Cytokine Variations After Positive Oral Food Challenges in Adult‐Onset Food Protein‐Induced Enterocolitis

To the Editor,

In recent years, an increase in the number of reports of adult‐onset food protein‐induced enterocolitis syndrome (FPIES) has been observed, reflecting increased awareness of FPIES in adult patients.

The diagnosis of FPIES is based on clinical history and oral food challenge (OFC), as no clinical biomarker exists [1, 2]. The pathophysiology of FPIES is not well known. Prior studies demonstrate that FPIES reactions are associated with systemic innate immune activation [3]. These findings were supported by Mehr et al. describing changes in innate inflammatory gene expression in peripheral blood after an FPIES OFC [4]. Additionally, Berin et al. reported that in the pediatric FPIES population, reactions were associated with a unique IL‐17 signature and activation of innate lymphocytes [5].

The aim of our study was to identify potential biomarkers implicated in the pathophysiology of adult‐onset FPIES that may be useful to assess disease activity versus resolution. We measured detectable cytokines and chemokines in adult‐onset FPIES patients undergoing a supervised OFC and compared this to a healthy adult population with no prior history of FPIES or food allergy.

Adults with a history of FPIES presenting in adulthood underwent a supervised OFC with the causative food. All patients included in the active group had at least 3 episodes suggestive of FPIES with the same or related foods and reported having tolerated the offending food in the past. Open OFCs were performed to confirm an FPIES diagnosis or to evaluate resolution, if no accidental exposures had occurred in the last 18 months.

The OFCs were performed in the day hospital unit. An adult serving size was divided into 2 doses; the first was 25% of the total amount, followed by a 2‐h observation. The remaining dose was given, followed by a 4‐h observation period [6]. The Ethics Committee of Alicante General University Hospital approved the study, and patients provided their informed consent prior to study enrolment.

Cytokine and chemokine variations were analyzed in the serum of OFC reactors and in two non‐reactor groups (patients with resolution of their FPIES and a control group with no prior history of FPIES or food allergy). The control group included patients avoiding a food due to non‐specific symptoms. The clinical history of these patients was not consistent with a food allergy, and specific IgE and skin tests were negative. OFCs in this control group were performed to confirm tolerance to the suspected foods.

Blood samples were obtained at baseline (T1) and 2 h following symptom onset (T2). In non‐reactors, T2 samples were taken prior to discharge, 4 h post‐challenge.

Serum cytokines and chemokines were analyzed at T1 (pre‐OFC) and T2 (post‐OFC) using the BD Cytometric Bead Array (CBA) kits: the Th1/Th2/Th17 kit for TNF‐α, IL‐10, IL‐6, IL‐17A, IFN‐γ, and IL‐4, and the Inflammatory Cytokine kit for IL‐12p70, IL‐1β, and IL‐8. Antibody‐coated beads were incubated with serum samples and detected by fluorescent antibodies. Cytokine levels were quantified by analyzing the fluorescent intensity on a BD FACSCanto II flow cytometer. Data were processed using FCAP Array TM software (v.3.0) (Becton Dickinson).

Demographic and clinical characteristics are outlined in Table 1. Serum samples were collected pre‐OFC (T1) and post‐OFC (T2) from FPIES OFC reactors (n = 38), FPIES OFC tolerators (n = 6) and control patients (n = 10) with similar demographic characteristics but no prior history of FPIES. Results were analyzed by comparing post‐challenge values (T2) to baseline (T1) in the three populations. In 20 FPIES OFC reactors, neutrophilia in excess of 1000 cells/mL was observed following a reaction during the OFC.

TABLE 1.

Demographic and clinical characteristics in patients with active and resolved adult‐onset FPIES. Descriptive analysis was performed using percentages and absolute numbers for qualitative data, and for quantitative data, median (interquartile range [IQR], p25‐p75) was used.

Characteristics	Control	Active FPIES	Resolved FPIES	p
Sex (n, %)
Male	2 (20)	6 (16)	1 (17)	0.999
Female	8 (80)	32 (84)	5 (83)
Age at OFC (years), median (IQ25–75)	38 (26–42)	39 (28–45)	42 (21–52)	0.910
Age at first reaction (years), median (IQ25–75)	—	32 (20–40.5)	31.5 (7–47)	0.590
Number of episodes, median (IQ25–75)	—	10 (5–10)	10 (10–12)	0.765
Time from the most recent reaction to OFC, (mo) median (IQ25–75)	—	28 (14–48)	32 (12–60)	0.987
Time to reaction from T1 to T2 (median) hours (IQ25–75)	—	4.2 (3.8–6.2)	—
Symptoms during OFC (n, %)
Abdominal pain		35 (92)	—
Nausea	—	27 (71)	—
Diarrhea	—	32 (84)	—
Vomiting	—	23 (60)	—
Lethargy	—	11 (29)	—
Atopy (n, %)	—	23 (60.5)	5 (83.3)	0.392
Allergic rhinitis	—	22 (57.9)	2 (33.3)	0.042 *
Asthma	—	10 (26.3)	1 (16.6)	0.999
Atopic dermatitis	—	5 (13.1)	2 (33.3)	0.238
IgE mediated food allergy to other foods	—	3 (7.9)	1 (16.6)	0.878
Eosinophilic esophagitis	—	2 (5.2)	1 (16.6)	0.755
Foods (n, %)
Shrimp	1	4 (10.2)	1 (16.6)	0.537
Fish	7	5 (13.1)	1 (16.6)	0.999
Cow's milk	—	6 (15.3)	1 (16.6)	0.999
Banana	—	4 (10.2)	1 (16.6)	0.537
Egg	—	11 (28.2)	2 (33.3)	0.999
Mushroom	—	2 (5.1)	0 (0)	0.999
Oat	2	4 (10.2)	0 (0)	0.999
Beef	—	1 (2.5)	0 (0)	0.999
Rice	—	2 (5.1)	0 (0)	0.999

Note: Comparisons of quantitative variables were performed with the nonparametric (Kruskal–Wallis test and Mann–Whitney test for non‐related samples). We compared active group with remission group for all the variables, and in the variables age and sex we compared all 3 groups (active, remission and control group).

^* Statistically significant p‐values were defined as ≤ 0.05.

Symptomatic FPIES OFCs were associated with a significant elevation in IL‐17 (p < 0.001) and IL‐6 (p < 0.05) at T2 versus T1. These cytokines were not increased in either patients whose FPIES resolved or in the control group. The remaining cytokines and chemokines did not show significant variations between the post‐OFC and baseline levels (Figure 1).

FIGURE 1.

Cytokines values in adult‐onset active FPIES (cases) and resolved FPIES (remission) and healthy controls. Biomarkers were measured from serum obtained at T1(baseline) and T2 (2‐h following symptom onset in FPIES reactors; 4‐h post‐OFC in non‐reactors: FPIES resolved or non‐food allergic controls). A within group analysis was performed using a paired t test. *p < 0.05, **p < 0.001.

Berin et al. had previously reported that following a symptomatic FPIES challenge in the pediatric population, a significant elevation of cytokines in the IL‐17 family was observed, specifically an increase in IL‐17A, IL‐17C, and IL‐22. However, the same cytokines were not increased in the FPIES non‐reactors or in patients with a history of IgE food allergy [5]. Our results in the adult population are similar to those reported by Berin et al. [5] in the pediatric population, supporting a common mechanism between pediatric and adult‐onset FPIES.

The IL‐17 cytokine family is not fully understood; however, IL‐17A has been well studied given its pivotal role in inflammatory processes and in autoimmunity. Additionally, elevation in IL‐17B has been observed in intestinal inflammation, promoting neutrophil migration following intraperitoneal administration, suggesting also a pro‐inflammatory role [7]. Production of pro‐inflammatory cytokines, such as IL‐1, IL‐6, and TNF‐α, has been shown to be produced de novo by neutrophils. Thus, the neutrophilia seen in an acute FPIES reaction can further explain the elevated IL‐6 levels, which further supports the acute production of neutrophils during inflammatory responses [8].

In conclusion, our findings provide supporting evidence about the pathophysiology of adult‐onset FPIES and support an IL‐17 inflammatory signature during an acute FPIES episode, similar to what has been previously reported in the pediatric FPIES population. These findings may help to facilitate our understanding of this elusive disease and provide evidence in identifying biomarkers aiding in the diagnosis.

Author Contributions

P.G.‐D. designed the work; P.G.‐D., S.A. wrote the manuscript; J.B., A.I.J. acquired the data; J.B. data analysis; P.G.‐D., S.A. wrote the draft of the manuscript; P.G.‐D., S.A., C.S.‐F., T.J., P.G.‐V., F.M.C., and J.B. revised the manuscript. All authors approved the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.