Factors associated with coinfections in invasive aspergillosis: a retrospective cohort study

Objectives To describe the coinfections in invasive aspergillosis (IA), to identify factors associated with coinfections, and to evaluate the impact of coinfection on mortality. Patients and methods We conducted a monocentric retrospective study of consecutive putative, probable, or proven IA that occurred between 1997 and 2017. All coinfections, with an onset within 7 days before or after the first sign of aspergillosis, were identified. Factors associated with coinfections and mortality were analysed by multivariable analysis. Results Among the 690 patients with IA included in the study, the median age was 57 years (range 7 days to 90 years). A coinfection was diagnosed in 272/690 patients (39.4%, 95%CI 35.8–43.2). The location of this coinfection was pulmonary only in 131/272 patients (48%), bloodstream only in 66/272 patients (24%) and other/multiple sites in 75/272 patients (28%). Coinfections were bacterial (110/272 patients, 40%), viral (58/272, 21%), fungal (57/272, 21%), parasitic (5/272, 2%) or due to multiple types of pathogens (42/272, 15%). Factors associated with a coinfection in adjusted analysis were: allogeneic haematopoietic stem-cell transplantation (OR 2.3 (1.2–4.4)), other haematological malignancies (OR 2.1 (1.2–3.8)), other underlying diseases (OR 4.3 (1.4–13.6)), lymphopenia (OR 1.7 (1.1-2.5)), C-reactive protein >180 mg/L (OR 1.9 (1.2–3.0)), fever (OR 2.4 (1.5–4.1)), tracheal intubation (OR 2.6 (1.5–4.7)), isolation of two or more different Aspergillus species (OR 2.7 (1.1–6.3)), and the presence of non-nodular lesions on chest computed tomography (OR 2.2 (1.3–3.7) and OR 2.2 (1.2–4.0)). Coinfections were independently associated with a higher mortality at week 12 (adjusted HR 1.5 (1.1–1.9), p < 0.01). Conclusions Coinfections are frequent in IA patients and are associated with higher mortality.