Recombinant allergen-based provocation testing

Abstract

Over the last 25 years, recombinant allergens from all important allergen sources have been cloned and are now available as recombinant proteins. These molecules can be produced in practically unlimited amounts without biological or batch-to-batch variability. It has been shown in provocation tests that recombinant allergens have similar clinical effects as their natural counterparts. With the help of these tools it is possible to reveal the precise reactivity profiles of patients and to uncover and differentiate cross-reactivity from genuine sensitization to an allergen source. Although it has been shown some time ago that it would be possible to replace crude allergen extracts with recombinant allergens for skin prick testing, and even though the use of allergen components can improve routine diagnosis, these tools are still not available for clinical routine applications. The use of provocation tests is a crucial step in the development of new, hypoallergenic vaccines for therapy of allergic disease. Here we describe important provocation methods (skin prick test, intradermal test, atopy patch test, nasal provocation, colonoscopic provocation test) and give an overview of the clinical provocation studies which have been performed with recombinant allergens so far.

1. Introduction

Over the past 25 years, recombinant forms of most important allergens have been produced and have been shown to be equal to their natural counterparts regarding their ability to bind IgE antibodies, stimulate T cells and induce allergic reactions [1,2]. Recombinant allergens are defined molecules which can be produced in a highly purified and controlled manner, thus avoiding batch-to-batch variability. This is opposed to natural allergen extracts, which contain most often several different allergens in varying concentrations [3–5] as well as a high number of non-allergenic components. The composition of allergen extracts depends on several not manipulable factors during the production of the natural source material [6,7]. Furthermore, standardization of allergen extracts can only be made for one major allergen, while the composition of other components remains unchanged [8]. Recombinant allergens have been shown to be able to complement or replace natural allergen extracts for diagnosis [9], and highly sophisticated in vitro diagnostic test systems have been developed which allow the precise analysis of the reactivity profiles of allergic patients [10–12]. At present, a combination of tests based on natural allergen extracts and component resolved testing is used for diagnosis of allergy in routine settings.

Specific immunotherapy is the only allergy treatment which is able to change the course of allergic disease [13,14]. Based on the knowledge of the precise immunological and structural properties of allergens and the location of IgE epitopes, recombinant allergen derivatives which have a reduced ability to induce effector cell degranulation have been produced, with the goal to improve treatment success and reduce side effects of immunotherapy [15,16]. The clinical characteristics of promising vaccine candidates need to be evaluated not only in in vitro test systems but also directly in allergic patients, using provocation testing [17–19]. According to a guideline by the European Medicines Agency (EMA) published on June 1st in 2009 (http://www.ema.europa.eu/pdfs/human/ewp/1850406enfin.pdf), new allergy vaccines do not need to undergo a classical phase I clinical study in healthy non-allergic subjects. This clinical phase is usually replaced by a provocation study in allergic subjects, e.g., a skin test study, which is immediately followed by a phase II study in allergic patients.

The number of published studies employing provocation testing with recombinant allergens has declined substantially over the past few years. This can be explained by the implementation Commission Directive 2003/94/EC (Medicinal Products for Human and Veterinary Use. Eudralex), which regulates the Good Manufacturing Practice (GMP) in the EU, thus prohibiting the approval of clinical studies with non-GMP produced recombinant allergens. As GMP production of allergens is both elaborate and costly, only few recombinant allergens meeting these criteria are currently available and only few provocation studies with recombinant allergens were initiated after that time point. However, it is to be expected that the number of recombinant allergens produced in GMP quality will increase over the next few years and thus the number of provocation studies with recombinant allergens will rise again. In particular, there remains the necessity of clinical evaluation of new recombinant allergen vaccines before they can be put into use for subcutaneous or sublingual treatment of allergic patients. This review will summarize provocation methods which can be and have been performed with recombinant allergens, mainly focussing on skin prick and intradermal testing, atopy patch testing, nasal provocation testing and colonic provocation testing. Bronchial and conjunctival provocation testing will both be only shortly addressed.

2. What can be gained by using allergen components for provocation testing?

With the wide availability of allergen components for measurement of allergen-specific IgE, the routine diagnostic spectrum in in vitro tests has changed. Component-resolved testing allows the identification of patients who are genuinely sensitized to an allergen source and those who have positive skin reactions merely because they are sensitized to a highly cross-reactive panallergen [20]. Although many skin test studies and other provocation studies have already been performed with recombinant allergens (Tables 1–3) and although the advantages of skin testing with recombinant allergens have been recognized many years ago [21], and even though standardization of allergen extracts has remained a difficult problem [8], 25 years after the first allergen was produced in a recombinant form even the most relevant allergen components are still not commercially available for biological testing. This can be attributed to the fact that test substances based on recombinant allergens legally need to undergo far more rigorous, elaborate and costly studies than those based on natural allergen extracts [22].

Table 1.

Skin prick test and intradermal test studies with recombinant allergens.

Allergensource		SPT/IDT	Number of patients	Number of controls	Concentration used (mg/l)	Reference
Mite	Der p 2	IDT	45	11	1	[95]
		SPT	230	12	1–100	[96,97]
	Der p 5	IDT	76	18	10−5–10	[95,98,99]
		SPT	135	12	1–100	[96,98,99]
	Der p 7	SPT	60	12	1–100	[96]
	Der f 11	IDT	21	0	1	[100]
	Blot t 5	SPT	82	0	5	[98,99]
		IDT	31	0	?	[99]
	Cra-A	SPT	22	15	100	[101]
	Lep d 2	SPT	44	38	1–100	[102]
	Tyr p 2	SPT	33	38	1–100	[102]
Pollen	Bet v 1	SPT	134	70	3–100	[23,55,103–105]
		IDT	99	8	10−6–10	[90,50,105]
	Bet v 1d	SPT	48	21	10	[93]
		IDT	48	21	10−2–10	[93]
	Bet v 2	SPT	91	62	3–50	[23,55,103,104]
	Ole e 1	SPT	33	20	0.1–100	[106]
	Fra e 1	SPT	30	0	1–10	[107]
	Pla a 1, 2	SPT	47	24	0.3–100	[53]
	Par j 1, 2	SPT	30	15	0.5–50	[49]
	Phl p 1	SPT	139	11	1–20	[23,24,55]
	Phl p 2	SPT	145	11	0.5–10	[23,24,55]
	Phl p 4	SPT	82	0	2–18	[24]
	Phl p 5	SPT	71	11	1–10	[23,55]
	Phl p 13	SPT	82	0	2–18	[24]
Fungi	Alt a 1	SPT	42	20	1–100	[54]
	Asp f I	SPT	40	0	100	[108]
		IDT	100	20	0.1–1	[109]
	Asp f I/a	SPT	70	39	100	[110–112]
		IDT	49	51	10−5–10	[110,111,113]
	Asp f 3	IDT	119	24	0.1–10	[109,114]
	Asp f 4	IDT	24	5	10−2–10	[115]
	Asp f 6	IDT	128	29	10−5–10	[109,115,116]
	Asp f 8	IDT	8	2	10−5–1	[117]
	Cop c 1	SPT	5	5	10−2–479	[52]
Bee venom	PLA 2	IDT	91	27	10−8–10−2	[118,119]
Dog	Can f 1, 2	SPT	25	11	0.0025–250	[120]
Latex	Hev b 2, 3, 5, 6, 7, 8	SPT	29	10	10−9–1000	[121]
Food	Api g 1	SPT	36	5	10–1000	[122]
		IDT	36	5	10−2–10	[122]
	Pho d 2	SPT	20	0	50	[38]
	Mal d 4	SPT	5	0	0.02–100	[123]
	Pru av 1, 3, 4	SPT	33	46	10–100	[124]
	Pru p 3	SPT	20	0	30	[38]
	Ara h 1, 2, 3	SPT	61	30	0.1–100	[57,58]
	Ara h 6	SPT	31	0	0.1–100	[58]
	Gad c 1	SPT	10	0	1000	[125]
	Sal s 1	SPT	10	0	1000	[125]
	The c 1	SPT	10	0	1000	[125]

Table 2.

Skin prick and intradermal tests performed with genetically modified allergens.

Allergen source		SPT/IDT	Nr of patients	Nrof controls	Concentration used (mg/l)	References
Mite	Der f 2 fragments
	Lep d 2 derivative	SPT	17	8	100	[126]
	Der p 1, Der p 2 hybrids	SPT	106	40	5–500	[127]
	Blo t 5 mutants	SPT	5	0	10	[128]
Pollen	Bet v 1 fragments, Bet v 1 trimer	SPT	52	22	1–100	[17,19]
	Bet v 1 fragments	IDT	52	22	10–4–1	[17,19]
	Ole e 1 peptides	SPT	14	0	?	[129]
	Par j 1, 2 hybrid	SPT	30	15	5–250	[130]
	Par j 1, 2 hybrid	SPT	5	0	5–50	[131]
	Par j 1 variants	SPT	10	0		[132]
	Par j 2 fragments	SPT	10	0	20	[133]
	Phl p 1 peptides	SPT	8	0	20–100	[134]
	Phl p 1, 2, 5, 6 hybrid	SPT	32	9	4–108	[56]
Latex	Hev b 6 mutant	SPT	4	0	0.1–1	[135]
Dog	Can f 1, 2	SPT	25	11	0.0025–250	[120]
Cow	Bos d 2 fragments	SPT				[136]
	Bos d 2 fragments	SPT			0.025–250	[137]
Food	Cuc m 2 mutants	SPT	13	6	50	[138]
	Parvalbumin mutant	SPT	1	0	1–32	[139]
	Mal d 1 mutant	SPT	14	0	0.02–100	[140]
	Mal d 1 mutant	SPT	2	0	0.02–100	[141]

Table 3.

Nasal, bronchial, ocular and colonoscopic provocation studies with recombinant allergens.

Site of provocation	Allergen(s) used	Nr of patients	Nr of controls	Dose range (μg/ml)	Application of allergen	Ref.
Nasal	Bet v 1	13	0	0.1–10	spray, better ventilated nostril	[77]
	Bet v 1, 2, Phl p 1, 2, 5	24	8	5–40	spray, alternating nostrils	[23]
	Bos d 2, Bos d 2 fragments	22	12	0.1–100	spray, unilateral	[80]
	Bet v 1, Bet v 1 fragments, Bet v 1 trimer	10	0	1–100	spray, right nostril	[18]
	Bet v 1	34	5	1–100	spray, unilateral	[78]
	Bet v 1, Bet v 1 fragments	19	0	0.04–50	spray, both nostrils	[83]
	Bet v 1, Bet v 2, Phl p 1, 2, 5	8	0	5–40	spray, both nostrils	[87]
	Art v 1	32	10	1–100	spray, unilateral	[79]
Bronchial	Bet v 1	13	10	0.1–10		[77]
Conjunctival	nBet v 1, rBet v1a, rBet v 1d	48	21	0.01–10		[93]
Colonoscopic	Bet v 1	34	5			[94]

Another important application of recombinant allergens or allergen components is to study the clinical relevance of allergen components. It has been shown previously that the IgE binding capacity of an allergen alone does not predict its ability to induce allergic responses [23]. This is of particular importance for the design of new allergy vaccines. In this context, it has been shown that group 4 and group 13 grass pollen allergens have ninefold smaller allergenic activity than other grass pollen allergens (group 1, 2 and 5 allergens) and are therefore not essential components of therapeutic vaccine formulations against grass pollen allergy [24]. Furthermore, provocation tests have proven to be valuable for the evaluation of new therapeutic vaccines which have altered IgE binding capacity and allergenic activity (Table 2).

A comparison of possible advantages and problems associated with the use of natural allergen extracts and recombinant allergens for biological testing can be found in Table 4.

Table 4.

Advantages and disadvantages of recombinant allergens for biological testing.

	Natural allergen extracts	Recombinant allergens
Advantages	Easy to prepare	Detection of cross-reactivity and genuine sensitization by component-resolved testing
	Inexpensive	Highly defined proteins
	Market authorization relatively easy to obtain	No contamination from other allergen sources
	Ideally contain all allergenic proteins	Precise amount and structural characteristics of proteins known
Disadvantages	Standardization difficult or impossible	More expensive and laborious preparation (when first set up)
	Quality depends on source material, batch-to-batch variability	Market authorisation more difficult to obtain
	Contains undefined components and may contain contaminations from other allergen sources	For some allergen sources no complete/representative panel of allergens yet available
	Endogenous degradation may cause low sensitivity	Some allergens need special expression systems
	Complex mixtures of proteins may lead to low assay specificity

3. Skin prick and intradermal testing

Skin prick tests (SPT) and intradermal or intracutaneous skin tests (ICT) were introduced by Blackley in 1865 [25] and have since then served as an important tool in the diagnosis of immediate-type allergic reactions. They are easy to perform, inexpensive, safe and allow a visualization of sensitization within 15–20 min. They are performed by introducing small amounts of allergen into the dermis [26]. In the skin of allergic subjects, effector cells are armed with allergen-specific IgE that is bound to their high affinity receptor, FcεRI. Upon contact with allergen, cross-linking of IgE occurs and leads to release of mediators (histamine, tryptase, TNF-α, prostaglandins, leukotriens, IL-4, and others [27,28]). The released mediators cause vasodilatation and increase vascular permeability of the skin, thus resulting in tissue edema and the development of the typical “wheal reaction” as well as localized erythema caused by vasodilatation. In skin prick tests, mainly the size of the wheal determines whether a skin prick test reaction is regarded as positive or negative, while the erythema is usually not accounted for [29,30]. A late phase reaction may occur one to two hours later, peaking at 6 to 12 h and usually diminishes within 48 h [31,32].

It needs to be borne in mind that results from skin prick testing and the measurement of allergen-specific IgE in the serum do not always correlate and that subjects with positive skin reactions do not necessarily suffer from allergic symptoms [33–35].

A number of recent skin test studies have explored the usefulness of three allergen components for the diagnosis of food allergy [36–39]. The use of recombinant allergen components would be particularly useful in food allergy as the detection of potential pollen-food cross-reactivity is important and food allergen extracts are often unstable and unreliable. In a study by Viera et al. [38], natural profilin (Phl p 2) from date palm extract, the major apple allergen, Mal d 1, from apple extract and a peach LTP commercial extract which was shown to lack other allergens were used for skin testing and compared with IgE reactivity to recombinant Bet v 1, Bet v 2, Phl p 12 and Pru p 3. The authors found that sensitization to pan-allergens in children with fruit and vegetable allergy was common and that using allergen components would be a simple and feasible way of improving allergy diagnosis. In another study, Asero et al. studied the clinical relevance of positive skin prick tests to the same three allergen components in pollen allergic patients [39]. The authors confirmed that the clinical relevance of hypersensitivity to pan-allergens is often limited in patients with respiratory allergy.

3.1. Methods of skin testing

3.1.1. Skin prick tests

Skin prick testing is a routine method which has recently been extensively reviewed [30,40] and will therefore not be described in detail in this review, which will focus on the particularities of skin prick testing with recombinant allergens. In short, a skin prick test is performed by applying the allergen solution on the volar forearm or, if this is not possible, the back of the patient. A lancet is passed through the drop and inserted into the skin. The wheal and flare reaction is interpreted after approximately 15 min.

3.1.2. Intradermal tests

Intradermal skin tests have been used for the biological evaluation of recombinant allergens and for validation of genetically engineered hypoallergenic derivatives (Tables 1 and 2). Before the test, patients are advised to stop the use of certain medications (Table 5). The recombinant allergen solution is injected intracutaneously from a 0.5 or 1.0 ml plastic syringe through a 26-gauge needle. Between 0.02 and 0.05 ml of the allergen solution is injected into the skin to produce an intradermal bleb approximately 3 mm in diameter. Wheal reactions less than 5 mm are regarded as negative [41,42]. In experienced hands, the intradermal skin test is more reproducible than SPT, but a higher level of technical skill is required [43,44]. The advantages of the intradermal test are a higher sensitivity, disadvantages are that the test is painful, more laborious to perform and more often produces false-positive reactions. Furthermore, it has an increased risk of systemic allergic reactions as compared to skin prick testing [41].

Table 5.

Potential interference of medications with provocation reactions.

	SPT, ICT	APT	NPT
Antihistamines
2nd generation H1-blocker	7 days	3 days	3 days
Ketotifen	5 days	–	–
Glucocorticosteroids, local
Skin in test area	>1 week	>1 week	–
Nasal	0	0	7 days
Inhaled	0	0	–
Glucocorticosteroids, systemic			7 days
Short-term < 50 mg/d Prednisolone-equivalent	3 days	–	–
Short-term > 50 mg/d Prednisolone-equivalent	1 week	–	–
Long-term < 10 mg/d Prednisolone-equivalent	0	–	–
Long-term > 10 mg Prednisolone-equivalent	3 weeks	–	–
Topical calcineurin inhibitors
Tacrolimus	>1 week	0
Pimecrolimus	>1 week	2 days
Omalizumab	>4 weeks	–	–
Antidepressants			3 days
Doxepin	7 days	–	–
Desipramine	3 days	–	–
Nasal alpha-Adrenergics	–	–	3 days
References	[40]	[59,70]	[84]

SPT: skin prick test, IDT: intradermal test, APT: atopy patch test, NPT: nasal provocation test; 0: no time interval necessary before provocation, – : no recommendations regarding interference found.

3.2. Safety of the skin prick and intradermal test

Fatal systemic reactions in response to skin prick testing are exceptionally rare, but have been observed when using natural allergen extract [45–57]. Intradermal tests have a higher propensity than skin prick tests to induce large immediate local reactions as well as systemic reactions and fatalities have been reported but are also extremely rare [46]. So far, no complications have been reported when recombinant allergens were used for skin prick or intradermal testing, however, these molecules have so far not been used in the broad range.

3.3. Doses used for skin testing with recombinant allergens

More than 40 studies reported results from skin prick tests with recombinant allergens. We have summarized these data regarding the allergen tested, number of allergic patients and control individuals, details of the skin test method and allergen concentrations used (Table 1). In many of these studies, end-point titration using 3- or 10-fold dilution series was used for a quantitative assessment of the allergen concentration necessary for inducing a positive skin reaction. As recombinant allergens are highly purified molecules and can therefore be applied in precisely quantifiable concentrations, this technique has yielded highly reproducible results. This technique is superior to using a fixed concentration in the context of scientific studies, as wheal sizes may be identical even when a 100-fold difference in potency exists [48]. Substantial differences were found regarding the biological activity of different allergen components. Using end-point titration, it was possible to find the concentration which detects all the patients sensitized to this molecule in a given population. As an example, among 51 patients with clinically relevant birch allergy who were tested by skin prick tests at 3, 10 and 50 μg/ml, it was demonstrated that all patients allergic to birch were detected with a solution at 10 μg/ml whereas only 27 out of 51 gave a significant positive test at 3 μg/ml [19]. For rPar j 1, 8 out of 30 patients did not react to a concentration of 5 μg/ml whereas all patients gave a positive reaction to rPar j 2 at that concentration [49]. In a study which compared skin test reactivity, nasal reactivity and specific IgE levels to three grass pollen allergens (Phl p 1, Phl p 2, Phl p 5) and two birch pollen allergens (Bet v 1, Bet v 2) significant differences were observed regarding the in vitro reactivity and the biological activity of the allergens [23]. These studies underline the necessity to determine for each recombinant allergen the correct concentration able to detect patients sensitized to this molecule.

In account of a great inter-individual variability of sensitivities to the molecular allergen, the starting dose should be low for safety reasons. Among 18 patients with a clinical history of birch pollinosis tested intradermally with 10-fold dilutions of rBet v 1 in order to determine the lowest concentration inducing a significant cutaneous response, we observed a broad range of skin sensitivity to rBet v 1 (from 10 to 10⁻⁵ μg/ml) [50,51]. This is also demonstrated by Brander et al. who showed that less than 2 pmol of Cop c 1 (a recombinant allergen from the basidiomycete coprinus comatus) was able to elicit a positive reaction in highly sensitized subjects [52].

In other studies the optimal concentration was selected by comparison with the wheals induced with the positive control (i.e. histamine 10 mg/ml) in a given population sensitized to the allergen source. For the two molecular allergens of plane tree pollen the median of the allergen concentrations that induced a wheal surface area of the same size as that of the histamine was 3 μg/ml for nPla a 1, 52.3 μg/ml for nPla a 2 and 20.7 μg/ml for rPla a 1 [53]. In another study the concentration of nAlt a 1 that induces a weal surface of the same size as the median weal size produced by histamine (10 mg/ml) was 2.91 μg/ml, and 5.81 μg/ml for ry Alt a 1 and 14.63 μg/ml for rb Alt a 1 [54]; these results point out the importance of the expression system for the production of recombinant allergens.

3.4. Replacing allergen extracts with mixtures of recombinant allergens

Some allergen sources contain only one major allergen which is recognized by virtually all patients who are sensitized to the allergen source. Two examples for this are birch pollen allergy (major allergen: Bet v 1) and ash allergy (major allergen: Fra e 1), where skin testing with the extract could be replaced or complemented with skin testing with the major allergen and relevant pan-allergens (profilin, polcalcin) which indicate cross-reactivity. For allergen sources containing several different allergen molecules, diagnostic tests which include the best indicators for specific sensitization to the allergenic sources could replace traditional extracts. In this context it has been demonstrated that a mix of Pla a 1 and Pla a 2 detected 100% of monosensitized patients allergic to Platanus acerifolia and 87.5% of polysensitized patients allergic to Platanus acerifolia; no false positive reaction was detected [53]. Heiss et al. skin tested patients allergic to grass and birch with 5 recombinant allergens (Phl p 1, Phl p 2, Phl p 5, Bet v 1, Bet v 2) [55] and detected 52/54 grass pollen allergic patients and 35/36 birch pollen allergic patients. A recombinant hybrid molecule consisting of four major timothy grass pollen allergens Phl p 1, 2, 5 and 6 detected 31 out of 32 patients allergic to grass pollen at the concentration of 36 μg/ml. A dose response relationship was established between the hybrid molecule concentrations (4, 12, 36, 108 μg/ml) and the areas of skin tests. Two late reactions persisting over 24 h occurred with the highest concentration (108 μg/ml), whereas no relevant local side effect was observed at the concentration of 36 μg/ml [56]. This study points out that for each recombinant allergen, a large number of patients and concentrations should be tested to give precise dose recommendations.

3.5. Can skin tests with recombinant allergens predict disease severity?

This is illustrated by using peanut recombinant allergens in skin tests [57,58]. In one study it was shown that co-sensitization to Ara h 2 and/or Ara h 3 is predictive of more severe reactions [57]. In another study the cumulative skin prick test response for the purified allergens at a concentration of 0.1 μg/ml showed a significant difference between patients with mild and severe symptoms. The differences were particularly evident with Ara h 2 and 6 at low concentrations and Ara h 1 and 3 at higher concentrations [58].

4. Atopy patch testing

While the IgE-mediated early allergic response can easily and reliably be evaluated using skin prick and/or intradermal testing, this is much less true for delayed facets of the allergic response which are mediated by allergen-specific T cells. Atopy patch tests (APTs) have been developed to investigate patients with atopic dermatitis (AD) whose skin lesions exacerbate after contact with certain respiratory or food allergens [59]. Recombinant allergens were used for atopy patch tests for the first time in 2002, when Johansson et al. used Malassezia furfur extract and recombinant Mal f 1, Mal f 5 and Mal f 6 for APT in 40 patients to study the role of circulating Th2 cells in atopic dermatitis [60]. The role of T cells in atopy patch test reaction has been investigated in several studies [61,62], but whether allergen-specific IgE antibodies also play a role was only recently established. In a study published by Campana et al., patch test reactivity to recombinant Bet v 1 and two fragments of Bet v 1 was compared in five birch pollen allergic patients with AD [63]. The two Bet v 1 fragments (aa1-74, aa75-160) almost completely lack IgE reactivity and allergenic activity, but retain the full T-cell epitope repertoire of the Bet v 1 allergen [64]. A positive atopy patch test reaction to Bet v 1 was observed in all 5 birch pollen allergic AD patients, while a positive reaction to the two fragments was also observed in four of these five patients. This study did thus clearly show that only allergen-specific T cell epitopes but not IgE epitopes are prerequisites for the induction of a positive APT reaction. Furthermore, this study may also indicate that a positive APT reveals a principal ability of a molecule to induce T cell mediated reactions. During the last 20 years, numerous vaccine candidates based on genetically modified recombinant allergens have been developed with the aim to improve specific immunotherapy of allergic disease [15,65]. While most of these molecules have been mainly designed to have a greatly reduced ability to induce effector cell degranulation while retaining their allergen-specific T cell reactivity, some recently developed molecules also have altered T cell epitopes [66–68]. The atopy patch test may potentially be helpful to predict delayed type reactions, however, this needs to be confirmed in further studies.

4.1. Methods of atopy patch testing

Recombinant allergen skin test solutions should be prepared freshly from lyophilized allergens diluted in H₂O. Use of the aqueous solutions mixed with petrolatum or directly applied on paper discs in 8 mm Finn chambers have both been described. If petrolatum is used as a carrier, it should be included as a negative control as positive reactions to petrolatum have been described in 1% of AD patients [69].

Before atopy patch testing, patients should be advised to stop the use of certain medications (Table 5). Testing is performed using aluminium cups (Finn Chambers, Epitest Ltd Oy, Tuusala, Finland), which are administered on the back of patients and kept in place using hypoallergenic adhesive tape. Both 8 mm and 12 mm cups are available, but the larger chambers have proven to yield more reliable results [70]. Only skin not currently affected by disease should be used for testing. It has been recommended that the skin should not be pre-treated by scratch test or acetone, however, in the studies performed with recombinant allergens so far, the skin was tape-stripped before testing [63,71].

4.2. Interpretation of APT results

The reactions should be read at 48 or 72 h [59,70]. As opposed to standard patch tests, atopy patch tests are regarded as positive also if the reaction disappears within 24 h after the test chambers have been removed [59]. The reaction should be graded according to the grading system developed by the European Task Force on Atopic Dermatitis (ETFAD, Table 6) [70], thus erythema alone is graded either as doubtful or negative and only palpable reactions (papules or infiltrated reactions) are of clinical relevance. Notably, positive atopy patch test reactions occur not only in patients with atopic dermatitis but also in allergic patients without skin disease, although they are substantially less frequent in this patient group. It has been reported that positive APT reactions can be induced in 20% of house dust mite allergic patients and in approximately 10% of pollen allergic patients not suffering from AD [72]. When atopy patch tests with recombinant allergens are performed in these patient groups without skin disease, it is important to include a relevant positive control in the test. This may either consist of a commercial test solution or of non-commercial solutions of the wildtype allergen, which have been described to yield better sensitivity and reproducibility than commercial tests [73]. Furthermore, a negative control needs to be included as false positive reactions may occur because of the irritation due to the occlusion of the test chambers.

Table 6.

Patch test reading according to ETFAD.

Grading of APT reactions (ETFAD)
–	Negative
?	Only erythema, questionable
+	Erythema, infiltration
++	Erythema, few papules (up to 3)
+++	Erythema, many or spreacing papules
++++	Erythema, vesicles

Revised key for APT according to the European Task Force on Atopic Dermatitis [70].

5. Nasal provocation tests with recombinant allergens

Assessing the clinical reactivity to specific allergens by nasal provocation testing harbours a number of advantages: The test is done in the organ which is most frequently affected by respiratory allergy and should thus correspond well to the effects of natural exposure. Furthermore, the nose is readily accessible and thus provocation tests are relatively easy to perform. Nasal provocation tests are also less hazardous than other forms of provocation tests and the results can be interpreted both by objective (e.g., measurement of nasal air flow) and subjective methods (recording of symptoms). In clinical practice, nasal provocation tests are used in case of discrepancies between test results and history of allergic rhinitis, to confirm occupational allergic rhinitis and for diagnosis of perennial allergy before immunotherapy, while for research purposes, nasal challenges have been used to study the local mechanisms of allergic rhinitis [74,75]. Recently, a number of studies have been published which showed that a substantial number of patients with symptoms of rhinitis but negative skin tests with allergens and normal allergen-specific serum IgE levels harbour IgE antibodies in the nose and thus suffer from local allergic rhinitis [76]. This affection can only be diagnosed by nasal provocation tests. Nasal provocation tests with recombinant allergens are currently not used in routine clinical settings but exclusively for research purposes, where the specific advantages of precise dosing of single allergen molecules for provocation can be exploited.

The first nasal provocation studies performed with a recombinant allergen was done with the major birch pollen allergen, Bet v 1, and revealed that similar clinical effects in an effector organ of allergy can be achieved with recombinant and natural allergens [77]. This equality of recombinant and natural allergen in nasal provocation was confirmed in another study for Bet v 1 [78] and has also been shown for a major mugwort allergen, Art v 1 [79]. The possibility of precise dosing of individual allergens inherent to this technique was exploited in a study which compared skin prick test, nasal provocation test and IgE reactivity to individual two birch pollen allergens (Bet v 1, Bet v 2) and three grass pollen allergens (Phl p 1, Phl p 2, Phl p 5) [23]. The study demonstrated that on a molecular level skin testing provides a better reflection of nasal sensitivity than measurement of specific IgE levels.

Comparisons between nasal provocation tests with recombinant allergens and allergen derivatives have been performed on a number of occasions, mainly in order to prove that the reduced allergenic activity was present not only in vitro and in skin tests, but also upon contact with an allergic “shock-organ”: Ruoppi et al. compared the reactivity to recombinant Bos d 2, a cow dander allergen which is important as a cause of respiratory allergy in rural Finland, with two different Bos d 2 fragments (aa1-131 and aa81-172) [80]. The authors found that both in vitro and in vivo, the reactivity to the fragments was far lower than to wild-type recombinant Bos d 2. Hage-Hamsten et al. compared nasal reactivity to wildtype recombinant Bet v 1 with two Bet v 1 derivatives (a recombinant trimer and two recombinant fragments) [18]. Patients exhibited greatly reduced nasal symptoms upon contact with the two recombinant Bet v 1 derivatives as compared with the wildtype molecule. In this study, no difference was found between the symptoms induced by nasal provocation with the Bet v 1 derivatives and diluent alone and no late reactions were observed. The recombinant Bet v 1 fragments and the Bet v 1 trimer were thus judged to have greatly reduced propensity to induce immediate-type allergic reactions and were subsequently used in a successful clinical trial for injection immunotherapy of birch pollen allergic patients [81,82]. The same Bet v 1 fragments described above were used in another nasal provocation study which compared nasal symptoms and systemic immune responses after nasal contact with wildtype recombinant Bet v 1 and the two Bet v 1 fragments [83]. The nasal provocation studies that have so far been performed with recombinant allergen molecules or allergen derivatives are listed in Table 3.

5.1. Methods

5.1.1. Nasal application of allergens

The different methods for nasal allergen challenge have been previously summarized [84,85], and recently different methods for nasal challenge with grass pollen extract and sampling of nasal secretions have been compared [86]. We briefly summarize the techniques for nasal provocation tests based on these documents and our own experience [23,82,83,87,88].

Medications which should be stopped before a nasal provocation is carried out are listed in Table 5. Application of allergens should be preceded by anterior rhinoscopy, to rule out any substantial nasal pathology (e.g., severe septal, deviation) precluding correct deposition of test solutions. Challenging both nostrils, the better aerated nostril or alternating nostrils starting with the better aerated nostril have all been described. If only one nostril is challenged, both sides should nevertheless be measured to avoid not recognizing nasal cycling effects [85]. Allergens may be applied to the nose using nasal sprays, drops, filter paper, or direct application of small amounts of test solution to the inferior turbinate using a pipette.

If the nostrils are sufficiently patent, a relatively large part of the nasal mucosa can be reached using a nasal spray. Subjects should be told to inspire before spraying of the test solution and to hold their breath during spraying to avoid entry of allergens into the lower airways.

Use of filter paper and use of a pipette for application of test solutions both allow a standardized application of allergens to the mucosa, however these application methods can only be used under direction vision and may irritate the mucosa. Nasal drops do not allow application of allergen solutions at a defined location and are therefore less well applicable for standardized provocation procedures.

5.1.2. Dosing of allergens for nasal provocation

Nasal provocation with allergens should always be preceded by challenging the nose with diluent. To identify the threshold dose, application of allergens is started at a low concentration, which is increased with each subsequent challenge every 10–20 min. We found in our own studies that a dose increase by a factor 10 is too big and suggest that the dose should best be increased by factor 3 or 5. The starting doses and dose ranges that have been described for nasal provocation testing with recombinant allergens are listed in Table 3.

5.1.3. Evaluation of the clinical response to challenge

As in allergic rhinitis, different symptoms including itching, sneezing, secretion and obstruction all play an important role, a scoring system which takes both subjective symptoms (i.e., itching of the nose, ears, runny nose) and objective measurements (i.e., counting sneezes, weighing of nasal secretions, measurement of changes in nasal airflow) into account as method for evaluating the response to nasal provocation with allergens. The different methods of measurement of the nasal airflow have been described by Malm et al. [85]. The most frequently used and most well standardized method is active anterior rhinomanometry, during which nasal flow (cm³/s) at a transnasal pressure of 150 Pascal is measured. Reproducibility of the peak nasal inspiratory flow has also recently been demonstrated [86]. For the assessment of symptoms, either categorized scoring systems [89] or a visual analogue scale [90,91] have been described.

5.1.4. Collection of nasal fluids

A variety of methods for collection of nasal fluids have been described, including nasal lavage, adsorptive filter strips or nasal sponges. Scadding et al. have compared different methods of nasal fluid collection and concluded that the most practical method of collecting nasal fluid for analysis of biomarkers was a synthetic polyurethane sponge, pre-cut into 20 × 15 × 5 mm pieces and sterilized by autoclaving prior to use [86]. The sponges were placed between the inferior turbinate and the nasal septum and left in place for 2 min. After removal, sponges are added to 2 ml centrifuge tubes with indwelling 0.22 μg cellulose acetate filters and are centrifuged. We have used the same method for collecting nasal fluids and found that a higher amount of nasal fluid can be collected if sponges are left in place for 5 or 10 min.

6. Bronchial provocation tests with recombinant allergens

In the first provocation study done with a recombinant allergen, skin prick testing, nasal provocation testing and bronchial provocation testing was performed with recombinant and natural Bet v 1 [77] (Table 3). Since then, no further bronchial provocation study with recombinant allergens has been performed. As this study and the methods of bronchial provocation testing with recombinant allergens have been previously described [92], this topic will not be covered in this article.

7. Conjunctival provocation tests with recombinant allergens

The conjunctival provocation test is a reliable, reproducible and safe method to evaluate allergic reactions in the eye. It is a common method for recording the success of treatment in specific immunotherapy trials. So far, conjunctival provocation tests with recombinant allergens have been performed only in one study [93] (Table 3). This study and the method of conjunctival provocation testing have been described in a previous issue of Methods [92], and no new guidelines for conjunctival provocation have been published since. Therefore a detailed description of conjunctival provocation tests will not be given in this article.

8. Colonoscopic Provocation with recombinant allergens

The colonoscopic allergen provocation test (COLAP) has been described by Bischoff and coworkers as a means to objectively differentiate whether gastrointestinal (GI) tract complaints (nausea, vomiting, abdominal pain or diarrhea) are caused by food allergy or other causes (e.g., irritable bowel syndrome). In one recent study, COLAP was performed with recombinant Bet v 1 in order to confirm birch-pollen associated food allergy [94]. The authors performed colonoscopic allergen provocation in 16 patients with a history of food allergy clearly associated with Bet v 1-associated foods (e.g., apples, hazelnuts, celery), 18 patients with GI symptoms without clear association with Bet v 1-related foods and 5 healthy control individuals. They found that all the patients with pollinosis and a history of intolerance of Bet v 1-related foods and three of the patients with GI symptoms but without clear history of Bet v 1-related food intolerance were positive in the COLAP test.

8.1. Methods

The COLAP test can only be performed by a physician trained to perform colonoscopy and experienced in injection of the gut mucosa.

The cecum is chosen for provocation because of its relatively minor peristaltic movements and the tolerability of prolonged examination of this portion of the gut as compared to the gut accessible by gastroduodenoscopy. The test should be performed only on macroscopically normal looking mucosa. A small number of different allergens can be tested simultaneously by injection into the mucosa with a fine needle manufactured for sclerosing of small blood vessels in the gut [94].

A volume of 0.2 ml of an allergen solution at a concentration of 10% of the concentration recommended for skin tests is used for provocation. Histamine at a concentration of 3.4 μg/ml serves as a positive control and induces a positive reaction in approximately two thirds of patients. 0.9% NaCl solution serves as negative control. Positive food allergen challenge in the COLAP test induces a wheal and flare reaction, similar to the reaction induced in skin prick tests [94].

9. Conclusion

Recombinant allergens from all important allergen sources are now available and can be produced in large amounts in consistent quality. Their biological activity has so far been investigated in skin prick and intradermal tests, atopy patch tests, as well as nasal, bronchial, conjunctival and colonoscopic provocation tests. As the value of component-resolved clinical tests has been proven in numerous studies and the use of panels of recombinant allergens for highly sophisticated in vitro tests has to improve diagnosis of allergy, broad availability of recombinant allergens for biological tests would improve allergy diagnosis. Currently, the most important application of provocation testing with recombinant allergens is for the evaluation of new vaccines with reduced allergenic potential and/or reduced allergen-specific T cell epitopes.