Managing cows' milk protein allergy in infants

Review the infant’s allergic clinical history

If an infant is suspected of having CMPA, the first step is to take an allergy-focused clinical history. This process will also inform further investigations (Table 1).

In general, clinicians should aim to identify:¹

• Presenting symptoms and signs that may indicate possible CMPA, including age of onset, timing of onset following ingestion, reaction duration, severity and frequency
• The infant’s feeding history, including identifying sources of cows’ milk protein in the infant’s and mother’s diet (if breastfed), the frequency of feeding (if breastfed), the quantity consumed (if bottle-fed), and any concerns regarding difficulty feeding or growth
• Details of any previous management, including medicines and the response to any treatment or dietary change
• Any history of atopic conditions in the infant or a first degree relative, i.e. parents or siblings

Interpreting the findings. The allergy-based clinical history should provide a clear distinction between IgE- and non-IgE mediated reactions, particularly in terms of the timing, nature of symptoms and family history:¹

• A history of immediate allergic symptoms in response to cows’ milk ingestion strongly supports a diagnosis of IgE-mediated CMPA
• Diagnosing non-IgE mediated CMPA is generally more difficult, particularly as the symptoms have a slower onset in relation to ingestion, they can vary significantly and may be localised to the gastrointestinal tract

CMPA is less prevalent in infants who are exclusively breastfed (0.5%) compared with the overall population (2 – 3%) as their only exposure to cows’ milk protein comes through the maternal diet, which is associated with a low risk of clinical allergy.¹³ Therefore, it is important to not over-interpret minor symptoms in exclusively breastfed children. Any family history of atopic conditions, however, makes a diagnosis of CMPA more likely.¹

For suspected non-IgE-mediated CMPA reactions:

A trial elimination of cows’ milk in the infants diet for two to four weeks is suitable in most infants with mild-to-moderate non IgE-mediated CMPA symptoms and signs.^{1, 13} This should encompass all possible sources of cows’ milk, including dairy products such as yoghurt and cheese. Breastfeeding should continue to be actively supported, with maternal avoidance of cows’ milk.¹³ If infants cannot continue exclusive breastfeeding, or they are already formula feeding, they will require a non-cows’ milk formula during this time (see: “Management of CMPA in infants following diagnosis”). For infants with suspected severe non-IgE reactions, consumption of cows’ milk should stop, and they should be urgently referred to secondary care.

If the infant’s symptoms clearly improve during an elimination trial of two to four weeks: re-introduce cows’ milk to observe whether symptoms return, thereby confirming CMPA; this can be done at home.¹ However, if there is still uncertainty, then more specific testing may be required (Table 1). Allergy testing (e.g. skin prick) is generally not necessary.

If the infant’s condition does not improve following an elimination trial of two to four weeks: consider other possible sources of cows’ milk, e.g. baked goods. However, tolerance to milk in processed products does not exclude CMPA as production methods such as heating and fermentation can diminish the amount of allergen present. In addition, alternative diagnoses should be investigated, and the infant may need to be referred to a paediatrician or allergy specialist.

For suspected IgE-mediated CMPA reactions:

For infants suspected of having an IgE-mediated CMPA reaction based on their allergic history, e.g. immediate urticaria following cows’ milk consumption, a challenge with cows’ milk is not generally not recommended in primary care.⁶ Cows’ milk should be immediately excluded from the infant’s diet and they should undergo allergy testing, e.g. serum allergen-specific IgE or skin prick testing, and referral to a paediatrician or allergy specialist should be considered.¹³ This history alongside a positive allergy test is sufficient to confirm a diagnosis of IgE-mediated CMPA. In all cases of anaphylaxis, the infant should be urgently referred to secondary care for management without the need for initial allergy testing.

Table 1. Clinical presentations and differential diagnosis of IgE-mediated and non-IgE-mediated CMPA.^{4, 8–11}

Immune type	Condition	Time to reaction following ingestion	Clinical features	Distinguishing features	Occurrence in exclusively breastfed infants	Differential diagnosis	Potentially useful investigations (some of which are performed in secondary care)
IgE	Acute allergic reaction (non-anaphylactic)	Immediate, up to two hours	Perioral or orbital angioedema/erythema, generalised urticaria, vomiting, diarrhoea	No recurrence if cows’ milk is completely avoided. Incidence approximately 2% in infants.	Possible	Idiopathic/viral induced urticaria	Skin prick test, IgE antibodies.† An oral inpatient challenge may be considered but is not required for a diagnosis if there is a clear correlation between consumption of milk and the onset of symptoms.
	Anaphylaxis	Immediate, up to two hours	Respiratory +/- cardiovascular involvement often associated with above features	As above. IM adrenaline preferably in the midpoint of the anterolateral thigh is the treatment of choice. Rare manifestation of CMPA.	Extremely rare	Sepsis, acute cardiovascular or respiratory compromise, seizures	Skin prick test, IgE antibodies.† An inpatient challenge is not indicated.
Mixed (IgE and non-IgE)	Atopic dermatitis (eczema)*	Highly variable: min/hours/days	Pruritic rash	Infants: cheeks, trunk, extensor surfaces Older children: flexure surfaces, face, eyelids. Approximately 70% of cases occur in infants aged 3–6 months, however, the incidence due to CMPA unknown.		A wide variety including seborrhoeic dermatitis, psoriasis, acrodermatitis enteropathica	Cows’ milk elimination and re-challenge
Non-IgE	Eosinophilic oesophagitis**	Days	Vomiting, irritability, feed refusal, failure to thrive, oesophageal dysmotility	Histological diagnosis, unresponsive to proton pump inhibitors. Extremely rare with variable age of onset (less likely in older children). A family history is common.	Extremely rare (case reports)	GORD, mucosal candidiasis, Crohn’s disease	Endoscopy, eosinophil count
	Food protein-induced enterocolitis syndrome (FPIES)	Vomiting 2 – 4 hours Diarrhoea 5 – 10 hours	Profuse vomiting +/- diarrhoea, sudden onset of pallor and floppiness. Around 20% present as hypovolaemic shock (with associated metabolic acidosis and methaemoglobinaemia)	Responds to fluid resuscitation, adrenaline not required. Half of affected children outgrow the condition by three years of age.	Extremely rare (case reports)	Sepsis, gastroenteritis, malrotation, intussusception, metabolic disorder	Clinical diagnosis with a focus on history, no laboratory markers available (a raised white blood cell count often adds to clinical confusion)
	Cows’ milk protein-induced GORD	Hours/days	Frequent regurgitation, poor feeding, feed aversion, failure to thrive, symptoms indicative of tissue damage e.g. haematemesis	Partially responsive to proton pump inhibitors when underlying mechanism related to CMPA. Up to 40% of infants with GORD have CMPA.		Idiopathic GORD, eosinophilic oesophagitis, malrotation	Clinical diagnosis. Requires endoscopy if haematemesis or significant failure to thrive.
	Enteropathy	Hours/days	Vomiting, diarrhoea, severe irritability, failure to thrive, iron deficiency anaemia, protein losing enteropathy	Receiving cows’ milk in diet. Unknown incidence due to CMPA.		Lactose intolerance, coeliac disease giardiasis, immune deficiencies, autoimmune enteropathy	Small bowel biopsy for histology, duodenal disaccharidases and microscopy of duodenal aspirate for giardia
	Proctocolitis	Hours/days	Low-grade rectal bleeding and mucus in stool in an otherwise well infant usually within the first three months of life	Normal perianal inspection, thriving. CMPA is the most common cause.		Necrotising enterocolitis, constipation with anal fissure, infantile inflammatory bowel disease, chronic granulomatous disease, juvenile polyp	Rectal biopsy only if atypical features or non-responsive to treatment (eosinophils predominate). If symptoms are prolonged monitor haemoglobin and ferritin levels.
	Heiner’s Syndrome (milk-induced pulmonary disease/ haemosiderosis)	Hours/days	Cough, wheezing, recurrent fever, nasal congestion, recurrent otitis media, failure to thrive, haemoptysis, dyspnoea, colic, anorexia, vomiting and diarrhoea, blood in stool	Extremely rare. Patients may have IgG antibodies to cows’ milk protein (and in some cases IgE antibodies will be present). Peripheral eosinophilia is often seen. Responds to dietary elimination.	Unknown	Idiopathic pulmonary haemosiderosis, chronic bronchopneumonia, aspiration pneumonia	Chest X-ray (looking for pulmonary infiltrates), cows’ milk elimination and re-challenge
Potentially Non-IgE unlikely to be isolated or a sole manifestation	Colic	Hours/days	Paroxysms of unexplained, inconsolable crying (more than 3 hours a day, more than 3 days per week for at least 3 weeks)	Responds to dietary elimination, more likely if it presents soon after the introduction of cows’ milk protein in the diet		Idiopathic colic, developmental disorders, urinary tract infection	Cows’ milk elimination and re-challenge
	Constipation	Hours/days	Passage of infrequent and/or hard stools	Responds to dietary elimination, more likely if it presents soon after the introduction of cows’ milk protein in the diet		Hirschsprung’s disease, slow transit constipation	Cows’ milk elimination and re-challenge in conjunction with laxative treatment. Consider rectal biopsy in infants with early-onset severe constipation.

^* Atopic dermatitis as a sole manifestation of CMPA is debated in the literature, and it is important to establish whether other stimuli are potential triggers, such as laundry powder or soap

^** There is a lack of expert consensus as to whether eosinophilic oesophagitis is a primary non-IgE-mediated reaction, or a mixed immune reaction. However, most guidelines define it as non-IgE-mediated.

^† Higher IgE levels are predictive of a longer duration of CMPA. CMPA, cows’ milk protein allergy; GORD, Gastro-oesophageal reflux disease; IM, intramuscular.

1. Exclude cows’ milk

Avoidance of cows’ milk is the only treatment for CMPA.⁵ Regardless of the underlying cause and clinical type of CMPA, any dietary consumption of cows’ milk should be eliminated following diagnosis; this includes both direct sources in the infant’s diet (e.g. cows’ milk, cows’ milk formula or solid foods containing cows’ milk) and potentially maternal cows’ milk consumption if the infant is breastfed (see: “The role of breastfeeding in the management of CMPA”). For infants with mild-to-moderate CMPA symptoms, solid foods containing milk may potentially still be able to be consumed because in some cases the production process significantly diminishes the amount of allergen present.

A nutritional assessment should ideally be performed in primary care following any decision to remove cows’ milk protein from an infant’s diet (or the mother’s) to ensure the amounts of protein, calories and micronutrients (e.g. calcium and vitamin D) are adequate. However, calcium supplementation is usually not required for the infant.⁵ Other dietary sources of calcium should be encouraged provided they are age appropriate and the infant does not have a proven allergy to them, e.g. some breakfast cereals, tofu, dark leafy vegetables. Maternal calcium supplementation may be needed in some cases, which can be achieved with 800 mL per day of a calcium fortified non-cows’ milk alternative, e.g. rice or oat milk, or with oral calcium supplements.⁵

2. The role of breastfeeding in the management of CMPA

Exclusive breastfeeding is recommended in infants until age six months, at which point it is not sufficient alone for growth and development.¹⁷ Infants generally should continue to be breastfed until they are aged at least 12 months, however, solids should be progressively introduced from around the age of six months (not before four months).¹⁷

Infants with CMPA should continue breastfeeding if possible:^{1, 13}

• Exclusively breastfed infants – eliminate cows’ milk protein from the maternal diet to avoid transmission via breastmilk
• Mixed feeding infants – eliminate cows’ milk protein from the infant’s diet first; the mother may be able to continue consumption of milk in some cases of delayed reactions; if there are residual symptoms, or if the infant has an immediate CMPA reaction, maternal cows’ milk protein exclusion is recommended

If maternal exclusion of cows’ milk does not resolve symptoms, or if breastfeeding is problematic or not solely adequate, or if the mother wishes to wean the infant before they are aged 12 months, infant formula will be required and the mother supported in this decision. However, prior to switching to formula, ensure that other reasons for the infant’s symptoms have been considered and if the mother wishes to continue breastfeeding that she has received additional support to do so.

3. Selecting an appropriate infant formula

Only three types of formula are recommended for infants with CMPA: (1) soy formula (not funded), (2) extensively hydrolysed formula (eHF) and (3) amino acid formula (AAF).¹⁸ Both eHF and AAF are fully funded with Special Authority approval (separate application criteria). Other formulas such as goats’-milk based, lactose-free and partially hydrolysed formula are not suitable for infants with CMPA due to the potential for cross-reactivity against proteins with a similar structure, or direct immune responses against conserved proteins. Alternative “milk beverages”, such as rice, oat or almond milk, are nutritionally inadequate and therefore not recommended as a substitute for breast or cows’ milk in infants.⁵ N.B. They are also not recommended as the primary milk option for children aged under five years.

Extensively hydrolysed formula (eHF) is the first-line choice formula for most infants with CMPA

The age of the infant and the clinical characteristics of the CMPA should determine the type of formula most appropriate as an initial option. However, in the majority of cases, eHF is recommended as the first-line choice of infant formula for CMPA (Table 2).^{1, 13} Hydrolysis of cows’ milk breaks down potentially allergenic proteins into less-reactive peptides, making an immune response less likely. As a result, CMPA symptoms resolve in approximately 90% of infants that transition to eHF.⁵

Soy can be discussed as an alternative to eHF in infants aged over six months

Soy-based formula is not funded in New Zealand but is comparable in price to standard cows’ milk formula. In some cases, infants may find soy formula more tolerable than eHF and therefore its use can be discussed as a possible alternative, e.g. non-anaphylactic CMPA associated with gastrointestinal symptoms (Table 2).⁶ However, soy formula is not considered to be suitable for children aged under six months as rates of concurrent soy allergy are higher in this group.¹⁹

Amino acid formula (AAF) should be reserved for complicated or severe conditions

AAF is considered to be the closest formula option to being “non-allergenic” as it consists of individual amino acids.²⁰

Despite being highly tolerable, its current use as a funded formula is controversial as it is approximately three times more expensive to produce than eHF, and eHF will be sufficiently hypoallergenic in most infants.^{21, 22} Therefore, access to funded AAF should not be based on a preference; there needs to be a medical requirement for it (Table 2).

The indications for AAF in infants with CMPA include:²⁰

• Anaphylaxis
• Eosinophilic esophagitis
• Severe intolerance, allergy or malabsorption on eHF
• Growth faltering, particularly with multisystem involvement and multiple food exclusions

Currently, many infants are inappropriately prescribed AAF. In children aged 12 months or older, the expected prescribing ratio for eHF to AAF should be approximately 3:1, however, under the current funding criteria it is estimated to be an almost 1:1 ratio.²² Most children have a reduced requirement for milk once they are aged over 12 months and should be able to progress to solids as their primary source of nutrition, meaning that dependence on AAF can be avoided or reduced. However, the timing of this dietary transition will differ between children, and there are no clear recommendations for weaning off formula. It may be appropriate to consult with a paediatrician or dietitian with expertise in the management of CMPA for further advice.

4. Long-term re-introduction of cows’ milk

In most cases, CMPA is a self-limiting condition; resolving between the ages of one to three years in many children.⁴ Therefore, in the long-term it is important to regularly review and consider a cows’ milk challenge to avoid unnecessary dietary restriction.^{1, 13} This is required for AAF Special Authority funded renewals. Reassessment for tolerance should be on a case-by-case basis, but in general a review every six months is reasonable.^{1, 13} However, for some infants with more severe IgE-mediated CMPA this recommendation will not be suitable, and it is best to discuss a tailored re-introduction strategy with their paediatrician.

Table 2: Appropriate choice of formula feed in infants with CMPA syndromes in primary care.^{1, 6}

^* Soy formula is not funded but may be used as an alternative to eHF for some infants with mild CMPA symptoms

^† eHF must first be trialled first for 2–4 weeks and found to be inappropriate due to severe intolerance, allergy or malabsorption. AAF, amino acid formula; eHF, extensively hydrolysed formula; GORD, gastro-oesophageal reflux disease.