Hong Kong J. Dermatol. Venereol. (2008) 16, 77-91

Review Article

The application of allergy tests in childhood allergy

MHK Ho , EYT Chan , TL Lee , YL Lau

The prevalence of allergic diseases in childhood has increased considerably in developed nationsin the last 2-3 decades, and accordingly the need of allergy testing has increased steadily. HongKong is also facing similar challenge. Allergy testing is an important prerequisite for both earlyidentification of infants at increased risk of allergic disease and for specific allergy treatments. Inthis review, we shall discuss the allergy tests currently practised in Hong Kong. We have madepractical recommendations to local practitioner for testing allergy related dermatological problems.It is desirable to strengthen the co-operation between the specialist sector/hospitals and generalpractitioners in order to let the right children get the right tests at the right time.

Keywords: Allergy testing, CAP-FEIA, IgE, patch test, radioallergosorbent test

CAP E

Department of Paediatrics and Adolescent Medicine,Queen Mary Hospital, The University of Hong Kong,Hong Kong

MHK Ho, MBBS, FHKAM(Paediatrics)TL Lee, MBBS, FHKAM(Paediatrics)YL Lau, MD, FHKAM(Paediatrics)

Department of Pathology, Queen Mary Hospital, TheUniversity of Hong Kong, Hong Kong

EYT Chan, FRCPath, FHKAM(Pathology)

Correspondence to: Dr. MHK Ho

Department of Paediatrics and Adolescent Medicine,Queen Mary Hospital, 102 Pokfulam Road, Hong Kong

Introduction

The prevalence of allergic diseases in childhood(atopic eczema, asthma, allergic rhinitis/conjunctivitis, and food allergy) has increasedconsiderably in developed nations in the last 2-3decades, and accordingly the need of allergytesting has increased steadily. In population-basedstudies of Hong Kong,1-3 the estimated prevalenceof allergic diseases in school age children was25-30%. Atopic eczema was found in 8-10% andthe prevalence of asthma, food allergy, allergic

MHK Ho et al78

rhinitis and conjunctivitis was 7-10%, 2-4% and20-30% respectively. Many of these childrensuffered from co-morbidities of these allergicdisorders (Figure 1). In the setting of anuniversity affiliated referral allergy clinic forchildren; we have been seeing a four-foldincrease in demand on allergy testing duringthe last 5 years. In an analysis of the presentingcomplaints of 300 consecut ive pat ients(unpublished data from author), they mainly fellinto these categories: 1) suspected IgE-mediatedfood allergy (47%); 2) poorly controlled eczema(30%); 3) recurrent angioedema/chronicurticaria (15%); 4) poorly controlled rhinitis/asthma (4%); 5) drug/latex allergy (3%) and6) suspected anaphylaxis (1%). Although theywere undoubtedly subjected to referral bias, theymight partly reflect health seeking behaviour ofpatients and their parents, and perhaps, to alesser degree, the genuine increase in diseaseburden. Among these referrals, most were madeby general practitioners, family physicians andpaediatricians in both private and public sectorsbut fewer than 3% were actually made bydermatologists. One in five referrals from theprivate sector already had some allergy testsdone in one form or the other. In this review,we shall focus on the allergy tests currentlypractised in Hong Kong. We try to makepractical recommendations that are relevant tolocal practitioners with special interest in allergyrelated dermatological problems.

Difference between atopy and allergy

The word allergy was coined in the 19th centuryfrom Greek words 'allon argon' and it meant "toreact differently". Atopy is also derived from Greek.'Topos' means a place and 'atopos' means out ofplace. In allergy scientific literature, atopy isgenerally defined as the genetic trait that confersan increased risk of sensitization to commonenvironmental allergens. Currently, atopy isassessed by looking for evidence of allergicsensitization which literally means the subject hasdetectable IgE directed against relevant allergens,but that may not always imply allergy disease. Inthis review, allergy is defined as hypersensitivityinitiated by immunological mechanism which canbe antibody or cell-mediated.4 In most patients,the antibody typically responsible for an allergicreaction belongs to the IgE isotype, and is thuscalled IgE-mediated allergy. In non-IgE mediatedallergy, different mechanisms may be responsible(e.g. IgG, immune complexes, cell mediated, etc.).In patients with IgE-mediated allergic disease,upon exposure to a provoking allergen, a biphasicresponse would develop. The early responseoccurs within 20 minutes, being characterisedby classic IgE-mediated immediate typehypersensitivity, and is followed by the latephase response after 3-6 hours. This lateresponse is characterised by involvement ofeosinophilic inflammation and is associated withprolonged symptoms and hyper-reactivity. In

Figure 1. The phenotypic heterogeneityof atopy. The outer box represents ageneral population: approximately40% is atopic as defined by skintest positivity. Allergic rhinitis has apopulation prevalence of 20%. There isa considerable overlap between asthmaand rhinitis; however, not all asthmaticsare atopic. Atopic eczema has also anoverlap with asthma and allergic rhinitis.Please note that figures are estimationsfor a population of children and adults.

Allergy testing for children 79

case of daily exposure to allergen, persistentinflammation causes functional and structuralchanges. The relationship between atopy andallergy is thus a good example of gene-environment interactions.

Allergy march

Various manifestations of atopy/allergy diseaseoften present in a characteristic sequence, referredto as the 'atopic march'. Allergic manifestationsare rare during the first month of life. The initialsigns of allergic disease are atopic eczema andfood allergies, which typically peaks during thefirst 2 years of life. The development of food allergyclosely parallels the pattern of atopic eczema, withthe highest rates of food sensitization also notedin the first 2 years of life. The peak incidence offood allergy is 4% to 8% at one year of age,followed by a gradual decrease to 1-2% throughadu l thood . Cor re spond ing l y , spec i f i cimmunoglobulin E (IgE) antibodies against milkand egg are most frequently detected during thefirst 1-2 years of life. The development of allergicsensitization to inhalant allergens begins to riseat 3 years of age and is often accompanied byallergic respiratory diseases. Most children withasthma experience disease onset during the first5 years of life, whereas IgE against inhalantallergen is predominant in school age children.Interestingly, IgE antibodies to allergens (hen's eggwhite, cow's milk) in infants predict sensitizationto inhalant allergen and allergy before 7 years ofage.5

Biology of immunoglobulin E

Normally present at very low level in plasma, IgEantibodies are produced primarily by plasma cellsin mucosa-associated lymphoid tissue and theirlevels are commonly elevated in patients sufferingfrom atopic condition like asthma, allergic rhinitisand atopic eczema. Production of allergen-specificIgE in atopic individuals is driven by geneticpredisposition as well as by environmental factors,

including chronic allergen exposure. The lineagecommitment by B cells to produce IgE involvesirreversible genetic changes at the immunoglobulinheavy chain gene locus and is very tightlyregulated. It requires both cytokine signals(interleukin (IL)-4, IL-13) and interaction of B cellsCD40 with its ligand on activated T cells.

IgE antibodies exert their biologic functions viathe high-affinity IgE receptor, FcεRI, and the low-affinity receptor, CD23. In the classic 'immediatehypersensitivity' reaction, the interaction ofpolyvalent allergens with IgE bound to mast cellsvia FcεRI triggers receptor aggregation, whichinitiates a series of signals that result in the releaseof various vasoactive and chemotactic mediators.In addition, IgE antibodies have a number ofimmunomodulatory functions. These include up-regulation of IgE receptors, enhancement ofallergen uptake by B cells for antigen presentation,induction of Th2 cytokine expression by mast cellsand these may all collaborate to amplify andperpetuate allergic response in susceptibleindividuals. Thus blockade of IgE effect, usingnovel anti-IgE therapies, may ultimately prove tohave a broad effect.6

Why we need allergy testing?

Allergy testing is a very important prerequisite forspecific allergy treatment. It is useful for earlyidentifications of infants at risk for laterdevelopment of allergic diseases and specificallergy treatment. The latter includes specificallergen avoidance measures (e.g. dust miteallergen-prove cover, food avoidance), relevantpharmacotherapy (e.g. autoinjectable adrenaline-EpiPen®) and immunotherapy (conventionalsubcutaneous or novel sublingual desensitisation).7

Who should be tested for allergy?

Generally, all individuals with severe, persistingor recurrent possible allergic symptoms andindividuals with need for continuous prophylactic

MHK Ho et al80

treatment should be tested for specific allergyirrespective of the age of the child (Table 1). Therange of allergy tests typically depends on age ofthe child, positive family history and symptomssuch as possible seasonal or diurnal variations.

Asthma. Allergy testing should be performed,where an early intervention (elimination of relevantallergens/treatments) may improve disease controland prevent progression of disease.

Atopic eczema. Symptoms not under good controlby topical steroids or when food allergy issuspected should also be tested.

Urticaria is rarely caused by allergy except in caseswith close relation to intake of specific food orexposure to specific allergens.

Insect sting reactions. Virtually non-existing inlocal children but occasional happens in expatriatechildren. Children should only be tested for allergy

to insect venoms (bee or wasp) in case of severesystemic reactions. Local reaction and urticariaare not indication for testing.

Allergy to drugs. Allergy testing should beconducted for symptoms such as itchy skinreactions, urticaria, angioedema, asthma oranaphylaxis. Maculopapular exanthema is not anindication for testing.

Latex allergy. Allergy testing is primarily indicatedfor at risk children group, e.g. spina bifida,urogenital malformation with recurrent catheterinsertions, early major surgery, atopic subjects orother patients who have clinical symptoms uponexposure to latex.

Anaphylaxis. Patients with anaphylaxis need athorough, comprehensive allergy-immunologyevaluation to diagnose the specific aetiology. Itaims at preventing exposure to the allergens andproviding strategies for dealing with episodes of

Table 1. Indications for allergy testingGastrointestinal symptoms: vomiting, Persisting or intermittent symptoms without any other known reason

diarrhoea, colic, failure to thrive

Atopic eczema Persisting symptoms or allergen related symptoms, particularly incase of other concurrent atopic symptoms

Acute urticaria/angioedema Severe cases and/or suspicion of specific allergy

Children <3 years of age with recurrent Persisting severe symptoms and need daily treatmentwheezing/asthma Children with long lasting cough/wheeze/dyspnoea, particularly

during play/physical activity and during the nightChildren with reduced level of activity or frequent pneumoniae

without other known causes should be allergy tested

Children >3 years with asthma Should always receive allergy testing for relevant allergenShould be investigated for rhinitis

Rhinitis Treatment resistant casesShould be investigated for concurrent asthma

Conjunctivitis Treatment resistant cases

Insect sting reaction Only severe systemic reactions should be tested

Anaphylaxis Should always be evaluated for allergy under special observation

Contact dermatitis Treatment resistant dermatitis, hand dermatitis, unusual distribution(consider patch testing)

Allergy testing for children 81

allergic anaphylaxis. The physician and familyneed to implement a written action plan detailingthe early recognition of signs and symptoms ofanaphylaxis and the use of an epinephrine auto-injector for self-administration as pre-hospitaltreatment. Because fatal anaphylaxis occursdespite timely and appropriate treatment,successful avoidance strategies and educationremains the mainstay of management.8,9

How do we select allergen panels?

In rank order, house dust mites (Dermatophagoidespteronyssinus and Dermatophagoides farinae) andcockroach are the commonest inhalant allergensin Hong Kong.10-12 Pets (dog and cat dander)should be tested according to exposure. Commonfood allergens are not much different from thoseof Western data, with cow's milk common ininfants. Hen's egg, codfish and peanuts arecommon in preschoolers. Shellfish is relativecommon in adolescents or young adults.13 Otherfoods like wheat and tree nuts are quiteuncommon and should only be tested when thereis a relevant history. In Table 2, proposals ofrelevant allergens testing at different ages aredescribed. When children are tested according toage for specific allergy diagnosis, additionaltesting might be considered for assessment ofsensitization reflecting the atopic constitution.

Children less than 3 yearsIn early infancy, food allergy with manifestationsfrom skin, gastrointestinal or respiratory tract isrelative common. Atopic eczema and recurrentwheezing are the common presenting symptoms.Children with long-lasting cough/wheeze/dyspnoea, particularly during play/physical activityand during night and children with reduced levelof activity or frequent pneumonia without otherknown causes should be allergy tested. Childrenwith food allergy almost always show symptomsfrom two or more organ systems concomitantly,but in some cases, e.g. atopic eczema, the childrenmay only show one severe persisting symptom.Cow's milk allergy is the most common food

allergy in young children followed by allergy tohen's egg, peanut and fish. A few children aresensitized to indoor inhalant allergens alreadyduring the first 1-2 years of life and in case ofpersisting asthmatic symptom, allergy testing forrelevant indoor allergens such as mite and catmay be useful.

Children above 3 yearsWith increasing age, allergy to inhalant allergensdevelops, including indoor allergens (house dustmites, pets, cockroaches) and outdoor allergens(pollen and moulds). During childhood a highfrequency of sensitization is seen in individualswith asthma, rhinitis or conjunctivitis. A greatproportion of asthmatics (>70%) also suffers fromrhinitis, and a great proportion of individuals withrhinitis also suffers from asthma, though frequentlynot detected. Typical hay fever is seldom seen inHong Kong. We rarely need to test patients onpollens. Mold/fungi allergy is also infrequent inlocal children.

Cross-reactivitiesCross-reactivities occur when two or moreallergens share epitopes, or in some cases havevery similar epitopes, and therefore bind to thesame IgE-antibodies. Thus, patients sensitized toone of the allergens may also react to the otherwithout previous exposure and sensitization. Theknowledge about cross-reactivities is importantwhen evaluating the need and extent of allergytesting (Table 3).

Practical pearls

Although positive skin prick test (SPT) and/orpositive specific IgE in serum indicate that a personhas antigen-specific IgE, these findings do notprove that exposure to the allergen in questioncauses clinical allergic symptoms. Theoretically,only controlled allergen challenge can confirm thecause-effect relationship. However, in daily clinicalpractice, it is common to use a positive SPT andor/presence in serum of specific IgE to relevantenvironmental allergens and a suggestive clinical

MHK Ho et al82

history as proof of allergy induced disease. Thehigher the specific IgE antibodies, the stronger theassociation with clinical disease. SPT and in vitrotests may differ in their accuracy (sensitivity,specif ici ty and reproducibil i ty) to detectsensitization depending on the quality of theextracts used and the methodology. Patientsreferred to tertiary centre and/or specialists tendto have a higher prevalence of sensitization,multiple sensitization and more severe clinicalsymptoms. Thus, extrapolating the data directly

Table 2. Allergy testing according to age and disease/symptoms. The allergen panel should be adjustedaccording to allergen related symptoms and local allergen exposure, indoor as well as outdoor (in Hong Kongwe seldom has positive tests to outdoor allergens)

Disease/symptoms What to test in relation to ageEczema <3 years of age >3 years of age

Foods (for IgE food allergy Foods (in case with severeassociated eczema) persisting eczema)• cow's milk (most common <1 year) • cow's milk• egg white • egg white• peanut, cod fish, • peanut• if with suggestive history test: • if with suggestive history test:

wheat, nuts, soy wheat, nuts, fish, shellfish, soyinhalant allergen inhalant allergen(to test the atopic risk) (allergen-associated eczema)

• house dust mites • house dust mites• cat, dog and cockroach • cat, dog and cockroach

Persistent and intermittent For allergen-specific diagnosisrunny and stuffy nose • house dust mitesand/or wheezing • cat, dog and cockroach

• +/-molds• +/-pollen

Table 3. Cross-reactivities between allergens

Symptom related allergens Frequently cross-reactive allergensCow's milk Goat's milk, sheep's milk, beefPeanut Tree nuts, soy beans, green beansShrimp Other shellfish e.g. clams, crabs, lobstersDust mites Shellfish (with tropomysin)Grass Potato, tomato, wheat , peanutBirch Apple, apricot, plum, peach, hazelnut, carrot, potato, celery, cherry, pearLatex Kiwifruit, mango, papaya, avocado, banana, chestnutLentils Peanut, soy

from tertiary centre without due consideration oflocal setting is not very desirable. In communitypractice, where a low level of mono-sensitizationwithout a suggestive clinical history may beencountered, the clinical relevance may be limited.

Allergy testing includes the following elements:• Case history• Determination of IgE-sensitization

♦ SPT♦ Total and specific IgE in serum

Allergy testing for children 83

• Allergen challenge♦ Food allergy (in particular)

• Other optional tests (e.g. basophil histaminerelease assay, environmental investigations).These tests are mainly for research purposesor in the hands of specialists

Skin prick test

Percutanous/epicutaneous (prick or puncture) withallergen extracts is the favoured method of in vivotesting for IgE-mediated sensitivity. It is safe, notpainful and easy to perform. It is less sensitive butmore specific than intradermal route. Intradermalskin test for food allergy is now consideredinappropriate due to higher risk of systemicreaction. It is, however, reserved for evaluation ofdrug allergy because haptens of drugs are weakallergens.

There is no age limit for doing skin test in childrenbut the number of skin tests should be kept to theminimum, especially for younger children. Theprerequisite is avoid short and long actingantihistamine for 3 and 7 days prior the testrespectively. Potent topical steroid and calcineurininhibitors over the tested area should also beavoided for 1-2 weeks. Positive control (histamine,10 mg/ml) and negative control (glycerated saline)solutions are used. SPT are performed on the backin infants and on the forearm in older children.The skin wheal diameter is measured between 10to 15 minutes later. Irregular shapes are recordedas the mean of two perpendicular diameters.Attention should be paid to the quality of extractsand the characteristics of the device used. Theresults should be documented in a quantitativemanner that can be interpreted by otherpractitioners. We prefer to use single head lancetthan multi-adaptors. The lancet gives clearerresult, whilst multi-adaptor is easier to perform incase of a non-cooperative child. However it tendsto give more false positives due to its larger gridwhich inflicts a bigger trauma to skin.

Positive skin test results are also useful for

education purposes to the patient and the patient'sfamily, and may improve treatment compliance.A positive response is often referred to as a whealof 3 mm or larger. SPT is most informative whennegative because the negative predictive value(NPV) of the test is very high (>95%). Conversely,the positive predictive value (PPV) is between30% and 50%. Thus positive skin tests must becorrelated with clinical history and foodchallenges. The size of wheal depends on manyfactors such as level of specific IgE, binding affinityof the IgE antibody, releasability of the patients'mast cell, reactivity of patients' skin to histamine,area of body used for testing (with back beingmore reactive than the arm) and sometimes age(baby younger than 4 months may have falsenegative results).

With standardised allergen reagent andtechnique, the diameter of wheals in SPT correlateswith clinical sensitivity but not severity, i.e. thelarger the wheal , the more likely clinical reactionwould develop upon exposure but it does notpredict how severe the reaction will be.14-20 In theanalysis of fruit allergy, we sometimes need toapply prick-by-prick technique with fresh fruit as"testing reagent" because some of the fruit proteinsare quite labile and their allergic potential readilylose after even brief storage and processing.

Total IgE

It is used as an indicator of atopy but not by itselfa diagnostic marker for allergic disease. There isa wide overlap in the total serum IgE levelsbetween atopic and non-atopic populations (e.g.parasitic infestation, skin diseases other thaneczema, drug induced conditions, graft versus hostdisease, Hyper IgE syndrome etc.).

Quantification of serum allergen-specific IgE

This is the most important analyte measured inthe clinical immunology laboratory for diagnosis

MHK Ho et al84

of allergic disease. It is synonymous withradioallergosorbent test (RAST). Classically it is atwo-stage non-competitive radioimmunoassay.Allergen-specific antibodies are bound to anallergosorbent. Bound IgE antibodies are detectedwith radio-iodinated anti-human IgE. An analysisagainst a standard calibration curve is performedin each assay to quantify the allergen-specific IgE.Now after several generations of modification,most practising allergists and accreditedlaboratories favoured the use of a FDA-approvedsystem, namely CAP-FEIA or ImmunoCap (Phadia,Pharmacia, Uppsala, Sweden). It is in fact widelyvalidated in different clinical settings and ethnicpopulations. Quantitative IgE levels to selectedfoods (milk, egg, fish, and peanut), if above apre-defined IgE antibody threshold (Table 4), mayeliminate the need for food challenges.21-24 It isinterchangeable with SPT for inhalant allergen inmost of the circumstances.

Specific in vitro IgE immunoassays may bepreferable to skin testing for patients and theirfamilies who: 1) have severe dermographism,or generalised eczema, 2) unable to stopantihistamines, 3) refuse skin testing or cannotcooperate with testing and 4) have a clinicalhistory suggestive of a higher risk of anaphylaxiswith skin testing to a particular allergen.However, RAST is rather expensive and of limited

availability. In general, we preferred SPT thanRAST in daily clinical practice.

Multi-allergen test

This is a quantitative in vitro measurement of apanel of common food or inhalant allergens(e.g. Phadiatop®, Phadiatop infant®, fx5®). It isoften applied to young infants as screening ofatopic constitution.25 Generally, we think it is abetter tool than total IgE as an indicator of atopy.Prediction of asthma development at age of 4-5years has been made by utilising the sum of IgEantibodies level in combination with the numberof allergens that elicit positive tests results. It mayrepresent a more efficient diagnostic tool and itswider application is being actively sought. Onerecent study suggested it might be used as analternative to SPT in the epidemiological settingto diagnose atopy.26 The results of the tests were:sensitivity 85.0% (95%CI 82.2-87.8%), specificity85.5% (95%CI 82.7-88.3%), positive predictivevalue 72.7% (95%CI 69.0-76.1%), negativepredictive value 92.7% (95%CI 90.6-94.7%) andaccuracy 85.3% (95%CI 82.3-88.0%); with betterperformance among the symptomatic groups.New emerging commercial blot kits, which couldscreen out a large number of allergens with minuteamount of blood, are in fact not well studied.

Table 4. Diagnosis decision points of skin tests and serum specific IgE levels for clinical food reactivity

Study population Age (year) Food Cut-off level*Skin prick tests Wheal size (mm)Australian ≤2 Cow's milk 6

Egg 5Peanut 4

Australian ≤16 Cow's milk 8Egg 7Peanut 8

Serum food-specific IgE# KUa/LUSA ≤14 Cow's milk 32

Egg 6Peanut 15

*95% positive predictive value of positive food challenge, it varies with study population and currently no Asian data#by (Pharmacia-CAP-FEIA)

Allergy testing for children 85

Patch test

Clinical features may be insufficient to distinguishallergic from irritant contact dermatitis andendogenous eczema. Patch tests are used indelayed hypersensitivity reactions (i.e. allergiccontact dermatitis). Patch testing is indicated inpatient with treatment-resistant dermatitis, handdermatitis, stasis dermatitis, suspected allergy totopical medicaments (e.g. otitis externa) andsuspected occupational exposure. Eczema withunusual distribution should preferably seekspecialist (e.g. dermatologist) opinion priorto performing such test. Three importantconsiderations before we coin the diagnosis ofcontact dermatitis: a) the index case has the clinicalfeature of contact dermatitis, b) the index caseshave environmental exposure to the allergen inconsideration and c) the allergen identified bypatch test explains the clinical feature and course.Unselected patch testing in people with eczemacan have many caveats.

The test involves placing a series of allergens onthe skin and leaving them in place for 48 hours.The precise technique varies between differentcentres but experience is essential for correctinterpretation. Allergen is applied in aluminiumFinn chambers (a disc with a rim), usually on theback. Allergen can be applied in an inert vehiclesuch as petroleum jelly or in an aqueous solution.Substances in petroleum jelly are applied directlyto the disc solution or on filter paper. The chambersare held in place with non-allergenic tape(Micropore® or equivalent) and are left for 48hours. During this time the patient is instructed tokeep the area dry and not to remove or tamperwith the tests. Bathing and showering should beavoided until after the final reading. Some wouldre-examine the skin at 72 and 96 hours if there isno reaction at 48 hours. It gives semi-quantitativescale which is usually reported on a three pointscale (+, ++, +++). Toxic/irritant reactions andquestionable results should be documented. Mostcentres use a standard series of the most frequentallergens for all patients and add on other

allergens according to patient's clinical history andlikely exposure.

Data on patch test findings in Hong Kong arescarce. In a retrospective analysis of all patch testsperformed on patients with suspected allergiccontact dermatitis in the Social Hygiene Service,a total of 2585 patients were patch tested withthe European standard series from January1995 to December 1999.27 One or morepositive responses were noted in 55% ofsubjects. The most common allergens werenickel sulfate (24.4%), fragrance mix (13.7%),cobalt chloride (8.7%), p-phenylenediamine(6.0%) and balsam of Peru (5.7%). Nickelsensitivity was more common in female patientsand dichromate sensitivity was more commonin male patients. A positive atopy history was asignificant risk factor for contact allergicsensitivity. We should be cautious about the factthat some atopic subjects may have falsenegative result due to skin anergy.

In children, allergic contact dermatitis is relativelyuncommon. As an infrequent user, we prefer usinga commercial kit (True-test®) for patch testing.There is a standardised reading method accordingto the manufacturer. The main problem is thatthe applied test may quite easily slough off. Wehave solved it by additionally applying Tegaderm®or equivalent over it, providing that the patienthas no allergic history to such material.

Atopy patch tests have been used in recentstudies28 for identification of cell mediatedreactions to foods, particularly in children withatopic eczema. It uses similar technique as FinnChamber method with food allergen soaked in afilter paper. A diagnostic algorithm (atopy patchtest constitutes part of it) has been proposed byseveral international working groups (Figure 2).29

A ready-to-use atopy patch test (Diallertest®) isnow available in Hong Kong. It is a commercialkit with standard interpretation defined by themanufacturer. It is used to identify cow's milkdelayed hypersensitivity. Patient with positive

MHK Ho et al86

Figure 2. Diagnostic algorithm for food related hypersensitivities.

Allergy testing for children 87

results is thus recommended to be fed withextensive hydrolysed formula (e.g. Neocate®). Themanufacturer of this test is also the manufacturerof the recommended special formula. There is sofar only one controlled study sponsored by thecompany, showing a sensitivity of 76%, a specificityof 90% and 90% PPV in comparing atopy patchtest with food challenge.30 We believe that suchtest should be properly evaluated in the localpopulation before wide adoption.

Provocation test

The provocation test or challenge test is recognisedas the gold standard against all other in vivo orin vitro tests. Inhalation allergen challenges appliedvia route of ocular, intra-nasal, respiratory tractare rarely performed nowadays particularly inchildren. Food and drug challenges are relativelycommon. Food challenge is by far the mostcommon procedure carried out in the paediatricallergy service under the supervision of trainedmedical personnel. Food challenge procedureshould be properly validated and standardised inadministration and documentation. This involvesgiving a child increasing amounts of a food overa period of about several hours and observing forany objective clinical allergic response within asetting with resuscitative facilities. The differentapproaches to food challenges are listed in Table 5.

Other tests

Histamine release testThis test measures the histamine release frombasophil granulocytes. The results of histaminerelease test were comparable to the results of SPTand specific IgE-tests. Histamine release test ismore complicated for daily clinical practice, butmay be a helpful tool in certain cases, e.g. testingfor infrequent allergies or drug allergies.

Assessment of indoor allergensDetermination of allergen content in house dustsamples (house dust mites, pets) may be useful in

order to prove clinically important exposure to theinvestigated allergens and to monitor avoidancemeasures.31 But the main problem of this test isthe difficulty in standardising the collection andquantification methodology.

Allergy testing application in clinicalscenarios

Atopic dermatitis and food allergyIn our paediatric dermatology clinic, overwhelmingmajority of cases are being consulted for poorlycontrolled eczema. Issues concerning food allergyand allergy testing are commonly brought up byparents. Self-initiated avoidance/elimination of awide range of foods is not uncommonly seen andsome may actually render their child with problemof malnutrition and poor growth. From the allergyperspective, literature reports suggested about athird of children with moderate to severe atopiceczema was affected by food allergy, though somerecent reports claimed an even higher proportion.32

The younger the child and the more severe theeczema, the more likely the child is affected by foodallergy.33-36 Egg, milk and peanut allergy accountfor about 80% of food allergy diagnosed by foodchallenge in children with eczema and egg allergyis the most common. Appropriate diagnosis of foodallergy and elimination of the culprit may lead tosignificant clearing of eczematous lesions in themajority of food allergy-associated eczema. Infantswith eczema and egg allergy are at high risk fordeveloping respiratory allergy.

The diagnosis of food allergy in eczema iscomplicated by several factors related to thedisease: 1) the immediate response to ingestionof causal food is apparently down-regulated withrepetitive ingestion, making obvious "cause-effect"relations by history difficult to establish; 2) otherenvironmental trigger factors (other allergen,irritants, infection) may play a role in the waxingand waning of the disease, obscuring the effectof dietary changes and 3) patients have the abilityto generate IgE to multiple allergens which mayor may not be associated with clinical symptoms,

MHK Ho et al88

making diagnosis based solely on laboratorytesting impossible.

The best general approach is to screen childrensuffering from moderate to severe persistenteczema for food sensitization: <1 year to milkand <3 years of age to eggs, cod fish, peanutsby using SPT or RAST, with only additional testingfor other suspected food when we obtain therelevant history. If RAST reveals specific IgE tocertain foods, then determining the quantity offood-specific IgE antibodies (e.g. CAP-FEIA,Pharmacia) to the positive allergens would indicatewhether food challenge are necessary. After this,the best initial treatment is elimination of suspectedfood from the diet, followed by a food challengeif indicated. Because symptoms are chronic ineczema and often a large number of foods arealleged, it is often necessary to perform diagnosticoral challenge.

Urticaria and angioedemaCauses of acute urticaria can only sometimesbe determined and are likely to involve IgE-mediated reactions, viral infections or insectbites and stings. If there is no apparent triggerand temporal relationship with food exposure,the positive yield of doing skin test or RAST isactually quite low. Chronic urticaria (>6 weeks)is typically idiopathic. Physical stimuli (pressure,

rubbing, thermal, solar) often contribute to thesymptoms. Urticaria must be distinguished fromurticarial vaculitis and sometimes skin biopsymay be required. In 40 consecutive localchildren referred to us for evaluation of chronicurticaria with a median follow up of 26 months,about 60% has a prolonged remission. Theyhave been investigated by a standard panel ofallergy test (common inhalation allergen andfood allergen screening by SPT or RAST) andautoimmunity screening (ANA, C3 C4, anti-thyroglobulin, anti-microsomal antibody), andfewer than 5% has any positive test of the abovepanel. C1-inhibitor assay has been studied in1/3 of patients before seeking our opinion forrecurrent angioedema and none actually showpositive result. In fact it is quite unnecessary forsuch test to be done if the case has concomitanturticaria, so called "itchy angioedema". Weagree to screen C1-inhibitior in case of painfulangioedema, laryngeal oedema, prolongedoedema with poor response to antihistamineand adrenaline and those with positive familyhistory (data from author, manuscript inpreparation). C3/4 is an alternative screeningtest as C1-inhibitor quantification alone maymiss a substantial number of those functionaldeficient cases. Adult studies suggested thatcontact allergen37 and food additive38 seems tobe relevant in some of the chronic urticaria

Table 5. Different approaches of food challenges

Open challenge• Scientifically less vigor• Useful in situation to refute the suspected history in which the chance of allergy is low• Useful for infants and young child in whom subjective symptom is rarely a problem

Single-blind challenge• Very useful in daily clinical allergy practice• Less time consuming than double-blind-placebo-controlled challenge• Often provide an excellent diagnostic aid in confirming or refuting histories of hypersensitivity reactions, in

particular circumstances where patient opinions or concerns may influence the outcome.

Double-blind, placebo-controlled challenge• 'Gold standard'• Designed to reproduce the individual's signs and symptoms• Tedious and time consuming, may take days to complete

Allergy testing for children 89

patients. However, similar study in children iscurrently lacking. Our clinical impression is thatfood additive may not be very relevant. Themajority of parents among our paediatricpatients have been doing such measures butthis seems unhelpful.

Anaphylaxis, systemic mastocytosis and mastcell tryptaseAnaphylaxis is a potentially fatal multi-systemsyndrome as a result of massive release ofinflammatory mediators from mast cells andbasophils. Typically, the symptoms can becutaneous, respiratory, gastrointestinal and/orcardiovascular. Cutaneous symptoms are themost commonly occurring symptoms in acuteanaphylaxis but the absence of cutaneoussymptoms does not preclude the diagnosis ofanaphylaxis. Strictly, quantification of mast celltryptase is not an allergen test, but rather todifferentiate a true anaphylactic event fromother medical emergencies.39 Tryptase is themost abundant protein in mast cells. It is alsofound in basophils in much smaller amount.Tryptase proforms are continuously releasedfrom the mast cells into the bloodstreamreflecting the number of mast cells. Elevatedbaseline tryptase levels serve as a risk markerfor cer ta in pat ien ts to ge t in to severeanaphylact ic react ions, especial ly af terparenteral introduction of substances such asinsect venoms and drugs. Pathological increasedlevels of tryptase proforms reflect the mastburden in certain haematological abnormalities,neoplasms and the severity grade in systemicmastocytosis.40 Mature tryptase is released intothe blood stream during mast cell activation,either by IgE or non-IgE mediated mechanisms.The transiently increased level of mature tryptase(>10 µg/L), detectable 1 to 4 hours after theonset of systemic symptoms with hypotension,serves as a c l in ical marker conf i rminganaphylaxis. The biological half-life for tryptaseis about 2 hours. The return of baseline shouldbe verified.40,41 Since the test availability is quitelimited, prior consultation should be made withlaboratory immunologist.

Should I develop allergy testingservice in my office practice?

The primary care sector is in a state of flux at themoment, owing to reposition of health bureau/Hospital Authority's service target and paralleldiscussion about the future shape of the healthcaresystem including building up a children's hospital.Whatever shape the new system takes, we can besure that there will continue to be a large numberof patients with allergic conditions who will wantto know more about their condition, the substanceto which they are reacting and what they can doto improve matters. Ideally, this advice should beeasily accessible to the patient. It should bepossible to provide targeted information andrecommended care pathway for individualpatients and their advisors in the healthcaresystem. Anyone who is advising on allergenavoidance or other targeted intervention oughtreally to be offering a diagnostic service to guidetheir decision-making. We have been trainingnurses in doing skin testing and many nursesspecialising in food allergy and asthma havealready received training in skin testing and allergyadvice. However the public sectors are reluctantto deploy them in fulfilling this task as thereare other calls on their time and no directreimbursement.

Skin testing is useful and relatively cheap, but maynot be an appropriate option for a practice thatonly tests a few patients each year. Easy access toand sensible use of immunological tests may bea better option for low-volume users (Table 6).Partnership with secondary/tertiary centre withallergy testing service is economically sound. It isrecommended to ensure and strengthen co-operation between specialist sector/hospitals,general practitioners and local home carer inorder to benefit the individual patient. The sharedcare covering both primary and secondary/tertiarysectors should ensure that the right children willget the right tests at the right time, along with thebest evidence-based treatment of their allergicdisease. Specialist referral should be made if thepatient failed to respond to standard therapy, or

MHK Ho et al90

where there are possible serious clinical or medico-legal issues (e.g. allergic reaction duringanaesthesia and potential anaphylactic reaction).High risk drug and food challenge should becarried out in appropriate clinical settings withtrained personnel and resuscitative facility as back-up. Specialist advice should also be sought beforestarting interventions such as immunotherapy thatmay have a major impact on patients' quality of life.

Conclusions

Allergy testing is an important pre-requisite forboth early identification of infants at increasedrisk of developing allergic disease and for specificallergy treatment. We recommend to strengthenthe co-operation between specialist sector/hospitals and general practitioners so that the rightchildren will get the right tests at the right time,along with the best evidence-based treatment oftheir allergic disease.

References

1. Worldwide variation in prevalence of symptoms ofasthma, allergic rhinoconjunctivitis, and atopic eczema:ISAAC. The International Study of Asthma and Allergiesin Childhood (ISAAC) Steering Committee. Lancet 1998;351:1225-32.

2. Asher MI, Montefort S, Björkstén B, Lai CK, StrachanDP, Weiland SK, et al. Worldwide time trends in theprevalence of symptoms of asthma, al lergicrhinoconjunctivitis, and eczema in childhood: ISAACPhases One and Three repeat multicountry cross-sectional surveys. Lancet 2006;368:733-43.

3. Lee SL, Wong W, Lau YL. Increasing prevalence ofallergic rhinitis but not asthma among children in HongKong from 1995 to 2001 (Phase 3 International Studyof Asthma and Allergies in Childhood). Pediatr AllergyImmunol 2004;15:72-8.

4. Muraro A, Dreborg S, Halken S, Høst A, Niggemann B,Aalberse R, et al. Dietary prevention of allergic diseasesin infants and small children. Part II. Evaluation ofmethods in allergy prevention studies and sensitizationmarkers. Definitions and diagnostic criteria of allergicdiseases. Pediatr Allergy Immunol 2004;15:196-205.

5. Hattevig G, Kjellman B, Björkstén B. Clinical symptomsand IgE responses to common food proteins andinhalants in the first 7 years of life. Clin Allergy 1987;17:571-8.

6. Ogettgen HC, Geha RS. Regulation and biology ofimmunoglobulin E. In: Leung DYM, Sampson HA, GehaRS, Szefler SJ, ed. Pediatric Allergy Principles andPractice. Philadelphia: Mosby, 2003:39-49.

7. Greenberger PA, Ballow M, Casale TB, Platts-Mills TA,Sampson HA. Sublingual immunotherapy andsubcutaneous immunotherapy: issues in the UnitedStates. J Allergy Clin Immunol 2007;120:1466-8.

8. Simons FE, Frew AJ, Ansotegui IJ, Bochner BS, GoldenDB, Finkelman FD, et al. Risk assessment in anaphylaxis:current and future approaches. J Allergy Clin Immunol2007;120(1 Suppl):S2-24.

9. Bock SA, Muñoz-Furlong A, Sampson HA. Furtherfatalities caused by anaphylactic reactions to food,2001-2006. J Allergy Clin Immunol 2007;119:1016-8.

10. Yuen AP, Cheung S, Tang KC, Ho WK, Wong BY, CheungAC, et al. The skin prick test results of 977 patientssuffering from chronic rhinitis in Hong Kong. Hong Kong

Table 6. Determination of allergen specific IgE

Skin testing SerologyMethod Prick test RASTAvailability Limited WidelyExpense Cheap ExpensiveResults Immediate DelayedRisk of anaphylaxis Rare Nil

AntihistamineInterference Extensive eczema High total IgE

DermatographismSensitivity 4+ 2+Specificty 4+ 2+PPV 2+ 2+NPV 4+ 3+Abbreviation: RAST=radioallergosorbent test; PPV=positive predictive value; NPV=negative predictive value; ++++extremely good; +++ very good; ++ good

Allergy testing for children 91

Med J 2007;13:131-6.11. Leung R, Ho P, Lam CW, Lai CK. Sensitization to inhaled

allergens as a risk factor for asthma and allergicdiseases in Chinese population. J Allergy Clin Immunol1997;99:594-9.

12. Leung TF, Lam CW, Chan IH, Li AM, Ha G, Tang NL, etal. Inhalant allergens as risk factors for the developmentand severity of mild-to-moderate asthma in Hong KongChinese children. J Asthma 2002;39:323-30.

13. Wu AY, Williams GA. Clinical characteristics and patternof skin test reactivities in shellfish allergy patients in HongKong. Allergy Asthma Proc 2004;25:237-42.

14. Ho MH, Heine RG, Wong W, Hill DJ. Diagnostic accuracyof skin prick testing in children with tree nut allergy. JAllergy Clin Immunol 2006;117:1506-8.

15. Ho MH, Wong WH, Heine RG, Hosking CS, Hill DJ,Allen KJ. Early clinical predictors of remission of peanutallergy in children. J Allergy Clin Immunol 2008;121:731-6.

16. Hill DJ, Heine RG, Hosking CS. The diagnostic value ofskin prick testing in children with food allergy. PediatrAllergy Immunol 2004;15:435-41.

17. Sporik R, Hill DJ, Hosking CS. Specificity of allergenskin testing in predicting positive open food challengesto milk, egg and peanut in children. Clin Exp Allergy2000;30:1540-6.

18. Astier C, Morisset M, Roitel O, Codreanu F, JacquenetS, Franck P, et al. Predictive value of skin prick testsusing recombinant allergens for diagnosis of peanutallergy. J Allergy Clin Immunol 2006;118:250-6.

19. Hill DJ, Hosking CS, Heine RG. Clinical spectrum offood allergy in children in Australia and South-East Asia:identification and targets for treatment. Ann Med 1999;31:272-81.

20. Hill DJ, Hosking CS, Reyes-Benito LV. Reducing the needfor food allergen challenges in young children: acomparison of in vitro with in vivo tests. Clin Exp Allergy2001;31:1031-5.

21. Shek LP, Soderstrom L, Ahlstedt S, Beyer K, SampsonHA. Determination of food specific IgE levels over timecan predict the development of tolerance in cow's milkand hen's egg allergy. J Allergy Clin Immunol 2004;114:387-91.

22. Sampson HA, Ho DG. Relationship between food-specific IgE concentrations and the risk of positive foodchallenges in children and adolescents. J Allergy ClinImmunol 1997;100:444-51.

23. Sicherer SH, Sampson HA. Peanut allergy: emergingconcepts and approaches for an apparent epidemic.J Allergy Clin Immunol 2007;120:491-503.

24. Knight AK, Shreffler WG, Sampson HA, Sicherer SH,Noone S, Mofidi S, et al. Skin prick test to egg whiteprovides additional diagnostic utility to serum egg white-specific IgE antibody concentration in children. J AllergyClin Immunol 2006;117:842-7.

25. Ballardini N, Nilsson C, Nilsson M, Lilja G. ImmunoCAPPhadiatop Infant--a new blood test for detecting IgEsensitisation in children at 2 years of age. Allergy 2006;61:337-43.

26. Garcia-Marcos L, Sanchez-Solis M, Martinez-Torres AE,Lucas Moreno JM, Hernando Sastre V. Phadiatopcompared to skin-prick test as a tool for diagnosingatopy in epidemiological studies in schoolchildren.

Pediatr Allergy Immunol 2007;18:240-4.27. Lam WS, Chan LY, Ho SC, Chong LY, So WH, Wong

TW. A retrospective study of 2585 patients patch testedwith the European standard series in Hong Kong (1995-99). Int J Dermatol 2008;47:128-33.

28. Heine RG, Verstege A, Mehl A, Staden U, Rolinck-Werninghaus C, Niggemann B. Proposal for astandardized interpretation of the atopy patch test inchildren with atopic dermatitis and suspected foodallergy. Pediatr Allergy Immunol 2006;17:213-7.

29. Høst A, Halken S, Muraro A, Dreborg S, Niggemann B,Aalberse R, et al. Dietary prevention of allergic diseasesin infants and small children. Pediatr Allergy Immunol2008;19:1-4.

30. Kalach N, Soulaines P, de Boissieu D, Dupont C. A pilotstudy of the usefulness and safety of a ready-to-use atopypatch test (Diallertest) versus a comparator (FinnChamber) during cow's milk allergy in children. J AllergyClin Immunol 2005;116:1321-6.

31. Sun BQ, Wu A, Chan A, Chik S, Wong D, Zhong NS.House dust mite allergen (Derp1 and Blot5) levels inasthmatics' home in Hongkong. Chin Med Sci J 2004;19:185-8.

32. Hill DJ, Heine RG, Hosking CS, Brown J, Thiele L, AllenKJ, et al. IgE food sensitization in infants with eczemaattending a dermatology department. J Pediatr 2007;151:359-63.

33. Hill DJ, Hosking CS, de Benedictis FM, Oranje AP,Diepgen TL, Bauchau V, et al. Confirmation of theassociation between high levels of immunoglobulin Efood sensitization and eczema in infancy: aninternational study. Clin Exp Allergy 2008;38:161-8.

34. Lowe AJ, Hosking CS, Bennett CM, Carlin JB, AbramsonMJ, Hill DJ, et al. Skin prick test can identify eczematousinfants at risk of asthma and allergic rhinitis. Clin ExpAllergy 2007;37:1624-31.

35. Lowe AJ, Abramson MJ, Hosking CS, Carlin JB, BennettCM, Dharmage SC, et al. The temporal sequence ofallergic sensitization and onset of infantile eczema. ClinExp Allergy 2007;37:536-42.

36. Hill DJ, Hosking CS. Food allergy and atopic dermatitisin infancy: an epidemiologic study. Pediatr AllergyImmunol 2004;15:421-7.

37. Guerra L, Rogkakou A, Massacane P, Gamalero C,Compalati E, Zanella C, et al. Role of contactsensitization in chronic urticaria. J Am Acad Dermatol2007;56:88-90.

38. Di Lorenzo G, Pacor ML, Mansueto P, Martinelli N,Esposito-Pellitteri M, Lo Bianco C, et al. Food-additive-induced urticaria: a survey of 838 patients with recurrentchronic idiopathic urticaria. Int Arch Allergy Immunol2005;138:235-42.

39. Schwartz LB. Diagnostic value of tryptase in anaphylaxisand mastocytosis. Immunol Allergy Clin North Am 2006;26:451-63.

40. Brockow K, Jofer C, Behrendt H, Ring J. Anaphylaxis inpatients with mastocytosis: a study on history, clinicalfeatures and risk factors in 120 patients. Allergy 2008;63:226-32.

41. Simons FE, Frew AJ, Ansotegui IJ, Bochner BS, GoldenDB, Finkelman FD, et al. Risk assessment in anaphylaxis:current and future approaches. J Allergy Clin Immunol2007;120(1 Suppl):S2-24.