Standardizing a Four-Marker Basophil Activation Test: Influence of Storage Time, Temperature, and IgE Levels

Introduction

Allergic diseases affect approximately 25% of the population worldwide.^1–3 Current diagnostics, such as skin prick testing (SPT) and allergen-specific IgE (sIgE) detection, lack sufficient specificity and correlate poorly with clinical symptoms.⁴ Provocation testing constitutes the gold standard for the definitive diagnosis of food and drug allergies. However, they are rarely performed in clinical practice due to being labor-intensive, costly, and posing potential safety risks to patients.^5,6

Accumulating studies have demonstrated the important role of basophils in both immediate allergic reactions and chronic inflammation.⁷ The basophil activation test (BAT), which measures IgE-mediated allergic reactions in vitro, has gained increasing attention in recent years. Following allergen-specific and non-specific stimulation, basophils rapidly degranulate, releasing histamine, leukotrienes, and cytokines. They also upregulate surface activation markers, most commonly CD63 and CD203c, which can be quantified by flow cytometry.⁸ This method demonstrates higher specificity in allergy diagnosis.^9–12

As an in vitro diagnostic method, BAT offers considerable safety advantages by eliminating direct allergen exposure while preserving diagnostic accuracy. BAT is recommended when routine clinical diagnostics (eg., SPT and specific IgE testing) yield ambiguous or clinically discordant results, or when in vivo testing poses unacceptable risks to patients, or when necessary reagents for conventional tests are unavailable.⁹ However, its widespread clinical adoption is hindered by two unresolved methodological barriers: (1) the absence of quantitative, cost-benefit evaluated guidance for gating strategy selection, and (2) conflicting, stimulus-specific storage recommendations that remain to be reconciled.

Regarding the first barrier to gating strategy selection, prior studies have compared gating strategies descriptively, without quantifying the incremental specificity gain relative to cost increase. While multi-marker strategies (eg., incorporating CD123, CD203c, CCR3, and HLA-DR) enhance sensitivity and specificity, their high reagent cost and complexity limit clinical feasibility. Simplified approaches using fewer markers risk contaminating the target population with other cell types.^13–16 Evidence-based optimization is needed to identify the minimal marker panel achieving acceptable specificity without prohibitive cost.

Regarding the second barrier for storage conditions, existing studies report inconsistent time-temperature combinations without distinguishing allergen-specific from non-specific responses. Some studies advocate processing within 4 h of collection at RT,^15,17 while others demonstrate 24 h stability at 4°C.¹⁸ Critically, no prior study has compared temperature effects across different stimuli within a unified framework, preventing evidence-based standardization for routine laboratory workflows involving specimen transportation and batch testing.

To address these methodological gaps, house dust mite (HDM) allergy was selected as a clinically relevant model. HDM was chosen because it is globally prevalent, often presents with discordant sIgE and clinical symptoms, and is commercially available in standardized formulations suitable for methodologically rigorous evaluation.¹⁹

This study addresses the lack of analytically validated, cost-efficient BAT workflows. Our primary objective is operational feasibility: to develop a minimal gating panel with defined cost-efficiency boundaries and unify stimulus-specific storage parameters within a single analytical framework. To achieve this, we conducted three evaluations: (1) quantitative comparison of two-, three-, and four-marker gating strategies to balance analytical specificity against reagent cost; (2) systematic assessment of storage time/temperature effects on basophil responses to allergen-specific (Df) and non-specific (anti-IgE, fMLP) stimuli; and (3) exploratory analysis of correlations between BAT outcomes and serological markers (such as sIgE, total IgE). These findings represent an analytically validated methodological framework for future diagnostic accuracy studies.

Methods

Study Population

Seventy-nine adult patients from the respiratory department were included. Participants were divided into two groups based on symptoms and serological testing. The Dermatophagoides farinae (Df) sensitized group: patients with symptoms of allergic rhinitis or asthma and serum Df specific IgE (Df sIgE) ≥0.35 KU/L.²⁰ Control group: respiratory patients without allergic disease (no allergic symptoms and Df sIgE <0.35 KU/L). Exclusion criteria included: anti-histamine use within 72 hours, systemic corticosteroid treatment within 2 weeks, acute infections, hematological malignancies, pregnancy, or lactation.

Establishment of Basophil Activation Test

BAT was performed according to the following procedure. Fresh venous blood was collected into heparin tubes. Immediately after collection, the tubes were gently inverted 8–10 times, and transported at RT to the laboratory within 30 minutes. Samples were maintained at RT and processed within 6 hours of collection.

For the assay, whole blood was incubated with Df extract (Glallergen Co., Ltd) at 37°C for 15–30 minutes. The optimal concentration of 10 µg/mL was selected based on preliminary titration experiments (serial dilutions from 0.01 to 10 µg/mL tested in 19 Df-sensitized patients) and previous literature on HDM-stimulated BAT.^21,22 Anti-IgE antibody (Sigma-Aldrich, USA) and N-Formyl-Met-Leu-Phe (fMLP; Sigma-Aldrich, USA) stimulation were used as positive controls, while reaction buffer alone served as the negative control.

Following stimulation, samples were stained with fluorescently conjugated antibodies (CD63, CD203c, CD123, and CD45) (Beckman Coulter, USA), subjected to erythrocyte lysis, and acquired by flow cytometry (Navios, Beckman Coulter, USA). Data were analyzed with Kaluza software (Beckman Coulter, USA).

Basophils were identified using three different gating strategies: (1) four-marker strategy: CD123⁺CD203c⁺HLA-DR⁻CD45⁺side scatter (SSC)^low, (2) three-marker strategy: CD123⁺CD203c⁺CD45⁺SSC^low, (3) two-marker strategy: CD203c⁺CD45⁺SSC^low. The four-marker gating strategy was employed as the reference for comparative evaluation within a five-color panel that also included the activation marker CD63. The three-marker and two-marker gating strategies were used within a four-color and a three-color panel, respectively, each also incorporating CD63 as the activation marker.

Basophil activation was quantified using two parameters: (1) the percentage of CD63 positive basophils (CD63⁺%), corrected by subtracting the negative control value, and (2) the stimulation index of CD63 (CD63SI), calculated as the ratio of the mean fluorescence intensity (MFI) of CD63 in stimulated samples to that in unstimulated samples. A positive response is defined as both CD63⁺% ≥5% and CD63 SI ≥2, based on previously established thresholds in BAT studies.^23,24

Patients whose basophils did not respond to anti-IgE stimulation but exhibited a response to fMLP were classified as non-responders according to Pascal M and excluded from activation analysis.²⁵ For Df-specific analyses, only sensitized patients with anti-IgE-responsive basophils were included, unless otherwise specified.

Quantification of Membrane-Bound IgE, Df sIgE, Total IgE, Basophil Count, and Eosinophil Count

We performed a comprehensive correlation analysis between BAT outcomes and key immunological parameters, including basophil membrane-bound IgE (mIgE), Df sIgE, total IgE (tIgE), basophil count, and eosinophil count. All quantitative measurements were carried out using standardized methods: tIgE and Df sIgE levels were quantified using the ImmunoCAP™ 250 system (Thermo Fisher Scientific, Uppsala, Sweden), a fluorescent enzyme immunoassay (FEIA) platform. Basophil and eosinophil counts were determined using automated hematology analyzers (BC-6800 Plus, Shenzhen Mindray Bio-Medical Electronics Co., Ltd., China). MIgE was assessed via flow cytometry using FITC-conjugated anti-human IgE antibodies.

The Effect of Storage Temperature and Duration on BAT Outcomes

To evaluate storage stability, a total of 21 Df-sensitized patient samples were used. The initial BAT analysis was performed within 6 hours after collection (designated as Day 0). Following this, aliquots from each sample were stored under two different conditions: 2–8 °C (low temperature, LT) and room temperature (RT, 18–25°C). Each of the 21 samples was tested at all subsequent time points (day 1: 24–28 h and day 2: 48–52 h) under both storage conditions, following a paired, repeated-measures design.

Statistical Analysis

Continuous data were presented as median and interquartile range (IQR). Categorical variables were expressed as frequencies and percentages. The normality of distribution was assessed using the Shapiro–Wilk test. As the data were not normally distributed, non-parametric tests were used for all group comparisons. Differences between two independent groups (eg., Df-sensitized vs. control) were analyzed using the Mann–Whitney U-test. Differences among paired samples (eg., across storage time points, cell counts) were analyzed using the Wilcoxon signed-rank test. Comparisons of categorical variables between groups were performed using the chi-square test or Fisher’s exact test, as appropriate. Associations between variables were evaluated using Spearman’s rank correlation coefficient (r), with coefficients of determination (r²) to indicate the proportion of explained variance. For these exploratory correlation analyses between BAT outcomes and multiple immunological parameters, p values are reported without formal correction for multiple comparisons, and results should be interpreted as hypothesis-generating. A two-sided p-value < 0.05 was considered statistically significant. Analyses of basophil counts included all participants (including non-responders), whereas analyses of basophil reactivity restricted to participants with responsive basophils. For Df-specific analyses, this subset was further limited to Df-sensitized patients. All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA). Agreement between gating strategies for qualitative BAT outcomes was assessed using Cohen’s kappa coefficient (κ).

Results

Characteristics of the Study Population

A total of 79 patients were included in this study, and 7/79 (8.86%) failed to respond to anti-IgE. Finally, 32 Df-sensitized patients and 39 non-allergic controls were analyzed. The baseline characteristics of the study population are summarized in Table 1. TIgE and Df sIgE levels were markedly higher in the sensitized group compared to the control group. Age and gender distribution were similar between the two groups.

Table 1 Characteristics of Df-Sensitized and Non-Sensitized Patients

Establishment of the Basophil Activation Test

Gating Strategies and Identification of Basophils

Flow cytometric identification of basophils is a prerequisite for measuring their activation in response to IgE-dependent or IgE-independent stimuli. Basophils were identified and gated as CD123⁺CD203c⁺CD45⁺SSC^low cells (Figure 1A and B). A distinct cell population was evident within the CD123⁺CD203c⁺ region (Figure 1A). To confirm basophil identity, we evaluated the absence of HLA-DR. Nearly all gated cells were negative for HLA-DR (Figure 1C), consistent with a basophil phenotype.

Figure 1 Representative gating strategy and immunophenotypic verification of human basophils. Basophils were identified using two distinct gating strategies: CD123+CD203c+CD45+SSlow (A and B) and CD203c+CD45+SSlow (G–H). Basophil identity was confirmed by absence of HLA-DR (C and I). Representative CD63 expression on basophils under negative control, anti-IgE, and Df stimulation is shown for the three-marker gate (D–F) and the two-marker gate (J–L), respectively.

To validate the accuracy of our three-marker gating panel (CD123⁺CD203c⁺CD45⁺SS^low), we compared it with a more extensive four-marker gating strategy (CD123⁺CD203c⁺HLA-DR⁻CD45⁺SSC^low) reported.²⁶ Basophil counts did not differ significantly between the two strategies under any of the stimulation conditions tested (Figure 2).

Figure 2 Comparison of basophil counts and CD63 expression between three-marker (CD123+CD203c+CD45+SSlow) and four-marker (CD123+CD203c+HLA−DR−CD45+SSlow) gating strategies across stimulations.

Given the high specificity of CD203c for basophils, we investigated a simplified, cost-effective gating strategy utilizing CD203c⁺CD45⁺SSC^low without CD123. This approach consistently identified a well-defined cellular population (Figure 1G–H), approximately 92% of which exhibited HLA-DR⁻ phenotype (Figure 1I). However, the basophil count obtained with CD123⁺CD203c⁺CD45⁺SSC^low strategy was significantly lower than that obtained with CD203c⁺CD45⁺SSC^low (two marker gating approach) (Table 2), confirming that the inclusion of CD123 helps exclude non-basophil populations.

Table 2 Comparison of Basophil Counts Obtained by CD123+CD203c+CD45+SSClow and CD203c+CD45+SSClow Gating Strategies

Changes of Molecular Expression Levels

The expression of these membrane markers was analyzed in basophils from all enrolled patients, excluding non-responders. Among markers evaluated, CD45, CD63 and CD203c (expressed as MFI), as well as side scatter, all increased significantly after anti-IgE stimulation compared to unstimulated conditions, whereas CD123 remained unchanged (Figure 3). CD63 exhibited the most pronounced response (32.5-fold increase), consistent with its established role as a principal activation indicator in BAT.¹⁷ Consequently, CD63 was selected as the activation marker for subsequent analyses. Flow cytometric gating resolved two discrete subpopulations upon stimulation: CD63-negative (resting) and CD63-positive (activated) (Figure 1D–F and J-L). Activation was quantified as CD63+% or CD63SI.

Figure 3 Basophil activation characteristics and evaluation of gating strategies. (A–E) Changes in CD63, CD203c, CD123, and CD45 on basophils and side scatter characteristics following anti-IgE stimulation compared with negative control assessed using the three-marker gating (CD123+CD203c+CD45+SSlow). (F–H) Comparison of CD63+% between three- and two-marker (CD203c+CD45+SSlow) gating strategies for anti-IgE, fMLP and Df stimulation. * p<0.05.

No differences in CD63+% and CD63 MFI were observed between the three- and the four-marker gating strategies, whereas the two marker-gating approach yielded significantly lower CD63+% values than the three-marker gating strategy (Figures 2 and 3).

Taken together, these findings indicate that the three-marker gating panel identifies basophils with purity comparable to the four-marker gating reference and higher than the two-marker gating approach, while maintaining equivalent activation readouts-representing a balance between analytical specificity and cost-effectiveness.

Agreement Between Gating Strategies for Qualitative BAT Classification

We evaluated agreement between the three-marker and four-marker gating strategies for qualitative classification (positive/negative) using the predefined threshold. Among the 19 Df‑sensitized patients with responsive basophils, the two gating strategies showed perfect concordance (κ = 1.00, 95% CI: 1.00–1.00), with 16/19 positive and 3/19 negative by both methods.

For comparison between the three- and two-marker gating strategies (n = 72), concordance was 97.2% (κ = 0.917, 95% CI: 0.832–1.000), with two samples classified as positive by three-marker but negative by two-marker, consistent with lower analytical sensitivity of the two-marker gating approach in this cohort.

These findings support analytical equivalence of the three-marker and four-marker gating strategies, with the three-marker gating strategy showing better concordance with the reference standard than the two-marker gating approach.

The Association Between Basophil CD63 Expression and Biomarkers of Allergic Inflammation

IgE levels and eosinophil counts are commonly used laboratory parameters for allergy diagnosis. We therefore examined correlations between these markers and BAT outcomes in Df-sensitized patients (Table 3). Df-induced basophil CD63 expression exhibited moderate correlations with Df sIgE levels (r=0.4352 for CD63 MFI, 95% CI: 0.1020–0.6805) and mIgE levels (r=0.6659 for CD63+%, 95% CI: 0.1925–0.8879; and r=0.6264 for CD63 MFI, 95% CI: 0.1263 to 0.8726), though confidence intervals were wide indicating substantial uncertainty. But a negative correlation was observed with tIgE (r=−0.3798 for CD63 MFI, 95% CI:-0.6496 to −0.02511). In contrast, basophil activation did not associate with basophil counts, eosinophil counts, or fMLP-stimulated CD63 expression.

Table 3 Association of IgE Levels, Eosinophil and Basophil Counts with Df-Induced CD63 Expression on Basophils

We further assessed the correlation between anti-IgE-induced CD63 expression and the aforementioned parameters in the entire study population, excluding non-responders. A weak positive correlation was observed between CD63+% and mIgE levels (r=0.363, 95% CI:0.0161 to 0.6600), whereas a weak inverse correlation was found between CD63SI and tIgE levels (r=−0.2854, 95% CI:-0.4945 to −0.0451). No significant associations were detected between anti-IgE-induced CD63 expression and other variables, including Df sIgE levels, basophil counts, or eosinophil counts (Table 4).

Table 4 Association Between IgE Levels, Eosinophil and Basophil Counts, and Anti-IgE Induced CD63 Expression on Basophils

Additionally, we analyzed the correlation between CD63 expression under different stimulation conditions. The correlation between Df- and anti-IgE induced basophil CD63 expression was also moderate (r=0.4483 for CD63+%, 95% CI: 0.1075 to 0.6949; r=0.5117 for CD63 MFI, 95% CI: 0.1881 to 0.7351), though confidence intervals were wide (Table 3).

The coefficient of determination (r²) indicated that tested markers explained only 8.15% to 44.34% of the total variability in BAT outcomes, leaving 55.66% to 91.85% unexplained.

Influence of Storage Temperature and Duration on BAT Outcomes

To assess the impact of storage time and temperature on basophil reactivity, we assessed responses to anti-IgE, fMLP and serially diluted Df extract under two storage conditions: RT (18–25°C) and LT (2–8°C), at three points post-collection: 0–6 hours (day 0), 24–28 hours (day 1), and 48–52 hours (day 2).

Quantitative Changes in CD63 Expression

We first analyzed the basophil activation magnitude by measuring CD63 MFI. Overall, quantitative analysis revealed significant temporal- and temperature-dependent declines in activation (Figure 4).

Figure 4 The effect of storage time and temperature on basophil CD63 expression stimulated by Df, anti-IgE and fMLP. * p<0.05.

For Df stimulation, CD63 MFI decreased significantly after 1 or 2 days of storage at LT relative to baseline (day 0). A similar but less pronounced trend was observed at RT. For anti-IgE stimulation, CD63 MFI declined significantly by day 2 under both storage conditions, with no significant decreases on day 1. When comparing the two storage conditions directly, both Df and anti-IgE stimulation exhibited a trend toward greater decline in CD63 expression from baseline under LT relative to RT, though this difference was not statistically significant. In contrast, for fMLP stimulation, CD63 MFI was significantly lower in LT-stored samples at both days 1 and 2 relative to day 0. Moreover, at each time point, LT storage led to significantly greater suppression than RT, indicating that fMLP-induced basophil activation is uniquely sensitive to LT storage.

Impact on Qualitative (Positive/Negative) Classification

We next evaluated whether the quantitative declines described above translated into qualitative classification using the predefined positivity threshold (CD63+% ≥5% and SI ≥2). Qualitative outcomes are detailed in Table 5.

Table 5 IgE and CD63 Values at Different Storage Time and Temperature for Df-Sensitized Patients

Anti-IgE stimulation showed high stability. Only 3 of 20 (15%) samples stored at RT and 3 of 21 (14.28%) samples stored at LT yielded negative on Day 1, with no further change on day 2. In contrast, fMLP-induced activation exhibited marked temperature sensitivity. 1 of 20 (5%) RT-stored samples turned negative on day 1, compared to 7 of 21 (33%) LT-stored samples, indicating that RT is required for fMLP control preservation. Df stimulation (10 μg/mL) showed initial stability with only one sample per group converting to negative on Day 1. By day 2, temperature-dependent divergence emerged. 5 of 21 (24%) RT-stored samples, but in only 2 of 21 (9.5%) LT-stored samples lost reactivity. This pattern was amplified under gradient-dilution Df challenge (1–0.01 ug/mL), with 8 of 21 (38%) RT-stored versus 5 of 21 (23.8%) LT-stored assays became negative on day 1.

These data indicated stimulus-specific storage patterns. Quantitatively, CD63 MFI declined significantly with storage. Qualitatively, LT better preserved Df positivity beyond 24 h, whereas fMLP required RT storage, and anti-IgE remained stable under either condition for 24–28 hours.

Discussion

This study describes a cost-efficient BAT workflow with demonstrated analytical validity through systematic comparison of gating strategy and definition of pre-analytical storage conditions. We caution that this establishes a methodological framework rather than diagnostic accuracy.

To date, no consensus gating protocol for basophil identification in BAT has been established, resulting in substantial inter-laboratory and inter-kit variability.^27–30 The analytical equivalence of three- and four-marker gating strategies aligns with prior work demonstrating that CD203c⁺CD123⁺HLA-DR⁻ gating improves specificity compared to CD123⁺HLA-DR⁻ alone.³¹ We extend these findings by showing that explicit HLA-DR staining is redundant when CD203c is combined with CD123, as >97% of CD123⁺CD203c⁺ cells are inherently HLA-DR⁻. Whether this analytical equivalence translates to diagnostic equivalence requires prospective validation (NCT06559319, currently underway).

Compared with the two-marker gating strategy (CD203c⁺CD45⁺SSC^low), the three-marker gating panel demonstrated higher phenotypic purity (97.2% vs. 91.8%) and detection agreement (κ=0.917). The 8.2% contamination rate in the two-marker gating approach suggests potential inclusion of non-basophil populations, supporting the mechanistic rationale for three-marker gating adoption.

Df-stimulated BAT outcomes correlated moderately with Df sIgE and mIgE (r²=0.19–0.44), indicating that the IgE parameters account for less than half of the variability in basophil reactivity. This substantial unexplained variance suggests that discordant profiles—such as elevated sIgE with low basophil reactivity—may represent asymptomatic sensitization rather than true allergy. Evidence indicates that basophil reactivity is governed by factors beyond IgE availability, including FcεRI surface density, SYK signaling pathway activity, and basophil priming status.^32–34 Thus, BAT provides functional information complementary to static IgE measurements.

Our observed Df sIgE correlation (r=0.44) is lower than reports in mite-allergic children (r=0.53–0.60)³⁵ and venom allergy (r=0.60–0.68).³⁶ These discrepancies likely reflect multiple factors: (1) age-related differences, as our elderly cohort contrasts with pediatric populations; (2) absence of IL-3 priming in our protocol, which upregulates FcεRI expression and enhances basophil responsiveness;³⁷ (3) allergen-specific variables, including extract composition differences between mite and venom allergy. Notably, tIgE showed weak negative correlations with Df- and anti-IgE stimulated CD63+%. This unexpected finding may reflect high polyclonal IgE occupancy of FcεRI limiting anti-IgE accessibility, though this mechanism requires direct experimental validation.³⁸

In this cohort, we observed that storage stability depended on both analytical purpose (quantitative/qualitative) and stimulus type. Quantitatively, CD63 MFI declined significantly with storage: fMLP responses were most sensitive to LT, followed by Df and anti-IgE. These differences may reflect stimulus-specific signaling lability: fMLP signals through G-protein coupled formyl peptide receptors (FPRs) that undergo irreversible or slowly reversible transitions between low- and high-affinity states,³⁹ whereas IgE-mediated FcεRI activation appears relatively preserved during short-term storage (24–28 h), though the molecular basis for this stability remains unclear. Df extract, with its complex composition, shows LT sensitivity distinct from anti-IgE.

Qualitatively, positive/negative classification based on the pre-defined positivity threshold proved more robust than quantitative measures. We observed differential storage stability: (1) Df: 2–8°C up to 48–52 hours (9.5% negative-conversion vs. 24% at RT); (2) fMLP: RT within 24–28 hours (5% negative-conversion vs. 33% at 2–8°C); (3) anti-IgE: stable under either condition for 24–28 h. These stimulus-specific requirements enable flexible batch processing.

This study systematically evaluates gating strategies, separates quantitative vs. qualitative storage outcomes across different stimuli, and integrates cost-effectiveness analysis. However, limitations must be acknowledged. First, our single-center clinic cohort, predominantly elderly, lacks healthy controls and detailed atopic histories, limiting generalizability. The absence of a clinical gold standard precludes diagnostic accuracy assessment. Second, the four-marker gating reference standard, though literature-supported, is not an absolute gold standard, and any systematic bias in this reference would affect both strategies. Third, no prospective power calculation was performed, and statistical power was limited by sample size. Post-hoc power analyses indicated 80% power to detect |r|≥0.48 (Df-specific) and ≥0.33 (anti-IgE). Weaker associations may have been missed, and findings should be interpreted as hypothesis-generating. Fourth, formal intra-/inter-assay variability was not assessed, though indirect indicators (perfect gating concordance, consistent storage trends) suggest acceptable reproducibility.

These methodological refinements offer several features relevant to clinical utility: cost reduction, stimulus-specific storage flexibility, and functional complementarity to IgE measurements. However, these features currently represent analytical and pre‑analytical feasibility. Prospective diagnostic validation (NCT06559319) is required to establish clinical utility.

Conclusions

This study describes an analytically validated BAT protocol using a three-marker gating strategy (CD123⁺CD203c⁺CD45⁺SSC^low) within a four-color panel that achieves 25% cost reduction while maintaining equivalence to four-marker gating reference. BAT outcomes showed only moderate correlations with IgE parameters, indicating its role as a complementary functional assay. For storage, quantitative analysis required processing within 4–6 h, whereas qualitative outcomes demonstrated stimulus-specific stability. These methodological refinements enhance feasibility for research applications, with prospective diagnostic validation underway.