COVID-19-associated Pulmonary Aspergillosis in Mechanically Ventilated Patients at 7 US Hospitals: Epidemiology and Estimated Likelihood of Invasive Pulmonary Aspergillosis-Results of the Prospective MSG-017 Study.

Publication date: Jul 01, 2025

There is no prospective, US multicenter study of COVID-19-associated pulmonary aspergillosis (CAPA). CAPA definitions do not differentiate invasive aspergillosis (IPA) from colonization. Validity of single mycologic test results is unclear. We performed a prospective 7-center US study of mechanically ventilated adults with COVID-19 (April 2021-May 2022). Mycoses Study Group (MSGERC) CAPA criteria include host and clinical factors, imaging and test results (histopathology; bronchoalveolar lavage [BAL] culture and/or BAL or serum galactomannan-immunoassay). Proven, putative, and unlikely IPA were defined by clinical criteria. CAPA-unlikely IPA criteria included survival or negative autopsy following no/limited antifungal treatment. IPA likelihood was estimated using sensitivity/specificity of tests from autopsy data. CAPA incidence was 7% (14/212). Independent CAPA risk factors were EORTC/MSGERC host factor and cavitary lesions. Seven percent, 79%, and 14% of CAPA patients had proven, putative, and unlikely IPA, respectively. Respective estimated IPA likelihoods were 84%, 7%-99%, and 1%-8%. Overall, median estimated IPA likelihood was 30%. Patients with CAPA-unlikely IPA had a single positive BAL galactomannan-immunoassay with other negative tests. CAPA mortality (71%) was not impacted by antifungal treatment or significantly different than without CAPA. CAPA incidence was 10% and 16% by European Confederation of Medical Mycology and Public Health Wales definitions, respectively. IPA was unlikely in 75% (6/8) and 57% (13/23) diagnosed by these definitions but not MSGERC. CAPA is associated with high mortality, but IPA’s contribution is unclear. Single positive tests are insufficient for diagnosing CAPA-IPA. IPA likelihood is best estimated by combining test results (both positive and negative).

Open Access PDF

Semantics

Original Article