Laboratory‐based allergy diagnosis: should we go with the flow?

Allergy diagnosis is an area fraught with difficulties. It is, however, of paramount importance not only to confirm the allergic status of the patient but also to reveal the ‘offending’ allergens in order to practice allergen avoidance and use specific immunotherapies. Careful clinical history taking is vital to diagnosis. However, objective tests are needed for confirmation. Skin tests are commonly used. These may not always be possible, can be contra-indicated in some individuals and may result in false positive or false negative results. Indeed both skin testing and other challenges, e.g. oral or inhaled, may cause anaphylactic reactions in some patients. Thus laboratory-based diagnostic tests are needed. Although total IgE concentrations are measured, specific IgE concentrations are of more value to the allergist/clinical immunologist. A series of other in vitro tests, such as leukotriene measurements using the cellular allergen stimulation test (CAST), basophil histamine release tests and basophil activation tests, are available and have been reviewed recently [1]. In the last 5 years, an increasing number of papers have appeared in the literature describing the potential use of flow cytometry in the diagnosis of allergic disease [2–13]. The majority of these reports have concentrated on allergen-induced basophil activation and the tests are commonly referred to using the acronyms BAT (basophil activation test [4,6,11,13]) or BASIC (basophil altered surface immunofluorescence detected by flow cytometry [8]). The basis of these flow cytometric tests is the demonstration of altered membrane phenotypes on activated basophils. Up-regulation in the expression of CD45 [2], CD11b, CD11c [14], down-regulation of CD62L [14] and decreased numbers of IgE-positive basophils following in vitro stimulation have been used to assess activation status. However, the most commonly used marker in basophil activation studies is CD63. This tetraspan granule protein is the only known marker that appears de novo on the basophil surface membrane during activation [15]. In addition, CD63 expression correlates strongly with degranulation [6,15], making CD63 an ideal surrogate marker of basophil activation. Some examples of how these assays have been applied to allergy diagnosis are provided below. Induction of CD63 by Dermatophagoides pteronyssinus (Der p) and Dactylis glomerate (Dac g) was reported, in patients with confirmed house dust mite and grass pollen allergy, by Cozon et al. [5] using a whole blood assay. The authors also demonstrated that increased assay sensitivity could be achieved by pre-incubation of blood with IL-3. Use of the commercially available BASOTEST® kit was described by Paris-Kohler and coworkers for the diagnosis of cypress pollen allergy [9]. The method provided a sensitive and specific test (91% and 100%, respectively) for cypress pollen allergy and showed superior diagnostic efficiency (95%) to that of specific IgE (88%) determinations. The article by Sanz and colleagues in this issue of Clinical and Experimental Allergy describes a well-conducted investigation into allergen-induced basophil activation [13]. The authors determined Der p – or Lolium perenne (Lol p) – induced CD63 expression on basophils from leucocyte preparations from patients who were sensitive to one or other of the above allergens, from patients who were sensitive to neither of the above allergens and from non-atopic control subjects. Importantly, the authors established the performance characteristics of the assay. Using receiver operator curve (ROC) analyses, a positive cut-off (> 15%) was determined and assay sensitivity (93%) and specificity (98%) calculated. Validation of the assay against a battery of conventional investigations (skin tests, total and specific IgE measurements, histamine release and leukotriene production assays) provided strong evidence for the diagnostic potential of the method. Application of the technique to the investigation of food allergy has been described by Moneret-Vautrin et al. [6]. From their study, the authors concluded that the flow cytometric assay gave comparable results to the leukotriene (LTC4) release test and was more efficient than the histamine release test in the diagnosis of food allergy. A comparative analysis of the BAT assay with clinical history, skin tests, specific IgE, histamine and leukotriene release was undertaken in 45 patients with local or systemic reactions to hymenoptera stings [11]. Positive correlations between the flow cytometric assay and histamine and leukotriene release were observed. Assay sensitivity and specificity were reported as 100% when referenced to clinical history, leading the authors to suggest that the technique should be considered as a suitable method for venom allergy diagnosis. Use of a three-colour flow cytometric assay (IgE, CD45, CD63) was reported by Monneret et al. [10] in four cases of muscle relaxant drug-induced anaphylaxis. Correlation of results with intradermal skin test positivity was observed. An extensive study of basophil activation by muscle relaxant drugs has been performed by Abuaf and coworkers [8] using the BASIC assay, which measures changes in fluorescence intensity as opposed to the percentage of positive cells. The authors demonstrated characteristic dose–response curves, but pointed out that at high concentrations these drugs may act as non-specific basophil activators. Furthermore, cross-reactivity between the drugs was observed. For diagnosis of muscle relaxant allergy, the BASIC test offered a high degree of specificity but the sensitivity of the assay was low. Clearly these assays are at a developmental stage. Consequently, there are a number of issues that should be addressed before these assays enter mainstream diagnostic use. It will be important, for instance, to standardize a number of technical procedures and to establish assay performance characteristics for respective allergens. The question as to whether whole blood or leucocyte suspensions offer the preferred starting material should be considered. There are benefits, in terms of overall assay time and of blood volumes required, in using whole blood-based assays. However, use of small blood volumes (typically 100 µL) inevitably results in low numbers of basophils (often < 500) being available for analysis. Using leucocyte suspensions allows a larger number of cells to be analysed and provides cells for other functional assays (e.g. histamine release), which can be performed concurrently. A comparative study of the whole blood verses leucocyte suspensions is awaited. There is general consensus that CD63 is the best marker for flow cytometric determination of basophil activation. However, is dual-colour flow cytometry, using CD63 and IgE (FcεRI), sufficient for identification of activated basophils? Monneret et al. [7] advocate the use of three-colour (CD63, CD45, FcεRI) flow cytometry. This may be beneficial given that platelets, which also express CD63, could be a contaminant of basophil/leucocyte preparations [15]. However, discrimination of basophils with adherent platelets would require additional platelet-specific markers (e.g. CD41) [4]. Furthermore, as IgE may alter FcεRI detection on basophils [15], the effect of high circulating IgE levels from atopic individuals requires clarification. How does priming with IL-3 affect basophil CD63 expression? Cozon and colleagues [5] have demonstrated that IL-3 is capable of potentiating allergen-induced CD63 expression in a manner analogous to that observed for histamine and leukotriene release. Thus, assay sensitivity may be increased and this may be important when the allergenic stimulus is weak. However, non-specific induction leading to false positive results needs to be borne in mind. This area requires further investigation. The range of allergens that could be investigated by this technique is potentially large and diverse. Consequently, optimization of assay conditions for individual allergens (e.g. allergen concentrations) is necessary. Characteristic dose–response curves have been demonstrated by several groups [2,7–9,11] which form the basis for choosing appropriate allergen concentrations. However, investigation of some drug reactions may be problematic due to cross-reactivity and non-specific basophil activation at high concentrations [8]. Furthermore, in other basophil activation systems, extracts from local fresh food have performed better than commercial extracts [16]. Finally, setting positivity levels, for individual allergens, requires careful assessment (see below). In our recent editorial [1], we listed three criteria to be fulfilled before a test should enter the diagnostic repertoire for allergy testing: 1 The test must offer good sensitivity, specificity, and positive and negative predictive values, which have been established in well-conducted clinical trials which include an accepted ‘gold standard’ method. 2 The test must be reproducible and technically straightforward. 3 The test must be cost-effective. The study described by Sanz et al. [13] fulfils the first criterion. In a group of 104 patients and 62 controls they achieved a sensitivity of 93%, a specificity of 98%, a positive predictive value of 99% and a negative predictive value of 89%. Although, a true ‘gold standard’ method for in vitro allergy diagnosis is lacking, the authors used the currently accepted methods for allergy diagnosis, i.e. clinical history, positive skin test and raised allergen-specific IgE. It also fulfils the second criterion in that the intra-assay variation between duplicates was 0.66 (n = 32) and the method is technically straightforward. While no information was provided regarding relative costs of the assay, other investigators [7] have indicated that reagent costs compare favourably with other basophil functional assays. However, the ultimate test of cost-effectiveness will depend on evaluation of the diagnostic potential of flow cytometric assays in clinical practice. We believe that the evidence available suggests that clinical immunologists and allergists should now go with the flow for allergy diagnosis. We recommend that the following studies, using a large number of patients and controls, be performed as a matter of urgency: 1 Comparison of basophil activation using whole blood samples and leucocyte suspensions from the same subjects. 2 Study of basophil activation in the presence and absence of IL-3 priming. 3 Establishment of positivity criteria for a range of common allergens, using ROC analyses. In order that these data be obtained in the shortest possible time, we recommend that the studies be performed in a multicentre trial.