METHODS: Case notes of 80 patients (all with severe asthma and peripheral blood eosinophilia >500/ul) suspected of ABPA were retrospectively analyzed. Total IgE, specific IgE (m3IgE) and specific IgG Aspergillus fumigatus (Gm3) were measured using ImmunoCAP. Patients were assigned into 3 groups based on Modified ISHAM criteria: definite, possible, not ABPA. Gm3 levels were excluded at this assignment.
RESULTS: Of 80 patients (48 male, age range 19-86 years), 30 (37.5%) had positive specific IgE against Aspergillus fumigatus (>0.35kUA/L). 13 patients had ‘definite’ ABPA, 15 ‘possible’ and 52 ‘not ABPA’ group. Median total IgE levels of patients with definite and possible ABPA were 2144kU/L and 2597kU/L respectively (non-signifcant), while median m3IgE levels were 4.35kUA/L and 1.47kUA/L (p<0.05). Mean Gm3 levels compared using ANOVA (Statgraphics Centurion Professional XV, Statpoint Inc) showed a significant difference between definite and other groups (p<0.001), that was confirmed using multiple range tests (Fisher's least significant difference procedure). Mean Gm3 level in definite ABPA was 125.17mgA/L (±SD 54.84, 95%CI 92.03-158.32), while mean Gm3 levels in possible and not ABPA were 18.61mgA/L and 30.05mgA/L respectively. Sensitivity for definite ABPA was 69% and 85% using 90mgA/L and 60mgA/L, with specificity 100% for both.
CONCLUSIONS: Positive Gm3 supports a diagnosis of ABPA but low level false positives are common. Using a higher cut-off of 60mgA/L gave good sensitivity while maintaining excellent specificity.