Improved Definition of Human Leukocyte Antigen Frequencies Among Minorities and Applicability to Estimates of Transplant Compatibility

An HLA map of the world: A comparison of HLA frequencies in 200 worldwide populations reveals diverse patterns for class I and class II

HLA frequencies show widespread variation across human populations. Demographic factors as well as selection are thought to have shaped HLA variation across continents. In this study, a worldwide comparison of HLA class I and class II diversity was carried out. Multidimensional scaling techniques were applied to 50 HLA-A and HLA-B (class I) as well as 13 HLA-DRB1 (class II) first-field frequencies in 200 populations from all continents. Our results confirm a strong effect of geography on the distribution of HLA class I allele groups, with principal coordinates analysis closely resembling geographical location of populations, especially those of Africa-Eurasia. Conversely, class II frequencies stratify populations along a continuum of differentiation less clearly correlated to actual geographic location. Double clustering analysis revealed finer intra-continental sub-clusters (e.g., Northern and Western Europe vs. South East Europe, North Africa and Southwest Asia; South and East Africa vs. West Africa), and HLA allele group patterns characteristic of these clusters. Ancient (Austronesian expansion) and more recent (Romani people in Europe) migrations, as well as extreme differentiation (Taiwan indigenous peoples, Native Americans), and interregional gene flow (Sámi, Egyptians) are also reflected by the results. Barrier analysis comparing D_ST and geographic location identified genetic discontinuities caused by natural barriers or human behavior explaining inter and intra-continental HLA borders for class I and class II. Overall, a progressive reduction in HLA diversity from African to Oceanian and Native American populations is noted. This analysis of HLA frequencies in a unique set of worldwide populations confirms previous findings on the remarkable similarity of class I frequencies to geography, but also shows a more complex development for class II, with implications for both human evolutionary studies and biomedical research.

Molecular Signatures of Natural Killer Cells in CMV-Associated Anterior Uveitis, A New Type of CMV-Induced Disease in Immunocompetent Individuals

Cytomegalovirus (CMV) causes clinical issues primarily in immune-suppressed conditions. CMV-associated anterior uveitis (CMV-AU) is a notable new disease entity manifesting recurrent ocular inflammation in immunocompetent individuals. As patient demographics indicated contributions from genetic background and immunosenescence as possible underlying pathological mechanisms, we analyzed the immunogenetics of the cohort in conjunction with cell phenotypes to identify molecular signatures of CMV-AU. Among the immune cell types, natural killer (NK) cells are main responders against CMV. Therefore, we first characterized variants of polymorphic genes that encode differences in CMV-related human NK cell responses (Killer cell Immunoglobulin-like Receptors (KIR) and HLA class I) in 122 CMV-AU patients. The cases were then stratified according to their genetic features and NK cells were analyzed for human CMV-related markers (CD57, KLRG1, NKG2C) by flow cytometry. KIR3DL1 and HLA class I combinations encoding strong receptor–ligand interactions were present at substantially higher frequencies in CMV-AU. In these cases, NK cell profiling revealed expansion of the subset co-expressing CD57 and KLRG1, and together with KIR3DL1 and the CMV-recognizing NKG2C receptor. The findings imply that a mechanism of CMV-AU pathogenesis likely involves CMV-responding NK cells co-expressing CD57/KLRG1/NKG2C that develop on a genetic background of KIR3DL1/HLA-B allotypes encoding strong receptor–ligand interactions.

Stem cell donor HLA typing improves CPRA in kidney allocation

The Organ Procurement and Transplantation Network (OPTN) Kidney Allocation System provides a priority to sensitized candidates based on the calculated panel reactive antibody (CPRA) value. The human leukocyte antigen (HLA) haplotype reference panel used for calculation of the CPRA by the United Network for Organ Sharing (UNOS), the OPTN contractor, has limitations. We derived a novel panel from the National Marrow Donor Program HLA haplotype data set and compared the accuracy of CPRA values generated with this panel (NMDP-CPRA) to those generated from the UNOS panel (UNOS-CPRA), using predicted and actual deceased donor kidney offers for a cohort of 24 282 candidates. The overall accuracy for kidney offers was similar using NMDP-CPRA and UNOS-CPRA. Accuracy was slightly higher for NMDP-CPRA than UNOS-CPRA for candidates in several highly sensitized CPRA categories, with deviations in linkage disequilibrium for Caucasians and the smaller size of the UNOS panel as contributing factors. HLA data derived from stem cell donors yields CPRA values that are comparable to those derived from deceased kidney donors while improving upon several problems with the current reference panel. Consideration should be given to using stem cell donors as the reference panel for calculation of CPRA to improve equity in kidney transplant allocation.

Development of a Calculated Panel Reactive Antibody Web Service with Local Frequencies for Platelet Transfusion Refractoriness Risk Stratification

BACKGROUND

Calculated panel reactive antibody (cPRA) scoring is used to assess whether platelet refractoriness is mediated by human leukocyte antigen (HLA) antibodies in the recipient. cPRA testing uses a national sample of US kidney donors to estimate the population frequency of HLA antigens, which may be different than HLA frequencies within local platelet inventories. We aimed to determine the impact on patient cPRA scores of using HLA frequencies derived from typing local platelet donations rather than national HLA frequencies.

METHODS

We built an open-source web service to calculate cPRA scores based on national frequencies or custom-derived frequencies. We calculated cPRA scores for every hematopoietic stem cell transplantation (HSCT) patient at our institution based on the United Network for Organ Sharing (UNOS) frequencies and local frequencies. We compared frequencies and correlations between the calculators, segmented by gender. Finally, we put all scores into three buckets (mild, moderate, and high sensitizations) and looked at intergroup movement.

RESULTS

2531 patients that underwent HSCT at our institution had at least 1 antibody and were included in the analysis. Overall, the difference in medians between each group's UNOS cPRA and local cPRA was statistically significant, but highly correlated (UNOS vs. local total: 0.249 and 0.243, ρ = 0.994; UNOS vs. local female: 0.474 and 0.463, ρ = 0.987, UNOS vs. local male: 0.165 and 0.141, ρ = 0.996; P < 0.001 for all comparisons). The median difference between UNOS and cPRA scores for all patients was low (male: 0.014, interquartile range [IQR]: 0.004-0.029; female: 0.0013, IQR: 0.003-0.028). Placement of patients into three groups revealed little intergroup movement, with 2.96% (75/2531) of patients differentially classified.

CONCLUSIONS

cPRA scores using local frequencies were modestly but significantly different than those obtained using national HLA frequencies. We released our software as open source, so other groups can calculate cPRA scores from national or custom-derived frequencies. Further investigation is needed to determine whether a local-HLA frequency approach can improve outcomes in patients who are immune-refractory to platelets.

Survival analysis of patients in the waiting list for kidney transplantation in terms of competing risks

Aim: to analyze the survival of patients on the waiting list for kidney transplantation and the results of transplantation depending on the duration of waiting.

Materials and methods. We performed a retrospective observational analysis that included 1,197 patients on the waiting list. The end point was exclusion from the waiting list (WL). The causes for exclusion (death, exclusion due to deterioration of the comorbid background or transplantation) were considered in terms of competing risks.

Results. In total, 72.5% of patients reached the end point: 21.1% of them died, 11% were excluded, and 40.4% underwent transplantation. Kaplan–Meier estimate showed that cumulative risk of death was 80.4% [95% CI 77.9; 88.6], of exclusion was 77.9% [95% CI 65.4; 88.2], of transplantation was 63.6% [95% CI 58.3; 69] after 10 years on the waiting list. However, such an assessment cannot be directly interpreted as a prediction of the relevant event risk of occurrence for the patient in the WL, because it does not take into account competing events. According to a balanced assessment of the competing risks (Fine and Gray estimate), cumulative incidence was 30.9% (95% CI 27.7; 34.2) for death, 18.2% [95% CI 15.5; 21.1] for exclusion and 49.4% [95% CI 46; 52.6%] for transplantation after 10 years on WL. The probability of transplantation was significantly higher than the risk of death up to and including 5 years of waiting (incidence rate ratio – IRR 1.769 [95% CI 1.098; 2.897]). When waiting 7 to 8 years, the probability of transplantation was less than the risk of death: IRR 0.25 (95% CI 0.093; 0.588; p = 0.0009). Of the 483 recipients, 61 died and 119 returned to dialysis. The risk of graft loss after 10 years was 68.5% [95% CI 57.5; 79.1] and the risk of death of a recipient with a functioning graft was 48.3% [95% CI 34.7; 63] according to Kaplan–Meier estimate. The cumulative incidence of the method was 30.8% [95% CI 23.3; 38.5%] and 55.7% [95% CI 46.6; 63.5%] according to Fine and Gray estimate, respectively. The risk of death after transplantation increases significantly when waiting for more than 6 years – IRR 4.325 [95% CI 1.649; 10.47], p = 0.0045 relative to a shorter waiting period. With an increase in the waiting period, the comorbid background (CIRS scale) deteriorates significantly, even adjusted for the initial patient condition: the partial correlation r = 0.735; p < 0.0001.

Conclusion. 1. In the context of competing risks, the Fine and Gray estimate gives a more balanced risk assessment compared to the Kaplan–Meier method. 2. Increasing the waiting time for transplantation significantly increases the risk of death of the candidate on the waiting list and reduces the probability of transplantation, as well as increases the risk of death of the recipient after transplantation. Apparently, this is mainly due to the deterioration of the comorbid background.

Strong Association of the HLA-DR/DQ Locus with Childhood Steroid-Sensitive Nephrotic Syndrome in the Japanese Population

Background Nephrotic syndrome is the most common cause of chronic glomerular disease in children. Most of these patients develop steroid-sensitive nephrotic syndrome (SSNS), but the loci conferring susceptibility to childhood SSNS are mainly unknown.Methods We conducted a genome-wide association study (GWAS) in the Japanese population; 224 patients with childhood SSNS and 419 adult healthy controls were genotyped using the Affymetrix Japonica Array in the discovery stage. Imputation for six HLA genes (HLA-A, -C, -B, -DRB1, -DQB1, and -DPB1) was conducted on the basis of Japanese-specific references. We performed genotyping for HLA-DRB1/-DQB1 using a sequence-specific oligonucleotide-probing method on a Luminex platform. Whole-genome imputation was conducted using a phased reference panel of 2049 healthy Japanese individuals. Replication was performed in an independent Japanese sample set including 216 patients and 719 healthy controls. We genotyped candidate single-nucleotide polymorphisms using the DigiTag2 assay.Results The most significant association was detected in the HLA-DR/DQ region and replicated (rs4642516 [minor allele G], combined P_allelic=7.84×10^-23; odds ratio [OR], 0.33; 95% confidence interval [95% CI], 0.26 to 0.41; rs3134996 [minor allele A], combined P_allelic=1.72×10^-25; OR, 0.29; 95% CI, 0.23 to 0.37). HLA-DRB1*08:02 (Pc=1.82×10^-9; OR, 2.62; 95% CI, 1.94 to 3.54) and HLA-DQB1*06:04 (Pc=2.09×10^-12; OR, 0.10; 95% CI, 0.05 to 0.21) were considered primary HLA alleles associated with childhood SSNS. HLA-DRB1*08:02-DQB1*03:02 (Pc=7.01×10^-11; OR, 3.60; 95% CI, 2.46 to 5.29) was identified as the most significant genetic susceptibility factor.Conclusions The most significant association with childhood SSNS was detected in the HLA-DR/DQ region. Further HLA allele/haplotype analyses should enhance our understanding of molecular mechanisms underlying SSNS.

Prevalence of pre-transplant anti-HLA antibodies and their impact on outcomes in lung transplant recipients

Background

Previous studies have suggested that antibodies against human leukocyte antigen (HLA) are associated with worse outcomes in lung transplantation. However, little is known about the factors associated with outcomes following lung transplantation in Asia. Accordingly, we investigated the prevalence of anti-HLA antibodies in recipients before transplantation and assessed their impact on outcomes in Korea.

Methods

A single-center retrospective study was conducted. The study included 76 patients who received a lung transplant at a tertiary hospital in South Korea between January 2010 and March 2015.

Results

Nine patients (11.8%) had class I and/or class II panel-reactive antibodies greater than 50%. Twelve patients (15.8%) had anti-HLA antibodies with a low mean fluorescence intensity (MFI, 1000–3000), 7 (9.2%) with a moderate MFI (3000–5000), and 12 (15.8%) with a high MFI (> 5000). Ten patients (13.2%) had suspected donor-specific antibodies (DSA), and 60% (6/10) of these patients had antibodies with a high MFI. In an analysis of outcomes, high-grade (≥2) primary graft dysfunction (PGD) was more frequent in patients with anti-HLA antibodies with moderate-to-high MFI values than in patients with low MFI values (39.4% vs. 14.0%, p = 0.011). Of 20 patients who survived longer than 2 years and evaluated for pBOS after transplant, potential bronchiolitis obliterans syndrome (pBOS) or BOS was more frequent in patients with anti-HLA antibodies with moderate-to-high MFI than in patients with low MFI, although this difference was not statistically significant (50.0% vs. 14.3%, p = 0.131).

Conclusions

The prevalence of anti-HLA antibodies with high MFI was not high in Korea. However, the MFI was relatively high in patients with DSA. Anti-HLA antibodies with moderate-to-high MFI values were related to high-grade PGD. Therefore, recipients with high MFI before lung transplantation should be considered for desensitization and close monitoring.

HLA-A, -B, -C and -DRB1 allele and haplotype frequencies in the Macedonian population based on a family study

AIM

The aim of this study was to determine HLA allele and 2-, 3- and 4-loci haplotype frequencies in a sample from Macedonian population with defined haplotypes based on family history.

MATERIAL AND METHODS

We analysed 286 unrelated individuals with Macedonian origin, parents of patients who needed stem cell transplantation, in the period of 01.01.2003 till 31.12.2016. Allele and haplotype frequencies, as well as Hardy-Weinberg equilibrium were calculated using the Arlequin3.5 software. Population comparison was calculated using the PHYLIP software.

RESULTS

We identified 18 HLA-A, 26 HLA-B, 13 HLA-C and 13 HLA-DRB1 allele group families. The most frequent allele groups in our population were HLA-A*02 (29.0%), HLA-A*24 (13.8%), HLA-B*35 (16.1%), HLA-B*51 (14.7%), HLA-B*18 (14.7%), HLA-C*07 (27.9%), HLA-DRB1*11 (25.5%) and HLA-DRB1*16 (14.8%). The most frequent four loci haplotype was HLA-A*01-B*08-C*07-DRB1*03 (2.7%). Our comparison showed that the Macedonian population is closely related to the neighbouring countries in the Balkan Peninsula.

CONCLUSION

This study provides data about the HLA diversity in the Macedonian population, which can be very important in the process of unrelated donor search, and in addition yields control group for future disease association studies in our population.

Human Leukocyte Antigen Sensitization in Solid Organ Transplantation: A Primer on Terminology, Testing, and Clinical Significance for the Apheresis Practitioner

The human leukocyte antigen (HLA) system is an important immunologic barrier that must be considered for successful solid organ transplantation. Formation of donor-specific HLA antibodies in solid organ transplantation is an important cause of allograft injury and may contribute to recipient morbidity and mortality. Therapeutic plasma exchange is often requested to lower HLA antibody levels prior to or after transplantation and for management of HLA antibodies in the context of organ rejection. In this review, we summarize the current terminology, laboratory testing, and clinical significance of HLA sensitization in the solid organ transplant population. Furthermore, to illustrate applications of HLA testing in clinical practice, we summarize our own lung and kidney institutional protocols for managing HLA antibodies in the peri-transplant setting.

Concordance and discordance in anti-HLA antibody testing

BACKGROUND

Correct identification of the specificity of antibodies directed against HLA using single antigen Luminex beads (SALB) is essential in current HLA laboratory practice for transplantation. The aim of this study was to investigate the magnitude of concordance and discordance among laboratories in testing for anti-HLA antibodies using SALB.

METHOD

35 sera were distributed by the ASHI Proficiency Testing Program to HLA laboratories worldwide. We analyzed 4335 test results submitted between April 2010 and April 2013 by participating laboratories.

RESULTS

SALB was used by approximately 94% of the participating laboratories, yet concordant assignment of antibody specificity was imperfect. For each serum, the assignment of an average of 10 antibody specificities was discordant. Disagreement was observed for antibodies directed against common as well as uncommon antigens. The assignment of an average of 15 antibody specificities in each "positive" serum appeared to be influenced by vendor-dependent causes. Inter-vendor concordance was lower than intra-vendor concordance, indicating that vendor dependent factors may be a central cause for disagreement.

CONCLUSIONS

Our study illustrates the prevalence of concordance and discordance, also affected by unpremeditated causes, in reporting SALB antibody results. Insufficient concordance and standardization in antibody testing may have practical implications for organ allocation and organ sharing programs.

HLA-B50 polymorphism in the Saudi population

The HLA-B50 serologic family is very frequent in people of Arabic origin. In Saudi Arabia, HLA-B50 is the most frequent HLA-B allele. The aim of this study was to investigate the distribution of HLA-B50 alleles in healthy Saudi individuals. A total of 162 healthy Saudi individuals were selected based on low-resolution HLA typing. DNA samples were typed by sequence-based typing method for exons 2, 3 and 4 of the HLA-B locus (Genome Diagnostics B.V.). The HLA-B*50 alleles were analysed using SBT engine software. HLA-B*50:01:01 was found in 161 of 162 individuals (99.4%), while HLA-B*50:09 was found in one individual (0.6%). HLA-B*50:01:01 is the most common HLA-B50 allele in Saudi Arabia.

Identification of class I HLA T cell control epitopes for West Nile virus

The recent West Nile virus (WNV) outbreak in the United States underscores the importance of understanding human immune responses to this pathogen. Via the presentation of viral peptide ligands at the cell surface, class I HLA mediate the T cell recognition and killing of WNV infected cells. At this time, there are two key unknowns in regards to understanding protective T cell immunity: 1) the number of viral ligands presented by the HLA of infected cells, and 2) the distribution of T cell responses to these available HLA/viral complexes. Here, comparative mass spectroscopy was applied to determine the number of WNV peptides presented by the HLA-A*11:01 of infected cells after which T cell responses to these HLA/WNV complexes were assessed. Six viral peptides derived from capsid, NS3, NS4b, and NS5 were presented. When T cells from infected individuals were tested for reactivity to these six viral ligands, polyfunctional T cells were focused on the GTL9 WNV capsid peptide, ligands from NS3, NS4b, and NS5 were less immunogenic, and two ligands were largely inert, demonstrating that class I HLA reduce the WNV polyprotein to a handful of immune targets and that polyfunctional T cells recognize infections by zeroing in on particular HLA/WNV epitopes. Such dominant HLA/peptide epitopes are poised to drive the development of WNV vaccines that elicit protective T cells as well as providing key antigens for immunoassays that establish correlates of viral immunity.

Human leukocyte antigen profiles of latin american populations: differential admixture and its potential impact on hematopoietic stem cell transplantation

The outcome of hematopoietic stem cell transplantation (HSCT) is shaped by both clinical and genetic factors that determine its success. Genetic factors including human leukocyte antigen (HLA) and non-HLA genetic variants are believed to influence the risk of potentially fatal complications after the transplant. Moreover, ethnicity has been proposed as a factor modifying the risk of graft-versus-host disease. The populations of Latin America are a complex array of different admixture processes with varying degrees of ancestral population proportions that came in different migration waves. This complexity makes the study of genetic risks in this region complicated unless the extent of this variation is thoroughly characterized. In this study we compared the HLA-A and HLA-B allele group profiles for 31 Latin American populations and 61 ancestral populations from Iberia, Italy, Sub-Saharan Africa, and America. Results from population genetics comparisons show a wide variation in the HLA profiles from the Latin American populations that correlate with different admixture proportions. Populations in Latin America seem to be organized in at least three groups with (1) strong Amerindian admixture, (2) strong Caucasian component, and (3) a Caucasian-African gradient. These results imply that genetic risk assessment for HSCT in Latin America has to be adapted for different population subgroups rather than as a pan-Hispanic/Latino analysis.

Evaluation of a non-viral vaccine in smallpox-vaccinated individuals and immunized HLA-transgenic mice

The current poxvirus vaccine is associated with rare, but serious adverse events. Therefore, we investigated a non-replicating approach to vaccine design. Peptides encoding potential HLA-binding motifs were derived from the orthopoxvirus genes, D8L, A27L, and C12L (the IL-18-binding protein [vIL18BP105]), all of which are preserved among poxviruses that infect humans, and which may be a target of host immunity. The peptides were tested with poxvirus-vaccinated human PBMC and serum for eliciting memory responses, as well as with splenocytes and serum from peptide-immunized, human HLA-DR04 transgenic (HLA tg) mice. vIL18BP105 induced 5-fold proliferation of vaccinated-donor PBMC over non-vaccinated (P<0.001), including IL-2-producing CD8+ cells. Serum IgG recognizing vIL18BP105 was detected (P<0.002 vs non-vaccinated) by ELISA. Viral peptides were conjugated to the HLA-targeting mAb, L243, for immunization of HLA tg mice. Splenocytes from vIL18BP105-L243-immunized mice proliferated upon exposure to vIL18BP105 (P<0.001). Proliferating splenocytes were interferon-γ-producing CD4(+)CD45RA(neg). vIL18BP105-L243-immunized mice generated IgG more rapidly than free-peptide-immunized mice. Peptide-specific antibody was also detected when different L243-peptide conjugates were combined. vIL18BP, by eliciting human memory responses, is a viable antigen for inclusion in a virus-free vaccine. The immunogenicity of peptides was boosted by conjugation to L243, whether administered alone or combined.

Application of calculated panel reactive antibody using HLA frequencies in Koreans

Background

Introduction of the Luminex panel reactive antibody (PRA)-single antigen (SA) assay has increased the detection rates of unacceptable antigens in sensitized patients; the calculated PRA (CPRA) level represents the percentage of actual organ donors that express 1 or more of these unacceptable antigens. We developed a CPRA calculator based on the HLA frequencies in Koreans to measure sensitization levels in Korean patients.

Methods

To develop the calculator, we obtained the HLA-A, HLA-B, and HLA-DR phenotypes of 1,622 Koreans, and compared these with previously reported frequencies in Koreans. Sera from patients awaiting kidney transplantation were tested for HLA antibodies by Luminex PRA-screen, PRA-identification (ID), and PRA-SA assays. The measured %PRA from the PRA-screen (N=55) and PRA-ID (N=71) were compared to the %CPRA for the unacceptable antigens obtained from PRA-SA.

Results

Phenotype frequencies used for the CPRA calculator agreed with previously reported data. The concordance rates among the 3 PRA methods for the detection of class I and class II antibodies were 76.1-81.8% (kappa, 0.519-0.636) and 72.7-83.6% (0.463-0.650), respectively. For the detection of broadly sensitized sera (>50% or >80%), the concordance rates were over 80%. In sera with 80-100% CPRA, 91.7% and 94.4% of the samples had concordant results (80-100% PRA) in the PRA-screen and PRA-ID assay, respectively.

Conclusions

Although further clinical studies are required to confirm the benefits of CPRA values, adoption of CPRA analysis based on HLA frequencies in Koreans may be useful for sensitization measurements and organ-allocation algorithms.

Calculated PRA: initial results show benefits for sensitized patients and a reduction in positive crossmatches

The calculated panel reactive antibody (CPRA), which is based upon unacceptable HLA antigens listed on the waitlist form for renal transplant candidates, replaced PRA as the measure of sensitization among US renal transplant candidates on October 1, 2009. An analysis of the impact of this change 6 months after its implementation shows an 83% reduction in the number of kidney offers declined nationwide because of a positive crossmatch. The increasing acceptance and utilization of unacceptable HLA antigens to avoid offers of predictably crossmatch-positive donor kidneys has increased the efficiency of kidney allocation, resulting in a significant increase in the percentage of transplants to broadly sensitized (80+% PRA/CPRA) patients from 7.3% during the period 07/01/2001-6/30/2002 to 15.8% of transplants between 10/1/09-3/31/10. The transplant rates per 1000 active patient-years on the waitlist also increased significantly for broadly sensitized patients after October 1, 2009. These preliminary results suggest that 'virtual' positive crossmatch prediction based on contemporary tools for identifying antibodies directed against HLA antigens is effective, increases allocation efficiency and improves access to transplants for sensitized patients awaiting kidney transplantation.

Primary sclerosing cholangitis in genetically diverse populations listed for liver transplantation: unique clinical and human leukocyte antigen associations

Primary sclerosing cholangitis (PSC) is well characterized in European populations. We aimed to characterize clinical characteristics and human leukocyte antigen (HLA) associations in a population of European American, Hispanic, and African American PSC patients listed for liver transplantation (LT). Population-stratified demographic, clinical, and HLA data from 6767 LT registrants of the United Network for Organ Sharing who had a diagnosis of PSC (4.7% of the registrants) were compared to data from registrants with other diagnoses. Compared to European Americans and Hispanics, African Americans were significantly younger (46.6 ± 13.7, 42.3 ± 15.9, and 39.7 ± 13.1 years, respectively; P = 0.002) and were listed with a higher Model for End-Stage Liver Disease score (15.2 ± 7.5, 14.9 ± 7.6, and 18.1 ± 9.3, respectively; P = 0.001); they were also less frequently noted to have inflammatory bowel disease in comparison with European Americans (71.4% versus 60.5%, P < 0.01). In multivariate analysis, African origin was a significant factor associated with listing for LT with PSC (odds ratio with respect to European Americans = 1.325, 95% confidence interval = 1.221-1.438). HLA associations in European Americans, Hispanics, and African Americans with PSC versus alcoholic liver disease were detected for HLA-B8, HLA-DR13, and protective HLA-DR4. However, HLA-DR3, which is in linkage disequilibrium with HLA-B8, showed associations only in European Americans and Hispanics. In conclusion, African Americans with PSC who are listed for LT differ clinically from European Americans and Hispanics. The association with HLA-B8 but not HLA-DR3 in African Americans should make possible the refinement of the HLA associations in PSC.

Statistical resolution of ambiguous HLA typing data

High-resolution HLA typing plays a central role in many areas of immunology, such as in identifying immunogenetic risk factors for disease, in studying how the genomes of pathogens evolve in response to immune selection pressures, and also in vaccine design, where identification of HLA-restricted epitopes may be used to guide the selection of vaccine immunogens. Perhaps one of the most immediate applications is in direct medical decisions concerning the matching of stem cell transplant donors to unrelated recipients. However, high-resolution HLA typing is frequently unavailable due to its high cost or the inability to re-type historical data. In this paper, we introduce and evaluate a method for statistical, in silico refinement of ambiguous and/or low-resolution HLA data. Our method, which requires an independent, high-resolution training data set drawn from the same population as the data to be refined, uses linkage disequilibrium in HLA haplotypes as well as four-digit allele frequency data to probabilistically refine HLA typings. Central to our approach is the use of haplotype inference. We introduce new methodology to this area, improving upon the Expectation-Maximization (EM)-based approaches currently used within the HLA community. Our improvements are achieved by using a parsimonious parameterization for haplotype distributions and by smoothing the maximum likelihood (ML) solution. These improvements make it possible to scale the refinement to a larger number of alleles and loci in a more computationally efficient and stable manner. We also show how to augment our method in order to incorporate ethnicity information (as HLA allele distributions vary widely according to race/ethnicity as well as geographic area), and demonstrate the potential utility of this experimentally. A tool based on our approach is freely available for research purposes at http://microsoft.com/science.