Practical Transfusion Medicine 4th Ed.

2. Essential immunology for transfusion medicine

Jaap Jan Zwaginga1 & S. Marieke van Ham2

1Jon J. van Rood Center for Clinical Transfusion Research, Sanquin, Leiden and the Department of Immunohematology and Bloodtransfusion, Leiden University Medical Center, Leiden, The Netherlands

2Department of Immunopathology, Sanquin Research and Landsteiner Laboratory, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands

Cellular basis of the immune response

Leucocytes from the myeloid and lymphoid lineage are the key effector cells of both the innate and adaptive immune system, and are differentiated from haemopoietic stem cells (HSC) in the bone marrow.

Innate immune cells

Phagocytes and antigen presenting cells (APCs)

Cells of the myeloid lineage include monocyte-derived macrophages, neutrophils (polymorphonuclear neutrophils, PMNs) and dendritic cells (DCs). All three function as phagocytes that remove dead cells and cell debris or immune complexes. Foremost, these cells act as the first line of innate defence, ingesting and clearing pathogens. Very important in this is their activation via specific receptors, termed PRR (pattern recognition receptors) by danger signals derived from pathogens or inflamed tissue. This triggers their differentiation and their expression and/or secretion of signalling proteins, which lead to further activation of the immune response. Some of these proteins (like IL-1, IL-6 and TNF) increase acute phase proteins that activate complement, while others (chemokines) attract circulating immune cells to the site of infection. DCs, and also macrophages, additionally serve as APCs that process and present digested proteins as antigen to specific T cells of the lymphoid lineage. PRR ligation in this setting induces maturation of APCs with acquisition of chemokine receptors, which allow their migration to the lymph nodes where the resting T cells reside. Simultaneously, mature APCs acquire costimulatory molecules and secrete cytokines. All are needed for T-cell activation and differentiation and eventually the immune response to the specific pathogen. The type of PRR ligation determines the formation of defined cytokines and with it an optimal pathogen class-specific immune answer, with minimal tissue damage.

NK lymphocytes

Natural killer (NK) cells are capable of killing virus-infected cells either specifically targeted by the presence of antibody on the cell's surface (antibody-dependent cell-mediated cytotoxicity – ADCC) or through the recognition of changes in the infected cell surfaces. Moreover, NK cells are normally kept from killing by expression of inhibiting receptors that recognize the presence of self-MHC molecules (see below) on autologous cells. Allogeneic cells with noncompatible MHC but also aberrant autologous cells (e.g. tumour cells) with lowered MHC expression lack sufficient of these NK inhibiting structures and trigger the default killing potential of NK cells.

Adaptive immune cells

T-lymphocytes

After migration of progenitor T cells to the thymus epithelium, billions of T cells are formed with billions of antigen receptor variants. Each lymphocyte expresses only one kind of heterodimeric T-cell receptor (TCR). For the large majority of T cells this is an alpha and a beta chain, which form a structure that is similar to the specific antigen binding site of immunoglobulin molecules. Immature T cells initially express a TCR receptor in complex with CD4 and CD8 molecules, which respectively interact with major histocompatibility complex class II and class I molecules. The presentation of self-antigens within such MHC molecules on thymic stromal cells determines the fate of the immature T cells. First of all, these interactions induce T-cell maturation into T cells that express only CD4 or CD8. Most important, however, is that these interactions are responsible for the removal of T cells that have a TCR with high binding affinity for the MHC complexes that express the self-antigen. The cells that survive this so-called ‘negative selection’ process migrate to the secondary lymphoid organs. There TCR specific binding to complexes of MHC can activate them with non-self (e.g. pathogen-derived) antigens on matured APCs. Interactions between the costimulatory molecules CD80 and CD86 on the APC with CD28 on the T cell subsequently drives the activated T cells into proliferation. Without this costimulation (e.g. by not fully differentiated APCs by insufficient or absent PRR ligation), T cells can become nonfunctional (anergized). The requirement of PRR-induced danger signals thus forms a second checkpoint of T-cell activation to prevent reactivity to self-antigens. Additionally, APC-released cytokines direct T-cell differentiation.

While immunoglobulins bind to amino acids in the context of the tertiary structure of the antigen, the TCR recognizes amino acids on small digested antigen fragments in the context of an MHC molecule. As indicated, there are two classes of MHC (called human leucocyte antigens of HLA in humans) molecules that are similar in their two polypeptide structure with an antigen binding groove (see Chapter 4). The MHC is polygenic determined, resulting in different sets of peptide binding specificities. Moreover, MHC genes are polymorphic, with many allelic variations in the population. Both MHC characteristics ensure endless protein/antigen binding capacities and thus adaptation of the immune response to new/rapidly evolving pathogens. MHC class I is expressed on all nucleated cells and presents so called ‘endogenous’ antigen constituting self-antigens, but also antigens from viruses and other pathogens that use the replication machinery of eukaryotic cells for their propagation. To be loaded on to MHC I, proteins need to be processed in smaller antigen parts by the proteasome. Subsequently, processed proteins are shuttled into the endocytoplasmatic reticulum for loading on to newly synthesized MHC class I molecules. Finally, this complex is cell surface expressed. CD8+ cytotoxic T cells (CTLs) recognize the MHC class I/antigen complex on the cells. Although viruses and parasites (like Plasmodium falciparum) can hide in red blood cells because the latter lack MHC, red cells also lack the DNA replication machinery for such pathogens.

MHC class II molecules of APCs present antigenic proteins that are ingested or endocytosed from the extracellular milieu. Upon cell activation these proteins are protease digested in acidified endocytic vesicles yielding smaller antigen fragments. Again after fusion with the MHC class II containing compartments, the antigen is loaded on to the MHC class II molecule and routed to the plasma membrane.

The described antigen expression routes, however, are not absolute. Specialized DC in this respect can also express viral and other extracellular-derived proteins on MHC class I to CD8+ CTLs while, vice versa, primarily cytosolic proteins via so-called autophagy can become localized in the endocytic system and become expressed in MHC class II. This so-called antigen cross-presentation adds flexibility to the adaptive immune response.

Paradoxically, having described the fact that T cells become activated only when the specific TCR recognizes antigen in the context of its own MHC (termed MHC restriction) seems to refute the condition that MHC/HLA mismatched tissue transplants are rejected. Many acceptor T cells, however, can be activated because their TCR perceives donor-specific MHC as foreign in itself. A large circulating pool of T cells reacting with non-self MHC is usually present and explains the acute CD8-dependent rejection of non-self MHC in transplant rejection that occurs without previous immunization.

T helper (Th) cells

Differentiation into T helper cells is dependent on signals (cytokines and/or plasma membrane molecules) derived from the APC. Different Th subsets can be characterized by their cytokine release and their extralymphatic action in infected tissues. Th1 cells release IFN gamma and IL-2 and bind and help macrophages to kill intracellular pathogens. In addition, Th1 cells support CTL function and are required for optimal CTL memory formation. Th17 cells releasing IL-17 and IL-6 probably enhance the early innate response by activating granulocytes and seem most needed for antifungal immunity. Both Th1 and Th17 are drivers from strong proinflammatory immune responses, which might explain why they are also associated with auto-immunity. Classically, Th2 cells are thought to be the main Th subsets to support B-cell differentiation and the formation of antibodies. The IL-4, -5 and -13 releasing Th2 cells, furthermore, help to kill parasites by inducing IgE production, which activates mast cells, basophils and eosinophils. Th2, however, is also associated with aberrant immunity, as observed in allergies. Finally, the recently defined follicular Thelper cells (Tfh) seem to be required for long-lived immunity and to induce antibody formation upon primary immunization and upon reactivation of memory B cells in the case of a re-encounter with an antigen.

Regulatory T cells (T regs)

These form a specialized CD4+ T-cell subset, which limits T-cell activation and autoimmunity. While naturally occurring T regs are already formed in the thymus, induced T regs are likely to be induced by APC that show suboptimal costimulation and that release anti-inflammatory cytokines (IL-10 and TGF-beta). These inhibit proinflammatory T-cell differentiation as well as many innate cell functions like DC differentiation.

B lymphocytes

In the bone marrow, progenitor B cells upon local cues from stromal cells, divide and are directed towards acquisition of their antigen-specific B-cell receptor (BCR). With initially millions of different binding affinities of this surface expressed immunoglobulin, immature B-cell clones that show self-antigen binding affinity are eliminated by premature stimulation. B cells mature in the peripheral lymphoid tissues where they respond to foreign antigens via activation of the BCR. Upon receipt of additional survival signals, they proliferate and differentiate into plasma cells that in their turn secrete immunoglobulins with identical binding specificities as the activated B cells they are derived from. Depending on their differentiation pathway, plasma cells secrete specific classes of effector antibodies (i.e. IgM, IgD, IgG, IgA and IgE). Long-lived IgG producing plasma cells migrate to the bone marrow where they can survive for many years. In addition to plasma cells, memory B cells are formed during the first antigen encounter, awaiting reactivation in a following infection.

B-cell activation and T-cell-dependent antibody formation

The binding of antigens by the BCR also leads to intracellular signalling and activation of the B cells upon antigen (re-)challenge. The antigen-BCR complex is internalized and processed into fragments, which can be re-expressed on MHC class II molecules at the B-cell surface. This APC function of B cells is first of all designed to recruit T helper cells that are activated by DCs that have presented the same antigen. This process ensures that T helper cells only support B-cell differentiation of those B cells that have become activated by the same pathogen, thus minimizing the risk of activation of autoreactive B cells. Activated B cells and activated T cells have an enhanced change to meet as both cells are induced to migrate to the border of the B–T cell zones in the secondary lymphoid organs upon activation. Activated Th cells express CD40L, which provides costimulation to the B cells. Ligation of the B cell via the CD40 costimulatory molecule together with cytokines secreted by the Th cells modulate the direction of B-cell differentiation. For example, Th1 support differentiation into IgG1 producing plasma cells and Th2 support formation of IgE producing plasma cells. Some pathogens that have a repetitive structure (called thymus-independent antigens) can activate B cells to produce IgM antibodies against mostly extracellular pathogens without T-cell help. This offers a fast response mechanism, but of low affinity. Higher affinity antibody formation requires T-cell helper interactions.

Humoral immune response

Immunoglobulins (Igs) are in fact the secreted form of the B-cell receptor. This specific effector molecule is secreted by plasma cells. The Ig's basic structure is a roughly Y-shaped molecule made up of two identical heavy chains with four domains and two identical (kappa or lambda) light chains with two domains. These heavy and light chains are interlinked by noncovalent and disulfide bonds. Two identical highly specific antigen binding sites (the arms of the Y) are formed by the amino terminus domains of the heavy and light chains and form the variable (Fab) domain of the Igs. The specificity and variability of these antigen binding sites is a result of two extra beta strands in these variable domains. Connected to the normal seven beta strands found in the ‘constant’ domains, these additional amino acid sequences form tertiary protein structures with an almost unending repertoire of different three-dimensional ‘binding locks’ for antigens (Figure 2.1). Both heavy chains combine via their carboxy terminal domains to the so-called constant (Fc) region (the trunk of the Y) of the Ig, which is more or less flexibly attached to the antigen binding part by a so-called hinge area in the heavy chains. The Fc region determines the Ig class and consequently the Ig effector function, which is different for each Ig class. Some effector Igs form higher order structures, with secreted IgA being a dimer and IgM a pentamer.

Fig 2.1 Basic structure of an immunoglobulin molecule. Domains are held in shape by disulfide bonds, though only one is shown. CH1–3, constant domain of an H chain; CL, constant domain of a light chain; VH, variable domain of an H chain; VL, variable domain of a light chain.

c02f001

Basis of antibody variability

The BCR/antibody variability via the variable region originates from random DNA recombination of two or three of many variable region gene segments. The endless recombinations are the pretranscriptional basis of the extremely variable tertiary structures of the immunoglobulin molecule (Figure 2.1) and the enormous B-cell repertoire in the bone marrow. Of these cells, each having one single antigen specificity, cells with self-reactive BCRs will be destroyed. From the remainder, clones are selected on the basis of their specific pathogen antigen binding abilities and these mature and expand into antibody producing plasma cells. This secondary diversification of the antibody repertoire takes place in extrafollicular tissues or in the germinal centres of the lymphoid organs and consists of:

·        Several sequential enzyme-driven steps leading to point mutations or so-called somatic hypermutations of the variable regions of both the heavy and light chains.

·        A process called affinity maturation leads to selection of B cells with a BCR type that has the highest affinity for the antigen. Random mutations are thus eventually directed towards plasma cells that secrete Igs with optimized antigen affinity.

·        Immunoglobulin (sub)class switching by helper T-cell-released cytokines that induce transcription of so-called switch regions. This process enables the first produced IgM by naïve B cells to evolve into IgG or IgE class antibodies. There are five immunoglobulin classes (isotypes) based on different genes that are used for the C domains of the H chain and an additional four and two immunoglobulin subclasses for IgG and IgA. These subclasses determine the effector functions of the Ig, as well as their serum half-life and their ability for placental transfer (Table 2.1).

·        The combination of somatic hypermutation/affinity maturation and class switching explains why during immune responses the first formed IgM Igs generally show low binding affinity to the antigen, while the later formed IgGs have undergone affinity maturation and indeed show enhanced antigen binding.

Table 2.1 Immunoglobulin classes and their functions.

Table02-1

Antibody effector functions

The class (isotype) of an Ig is determined by coupling variable domains with identical binding specificities to different constant regions of the heavy (H) chain. While IgM only functions in circulation, other classes also function in other body compartments. IgA in this respect is mostly localized in epithelial tissues like the gut. As a dimer, IgA can be excreted into the intestines and into exocrine (e.g. milk, saliva and tear producing) glands and act there as an early defence to pathogen invasion of these tissues and of the newborn via transfer in mother's milk. Although antibodies can neutralize toxins and pathogens, e.g. by blocking their adhesion to cell surfaces, definite clearing them from the body is achieved by the following processes.

·        For pathogens, mostly by phagocytes, which are triggered to ingest and destroy antibody-coated structures by their Fc receptors crosslinking Fc tails of antigen-bound Igs. This IgG-mediated process is responsible for the clearance of antigen–Ig complexes from circulation in the spleen and liver.

·        For parasites, by exocytosis of stored mediators, e.g. from mast cells that are triggered by their Fcɛ receptor recognizing the Fc region of IgE.

·        Activation of the complement cascade (see Figure 2.2). The system is part of innate immunity but is also vital to the effector functions of complement-fixing immunoglobulin isotypes. Central to the complement's function is the activation of C3 by three routes:

a. The classical pathway: this pathway can be powerfully activated by IgM but in decreasing order by IgG3, 1 and 2 as well and consists of four numbered components (C1–C4) and two regulatory proteins (C1 inhibitor, C4 binding protein). The first component (C1) consists of three subcomponents, C1q, C1r and C1s. It is the interaction between C1q and aggregated IgG or IgM bound to antigen that initiates activation of the classical complement sequence. The fixation of C1q activates C1r and C1s. C1s cleaves C4 and C2, whose active fragments C4b and C2a form the classical pathway C3 convertase.

b. The alternative pathway, consisting of C3b, factor B and factor D, and the regulatory proteins, properdin, factors H and I. Factor B binds to a cleavage fragment of C3, C3b, to form C3bB. Factor D cleaves the bound factor B to form the alternative pathway C3 convertase (C3bBb). It activates C3 in a fashion similar to the C3 convertase of the classical pathway, C4b2a. Properdin acts to stabilize this alternative pathway C3 convertase, as do carbohydrate-rich cell surfaces, by partially shielding the convertase from inhibitors. The alternative pathway is inhibited by default and always shows some activation.

c. The lectin pathway is initiated by soluble proteins like mannose binding lectin and ficolins that are structurally related to C1q and that recognize and bind carbohydrates on the surface of microorganisms. Serine proteases associated with these recognition proteins activate C4, which, similar to C1r and C1s, leads to the same outcome, namely generation of C3 convertase.

Fig 2.2 The different pathways for complement activation. MBL, mannan-binding lectin; MASP, MBL-associated serine protease.

c02f002

The fate of antibody-coated cells (e.g. red blood cells in auto- or alloimmune haemolysis) is dependent on whether there is partial or total activation downstream from C3. Total activation in this respect generates the membrane attack complex with the formation of the trimolecular complex of C4b2a3b or C5 convertase. This complex cleaves C5 into two fragments, C5a and C5b. C5b forms a complex with C6, C7 and C8, which facilitates the insertion of a number of C9 molecules in the membrane. This so-called membrane attack complex (MAC) creates lytic pores in the membrane that destroys the target cell. IgM mediates this process especially well. MAC can also be transferred to cells close by and leads to so-called bystander lysis.

Partially activated complement, in contrast, recruits and activates phagocytes to sites of infection, but moreover it can mediate homing and clearance of complement-coated cells in macrophage areas of the spleen or liver, which next to Fc receptors also carry complement receptors.

Red blood cell antibodies illustrating the above principles

Several hundreds of red cell transfusion-related antigens have been identified. Blood group antigens are inherited polymorphic structures located on proteins, glycoproteins and glycolipids (Chapter 3). Alloimmunization can happen after contact with non-self blood antigens by transfusion or during pregnancy and delivery. The responsible cellular mechanisms are still largely unclear but powerful animal models will generate definite knowledge in this respect [1]. The humoral response, however, is much easier to investigate. IgM class antibodies are formed first but are usually shortly present after immunization, as T-cell-independent B-cell activation and IgM production does not generate memory B cells and IgM producing plasma cells are often short lived. As so called naturally occurring antibodies, IgM, however, can also be present permanently. The binding affinity of antibodies increases in germinal centre reactions, coinciding with the switch from IgM to IgG. The chemical nature of the antigen itself also determines the elicited humoral response. Blood cell antibodies against carbohydrate antigens are generally IgM or IgG2 and IgG4, or a combination of these. Antibodies against protein blood group antigens are typically of the IgG class with predominantly IgG1 and IgG3.

Best known of the so-called naturally occurring IgM antibodies are those directed against the A or B blood group antigens. Some antigens, mostly bacterial polysaccharides A or B, can stimulate subsets of mature B cells directly due to the ability of these molecules to crosslink several BCRs and, concomitantly, FcR and complement receptors if antigen is opsonized. It is in this respect supposed that gastrointestinal bacteria trigger this so-called anti-A, B isoagglutinin production. This explains their presence from the first months of life. They are of the IgM class because T cells – as a necessary trigger of isotype switching – are not involved. Although some IgM to IgG switching does occur for A and B antigens, the T-cell-independent antibody formation for these carbohydrate antigens, however, has to be discerned from the T-cell-dependent high affinity IgG forming mechanisms for the polypeptide blood groups. Immunization and antibody formation can occur after allogeneic blood contact, e.g. by transfusion, transplantation or associated with pregnancy. This can cause direct or later haemolysis of subsequent transfused mismatched red blood cells but also haemolytic disease of a (next) child with incompatible fatherly antigens. While IgM cannot pass the placenta, motherly IgG directed against antigens on the fetus' blood cells can. Fortunately A and B antigens are only expressed at low levels on fetal red blood cells. Therefore, the anti-A or -B IgG transferred from the mother's blood usually does not lead to haemolysis in the fetus. On the other hand, organ transplants from donors that are AB incompatible can be acutely rejected by recipient anti-A or -B directed to A or B expression on the organ vasculature.

The mechanisms responsible for gradually increasing antibody specificities and subclass changes can also be witnessed by studies on the molecular structure of the V domains (like the predominant use of IGHV3 superspecies genes) of monoclonal antibodies against the protein antigen RhD on the red cell membrane. This selective use of V genes in antibody production against a certain antigen was found in pregnancy-induced RhD immunized females who volunteered for further immunization with RhD [2]. These hyperimmunization programmes are of great importance in the acquisition of therapeutic quantities of anti-D antibodies.

Antibody and complement-mediated blood cell destruction

A transfusion into a recipient with circulating antibodies against transfused antigens can cause an acute (within 24 hours posttransfusion) haemolytic syndrome (Chapter 7). This clinical picture can be life threatening, especially when intravascular haemolysis is induced. The delayed form of haemolysis occurs between 1 and 7 days after an antigen-incompatible transfusion and is typically less severe. The latter namely depends on boosting of a previous, but at the moment of transfusion, undetectable memory immunization [3].

Most red blood group allo- and autoantibodies of the IgG isotype bring about lysis via the interaction of the IgG constant domain with Fcγ receptors on cells of the mononuclear phagocytic system. Several receptor types are described.

·        FcγRI is the most important receptor that causes blood cell destruction. This is a high-affinity receptor found predominantly on monocytes. The consequence of adherence of IgG-coated red cells to FcγRI-positive cells is phagocytosis and lysis. This is usually extravascular and takes place in the spleen. The lysis can be demonstrated in vitro as ADCC.

·        FcγRII is a lower affinity receptor found on monocytes, neutrophils, eosinophils, platelets and B cells.

·        FcγRIII is also of relatively low affinity and found on macrophages, neutrophils, eosinophils and NK cells. It is responsible for the ADCC demonstrable in vitro with NK cells.

·        There is also an FcRn (neonatal) on the placenta and other tissues of a different molecular family, which mediates the transfer of IgG into the fetus and is involved in the control of IgG concentrations.

The severity of haemolysis by IgG antibodies is determined by the concentration of antibody, its affinity for the antigen, antigen density, the IgG subclass and their complement activating capacity (see below). IgG2 antibodies generally do not reduce red cell survival, whilst IgG1 and IgG3 do. There is ample evidence in patients with warm-type autoimmune haemolytic anaemia that IgG1 and IgG3 are more effective in causing red cell destruction than IgG2. The level of IgG1 coating of red cells needs to exceed a threshold of approximately 1000 molecules per red cell to cause cell destruction. For a long time, it has been speculated that polymorphisms in the genes of the family of FcγRs might be significant in causing differences of severity of blood cell destruction observed between patients with apparently similar levels of IgG coating. In line with this, a single amino acid polymorphism of the FcγRIIa receptor dramatically alters the affinity for human IgG2 and additional polymorphisms might have an effect on the interaction with IgG1 and IgG3.

The complement system, either working alone or in concert with an antibody, plays an important part in immune red cell destruction. In contrast to extravascular FcγR-mediated destruction, complement-mediated lysis occurs in the intravascular compartment. The ensuing release of anaphylatoxins such as C3a and C5a contributes to acute systemic effects. IgM antibodies against the A and B antigens are mostly known for this rapid complement-mediated destruction and intravascular haemolysis of incompatible red cells. These events are due to incorrectly typed transfusion units or misidentification of product or recipient and, although rarely happening, remain the most important cause of transfusion-related mortality and morbidity.

Apart from intravascular lysis, blood cells coated with C3b will bind to cells carrying receptors for C3b (CR1 or CD35). This leads to extravascular cell destruction mainly in the liver. If, however, the bound C3b degrades to its inactive components iC3b and C3dg before the cell is lysed, then the cell is protected from lysis. Membrane-bound molecules such as decay accelerating factor (DAF) and membrane inhibitor of reactive lysis (MIRL) also protect red cells from lysis in this way.

Clinical aspects related to alloimmunization against blood cell antigens

Although platelet-directed antibodies (both directed against HLA class I and human platelet antigen or HPA) can lead to refractoriness to platelet transfusions (see Chapters 4 and 5), the consequences of red cell antibodies in the case of incompatible transfusions are much more significant. The effects not only lead to direct (destruction of the allogeneic red cells) but also to indirect (haemolysis-dependent thrombosis, multiple organ damage) morbidity and sometimes mortality (see Chapter 7). The latter can and has to be prevented by determining and recording alloimmune antibody specificities and matching further transfusions for the absence of reactive antigens.

Red cell alloimmunization is reported with large variations between 2 and 21%, with a strong predominance for specific antigens. This reported variation is certainly influenced by the quantity of alloexposures (the number of transfusions) [4]. On the other hand, medication-suppressed immunity or an activated immune system by the presence of autoimmune disorders, infection [5] or pre-existing haemolysis priming APCs with danger signals are all likely to influence auto- and alloimmunization efficacy. Such factors might protect and also be critical as additional triggers of red cell alloimmunization because most peptide red cell antigens only differ by single amino acid substitutions. In contrast, the much more immunogenic red cell antigens like the A, B and the RhD antigens, which are present or absent, are routinely matched for and therefore of less importance for transfusion-mediated alloimmunization. Finally, alloimmunization efficacy is influenced by factors involving the compatibility between donor and patient, e.g. the extent of their genetic or ethnic differences. The latter is not only the case for red cell blood groups themselves but also for HLA differences between donor and recipient. Certain HLA types are associated with a higher red blood cell alloimmunization risk, suggesting specific HLA restriction for the presentation of some red cell antigens [6,7]. Interestingly, a first alloimmunization increases the risk of further antibody formation. This again might indicate a subgroup of so-called responder patients that have an intrinsic higher risk for alloimmunization [8]. On the other hand, the observation that preventative matching of donor blood for highly immunogenic antigens like Rh E and Kell prevents alloimmunization against other antigens might alternatively indicate that a first immunization itself changes the susceptibility for subsequent events [9]. Better identification of clinical or genetic patient factors and possibly also product (like storage) risk factors for red cell antigen alloimmunization will be of great importance; this might enable a cost-effective matching in specific high-risk conditions. This would prevent antibody formation that is especially important in later acute conditions when only low level matching is possible.

Although alloimmunization against red cell antigens – because of the many patients who undergo repetitive exposure to obligatory more or less mismatched blood – is important enough, (co-)transfused platelets and leucocytes, respectively expressing MHC class I and both class I and II, are more effective in inducing alloimmunization. This is first clear from the formation of non-self-directed transfusion-dependent HPA and HLA and other leucocyte-directed antibodies. Leucocyte reduction in this respect, first introduced in the late 1990s to prevent the merely theoretical transmission of variant Creutzfeldt–Jacob disease by leucocytes in blood components, did decrease transfusion-associated primary HLA antibody formation. Antigens of the human leucocyte antigen (HLA) system can be recognized directly or indirectly by the cells of the immune system. The direct pathway involves the direct recognition of non-self HLA molecules on the donor antigen presenting cells (APC) in blood or in the transplanted tissues and involves activation of immunologically naive T cells. The indirect pathway involves processing and presentation of donor-derived HLA peptides by the host APC. Interesting, alloimmunization, e.g. against red cells, is associated with the additional presence of HLA antibodies [10,11]. Again, it is unclear if this observation involves the so-called (more easily immunized) responder patients or that it merely indicates a general non-self antigen exposure, which is inherent to allogeneic blood exposure.

HLA and HPA antibodies are associated with various subsequent effects. First of all, recipient HLA antibodies can cause refractoriness to platelet transfusions because donor platelets express (although varying amounts of) incompatible HLA class I molecules. The same can happen on the basis of alloimmunization against inherited polymorphisms in platelet-specific antigens (HPA variants). These HPA antibodies can cause platelet transfusion refractoriness, but in contrast to HLA antibodies also neonatal alloimmune thrombocytopenia (NAITP) and sometimes fatal bleeding complications (see Chapter 5). HLA antibodies in this respect do not seem able to cross the placental barrier as can HPA antibodies. Second, donor HLA (but also other leucocyte binding) antibodies are instrumental in at least part of transfusion-related acute lung injury (TRALI) cases (see Chapter 9). Third, HLA antibodies in the recipient itself can cause cytokine-induced febrile nonhaemolytic transfusion reactions (FNHTR) when reacting with and destroying donor leucocytes in transfusion products (see Chapter 8).

Finally, hyperhaemolysis and posttransfusion purpura are associated with antibodies acting against transfused blood components. Posttransfusion purpura is a rare bleeding complication typically 5–10 days after transfusion [12] (see Chapter 12). Hyperhaemolysis is a similar event but then lowering the haemoglobin below pretransfusion levels due to destruction of transfused and autologous red blood cells. Hyperhaemolysis is most seen in transfused haemoglobinopathy patients with a background of active haemolysis [13] (see Chapter 7).

Next to the alloimmunization against red blood groups, HPA, HLA and other leucocyte antigens, it seems only logical that transfusion of various amounts and degrees of functional and viable leucocytes can have additional effects on the immune system of the recipient. This transfusion-related immunomodulation (TRIM) (see Chapter 10) is mostly studied by comparing outcomes of leucocyte-rich and leucocyte-reduced transfusion products [14]. Patients with cancer and infections – disorders that for their outcome are determined by immune competence – are most likely to be affected by possible TRIM effects.

Key points

1. Transplants and allogeneic blood are intrinsically non-self and capable of eliciting an immune response; additional danger signals (as in inflammatory conditions) are needed to prime and activate the blood cells that are most important for alloantibody formation.

2. The ability of antibodies to bring about erythrocyte or platelet destruction varies according to their isotype and their antigenic, Fc receptor and complement binding and activating capacities.

3. Most clinical problems encountered in transfusion medicine are antibody-based; also the responsible and modulating cellular mechanisms, still need more elucidation [15].

4. Better identification of high risk patients (responders) and conditions together with increasing possibilities to perform extensive (genotypic) typing for RBC and platelet antigens will enable selective preventative and cost-effective donor–recipient matching [15].

5. High alloimmunization risk patients and conditions might additionally benefit from immunomodulatory therapies that aim at the prevention of allo-specific B-cell activation and plasma cell differentiation.

References

1. Gilson CR & Zimring JC. Alloimmunisation to transfused platelets requires priming of CD4+ T cells in the splenic microenvironment in a murine model. Transfusion 2011; 52: 849–859.

2. Dohmen SE, Verhagen OJHM, Muit J, Ligthart PC & van der Schoot CE. The restricted use of IGHV3 superspecies genes in anti-Rh is not limited to hyperimmunized anti-D donors. Transfusion 2006; 46: 2162–2168.

3. Vamvakas EC, Pineda AA, Reisner R, Santrach PJ & Moore SB. The differentiation of delayed hemolytic and delayed serologic transfusion reactions: incidence and predictors of hemolysis. Transfusion 1995; 35: 26–32.

4. Zalpuri S, Zwaginga JJ, le Cessie S, Elshuis J, Schonewille H & van der Bom JG. Red-blood-cell alloimmunisation and number of red-blood-cell transfusions. Vox Sanguinis 2011, July: 6.

5. Hendrickson JE, Desmarets M, Deshpande SS, Chadwick TE, Hillyer CD, Roback JD & Zimring JC. Recipient inflammation affects the frequency and magnitude of immunization to transfused red blood cells. Transfusion 2006; 46: 1526–1536.

6. Hoppe C, Klitz W, Vichinsky E & Styles L. HLA type and risk of alloimmunisation in sickle cell disease. Am J Hematol 2009; 84: 462–464.

7. Noizat-Pirenne F, Tournamille C, Bierling P, Roudot-Thoraval F, Le Pennec PY, Rouger P & Ansart-Pirenne H. Relative immunogenicity of Fya and K antigens in a Caucasian population, based on HLA class II restriction analysis. Transfusion 2006; 46: 1328–1333.

8. Higgins JM & Sloan SR. Stochastic modelling of human RBC alloimmunisation: evidence for a distinct population of immunologic responders. Blood 2008; 112: 2546–2553.

9. Schonewille H, Rene RP, de Vries RRP & Brand A. Alloimmune response after additional red blood cell antigen challenge in immunized hemato-oncology patients. Transfusion 2009; 49: 453–457.

10. Buetens O, Shirey RS, Goble-Lee M, Houp J, Zachary A, King KE & Ness PM. Prevalence of HLA antibodies in transfused patients with, without red cell antibodies. Transfusion 2006; 46: 754–756.

11. Sanz C, Ghita G, Franquet C, Martínez I & Pereira A. Red-blood-cell alloimmunisation and female sex predict the presence of HLA antibodies in patients undergoing liver transplant. Vox Sanguinis 2010; 99: 261–266.

12. Hendrickson JE & Hillyer CD. Noninfectious serious hazards of transfusion. Anesth Analg 2009; 108: 759–769.

13. Win N. Hyperhemolysis syndrome in sickle cell disease. Expert Rev Hematol 2009; 2: 111–115.

14. Hebert PC, Fergusson D, Blajchman MA, Wells GA, Kmetic A, Coyle D, Heddle N, Germain M, Goldman M, Toye B, Schweitzer I, van Walraven C, Devine D & Sher GD. Leukoreduction Study Investigators: clinical outcomes following institution of the Canadian universal leukoreduction program for red blood cell transfusions. J Am Med Assoc 2003; 289: 1941–1949.

15. Zimring JC, Welniak L, Semple JW, Ness PM, Slichter SJ & Spitalnik SL, for the NHLBI Alloimmunisation Working Group. Current problems and future directions of transfusion-induced alloimmunisation: summary of an NHLBI working group. Transfusion 2011; 51: 435–441.

Further reading

Murphy K. Janeway's Immunobiology, 8th edn. London and New York: Garland Science; 2011.