Relationships between cetuximab‐induced anaphylaxis and specific antibodies against allergen and tick‐transmitted infections

The principal antigen responsible for cetuximab‐induced anaphylaxis is galactose‐α‐1,3‐galactose (α‐gal), expressed on the Fab portion of the monoclonal antibody. Recent studies have shown that cetuximab allergy is associated with a history of tick bite. We explored factors predicting cetuximab‐induced anaphylaxis, focusing on the relationship thereof to a history of tick bite.

found that immunoglobulin E (IgE) antibodies against cetuximab were present prior to therapy in most patients who developed allergic reactions during cetuximab administration.
The antibodies were specific for galactose-a-1,3-galactose (a-gal), which is present on the Fab portion of the cetuximab heavy chain. 1 In 2009, Commins et al reported that patients with IgE antibodies to a-gal experienced delayed anaphylaxis after eating red meat, consistent with the abundant expression of the a-gal epitope in the cells and tissues of nonprimate mammals. [2][3][4] These authors further suggested that IgE antibodies to a-gal were produced in response to tick bites, because the anti-a-gal IgE antibodies reacted with proteins of Amblyomma americanum, a tick species common in the southeastern United States. Thus, IgE antibodies to a-gal produced by repeated tick bites may be associated with the development of anaphylaxis, both during cetuximab infusion and after consuming red meat. 5 High incidences of both cetuximab-and red meat-induced anaphylaxis have been reported in the southern United States, where tick-mediated Rocky Mountain spotted fever is also common. 5,6 In Japan, Chinuki et al. reported a relatively high frequency of cetuximab anaphylaxis in eastern Shimane prefecture, where the prevalence of tick-mediated Japanese spotted fever is also high. 7 In the cited study, specific IgE antibodies against beef, pork, and cetuximab were detected in patients with cetuximab-induced anaphylaxis. 8 Thus, measurement of the levels of IgE antibodies against beef, cetuximab, and a-gal may be useful to avoid cetuximab anaphylaxis in patients given this monoclonal antibody. 8 As relatively large numbers of tick bite cases are reported in Ehime Prefecture, Japan, we explored whether cetuximab-induced anaphylaxis was associated with the presence of serum antibodies against red meat and/or a-gal. In addition, we measured the levels of antibodies against Rickettsia japonica, Fancisella tularensis, and severe fever with thrombocytopenia syndrome virus (SFTSV), which cause Japanese spotted fever, tularemia, and severe fever with thrombocytopenia syndrome (SFTS), respectively; these conditions indicate a history of tick bite.

| Patients
Patients admitted to Ehime University Hospital or the Shikoku Cancer Center Gastrointestinal Medical Oncology Department between 2012 and 2014, and who were treated with cetuximab, were eligible for inclusion. All patients gave written informed consent and underwent blood testing. The institutional review boards of both institutions approved the study.

| Measurement of specific IgE antibodies
The immunoCAP kit (Thermo Fisher Diagnostics, Tokyo, Japan) was used to measure the levels of specific IgE antibodies against beef, pork, chicken, dog and cat dander, and a-gal; the results were grouped into seven classes (0-6). The standard positive cutoff was 0.35 U A /mL, and the limit of detection was 0.1 U A /mL. An IgE level <0.35 U A /mL was graded as class 0, 0.35-0.7 U A /mL as class 1, 0.7-3.5 U A /mL as class 2, 3.5-17.5 U A /mL as class 3, 17.5-50 U A /mL as class 4, 50-100 U A /mL as class 5, and >100 U A /mL as class 6.
Anticetuximab IgE levels were measured using an enzyme-linked immunosorbent assay (ELISA) (Somru BioScience Inc., Charlottetown, PE, Canada), according to the manufacturer's protocol. The cutoff was 0.055, obtained by calculating by the average absorbances of six negative samples and adding two standard deviations to the mean.
Antibodies to rickettsial and viral antigens were measured using the indirect immunoperoxidase (IP) test. 9,10 The antigen was spotted on a slide, air-dried for 30 min, and then fixed in acetone for 10 min at room temperature. Air-dried antigen-slides were either used immediately or stored at À20°C. Sera initially diluted to 1:40 with 0.01 M phosphate-buffered saline (PBS), pH 7.2, were subjected to serial twofold dilution in microtiter plates. Antigen-bearing slides spotted with 0.01 mL of each dilution were incubated for 30 min at 37°C in a humidified chamber, and washed twice for 5 min in PBS.

| Western blotting
Cetuximab (Bristol-Myers Squibb, Tokyo, Japan) was diluted in buffer to 10 lg/mL and subjected to polyacrylamide gel electrophoresis.
The protein was transferred by blotting the gel onto a polyvinylidene NAMBA ET AL. | 59 difluoride membrane and reacted with patient sera. Antigen-specific IgE in serum was detected using a peroxidase-labeled anti-human IgE antibody (KPL, Gaithersburg, MD, USA) and an ECL Prime kit (GE Healthcare, Tokyo, Japan).

| Statistical analysis
Student's t-test and Fisher's exact test were used to analyze the results. All analyses were performed with the aid of STATFLEX software (ver. 4.2; ARTEC Inc, Osaka, Japan). A two-sided P value <.05 was considered to indicate statistical significance.

| Baseline patient characteristics
The 52 Japanese patients were enrolled in this study. The mean age of the 52 patients (40 men, 12 women) was 66 years (range: 27-90 years). Blood samples were taken before cetuximab administration in three of the 52 patients, and after administration (range: 1-497 days; median = 14 days, average = 94.5 days) in the remaining 49 patients. Diagnoses were based on the World Allergy Organization guidelines for assessment and management of anaphylaxis. 12 The anaphylaxis guidelines of the Japanese Society of Allergology were used for grading. 13 Anaphylaxis developed in five (four men, one woman) of the 52 patients during the first course of cetuximab (Table 1). Four cases were of grade 3 and one was of grade 2 ( Table 1). All five patients then stopped cetuximab therapy. Chlorpheniramine maleate (5 mg) and dexamethasone sodium phosphate (6.6 mg) were given before cetuximab treatment. Of the 52 patients, two had histories of tick bite and one (case 1) developed anaphylaxis to cetuximab (Table 2). Of the five patients developed anaphylaxis to cetuximab, four had B-negative blood and one (case 4) had blood type B (Table 1).

| Detection of specific IgE antibodies
We measured the serum levels of specific IgE antibodies against beef, pork, chicken, and dog and cat dander. Specific IgE antibodies against pork were not detected in any patient (P = 1.00) ( Table 3).
Specific antibeef IgE antibodies of class 1 and higher were detected in two of the 52 patients, neither of whom had cetuximab-induced anaphylaxis (P = .81) ( Table 3).
Two patients had specific IgE class 1 antibodies against a-gal; both exhibited cetuximab-induced anaphylaxis (P = .0075) (Tables 2   and 3). The other three anaphylactic patients were of specific IgE anti-a-gal class 0. In two of those patients, the levels were over the detection limit but still within the class 0 range ( Table 2). The 47 patients who did not have cetuximab-induced anaphylaxis had no specific IgE antibodies against a-gal. Thus, positive antibody status was significantly associated with cetuximab-induced anaphylaxis.
Western blotting was performed to detect specific IgE antibodies against cetuximab, but bands specific for the monoclonal antibody were not apparent in any of the five patients with anaphylaxis. Upon

| DISCUSSION
Cetuximab is a chimeric mouse-human monoclonal IgG antibody against the epidermal growth factor receptor (EGFR), and is approved for use in patients with EGFR-positive unresectable colon and rectal cancer, and head-and-neck cancer. In addition to side effects involving the skin, severe anaphylaxis following first-time administration of the drug has been reported; the frequency is 1.4%-4.5%. [14][15][16][17][18] T A B L E 1 Patients' characteristics and symptoms at the anaphylaxis due to cetuximab Antibodies produced against a-gal following a tick bite induce anaphylaxis in patients given cetuximab, the structure of which contains a-gal. Although infliximab, palivizumab, and abatacept also contain the a-gal structure, their interactions with a-gal-specific IgE are not sufficient to elicit an immune response. 19 The Fab domain of cetuximab contains predominantly bi-a-galactosylated bi-antennary complex glycans. Other therapeutic monoclonals contain principally mono-a-galactosylated glycans in the Fc domain; these have a lower affinity for a-gal IgE antibodies than do Fab-domain glycans, 19 explaining the more frequent development of a-gal-specific IgEinduced anaphylaxis following cetuximab administration.
The annual incidence of tick-mediated Japanese spotted fever in Ehime is 12-17 individuals per 1 400 000. 20 In addition, 0.7%-12% of the Japanese population is positive for antibodies against R. japonica 21,22 ; these individuals live predominantly in areas with high rates of Japanese spotted fever. 23 We therefore speculated that measurement of anti-R. japonica antibodies would be useful to verify a history of tick bite. However, we were unable to define the regionality of cetuximab-induced anaphylaxis, or to evaluate the risk of such anaphylaxis by measuring serum antibody titers to tick-transmitted infections.
A significant relationship with cetuximab-induced anaphylaxis was evident only for IgE antibodies targeting a-gal, which were detected in class 1 in two of the five of the patients with anaphylaxis. This association between the detection of specific IgE antibodies and a history of anaphylaxis to cetuximab was significant (P = .0075) ( Table 3). In addition, specific anti-a-gal IgE antibodies were detected in two of the remaining three patients with anaphylaxis, at levels just above the limit of detection (class 0), whereas no anti-a-gal IgE antibodies were detected in any of the 47 patients who did not develop cetuximab-induced anaphylaxis. Our data suggest that although measurement of a-gal-specific IgE antibodies may be useful to avoid cetuximab-induced anaphylaxis, it may be necessary to increase the sensitivity of the test.
We found that the presence of specific IgE antibodies against beef, Our study suggested that ELISA-based measurement of anticetuximab antibodies is not useful to prevent anaphylaxis, as three of our patients with anticetuximab IgE antibodies did not develop anaphylaxis following treatment with cetuximab. However, it should be noted that blood samples were taken after, but not before,  26 It is therefore necessary to review our data while considering pretreatment sera.
The fact that anti-a-gal IgE antibodies were present in two of the five patients with cetuximab-induced anaphylaxis, but not in any patient without anaphylaxis, indicates that the anti-a-gal antibody is an essential contributor to the allergic response. The patient with cetuximab-induced anaphylaxis who lacked the anti-a-gal IgE antibody (case 5) may have experienced an infusion reaction (a frequent complication during the first infusion of a monoclonal antibody 27 ), not anaphylaxis. The mechanism underlying the infusion reaction is unclear, but may involve cytokine release and induction of systemic symptoms by mechanisms that differ from those of IgE-mediated hypersensitivity reactions. 27 Although the clinical symptoms of the infusion reaction resemble those of anaphylaxis, the latter is rarely accompanied by fever. Case 5 had a fever of 39°C (Tables 1 and 2), whereas the other four patients had no fever. Measurement of total tryptase and histamine levels can distinguish between anaphylaxis and an infusion reaction, 12 but this was not performed in this patient.
The involvement of factors other than a-gal hypersensitivity in cetuximab-induced anaphylaxis cannot be ruled out. Cetuximab preparations contain the additive polysorbate 80, a known inducer of anaphylaxis.
Urticaria has occurred in patients given adalimumab and ustekinumab; eyelid angioedema, rhinitis, conjunctivitis, dyspnea, and wheezing have developed in recipients of the human papilloma vaccine. 28,29 All three drugs contain polysorbate 80. Thus, it is possible that polysorbate caused anaphylaxis in our patient without anti-a-gal IgE antibodies.
In conclusion, cetuximab-induced anaphylaxis may be associated with a-gal hypersensitivity, as suggested by the presence of IgE antibodies in two of our five patients who developed anaphylaxis in response to the drug. In addition, specific anti-a-gal IgE antibodies were detected in two of the remaining three patients with anaphylaxis, at levels just above the limit of detection (class 0), whereas no anti-a-gal IgE antibodies were detected in any of the 47 patients who did not develop cetuximab-induced anaphylaxis. However, the measurement of antibodies against tick-transmitted infectious agents did not yield proof of any relationship between tick bite and cetuximab-induced anaphylaxis. Moreover, anticetuximab antibodies were detected in only one of the five patients with cetuximab-induced anaphylaxis. Our results suggest that measurement of anti-a-gal IgE antibody titers could prevent the development of cetuximab-induced anaphylaxis, but the test sensitivity must be improved.