Abstract
When one examines reviews regarding diagnostic allergy testing, it generally is noted that the advantage of skin testing is that it uses testing to an end organ, is less expensive, is more sensitive, and provides immediate results, which are easily visualized by the patient. In the case of in vitro testing, it is thought to be more helpful for patients with dermographia or widespread dermatitis, is not affected by H1 or H2 antagonists, and is thought to be safer than skin testing. This may not be entirely correct because studies actually demonstrated phlebotomy to be associated with more morbidity than skin-prick testing. Other concerns regarding both tests include allergen quality, the need from proficiency testing, and the fact that these tests should be used for corroborating a diagnosis as opposed to a screening test for allergy.
When one examines reviews regarding diagnostic allergy testing, it is generally noted that the advantage of skin testing is that it uses testing to an end organ, is less expensive, and is more sensitive, and it provides immediate results that are easily visualized by the patient. In the case of in vitro testing, it is thought to be more helpful for patients with dermographia or widespread dermatitis, is not affected by H1 or H2 antagonists, and is thought to be safer than skin testing. This may not be entirely correct, as demonstrated here.
When considering cost, the Medicare payment schedule (2015) for skin testing is $6.44, versus $7.10 for in vitro testing (per allergen). This difference is likely to be less of an issue, however, in the future because new chip technology may reduce the cost for in vitro testing in the future. 5
There have been several reports in the lay press, e.g., The New York Times article entitled “Allergies, a single shot may suffice,” which highlighted the superiority of serum immunoglobulin E (sIgE) testing compared with skin testing. 6 In this article, the authors describe the pain associated with the multiple pricks needed when performing skin testing and state that the advantage of in vitro testing is, “It was one needle prick and then it was all over.” 6 Interestingly, the literature would paint a very different picture. In a study by the College of American Pathologists that examined patient satisfaction and complications in 630 institutions by examining 23,783 patients after phlebotomy, they found that, although the mean time was 18.8 minutes, >2% of patients required >1 hour and up to 11 attempts to attain a successful blood draw. Furthermore, they demonstrated that >16% of patients experienced ecchymosis, with a mean size of 15.1 mm. 7
In the National Health and Nutrition Examination Survey II survey that examined 16,204 patients (range, 6–74 years of age) who underwent several procedures, including eight skin-prick tests (SPT) as well as venipuncture, they, surprisingly, found that it was venipuncture that was associated with more adverse effects. Specifically, they found that, after venipuncture, 26 of the patients had syncope, 39 had near syncope, and 12 experienced malaise compared with 6, 0, and 1 subjects, respectively, after skin testing. Interestingly only one subject had what was categorized as an “allergic event” (an asthma attack), and it was after venipuncture, with no such events after skin testing. 8
We have long known that extract potency is of great importance when considering allergy diagnostic testing. This is exemplified in a study by Meiser and Nelson 9 that compared an unstandardized dog extract, which contained ∼5 μg/mL of the major allergen Can d 1, with alum precipitated (A-P) dog (Hollister stick), which contained ∼180 μg/mL. They found that, although 59% of patients had SPT positive results, with a mean wheal of 6.9 mm to the A-P dog, only 35 (59% of the A-P positive group) were positive to the less potent dog extract, which demonstrated a mean wheal size of only 3.4 mm. 9 Just as in the case of skin testing, the allergen used in the solid phase of in vitro testing is of great importance. It is key to ensure that antigenic epitopes are not lost in preparing the solid-phase antigenic epitopes and should be in excess to maximize binding of the immunoglobulin E (IgE) antibody. If not in excess, then there can be competitive binding with spurious reduction in sensitivity from antibodies of other isotypes, specifically IgG.3,10
As in the performance of any diagnostic test, it is imperative that proficiency testing be performed. In the case of in vitro testing, the National Committee for Clinical Laboratory Standards recommends daily performance of quality control measures and has established minimal performance targets for these assays. It recommends that the interassay coefficient of variation in IgE antibody assays should not exceed 15%. They further encourage laboratories to participate in a program of interlaboratory proficiency testing. 10 Such proficiency testing standards give us a sense of comfort when ordering in vitro allergy tests; however, even with such calibration and increased automation, in vitro assays still have flaws. In a study by Williams, 11 six large commercial laboratories were sent blinded samples of the same sera, both diluted and undiluted. Each sample was analyzed for 17 different allergens. When compared with the curve expected from an ideal assay, sadly only two of the laboratories demonstrated precision and accuracy. 11
Likewise, in a study by Wood et al., 12 60 samples for peanut and 20 for soy and mouse-human chimeric IgE antibodies specific for the Bet v 1 and Der p 2 were submitted for sIgE measurement by using three different systems: ImmunoCAP, Immulite, and Turbo radioaller-gosorbent test (RAST), with ImmunoCAP used as the reference. There was poor agreement among the three systems for soy and peanut. The Turbo RAST system showed the most variability and failed to identify 11 subjects with peanut positive results and 5 subjects with soy positive results. With regard to the chimeric antibodies, there were widely disparate results among the three assays, with Immunolite generally overestimating sIgE and Turbo RAST generally underestimating sIgE. 12 Similar variability in results among the assays were demonstrated by Wang et al., 13 which thus reinforced that just because two systems present their results in the same units does not mean that the results are necessarily correct or interchangeable.
In all fairness however, skin testing has also been shown to have the potential for variability. To begin with, the devices for percutaneous testing vary in the degree of trauma that they impart to the skin. Therefore, they differ in the size of positive reactions and also in the likelihood of producing a reaction at the site of the negative control. For this reason, they require different criteria for what constitutes a positive reaction. Comparisons of percutaneous devices have been reviewed elsewhere in greater detail.14–24 Likewise, when comparing the size of skin tests when using the same extract and device, significant variability can be seen among skin test technicians. 22 This should be of no surprise because skin test size varies on the depth of penetration of the cutis vera. 23 What is most bothersome regarding skin testing is the potential for variability in the depth of skin testing by the same technician during the performance of skin tests on an individual patient. 22 This will result in a significant coefficient of variation. For these reasons, it is imperative that all technicians performing skin testing undergo proficiency testing on a regular basis. The procedure for skin testing proficiency has been reviewed else-where. 24
When exploring direct comparison between skin testing and in vitro testing in the evaluation of the sensitivity (low false negative). aeroallergens, one must separate in vitro studies before 1990 to those after because earlier assays that assessed sIgE involved antihuman IgE detecting antibody of variable quality and specificity (low false positive). More modern assays involve the use of monoclonal or mixed polyclonal antibodies with complementary dose-response characteristics, which have ameliorated this concern. 25
In a study by Brand et al., 26 which examined 274 subjects with obstructive lung disease who, via questionnaire, were deemed to be allergic to house-dust mite and compared skin testing, via the intradermal method, with sIgE with ImmunoCAP testing, the investigators found that skin testing had a greater diagnostic value (a calculation based on both sensitivity and specificity) compared with in vitro testing. 26 In a study by Wood et al., 27 patients with cat allergy (proven via a cat exposure model) underwent a comparison of skin testing versus sIgE testing (Phadezyme assay was used initially, and the Pharmacia CAP-RAST was used later in the study). They found the overall sensitivity to be greater with SPT (79.2 ± 8.9 versus 69.2 ±10.5), whereas specificity was higher with in vitro testing (100 versus 90.6 ± 6.4). Their conclusion was that sIgE is highly specific but somewhat less sensitive than SPT.
In a more recent study, Sharma et al. 28 examined a group of laboratory workers who work with mice, sensitive to mouse allergen as confirmed by nasal allergen challenge and found that 83% of these workers demonstrated a positive SPT response, whereas only 70% of this group had positive sIgE (ImmunoCAP assay) to mouse. 28 Furthermore, specificity was also higher for SPT (94%) when compared with sIgE (91%). They thus concluded that mouse-specific IgE appeared to be less useful than SPTs in the diagnosis of mouse allergy.
Finally, a cross-sectional study was carried out with a sample of 2167 subjects, who underwent SPT and ImmunoCAP testing to a group of perennial and seasonal allergens. The subjects’ nasal allergy symptoms were then stratified into the following: symptoms after exposure to indoor allergens only, symptoms after exposure to outdoor allergens only, and symptoms after exposure to both indoor and outdoor allergens. 29 They found that both skin test and sIgE reactivity to indoor and outdoor allergens were significantly related to their corresponding nasal symptom groups. Odds ratios increased with an increasing number of positive skin test results or increasing levels of sIgE to allergens in all three nasal symptom groups. For each allergen, a positive skin test result together with a positive sIgE measurement were the strongest predictors of nasal symptoms. Specifically, they demonstrated that SPT was slightly more sensitive for indoor allergy versus sIgE (39.1% versus 36.6%) as well as outdoor allergy (43.7% versus 41.5%), but, interestingly, the greatest sensitivity was seen when using these two tests in tandem (44.6% for indoor allergy and 48.6% for outdoor allergy), which indicates that both skin testing and sIgE measurement may be considered complementary to one another in diagnosing allergic rhinitis. Although beyond the scope this discussion regarding aeroallergy testing, this has also been demonstrated in Hymenoptera sensitivity. 30
As mentioned earlier, innovation in technology regarding in vitro testing is on the horizon. Specifically, biochip technology has advanced in recent years, which led to the development of DNA microarrays, which is a powerful technology that makes it possible to monitor the expression levels of a multitude of genes on miniaturized formats. Subsequently, protein microarray technologies, which are technically more challenging (due to the heterogeneity of the molecules and the need to preserve their tertiary structure after binding to the chip), has developed protein microarrays designed for the detection and quantification of proteins and their function. 31
As a result of the progress in molecular allergy and biochip technology, a miniaturized allergy test has been developed. The principle of this allergen chip is a reverse immunoassay that consisted of recombinant allergens in nanoliter quantities as capture molecules immobilized on glass slides. With this, low volumes of serum can be screened for allergen-sIgE by fluorescence-labeled anti-IgE. This allows simultaneous analysis of a multitude of single allergens to be probed for IgE binding, which allows the identification of the disease-causing molecules, i.e., also the target molecules for specific immunotherapy in a single test. 32 The use of these component resolved diagnostics (CRD), have become a major addition to our evaluation of patients with peanut sensitivity. 33 Similarly, they appear useful in the evaluation of aeroallergens. The results of these analyses could be the basis for patient-tailored allergen preparations to improve the specificity and efficacy of specific immunotherapy and reduce the risk of sensitizing patients to other allergens present in an allergen extract to which they are not yet sensitized.
As an example, there has long been concern that sensitization to profilins and other cross-reacting molecules could impede the proper allergy immunotherapy prescription in patients who are polysensitized patients and have pollen-related allergic rhinitis. In these patients, the use of CRD might optimize immunotherapy prescriptions by more finely identifying the disease-eliciting pollen sources. A recent study by Stringari et al. 34 explored this issue by examining the effect of CRD on immunotherapy prescriptions in children with pollen-related AR. In their study, >600 children with significant pollen associated allergic rhinitis underwent SPT to grass, cypress, olive, mugwort, pellitory, and/or Betulaceae pollen. The investigators considered positive skin test results to be clinically relevant if symptoms occurred during the corresponding peak pollen season. They then measured IgE to Phl p 1, Phl p 5, Bet v 1, Cup a 1, Art v 1, Ole e 1, Par j 2, and Phl p 12 (profilin) by using ImmunoCAP. 34 When comparing the prescription based on skin test and/or symptom responses to those based on CRD results according to GA 2 LEN-European Academy of Allergology and Clinical Immunology guidelines, the addition of CRD significantly changed the composition of allergen immunotherapy, which resulted in a change in 42% of the immunotherapy components. As noted by the investigators, further study is needed to see if the hypothesis that CRD-guided prescription improves allergy immunotherapy efficacy is true. It is certain that much more data will be published in the arena of CRD in the upcoming years. For those wishing more information, a suggested resource is the WAO-ARIA-GA 2 LEN consensus document on molecular-based allergy diagnos-tics. 35
In conclusion, when one analyzes the available data regarding skin testing versus in vitro testing in the diagnostic evaluation of aeroallergy, it would appear that, presently, skin testing is more sensitive and cheaper; however, future innovation in the field of chip technology will likely have a great impact on this comparison. Although many believe that in vitro testing has less variability and is less painful, the literature would bring this belief into question. In the end, no diagnostic allergy test is perfect, and clinical history must be factored into the equation because a positive test result solely indicates sensitivity, which is not the same as clinical allergy. This debate will need to be reevaluated with regularity in the upcoming years.