BLOOD GROUP GENOMICS Red blood cell antigen genotype analysis for 9087 Asian, Asian American, and Native American blood donors Meghan Delaney,1,2 Samantha Harris,1 Askale Haile,3 Jill Johnsen,4,5 Gayle Teramura,1 and Karen Nelson1
BACKGROUND: There has yet to be a comprehensive analysis of blood group antigen prevalence in Asian Americans and Native Americans. There may be ethnic differences in blood group frequencies that would result in clinically important mismatches through transfusion. STUDY DESIGN AND METHODS: Blood donors who self-identified as Asian or Native American were tested using a single-nucleotide polymorphism (SNP) DNA array (HEA BeadChip kit, Bioarray Solutions Ltd) that predicts expression of 38 human erythrocyte antigens (HEAs) and by serology for ABO, D, C, M, N, Jka, and Jkb. The prevalence of blood group antigens was compared to published European prevalence. Discrepancies between SNP-predicted and serologydetected antigens were tallied. RESULTS: A total of 9087 blood donors were tested from nine Asian and Native American heritages. The predicted prevalence of selected antigens in the RHCE, JK, FY, MNS, LU, CO, and DO blood group systems were variable between Asian populations, but overall not significantly different than Europeans. Compared to European frequencies, Kell blood group allele frequencies were significantly different in the Chinese, Native American, Hawaiian/Pacific Islander, South Asian, and Southeast Asian heritage blood donors; Diego antigens Dia and Dib were different in donors of Native American and South Asian ancestries (p < 0.05). Of the donors tested, 4.5% showed a SNP-serology discrepancy that segregated within specific ethnic groups. CONCLUSION: This study provides HEA allele frequency and antigen prevalence data in a cohort of Asian and Native Americans donors. Several ethnic groups exhibited differences in HEA frequencies compared to Europeans. Genotype-serotype discrepancies were detected in all systems studied.
T
oday, blood centers and hospitals have access to genetic techniques to determine the predicted red blood cell (RBC) phenotype with molecular testing.1,2 Critical to the implementation of technological advancements in RBC typing technology is the understanding of blood group gene frequencies within different ethnic groups.3,4 The ethnic background of a transfusion recipient and blood donor pair can affect transfusion outcomes, as is evidenced by African American patients with sickle cell disease.5 This group of patients has the highest rate of RBC alloimmunization (up to 47%), which is at least partly attributed to the high degree of European American blood donors in the United States.6-8 The resulting minor blood group antigen mismatches lead to RBC alloantibody formation over the course of years of transfusion therapy. Study of the extended blood group phenotype/genotype in the African descent population has led to improvements in transfusion care of African descent transfusion recipients.9-13
ABBREVIATIONS: HEA 5 human erythrocyte antigen; SNP(s) 5 single-nucleotide polymorphism(s). From the 1Specialty Diagnostics, Red Cell Genomics Laboratory, and the 3Red Cell Reference Laboratory, Puget Sound Blood Center; the 2Department of Laboratory Medicine and the 5Department of Medicine, Division of Hematology, University of Washington; and the 4Puget Sound Blood Center Research Institute, Seattle, Washington. Address reprint requests to: Meghan Delaney, DO, MPH, Puget Sound Blood Center, 921 Terry Avenue, Seattle, WA 98104; e-mail: [email protected]. Funded by the Telemedicine & Advanced Technology Research Center (TATRC); US Army Medical Research Acquisition Activity (UWAMRAA), W81XWH-06-0052; and GSA.gov Contract GST0906BHM0529 Received for publication April 14, 2014; revision received March 14, 2015; and accepted April 10, 2015. doi:10.1111/trf.13163 C 2015 AABB V
TRANSFUSION 2015;55;2369–2375 Volume 55, October 2015 TRANSFUSION 2369
DELANEY ET AL.
TABLE 1. ABO and D percent phenotype prevalence from serologic testing (n 5 8453 with ABO and D results available)* Number of samples with ABO/D serology
Ethnicity Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Total European/Caucasian (published)2
1699 1280 897 1003 999 260 472 915 928 8453
O
A
B
AB
D1
D–
0.437 0.448 0.536 0.325 0.307 0.554 0.451 0.352 0.408 0.413 0.44
0.255 0.266 0.349 0.376 0.330 0.323 0.337 0.261 0.212 0.293 0.43
0.245 0.234 0.089 0.213 0.258 0.088 0.148 0.311 0.303 0.228 0.09
0.064 0.052 0.026 0.086 0.104 0.035 0.064 0.075 0.077 0.067 0.04
0.992 0.989 0.876 0.985 0.995 0.892 0.977 0.927 0.971 0.965 0.85
0.008 0.011 0.124 0.015 0.005 0.108 0.023 0.073 0.029 0.035 0.15
* Percentages are calculated within each ethnicity code (per row).
However, recent evidence supports that even racially matched blood transfusions in the sickle cell disease population do not protect from allosensitization and that genotypic matching may be the next step for improved patient outcomes.14 Blood banks must be prepared to support the transfusion needs of all patients and thus must be knowledgeable about the prevalence of blood group antigens in the population(s) they serve. Some ethnic backgrounds are disproportionately affected by genetically derived medical illnesses, such as sickle cell disease, thalassemia, and glucose-6-phosphate dehydrogenase deficiency.15 Manufacturers of RBC panels must develop and sustain reagents capable of detection of antibodies directed to blood group antigens that are represented at different frequencies in certain ethnic populations, such as Jsa, V, VS, and Mia. There has yet to be a comprehensive analysis of blood group antigens drawn from a large cohort of Asian American and Native American ancestral background people in the United States. Due to our center’s geographical location, we serve a large population of blood donors and patients of Asian American and Native American heritage. An extensive study was undertaken to better understand the blood type distributions in these populations and to test the robustness of our molecular blood typing technologies. RBC genotyping and blood type serology for selected blood group systems was performed in a large cohort of self-identified Asian American and Native American blood donors. Blood group allele frequencies were compared to published European ancestry RBC antigen prevalence, and concordance between serologic phenotyping and genotyping was determined for the RBC antigens C, M, N, Jka, and Jkb.
MATERIALS AND METHODS Discarded blood samples were collected between September 2007 and July 2010 from volunteer blood donors who 2370 TRANSFUSION Volume 55, October 2015
consented and self-identified to be of ethnic heritages that were Asian American or Native American. Samples from donors who self-selected as African American (choices: African, African American, Caribbean, South or Central American), European/Caucasian (choices: Middle East or North Coast Africa, North American, or European), Hispanic/South American (choices: Caribbean, Mexican or Chicano, South or Central American), multiracial, other, or “do not ask” were excluded. Donors were tested one time; repeat donors were not tested. A total of 8453 donor samples were from the Puget Sound Blood Center and 634 samples were from the Blood Bank of Alaska. Study approval was obtained from the Western Institutional Review Board. DNA isolation from EDTA peripheral blood was done using a DNA purification kit (Puregene blood kit, Qiagen, Valencia, CA) and an automated DNA extraction kit (Bee Robotics blood DNA extraction kit, QuatroProbe DNA robot, GTI, Waukesha, WI). Blood antigen molecular testing was carried out using an in-vitro diagnostic test for molecular determination of the allelic variants that indicate human erythrocyte antigen (HEA) phenotypes (HEA BeadChip kit, Bioarray Solutions Ltd, Immucor, Warren, NJ).16 Results were rendered by Web-based software (BASIS, Bioarray Solutions Ltd). Indeterminate results were provided when the calculation for the specific antigen signal intensity is too close to the cutoff to provide high confidence results.16 Each indeterminate result was investigated by first repeating the assay, as suggested by the manufacturer; if indeterminate results were again obtained, the authors, in collaboration with scientists from the manufacturer of the HEA BeadChip utilized relative fluorescence intensities to make calls. If an indeterminate result was not able to be resolved, they were recorded as such. Discrepancy was defined as either positive genotype results with negative serology, or negative single-nucleotide polymorphism (SNP) result and positive serology.
ASIAN AMERICAN RBC GENOTYPE
TABLE 2. RBC antigen allele frequencies by SNP analysis in blood donors of different self-identified Asian or Native American heritages and European/Caucasian percent prevalence of blood antigens* C
E
c
e
K
k
Kpa
Kpb
Jsa
Jsb
Ethnicity
Total
RH*2
RH*3
RH*4
RH*5
KEL*1
KEL*2
KEL*3
KEL*4
KEL*6
KEL*7
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.895 0.923 0.645 0.864 0.816 0.654 0.825 0.842 0.900 0.818 0.68
0.395 0.257 0.411 0.459 0.514 0.520 0.348 0.196 0.308 0.379 0.29
0.502 0.395 0.815 0.582 0.624 0.804 0.592 0.584 0.482 0.598 0.80
0.953 0.978 0.932 0.939 0.893 0.910 0.952 0.987 0.966 0.946 0.98
0.002 0.011 0.059 0.014 0.004 0.053 0.025 0.014 0.007 0.021 0.09
1.000 1.000 0.999 1.000 1.000 0.998 1.000 1.000 0.999 1.000 0.998
0.001 0.001 0.013 0.014 0.000 0.008 0.000 0.002 0.000 0.004 0.02
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
0.003 0.000 0.000 0.000 0.000 0.002 0.000 0.000 0.000 0.001 < 0.0001
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
Fya
Fyb
Jka
Jkb
M
N
S
s
Lua
Lub
Ethnicity
Total
FY*1
FY*2
JK*1
JK*2
MNS*1
MNS*2
MNS*3
MNS*4
LU*1
LU*2
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.982 0.965 0.760 0.973 0.992 0.860 0.925 0.871 0.968 0.922 0.66
0.128 0.270 0.705 0.292 0.171 0.576 0.462 0.563 0.281 0.383 0.83
0.697 0.721 0.758 0.728 0.740 0.786 0.779 0.820 0.764 0.755 0.77
0.770 0.744 0.751 0.762 0.784 0.741 0.648 0.675 0.712 0.732 0.74
0.801 0.790 0.839 0.786 0.759 0.892 0.835 0.852 0.860 0.824 0.78
0.652 0.701 0.635 0.692 0.732 0.554 0.642 0.631 0.607 0.650 0.72
0.096 0.163 0.503 0.166 0.105 0.490 0.246 0.487 0.214 0.274 0.55
0.988 0.980 0.911 0.983 0.987 0.899 0.967 0.880 0.947 0.949 0.89
0.002 0.013 0.057 0.010 0.005 0.043 0.035 0.004 0.000 0.019 0.08
0.999 1.000 0.999 1.000 0.999 0.997 0.998 0.999 1.000 0.999 0.998
Dia
Dib
Coa
Cob
Doa
Dob
Hy
Jo(a)
LWa
LWb
Sc1
Sc2
Ethnicity
Total
DI*1
DI*2
CO*1
CO*2
DO*1
DO*2
DO*4
DO*5
LW*5
LW*7
SC*1
SC*2
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiin/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.044 0.010 0.019 0.079 0.105 0.010 0.010 0.003 0.022 0.034 0.0001
0.992 0.995 0.997 0.993 0.989 0.997 0.998 0.994 0.993 0.994 1.00
1.000 1.000 0.999 1.000 0.999 0.998 1.000 0.998 1.000 0.999 0.995
0.003 0.010 0.068 0.016 0.007 0.039 0.021 0.012 0.007 0.020 0.10
0.206 0.228 0.591 0.254 0.209 0.588 0.383 0.623 0.334 0.380 0.67
0.987 0.978 0.858 0.977 0.987 0.839 0.960 0.841 0.944 0.930 0.82
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1.000 0.999 1.000 1.000 1.000 1.000 1.000 0.999 0.998 1.000 1.00
0.000 0.004 0.005 0.000 0.000 0.010 0.000 0.000 0.000 0.002 < 0.01
1.000 0.999 0.999 1.000 1.000 1.000 1.000 0.999 0.997 0.999 > 0.99
0.001 0.002 0.004 0.003 0.004 0.002 0.000 0.002 0.000 0.002 0.01
* Note: Kpa, Kpb, Jsa, Jsb were introduced with the second generation (Version 1.2) of the genomic testing platform (n 5 4962).
For the first 4125 samples, the HEA BeadChip Version 1.1 platform was utilized, which assigns 34 blood group antigens (in 11 systems) using 19 SNPs. During the course of the study, the HEA BeadChip platform Version 1.2 became available, which assigns 38 blood group antigens in 11 systems using 24 SNPs. The Version 1.2 platform included five additional SNPs: two for additional antigens Kpa, Kpb, Jsa, and Jsb (KEL*03, KEL*04, KEL*06, KEL*07 alleles) and three SNPs to improve the coverage of RHCE designations (733C>G, 1006G>T, 109Ins).16 This version was employed for the subsequent 4962 samples. The genotype-predicted blood group antigen results were grouped by ethnic heritage selected by the blood donor (Microsoft Excel, Microsoft Corp., Redmond, WA). Parallel serologic antigen typing was
conducted for C (Seraclone human monoclonal, MS24, Bio-Rad/BioTest, Hercules, CA), Jka, Jkb (polyclonal, Immucor Gamma, Norcross, GA), M, and N (murine monoclonal, gamma clone by tube test, Immucor Gamma; and Seraclone murine monoclonal, BS57 and BS41, Biotest). Concordance was calculated for these antigens between genotype-predicted blood type and serology. Genotype-serology discrepancies found for C, Jka, Jkb, M, and N were tabulated by ethnic heritage. ABO and D serologic testing was carried out using a blood grouping reagent system (Olympus PK7200/PK7300 Beckman Coulter, Fullerton, CA; and Gallileo, Immucor). Using chi-square analysis, the prevalences of the blood group antigens in each ethnic group were compared to Volume 55, October 2015 TRANSFUSION 2371
DELANEY ET AL.
TABLE 3. Samples found to have M and N serology and genotype discrepancy, with Glycophorin A (GYPA) and Glycophorin B (GYPB) genotyping results Number of samples 3 31 153 1 2 3 3 3 2 1 2
Serology
Predicted antigen expression (genotype, by SNP)
GYPA 59C>T
GYPB 143T>C
GYPB Ex5 230C>T
GYPB In5 15g>t
N1 N1 N1 N1 N– N– N– M1 M– M– M–
N– N– N– N– N1 N1 N1 M– M1 M1 M1
C/C C/C C/C C/C C/T C/T C/T T/T C/T C/T C/T
T/T T/C C/C C/C T/C T/T C/C C/C T/C T/T C/C
C/C C/C C/C C/T C/C C/C C/C C/C C/C C/C C/C
G/G G/G G/G G/G G/G G/G G/G G/G G/G G/G G/G
those of published prevalence of antigens in European/ Caucasian ancestry people to identify significant differences, or mismatches, which may be encountered in transfusion.2 p values were considered significant if they were less than 0.05 and not significant if more than 0.05. SNP-serology concordance rates were calculated using total typed samples with both SNP data and serology.
RESULTS A total of 9087 blood donors from nine populations of Asian American and Native American heritage were selected for study, representing 4.1% of total blood donors collected over 35 months. A total of 8955 (98.5%) were serologically tested in parallel for five antigens, C, M, N, Jka, and Jkb. The ABO, D (serology), and Genotype-predicted blood group antigen prevalence(s) were calculated (Tables 1 and 2). There were nine rare blood donors found (defined as being negative for a “highincidence” antigen < 1:1000). These donors had the following types: one Co(a–) (Native American), six Di(b–) (Korean [4], Japanese [1], Southeast Asian [1]), two k– (Native American [1], Alaska Native/Aleut [1]). Additionally, there were 589 that were uncommon blood donors (defined as lacking more than five common blood group antigens).
RhCE Although the predicted prevalence of expression of the RhCE (C, c, E, e) blood group antigens were variable between ethnic groups, overall they were not significantly different than those published in European/Caucasian populations (data not shown, p 5 0.9). Using the SNPs (733C>G, 1006G>T) that are associated with expression of V and VS antigens in the Rh blood group system, three donors were found that are predicted to be VS1 and V– (one Chinese, one Alaska Native/Aleut, one Korean). Fifteen donors were found to be positive for both VS and V (six American Indian, three Alaska Native/Aleut, two Chinese, two Filipino, one Korean, one Hawaiian/Pacific Islander). There were 25 donors with discrepancies between C serology and genotype-predicted C expression. Eighty-six donors had indeterminate C genotyping results (307C>T, 109Ins); 79 were positive and seven were negative by serology. For c, e, and E there were 35, 14, and 18 indeterminate genotype results, 307C>T, 676G>C), respectively.
Kell, Kidd, and Duffy The predicted frequency of expression for the Kell blood group antigens (K, k, Kpa, Kpb, Jsa, Jsb) was found to be different (p < 0.05) than published European/Caucasian frequencies for donors of Chinese, Native American, Hawaiian/Pacific Islander, South Asian, and Southeast
TABLE 4. Concordance for five blood group antigens between using serology and molecular typing* C
Jka
Jkb
M
N
Total
Discordant Concordance (%)
13 99.7
13 99.7
69 98.3
7 99.8
74 98.2
178 95.7
Discordant Concordance (%)
10 99.8
14 99.7
88 99.7
0 100.0
121 97.6
233 95.3
Version 1.1
1.2
* HEA BeadChip Version 1.1, n 5 4125; HEA BeadChip, Version 1.2, n 5 4966.
2372 TRANSFUSION Volume 55, October 2015
ASIAN AMERICAN RBC GENOTYPE
Fig. 1. Molecular (SNP) and serologic (Ser) discrepancies detected grouped by donor ethnic heritage. Discrepancy is defined as SNP1/Ser– or SNP–/Ser1.
Asian heritage. Indeterminate genotyping results were rare; only one donor was indeterminate for KEL (578T>C). The predicted frequency of the Kidd (Jka, Jkb) blood group antigens was not found to be significantly different than those from published European/Caucasian populations (p > 0.05, data not shown). There were 10 donors with genotype-serology discrepancies found to be serologically Jk(a–b–) but genotypically Jk(a1b1) using the SNP which detects SLC14A1 838G>A. The ethnicities of these donors were Hawaiian/Pacific Islander (six) and Filipino (four). There were 27 Jka and 157 Jkb donors with discrepancies between serologic and genotyping results.17,18 Six donors exhibited indeterminate genotyping results for Jka (two positive, four negative by serology) and seven Jkb indeterminate results (five positive and two negative by serology). SNPs that encode a common Duffy silencing mutation 267T>C (FY*02N.01) were detected in seven donors of Chinese (five), Filipino (one), and Native American (one) backgrounds. The FY*02 results ranged in frequency from 0.171 to 0.576 within the ethnic groups (Table 2). There were 39 individuals with 265C>T, FY*02M.01 (predicted phenotype of Fyx). There were 33 Fya and 34 Fyb donors with indeterminate results.
MNS The MNS blood group system genotype-predicted antigen frequencies were not significantly different in the overall study population compared to published European/Caucasian prevalence (p > 0.05, data not shown). Interestingly, the MNS*3 results ranged in frequency from 0.096 to 0.503 within the ethnic groups represented by the study samples (Table 2). There were no donors predicted to be
U– or variant Uvar (Ex5 230C>T, In5g>t) polymorphisms detected that are interrogated by the genotyping platform. There were eight donors with genotype-serology discrepancies for the M antigen, 188 with N discrepancies who were predicted by genotype to be N– who were serologically N1, while there were eight that were predicted by genotype to be N1 who were serologically N– (Table 3). Of the N discrepancies, 69 were found while the HEA BeadChip Version 1.1 platform was in use, and 121 were found while using Version 1.2. Within the MNS system, there were also numerous indeterminate genotype results: 11 M (seven positive, four negative by serology), 20 N (18 positive, two negative by serology), and 90 and 89 for the S and s antigens, respectively.
Other blood group systems The Diego blood group antigens (Dia, Dib) showed significant differences in genotype-predicted antigen expression frequencies for Native American (p < 0.05) and South Asian (p 5 0.003) compared to reference European/Caucasian frequencies (Table 2). None of the other blood group systems tested (Lutheran, Dombrock, Landsteiner-Wiener, Colton, or Scianna) had significant genotype-predicted frequency differences compared to European/Caucasian published frequencies. The CO*2 frequency results ranged in frequency from 0.003 to 0.039 within the ethnic groups represented by the study samples (Table 2).
Genotype-serology blood type concordance and discrepancies The concordances between genotype-predicted blood type and serology-detected blood group antigens were 95.7% (Version 1.1) and 95.3% (Version 1.2) for the five antigens with serology testing (M, N, C, Jka, Jkb). There Volume 55, October 2015 TRANSFUSION 2373
DELANEY ET AL.
were 412 total discrepancies (Table 4). To our knowledge, the N antigen did not have changes in the diagnostic SNP from Version 1.1 to Version 1.2; there was a slight increase from 0.016 discrepancy rate to 0.024 discrepancy rate between the two version, respectively. Since the testing took place over the course of several years, there was an N antisera change over the course of the study, but this did not appear to have an impact on the results. There were patterns of genotype-serology blood type assignment discrepancies that were disproportionately higher in certain donor ethnic groups (Chinese, Filipino, Hawaiian/Pacific Islander, Southeast Asian) for specific blood group antigens, which suggests the genotype accuracy may be confounded by undetected underlying genetic variation in these populations, such as glycophorin hybrid alleles known to be present in people of Asian descent19 (Fig. 1).
CONCLUSIONS This report describes the results of a survey of the prevalence of genotype-predicted blood antigens in a large cohort of self-identified Asian American and Native American blood donors. These data provide reference blood group frequencies that had previously not been well defined in these ethnic groups. Moreover, many of the blood group antigens tested do not have Food and Drug Administration (FDA)licensed antisera, which has made population studies of blood group antigen frequencies difficult before the advent and subsequent FDA approval of RBC genotyping. The information from this study may help inform and support transfusion care of patients of Asian American and Native American ancestral backgrounds because it illustrates potential mismatches between European descent and Asian or Native American-descent donor/patient pairs that can occur during transfusion and lead to RBC alloimmunization.20 Although this study focused on RBC typing, studies that point to blood group antibodies in the Asian descent population illustrate differences in antibody development. For instance, Miltenberger and the antibody antiMia and anti-Mur (often found together) are common in Asian transfusion recipients, but are rare to unknown in transfused patients of European ancestry. For example, anti-Mia was found in 29% of Chinese liver transplant patients,21 while a review of 28,303 mostly Chinese patients’ antenatal antibody screens found that 57.6% of antibodies were anti-Mia, making it the most common RBC antibody specificity detected.22-25 The overall mean concordance between molecular and serology testing was more than 95% for each of the five antigens (C, Jka, Jkb, M, N) that were separately tested by serology and genotyping analysis. The discrepancies that were found may be due to underlying genetic variation that is not targeted by the RBC genotyping array and therefore resulted in incorrect blood type assignment. In particular, it would be predicted that SNP testing alone cannot identify 2374 TRANSFUSION Volume 55, October 2015
the hybrid glycophorins (such as Gp.MUR), as well as JK alleles that are silenced or encode weakened antigen expression and therefore may not be sufficient to type these blood group systems in these populations without the addition of serology or additional genotyping.20-24,26 Additionally, the study may underestimate the rate of genotypeserology discrepancies, as every blood group antigen predicted by SNP testing was not able to be tested serologically. It is possible that the N antisera may not be highly specific for the N antigen; as such, serologic reagent specificity may be contributory to the level of N discrepancies detected. Unfortunately, because fresh RBCs are no longer available, serologic testing for the presence of rare or hybrid MNS proteins that can produce novel antigens, such as Mia or Stones antigen (Sta), was not possible. Overall, this study has shown that molecular genotyping for many blood group antigens can be used to determine the predicted prevalence of many blood group antigens. However, the study also finds a significant number of discrepancies that are not easily resolved by the approach employed in this study of Asian American and Native American populations. This may be due to a bias in the genotyping platform development that focused on alleles described and documented in other ethnic or African American populations. Further investigation is warranted for blood group genotyping in Asian and Native American descent populations. In conclusion, this study provides a large populationlevel survey of extended blood group frequencies in selfidentified Asian and Asian-descent blood donors in the United States. The genotyping approach for blood group antigen prediction is a useful technological advance for the field of transfusion medicine, some known data sets. This study also supports the premise that ethnic diversity can lead to important differences in the distribution of blood groups with clinical significance for transfusion. ACKNOWLEDGMENTS We are grateful to the many people who contributed to this project from the Red Cell Reference Laboratory at Puget Sound Blood Center, especially Rosalind Armour. Other contributors include Sandy Linauts, Shelley Nakaya Fletcher, Haley Huston, Wendy Hofeling, Prashant Gaur, Lakshmi Gaur, and the Blood Bank of Alaska.
CONFLICT OF INTEREST MD has received travel support and honoraria from Bioarray Solutions. The other authors have disclosed no conflicts of interest.
REFERENCES 1. Klapper E, Zhang Y, Figueroa P, et al. Toward extended phenotype matching: a new operational paradigm for the transfusion service. Transfusion 2010;50:536-46.
ASIAN AMERICAN RBC GENOTYPE
2. Reid ME. Applications of DNA-based assays in blood group antigen and antibody identification. Transfusion 2003;43: 1748-57. 3. Daniels G. Human blood groups. 3rd ed. Oxford: Blackwell Science; 2013. 4. Reid ME, Lomas-Francis C, Olsson ML. The blood group antigen factsbook. 3rd ed. London: Academic Press (Elsevier); 2012. 5. Miller ST, Kim HY, Weiner DL, et al. Red blood cell alloimmunization in sickle cell disease: prevalence in 2010. Transfusion 2013;53:704-9. 6. Godfrey GJ, Lockwood W, Kong M, et al. Antibody development in pediatric sickle cell patients undergoing erythrocytapheresis. Pediatr Blood Cancer 2010;55:1134-7. 7. Aygun B, Padmanabhan S, Paley C, et al. Clinical significance of RBC alloantibodies and autoantibodies in sickle cell patients who received transfusions. Transfusion 2002;42:37-43. 8. Vichinsky EP, Luban NL, Wright E, et al. Prospective RBC phenotype matching in a stroke-prevention trial in sickle cell anemia: a multicenter transfusion trial. Transfusion 2001;41:1086-92. 9. Wilkinson K, Harris S, Gaur P, et al. Molecular blood typing augments serologic testing and allows for enhanced matching of red blood cells for transfusion in patients with sickle cell disease. Transfusion 2012;52:381-8. 10. Afenyi-Annan A, Brecher ME. Pre-transfusion phenotype matching for sickle cell disease patients. Transfusion 2004; 44:619-20. 11. Singleton BK, Green CA, Avent ND, et al. The presence of an RHD pseudogene containing a 37 base pair duplication and a nonsense mutation in Africans with the Rh D-negative blood group phenotype. Blood 2000;95:12-8. 12. Chou ST, Westhoff CM. Molecular biology of the Rh system: clinical considerations for transfusion in sickle cell disease. Hematology Am Soc Hematol Educ Program 2009;178-84. 13. Meny GM. The Duffy blood group system: a review. Immunohematology 2010;26:51-6. 14. Chou ST, Jackson T, Vege S, et al. High prevalence of red blood cell alloimmunization in sickle cell disease despite
transfusion from Rh-matched minority donors. Blood 2013; 122:1062-71. 15. Kaushansky K, Lichtman M, Beutler E, et al. Williams hematology. 8th ed. Part VI, The erythrocyte. New York: McGrawHill Professional; 2010. 16. HEA BeadChipTM kit package insert, IVD. Warren (NJ). Bioarray Solutions; 2007–2010. 17. Reid ME. MNS blood group system: a review. Immunohematology 2009;25:95-101. 18. Posadas J, Shnyreva M, Gaur P, et al. Distribution of JK silent alleles in Asian American and Pacific Islander populations (SP229) [abstract]. Transfusion 2010;50(Suppl): 143A. 19. Gaur LK, Posadas J, Teramura G, et al. Novel Kidd polymorphisms may address serological discrepancies (S32-020D) [abstract]. Transfusion 2008;48(Suppl):13A. 20. Murphy EL, Shaz B, Hillyer CD, et al. Minority and foreignborn representation among US blood donors: demographics and donation frequency for 2006. Transfusion 2009;49: 2221-8. 21. Au WY, Liu CL, Lo CM, et al. Red blood cell alloantibodies and liver transplantation in Chinese patients. Transplantation 2003;76:324-6. 22. Westhoff CM. The structure and function of the Rh antigen complex. Semin Hematol 2007;44:42-50. 23. Irshaid NM, Henry SM, Olsson ML. Genomic characterization of the Kidd blood group gene: different molecular basis of the Jk(a-b-) phenotype in Polynesians and Finns. Transfusion 2000;40:69-74. 24. Guo Z, Wang C, Yan K, et al. The mutation spectrum of the JK-null phenotype in the Chinese population. Transfusion 2013;53:545-53. 25. Lee CK, Ma ES, Tang M, et al. Prevalence and specificity of clinically significant red cell alloantibodies in Chinese women during pregnancy—a review of cases from 1997 to 2001. Transfus Med 2003;13:227-31. 26. Vege S, Westhoff CM. Molecular characterization of GYPB and RH in donors in the American Rare Donor Program. Immunohematology 2006;22:143-7.
Volume 55, October 2015 TRANSFUSION 2375
BACKGROUND: There has yet to be a comprehensive analysis of blood group antigen prevalence in Asian Americans and Native Americans. There may be ethnic differences in blood group frequencies that would result in clinically important mismatches through transfusion. STUDY DESIGN AND METHODS: Blood donors who self-identified as Asian or Native American were tested using a single-nucleotide polymorphism (SNP) DNA array (HEA BeadChip kit, Bioarray Solutions Ltd) that predicts expression of 38 human erythrocyte antigens (HEAs) and by serology for ABO, D, C, M, N, Jka, and Jkb. The prevalence of blood group antigens was compared to published European prevalence. Discrepancies between SNP-predicted and serologydetected antigens were tallied. RESULTS: A total of 9087 blood donors were tested from nine Asian and Native American heritages. The predicted prevalence of selected antigens in the RHCE, JK, FY, MNS, LU, CO, and DO blood group systems were variable between Asian populations, but overall not significantly different than Europeans. Compared to European frequencies, Kell blood group allele frequencies were significantly different in the Chinese, Native American, Hawaiian/Pacific Islander, South Asian, and Southeast Asian heritage blood donors; Diego antigens Dia and Dib were different in donors of Native American and South Asian ancestries (p < 0.05). Of the donors tested, 4.5% showed a SNP-serology discrepancy that segregated within specific ethnic groups. CONCLUSION: This study provides HEA allele frequency and antigen prevalence data in a cohort of Asian and Native Americans donors. Several ethnic groups exhibited differences in HEA frequencies compared to Europeans. Genotype-serotype discrepancies were detected in all systems studied.
T
oday, blood centers and hospitals have access to genetic techniques to determine the predicted red blood cell (RBC) phenotype with molecular testing.1,2 Critical to the implementation of technological advancements in RBC typing technology is the understanding of blood group gene frequencies within different ethnic groups.3,4 The ethnic background of a transfusion recipient and blood donor pair can affect transfusion outcomes, as is evidenced by African American patients with sickle cell disease.5 This group of patients has the highest rate of RBC alloimmunization (up to 47%), which is at least partly attributed to the high degree of European American blood donors in the United States.6-8 The resulting minor blood group antigen mismatches lead to RBC alloantibody formation over the course of years of transfusion therapy. Study of the extended blood group phenotype/genotype in the African descent population has led to improvements in transfusion care of African descent transfusion recipients.9-13
ABBREVIATIONS: HEA 5 human erythrocyte antigen; SNP(s) 5 single-nucleotide polymorphism(s). From the 1Specialty Diagnostics, Red Cell Genomics Laboratory, and the 3Red Cell Reference Laboratory, Puget Sound Blood Center; the 2Department of Laboratory Medicine and the 5Department of Medicine, Division of Hematology, University of Washington; and the 4Puget Sound Blood Center Research Institute, Seattle, Washington. Address reprint requests to: Meghan Delaney, DO, MPH, Puget Sound Blood Center, 921 Terry Avenue, Seattle, WA 98104; e-mail: [email protected]. Funded by the Telemedicine & Advanced Technology Research Center (TATRC); US Army Medical Research Acquisition Activity (UWAMRAA), W81XWH-06-0052; and GSA.gov Contract GST0906BHM0529 Received for publication April 14, 2014; revision received March 14, 2015; and accepted April 10, 2015. doi:10.1111/trf.13163 C 2015 AABB V
TRANSFUSION 2015;55;2369–2375 Volume 55, October 2015 TRANSFUSION 2369
DELANEY ET AL.
TABLE 1. ABO and D percent phenotype prevalence from serologic testing (n 5 8453 with ABO and D results available)* Number of samples with ABO/D serology
Ethnicity Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Total European/Caucasian (published)2
1699 1280 897 1003 999 260 472 915 928 8453
O
A
B
AB
D1
D–
0.437 0.448 0.536 0.325 0.307 0.554 0.451 0.352 0.408 0.413 0.44
0.255 0.266 0.349 0.376 0.330 0.323 0.337 0.261 0.212 0.293 0.43
0.245 0.234 0.089 0.213 0.258 0.088 0.148 0.311 0.303 0.228 0.09
0.064 0.052 0.026 0.086 0.104 0.035 0.064 0.075 0.077 0.067 0.04
0.992 0.989 0.876 0.985 0.995 0.892 0.977 0.927 0.971 0.965 0.85
0.008 0.011 0.124 0.015 0.005 0.108 0.023 0.073 0.029 0.035 0.15
* Percentages are calculated within each ethnicity code (per row).
However, recent evidence supports that even racially matched blood transfusions in the sickle cell disease population do not protect from allosensitization and that genotypic matching may be the next step for improved patient outcomes.14 Blood banks must be prepared to support the transfusion needs of all patients and thus must be knowledgeable about the prevalence of blood group antigens in the population(s) they serve. Some ethnic backgrounds are disproportionately affected by genetically derived medical illnesses, such as sickle cell disease, thalassemia, and glucose-6-phosphate dehydrogenase deficiency.15 Manufacturers of RBC panels must develop and sustain reagents capable of detection of antibodies directed to blood group antigens that are represented at different frequencies in certain ethnic populations, such as Jsa, V, VS, and Mia. There has yet to be a comprehensive analysis of blood group antigens drawn from a large cohort of Asian American and Native American ancestral background people in the United States. Due to our center’s geographical location, we serve a large population of blood donors and patients of Asian American and Native American heritage. An extensive study was undertaken to better understand the blood type distributions in these populations and to test the robustness of our molecular blood typing technologies. RBC genotyping and blood type serology for selected blood group systems was performed in a large cohort of self-identified Asian American and Native American blood donors. Blood group allele frequencies were compared to published European ancestry RBC antigen prevalence, and concordance between serologic phenotyping and genotyping was determined for the RBC antigens C, M, N, Jka, and Jkb.
MATERIALS AND METHODS Discarded blood samples were collected between September 2007 and July 2010 from volunteer blood donors who 2370 TRANSFUSION Volume 55, October 2015
consented and self-identified to be of ethnic heritages that were Asian American or Native American. Samples from donors who self-selected as African American (choices: African, African American, Caribbean, South or Central American), European/Caucasian (choices: Middle East or North Coast Africa, North American, or European), Hispanic/South American (choices: Caribbean, Mexican or Chicano, South or Central American), multiracial, other, or “do not ask” were excluded. Donors were tested one time; repeat donors were not tested. A total of 8453 donor samples were from the Puget Sound Blood Center and 634 samples were from the Blood Bank of Alaska. Study approval was obtained from the Western Institutional Review Board. DNA isolation from EDTA peripheral blood was done using a DNA purification kit (Puregene blood kit, Qiagen, Valencia, CA) and an automated DNA extraction kit (Bee Robotics blood DNA extraction kit, QuatroProbe DNA robot, GTI, Waukesha, WI). Blood antigen molecular testing was carried out using an in-vitro diagnostic test for molecular determination of the allelic variants that indicate human erythrocyte antigen (HEA) phenotypes (HEA BeadChip kit, Bioarray Solutions Ltd, Immucor, Warren, NJ).16 Results were rendered by Web-based software (BASIS, Bioarray Solutions Ltd). Indeterminate results were provided when the calculation for the specific antigen signal intensity is too close to the cutoff to provide high confidence results.16 Each indeterminate result was investigated by first repeating the assay, as suggested by the manufacturer; if indeterminate results were again obtained, the authors, in collaboration with scientists from the manufacturer of the HEA BeadChip utilized relative fluorescence intensities to make calls. If an indeterminate result was not able to be resolved, they were recorded as such. Discrepancy was defined as either positive genotype results with negative serology, or negative single-nucleotide polymorphism (SNP) result and positive serology.
ASIAN AMERICAN RBC GENOTYPE
TABLE 2. RBC antigen allele frequencies by SNP analysis in blood donors of different self-identified Asian or Native American heritages and European/Caucasian percent prevalence of blood antigens* C
E
c
e
K
k
Kpa
Kpb
Jsa
Jsb
Ethnicity
Total
RH*2
RH*3
RH*4
RH*5
KEL*1
KEL*2
KEL*3
KEL*4
KEL*6
KEL*7
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.895 0.923 0.645 0.864 0.816 0.654 0.825 0.842 0.900 0.818 0.68
0.395 0.257 0.411 0.459 0.514 0.520 0.348 0.196 0.308 0.379 0.29
0.502 0.395 0.815 0.582 0.624 0.804 0.592 0.584 0.482 0.598 0.80
0.953 0.978 0.932 0.939 0.893 0.910 0.952 0.987 0.966 0.946 0.98
0.002 0.011 0.059 0.014 0.004 0.053 0.025 0.014 0.007 0.021 0.09
1.000 1.000 0.999 1.000 1.000 0.998 1.000 1.000 0.999 1.000 0.998
0.001 0.001 0.013 0.014 0.000 0.008 0.000 0.002 0.000 0.004 0.02
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
0.003 0.000 0.000 0.000 0.000 0.002 0.000 0.000 0.000 0.001 < 0.0001
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
Fya
Fyb
Jka
Jkb
M
N
S
s
Lua
Lub
Ethnicity
Total
FY*1
FY*2
JK*1
JK*2
MNS*1
MNS*2
MNS*3
MNS*4
LU*1
LU*2
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiian/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.982 0.965 0.760 0.973 0.992 0.860 0.925 0.871 0.968 0.922 0.66
0.128 0.270 0.705 0.292 0.171 0.576 0.462 0.563 0.281 0.383 0.83
0.697 0.721 0.758 0.728 0.740 0.786 0.779 0.820 0.764 0.755 0.77
0.770 0.744 0.751 0.762 0.784 0.741 0.648 0.675 0.712 0.732 0.74
0.801 0.790 0.839 0.786 0.759 0.892 0.835 0.852 0.860 0.824 0.78
0.652 0.701 0.635 0.692 0.732 0.554 0.642 0.631 0.607 0.650 0.72
0.096 0.163 0.503 0.166 0.105 0.490 0.246 0.487 0.214 0.274 0.55
0.988 0.980 0.911 0.983 0.987 0.899 0.967 0.880 0.947 0.949 0.89
0.002 0.013 0.057 0.010 0.005 0.043 0.035 0.004 0.000 0.019 0.08
0.999 1.000 0.999 1.000 0.999 0.997 0.998 0.999 1.000 0.999 0.998
Dia
Dib
Coa
Cob
Doa
Dob
Hy
Jo(a)
LWa
LWb
Sc1
Sc2
Ethnicity
Total
DI*1
DI*2
CO*1
CO*2
DO*1
DO*2
DO*4
DO*5
LW*5
LW*7
SC*1
SC*2
Chinese Filipino American Native Japanese Korean Alaska Native/Aleut Hawaiin/Pacific Islander South Asian Southeast Asian Mean (study samples) European/Caucasian (published)2
1715 1333 970 1022 1033 621 522 922 942 9080
0.044 0.010 0.019 0.079 0.105 0.010 0.010 0.003 0.022 0.034 0.0001
0.992 0.995 0.997 0.993 0.989 0.997 0.998 0.994 0.993 0.994 1.00
1.000 1.000 0.999 1.000 0.999 0.998 1.000 0.998 1.000 0.999 0.995
0.003 0.010 0.068 0.016 0.007 0.039 0.021 0.012 0.007 0.020 0.10
0.206 0.228 0.591 0.254 0.209 0.588 0.383 0.623 0.334 0.380 0.67
0.987 0.978 0.858 0.977 0.987 0.839 0.960 0.841 0.944 0.930 0.82
1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.00
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1.000 0.999 1.000 1.000 1.000 1.000 1.000 0.999 0.998 1.000 1.00
0.000 0.004 0.005 0.000 0.000 0.010 0.000 0.000 0.000 0.002 < 0.01
1.000 0.999 0.999 1.000 1.000 1.000 1.000 0.999 0.997 0.999 > 0.99
0.001 0.002 0.004 0.003 0.004 0.002 0.000 0.002 0.000 0.002 0.01
* Note: Kpa, Kpb, Jsa, Jsb were introduced with the second generation (Version 1.2) of the genomic testing platform (n 5 4962).
For the first 4125 samples, the HEA BeadChip Version 1.1 platform was utilized, which assigns 34 blood group antigens (in 11 systems) using 19 SNPs. During the course of the study, the HEA BeadChip platform Version 1.2 became available, which assigns 38 blood group antigens in 11 systems using 24 SNPs. The Version 1.2 platform included five additional SNPs: two for additional antigens Kpa, Kpb, Jsa, and Jsb (KEL*03, KEL*04, KEL*06, KEL*07 alleles) and three SNPs to improve the coverage of RHCE designations (733C>G, 1006G>T, 109Ins).16 This version was employed for the subsequent 4962 samples. The genotype-predicted blood group antigen results were grouped by ethnic heritage selected by the blood donor (Microsoft Excel, Microsoft Corp., Redmond, WA). Parallel serologic antigen typing was
conducted for C (Seraclone human monoclonal, MS24, Bio-Rad/BioTest, Hercules, CA), Jka, Jkb (polyclonal, Immucor Gamma, Norcross, GA), M, and N (murine monoclonal, gamma clone by tube test, Immucor Gamma; and Seraclone murine monoclonal, BS57 and BS41, Biotest). Concordance was calculated for these antigens between genotype-predicted blood type and serology. Genotype-serology discrepancies found for C, Jka, Jkb, M, and N were tabulated by ethnic heritage. ABO and D serologic testing was carried out using a blood grouping reagent system (Olympus PK7200/PK7300 Beckman Coulter, Fullerton, CA; and Gallileo, Immucor). Using chi-square analysis, the prevalences of the blood group antigens in each ethnic group were compared to Volume 55, October 2015 TRANSFUSION 2371
DELANEY ET AL.
TABLE 3. Samples found to have M and N serology and genotype discrepancy, with Glycophorin A (GYPA) and Glycophorin B (GYPB) genotyping results Number of samples 3 31 153 1 2 3 3 3 2 1 2
Serology
Predicted antigen expression (genotype, by SNP)
GYPA 59C>T
GYPB 143T>C
GYPB Ex5 230C>T
GYPB In5 15g>t
N1 N1 N1 N1 N– N– N– M1 M– M– M–
N– N– N– N– N1 N1 N1 M– M1 M1 M1
C/C C/C C/C C/C C/T C/T C/T T/T C/T C/T C/T
T/T T/C C/C C/C T/C T/T C/C C/C T/C T/T C/C
C/C C/C C/C C/T C/C C/C C/C C/C C/C C/C C/C
G/G G/G G/G G/G G/G G/G G/G G/G G/G G/G G/G
those of published prevalence of antigens in European/ Caucasian ancestry people to identify significant differences, or mismatches, which may be encountered in transfusion.2 p values were considered significant if they were less than 0.05 and not significant if more than 0.05. SNP-serology concordance rates were calculated using total typed samples with both SNP data and serology.
RESULTS A total of 9087 blood donors from nine populations of Asian American and Native American heritage were selected for study, representing 4.1% of total blood donors collected over 35 months. A total of 8955 (98.5%) were serologically tested in parallel for five antigens, C, M, N, Jka, and Jkb. The ABO, D (serology), and Genotype-predicted blood group antigen prevalence(s) were calculated (Tables 1 and 2). There were nine rare blood donors found (defined as being negative for a “highincidence” antigen < 1:1000). These donors had the following types: one Co(a–) (Native American), six Di(b–) (Korean [4], Japanese [1], Southeast Asian [1]), two k– (Native American [1], Alaska Native/Aleut [1]). Additionally, there were 589 that were uncommon blood donors (defined as lacking more than five common blood group antigens).
RhCE Although the predicted prevalence of expression of the RhCE (C, c, E, e) blood group antigens were variable between ethnic groups, overall they were not significantly different than those published in European/Caucasian populations (data not shown, p 5 0.9). Using the SNPs (733C>G, 1006G>T) that are associated with expression of V and VS antigens in the Rh blood group system, three donors were found that are predicted to be VS1 and V– (one Chinese, one Alaska Native/Aleut, one Korean). Fifteen donors were found to be positive for both VS and V (six American Indian, three Alaska Native/Aleut, two Chinese, two Filipino, one Korean, one Hawaiian/Pacific Islander). There were 25 donors with discrepancies between C serology and genotype-predicted C expression. Eighty-six donors had indeterminate C genotyping results (307C>T, 109Ins); 79 were positive and seven were negative by serology. For c, e, and E there were 35, 14, and 18 indeterminate genotype results, 307C>T, 676G>C), respectively.
Kell, Kidd, and Duffy The predicted frequency of expression for the Kell blood group antigens (K, k, Kpa, Kpb, Jsa, Jsb) was found to be different (p < 0.05) than published European/Caucasian frequencies for donors of Chinese, Native American, Hawaiian/Pacific Islander, South Asian, and Southeast
TABLE 4. Concordance for five blood group antigens between using serology and molecular typing* C
Jka
Jkb
M
N
Total
Discordant Concordance (%)
13 99.7
13 99.7
69 98.3
7 99.8
74 98.2
178 95.7
Discordant Concordance (%)
10 99.8
14 99.7
88 99.7
0 100.0
121 97.6
233 95.3
Version 1.1
1.2
* HEA BeadChip Version 1.1, n 5 4125; HEA BeadChip, Version 1.2, n 5 4966.
2372 TRANSFUSION Volume 55, October 2015
ASIAN AMERICAN RBC GENOTYPE
Fig. 1. Molecular (SNP) and serologic (Ser) discrepancies detected grouped by donor ethnic heritage. Discrepancy is defined as SNP1/Ser– or SNP–/Ser1.
Asian heritage. Indeterminate genotyping results were rare; only one donor was indeterminate for KEL (578T>C). The predicted frequency of the Kidd (Jka, Jkb) blood group antigens was not found to be significantly different than those from published European/Caucasian populations (p > 0.05, data not shown). There were 10 donors with genotype-serology discrepancies found to be serologically Jk(a–b–) but genotypically Jk(a1b1) using the SNP which detects SLC14A1 838G>A. The ethnicities of these donors were Hawaiian/Pacific Islander (six) and Filipino (four). There were 27 Jka and 157 Jkb donors with discrepancies between serologic and genotyping results.17,18 Six donors exhibited indeterminate genotyping results for Jka (two positive, four negative by serology) and seven Jkb indeterminate results (five positive and two negative by serology). SNPs that encode a common Duffy silencing mutation 267T>C (FY*02N.01) were detected in seven donors of Chinese (five), Filipino (one), and Native American (one) backgrounds. The FY*02 results ranged in frequency from 0.171 to 0.576 within the ethnic groups (Table 2). There were 39 individuals with 265C>T, FY*02M.01 (predicted phenotype of Fyx). There were 33 Fya and 34 Fyb donors with indeterminate results.
MNS The MNS blood group system genotype-predicted antigen frequencies were not significantly different in the overall study population compared to published European/Caucasian prevalence (p > 0.05, data not shown). Interestingly, the MNS*3 results ranged in frequency from 0.096 to 0.503 within the ethnic groups represented by the study samples (Table 2). There were no donors predicted to be
U– or variant Uvar (Ex5 230C>T, In5g>t) polymorphisms detected that are interrogated by the genotyping platform. There were eight donors with genotype-serology discrepancies for the M antigen, 188 with N discrepancies who were predicted by genotype to be N– who were serologically N1, while there were eight that were predicted by genotype to be N1 who were serologically N– (Table 3). Of the N discrepancies, 69 were found while the HEA BeadChip Version 1.1 platform was in use, and 121 were found while using Version 1.2. Within the MNS system, there were also numerous indeterminate genotype results: 11 M (seven positive, four negative by serology), 20 N (18 positive, two negative by serology), and 90 and 89 for the S and s antigens, respectively.
Other blood group systems The Diego blood group antigens (Dia, Dib) showed significant differences in genotype-predicted antigen expression frequencies for Native American (p < 0.05) and South Asian (p 5 0.003) compared to reference European/Caucasian frequencies (Table 2). None of the other blood group systems tested (Lutheran, Dombrock, Landsteiner-Wiener, Colton, or Scianna) had significant genotype-predicted frequency differences compared to European/Caucasian published frequencies. The CO*2 frequency results ranged in frequency from 0.003 to 0.039 within the ethnic groups represented by the study samples (Table 2).
Genotype-serology blood type concordance and discrepancies The concordances between genotype-predicted blood type and serology-detected blood group antigens were 95.7% (Version 1.1) and 95.3% (Version 1.2) for the five antigens with serology testing (M, N, C, Jka, Jkb). There Volume 55, October 2015 TRANSFUSION 2373
DELANEY ET AL.
were 412 total discrepancies (Table 4). To our knowledge, the N antigen did not have changes in the diagnostic SNP from Version 1.1 to Version 1.2; there was a slight increase from 0.016 discrepancy rate to 0.024 discrepancy rate between the two version, respectively. Since the testing took place over the course of several years, there was an N antisera change over the course of the study, but this did not appear to have an impact on the results. There were patterns of genotype-serology blood type assignment discrepancies that were disproportionately higher in certain donor ethnic groups (Chinese, Filipino, Hawaiian/Pacific Islander, Southeast Asian) for specific blood group antigens, which suggests the genotype accuracy may be confounded by undetected underlying genetic variation in these populations, such as glycophorin hybrid alleles known to be present in people of Asian descent19 (Fig. 1).
CONCLUSIONS This report describes the results of a survey of the prevalence of genotype-predicted blood antigens in a large cohort of self-identified Asian American and Native American blood donors. These data provide reference blood group frequencies that had previously not been well defined in these ethnic groups. Moreover, many of the blood group antigens tested do not have Food and Drug Administration (FDA)licensed antisera, which has made population studies of blood group antigen frequencies difficult before the advent and subsequent FDA approval of RBC genotyping. The information from this study may help inform and support transfusion care of patients of Asian American and Native American ancestral backgrounds because it illustrates potential mismatches between European descent and Asian or Native American-descent donor/patient pairs that can occur during transfusion and lead to RBC alloimmunization.20 Although this study focused on RBC typing, studies that point to blood group antibodies in the Asian descent population illustrate differences in antibody development. For instance, Miltenberger and the antibody antiMia and anti-Mur (often found together) are common in Asian transfusion recipients, but are rare to unknown in transfused patients of European ancestry. For example, anti-Mia was found in 29% of Chinese liver transplant patients,21 while a review of 28,303 mostly Chinese patients’ antenatal antibody screens found that 57.6% of antibodies were anti-Mia, making it the most common RBC antibody specificity detected.22-25 The overall mean concordance between molecular and serology testing was more than 95% for each of the five antigens (C, Jka, Jkb, M, N) that were separately tested by serology and genotyping analysis. The discrepancies that were found may be due to underlying genetic variation that is not targeted by the RBC genotyping array and therefore resulted in incorrect blood type assignment. In particular, it would be predicted that SNP testing alone cannot identify 2374 TRANSFUSION Volume 55, October 2015
the hybrid glycophorins (such as Gp.MUR), as well as JK alleles that are silenced or encode weakened antigen expression and therefore may not be sufficient to type these blood group systems in these populations without the addition of serology or additional genotyping.20-24,26 Additionally, the study may underestimate the rate of genotypeserology discrepancies, as every blood group antigen predicted by SNP testing was not able to be tested serologically. It is possible that the N antisera may not be highly specific for the N antigen; as such, serologic reagent specificity may be contributory to the level of N discrepancies detected. Unfortunately, because fresh RBCs are no longer available, serologic testing for the presence of rare or hybrid MNS proteins that can produce novel antigens, such as Mia or Stones antigen (Sta), was not possible. Overall, this study has shown that molecular genotyping for many blood group antigens can be used to determine the predicted prevalence of many blood group antigens. However, the study also finds a significant number of discrepancies that are not easily resolved by the approach employed in this study of Asian American and Native American populations. This may be due to a bias in the genotyping platform development that focused on alleles described and documented in other ethnic or African American populations. Further investigation is warranted for blood group genotyping in Asian and Native American descent populations. In conclusion, this study provides a large populationlevel survey of extended blood group frequencies in selfidentified Asian and Asian-descent blood donors in the United States. The genotyping approach for blood group antigen prediction is a useful technological advance for the field of transfusion medicine, some known data sets. This study also supports the premise that ethnic diversity can lead to important differences in the distribution of blood groups with clinical significance for transfusion. ACKNOWLEDGMENTS We are grateful to the many people who contributed to this project from the Red Cell Reference Laboratory at Puget Sound Blood Center, especially Rosalind Armour. Other contributors include Sandy Linauts, Shelley Nakaya Fletcher, Haley Huston, Wendy Hofeling, Prashant Gaur, Lakshmi Gaur, and the Blood Bank of Alaska.
CONFLICT OF INTEREST MD has received travel support and honoraria from Bioarray Solutions. The other authors have disclosed no conflicts of interest.
REFERENCES 1. Klapper E, Zhang Y, Figueroa P, et al. Toward extended phenotype matching: a new operational paradigm for the transfusion service. Transfusion 2010;50:536-46.
ASIAN AMERICAN RBC GENOTYPE
2. Reid ME. Applications of DNA-based assays in blood group antigen and antibody identification. Transfusion 2003;43: 1748-57. 3. Daniels G. Human blood groups. 3rd ed. Oxford: Blackwell Science; 2013. 4. Reid ME, Lomas-Francis C, Olsson ML. The blood group antigen factsbook. 3rd ed. London: Academic Press (Elsevier); 2012. 5. Miller ST, Kim HY, Weiner DL, et al. Red blood cell alloimmunization in sickle cell disease: prevalence in 2010. Transfusion 2013;53:704-9. 6. Godfrey GJ, Lockwood W, Kong M, et al. Antibody development in pediatric sickle cell patients undergoing erythrocytapheresis. Pediatr Blood Cancer 2010;55:1134-7. 7. Aygun B, Padmanabhan S, Paley C, et al. Clinical significance of RBC alloantibodies and autoantibodies in sickle cell patients who received transfusions. Transfusion 2002;42:37-43. 8. Vichinsky EP, Luban NL, Wright E, et al. Prospective RBC phenotype matching in a stroke-prevention trial in sickle cell anemia: a multicenter transfusion trial. Transfusion 2001;41:1086-92. 9. Wilkinson K, Harris S, Gaur P, et al. Molecular blood typing augments serologic testing and allows for enhanced matching of red blood cells for transfusion in patients with sickle cell disease. Transfusion 2012;52:381-8. 10. Afenyi-Annan A, Brecher ME. Pre-transfusion phenotype matching for sickle cell disease patients. Transfusion 2004; 44:619-20. 11. Singleton BK, Green CA, Avent ND, et al. The presence of an RHD pseudogene containing a 37 base pair duplication and a nonsense mutation in Africans with the Rh D-negative blood group phenotype. Blood 2000;95:12-8. 12. Chou ST, Westhoff CM. Molecular biology of the Rh system: clinical considerations for transfusion in sickle cell disease. Hematology Am Soc Hematol Educ Program 2009;178-84. 13. Meny GM. The Duffy blood group system: a review. Immunohematology 2010;26:51-6. 14. Chou ST, Jackson T, Vege S, et al. High prevalence of red blood cell alloimmunization in sickle cell disease despite
transfusion from Rh-matched minority donors. Blood 2013; 122:1062-71. 15. Kaushansky K, Lichtman M, Beutler E, et al. Williams hematology. 8th ed. Part VI, The erythrocyte. New York: McGrawHill Professional; 2010. 16. HEA BeadChipTM kit package insert, IVD. Warren (NJ). Bioarray Solutions; 2007–2010. 17. Reid ME. MNS blood group system: a review. Immunohematology 2009;25:95-101. 18. Posadas J, Shnyreva M, Gaur P, et al. Distribution of JK silent alleles in Asian American and Pacific Islander populations (SP229) [abstract]. Transfusion 2010;50(Suppl): 143A. 19. Gaur LK, Posadas J, Teramura G, et al. Novel Kidd polymorphisms may address serological discrepancies (S32-020D) [abstract]. Transfusion 2008;48(Suppl):13A. 20. Murphy EL, Shaz B, Hillyer CD, et al. Minority and foreignborn representation among US blood donors: demographics and donation frequency for 2006. Transfusion 2009;49: 2221-8. 21. Au WY, Liu CL, Lo CM, et al. Red blood cell alloantibodies and liver transplantation in Chinese patients. Transplantation 2003;76:324-6. 22. Westhoff CM. The structure and function of the Rh antigen complex. Semin Hematol 2007;44:42-50. 23. Irshaid NM, Henry SM, Olsson ML. Genomic characterization of the Kidd blood group gene: different molecular basis of the Jk(a-b-) phenotype in Polynesians and Finns. Transfusion 2000;40:69-74. 24. Guo Z, Wang C, Yan K, et al. The mutation spectrum of the JK-null phenotype in the Chinese population. Transfusion 2013;53:545-53. 25. Lee CK, Ma ES, Tang M, et al. Prevalence and specificity of clinically significant red cell alloantibodies in Chinese women during pregnancy—a review of cases from 1997 to 2001. Transfus Med 2003;13:227-31. 26. Vege S, Westhoff CM. Molecular characterization of GYPB and RH in donors in the American Rare Donor Program. Immunohematology 2006;22:143-7.
Volume 55, October 2015 TRANSFUSION 2375
