On the immunologic basis of Rh immune globulin (anti-D) prophylaxis



This article corrects:

  1. On the immunologic basis of Rh immune globulin (anti-D) prophylaxis Volume 46, Issue 8, 1271–1275, Article first published online: 28 July 2006

  • Kumpel BM. On the immunologic basis of Rh immune globulin (anti-D) prophylaxis. Transfusion 2006;46:1271-1275. The Publisher regrets several errors in the Editorial, primarily the incorrect use of “ABO-matched pregnancies” within the second paragraph on page 1271. The correct sentence should read:
    In Liverpool, IgG anti-D was given to clear D+ RBCs from the circulation1 because it was known that destruction of fetal RBCs by maternal anti-A and/or anti-B in ABO incompatible pregnancies reduced the incidence of RhD immunization2,3.
    The complete, corrected editorial has been republished below.

The administration of Rh immune globulin (RhIG) to D– women after parturition is a remarkably successful therapy and has prevented many thousands of D+ infants worldwide suffering from hemolytic disease of the fetus and newborn (HDFN) since its introduction nearly 40 years ago.

In 1960, clinicians on both sides of the Atlantic independently thought it might be possible to prevent RhD immunization with passive anti-D. There is no claim for priority of one group over the other (J.C. Woodrow, personal communication, April 2006). Their rationales were different. In Liverpool, IgG anti-D was given to clear D+ RBCs from the circulation1 because it was known that destruction of fetal RBCs by maternal anti-A and/or anti-B in ABO-incompatible pregnancies reduced the incidence of RhD immunization.2,3 The approach in New York was based on antibody-mediated immune suppression (AMIS)4 after early work on vaccination showed suppression of immune responses to diphtheria toxoid by excess antitoxin5 and lack of anti-D responses to anti-D–coated RBCs in volunteers.6 Reviews detail the pathogenesis of HDFN and the experimental studies, clinical development, and current usage of RhIG (anti-D) prophylaxis and clinical management of affected cases of HDFN.7–10

Despite the routine use of antenatal and postnatal RhIG prophylaxis, its mechanism of action is still unclear.11 Recent work by Branch and colleagues12 published in the August 2006 issue of TRANSFUSION suggests a role for immunomodulatory cytokines. Apart from RBC clearance and AMIS mentioned above, other theories (anti-idiotypic networks, masking of epitopes) are unlikely and unproven.13 Animal models have limitations.

Consideration of the characteristics of the anti-D immune response and its prevention by passive anti-D in the light of recent advances in immunology may enable a reappraisal of how prophylactic anti-D may work.

RhD immunization occurs after fetomaternal hemorrhages (FMH)14 or transfusion of D+ blood. Importantly, approximately 5 to 15 weeks elapse before anti-D is detectable,15 very much slower than typical 5- to 10-day antibody responses to foreign antigens such as microorganisms. IgG is often the only immunoglobulin class detected and is usually high-affinity16 IgG1 and/or IgG3, the typical IgG subclass response to protein antigens, indicating T-helper (Th) cell involvement. The likelihood of an anti-D response occurring is roughly proportional to the volume of RBCs and to the frequency of occurrence and is greater toward homozygous than heterozygous D+ RBCs.17 No significant HLA association with anti-D immunization has been documented.18

Effective suppression requires approximately 200 IgG anti-D molecules per RBC although with the doses used in clinical practice more anti-D is bound.19 Prophylaxis should be given within 72 hours of delivery, this timing being adopted after early successful studies using prison volunteers.4 Anti-D given 13 days after injection of RBCs did not prevent RhD immunization in some subjects.20 AMIS was not effective once the anti-D response was established.21 Nevertheless, for women in whom postnatal prophylaxis failed, levels of immune anti-D and severity of HDFN in subsequent pregnancies were less than in untreated patients, suggesting a long-lasting regulatory effect of passive anti-D on the immune response.22 For protection against immunization, the D+ RBCs must be removed from the circulation within 5 days of administration of anti-D.19 In countries where the FMH is monitored, patients with large FMH are followed up for 1 to 2 days after injection to check whether the fetal cells have cleared and, if not, additional anti-D is given.23 Clearance of RBCs is the only early clinical indicator of effectiveness.

Anti-D–coated RBCs are sequestered in the spleen24 (Fig. 1). The functions of the spleen are 1) clearance of senescent blood cells and foreign microorganisms by filtration and 2) development of immune responses to blood-borne antigens. These activities are anatomically separated, taking place in the red pulp and white pulp, respectively.

Figure 1.

Diagrammatic representation of the human spleen. Blood entering the red pulp flows through a reticular endothelial network of splenic cords and vascular sinusoids lined with macrophages. Receptors on the macrophages detect old or foreign cells mediating their phagocytosis. Macrophages are excellent “garbage collectors” but poor antigen-presenting cells (APCs). Apoptotic blood cells (senescent cells destined for recycling) have externalized phosphatidyl serine (PS) on the cell membrane, which is recognized by a specific macrophage receptor (PS receptor) triggering phagocytosis25 and production of TGF-β, an anti-inflammatory cytokine.26 This results in their “silent death.” Microorganisms are detected by the innate immune system using a variety of pattern recognition receptors (PRRs) including mannose and lectin receptors (recognizing species-specific sugar structures and microbial cell wall glycoproteins and glycolipids),27 scavenger receptors (for Gram-positive and -negative bacteria), and some toll-like receptors (TLRs), targeting microbial cell products.28 PRR interactions result in phagocytosis of microbes and release of proinflammatory cytokines that enhance immune responses to these pathogens and subsequent antibody production. Finally, macrophages express several IgG Fc receptors (FcγR) that form the link between innate and adaptive immunity.29 These receptors recognize cell-bound IgG “labeling” target cells to which immunity has developed, mediating their accelerated phagocytosis and destruction. This is a noninflammatory process, avoiding unnecessary collateral tissue damage. Immune responses to blood-borne antigens take place in the white pulp. Small arterioles are surrounded by periarteriolar lymphoid sheaths (PALS) containing many T cells in which are scattered lymphoid follicles and germinal centers comprising mostly B cells. Marginal zones adjacent to PALS have DCs, B cells, and CD4+ Th cells, while antibody-secreting plasma cells reside in the red pulp. The margins of these regions are diffuse and cells traffic widely.30 DCs, the most efficient APCs, are poorly phagocytic but acquire antigen nonspecifically by endocytosis during surveillance of their environment. After proteolysis of ingested material in endosomes, peptides are presented on HLA Class II molecules for recognition by T cells that are very motile and briefly contact many DCs. Proinflammatory signals increase cytokine production and the expression of costimulatory molecules on DCs. If there is a good fit between peptide-HLA Class II complexes on DCs and T-cell receptors (TCRs) on Th cells together with sufficient costimulation and IL-12 production, the DCs will activate the contacting T cells.31 Activated antigen-specific Th cells then clonally proliferate and migrate30 to search for B cells expressing the same peptide-HLA Class II complex as the DCs. B cells acquire antigen by B-cell receptor (BCR) mediated endocytosis that requires conformational recognition of antigen through membrane IgM (the BCR). The TCRs of Th cells then recognize the peptide-HLA Class II complexes and activate the contacted B cells through costimulatory molecules and IL-4 leading to B-cell proliferation, migration away from the follicles, and antibody secretion.32 Reproduced, with permission, from Roitt I, Delves P. In: Roitt’s Essential Immunology, 10th ed. Blackwell Publishing, Oxford, UK, 2001, Chapter 8, Fig. 08-08 (taken from the associated Web site: http://www.roitt.com/).

These roles of the spleen may illuminate the processes involved in the primary anti-D response and its prevention by passive anti-D. Because macrophages do not have receptors for blood groups, allogeneic RBCs will not be recognized and will circulate normally until they become senescent and/or apoptotic. This may take several weeks because the mean RBC life span is approximately 120 days. They will then be phagocytosed by macrophages either through the phosphatidylserine (PS) receptor binding to PS on apoptotic cells25 or by FcγR recognition of IgG autoantibodies bound to senescent RBCs.33 Although these mechanisms of allogeneic RBC removal would be noninflammatory,26 there may be cases when sufficient RhD peptides are presented to stimulate an anti-D response. This conjecture is consistent with the fact that not all immunized D– individuals make anti-D. If, however, RhIG prophylaxis is given before the anti-D response occurs, passive anti-D binds to the D+ RBCs ensuring their rapid FcγR-mediated phagocytosis and subsequent anti-inflammatory destruction.34 The effect of passive anti-D is to route the IgG-coated RBCs to red pulp macrophages for quick disposal and “environmentally friendly” recycling, preventing aging of the RBCs when an anti-D response may be initiated. This hypothesis is an immunologic explanation for the role of accelerated RBC clearance in RhIG prophylaxis.

Experimental studies on AMIS using in vitro or in vivo murine models have not given much insight into the way RhIG prophylaxis might work. Mice do not have blood groups and only express a RhCE-like protein but lack RhD.35 Sheep RBCs (SRBCs) have generally been used for immunization of mice but these generate a very rapid (4-day) anti-species IgM complement fixing antibody response, quite unlike very slow IgG anti-D production. Phagocytosis of SRBCs is likely to be mediated by innate immune receptors to foreign sugars on the xenogeneic RBCs, generating pro- rather than anti-inflammatory responses as occurs with allogeneic RBCs. Passive antibody to SRBCs gives a dose-dependent reduction in the plaque-forming anti-SRBC response (to a maximum of 99%), in contrast to the all-or-nothing effect of prophylactic anti-D, but is only achieved at doses of IgG:SRBCs approximately 1000 times that of anti-D.36,37 Epitope masking may provide an explanation for prevention of the SRBC antibody response by IgG, IgE, and F(ab′)2 fragments of anti-SRBCs36 and also by methoxypolyethylene glycol–coated SRBCs,38 because AMIS was effective in mice deficient in FcγR.36 AMIS could also be due to clearance of the SRBCs to the liver, a nonimmunogenic organ. Rabbits have a triallelic blood group system (HgA,E,F) and were used in early experiments on AMIS.39–41 The kinetics of clearance of allogeneic RBCs in rabbits and humans are not identical and AMIS in rabbits was found to be epitope-specific,39 not particle-specific, as in RhIG prophylaxis.42 Recently a rabbit model of HDFN has been established and used to investigate novel therapies.43

A role for direct cellular inhibition cannot be ruled out. Most immune signaling pathways have regulatory mechanisms. Many receptors including FcγRs, B-cell receptors (BCRs), and T-cell receptors (TCRs) contain activatory (immunoreceptor tyrosine–based activation motif, ITAM) or inhibitory (ITIM) modules in their cytoplasmic domains or associated γ-chains.44 Inhibitory FcγRs are expressed on B cells, mast cells, monocytes, and dendritic cells (DCs) and likely act in concert with activatory receptors to set a balance between positive and negative signals to fine tune immune responses.45 When the ITIM-bearing FcγRIIb on B cells is co–cross-linked with the BCR (with ITAM in subunit) by cell-bound IgG and antigen, respectively, these receptors localize to lipid rafts (membrane microdomains) to initiate an intracellular inhibitory signaling cascade resulting in suppression of B-cell activation, proliferation, and antibody synthesis.46 This might occur with prophylactic anti-D–coated D+ RBCs and immature B cells.13 On DCs, pro- and anti-inflammatory mediators (including IFN-γ and monomeric IgG) regulate expression of activatory and inhibitory FcγR, with ligation of these receptors differentially influencing DC maturation and hence their potential for immunogenicity or tolerance.47

Thus the role of cytokines may be very important, especially at the cellular level. Branch and colleagues report in the August 2006 issue that after antenatal administration of 300 µg RhIG to D– women there was an increase in two anti-inflammatory cytokines, TGF-β and prostaglandin E2 (PGE2), in maternal plasma.12 This response to prophylactic anti-D contrasts with the effect of much higher doses of RhIG given to treat D+ ITP patients (25-50 µg/kg to children;48 50-75 µg/kg to adults49) when transient production of proinflammatory cytokines predominated, sometimes accompanied by chills. An explanation for these apparently conflicting data may be that rapid sequestration of large numbers of autologous RBCs with accompanying hemolysis may have the opposite effect to clearance of a few allogeneic RBCs. There was no identification of FMH in the antenatal study,12 and so it is not known if the cytokine response resulted from anti-D interacting with D+ RBCs. Therefore, the effect of nonspecific (non–anti-D) IgG in RhIG cannot be overlooked, being approximately 100 times more than the anti-D (25-40 mg nonspecific IgG per 300 µg anti-D), and it would behave like IVIG. IVIG stimulated release of the anti-inflammatory cytokines IL-1Ra50 and IL-1049 into plasma of hypogammaglobulinemic or ITP patients. Thus the IVIG-like component of RhIG may have contributed to the anti-inflammatory cytokine response observed by Branch and coworkers.12

In summary, the evidence indicates that prevention of RhD immunization by prophylactic anti-D might occur both by rapid macrophage-mediated clearance of anti-D–coated RBCs and/or by down regulation of immature DCs or anti-D–specific B cells before development of the anti-D response occurs. But after consideration of the anatomy and function of the spleen, where the flow of RBCs is predominantly through the red pulp (the site of destruction of IgG-coated RBCs) and few RBCs enter the white pulp (areas of immune response generation) the hypothesis of accelerated clearance and destruction of D+ RBCs is, in the author’s opinion, likely to be the main mechanism of action of passive anti-D.

Finally, in the current era of stringent regulatory controls, it is unlikely that such an innovative medicine as RhIG prophylaxis could now be given to pregnant women without accurate knowledge of its mechanism of action or disease etiology. Were it not for the foresight and drive of the original workers to introduce prophylactic anti-D many years ago, there would be a continual toll from RhD hemolytic disease.