PRODUCT RISK FACTORS Anti-A and anti-B: what are they and where do they come from? Donald R. Branch
Intravenous immunoglobulin (IVIG) is made from thousands of donors having a variety of blood groups. All of the donors being used for IVIG production, with the exception of group AB donors, have in their plasma antibodies of variable titer commonly known as isohemagglutinins or ABO antibodies. As blood groups O and A are the most commonly found in the world population, most of the plasma used in IVIG production is from donors having these blood groups, with group B and group AB donors being fewer in number. Consequently, all batches of IVIG contain antibodies that are reactive with individuals of group A, group B, and group AB. These antibodies were originally discovered by Dr Karl Landsteiner in the early 1900s and are now known to consist of immunoglobulin (Ig)M, IgG, and IgA classes. As the process for producing IVIG results in almost exclusively IgG, isohemagglutinins contained in IVIG are of this immunoglobulin class. ABO antibodies are highly clinically significant and, because of this, blood bank cross-matching is done to ensure that blood of the correct type is transfused into recipients to avoid a socalled major mismatch or major incompatibility that can cause significant morbidity and often death. Administration of IVIG, which contains ABO antibodies, is often infused into individuals who have the corresponding ABO antigens, commonly called a minor mismatch, and although not as significant as a major mismatch, the isohemagglutinins contained in the IVIG have some risk for a significant transfusion reaction due to the ABO incompatibility. Indeed, currently there is no way to match IVIG to recipients according to blood type, so when IVIG is administered to group A, B, or AB recipients, there is potential for transfusion reactions analogous to a blood transfusion mismatch. For this reason, strict guidelines have been put into place to restrict the titers of the ABO antibodies contained in IVIG. This review will provide background information about the discovery and biochemistry of the ABO antigens and discuss the various isohemagglutinins that are found in plasma of the different ABO blood types and their potential clinical significance. In addition, a brief discussion of the controversial topic of the origins of these antibodies will conclude this review.
DISCOVERY OF THE ABO BLOOD GROUP SYSTEM
T
he 1930 Nobel Prize in Physiology or Medicine was awarded to Dr Karl Landsteiner “for his discovery of human blood groups.” Dr Landsteiner (Fig. 1) performed a simple experiment using blood from himself and his available staff. Published in 1901,1 Landsteiner found that when he mixed different combinations of red blood cells (RBCs) and plasma he obtained agglutination but with different patterns depending on the RBCs and plasma that were being mixed. One pattern of agglutination resulted in what Landsteiner called “A” and a different pattern he called “B.” There were two individuals’ RBCs that did not agglutinate with any of the plasma, which Landsteiner termed “C.” The “C” was eventually replaced with the letter “O” (that may, initially, have been a numerical zero to indicate zero reaction with plasma) and thus was born the ABO blood group system. One year later, colleagues in Landsteiner’s lab identified the AB blood group where the RBCs agglutinated with most plasma tested but the plasma did not agglutinate other RBCs.2 Although Landsteiner was awarded the Nobel Prize for his discovery of the ABO antigens and antibodies, he was not pleased with this accolade as he viewed his other work on viruses and haptens as more substantial and important.
LANDSTEINER’S LAW Landsteiner’s findings of four different “types” of blood based in good part on antibodies contained in plasma led
From the Centre for Innovation, Canadian Blood Services; and the Departments of Medicine and Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada. Address reprint requests to: Dr Donald R. Branch, Canadian Blood Services, 67 College Street, Toronto, Ontario, M5G 2M1, Canada; e-mail: [email protected]. doi:10.1111/trf.13087 C 2015 AABB V
TRANSFUSION 2015;55;S74–S79 S74
TRANSFUSION Volume 55, July 2015
ABO ISOHEMAGGLUTININS
him to postulate what has become known as “Landsteiner’s Law.”2,3 Simply put, Landsteiner’s Law states that whichever ABO antigens are lacking on a given person’s RBCs, that person will always have the corresponding antibody or isohemagglutinin.3 Thus, Group A individuals, lacking B antigen, always have anti-B in their plasma, while group B, lacking A antigen, always have anti-A. Group O, lacking both A and B antigens, always have both anti-A and anti-B, and group AB, having both A and B antigens, do not have any isohemagglutinins. The very rare Bombay phenotype has anti-A, -B, and -H, as this rare ABO blood group lacks A and B antigens but also lacks the H antigen (see below). In 1952, Dodd4 described an unexpected isohemagglutinin in serum that was only found in group O individuals. This antibody appeared to recognize both A and B antigens and was called a linked anti-A and anti-B.2,4 This antibody is now known as anti-A,B2,5-7 and reacts with a recognition determinant that is found on both A and B RBCs and can be identified only by absorption-elution studies.2,5,6 The anti-A,B follows Landsteiner’s Law so that it is found only in group O or Bombay individuals (Fig. 2).
ANTIGENS OF THE ABO BLOOD GROUP SYSTEM The antigens of the ABO system are controlled by the action of transferases.2,8 These transferases are enzymes that are responsible for the addition of specific sugars to an oligosaccharide precursor chain having a terminal galactose (Fig. 3). Group O individuals have an H gene located on chromosome 19 that encodes for a fucosyltransferase that adds a fucose sugar in a1,2-linkage to the terminal galactose on the precursor oligosaccharide, forming the so-called H antigen. H antigen forms a precursor oligosaccharide necessary to form A antigen and B antigen. Group A is formed by a gene located on Chromosome 9 that encodes for another transferase, a1,3N-acetylgalactosaminyltransferase, that adds N-acetylgalactosamine in a1,3 linkage to the terminal H-antigen. Group B gene, also encoded by Chromosome 9, produces a1,3-galactosyltransferase, which adds a galactose in a1,3 linkage to the H antigen. If both A and B genes are present, some H antigen will be converted to A antigen and some to B antigen creating the blood group AB. If the H gene is absent or defective, no H antigen can be formed and therefore no A or B antigen. These rare individuals are called Bombay (Fig. 3).
ABO ISOHEMAGGLUTININS: RELEVANCE TO INTRAVENOUS IMMUNOGLOBULIN ABO isohemagglutinins are mostly immunoglobulin (Ig)M antibodies but IgG and IgA classes also exist in plasma. The fractionation and manufacturing process for the pro-
Fig. 1. Dr. Karl Landsteiner (circa 1920): Father of the ABO blood group system and 1930 Nobel Prize winner for this pioneering discovery.
duction of intravenous immunoglobulin (IVIG) basically removes the majority of the IgM and IgA antibodies, leaving most of the IgG isohemagglutinins. As IVIG is fractionated from plasma pools containing a high proportion of group O donor plasma, IVIG must contain, in addition to anti-A and anti-B, anti-A,B, as this antibody is mostly found as IgG in group O people.2,5 IVIG is not expected to contain anti-H, given the rarity of the Bombay phenotype. There may be, however, some low levels of anti-A1, an antibody that specifically recognizes a majority of group A people and is made, again, according to Landsteiner’s Law, from a subset of group A individuals who lack the A1 phenotype2,5 (Fig. 2). All ABO isohemagglutinins have the potential for clinical significance.2,5 Both IgM and IgG classes can activate complement and cause intravascular lysis.2,5 IVIG contains almost exclusively IgG without IgM or IgA; thus, there is potential for clinical sequelae when IVIG is administered if the recipient is group A, B, or AB. Indeed, extravascular hemolysis has been shown to occur in ABO hemolytic disease of the fetus and newborn (HDFN) when anti-A, -B, or -A,B cross the placenta.5 Of note, anti-A,B can often be eluted from infants’ RBCs born to group O mothers and has been shown to be a highly clinically significant antibody.5 In one study, six of nine babies required treatment due to anti-A,B hemolytic disease of the fetus and newborn, whereas in babies having only Volume 55, July 2015 TRANSFUSION S75
BRANCH
Fig. 2. The five major ABO blood groups. Type A is found in approximately 35% to 36% of Caucasians. It follows Landsteiner’s Law by always having anti-B in the serum or plasma. Certain subtypes of A, that lack A1, can also have anti-A1. Type B is found in approximately 7% to 9% of Caucasians and follows Landsteiner’s Law by always having anti-A in the serum or plasma. Type AB is found in approximately 2% to 4% of Caucasians and has no isohemagglutinins in the serum or plasma. Type O is the most common ABO blood group comprising approximately 37% to 40% of Caucasians and follows Landsteiner’s Law by always having anti-A, anti-B, and anti-A,B in the serum or plasma. Finally, Type Bombay lacks the ability to add fucose to the oligosaccharide backbone and, thus, lacks H antigen, making it impossible to generate A or B antigens. Bombay is very rare, comprising less than one in 250,000 of the Caucasian population but can be found in one in 10,000 people in India. It follows Landsteiner’s Law by having not only anti-A, anti-B, and anti-A,B in the serum or plasma but also anti-H, as it lacks this ABO antigen.
anti-A or anti-B on their RBCs, only five of 14 required treatment.5 Although the contribution of anti-A,B to hemolysis after administration of IVIG to either group A or group B individuals has not been addressed, a significant number of ABO-associated hemolysis cases have been reported after IVIG administration and can be severe.9-12 One way manufacturers and regulatory agencies attempt to avoid the potential for hemolysis after IVIG therapy is to control the levels of anti-A and anti-B in the product.13 This can be achieved by limiting the potency of these antibodies by avoiding high titers. Thus, the Food and Drug Administration (FDA) has required IVIG not to have titers of anti-A and/or anti-B in IVIG higher than 64.13,14 In practice most manufacturers’ isohemagglutinin titers are lower than the FDA requirements. One potential problem with this approach is that all manufacturers and the FDA have ignored the anti-A,B isohemagglutinin conS76
TRANSFUSION Volume 55, July 2015
tained in IVIG. Indeed, when performing titration studies of IVIG against either A or B RBCs, there is no certainty that the resulting endpoint is due to anti-A or anti-B as at least the possibility exists of anti-A,B reactivity. This is because anti-A,B reacts with an as-yet-undefined, shared epitope on either A or B cells. This is an important concept as new-generation IVIG products have slightly higher titers, but within FDA guidelines, of anti-A and anti-B13 but the titers of anti-A,B are unknown. To obtain the specific titer of anti-A or anti-B in IVIG one would have to first adsorb the IVIG using group B cells to obtain an accurate titration of anti-A and, likewise, adsorb the IVIG with group A1 cells to obtain and accurate titration of the antiB contained in the IVIG. If one compares the titer results of the adsorbed against unadsorbed IVIG, one could get an idea of the titer of the anti-A,B in the IVIG preparation. More recently, manufacturers are turning to affinity chromatography to remove anti-A and anti-B from
ABO ISOHEMAGGLUTININS
Fig. 3. Schematic representation of the biochemical determinants of the ABO blood group system. The four blood groups are represented based on the addition of specific sugars by enzyme transferases to the backbone precursor oligosaccharide: Bombay lacking transferase activity for addition of fucose to the backbone oligosaccharide; group O, which requires a fucosyltransferase to add fucose to the oligosaccharide backbone to create the H antigen; Group A, requiring N-acetylgalactosaminyltransferase to add an N-acetylgalactosamine to the H antigen; and group B, which requires a galactosyltransferase to add a galactose to the H antigen. Red diamond 5 galactose; blue square 5 N-acetylglucosamine; black square 5 fucose; green diamond 5 Nacetylgalactosamine.
IVIG.13,15 Interestingly, complete exhaustion of these isohemagglutinins contained in IVIG preparations appears to be difficult if not impossible, despite multiple passages through the affinity columns.15 One can speculate that this is due to residual anti-A,B, which recognizes a chimeric antigen of both A and B components. The recognition determinate for anti-A,B has yet to be adequately elucidated so that currently available synthetic blood group antigens used in affinity chromatography do not contain the anti-A,B recognition carbohydrate sugars. The importance of anti-A,B cannot be overstated as many investigators have reported that this is a highly clinically significant isohemagglutinin that is almost exclusively of
the IgG subclass.2,5 Ongoing clinical studies of IVIGs having ABO isohemagglutinins reduced by donor selection and/or affinity chromatography will be helpful in deciding on the significance of anti-A,B contained in IVIG products. Most ABO isohemagglutinins, even if IgG, have the potential of causing complement-mediated, severe intravascular hemolysis. These ABO isohemagglutinins are known as ABO hemolysins.2,5 The fact that the antigen site density is very high for ABO antigens (Fig. 2) and that IgG antibodies can arrange themselves in close enough proximity for complement activation means that intravascular hemolysis is possible after administration of IVIG.9 It is possible that rare IVIG-associated hemolysis is not a result of low-titer anti-A, anti-B, or anti-A,B but high-titer hemolysin activity. Hemolysin activity is easy to test2 but is not routinely done for IVIG. However, hemolysin testing could be performed on individual lots of IVIG before release and may provide for an extra measure of safety. Briefly, using a two-stage method, one would add group A1, B, and O RBCs into separate tubes; add IVIG to each tube; and incubate at 37 C for 30 minutes. The tubes are removed and centrifuged, the supernatant is removed, an equal volume of freshly drawn normal group AB serum (can be previously fresh-frozen) is added to each tube as a source of complement, and the tubes are incubated again at 37 C for 30 minutes. The tubes are centrifuged and observed for reduction in size of the cell pellet and for free hemoglobin (Hb; pink to red supernatant), which can be quantified if desired. The group O tube is the negative control and should not show any evidence of RBC pellet reduction or free Hb. Any level of cell pellet reduction or pink color in either the A1 or the B tube and a negative finding in the O tube is considered a positive result, with the strength of the hemolysin activity dependent on the remaining cell pellet and intensity of the red color of the supernatant.2
ORIGINS OF ABO ISOAGGLUTININS The origin of the ABO isohemagglutinins is not really resolved. The question is whether these antibodies are produced through some inherited, “natural” innate mechanism, not requiring antigenic stimulation or, instead, follow classical adaptive immune-mediated mechanisms.2,16,17 An answer to this was attempted in the late 1950s and early 1960s when Springer and his associates17-19 published a series of landmark papers that suggested the ABO isohemagglutinins were produced by classical adaptive immunity in response to bacterial antigens. They showed that chickens kept in a germ-free environment would produce anti-B but not anti-A when fed bacteria expressing high levels of a B-like antigen and lower levels of an A-like antigen.17 Later Springer and colleagues18 showed that bacteria reacted with anti-A and Volume 55, July 2015 TRANSFUSION S77
BRANCH
TABLE 1. Comparison of natural and immune antibodies B-cell subset
Location
Innate
B1 (CD51)
Restricted
Adaptive
B2 (CD5-)
Systemic
Immune arm Natural antibody Immune antibody
Immunoglobulin class IgM IgG (no IgA, IgD, or IgE) IgG > IgM > IgA IgD, IgE
anti-B suggesting a cross-reactive antigen on bacteria responsible for induction of anti-A and anti-B via bacterial exposure. However, Springer and colleagues17,19 were only ever able to show weak anti-A production despite their many studies. Thus, the origin of these isohemagglutinins continued to perplex. Indeed, many studies before and after the work of Springer and colleagues have found antibodies, including ABO isohemagglutinins, detectable in neonates and newborns very close to birth that are not of maternal origin due to crossing the placenta and that cannot be adequately explained by an adaptive immune response to environmental stimuli.16,20-25 Even Landsteiner believed that ABO isohemagglutinins were spontaneously produced.2 With the discovery of two distinct subsets of B lymphocytes,16,26,27 one termed B1 (CD51), T-cell independent that does not require prior stimulation, responsible for “natural” antibody production, and the B2 subset (CD52), T-cell dependent and responsible for humoral responses via the adaptive immune system, the question of the origins of ABO isohemagglutinins has become more complex. Table 1 shows a comparison of “natural” versus immune-mediated antibody properties. Natural antibodies produced by B1 B lymphocytes have been proposed to include the ABO isohemagglutinins in addition to other antibodies produced to carbohydrate antigens as part of the innate immune system.16,28-31 Some investigators believe that ABO isohemagglutinins are initially produced “spontaneously” from a fixed set of ancestral germline genes found in B1 B lymphocytes.32 Indeed, newborns have mostly B1 B lymphocytes and produce only IgM antibodies, including ABO isohemagglutinins, early in life.16,20-25 Involvement of B2 B lymphocytes requires adaptive immunity, and it is known that IgG isohemagglutinins can be simulated, perhaps in response to bacterial and/or food antigens.16,18,33 Wuttke and coworkers16 found that IgM ABO isohemagglutinins of endogenous origin were always found in newborns by 8 months and usually much earlier, but by 8 months both B1 and B2 B lymphocytes were producing antibodies. The ABO antibodies found in IVIG are due to B2 B lymphocytes and an adaptive immune response.
T-cell dependent?
Antigenic stimuli
Plasma cells
Germline genes
No
Not required
No
Restricted germline
Yes
Requires antigen presentation and T-cell help
Yes
Rearrangement; hypermutation
the early 1900s with four major blood group antigens, A, B, AB, and O. The ABO isohemagglutinins, also discovered in the early 1900s, continue to follow Landsteiner’s Law without exception so that in individuals lacking an ABO antigen the isohemagglutinins corresponding to that antigen are always found in the plasma. It is plasma containing these isohemagglutinins that is used in the manufacture of IVIG and that can be potentially clinically relevant when IVIG is infused into non–group O individuals. Although the biochemistry of the ABO blood group system has been mostly worked out, elements still exist such as the epitope recognized by anti-A,B that still eludes investigators. Also what test may be the best predictor of potential clinical significance of ABO antibodies contained in IVIG or other plasma-based blood products remains a question: titration cutoff, hemolysin activity, anti-A,B levels. Although ABO antibodies can be IgM and IgG, elucidation of the origin of natural, IgM ABO antibodies compared to immune-mediated, IgG antibodies remains controversial.
CONFLICT OF INTEREST The author has disclosed no conflicts of interest.
REFERENCES 1. Landsteiner K. Ueber Agglutinationserscheinungen normalen menschlichen Blutes. Wien Klin Wochschr 1901;14: 1132-34. 2. Mollison PL. Blood transfusion in clinical medicine. 5th ed. Oxford: Blackwell Scientific Publications; 1972. 3. Springer GF, Tegtmeyer H. Apparent violation of “Landsteiner’s law.” Klin Wochenschr 1974;52:295-7. 4. Dodd BE. Linked anti-A and anti-B antibodies from group O sera. Br J Exp Pathol 1952;33:1-18. 5. Mollison PL. Blood transfusion in clinical medicine. 7th ed. Oxford: Blackwell Scientific Publications; 1983. p. 290-1. 6. Kajii E, Usuda S, Ikemoto S. Characterization of a monoclonal crossreacting anti-A,B antibody. Nihon Hoigaku Zasshi
SUMMARY The ABO blood group system has been studied for more than 110 years. The nomenclature has not changed since S78
TRANSFUSION Volume 55, July 2015
1990;44:227-33. 7. Moore S, Chirnside A, Micklern LR, et al. A mouse monoclonal antibody with anti-A,(B) specificity which agglutinates Ax cells. Vox Sang 1984;47:427-34.
ABO ISOHEMAGGLUTININS
8. Sheffield WP, Tinmouth A, Branch DR. Blood group biochemistry: a Canadian Blood Services research and development symposium. Transfus Med Rev 2005;19:295-307. 9. Michelis FV, Branch DR, Scovell I, et al. Acute hemolysis after intravenous immunoglobulin amid host factors of ABOmismatched bone marrow transplantation, inflammation, and activated mononuclear phagocytes. Transfusion 2014; 54:681-90. 10. Desborough MJ, Miller J, Thorpe SJ, et al. Intravenous immunoglobulin-induced haemolysis: a case report and review of the literature. Transfus Med 2014;24:219-26. 11. Berard R, Whittemore B, Scuccimarri R. Hemolytic anemia following intravenous immunoglobulin therapy in patients treated for Kawasaki disease: a report of 4 cases. Pediatr Rheumatol Online J 2012;10:10. 12. Markvardsen LH, Christiansen I, Harbo T, et al. Hemolytic anemia following high dose intravenous immunoglobulin in patients with chronic neurological disorders. Eur J Neurol 2014;21:147-52.
20. De Biasi B. Studies on iso-agglutinins in the blood of the new-born. JAMA 1923;81:1776-8. 21. Thomaidis T, Fouskaris G, Matsaniotis N. Isohemagglutinin activity in the first day of life. Am J Dis Child 1967;113: 654-7. 22. Chattoraj A, Gilbert R, Josephson AM. Serological demonstration of fetal production of blood group isoantibodies. Vox Sang 1968;14:289-91. 23. Godzisz J. [Synthesis of natural allohemagglutinins of the ABO system in healthy children aged 3 months to 3 years]. Rev Fr Transfus Immunohematol 1979;22:399-412. French. 24. Mencarini L, Tozzi C, Arachi S, et al. [Autochthonous anti-A/ B allohemaggluntinins in umbilical cord blood]. Minerva Pediatr 1982;34:83-5. Italian. 25. Merbl Y, Zucker-Toledano M, Quintana FJ, et al. Newborn humans manifest autoantibodies to defined self molecules detected by antigen microarray informatics. J Clin Invest 2007;117:712-8. 26. Martin F, Kearney JF. B1 cells: similarities and differences
13. US Food and Drug Administration. Public workshop—strat-
with other B cell subsets. Curr Opin Immunol 2001;13:195-
egies to address hemolytic complications of immune globulin infusions. Silver Spring (MD): FDA; 2014 [accessed 2014
201. 27. Montecino-Rodriguez E, Leathers H, Dorshkind K. Identifi-
Oct 5]. Available from: http://www.fda.gov/biologicsbloodvaccines/newsevents/workshopsmeetingsconferences/ ucm378388.htm 14. Human normal immunoglobulin for intravenous administration. In: European pharmacopoeia. 3rd ed. Strasbourg: Council of Europe; 1997. p. 963-5. 15. Hoefferer L, Glauser I, Gaida A, et al. Isoagglutinin reduction by a dedicated immunoaffinity chromatography step in the manufacturing process of human immunoglobulin products. Transfusion 2015;55 (Suppl 2):S117-S121. 16. Wuttke NJ, Macardle PJ, Zola H. Blood group antibodies are made by CD51 and by CD5- B cells. Immunol Cell Biol 1997;75:478-83. 17. Springer GF, Horton RE, Forbes M. Origin of anti-human blood group B agglutinins in white Leghorn chicks. J Exp Med 1959;110:221-44.
cation of a B-1 B cell-specified progenitor. Nat Immunol. 2006; 7:293-301. 28. Parker W, Lundberg-Swanson K, Holzknecht ZE, et al. Isohemagglutinins and xenoreactive antibodies: members of a distinct family of natural antibodies. Hum Immunol 1996;45: 94-104. 29. Holodick NE, Tumang JR, Rothstein TL. Immunoglobulin secretion by B1 cells: differential intensity and IRF4dependence of spontaneous IgM secretion by peritoneal and splenic B1 cells. Eur J Immunol 2010;40:3007-16. 30. Baumgarth N, Tung JW, Herzenberg LA. Inherent specificities in natural antibodies: a key to immune defense against pathogen invasion. Springer Semin Immunopathol 2005;26: 347-62. 31. Bovin NV. Natural antibodies to glycans. Biochemistry (Mosc) 2013;78:786-97.
18. Springer GF, Williamson P, Brandes WC. Blood group activity of gram-negative bacteria. J Exp Med 1961;113:1077-93.
32. Arend P. Ancestral gene and “complementary” antibody dominate early ontogeny. Immunobiology 2013;218:755-61.
19. Springer GF, Horton RE. Blood group isoantibody stimula-
33. Yamamoto S. The occurrence of materials cross-reacting
tion in man by feeding blood group-active bacteria. J Clin Invest 1969;48:1280-91.
with anti-A and -B agglutinins in fruit or seed extracts of higher plants. J Immunogenet 1977;4:325-30.
Volume 55, July 2015 TRANSFUSION S79
Intravenous immunoglobulin (IVIG) is made from thousands of donors having a variety of blood groups. All of the donors being used for IVIG production, with the exception of group AB donors, have in their plasma antibodies of variable titer commonly known as isohemagglutinins or ABO antibodies. As blood groups O and A are the most commonly found in the world population, most of the plasma used in IVIG production is from donors having these blood groups, with group B and group AB donors being fewer in number. Consequently, all batches of IVIG contain antibodies that are reactive with individuals of group A, group B, and group AB. These antibodies were originally discovered by Dr Karl Landsteiner in the early 1900s and are now known to consist of immunoglobulin (Ig)M, IgG, and IgA classes. As the process for producing IVIG results in almost exclusively IgG, isohemagglutinins contained in IVIG are of this immunoglobulin class. ABO antibodies are highly clinically significant and, because of this, blood bank cross-matching is done to ensure that blood of the correct type is transfused into recipients to avoid a socalled major mismatch or major incompatibility that can cause significant morbidity and often death. Administration of IVIG, which contains ABO antibodies, is often infused into individuals who have the corresponding ABO antigens, commonly called a minor mismatch, and although not as significant as a major mismatch, the isohemagglutinins contained in the IVIG have some risk for a significant transfusion reaction due to the ABO incompatibility. Indeed, currently there is no way to match IVIG to recipients according to blood type, so when IVIG is administered to group A, B, or AB recipients, there is potential for transfusion reactions analogous to a blood transfusion mismatch. For this reason, strict guidelines have been put into place to restrict the titers of the ABO antibodies contained in IVIG. This review will provide background information about the discovery and biochemistry of the ABO antigens and discuss the various isohemagglutinins that are found in plasma of the different ABO blood types and their potential clinical significance. In addition, a brief discussion of the controversial topic of the origins of these antibodies will conclude this review.
DISCOVERY OF THE ABO BLOOD GROUP SYSTEM
T
he 1930 Nobel Prize in Physiology or Medicine was awarded to Dr Karl Landsteiner “for his discovery of human blood groups.” Dr Landsteiner (Fig. 1) performed a simple experiment using blood from himself and his available staff. Published in 1901,1 Landsteiner found that when he mixed different combinations of red blood cells (RBCs) and plasma he obtained agglutination but with different patterns depending on the RBCs and plasma that were being mixed. One pattern of agglutination resulted in what Landsteiner called “A” and a different pattern he called “B.” There were two individuals’ RBCs that did not agglutinate with any of the plasma, which Landsteiner termed “C.” The “C” was eventually replaced with the letter “O” (that may, initially, have been a numerical zero to indicate zero reaction with plasma) and thus was born the ABO blood group system. One year later, colleagues in Landsteiner’s lab identified the AB blood group where the RBCs agglutinated with most plasma tested but the plasma did not agglutinate other RBCs.2 Although Landsteiner was awarded the Nobel Prize for his discovery of the ABO antigens and antibodies, he was not pleased with this accolade as he viewed his other work on viruses and haptens as more substantial and important.
LANDSTEINER’S LAW Landsteiner’s findings of four different “types” of blood based in good part on antibodies contained in plasma led
From the Centre for Innovation, Canadian Blood Services; and the Departments of Medicine and Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada. Address reprint requests to: Dr Donald R. Branch, Canadian Blood Services, 67 College Street, Toronto, Ontario, M5G 2M1, Canada; e-mail: [email protected]. doi:10.1111/trf.13087 C 2015 AABB V
TRANSFUSION 2015;55;S74–S79 S74
TRANSFUSION Volume 55, July 2015
ABO ISOHEMAGGLUTININS
him to postulate what has become known as “Landsteiner’s Law.”2,3 Simply put, Landsteiner’s Law states that whichever ABO antigens are lacking on a given person’s RBCs, that person will always have the corresponding antibody or isohemagglutinin.3 Thus, Group A individuals, lacking B antigen, always have anti-B in their plasma, while group B, lacking A antigen, always have anti-A. Group O, lacking both A and B antigens, always have both anti-A and anti-B, and group AB, having both A and B antigens, do not have any isohemagglutinins. The very rare Bombay phenotype has anti-A, -B, and -H, as this rare ABO blood group lacks A and B antigens but also lacks the H antigen (see below). In 1952, Dodd4 described an unexpected isohemagglutinin in serum that was only found in group O individuals. This antibody appeared to recognize both A and B antigens and was called a linked anti-A and anti-B.2,4 This antibody is now known as anti-A,B2,5-7 and reacts with a recognition determinant that is found on both A and B RBCs and can be identified only by absorption-elution studies.2,5,6 The anti-A,B follows Landsteiner’s Law so that it is found only in group O or Bombay individuals (Fig. 2).
ANTIGENS OF THE ABO BLOOD GROUP SYSTEM The antigens of the ABO system are controlled by the action of transferases.2,8 These transferases are enzymes that are responsible for the addition of specific sugars to an oligosaccharide precursor chain having a terminal galactose (Fig. 3). Group O individuals have an H gene located on chromosome 19 that encodes for a fucosyltransferase that adds a fucose sugar in a1,2-linkage to the terminal galactose on the precursor oligosaccharide, forming the so-called H antigen. H antigen forms a precursor oligosaccharide necessary to form A antigen and B antigen. Group A is formed by a gene located on Chromosome 9 that encodes for another transferase, a1,3N-acetylgalactosaminyltransferase, that adds N-acetylgalactosamine in a1,3 linkage to the terminal H-antigen. Group B gene, also encoded by Chromosome 9, produces a1,3-galactosyltransferase, which adds a galactose in a1,3 linkage to the H antigen. If both A and B genes are present, some H antigen will be converted to A antigen and some to B antigen creating the blood group AB. If the H gene is absent or defective, no H antigen can be formed and therefore no A or B antigen. These rare individuals are called Bombay (Fig. 3).
ABO ISOHEMAGGLUTININS: RELEVANCE TO INTRAVENOUS IMMUNOGLOBULIN ABO isohemagglutinins are mostly immunoglobulin (Ig)M antibodies but IgG and IgA classes also exist in plasma. The fractionation and manufacturing process for the pro-
Fig. 1. Dr. Karl Landsteiner (circa 1920): Father of the ABO blood group system and 1930 Nobel Prize winner for this pioneering discovery.
duction of intravenous immunoglobulin (IVIG) basically removes the majority of the IgM and IgA antibodies, leaving most of the IgG isohemagglutinins. As IVIG is fractionated from plasma pools containing a high proportion of group O donor plasma, IVIG must contain, in addition to anti-A and anti-B, anti-A,B, as this antibody is mostly found as IgG in group O people.2,5 IVIG is not expected to contain anti-H, given the rarity of the Bombay phenotype. There may be, however, some low levels of anti-A1, an antibody that specifically recognizes a majority of group A people and is made, again, according to Landsteiner’s Law, from a subset of group A individuals who lack the A1 phenotype2,5 (Fig. 2). All ABO isohemagglutinins have the potential for clinical significance.2,5 Both IgM and IgG classes can activate complement and cause intravascular lysis.2,5 IVIG contains almost exclusively IgG without IgM or IgA; thus, there is potential for clinical sequelae when IVIG is administered if the recipient is group A, B, or AB. Indeed, extravascular hemolysis has been shown to occur in ABO hemolytic disease of the fetus and newborn (HDFN) when anti-A, -B, or -A,B cross the placenta.5 Of note, anti-A,B can often be eluted from infants’ RBCs born to group O mothers and has been shown to be a highly clinically significant antibody.5 In one study, six of nine babies required treatment due to anti-A,B hemolytic disease of the fetus and newborn, whereas in babies having only Volume 55, July 2015 TRANSFUSION S75
BRANCH
Fig. 2. The five major ABO blood groups. Type A is found in approximately 35% to 36% of Caucasians. It follows Landsteiner’s Law by always having anti-B in the serum or plasma. Certain subtypes of A, that lack A1, can also have anti-A1. Type B is found in approximately 7% to 9% of Caucasians and follows Landsteiner’s Law by always having anti-A in the serum or plasma. Type AB is found in approximately 2% to 4% of Caucasians and has no isohemagglutinins in the serum or plasma. Type O is the most common ABO blood group comprising approximately 37% to 40% of Caucasians and follows Landsteiner’s Law by always having anti-A, anti-B, and anti-A,B in the serum or plasma. Finally, Type Bombay lacks the ability to add fucose to the oligosaccharide backbone and, thus, lacks H antigen, making it impossible to generate A or B antigens. Bombay is very rare, comprising less than one in 250,000 of the Caucasian population but can be found in one in 10,000 people in India. It follows Landsteiner’s Law by having not only anti-A, anti-B, and anti-A,B in the serum or plasma but also anti-H, as it lacks this ABO antigen.
anti-A or anti-B on their RBCs, only five of 14 required treatment.5 Although the contribution of anti-A,B to hemolysis after administration of IVIG to either group A or group B individuals has not been addressed, a significant number of ABO-associated hemolysis cases have been reported after IVIG administration and can be severe.9-12 One way manufacturers and regulatory agencies attempt to avoid the potential for hemolysis after IVIG therapy is to control the levels of anti-A and anti-B in the product.13 This can be achieved by limiting the potency of these antibodies by avoiding high titers. Thus, the Food and Drug Administration (FDA) has required IVIG not to have titers of anti-A and/or anti-B in IVIG higher than 64.13,14 In practice most manufacturers’ isohemagglutinin titers are lower than the FDA requirements. One potential problem with this approach is that all manufacturers and the FDA have ignored the anti-A,B isohemagglutinin conS76
TRANSFUSION Volume 55, July 2015
tained in IVIG. Indeed, when performing titration studies of IVIG against either A or B RBCs, there is no certainty that the resulting endpoint is due to anti-A or anti-B as at least the possibility exists of anti-A,B reactivity. This is because anti-A,B reacts with an as-yet-undefined, shared epitope on either A or B cells. This is an important concept as new-generation IVIG products have slightly higher titers, but within FDA guidelines, of anti-A and anti-B13 but the titers of anti-A,B are unknown. To obtain the specific titer of anti-A or anti-B in IVIG one would have to first adsorb the IVIG using group B cells to obtain an accurate titration of anti-A and, likewise, adsorb the IVIG with group A1 cells to obtain and accurate titration of the antiB contained in the IVIG. If one compares the titer results of the adsorbed against unadsorbed IVIG, one could get an idea of the titer of the anti-A,B in the IVIG preparation. More recently, manufacturers are turning to affinity chromatography to remove anti-A and anti-B from
ABO ISOHEMAGGLUTININS
Fig. 3. Schematic representation of the biochemical determinants of the ABO blood group system. The four blood groups are represented based on the addition of specific sugars by enzyme transferases to the backbone precursor oligosaccharide: Bombay lacking transferase activity for addition of fucose to the backbone oligosaccharide; group O, which requires a fucosyltransferase to add fucose to the oligosaccharide backbone to create the H antigen; Group A, requiring N-acetylgalactosaminyltransferase to add an N-acetylgalactosamine to the H antigen; and group B, which requires a galactosyltransferase to add a galactose to the H antigen. Red diamond 5 galactose; blue square 5 N-acetylglucosamine; black square 5 fucose; green diamond 5 Nacetylgalactosamine.
IVIG.13,15 Interestingly, complete exhaustion of these isohemagglutinins contained in IVIG preparations appears to be difficult if not impossible, despite multiple passages through the affinity columns.15 One can speculate that this is due to residual anti-A,B, which recognizes a chimeric antigen of both A and B components. The recognition determinate for anti-A,B has yet to be adequately elucidated so that currently available synthetic blood group antigens used in affinity chromatography do not contain the anti-A,B recognition carbohydrate sugars. The importance of anti-A,B cannot be overstated as many investigators have reported that this is a highly clinically significant isohemagglutinin that is almost exclusively of
the IgG subclass.2,5 Ongoing clinical studies of IVIGs having ABO isohemagglutinins reduced by donor selection and/or affinity chromatography will be helpful in deciding on the significance of anti-A,B contained in IVIG products. Most ABO isohemagglutinins, even if IgG, have the potential of causing complement-mediated, severe intravascular hemolysis. These ABO isohemagglutinins are known as ABO hemolysins.2,5 The fact that the antigen site density is very high for ABO antigens (Fig. 2) and that IgG antibodies can arrange themselves in close enough proximity for complement activation means that intravascular hemolysis is possible after administration of IVIG.9 It is possible that rare IVIG-associated hemolysis is not a result of low-titer anti-A, anti-B, or anti-A,B but high-titer hemolysin activity. Hemolysin activity is easy to test2 but is not routinely done for IVIG. However, hemolysin testing could be performed on individual lots of IVIG before release and may provide for an extra measure of safety. Briefly, using a two-stage method, one would add group A1, B, and O RBCs into separate tubes; add IVIG to each tube; and incubate at 37 C for 30 minutes. The tubes are removed and centrifuged, the supernatant is removed, an equal volume of freshly drawn normal group AB serum (can be previously fresh-frozen) is added to each tube as a source of complement, and the tubes are incubated again at 37 C for 30 minutes. The tubes are centrifuged and observed for reduction in size of the cell pellet and for free hemoglobin (Hb; pink to red supernatant), which can be quantified if desired. The group O tube is the negative control and should not show any evidence of RBC pellet reduction or free Hb. Any level of cell pellet reduction or pink color in either the A1 or the B tube and a negative finding in the O tube is considered a positive result, with the strength of the hemolysin activity dependent on the remaining cell pellet and intensity of the red color of the supernatant.2
ORIGINS OF ABO ISOAGGLUTININS The origin of the ABO isohemagglutinins is not really resolved. The question is whether these antibodies are produced through some inherited, “natural” innate mechanism, not requiring antigenic stimulation or, instead, follow classical adaptive immune-mediated mechanisms.2,16,17 An answer to this was attempted in the late 1950s and early 1960s when Springer and his associates17-19 published a series of landmark papers that suggested the ABO isohemagglutinins were produced by classical adaptive immunity in response to bacterial antigens. They showed that chickens kept in a germ-free environment would produce anti-B but not anti-A when fed bacteria expressing high levels of a B-like antigen and lower levels of an A-like antigen.17 Later Springer and colleagues18 showed that bacteria reacted with anti-A and Volume 55, July 2015 TRANSFUSION S77
BRANCH
TABLE 1. Comparison of natural and immune antibodies B-cell subset
Location
Innate
B1 (CD51)
Restricted
Adaptive
B2 (CD5-)
Systemic
Immune arm Natural antibody Immune antibody
Immunoglobulin class IgM IgG (no IgA, IgD, or IgE) IgG > IgM > IgA IgD, IgE
anti-B suggesting a cross-reactive antigen on bacteria responsible for induction of anti-A and anti-B via bacterial exposure. However, Springer and colleagues17,19 were only ever able to show weak anti-A production despite their many studies. Thus, the origin of these isohemagglutinins continued to perplex. Indeed, many studies before and after the work of Springer and colleagues have found antibodies, including ABO isohemagglutinins, detectable in neonates and newborns very close to birth that are not of maternal origin due to crossing the placenta and that cannot be adequately explained by an adaptive immune response to environmental stimuli.16,20-25 Even Landsteiner believed that ABO isohemagglutinins were spontaneously produced.2 With the discovery of two distinct subsets of B lymphocytes,16,26,27 one termed B1 (CD51), T-cell independent that does not require prior stimulation, responsible for “natural” antibody production, and the B2 subset (CD52), T-cell dependent and responsible for humoral responses via the adaptive immune system, the question of the origins of ABO isohemagglutinins has become more complex. Table 1 shows a comparison of “natural” versus immune-mediated antibody properties. Natural antibodies produced by B1 B lymphocytes have been proposed to include the ABO isohemagglutinins in addition to other antibodies produced to carbohydrate antigens as part of the innate immune system.16,28-31 Some investigators believe that ABO isohemagglutinins are initially produced “spontaneously” from a fixed set of ancestral germline genes found in B1 B lymphocytes.32 Indeed, newborns have mostly B1 B lymphocytes and produce only IgM antibodies, including ABO isohemagglutinins, early in life.16,20-25 Involvement of B2 B lymphocytes requires adaptive immunity, and it is known that IgG isohemagglutinins can be simulated, perhaps in response to bacterial and/or food antigens.16,18,33 Wuttke and coworkers16 found that IgM ABO isohemagglutinins of endogenous origin were always found in newborns by 8 months and usually much earlier, but by 8 months both B1 and B2 B lymphocytes were producing antibodies. The ABO antibodies found in IVIG are due to B2 B lymphocytes and an adaptive immune response.
T-cell dependent?
Antigenic stimuli
Plasma cells
Germline genes
No
Not required
No
Restricted germline
Yes
Requires antigen presentation and T-cell help
Yes
Rearrangement; hypermutation
the early 1900s with four major blood group antigens, A, B, AB, and O. The ABO isohemagglutinins, also discovered in the early 1900s, continue to follow Landsteiner’s Law without exception so that in individuals lacking an ABO antigen the isohemagglutinins corresponding to that antigen are always found in the plasma. It is plasma containing these isohemagglutinins that is used in the manufacture of IVIG and that can be potentially clinically relevant when IVIG is infused into non–group O individuals. Although the biochemistry of the ABO blood group system has been mostly worked out, elements still exist such as the epitope recognized by anti-A,B that still eludes investigators. Also what test may be the best predictor of potential clinical significance of ABO antibodies contained in IVIG or other plasma-based blood products remains a question: titration cutoff, hemolysin activity, anti-A,B levels. Although ABO antibodies can be IgM and IgG, elucidation of the origin of natural, IgM ABO antibodies compared to immune-mediated, IgG antibodies remains controversial.
CONFLICT OF INTEREST The author has disclosed no conflicts of interest.
REFERENCES 1. Landsteiner K. Ueber Agglutinationserscheinungen normalen menschlichen Blutes. Wien Klin Wochschr 1901;14: 1132-34. 2. Mollison PL. Blood transfusion in clinical medicine. 5th ed. Oxford: Blackwell Scientific Publications; 1972. 3. Springer GF, Tegtmeyer H. Apparent violation of “Landsteiner’s law.” Klin Wochenschr 1974;52:295-7. 4. Dodd BE. Linked anti-A and anti-B antibodies from group O sera. Br J Exp Pathol 1952;33:1-18. 5. Mollison PL. Blood transfusion in clinical medicine. 7th ed. Oxford: Blackwell Scientific Publications; 1983. p. 290-1. 6. Kajii E, Usuda S, Ikemoto S. Characterization of a monoclonal crossreacting anti-A,B antibody. Nihon Hoigaku Zasshi
SUMMARY The ABO blood group system has been studied for more than 110 years. The nomenclature has not changed since S78
TRANSFUSION Volume 55, July 2015
1990;44:227-33. 7. Moore S, Chirnside A, Micklern LR, et al. A mouse monoclonal antibody with anti-A,(B) specificity which agglutinates Ax cells. Vox Sang 1984;47:427-34.
ABO ISOHEMAGGLUTININS
8. Sheffield WP, Tinmouth A, Branch DR. Blood group biochemistry: a Canadian Blood Services research and development symposium. Transfus Med Rev 2005;19:295-307. 9. Michelis FV, Branch DR, Scovell I, et al. Acute hemolysis after intravenous immunoglobulin amid host factors of ABOmismatched bone marrow transplantation, inflammation, and activated mononuclear phagocytes. Transfusion 2014; 54:681-90. 10. Desborough MJ, Miller J, Thorpe SJ, et al. Intravenous immunoglobulin-induced haemolysis: a case report and review of the literature. Transfus Med 2014;24:219-26. 11. Berard R, Whittemore B, Scuccimarri R. Hemolytic anemia following intravenous immunoglobulin therapy in patients treated for Kawasaki disease: a report of 4 cases. Pediatr Rheumatol Online J 2012;10:10. 12. Markvardsen LH, Christiansen I, Harbo T, et al. Hemolytic anemia following high dose intravenous immunoglobulin in patients with chronic neurological disorders. Eur J Neurol 2014;21:147-52.
20. De Biasi B. Studies on iso-agglutinins in the blood of the new-born. JAMA 1923;81:1776-8. 21. Thomaidis T, Fouskaris G, Matsaniotis N. Isohemagglutinin activity in the first day of life. Am J Dis Child 1967;113: 654-7. 22. Chattoraj A, Gilbert R, Josephson AM. Serological demonstration of fetal production of blood group isoantibodies. Vox Sang 1968;14:289-91. 23. Godzisz J. [Synthesis of natural allohemagglutinins of the ABO system in healthy children aged 3 months to 3 years]. Rev Fr Transfus Immunohematol 1979;22:399-412. French. 24. Mencarini L, Tozzi C, Arachi S, et al. [Autochthonous anti-A/ B allohemaggluntinins in umbilical cord blood]. Minerva Pediatr 1982;34:83-5. Italian. 25. Merbl Y, Zucker-Toledano M, Quintana FJ, et al. Newborn humans manifest autoantibodies to defined self molecules detected by antigen microarray informatics. J Clin Invest 2007;117:712-8. 26. Martin F, Kearney JF. B1 cells: similarities and differences
13. US Food and Drug Administration. Public workshop—strat-
with other B cell subsets. Curr Opin Immunol 2001;13:195-
egies to address hemolytic complications of immune globulin infusions. Silver Spring (MD): FDA; 2014 [accessed 2014
201. 27. Montecino-Rodriguez E, Leathers H, Dorshkind K. Identifi-
Oct 5]. Available from: http://www.fda.gov/biologicsbloodvaccines/newsevents/workshopsmeetingsconferences/ ucm378388.htm 14. Human normal immunoglobulin for intravenous administration. In: European pharmacopoeia. 3rd ed. Strasbourg: Council of Europe; 1997. p. 963-5. 15. Hoefferer L, Glauser I, Gaida A, et al. Isoagglutinin reduction by a dedicated immunoaffinity chromatography step in the manufacturing process of human immunoglobulin products. Transfusion 2015;55 (Suppl 2):S117-S121. 16. Wuttke NJ, Macardle PJ, Zola H. Blood group antibodies are made by CD51 and by CD5- B cells. Immunol Cell Biol 1997;75:478-83. 17. Springer GF, Horton RE, Forbes M. Origin of anti-human blood group B agglutinins in white Leghorn chicks. J Exp Med 1959;110:221-44.
cation of a B-1 B cell-specified progenitor. Nat Immunol. 2006; 7:293-301. 28. Parker W, Lundberg-Swanson K, Holzknecht ZE, et al. Isohemagglutinins and xenoreactive antibodies: members of a distinct family of natural antibodies. Hum Immunol 1996;45: 94-104. 29. Holodick NE, Tumang JR, Rothstein TL. Immunoglobulin secretion by B1 cells: differential intensity and IRF4dependence of spontaneous IgM secretion by peritoneal and splenic B1 cells. Eur J Immunol 2010;40:3007-16. 30. Baumgarth N, Tung JW, Herzenberg LA. Inherent specificities in natural antibodies: a key to immune defense against pathogen invasion. Springer Semin Immunopathol 2005;26: 347-62. 31. Bovin NV. Natural antibodies to glycans. Biochemistry (Mosc) 2013;78:786-97.
18. Springer GF, Williamson P, Brandes WC. Blood group activity of gram-negative bacteria. J Exp Med 1961;113:1077-93.
32. Arend P. Ancestral gene and “complementary” antibody dominate early ontogeny. Immunobiology 2013;218:755-61.
19. Springer GF, Horton RE. Blood group isoantibody stimula-
33. Yamamoto S. The occurrence of materials cross-reacting
tion in man by feeding blood group-active bacteria. J Clin Invest 1969;48:1280-91.
with anti-A and -B agglutinins in fruit or seed extracts of higher plants. J Immunogenet 1977;4:325-30.
Volume 55, July 2015 TRANSFUSION S79
