Clinical features and risk factors for viral pneumonia complicated with invasive pulmonary aspergillosis in adult patients

Viral pneumonia has recently been defined as a risk factor for invasive pulmonary aspergillosis (IPA). Wauters et al[1] reported that 23% of critically ill patients who were admitted to the intensive care unit (ICU) with swine-origin influenza A (H1N1) viral infection had developed invasive aspergillosis. IPA was diagnosed in 83 of 432 (19%) patients who were admitted with influenza; for patients with influenza who were immunocompromised, the incidence of IPA reached up to 32% (38 of 117 patients), whereas in the non-immunocompromised influenza group, the incidence was 14% (45 of 315 patients).[2] The 90-day mortality was 51% in patients in the influenza cohort with IPA and 28% in the influenza cohort without IPA.[2] The prevalence of IPA is reported between 4% and 9% in critically ill patients with coronavirus disease.[3] Magira et al[4] found that respiratory viral infections caused by both parainfluenza virus and respiratory syncytial virus frequently preceded IPA in patients with leukemia and those who underwent hematopoietic stem cell transplantation. Considering the high incidence and mortality of viral pneumonia complicated with IPA, the risk factors for it should be identified. Hence, we performed a study to explore the risk factors for viral pneumonia complicated with IPA. We retrospectively screened data of hospitalized patients with viral pneumonia from the microbiology laboratories of six hospitals in China between August 2016 and December 2019. The inclusion criteria were as follows: (1) All patients were defined as the presence of a new or progressive infiltrate using chest radiography complicated with two or more of the following symptoms: fever, cough, expectoration, rhinorrhoea, sore throat, or dyspnea, or a diagnosis of pneumonia by the attending physician. (2) Patients who had a positive result of reverse transcription polymerase chain reaction (PCR) for respiratory virus nucleic acids in nasopharyngeal swab, sputum, or bronchoalveolar lavage (BAL) fluid. (3) Patients who received galactomannan (GM) test or sputum fungal culture. The following patients were excluded from this study: (1) aged <18 years; (2) patients who had no symptoms of infection; (3) non-pneumonic patient (such as upper respiratory tract infections); and (4) the GM or sputum fungal culture result was not available. The Ethics Committee of China-Japan Friendship Hospital (No. 2015-86) approved this retrospective study and orchestrated centralized collaboration and approval of all participating institutions. The written informed consent was waived. Diagnostic criteria were set according to the revised definitions of invasive fungal diseases by the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group.[5] IPA diagnosis was defined as having clinical features and microbiological evidence of IPA. Clinical features included any of the following computed tomography (CT) or chest X-ray manifestations: (1) consolidation with or without a halo sign; (2) air crescent sign; (3) cavity; and (4) wedge-shaped, segmental, peribronchial, or lobar consolidation. Microbiological evidence included either (1) positive result from Aspergillus culture or microscopic examination with lower respiratory tract specimens, or (2) positive result from GM test of serum or BAL specimens. All patients with IPA were given antifungal treatment, such as voriconazole, caspofungin, or amphotericin B. Respiratory viruses such as respiratory syncytial virus, influenza virus (IFV) type A, IFV type B, parainfluenza virus, human rhinovirus, human metapneumovirus, and adenovirus, were tested using nasopharyngeal swabs, sputum, endotracheal aspirate (ETA), or BAL fluid by reverse-transcription real-time PCR (Shanghai Zhijiang Biological Technology, Shanghai, China). Besides, the bacterial and fungal pathogens in sputum, ETA, and BAL fluid were also tested. The Aspergillus GM antigen in the samples was detected by the platelia Aspergillus test (Bio-Rad Laboratories, Marnes-la-Coquette, France). Data were analyzed using SPSS (SPSS Inc., Version 23.0, Chicago, IL, USA). In univariable analysis, we compared categorical variables by Fisher's exact test or chi-squared test; continuous variables were compared by the Student's t-test or Mann–Whitney U test, as appropriate. Binary analysis was performed using backward and forward stepwise logistic regression, with explanatory variables including known risk factors for IPA, and baseline factor significantly different between the IPA and non-IPA groups. Two-tailed P value less than 0.05 was considered indicative of statistical significance. Among the six hospitals, 2500 inpatients were positive for viral nucleic acid, of which 1883 patients were excluded for not having pneumonia or unavailable results of GM test or sputum fungal culture. A total of 617 adult patients with viral pneumonia were included in the final analysis, and the IPA and non-IPA groups had 91 and 526 cases, respectively. Compared with non-IPA group, IPA group had higher incidence of dyspnea (79.1% [72/91] vs. 65.8% [346/526]; χ2 = 6.320; P = 0.012), white blood count (10.7 [6.9, 13.8] × 109/L vs. 7.5 [5.4, 10.8] × 109/L; U = 4.811; P <0.001), neutrophil count (8.0 [5.4, 12.3] × 109/L vs. 5.5 [3.5, 8.9] × 109/L; U = 5.106; P <0.001), percentage of patients with diabetes mellitus (35.1% [32/91] vs. 21.9% [115/526]; χ2 = 7.563; P = 0.006); proportion of solid organ transplant recipients (18.7% [17/91] vs. 6.3% [33/526]; χ2 = 16.037; P <0.001), pneumonia severity index score (86.0 [64.0, 107.0] vs. 74.0 [56.8, 99.0]; U = 2.843; P = 0.004). And compared with non-IPA group, the CT result of IPA group showed higher manifestation of halo sign (4.4% [4/91] vs. 0% [0/526]; P <0.001), air crescent sign (4.4% [4/91] vs. 0.4% [2/526]; P <0.001), cavity (15.4% [14/91] vs. 4.4% [23/526]; χ2 = 16.409; P <0.001), peribronchial consolidation or nodular opacities (19.8% [18/91] vs. 1.9% [10/526]; χ2 = 56.601; P <0.001), and tree-in-bud (13.2% [12/91] vs. 3.8% [20/526]; χ2 = 13.657; P <0.001). Besides, the rates of glucocorticoid use before admission (49.5% [45/91] vs. 30.0% [158/526]; χ2 = 13.150; P <0.001), glucocorticoid use after admission (49.5% [45/91] vs. 28.5% [150/526]; χ2 = 15.727; P <0.001), non-invasive ventilation (24.2% [22/91] vs. 13.3% [70/526]; χ2 = 7.222; P = 0.007), invasive ventilation (51.6% [47/91] vs. 26.6% [140/526]; χ2 = 23.015; P <0.001), respiratory failure (69.2% [63/91] vs. 44.5% [234/526]; χ2 = 19.026; P <0.001), and in-hospital mortality (34.1% [31/91] vs. 18.4% [97/526]; χ2 = 11.519; P = 0.001) were significantly higher in IPA group compared with non-IPA group [Table 1]. Table 1 - Clinical characteristics of viral pneumonia with invasive pulmonary aspergillosis or without invasive pulmonary aspergillosis. Variables IPA group (n = 91) Non-IPA group (n = 526) χ 2/U P values Sex (female) 24 (26.4) 201 (38.2) 4.693 0.030 Age (years) 62.0 (54.0, 68.0) 62.0 (49.0, 72.0) −0.186∗ 0.852 Current smoker or ex-smoker 49 (53.8) 195 (37.1) 9.696 0.002 Symptoms and signs Fever 67 (73.6) 385 (73.2) 0.007 0.931 Expectoration 91 (100.0) 494 (93.9) 5.839 0.016 Dyspnea 72 (79.1) 346 (65.8) 6.320 0.012 Laboratory examinations White blood cell (×109/L) 10.7 (6.9, 13.8) 7.5 (5.4, 10.8) 4.811∗ <0.001 Neutrophils (×109/L) 8.0 (5.4, 12.3) 5.5 (3.5, 8.9) 5.106∗ <0.001 Lymphocyte (×109/L) 0.9 (0.5, 1.5) 1.0 (0.6, 1.6) −0.851∗ 0.395 Procalcitonin (ng/mL) 0.4 (0.2, 1.2) 0.3 (0.2, 0.7) 1.572∗ 0.116 Severe pneumonia index score 86.0 (64.0, 107.0) 74.0 (56.8, 99.0) 2.843∗ 0.004 CURB65 score>1 35 (38.5) 161 (30.6) 2.207 0.137 Comorbidities Interstitial lung disease 32 (35.2) 116 (22.1) 7.315 0.007 Chronic obstructive pulmonary disease 15 (16.5) 45 (8.6) 5.555 0.018 Diabetes mellitus 32 (35.1) 115 (21.9) 7.563 0.006 Kidney disease 1 (1.1) 12 (2.3) 0.833† Connective tissue disease 11 (12.1) 59 (11.2) 0.059 0.809 Tumor 4 (4.4) 37 (7.0) 0.871 0.351 Solid organ transplantation 17 (18.7) 33 (6.3) 16.037 <0.001 Immune deficiency 41 (45.1) 161 (30.6) 7.353 0.007 Imaging features Multi-lobe or segment infiltrates 89 (97.8) 492 (93.5) 2.263 0.132 Pleural effusion 17 (18.7) 74 (14.1) 1.225 0.268 Ground-glass opacity 46 (50.5) 216 (41.1) 2.624 0.105 Consolidation 54 (59.3) 298 (56.7) 0.144 0.705 Reticular pattern 25 (27.5) 108 (20.5) 2.070 0.150 Halo sign 4 (4.4) 0 (0) <0.001† Cavity 14 (15.4) 23 (4.4) 16.409 <0.001 Air crescent sign 4 (4.4) 2 (0.4) <0.001† Peribronchial consolidation or nodular opacities 18 (19.8) 10 (1.9) 56.601 <0.001 Tree-in-bud 12 (13.2) 20 (3.8) 13.657 <0.001 Treatment measures Bronchoalveolar lavage 82 (90.1) 391 (74.3) 10.790 0.001 Glucocorticoids use before admission 45 (49.5) 158 (30.0) 13.150 <0.001 Immunosuppressants before admission 18 (19.8) 87 (16.5) 0.577 0.448 Glucocorticoids use after admission 45 (49.5) 150 (28.5) 15.727 <0.001 Noninvasive ventilation 22 (24.2) 70 (13.3) 7.222 0.007 Invasive ventilation 47 (51.6) 140 (26.6) 23.015 <0.001 Complications Respiratory failure 63 (69.2) 234 (44.5) 19.026 <0.001 Septic shock 37 (40.7) 91 (17.3) 25.745 <0.001 ICU admission 54 (59.3) 213 (40.5) 11.027 0.001 In-hospital mortality 31 (34.1) 97 (18.4) 11.519 0.001 Data are presented as n (%) or median (Q1, Q3). ∗U values; †Fisher's exact test. CURB-65: Confusion, Urea, Respiratory Rate, blood pressure, Age≥65 years; ICU: Intensive care unit; IPA: Invasive pulmonary aspergillosis. The risk for viral pneumonia complicated with IPA included glucocorticoid use after admission (odds ratio [OR], 1.981; 95% confidence interval [CI]: 1.194–3.286; P = 0.008), invasive ventilation (OR, 3.115; 95% CI: 1.905–5.094; P <0.001), solid organ transplantation (OR, 3.142; 95% CI: 1.506–6.554; P = 0.002), diabetes (OR, 1.755; 95% CI: 1.048–2.938; P = 0.032), chronic obstructive pulmonary disease (COPD) (OR, 2.084; 95% CI: 1.055–4.117; P = 0.034), and IFV infection (OR, 1.981; 95% CI: 1.194–3.286; P = 0.008). In general, glucocorticoid use, invasive mechanical ventilation, COPD, solid organ transplantation, and diabetes mellitus are classic risk factors for IPA. Corticosteroid use prior to influenza infection has been evidenced to be associated with IPA.[2] The use of corticosteroids before ICU admission was considered as an independent risk factor for viral pneumonia complicated with IPA.[3] Apostolopoulou et al[5] evidenced that cumulative prednisone doses >140 mg within 7 days and pneumonia were independent risk factors for IPA with non-influenza respiratory viral infections. Respiratory viruses can damage the epithelial layer, leading to denudation, basement membrane exposure, and subsequent pathogen adherence and invasion. Steroids, particularly at higher doses, can compromise alveolar macrophage phagocytosis, predisposing patients to Aspergillus invasion.[5] Our study showed that the incidence of IPA in adult patients with IFV pneumonia was higher than that in non-IFV patients, but it is unclear why influenza patients have higher risk for IPA than patients with other virus infection; respiratory epithelium damage and mucociliary clearance dysfunction might facilitate the invasion of Aspergillus. Moreover, influenza-induced acute respiratory distress syndrome and hypoxia might cause immune paralysis.[2] In conclusion, the incidence of IPA in IFV pneumonia patients was higher than that in non-IFV pneumonia patients. The risk factors for IPA in adult patients with viral pneumonia are glucocorticoid use after admission, invasive ventilation, solid organ transplantation, diabetes, COPD, and IFV infection. Funding This work was supported by a grant from the China National Key Research and Development Program (No. 2018YFE0102100). Conflicts of interest None.