 males), treated up-front with novel agents-based induction treatment, followed by consolidation with one autologous stem cell transplantation and randomly assigned to lenalidomide maintenance until progression, we evaluated the role of KIR expression and KIR haplotype AA/Bx on progression free survival (PFS1) after first (PFS2) or second (PFS3) line of treatment. In HRY-MM patients, predictors of outcome were complete remission achievement after induction treatment, LDH and consolidation treatment (p<0.01), but not maintenance and monoclonal component at diagnosis. Lacking KIR 3DL1-edu was more frequent in MM than healthy subjects (76.7% vs 43.5%, p<0.0023; CI=95%). After median follow-up of 45 months, HRY-MM-patients carrying on 3DL1-edu had longer PFS than those without 3DL1edu (53.8 versus 41.5 months, p=0.08), although no significant differences and overall survival (OS). There weren't any significant differences in KIR haplotype AA/Bx frequencies in MM versus healthy patients. However, in HRY-MM-patients, Bx haplotype was associated with longer PFS (56.4 versus 27.8 months, p=0.06), without any effects on OS. In multivariate analysis, only complete remission achievement after induction and consolidation treatment were predictors of outcome, independently from KIR and haplotype genetics. KIR 3DL1-edu or Bx haplotype are associated to clinical outcome in HRY-MM-patients, implying a role of NK cytotoxicity against MM residual cells. Our data need to be confirmed in larger prospective series. Psoriasis is a multifactorial disease with a strong genetic susceptibility component. The HLA-C*06:02 allele has been found strongly associated in most populations worldwide, but functional data supporting a pathogenetic role for this allele in psoriasis are still lacking. In recent years, killer-cell immunoglobulin-like receptors (KIRs) expressed on the cell surface of natural killer (NK) cells have aroused considerable interest because of their binding affinity with specific HLA-C molecules. In the present study, we compared the distribution of HLA-C and KIR genotypes (KIR2DS1, KIR2DS2, KIR2DL1, KIR2DL2 and KIR2DL3) between 340 Sardinian psoriatic patients and 202 healthy controls. Statistical analysis of the comparisons performed between the two groups revealed a significantly higher frequency for KIR2DS1 in the group of psoriatic patients (51.2% vs 40.7%; P=0.02). Also in our study, HLA-C*06:02 had a higher frequency in patients compared to the controls (25.6% vs. 16.0%; P = 0.01). This prompted us to apply a gene-gene interaction test. The results showed that each locus was independently associated. Moreover, KIRD2S1 association was not influenced by clinical parameters such as onset age, psoriasis area and psoriasis severity index (PASI) or a family history of psoriasis. These results support the hypothesis that KIR2DS1 is an independent psoriasis susceptibility factor with a different mechanism of action than that of HLA-C*06:02. Relapsing-remitting multiple sclerosis (RRMS) is a chronic inflammatory disorder of the central nervous system which in Sardinia has a much higher incidence (5/105) and prevalence (1/103) than in mainland Italy. In the literature, the disease has been found associated with alleles coding for HLA class II molecules. In particular, it would seem that haplotypes containing the DRB1*04:05~DQA1*05:01~DQB1*03:01 alleles have a predisposing effect in the Sardinian population. In this study, we evaluated the role of natural killer cell immunoglobulin-like receptors (KIRs) and their HLA class I ligands in susceptibility to RRMS. A homogeneous group of 191 patients affected by RRMS was compared with a group of 215 healthy controls, all of certain Sardinian origin. Both groups were evaluated for the presence/absence of single KIR genes and their ligands as well as for the activating and/or inhibitory KIR genes present in the respective genotypes. In our cohort, RRMS was strongly associated with the HLA-B18, Cw5, DR3 haplotype (P < 0.0002). Analogous to the findings of Fusco et al., the KIR2DS1 activating KIR gene and the high affinity HLA-C2 ligands were observed with a significantly higher frequency in patients than in controls. [OR = 1.64, 95% CI =1.1 2.5, P =0.016]. Our study shows that KIRs and HLA class I molecules are involved in the onset of RRMS. However, the mechanisms underlying the activation or inhibition of natural killer cells remain to be clarified. Further research will be required to shed light on the precise role of KIRs and their HLA class I ligands within this context. KIRs modulate inhibition/activation of NK cells through the interaction with their HLA ligands. As an example, cytotoxicity and cytokine production of the polymorphic KIR3DL1 + NK cells depend on the allotype and/or HLA-A/B/Bw4 targets. Two lineages of 3DL1 allotypes and one of 3DS1 evolved under balancing selection and are present in all populations. Mestizos are the result of a tri-racial admixture comprised by Mediterranean genes brought by the Spaniards, an African component that came from Bantu slaves and native genes from the 62 different ethnic groups inhabiting Mexico. The aim of this study was to determine the 3DL1-S1 profile in Mestizos. We typed by SBT, KIR3DL1/ S1 that segregate as alleles from the same locus, as well as HLA class I alleles. Mexican Mestizo donors (n=100) from our Unrelated Mexican Registry DONORMO were chosen (X=35 years, 69% females). A KIR typing protocol was followed according to P. Norman. SBT was performed using a 3500 Sequencer. HLA typing was done with a Luminex SSOP method. KIR Allele typing was analyzed with the Assign SBT v4.7.1 software and allele frequency (AF) was obtained by direct counting. HLA analysis was done with Match it DNA v1.2.0 software. The presence of 3DL1 was shown in 95% of the subjects and 3DS1 in 40% (3DL1+/S1-=61%, 3DL1+/S1 += 34%, 3DL1-/S1+=5%). Eighteen different alleles were found, the most frequent ones being: 3DL1*01502= 22.2%, 3DS1*01301=21.0%, 3DL1*00101=14.2%, 3DL1*00501= 11.9%, 3DL1*00401=8.5%, 3DL1*002= 4.5%,3DL1*029 01=4.0%. 3DL1-Bw4 ligands were present in 60%, of which, 26% carried HLA-Bw4 alleles, 19% HLA-A/B/Bw4 alleles and 15% HLA-A/Bw4. Interestingly, a new allele was detected: KIR3DL1*00401, SNP 5C>T. Similar AF were observed for KIR3DL1*01501, *01502, *001, *01301 between Mexican Mestizos and USA Hispanics who differ for *008, *023, *00402, *009. Mexicans showed clear differences from Venezuelan Warao, Bari and Yucpa Indians. Mexicans vary as compared to Africans, Asians, European & Middle Eastern; Tanzania and Zimbabwe groups share similarities, since they are Bantu. The differences reflect that the diversity for the KIR3DL1/S1 NK repertoire is under natural selection, probably due to the exposure to distinct pathogens in the particular environment of populations and the immunological response evolved to infection in each ethnicity. NK cells of donor origin can recognize and kill residual leukemic cells and cure malignant patients in HSC transplant setting. NK cell function is regulated by killer cell immunoglobulin-like receptors (KIR) that recognize cognate locus-specific HLA class I molecules on target cells depending on specific amino acid residues at positions 77 and 80 or 77-83 motifs in HLA class I molecules. These HLA motifs are crucial for the prediction of donor NK cell capacity in malignant recipients. Due to HLA polymorphism, recombination and mutation events, the polymorphic HLA molecular interface is encountered by KIR molecules (some HLA-B molecules express C1 motifs, some HLA-A molecules express Bw4 motifs, C1 and Bw6 77-83 motifs are identical, but only C1 ligands bind KIR, etc.). To figure out the role of HLA motifs in binding KIRs we analyzed HLA: KIR crystallographic data and compared physically involved amino acid sequences in 133 common HLA class I molecules. In crystallographic data five short linear sequence motifs (14-19, 66-76, 77-84, 88-92, 142-151) were described that physically bound KIRs (whole HLA: KIR interface). In 133 common HLA class I molecules we found 54 motif combinations including: A3/11-3 combinations in 4 molecules (3/4); Bw4(80-I)-10/22; Bw4(80-T)-4/10; C1-6/19; C2-8/15; Bw6-10/38 and A non-KIR-ligating(A noL)-13/25. Exceptions included: 1) all motifs in B*46:01 matched to C1; 2) B*73:01 differed from C1 only at positions 66-77; 3) several A noL molecules (A30, A68, A69, A34, A66) were identical to A3/11 at motifs 77-84 and 66-76 but differed at 142-151; 4) C1 and Bw6 were identical at 77-84 but differed at 66-76 motifs; 5) Bw4 positive HLA-A molecules have isoleucine instead of threonine at position 80, a substitution known to increase Bw4: KIR3DL1 affinity; etc. These regularities and exceptions can be used for predicting KIR ligation among less frequent HLA molecules based on their amino acid motifs of whole HLA:KIR interfaces. The prediction might improve selecting optimal HSCT donors for malignant patients. The success of cancer therapy depends not only on adequate clinical procedures that aim to eliminate tumor cells, but also on the functional state of the patient immune system. NK cell cytotoxicity is regulated by the interaction between killer immunoglobulin-like receptors (KIRs) and human leukocyte antigen (HLA) class I ligands on target cells and the different binding affinity of the Fcγ receptor IIIA (CD16A) for IgGcoated tumor cells. Thus, it is conceivable that KIR and CD16A gene contents may contribute to the function of NK cells by modulating an immune response in breast cancer (BC). BC progression and metastases have been linked to anti-tumor immune response inefficacy. Therefore, we have investigated the combined effect of the KIR genes and their HLA-C ligands together with CD16A on the susceptibility and progression to development of BC in Italian patients and matched controls (Ctrs). The gene polymorphisms were investigated by PCR-SSP methods for typing of KIR receptors and by PCR-SBT for HLA-C and FCGR3A alleles. Our results typed in 47 BC pts and 39 Italian female Ctrs showed a predisposing effect of KIR2DS1 to more aggressive neoplasms (stage III-IV), both considering the gene alone (III-IV stage 63.2% vs 0-II 32.1%, p=0.043 OR=3.619) and in the KIR2DS1/HLA-C2+ combination (42.9% vs 6.3%, p=0.030). On the contrary, the KIR2DL2 and KIR2DS4ins genes, which are linked on the same haplotype, seemed to be protective for a more aggressive neoplasm (36.8% vs 60.7%, p=0.04; 10.5% vs 35.7% vs, p=0.06, respectively). The synergic absence of these two KIR genes increased the occurrence of more advanced BC by 5.7-fold (63.2% vs 23.1% in Ctrs P=0.0071). Regarding FCGR3A48 A/T/G and 158 G/T gene polymorphisms, we described a higher frequency of 48T-158G and 48T-158T haplotypes in both BC and Ctrs groups, even though the 48TT-158TT (LL/FF) genotype was present mainly in controls (3.6% vs 11.0%). Our findings evidence a potential role of KIR-gene content together with their HLA-C ligands in BC tumoral progression, suggesting that the development of breast cancer can result from disorder of immune regulation. In this study we analyzed the KIR haplotype(s) carrying KIR2DS2*005, an allele originated by unequal crossing over between KIR2DS2 and KIR2DS3 that codes for a surface receptor sharing the ectodomains with KIR2DS2 and the transmembrane and cytoplasmic regions with KIR2DS3. To this aim, using an SSP-PCR approach, we firstly determined the KIR repertoire in a cohort of 607 individuals. Then, we investigated the KIR2DS2*005 presence in the donors characterized by Cen-B regions lacking KIR2DL2 and/or KIR2DL5B (i.e. the genes located between KIR2DS2 and KIR2DS3) as well as in the donors having two Cen-B regions (since the presence of a true Cen-B1 region in one haplotype might mask the presence of a contracted KIR2DS2*005pos Cen-B region in the other one). To detect KIR2DS2*005 we used a set of primers that specifically amplified the KIR2DS2/2DS3 hybrid gene. Thereafter, sequencing the KIR2DS2*005 specific PCR products, we demonstrated that the junction point between KIR2DS2 and KIR2DS3 was identical in all the KIR2DS2*005pos donors identified. Finally, as the majority of the individuals of the cohort were parents of children (not included in the cohort), we performed KIR haplotype evaluation combining family segregation with well-defined KIR haplotype features. Our analysis clearly indicated that KIR2DS2*005 was present in two different haplotypes sharing the same centromeric region but differing in the telomeric region. In particular, we found 5 individuals with the KIR2DS2*005pos haplotype characterized by Tel-B motif carrying KIR2DS5 (as previously described) but also 4 donors with the KIR2DS2*005pos haplotype characterized by Tel-A region. Given the crucial role of KIR variability in modulating NK cell education and consecutively their response during viral infection, pregnancy and allogeneic transplantation, we think that the KIR2DS2*005 detection, improving the accuracy in KIR repertoire analysis, might be routinely included in KIR genotype studies. Compared to other mammalian species, pigs are thought to possess relatively few natural killer (NK) cell receptor genes. These include a minimal set of killer lectin-like receptors (KLR), a single killer immunoglobulin (Ig)-like receptor (KIR) of questionable functionality, and only three described leukocyte Ig-like receptors (LILR). Both the KIR and the LILR are encoded in the leukocyte receptor complex (LRC), which also contains several other Ig-like receptors, many of which are involved in innate immunity. These characterizations were based on early genome assemblies, including the previous pig reference genome, Sscrofa10.2, assembled using short-read sequencing technology which is well known to introduce errors over immune gene complexes. Recently, the updated assembly Sscrofa11.1 was released based on long-read sequencing technology. Our analyses found that the LRC in this new assembly is almost contiguous with only a single gap, compared to 14 gaps in Sscrofa10.2 and contains a total of 17 LILR genes and gene fragments. Most of these new genes are missing in the previous assembly. Elsewhere in the LRC, we identified a novel gene possessing a long intracellular tail and two Ig-like domains. We also identified a second, highly similar pseudogene with a short intracellular tail in some haplotypes. Phylogenetically, these novel genes are related to others in the LRC yet form a distinct clade. The novel inhibitory gene and variable numbers of short-tailed genes were identified in the genome assemblies of a diverse range of species, including a pseudogene in humans. Transcriptome analysis revealed that the novel inhibitory gene and six functional LILR genes are highly expressed in porcine peripheral blood. In contrast, the single KIR gene, KIR2DL1, is barely expressed, confirming its lack of functionality. This discovery of several new leukocyte receptors in pigs will help inform future investigations into the function of porcine immune cells. The aim of this study was to analyze the influence of KIR2DL2 and KIR2DL3 in alcoholic cirrhosis patients by assessing the coexistence of viral infection. Genomic DNA was extracted from peripheral blood by automatic extraction. KIR genes were determinate in 281 alcoholic cirrhosis patients and 319 controls using a PCR-SSP method. Patients were grouped according to whether or not they had viral infections (n=68 and n=213, respectively) and classified according to zygosity for KIR2DL2/L3. Our results showed no significant differences in the frequency of KIR genes in controls and patients, except for the KIR2DL2 gene, whose frequency was significantly lower in patients than controls ( The cause and pathogenesis of autoimmunity diseases are influenced by environmental and genetic risk factors. The HLA system has been shown to play an important role in susceptibility to various diseases, particularly those with an immune or autoimmune base. Rheumatic diseases are very prevalent among the population, estimating that, in Spain, they affect 1 in 4 people over 20 years of age. In particular, in psoriasis, HLA-C*06 has been shown to be associated with an increased risk and lower age of onset of the disease. The aim of this work was to evaluate the frequency HLA-C in patients with various rheumatic diseases. 104 patients and 250 healthy people were typed for HLA-C. The DNAbased HLA-C typing was performed by PCR-SSOP Luminex method using Lifecodes HLA typing kits (Gen-Probe) in accordance with the manufacturer's instructions using Luminex. We compared this group of patients with the group of healthy people. The significance values (p) for all results were obtained with the use of Chi-Square test and confidence intervals at 95%. The comparisons between healthy people and patients with rheumatic diseases showed significant differences for HLA-C*15 (p = 0.016) and HLA-C*16 (p = 0.006). HLA-C*15 was detected in 5.2% of healthy people and 12.5% in patients. HLA-C*16 was detected in 18.0% of healthy people and 6.7% in patients. When patients were stratified by epitopes C1 and C2 we didn't find association. In summary, our study suggests that HLA-C*15 could be a risk factor for rheumatic disease and HLA-C*16 a protective factor. However, more studies will be needed in the future to confirm these results. This study aims to evaluate which of the psoriasis related alleles are more frequent in Calabrian population and if affected patients have other alleles not usually related to this pathology. Our additional purpose is to understand whether HLA typing is helpful to clinicians in order to choose the more effective treatment. This study -still in progress -has been carried out on 30 Calabrian psoriasis patients aged from 25 to 80 years; for each one of them we have collected clinical information, particularly focused on clinical phenotype and signs and best treatment outcome. 728 unrelated Calabrian individuals, aged from 18 to 50 years, enrolled as volunteer potential bone marrow donors in the Calabrian Registry for IBMDR (Italian Bone Marrow Donors Registry), have been selected to estimate HLA-C locus frequency in the population as reliable control samples in distribution of allelic variability of HLA alleles. The HLA typing of the C locus has been performed using Luminex SSO technology for low-medium definition and PCR-SSP method for high resolution. Allele frequencies have been obtained by direct count. Results show that the Calabrian population is not homogeneous for genetic susceptibility: C*06:02, closely related to worse symptoms, has a 21% frequency and C*07:01 a 40% frequency. The study of HLA from a translational point of view could be in the future a valid tool to identify patients with psoriasis at high risk of developing arthritis and undertake the earliest and most effective individual therapeutic strategies, reducing disability and improving patients' life quality. Tumor necrosis factor (TNF) alpha is a major proinflammatory cytokine involved in the immune response in IBD. Moreover, the use of biologic agents, particularly anti-TNF-alpha, has revolutionized the treatment of IBD, including CD. The aim of this study was to investigate the influence of two single nucleotide polymorphisms (SNPs) of TNF-alpha gene promoter on the risk and phenotype of Crohns disease in a Romanian population. We genotyped 57 CD patients and 160 healthy controls for TNF-alpha rs1800629 (-308G/A) and rs361525 (-238G/A) using Taq-Man Allelic Discrimination Assays (Real-time PCR System, Applied Biosystems by Thermo Fisher Scientific, USA). Association tests for each polymorphism were performed with the software package PLINK v 1.07. Controls and patients were in Hardy-Weinberg equilibrium for the investigated SNPs. The frequency of the minor allele TNF-238A was significantly higher in CD patients (7%) than in controls (2.2%): p=0.03, OR 3.375. The carriers of this allele (GA+AA genotypes) were more frequent among patients (14% vs. 4.5%, p=0.01, OR 3.569) . The minor allele TNF-308A was overrepresented in patients with EIM (27% vs. 5.2%, p=0.006, OR 6.632). Our results show that TNF-alpha gene polymorphisms influence the risk and the phenotype of Crohns disease in Romanian population. Psoriasis is a chronic, inflammatory, relapsing-remitting disease with a multifactorial etiology, spacing from the genetic predisposition to several triggering environmental factors. The class I HLA complex is known to be the chief susceptibility factor and in particular certain determinants are associated with an early onset and a more severe grade of the disease. Moreover, several observations suggest that susceptibility to psoriasis could be determined by the imbalance between the signaling of activating and inhibitory KIR receptors, whose HLA antigens are considered the natural ligands. To investigate the association between HLA class I KIR ligands and predisposition to psoriasis we used 103 psoriatic patients, genotyped by PCR-SSO a technique, and 227 healthy controls. Although HLA-C*06 was significantly higher in patients (15.05% vs. 7.05%, p=0.001), there was no difference in the frequency of HLA-C1 and HLA-C2 epitopes between the two cohorts. HLA alleles with a Bw4 motif were increased in individuals with Psoriasis compared to controls (48% vs 40.3% p=0.062). This data was reproduced when we examined HLA-Bw4 80I/Bw4 80T as KIR ligands, albeit with no statistical difference (67% vs 61% p=0.28). In particular HLA-Bw4 80I was present in 54.37% of patients with respect to 45.37% of controls (p=0.12). Interestingly psoriatic patients showed a significant decrease of HLA-A alleles (A3 and A11), ligands of the inhibitory KIR3DL2 (27.18% vs. 37.88%, p=0.058 OR=0.61). The inhibitory KIR3DL1 binds HLA-Bw4 motif, and the molecules with Bw4 80I may be better ligands. Considering that KIRD3DS1 may also recognize some Bw4 allotypes and that, in psoriasis, NK cells mostly carry the inhibitory KIR3DL1 receptors of low expression type, it is evident that the disease is induced and sustained by an intrinsic inflammatory state. Celiac Disease is an intolerance situation against gluten in genetically susceptible individuals. Chronic inflammation is observed in intestinal mucosa in these patients. Consequently, villous atrophy, hyperplasia, and lymphocyte infiltration occur during the course of the disease. Celiac disease is derived from variations in a number of genes and environmental factors. In other studies, it was observed that the disease was associated with especially HLA-DQA1, DQB1, and DRB1 alleles. In this study, we aimed to compare HLA-DQA1, DQB1, and DRB1 alleles of Turkish celiac patients and healthy individuals. In addition, the relation between these alleles and Celiac Disease was examined. In total, 97 celiac patients and 148 healthy individuals were included. HLA-DQA1, DQB1, and DRB1 typing were performed by low-resolution sequence specific oligonucleotide probe (SSOP) methods. The results were analyzed by Pearson chi-square (Statistical Package for Social Sciences for Windows Version 22.0) test. DQA1*02 and DQA1*05 alleles were significantly higher and DQA1*01 allele was significantly lower in patients than control group (p<0.05). According to relative risk analysis, the individuals with DQA1*02 (1.70) and DQA1*05 (1.72) alleles have high risk of disease development. The DQB1*02 allele was significantly higher in patients than healthy individuals (p<0.05). According to relative risk analysis, the individuals with DQB1*02 (3.44) allele have high risk of disease development. DRB1*03 was found significantly higher in patients than the control group (p<0.05). According to relative risk analysis, the individuals with DRB1*03 (5.50), DRB1*07 (1.69), and DRB1*01 (1.10) alleles have high risk of disease development, respectively. In conclusion, there was a statistically significant association between celiac disease and DQA1*02, DQA1*05, DQA1*01, DQB1*02, DRB1*03, DRB1*07, and DRB1*01 alleles. The data were compatible with other literature. TINU is an autoimmune disease characterized by the simultaneous presence of tubule-interstitial nephritis and uveitis. It is a rare disease that may be responsible for kidney failure. It affects young people aged 11 to 15 years. Mostly observed in females with ratio 3:1 compared to males. It seems linked to a defect in cell-mediated immunity and to delayed hypersensitivity phenomena with the appearance of a T-mediated inflammatory lymphocyte state; such mechanisms would seem to involve alterations of the C-reactive protein (mCRP), autoantigen common to both uvea and renal tubular cells. TINU seems strongly associated with some HLA subtypes such as DQA1*01, DQB1*05, and DQB1*01. No specific risk factors were identified. In addition to interstitial nephritis and uveitis, patients with TINU may also have systemic symptoms such as fever, weight loss, asthenia, malaise, anorexia, abdominal and lumbar pain as well as skin rashes. We report a case of an 11 year old child arriving in Pediatric E.R. for fever and uveitis. The general conditions appear discrete. There is asthenia, skin pallor and a body temperature of 37.5 C. Laboratory tests showed anemia, high levels of VES and PCR. Ocular examination was required, confirming uveitis. Multistick urine tests detected proteinuria andglycosuria requiring urgent hospitalization. During admission the general conditions were stable, with some feverish manifestations. Laboratory tests were performed to exclude viral infection (TORCH), with negative results; 24-hour urine exams (twice during the stay) and multistick test every morning. A Normal abdominal echo was observed. A pharyngeal swab showed positivity to SBEGA and antibiotic therapy was initiated. Persisting normoglycemic glycosuria, associated with bilateral uveitis, with increased microalbuminuria and urinary globulins led us to suspect TINU. Typing of HLA DQA1*01, DQB1*05 and DQB1*01 confirmed the suspected diagnostic. Behcets Disease was discovered by Turkish Dermatology Professor Hulusi Behcet in 1937. The disease had been described by three symptoms with recurrent oral aphthae, genital wounds, and uveitis. It was later revealed that it shows a systemic course with joint, vascular, intestinal, pulmonary, and neurological involvement. In this study, we aimed to investigate HLA-B*51 sub-alleles with high resolution, to identify the most frequently observed alleles in patients and to evaluate the association between HLA-B*51 sub-alleles and disease pathogenesis in patients who applied to Izmir Tepecik Education and Research Hospital Tissue Typing Laboratory. For this purpose, sub-allele typing of 24 HLA-B*51 positive patients were performed by sequence-specific primer polymerase chain reaction (PCR-SSP). Seventy-three healthy and HLA-B*51 positive bone marrow donors were used as a control group. Data of patients and healthy individuals were compared. The results were evaluated by Pearson chi-square analysis. Association of HLA-B*51:01 allele frequency between patient and control group was not statistically significant (p<0.05; p=0.457), whereas it was found to be significant for HLA-B*51:08 (p<0.05; p=0.003). When the association between 51:01/51:08 sub-alleles and major/minor organ involvement, it was revealed that, these two sub-alleles were not significantly different for organ involvement. The aim of the study was to analyze the correlation between cytogenetic profiles and some genotypes in 24 Caucasoid patients with MM from the North-West region of Russia. Genotyping IL-4, TGF-b1, IL-1a, IL-1b was performed by PCR-SSP; study of cytogenetic abnormalities was performed by standard GTG-method and interphase FISH analyses: LSI 13(RB1)13q14, IGH/CCND1, IGH/FGFR3, LSITP53 (17q13.1); p-values less than 0.05 were considered statistically significant. Previous studies on cytokine polymorphisms allow us to describe some genotypes as negative prognostic markers associated with the development of MM (IL-1a -889 TT, IL-1b +3962 TT, IL-6 -174 GG, IL-6 nt565 GG; 1st gr.) and IL-4 -33 CC and TGF-b1 codon 25 GG as positive prognostic markers -2nd gr. In some patients, the presence of negative and positive markers together (3rd gr.) was registered. Also, in the studied patients normal and abnormal cytogenetic profiles were observed. The frequency of abnormal cytogenetic transformations in the 2nd gr. was noticeably lower compared to the 1st and 3rd gr. (0.11 vs 0.78 and 0.67). Differences in the frequency between patients with positive prognostic markers and normal cytogenetic profile (0.89) compare to patients with negative (0.22) or mixed (0.33) genotypes and normal cytogenetic profiles were registered (p<0.05). In the 1st gr. the frequency of cytogenetic abnormalities was remarkably higher compared to patients with the normal profile (0.78 vs 0.22). In patients from the 2nd gr. the frequency of normal cytogenetic profiles was noticeably higher than in patients with aberrations (0.89 vs 0.11). Our results allow to describe IL-1a -889 TT, IL-1b +3962 TT, IL-6 -174 GG, IL-6 nt565 GG as markers associated with the presence of cytogenetic abnormalities; IL-4 -33 CC and TGF-b1 codon 25 GG indicate the normal cytogenetic profile in patients with MM. Although, if patients have both negative and positive prognostic markers it seems that the chance of finding cytogenetic abnormalities is much higher compared to patients with positive prognostic markers only. Reports of association of genetic variants of interleukin-6 (IL6) and its receptor (IL6R) with psychiatric disorders are inconsistent, and there are few population-based studies thus far in bipolar disorder (BD). We genotyped the IL6 rs1800795 and IL6R rs2228145 polymorphisms in two independent samples exposed to different environmental stimuli -French and south Indian Tamil BD patients with quantitation of circulating plasma IL-6 levels in the latter sub-sample. In both populations, allele and genotype frequencies did not differ significantly between cases and controls for either polymorphism. Upon stratifying based on age at onset, we found no associations with the IL6 rs1800795 variant. However, the IL6R rs2228145 C allele (P=0.01, OR=1.73, 95% CI= 1.14-2.63) and CC genotype (P=0.01, OR=1.73, 95% CI= 1.14 -2.63) were associated with early onset of disease in the French sample when compared to late-onset BD. A similar trend was observed in the Indian population where we also found that plasma IL-6 levels were significantly higher in BD and also in patients who were in residual phase or remission both as compared to controls (P=0.007 and P=0.005 respectively). Our findings are in favor of a possible trans-ethnic implication of the IL6R genetic diversity in BD and reinforce the notion that IL-6 is an important marker of the operating inflammatory processes in the disease. A linked between human parvovirus B19 (hPVB19) infection and development of some chronic diseases including inflammation, autoimmune diseases and cancer is being investigated. The Non-Structural Protein (NS1) of hPVB19 is considered as a double-edged sword in its pathogenesis. We investigated the possible role of hPVB19 NS1 in the modulation of inflammatory response which is a trigger for the development of cancers. The HEK-293T cell line was used as a common epithelial cell in the biological research. A plasmid containing a fully sequenced NS1 gene named pCMV6-AC-GFP-NS1 was transfected into HEK-293T cells using Lipofectamine 3000. To evaluate the transfection efficiency, expression of Green Fluorescent Protein (GFP) was evaluated by fluorescent microscopy during 24, 48, and 72 hours post-transfection. Mock (pCMV6-AC-GFP) transfected cells were used as a control group. The percentage of apoptotic cells were measured by flow cytometry. Furthermore, cellular supernatants were collected at 24, 48 and 72 hours post-transfection to determine the type and the quantity of cytokines produced by treated and untreated HEK-293T cells using flow cytometry. The commercially available LEGENDplex Human Thelper Cytokine Panel was used. The percentage of apoptosis had no significant changes in treated cells with mock and pCMV6-AC-GFP-NS1 in 24, 48, and 72 hours posttransfection. The concentration of released cytokines such as Interleukin (IL)-2, IL-6, IL-9, IL-10, IL-13, IL-17A, IL-21, IL-22, IFN-gamma, and TNF-alpha has increased in 24, 48, and 72 hours post-transfection. Overall, our results suggest that pro-inflammatory cytokine levels are increased following expression of NS1 in HEK-293T. They are involved in the up-regulation of inflammatory reactions. Therefore, hPVB19 NS1 function may play a role in progression of some chronic and inflammatory diseases as well as cancers. Schizophrenia is a multifactorial chronic disease characterized by heterogeneity of psychiatric manifestations. Its aetiology is still not clear, although several recent studies agree on relating schizophrenia to a low-grade chronic inflammatory disease, determined by infections and nonpsychiatric autoimmune disorders such as systemic lupus erythematosus, psoriasis and celiac disease. Moreover, the use of specific anti-psychotic drugs reduces inflammation and patients affected with autoimmune diseases undergoing biological treatment show a decrease in psychiatric symptoms. Several GWAS founded genetic variants nearby and inside the MHC region and confirmed the pivotal implication of HLA as a key regulator in the immune response; other HLA alleles have been associated both with better responses and adverse drug reactions, while other genes are likely to be linked to schizophrenia. Our study, still in progress, aims to evaluate if and which autoimmune diseases and/or ADR related HLA alleles our schizophrenic patients have, and if other alleles/haplotypes have a particular frequency. To date, 15 patients with different schizophrenic profiles have been typed for HLA-A, -B, -C, -DRB1 loci using low-medium resolution SSO-luminex technology. Results obtained have highlighted the presence of at least one autoimmune disease related allele in all of the patients. Also, ADR related alleles were observed. One autoimmune haplotype was found and another allele frequency registered. It has not been possible to obtain specific data on eventual autoimmune comorbidity, an extended research to autoantibodies and specific disease markers, attention to individual drugs reaction might complete the results. Despite the low number of patients typed, the data are interesting and if confirmed by a larger individuals sample, HLA typing in schizophrenic patients in the near future could become a valid clinic tool for a more accurate and personalized therapeutic route.

GENETIC VARIABILITY IN TLR10 GENE MAY  MODULATE THE SUSCEPTIBILITY TO  RHEUMATOID ARTHRITIS AND RESPONSE TO  TREATMENT WITH ITNF- Response to treatment with iTNF-alpha was more effective after 24 weeks as compared to 12 weeks (p=0.001), especially in female patients (p=0.05). Women carrying the A allele responded better to treatment after 24 weeks of observation as compared to those with the CC genotype (p=0.053). Response to iTNF-alpha was more effective in patients with low stage of disease (p=0.01), with rheumatoid factor (RF) positivity (p=0.01) and with double positivity against citrulinated (CCP) protein and RF (p=0.003). RF-positive patients (especially women, p=0.001) characterized with a higher degree of the disease as compared to RF-negative cases (p=0.01). Men had a higher activity of the disease before iTNF-alpha treatment (p=0.05), therefore the remission of the disease was more common in women (p=0.04). TLR10 polymorphism has an important role in RA and may potentially influence risk of the disease, iTNF-alpha therapy effectiveness and patient outcomes.

Francisco Perea 1 , Abel Sánchez-Palencia 1 , Mercedes Gómez-Morales 2 , Míguela Méndez García 1 , Angel Concha 3 , Amanda Rocío González-Ramírez 4 , Teresa Cabrera 5 , Federico Garrido 1 , Francisco Ruiz-Cabello 1 , Natalia Aptsiauri 5 1 Hospital Universitario Virgen de las Nieves, Granada, Spain, 2 Universidad de Granada,, Granada, Spain, 3 Complejo Hospitalario Universitario, La Coruña, Spain, 4 Fundación de Investigación Biosanitaria Alejandro Otero, Granada, Spain, 5 Universidad de Granada, Granada, Spain Correspondence: naptsiauri@ugr.es

We analyzed 59 samples of non-small cell lung cancer (NSCLC) for simultaneous expression of PD-L1 and HLA class I (HLA-I) molecules. The objective was to demonstrate the impact of two independent mechanisms of cancer immune escape, such as the loss of tumor HLA-I and the upregulation of PD-L1. Although there was no direct association between tumor stage/grade and HLA-I or PD-L1 expression analyzed separately, we obtained statistically significant results when we combined HLA-I and PD-L1 expression and classified tumors into four groups: 1) HLA-I +/PD-L1+, 2) HLA-I+/PD-L1-, 3) HLA-I-/PDL1+, and 4) HLA-I-/PD-L1-. The expression of PD-L1 appeared to be an important factor associated with bigger tumor size in HLA-I negative tumors. The majority of HLA-I-/PD-L1+ tumors had more advanced stages T3+T4 in comparison with HLA-I-/PD-L1-(p=0.006) and HLA-I+/PD-L1tumors. Thus, HLA-I-/PD-L1+ tumors have more aggressive immune escape phenotype. Interestingly, HLA-I positive tumors lacking PD-L1 expression (immune rejection phenotype) showed less advanced stages as compared to HLA-I+/PD-L1+ tumors. Notably, we also found an association between the expression of HLA-I and PD-L1 and lymphatic spread (N): a high proportion of HLA-I-/PD-L1+ tumors (40%) had involvement of one or more lymph nodes (N1 + N2), as compared to only 6% detected among HLA-I +/PD-L1-tumors (p=0.028). In addition, HLA-I-/PD-L1+ tumors had the lowest grade of CD8+T-cells infiltration and a "non-permissive" structure with tumor encapsulated by a stroma. Interestingly, patients with HLA-I+/PD-L1-tumors had a tendency to better survival than patients with tumors positive for both markers. In conclusion, we have demonstrated that lack of HLA-I and positive expression of PD-L1 represents an aggressive tumor phenotype characterized by greater tumor size; more advanced stage with lymph node involvement, and reduced tumor immune infiltration with a "non-permissive" tumor tissue organization Many studies show that there is a genetic component to psoriasis. Patients with a first-degree relative with psoriasis have a five-fold increased risk of developing the disease and many children with psoriasis have a first-degree relative to the disease. In particular it is associated with HLA-C*06. On the basis of these studies we conducted research aimed at identifying the HLA alleles involved in our region. For this purpose, we have focused our attention on patients with plaque psoriasis. The study was conducted on a total number of 949 patients, divided as follows: 326 in 2015, 323 in 2016 and 300 in 2017. The most common alleles in these control patients were the C*04 and C*12 alleles, while we observed the association between the C*07 allele and psoriasis. In 2015 60 patients (44 M and 16 F) affected by plaque psoriasis were subjected to HLA-C typing, of which 25 presented the C*06, 25 C*07 and 10 C*04 alleles. In 2016, we typed 38 patients with plaque psoriasis (21 M and 17 F), of these 11 presented C*06 and 27 C*07. In 2017, 20 people with psoriasis (11 M and 9 F) were typed: 15 positive results for C*06 and 5 for C*07. Although the association between psoriasis and the C*06 allele is widely reported in previous studies, the association with the C*07 allele also appears to be important in our study. It therefore seems appropriate not to restrict the immunogenetic study to the C*06 allele alone but to extend it to the whole C locus. Analyzing the sequential nucleotides of the HLA-C alleles associated with psoriasis showed that the C*06 and C*07 alleles share some nucleotide sequences. In particular in position 73, the presence of alanine in place of threonine is found in both the alloantigens. It has been hypothesized that alanine may play an important role in the association between HLA and psoriasis being located in the helix domain binding site for peptides, involved in the presentation of the antigen to lymphocytes. In order for lymphocyte activation and the production of the histological cytokines of psoriasis to take place, a promoter must be presented to the CD4 + or CD8 + lymphocytes in association with the HLA antigen. This could explain the need for the association between skin disease and a particular HLA antigen. Bladder cancer (BCA) is one of the most common cancers. In terms of incidence, bladder cancer is second only to prostate cancer among urinary tract cancers and the incidence continues to rise. BCA is more common in men and elderly age patients. In the pathogenesis of bladder cancer, the inflammatory process mediated by cytokines play an important role. Recently, attention is also drawn to the importance of sub-population of Th17 T helper cells either in the tumor growth and progression or in the response to BCG therapy. Given the crucial role of impaired cytokine expression in predisposition to bladder cancer in the present study we analyzed polymorphisms in IL-17 gene and their impact on the pathogenesis of bladder cancer. Single nucleotide polymorphisms (SNP) rs2275913 (A>G) of IL-17A and rs763780 (T>C) of IL-17F genes were genotyped in 174 patients (male/female, 134/40) with bladder carcinoma (BCA) and in 118 healthy subjects (male/female, 88/30). Genotyping was performed using allelic discrimination methods with the TaqMan SNP Genotyping Assay. There were no differences in the frequency of alleles and genotypes between individuals with and without BCA for the evaluated polymorphisms (P>0.05). However, we found a trend to a lower frequency of rs2275913 [AA] homozygous in compared to the individuals carrying the G allele (GA +GG genotypes) in bladder cancer patients than in controls (p=0.07, OR 0.52, 95% CI 0.25-1.05). Moreover, we observed that patients with rs763780 [CT] genotype were less prone to bladder cancer than patients with [TT] genotype, but differences in genotype frequencies did not reach statistical significance (OR 0.56, 95%CI 0.25-1.25, p=0.15). When stratified by gender, we noted that the rs2275913 [AA] genotype was less frequent in female BCA patients compared to healthy woman, but differences did not reach statistical significance (OR 0.31, 95%CI 0.08-1.14, pcorrected=0.13). For SNP rs763780, male subjects with [CT] genotype were more often represented in the control group in comparison to men with BCA (OR 0.44, 95% CI 0.16-1.15, p=0.10). Our results do not strictly confirm the association of investigated polymorphisms with susceptibility to BCA, but they indicated that studies should be extended to larger group of patients. Worldwide studies demonstrated HLA-DRB1*03 and *04 as major genetic susceptibility factors for Type 1 Diabetes (T1D), an organ specific autoimmune disease. However, not all DR3/DR4 haplotypes predispose equally to T1D. HLA-G, a non-classical class I molecule is involved in induction and maintenance of tolerance and therefore implicated in the pathogenesis of autoimmune diseases and transplantation. The HLA-G 3´UTR 14 bp insertion/deletion polymorphism is associated with mRNA post-transcriptional regulation. Induction of HLA-G expression on dendritic cells from diabetics mediated the inhibition of autologous T cell activation, indicating role in diabetes pathogenesis. Conditional analysis among~2400 T1D affected families revealed a novel association around the HLA-G locus. In the present study we evaluated HLA-G 3´UTR 14 bp indel polymorphism among 78 T1D cases from North India and 70 ethnically matched healthy controls (HCs). The +14 bp allele was more frequent in T1D cohort (53% vs 47% in HCs). Genotype analysis revealed higher percentage of T1D cases to be ins/ins homozygous. Segregation of the study cohort on basis of the presence or absence of the T1D susceptibility allele (HLA-DR3) revealed interesting results. Among the DR3+ve group, the low sHLA-G secretary genotype (+/+14 bp) was more frequent in patients whereas high secretary genotype (-/-14 bp), was more frequent in the HCs. Among DR3-ve cohort, the low secretary +/+ genotype was more frequent in T1D cases (43% vs 22%, P=0.04). These findings suggest i) presence of high sHLA-G secretary genotype probably inhibited the autoreactive T cells in DR3+ve controls, ii) in DR3-ve T1D cohort, this inhibition was probably not possible due to presence of low secretary 14 bp genotype. The preliminary findings suggest that HLA-G 3´UTR polymorphism may influence diabetes pathogenesis. We are presently estimating sHLA-G levels in our cohort and its correlation with above findings may reveal further clinically relevant information. IL-23 is a pro-inflammatory cytokine belonging to the IL-12 cytokine family. This cytokine is necessary for the differentiation of Th17 lymphocytes, a subtype of T lymphocytes involved in chronic inflammatory diseases. The interleukin-23 (IL-23R) is one of the key elements in the inflammatory process mediated by T-helper cells 17, which also plays an important role in the pathogenesis of cancer. Interleukin-23 (IL-23) is produced predominantly by activated DCs and macrophages. It is believed that several single-nucleotide polymorphisms (SNPs) of IL-23R have a significant effect on the susceptibility to many tumors. Since chronic inflammation contributes to the pathogenesis of bladder cancer (BCA), the polymorphisms in the IL-23R gene may be associated with susceptibility to this disease. To assess whether polymorphisms of the interleukin-23 receptor gene are associated with the risk of bladder cancer, the following SNPs in the IL-23R gene: IL-23RC> A [CC] (rs10889677), IL-23Rc.1142G> A (rs11209026) were tested by allelic discrimination methods using the Genotyping SNP TaqMan Assay in 175 bladder cancer patients and 119 healthy subjects from the same geographic region in Poland. We found that patients with the IL-23RC>A genotype [CC] (rs10889677 CC) were less susceptible to BCA than those possessing A allele [AC + AA genotype] (OR 0.62, 95% CI 0.38-0.99, p = 0.04). What's more, we noticed that the frequency of allele IL-23RC>A [A] was higher in BCA patients than in healthy controls, but the differences did not reach statistical significance in investigated group of BCA patients and controls (0.30 vs 0.23, p=0.06). The distributions of alleles and genotypes for SNP IL-23Rc.1142G> A (rs11209026) were similar in BCA patients and controls. The haplotype analysis showed that haplotype rs10889677A, rs11209026G (haplotype AG) was more frequently represented in patients with BCA and increased the risk of BCA by 45% (p=0.05, OR 1.45, 95% CI 0.99-2:11). Our results indicate that the polymorphism of rs10889677 in the IL-23R gene may be considered as a potentially low-risk factor for bladder cancer, but the results must be confirmed in further studies. B and T lymphocyte attenuator (BTLA) is one of the members of the immunoglobulin superfamily which, like CTLA-4 and PD-1, negatively regulates immune responses. A high expression of BTLA and its ligand, herpes virus entry mediator (HVEM), was observed in hematological cancer cells. Although BTLA plays the important role in maintaining immune homeostasis, there are no published data on the relationship of polymorphisms in the gene encoding BTLA with susceptibility to multiple myeloma (MM). The aim of this study was to evaluate the association between three polymorphisms of BTLA gene, selected based on the results of our previous work on B-cell chronic lymphocytic leukemia, and susceptibility to multiple myeloma in the Polish population. The following SNPs in the BTLA gene: rs9288953 located in intron 1, rs1982809 in the 3 0 near gene position and rs2705511 in the intragenic region, were investigated. Genotyping was done using allelic discrimination methods with the TaqMan SNP Genotyping Assays in 206 patients with MM and in 468 healthy controls from the same region in Poland. The distributions of the alleles and genotypes in the BTLA SNPs were similar in MM patients and controls. However, we noted that patients with AA genotype in rs2705511 were less prone to MM than patients carrying the C allele [CA+CC genotype], but differences did not reach statistical significance (p=0.19, OR 0.80, 95% CI 0.58-1.12). Moreover, haplotype estimation analysis indicated that the haplotype of rs2705511A, rs1982809G, rs9288953C (AGC) was represented more frequently in MM patients and significantly increased the risk of MM (p=0.005, OR 3.20, 95%CI 1.36-7.54). The results of present study do not strictly confirm the association of investigated BTLA gene polymorphisms with susceptibility to multiple myeloma but indicated that haplotype rs2705511A, rs1982809G, rs9288953C (AGC) might be the risk factor of the susceptibility to MM. The main objective of our study was to analyze the mechanisms responsible for the loss of HLA class I (HLA-I) expression in lung cancer. We extracted DNA and RNA for genetic analysis from tumor cells isolated from HLA-I negative tumors by laser microdissection. We found that the loss of heterozygosity (LOH) at the chromosome 6 (6p21.3 region) (LOH-6) is a major cause of the loss of HLA-I haplotype in a significant number of cases analyzed. In addition, the RT-PCR analysis revealed a decrease in the expression of β2m, heavy chain (HC HLA) and the antigen processing machinery (APM) genes LMP2 and LMP7. This study concludes that the total loss of HLA-I expression in lung cancer is caused by two mechanisms: LOH-6 affecting HLA genes, and the decrease in the expression of APM genes. According to our study the frequency of LOH can be very high in lung cancer and it may even be underestimated. In fact, in our study 100% of the tumors classified as HLA-I negative presented LOH-6, but we don't know the frequency of LOH-6 in HLA-I positive tumors, which might be also high. Consequently, we also analyzed the incidence of LOH-6 in six human lung cancer cell lines with apparently normal cell surface expression of HLA-I molecules. Genomic typing of HLA-I in three of them revealed the existence of homo/hemizygosity compatible with a loss of HLA haplotype. Another cell line was also homozygous at HLA-B and C loci. Using GCH array analysis of the cell lines we verified the results of the genomic typing. Thus, SKLU-1 and SKMES cell lines showed a large loss in chromosome 6 and the A427 cell line presented a deletion of the short arm of chromosome 6 affecting the B and C loci. In conclusion, our study shows that HLA-I alterations have high incidence in lung cancer and that may be undetected if only traditional methods of analysis are used, including immunohistochemistry, flow cytometry and sequencing, all of which cannot detect large genetic deletions. The implication of the complement pathway in neurodevelopmental and innate immune processes suggest its involvement in neuropsychiatric disorders (increased C3, C4 in Schizophrenia [SZ], low C3 and C4 in systemic lupus erythematosus [SLE] ). This notion was recently demonstrated by reported correlations between C4 copy number variations (CNVs), structural indels of the HERV-K retrovirus, high expression of C4 molecules and risk of SZ. There is no data on bipolar disorder (BD). The C4 component corresponds to the functionally different C4A and C4B isoforms encoded by polymorphic loci which leads to functionally important CNVs and structural size variations (indels of HERV-K). Thus, we analyzed such genetic diversity in bipolar disorder (BD), SZ and SLE under a case control design. 146 BD, 151 SZ, 86 SLE and 106 healthy subjects, enrolled at JIPMR, India, were included in the study. Q-PCR and was used to determine the C4A, C4B, C4L and C4S CNVs and their distribution compared by the Chi square test. We found that (i) among all the tested groups, the most common C4 CNV variant was two copy allele, (ii) the frequency of the C4A one copy and C4B, C4L >3 copy alleles were higher in BD as compared to HC (P=0.01, P=0.01, P=0.002), (iii) the frequency of the C4L >3 copy and C4S 2 copy allele were higher in SLE compared to HC (P=0.0003, P=0.02 respectively), (iv) the frequency C4L two copy allele was higher in HC than in BD (P=0.0003), SZ (P=0.03) and SLE (P=0.0004) while the frequency of C4S one copy allele was higher in controls as compared to SLE (P=0.00005). Our preliminary data showed varying C4 profiles among the studied groups, especially higher copy numbers in disease groups as compared to HC. Ongoing analysis including C4 diversity, circulating C4 levels and fine genetic dissection of the cluster along with descriptive clinical subsets may provide insights into the biology of C4 in psychiatric and autoimmune conditions Classical autoimmune hepatitis (AIH), together with primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC), form the three main categories of autoimmune liver disease. The prevalence of AIH is estimated to range between 0.5 and 1/100,000 inhabitants. Despite the abundance of information in the literature on the immunopathogenetic mechanisms of AIH, very few reports have attempted to unravel the role of HLA-G molecules in the development of the disease. It is well known that HLA-G molecules are involved in modulating immune response and are capable of promoting strong tolerogenic effects. This prompted us to retrospectively analyze a total of 114 Sardinian patients diagnosed with type I AIH. Metabolic and genetic causes of hepatopathy were excluded in all patients. Patients and controls were typed at high resolution (Next Generation Sequencing) for the alleles and polymorphisms at the 3´UTR (UTR-1, -2, -3, -4, -5, -6, -7, -8, and rare) of the HLA-G locus. The immunogenetic characteristics of the AIH patients were compared with those of 156 healthy individuals from the Sardinian Voluntary Bone Marrow Donor Registry.No differences were observed for HLA-G allele frequencies between patients with AIH Type I and the healthy controls. Instead the frequency of the UTR-1 haplotype was much higher in the patients compared to the controls (25.6% vs 40.3%, OR 1.96, 95% CI 1.3 2.9, P=0.0004). The higher frequency of this haplotype corresponded to an increased expression of HLA-G molecules in patients affected by AIH-1. It is therefore possible to hypothesize that the HLA-G molecules expressed in AIH-1 patients, as well as having a tolerogenic function, are capable of exerting other functions. The endoplasmic reticulum aminopeptidases ERAP1 and ERAP2 trim HLA class I -binding precursors to optimal length antigenic peptides for loading onto HLA class I molecules, transport to the cell surface and presentation to CD8+ T cells and NK cells. HLA-C molecules are expressed by placental trophoblast and may interact with killer immunoglobulin-like receptors (KIR) on the decidual NK cell surface. Therefore, we investigated whether polymorphisms of KIR, HLA-C, ERAP1 and ERAP2 genes can predispose to recurrent spontaneous abortion. We tested 285 women with at least two miscarriages and 223 fertile control women. We used the TaqMan SNP Genotyping Assays for ERAP1 (rs26653, rs2287987, rs30187, rs27044, rs27524) and ERAP2 (rs2248374) typing. HLA-C and KIR genes were determined by PCR-SSP. The statistical significance of differences in genotype and allele frequencies between the control group and patients was estimated using the two-sided Fishers exact test and by the chi-square test. The homozygous women of ERAP1 rs30187TT genotype were prevalent in the recurrent miscarriage group (p = 0.045, OR = 1.84). Patients carrying rs30187TT and positive for HLA-C C2, and also patients carrying rs30187TT and KIR Bx were in higher frequency in the miscarriage group than in fertile women, however not significant (p = 0.078, OR = 1.96; p = 0.072, OR = 1.89, respectively). We found also the association of rs26618TT genotype in patients positive for HLA-C C1 group (p = 0.02, OR = 1.58) with recurrent miscarriage. Moreover, we observed higher frequency of the combined KIR Bx and rs26618TT genotype in patients with miscarriage (p = 0.053, OR = 1.53), while KIR Bx and rs26618CT genotype was decreased in patients group in comparison to fertile women (p = 0.048, OR = 0.64). In conclusion, some SNPs of ERAP1 in the context of HLA-C and KIR Bx genotype could be associated with susceptibility to recurrent miscarriage. IgA deficiency is the most common primary immunodeficiency in humans, with an estimated prevalence of 1/500-1/700. Positive and negative associations of HLA alleles with IgA deficiency have previously been reported, suggesting the involvement of genetic factors in the development of this disease. The aim of this study was to evaluate the HLA allele distribution in a cohort of IgA deficient patients from Cantabria (northern Spain) and to study the prevalence of commonly associated diseases. HLA-A, -B, -C, -DRB1, -DQA1 and -DQB1 alleles were studied in 156 patients, 77 males & 79 females, with IgA deficiency and in 262 healthy controls. Chi square or Fisher exact test were used for statistical analysis. In our IgA deficient patients, a significant positive association with HLA-A*01, A*33, B*08, B*14, C*08, DRB1*01, DRB1*03, DRB1*07, DQA1*02, DQA1*05 and DQB1*02 alleles was found. By contrast, HLA-A*24, B*07, B*15, C*01, C*02, C*03, DRB*11, DRB*15, DQA1*03, DQB1*03 and DQB1*06 alleles were significant more frequent in the control group. Among the 156 patients with IgA deficiency, 34 were asymptomatic (21.8%), 82 had recurrent infections (52.6%), 29 had celiac disease (18.6%), other autoimmune disorders (hypothyroidism, rheumatoid arthritis, vitiligo, scleroderma, Crohns disease, idiopathic thrombocytopenic purpura, and insulin-dependent diabetes mellitus) were documented in 23 cases (14.7%). Finally, allergic manifestations including rhinitis, asthma and atopic dermatitis were found in 35 patients (22.4%). Our results regarding IgA deficiency-HLA association are coincident but not at all with previous studies performed in other Spanish populations. For the first time, a weak protective role for B*15, C*01, C*02, C*03, DRB1*11 and DQA1*03 alleles was found. On the other hand, the clinical profile of our IgA deficiency population, was different from other previously reported Spanish populations. In coeliac disease, specific HLA haplotypes have the ability to efficiently present gliadin peptides to reactive T cells in the gut. Approximately 90% of celiac patients express HLA-DQ2 in cis (DRB1*03:01~DQA1*05:01~DQB1*02:01) or in trans (DRB1*07~DQA1*02:01~DQB1*02:02; DRB1*11/12~DQ A1*05:05~DQB1*03:01). The remainder of the patients carry HLA-DQ8 (DRB1*04~DQA1*03:01~DQB1*03:02). In this study 99 (66 female and 33 male) unrelated celiac patients were diagnosed in a single center in Northern Greece and fulfilled the ESPGHAN diagnostic criteria. They were subjected to analysis in order to analyze the HLA-DQ distribution and determine the disease risk conferring by the different HLA-DQ types. The control group was consisted of 120 healthy unrelated individuals. Disease risk was expressed as 1:N, where N is the number of subjects among which is present one patient. Considering a disease prevalence of 1:100 in the general population, N was calculated as a percentage of controls with that particular haplotype multiplied by 100 and divided by percentage of patients with the same HLA-DQ type. HLA-DQ2 (cis or trans) and/or DQ8 were detected in 98 patients (99 %) and in 44 (37 %) controls. The only patient who did not carry neither DQ2 nor DQ8 was homozygous for DR11-DQ7. Based on the HLA-DQ genotype, we obtained a gradient of disease risk ranging between 1:7 and 1:5445. The highest risk was for subjects carrying HLA-DR3~DQ2/ DR4~DQ8 (1:7) followed by: DQ8/B1*02 (1:10), DR3~DQ2/B1*02/*02 (1:16) and DQ2 trans (DR7~DQ2/ DR11~DQ7) 1:30. The presence of DQ8 only without B1*02 led to the disease likelihood of 1:108 while the presence of beta chain *02 (usually presented in haplotype DR7-DQ2) only without any of other susceptible alleles resulted in 1:208 disease risk. The second most predisposing genotype was DQ8/B1*02 (1:10) presenting an effect of DQ8/B1*02 also evident in absence of DQA1*05 as the occurrence of DQB1*02 allele in DQ8+ subjects raised the risk from 1:108 to 1:10. We suggest that asymptomatic relatives of patients carrying DQ2/DQ8 and DQ8/B1*02 should be monitored more often because of the highest disease risk with these haplotypes for early diagnosis. Behcet disease is an autoimmune vasculitis disorder which is characterized by a triad of recurrent oral ulcers, genital ulcers, and ocular involvement that might lead to blindness of patients suffering from this disease. Prevalence of Behcet disease has been reported historically to be distributed among geographical areas along the ancient Silk Road. It has been long known that there is a strong association between the incidence of Behcet disease and class I HLA-B alleles, mainly HLA-B*51. To clarify the disease association between HLA and Behcet disease we participated in the 17th IHIW in the HLA NGS component for disease association. 60 Behcet disease Egyptian patients and 40 healthy Egyptian subjects as control group were HLA typed at low resolution using INNO-LiPA (Fujirebio) SSO strip assay for two loci (HLA-A and HLA-B) and at allelic level resolution for 11 loci (HLA-A, -B, -C, -DRB1, -DRB3/4/5, -DQA1, -DQB1, -DPA1, -DPB1) using the NGSgo Workflow (GenDx) on an Illumina MiSeq platform by paired-end sequencing (2x150 bp), and HLA typing with NGSengine software (GenDx). The typing data that were collected for the IHIW have now been studied and the results are presented here. Both studies show that Behcet disease is highly associated with HLA class I alleles, of which HLA-B*51:08:01 and HLA-B*51:01:01 (p<0.001, p<0.001) are the alleles with highest significance. When comparing the patients genotype with their clinical manifestations again HLA-B*51 was associated with severe eye disease (p<0.001) and blindness (p<0.005). It is highly recommended for patients with origins linked to the historic silk road to perform HLA-B*51 genotyping as risk assessment for blinding eye disease in an attempt to guide treatment options to salvage patients' eyesight. The contribution of the immune system to the pathogenesis of multiple myeloma (MM) has been receiving growing attention. The important role of co-inhibitory receptors in these processes has been considered. One of these receptors is inducible costimulatory molecule (ICOS) -a member of CD28-superfamily which is expressed on activated T cells. We extended our previous work which was undertaken to evaluate the association between ICOS gene polymorphisms and susceptibility to MM in the Polish population with a new ICOS gene polymorphism -ICOSc.2373G>C (rs4675379). Genotyping was carried out using allelic discrimination methods with the TaqMan SNP Genotyping Assay in 392 individuals, including 205 MM patients. Moreover, we performed haplotype estimation analysis concerning previously described polymorphisms: ICOSISV1 +173T>C (rs10932029), ICOSc.1624C>T (rs10932037), ICOSc.602A>C (rs10183087), ICOSc.1564T>C (rs4404254) and newly investigated ICOSc.2373G>C (rs4675379). We noted that the frequency of ICOSc.2373G>C[G] allele was higher in MM patients than in healthy controls, but differences did not reach statistical significance ( There is a growing evidence that endothelial function could be involved in the pathogenesis of multiple sclerosis. Previous reports have shown its role in other autoimmune disorders like rheumatoid arthritis. HLA-DRB1*15 induces a double risk for having multiple sclerosis. Furthermore, in relapsing-remitting multiple sclerosis several markers of inflammation have been associated with the severity of disease. For this reason, we have investigated the endothelial damage in multiple sclerosis patients measuring carotid intima-media thickness and its correlation with the levels of inflammation biomarkers. 37 patients with relapsing-remitting multiple sclerosis, aged 40.90, median EDSS= 2.0 according to McDonald 2010, criteria and without known vascular risk factors, were recruited. Carotid intima-media thickness was measured in these patients by ultrasound scan. DNA was obtained from peripheral blood by using a semi-automated method. HLA antigens were identified by polymerase chain reaction with sequence-specific oligonucleotides and Luminex analysis.

Several biomarkers of inflammation were also analyzed with the Milliplex system on a Luminex 200 platform. The Chi 2 method and Student T test were used for statistical analysis. Patients with the HLA-DRB1*15 antigen (n=14) had carotid intima-media thickness thicker (mean 1.044 mm) than those lacking this antigen (mean 0.928 mm; p=0.041). There was no correlation between the HLA-DRB1*15 allele and several biomarkers of inflammation (reactive C protein, erythrocyte sedimentation rate, fibrinogen or von Willebrand Factor). For the first time, we show an association of HLA-DRB1*15 with endothelial damage measured by using the carotid intima-media thickness in patients having relapsing-remitting multiple sclerosis. We hypothesized that endothelial damage could be correlated with inflammation and disease activity. Further and larger studies are required to confirm this association.

Katsushi Tokunaga 1 , Xiaoyuan Jia 1 , Tomoko Horinouchi 2 , Yuki Hitomi 1 , Akemi Shomo 2 , Seik-Soon Khor 1 , Yosuke Omae 1 , Kaname Kojima 3 , Yosuke Kawai 3 , Masao Nagasaki 3 , Kandai Nozu 2 , Kazumoto Iijima 2 1 University of Tokyo, Graduate School of Medicine, Tokyo, Japan, 2 Kobe University Graduate School of Medicine, Kobe, Japan, 3 Tohoku Medical Megabank Organization, Sendai, Japan Correspondence: tokunaga@m.u-tokyo.ac.jp Nephrotic syndrome is the most common cause of chronic glomerular disease in children. A genome-wide association study (GWAS) was conducted in the Japanese population. A total of 224 patients with childhood steroid-sensitive nephrotic syndrome (SSNS) and 419 adult healthy controls were genotyped using Affymetrix Japonica Array in the discovery stage. Imputation for HLA-A, -C, -B, -DRB1, -DQB1, and -DPB1 genotypes was conducted based on Japanese-specific references. Genotyping for HLA-DRB1 and -DQB1 was performed using sequence-specific oligonucleotide probing method on a Luminex platform. Wholegenome SNP imputation was conducted using a phased reference panel of 2,049 healthy Japanese individuals. Replication was performed in an independent Japanese sample set including 216 cases and 719 healthy controls. Candidate single nucleotide polymorphisms were genotyped using DigiTag2 assay in the replication stage. The most significantly associated SNP was detected in the HLA-DR/DQ region and successfully replicated (rs3134996, combined P =1.720 −25 , OR= 0.29). HLA typing and imputation identified HLA-DRB1*08:02 (P=1.820 −9 , OR=2.62), and HLA-DQB1*06:04 (P=2.090 −12 , OR=0.10) as most strongly associated alleles with the disease. Moreover, HLA-DRB1*08:02-DQB1*03:02 (P=7.010 −11 , OR=3.60) was identified as the most significant susceptible genetic factor, showing more significant and stronger association than single alleles. Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland, 2 Wroclaw Medical University, Wrocław, Poland, 3 Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland Correspondence: bogunia@iitd.pan.wroc.pl Rheumatoid arthritis (RA) is an autoimmune disease characterized by inflammation of joint synovial tissue, resulting in the progression of cartilage and bone tissue damage, ultimately leading to disability. Our previous studies showed that IL17A, IL17F and TNFA genetic variants may affect RA susceptibility or treatment outcome. The present study aimed to assess the relationships between serum levels of pro-inflammatory cytokines, their genetic variability and progression of the disease. Serum levels of selected cytokines were examined using monoplex (IL-17F) and 7xplex ProcartaPlex Multiplex (IL-17A, IL-1beta, IL-6, IL-22, IL-23, sRANKL, TNF-alpha) Immunoassays from Invitrogen Inc., on a Luminex 200 system. The measurements were performed for 127 RA patients at three time points: before, three and six months after anti-TNF treatment and 73 controls. Genotyping was performed employing either Taqman or LightSNiP assays. As compared to controls, patients before therapy presented with significantly higher serum levels of IL-6 (p<0.001), IL-1beta (p=0.002) and TNFalpha (p<0.001). The presence of the IL6 rs1800795 CC homozygous genotype was associated with the highest IL-6 concentrations before treatment (p=0.026). This homozygosity was also more frequently detected in patients with more active disease (DAS28>3) (p=0.048). IL1B genotypes were not found to be significantly associated with serum level of this cytokine. RA patients possessing the IL1B rs16944 C allele are characterized with lower CRP levels after three months of anti-TNF therapy (p=0.035).

As for TNFA polymorphisms (rs1800629, rs1799724, rs361525) studied, a significant relationship was identified between TNFA rs1799724 CC homozygosity the presence of rheumatoid factor (p=0.033). These results show that patients and controls differ with respect to the IL-6, IL-1beta and TNF-alfa levels. The IL6 and TNFA CC homozygosity seem to play rather unfavorable role in RA. Recurrent implantation failure (RIF) affects approximately 10% of women who have undergone several in vitro fertilization embryo transfers (IVF-ETs). KIR and HLA-C interactions between NK cells and trophoblasts have been proposed to influence the IVF-ET outcome. The aim of this study was to evaluate an association of genes encoding KIR and its ligand -HLA-C with susceptibility to RIF. A total of 501 couples from the Polish population were recruited to this study. Among them 182 couples suffered from recurrent implantation failure after IVF-ET, and 319 were fertile couples after natural conception. We used PCR-SSP for detection of KIR and HLA-C genes. Statistical analysis was performed using two-tailed Fishers exact test. We found higher frequency of KIR2DL2 and KIR2DS2 genes in female patients with RIF in comparison to fertile women, but results were not significant (p = 0.06, OR = 1.43 and p = 0.07, OR = 1.41, respectively). Surprisingly, KIR2DL2, KIR2DS2, KIR2DS3, and KIR2DL5 were significantly more frequent in male partners from the RIF group in comparison to partners from the control group (p = 0.0005, OR = 2.09; p = 0.0012, OR = 1.96; p = 0.0013, OR = 2.02; p = 0.021, OR = 1.62, respectively). There were no differences in the frequencies of HLA-C C1 and C2 alleles and genotypes between RIF patients and fertile women, and also between their partners. We observed no association of the combined female KIR and partner HLA-C genotype with RIF. In conclusion, polymorphisms of KIR genes in male partners could be associated with recurrent implantation failure of their females. Celiac disease in one of the most prevalent genetically determined autoimmune diseases. Furthermore, it is one of the diseases with the strongest association with particular HLA haplotypes, specifically with DQ2 and DQ8. It has been demonstrated that the presence of these haplotypes confers a risk to the development of the disease through ingestion of gluten that would lead to a broad range of clinical signs and symptoms with histologic alterations in small bowel. Usually, diagnosis of celiac disease includes determination of IgA or IgG antibodies, anti-tissue transglutaminase type 2, anti-endomysial, and anti-deamidated forms of gliadin peptides. In most cases, a definitive diagnosis requires a jejunal biopsy showing typical histologic abnormalities. Regarding the risk of the HLA haplotypes, several discrepancies among populations have been found, mainly due to no take into account the difference between the HLA DQ2.2 and DQ2.5 molecules. For this reason, we genotyped 781 celiac disease patients (492 children < 15 years old, and 289 adults) who fulfilled the ESPGHAN criteria for this disease by using polymerase chain reaction and hybridization with sequence specific oligoprobes on a Luminex platform. In our population, when we analyzed both children and adults, we found some differences in the frequency of the haplotypes associated with an increased susceptibility to celiac disease compared with previously reported populations and in the risk that these haplotypes conferred the disease, mainly in the distribution of DQ2.2/DQ2.2, DQ2.2/x, DQ2.2/DQ7 and DQ8 either in homozygous or heterozygous states. Interestingly, half of the DQ2.2/x genotypes carried DQA1*04 in trans position. This difference did not change substantially when we analyzed them separately. We think that the main reasons of these differences are: a) the different prevalence of these haplotypes among populations and b) lack of separate analysis between DQ2.2. and DQ2.5. the HLA-G gene, which potentially is related to decreased HLA-G expression levels. The association of celiac disease (CD) and HLA class II genes is one of the best established and most documented disease associations of HLA system. The aim of our retrospective study was to determine whether there is a difference in HLA class II allele distribution among individuals who have a higher risk of developing CD (family members of confirmed CD patients or individuals with unconfirmed diagnosis who were referred to our Tissue Typing Centre for HLA typing as a part of the diagnostic procedure) and healthy controls with no history of CD in their family background. For that purpose, we analyzed the results of HLA-DQ typing performed using the PCR-SSP high resolution method (Olerup GmbH, Vienna, Austria) carried out for patients with CD (N=118), family members of CD patients (N=206), subjects with unconfirmed CD (N=1685) and healthy controls (HC) (N=985). Results show that genotype DQ2.5 cis in both single or double dose was statistically more frequent among all the groups when compared with HC (P<0.0001, each). Also, it was more frequent among patients with CD than family members (P<0.0001 both single and double dose) and unconfirmed patients (P=0.0003 and P=0.0057, respectively). The trans configuration of this genotype also showed significantly higher incidence among patients when compared with HC, but also unconfirmed patients and family members (P=0.0002, P=0.0032 and P=0.0160, respectively). No difference was seen between unconfirmed patients and family members when compared with controls. The DQ8 genotype did not show this specific pattern. No statistical significance was observed in any of the groups both for single or double doses. These results show that the highest risk for CD was indicated by the presence of the HLA-DQ2.5 heterodimer (HLA-DQA1*05-DQB1*02) both in single or double dose combination for all CD patients and underline the importance of HLA-DQB1 typing for assessing the susceptibility to CD.

Manuela Testi 1 , Giulia Greco 2 , Maria Troiano 3 , Emanuele Rastelli 2 , Eugenio Pompeo 4 , Giovanni Antonini 5 , Marco Andreani 3 , Roberto Massa 2 1 San Camillo-Forlanini Hospital, Rome, Italy, 2 Neuromuscular Unit, Rome, Italy, 3 Laboratory of Immunogenetics and Transplant Biology -IME Foundation at Polyclinic of Tor Vergata, Rome, Italy, 4 Thoracic Surgery Unit, Rome, Italy, 5 Department NESMOS, Rome, Italy Correspondence: m.andreani@fondazioneime.it

We assessed the relative prevalence of Myasthenia Gravis (MG) subtypes, their differential clinical features and their association with HLA class II alleles within a large series of unselected Italian patients. In 230 consecutive patients, age at onset, presence/absence of serum anti-acetylcholine receptor (AChR), or anti-muscle-specific kinase (MuSK) antibodies (Ab), presence of thymoma were considered for stratification in MG subgroups. Moreover, we performed (n=179) HLA-DRB1 LR l and HLA-DQB1 HR typing, in order to extend our preliminary findings suggesting associations of particular alleles with MG subtypes. Mean age at onset was lower for females (51.9%) and for thymomatous (TMG) patients. Comorbidity for other autoimmune disorders was frequent (29%), with thyroiditis being most common (58%). TMG patients were more often associated with higher MGFA grades and showed significantly higher anti-AChR Ab titer than non-thymomatous (NTMG) patients. Among the latter, those with late onset (LO-NTMG) showed significantly higher Ab titers, but less often reached high Myasthenia Gravis Foundation of America (MGFA) grade than the early-onset (EO-NTMG) patients. Significant associations were found between HLA DQB1*05:01 and TMG patients and between DQB1*05:02 and DRB1*16 alleles and LO-NTMG. Therefore, LO-NTMG is the most frequent disease subtype, with an increasing incidence in Italy. The association between high anti-AChR Ab titers and high MGFA grades infers a prognostic value to Ab serum concentration. TMG patients tend to reach a higher clinical severity as well as higher antibody titers compared to NTMG ones. Our findings also suggest that two distinct cutoffs (<40 and >60 years at onset) may better define MG age subgroups, with different clinical and serological characteristics. Moreover, our results show that LO-NTMG and TMG differ in their HLA-DQ association, adding to the existing evidence that suggests these two forms likely have different pathogenic mechanisms.

Dentistry, Palacky University and University Hospital, Olomouc, Czech Republic Correspondence: martin.petrek@fnol.cz Sarcoidosis is a multi-systemic inflammatory disease of unknown aetiology mostly affecting lungs. A role of environmental and genetic factors of polygenic nature has been suggested in disease etiopathogenesis. For this study, we selected 18 SNPs located in different genes with major roles in the immune system and inflammation to investigate their association with pulmonary sarcoidosis in the Greek population. The selected SNPs were investigated in 77 Greek sarcoidosis patients [mean age in years standard deviation, age range, male:female ratio) 56.6 11.6, 28-85, 22:55 compared to 301 healthy control subjects (33.5 10.8, 16-79, 134:167] of West-Slavonic (Czech, Slovak, and Polish) origin. All SNPs were genotyped using a multiplexed MassARRAY iPLEX assay based on MALDI-TOF mass spectrometry (Agena Bioscience, CA, USA). Each SNP was tested for HWE (> 0.01) using Pearson's chi-square test or Fishers exact test, as applicable. Genetic associations with odds ratio (OR) and 95% confidence intervals (CI) were estimated using allelic model with Fishers exact test. The analysis revealed associations of functional ANXA11 rs1049550*A Arg230Cys (OR [95% CI], p-value as 0.67 [0.98-0.46], 4.01E-02) and intronic TGFB2 rs1891467*A (1.66 [2.46-1.12], 1.40E-02) variants with sarcoidosis in Greek patients at a primary level. These preliminary findings extend previous studies performed in other Europeans to the Greek population. Further analysis of investigated gene variants in an extended patient cohort and most importantly an inclusion of Greek healthy control subjects for results verification are in progress in our laboratory.

Kalliroi Tsalimalma 1 , Kalliopi Adam 1 , Athanasios Kontos 2 , Dimitris Maltezas 2 , Nikolitsa Kafassi 1 1 General Hospital Athens Laiko, Athens, Greece, 2 National University of Athens, Athens, Greece Correspondence: kalliopi58@live.com

The CD4/CD8 ratio normalization has emerged as a strong predictor for immune reconstitution in virologically suppressed HIV-1 (+) patients. However, most HIV-infected individuals fail to normalize their CD4/CD8 despite CD4 recovery after antiretroviral treatment (ART). Although the characteristics of this phenotype have been extensively studied, they still have not been elucidated. HLA class I have been consistently linked to the outcome of HIV infection but to our knowledge, no data have been reported for HLA class II alleles in immune recovery after ART initiation. In the present study, we selected 69 HIV infected patients on ART with sustained undetectable viral load (<50 copies per ml) and classified them in two groups according to the time from ART initiation till CD4/CD8 normalization (equal to one or more in at least three consecutive values). One group (A), 32 individuals reached CD4/CD8 normalization more rapidly in the first 74 months of ART. In the other group of 37 individuals (B), 6 patients showed a later ratio normalization, 95-163 months after ART and the rest (31) retained CD4/CD8 <1 until their last follow up, 8-20 years from ART. All patients were typed for HLA-DR low resolution with PCR-SSP. Between the patient groups, significant differences were observed in the frequencies of DRB1 alleles. Individuals from (A) were presented with lower DRB1*07 and higher DRB1*11 allele (1.6 vs 9.5% p=0.04 and 35.9 vs 20.3% p=0.03). The Kaplan-Meier analysis showed a higher probability of achieving CD4/CD8 normalization for individuals without the DRB1*07 and lower for those having the DRB1*11 allele. Our results ascribe a prognostic character to HLA-DRB1 typing for individuals with a persistently suboptimal CD4/CD8 after long term therapy. Patients negative for DRB1*11 or positive for DRB1*07 allele, may be proven appropriate candidates for extended therapeutic interventions. Since HLA-B*57:01 typing is anyway performed at the beginning of HIV, DRB1 could be additionally included. Certainly, more data from larger sample size are needed to validate these results. Celiac disease (CD) is a prevalent immune-mediated enteropathy involving a permanent intolerance to dietary gluten in genetically susceptible individuals. Although gluten is the environmental driver of CD, several other features such as a strong HLA class II gene association (HLA-DQA1 and HLA-DQB1), auto-antibodies and the presence of CD4+ T cells are also correlated with this disease. Thus, CD provides an excellent example of the likely benefits of incorporating genetic testing in patient management. Most of the CD patients (~90-95%) carry the DQ2 allele encoded by HLA-DQA1*05 and HLA-DQ*02 while the remainder carry DQ8 allele encoded by DQA1*03 and DQB*03:02.

We present here the evaluation and summary of the HLA-DQ genotypes observed in patients referred for celiac disease testing at our referral HLA laboratory. Patient DNA samples were typed for DQA1 and DQB1 genes by PCR-SSP (Sequence Specific Primer) and/or SSO (Sequence Specific Oligonucleotide) methods. Patients carrying the DQA1*05:01/05~DQB1*02 haplotype were considered as HLA-DQ2.5 heterodimer positive, DQA1*02:01~DQB1*02 haplotype were considered as HLA-DQ2.2 heterodimer positive and DQA1* 03~DQB1*03:02 were considered as HLA-DQ8 heterodimer positive while the remaining haplotypes were considered as others. Of 387 samples tested, 226 (58%) were found to be DQ2.5 positive, 34 (9%) were DQ2.2 positive, 15 (4%) were DQ8 positive and six (2%) were both DQ2 and DQ8 positive, while, 102 (26%) cases were amongst others. Based on this study, it can be observed that majority of the patients referred for CD testing carry the permissive gene pairs (heterozygous for DQ2.5 or DQ8) while a proportion of patients also carry haplotypes other than DQ2 and DQ8 which implies the significance of HLA typing technique (by molecular methods including both DQA1 and DQB1 locus) for accurate HLA genotyping and its benefits in CD patient management.

HLA ASSOCIATED DISEASES. WHAT SHOULD BE ANALYZED?

Kirsten Kronenberg 1 , Stela Radojska 1 , Melanie Stoermer 1 , Birgit Gathof 1 1 University of Cologne, Cologne, Germany Correspondence: kirsten.kronenberg@uk-koeln.de Immune recognition is the important function of the human major histocompatibility complex (MHC), one of the most polymorphic genetic systems in humans. Many HLA genes are reported to be associated with complex autoimmune and infectious diseases. This aim of this study was to re-evaluate HLA typing for disease association in a university hospital. The international scientific literature and different medical societies guidelines were analyzed for diagnostically relevant disease associations. Requests for disease associations to our laboratory were analyzed between 2013-2017. While rheumatic diseases are the most common HLA associated disorders, HLA-B*27 (14%) requests declined over time. For coeliac disease HLA-DQA1*05:01/DQB1*02:01 and HLA-DQA1*03:01/DQB1*03:02 (22%) are associated with an immunologically mediated disease of the small intestinal mucosa, related to ingestion of gluten. Recently HLA associations with drug intolerance such HLA-B*57:01 (53%) and Abacavir, have become diagnostically relevant to prevent serious hypersensitivity reaction. The strongest association in HLA class II (90-100% Caucasian patients) is narcolepsy, related to HLA-DRB1*15:01/DQB1*06:02 (2%). In future, for prediction or early diagnosis of diabetes Type 1 HLA-DQB1*02:01/DQB1*03:02 alleles might be discussed. More than 100 diseases were reported to be HLA associated in literature. Which ones to determine in clinical and preventive contexts will be a project for the next years. The development of new typing technologies and innovations in bioinformatics will lead to better understanding of the structure of the MHC region and its role in disease associations. Genome wide, large scale and multicenter studies are necessary to look behind the mechanism of the relevant HLA associated diseases.

BETA-2-MICROGLOBULIN: A POTENTIAL BIOMARKER FOR BIPOLAR DISORDER AND SCHIZOPHRENIA?

Romain Sayous 1 , Céline Manier 1 , Nora Hamdani 2 , Alix Romier 1 , Aparna Sundaresh 1 , Christina Mariaselvam 1 , Rémi Gadel 3 , Céline Hebbache 3 , Jean-Romain Richard 3 , Marion Leboyer 2 , Ryad Tamouza 2 1 Inserm U955, Créteil, France, 2 Inserm U955/FondaMental Foundation/AP-HP, Créteil, France, 3 Inserm U955/Fonda-Mental Foundation, Créteil, France Correspondence: r@sayous.org Bipolar disorder (BD) and schizophrenia (SZ) are staging disorders with progressive cognitive and functioning decline. Among etiopathogenic hypothesis, immunoinflammatory dysfunction re-emerged recently, setting up the innovative field of immunopsychiatry. Beta-2-microglobulin (B2m) is a potential biomarker for cognitive decline: (i) part of MHC-I molecules at cell surface, its circulating concentration reflects cellular immunity activation; (ii) while MHC-I has a pivotal role in neurodevelopment/plasticity, B2m is implicated in cognitive ageing processes. However, its relationship with psychiatric condition remains unclear. We hypothesized that serum B2m levels may reflect acute episodes and functioning impairment in BD and SZ. We analyzed B2m levels in a cohort of 128 BD / 59 SZ acute episode inpatients assessed twice (admission & discharge), compared with 46 BD / 46 SZ stabilized outpatients and 115 healthy controls. Subjects were clinically assessed by standardized interview: MADRS, YMRS, PANSS for symptom dimensions and GAF, CGI, FAST for global functioning. Data analysis compared serum B2m between groups (ANOVA / groupby-group t-test) then admission vs discharge (paired t-test). Finally, we explored correlations with clinical scores. Significant B2m level differences appeared between groups (p<.0001) but remained stable on discharge vs admission. In bipolar disorder, B2m levels were higher on acute episode (1.88 mg/L) compared to euthymia (1.65 mg/L, p<.005) and controls (1.58 mg/L, p<.00001). In schizophrenia, patients had higher B2m levels (1.84 0 mg/L, p<.002) compared with controls, correlated with PANSS disorganization sub-score (N7+G11+G10+P2+N5 items, p<.001). These preliminary data suggest serum B2m as a state biomarker of acute phase bipolar disorder and a disorganization severity biomarker in schizophrenia. Further analysis will investigate it on a staging and functioning perspective. Transfusion of red blood cell (RBC) products is one of the most relevant causes of immunization against RBC alloantigens. Association between the presence of monospecific RBC alloantibodies and particular HLA-DRB1 variants was reported in several studies. Moreover, we and others have shown that the RBC multiresponder status is generally associated with the HLA-DRB1*15 allelic group. It is apparent that only a minority of individuals are prone to develop RBC alloantibodies because there is high percentage of patients who do not form alloantibodies despite repeated transfusion therapy. The aim of this study was to investigate possible association between the unresponsiveness to RBC antigens and HLA class II variants. In total, 109 Czech patients without RBC alloantibodies who were administered with at least 10 RBC units (referred to as non-responders below) were enrolled into the study. 96 non-responders (88.1%) had RhD positive and 13 (11.9%) RhD negative status. HLA-DRB1 and -DQB1 variants were determined by a PCR-SSO method and their frequencies were compared between the group of non-responders and 375 ethnically and regionally matched controls. By comparison of 13 DRB1 and 5 DQB1 allelic groups between the nonresponders and healthy control group no statistically significant differences in the frequencies of investigated HLA class II variants were revealed. In conclusion, we did not find any association between HLA class II and the unresponsiveness to RBC antigens. Further work is needed to analyze HLA class II variants in larger non-responder subgroups according to their RBC antigen status which may interfere with the susceptibility to alloimmunization by particular RBC antigens. are published and/or accessible from GenBank, EMBL Sequence Database (EMBL), or DNA Databank of Japan (DDBJ), and provides an important tool for investigators in the field of biomedical, evolution and conservation research. In 2017, SFHI proposed to its members an EPT program for HLA antibody (Ab) identification with the SA techniques. A total of 24 sera distributed in April were tested in May and October (six per class at each time point). The originality of this EPT is to analyze the results at three levels: 1. the classical, qualitative level (consensus for positive and negative specificities); 2. a quantitative study of mean fluorescence intensity (MFI) of each bead for each serum; 3. the clinical interpretation made from the Ab profiles for the participants affiliated to the French biomedical agency for organ allocation (Agence de la Biomedicine): list of antigens that would be declared as forbidden, permissible and grey zone, according to CRISTAL, the French database for kidney transplantation allocation. Participants recorded their results via the dedicated web platform CQXplore(TM), easily allowing for the copy of the MFI data from the Fusion (One Lambda) and Match-It (Immucor) softwares as well as for the input of the positive/negative and forbidden/permissible/grey zone specificities. This three level analysis allows the performances between laboratories and providers to be anonymously compared and practices to be discussed through the SFHI antibody and organ transplantation working group, as consequences of distinctly reporting anti-allele, anti-DQalpha or anti-DR51/52/53 antibodies into CRISTAL can have dramatic consequences on transplant access. In 2017, 26 francophone laboratories (25 from France) and two reagent kit providers were enrolled. Data analysis and performance certificates were returned within 15 days after the data collection deadline. Detailed results are accessible to members on the SFHI website. In 2018, the EPT will be extended to anti-HLA Ab screening, and chimerism, and HLA Typing is scheduled for 2019. The EPT is registered at EFI and should be open to all European laboratories from 2019 on. Cytokines are constitutively expressed and influence both the development and function of the nervous system. In particular, interleukin-1beta (IL1beta) has been recognized to play an important role in normal development of the brain as a regulator of synaptic function and neuronal plasticity underlying learning and memory. Imbalanced cytokine production, signaling, and/or regulation can therefore have a wide range of neurological consequences. The cytokine's production is under genetic control and polymorphisms have been identified in these genes. In intellectual disability (ID) such as already demonstrated in autism, cytokine dysregulation could contribute to neuronal dysfunction influencing cognitive development. To investigate on this, the polymorphisms of several cytokines and their serum levels in 45 ID patients and 23 controls were studied. We demonstrated that IL-1beta -511 (CT) and +3962 (CT) polymorphisms are involved in ID, showing a positive association of CC genotype (-511 p=0.04; +3962 p=0.04) and a TT genotype negative association (-511 p=0.0009; +3962 p=0.025). The gender-related difference analysis in the disease onset confirmed results only for the -511 polymorphism. Specifically, the predisposing association of CC genotype was established in the females and the protective one of TT genotype in the males. The gender impact was still more evident considering both polymorphisms. In fact, the predisposing CC-CC combination was more frequent in females (p=0.02). Analyzing the IL-1beta serum levels, increased levels were seen in patients of both groups with predictive CC -511 genotype (p<0.05). It is interesting to note that the highest serum levels have been found in females with severe ID. Overall, our results could help to identify factors causing abnormal CNS function in child with ID to develop correct interventions and treatments. extraction of 90 blood samples and six buccal swab samples was performed with four different methods: QiaSymphony (32 samples), manual Qiagen columns (27 samples), EZI Tissue Kit (six swab samples) and Quatrobrobe (31 samples). The DNA concentration used to carry out HLA HD typing was of 12.5-25 ng/μl with both kits. Data sheets were elaborated with Fusion and Histomatch softwares, according to the IPD-IMGT/HLA Database version 3.27.0. We found 100% concordance in final identification of HLA-A, -B, and -DRB1 Common/Well Documented alleles between the kits but the number of the ambiguities was different from one kit to the other. Regarding the analytical performances, in Luminex session 16 HLA-A, -B, -C, -DRB1 and -DQB1 + supplemental typings were performed, allowing us to delete a great number of null alleles (e.g. B*51:11N). In Bag session only 16 HLA-A, -B and -DRB1 +Extend samples were typed but in a shorter time and with complete automation. These preliminary data suggest that Bag HISTO SPOT Extend is a very useful method for HD resolution in a small to medium size H&I laboratory. On the other hand, for high resolution typing a different approach is needed such as SBT plus SSOR or NGS technology.

before abacavir treatment the detection of the allele is important in sensitive patients. The aim of this study was the detection of HLA-B*57:01 prevalence in HIV infected patients in the Turkish population. 13 women, 165 men of 178 HIV-1 positive patients were analyzed for HLA-B*57:01 in the Aegean region. A control group consisted of 2946 bone marrow patients that were tested for HIV and found to be negative. HLA genotyping was performed by PCR-SSP (HLA-B CombiTray Kit, Olerup, Sweden) method. When the B*57 allele was detected, HLA-B*57:01 allele presence/absence was tested by using HLA-B*57:01 CombiTray Kiti Olerup, Sweden. HLA*B57 alleles were detected in five patients. The HLA-B*57:01 subtype was detected in four patients (2.26%). The prevalence of HLA-B*57:01 of the control group was detected as 2.92%. HLA-B*57:01 detection is a pharmacogenetic test example to detect the adverse effects of a drug

Endometriosis is a disease that affects 10% of women of reproductive age. It relates to the occurrence of ectopic endometrium outside the uterus. The etiology is still unknown. It was found that the decreased activity of natural killer cells has an impact on endometriosis. Killer cell immunoglobulin-like receptor, KIR2DL4, and leukocyte immunoglobulin-like receptor B (LILRB) are expressed on their surface and may regulate their activity. KIR2DL4, LILRB1 and LILRB2 recognize human leukocyte antigen-G (HLA-G) localized on ectopic endometrial cells. The aim of our study was to evaluate the association of haplotypes composed of KIR2DL4, LILRB1 and LILRB2, as well as haplotypes of their ligand -HLA-G -with endometriosis. The study included 590 Polish women: 276 patients suffering from endometriosis and 314 fertile control women. Patients were classified according to the severity of disease. We tested KIR2DL4 rs649216:T>C, LILRB1 rs41308748: G>A, LILRB1 rs1061680:T>C, LILRB2 rs383369:G>A, LILRB2 rs7247538:C>T, HLA-G rs1632947:G>A, HLA-G rs1233334:G>C/T, HLA-G rs371194629:ins/del. We used different methods: PCR, PCR-RFLP, and allelic discrimination technique with TaqMan SNP Genotyping Assays. The presence of 18 different haplotypes has been detected. We found the association of KIR2DL4-LILRB1-LILRB2 T-AT-AT, C-GT-GT and C-GT-GC haplotypes with susceptibility to endometriosis (p = 0.000855, p = 0.0145, p = 0.0105, respectively). In contrast, we observed protective effects of T-GT-GC (p = 0.00914) and T-AT-AC (p = 0.0371) haplotypes against disease. The A-C-del haplotype of HLA-G predisposed to disease (p = 0.012), while the G-C-del haplotype exerted a protective effect (p = 0.0149). In conclusion, these results support the role of tested polymorphisms and haplotypes of KIR2DL4, LILRB1, LILRB2, and HLA-G in susceptibility to endometriosis and its progression. HLA-G is an immune modulating molecule presented by fetal extravillous trophoblasts (EVTs) at the fetal-maternal interface to facilitate placentation and favorable pregnancy outcome. In addition, soluble isoforms of HLA-G have been found in the maternal circulation and amniotic fluid. Several polymorphisms are present in the 3-prime untranslated region (3´UTR) of the HLA-G gene. Since the 3´UTR is targeted by microRNAs, polymorphisms in this region may have an influence on the level of HLA-G expression, and consequently on pregnancy outcome. We studied 23 women with recurrent miscarriages (RM), defined as a history of three or more consecutive miscarriages, and 46 women with uncomplicated pregnancy (controls). Genomic DNA was isolated from EDTA blood to sequence a 699/713-bp fragment covering the 3´UTR of exon 8. Besides investigating single nucleotide polymorphisms in this region, we composed UTR haplotypes based on the combination of 8 SNPs.To study local HLA-G expression, tissue homogenates from term placentas were processed for mRNA quantification of HLA-G and PRG2 (EVT marker) by real-time qPCR. We found a higher frequency of the HLA-G 3'UTR 14 bp insertion allele in women with a history of RM compared to controls (82.6% versus 56.8%, p=0.035). This 14 bp HLA-G insertion has been shown to be related to lower soluble HLA-G levels. We did not find a difference in HLA-G haplotypes between women with RM and controls, neither did we find a difference in the HLA-G genotype and haplotype distribution of their children. Corresponding with this, HLA-G/PRG2 ratios at the mRNA level in term placentas did not differ between groups. In conclusion, women with RM more often have the 3'UTR 14 bp insertion present in Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wrocław, Poland, 2 Wroclaw Medical University, Wrocław, Poland Correspondence: piotr.lacina@iitd.pan.wroc.pl CD147 (also known as basigin, BSG) is a multifunctional transmembrane glycoprotein and a member of the immunoglobulin superfamily. It plays a major role in energy metabolism, specifically lactate transport. CD147 is overexpressed in many cancers. Its overexpression drives cancer cell survival and proliferation due to more efficient lactate transport, as malignant cells have greatly increased lactate production through the glycolytic pathway. Importance of CD147 remains unknown in acute myeloid leukemia (AML), although it is known that it is present in AML myeloblasts. This prompted us to investigate whether genetic polymorphism in the gene coding for CD147 has an effect on disease progression or survival in AML. Following an in silico analysis, we selected two CD147 single nucleotide polymorphisms (SNPs), rs4919859 and rs8259, expected to be located in transcription factor and microRNA binding sites, respectively. Genotyping was performed on DNA samples isolated from AML patients (n=93) and healthy individuals (n=216) using Taqman assays. We observed that patients with the rs4919859 C allele had lower overall survival than GG heterozygotes (p=0.016). The C allele prevailed in patients with CRP levels suggesting significant inflammation and bacterial invasion (>50 mg/l; p=0.067) and was slightly more frequent in patients with no CBFB-MYH11 rearrangement (p=0.087). It was also more common in patients older than 70 at diagnosis (p=0.026). The rs8259 A allele was significantly more frequently detected in patients with acute myeloblastic leukemia (subtypes M0, M1 and M2 in the FAB classification) than in other AML subtypes (p=0.007). We also observed that absence of rs8259 A tended to be more common in patients with very high WBC count (>30,000 cells/ul, p=0.075) as compared to patients with lower levels. Our results show that CD147 SNPs may affect the course of AML. Clinicians know that HLA typing does not make diagnosis by itself; however, here we describe how HLA typing can help physicians with prognosis with knowledge of s disease severity, different co-morbidities and relative risks. Starting from a meta-analytic approach, we undertook systematic searches on the main online repositories of the literature on the association between HLA alleles and 8 immunemediated pathologies: coeliac disease (CD), juvenile idiopathic arthritis (JIA), narcolepsy (N), acute anterior aveitis (AAU), birdshot chorioretinopathy (BC), Behcet disease (BD), sympathetic ophthalmia (SO), Vogt-Koyanagi-Harada (VKH). A multidisciplinary team (clinicians, immunogeneticists and biostatisticians) reviewed 4,312 works by clinical, laboratory and methodological criteria, and included 58 full articles describing case-control, cohort and cross-sectional studies for a total of 740 CD patients (943 controls), 2473 JIA (9543 controls), 2312 N (7802 controls), 172 AAU (315 controls), 65 BC (4342 controls), 1233 BD (2719 controls), 42 SO (101 controls), 157 VKH (387 controls). The following alleles are definitive markers of severe forms: B*27 in AAU, A*26 in BD, DQB1*06:02 in N, DRB1*04 in SO and VKH. The following alleles discriminate different forms: DRB1*11/08 for pauciarticular JIA, DRB1*16 for systemic JIA. Early onset related with B*51 in BD, DQB1*02:01 homozygosity in CD. Negative predictive value in first degree-relatives of CD patients who do not carry DQB1*02:01. Co-morbidities found for carriers of DRB1*04 who can show SO together with VKH and rheumatoid arthritis. A gene-dose effect augmenting disease risk has been found for DQB1*02:01 in CD, DQB1*06:02 in N, B*51 in BD. Clinical translation is the first goal for a biomedical researcher, consequently we think that the correlations between HLA alleles and prognostic factors can be really helpful to clinicians in predicting the course of the disease and identifying tailored therapeutic strategies. We will collect further information from a huge number of patients affected by the major immune-mediated diseases (including all major ethnic groups) with the aim of creating an easy-to-handle table that links the different HLA predisposing alleles with prognostic factors and personalized medicine. Non-classical human leukocyte antigen-G (HLA-G) expression promotes cancer invasiveness and metastatic progression. HLA-G can exist as cell surface molecule or in soluble forms (sHLA-G) including secreted or shed molecules or released molecules via extracellular vesicles (EVs). HLA-G expression is associated with regulatory elements targeting certain single nucleotide polymorphisms (SNPs) in the HLA-G 3´untranslated region (UTR) which arrange as haplotypes. In this study, we addressed the question how the HLA-G 3´UTR haplotypes and sHLA-G impact the clinical parameters and disease outcome in epithelial ovarian cancer (EOC). For this, we (i) sequenced the HLA-G 3Ú TR in histologically confirmed EOC patients encompassing nine SNPs located between +2960 and +3227, (ii) defined the 3´UTR haplotypes, (iii) inferred the total amount of sHLA-G (sHLA-Gtot) and vesicular sHLA-G (sHLA-GEV) and (iv) analyzed the relationship of sHLA-G and 3´UTR haplotypes to EOC. Levels of sHLA-Gtot and sHLA-GEV were significantly increased in EOC patients compared to healthy donors (p<0.0001). In EOC patients, elevated levels were associated with advanced disease stage (p<0.0001) mirroring the tumor burden. Strikingly, release of sHLA-GEV was promoted in EOC (p=0.0003) and strongly associated (p<0.01) with the presence of circulating tumor cells (CTC). Although sHLA-Gtot and sHLA-GEV levels were independent of HLA-G 3´UTR haplotypes, the homozygous UTR-2 genotype was associated with the presence of CTC prior to therapy (p=0.026). Patients carrying UTR-7 had a significantly improved overall survival (p=0.016) compared to UTR-7 negative ones. This study gives evidence that EOC burden is associated with an enhanced expression of sHLA-G and that EOC favors the release of vesicular sHLA-G. It seems that regulatory elements targeting SNPs specific for UTR-2 or UTR-7 haplotypes are operative in the course of disease of EOC.

Gaurav Sharma 1 , Gurvinder Kaur 1 , Narinder Kumar Mehra 1 , Surendra Kumar Sharma 1 1 All India Institute of Medical Sciences, Delhi, India Correspondence: drgauravsharma@aiims.ac.in Population specific genetic correlations influence HIV/AIDS vulnerability and are well supported by studies of specialized cohorts e.g. viremic controllers, elite controllers, exposed uninfected individuals, rapid and slow progressor individuals. The differential genetic influence of cytokine genes on immune response networks is known to affect the establishment and spread of HIV infection, however information is limited on Indian populations. Therefore, the present study was conducted to explore the largely unknown genetic influence of various pro (IL-1, IL-2, IFNγ etc.) and anti-inflammatory (IL-4, IL-10, TGF-β etc.) cytokine gene polymorphisms on HIV infection, in a North Indian population. A total of 157 HIV patients, 130 healthy and 27 exposed uninfected healthy controls were enrolled and informed consent was obtained. Genotyping was performed using a PCR-SSP based approach, to evaluate 22 functionally relevant single nucleotide polymorphisms in 13 cytokine and their receptor genes. Frequencies of most of the analyzed variants were statistically indifferent between HIV patients and uninfected healthy controls at the allelic, genotypic and haplotypic levels. Significantly higher allelic frequencies of IL-1α -889 T and IL-4 -1098 T were observed in HIV patients, while IL-1α -889 CC, IL-4 -1098 GG and IL-6 nt565 AA genotypes were observed significantly lower compared to healthy uninfected controls. This study represents the first report that highlights the influence of various cytokine gene polymorphisms in context of HIV infection in the North Indians. The MHCs of various non-human primates have been analyzed in recent decades, mostly because of their role as animal models in research on human diseases. The sequenced primate genes are assembled in the MHC-NHP database (www.ebi.ac.uk/ipd/mhc/group/NHP), which is part of the Immuno Polymorphism database (IPD). The NHP-part of the IPD covers Apes, Old and New World monkeys and comprises at present more than 7000 sequences of 51 different species. The first version was released online in March 2002. In 2017 an improved database was released, that introduced a new type of organization of data and a new submission pipeline, to keep up with the current processing of data. Since then we have seen the inclusion of more than 1,150 new alleles from over 1,450 submissions. This rapid increase in numbers represents approximately 13% of named alleles, which have been processed since the new release. In addition, the new version includes a multiple alignment tool that allows displaying single gene alignments, as well as inter-and intra-species gene alignments from all groups within the IPD-MHC database. The MHC-NHP database endeavors to give an up to date overview of annotated and quality-controlled MHC sequences, which Several mutations in the HFE gene are associated with hereditary haemochromatosis (HH). The majority of patients with HH are homozygous for the p.C282Y (Cys282Tyr) variant, but p.H63D (Hist63Asp) and p.S65C (Ser65Cys) have also been implicated. UK NEQAS for H&I have operated an HFE external quality assessment (EQA) scheme since 1999. Here we report the results of HFE testing from 2011-2017. Ten blood samples are distributed yearly. Participants may register for assessment of C282Y +/-H63D +/-S65C. Results reported by at least 75% of participants are taken as the consensus result for assessment. The number of participating laboratories has varied from 51-58. All labs tested for p.C282Y and p. H63D while 21-29 also reported results for p.S65C. The majority of labs tested using PCR-SSP, real time PCR, melt curve analysis or PCR-RFLP although other techniques (e.g. SBT) were also used. The 70 samples distributed included seven different 3-codon genotypes. 1-3 samples per year were homozygous for the p.C282Y mutation. A total of 9582 results were assessed over the seven-year period (3940 for codon 282, 3924 for codon 63, and 1718 for codon 65). There were 22 errors involving 15 different labs indicating an overall accuracy rate of 99.8%: nine errors for codon 282, eight errors for codon 63 and five for codon 65. Three of the codon 282 errors incorrectly assigned samples as homozygous for the p.C282Y mutation and four errors missed samples homozygous for p.C282Y. The number of errors had shown a decreasing trend from six in 2011/2012 to zero in 2015, however there were five errors in 2017. The causes of the incorrect results were reported for 21/22 errors; 11 were due to transcription errors, eight due to technical incidents (e.g. primer failures, gel loading errors) and two due to a sample mix up in 2017 involving two EQA samples. Although the overall accuracy rate is >99% the errors detected emphasize the need for ongoing participation in EQA schemes to help ensure high quality testing. Psoriasis is a systemic skin disease orchestrated by different immune cells such as T lymphocytes, dendritic cells and inflammatory cytokines. Despite the availability of several new systemic agents, methotrexate remains the gold standard for the treatment of moderate to severe psoriasis. The HLA-C*06:02 allele is a major susceptibility allele, which affects disease onset, clinical phenotype and severity of psoriasis. The frequency of this allele in patients varies from 10 to 77%, as reported in different populations. Despite the known role of C*06:02, little is known on its effect on methotrexate treatment. To our knowledge, two clinical studies so far reported a better response in patients carrying C*06. We examined the association between C*06:02 and response to methotrexate in a case-control study including 131 patients with plaque psoriasis and 164 healthy controls typed for HLA-B and C alleles. While 56% out of all patients responded to the treatment with methotrexate (RMTX), 44% were non-responders (NRMTX). Our own computer program based on two tailed Fisher exact test was used for statistical analysis. As expected, C*06:02 was significantly more frequent (f=49%) in patients with psoriasis compared to controls (p=7.000 -9 ). Surprisingly, the same allele C*06:02 was over-represented in both groups of patients ) and RMTX (p=2.010 -4 ) compared to controls, respectively. However, no significant difference in HLA-C*06:02 frequencies was observed between RMTX and NRMTX. On the other hand, the B*08 was significantly more frequent in the RMTX than in NRMTX (p=2.110 -3 ). Although C*06 and B*08 are not in linkage disequilibrium, we performed stratification by C*06:02, resulting in lower number of patients included in the analysis and the lower significance for B*08 (p=0.09). We conclude that further investigation is required to confirm the effects of HLA on the response to methotrexate and to establish the predictive value of this finding. The number of participating labs has varied from 7-11. With the exception of one lab that did not report results for KIR2DP1 and KIR3DP1, all labs reported the presence/ absence of all KIR genes. The majority of labs tested using PCR-SSP, but some tested using PCR-SSO. The 25 samples distributed included 12 different KIR genotypes. A total of 205 genotypes have been reported since the scheme was initiated. There have been five errors involving two labs, indicating an overall accuracy of 97.6%. There was one error in 2015 where KIR2DL3 was reported as positive (consensus negative), and four errors in 2016 where the same laboratory reported 4 samples as negative for KIR3DS1 (consensus positive). There were no errors in 2017. The introduction of new EQA schemes in line with clinical developments is crucial to help ensure the accuracy of all laboratory test results that can affect patient care. The errors identified in KIR genotyping in the past three years highlight the need for this EQA scheme. Correspondence: tamouza.ryad@gmail.com Bipolar Disorder (BD) is characterized by deregulated immune processes including chronic inflammation and autoimmunity both suggesting intricacies between innate and adaptive immune responses. As recent genome wide association studies (GWAS) clearly implicate the major histocompatibility complex (MHC) region in several psychiatric conditions including BD, we addressed here the potential influence of the human leukocyte antigen (HLA) genetic diversity on BD risk and/or clinical presentations thought to be related with inflammatory processes. Patients meeting DSM-IV criteria for BD and HC (475 and 195 respectively) were genotyped at medium resolution. Comparisons of genotype, allele frequencies and haplotype analysis were performed using Chi-square testing and Estihaplo software. We found that (i) the HLA-A*02~B*44~DRB1*07 sub-haplotype is less prevalent in BD as compared to HC (pc=2.4x10 -2 ), (ii) the 57.1 and the 8.1-derived ancestral haplotypes (HLA-A*02~B*57~Cw*06~DRB1*07~DQB1*09 and HLA-A*02~B*08~Cw*07) are both associated with rapid cycling (pc=1.9x10 -3 and 1.05x10 -2 respectively), (iii) the 8.1AHderived HLA class II-DRB*03~HLA-DQB1*02 subhaplotype is more frequent in BD patients with history of suicidal behavior (pc=2.1x10 -2 ) and (iv) the 7.1 AH-derived A*03~B*07~DRB1*15 sub-haplotype is more frequent in BD patients starting the disorder with an hypomanic episode (pc=8.5x10 -3 ), while the HLA-A*02~B*07~DRB1*15 subhaplotype is more frequent in BD patients presenting psychotic symptoms during the first episode (pc=4.0x10 -2 ). According to the well-known implication of the abovementioned HLA haplotypes and sub-haplotypes in the development of common immune disorders, our findings converge toward likely operating HLA-mediated pro-inflammatory processes in BD. Since 2012 we used One Lambda LabType HD exon 4-7 Supplemental Luminex technology (with WHO nomenclature and P/G codes) to perform HD HLA-A, -B, -C,-DRB1 and -DQB1 typing, to bone marrow registry donors and to new hematological patients. The aim of this study was to evaluate the recent Bag HISTO SPOT Extend kit performances by comparing them with our routine analysis. DNA