Cryopreservation of Genotyped ABO Subgroup Rbcs for Quality Assurance of ABO Grouping Reagents

Available online at www.annclinlabsci.org 216 Annals of Clinical & Laboratory Science, vol. 48, no. 2, 2018 Cryopreservation of Genotyped ABO Subgroup RBCs for Quality Assurance of ABO Grouping Reagents

Sinyoung Kim1, Sungwook Song2, and Hyun Ok Kim1

1Department of Laboratory Medicine, Yonsei University College of Medicine, Seoul and 2Daejeon-Sejong-Chungnam Blood Center, Korean Red Cross, Daejeon, Korea

Abstract. Background. Quality assurance of newly developed or manufactured blood grouping reagents with reagent red blood cells (RBCs) is crucial in the process of product approval by governmental agency. However, RBCs with rare blood group are not easily available in the fresh state. We investigated the feasibil- ity of cryopreserved and genotyped ABO subgroup RBC reagents for quality assurance purpose. Methods. We obtained RBCs from 10 volunteers with ABO subgroup phenotypes. The ABO genotypes and alleles were analyzed by the sequencing of ABO exon 6 and 7. Using the 40% wt/vol high-glycerol method, RBC units were cryopreserved as reagent RBCs into separate cryovials at -80°C. The potency titrations were performed before and after cryopreservation for 6 months and 2 years to evaluate the stability of ABO antigens. Results. ABO genotypes of cryopreserved RBCs were cis-AB01/O01 (n=2), cis-AB01/B101 (n=1), Aw10/B101 (n=2), A201/A201 (n=1), A205/B101 (n=1), A102/B112 (n=1), and A101/B306 (n=1). These ABO subgroup alleles are exclusively present in East-Asian population except for the known ABO*A201 allele. The potency titers of cryopreserved RBCs were not significantly different between pre-freezing and post-freezing. Conclusions. The national performance evaluation of ABO blood grouping reagents could be performed with cryopreserved and genotyped reagent RBCs.

Introduction blood grouping reagents including detailed technical specification. In the EU, blood grouping re- With the accumulation of knowledge on the mo- agents are classified as in vitro diagnostic medical lecular background of blood group genetics, the device which should comply with regulations speci- replacement of conventional serologic methods for fied in the common technical specifications for in pre-transfusion testing by molecular methods has vitro Diagnostic Medical Devices (2009/108/EC) begun in limited clinical settings [1,2]. Although [5]. According to this common technical specifica- blood group genotyping in pre-transfusion testing tion, newly developed ABO grouping reagents will be more widely used in the near future, con- should be tested on at least 3,000 samples and spe- ventional serologic grouping methods are still im- cific specimens should be included to reflect variant portant [1,3,4]. Furthermore, conventional sero- or weak antigen expression in this performance logic techniques, including blood grouping, evaluation. Even before the batch release, each unexpected antibody screening, and identification blood grouping reagent should undergo a specifici- would be continuously modified and improved. In ty test which examines its reactivity with samples of conventional serologic blood grouping, the quality A1,A2B, Ax, B, O blood groups. Similarly, in the and quality assurance of blood grouping reagents USA, manufacturers of ABO grouping reagents are utmost important in identifying accurate blood should confirm the reactivity with samples of A1, groups of pre-transfusion specimens, especially for A2B, Ax blood groups for new product license ap- the ABO blood group. plications and include a panel of appropriate rare blood group cells in field trials [6-8]. Each country has its own guidelines and/or standards of production, quality, and release of ABO For the performance evaluation of newly developed or manufactured ABO grouping reagent, fresh red Address correspondence to Hyun Ok Kim, MD, PhD; Department blood cells (RBCs) samples should be prepared and of Laboratory Medicine, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul, 03722, Korea; phone: +82 2 2228 tested with an appropriate method. However, fresh 2441; fax: +82 2 313 0956; e mail:[email protected] RBCs samples with rare ABO subgroups could not

Table 1. Oligonucleotide primers used for amplification and sequencing of ABO genes.

Primer name Primer Primer sequence Primer Amplified ABO region pair location mo57a A GGGTTTGTTCCTATCTCTTTGC Intron 5 Exon 6-7 mo71R GGGCCTAGGCTTCAGTTACTC 3'-UTR ABO-261F-C B AGGAAGGATGTCCTCGTGGTG Exon 6 Exon 7 (non-O1) mo71R GGGCCTAGGCTTCAGTTACTC 3'-UTR ABO-261F-C C AGGAAGGATGTCCTCGTGGTG Exon 6 Exon 7 (B-specific) ABO-1096R-B GAGGGGGACGGGGCTGCT 3'-UTR ABOex6-F CATGTGACCGCACGCCT Intron 5 ABOex6-R TCGGCCACCTCACTGACTTA Intron 6 For sequencing ABOex7-F2 GGAGAACGTGGTGCATCTG Intron 6 ABOex7-R2 GGCGCTCGTAGGTGAAGG Exon 7 ABOex7-F3 AGATCCTGACTCCGCTGTTC Exon 7 ABOex7-R3 CCCGTTCTGCTAAAACCAAG 3'-UTR

Table 2. Phenotypes and genotypes of 10 cryopreserved ABO subgroup RBCs.

No Phenotype Genotype Potency titersa A antigen B antigen

1 A2B3 cis-AB01/O01 768 (128-1,024) 12 (8-32) 2 A2B3 cis-AB01/O01 512 (128-1,024) 6 (4-8) 3 A2B cis-AB01/B101 24 (16-64) 1,024 (512-1,024) b 4 AweakB Aw10/B101 0 (0-1) 1,024 (512-1,024) b 5 AweakB Aw10/B101 1 (0-2) 1,024(1,024) 6 A2 A201/A201 256 (256-1,024) 0 (0) 7 A2B A205/B101 512 (256-1,024) 768 (512-1,024) 8 A1Bweak A102/B112 1,024 (1,024) 320 (128-512) 9 A1Bweak A101/B306 1,024 (512-1,024) 0.5 (0-1) c 10 Bweak B101 /O02 0 (0) 128 (16-256) aMedian (Min-Max); bMacroscopic agglutination can be detected only by the two kinds of anti-A reagents.; cSample #10 showing Bweak phenotype could not be resolved by direct sequencing of ABO gene including exons 6 and 7. be readily available for the performance evaluation. the genotyped and cryopreserved RBC panels of In Korea, from 2007 to 2009, total 1,771 (0.052%) rare ABO subgroup for national inspection of ABO donors showed diverse phenotypes of ABO sub- blood grouping reagents. group among 3,397,983 healthy blood donors [9]. Common ABO subgroups were A2B (n=496, Materials and Methods 28.0%), A2B3 (n=457, 25.8%), and A1B3 (n=362, Whole blood collection. All donations were from eligi- 20.4%). However, A2 phenotype which was preva- lent in Western countries was found only in 81 ble, voluntary, non-remunerated volunteers screened ac- blood donors (4.6% of blood donors with ABO cording to the standards of the Korean Red Cross Blood subgroup, 0.002% of all blood donors). Service. Whole blood units were collected into blood collection bag containing citrate phosphate dextrose ad- enine (CPDA)-1 anticoagulant solution from 10 donors Ideally, materials used for quality control need to who were known to have ABO subgroup phenotype at be stable, available in aliquots that can be periodi- four blood centers of the Korean Red Cross. Subsequent cally analyzed over a long time. Since fresh RBCs RBC concentrates production was carried out at each with rare ABO subgroups do not fulfill above-men- blood center using standard operating procedures and tioned criteria of control materials, we established transported to Severance Hospital. This study was 218 Annals of Clinical & Laboratory Science, vol. 48, no. 2, 2018

approved by the Institutional Review Board and Ethics glycerolized RBCs were aliquoted into separate cryovials Committee of Severance Hospital and written informed (2 mL in each vial) and frozen with a slow drop in tem- consent was obtained from each donor before whole perature, and stored at -80°C in mechanical freezers. The blood donation. freezers used for this study was monitored electronically and manually and no failures in temperature mainte- Serological typing of ABO blood group. Serologic ABO nance occurred during the storage of the RBC samples. blood grouping of RBC concentrates was carried out with standard hemagglutination technique, as previous- After 6 months or 1 year of storage, the cryopreserved ly described [9-11]. Following four commercial anti-A vials were thawed in a 37°C water bath and thawed and anti-B monoclonal reagents were used: Novaclone RBCs were quickly transferred to a 5 mL test tube, then (Dominion Biologicals, Dartmouth, Canada), BioClone 2 mL of warm 9% NaCl solution was added dropwise (Ortho-Clinical Diagnostics, Raritan, NJ, USA), Mono- with gentle mixing. After the incubation at room tem- type (Medion Diagnostics AG, Bonnstrasse, Germany), perature for 1 minute, the test tube was filled with 2.5% and Bioscot (Millipore, Livingstone, UK). All reagents NaCl solution. The thawed RBCs were repeatedly cen- were used according to the manufacturers’ instruction. trifuged at 1,000 g for 15 seconds and washed with 0.9% NaCl solution until the supernatant appeared clear. Genotyping of ABO blood group. Genomic DNA was Deglycerolized RBCs were then suspended in Alsever’s extracted from whole blood by QIAmp DNA Blood solution (Sigma Chemicals, St Louis, MO) and stored at Mini Kit (Qiagen GmbH, Hilden, Germany). We per- 4°C before the potency titration. formed direct sequencing of ABO exon 6 and 7 for ABO genotyping, as described previously [12,13]. A 2237-bp Evaluation of ABO antigen stability. We performed length fragment containing exon 6 and exon 7 was am- the potency titrations for pre-freezing and post-thawed plified using primer pair A (Table 1). The polymerase RBCs to evaluate the stability of ABO antigen as previ- chain reaction (PCR) was performed in a total volume ously reported [7,17,18]. Briefly, the potency titer was of 20 µL containing 10 pmol of each primer, 100 ng of measured as follows: beginning with undiluted blood genomic DNA, 5 nmol of each dNTP, and 1 U TaKaRa grouping reagent, two-fold serial dilutions to 1 in 1024 Ex Taq HS (Takara, Shiga, Japan). Thermal cycling was were prepared. One volume of each dilution of the re- performed using a GeneAmp PCR system 9700 (Applied agent placed in a separate test tube and one volume of Biosystems, Foster City, CA) in the following condi- 2–3% RBCs suspension was added to each tube. The tions: an initial denaturation step of 10 minutes at 96°C, contents of each tube were mixed thoroughly and incu- 15 cycles consisting of 20 seconds at 94°C, and 2.5 min- bated for 5 minutes at room temperature. After centrifu- utes at 65°C, then 25 cycles of 20 seconds at 94°C, 30 gation at 1,000 g for 15 seconds, the cell buttons of each seconds at 61°C, 2.5 minutes at 72°C, and a final exten- tube were examined macroscopically with gentle agita- sion for 5 minutes. The amplified fragments were puri- tion. The potency titer was the reciprocal of the highest fied using QIAquick gel extraction kit (Qiagen) and se- dilution of the reagent that showed visible quenced bidirectionally with a total of 6 primers. The agglutination. Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems) and an ABI PRISM 310 Genetic Analyzer Results (Applied Biosystems) were utilized according to manufacturers’ instructions. In addition, allele-specific prim- ABO phenotype and genotype of cryopreserved ers were used to ensure that sequencing results obtained RBCs. The phenotyping and genotyping results of for exons 6 and 7 reflected only the relevant allele. All cryopreserved RBCs are shown in Tables 2&3. obtained nucleotides sequences were compared with known ABO polymorphisms from the Blood Group Two A2B3 (#1 and #2) and 1 A2B (#3) phenotype Antigen Gene Mutation Database (www.ncbi.nlm.nih. samples were determined as ABO*cis-AB01/O01 gov/gv/mhc/xslcgi.cgi?cmd=bgmut). and ABO*cis-AB01/B101, respectively. This ABO*cis-AB01 allele differs from ABO*A101 allele Cryopreservation and thawing of frozen RBCs. For by 2 nucleotide polymorphisms (c.467C>T and the cryopreservation of RBCs, we used the high-glycerol c.803G>C) and is relatively more common in the method which was modified from the original Meryman- Korean and Japanese population, as previously re- Hornblower method for clinical practice [14-16]. ported [19,20]. Within 5 days after collection, Glycerolyte 57 solution (57% wt/vol glycerol; Baxter Healthcare, Deerfield, IL) An A antigen of two A B phenotype samples (#4 was added into RBCs to achieve a final glycerol concen- weak tration of 40% (wt/vol) with constant mixing. The and #5) was weakly detected only by the two kinds Cryopreservation of genotyped ABO subgroup RBCs 219 of monoclonal anti-A reagents 96 A A A (BioClone and Mono-type) among 10 the 4 tested monoclonal anti-A re-

62 agents. The nucleotide sequence of

10 these samples was identical to that of ABO*A101 allele except for a o 61 nucleotide substitution (c.784G>A - 354 Pr 10 in exon 7) and is compatible to ABO*Aw10 allele. This variant al- g NA. 09 G 33 Ar Gly 10

cD lele was exclusively found in

O Koreans and showed the allelic en- u 0 0 AB 31 Le hancement [20,21]. The A allele of A A A 93 of samples showing A (#6) and A B g 2 2 l 7 (#7) phenotypes were genotyped as A 27 Va nnin 829 typical ABO*A201 allele and rare gi be 8

3 ABO*A205 allele, respectively. a e 26 C 80 Al th ABO*A205 allele (c.467C>T and t

om c.1009A>G) is the common cause 6 6 u Gly fr ACACAC Me

26 of A B phenotype in the Chinese 79 Le 2 ng

ti Han population and also reported 2 p n ar 4

st in other East Asian countries 26 As A As 78

ed [22-24]. er 7 1 o mb 25 Pr 77 Two samples (#8 and #9) showing nu e

y A B phenotype were genotyped r 3

ar 1 weak A A A Gl Se 70 as ABO*A102/B112 and ns io 7 y. o ABO*A101/B306. An ABO*B112 7 235 it 1 Pr ud AT 22 allele, which was first identified by 68 pos st is

de Zhu et al. [23], had the same nucle- th ti 7 9 otide sequences as ABO*B101 allele in His 65 21 T T cleo es except for c.559C>T nucleotide el nu e 6 6 all changes and Arg187Cys amino acid A 21 Ph 64 Th e O

. changes. The nucleotide sequence s AB ce g 7 9 of ABO*B306 allele was identical to of en 18 Ar Cy 55

es that of the ABO*B101 allele except qu se p n 3 7 for a single nucleotide substitution e erenc 18 As As 54 nc c.547G>A, causing Asp183Asn di ff re fe

6 substitution. A sample (#10) show- ce g 6 re G GTTGA en 52 17 Ar e ing the Bweak phenotype could not qu th

se be resolved by direct sequencing of 7 as u Gly o 6 T T T T id 46 exons 6 and 7 of ABO gene. Le Pr ed 15 ac us s 7 ino r Effect of cryopreservation on ABO 29 wa G G G am Th 99 6 d

ce antigen stability. The potency ti- l an 1 en

s ters of ABO subgroup RBCs were 26 GACCGCTCGGCGCGGGACGG- -G87 Va qu de

se compared before and after cryo- ti eo

01 preservation for 6 months and 2 cl 01 A1 ge id years (Figure 1). All cryopreserved Nu de an AB 1 10 ac O* 01 05 06 01 02 01 on ti 01 12 3.

s- RBCs showed no difference more o on on eo e Ex A1 A1 Aw A2 A2 B1 B1 B3 ci O0 O02 A1 Ch AB ti ti in cl bl

si si than 2 titers of pre-freezing Nu Ta po Am po Th e samples. 220 Annals of Clinical & Laboratory Science, vol. 48, no. 2, 2018

Figure 1. Four representative results of the potency titers before freezing and after cryopreservation for 6 months and 2 years. Anti-A (●, closed circle), anti-B (○, open circle). (A) Sample #1 (cis-AB01/O01). (B) Sample #5 (Aw10/B101), the potency titers cannot be illustrated due to low median titer value. (C) Sample #3 (cis-AB01/B101). (D) Sample #6 (A201/ A201).

Discussion Testing with RBC antigens that have weak reactivity is an important step in quality control of blood Although ABO grouping is the most important grouping reagents. Unfortunately, the low preva- pre-transfusion serological test, the importance of lence of these blood types in general population ABO grouping reagents is often overlooked in rou- precludes easy procurement of the antigens. tine clinical practice. After the development of ABO grouping reagent production from monoclo- The reported frequency of ABO subgroup pheno- nal antibodies in the mid-’80s, qualified ABO types in Korea is 0.052-0.084% among the Korean grouping reagents have been produced in a stan- blood donors. A2B, A2B3, and A1B3 phenotypes dardized manner in terms of affinity, avidity, speci- were frequently found in ABO subgroup pheno- ficity, agglutination, etc. The performance of types with a rate of 28.0%, 25.8%, and 20.4%, re- monoclonal reagents depends on the following fac- spectively. These phenotypes were probably due to tors: the immunological and physicochemical char- cis-AB allele, which is relatively common in the acteristics of antibodies, the medium, the technol- Korean population and produces multiple different ogy used, and the condition of reaction [25]. So, phenotypes such as A2B3,A1B3,A2B, A1Bx, and even if the same manufacturer’s monoclonal ABO etc [20]. Cho reported genotypes of 194 donors grouping reagents are used, variability in antigen- (0.114%) with ABO subgroup among 169,605 antibody interactions could not be eliminated. blood donors in southwest Korea [20]. Of these More importantly, less common antigens such as 194 donors, cis-AB01 alleles were detected in 60 weak A or weak B may produce insufficient reagent donors (30.9%). The genotypes of cis-AB01/O01 activity and confound the interpretation of results. and cis-AB01/O02 produced the A2B3 phenotype, Cryopreservation of genotyped ABO subgroup RBCs 221

References and cis-AB01/B101 consistently produced A2B phenotype. However, the cis-AB01/A102 genotype 1. Quill E. Medicine. Blood-matching goes genetic. Science was associated with 3 different phenotypes (A1B3, 2008;319:1478-9. A B , and A B ). 2. Anstee DJ. Red cell genotyping and the future of pretransfu- 1 x or el int 3 sion testing. Blood 2009;114:248-56. 3. Denomme GA. Prospects for the provision of genotyped In our study, RBCs with cis-AB01/O01 (n=2) and blood for transfusion. Br J Haematol 2013;163:3-9. cis-AB01/B101 (n=1) were collected and cryopre- 4. Flegel WA, Gottschall JL, Denomme GA. Integration of red cell genotyping into the blood supply chain: a population- served for reagent RBCs using the high glycerol based study. Lancet Haematol 2015;2:e282-9. method. Additionally, 2 cases (#4 and #5) of Aw10/ 5. European Commission. Commission Decision of 3 February 2009 amending Decision 2002/364/EC on common techni- B101 genotype, 1 case (#6) of A201/A201 geno- cal specifications for in vitro-diagnostic medical devices type, and 1 case (#7) of A205/B101 genotype were (2009/108/EC). Official Journal of the European Union cryopreserved. For the variant of B allele, we col- 2009;L39/34-L39/49. 6. U.S. Food and drug administration. Part 660. Additional lected and cryopreserved 1 case (#8) of A102/B112 standards for diagnostic substances for laboratory tests. genotype and 1 case (#9) of A101/B306 genotype. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/ Our study observed no significant difference in po- CFRSearch.cfm?CFRPart=660&showFR=1. Accessed 14 Juli 2017. tency titers before freezing and after cryopreserva- 7. U.S. Food and drug administration. Recommended methods tion for 6 months and 2 years; therefore, the high for blood grouping reagents evaluation, 1992. https://www. fda.gov/downloads/BiologicsBloodVaccines/Guidance glycerol cryopreservation method appears to be a ComplianceRegulatoryInformation/ useful technique as it allows frozen-thawed reagent OtherRecommendationsforManufacturers/UCM062695. RBCs to be used for quality assurance of blood pdf. Accessed 14 Juli 2017. 8. U.S. Food and drug administration. Points to consider in the grouping reagents. design and implementation of field trials for blood grouping reagents and anti-human globulin, 1992. https://www.fda. In each country, ABO blood grouping procedures gov/downloads/BiologicsBloodVaccines/ GuidanceComplianceRegulatoryInformation/Guidances/ should be standardized based on the population- Blood/UCM080919.pdf. Accessed 14 Juli 2017. based data including the phenotypic frequency and 9. Lee JY, Oh DJ, Park YM. The frequency and distribution of the ABO subgroups in Korean blood donors. Korean J Blood genetic basis of ABO subgroup. The cis-AB blood Transfus 2010;21:223-9. group is the most frequently observed ABO sub- 10. Seltsam A, Hallensleben M, Eiz-Vesper B, Lenhard V, group in Korea, and group O RBCs and group AB Heymann G, Blasczyk R. A weak blood group A phenotype caused by a new mutation at the ABO locus. Transfusion plasma or platelet products are recommended for 2002;42:294-301. the transfusion support of patients with cis-AB 11. Seltsam A, Hallensleben M, Kollmann A, Blasczyk R. The blood group [26]. Currently, the cryopreserved and nature of diversity and diversification at the ABO locus. Blood 2003;102:3035-42. genotyped ABO subgroup RBC panels including 12. Olsson ML, Irshaid NM, Hosseini-Maaf B, Hellberg A, cis-AB blood group have been accepted and used as Moulds MK, Sareneva H, Chester MA. Genomic analysis of clinical samples with serologic ABO blood grouping discrep- one of the standardized materials for the approval ancies: identification of 15 novel A and B subgroup alleles. of ABO grouping reagents by Ministry of Food and Blood 2001;98:1585-93. Drug safety in Korea. Furthermore, using the same 13. Seltsam A, Hallensleben M, Kollmann A, Burkhart J, Blasczyk R. Systematic analysis of the ABO gene diversity strategy, national inspection and performance eval- within exons 6 and 7 by PCR screening reveals new ABO al- uation on blood grouping reagents for Rh and oth- leles. Transfusion 2003;43:428-39. er minor blood groups can be performed. 14. Meryman HT, Hornblower M. A method for freezing and washing red blood cells using a high glycerol concentration. Transfusion 1972;12:145-56. Acknowledgements 15. Hogman CF, Hornblower ML, Flodin M, Gillberg G, Meryman HT, Säfwenberg J. A simple method for high-qual- This study was supported by a grant of the Korea FDA ity frozen red cells in blood group serology. Transfusion (07102 International Collaboration Research Project 1986;26:434-6. 735), Republic of Korea and by a faculty research grant 16. Lelkens CC, Noorman F, Koning JG, Truijens-de Lange R, of Yonsei University College of Medicine Stekkinger PS, Bakker JC, Lagerberg JW, Brand A, Verhoeven (6-2011-0089). AJ. Stability after thawing of RBCs frozen with the high- and low-glycerol method. Transfusion 2003;43:157-64. 222 Annals of Clinical & Laboratory Science, vol. 48, no. 2, 2018

17. Thorpe SJ, Fox B, Heath AB, Scott M, de Haas M, Kochman 21. Cho D, Lee SY, Ryang DW. ABO subgroup studies in Korea. S, Padilla A. International standards for minimum potency of ISBT Science Series 2015;10:332-335. anti-A and anti-B blood grouping reagents: evaluation of can- 22. Ogasawara K, Yabe R, Uchikawa M, Bannai M, Nakata K, didate preparations in an international collaborative study. Vox Takenaka M, Takahashi Y, Juji T, Tokunaga K. Different alleles Sang 2006;91:336-44. cause an imbalance in A2 and A2B phenotypes of the ABO 18. Joint United Kingdom (UK) Blood Transfusion and Tissue blood group. Vox Sang 1998;74:242-7. Transplantation Services Professional Advisory Committee. 23. Zhu F, Tao S, Xu X, Ying Y, Hong X, Zhu H, Yan L.Distribution Guidelines for the blood transfusion services in the UK. http:// of ABO blood group allele and identification of three novel al- www.transfusionguidelines.org.uk/red-book. Accessed 14 Juli leles in the Chinese Han population. Vox Sang 2010;98:554-9. 2017. 24. Chang CS, Lin KT, Chang JG, Lin CW, Hsieh LL, Yeh CJ, Liu 19. Yamamoto F, McNeill PD, Kominato Y, Yamamoto M, TC. Molecular basis of the A2B in Taiwan. Int J Hematol Hakomori S, Ishimoto S, Nishida S, Shima M, Fujimura Y. 2008;88:127-33. Molecular genetic analysis of the ABO blood group system: 2. 25. Rouger P, Noizat-Pirenne F, Le Pennec PY. Advances in the use cis-AB alleles. Vox Sang 1993;64:120-3. of monoclonal antibodies for blood group testing. Transfus Clin 20. Cho D, Kim SH, Jeon MJ, Choi KL, Kee SJ, Shin MG, Shin Biol 1997;4:345-9. JH, Suh SP, Yazer MH, Ryang DW. The serological and genetic 26. Yoon SZ, Lee HM, Kim HS, Kim SD, Oh AY, Kim CS. basis of the cis-AB blood group in Korea. Vox Sang Transfusion for a patient of cis-AB blood type undergoing a 2004;87:41-3. redo cardiac operation. Br J Anaesth 2005;94:733-4.