Cow’s milk allergy: evidence-based diagnosis and management for the practitioner

The topic of definition still causes confusion among physicians. Words such as “allergy,” “intolerance,” and “hypersensitivity” are used interchangeably. The accepted definition of allergy is “a hypersensitivity reaction triggered by specific immunologic mechanisms” [7, 28, 29] There is no such thing as “allergy to lactose” but rather lactose intolerance.

Conclusions from a 2010 systematic review concluded that the evidence for the prevalence of food allergy is greatly limited by a lack of uniformity of the criteria for making a diagnosis. Consequently, it remains unclear whether the prevalence is increasing [10], although some data suggest it [9]. The prevalence of CMA in children living in the developed world is approximately 2 to 3 % [25, 56], making it the most common cause of food allergy in the pediatric population. Only among breastfed infants is the prevalence lower (0.5 %) [25]. These numbers most likely refer to IgE-mediated CMA, while the prevalence of non-IgE-mediated CMA is not well known.

The major cow’s allergens belong to the casein fraction of proteins (αs1-, αs2-, β-, and κ-casein) and to whey proteins (α-lactalbumin and β-lactoglobulin) [62]. There is some cross-reactivity with soy protein, particularly in non-IgE-mediated allergy. There are immune and non-immune-mediated allergic phenomena. Immune-mediated adverse food reactions can be classified into four major categories: IgE-mediated, non-IgE-mediated, mixed, and cell-mediated reactions [7]. CMA is most frequently caused by a non-IgE-mediated mechanism.

Two basic mechanisms explain allergic reactions to cow’s milk as well as to other food allergens: those mediated by IgE and those not mediated by IgE. The most common IgE-mediated manifestations of CMA are acute urticaria and angioedema. The most common non-IgE-mediated manifestations of CMA are involving skin and the gastrointestinal tract. At the level of the gastrointestinal tract, presentations include the following: (1) CM-induced enterocolitis syndrome which involves the entire gastrointestinal tract, (2) CM-induced enteropathy that involves only the small bowel, and (3) CM-induced proctitis and proctocolitis, involving the rectum and colon [18].

CMA is mostly a disease of infancy and early childhood. Affected infants present usually within the first 6 months of life, and one review reported that the majority of infants develop symptoms before 1 month of age, often within 1 week after the introduction of cow’s milk proteins to their diet [25]. However, breastfed infants can also be affected by dairy products ingested by the mother and eliminated in her breast milk. Rare is the onset of symptoms after 12 months of age [36]. The majority of affected children have one or more symptoms involving one or more organ systems, mainly the gastrointestinal tract and/or skin. One recent review suggests that gastrointestinal food allergies are commonly associated with a wide range of extra-intestinal manifestations such as fatigue, allergic shiners, mouth ulcers, joint pain/hypermobility, poor sleep, night sweats, headache, and bed wetting [13].

Symptoms of non-IgE-mediated CMA are mostly delayed reactions that occur beyond 2 h following ingestion) and usually involve the gastrointestinal tract and/or skin [54]. Symptoms such as urticaria and/or angioedema with vomiting and/or wheezing are suggestive of IgE-mediated CMA, which generally occur within minutes and up to 2 h of cow’s milk protein ingestion. The skin is frequently involved followed by the gastrointestinal tract and, least frequently, the respiratory and/or cardiovascular systems. The majority of reactions are mild to moderate, but life-threatening anaphylaxis (1–2 %) can also occur [36, 53] (Table 1). Together with peanuts and tree nuts, cow’s milk is one of the most common foods capable of causing anaphylactic reactions [27]. Evidence of sensitization (presence of specific IgE) is typical [53].

Other IgE-mediated disorders include food protein-induced enterocolitis syndrome (the entire gastrointestinal tract is involved), food protein-induced enteropathy (small bowel), food protein-induced proctitis and proctocolitis (rectum and colon), and food-induced pulmonary hemosiderosis (Heiner’s syndrome) [53]. Mixed IgE- and non-IgE-mediated reactions, involving humoral and/or cell-mediated mechanisms, also manifest themselves at the level of the skin and/or gastrointestinal tract. Such entities include allergic eosinophilic gastrointestinal disorders and atopic dermatitis (eczema).

CMA is generally outgrown during early childhood or, at the latest, in adolescence. Overall, the chances of outgrowing an allergy are better for non-IgE-mediated CMA. Children at risk of not resolving the problem are those affected with IgE-mediated CMA who have high levels of milk-specific IgE antibodies, multiple food allergies, and/or concomitant asthma and allergic rhinitis. Such children are more likely to have a more prolonged persistence of sensitization [36, 51]. Greater chances of developing tolerance to cow’s milk were found in children with low levels of IgE binding to cow’s milk and specific IgE binding to α-lactalbumin, β-lactoglobulin, κ-casein, and αs1-casein [1]. Resolution of CMA within the first 5 years of life could be predicted by milk-specific IgE levels, skin prick test results, and the severity of atopic dermatitis [63]. A web-based calculator to determine the prognosis of children with CMA is available at www.​cofargroup.​org. Validation studies are still needed [36].

Among other organizations, ESPGHAN has developed an algorithm for the evaluation of infants and children with symptoms compatible with the diagnosis of CMA (Fig. 1). In addition to the detailed medical history and physical examination, diagnostic elimination diets, skin prick tests (SPTs), specific IgE (sIgE) measurements, and oral food challenges are part of the routine work-up [7, 18, 31, 53]. If the patient is in the appropriate age range and the history and symptoms are consistent with the diagnosis of CMA, an open or single-blind challenge is often sufficient to make the diagnosis. However, a double-blind, placebo-controlled oral food challenge is still the gold standard for the diagnosis of food allergy [50]. When cow’s milk protein is the only suspected allergen, the diagnosis is simpler than on cases where the child is already ingesting a variety of foods. When multiple food allergies are suspected, published standards for office-based oral food challenges [40], as well as for double-blind, placebo-controlled, oral food challenges, should be followed [49].

A systematic review published recently looked into the specificity and sensitivity of tests employed for the diagnosis of food allergy. Results indicate that the existing evidence regarding the accuracy of such tests is limited, making their interpretation problematic [58]. In IgE-mediated food allergy, determination of SPT and sIgE seems to be sensitive, albeit not specific. For the specific case of IgE-mediated CMA, the values are as follows: sIgE, sensitivity 87 % (75 to 94) and specificity 48 % (36 to 59); SPT, sensitivity 88 % (76 to 94) and specificity 68 % (56 to 77). In non-IgE-mediated CMA, the value of tests is far more limited, and the clinician needs to rely on history, physical examination, and the results of the elimination diet and relapse upon milk challenge. The offending food challenge is the diagnostic gold standard [36]. Screening tests, such as SPTs, sIgE tests, and atopy patch tests, have been shown to lack specificity and sensitivity [7]. On the other hand, tests such as Vega (electrodermal), or cytotoxicity, iridology, kinesiology, food-specific IgG, pulse, and hair analysis are not recommended for the diagnosis of allergy because of the lack of any scientific evidence and reliability and reproducibility [9, 36]. Fecal studies for food particles or immune components are not reliable either.

Avoidance of cow’s milk protein in any form is the only available treatment [14, 21, 36]. In the case of breastfed infants, the mother should eliminate all dairy products from her own diet. It has to be considered that it may take up to 72 h to clear breast milk antigens ingested by the lactating woman. Calcium supplements should be added to the mother’s diet to replace milk intake [18, 31, 36]. From the practical standpoint, treatment of CMA imposes less sacrifices on the mother if the child was not being breastfed, provided that the family is given access or can afford the cost of special infant formulas. For infants 6 months old or younger, the recommended formulas for treatment of CMA are extensively hydrolyzed protein or amino acid-based formula. In infants older than 6 months, soy formula could be tried particularly in IgE-mediated cases.

Calcium supplementation (also phosphorus and vitamin D) is not generally necessary in infants ingesting sufficient amounts of special formula. Whenever milk intake is below 500 ml, assessment by a pediatric dietitian is recommended, and Ca supplement may be needed [36].

Cow’s milk exclusion diets without appropriate substitution may lead to nutritional deficiencies and poor growth [38]. Discomfort from the underlying illness such as atopic dermatitis or feeding difficulties due to esophageal dysmotility in eosinophilic esophagitis may further contribute to inadequate nutrient intake. Considering these concerns, scientific organizations recommend the use of an age-appropriate milk substitute in children younger than 2 years of age and food counseling. Isolauri et al. analyzed 100 infants with a mean age of 7 months with a diagnosis of atopic dermatitis and challenge-proven CMA who were evaluated for growth during the therapeutic elimination diet [26]. Although clinical control of symptoms was achieved in all patients, mean length SD score and weight-for-length index of patients decreased compared with those of healthy age-matched children, p < 0.0001 and p = 0.03, respectively. In addition, low serum albumin was seen in 6 %, abnormal urea concentration in 24 %, and low serum phospholipid docosahexaenoic acid in 8 %. The delay in growth was more pronounced in a subgroup of patients with early onset than in those with later of symptoms.

Once an infant is diagnosed as having CMA and is placed on an exclusion diet, reevaluation needs to be performed every 6 months if the child is under 1 year of age and every 6–12 months from 1 year of age onward, to determine if the child is a candidate for reintroducing cow’s milk. The BSACI has suggested an escalation of products, a so-called “milk ladder,” starting with baked milk products, as thermal processing reduces allergenicity [36]. If well tolerated, more allergenic products can be reintroduced progressively leaving for the end uncooked cheese and fresh cow’s milk, which should only be introduced in children with demonstrated full tolerance to baked milk products.

The WAO recently concluded that, as of today, no single probiotic supplement or combination of them has shown to dramatically influence the course of allergic manifestations or long-term outcome in a permanent way [19]. One randomized controlled trial (RCT) published subsequently to the WAO document found that the addition of Lactobacillus rhamnosus GG (LGG) to the therapeutic formula has an impact on acquisition of tolerance. In this trial particpants were randomly assigned to receive one of the following formulas: extensively hydrolyzed casein, extensively hydrolyzed casein with LGG, hydrolyzed rice, soy, or amino acid-based [4]. The rate of oral tolerance after 1-year treatment determined by food challenge was significantly higher in the groups that received extensively hydrolyzed casein formula whether it was with LGG (78.9 %) or without (43.6 %) compared with the other groups: hydrolyzed rice formula (32.6 %), soy formula (23.6 %), and amino acid-based formula (18.2 %) Repeat studies are needed.

At present, there are no established guidelines or protocols on how to proceed with this aspect of treatment. Once an elimination diet is in place and the patient improves, the next major challenge is the induction of tolerance. Reasoning behind the use of the oral route is to expose the immune system to either low doses of antigen or to antigenically modified molecules, capable of inducing a response of immunotolerance without one of allergy. As shown in two meta-analyses, compared to an elimination diet alone, oral immunotherapy for IgE-mediated CMA showed improved chances of achieving CM’s tolerance [8, 64]. However, those two meta-analyses also showed that development of long-term tolerance was unlikely. Oral immunotherapy, however, poses the risk of severe adverse reactions. Experience, however, indicates that when reactions occur, these are generally mild and short lasting. A possible form of oral immunotherapy could be the use of extensively heated milk as well as egg protein because studies have shown that the protein treated in such manner may be tolerated by children who react to raw cow milk [30, 35]. Although studies are still limited, experts have suggested that an oral challenge under professional supervision using heated milk could be tried in children with CMA. Guidelines still do not recommend the use of baked milk products for desensitization in routine clinical practice.

There is a potential role for probiotics in inducing immunotolerance. A study of the effect of certain probiotic strains on tolerance acquisition in children with CMA gave negative results [24]. However, Berni-Canani et al. [3] randomly allocated infants with CMA while still receiving intact protein formula to a group that received either extensively hydrolyzed casein formula or the same EHCF containing Lactobacillus GG. After 6 months of an exclusion diet, a double-blind placebo-controlled milk challenge was performed in 55 patients, and evidence of tolerance was seen in 21.4 and 59.3 %, respectively. The difference in acquisition of immunotolerance was significant only for those children with non-IgE-mediated CMA (p = 0.017).

It could be hypothesized that allergen avoidance during the first few months of life, period of immune immaturity, could be beneficial in allergy prevention. However, evidence points otherwise.

One recent systematic review found no robust studies investigating the effectiveness of adrenaline (epinephrine), H1 anti-histamines, systemic glucocorticosteroids, or methylxanthines in the management of anaphylaxis [12].

Even if the evidence is limited, the first-line treatment for anaphylaxis is epinephrine (both in the outpatient setting [autoinjector] and in a hospital setting). Other medications used in the management of anaphylaxis include anti-histamines or anti-inflammatory drugs (systemic or topical steroids) [7]. The latter are the main therapy in cases of eosinophilic esophagitis or gastroenteritis in which dietary restriction was not feasible or had failed to improve the disease [48].

Infants suspected of having CMA who are not breastfed are placed on a special infant formula for up to 6 weeks (depending on the symptoms) and then challenged. If they do not relapse, they are considered to be either cured or that the diagnosis was incorrect. However, CMA may relapse with symptoms that are different from those seen at presentation. One form of CMA is enteritis which may lead to nutrient malabsorption. It is recommended that infants be followed closely for growth parameters following reintroduction of cow’s milk to their diet.

Another caveat is that infants with CMA may have delayed gastric emptying and present with vomiting hours after having ingested milk or food. In evaluating infants experiencing vomiting and considering gastroesophageal reflux, it has to be kept in mind that in simple gastroesophageal reflux, vomiting occurs during or immediately after a meal (30 min), while vomiting that occurs hours after a meal may be associated to allergy. Ravelli et al. described that in sensitized infants, cow’s milk induces severe gastric dysrhythmia and delayed gastric emptying, which, in turn, may exacerbate GER and induce reflex vomiting [47].

In this article, we review current recommendations regarding CMA, which is the most common food allergy in infants and young children. A medical history including history of allergy in close relatives, physical examination, and diagnostic elimination diets are the first steps for the accurate diagnosis and management of these patients. sIgE measurements, SPTs, and oral food challenges are usually performed to determine if the problem is IgE-mediated or not. Strict avoidance of the offending allergen is the only therapeutic option. Recommendations for the primary prevention of allergy include exclusive breastfeeding for at least 4 months and up to 6 months if possible. Infants with a documented hereditary risk of allergy (i.e., an affected parent and/or sibling) who cannot be exclusively breastfed should receive a formula with confirmed reduced allergenicity, i.e., a partially or extensively hydrolyzed as a way to minimize the risk of allergic reactions, primarily atopic dermatitis. There is no evidence that avoidance or delayed introduction of solid foods beyond 4–6 months has a positive effect for allergy prevention.