Allergy screening with extract‐based skin prick tests demonstrates higher sensitivity over in vitro molecular allergy testing

Abstract Background As extract‐based skin testing as well as in vitro tests for major allergens have their own advantages, both procedures are usually performed in routine settings. In times of shortages in medical staff and supplies, we asked ourselves, how many patients would be underdiagnosed, if only one test could be used. Methods In a retrospective analysis, we investigated a cohort of 2646 patients seen by a single physician in a large Austrian outpatient allergy clinic in 2018. Only patients with an allergen source‐specific history and pairs of extract‐based skin prick (SPT) and in vitro molecular allergy tests to major allergens were included. Results For all tested allergen sources, sensitivity was higher for SPT than for sIgE‐based molecular allergy testing. Concerning 1006 birch pollen‐allergic patients, 791 (78.6%) had positive results with both tests, while 153 (15.2%) only with the SPT and 62 (6.2%) only with the sIgE to Bet v1. The other allergen sources showed similar results: For house dust mite 816/1120 (72.9%), grass pollen 1077/1416 (76.1%) and cat 433/622 (69.6%) remained test‐positive with both procedures, whereas in 276 (24.6%), 224 (15.8%) and 173 (27.8%) times only the SPT and 28 (2.5%), 115 (8.1%) and 16 (2.6%) times only the sIgE to Der p1/2/23, Phl p1/5 and Fel d1 showed a positive result. Each comparison was statistically significant (each p < 0.0001, Chi‐squared test). Conclusions Screening for allergy with major molecular allergens has lower sensitivity when compared with extract‐based skin tests. A combination of both is required for an optimal sensitivity.


| INTRODUCTION
In allergology, allergen skin and in vitro tests are widely used for screening purposes. Hence, reliability that is, sensitivity and specificity are important to reliably confirm or rule out IgE-mediated allergy in large patients' cohorts. In the past, there were studies suggesting a better sensitivity of extract-based skin prick tests (SPTs) than extract-based in vitro tests. 1 Skin testing has always been a standard procedure in allergy diagnosis. 2 It offers the advantage of giving immediate results that can be discussed with patients without having to wait for laboratory results and without requiring a second appointment for discussing the clinical relevance of the allergy tests. Also, the results of skin tests are quite instructive to patients especially concerning emotional allergens such as pet animals. However, skin testing is an expensive procedure requiring extra-work force by skilled personal. Urticaria factitia (dermatographism) and clinically irrelevant sensitization to pan-allergens (e.g., profilin) may hamper the reading in a small number of patients. Also, some drugs such as antihistamines, mirtazapine and quetiapine may be the cause for false negative results. 3 The withdrawal of commercially available SPT reagents has also been a problem reducing availability of important skin test reagents, for example, latex extract. 4 However, also in vitro tests have been voluntarily withdrawn from the market by manufacturers and supply problems have affected also important molecular allergens for example, the major latex allergens Hev b5 and Hev b6. High total IgE levels might lead to false positive and very low total IgE levels to false negative in vitro tests. 5 Austria is a fortunate high-income country, where double allergy testing is available and reimbursed by public healthcare. In this study, we asked ourselves, how many allergy patients would be falsely labeled as NOT suffering from allergy if only one of the methods would be available: (a) extract-based SPT or (b) molecular allergy based on specific IgE (sIgE) with the major allergens of the most important inhalant allergen sources. We compared pairs of SPT and sIgE to house dust mite, cat, birch and grass pollen concerning sensitivity in patients with a clear-cut history of clinically relevant allergy to any of the aforementioned allergen sources.

| Patients & ethics committee vote
A cohort of patients presenting themselves at the Floridsdorf Allergy Center (FAZ) between January 2018 and December 2018 with a history compatible with an IgE-mediated type 1 allergy to inhalant allergens was included into this study. All patients were seen by a single physician to avoid inter-observer variation. Only patients with complete pairs of extract-based SPT and sIgE to the major allergens of cat (Fel d1), birch pollen (Bet v1), grass pollen (Phl p1 & 5), and house dust mite (Der p1/2/23) were included into this study. We chose cat and birch pollen as representants for allergen sources with dominating single major allergens (Fel d1 and Bet v1 respectively) and grass pollen and house dust mite (Dermatophagoides pteronyssinus) as representants for complex allergens with more than two dominating major allergens . To exclude double entries of the same   patients, only the first visit was considered in patients with several   presentations in the year 2018. Of the 5857 available data sets, 2646 patients were finally included into the study (Figure 1). The patients' data had been retrospectively retrieved from routine patients records and entered into a spreadsheet by TG/BH/ LK as part of their medical diploma thesis at the Medical University of Vienna (Ethics committee vote #2213/2018). The informed consent form is provided in the online repository (available in German as S1). The details were already described elsewhere. 6

| Diagnostics
The diagnostic procedure started by obtaining a careful allergenspecific history by an experienced allergologist. Allergen-specific details were recorded and allocated to the referring allergen (e.g.,

| Patients
There were 2646 patients, who met the inclusion criteria (Table 1). 53.7% were female and the arithmetic mean age was 32.7 years (σ = 18.0 years). Thirty-five percent of them suffered from allergic rhinoconjunctivitis, 11% from bronchial asthma or allergic cough and 7% from atopic dermatitis. We compared the results of the SPT with the molecular allergy diagnostics from 1281 patients regarding birch pollen allergy, from 1362 patients regarding dust mite allergy, from 1577 patients regarding grass pollen allergy and from 709 patients regarding cat allergy. Some patients suffered from sensitizations to more than one allergen (details in Figure 1). In total, the tests of 2646 patients were compared this way.

| Comparing test sensitivity between extractbased skin tests and molecular allergy testing by assessing specific IgE to four important allergens sources
Only patients with at least one positive result in either SPT or sIgE were included for comparing the sensitivity of both methods. Con-   Figure 2D).

| Correlation between wheal size and antibody level
The correlation between wheal-size categorized according to Ruëff et al. 7 and sIgE levels was highest in dust-mite, cat-and birch pollen sensitization (Spearman's Rho 0.54, 0.47 and 0.45, p < 0.01) and lowest in grass-pollen sensitized individuals (0.35), while still retaining significance (p < 0.01). These data sets were also visualized with box plots, by which a positive correlation was visible for all four allergens (Figure 3).

| The diagnostic value of adding Der p23 for diagnosing house dust mite allergy
House dust mite extracts contain a lot of known allergens but only some are described as major allergens, namely Der p1, Der p2 and lately Der p23. 8

| DISCUSSION
In this study we could demonstrate that the sensitivity of extractbased SPT is superior to measurement of sIgE against the major allergens (also known as allergen "components") of the most important inhalant allergen sources. This was true for complex allergen sources such as house dust mite with three (Der p1/2/23) and grass pollen with five (Phl p1/2/4/5/6) major allergens but also for less complex allergen sources with single dominating major allergens such as cat (Fel d1) or birch pollen (Bet v1).
Many allergologist regard in vitro testing with major allergens as the most modern approach of diagnosing type 1 allergy. 9 However, allergen sources are often complex and consist of mixtures of major, minor and some pan-allergens. If screening includes only the most prevalent variants of major allergens, allergic patients with sensitizations to minor allergens will produce false negative tests. Also, molecular allergens can be missing (unknown or commercially unavailable) for many rare allergens, for example, storage mites or shrimp. 10,11 This, after many years, keeps the door wide open for the standard allergy diagnostic pathway based on extracts from natural sources for skin tests but also in vitro tests. Extracts from natural sources offer the advantage of including a lot of minor allergen components and variants of the major allergens that are sometimes missing even in very well-known allergens; for example, the major birch pollen allergen Bet v1 is currently listed in 28 isoforms on the IUIS Allfam Server. 12 Also, most rare allergens are only available in extract form. 13 Insufficient sensitivity of molecular allergology has always been an issue from the very beginning and has not been completely resolved for all allergens yet. 14 However, extracts bear inherent disadvantages over molecular allergy diagnosis: (A) they

F I G U R E 3
Box plots comparing antibody-levels relative to the wheal size of the patients' skin prick test.

F I G U R E 4
Pie chart demonstrating the low additional value of adding the major allergen Der p23 to the molecular allergen panel in house dust mite-sensitized patients.
could show, that if molecular allergens are used for skin and in vitro tests side-by-side, sensitivity and specificity in LTP allergy to peach (Pru p3) is nearly the same, independently from the method used. 21 However, this is the rare exception where allergen components are available for skin testing.

| Dust mites
Bousquet et al. 22  another much smaller study from Italy, SPTs were compared with the ISAC112 platform from the same manufacturer as in our study using a different technical platform but the same molecular allergens. 23 The patients differed from our study as the Italian cohort also contained a lot of food allergic and urticaria patients that had been excluded in our study focusing on inhalant allergy. Also, Der

| Cat
Concerning cat sensitization, Bousquet et al. 22  EAACI survey showed that around 90% of European allergologists regard the SPT as the most important diagnostic tool. 29 Germany had been the first European country to implement the European guideline into national law but in recent years, many European countries have followed the German example putting even more pressure into this difficult matter. Hence, the EAACI created a task force aiming at keeping a close eye on not to lose the prerequisite for one of the allergologist's most important tools: commercially available SPT solutions to frequent and rare allergens of good quality and constant availability. 30

| CONCLUSION
This study reveals that screening for inhalant allergy with even the best characterized molecular allergens has insufficient sensitivity and cannot simply replace extract-based testing. Therefore, a combination of both-extract-based skin test supplemented by in vitro testsis recommended for retaining sufficient sensitivity in routine clinical practice.

CONFLICT OF INTEREST
SW declares having received lecture fees from ThermoFisher Scientific. The other authors have nothing to declare.