Introduction

The culprit agents in IgE-mediated hypersensitivity reactions are often detected by careful clinical histories and confirmed by skin prick testing, serum IgE measurements, and, if necessary, provocative challenges. There are many instances, however, when these in vitro and in vivo assays are contradictory, inaccurate, or unethical to perform. Thus, there is a need for a reliable in vitro functional assay to shed light on these difficult cases and to avoid potentially dangerous provocative challenges.

Blood basophils, leukocytes that typically comprise <1 % of the circulating white blood cells, are a substrate for such an in vitro assay. Basophils are similar to mast cells in that they express the high-affinity IgE receptor and secrete cytokines and inflammatory mediators upon stimulation. Early studies of the role of basophils in allergic disease focused on mediator release, such as histamine [1, 2] and leukotrienes [3]. The development of a flow cytometry-based functional basophil assay became possible in 1991, when Knol et al. demonstrated that a surface marker, CD63, is up-regulated on basophils at the same time as basophil degranulation [4]. Over the past 20 years, many advances have been made in improving the basophil activation test and broadening its application in the clinical realm. The aim of this paper is to review some of the recent advances in the performance and application of these functional assays.

Principle of the Basophil Activation Test

Human basophils have secretory granules containing histamine and are capable of secreting cytokines, chemokines, and lipid mediators. Activation of basophils occurs upon cross-linking of the high-affinity IgE receptor (FcεRI) by allergen or artificial cross-linkers. Basophils can also be activated via complement and chemokine receptors [5]. Electron microscopy has demonstrated two distinct pathways of basophil degranulation: piecemeal and anaphylactic degranulation. In anaphylactic degranulation, the cells undergo rapid morphologic changes and exocytosis of intracellular granules containing preformed mediators. In contrast, in piecemeal degranulation, cells secrete granule contents without exocytosis [6, 7]. Basophils express unique surface markers depending upon whether the cells undergo anaphylactic or piecemeal degranulation, which can then be measured by flow cytometry.

CD63, also known as lysosomal-associated membrane glycoprotein-3 (LAMP-3), is a 53-kDa member of the transmembrane-4 superfamily (tetraspanins). In a resting basophil, this protein is located on the membrane of intracellular secretory granules. After stimulation by FcεRI, these granules fuse with the plasma membrane and thus, CD63 is expressed on the surface of degranulated basophils [4]. Early studies suggested that this marker is associated with granules containing histamine, suggesting that CD63 up-regulation can be used as an indirect marker of histamine release [4, 8]. However, recent studies suggest that histamine release may be the sum of both anaphylactic and piecemeal degranulation, and up-regulation of CD63 may be representative of only anaphylactic degranulation [9].

Buhring et al. described an additional basophil activation marker, CD203c, a glycosylated type II transmembrane molecule that belongs to the family of ectonucleotide pyrophosphatsase/phosphodiesterase enzymes and is present on CD34+ progenitor cells, basophils, and mast cells. CD203c is constitutively expressed at low levels on the basophil surface membrane and is quickly up-regulated upon cell activation via allergen or more slowly via IL-3 [10-12]. This up-regulation differs from CD63 in terms of the inciting stimuli and early signaling events, which suggests that CD203c may be associated with piecemeal degranulation [13-15].

An alternate measure of basophil activation includes the degree of phosphorylation of intracellular molecules, such as p38 MAPK (mitogen-activated protein kinase) [16]. Furthermore, three new basophil activation markers (CD164, CD13, CD107a) have been identified through screening of monoclonal antibodies against resting and activated CD203c basophils [15]. Only CD164 has been validated in a study of pollen allergy, but further correlative and mechanistic studies need to be performed [17].

Update on the Performance of Basophil Activation Tests

Despite the fact that the BAT has been validated for a wide-range of IgE-mediated conditions, there is still considerable variation in the performance of this test. The European Interest Group for the Evaluation of BAT in clinical routine (EuroBAT), a working group of the European Academy of Allergy and Clinical Immunology (EAACI), is attempting to establish standardized techniques. Below, we summarize some of the recent findings regarding the technical performance of the BAT.

Sampling and Preservation of Basophils

Until recently, it was unclear how quickly basophils needed to be processed for optimal expression of CD63 and CD203c. In 2009 and 2010, Sturm et al.(•) performed two studies, where they demonstrated that storing basophils for 24 and 48 h at 4 °C resulted in decreased CD63 and CD203c expression. Furthermore, the authors noticed a decrease in CD63 and CD203c expression after storage for just 4 h, at the lower concentrations of anti-IgE stimulation [18, 19•]. It was thus recommended that basophils undergo processing within 3 h of sampling in order to optimize viability and functionality. In contrast, Sousa et al. found that CD63 expression was diminished only after storage for 48 h at 4 ° C, but this may be due to the small sample size (n = 3). Similar shifts were seen in CD63 expression between blood preserved in acid-citrate dextrose (ACD) or EDTA tubes [20].

Basophil Specific Markers

A common strategy to identify basophils in peripheral blood samples by flow cytometry is based on their surface expression of IgE. However, IgE surface levels vary considerably between individuals [21], and it has been demonstrated that up to 50 % of peripheral leukocytes expressing IgE are actually monocytes [22]. However, the search for new identification markers has resulted in many potential candidate markers: CCR3, CRTH2, CD203c, and CD123.

CCR3, or eotaxin CC chemokine receptor-3, is expressed on basophils, mast cells, and Th2 lymphocytes. This marker is the basis of the commercial Flow2-CAST assay, in contrast to Flow-CAST, which identifies basophils based on IgE. In a study on patients with confirmed beta-lactam allergy, Eberlein et al. found that the Flow2-CAST assay had similar specificity but slightly higher sensitivity (55 vs. 53 %) compared to the original Flow-CAST [23]. As CCR3 is also expressed on T cells, the addition of CD3 was recommended to eliminate contamination [24]. However, a different study found that CCR3 had less inter-individual variability as a single marker for basophils relative to IgE and CD123, even without an additional T-cell marker [21].

CRTH2 is also expressed on basophils, eosinophils, and Th2 lymphocytes. Basophils can be further identified within this subset on the basis of side-scatter (to differentiate from eosinophils) and the secondary marker, CD3 (to differentiate from T cells) [25]. Boumiza et al. validated this gating strategy in 18 patients with latex or dust mite allergy [26], but to our knowledge, further studies with this technique have not been published.

Finally, CD123, a subunit of the IL-3 receptor, is highly expressed on basophils and plasmacytoid dendritic cells in the peripheral blood. Plasmacytoid dendritic cells can be further excluded on the basis of either an anti-HLA-DR or anti-BDCA-2 antibody. In comparing approaches to identify basophils based on the markers CD123, CCR3, or IgE, CD123 was found to have similar expression levels to CCR3 and superior to that of IgE. However, the variability of CD123 expression was significantly larger than that of CCR3, which led the authors to conclude that CCR3 is a superior basophil identification marker [21].

Basophil Activation Markers: Comparative Studies of CD63 v. CD203c

As mentioned above, most basophil activation tests are based on the activation markers CD63 and CD203c. A few recent studies have evaluated the differential expression of these two markers in various allergic conditions. In confirmed IgE-mediated amoxicillin allergy, the sensitivity for CD203c was found to be far superior to that of CD63 (60 vs. 20 %) [22]. In contrast, the same group demonstrated that CD63 expression was up-regulated more frequently than CD203c in patients with non-allergic NSAID hypersensitivity [27]. In patients with venom hypersensitivity, CD203c and CD63 were found to have similar kinetics, with maximum expression detected after 20 min of allergen stimulation. Furthermore, the addition of 300 pM of IL-3 for 10 min prior to stimulation was found to optimize CD63 expression but actually decreased CD203c expression to antigen stimulation in a dose-dependent manner. It was also demonstrated that neither CD63 nor CD203c expression was influenced by prior in vivo ingestion of the antihistamine desloratidine [18, 19•]. This was further confirmed in vivo with CD63 expression in pollen allergic patients [28]. While CD63 and CD203c may represent different pathways of basophil activation and degranulation, they have both been validated as acceptable markers for the basophil activation test.

Update on the Application of Basophil Activation Tests

The basophil activation test has been shown to be sensitive and specific for the diagnosis of IgE-mediated hypersensitivity reactions to hymenoptera venom, pollens, foods, natural rubber latex allergy, and drugs. Many of these validation studies have been extensively reviewed elsewhere. Recent advances over the past 3 years in the application of the basophil activation test to allergic disease are summarized in Table 1.

Table 1 Clinical application of the basophil activation test in allergic disease
Full size table

Drug Allergy

Aspirin and NSAIDs

Aspirin and NSAID intolerance is a very heterogeneous disorder, manifested by urticaria, angioedema, sinusitis, asthma, rhinoconjunctivitis, and anaphylaxis. It is accepted that many of these reactions are not IgE-mediated but rather result from inhibition of cyclooxygenase-1 (COX-1) and thus unregulated synthesis of cysteinyl leukotrienes and mediator release from mast cells, basophils, and eosinophils [29]. This has made the study of basophil activation tests particularly problematic, in that subjects are often included who do not have true IgE-mediated reactions to NSAIDs. Abuaf et al. demonstrated this clearly when they looked at 60 patients with NSAID hypersensitivity, characterized by angioedema and urticaria. In addition, 22 of these patients had hypotension, laryngeal edema, dyspnea, abdominal pain, vomiting, or diarrhea after NSAID intake. They found that the sensitivity among those patients without the visceral symptoms was only 21 %, but this increased to 64 % among those with more severe reactions [27]. Similarly, Korosec et al. demonstrated that the BAT among NSAID-intolerant patients may only have diagnostic value for those with anaphylactoid reactions rather than those with asthma/rhinitis symptoms [30]. It has been further shown by Gomez et al. that the timing of the basophil activation test impacts the sensitivity in the evaluation of NSAID allergy. In a study of 51 patients with allergy to pyrazolones, they demonstrated an overall sensitivity of 54.9 %, but this increased to 85.71 % among patients with positive skin tests. However, after 6 months, 60 % of those with previously positive BATs were now negative [31]. Thus, it appears that the BAT may be useful in the evaluation of severe NSAID hypersensitivity, but it has limited value in assessing non-IgE-mediated reactions or milder reactions.

Radiocontrast Media (RCM)

The incidence of hypersensitivity reactions to radiocontrast media has decreased with the introduction of nonionic formulations. However, even with these newer agents, 0.7–3.1 % of patients will experience a mild reaction and approximately 0.04 % will experience a severe immediate hypersensitivity reaction [32, 33]. It was previously thought that all of these reactions were the result of non-IgE-mediated mechanisms; however, recent studies have shown the benefit of skin prick testing in identifying a causative agent in many of these patients [34]. Thus, a few recent studies have evaluated the utility of using the basophil activation test in the diagnosis of RCM allergy. One case series described three patients who all experienced anaphylaxis coincident with the administration of gadolinium-derived agents. In each case, the causative agent was determined by a positive skin test and BAT [35]. Finally, in 2011, Pinnobhun et al. performed BATs to five different radiocontrast media agents at two different dilutions in 26 patients with a history of RCM allergy. They found that the BAT sensitivity ranged from 42.6 to 61.5 % and specificity ranged from 88.4 to 100 % [36], which is comparable to that of many other drugs that have been validated through BAT.

Fluoroquinolone Antibiotics

Hypersensitivity reactions to fluoroquinolone antibiotics are difficult to diagnose because of the false positive results of skin testing. Three recent studies have looked at the use of basophil activation tests to assist in the diagnosis of these reactions. In the first study, Ben Said et al. found that skin tests as well as BATs were positive in five of five patients with confirmed quinolone allergy [37]. Aranda et al. then looked at 38 patients with confirmed quinolone allergy and performed BATs to ciprofloxacin, levofloxacin, and moxifloxacin. They found this test to have a sensitivity of 41.7–45 %, but this increased to 79.5 % when ciprofloxacin BATs were included with moxifloxacin in those patients allergic to moxifloxacin [38]. In addition, Rouzaire et al. performed BATs in 34 patients who were being evaluated for quinolone hypersensitivity. They found that 50 % of their subjects had negative BATs. Given that information, they were able to successfully reintroduce quinolones in 15 of the 17 patients [39]. These studies demonstrate that the BAT appears to be a useful confirmatory test in patients being evaluated for quinolone hypersensitivity.

Other Drugs

Case reports in which the basophil activation test was used to confirm the culprit drug involved in hypersensitivity reactions have recently been published for atropine [40], glatiramer acetate [41], methylprednisone [42], and antihistamines [43]. Not only was the culprit agent identified, but in some of these cases, alternate therapies were chosen because of their negative BATs.

Food Allergy

Wheat-dependent Exercise Induced Anaphylaxis (WDEIA)

It was recently discovered that patients with WDEIA could be sensitized to two different wheat proteins: ω-5 gliadin [44] and hydrolyzed wheat protein (HWP) [45]. Chinuki et al. performed BATs in ten individuals with WDEIA, in which five were sensitized to ω-5 gliadin and five were sensitized to HWP. They found that basophils from subjects sensitized to ω-5 gliadin had increased CD203c in a concentration-dependent manner when exposed to purified ω-5 gliadin, but not HWP. Those who were sensitized to HWP similarly had positive BATs to HWP but did not react to ω-5 gliadin [46]. It is thus possible that the BAT may be used as an adjunctive test to diagnose subtypes of WDEIA.

Monitoring Tolerance in Cow’s Milk Allergy

Currently, skin tests and specific IgE are used to predict whether an individual has outgrown their food allergy, but these tests are not always reliable. In a recent study of 112 children with cow’s milk allergy presenting for oral food challenges (OFC), Rubio et al. found that the BAT was significantly higher in children who failed OFCs compared to those who passed. The BAT was found to have a sensitivity of 91 % and a specificity of 90 %, which were higher than those for both skin tests and specific IgE [47]. Ford et al. similarly demonstrated that children who tolerated straight milk and baked milk had lower basophil reactivity to serial dilutions of milk protein than those who were unable to tolerate baked milk in an oral food challenge [48]. These findings suggest that BATs may be used in the diagnostic algorithm of when to safely perform an oral food challenge.

Differentiating Sensitization From True Food Allergy

It is commonly accepted that the presence of IgE or a positive skin test to a food does not always indicate a true food allergy. In 2009, Ocmant et al. performed basophil activation tests on 63 children with documented peanut or egg allergy, 28 children who were asymptomatically sensitized to these foods, and 51 controls. They found that the basophils from the food-allergic children had significantly higher levels of activation to in vitro allergen exposure than the other two groups. Furthermore, only 2 of the asymptomatically sensitized children had activated basophils [49].

Testing the Allergenicity of Foods

The basophil activation test can be used in vitro to measure the potential allergenicity of foods. Sebato et al. published two studies in which they used this technique to study thermally processed peanuts. In the first study, they performed BATs in ten patients with severe peanut allergy to five different thermally processed peanut varieties. They found that the BATs varied considerably between patients and varieties, and these results did not correlate with the results predicted from IgE immunoblotting [50]. In their second study, they sought to determine whether the BAT could be used to detect trace amounts of peanut in food. They performed BATs on five peanut-allergic patients and five controls with peanut-spiked biscuits and chocolates, and they demonstrated that the peanut-allergic basophils responded to the peanut-spiked foods, whereas those of the controls did not [51].

In addition, Dolle et al. used the BAT to investigate the allergic potential of two tomato varieties: Reisetomate and Matina. They performed the BAT among 6 patients with confirmed tomato allergy and discovered that there was greater basophil reactivity to the Matina variety, which correlated to increased clinical reactivity, as measured by increased positivity on skin prick testing and oral food challenges, in 25 patients with tomato allergy [52]. These results, taken together, demonstrate that the BAT may in fact be a useful tool to detect the allergic potential of food.

Other Foods

Over the past 3 years, case reports have been published in which the basophil activation test was used to confirm the culprit food involved in a hypersensitivity reaction. These foods have included sesame [53], ginseng [54], sulfite [55], and beef [56].

Venom Hypersensitivity

Identifying the Culprit Venom with Inconclusive Skin Tests and Serum IgE

Approximately 80–100 % of patients with venom hypersensitivity can undergo successful desensitization with venom immunotherapy [57], but this success relies on the correct identification of the culprit insect. This is often difficult, as venom skin tests and specific IgE can be discordant or inconclusive. In 2009, Ebo et al. found that 54 of 118 (45 %) patients presenting for venom hypersensitivity evaluation had inconclusive in vitro tests, with most of these patients having positive IgE to both wasp and honeybee venom. BATs were performed in all patients and the culprit venom was detected in 38 of those patients with inconclusive results [58]. Similarly, Korosec et al. examined 47 patients with a convincing history of insect venom allergy but negative specific IgE and SPT. They performed BATs and intradermal skin testing in 37 of these patients, and found significantly higher diagnostic sensitivity for the BAT compared to intradermal testing (76 vs. 46 %) [59]. In contrast, in a study looking at seven patients with systemic mastocytosis, a convincing history of venom hypersensitivity, and negative venom skin tests, the BAT was only positive to the negative control in one patient and thus did not provide any useful information [60]. The basophil activation test may be a helpful diagnostic tool for identifying the culprit venom in patients with inconclusive conventional in vitro tests, but further studies in those with systemic mastocytosis need to be performed.

Testing the Allergenicity of Cross-reactive Carbohydrate Determinants (CCDs)

It is common for patients being evaluated for venom hypersensitivity to have IgE against both honeybee and yellow jacket venom, which may be due to true co-sensitization or cross-reactivity to carbohydrate determinants found on both venoms. In 2010, Mertens et al. (••) sought to answer whether these IgE antibodies against CCDs had biologic activity through the use of the basophil activation test. Skin prick testing, venom, CCD-specific IgE measurements, and BATs to both native venom and CCD-depleted venom, were performed in 62 patients with hymenoptera venom hypersensitivity. In the patients who were only positive to one venom on skin testing, but had IgE to CCD, the BAT was positive to both native venoms but only positive in the CCD-depleted skin test venom in 67 % of patients. Conversely, in patients who were only positive to one venom on skin test, but did not have IgE to CCD, the BAT was only positive to the skin test venom in both native and CCD-depleted forms. The authors thus concluded that the CCDs do in fact have biologic activity, but they are likely clinically irrelevant because the skin tests were only positive to one venom [61••]. Recently, Eberlein et al. demonstrated that performing the BAT with horseradish peroxidase, a marker for CCD sensitization, has a sensitivity of 92 % for basophil reactivity to CCDs [62], which may be helpful in distinguishing those patients who are truly sensitized to more than one venom.

Pollen Allergy

The basophil activation test has been used to determine the allergenicity of individual pollens, but few studies have compared the basophil activation test to skin testing and specific IgE in the diagnosis of allergy to multiple inhalants. In 2012, Khan et al. compared the BAT to skin test and sIgE to nine allergens. They calculated a sensitivity of 57–84 % and specificity of only 73–81 %, and they found a large number of false positives with BAT among atopic patients. They thus concluded that the BAT is not sensitive enough to use for the routine diagnosis of individual pollen allergy, which they believe may have been due to non-specific IgE cross-linking in the performance of the CD63 basophil activation test [63]. In contrast, Ozdemir et al. found a sensitivity of 77–100 % and specificity of 100 % when performing the BAT with CD203c expression in grass-allergic patients, and they concluded that this is a reliable tool in the diagnosis of pollen allergy [64].

Monitoring Immunotherapy Responses

Many recent studies have looked at the basophil activation test as a means of monitoring response to immunotherapy. Erzen et al. (••) demonstrated that patients who had a negative sting challenge after completing venom immunotherapy had decreased BATs 1 year after VIT. In comparison, the one patient who did not pass the sting challenge in their study did not have a change in their BAT at the same time point [65••]. Zitnik et al. similarly showed that children who underwent VIT to honeybee had a decrease in their baseline BAT to the lowest concentration of honeybee venom starting at just 6 months of VIT [66]. In contrast, Chicocka-Jarosz et al. examined the use of the BAT in monitoring rush VIT in children with honeybee allergy, and they did not see a difference in the baseline BAT after 40 days [67].

Similar findings were found in studies of SLIT. In patients with pollen allergy, Van Overtvelt et al. did not find a significant decrease in the BAT to various grass allergens after 2 or 4 months of SLIT [68]. In looking at SLIT for natural rubber latex allergy, the BAT to Hev b 6.01 and Hev b 6.02 decreased at 6 months but then returned to pre-treatment levels after 1 year of therapy [69]. Finally, in a Phase I study of timothy grass and dust mite dual-SLIT for pollen allergy, it was demonstrated that the BAT to those two allergens decreased after 24 months of SLIT compared to baseline values [70]. Given these findings, it is likely that the BAT may be a useful marker for immunotherapy success; however, the changes may not be evident until approximately 6 months, and the duration of this effect demands further study.

Similar studies have also recently been performed for food allergy. It has been shown that basophils of patients treated with FAHF-2 were found to have a significant reduction in their CD63 expression after 6 months of treatment [71]. In contrast, basophils from patients with pollen food allergy syndrome who ate increasing amounts of apple over an 8-month period were not found to have any change in reactivity [72]. Further mechanistic studies on basophils in the induction of food tolerance need to be performed to clarify these observations.

Conclusions

The basophil activation test is an in vitro assay in which the activation of basophils upon exposure to various stimuli is measured by flow cytometry. Though this technique has not yet been standardized, IgE, CCR3, and CD123 are commonly used to identify basophils in the peripheral blood samples, whereas CD63 and CD203c are used as markers of IgE receptor activation. The basophil activation test has been validated in many IgE-mediated conditions, including drug allergy, food allergy, venom hypersensitivity, and pollen allergy. Furthermore, in recent years, the application of this test has been expanded to include quinolone and NSAID drug allergy, differentiating sensitization from true allergy, determining the allergenic potential of CCD determinants, and monitoring the success of immunotherapy. The basophil activation test continues to be a useful in vitro tool for the study of allergic disease.