Clinical characteristics and prognostic risk factors of mortality in patients with interstitial lung diseases and viral infection: a retrospective cohort study

Introduction Patients with interstitial lung disease (ILD) who subsequently develop a viral infection have high rates of morbidity and mortality. Hypothesis/Gap statement Few large-scale epidemiological studies have investigated potential prognostic factors for morbidity and mortality in this patient group. Aim To evaluate the risk factors for morbidity and mortality in hospitalized patients with ILD and viral infection, as well as the clinical characteristics. Methodology This retrospective cohort study included patients with ILD who were hospitalized for a viral infection in two tertiary academic hospitals in China, between 1 January 2013 and 31 December 2019. We analysed the prevalence of comorbidities, clinical characteristics, 30 day mortality rates, and prognostic risk factors. Results A total of 282 patients were included; 195 and 87 were immunocompromised and immunocompetent, respectively. The most common underlying interstitial diseases were idiopathic pulmonary fibrosis (42.9 %) and connective tissue disease (36.9 %). The 30 day mortality rate was 20.6 %. During the influenza season, an increase in influenza virus (IFV) (25.7 %), respiratory syncytial virus (14.9 %) and cytomegalovirus (CMV) (11.3 %) cases was observed in the immunocompromised group. The most frequently detected virus in the immunocompetent group was IFV (44.8 %), followed by respiratory syncytial virus (11.5 %), and human rhinovirus (9.2 %). During the non-influenza season, CMV (34.4 %) was the main virus detected in the immunocompromised group. The 30 day mortality rates of non-IFV patients were higher than those of IFV patients. Older age (>60 years), respiratory failure, persistent lymphocytopenia, invasive mechanical ventilation and non-IFV virus infection were significantly associated with increased 30 day mortality. Conclusion Patients with ILD who develop viral infection have high rates of morbidity and mortality, which are associated with increased age (>60 years), respiratory failure, mechanical ventilation, persistent lymphocytopenia and non-IFV virus infection. These risk factors should be carefully considered when determining treatment strategies for this patient population.


INTRODUCTION
Few studies have evaluated the impact of viral infections on the acute exacerbation of idiopathic pulmonary fibrosis (IPF) and/or non-IPF interstitial lung disease (ILD). Saraya et al. documented respiratory virus infections in 19.2 % of patients with acute exacerbation of interstitial pneumonia; no difference was observed between patients with IPF and non-IPF ILD [1]. In another study in which bronchoalveolar OPEN ACCESS The acute exacerbation of IPF is a dangerous condition and has a mortality rate of over 50 % [3]. Some reports have documented 1 year mortality rates of almost 100 % in patients with an acute exacerbation of IPF [4,5]. Weng found that 60 % of samples collected from patients with an acute exacerbation of IPF were virus positive [6]. Drake et al. concluded that patients with ILD, particularly those with poor lung function and obesity, are at an increased risk of death from coronavirus disease [7]. However, there is a current lack of large-scale epidemiological studies that have investigated viral infections and prognosis in patients with ILD. Therefore, the purpose of this study was to evaluate potential risk factors for mortality in hospitalised patients with ILD and viral infections, as well as clinical characteristics.

Study design and participants
We retrospectively recruited patients with an acute exacerbation of ILD and viral infection, who were hospitalized between 1 January 2016 and 31 December 2019, at two secondary and tertiary academic hospitals in China. IPF was defined by the 2007 American Thoracic Society/ European Respiratory Society criteria [8]; the definition was broadened to include patients with previously known or established fibrotic disease at admission [9]. We enrolled patients who had usual interstitial pneumonia patterns on their radiological examination, meaning those with an acute exacerbation of connective tissue disease (CTD)-associated interstitial pneumonia and unilateral lung transplantation for ILD. The inclusion criteria were as follows: (1)  †The reason of unilateral lung transplantation was interstitial lung disease. ‡Other interstitial pneumonia includes non-specific interstitial pneumonia, organizing pneumonia, allergic pneumonia, radiation pneumonia, drug-induced interstitial pneumonia, etc.

Study quality control
Key investigators, including clinicians, statisticians, microbiologists and radiologists, worked together to draft the protocol and create a single formatted case report form (CRF) used by all centres. Before study initiation, all investigators from the six centres received training related to the study protocol, including the screening process, definitions of underlying diseases, and the formatted CRF. After the data were collected, CRFs were reviewed by a trained researcher to ensure completeness and data quality. The study was approved by the Ethics Committee of China-Japan Friendship Hospital. There was a centralized collaboration between all participating hospitals, which included anonymized data submission and collection.

Data collection
The following data were collected from the medical records of patients during their hospitalisation: (1) demographics; (2) clinical symptoms; (3) initial vital signs and lung examination findings; (4) severity of disease (indicated by intensive care unit [ICU] admission, use of invasive or non-invasive mechanical ventilation, pneumonia severity index (PSI) score and/or confusion-urea-respiratory rate-blood pressure-65 (CURB-65) score [10,11]; (5) laboratory and microbiological data (blood, sputum and/or BALF samples, bacterial or fungal cultures, viral nucleic acid detection and antibiotic susceptibility patterns); (6) treatment information, including use of vasoactive agents, antimicrobials, glucocorticoids and/or other immunosuppressants; and (7) survival status 30 days after admission. High-dose steroid use within 30 days of admission was defined as a prednisolone or glucocorticoid dose of at least 30 mg/day. Persistent lymphocytopenia was defined as a peripheral blood lymphocyte count of <1×10 9 l −1 for more than 7 days.

Diagnostic procedures
A viral aetiology was confirmed based on the following criteria: reverse transcription real-time (RT)-PCR (Shanghai Zhijiang Biological Technology, China) detection of Enterobacter aerogenes *Not all bacterial strains had drug-sensitivity results.  Table 3. Comparative analysis of different viral pneumonia in patients with interstitial lung disease Variables CMV N=64

Pathogen-specific diagnostic information
A diagnosis of pneumonia caused by Aspergillus required one or more of the following criteria: (1) histopathologic or direct microscopic evidence of dichotomous septate hyphae with a tissue culture positive for Aspergillus; (2) a positive Aspergillus culture from BALF; (3) a galactomannan optical index in BALF of ≥1; (4) a galactomannan optical index in serum of ≥0.5; and (5) Aspergillus species identified by culture and microscopic characteristics [12,13].
The diagnosis of Pneumocystis jirovecii pneumonia (PCP) was based on one of the following criteria: (1) high-resolution computed tomography imaging showing diffuse ground-glass opacity with a patchy distribution; (2) mycological criteria (microscopic examination of the respiratory sample revealing the presence of Pneumocystis cystic or trophic forms); and (3) a positive PCR test for Pneumocystis deoxyribonucleic acid [14].
Co-infection was documented if bacteria or fungi were isolated from lower respiratory tract specimens (qualified sputum, endotracheal aspirate and BALF) and/or blood samples within 48 h of hospitalization. Nosocomial infection was diagnosed based on clinical signs or symptoms of nosocomial pneumonia, bacteremia, and a positive culture of a new pathogen obtained from lower respiratory tract specimens and/or blood samples obtained ≥48 h after admission.

Statistical analysis
Patient demographics, clinical characteristics and pathogen testing results are expressed as mean (±standard deviation), median (interquartile range) or number (percentage). Group comparisons were conducted using Student's t-test or the Wilcoxon rank-sum test for continuous variables with and without normal distributions, respectively. Categorical variables of the two groups were compared using the χ2 test. Cox regression analysis was used to examine independent predictors of mortality, and its results were reported as hazard ratio (HR) and 95 % CI. Kaplan-Meier survival curves were used to compare the 30 day survival rate for patients by the log-rank test.

≥Two viruses N=41 P-Value
Septic shock during hospitalization  Table 3. Continued

Continued
Statistical analyses were performed using SPSS version 19.0 (SPSS, Chicago, IL, USA). All tests were two-sided, and P-values<0.05 were considered statistically significant.

Patient and public involvement
Neither patients nor the public were involved in the development of the research question, study design, patient recruitment or the conduct of the study. Ninety-five (43.3 %) patients were admitted to the ICU for treatment, with 23.8% and 24.8 % having received non-invasive and invasive ventilation, respectively. The 30 day mortality rates were 20.6 %, respectively. A total of 195 patients were immunocompromised, and 87 patients were immunocompetent. The following parameters were significantly higher in the immunocompromised group than in the immunocompetent group: proportion of patients with persistent lymphocytopenia, diabetes and CTD; use of anti-Pseudomonas drugs, anti-Aspergillus drugs, ganciclovir and sulfonamides; requirement for ICU admission, non-invasive ventilation, invasive mechanical ventilation and/or extracorporeal membrane oxygenation; adverse outcomes including respiratory failure and septic shock; peripheral blood leucocyte, neutrophil and lymphocyte counts; and lactate dehydrogenase, urea nitrogen, d-dimer and procalcitonin levels (P<0.05). Age, haemoglobin levels and the proportion of patients with cough symptoms and IPF were significantly lower in the immunocompromised group than in the immunocompetent group (Table 1).

RESULTS
During the influenza season (November, December, January, February), an increase in IFV (25.7 %), RSV (14.9 %) and CMV (11.3 %) cases was found in the immunocompromised group. The most frequently detected virus in the immunocompetent group was IFV (44.8 %), followed by RSV (11.5%) and HRV (9.2 %). During the non-influenza season, CMV (34.4 %) was the main virus detected in the immunocompromised group.
Patients with PIV had the highest average age (75 years) and the lowest incidence of fever (30.8 %). Patients with CMV and two or more virus groups had higher neutrophil and lactate dehydrogenase levels and lower lymphocyte counts than other viruses. Patients with CMV had a lower oxygenation index (P<0.05). Patients with CMV, HRV, PIV or two or more virus groups had more frequently required non-invasive mechanical ventilation, invasive mechanical ventilation and ICU care, and had higher rates of respiratory failure, septic shock and 30 day mortality (Table 3).  Table 4. Continued The following parameters were significantly higher in the nonsurvivors' group than in the survivors' group: age, underlying connective tissue disease, proportion of fever and dyspnoea, peripheral blood leukocytes, neutrophils, lactate dehydrogenase, urea nitrogen, d-dimer on the first day of admission, patients with persistent lymphocytopenia, consolidation on CT image, PSI score and CURB-65 score >1,CMV infection, PCP infection, non-IFV infection, nosocomial bacterial infection, requirement for ICU admission, non-invasive ventilation, invasive mechanical ventilation and/or extracorporeal membrane oxygenation; respiratory failure; (P<0.05). Lymphocytes, haemoglobin, and albumin were significantly lower in the non-survivors' group than in the survivors' group (Table 4).
Multivariate Cox regression analysis indicated that the following factors were independent predictors of 30 day mortality in patients with ILD: age >60 years, respiratory failure, persistent lymphocytopenia, invasive mechanical ventilation and non-IFV type A infection (Table 5, Fig. 1).

DISCUSSION
This study was a large-scale retrospective investigation of the clinical characteristics and prognostic risk factors of mortality in hospitalized patients with ILD who developed viral infection. The main findings are summarized as follows: (1) patients with ILD who developed viral infection had a higher mortality, with the 30 day rates being 20.6 %, respectively; (2) the distribution of virus types in immunocompromised patients differed between influenza and non-influenza seasons; (3) the disease severity and mortality in non-IFV patients were higher than those of IFV patients; and (4) independent risk factors for mortality included age >60 years, respiratory failure, persistent lymphocytopenia, invasive mechanical ventilation and non-IFV infection.
Previous studies have shown that viruses, especially respiratory viruses, may be co-factors for the development or exacerbation of lung fibrosis [15]. One such study, which conducted  [19]. Our large-scale epidemiological study of patients with ILD and viral infection found that IFV and RSV were the main pathogens during the influenza season, followed by CMV. During the non-influenza season, CMV was the main pathogen in immunocompromised patients, followed by IFV, RSV, PIV and HRV. Therefore, in the case of patients with suspected interstitial disease complicated with virus infection, we suggest that the viral nucleic acid test should be performed as early as possible to confirm the etiological diagnosis.
The disease severity, complications, and outcomes of immunocompetent patients with community-acquired pneumonia were similar between IFV and non-IFV-related respiratory diseases [20][21][22]. For elderly hospitalized patients with respiratory symptoms, RSV, human metapneumovirus and PIV have been associated with higher mortality [23][24][25][26] and more complications [25] than influenza. Our study showed that disease severity and mortality in non-IFV patients were higher than those in IFV patients. This result can be attributed to the following reasons: (1) the early use of oseltamivir in patients with influenza; (2) the lack of a specific drug for HRV and PIV; and (3) CMV was closely related to immunocompromised patients and had high mortality [27,28]. Thus, when patients with ILD develop symptoms of a viral infection, an increased vigilance is warranted for the detection of non-IFV infections.
Factors identified by previous studies as being associated with a poor prognosis in patients with ILD include a lower baseline forced vital capacity and carbon monoxide diffusing capacity; more extensive abnormalities on computed tomography at the time of acute exacerbation; and poor oxygenation and BALF neutrophil and lymphocyte percentages [29,30]. Viral infections, mostly CMV and human herpesvirus 7, have been identified in patients with acute exacerbation of IPF and non-IPF ILDs; however, virus infection was not found to be an independent predictor of 60 day survival in a simple logistic regression analysis [5]. Moua et al. suggested that the following factors were predictive of increased in-hospital mortality: male sex, acute exacerbation, longer duration of hospitalization, ICU admission, mechanical ventilation, use of bronchoscopy in an ICU setting and the intravenous administration of high-dose steroids [1]. In our study, we did not find a close relationship between high-dose hormone administration and poor prognosis, but we found that lymphopenia was directly related to poor prognosis, similar to the finding of other viral infection studies [31]. We also found that non-IFV virus infection was closely related to poor prognosis. Therefore, we must pay attention to the higher mortality rates due to viral infections such as CMV, HRV, PIV and mixed virus infections.
This study had several limitations. First, it utilized a retrospective observational design. Second, lung-function tests were not performed, as many of the patients could not undergo these tests. Third, we did not re-evaluate patient prognosis at a 1 year follow-up; therefore, it was impossible to suggest that viral infection was associated with poor long-term prognosis of ILD.

CONCLUSIONS
Patients with ILD who subsequently developed viral infection had high rates of morbidity and mortality, which were associated with increased age (>60 years), respiratory failure, mechanical ventilation, persistent lymphocytopenia and non-IFV virus infection. These risk factors should be carefully considered when determining treatment strategies for this patient population.

Author contributions
Study design: L.L. and G.Y. Data collection: L.L., C.W., L.S .and X.Z. Statistical analysis: L.L. Writing: L.L. and G.Y. All authors take full responsibility for the study design, data analysis and interpretation, and preparation of the manuscript. All authors approved the final draft of the manuscript.