Rostrum: Human IgE-independent systemic anaphylaxis

Abstract

Anaphylaxis is a rapidly developing, life-threatening, generalized or systemic allergic reaction that is classically elicited by antigen (Ag) crosslinking of Ag-specific IgE bound to the high affinity IgE receptor, FcεRI, on mast cells and basophils. This initiates signals that induce cellular degranulation, with release and secretion of vasoactive mediators, enzymes, and cytokines. IgE-independent mechanisms of anaphylaxis, however, have been clearly demonstrated in experimental animals. These include IgG-dependent anaphylaxis, which involves the triggering of mediator release by IgG/antigen complex crosslinking of FcγRs on macrophages, basophils and neutrophils; anaphylaxis mediated by the binding of complement-derived peptides, C3a and C5a, to their receptors on mast cells, basophils and other myeloid cells; and direct activation of mast cells by drugs that interact with receptors on these cells. Here, we review the mechanisms involved in these IgE-independent forms of anaphylaxis and the clinical evidence for their human relevance. We conclude that this evidence supports the existence of all three IgE-independent mechanisms as important causes of human disease, although practical and ethical considerations preclude their demonstration to the degree of certainty possible with animal models. Further, we cite evidence that different clinical situations can suggest different mechanisms as having a primal role in anaphylaxis and that IgE-dependent and distinct IgE-independent mechanisms can act together to increase anaphylaxis severity. As specific agents become available that can interfere with mechanisms involved in the different types of anaphylaxis, recognition of specific types of anaphylaxis is likely to become important for optimal prophylaxis and therapy.

Introduction

Anaphylaxis is a rapidly developing, life-threatening, generalized or systemic allergic reaction¹. Foods, drugs, and insect stings are the most common causes of this disorder². Classically, anaphylaxis is induced by antigen (Ag) crosslinking of Ag-specific IgE that has bound to the high affinity IgE receptor (FcεRI) on mast cells and basophils³. Crosslinking of IgE and its receptor induces a signaling cascade that results in mast cell degranulation with release of mediators, including histamine, as well as preformed cytokines and proteases, and synthesis and secretion of additional cytokines as well as lipid mediators, such as platelet activator factor (PAF), leukotrienes and prostaglandins⁴. Passive immunization studies in which mice were sensitized by injecting an Ag-specific IgE antibody (Ab), followed by enteral or parenteral exposure to that Ag, support the importance of IgE and mast cells in Ag-induced shock⁵. Indeed, both passive and active immunization studies in which mice were challenged orally with the appropriate Ag have generally demonstrated that genetic or Ab elimination of IgE, mast cells, or the IgE-binding chain of FcεRI, FcεRIα, completely suppresses anaphylaxis development^6-8. In contrast, studies in which mice were actively immunized with an Ag, followed by parenteral challenge with the same Ag, have often revealed that anaphylaxis can occur in the absence of the classical IgE/FcεRI/mast cell pathway and demonstrated that a disorder that closely resembles IgE-mediated systemic anaphylaxis can be mediated by mechanisms that involve IgG rather than IgE^9-11. Consistent with this, mice that are passively immunized with an IgG1, IgG2a, or IgG2b (but not IgG3) monoclonal Ab (mAb) specific for the hapten trinitrophenyl (TNP) develop anaphylaxis, that is nearly indistinguishable clinically from IgE-mediated anaphylaxis, when challenged parenterally, but not enterally, with a TNP-protein conjugate^{5, 6, 8}. These observations, coupled with several human clinical observations, suggest that IgE-independent anaphylaxis may be clinically important. Here, we will first review observations that prove the existence of IgG-mediated anaphylaxis in mice and describe differences in the mechanisms behind the classical, IgE-mediated pathway and the alternative IgG-mediated pathway in this species, as well as the clinical implications of these differences. Next, we will review observations that support the existence of IgG-mediated anaphylaxis in humans, as well as the implications and limitations of these observations. Finally, we will discuss the evidence and its limitations for other, Ab-independent mechanisms of anaphylaxis in both mice and humans.

Murine evidence for IgG-mediated anaphylaxis

Evidence for IgE-independent, IgG-dependent anaphylaxis was provided by studies in which mice were immunized, then parenterally challenged with a potent antigen¹¹. In some of these active immunization models, disease developed even if mice were first treated with an anti-IgE monoclonal antibody, but was suppressed if mice were instead treated with a rat IgG2b monoclonal Ab (mAb), 2.4G2. This mAb binds to and triggers, but then blocks the inhibitory low affinity IgG receptor, FcγRIIB, and the stimulatory low affinity IgG receptor, FcγRIII, and indirectly blocks the other murine FcγRs: FcγRI and FcγRIV^{11, 12}. The existence of IgE-independent anaphylaxis in actively immunized mice was demonstrated most conclusively by studies that: (1) induced severe anaphylaxis in actively immunized IgE- or FcεRIα-deficient mice, but not in actively immunized mice that lacked all stimulatory FcRs (i.e.; FcRγ-deficient mice); and (2) demonstrated reduced severity or absence of anaphylaxis in actively immunized mice that lacked function of one or more of the stimulatory murine FcγRs^11-14.

Subsequent passive immunization studies demonstrated that an anti-IgE mAb would block anaphylaxis when mice were sensitized with an Ag-specific IgE mAb, but not when mice were sensitized with an Ag-specific IgG1, IgG2a, or IgG2b mAb, while reciprocal results were found when passively immunized mice were treated with 2.4G2^{10-12, 15, 16}. The severity of systemic anaphylaxis in these IgG passive immunization models was normal or increased in mice that that were deficient in FcεRIα¹³. In contrast, anaphylaxis in mice passively sensitized with an Ag-specific IgG1 mAb was totally absent in mice deficient in FcγRIII (the only stimulatory murine FcγR that binds mouse IgG1), while total suppression of anaphylaxis in mice sensitized with an IgG2a mAb (which binds to all three stimulatory murine FcγRs) required the deletion or blocking of all of these receptors^{10, 12}. The importance of FcγRs in murine IgG-dependent anaphylaxis was also shown by the unique inability of IgG3, among the murine IgG isotypes, to mediate anaphylaxis, which correlates with the observation that IgG3 is the only murine IgG isotype that fails to bind to any stimulatory murine FcγR^{17, 18}.

Clinical implications of studies of murine IgG-mediated anaphylaxis

Studies of murine IgG-mediated anaphylaxis by several groups have evaluated the mediators involved, the responsible cell types, and the quantities of Ag required to induce shock. Nearly all studies have identified PAF, rather than histamine, as the mediator most important in IgG-mediated anaphylaxis in actively immunized mice^{11, 19, 20}, although this has not been investigated thoroughly in passively immunized mice. In contrast to agreement about the importance of PAF in IgG-mediated anaphylaxis in actively immunized mice, different studies have identified monocyte/macrophages, basophils, or neutrophils as the critical cell type in IgG-mediated anaphylaxis^{11, 19, 20}. All of these cell types express FcγRIII and FcγRIV in mice and all are capable of producing PAF in response to appropriate stimuli^{18, 20-24}. Differences in cell types that appear to be responsible for IgG-mediated anaphylaxis may result from differences in mouse strains used, stimuli that elicit anaphylaxis, endogenous bacterial flora and/or animal husbandry practices.

Results of studies that compared the doses of antigen required to induce IgE- vs. IgG1-mediated anaphylaxis suggest that the dose of challenge Ag determines when IgG-mediated anaphylaxis can occur. In mice that were passively sensitized with high affinity IgE or IgG Abs to TNP, 100- to 1,000-fold less TNP-conjugated protein was required to induce shock in IgE- than in IgG-sensitized mice¹⁵. This was true regardless of the extent of TNP-labeling of the TNP-conjugated protein, although less TNP-conjugate was required to induce either IgE- or IgG-mediated anaphylaxis when the protein was heavily labeled¹⁵. These observations are consistent, respectively, with the much higher affinity of FcεRI than FcγRIII, the much higher ratio of cell-bound to serum IgE than IgG, and the better crosslinking of an Ag-specific mAb by an Ag that has multiple copies of the epitope bound by that mAb.

Because IgG-mediated anaphylaxis requires a much larger dose of Ag than IgE-mediated anaphylaxis, anaphylaxis induced by parenteral administration of a small quantity of Ag (e.g.; insect sting) is much more likely to be IgE-mediated. Anaphylaxis induced by Ag ingestion (e.g.; food allergy), similarly, always appears to be IgE-mediated^{6, 7}, because induction of anaphylaxis in food allergy models requires systemic absorption of ingested Ag and only a very small percentage of ingested Ag is absorbed with all epitopes intact^{7, 25}. In contrast, both IgE- and IgG-mediated anaphylaxis can be induced by parenteral administration of a relatively large quantity of Ag¹⁵ (e.g.; infusion of a therapeutic Ab or drug), particularly an Ag that has multiple iterations of an antibody-reactive epitope (e.g.; a carbohydrate Ag, such as dextran).

The difference in Ag dose requirement for IgE vs. IgG-mediated anaphylaxis allows IgG to act both as a mediator of anaphylaxis and a blocker of IgE-mediated anaphylaxis, depending on Ab and Ag concentrations (Figure 1). In the presence of Ag-specific IgE, Ag-specific IgG Ab will block anaphylaxis that would otherwise be induced by a low dose of Ag by intercepting Ag before it can bind to mast cell-associated IgE and by interacting with the inhibitory receptor, FcγRIIB^{15, 26} (Figure 1A and B), but mediate anaphylaxis induced by a higher Ag dose (Figure 1C and D). The ability of IgG to both block Ag access to mast cell-associated IgE and to mediate anaphylaxis through IgG/Ag complex binding to stimulatory FcγRs can create the counter-intuitive situation where an intermediate dose of Ag will induce IgG-, but not IgE-mediated anaphylaxis in the presence of both Ag-specific IgE and IgG¹⁵ (Figure 1C).

Figure 1. Relative concentrations of Ag and Ab determine the roles of IgE and IgG Abs in Ab-mediated anaphylaxis.

IgE-mediated anaphylaxis requires considerably less Ab and Ag than IgG-mediated anaphylaxis. Consequently, when Ab levels are low (A), only IgE-mediated anaphylaxis can occur. When Ag levels are low but Ab levels are high (B), IgG “blocking” Abs prevent IgE-mediated anaphylaxis by intercepting Ag before it can bind to FcεRI-associated IgE and by binding to the inhibitory receptor, FcγRIIB, but the quantity of IgG/Ag complexes is too low to trigger IgG-mediated anaphylaxis. Consequently, anaphylaxis does not occur. When Ag and Ab levels are both high, but Ab is in excess to Ag (C), IgG Abs block the binding of Ag to FcεRI-bound IgE but IgG/Ag complexes can bind to FcγRs; consequently, only IgG-mediated anaphylaxis occurs. When Ag and Ab levels are both high, but Ag is in excess (D), IgG/Ab complexes are sufficient to trigger IgG-mediated anaphylaxis and enough Ag escapes IgG blockade to bind to FcεRI-associated IgE and trigger IgE-mediated anaphylaxis.

Limitations of studies of mouse IgG-mediated anaphylaxis

Although experimental evidence that IgG-mediated anaphylaxis can occur in mice is unequivocal, there are concerns about the interpretation of studies that identify the importance of different cells and receptors in this process. Nearly all studies that analyze the importance of specific receptors use gene deletion to eliminate specific FcγRs or Abs to block these receptors, while studies that analyze the importance of specific cell types either use Abs or drugs that eliminate these cell types or transfer a specific cell type to a recipient mouse. Although gene deletion can cause complete deficiency of a specific receptor, the elimination of one receptor under at least some circumstances can increase the expression and/or signaling capacity of the remaining receptors^{15, 24}. This may lead investigators to exaggerate the importance of the remaining receptors. IgG mAbs to a specific receptor may influence non-targeted receptors by signaling through the targeted receptor and/or binding of the Fc part of the IgG mAb to non-targeted receptors (e.g.; mAb 2.4G2, which binds to FcγRIIB and FcγRIII, at least partially suppresses the expression and function of FcγRI and FcγRIV¹²). Techniques used to delete macrophages, such as silica, clodronate liposomes, and gadolinium, may activate these cells before killing them²⁷; cytokines or mediators produced by the activated cells may influence the ability of remaining cell types to contribute to anaphylaxis. Antibodies used to delete specific cell types, including neutrophils, basophils and platelets, may activate and deplete complement upon binding the targeted cell type or form immune complexes with cell membrane antigens that bind to FcγRs on other cells; both may influence the ability of these other cells to contribute to anaphylaxis. In this regard, for example, we have found that IgG-mediated anaphylaxis is suppressed when platelets are eliminated with IgG anti-platelet antibodies, but not when platelets are eliminated with neuraminidase¹¹. Abs that appear to eliminate a cell type, when that cell type is studied in one organ (e.g.; blood), may actually cause redistribution of that cell type to another organ (e.g.; spleen). Finally, because cells, such as neutrophils, may be partially activated by in vitro purification procedures, it is possible that transfer studies with purified neutrophils¹⁹ may exaggerate the importance of this cell type in IgG-mediated anaphylaxis. Although none of these techniques are used to evaluate whether IgG-mediated anaphylaxis exists in humans, concerns about their use in mice affects hypotheses about which receptors and cell types might contribute to putative human IgG-mediated anaphylaxis.

Theoretical considerations about the possibility of human IgG-mediated anaphylaxis

Because the kinds of experiments that have proven the existence of IgG-mediated anaphylaxis in mice would not be appropriate in humans, even when possible, evidence for human IgE-mediated anaphylaxis is typically anecdotal and correlative, rather than definitive. In addition, important differences between mice and humans, including differences in the properties of their IgG isotypes, differences in their FcγRs, differences in cellular FcγR distribution, and differences in the properties of FcγR-expressing cells themselves raise questions about the applicability of observations about IgG-mediated anaphylaxis in mice to humans. However, consideration of each of these differences fails to provide a reason for thinking that human IgG-mediated anaphylaxis is unlikely.

Human IgG1 and IgG3, and possibly IgG4, bind to human FcγRs with an affinity range similar to what is observed in mice²⁴. Although humans lack FcγRIV, they express FcγRI and FcγRIII and have stimulatory FcγRs, FcγRIIA and FcγRIIC, which are not present in mice²⁴. Activation of human basophils, monocyte/macrophages and neutrophils can cause these cells to produce PAF, which has been associated with human anaphylaxis^{28, 29}. In addition, as noted earlier, human neutrophils can mediate IgG-dependent anaphylaxis when infused into mice¹⁹.

Differences in cellular FcγR expression may actually make it more likely for IgG to mediate anaphylaxis in humans than in mice, because while both human and mouse mast cells express a stimulatory FcγR (FcγRIII in mice, FcγRIIA and possibly, FcγRIIC in humans), human mast cells express relatively little or no inhibitory FcγRIIB, while mouse mast cells express relatively large amounts of this receptor^{24, 30, 31}. Similarly, the much larger number of granules in human than in murine basophils³² makes degranulation of these cells more likely to induce anaphylaxis in the former species. Taken together, there is no reason to believe that differences in IgG isotypes, FcγRs, cellular distribution of these receptors, or the physiology of FcγR-expressing cells make IgG less likely to mediate anaphylaxis in people than in mice.

Clinical evidence for human IgG-mediated anaphylaxis

Several clinical observations support the importance of IgG-mediated anaphylaxis in humans, although each of these observations is open to alternative conclusions. Multiple researchers have described anaphylaxis in patients who were treated with a biologic therapeutic and had IgG, but not detectable IgE Ab that was specific for that therapeutic³³. This has been reported in transfused and intravenous immunoglobulin-treated IgA-deficient individuals (who had developed IgG anti-IgA Abs)³⁴; individuals treated with a variety of chimeric, humanized and even fully human mAbs³⁵; individuals treated with dextran³⁶ or aprotinin³⁷; and von Willebrand factor-deficient individuals who have been infused with von Willebrand factor³⁸. It is noteworthy that all of these examples of putative IgG-mediated anaphylaxis involve the parenteral administration of a large quantity of an Ag – precisely the condition that favors IgG-mediated anaphylaxis in mice. It remains possible, however, that the individuals who developed anaphylaxis may have had cell-bound, FcεRI-associated therapeutic-specific IgE without detectable Ag-specific IgE in serum. This is possible because the high affinity of FcεRI for IgE and the relatively low level of FcεRI crosslinking required to induce mast cell degranulation allow sufficient IgE to bind to mast cells to mediate their activation, even when serum IgE levels are very low. Similarly, other evidence that supports the existence of IgE-independent anaphylaxis, such as anaphylaxis without evidence of basophil activation, anaphylaxis in the absence of increased serum tryptase³⁹, and anaphylaxis in skin test-negative individuals might be explained by a lack of sensitivity of the tests used, the small time window in which a test reflective of IgE-mediated anaphylaxis remains positive, or by restrictions in the location and properties of mast cells responsible for IgE-mediated anaphylaxis (e.g.; Ag-specific IgE might be bound to vascular mast cells, but not to skin mast cells).

Additional evidence in favor of IgG-mediated anaphylaxis comes from a study of individuals treated with the chimeric mAb infliximab, which demonstrated that the presence of IgG anti-infliximab Ab levels ≥ 8 μg/ml was associated with a relative risk of anaphylaxis of 2.4^{40, 41}. Although this association suggests that IgG may have been involved in anaphylaxis pathogenesis, it is also possible that higher IgG antibody levels were a marker for higher IgE Ab levels.

Because PAF is more strongly associated with IgG- than with IgE-mediated anaphylaxis in mice, reports that serum PAF levels are higher in patients undergoing anaphylaxis than in a control group of patients and that serum concentrations of PAF acetylhydrolase, the enzyme that breaks down PAF, correlate inversely with anaphylaxis severity^{28, 29}, are also consistent with the existence of human IgG-mediated anaphylaxis. However, because PAF can also be produced by mast cells and basophils in response to IgE crosslinking^{42, 43}, and because human myeloid cells other than mast cells and basophils can express FcεRIα²⁴, these observations may instead reflect a role for PAF in human IgE-mediated anaphylaxis. This alternative explanation is somewhat refuted by evidence that human neutrophils, monocyte/macrophages and basophils can produce PAF in response to FcγR crosslinking and that human neutrophils can mediate anaphylaxis in mice¹⁹; however, it could be argued that there is no direct evidence that the amount of PAF or other mediators produced by these cells is sufficient to induce anaphylaxis in humans.

One last, intriguing piece of evidence in favor of human IgG-mediated anaphylaxis comes from a study demonstrating increased frequency of a gain of function allele of the stimulatory FcγR, FcγRIIA, in common variable immunodeficiency patients who have IgG anti-IgA antibodies and develop anaphylaxis after IVIG infusion⁴⁴. The impact of this elegant work, however, is limited by the small number of patients studied, the possibility that the mechanism that associates increased FcγRIIA activity with anaphylaxis may be indirect (e.g.; increased FcγRIIA function might promote an IgE response), and the lack of other reported associations of FcγR polymorphisms with human anaphylaxis.

Taken together, these observations appear to us to make the existence of human IgG-mediated anaphylaxis highly likely, particularly when anaphylaxis occurs in the presence of relatively high titers of specific IgG Ab and undetectable specific IgE in individuals who have been injected or infused with relatively large quantities of the recognized Ag. We concede, however, that these observations do not provide absolute proof of the existence or clinical importance of human IgG-mediated anaphylaxis.

Murine complement-mediated anaphylaxis

Studies in mice demonstrate that C3a and C5a, small peptides derived from C3 and C5, respectively and known as anaphylatoxins, can activate mast cells and other myeloid cells⁴⁵; however, there is a lack of convincing evidence that they are either required for antibody-mediated anaphylaxis or can produce shock in the absence of other factors in this species. Passive anaphylaxis studies have demonstrated that ligation of mast cell C3aR or C5aR are required for the induction of skin swelling by injected C3a and C5a, respectively, and that both anaphylatoxins stimulate mast cell degranulation⁴⁶. Studies in which mice were injected with soluble peanut extract demonstrated that components of that extract activated complement through both the classical and the lectin pathways^{47, 48}. C3a produced in this manner could stimulate hypothermia through a macrophage and PAF-dependent process, but only when mice were treated with a β-adrenergic antagonist and/or a long-acting formulation of IL-4 to increase their responsiveness to PAF⁴⁷. Perhaps more importantly, complement activation by peanut extract acted synergistically with IgE-mediated mast cell activation to cause shock in the absence of exogenous β-adrenergic antagonist or IL-4⁴⁷. This suggested that combined IgE-dependent mast cell activation and complement activation by peanuts might be one explanation for the severity of peanut-induced anaphylaxis. The suggestion that complement activation exacerbates anaphylaxis induced by other mechanisms, but is insufficient to independently induce murine systemic anaphylaxis, is also consistent with the inability of Ag-specific mouse IgG3, which efficiently activates complement, but does not bind to FcγRs, to sensitize mice to develop anaphylaxis following relevant Ag challenge. Murine studies may, however, underestimate the importance of complement-derived anaphylatoxins in human anaphylaxis, because complement components are less capable of inducing anaphylaxis in another rodent (the rat) than in some larger mammals, such as dogs and pigs⁴⁹.

These observations suggest that Ab-mediated anaphylaxis should be less severe in the absence of C3 or anaphylatoxin receptors when anaphylaxis is mediated by a complement-activating isotype, such as mouse IgG2a. Even IgE-mediated anaphylaxis might be expected to be less severe in the absence of complement or anaphylatoxin receptors, if IgE-mediated mast cell activation results indirectly in anaphylatoxin production. However, it is also possible that decreased stimulation of the G-protein-dependent anaphylatoxin receptors increases the responsiveness of other G-protein-dependent receptors, such as the histamine receptors, that mediate anaphylaxis, (this would be analogous to the increased signaling through FcγRIII that is observed in the absence of FcεRIα⁵⁰). Studies are required to evaluate these possibilities.

Complement-mediated anaphylaxis: Observations in humans

The potential for complement-mediated human anaphylaxis is suggested by studies showing expression of one or both anaphylatoxin receptors on human mast cells, basophils, other myeloid cells, and vascular endothelial cells^51-54. A role for complement in human Ab-mediated anaphylaxis is suggested by a correlation between the severity of anaphylaxis and serum anaphylatoxin levels, although the risk associated with elevated anaphylatoxin levels is not as high as the risk associated with elevated tryptase or histamine levels⁵⁵. Complement may have a particularly important anaphylaxis-enhancing role in vespid toxin-induced anaphylaxis, in which complement activation by proteases in vespid toxins is likely to exacerbate disease caused by IgE antibodies to vespid toxin antigens⁵⁶. There is also considerable clinical evidence for the induction of anaphylaxis by agents that directly activate complement in the absence of agent-specific IgE or IgG Abs. This can be observed in association with hemodialysis (particularly during the first use of a new dialysis membrane); protamine neutralization of heparin, liposomal drug infusion; infusions of drugs that are dissolved or suspended in certain lipid vehicles, such as Cremophor EL, and polyethylene glycol infusion⁴⁹. Two limitations of these correlative studies are that: 1) complement might not be the only factor activated that could contribute to shock; activation of the contact/kinin system is just one alternative possibility); and 2) no human studies have been performed to try to prevent anaphylaxis in any of these situations with inhibitors of complement activation or anaphylatoxin receptors. Taken together with the mouse data, the most likely interpretation of the clinical studies is that acute complement activation can induce anaphylaxis, particularly when other factors are present (e.g.; FcεRI or FcγR crosslinking, pre-existing vasculopathies) that can add to or synergize with anaphylatoxin effects.

Ab- and complement-independent anaphylaxis

Several drugs have been associated with anaphylaxis in susceptible individuals in the absence of a direct Ig-mediated mechanism or complement activation. These drugs include over-sulfated heparin, aspirin and non-steroidal anti-inflammatory drugs (NSAIDs), antibiotics, including vancomycin and the fluoroquinilones, opiates, and drugs used in general anesthesia, particularly neuromuscular blocking agents^{39, 57-59}. Different mechanisms for anaphylaxis induction have been implicated for these different drugs in in vitro studies with human cells and plasma and in vivo animal studies. Over-sulfated heparin directly activates the kinin system, with increased production of bradykinin⁵⁹. NSAIDs, including aspirin, block cyclooxygenases, which are essential for prostaglandin production. This results in decreased levels of prostaglandin E2, which can suppress anaphylaxis, and increased levels of cysteinyl leukotrienes, which, among other effects, increase pulmonary smooth muscle contraction and vascular permeability³⁹. Aspirin, unlike other NSAIDs, has also been reported to increase FcεRI-mediated basophil activation by enhancing phosphorylation of the signaling molecule, Syk.⁵⁸ Vancomycin activates mast cells to release histamine and other mediators through a mechanism that is calcium-, phospholipase C- and phospholipase A2-dependent, but is otherwise unknown⁶⁰. Opiates also induce histamine release, through a mechanism that involves binding to central opioid receptors³⁹. Fluroquinolone antibiotics and nicotinic receptor antagonist non-steroidal neuromuscular blocking agents, such as tubocurare, which have a tetrahydroisoquinalone motif, directly activate mast cells by binding to MRGPRX2, a G-protein-coupled receptor⁵⁷. Taken together, these observations provide considerable reason to believe that the direct effects of these drugs on mast cells and basophils contribute to anaphylaxis. However, Ab/FcR interactions can also contribute to the ability of at least some of these drugs to induce anaphylaxis. With the exception that opioid-induced anaphylaxis has been reversed in humans by administration of opioid receptor antagonists^{39, 61}, the importance of direct mast cell activation for anaphylaxis induction by these drugs has not been proven in humans in vivo. For anaphylaxis associated with other drugs, such as iodinated radiological contrast media, the relative roles of IgE Abs, IgG Abs, complement, and direct effects on myeloid cells are still debated³⁹. It also remains to be determined whether any anaphylaxis-associated drugs cause disease solely through direct effects on mast cells or as part of a two- or multi-hit mechanism, and why some people are much more susceptible than others.

Conclusion

Although it is currently impossible to prove beyond a doubt that non-IgE-mediated anaphylaxis is clinically relevant, considerable evidence supports the occurrence and clinical importance of human IgE-independent anaphylaxis that is mediated by IgG, complement, or direct basophil and mast cell activation. Clinical situations that have been associated with the four different putative types of anaphylaxis are summarized in Table I, which also summarizes the cells, receptors and mediators that are thought to contribute to the pathogenesis of each form of this disorder. IgG-mediated anaphylaxis should be suspected when there are large infusions of Ag and high titers of IgG Ab specific for the infused Ag, while complement-mediated anaphylaxis and direct mast cell/basophil activation should be suspected in patients who have received drugs, biologicals, or excipients that are known to have the appropriate complement or mast cell/basophil-activating properties. Identifying the probable cause of a specific episode of anaphylaxis is likely to increase in importance as therapeutics that block a specific pathway become available; pretreatment with such therapeutics may also be useful as prophylaxis for patients who require specific anaphylaxis-associated drugs. Finally, it is important that a single episode of anaphylaxis may involve more than one mechanism. Simultaneous occurrence of IgG- and IgE-mediated anaphylaxis has been demonstrated in mice⁶², as has synergy between anaphylatoxin- and IgE-mediated anaphylaxis⁴⁷; IgG/Ag complexes that bind to stimulatory FcγRs may also initiate anaphylaxis by activating complement, and direct mast cell activation by drugs is likely to act additively or synergistically with Ab- or complement-dependent activation of these and other myeloid cells to increase anaphylaxis severity.

Table I.

Etiologic mechanisms of anaphylaxis and their distinguishing characteristics

Type	Inciting Agents	Cells	Receptors	Mediators
IgE-Mediated	Food allergy Insect sting allergy Drug allergy	Mast cells Basophils	FcεRI	Histamine PAF
IgG-Mediated	Biologicals Drugs Dextrans Aprotinin Transfusions	Macrophages Neutrophils Basophils	FcγRIII FcγRI FcγRIV (mouse) FcγRIIA (human)	PAF Histamine
Complement-Mediated	Lipid incipients Micellar drugs Liposome Other nanoparticles Polyethylene glycol Cellulose membranes	Macrophages Mast cells	C3aR C5aR	PAF Histamine
Direct Mast Cell Activation (mechanisms differ for  different agents)	NSAIDs, including aspirin Vancomycin Opiates Local Anesthetics Fluoroquinolone antibiotics Neuromuscular blockers Octreotide Leuprolide	Mast cells Other myeloid cells	MRGPRX2 Other receptors	Cysteinyl leukotrienes Histamine

Abbreviations

Ab

antibody

Ag

antigen

FcεRI

high affinity IgE receptor

FcεRIα

α polypeptide chain of FcεRI

FcγR

IgG receptor

mAb

monoclonal antibody

PAF

platelet activating factor

TNP

trinitrophenyl