Bacterial and Viral Coinfection in IPF Patients

Methods

Study Population

In this retrospective cohort study, 67 nasopharyngeal lavage (NPL) and bronchoalveolar lavage (BAL) samples were collected from IPF patients referred to hospitals of the Kermanshah University of Medical Sciences (KUMS) between June 2017 and September 2018. All acute exacerbated patients were hospitalized. Inclusion criteria were: radiological, spirometry, therapeutic, and biological data, which are considered for IPF patients by the clinical team of KUMS hospitals.^[2] Exclusion criteria were: Patients with chronic hypersensitivity pneumonitis, connective tissue disease or asbestosis. The Ethical Committee of KUMS approved this study.

Nucleic Acid Extraction

DNA and RNA extraction were performed using 200 ml of NPL and/or BAL specimens by QIAamp MinElute Virus Spin Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions. Extracted genomic DNA/RNA was stored at − 80 °C before use.

DNA Array Assay for Detection of Viruses and Bacteria

The CLART® Pneumovir DNA array assay (Genomica, Coslada, Madrid, Spain) was used to detect a number of viruses in this study (RSV, influenza viruses, HPIV, rhinovirus, adenovirus and coronavirus) according to manufacturer's instructions and as described by us previously.^[4] Detection of S. pneumonia, S. aureus and H. influenza was also carried out using this method.

Polymerase Chain Reaction (PCR) for Detection of Bacteria

Detection of P. aeruginosa was carried out by PCR analysis as described by Tyler et al. previously.^[14] Detection of K. pneumoniae was confirmed by PCR based on a study by Turton et al..^[15]

Statistical Analysis

In this study, all data was presented as the mean ± standard deviation for continuous variables. Categorical variables are presented as N (%). A normality test performed for the continuous variables using Kolmogorov–Smirnov test. A Student's t-test (parametric) or the Mann Whitney test (non-parametric) was used to test for statistical significance (two-tailed) between two experimental groups. Two-sided Chi square/Fisher's exact tests were used to assess the associations between IPF and the categorical variables. Principal component analysis was used in order to investigate the pattern of bacterial and virus infections, and coinfection in the patients. Kaplan–Meier survival curve analysis and log rank test were used to test time-to-death between non-infected, and bacterial-, viral- and co-infection groups of IPF patients. False discovery rate (FDR) was corrected using the Benjamini–Hochberg correction method for multiple comparisons. All statistical analysis were analyzed using R software version 3.5.1 and STATA software versions 11.2. Statistical significance was recorded at p < 0.05.

Table 1. Baseline characteristics of the IPF patients (n = 67)
Characteristics
Age (year)*	62.8 ± 12.44
FVC*	75.3 ± 4.37
FEV₁*	70.9 ± 4.30
DL_CO*	70.5 ± 3.96
Sex⁺
Male	40 (59.7)
Female	27 (40.3)
Disease status⁺
Acute exacerbation	55 (82.1)
(AE-IPF) stable	12 (17.9)
Immunosuppression drugs⁺
Pred 5 mg	2 (3)
Pred 10 mg	5 (7.5)
Pred 20 mg	2 (3)
CyA 125 mg	2 (3)
Unknown	8 (11.9)
Surgical lung biopsy⁺
No	63 (94)
Yes	4 (6)
History of fibrosis in family⁺
No	55 (82.1)
Yes	12 (17.9)
Bacterial infection No. 7 (10.4%)
P. aeruginosa ⁺	17 (27)
S. pneumonia ⁺	17 (27)
S. aureus ⁺	16 (25.4)
K. pneumoniae ⁺	16 (25.4)
H. influenzae ⁺	29 (46)
Viral infection No. 12 (17.9%)
Parainfluenza⁺	12 (19)
Influenza⁺	12 (19)
Adeno⁺	10 (15.9)
Rhinovirus⁺	25 (39.7)
RSV⁺	15 (23.8)
Coronavirus⁺	9 (14.3)
Co-infection⁺ No. 40 (59.7%)
Death⁺
No	49 (73.1)
Yes	18 (26.9)

Ref. considered as the reference level for each categorical variable, NA not available, forced vital capacity, FVC forced expiratory volume 1, FEV₁, diffusing capacity of lung for carbon monoxide: DL_C0
*Indicated as mean ± standard deviation
⁺Indicated as "n" (%)

Table 2. Comparison of pulmonary function indices between non-infected, bacterial-, viral- and co-infected IPF patients, respectively
PFT index	Non-infected Group 1 (n = 8)	Bacterial Infection Group 2 (n = 7)	Viral Infection Group 3 (n = 12)	Coinfection Group 4 (n = 40)	R, Adj. p*	R, Adj. p ⁺	R, Adj. p ^$
FVC	79.7 ± 3.30	76.6 ± 5.19	77.8 ± 2.88	73.7 ± 4.08	0.961, 0.368	0.976, 0.652	0.924, 0.013
FEV₁	73.5 ± 5.0	73.1 ± 5.55	73.1 ± 3.39	69.5 ± 3.82	0.994, 0.996	0.994, 0.992	0.945, 0.152
DL_CO	76.5 ± 1.29	71.7 ± 4.64	71.2 ± 2.73	69.3 ± 3.70	0.937, 0.076	0.93, 0.030	0.905, 0.001

Data represents Mean ± SD
IPF uninfected patients are considered as the reference group (Group 1). The Adj. P is based on the marginally adjusted p values by the Benjamini–Hochberg-FDR correction at α = 0.05. Bold values indicated as statistically significant at p < 0.05 level
PFT pulmonary function test, R ratio
*Comparison between group 2 versus group 1
⁺Comparison between group 3 versus group 1
^$Comparison between group 4 versus group 1

Table 3. Comparison of disease status, survival time and death occurrence between non-infected, bacterial-, viral- and co-infected IPF patients, respectively
Characteristics	Non-infected Group 1 (n = 8)	Bacterial Infection Group 2 (n = 7)	Viral Infection Group 3 (n = 12)	Coinfection Group 4 (n = 40)	p*	p ⁺	p ^$
Disease status⁻ (acute vs. stable)	0	1 (14.3)	1(8.3)	22 (55)	0.675	0.930	< 0.001
Death status⁻ (death vs. survive)	0	2 (28.5)	0	15 (37.5)	0.40	0.622	0.043
Survival time⁺ (months-to-death)	42.5 ± 6.55	38.5 ± 7.02	34.4 ± 2.83	32.9 ± 9.12	0.426	0.193	0.026

Disease- and death-status are indicated as "n" (%). Survival time is indicated as median ± IQR
IPF uninfected patients are considered as the reference group (Group 1). Bold values indicated as statistically significant at p < 0.05 level. Follow up period is 60 months
*Comparison between group 2 versus group 1
⁺Comparison between group 3 versus group 1
^$Comparison between group 4 versus group 1