Williams Hematology



Definitions and History
Blood Group Systems

Abo Blood Group System

Rh Blood Group System

Other Blood Group Systems
General Immunology of Blood Group Antigens

Antigen Expression

Biochemistry of Erythrocyte Antigens

Carbohydrate Antigens

Protein Antigens

Effect of Enzymes and Other Chemicals on Erythrocyte Antigens
Genetics of Erythrocyte Antigens


Gene Complexes

Silent Alleles

Gene Frequencies
Distribution of Red Cell Antigens in Health and Disease

Expression of Red Cell Antigens in Other Body Tissues and Fluids

Associations of Red Cell Antigens with Disease
Antierythrocyte Antibodies

Immunology of Red Cell Antibodies

Clinical Significance of Erythrocyte Antibodies
Serologic Detection of Erythrocyte Antigens and Antibodies



Extended Antigen Phenotyping

Antibody Screen

Direct Antiglobulin Test

Compatibility Testing

Antibody Identification
Chapter References

Human red blood cells (RBC) bear numerous cell surface structures that can be recognized as antigens by the immune system of individuals who lack that particular structure. The characterization of RBC antigens and antibodies has been the basis of compatibility testing in the blood transfusion laboratory, thereby minimizing the risk of hemolytic transfusion reactions. Such knowledge has also provided the scientific basis for understanding hemolytic disease of the newborn and autoimmune hemolytic anemias. In recent years, the biochemical and molecular bases for many erythrocyte antigens have been elucidated, and this has led to further definition of their biologic functions. Blood group antigens play a critical role in susceptibility to infection by malarial parasites and also in some viral and bacterial infections. The absence of certain RBC antigens is associated with specific clinical disorders, and the recognition of these associations has led to an improved understanding of the function of antigens in the RBC membrane. Diverse inherited and acquired disorders are associated with alteration of RBC antigen expression, and these alterations often play a critical role in the clinical manifestations of these disorders. Thus, although erythrocytes have traditionally been considered relatively inert cellular containers of hemoglobin, they are in fact active in a variety of physiologic processes.

Acronyms and abbreviations that appear in this chapter include: AET, 2-aminoethyl-isothiouronium bromide; DTT, dithiothreitol; GP, glycoprotein; GPC, glycophorin C; GPD, glycophorin D; HLA, human leukocyte antigens; HEMPAS, hereditary erythroblastic multinuclearity with a positive acidified serum test; 2-ME, 2-mercaptoethanol; RBC, red blood cell; Rh, Rhesus.

A blood group system is a group of antigens encoded by alleles at a single gene locus or at gene loci so closely linked that crossing over does not occur or is very rare. An antigen collection is a group of antigens that are phenotypically, biochemically, or genetically related, but their genes are not known to be allelic.1
The placement of a blood group antigen into a system or collection follows a natural progression.2 First, its antibody is discovered, usually in the serum of a multiparous woman or a multiply transfused recipient, and found to have unique specificity. Using traditional serologic methods, the antibody can be used to study basic biochemical properties of the corresponding antigen, its pattern of inheritance, and its gene frequency, and to search for an antithetical antigen. The biochemistry and structure of the newly recognized antigen can also be evaluated using modern biochemical and molecular genetic methods. Identified characteristics are then compared to known systems and collections, and an appropriate assignment is made.
The initial naming of blood group antigens does not always follow the classical convention wherein dominant traits are given capital letters and recessive traits are designated with lower case letters. The gene in the ABO blood group system, for example, that determines the recessive O phenotype is designated O, while the genes S and s in the MNS system are codominant. Typically, red cell antigens are given alphabetical designations (e.g., CDE of the Rh system) and/or are named after the family of the antibody producer (e.g., Kell for Mrs. Kellacher and Fy for Mr. Duffy).
To help standardize red cell blood group terminology, the International Society of Blood Transfusion (ISBT) now uses a numerical system based on nomenclature first proposed by Rosenfield et al. Each system and collection has been given a number and letter designation, and each antigen within the system is numbered sequentially in order of discovery. To date, 23 blood group systems and 5 antigen collections are defined (Table 137-1).1,3,4 and 5 High-incidence (or public) antigens and low-incidence (or private) antigens that are not associated with known systems or collections also are grouped into numbered series.


See Table 137-2 for a summary of the characteristics of common erythrocyte antigens and Refs. 4 through 9 for comprehensive details.


The ABO blood group system was the first system to be described and remains the most significant one in transfusion medicine. Erythrocytes from most normal individuals phenotype as A, B, AB, or O, the latter indicating a lack of A and B antigen. The sugars defining A and B antigen are found on precursor carbohydrate chains carrying the H antigen, which is produced by a gene product from another chromosome. Because H is a required precursor that becomes “hidden” when A or B sugar is added, group A or B erythrocytes appear to have less H than group O cells. Nonetheless, H is found on all human erythrocytes except those in rare individuals of the Oh (Bombay) phenotype, who lack the H (or FUT1) gene. This system is therefore commonly referred to as ABH.
Normal individuals who lack either A or B antigen consistently make anti-B or anti-A, respectively, within several months after birth. These antibodies can cause intravascular hemolysis of ABO incompatible erythrocytes and are associated with severe acute hemolytic transfusion reactions and death. Because these antigens also are expressed on most tissue cells, ABO compatibility is a significant consideration in solid organ transplantation. However, ABO incompatibility only rarely causes clinical hemolytic disease of the newborn, presumably because antibodies directed against the A and B antigen are predominantly IgM, which do not cross the placenta, and because A and B antigens are not well developed at birth.
The Rh (Rhesus) system is the second most important blood group system in transfusion medicine because antigen-positive erythrocytes so frequently immunize antigen-negative individuals through transfusion and pregnancy.
The inheritance of Rh antigens is determined by a complex of two closely linked genes: one encodes the protein carrying D antigen (RHD), the other encodes the protein carrying C or c and E or e specificity (RHCE). People who are Rh- positive have both RHD and RHCE genes, whereas Rh-negative individuals have only the RHCE gene. Depending on the Rh genes present on a chromosome, eight common antigen combinations or haplotypes are possible: Dce (Rh0), DCe (Rh1), DcE (Rh2), DCE (Rhz), ce (rh), Ce (rh’), cE (rh”), and CE (rhy). The letter “d” is commonly used to designate the lack of D, but there is no d antigen and anti-d has never been found.
Several nomenclatures can be used to describe Rh genes and antigens. Fisher-Race nomenclature, which uses CDE terminology, more commonly is used for antigens; Wiener nomenclature, which uses Rh designations, is favored for haplotypes and gene complexes. A person who inherits r (ce gene) from one parent and R1 (D and Ce genes) from the other parent expresses D, C, c, and e antigens on his or her erythrocytes.
Rh is the largest blood group system, but its most important and immunogenic antigen is D (Rh0 in Weiner terminology, referring to its discovery using a Rhesus monkey antibody to human erythrocytes). For most clinical purposes, it is sufficient to test individuals for the D antigen and classify them as D+ or Rh-positive, or D– or Rh-negative. Approximately 85 percent of the Caucasian population is Rh positive, and 15 percent is Rh- negative. Because most Rh-negative recipients will produce anti-D if they receive a unit of Rh-positive blood and because anti-D can cause acute hemolytic transfusion reactions and severe hemolytic disease of the newborn, all donors and recipients are routinely typed and matched for D to avoid sensitization from transfusion. Also, Rh immunoglobulin specific for D is given routinely to Rh-negative mothers bearing Rh-positive infants to prevent immunization to the D antigen.
The antigens C, c, E, and e are less immunogenic and become important in patient care only after the corresponding antibody develops or when basic Rh haplotypes must be determined. The remaining 40 antigens represent other Rh protein epitopes whose corresponding antibodies are seldom encountered. Some are encoded by variant Rh alleles and appear as antithetical antigens to C, c, E, or e or as related “extra” antigens; others are referred to as “compound” antigens or cis gene products. For example, the protein produced by the gene ce encodes c, e, and f (or ce) antigen; other compound specificities include Ce (rhi), cE, CE, V (ces), and Ces. Still other Rh antigens are related to the complex “mosaic” nature of D and e antigen. If immunized, individuals who lack a subpart of D or e and make antibody to the portion they lack can present with a challenging serologic picture. For example, the D+ person who lacks part of the D epitope and makes an antibody to the missing portion appears to make alloanti-D because normal D+ erthrocytes carry all D epitopes.
Some, but not all, individuals who lack part of the D antigen (partial D) have a weak expression of D on their red cells that is only detected with more sensitive antiglobulin testing. Having a C gene in transposition to a D gene (e.g., Dce/Ce or DCe/Ce genotypes) also can weaken the expression of D in some individuals. A third type of weak D expression results from inheriting a D gene that encodes all epitopes of D but in less than normal quantity.
The other blood group systems and their antigens become important when antibody develops and transfusion is needed or when hemolytic disease of the newborn is a concern. Blood bank laboratories identify the specificity or characterize the reactivity of all antibodies detected in routine testing. Once this basic information is known, the blood bank assesses the clinical significance of the antibody and selects the most appropriate blood for transfusion. Occasionally, disease or observed red cell anomalies warrant studying red cell antigen expression in more detail. More extensive antigen typing sometimes offers clues to broader diagnoses.
An antigen is a substance that can evoke an immune response when introduced into an immunocompetent host and that can react with the antibody produced from that immune response. Its structure and stereochemical fit with its antibody are key to its specificity. An antigen can have several epitopes, or antigenic determinants, each of which is capable of eliciting an antibody response.
The ability of an antigen to stimulate an immune response is called immunogenicity, and its ability to react with an antibody is called antigenicity. These primary characteristics are affected by antigen size, shape, rigidity, and the number and location of the determinants on the red cell membrane. The ability of a host to recognize and respond to foreign antigens is influenced also by its HLA-DR alleles, as well as by genes outside the HLA system.6
The number of antigen sites per erythrocyte has been estimated by measuring the uptake of125 I-labeled antibody or of ferritin-conjugated anti-IgG.7 Numbers vary widely among blood group systems: from 1 to 2 × 106 sites for ABH to 2 to 6 × 103 for Lea, K, and Lub.
Most erythrocyte antigens can be detected early in fetal development (ABH at 5–6 weeks’ gestation and other specificities by 12 weeks), but not all are fully developed at birth. ABH, I, P1, Lua, Lub, Yta, Xga, Vel, Bg, Knops, and Dombrock antigen expression is considerably weaker on cord erythrocytes than in adults, and Lea, sometimes Leb, Ch/Rg, AnWj, and Sda are not routinely detected, although 50 percent of cord samples type Le(a+) with more sensitive test methods.6 About 2 years may pass before adult expression of ABH, I, and Lewis antigens is detected, and 7 years or more for P1 and Lutheran antigens.
Individuals who are homozygous for an allele typically have a greater number of antigen sites than do individuals who are heterozygous. Consequently, their erythrocytes can react more strongly with antibody. This difference in expression and antigen-antibody reactivity because of zygosity is known as dosage. For example, red cells from a homozygous MM individual carry a double dose of M antigen and react more strongly with anti-M than do red cells from a MN heterozygous individual carrying only a single dose of M. Antithetical antigens Cc, Ee, Kk, MN, Ss, and JkaJkb commonly show dosage.
Dosage is less obvious with D and LuaLub antigens. It may be very apparent within a family but not between families because D and Lu expression and reactivity can be unique family traits. Dosage within the Duffy system also may not be serologically obvious because Fy(a+b–) or Fy(a–b+) phenotypes are seen in either homozygous (FyaFya or FybFyb) or heterozygous (FyaFy or FybFy) individuals.
Some blood group antigens are inherited as very closely linked genes or haplotypes. Haplotype pairings and gene interaction (either cis or trans) also can effect phenotypic expression. For example, the pairing of C in trans position to D can result in weak expression of D (see discussion of Rh Blood Group System above), but having E in cis position with D is associated with strong D expression. Indeed, R2R2 red cells carry the strongest expression of D. In the Kell system, Kpa is associated with weakened expression of cis k and Jsb.
Still other antigens are affected by regulator genes.10 In(Lu), also called SYN-1B, is a dominant inhibitor gene that suppresses the expression of Lutheran antigens, P1, i, and many other antigens;11 the dominant inhibitor In(Jk) suppresses the expression of Jka and Jkb.12 Rare variants of the Rh regulator gene X1 can depress or shut off the expression of the Rh antigens (see “Rhnull Syndrome”).
Immunogenicity depends on many antigen characteristics, not just the number of antigen sites. For example, a K+Fy(a+) erythrocyte carries 7 to 17 × 103 Fya antigen sites on its membrane, but only 3 to 6 × 103 K sites. Yet such cells are nine times more likely to stimulate the formation of anti-K than of anti-Fya.6
Relative immunogenicity is estimated by comparing the actual frequency with which an antibody is found to the calculated frequency of a possible immunizing event. Although numbers vary, researchers agree that after A and B, the D antigen is most immunogenic (>80 percent of Rh-negative individuals produce anti-D after receiving a single Rh-positive unit9), followed by K, which stimulates anti-K in 10 percent of cases.9), followed by K, which stimulates anti-K in 10 percent of cases.6 The antigens c and E are three times less immunogenic than K, Fya is 25 times less potent, and Jka is 50 to 100 times less potent.13
An antibody typically recognizes an epitope consisting of four to five amino acids or one to seven sugar residues on linear polypeptides or polysaccharides, respectively. Alternatively, the antibody-binding site may encompass a more complex three-dimensional structure with branches or folds, and recognition may depend on both amino acid and sugar moieties. Table 137-2 and Table 137-3 and Fig. 137-1 provide a summary of blood group biochemistry and antigen structure.5,6,7 and 8,14


FIGURE 137-1 Erythrocyte membrane structures carrying blood group activity. (Modified from PD Issitt and DJ Anstee,4 and PL Mollison.6)

Polysaccharides with blood group activity are made by the sequential addition of specific sugars to specific precursors by specific transferase enzymes encoded by genes. Sugars commonly involved are D-galactose (Gal), N-acetyl-D-galactosamine (GalNAc), N-acetyl-D-glucosamine (GlcNAc), L-fucose (Fuc), and N-acetyl-neuraminic acid (NeuNAc, or sialic acid).
ABO, Lewis, and P blood group specificity depends on a terminal or immunodominant sugar, the polysaccharide to which the sugar is attached, and the type of linkage involved. I/i specificity is defined by a series of sugars on the inner portion of ABH oligosaccharide chains. The presence of at least two repeating Gal(b®4)GlcNAc(b® 3)Gal units in a linear structure defines i activity. I activity involves these same sugars in branched form. The I gene may actually encode the transferase responsible for branching [b(1–6)glucosaminyltransferase].
Oligosaccharide chains are attached to glycoproteins in secretion, to glycolipids in plasma, and to both on the erythrocyte membrane. About 70 percent of ABH-I antigens on the membrane are carried on glycoproteins, primarily the anion transporter band 3, but also on the glucose transporter band 4.5, the Rh glycoprotein, and others. About 10 pecent are on NeuNAc-rich glycoproteins; 5 percent on simple glycolipids, and the remainder on polyglycosyl ceramide.6 P, Pk, and P1 antigens is found on glycolipids both on the membrane and in plasma; whether they are carried on membrane glycoproteins is disputed.4
Lewis antigens are unique in that they occur only on type 1 polysaccharide chains, precursors found in plasma and secretions but not made by erythrocytes. Hence, they are plasma and secretion antigens and only exist on red cells by adsorption of Lewis substance from plasma. The Le (or FUT3) gene encodes an a(1–4)fucosyltransferase. Whether the resulting antigen is Lea or Leb depends on the secretor gene, Se (or FUT2), that encodes an a(1–2)fucosyltransferase.
Protein structures that carry blood group antigens can be grouped into three categories: (1) those that make a single pass through the erythrocyte membrane, (2) those that make multiple passes through the membrane, and (3) those inserted into the membrane through a covalently linked lipid.
Single-pass proteins include glycophorin A with its MN antigens, glycophorin B with SsU antigens, glycophorin C and D with Gerbich antigens, and the proteins encoded by Kell, Lutheran, LW, Indian, Knops, and Xg genes. Most of these proteins have an extracellular amino terminus and an intracellular carboxyl terminus (referred to as type I). An exception to this is Kell glycoprotein, where the terminal positions are reversed: the carboxyl terminus is extracellular, and the amino terminus is intracellular (type II).
Most proteins that make multiple passes through the erythrocyte membrane have both carboxyl- and amino-terminal ends that are intracellular, are very hydrophobic, and have a transport function. Rh, Diego, Colton, Kidd, and Kx proteins are included in this category. The gene product of the Duffy gene is also a multipass protein, but it has an extracellular amino terminus and has homology with a family of cytokine receptors.
Lipid-linked proteins have their carboxyl terminus replaced with the lipid glycosylphosphoinositol and are said to be GPI-linked or anchored. Cromer, Cartwright, Dombrock, and JMH proteins belong to this category. GPI-linked proteins are of special interest to hematologists because defective synthesis of the GPI anchor is responsible for paroxysmal nocturnal hemoglobinuria.15
The biochemical structure of an antigen and its location on the RBC membrane affect its expression on erythrocytes treated with enzymes and other chemicals. These reagents are used in serologic testing to help identify complex mixtures of antibodies and to help characterize antibody specificity when identity is not apparent.
Common proteolytic enzymes, such as ficin, papain, and bromelin, cleave protein from the erythrocyte membrane. This action consequently destroys accessible protein antigens and allows carbohydrate and more protected protein antigens to react more strongly with their antibody. The reactivity of ABH, I, P, Lewis, Rh, and Kidd antigens is strongly enhanced with enzyme treatment, while MN, Fya, Fyb, and many minor antigens (Xga, Ch, Rg, JMH, Indian, Pr, Tn, Ge2, Ge4, and some examples of Yta ) are destroyed. Ss can be destroyed with very strong enzyme treatment; Kell and Lutheran antigens are relatively unaffected.4,5
Reagents that reduce disulfide bonds, such as 2-mercaptoethanol (2-ME), dithiothreitol (DTT), and 2-aminoethylisothiouronium bromide (AET), destroy Kell blood group antigens but enhance Kx. Some researchers report that reducing reagents also denature the minor antigens LW, Scianna, Indian, JMH, and Yta, and weaken Lutheran, Dombrock, Cromer, Knops, AnWj, and MER2 antigens.4,5
Chloroquine treatment of erythrocytes at room temperature has little effect on most antigens. However, treatment for 30 min at 37°C (98.6°F) can weaken the expression of many antigens, including Fyb, Lub, Yta, JMH, and those in the Rh, Dombrock, and Knops systems.5
Protein antigens are considered direct gene products: the gene encodes a specific protein that expresses one or more antigenic epitopes. Carbohydrate antigens, made by transferase action, are considered indirect gene products. Most blood group genes are located on autosomes; only two, Xg and Xk, are located on the X chromosome (see Table 137-1 for gene and chromosome location).
Most genes that encode blood group antigens have two or more alleles. Individuals who inherit two identical alleles are homozygous and make a double dose of a single gene product, while those who inherit two different alleles are heterozygous and make two gene products. Males are hemizygous for the genes located on their single X chromosome.
Alleles commonly arise from DNA base-pair mutations or deletions. For example, A and B alleles differ from one another by seven DNA base substitutions, which result in four amino acid substitutions in their respective transferases.4,5 The common O allele is similar to A except for a single base deletion at nucleotide 261 that shifts the reading frame during RNA translation. The resulting protein is truncated and has no transferase activity. Another variant O allele encodes a transferase identical to that of B except it has arginine instead of alanine at amino acid position 268, which also may block enzyme activity. Alleles in other blood group systems arise in similar fashion.
Some blood group genes are complexes of several closely linked genes or loci that evolved through duplication of an ancestral gene. The antigens they encode are inherited within families as a packet or haplotype with few or no cross-overs. Blood group examples include the Rh system, with its genes RHD and RHCE, which encode D and CE proteins, respectively, and the MNS system, with its genes GYPA and GYPB, which encode the proteins glycophorin A and glycophorin B.
The ancestral gene for Rh is not known, but duplication appears to have taken place in early evolution in a common man-ape lineage; RHD and RHCE show remarkable homology. GPYA and GPYB probably arose by duplication of an ancestral GPYA gene encoding the N antigen,16 but there is less homology between the two linked genes. The most common MNSs complex is Ns, followed by Ms, MS, and NS.
In both Rh and MNS systems, other antigens arose by further mutations, deletions, and rare cross-over within the gene complex. Unequal pairing of GYPA and GYPB during meiosis, with subsequent recombination, has resulted in hybrids such as GYP(A-B) (called Lepore type, in analogy with a similar hemoglobin hybrid), which encodes a protein with the amino-terminal end of glycophorin A but the carboxyl-terminal end of glycophorin B. Anti-Lepore–type hybrids GYP(B-A-B) and GYP(A-B-A) are also known. Within the Rh complex, hybrids of RH(D-CE-D) and RH(CE-D-CE) have been identified.4,5 Such hybrids can result in altered antigen expression and new antigen epitopes.
Kell and Lutheran antigens also were thought to arise from large gene complexes of four or more loci, each having two or more alleles: K/k, Kpa/Kpb/Kpc, Jsa/Jsb, and K11/K17 for Kell and Lua/Lub, Lu6/Lu9, Lu8/Lu14, and Aua/Aub for Lutheran. It is more appropriate to now regard Kell and Lutheran proteins as single gene products that carry multiple antigenic epitopes. The early and still most common alleles in humans (kKpbJsbK11 and LubLu6Lu8Aua) encode antigens having very high frequencies in the population. Antigens of lower frequency (K, Kpa/Kpc, Jsa, Lua, Lu9, Lu14, and Aub) most likely arose from point mutations in different populations.
Some blood group alleles are said to be amorphic, or silent; that is, they do not produce a recognizable erythrocyte antigen, although they may indeed encode a product that is simply not detected with standard red cell test methods. As already discussed with regard to the ABO system, A and B genes produce transferases that add GalNAc or Gal, respectively, to the same precursors, but O produces no active enzyme. AB individuals express both A and B antigen, but AA and AO individuals express only A, and BB and BO individuals express only B. Amorphic alleles are only recognized in a homozygous state (i.e., group O individuals are OO), and the result is a “null” phenotype. Null phenotypes exist in all blood group systems; group O is the most common, followed by Fy(a–b–) and Le(a–b–) in Africans. All other nulls are quite rare.
The Fy(a–b–) phenotype is especially interesting. Although it was once thought that this phenotype represented the inheritance of two silent alleles, FyFy, it has since been shown that Fy(a–b–) Africans commonly have Fyb genes that express normal Fyb glycoprotein on tissue cells but not on erythrocytes. A mutation that disrupts the GATA-1 binding site for red cell transcription has been identified in these individuals, which helps explain why many Fy(a–b–) Africans do not make Duffy antibodies despite exposure to antigen-positive erythrocytes from transfusion.
Gene and phenotype frequencies vary widely with race and geographical boundaries.5,9,17 This information is needed when estimating the availability of compatible blood and the probability of hemolytic disease of the newborn, and in paternity and forensic investigations.
Antigens in the Rh and Kidd blood group systems are present only on erythrocytes and have not been detected on platelets, lymphocytes, or granulocytes or in plasma, other body tissues, or secretions (saliva, milk, amniotic fluid, etc.).4,5 and 6 MNSs, Lutheran, Kell, and Duffy antigens also are erythrocyte specific but have been found in other body tissues (see Table 137-2).
In contrast, ABH antigens have broad tissue distribution. In young embryos, they can be detected on all endothelial cells and all epithelial cells except those of the central nervous system. ABH, Lewis, I, and P blood group antigens are in plasma and on platelets and lymphocytes; granulocytes carry I antigen but no ABH.6 ABH on platelets and lymphocytes may be acquired at least in part by adsorption of plasma antigen; Lewis antigen is acquired only by adsorption. Body secretions (saliva, milk, etc.) contain ABH, I, and Lewis antigen but no P system antigens. Sda antigen is found in most body secretions, with the greatest concentrations seen in urine.4,6
Some blood groups are statistically associated with medical conditions or disease (Table 137-4).4,5 and 6, 10,18,19,20,21 and 22 For example, blood group A is more common in persons with cancer of the salivary glands, stomach, colon, or ovary and with thrombosis (due to higher levels of factors VIII, V, and IX). Blood group O is more common in patients with duodenal and gastric ulcers, rheumatoid arthritis, and von Willebrand disease. The adult i phenotype, especially in Asians, appears genetically linked to congenital cataracts.4,18,20 These statistical observations may not be of clinical significance.


Associations with infection arise when microorganisms carry structures with blood group activity. Yersinia pestis carries H-like antigen, and the smallpox virus is associated with A-like antigen, making group O and A individuals, respectively, more susceptible. The presence of blood group antibody and/or soluble blood group antigen in secretions may help confer protection. Having anti-B may offer protection against Salmonella, Shigella, Neisseria gonorrhoeae, and some Escherichia coli O86 infections. There is an association between nonsecretion of ABH antigen and susceptibility to Candida albicans, Neisseria meningitidis, Streptococcus pneumoniae, and Haemophilus influenzae.20
A number of disease associations with globoside structures have been identified. Streptococcus suis, which can cause meningitis and septicemia in humans, binds exclusively to Pk antigen. A class of toxins secreted by Shigella dysenteriae, Vibrio cholerae, and Vibrio parahaemolyticus also have binding specificity for Gal(a®4)-Galb(®4). In addition, globoside is the receptor of human parvovirus B19.8 Some strains of E. coli use a disaccharide receptor, Gal(a®4)-Galb, on uroepithelial cells to gain entry to the urinary tract receptors associated with P1, P, and Pk antigens.4,20,24,25 People with the rare p phenotype (P null) lack this disaccharide and are not susceptible to acute pyelonephritis from such E. coli strains.
Erythrocytes lacking Fya and Fyb antigen are not infected by the malarial parasite Plasmodium vivax and the simian parasite Plasmodium knowlesi. The parasites attach to the red cell membrane, but junction (the joining of merizoite and erythrocyte membrane) and penetration cannot take place. The Fy6 antigen, which Fy(a–b–) red cells lack, appears to be the critical receptor for P. vivax penetration.4,21,23
Plasmodium falciparum invasion is not dependent on Duffy antigen; instead, it is associated with red cell glycophorins and their O-linked oligosaccharides (carrying NeuNAc). Red cells with the following phenotypes have shown a decreased rate of infection: M-N- (GPA deficient), S-s-U- (GPB deficient), Ge- (Leach type or GPC deficient), and Cad- and Tn-positive red cells (with abnormal O-linked sugars).21
Altered antigen expression can occur with inherited and acquired disease. Inherited changes are fixed and consistent; acquired changes can disappear with remission or recovery. In some diseases, antigen expression weakens; in others, it increases or new antigens appear.
Weakened ABH expression on red cells has been noted in acute myeloid leukemias and may be due to reduced transferase activity.4,6 Normal antigen expression returns with disease remission. Transient weakened expression of target antigen also has been reported in some cases of autoimmune hemolytic anemia. Weak Rh, Kell, and Kidd blood group activity has been reported with concurrent autoantibody.4,6
Increased expression of i on red cells is associated with inherited disorders, such as thalassemia, sickle cell disease, Diamond-Blackfan syndrome, or hereditary erythroblastic multinuclearity with a positive acidified serum test (HEMPAS). Increased i expression also is noted with acquired conditions that decrease the red cell maturation time in the marrow, such as myeloblastic or sideroblastic erythropoiesis, refractory anemia, or excessive phlebotomy.6,19,20
Expression of the crypt-antigen Tn is caused by a galactosyltransferase deficiency acquired by somatic mutation in a population of stem cells. The antigen is present on all red cells, platelets, and granulocytes arising from these stem cells. This condition (seen as persistent mixed-field agglutination because of the presence of both normal and abnormal cell clones) causes other red cell abnormalities, such as depressed MN expression, enhanced H, and reduced NeuNAc content. It is associated with preleukemia and acute myelomonocytic leukemia.6,20 Other crypt-antigens (T, Tk) are seen as a result of infection when microbes produce enzymes that remove some sugars and expose new ones. Group A individuals can appear to acquire a B antigen when bacterial deacetylase removes the acetyl group on GalNAc.26,27 This phenomenon is associated with severe infection, gastrointestinal lesions, or malignancies.
Red cells also may acquire new blood group activity when they adsorb membrane material from certain microorganisms.19,20 Group B activity has been associated with E. coli86 and Proteus vulgaris infection and K activity with Enterococcus faecium. Acquired Jkb-like activity has been associated with E. faecium and Micrococcus infections, although the mechanism is not clear.
Rhnull Syndrome The Rhnull phenotype is associated with hereditary stomatocytosis, hemolytic anemia (usually mild), and a lack of proteins carrying Rh antigen. Rh protein resides in the red cell membrane, interacts with other membrane glycoproteins and possibly the cytoskeleton, and may help regulate or organize the lipids within the red cell membrane bilayer.22 Hence, it is important to membrane shape as well as the expression of other antigens. Rhnull cells have depressed expression of SsU, LW, and Fy5 antigens.
Most Rhnull red cells are stomatocytes or occasionally spherocytes and demonstrate an increased osmotic fragility, increased potassium permeability, and higher potassium pump activity. They have reduced cation and water content and a relative deficiency of membrane cholesterol.22 While it is assumed that these abnormalities contribute to shortened in vivo survival, Rhnull red cells survive normally in splenectomized patients, suggesting their removal is related more to splenic clearance because of shape rather than some other intrinsic factor.22
Two genetic mechanisms account for the Rhnull phenotype. Persons with the amorphic type are homozygous for the silent Rh gene
that encodes no Rh protein. Individuals with the more common regulator type of Rhnull have normal Rh genes but are thought to be homozygous for X0r, a rare allele of the normal Rh regulator gene Xlr that switches off Rh gene expression.10,22
Individuals with the Rhmod phenotype have similar membrane and clinical anomalies associated with Rhnull syndrome but demonstrate some Rh antigen expression. Their hypothesized regulator gene XQr may represent the same recessive regulator gene of Rhnull individuals but with less penetrance.22
McLeod Phenotype More than 60 males and no females with the McLeod phenotype have been identified. Each of these individuals had acanthocytosis, decreased red cell survival, very weak expression of Kell blood group antigens, and a lack of Kx antigen on their erythrocytes. Their hemolytic anemia may be well compensated.20,22
Kx, a 37-kD protein encoded by the Xk gene on the X chromosome, interacts with the cytoskeleton and helps stabilize the membrane, much as does Rh protein. The absence of Kx is associated with a lipid deficiency in the membrane bilayer that may be critical to the Kell antigen protein and general membrane shape. Red cells with the McLeod phenotype also show a defect in water transport, increased mobility of phosphatidylcholine across the membrane, and increased phosphorylation of protein band 3 and the B band of spectrin.28
After age 40, patients with the McLeod phenotype develop a slowly progressive form of muscular dystrophy that is associated with areflexia, choreiform movements, and cardiomegaly, leading to cardiomyopathy. They have elevated levels of serum creatine kinase and carbonic anhydrase III. Some patients with the McLeod phenotype and X-linked granulomatous disease apparently have sustained a deletion of both the Xk and the Phox-91 X-linked genes (see Chap. 72). All these anomalies may be related to a spectrum of deletions on the X chromosome near the Xk gene at position Xp21.
Gerbich Negative Phenotype The GPC gene on chromosome 2 encodes two proteins: glycophorin C (GPC), with antigens Ge3 and Ge4 (the Ge2 portion is “hidden” by the Ge4-bearing terminal end), and its shorter partner, glycophorin D (GPD), with antigens Ge2 (now exposed) and Ge3. GPC and GPD interact with membrane skeleton proteins 4.1 and p55, which are involved in cell deformability and membrane stability.
Gerbich-negative erythrocytes of the Leach type (Ge–2, –3, and –4) lack both GPC and GPD, have reduced protein 4.1 and elliptocytosis, but exhibit normal survival in vivo.20,22 Leach-type null red cells also have weakened expression of Kell blood group antigens.
Other Null Phenotypes The other rare null phenotypes are not associated with red cell shape changes or hemolytic anemia.20 However, patients with null phenotypes can develop erythrocyte antibodies that make it difficult to find compatible blood and that cause serious hemolytic transfusion reactions. For example, people with the Bombay phenotype (Oh or H null) demonstrate no red cell abnormality but make potent hemolytic anti-H as well as anti-A and -B. These antibodies are incompatible with all red cells except those from other persons with the Bombay phenotype. Likewise, p individuals (PP1Pk-negative) or Pk individuals (P-negative) can make hemolytic antibodies to the antigens they lack. Anti-PP1Pk and anti-P also are associated with spontaneous abortions in the first trimester.6 Women with such antibodies (notably IgG anti-P), even with a history of spontaneous abortions, have delivered viable infants after plasmapheresis.29
Null phenotypes in the MNSs and Lutheran systems are interesting because several types of null phenotypes are known. Within the MNSs blood group system, people may lack normal GPA [En(a–) or MN-negative], normal GPB (SsU-negative), or both (MkMk phenotype).12,18 The rare Lu(a–b–) phenotype is caused by a dominant inhibitor called In(Lu) (or SYN1–B), the homozygous pairing of the silent allele Lu, or a recessive sex-linked inhibitor XS2.4,6 Only the LuLu-type null is associated with high-prevalence antibody because the inhibitor type nulls produce very small amounts of Lutheran antigen. In(Lu) type Lu(a–b–) red cells have been associated with varying degrees of poikilocytosis and acanthocytosis and may hemolyze more quickly during storage, even though they demonstrate normal osmotic fragility.30
The Jk(a–b–) phenotype is caused by the silent alleles JkJk or the dominant inhibitor In(Jk). Erythrocytes having the Jk(a–b–) phenotype resist lysis in 2M urea, a solution commonly used in automated platelet counting systems. No significant clinical abnormalities have been identified to date, although Jk(a–b–) individuals have reduced ability to concentrate urine.
Blood group antibodies are classified as autoantibodies if they are specific for self-antigens that are present on the patient’s own RBC. Blood group antibodies are called alloantibodies if they react with alloantigens present on the RBC of other individuals. These antibodies also may be classified according to their mode of sensitization: naturally occurring or immune (following sensitization). A summary of common antierythrocyte antibodies is provided in Table 137-5.4,5,6 and 7,31


IgG IgG is the predominant antibody made in an immune response and constitutes about 80 percent of total serum immunoglobulin (see Chap. 83). When specific for erythrocyte antigens, these molecules can direct red cell destruction. Receptors on macrophages in the liver and spleen allow these cells to remove IgG-coated erythrocytes from circulation. IgG blood group antibodies also are capable of fixing complement, although some subclasses do so less efficiently than others: IgG3 > IgG1 > IgG2 > IgG4. How well an IgG antierythrocyte antibody binds complement depends also on the surface density and location of the recognized antigen. This is because the initiator of the classic complement cascade, C1q, requires that at least two IgG molecules bind the red cell within a span of 20 to 30 nm to initiate the complement cascade.6 For example, IgG anti-D rarely binds complement, presumably because most D sites are spaced too far apart and the epitope is so small that only one anti-D molecule per epitope can bind.6,32 Most IgG blood group antibodies do not agglutinate saline-suspended red cells, presumably because the IgG molecule is too small to span the distance between erythrocytes, although some exceptions are known (potent IgG examples of anti-A, -B, -M, and -K). Instead, most IgG antibodies sensitize red cells at 37°C (98.6°F) and are detected with the help of an antiglobulin reagent.
IgM IgM is a pentamer of five basic units (having µ: heavy chains plus a short J, or joining, chain) and makes up only about 4 percent of total serum immunoglobulin (see Chap. 83). IgM is the first class of immunoglobulin produced by a fetus and is the predominant antibody seen in an early primary immune response, but it does not cross the placenta. Because of their pentameric structure, even low-affinity IgM blood group antibodies can agglutinate red cells and activate complement very effectively. Both hemolyzing and agglutinating abilities are destroyed by the reducing reagent 2-ME or DTT. IgM blood group antibodies of very low affinity may agglutinate red cells only at temperatures below 37°C (98.6°F). Such antibodies still may fix complement onto the red cell membrane in vivo, presumably by binding to red cells at the lower temperatures of the extremities and activating the complement cascade. Because such IgM antibodies dissociate from erythrocytes at higher temperature, their reactivity may be detected in routine antiglobulin tests by virtue of the complement components that remain bound to the red cell membrane.
IgA IgA is the primary immunoglobulin in body secretions, where it exists predominantly as a dimer with a secretory component (see Chap. 83). IgA does not cross the placenta or fix complement, but aggregated IgA can activate the alternative pathway of complement, and IgA can trigger cell-mediated events. Multimeric IgA antibodies in serum are seen as hemagglutinins in blood bank tests and most often are associated with those having A, B, or Lutheran activity.
Young fetuses acquire low levels of maternal IgG, probably by diffusion across the placenta. These levels rise significantly between 20 and 33 weeks’ gestation as a selective transport system matures and maternal IgG is actively transported across the placenta. Thus, almost all blood group antibodies detected in the fetus and newborn are from the mother and disappear within the first few months of life.
Actual fetal antibody production begins shortly before birth with low levels of IgM, followed by IgG and IgA several weeks after birth. Antibody production continues to increase with age until adult levels are reached. Anti-A and anti-B are usually detected after 2 to 6 months.
Because of this late immune response in the newborn and because maternal antibody is so predominant at birth, blood bank standards permit abbreviated testing on neonates less than 4 months old.33 If available, the mother’s serum is used (and preferred) for identifying antibodies in a newborn and for cross-matching units of blood.
Naturally Occurring Antibodies in Development An antibody is naturally occurring when it is found in the serum of an individual who has never been exposed to the antigen through transfusion or pregnancy. These antibodies are most likely heteroagglutinins, produced in response to substances in the environment that are similar to erythrocyte antigens.
Evidence supporting this concept has come from studies on the formation of anti-B in chickens.34 Chicks raised in a normal environment made anti-B within the first 30 days of life, while chicks raised in a germ-free environment did not make anti-B by day 60. Naturally occurring alloanti-A and -B in humans, also called isoagglutinins, can be shown to increase in titer following ingestion or inhalation of suitable bacteria.35
However, since many antigens that are unlikely to be present in the environment also have been associated with naturally occurring antibodies, the stimulus for naturally occurring antibodies is not clearly known and may well be spontaneous, with no stimulation.6
Blood Group Associations and Frequency of Naturally Occurring Antibodies Naturally occurring alloantibodies are commonly associated with the carbohydrate antigens of the ABO, Lewis, and P blood group systems. Anti-A and -B are expected in people who lack the corresponding antigens, as are antibodies specific for H, PP1Pk, or P specificities. Naturally occurring antibodies reactive with A1, Lea, Leb, or P1 determinants also are seen frequently. Carbohydrate antigens, especially those with repetitive epitopes, can stimulate B cells to make specific antibody without the aid of helper T cells. Such thymus-independent immune responses typically result in antigen-specific antibodies of the IgM class.36
Within other systems,6 anti-Sda is found in 1 to 2 percent of normal people, and anti-Vw and anti-Wra in about 1 percent. Other less common antibody specificities are listed in approximate order of descending frequency: M, S, N, Ge, K, Lua, Dia, and Xga. Rh antigens are thought to reside only on red blood cells, but naturally occurring anti-D has been reported in 0.15 percent of Rh-negative donors and anti-E in over 0.1 percent of Rh-positive donors when more sensitive enzyme detection methods are used. Examples of naturally occurring anti-C, -Cw, and -Cx also have been described.
Some naturally occurring antibodies exist as autoagglutinins (anti-H and anti-I). Patients with autoimmune hemolytic anemia have been reported to produce many antibodies to low-frequency antigens with no specific stimulus, in addition to autoantibody.4,6
Characteristics of Naturally Occurring Antibodies Most naturally occurring antibodies are IgM, but some have an IgG component, and a few are predominantly IgG. Some anti-A or anti-B antibodies may even be of the IgA class. Antibodies that cause direct agglutination of saline-suspended red cells most commonly are of the IgM class. However, even IgG antibodies may cause agglutination of red cells when they bind erythrocyte antigens that are present at high density and number on the erythrocyte membrane, such as the ABO or MN antigens. With the exception of anti-A and anti-B, most common naturally occurring antibodies do not react at body temperature and are considered clinically insignificant. However, if they are found to react at 37°C (98.6°F), it is prudent to provide cross-match compatible blood for transfusion.
Immune Antibodies in Development Immune antibodies are produced following an exposure to foreign erythrocyte antigens through pregnancy or transfusion. The primary immune response is seen several weeks to several months after the first exposure to antigen. IgM usually is associated with early primary responses, but whether it is always the first antibody class made is unclear. In most individuals, IgG soon predominates. This is characteristic of a thymus-dependent immune response, where T cells help induce B cells to undergo isotype switching from IgM to IgG.36
In a secondary or anamnestic response, antibody concentration starts to increase several days to several weeks following exposure and may rise to very high IgG levels. Some IgG antibodies remain detectable up to 30 years after a stimulus. Others, especially Kidd antibodies, disappear after several months and are more commonly associated with delayed hemolytic transfusion reactions.4,6
Blood Group Associations and Frequency of Immune Antibodies Immune antibodies are found more commonly in individuals who have been multiply transfused than in multiparous women. This is because in pregnancy the immunizing dose of red cells is often too small to elicit a primary response, and the foreign antigens are limited to those of the father.6
Anti-D used to be the most common immune antibody found, but with the advent of Rh-matching of donors and recipients in the 1940s and the use of Rh immunoglobulin prophylaxis in the 1970s, its frequency has sharply decreased. Recent figures show anti-D to be in 0.27 to 0.56 percent of transfusion recipients, 0.10 to 0.20 percent of pregnant women, and 0.16 to 0.25 percent of healthy blood donors.6
In contrast, the frequency of immune antibodies other than anti-D has increased. Specificities other than D have been reported in about 0.6 percent of transfusion recipients, 0.14 percent of pregnant women, and 0.19 percent of healthy blood donors. Pooled data from three 5-year periods and approximately 300,000 patients suggest the absolute frequency of Rh antibodies other than anti-D is 0.22 percent; anti-K, 0.19 percent; anti-Fya, 0.05 percent; and anti-Jka, 0.035 percent.6
The rate of alloimmunization in sickle cell anemia was 18.6 percent in one survey, and 55 percent of these immunized patients made more than one antibody. The most common specificities were anti-C, -E, and -K.6
Characteristics of Immune Antibodies Immune antibodies are most often IgG but may be IgM and are sometimes IgA (notably Lua). Most immune antibodies react at body temperature and are considered clinically significant, except for those directed against Bg, Yka, Csa, McCa, Kna, JMH, and sometimes Lutheran antigens.
Clinically significant antibodies are those that are capable of destroying transfused red cells in vivo. The severity of the reaction will vary with antigen density and antibody characteristics.
Antibodies commonly associated with intravascular hemolysis include anti-A, -B, -Jka, and -Jkb. ABO incompatibility is the most potent cause of immediate hemolytic reactions because A and B antigen is so strongly expressed on erythrocytes and the antibodies so efficiently bind complement. Kidd antibodies are associated more often with delayed hemolytic reactions because they typically are difficult to detect and disappear quickly from circulation. IgG anti-Jka appears to bind complement only when traces of IgM anti-Jka are also present.6 Anti-PP1Pk, -Vel, and -Lea also have been associated with hemolysis, but such examples are rare.
Extravascular hemolysis occurs with IgG1 and IgG3 antibodies that react at body temperature, that is, immune antibodies reactive with Rh, Kidd, Kell, Duffy, or Ss antigens. Indeed, these make up the bulk of clinically significant antibodies. Antibodies not expected to cause red cell destruction are those that react only at temperatures below 37°C (98.6°F) and IgG antibodies of the IgG2 or IgG4 subclass.
Hemolytic disease of the newborn is caused by blood group incompatibility between a sensitized mother and her fetus (see Chap. 58). Antibodies most significant in hemolytic disease of the newborn are those that cross the placenta (IgG1 and IgG3), react at body temperature to cause red cell destruction, and are directed against well-developed red cell antigens. ABO incompatibility most commonly is seen, but ABO hemolytic disease of the newborn is clinically mild, presumably because the antigens are not fully expressed at birth. Antibodies directed against the D antigen can cause severe hemolytic disease of the newborn, and fetal health should be carefully followed when anti-D titers are greater than 1:16. Hemolytic disease of the newborn severity is less predictable with other blood group antibodies and can vary from mild to severe. For example, anti-K not only causes red cell hemolysis, but it may also suppress eyrthropoiesis.
Autoimmune hemolytic anemia is caused by the production of “warm-” or “cold-”reactive autoantibodies directed against red cell antigens (see Chap. 55 and Chap. 56). Production can be triggered by disease, viral infection, or drugs; from a breakdown in immune system tolerance to self antigens; or from exposure to foreign antigens that induce antibodies that cross-react with self red cell antigens.20 Autologous specificity is not always obvious, since antigen expression can be depressed when autoantibody is present.
Warm autoantibodies react best at 37°C (98.6°F) and are primarily IgG (rarely IgM or IgA). Most are directed against the Rh protein, but Wrb, Kell, Kidd, or MNSU blood group specificity also has been reported.
Cold-reactive autoantibodies are primarily IgM. They react best at temperatures below 25°C (77°F) but can agglutinate cells or activate complement at or near 37°C (98.6°F), causing hemolysis or vascular occlusion upon exposure to cold.6 Most cold-reactive autoantibodies have anti-I activity. Reactivity with i, H, Pr, P, or other antigenic specificities is much less common.
The biphasic cold-reactive IgG antibody associated with paroxysmal cold hemoglobinuria (the “Donath-Landsteiner” antibody) typically reacts with the high-frequency antigen P. It attaches to cells in the cold and very efficiently activates complement before it dissociates at warmer temperatures.
Table 137-4 lists diseases associated with specific antibody production. These antibodies only cause autoimmune hemolytic anemia if the patient carries the corresponding antigen.
ABO grouping is the single most important test performed in the blood bank because it is the fundamental basis for determining blood compatibility. ABO grouping is determined by reacting erythrocytes with licensed antisera to identify the A or B antigens they carry (the forward, or cell, grouping) and by reacting the corresponding serum or plasma with known A and B cells to identify the antibodies present (the reverse, or serum, grouping). Positive reactions are seen as hemagglutination or hemolysis using common test methods, and the results of one test should confirm those of the other.
If results are discrepant or reactions are weaker than expected, the cause must be investigated before the ABO group can be interpreted with confidence. Discrepancies can be related to red cell anomalies, serum anomalies, or both and may be associated with disease.4,6,9,20,31 Common causes, excluding clerical and technical error, are listed in Table 137-6.


The Rh or D type is the next most important test performed for blood compatibility. Individuals who type D+ are called Rh positive, and those who type D– are called Rh negative, provided reagent and cell controls are acceptable. Blood donors and pregnant women who type D– using standard typing sera are tested further for weak D expression using more sensitive methods, such as an antiglobulin test. Those with weak D antigen are considered Rh positive. Testing for weak D is optional for transfusion recipients.33
Reagent antisera to detect other common antigens (CcEe, MNSs, Kell, Duffy, Kidd, etc.) are available but used only when identifying the red cell phenotype is essential to antibody identification, blood compatibility, or paternity or forensic issues. Extended phenotyping is especially important to patients who are at high risk of alloimmunization from chronic blood transfusion, for example, those with sickle cell anemia or thalassemia. Ideally, the red cell phenotype of these patients should be determined prior to the initiation of transfusion therapy.
The antibody screen, or indirect antiglobulin test, detects “atypical” or “unexpected” antibodies in the serum (i.e., other than anti-A and anti-B) using group O reagent red cells that are known to carry most common antigens. The methods used must be able to detect clinically significant antibodies. Typically, serum or plasma and screening cells are incubated at 37°C (98.6°F) with an additive to potentiate antibody-antigen reactions; then an indirect antiglobulin test is performed. Hemagglutination or hemolysis at any point is a positive reaction, indicating the presence of naturally occurring or immune alloantibody or autoantibody. The antibody screen will not detect all atypical antibodies in serum, such as antibodies to low-incidence antigens not present on screening cells and antibodies that are not apparent at 37°C (98.6°F) and in the antiglobulin phase.
The direct antiglobulin test (direct Coombs’ test) detects antibody or complement bound to erythrocytes in vivo. Red cells are washed free of serum and then mixed with antiglobulin reagents that agglutinate cells coated with IgG or the C3 component of complement.
Positive direct antiglobulin test results are associated with the following: (1) transfusion reactions, where recipient alloantibody coats transfused donor red cells or transfused donor antibody coats recipient cells; (2) hemolytic disease of the newborn, where maternal antibody crosses the placenta and coats fetal red cells; (3) autoimmune hemolytic anemias, where autoantibody coats the patient’s own erythrocytes; (4) drug or drug-antibody complex interactions with red cells that can sometimes lead to hemolysis; (5) passenger lymphocyte syndrome, where transient antibody produced by passenger lymphocytes from a transplanted organ coats recipient red cells; and (6) hypergammaglobulinemia, where immunoglobulins nonspecifically adsorb onto circulating red cells.
A positive direct antiglobulin test results does not always indicate decreased red cell survival. As many as 10 percent of hospital patients and between 1 in 1000 to 1 in 9000 blood donors have a positive direct antiglobulin test result with no clinical indication of hemolysis.9
Compatibility testing refers to a collection of donor and recipient tests that are performed prior to red cell transfusion. Donors are tested for ABO, Rh, and unexpected antibody by the collecting facility. However, transfusing hospitals retest the ABO (and Rh on Rh-negative units) to verify the accuracy of the blood label.33 Routine recipient testing includes an ABO, Rh, and antibody screening on a blood sample collected within 3 days of the intended transfusion. Results are checked against historical records to verify ABO, Rh, and antibody status.
If the recipient has a negative antibody screening test result and no history of clinically significant antibodies, a serologic immediate spin cross-match between recipient serum and donor red cells or a “computer cross-match” (wherein computer software compares the ABO test results of both donor and recipient) is required to confirm ABO compatibility.9,33
If clinically significant antibodies are detected in recipient serum or they have been previously identified, red cell units should test negative for the offending antigens and should be cross-match compatible at 37°C (98.6°F) and the antiglobulin phase. The frequency of finding compatible units usually reflects the antigen frequency in the general population, that is, 91 percent of units should be compatible with a patient making anti-K because 9 percent of the population is K+. This reasoning will not be valid if the local donor population varies significantly from that of the general population. To calculate the frequency of compatible units when more than one antibody is present, one must multiply the frequencies of antigen-negative units for each specificity. For example, only 21 percent of units will be compatible for the recipient having both anti-K and anti-Jka: [0.91 for K–] × [0.23 for Jk(a–)] = 0.21.
When multiple clinically significant antibodies or antibodies directed against high-frequency antigens are present, finding compatible units may be extremely difficult. Antibody producers should be encouraged to give autologous units prior to their elective blood needs. If they are not candidates for autologous donation, compatible units may be found by testing the patient’s siblings or by asking regional blood suppliers to check their rare donor inventories and files. Such procurement takes additional time.
Repeat donor testing and cross-matching are not performed for plasma and platelet components, but the recipient’s ABO and Rh phenotypes must be known for appropriate selection of components. General ABO-Rh compatibility guidelines are given in Table 137-7.


All unexpected antibodies are investigated. Those detected in serum or plasma as an ABO discrepancy, a positive antibody screening result, or an incompatible cross-match are identified using a panel of 8 to 16 different group O cells that have been typed for common clinically significant antigens. Serum reactions with these cells are compared to their antigen typing to determine specificity.9 For example, an antibody that reacts with all K+ cells but not with K– cells is most likely anti-K.
A control of autologous cells and serum is tested concurrently with panel cells. Absence of reactivity with autologous cells implies that the antibody is an alloantibody, while a positive result suggests autoantibody or a positive direct antiglobulin test result. Once the antibody specificity is identified, the subject’s erythrocytes are tested for the corresponding antigen. If the alloantibody is anti-K, the cells should type K–. Such antigen typing helps to confirm serum findings.
When antibody is detected on both red cells (a positive direct antiglobulin test result) and in serum, only the antibody in serum is identified unless a review of the medical or transfusion history offers evidence that they might be different. When antibody is detected only on red cells and hemolysis is suspected, the antibody can be eluted and tested against panel cells to identify the specificity.

Lewis M, Anstee DJ, Bird GWG, et al: Blood group terminology 1990, from the ISBT Working Party on Terminology for Red Cell Surface Antigens. Vox Sang 58:152, 1990.

Rippee C, Myers J, Gindy L: Blood groups, in Clinical Practice of Transfusion Medicine, 3rd ed, edited by LD Petz, SN Swisher, S Kleinman, et al. Churchill Livingstone, New York, 1996.

Daniels GL, Anstee DJ, Cartron JP, et al: Blood group terminology, ISBT Working Party on Terminology for Red Cell Surface Antigens. Vox Sang 69:265, 1995.

Issitt PD, Anstee DJ: Applied Blood Group Serology, 4th ed. Montgomery Scientific, Miami, 1998.

Reid ME, Lomas-Francis C: The Blood Group Antigen Facts Book. Academic Press, San Diego, 1997.

Mollison PL, Engelfriet CP, Contreras M: Blood Transfusion in Clinical Medicine, 10th ed. Blackwell Scientific, Oxford, 1997.

Daniels GL: Human Blood Groups. Blackwell Scientific, Oxford, 1995.

Cartron JP, Rouger P (eds): Blood Cell Biochemistry, vol 6, Molecular Basis of Human Blood Group Antigens. Plenum Press, New York, 1995.

Vengelen-Tyler V (ed): Technical Manual, 12th ed. American Association of Blood Banks, Bethesda, MD, 1996.

Tippett P: Regulator genes affecting red cell antigens. Transfus Med Rev 4:56, 1990.

Marsh WL, Johnson DL, Mueller KA: Proposed new notation for the In(Lu) modifying gene. Transfusion 24:371, 1984.

Okubo Y, Yamaguchi H, Nagao N, et al: Heterogeneity of the phenotype Jk(a–b–) found in Japanese. Transfusion 26:237, 1986.

Giblett ER: A critique of the theoretical hazard of inter vs. intra-racial transfusion. Transfusion 1:23, 1961.

Clausen H, Hakomori S: ABH and related histo-blood group antigens: Immunochemical differences in carrier isotypes and their distribution. Vox Sang 56:1, 1989.

Yomtovian R, Prince GM, Medof ME: The molecular basis for paroxysmal nocturnal hemoglobinuria. Transfusion 33:852, 1993.

Cartron JP, Rahuel C: Human erythrocyte glycophorins: Protein and gene structure analyses. Transfus Med Rev 6:63, 1992.

Mourant AE, Kopec AC, Domaniewska-Sobczak K: The Distribution of Human Blood Groups and Other Polymorphisms, 2d ed. Oxford University Press, London, 1976.

Mourant AE, Kopec AC, Domaniewska-Sobczak K: Blood Groups and Diseases. Oxford University Press, London, 1977.

Garratty G (ed): Blood Group Antigens and Disease. American Association of Blood Banks, Arlington, VA, 1983.

Reid ME, Bird GWG: Associations between human red cell blood group antigens and disease. Transfus Med Rev 4:47, 1990.

Hadley TJ, Miller LH, Haynes JD: Recognition of red cells by malaria parasites: The role of erythrocyte-binding proteins. Transfus Med Rev 5:108, 1991.

Issitt PD: Null red blood cell phenotypes: Associated biological changes. Transfus Med Rev 7:139, 1993.

Nichols ME, Rubinstein P, Barnwell J, et al: Identification of an erythrocyte membrane protein complex carrying Duffy blood group antigenicity: Immunogenetics and association with susceptibility to Plasmodium vivax. J Exp Med 166:776, 1987.

Kallenius G, Mollby R, Svenson SB, et al: Occurrence of P-fimbriated Escherichia coli in urinary tract infection. Lancet 2:1369, 1981.

Kallenius G, Svenson S, Mollby R, Cedergren B, Hultberg H, Winberg J: Structure of carbohydrate part of receptor on human uroepithelial cells for pyelonephritogenic Escherichia coli. Lancet 2:604, 1981.

Gerbal A, Maslet C, Salmon C: Immunological aspects of the acquired B antigen. Vox Sang 28:398, 1975.

Gerbal A, Ropars C, Gerbal R, et al: Acquired B antigen disappearance by in vitro acetylation associated with A1 activity restoration. Vox Sang 31:64, 1976.

Marsh WL, Redman CM: The Kell blood group system: A review. Transfusion 30:158, 1990.

Shirey RS et al: Plasmapheresis and successful pregnancy after fourteen miscarriages in a P1k with anti-P [abstr]. Transfusion 24:427, 1984.

Udden MM, Umeda M, Hirano U, et al: New abnormalities in the morphology, cell surface receptors, and electrolyte metabolism in In(Lu) erythrocytes. Blood 69:52, 1987.

Quinley ED (ed): Immunohematology: Principles and Practice, 2d ed. Lippincott, Philadelphia, 1998.

Lomas C, Tippett P, Thompson KM, et al: Demonstration of seven epitopes on the Rh antigen D using human monoclonal anti-D antibodies and red cells from D categories. Vox Sang 57:261, 1989.

Menitove JE (ed): Standards for Blood Banks and Transfusion Services, 18th ed. American Association of Blood Banks, Bethesda, MD, 1997.

Springer GF, Horton RE, Rorbes M: Origin of anti-human blood group B agglutinins in white leghorn chicks. J Exp Med 110:221, 1959.

Springer GF, Horton RE: Blood group isoantibody stimulation in man by feeding blood group-active bacteria. J Clin Invest 48:1280, 1969.

Silberstein L, Spitalnik SL: Blood group antigens and antibodies, in Principles of Transfusion Medicine, edited by W Rossi, TL Simon, GS Moss. Williams & Wilkins, Baltimore, 1991.
Copyright © 2001 McGraw-Hill
Ernest Beutler, Marshall A. Lichtman, Barry S. Coller, Thomas J. Kipps, and Uri Seligsohn
Williams Hematology



  1. Execelent information and facts my friend, mujeres casadas que buscan hombres I just did not know what you published, great share. mujeres casadas que buscan hombres

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: