Surveillance for COVID-19-associated pulmonary aspergillosis.

In The Lancet Microbe, Paul Verweij and colleagues highlighted the challenges in diagnosing COVID-19-associated pulmonary aspergillosis. In this Correspondence, we report our insight.

Several European case series have reported a high incidence of putative COVID-19-associated pulmonary aspergillosis (19–33% of intubated patients).2, 3, 4 In response, we initiated a surveillance strategy for secondary fungal infections in intubated patients with COVID-19 at our tertiary referral hospital in London, UK. Weekly serum 1–3-β-D-glucan (Fungitell, CapeCod, MA, USA) and galactomannan (Platelia Aspergillus EIA, BioRad, Redmond, WA, USA) testing was done. Given the need to limit aerosol-generating procedures, deep respiratory sampling was only done at the physician's discretion. Any bronchoalveolar lavage or endotracheal aspirates were tested for galactomannan as per the manufacturer's recommendations for bronchoalveolar lavage. If positive (galactomannan index [GMI] ≥1·0), Aspergillus PCR was done from 1 mL of endotracheal aspirates deposit, which was ribolysed, and DNA was extracted using a DNA tissue method (EZ1, Qiagen, Hilden, Germany). Real-time PCR was done using AsperGenius assay (Bruker, Bremen, Germany).

From April 12 to May 26, 2020, 62 patients were screened (51 [82%] men, 11 [18%] women, median age 59 years, range 24–74 years). During this period, 24 (39%) of 62 patients died and seven (11%) patients had one or more positive fungal biomarkers (appendix). All serum galactomannan tests were negative (GMI <0·5). 1–3-β-D-glucan was positive (>80 pg/mL) in three (5%) patients: one patient with candidaemia, one patient with suspected candidiasis, and one patient with no diagnosis of invasive fungal infection. 85 endotracheal aspirates were tested for galactomannan; six (10%) of 62 patients had positive results, of which five (83%) of the samples tested positive by Aspergillus PCR and two (33%) of the samples grew Aspergillus fumigatus in culture.

Because of the absence of consensus criteria for COVID-19-associated pulmonary aspergillosis, definitions for post-influenza aspergillosis were applied. None of our patients met this case definition for COVID-19-associated pulmonary aspergillosis, which was problematic for two reasons. First, all patients with severe COVID-19 have pulmonary infiltrates; therefore, the radiological criterion is not useful. Second, bronchoalveolar lavage samples were not available and whether galactomannan positivity of endotracheal aspirates is an acceptable substitute criterion is unclear; this could reflect upper airway colonisation rather than disease. Overall, we clinically suspected two (3%) of 62 patients screened to have COVID-19-associated pulmonary aspergillosis, a substantially lower proportion than in previously reported case series. Notably, both patients had host factors (defined by the European Organisation for the Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium) for invasive aspergillosis (appendix).

Our experience further highlights the difficulty in defining and diagnosing COVID-19-associated pulmonary aspergillosis. There is clearly a need for ongoing surveillance and improved diagnostic strategies for this complication.