The alpha-gal syndrome: Understanding the role of tick bites, and the delays in severe anaphylaxis

Abstract

The presence of serum immunoglobulin E (IgE) antibodies to the oligosaccharide galactose-alpha-1,3-galactose (α-gal) was first recognized during the investigation of allergic reactions to the monoclonal antibody cetuximab. The availability of an assay for IgE to α-gal made it possible to show that the patients who reported delayed allergic reactions to meat also had IgE to α-gal. This syndrome of delayed reactions to meat in patients with IgE to an oligosaccharide that was present in nonprimate mammals represented a complete contrast to the other forms of food allergy. However, the realization that the sensitization resulted from bites of the lone star tick, Ambloymma americanum, provided an explanation for the increase in cases in an area of the United States where there had been a major increase in the deer population, which is the primary breeding host of this tick. In contrast to almost all other IgE-mediated reactions, there is no relationship between the delay after eating red meat and the severity of allergic reactions. There is also increasing evidence that serum tryptase in the range of 20–90 ng/mL can occur after severe reactions related to α-gal. In addition, a recently reported fatal anaphylaxis, which started 4 hours after the individual ate a hamburger, included a postmortem tryptase of >2000 ng/mL. A working hypothesis to explain the delay in reactions relates to the time taken to process glycolipids in food derived from mammals, first into chylomicrons and, finally, to low-density lipoproteins. However, the quantitative presence of α-gal that remains on these particles may have major individual differences. In contrast to the patients who have symptoms related to eating meat, there are many individuals who are sensitized and who are not aware of any symptoms. In addition, individuals who are sensitized to alpha-gal can react rapidly to intravenous injections to the monoclonal antibody cetuximab, or polyclonal antibodies to venom, or the newer antivenom digested to Fab or Fab’2.

The early investigations and observations that led to the definition of the galactose-alpha–1,3-galactose (α-gal) syndrome (AGS) had to incorporate a range of features that did not seem to be connected, which included the following:

• IThe target of the immunoglobulin E (IgE) antibodies to cetuximab was an oligosaccharide, which was already known as a major transplantation antigen.^1,2
• IIThe reactions to food derived from mammals that carried this specific oligosaccharide were reported in most patients to be delayed for 3 to 5 hours.³
• IIIThat sensitization to the oligosaccharide followed bites from a limited number of tick species, but, in the United States, this was primarily lone star ticks (Amblyomma americanum), whose bites can give rise to prolonged itching.^4–6
• IVThat the severity of the reactions was not closely related to the length of the delay before symptoms started or to the titer of specific IgE (sIgE) antibodies.^7,8

We propose here to focus on the state of research in relation to two major areas:

What is the effect of tick bites, why are these effects not seen with some other species of ticks and also not seen with mites, e.g., scabies, or with insects, e.g., fire ants, bed bugs, or mosquitos?^9,10

Why can these reactions to eating meat, which are delayed for many hours, still be severe or very severe? If this is a function of the nature of the food eaten, how does that help to explain the factors that influence severity?¹⁰

THE HISTORY OF IgE TO OLIGOSACCHARIDES AND THE EVIDENCE THAT ALLERGIC REACTIONS TO THE MONOCLONAL ANTIBODY CETUXIMAB ARE DUE TO PREEXISTING IgE ANTIBODIES TO α-GAL

That human IgE antibodies can bind to oligosaccharides has been well known since the 1980s when Rob Aalberse, Ronald Van Ree, and their colleagues identified that many of the sera with IgE to plant proteins also had IgE antibodies that bound to oligosaccharides on the proteins.¹¹ However, these IgE antibodies specific for plant-derived antigens did not seem to be related to allergic symptoms. The relevant sugars, most commonly MUXF3 and MMXF3, are present on a wide range of plant proteins and were named cross-reactive carbohydrate determinants (Fig. 1).^11,12 In some tropical countries, including Nigeria, Zimbabwe, and Honduras, IgE antibodies to carbohydrate determinants can cause major confusion by creating false-positive IgE results with plant-derived allergen extracts.^13–16

Figure 1.

Oligosaccharides that can be targets of naturally occurring human immunoglobulin E (IgE) antibodies. (A) The classic cross-reactive carbohydrate determinants (CCD) on asparagine (Asn) in plant proteins: MMXF3 on horseradish peroxidase (HRP) and MUXF3 on bromelain from pineapple; (B) Galactose-alpha-1,3-galactose (α-gal) is a major glycosylation of proteins and lipids in nonprimate mammals, which cannot be produced in primates; and (C) α-Gal is a major glycosylation of lipids in nonprimate mammals and is commonly O-linked to ceramide.

The first evidence that the IgE antibodies to α-gal were common in a large area of the United States came from the investigation of rapid allergic reactions to cetuximab.^1,17 Those studies involved a close collaboration with ImClone (Branchburg, NJ), the company that developed the monoclonal antibody that is specific for the epidermal growth factor receptor. That company provided the recombinant molecule that is used to treat cancer and is produced in the mouse cell line SP20. They also provided this antibody with the same amino acid sequence produced in a Chinese hamster ovary cell line. The antibody produced in SP20 was shown to bind to sIgE antibodies found in human serum effectively, whereas “the same” molecule produced in Chinese hamster ovary cells did not bind IgE antibodies. That finding provided the best evidence that the target of the IgE antibodies on cetuximab was post-translational.¹ In addition, it was Dr. Zhou and his colleagues working at ImClone who published the full glycosylation of cetuximab in 2007.¹⁸ However, it was Drs. Chung, Berlin, and their colleagues at Vanderbilt who provided the pretreatment sera from patients treated with cetuximab as well as closely matched controls, which made it possible to prove that these IgE antibodies were common in central Tennessee and were responsible for the reactions that occurred during the first treatments with cetuximab.¹

Once we had identified the specificity of the IgE responses to cetuximab, it was clear that this glycoprotein could be bound to a Streptavidin ImmunoCAP (Phadia/Thermo Fisher, Kalamazoo Michigan) and provided a reliable assay for IgE to α-gal. This assay was used for the next 10 years and was finally replaced by the U.S. Food and Drug Administration approved assay that uses beef thyroglobulin on the ImmunoCAP (Fig. 2).¹⁹

Figure 2.

Immunoglobulin E (IgE) antibodies specific for galactose-alpha-1,3-galactose (α-gal) can be measured by using either cetuximab or beef thyroglobulin attached to the ImmunoCAP. The initial assay was made by Phadia in 2023 by using cetuximab. Subsequently, the assay uses beef thyroglobulin, o215, and was approved by the U.S. Food and Drug Administration (FDA) in 2018. The assay results shown here used a cetuximab cap created by Phadia in 2003. The results show that the assay with cetuximab on the ImmunoCAP that had been stored in our laboratory for 20 years, correlated closely with the current o215 assay and gave a Spearman rank coefficient of R² = 0.93, p < 0.001.

THE ROLE OF TICK SALIVA IN ESTABLISHING sIgE RESPONSES TO α-GAL

Investigations of the immune response induced or altered by tick bites have involved the following:

• iStudies on antibodies and cells in the peripheral blood of patients with AGS.
• iiStudies on the immune response induced by injections of tick extracts into the skin of mice
• iiiDetailed studies on the cellular response in human skin after a tick bite.

The first thing to remember is that all humans have some and usually large quantities of IgG, IgA, and IgM antibodies to α-gal before they receive tick bites.² Early on, we considered whether the tick bites could establish an IgE response unrelated to the preexisting “natural” response in the intestine.² However, collective recent results make it much more likely that the primary effect of tick saliva is to produce an isotype switch to IgE in plasma blasts specific for IgG to α-gal. In this model, the fact that the response is dominantly to α-gal could be considered to be inevitable because that is the best recognized natural antibody response to a nonhuman oligosaccharide. The next question was whether some patients make IgE to tick bite–related protein antigens before they induce the switch of the α-gal response to IgE. In this respect, there is evidence that IgE responses occur against tick proteins, e.g., vitellogenin, which are glycosylated and also to other proteins, e.g., serpins, that are not glycosylated.²⁰ Apostolovic, Van Hage, and their colleagues in Stockholm also provided good evidence that the IgE response to tick saliva includes IgE antibodies to both proteins and α-gal.^20,21 Crispell et al.²² provided important data about the contents of saliva during feeding in a variety of tick species. Experiments with α-gal Knock Out mice by several groups established that extracts made from ticks that were actively feeding or to extracts with added α-gal to the extract, can induce high titers of sIgE to α-gal.²³ In some of these studies, it was also demonstrated that the sensitized mice can get allergic symptoms when challenged.^24,25

Finally, there are interesting data about the cellular basis of the skin reaction to α-gal. These data may help to explain why the lone star tick gives such profound and prolonged itching, whereas bites from many other ticks, particularly Ixodes scapularis, which carries Borrelia burgdorferi, in most patients do not give rise to local itching (Fig. 3).⁶ An important question in this regard relates to the relevance of basophils. The original work by Philip Askenase and others on basophils at the site of a tick bite was designed to understand acquired tick resistance and was primarily studied in guinea pigs.^26,27 Those results made a strong case that the basophils can play a role in controlling or killing ticks.^27,28 Subsequently, a research group in Japan documented the increasing prevalence of basophils in the skin of patients who had repeated bites from Amblyomma testudinarium.²⁹ The presence of both basophils and eosinophils in the skin has interesting parallels with the effect of a dust-mite patch test on the skin of patients with atopic dermatitis.³⁰ What matters here is that basophils, in addition to producing interleukin 4, are also known to produce interleukin 31, which is an important mediator of itch in the skin.^31,32 Thus, we would propose that the basophils in the skin can play a significant role in the local pruritus that is recognized as having a strong correlation with the production of sIgE to α-gal.⁶

Figure 3.

Confocal microscopy of the ventral side of the mouth parts of two different ticks: (A) Ambloymma americanum, the lone star tick, and (B) Ixodes scapularis, the deer tick. For each tick, the pigment was bleached with peroxide, cuticle stained with Congo red and calcofluor white, and imaged on Zeiss LSM (Zeiss, Oberkochen, Germany) at ×100 magnification. Images are courtesy of Igor Siwanowicz, Janelia Research Campus of the Howard Hughes Medical Institute, Ashburn, Virginia.

TICK BITES, IgE TO α-GAL, AND THE CLINICAL SYNDROME

With the assay for sIgE to α-gal available through our collaboration with Phadia/Thermo Fisher (Kalamazoo, Michigan), it rapidly became clear that the patients who reported delayed reactions to red meat also had positive IgE assays for this specific epitope.^1,3 In addition, it became clear that there were cases in Fayetteville, AK, Tulsa, Oklahoma, Springfield, Missouri, Memphis, Tennessee, and Durham, North Carolina, as well as in several clinics in central Virginia.³³ This geographic range was similar to the Centers for Disease Control and Prevention map of Rocky Mountain spotted fever and also to the known distribution of reactions to cetuximab. These observations led to questions about ticks, and the subsequent results of asking patients in Virginia about tick bites were astonishing.⁶ However, from personal experience and from hundreds of patients, it became clear that the patients were also being bitten by small, black or brown arthropods that gave rise to intense pruritis but were too small to be identified with the naked eye. These animals were most often described as “chiggers” and could be present in the hundreds. The ones that Dr. Platts-Mills had on his legs in August 2007 were sent to Dr. Robert Lane at Cal-Berkeley, who identified them definitively as larval lone star ticks.³⁴ This, in turn, led to the definition of AGS in the United States, which is a condition in which bites from larval or more mature lone star ticks give rise to sIgE antibodies to α-gal, which could cause immediate reactions to drugs such as cetuximab or Zostavax as well as delayed allergic reactions to food derived from nonprimate mammals.^7,33 By 2011, we were hearing reports of similar findings from several other countries, in particular, Sweden, Australia, Japan, Germany, and France.^5,8,9,35

There were two things that had to come together to establish the AGS. First, the specificity of the antibodies to cetuximab that were responsible for the infusion reactions, and, second, the evidence that saliva from certain ticks could induce these IgE antibodies.^5,22,29 However, once it was clear that the target of the IgE antibodies was α-gal, which is a major blood group substance of the nonprimate mammals, it made sense of the increasingly well-recognized fact that the patients were reporting reactions to meat, dairy, or organs derived from a wide variety of nonprimate mammals. The identity of the oligosaccharide α-gal leads to another area of research, because Uri Galili in Chicago and others had established that this antigen was the major transplantation barrier between pigs and humans.^2,36 Indeed, α-gal knockout mice and pigs had been established before we identified IgE to α-gal. They had already confirmed that all primates had IgG and IgM antibodies specific for α-gal, which were responsible for the hyperacute rejection of pig kidneys, which occurred when they were transplanted into baboons.³⁷ It is important to remember that α-gal is the “B like antigen” that Landsteiner³⁸ reported in 1928. Thinking about this focused us on the question: why do all humans have significant or large quantities of sIgG antibodies to α-gal, but that, at least in nontropical countries, sIgE only becomes part of that response when we are bitten by ticks. In the United States, two species of ticks have been associated with large numbers of cases of patients with sensitization to α gal. They are Ambloymma americanum (the lone star tick) and Dermacentor variabilis (the dog or wood tick). By far the largest area in the United States where cases of patients with AGS are common relates to the lone star tick.³³ This tick is certainly moving north in New England,³⁹ but there is also an area of northern Minnesota from the North Dakota border across to Duluth where cases of patients with AGS are present but the lone star tick is not present.^33,40 In this area, the dominant tick is called the “wood tick,” which is physically very similar to the tick called a “dog tick” in the South. There have been case reports of patients who may have been sensitized to α-gal by Ixodes pacificus or I. scapularis but those cases are so rare that they are still reportable.^41,42

In 2015 when we started interviewing allergists in practice about the number of patients with AGS they had seen, we found very strong confirmation of the main area where cases were common. However, in addition, we found an astonishing lack of cases in Texas and also in several areas along the Gulf coast where the lone star tick had been reported. This did not make sense at first, however, in 2018, at the Tick Summit in Maryland, Colonel Ting who had been stationed in San Antonio told us that the ticks had been killed by fire ants. In fact, there is extensive evidence from the 1960s that fire ants can either kill or disrupt the breeding of the ticks. Indeed, there is good evidence for this both in southern Alabama and Texas.^43,44 In a second phase of our interviews, we established that there was a highly significant inverse correlation between patients with AGS and the number of patients with fire ant anaphylaxis seen in the clinics of the same allergists.³³ The implication is that the major area of AGS, which is known to be spreading north may also be progressively restricted from the South by fire ants.

Additional Forms of Sensitization That Can Masquerade as AGS or Idiopathic Anaphylaxis

In addition to being able to identify causes in patients with IgE to α-gal, the assay has helped us to identify other causal IgE responses in patients previously classified as having idiopathic anaphylaxis. Thus, the ability to exclude AGS among patients with an IgE to α-gal of <0.1 IU/mL allowed us to focus on some interesting cases. First is of patients who described reactions, sometimes delayed, to pork, who were found to have IgE to cat serum albumin, which cross-reacts with pork serum albumin (Sus s 1). One of these patients actually had a titer of IgE to cat serum albumin of 100 IU/mL and she had the pork-cat syndrome.^45,46 Secondly we investigated a patient who had a severe episode of anaphylaxis shortly after receiving the old Zostavax vaccine that included significant gelatin. Initially this was assumed to be due to AGS. However, her IgE to α-gal was <0.1 IU/mL and her sIgE level to pork gelatin was 63 IU/mL.^47,48 The case we were asked to investigate of a woman who had prolonged anaphylactic reactions after consuming milk. In this case, assays on her IgE to milk proteins showed an IgE level to bovine serum albumin of 48 IU/mL and, again, her IgE level to α-gal was < 0.1 IU/mL (Table 1).

Table 1.

The oligosaccharide α-gal and mammalian proteins that can masquerade as idiopathic anaphylaxis

α-Gal = Galactose-alpha-1,3-galactose; FDA = U.S. Food and Drug Administration; IgE = immunoglobulin E; sIgE = specific IgE.

^*The initial report of an anaphylactic reaction of this kind in Paris was fatal.

^#Zostervax has been replaced by Shingrix (GlaxoSmithKline Biologicals, Rixensart, Belgium) in the United States but is available in some countries.

The important questions related to the antibody response to α-gal are the following:

• iWhy does the natural antibody response never include sIgE antibodies to α-gal?
• iiDoes the response to ticks start with IgE antibodies to tick-derived proteins before the switch of the natural IgG response to α-gal to produce IgE to α-gal⁴⁹?
• iiiDoes an effective sIgE response to tick bites require both tick-derived proteins that allow prolonged attachment of the tick and changes in gene expression, such that the salivary glands of the tick produce α-gal.²³

Results of recent work by Belkaid et al.⁵⁰ on the mechanism of the T-cell response in the gut of a mouse suggest that some aspects of the response are driven by products of the bacteria in the gut, which include the presence of unmethylated DNA on bacterial cell walls.⁵¹ Strikingly, unmethylated DNA is an important element of the traditional complete Freund’s adjuvant. That adjuvant was shown in the 1970s to suppress IgE production in both mice and rabbits.⁵² Thus, it may not be surprising that the “natural” antibody response to α-gal, which we assume is produced in the intestines, does not include IgE.^2,36,37 It seems reasonable, and has been widely observed, that many or most patients with AGS lose their IgE progressively if they have no further tick bites.⁵³ Despite this, there is no current evidence that they lose their preexisting “natural” IgG and IgM responses.

The Possible Explanation for the Delay between Eating Meat and the Beginning of Symptoms in Subjects with IgE to α-Gal

The delay between eating red meat and symptoms among individuals who had first experienced tick bites was described specifically for her family and a group of friends by Sandra Latimer in Athens, Georgia, as early as 1989 (Deutch A. Latimer S. Personal communication from Dr. Anthony Deutch and Mrs. Sandra Latimer of Athens GA.2208). The same phenomenon was reported to the Allergy Society in Sydney, Australia, by Van Nunen et al.⁴ in 2007 and published in 2009. In addition, we had seen several patients in central Virginia before 2006 who reported delayed reactions to red meat. However, this phenomenon was such a contrast from previous models of food allergy that you could only believe it after hearing the story from “at least four” individuals. However, none of those observations could have led to the identification of the specificity of the immune response that was responsible for the reactions. Needless to say, almost all of the early patients who reported reactions to red meat were ignored by both their family physicians and the allergists who saw them. This was inevitable because the possibility that a patient could become allergic to meat after many years of complete tolerance was confusing enough. Before the assay was available, the possibility that what seemed to be a form of IgE-mediated food allergy could start after a delay as long as 4 hours was “a step too far.”

Once the assay for sIgE to α-gal was established, and we had seen at least 20 cases, we had to face the reality that the allergic reactions to meat were, in most patients, taking between 3 and 5 hours to start. In addition, it became clear, both in Virginia and Sweden, that the severity of the reactions was not related to the length of delay and only vaguely related to the titer of IgE to α-gal.^54,55 The only explanation that made any sense in trying to understand the delay was the processing of glycolipids (Fig. 4).⁴⁹ In this respect, the most important study was reported by Roman-Carrasco together with Dr. Swoboda in Vienna.⁵⁶ That study used an in vitro model of small intestine absorption, and they showed convincingly that glycolipid molecules can transfer across the membrane and form lipid particles that can activate basophils from subjects with IgE to α-gal. By contrast, glycoprotein molecules were not transferred in an active form.⁵⁶ This observation fits well with the view that α-gal attached to a lipid in the form of chylomicrons enters the circulation via the thoracic duct ∼1 hour after being eaten. However, chylomicrons are generally in a size range between 1 μm and 400 nm. This size is far too large to escape from the healthy endothelium in the circulation and, actually, they are relatively small in number. The breakdown of these particles takes 2 to 3 hours, by which time they are in the form of low-density lipoproteins (LDL). By 4 hours after eating a steak or a hamburger, LDL particles are ∼20 nm in diameter, which is small enough to escape the circulation and to activate mast cells if the LDLs carry α-gal and the mast cells are sensitized with IgE to α-gal. Thus, we propose that LDLs could activate both basophils and mast cells, which gives rise to hives, angioedema, or more severe reactions. In addition, if LDLs activate mast cells, then that could increase serum tryptase levels (Fig. 4). In our initial challenge studies on patients with α-gal and controls, activation of basophils in vivo was seen in both patients and a small number of the controls. We think that the activation of basophils reflects both the presence of sIgE on the basophils and the presence of LDL receptors on basophils. By contrast, none of the controls had symptoms, a skin response, or an increase in serum tryptase.⁵⁵

Figure 4.

The time taken for glycolipids in the diet to be absorbed through the intestine wall and reassembled in lacteals to make chylomicrons, which pass through the thoracic duct into the subclavial vein is 1 hour. In the circulation, the process of creating smaller (very low-density lipoproteins [VLDL]) and smaller (low-density lipoproteins [LDL]) lipid particles takes several hours. As LDL, these particles can leave blood vessels because they are as small as 20 nm in diameter. This is assuming that the mast cells carry specific immunoglobulin E (sIgE) to galactose-alpha-1,3-galactose (α-gal) and that the LDLs carry α-gal. These processes could explain why allergic reactions occur 3–5 hours after eating meat.

Possible Cardiac Complications of Chronic Exposure to Mammalian Food

Initially two things lead us to consider a possible role of sIgE to α-gal in cardiovascular disease. First, we became aware that a surprising number of the adult male patients with AGS reported having heart attacks; second, there are reports of reactions to cardiac valves derived from mammals, which included rapid deterioration of the valves and antibody responses to α-gal in individuals who had mammalian valves.^57,58 Subsequently, the hypothesis that LDLs could enter into the walls of coronary arteries and trigger mast cells provided a plausible explanation about how mammalian food could initiate or enhance atherosclerosis.^57,58 In 2018, we started collaborating with Dr. McNamara and Dr. Taylor who had been investigating possible cases of Coronary Artery Disease (CAD) by using intravascular ultrasound. We assayed 118 sera from patients who had intravascular ultrasound results for sIgE to α-gal. The results showed a significant association between sIgE to α-gal and atherosclerosis that was strongest among patients ages < 55 years and was particularly associated with the unstable forms of atherosclerosis.⁵⁹ Strikingly, a major cardiology group in Sydney, Australia, confirmed our findings in a larger cohort being evaluated for unexpected causes of CAD. They used cardiac computed tomography monitoring to evaluate the extent of CAD, but, in addition to finding a positive association with the atherosclerosis score, they also found a positive association between sIgE to α-gal and ST Elevated Myocardial Infarction.⁶⁰ One of the factors that makes interpretation of these results difficult is that diets based on mammalian products, including both dairy and meat, have long been thought to be a risk factor for CAD and that all humans have IgG, IgA, and/or IgM antibodies to α-gal, thus, it is quite possible that antibodies to α-gal other than sIgE can contribute to a CAD risk. In fact, an earlier article from the United Kingdom reported an association between CAD and IgG antibodies to α-gal.⁶¹ Analysis of our preliminary data also suggests that sIgG3 to α-gal could also contribute.⁶² More recently, we collaborated with Corinne Keet and published data that show that sIgE to cow’s milk is associated with coronary disease mortality.⁶³ Overall, there are good reasons for continuing to study the relevance of sIgE to α-gal to CAD; however, the hypothesis should be that immune responses to antigens in both meat and cow’s milk can contribute to the cardiac risk associated with a diet rich in mammalian foods.

Severity of Allergic Reactions in Patients with AGS

In the early investigation of AGS, we were aware of patients who reported cases with anaphylaxis severe enough to require treatment in the emergency department and even occasionally requiring admission to the hospital. However, the severity of these reactions was not related to the delay after eating meat or to the titer of sIgE to α-gal.⁷ Indeed, the highest level of sIgE that we saw during the first 4 years of the assay was 400 IU/mL; this patient, who lived in Little Rock, Arkansas, by report of the local allergist, experienced only mild or moderate episodes of hives, which started 3–4 hours after eating meat. Interestingly, over the past 10 years, we have had a few opportunities to measure serum tryptase in patients admitted to the emergency department and admitted to the hospital. Of three cases of patients who spent time in the Medical Intensive Care Unit (MICU), one was admitted overnight and the other two for longer, and they had tryptase levels of 30 ng/mL, 34 ng/mL, and 90 ng/mL, respectively. In each of these patients, the tryptase levels were normal once they returned to baseline. It is important to remember that the emergency department’s definition of anaphylaxis is not the same as an allergist’s definition. This fact has two dimensions: first, most emergency departments will not recognize the diagnosis of anaphylaxis unless the patient is in shock, which, in general, requires a systolic blood pressure < 80 mm Hg; and, second, once the patient is in shock, observations about skin findings, such as hives or angioedema, assume low priority.

The bulk of the clinical issues with AGS relate to delayed allergic reactions after eating mammalian products. However, this form of sensitization was first identified during the investigation of allergic reactions that occurred during the first infusion of cetuximab in patients with cancer. These reactions were rapid and often severe. Indeed, we are aware of the details of three deaths that occurred rapidly after starting the first infusion. Two of the deaths occurred before we had published the New England Journal of Medicine article.¹ By contrast, the third case occurred in a patient in Charlottesville 4 years after that publication.¹ The patient collapsed rapidly but was admitted to a local intensive care unit, and we received his serum, which showed sIgE to α-gal of 3.5 IU/mL. Unfortunately, his condition was very severe when he was admitted and he did not survive. Recently, we have become aware of another drug that can cause severe reactions in patients who have preexisting sIgE to α-gal. This is one of the partially digested forms of polyvalent antibodies to snake venom. The drug, Anavip (Rare Disease Therapeutics, Inc, Franklin Tennessee), is derived from horse antiserum to snake venom and is made into Fab′2 fragments by digestion with pepsin.⁶⁴ Included in these results were the results of basophil activation tests, which compared the effects of Anavip with those for CroFab (SERB Pharmaceuticals, West Conshohocken, PA), the alternative antivenom, which is an FAB fragment of sheep antiserum. These results confirmed that Anavip was at least 10-fold more potent in basophil activation tests than CroFab. Subsequent to the publication about the digested antivenoms, there was a report of severe anaphylaxis that occurred with Anavip in North Carolina. In this patient, the blood pressure fell over 30 minutes to 54/23 mm Hg, his sIgE to α-gal was 3.2 IU/mL, and his serum tryptase level was 20 ng/mL, and he recovered fully after transfer to the emergency department at Duke for treatment with a different antivenom.⁶⁵

A second development that may be relevant to severity is that a significant number of patients with AGS present to gastroenterologists with symptoms similar to irritable bowel syndrome.⁶⁶ Other patients present to the emergency department with severe abdominal pain and are subsequently diagnosed with AGS.¹⁰ In February 2025, we became aware of a death that had occurred in New Jersey in September 2024, which occurred 4 hours after the patient ate a hamburger.⁶⁷ In this patient, there was a history of severe abdominal pain that had occurred 3 weeks earlier. After that episode, despite the severity of the pain, the patient did not consider that the episode of pain could be anaphylaxis and did not present to a medical facility. Both the fatal event and the earlier abdominal pain had started 4 hours after eating beef; however, the postmortem of the 47-year-old man concluded that there was “no apparent cause of death.”⁶⁷ This left his wife in an impossible situation in which her healthy, young husband and father of their children had died “for no reason.” She contacted a friend, Dr. Erin McFeely, and together they contacted our group in Virginia to inquire if this could be a case of AGS. We received the postmortem report and together we arranged to obtain the postmortem serum. Although his IgE to α-gal was low, i.e., 0.57 IU/mL, this was 3.5% of the total IgE level, and the serum level was completely negative, i.e., <0.1 IU/mL, for a wide range of IgE antibodies to other allergens that could have contributed to his death.⁶⁷ The value obtained for his postmortem tryptase, after dilution of the serum, was >2000 ng/mL, a value that is diagnostic of fatal anaphylaxis.^68–70

Subsequent to the report of the death in New Jersey, we became aware of two other patients who had died within several hours after eating meat. In each case, the diagnosis of AGS was already known; accordingly, the medical examiners sent postmortem blood for tryptase assays and the results in each case where >1,000 ng/mL. In our current view, the postmortem results make it unlikely that mastocytosis was the underlying cause of the severity of the anaphylaxis because none of the three had symptoms suggestive of that disease before their death.

The implication that elevations of serum tryptase are a feature of delayed severe reactions to meat or other constituents derived from cows, pigs, or sheep is further supported by the results of a case included in our initial report on challenges with beef.⁵⁵ One of the patients we challenged had no symptoms for 3.5 hours, at that time her skin flushed and she started to develop abdominal symptoms. The abdominal pain became severe and continued for >4 hours. In addition, at the 3-hour point, the basophils were activated as judged by CD63 staining ex vivo and her serum tryptase had risen to 20 ng/mL from a baseline of 4 ng/mL.⁵⁵

Summary with Regard to Severity.

We have seen significant elevations of tryptase in relation to AGS either in naturally occurring severe episodes after consumption of mammalian products, or in postmortem sera. In addition, both the challenge test and the fatal episode involved severe abdominal pain that started 3 or 4 hours after eating beef. Clearly, the tryptase results are dramatically different from the results seen with peanut-related reactions. Importantly, all our results confirm that the delayed reactions can be severe or very severe. At this point, we think that there are good reasons for including severe abdominal pain as an important form of anaphylaxis.

CONCLUSION

The awareness of IgE antibodies specific for the α-gal present in sera of cancer patients and controls came from the investigation of reactions to the monoclonal antibody cetuximab. This led to the availability of the assay for these antibodies and the recognition that some patients with otherwise idiopathic anaphylaxis also had the same IgE antibody.¹ This, in turn, led to the recognition that almost all the patients who reported clinical reactions 3–5 hours after eating meat or organs from nonprimate mammals had serum IgE to α-gal.

The possibility of tick bites as a cause arose because none of the well-known causes of IgE responses made sense. However, it was the map of where the allergic reactions were occurring that led to a detailed investigation of the relevance of A. americanum, the lone star tick. This, in turn, focused attention on the massive rise in the population of white-tailed deer, which is directly responsible for the increase in tick bites since 1950 but has been particularly dramatic since 1980. The deer population of New Jersey has recently been described as an “unsustainable statewide emergency.” The lone star ticks breed on these deer and all forms of this tick, i.e., the larvae (which are commonly known as “seed ticks” or “chiggers”), nymphs, and adults are known to bite humans.

Finally, the severe cases of anaphylaxis in patients with AGS are consistently associated with increased tryptase levels, up to 30 ng/mL or even 90 ng/mL. In addition, the three postmortem tryptase assays that we are aware of from patients who died of delayed anaphylaxis after eating beef had values of tryptase >1000 ng/mL. These are values that have not been associated with severe or fatal reactions to peanut. Thus, we now have a model of anaphylaxis in individuals who remained asymptomatic for 3–5 hours after eating a hamburger and can then have very severe or catastrophic anaphylaxis characterized by very high postmortem tryptase.⁶⁷ This adds yet another feature of AGS that is completely unlike the patients with severe and fatal anaphylaxis related to protein allergens such as tree nut, peanut, milk, or egg.