Results
The results of two patients were excluded from the analysis: one patient with an absolute basopenia (excluded from BAT analysis) and one patient with a deep lymphopenia (<300/mm3, excluded from LST analysis). Negative and positive controls were acceptable for BAT (median 5.19%, range 4.16–5.50, and 85.92%, range 44.99–97.85, respectively) and LST [median 1.52% (1.10–1.97) and 65.39% (31.81–86.89) for CD8 LST, mean 1.29% (1.00–1.80) and 68.32% (45.89–85.73) for CD4 LST]. Status of lung transplantation did not influence the level of basophil activation in anti-RFcεI positive control (median 18.64, range 6.48–21.23 and median 18.54, range 11.96–21.90, p = 0.53, for transplanted and non-transplanted patients, respectively) and with all the mold extracts (median 1.68, range 0.69–4.28 and median 1.35, range 0.62–2.11, p = 0.31 for Pen; 1.35, range 0.66–6.71 and 1.38, range 0.77–20.37, p = 0.07 for Af; 1.29, range 0.60–4.48 and 1.27, range 0.72–7.77, p = 0.44 for Alt).
Figure 1A shows that 15 patients (56%) had a positive BAT with at least one mold extract: 10 BAT Af, 6 BAT Pen, and 4 BAT Alt positive. Three patients and one patient had a positive BAT with two or all the three extracts, respectively. BAT showed low levels of activation in controls, in Af-sensitized patients, and in mold-colonized patients. The ABPA group, in which no patient was not lung transplanted, responded with higher basophil activation to Af as compared to other groups, but statistical significance was not reached due to the small sample size (three patients, p = 0.066). Af-induced basophil activation in these patients was higher than responses to Pen and Alt (mean 18.48, range 17.00–21.58 for Af vs. 1.02, range 0.54–1.69 for Pen and 0.97, range 0.60–1.21 for Alt), although the significance level was not reached either (p = 0.066). These patients had no positive responses with Pen and Alt extracts. When specific IgE to Af was positive (>0.10 kUA/L), BAT Af was significantly higher than BAT Pen and Alt (p = 0.008).
Figure 1  Results of the functional cytometric tests with mold extracts expressed as the stimulation index: BAT results (A), T CD8 LST results (B), T CD4 LST results (C), and correlation matrix of functional cytometric tests with mold extracts (D). p-value of the Kruskal–Wallis test, which compares results from Pen, Af, and Alt extracts, is written above each group of patients. ABPA, Allergic bronchopulmonary aspergillosis; Af, Aspergillus fumigatus extract; AF-S, Af-sensitized patients; Alt, Alternaria extract; BAT, Basophil activation test; Control, Control patients (without any Aspergillus-related disease); Fungal Colo, Fungal colonized-patients; LST, Lymphocyte stimulation test; Pen: Penicillium extract. Figures 1B,C describes results of lymphocyte activation, which was an infrequent finding with any mold extract and in all patient groups: mean for CD8 LST 3.40 (0.99–7.31), 3.26 (2.84–3.67), 1.87 (1.12–2.67), and 1.82 (1.04–3.38); mean for CD4 LST 1.71 (0.57–3.17), 3.39 (1.76–5.02), 1.23 (0.84–1.49), and 1.50 (0.50–2.69) in control, AF-S, ABPA, and fungal colonization groups, respectively.
Levels of CD4 and CD8 LST did not differ as a function of lung transplantation status or with the level of specific IgE to Af (data not shown). However, the only patient who displayed a strongly positive CD8 and CD4 LST with all the extracts is worth of notice. This 29-year-old woman, lung transplanted with no detectable sIgE to Af presented during the initial investigation with pulmonary micronodules, mucus plugging, and ground glass appearance (thickening and impaction of bronchioles) as evidenced by high-resolution computed tomography chest. Despite this evocative presentation, ABPA was not diagnosed because of the absence of humoral IgE and IgG responses to Af. However, she developed an ABPA 3 months after.

Correlation Analysis
The correlation matrix at Figure 1D showed a strong positive correlation of both CD4 and CD8 responses to each mold extract (correlation coefficient r-values ranging from 0.56 to 0.92, p < 0.001). In contrast, neither basophil responses to distinct mold extracts nor basophil and lymphocyte responses to a given mold extract were statistically significantly correlated (r-values ranging from −0.24 to 0.09).