To diagnose invasive pulmonary aspergillosis (IPA), galactomannan (GM) detection in serum or bronchoalveolar lavage fluid (BALF) is widely used. However, the utility of proximal airway GM test (from induced sputum or tracheal aspirate) has not been well elucidated.

In this retrospective cohort study, we evaluated the diagnostic performance of proximal airway GM in diagnosis of IPA including COVID-19 associated pulmonary aspergillosis (CAPA). Between January 2022 and January 2023, patients who had been tested for GM with clinical suspicion or for surveillance from any specimen (serum, induced sputum, tracheal aspirate, and BALF) were screened. IPA was diagnosed using EORTC/MSGERC criteria, and CAPA was diagnosed following the 2020 ECMM/ISHAM consensus criteria.

Of 624 patients with GM results, 70 met the criteria for proven/probable IPA and 427 had no IPA. The others included possible IPA and chronic form of aspergillosis. The sensitivities and specificities of serum, proximal airway, and BALF GM for proven/probable IPA versus no IPA were 78.9% and 70.6%, 93.1% and 78.7%, and 78.6% and 91.0%, respectively. Areas under the receiver operating characteristic curve (AUCs) were 0.742 for serum GM, 0.935 for proximal airway GM, and 0.849 for BALF GM (serum GM vs proximal airway GM, p = 0.014; proximal airway GM vs BALF GM, p = 0.334; serum GM vs BALF GM, p = 0.286).

This study demonstrates that the performance of GM test from non-invasive proximal airway samples is comparable or even better than those from serum and distal airway sample (BALF).