Abstract

Background and aim:

Patients with influenza or Coronavirus disease 2019 (COVID-19) are at risk of developing invasive pulmonary aspergillosis (IPA). Whether the clinical profiles and outcomes differ between influenza-associated pulmonary aspergillosis (IAPA) and COVID-19-associated pulmonary aspergillosis (CAPA) remains unclear, hindering efforts to optimize management. This study aimed to compare the demographic, clinical, laboratory characteristics, and outcomes of patients with IPA following influenza A/B versus COVID-19 during the same period.

Methods:

We conducted a single-center retrospective cohort study in China from December 1, 2022, to September 1, 2024. The study included 45 patients with IAPA and 82 patients with CAPA. We compared demographics, clinical features, management, and mortality between the IAPA and CAPA patients. Group comparisons utilized Student’s t-test or the Mann-Whitney U test for continuous variables, the Chi-square or Fisher’s exact test for categorical variables, and Kaplan-Meier survival curves, and time-dependent Cox proportional hazards model for survival data.

Results:

Compared to IAPA patients, CAPA patients had significantly lower lymphocytes, especially in CD4+ T cells, CD8+ T cells, and B cells (all p < 0.05). Corticosteroid use was more frequent among CAPA patients than IAPA patients. The median time from viral diagnosis to IPA detection was longer in CAPA patients than in IAPA patients. Respiratory co-infections (bacterial) were more common in the CAPA group (p = 0.030). After adjusting for confounders, the risk of death within the first 14 days following IPA diagnosis was 4.92 times higher in the CAPA group than in the IAPA group (HR = 4.92, 95% CI: 1.35–18.01, p = 0.016).

Conclusion:

CAPA was independently associated with an approximately five-fold increase in the risk of death within the first 14 days following IPA diagnosis. This early hazard, together with the higher frequency of corticosteroid use, respiratory co-infections (bacterial), and severe lymphopenia, underscores a critical window for early therapeutic intervention in patients with CAPA.

Introduction

Influenza and COVID-19 are major causes of severe viral pneumonia. They impose a substantial burden on global health systems (1, 2). Both viruses can induce profound pulmonary inflammation and immune dysregulation, increasing susceptibility to respiratory co-infections (3–5). Notably, invasive pulmonary aspergillosis (IPA) has emerged as a critical and lethal complication (6, 7).

Influenza-associated pulmonary aspergillosis (IAPA) is well-recognized, affecting approximately 16–28% of critically ill influenza patients in the intensive care unit (ICU) and contributing to high mortality (8–10). Similarly, during the COVID-19 pandemic, COVID-19-associated pulmonary aspergillosis (CAPA) was frequently identified as a leading fungal infection among severely ill patients, significantly worsening clinical outcomes (11–16). Both IAPA and CAPA highlight the interplay between viral lung injury and invasive fungal disease. However, it remains unclear whether the clinical presentation, risk factors, disease course, and outcomes of IPA differ substantially depending on the preceding viral pathogen. The host immune response, including the degree of lymphopenia and the therapeutic use of immunomodulators like corticosteroids, differs between influenza and COVID-19. This variation may in turn shape the phenotype of subsequent IPA (7, 17–19). Understanding these distinctions is critical for tailoring surveillance, diagnostic timing, and management strategies.

To date, direct comparative studies of IAPA and CAPA remain scarce. Existing research often conducted in different epidemiological periods, suggests potential differences in their clinical presentation, such as the degree of immunosuppression and the timing of aspergillosis onset (20–24). However, these studies have predominantly focused on risk factors for development, with limited concurrent, head-to-head comparison of clinical features and, more importantly, factors associated with mortality within the same timeframe (20–26). Therefore, we conducted a single-center retrospective cohort study. Our goal was to directly compare the demographic, clinical, laboratory, and outcome characteristics of patients with IPA following influenza A/B versus COVID-19 during the same period. We aimed to identify pathogen-specific patterns to inform more precise clinical management.

Materials and methodsEthics approval

This retrospective study received ethical approval from the institutional review board of ethics committee at Beijing Chaoyang Hospital, affiliated with Capital Medical University (Approval No.: 2026-KE-10) and was conducted in alignment with the Declaration of Helsinki (1964) and its subsequent amendments or comparable ethical standards. Given the retrospective nature of the study, which involved the analysis of anonymized clinical data, the ethics committee formally waived the requirement for informed consent.

Study participants

We performed a single-center retrospective cohort study involving consecutive patients admitted to Beijing Chao-Yang Hospital, Capital Medical University, from December 1, 2022, to September 1, 2024. Our study adhered to the STROBE guidelines.

Inclusion criteria: (1) a positive polymerase chain reaction (PCR) test for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or influenza A/B obtained from nasopharyngeal or throat swabs; (2) evidence of lower respiratory tract infection (fever, dyspnea, and radiographic lung infiltrates); (3) fulfillment of the diagnostic criteria for invasive pulmonary aspergillosis (IPA) as defined by the European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) (25); (4) invasive pulmonary aspergillosis following influenza or COVID-19.

Exclusion criteria: (1) individuals aged less than 18 years; (2) patients with pre-existing Aspergillus infection or IPA before influenza or COVID-19; (3) incomplete clinical data.

All respiratory samples were collected under aseptic conditions and processed immediately in the microbiology laboratory. Culture positivity was confirmed by standard mycological techniques, and growth of Aspergillus was considered significant only when accompanied by compatible clinical and (where available) radiological findings.

Clinical data collection

Data were retrospectively extracted from electronic medical records (EMRs) and included: baseline characteristics (age, sex, body weight), comorbidities (Arterial hypertension, Diabetes mellitus, Coronary heart disease, cancer, etc.), clinical course and interventions (length of hospital stay, ICU admission, oxygen therapy requirements, utilization of mechanical ventilation (MV), continuous renal replacement therapy (CRRT), etc.), as well as microbiological and laboratory data (results from galactomannan (GM) tests, fungal cultures, and other pertinent laboratory parameters). All data were obtained from the hospital’s EMR system. Diagnostic assessments (PCR, GM assays, cultures, microscopy) were performed as part of standard clinical care in accordance with established hospital protocols.

Diagnostic criteria

In our study, the diagnostic criteria were as follows: IAPA utilized the IAPA expert consensus criteria by Verweij and colleagues (5), while CAPA adopted the CAPA criteria established by the European Confederation of Medical Mycology (ECMM) and the International Society for Human and Animal Mycology (ISHAM) (26). It is important to note that histopathological examination was not included in this investigation, which may have constrained the pathological confirmation of IPA. The patients with probable IAPA and CAPA included in the study all had imaging evidence of pulmonary infiltrates, but this was not limited to chest CT; due to the severity of their illness, some patients only had chest X-ray results available.

Probable IAPA is defined based on recent expert consensus (5), necessitating the presence of pulmonary infiltrates along with at least one of the following: a serum GM index greater than 0.5, a bronchoalveolar lavage fluid (BALF) GM index over 1.0, or a positive Aspergillus culture from BALF. Alternatively, a cavitating infiltrate that cannot be attributed to another etiology, along with a positive culture from sputum or BALF, also qualifies as probable IAPA.

Probable CAPA: this was defined according to the modified 2020 ECMM/ISHAM consensus criteria (26). The presence of pulmonary infiltrates, which may or may not exhibit cavitation, coupled with clinical worsening and mycological indicators-such as a positive culture from BALF, a serum GM index exceeding 0.5, or a BALF GM index greater than 1.0-constitutes a diagnostic criterion for IPA.

Respiratory co-infection is usually defined as the presence of two or more different pathogens (including viruses, bacteria, fungi, and atypical pathogens) detected simultaneously or sequentially within a short period in the same acute respiratory infection event occurring in an individual, using reliable molecular diagnostic techniques (27–32). In this study, respiratory co-infection was defined by the identification of a clinically significant bacterial or fungal pathogen in BALF, sputum, or blood culture, prompting the initiation of targeted antibiotic or antifungal therapy, respectively.

Respiratory failure (33) was defined based on admission criteria according to the following: (1) partial pressure of oxygen (PaO2) < 60 mmHg or oxygen saturation (SpO2) < 90% on room air, or (2) PaO2/fraction of inspired oxygen (FiO2) ratio ≤ 300 mmHg, as assessed upon hospital admission.

Statistical analysis

Statistical analysis was conducted where categorical variables were presented as counts and percentages, with comparisons made using the Chi-square test or Fisher’s exact test, as appropriate. Continuous variables were reported as mean ± standard deviation (SD) if they followed a normal distribution and compared using independent samples t-tests. Non-normally distributed data were expressed as medians with interquartile ranges (IQR, 25–75%) and assessed using the Mann-Whitney U test. Kaplan-Meier survival curves for various subgroups were constructed and analyzed using the log-rank test. In the Kaplan-Meier survival curves “day 0” on the x-axis is defined as the date of IPA diagnosis. To address potential violations of the proportional hazard assumption over time, time-dependent Cox proportional hazards model was used to evaluate the effect of exposure on the outcome. The model was adjusted for covariates including age, sex, lymphocytes, chronic pulmonary disease, and immunocompromised status. The proportional hazards assumption was assessed using tests based on Schoenfeld residuals. All statistical tests were two-sided, with a p-value of less than 0.05 deemed statistically significant. The data analysis was executed using Free Statistics software (version 1.9.2).

Results

In total, this study evaluated 438 patients who tested positive for influenza virus nucleic acid and 1470 patients who tested positive for coronavirus nucleic acid. Among these, the positivity rates for IPA were 73 out of 438 (16.7%) and 125 out of 1470 (8.5%), respectively, revealing a statistically significant difference in incidence rates between the two groups (p < 0.001). Following the inclusion and exclusion criteria of this study, a total of 127 patients were enrolled, comprising 45 IAPA patients and 82 CAPA patients (see in Figure 1). Details regarding the baseline immune status of the study population are available in Supplementary material. In this cohort, no patients received antifungal prophylaxis prior to diagnostic sampling. Empirical antifungal therapy was initiated only after sampling in the vast majority of cases.

Flowchart of the study including patients. IPA: invasive pulmonary aspergillosis; IAPA: influenza- associated invasive pulmonary aspergillosis; CAPA: COVID-19-associated invasive pulmonary aspergillosis; GM: galactomannan; BALF:bronchoalveolar lavage fluid.

Clinical characteristics and test results

The clinical characteristics, demographic data, and underlying conditions of the patients are summarized in Table 1. A significant age difference (p = 0.012) was noted between IAPA patients (62.4 ± 16.5 years) and CAPA patients (69.0 ± 12.5 years). No notable difference was found between the two groups concerning other clinical characteristics. Among the underlying conditions, CAPA patients exhibited a significantly higher prevalence of chronic kidney disease (p = 0.028) and chronic liver disease (p = 0.032). Conversely, the proportion of patients with chronic pulmonary disease was significantly lower (p < 0.001) in the CAPA cohort (35.4%) compared to the IAPA cohort (68.5%). Additionally, the lymphocytes were also assessed. The comparison of immune cell counts, specifically CD4+T cells, CD8+T cells, and B cells, revealed significant differences between patients with CAPA and IAPA, as outlined in Table 1. Notably, CAPA patients exhibited lower immune cell counts. Additionally, fever and diarrhea were observed to be more prevalent among CAPA patients, with p-values of 0.015 and 0.019, respectively. Sensitivity results for patients after excluding individuals with prior immunosuppression are detailed in Supplementary Table S1.

VariablesTotal
(n = 127)CAPA patients
(n = 82)IAPA patients
(n = 45)p-valueDemographicsMale, n (%)84 (66.1)59 (72.0)25 (55.6)0.062Age, Mean ± SD, years66.7 ± 14.469.0 ± 12.562.4 ± 16.50.012BMI, Mean ± SD, kg/m223.3 ± 3.823.3 ± 4.223.3 ± 3.20.982Underlying conditions, n (%)Arterial Hypertension64 (50.4)42 (51.2)22 (48.9)0.802Coronary artery disease59 (46.5)43 (52.4)16 (35.6)0.068Cerebrovascular disease13 (10.2)7 (8.5)6 (13.3)0.542Diabetes mellitus50 (39.4)30 (36.6)20 (44.4)0.386Chronic kidney disease17 (13.4)15 (18.3)2 (4.4)0.028Chronic pulmonary disease60 (47.2)29 (35.4)31 (68.9)< 0.001Chronic liver disease13 (10.2)12 (14.6)1 (2.2)0.032Hematological malignancy9 (7.1)7 (8.5)2 (4.4)0.490Solid organ transplantation7 (5.5)6 (7.3)1 (2.2)0.420Autoimmune disease15 (11.8)7 (8.5)8 (17.8)0.123Solid organ malignancy11 (8.7)6 (7.3)5 (11.1)0.518Thrombosis18 (14.2)13 (15.9)5 (11.1)0.464Vaccine, n (%)29(22.8)17(20.7)12(26.7)0.446Immunocompromised status17(13.4)13(15.9)4(8.9)0.270Symptoms, n(%)Fever111 (87.4)76 (92.7)35 (77.8)0.015Cough115 (90.6)73 (89.0)42 (93.3)0.537Hemoptysis22 (17.3)13 (15.9)9 (20.0)0.555Diarrhea14 (11.0)13 (15.9)1 (2.2)0.019Vomit8 (6.3)7 (8.5)1 (2.2)0.258Sore throat20 (15.7)13 (15.9)7 (15.6)0.965Laboratory tests at admissionLeukocyte count, median (IQR), × 109/L8.6(5.7, 11.8)9.0(5.6, 11.8)7.9(6.1, 10.8)0.692Neutrophils, median (IQR), × 109/L6.7(4.5, 10.4)7.5(4.6, 10.7)6.2(4.4, 9.3)0.426Lymphocytes, median (IQR), × 109/L0.6(0.3, 1.2)0.5(0.3, 0.9)0.8(0.6, 1.6)< 0.001Hemoglobin, median (IQR), g/L121.0 ± 22.1119.4 ± 24.0124.0 ± 17.90.269CRP, median (IQR), mg/L50.4(18.7, 107.0)55.0(13.4, 107.8)48.0(23.0, 103.0)0.579Creatinine, median (IQR), μmol/L62.3(51.5, 87.1)64.2(52.4, 88.1)59.5(48.5, 80.0)0.400PCT, median (IQR), ng/mL0.2(0, 0.9)0.2(0, 1.0)0.1(0, 0.8)0.945CD4+Tcells, median (IQR), cell/μL172.0(77.0,346.5)138.5(69.8,262.)257.0(137.0,571.0)0.001CD8+ T cells, median (IQR), cell/μL108.0(56.0,252.5)101.0(41.2,152.0)183.0(93.0, 392.0)0.001NK cells, median (IQR), cell/μL59.0(30.5, 138.0)53.5(29.0, 123.2)77.0(35.0, 212.0)0.129B cells,median (IQR), cell/μL76.0(18.5, 146.5)49.0(17.0, 129.0)98.0(27.0, 164.0)0.049D- -Dimer, median (IQR), ng/L1.2(0.6, 2.7)1.2(0.8, 3.0)1.0(0.5, 2.0)0.209

Baseline features of the study patients.

Data are presented as number (%) or median (IQR) unless otherwise indicated. Bold indicated data with a significant difference. BMI, body mass index; CRP, C-reaction protein; PCT, procalcitonin; NK, natural killer cell; MV, mechanical ventilation.

Laboratory diagnostics and distribution of invasive pulmonary aspergillosis

In terms of laboratory diagnostics for IPA, the methods employed included BALF and sputum cultures. In our study, there were cases of repeated sample submissions; therefore, only the results of the first culture from each patient were included in the statistical analysis (for both BALF and sputum), and presented in Table 2. The positivity rates of the GM tests were 40.0% in blood serum and 76.2% in BALF. In the CAPA group, 35 cases underwent both sputum culture and BALF culture, while in the IAPA group, 16 cases underwent both sputum culture and BALF culture. Comparatively, patients with IAPA demonstrated a higher sputum culture positivity rate than those with CAPA. However, the positivity rates of GM tests in both serum and BALF did not show significant differences between CAPA and IAPA patients. Sensitivity results for patients after excluding individuals with prior immunosuppression are detailed in Supplementary Table S2.

VariablesTotal
(n = 127)CAPA patients
(n = 82)IAPA patients
(n = 45)p-valueBALF culture, n966432BALF culture (+), n (%)71(74.0)45(70.3)26(81.3)0.250Sputum culture, n603723Sputum culture (+), n (%)16(26.7)6(16.2)10(43.5)0.020Mircoscopy, n554015Mircoscopy (+), n (%)12(21.8)6(15.0)6(40.0)0.046Serum GM, n1158035Serum GM (+), n (%)46(40.0)32(40.0)14(40.0)1.000BALF GM, n422616BALF GM (+), n (%)32(76.2)22(84.6)10(62.5)0.102

Laboratory diagnositics of Aspergillus of the 127 enrolled patients.

Data are presented as number (%). Bold indicated data with a significant difference. There were cases of repeated sample submissions per patient; however, to maintain statistical independence and avoid inflation of the incidence rate, only the results of the first culture from each patient were included in the final statistical analysis.

Aspergillus cultures identified a total of 140 Aspergillus isolates. Among these, the following were identified: 85 Aspergillus fumigatus (A. fumigatus), 32 Aspergillus flavus (A. flavus), 14 Aspergillus niger (A. niger), 3 Aspergillus terreus (A. terreus), 2 Aspergillus nidulans (A. nidulans), 1 Aspergillus versicolor (A. versicolor), and 3 Aspergillus spp. (culture) as illustrated in Figure 2A. Among them, 13 patients (10.2%) had mixed infections with multiple Aspergillus species. A. fumigatus was the most frequently identified species among both IAPA and CAPA patients, accounting for 66.7 and 57.0%, respectively, no statistically significant difference was observed between the two groups, as detailed in Figure 2B.

Aspergillus isolated from the 127 enrolled patients (45 IAPA patients and 82 CAPA patients). Aspergillus spp (culture) refers to culture-positive isolates that were not identified to the species level. (A) The absolute sample sizes of each Aspergillus species in IAPA and CAPA patients. (B) The relative proportions of IAPA and CAPA cases for each species.

Clinical outcomes

It is noteworthy that 64 out of 82 CAPA patients received steroid treatment for SARS-CoV-2, resulting in an in-hospital mortality rate of 40.6% (26 out of 64). Conversely, the remaining 18 patients who did not receive steroids had a significantly lower in-hospital mortality rate of 22.2% (4 out of 18). Furthermore, a higher proportion of corticosteroid use was noted in the CAPA group compared to the IAPA group (p = 0.031). CAPA patients had a significantly longer interval from positive viral test to aspergillosis diagnosis compared to IAPA patients, with a median of 12.0 days (IQR: 8.0, 20.8) versus 2.0 days (IQR: 0.0, 6.0) (p < 0.001), as detailed in Table 3. Sensitivity results for patients after excluding individuals with prior immunosuppression are detailed in Supplementary Table S3.

VariablesTotal
(n = 127)CAPA patients
(n = 82)IAPA patients
(n = 45)p-valueAntifungal treatment, n (%)124(97.6)82(100.0)42(93.3)0.079Antiviral treatment, n (%)119(93.7)76(92.7)43(95.6)0.711Mortality, n (%)39(30.7)30(36.6)9(20.0)0.053In-put time median (IQR), days18.0(11.0, 28.5)20.0(11.0, 31.0)17.0(12.0,21.0)0.197Corticosteroids use 7 days before and after ICU admission, n (%)91(71.7)64(78.0)27(60.0)0.031Time from virus positive to aspergillus diagnosis, median (IQR), days9.0 (3.0, 16.0)12.0 (8.0, 20.8)2.0 (0.0, 6.0)<0.001Respiratory failure at admission, n (%)91(71.7)64(78.0)27(60.0)0.031MV, n (%)74(58.3)51(62.2)23(51.1)0.226ICU admission, n (%)68(53.5)45(54.9)23(51.1)0.684ICU mortality, n (%)24(18.9)18(22.0)6(13.3)0.235

Clinical outcome of 82 CAPA patients and 45 IAPA patients.

Data are presented as number (%) or median (IQR) unless otherwise indicated. Bold indicated data with a significant difference.

Additionally, respiratory co-infections (bacterial) were assessed among the patients, revealing that 86 out of 127 patients (67.7%) tested positive for bacterial infections. Moreover, a statistically significant difference was observed in respiratory co-infections (bacterial) between IAPA and CAPA patients (χ2 = 4.715; p = 0.030). A total of 136 unique clinical strains were isolated from 127 patients, which included Enterobacteriales (n = 37), Acinetobacter spp. (n = 31), Pseudomonas spp. (n = 32), Staphylococcus spp. (n = 11), and other bacteria (n = 25) as detailed in Table 4. Sensitivity results for patients after excluding individuals with prior immunosuppression are detailed in Supplementary Table S4.

VariablesTotal
(n = 127)CAPA patients
(n = 82)IAPA patients
(n = 45)p-valueRespiratory co-infections (bacterial), n (%)86(67.7)61(74.4)25(55.6)0.030Total isolate strains, n1369739Enterobacteriales, n (%)37(27.2)28(28.9)9(23.1)0.493Acinetobacter spp., n (%)31(22.8)21(21.6)10(25.6)0.616Pseudomonas spp., n (%)32(23.5)23(23.7)9(23.1)0.937Staphylococcus spp., n (%)11(8.1)7(7.2)4(10.3)0.810Others, n (%)25(18.4)18(18.6)7(17.9)0.934

Respiratory co-infections (bacterial) of 82 CAPA patients and 45 IAPA patients.

Data are presented as number (%) or median (IQR) unless otherwise indicated. Bold indicated data with a significant difference. IAPA, influenza-associated pulmonary aspergillosis; COVID-19, coronavirus disease-19; CAPA, COVID-19-associated pulmonary aspergillosis.

Survival analysis conducted 30 days post-diagnosis indicated a trend toward higher mortality in CAPA patients, though this did not reach statistical significance (p = 0.071); however, by 60 days, CAPA patients demonstrated a significantly increased mortality rate compared to IAPA patients (p = 0.043), as shown in Figure 3. Time-dependent Cox regression stratified into intervals of 0–14 days, 14–28 days, and 28–60 days, was performed. As shown in the