Answers are grounded in this paper and cite their sources.
RhD hemolytic disease is caused by maternal immunization to the D antigen, followed by subsequent transfer of maternal immunoglobulin G (IgG) across the placenta resulting in immune lysis of fetal red blood cells (RBCs). Severe RhD disease is still a source of clinical concern despite the introduction of prophylactic RhD immune globulin, and even in developed countries such as the United States and the European Union, the incidence of maternal alloimmunization remains at 1 to 1.5 percent of at-risk D– women, due mainly to failures of adequate RhD immune globulin administration or occult antepartum transplacental hemorrhage. Some 10 percent of these immunized women have a baby severely affected in utero, and the management of such pregnant women requires intensive monitoring to identify which fetuses are at risk of hemolytic disease of the fetus and newborn (HDFN) and identifying the optimal time for intervention to avoid fetal anemia and the risk of hydrops fetalis.1, 2 Over the past 10 years, with improved medical technology, and the introduction of Doppler examination of fetal cerebral blood flow, the accuracy of diagnosing fetal anemia has improved in the best treatment centers, thus reducing the number of invasive tests for diagnosis of HDFN.1, 2 Conventional management of seriously affected cases involves controlled early delivery (with risks of prematurity) and fetal intravascular transfusion (with risks of fetal exsanguination and boosting of maternal anti-D level). Although fetuses requiring intrauterine transfusion have an overall survival rate of greater than 90 percent in specialist units, fetal transfusions cannot be given safely before 20 weeks of gestation because of technical limitations and are associated with 1 to 3 percent fetal loss due to the procedure. Management of severe cases in early pregnancy is problematic with up to 20 percent fetal mortality in the presence of hydrops,1-3 and less conventional approaches have been attempted to prolong gestation prior to the need for intrauterine transfusions. Intensive maternal plasma exchange to remove IgG anti-D4 would seem logical, but has been abandoned as ineffective in the absence of concomitant immunosuppression due to rebound of antibody levels and the potential risks to the mother of repeated procedures.5 Maternal infusion of high-dose intravenous IgG immunoglobulin (IVIG) has also been used with some success in preventing severe HDFN,6, 7 similar to the use of IVIG in the management of neonatal alloimmune thrombocytopenia (NAIT). More recently, combination treatment with plasma exchange and IVIG to achieve temporary immunomodulation in cases of very severe maternal alloimmunization has also been used with some success.8 The mode of action of IVIG is uncertain, but may include increased catabolism of maternal IgG, blocking the transfer of pathologic anti-D across the placenta due to receptor competition, or inhibiting the hemolytic activity of the anti-D in the fetus by Fc receptor blockade of the fetal reticuloendothelial system. There is experimental evidence9 that the level of maternal IgG needs to be maintained above 20 g per L by IVIG administration to prevent placental transfer of anti-D, and a higher dose regime of IVIG has been shown to be more effective in reducing the morbidity associated with NAIT due to anti-HPA-1a.10 This becomes an expensive and labor-intensive process, however, and alternative noninvasive approaches to the management of HDFN would be welcome. A novel approach is proposed in this volume of TRANSFUSION by Nielsen and colleagues,11 who report in vitro results of recombinant mutant IgG anti-D deficient in hemolytic activity that are designed to block the hemolytic activity of maternal anti-D responsible for HDFN. A similar approach has been reported by Armour and colleagues,12, 13 although the ultimate target of this research was the management of NAIT, and the hypothesis was tested with recombinant mutant anti-D. For such mutant antibodies to be effective therapeutically, they must be given to the mother safely in adequate amounts, cross the placenta, bind effectively to D+ fetal RBCs, and block the binding of hemolytic maternal anti-D in the fetus. Is this a feasible proposition? Immunoglobulin class–specific transfer of maternal antibodies to the fetus occurs because maternal IgG (both IgG1 and IgG3 subclasses), but not IgM or IgA, crosses selectively into the fetal circulation. Transfer is an active process and IgG1 crosses the placenta early in pregnancy, and by 20 weeks of gestation maternal IgG1 is detectable in fetal blood at levels equal to those in the mother. At delivery, both IgG1 and IgG3 are found to be slightly higher in cord serum than in maternal serum (reviewed by Saji et al.14). IgG anti-D associated with HDFN is produced only as subclasses IgG1 and/or IgG3, and both are transferred equally efficiently across the placenta.15 It is now known that both the prolonged half-life of IgG (21 days) compared to other plasma proteins, and the ability to cross the placenta, is controlled by a unique receptor, FcRn, which is expressed in many human tissues including endothelium, epithelium, and placental syncytiotrophoblast cells (reviewed by Ghetie and Ward16 and Simister17). FcRn binds IgG in a pH-dependent manner with efficient binding at acidic pH (<6.5) but not at physiologic pH (>7.0). This binding is believed to take place in acidic vesicles after plasma proteins have been taken up by normal fluid-phase endocytosis. In endothelial cells, FcRn prevents the degradation of IgG by lysosomal enzymes by efficient binding at acidic pH in early endosomes, followed by recycling to the cell surface where IgG is subsequently released into the circulation at pH 7.4. In placental transport, it is believed that similar IgG uptake occurs via binding to FcRn present in syncytiotrophoblast granules, allowing transport across the syncytiotrophoblast, where the antibody is released into the villous stroma and subsequently translocated into the fetal circulation.16, 17 The FcRn interaction site on IgG is known from X-ray crystallography to be located in the hinge region at the CH2-CH3 interface, and site-directed mutagenesis has shown in animal models that the critical residues for maintaining plasma half-life and placental transfer are a group of highly conserved histidine residues (histidine 310, histidine 433, and histidine 435). The critical residue His435 has also been shown to determine efficient IgG1 transfer in a human placenta perfusion model.18 Therefore, any mutations of recombinant IgG anti-D designed to block fetal RBC hemolysis must avoid interfering with the FcRn-binding properties; otherwise the mutated anti-D would not survive in the circulation for the normal plasma half-life of 21 days and would not transfer efficiently across the placenta. Pathologic anti-D causing HDFN are generally IgG1 and/or IgG3 subclasses, which effectively opsonize fetal RBC for removal in the spleen by phagocytosis. Extensive in vitro immunologic studies indicate that the Fc portion of IgG interacts with mononuclear cells via specific receptors (FcγR) of which three classes are expressed on splenic macrophages, FcγRI, FcγRIIa, and FcγRIIIa.19FcγRI is a high-affinity receptor and low levels of serum IgG can inhibit effector function of pathologic anti-D; FcγRIIa is low affinity and does not bind RBC coated with physiologic levels of anti-D. In contrast, FcγRIIIa is highly effective in mediating opsonized RBC lysis in the presence of monomeric (plasma) IgG and is likely to be the primary receptor for in vivo lysis of IgG anti-D–coated RBC in the fetal spleen. IgG2 and IgG4 antibodies, in contrast to IgG1 and IgG3, are ineffective at inducing RBC destruction. The approach taken by Armour and colleagues12, 13 was to design an IgG1 hybrid Fc region, which would be inactive in binding to receptors FcγRI-III and complement C1q by incorporating the CH2 sequences 233 to 236 from IgG2, substituting IgG1 residues at 327, 330, and 331 with those from IgG4 and three additional modifications to remove allotypic residues, which might result in an immune response in some recipients. The hybrid IgG1 Fc was combined with the Fab antigen–binding variable region of an anti-D (Fog-1), which had previously been used in human studies of monoclonal anti-D, to create an antibody (Fog-1 G1Δnab) that lacked Fc effector functions in vitro and that also demonstrated the ability to block other hemolytic anti-D in vitro. This mutated anti-D was tested in vivo in D+ volunteers with autologous RBCs coated with the anti-D, and compared to polyclonal anti-D, clearance of RBCs was slower, incomplete, and transient, which was unexpected, and unexplained, although there was no evidence of hemolysis of the infused anti-D–coated RBCs.13 Although the mutated residues in Fog-1 G1Δnab are predicted not to interfere with the FcRn-binding regions, plasma half-life has not been examined, and formal placenta transfer experiments have not been reported, so it is not yet possible to predict whether this mutated antibody would be of clinical value in treating HDFN. A different approach was taken by Nielsen and colleagues,11 who used a mutant IgG3 anti-D with the long hinge region of the heavy chain deleted and a further mutation (cysteine131serine) to link the light chains via a disulfide bond to maintain stability.20 Loss of the hinge region removes complement-mediated RBC lysis, and Fcγ receptor–mediated hemolysis and phagocytosis of RBCs, and in the current studies the mutant IgG3 anti-D was effective in blocking the hemolytic activity of recombinant and polyclonal anti-D in vitro. In an attempt to retain FcRn-binding activity, a histidine 435 mutation was introduced into the hinge-deleted IgG3, but unfortunately the H435 mutant was less effective at blocking hemolysis. The circulating half-life of this antibody is also unknown, but could be predicted as short, because the half-life of normal IgG3 is shorter (7 days) than that of IgG1 (21 days) and the hinge deletion is likely to introduce major conformational change to the Fc region responsible for FcRn binding. Although the above reports are of considerable interest, unresolved questions remain. Because the plasma half-life is not known, it is not possible to predict whether sufficiently high, sustained levels of mutant anti-D can be achieved in vivo to be of therapeutic value. Because repeated doses would be required, there is the possibility of inducing an anti-IgG immune response that could interfere with efficacy. It is also not clear whether in vitro demonstration of competitive inhibition by nonhemolytic anti-D in artificial bioassays will be translated into effective blockade of hemolysis in vivo in the fetus, assuming that sufficient mutant anti-D crosses the placenta to bind to fetal RBCs. It will also be a challenge to persuade regulators to allow clinical studies with the mutant recombinant anti-D in pregnant women, given the difficulties already encountered with human studies of monoclonal anti-D for immunoprophylaxis.21 IgG of whatever type will only produce a temporary effect (“passive immunity”) and must be given repeatedly if adequate levels are to be sustained during pregnancy to achieve therapeutic effect. An alternative approach, currently under clinical trial in several immunologic disorders,22 would be to induce an active immune response to suppress antibody formation—so-called “tolerance.” After the immunologic mechanism of tolerance to “switch off” an antibody response to RBC was first demonstrated in animal models by the oral administration of antigen,23 early attempts were made at inducing tolerance in at-risk alloimmunized D– women by oral administration of D antigen in the form of encapsulated D+ RBC membrane fragments.24, 25 It was shown conclusively in a subsequent clinical trial, however,26 that oral “desensitization” was ineffective, and indeed increased maternal anti-D levels, so this labor-intensive experimental treatment was abandoned. When the immunologic mechanisms of mucosal tolerance and the control of the immune response became better understood, and the structure of the RhD protein identified, interest was renewed in tolerization as a form of specific immunosuppression to switch off the immune response to RhD. Most IgG antibody responses are dependent on T-cell help, and the production of antibodies specific for RBCs, including anti-D, is no exception. Antigen-specific T-helper (Th) cells recognize short antigen-derived peptides displayed by HLA Class II molecules expressed on the surface of specialized antigen-presenting cells (APCs). In the presence of costimulation, the Th cells become activated and induce antigen specific B cells to produce antibody directed against that antigen. It is now clear that the context in which such T cell-APC recognition takes place determines whether a specific immune response is activated or tolerized; for example, an epitope that induces an immune response when given parenterally can, paradoxically, induce tolerance if given via the mucosal route. An analysis of T-cell epitopes on the RhD protein using synthetic linear peptides, derived from the known amino acid sequence, has shown that a relatively small number of peptides are immunodominant and induce a Th-cell response in vitro in most individuals who have produced anti-D.27 These peptides stimulate antigen-specific memory T cells in vitro that are restricted primarily by HLA Class II (primarily DR15), and the level of anti-D in the donors correlates with the number of peptides that will stimulate Th-cell proliferation. This provides evidence that the D antigen is so immunogenic because of the number of antigen-specific peptides that can be generated in vivo in D– individuals. Control of an established immune response is least studied, but regulatory CD25+ T cells (Tr1) have been identified that inhibit allogeneic reactions through the production of IL-1028 and serve to “switch off” the antibody response in vivo. Studies of the nature of the immune response to RhD and its regulation have been limited by lack of a suitable animal model because wild-type rodents do not produce RhD-specific antibody. A humanized HLA-DR15–transgenic mouse model has recently been developed, however, which shows a number of similarities with the human IgG response to the RhD protein.29 The immunodominant RhD peptides identified in alloimmunized humans have been shown to switch off (tolerize) the immune response to the D protein in the humanized transgenic mice if given by a tolerogenic (nasal) route. This opens the possibility of mucosal tolerance to RhD as a novel noninvasive way of suppressing the anti-D levels in women at risk of HDFN by mucosal administration of synthetic RhD peptides, and clinical trials are currently under consideration.
No takes yet. Share an insight, caveat, or question.
S. J. Urbaniak (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: