Gastrointestinal-Isolated Distress is Common in Alpha-Gal Allergic Patients on Mammalian Meat Challenge

Abstract

BACKGROUND AND AIMS:

Alpha-gal allergy causes a delayed reaction to mammalian meats and has been reported worldwide. Patients with the allergy may present with isolated gastrointestinal (GI) symptoms, but this phenotype is poorly understood.

METHODS:

We pooled and analyzed symptoms and demographics of patients from two prospective cohorts of patients with diagnosis of alpha-gal allergy who reacted after eating mammalian meat under observation. We compared characteristics of patients who demonstrated GI-isolated symptoms on challenge with those who exhibited symptoms outside the GI tract (skin, respiratory, circulatory).

RESULTS:

Among the 91 children and adult alpha-gal allergic patients who exhibited symptoms after oral challenge with mammalian meat, 72.5% experienced GI distress with one or more GI symptoms, which was the most frequent class of symptoms, compared to skin changes in 57.1% and respiratory distress in 5.5%. The most common GI symptoms were abdominal pain (71%) and vomiting (22.0%). GI-isolated symptoms occurred in 37 patients (40.7%) who reacted, and those patients reacted more quickly than patients who exhibited systemic symptoms (median onset of symptoms in GI-isolated group 90 minutes vs. 120 minutes) and were more likely to be children than adults (RR=1.94, 95% CI 1.04–3.63).

CONCLUSIONS:

Isolated GI distress occurred in 4 in every 10 alpha-gal allergic individuals who developed symptoms on oral food challenge with mammalian meat. Alpha-gal allergic patients, particularly children, may exhibit GI distress alone, and adult and pediatric gastroenterologists should be aware of the diagnosis and management of the allergy.

INTRODUCTION

Abdominal pain, vomiting and diarrhea, often without clear organic etiology, are common gastrointestinal (GI) symptoms that account for substantial suffering and healthcare expenditures.¹ Limited research has suggested a causal link of allergies to abdominal pain conditions such as irritable bowel syndrome.^2–4 Alpha-gal allergy is characterized by a reaction to beef, pork and other mammalian meat and derived products, typically hours after ingestion, and is diagnosed by symptoms, elevated serum alpha-gal Immunoglobulin(IG)E antibodies, and response to diet changes.^5–10 Gastrointestinal (GI) symptoms without rash and angioedema have been reported among GI clinic patients with elevated alpha-gal IgE antibodies.^{8, 9} GI-isolated alpha-gal allergy, a term that describes alpha-gal allergy that manifests with GI symptoms alone and no skin or cardiopulmonary symptoms, is poorly understood.

We aimed to better understand GI-isolated allergic reactions to alpha-gal by pooling two cohorts of alpha-gal allergic patients who reacted to mammalian meat ingestion in prospective, monitored food challenges and characterizing those with isolated GI reactions.^{5, 6}

METHODS

Mammalian Meat Challenges

We pooled results of mammalian meat oral challenges conducted by allergists in two separate cohorts of patients with alpha-gal allergy,^{5, 6} and described GI symptoms on challenge. We compared patients who developed isolated GI symptoms vs. patients who developed other symptoms (skin, respiratory, or cardiovascular) with or without GI symptoms.

The challenges took place in Eastern Cape, South Africa⁵ in 2017 and Charlottesville, Virginia,⁶ and Durham, North Carolina in 2013.⁶ Participants with oral challenge-proven alpha-gal allergy had elevated alpha-gal IgE titers, a history of reactions to mammalian meat, and reaction on food challenge. Regarding inclusion/exclusion criteria, the African site included patients with history of skin or GI allergic symptoms, whereas the U.S. site required a history of urticaria. Both sites excluded patients with uncontrolled chronic medical conditions.

In the studies, patients ate either beef or pork sausage and were monitored for several hours by coordinators and physicians who documented all symptoms over an 8-hour period. Participants gave baseline demographic information and were able to drink water during the challenge, but not exercise or consume other foods until at least 5 hours into the challenge. The sausages contained salt, spices, and sugar, but no garlic, peppers, onions or other FODMAPs. No participants apart from one South African participant endorsed a chronic GI illness (the participant had a history of gastritis). No South African participants had other food allergies, while food allergies in the Virginia group were not elicited, apart from allergy to dairy, which none of the subjects reported.

In both studies, abdominal pain was only counted as a positive reaction if it was severe and persistent. In the South Africa study, this was defined as subjective complaints of pain combined with abnormal stillness or doubling over that persisted for over 3 minutes. Alpha-gal IgE levels were measured using commercially-available ImmunoCAP systems (Thermo Fisher in South Africa and Phadia US, Portage, Mich in the US). Prior studies have found correlation coefficients >0.95 among different assays for alpha-gal. Severe reactions were defined as those that caused respiratory compromise (stridor, wheezing, inability to speak, dysphonia) or circulatory compromise (hypotension or syncope).

Statistical Analyses

We compared characteristics of patients with GI-isolated symptoms and those who had systemic symptoms by using chi-square/Fisher’s exact test and Mann-Whitney rank sum test for categorical variables and numerical variables, respectively. As GI-isolated reactions were common, we calculated the relative risk of having a GI-isolated reaction for children compared to adults using Poisson regression with robust error variance. We described results using proportions for categorical variables and medians with ranges and interquartile ranges for numerical values.

For multivariable analysis we constructed a logistic regression model with the outcome of isolated GI symptoms versus systemic symptoms. We performed univariate analysis as well as a forward selection analysis, where the variable with the lowest Akaike information criterion was included first and subsequent variables were added at a retention threshold of p < 0.05.¹¹ For the univariate analysis, we grouped ages into child (<18), young adult (18–35) and older adult >35 as those are roughly typical groupings for analysis. We received data in Microsoft excel and exported for statistical analysis using STATA Version 12.1 (College Station, TX). A two-sided P-value of less than or equal to 0.05 was considered to be statistically significant.

RESULTS

A total of 126 patients were challenged in the two studies, 114 Black Africans in South Africa, and 12 White Americans in Virginia/North Carolina. Among the 91 challenge-proven alpha-gal allergic patients, two-thirds of patients were female (63.7%) and the median alpha-gal IgE titer was 12.6 U/L (range 0.7–344.5, IQR 4.48–29.1). This was a predominantly pediatric cohort, with 61.5% of patients (n=56) with an age under 18, and an age range of adults from 18 to 65. We display patient characteristics by site in Table 1.

Table 1.

Patient characteristics by site.

Characteristic	South Africa Site (n=81)	U.S. Site (n=10)
Gender, n(%)
Female	55 (68%)	3 (30%)
Male	26 (32%)	7 (70%)
Age, Median  range; IQR	12 4–65 (8–26)	37 19–54 (26–51)
GI-Isolated Symptoms (%)	36 (44%)	1 (10%)
GI & Skin Symptoms, not severe (%)	24 (30%)	1 (10%)
Skin Symptoms alone	17 (21%)	6 (60%)
Severe Reactions	4 (5%)	2 (20%)
Age< 18 (%)	56 (69%)	0
Alpha-gal IgE U/L, median  range; IQR	10.6 0.7–345 (4–29)	24 9–32 (13–27)
History of Tick Bite	9 (11%)	10 (100%)

IQR, Interquartile ratio; U/L Units per liter

Reaction types are noted in the study flowsheet in Figure 1. GI symptoms were the most common type of reaction, and 72.5% (n=66) of the 91 alpha-gal allergic patients on challenge experienced at least one GI symptom. GI-isolated symptoms occurred in 40.7% (n=37). Patients who exhibited both skin and GI symptoms with no severe reaction were 27.5% (n=25), skin-isolated symptoms such as hives, urticaria and angioedema without severe reactions occurred in 25.2% (n=23) and severe reactions, all accompanied by GI and/or skin reactions, occurred in 6.6% (n=6) of patients. Figure 2 shows skin reactions among patients. All patients with circulatory or respiratory distress received injectable epinephrine, and all responded.

Figure 1.

Study flowsheet.

Figure 2.

Representative urticaria appearing in a patient in South Africa (A) and Virginia (B).

All patients with GI-isolated reactions experienced abdominal pain, whereas 40.5% (n=15) of them vomited, 18.9% (n=7) developed diarrhea and 24.3% (n=9) experienced nausea. The frequencies of individual GI symptoms (abdominal pain, vomiting, diarrhea) were similar among those with GI-isolated reactions compared to those who had both GI and skin or severe symptoms.

Characteristics of patients with GI-isolated symptoms vs. patients with systemic symptoms (skin, respiratory, circulatory) are presented in table 2. Patients with GI-isolated symptoms reacted significantly faster than patients with systemic symptoms (median time to reaction 90 minutes for GI-isolated group vs. 120 minutes in systemic group, p=0.03). Children <age 18 were about twice as likely as adults to experience GI-isolated reactions (RR=1.94, 95% CI 1.04–3.63, absolute risk 75.7%). There was a trend toward median alpha-gal IgE being higher in the GI-isolated group (18.7 U/L in GI-isolated group vs 10.5 in the systemic group, p=0.23). The distribution of sex and self-reported history of tick bites were also similar in the GI-isolated group and those with systemic symptoms.

Table 2.

Characteristics of patients with GI-isolated symptoms and those with systemic symptoms on oral food challenge.

Characteristic	GI-Isolated Symptoms (n=37)	Systemic Symptoms (n=54)	P Value
Gender, n(%)
Female	25 (67.6%)	33 (61.1%)	0.66*
Male	12 (32.4%)	21 (38.9%)
Age, Median range; IQR	11 (5–57; 8–13)	17 (4–65; 9–38)	0.03**
Age< 18 (%)	28 (75.7%)	28 (51.8%)	0.03*
Onset of symptoms in minutes, median (range; IQR)	90 (45–330; 75–120)	120 (45–375; 100–185)	<0.003**
Alpha-gal IgE U/L, median (range; IQR)	18.7 (0.7–74.7; 4–35)	10.45 (0.9–344.5; 4.5–25)	0.23**
History of Tick Bite	7 (18.9%)	12 (22.2%)	0.8*

IQR, Interquartile ratio; U/L Units per liter

^*P-values calculated with Fisher’s exact

^**Mann-Whitney ranksum.

On multivariable analysis (Table 2), age <18 and faster onset of symptoms were independently and significantly associated with GI-isolated symptoms versus systemic symptoms, whereas site/race, gender, alpha-gal IgE titer and history of tick bites were not. On univariate analysis, the odds of GI-isolated symptoms vs. systemic symptoms among children were 5 times that of adults 35 years or older (odds ratio 5.33, 95% confidence interval 1.40–20.4). Longer symptom onset made GI-isolated symptoms less likely (odds of GI-isolated symptoms 0.87 per 15-minute increase in symptom onset, 95% confidence interval 0.78–0.97). Univariate analysis results are shown in Figure 3.

Figure 3.

Characteristics that predicted a GI-Isolated reaction to mammalian meat challenge, by univariate analysis.

DISCUSSION

GI distress was common and GI-isolated manifestations of alpha-gal allergy were frequent among patients who reacted in our pooled cohort, which includes to our knowledge all patients with oral food challenge-proven alpha-gal allergy, apart from 4 in Germany.¹² Internists, gastroenterologists, emergency room physicians and especially general pediatricians and pediatric gastroenterologists should be aware of the GI phenotype of alpha-gal allergy.¹³

The study demonstrates that an allergic condition outside of eosinophilic esophagitis, a well-recognized GI allergic condition, can cause isolated GI symptoms. Prior research on GI-isolated allergies outside of eosinophilic esophagitis is limited. One study performed endoscopy-delivered food challenges to the duodenum during endomicroscopy in patients with irritable bowel syndrome, and found that a majority of these patients had real-time increases in intestinal permeability at food antigen exposure, suggesting these patients had non-IgE food allergy.²

The GI-isolated phenotype was a common one, suggesting that alpha-gal reactions may frequently be a local hypersensitivity reaction, with reactions at the site of antigen contact. Mast cells are known to be present in the GI tract and, in the setting of alpha-gal allergy, populated with antigen-specific IgE that can recognize and bind to alpha-gal prior to systemic antigen absorption.¹⁴ Differences in GI mast cell number and/or high-affinity IgE receptor expression levels may explain why some patients have more robust GI symptoms than others. Moreover, gut microbiota stimulate the production of IgA, IgM and IgG to alpha-gal, which are present in all humans, and there are known interpersonal variations in the amount of these immunoglobulins.^15–18 Thus, a second explanation for prominent GI symptoms in alpha-gal allergy could be that in patients with lower levels of IgA, IgM and/or IgG anti-alpha-gal there is an increase in available alpha-gal for IgE binding and this leads to a heightened local (GI) allergic response.^{5, 18} Third, variations in the numbers of gut microbes responsible for stimulating the endogenous anti-alpha-gal IgA, IgM and IgG responses may predispose patients to GI-specific alpha-gal symptoms.¹⁹ These hypotheses are actively being explored by our groups.

Importantly, children were more likely to have isolated-GI alpha-gal allergy, as demonstrated in the cohort. GI isolated alpha-gal allergy in children is likely underdiagnosed and has been rarely reported. A case report describes a 6-year-old male from Missouri who presented to his pediatrician with abdominal pain and changes in bowel habits for 3 weeks, but alpha-gal allergy was not diagnosed until 11 months later, when he had urticaria hours after eating hamburger.²⁰ In a retrospective cohort of 42 alpha-gal allergic children in Missouri, most (78.0%) had a prior diagnosis of allergic rhinitis, eczema, or asthma, and the most common presenting symptoms were urticaria (78.0%), anaphylaxis (29.3%) angioedema (21.9%) and GI distress (17.0%).²¹

The delayed nature of alpha-gal reactions may make diagnosis of GI-isolated alpha-gal particularly challenging for both patients, parents and physicians, though GI-isolated patients did react more quickly than patients with systemic symptoms. Our cohort reacted after a minimum of 45 minutes to a maximum of over six hours. The delay allows for the allergy sufferer to be exposed to other potential food and other allergens and hinders identification of the cause-and-effect relationship which is usually obvious in food allergy.

Our study has some limitations. Challenges in our cohort studies were open-label, whereas experts consider double-blinded studies to be the gold standard for food challenges. Double-blinded challenges were not feasible as alpha-gal-free mammalian meat did not exist at the time of the challenges, and other meats have different flavors and texture. Also, though our study was conducted on two continents, doctors from Europe, South America, Australia and Asia have also reported alpha-gal allergy, and the challenged individuals in this study may be different from other alpha-gal allergic individuals.

However, the study also has strengths. Allergists from Africa and North America with a tertiary care referral basis of two different ethnic populations conducted the challenges. We systematically collected symptom data and serum IgE levels. Disease definitions and symptom manifestations were defined in the same ways.

In conclusion, GI-isolated distress and GI distress were common among alpha-gal allergic participants at food challenge, particularly children. Further prospective studies are necessary to better understand the epidemiology, pathophysiology, and protean manifestations of the GI phenotype of alpha-gal allergy.

Table 3.

Characteristics to predict isolated GI symptoms versus systemic symptoms in univariate and multivariable analyses.

Characteristic	Unadjusted odds ratio (95% confidence limits)	Multivariate odds ratio* (95% confidence limits)
Age:
< 18 years	5.33 (1.40 – 20.4)	4.78 (1.21 – 18.8)
18 – 35 years	3.20 (0.65 – 15.8)	3.42 (0.65 – 17.9)
≥ 35 years	Referent	Referent
Onset of symptoms - per 15 minutes	0.87 (0.78 – 0.97)	0.88 (0.88 – 0.99)
Gender - female versus male	1.33 (0.55 – 3.19)
Alpha-gal IgE - per 10 unit / liter increase	1.01 (0.90 – 1.11)
History of tick bite versus no history	0.82 (0.29 – 2.32)
Site - Virginia versus South Africa	0.14 (0.02 – 1.15)

^*,forward selection, U/L, Units per liter

Abbreviations:

GI

Gastrointestinal

Ig

Immunoglobulin