Food avoidance strategies in eosinophilic oesophagitis

4.1 Elemental diets

4.2 Elimination diets

Elimination diets are defined as therapeutic interventions in which one or more foods are restricted. The most common foods that cause Eoe are milk and wheat, followed by soy and egg as listed in Table 2. In EoE, we still talk about food and not single food antigens, as the specific food antigens responsible for EoE inflammation have not been identified. For example, it is under debate if wheat-free diet means wheat free or need to be gluten free.64 At present, advice is also to avoid foods in both their fresh and baked forms. Various empirical and test directed strategies have been studied to establish standardized ways to perform an elimination diet. The efficacy of a dietary change is determined entirely by assessment of histologic remission on oesophageal biopsy and control of EoE symptoms. Symptoms alone have been shown to be not reliable in assessing inflammation activity.65, 66 Various types of elimination have been used and described, with similar efficacy in both adults and children. In general, more foods are eliminated more likely is to have remission at first OGD, but most individuals are allergic really to only 1-3 foods.34 The different dietary options are typically presented to the patients.

Step-down (start from a more restrictive diet and reintroduce progressively all the eliminated foods) and step-up (eliminate 1-2 foods and if not effective progressively eliminate more foods) approaches are both acceptable. Consideration on specific patient characteristics when deciding to use either of those approaches are the following: severity of the disease, nutritional status, eating habits, ability to cook, economical resources, necessity to eat out of the house and reliance of pre-prepared foods. In an adult patient with significant disease (significant trouble swallowing, frequent vomiting etc.), with a normal to high BMI, a baseline diet rich in fruit and vegetable, time and resources to prepare own food a more restrictive diet may be easy to implement and will have higher chance to induce remission at first OGD. However, a step-up approach may be preferable in an individual with mild-moderate symptoms of long duration, who has a limited non-various diet rich in milk and wheat, has low acceptance for new foods, or relies on pre-prepared food and restaurant eating for most meals.

Various types of diets have been described in the literature in prospective and retrospective studies.

5.1 Milk avoidance

The milk antigen that causes EoE is not known. In general, following similarities with IgE-mediated food allergies all mammalian milk and partially and extensively hydrolysed formulas should be restricted.37, 82-84 During milk avoidance, all milk sources included baked milk should be avoided especially when the diet and the food triggers need to be established.43, 87 Later, when is known that the patient is allergic to milk, baked milk can be tried in the diet followed by OGD, as some patients maybe able to tolerate it.43, 87 Enriched soy milk can substitute for cow's milk, as it provides protein, calcium, vitamin D and riboflavin.84 As a general principle, caution needs to be exercised when substituting a frequent trigger such as milk with another frequent culprit like soy. Authors have seen some patients switch to soy milk develop EoE from soy unrecognized prior as diet had limited soy (unpublished data). Enriched tree nut based beverages are also available. However, they would not be permitted on an SFED. In this instance, enriched “milks” or “beverages” made from rice, coconut, oat, potato, flax/sunflower and pea may be considered.82-84 A variety of enriched products can also help meet calcium requirements for some patients. Most of these are lower in calories and protein than whole cow's milk, and the remainder of the diet should be carefully reviewed to ensure that nutritional requirements can be met with other foods.82-84 These products are not appropriate for younger children especially under the age of 2 years; severe malnutrition and rickets have been reported in toddlers provided rice beverages as a substitute for cow's milk.88 Table 7 provides information on some currently available dairy alternatives.

5.2 Wheat avoidance

Wheat and enriched wheat products provide iron, niacin, riboflavin, thiamin, folate and fibre. Wheat-free grains are not usually enriched with the above vitamins and minerals, and use of refined flours can adversely affect intake of dietary fibre.82-84

Whole-grain wheat substitutes such as amaranth, buckwheat, gluten-free oats, quinoa, teff, millet and brown rice are better sources of dietary fibre and micronutrients than refined wheat-free flours.82-84 Patients should also be encouraged to increase fruits and vegetables as able.

The question of whether to remove wheat or all gluten-containing grains (wheat, rye, barley and standard oats, which does not contain gluten but can be contaminated by gluten if processed in the same facility as gluten-containing grains) has been raised.64 There are currently no studies suggesting improved histologic outcomes with the latter approach. Although it was unclear from past studies if participants were actually eating rye, barley and standard/non-gluten free oats, as for example in the United States most processed food that is wheat free is also gluten free.64

5.3 Egg avoidance

In an egg-free diet, both egg whites and egg yolks must be avoided. Baked egg should be removed as well, especially while the diet is being established. Eggs provide protein, choline, vitamin A, riboflavin, pantothenic acid, biotin and selenium. Most of these can usually be provided by other foods, depending on what remains in the diet after egg is removed.82-84 Eggs are an important ingredient in baked goods, providing leavening and structure. Instructions on how to replace eggs in baking should be provided to the patient, for example, egg can be replaced with apple sauce, mashed banana and flax seeds.82-84 Vegan alternatives (eg VeganEgg) based on algae are also suitable alternatives for baking and making scrambled eggs. Cholesterol-free egg substitutes such as Egg Beaters (ConAgra Foods, Inc., Omaha, NE, USA) contain egg whites and are not appropriate for use.

5.4 Soy avoidance

Soy provides protein, thiamin, riboflavin, pyridoxine, folate, calcium, phosphorus, magnesium, iron and zinc. Its removal likely presents minimal nutritional risk unless it has been a significant part of the diet pre-diagnosis (ie milk-allergic infant or child on soy formula, vegetarian family). Refined soy oil and soy lecithin are tolerated by most soy-allergic individuals and are typically not avoided on a soy-elimination diet.84, 89

5.5 Peanut, tree nuts, fish and shellfish

Peanuts and tree nuts can be replaced with seeds (and seed butters). Major nutrients in fish and shellfish can be replaced with meats, along with seeds and vegetables oils high in omega-3 fatty acids.82-84

5.6 Label reading

Label reading will become a crucial skill for patients and all involved in food selection and preparation during an elimination diet. The specific legislation covering both the United States and the European Union is summarized in Table 8. These laws apply to conventional food products, dietary supplements, infant formulas and medical foods. Manufacturers must also list the specific tree nut, fish or crustacean shellfish used as an ingredient. Major food allergens must also be declared in spices, flavourings, colourings or additives, or if used to aid in processing. These regulations apply only to ingredients derived from the allergens listed in Table 8. Individuals who need to avoid ingredients not covered under the regulations must contact the manufacturer to confirm product safety. Ingredients and manufacturing processes can change over time, and labels should be read each time a food is purchased or consumed, even if the food has been used safely in the past.

An ingredient used in some European products is codex wheat starch, which is rendered gluten free through extensive washing. It still contains other allergenic wheat proteins and therefore needs to be avoided during a wheat elimination diet for EoE. As per EU regulations, wheat will still be emphasized on the ingredients list, and education should be provided to ensure that gluten-free products are also checked to be wheat free.

In both the United States and EU, precautionary allergen labelling (PAL or “may contain” statements) is not a legal requirement and therefore the risk of contamination from allergens from foods bearing these warnings is variable. Accordingly, there is also a risk of cross-contamination on foods without any warnings.90 Individualized advice should be provided to families regarding PAL, particularly in the context of IgE-mediated food allergies. It is not yet known at this time whether trace exposure can trigger EoE but it is prudent at least in the initial stages to minimize all exposure.83

5.7 Nutritional aspects relating to children

Dietary avoidance, particularly long-term avoidance, can lead to nutritional inadequacies.91, 92 Nutritional deficiencies in food-allergic conditions have been well described in food allergy, indicating deficiencies in calories, protein, fat, vitamins (particularly B vitamins) and minerals (particularly calcium and iron).93 A recent study94 indicates that a number of nutrients may be suboptimal in children with food allergies such as protein, carbohydrate, zinc, calcium, iron and vitamin E.

As mentioned, milk is now considered the number one trigger food in EoE. Milk and dairy products are the main sources of calcium and contribute to vitamin D intake in most people's diets. Either milk substitutes with added calcium or a calcium supplement should be suggested in these cases. A recent study from the UK, focusing on children with non–IgE-mediated food allergies,95 concluded that these children are at risk of vitamin D insufficiency and deficiency. Similar data were reported in terms of suboptimal intake of Vitamin D and calcium in children with EoE from the USA, reporting that Vitamin D and calcium intake (mean intake of 606 mg/d) was below the daily recommended intake (mean intake of 112 IU/d),82 despite normal serum levels above 30 ng/mL. None of these children were previously assessed by a RD, despite more than a 12 months difference between the mean age of diagnosis (3.4 years) and age of recruitment (4.8 years).

In one study, 10% children with EoE, presented with moderate malnutrition at baseline96 but no further weight loss was reported during the dietary intervention. Two studies from the same US-based centre indicated that previously 21% of children with EoE were diagnosed with failure to thrive,85 and a more recent study indicates that the BMI z score for children with EoE was 0.13 (45th percentile).82

5.8 Nutritional aspects relating to adults

Food allergy data from adult studies indicate insufficient intake of calcium, and that intake of energy, protein, lipid, phosphorus, zinc, vitamin B1, B2, niacin and cholesterol decreased as the number of food allergens increased. Intake of vitamin A, iron and vitamin B6 was also compromised.91, 97 A recent study from the United Kingdom however indicated that this low intake of nutrients may not be particularly related to food allergies, but general suboptimal nutrient intake in adolescents and adults; even taking nutrient supplements did not correct nutrient in (100% of the United Kingdom recommended nutrient intake]) for nutrients such as potassium, selenium and magensium.98

To date, few studies of dietary intervention have focused on nutritional outcomes in adult EoE. One major concern is that dietary restriction can lead to weight loss. Two studies of an elemental diet reported weight loss,45, 46 and one trial of the SFED reported no significant change in weight.24 However, none of these studies report baseline BMI of participants and other dietary intervention studies do not mention BMI or weight loss at all. A recent UK pilot study found that there was a statistically significant reduction in BMI postdietary intervention; however, for no participants were weight loss considered clinically significant and no BMI dropped below the normal range.99

A comparison of nutrient intake between food-allergic adults and non-allergic controls found micronutrient deficiencies in both groups thought to be due to poor food choices, with the food-allergic group faring slightly better.100 Nutritional deficiencies have also been shown to occur during dietary intervention for EoE; however, in this study these were also present at baseline and for all participants improved during the intervention.99 This is thought to be due to dietary counselling as well as decreased intake of convenience foods. Given the restrictive nature of prolonged elimination diets, further investigation into nutritional parameters is warranted.

In adults, lifestyle or social factors may affect adherence to dietary intervention, such as work, socializing and travel. Difficulty with adherence in particular with an elemental diet is reflected by high withdrawal rates in adult studies45, 46 and the overall low proportion of studies that have been completed in adults compared to children.47 Access to an experienced dietitian may also be a limiting factor due to lack of adult specialists in this area.

5.9 Ongoing evaluation

Dietitians should play a role in the assessment of nutritional status by taking a detailed allergy focused diet history, perform and interpret anthropometric measurements, and consider laboratory tests to confirm or rule out suspected nutritional deficiencies (albumin, total protein, electrolytes, blood urea nitrogen and creatinine). Micronutrient assessment of serum vitamin D (25-hydroxy vitamin D), iron (serum ferritin, iron and total iron binding capacity), zinc, vitamin B12, selenium, and folate levels and others may be suggested.83 These assessments should ideally be carried out every 6-12 months and more often when a deficiency was confirmed and is being treated. Adequate nutrition should ideally be provided by food rather than nutritional supplements. However, in some cases where it is difficult to find suitable alternative (allergen-free) foods due to limited access to foods or feeding aversions or food refusal, the use of nutritional supplements and amino acid-based formulas/beverages should be considered.

As we know more about the disease, we have learned that EoE can change over time. Even when a patient is on a stable effective diet period, OGD (once every 1-2 year) is recommended to be sure that the remission continues. Indeed, for example, dietary change may lead to increase consumption of allergenic foods rarely consumed before starting the diet (ie soy, pea proteins) that may become allergenic and induce EoE or patient adherence may reduce. On the other hand, occasionally patients can outgrow their food allergies so every few years reintroduction of the allergenic foods may be tried.101

Nutritional risks should decrease as foods are successfully reintroduced, and dietary variety expands. Periodic re-assessment of the diet via 3-day food record can be helpful in adjusting micronutrient supplementation. Families must be educated appropriately to enhance success with adherence to specific diet plans. Children must be monitored closely to ensure nutritional needs are met and adequate growth occurs.102-104 Quality of life can be negatively affected by elimination diets.105, 106 The impact of nutritional therapy for both the patient and family should be assessed on an ongoing basis. We mostly monitor blood levels only of Vitamin D as the level of other micronutrients has not shown to reflect real nutritional status

Periodic yearly Dexa scan is often used in prepubertal patients, postpubertal women and in all patient at risk of osteoporosis to monitor bone health, especially in patient on milk-free diet and on drugs such as PPI and steroids that can negatively affect bone mass accrual. No guidelines exist on the use of Dexa; therefore, its use is based on clinical judgment and feasibility of the procedure.

Ongoing monitor should also focus on compliance, and lack of compliance is frequent in dietary treatment not associated with acute symptoms.82

5.10 Reintroduction

Once a patient has a normal endoscopy, foods are reintroduced either singly or in groups to try to determine culprit(s). After each food is reintroduced, an OGD must be performed. The decision regarding which food(s) to reintroduce and in which order is often a clinical decision. Starting from either low- or high-risk foods are both acceptable options, and family preference often directs different reintroduction approaches.

In general, for high-risk food (milk, wheat, soy, egg) which are also the most nutritionally important one food at the time reintroduction is recommended followed by OGD after 6-12 weeks. One to three medium-risk (legumes, meats, seafood, peanut/nuts and grains) foods can be reintroduced at one time before repeating OGD depending of previous patient history. Typically, low-risk foods can be reintroduced in groups (fruit and vegetables)

5.11 Comorbidity with IgE-mediated food allergies

Patients with EoE have in 20%-30% of cases IgE-mediated food allergies,31 which add significant complexity in dietary management of EoE.

Nutritional deficit and difficulties in achieving food diversity increase as more foods are restricted. Children with IgE-mediated food allergy that reintroduce a food as they outgrow IgE-mediated food allergy or because they underwent immunotherapy for such food are at risk of developing EoE,48, 107 and children who have serum IgE specific for foods allergens but are able to tolerate the food may develop IgE-mediated food allergies including anaphylaxis after a period of prolonged avoidance108, 109 Therefore, each individual with a possible food allergy (at the present time or in the past) undergoing a diet for EoE should be evaluated by an allergist.

Eosinophilic oesophagitis that develops after previously outgrown IgE-mediated food allergy or oral immunotherapy107, 110 may explain why for high-risk foods like milk and egg the positive predictive value (but not the Negative predictive value) was elevated.75, 76 SPTs therefore may be useful in EoE patients with a possible past history of milk or egg allergy, as positives may point to a previous IgE-mediated food allergy to those foods that may now trigger EoE.

5.12 What if dietary treatment fails?

If a 6-food elimination diet is unsuccessful, one possible approach is to look at the patient's current diet. If there is a large consumption of rich in protein foods such as legumes (including pea protein now a popular cow's milk substitute) or meats, removal of those foods is advisable, before calling the patient non-responsive to dietary intervention.

However, about 20% of Children and adult treated with elimination diet do not respond to dietary therapy, even after step-up approach to remove mostly eaten foods and SFEDs approach is used.43, 46 When this happens is paramount to establish adherence to diet therapy, which is known to be poor in EoE especially when diet is used for long periods of time. In general, compliance is a significant problem when patients feel better in any therapeutic approach, even if there is no data on long-term compliance with diet or steroids. However, delayed in starting therapy leads to increased stenosis with every 10 years of age at diagnosis, the odd ratio increases 2-fold for strictures and 7-fold for dysphagia, suggesting that compliance should be encouraged to avoid long-term consequences.111

Once compliance is established, two options can be offered to patients. One is to start combination therapy with diet and PPI (especially if PPI were not used initially as first line therapy) (Figure 1), and second is to add steroids and slowly reintroduce all the eliminated food to see if steroids alone can maintain disease remission.36, 37

5.13 Future in EoE food allergy

Despite the clear evidence for efficacy of diet in EoE, the pathological process that leads to food allergy in EoE is not completely elucidated. The mechanism of food allergy-induced inflammation is probably complex and that is why in vitro testing for food allergies have been disappointing.

Measurement of IgG4 and T cell in vitro activation with specific food allergens23 is under active investigation to establish accuracy in predicting EoE-specific food allergies.49

IgG4 has been found elevated at the site of inflammation in EoE suggesting that it could be an IgG4-mediated disease.49 However, IgG4 is antibodies produced during a Th2 inflammation and have demonstrated the ability to reduce such inflammation as they are not able to effectively stimulate an immune response when bound to an antigen and they reduce the ability of other antibodies to bind to the antigen.51 Specific IgG4 to food responsible for EoE has been also been detected, but there is a large overlap with levels in healthy individuals, and therefore, they cannot be used clinically.49, 51

T cells from patients with EoE triggered by milk when stimulated with milk either in vitro or in vivo express high level of activation marker CD154 + , suggesting potential use in the future to predict specific food allergy in single individuals, but prospective studies will be needed to establish specificity and sensitivity of the assay for specific food allergens.23

Recently, Warners et al112 injected during OGD six allergens (wheat, milk, soy and three allergens based on the patients’ history) in the oesophagus of eight patients with EoE and three patients without EoE (controls). Five of the eight patients with EoE had evidence for an acute response (luminal obstruction and mucosal blanching); two other patients had a delayed wheal or flare reaction, during an endoscopy obtained 24 hours after the first one. No responses were observed in controls. The authors conclude that oesophageal mucosal food allergen injections induce acute and/or delayed responses in patients with EoE but not controls. No clinical relevance of such positive oesophageal prick test was established in the study112

Finally, minimally invasive and cheaper methods to obtain oesophageal tissue such as the oesophageal string test113 or cytosponge test114 are under active investigation and will make empiric diet more feasible if validated.