Objectives: To further understand the pathogenic mechanisms of AERD and investigate the potential biomarkers for AERD using metabolomics.
Subjects and Method: Forty five AERD, 44 aspirin tolerant asthma (ATA) and 28 normal controls were enrolled. Urine and serum samples were collected from all asthma subjects before and after lysine-aspirin bronchoprovocation test (Lys-ASA BPT). The metabolites of LTE4 and PGF2α were analyzed using liquid chromatography with mass spectrometry.
Results: Baseline serum and urine LTE4 levels were significantly higher in AERD than ATA (18.14 ± 7.19 pg/ml vs. 14.65 ± 4.39 pg/ml, P = 0.007; 7.36 ±13.19 pmol/mg creatinine vs. 2.69 ±3.62 pmol/mg creatinine, P=0.047, respectively). The ROC curve analysis that discriminates AERD from ATA indicated the cutoff value of the serum LTE4 was 14.5 pg/ml, with 71.1% sensitivity and 52.3% specificity (AUC = 0.651, P = 0.014) which was improved further if combined with serum periostin level (81.4% sensitivity, 63.4% specificity, AUC = 0.761, P < 0.001). However, the ROC curves were not formed using urine LTE4 level. There were no significant correlations between serum and urine levels of two metabolites. Urine LTE4 levels were significantly correlated with those of PGF2α before and after Lys-ASA BPT (r = 0.562, P < 0.001; r = 0.627, P < 0.001). Baseline urine LTE4 and PGF2α were significantly correlated with fall of FEV1 % after Lys-ASA BPT (r = 0.463, P = 0.008, and r = 0.610, P <0.001, respectively).
Conclusion: Serum LTE4 level can be a useful serum biomarker for representing the phenotype of AERD considering serum collection is simpler than any other sample collections. Overproduction of LTE4 and mast cell activation can contribute to induce bronchoconstriction to inhaled aspirin in the pathogenic mechanisms of AERD.