**4. Discussion**

FPIES was recognized and formally defined in the mid-1970s [21]. Until recently, both acute and chronic FPIES was considered very rare or underdiagnosed [22,23] due to symptoms that can easily be confused with other diseases, especially with sepsis in the instance of acute FPIES [1,7,20]. Intensive research in recent years has shown that FPIES is much more common than it seemed. Studies conducted in centers in Israel and Spain indicated that the cumulative incidence of FPIES in the birth cohort ranges from 0.34% to 0.7% [24,25]. Population studies in the USA have shown that physicians diagnose FPIES in 0.28% of children (<1 and 0.11% of infants) and in 0.22% of adults [26]. Obviously, the data regard two forms of the syndrome and its incidence not only after consumption of milk but of other foods as well.

In the group under study the first symptoms of chronic milk-FPIES occurred shortly after breastfeeding (Table 1). Most mothers decided to continue breastfeeding for a short time, i.e., one month. After 2–4 weeks of infant formula administration, vomiting and diarrhea occurred. The median age of chronic milk-FPIES diagnosis and of the initiation of milk-free diet were low—2.2 months. What is typical for FPIES induced by cow's milk is the appearance of symptoms during formula feeding. Only a few cases of this syndrome have been described in breastfed babies when the allergens were transported with breast milk [27,28]. Researchers in Japan report FPIES in 10% of breastfed babies [29].

The major symptoms of chronic milk-FPIES are intermittent vomiting and watery diarrhea that rapidly lead to weight and growth deficits. In the study group, as many as half of infants had low BMI < 10 c, and in every fifth child the index was extremely low at BMI ≤ 3 c (Table 1).

Although FPIES is a form of IgE-independent food allergy, patients often have atopy, including atopic dermatitis and/or food IgE sensitization. In the study group of children, AD was diagnosed in 18% of children (Table 2). Studies report more frequent AD coassociation (31% to 57%) in the United States and Australia while in Korea, Israel, and Italy it is rarer (up to 9%) [7].

In chronic milk-FPIES, food IgE sensitization defined as positive food SPT or serum food sIgE levels occurs in from 4% to 30% of patients [5,7]. In the study group, positive milk SPT and serum milk-sIgE were found in a quarter of children (Table 1). Only in 4 children did the sIgE concentration for milk exceed 0.7 kU/L. Nevertheless, during the 1. OFC in none of them immediate reactions after milk supply were observed. The milk-sIgE concentrations decreased systematically with age. FPIES caused by milk for which there is an elevated sIgE concentration is called atypical FPIES [7,8]. FPIES caused by milk in a child without the presence of milk sIgE is called classical FPIES.

Caubet et al. reported that children with chronic milk-FPIES who had increased milk sIgE were more likely to have persistence of chronic milk-FPIES after 3 years of age compared with those without sIgE. [9]. We did not observe this phenomenon in the study group of children. We also found that the presence of allergies to other foods did not delay the development of milk tolerance.

Only 3 children with elevated milk sIgE levels developed symptoms of IgE-mediated milk allergy during subsequent challenge (Table 2). Thus, a transition of FPIES—IgEindependent milk allergy into IgE-mediated allergy was observed in these children. This is a typical phenomenon in children with FPIES and does not occur in other forms of non-IgEmediated food allergy. In the study group, we observed such a reaction in 3 children during the third and fourth OFC in 18, 25, and 26 month-olds. During OFC (18–48 mL of milk), all of them developed extensive urticaria. Two of them also developed also bronchospasms (Table 2). In the following years, their sIgE of milk and casein were elevated. Over the 4 years of observation, we did not perform any provocation tests waiting for the period when milk sIgE would start decreasing.

Atopic sensitization to foods other than milk was observed less frequently (25% vs. 18%) (Table 1). Those foods were well tolerated by most children. However, two of them developed symptoms of IgE-mediated allergy (mild urticaria) already in infancy (wheat, egg white) (Table 3). In the following years, extensive urticaria occurred for two more foods—peanuts and soya (Table 3). The coexisting IgE-mediated food allergy at presentation or on follow-up assessment was also reported by other researchers in 20% to 40% of patients [9,11,12]. Children with FPIES are also more likely to be allergic (2–20%) to foods that typically cause FPIES. Cereals (rice, oat), vegetables (sweet potato, green beans, peas), and poultry meats (chicken and turkey), which are typically treated as a potential weak allergen, must be considered in the follow-up of FPIES, particularly in infants with FPIES to cow's milk or soy. Infants with FPIES are at risk of developing hypersensitivity to many food proteins [7,9,24]. In the study group of children, during the introduction of new foods, symptoms of IgE-independent allergy to 6 foods were observed in 6 infants (11% of children). These were rice, apple, turkey, and chicken meat.

Due to significant weight loss, a one-fifth of FPIES patients met the criteria for severe form of CMA and were treated with an elementary diet. The others received lactose-free casein hydrolysates (Table 1). Milk and baked egg were not used during treatment of chronic milk-FPIES primarily because of the short duration of the disease (at median 25 months of age 87% children had already developed tolerance to milk) and the lack of such recommendations in non-atopic allergies [1,7].

For infants with chronic milk-FPIES, avoidance of all forms of milk, including baked and processed milk, is recommended due to a lack of sufficient studies, although tolerance to baked cow's milk has been reported in a small case series [12,20,30]. Some researchers report that some infants with FPIES in OFC tolerate even more than 100 mL of cow's milk [24]. Most researchers believe that the tolerable dose is low [7].

In infants it is recommended to continue breastfeeding. In the study group only 13% of children were breastfed for 6 months, then received eHF.

The introduction of soy milk is not recommended because of the frequent co-occurrence of allergies to both these foods (in about 20–40% of US patients). This coincidence was not observed in patients from other countries such as Italy, Australia, or Israel [12,24,31]. In the study group of children symptoms of allergy to soya occurred only in 1 child in the second year of life in the form of IgE-dependent allergy (urticarial) (Table 3).

Goat and sheep milk are also not recommended in patients with chronic-milk-FPIES due to high homology of the protein sequences in these milks to cow's milk [32].

As regards patients with FPIES, any modification of their diet requires strict supervision, especially when it involves the introduction of grains (rice, oats), poultry and legumes, as they often cause symptoms of acute FPIES, as described in the literature [7,11]. According to the recommendations, in the presented group of children those foods were administered after 8 months of age (Table 4). Simultaneously, meat and fish was included as well. The order of the types of meat to be introduced was changed. Rabbit meat was given first, followed by pork and then turkey and chicken meat and beef. Eggs and pork were given at the same time as it is recommended for children without FPIES (7 and 8 months of age) [1,7].

Despite the delayed time of introducing new foods into the infants' diet, we observed IgE-independent allergy symptoms to 6 foods in 6 children (11%) (Table 3). They were skin lesions or loose stools occurring from 8 to 24 h after consumption of these foods: rice, apple, chicken, and turkey meat. No infant developed FPIES dependent on solid foods.

Although children with chronic milk-FPIES are more likely to react to solid food, most commonly rice or oat, current feeding guidelines do not recommend delay in introducing complementary foods past 6 months of life due to FPIES [5,33,34]. It is recommended that parents introduce a new food as a single ingredient and wait at least 4 days before introducing another food to observe for the development of a reaction [19].

Delayed intake of these foods is more likely to trigger allergy symptoms [5,14]. It is believed that when an infant tolerates the first few foods introduced, dietary expansion may be more tolerable [7]. The use of an elimination diet is always associated with a risk of nutritional deficiencies, so a dietary consultation is recommended [7,35],

OFCs performed in patients with FPIES are high-risk procedures and therefore require medical supervision. In patients with acute FPIES, intensive vomiting quickly leads to dehydration, acidosis, and consciousness disorders that require intensive intravenous treatment. In addition, in patients with chronic FPIES, the supply of harmful food after a period of elimination triggers acute symptoms (e.g., after the first elimination diet period). In all the FPIES patients with elevated milk sIgE levels, who have been treated with a milk-free diet, there is a risk of a sudden reaction during the follow-up OFC.

Therefore, in clinical practice, OFCs are only used in the initial diagnostic evaluation in cases when the patient's history is not clear, symptoms persist despite the elimination of the potential trigger food, or the time course of symptoms is atypical [7]. In an instance of a typical history and improvement on a dairy-free diet, OFC is not performed because the risk of complications following OFC might outweigh its benefits. OFCs are necessary when there is a need to find out whether the child has already developed tolerance to the eliminated food. In chronic FPIES we perform them more often, on average every 6 months, because the disease recedes faster. In acute FPIES, which regresses more slowly, OFCs are performed less frequently, every 12–18 months.

During 1. OFC, vomiting, diarrhea, and pallor occurred in all children in the study group (Table 2). As many as 42% of children required intravenous hydration and 7% received ondansetron. In the next challenge, the percentage of children with diarrhea, pallor, and dehydration decreased. Vomiting was the predominant symptom (Table 1). Dehydration was less frequent and was treated with oral rehydration.

An increase in leukocytosis and neutrophils (>1500 cells/mL) was also observed during OFCs, as reported by other researchers [7,9,21].

During the 3-year observation, milk tolerance developed in 89% of children with chronic milk-FPIES. In the twenty-fifth month of life the proportion was 87%, while at 12 months of life it was recorded at 62% (Table 2). In three children chronic milk-FPIES transformed into IgE-dependent milk allergy. Three other children dropped out of the study after 18 months of follow-up. Similar results were obtained by Hwang at al [36]. Population studies in Israel have shown that 60% of children with cow's milk-induced FPIES develop tolerance by age of 1 year, 75% by age of 2, and 85% by age of 3 [24]. The retrospective US studies revealed that the tolerance was reached by 35% of children by age of 2, 70% by age of 3, and 85% by age of 5 years [10]. The median age of tolerance

was 6.7 years. The Korean research shows that tolerance may occur sooner, after 6 or 12 months [36].
