Abstract
Background
The differential diagnosis of bilateral lung infiltrates and prognosis prediction can be challenging for clinicians in the intensive care unit (ICU). We analysed the proteome from bronchoalveolar lavage fluid (BALF) and determined its usefulness for evaluating the infectious causes and mortality associated with bilateral lung infiltrates.
Methods
In the ICU cohort, 136 patients with bilateral infiltrate on chest radiographs were selected, and bronchoscopy with bronchoalveolar lavage (BAL) was performed. Proteomic profiling of the exosomes in the BALF (n=20) was conducted to identify candidate protein biomarkers potentially associated with infection or mortality. The BAL samples (n=116) were used to measure the candidate biomarker levels.
Results
The candidate biomarkers, CD20, CLIC4, SCFD1 and TAP1, were selected for the differential diagnosis of infection or mortality. The levels of CD20 were significantly elevated in patients with non-infectious causes, compared with those with infectious causes (248.6±154.5 versus 177.6±150.9 ng·mL−1, p=0.014). The levels of CLIC4, SCFD1 and TAP1 did not differ between the two groups. As per the receiver operating characteristic analysis, CD20 was a significant predictor of non-infectious causes (area under curve 0.668; 95% confidence interval 0.567–0.769; p=0.002; cut-off value 167.6 ng·mL−1; sensitivity 74.1%; specificity 63.2%). There were no significant differences in the concentrations of the biomarkers between survivors and non-survivors.
Conclusions
Our results suggest that CD20 levels in BALF might be a useful biomarker for differentiating non-infectious and infectious diseases in patients with bilateral lung infiltrates.
Shareable abstract
Bilateral lung infiltrates have diverse clinical causes and courses in critically ill patients with respiratory failure. CD20 levels in bronchoalveolar lavage fluid might be a useful biomarker for differentiating non-infectious and infectious diseases. https://bit.ly/3XivxRM
Introduction
Bilateral lung infiltrates are a common presentation in critically ill patients with respiratory failure [1]. Patients with such infiltrates may have diverse clinical causes and courses, and they can progress to respiratory failure, with high rates of morbidity and mortality. Many aetiologies, both infectious and non-infectious, can cause bilateral infiltrates. However, the differential diagnosis of bilateral lung infiltrates can be complex and challenging for clinicians in the intensive care unit (ICU) setting, and the difficulty of diagnosis is associated with a poor prognosis.
Many studies have presented potential biomarkers associated with diagnosis or prognosis using bronchoalveolar lavage fluid (BALF) or plasma in patients with bilateral lung infiltrates [2–4]. For instance, a previous study showed that soluble triggering receptor expressed on myeloid cells-1 in BALF is a useful biomarker for bacterial lung infections in ICU patients [5]. Both serum C-reactive protein (CRP) and procalcitonin (PCT) are helpful biomarkers to distinguish infection from non-infectious systemic inflammation in ICU patients. Meta-analyses suggest similar sensitivities for both markers in the diagnosis of infection (75% for CRP versus 77% for PCT), but PCT has a slightly higher specificity (67% for CRP versus 79% for PCT) [6, 7]. However, these biomarkers are elevated in an immunologically medicated inflammatory disease, including aspiration pneumonitis, and their diagnostic value declines in localised infections. Thus, there is an increasing need for new biomarkers for predicting mortality or infectious causes in the clinical setting [8, 9].
Exosomes are major cell-to-cell communicators present in all the body fluids; they transport proteins, lipids and nucleic acids [10]. They can be applied as a biomarker for malignancy and neurologic, cardiovascular, infectious and respiratory diseases [11], such as COPD, idiopathic pulmonary fibrosis, lung cancer and acute lung injury [12–14]. In a previous study, the bronchoalveolar lavage (BAL) exosome level in patients with acute respiratory distress syndrome was positively correlated with hypoxia severity and infectious cause [15].
Recently, the analysis and quantification of proteins by mass spectrometry (MS)-based methods have been adapted for high-throughput analyses of thousands of proteins in the cells or body fluids [16, 17]. Bottom-up proteomics is used to identify and quantify proteins based on proteolytic peptides to discover protein markers associated with a disease. These data can be useful to evaluate the markers of the disease process or to give insight into the biological pathways.
To find new biomarkers that diagnose the cause of a lung injury and predict the prognosis, we evaluated candidate protein biomarkers using proteome profiling of exosomes extracted from BALF and measured the concentrations of the candidate proteins for validation.
Methods
Study population
Patients with acute lung injury were selected for the Asan sample cohort. Enrolled patients presented with bilateral lung infiltrates on a chest radiograph, and bronchoscopy with BAL was performed for differential diagnosis. The samples from the critically ill patients in the ICU cohort were prospectively collected. For all patients, the following data were recorded: baseline characteristics, diagnosis, leukocyte count, serum levels of CRP and PCT, ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen on the day of the bronchoscopy, Sequential Organ Failure Assessment score and survival outcome. The patients provided written informed consent for participation in the research (IRB number 2011-0001).
Confirmation of the diagnosis
Two pulmonologists reviewed the medical records of the patients with bilateral lung infiltrate and independently classified the diagnosis as infectious or non-infectious. In total, 136 patients for whom a consensus was achieved concerning the diagnosis participated in this study. The diagnosis of infection was made based on a combination of clinical manifestation with microbiological evidence and imaging findings. The disease was diagnosed as non-infectious when another cause was established for the pulmonary infiltrates and in the absence of significant bacterial growth in the culture of the BALF or endobronchial aspirated sputum.
Exosome extraction and proteomic profiling
The proteomic profiling of the exosomes in the BALF from 20 patients with bilateral lung infiltrates was conducted to identify candidate protein biomarkers potentially associated with infection or mortality. The patients were divided into four groups for proteomic profiling: infectious/survivor, infectious/non-survivor, non-infectious/survivor and non-infectious/non-survivor. Five patients from each group who exhibited high protein content in their exosome fractions were selected for the initial proteomic analysis.
The exosomes were extracted using differential ultracentrifugation and according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA), with some modifications. The exosome samples were buffered with lysis buffer (8 M urea, 50 mM ammonium bicarbonate, 1×Halt protease inhibitor cocktail, 0.1% benzonase nuclease and pH 8.0), lysed for 1 min with a probe sonicator (VCX-130, Sonics and Materials Inc., Newtown, CT, USA) at an amplitude of 28% in pulse mode (1 s on/2 s off) and centrifuged at 18 000×g for 10 min at 4 °C. The supernatant containing the extracted proteins of the samples was measured using a BCA protein quantification kit (Pierce BCA Protein Assay Kit, catalogue no 23225, Thermo Fisher Scientific, Waltham, MA, USA). Then, 300 µg of proteins were dissolved in 300 µL of lysis buffer, mixed 1:1 with 20 mM tris(2-carboxyethyl)phosphine hydrochloride (final concentration, 10 mM) and incubated for 1 h at 25 °C. Then iodoacetamide was added to reach a final concentration of 40 mM, and the samples were alkylated by incubation for 1 h in the dark at room temperature. The samples were diluted at 1:10 with 50 mM of ammonium bicarbonate and incubated with a mixture of trypsin/Lys-C (enzyme:sample, 1:25) for 16 h at 37 °C. The digestion reaction was stopped by adding formic acid at a final concentration of 0.3%. The peptide mixtures were desalted with a Sep-Pak C18 cartridge (Waters, Milford, MA, USA), lyophilised with a cold trap (CentriVap Cold Traps, Labconco, Kansas City, MO, USA) and stored at −80 °C until further use. The digested peptide was measured twice using the Dionex UltiMate 3000 RSLCnano system coupled with a Q Exactive mass spectrometer (Thermo Fisher Scientific). The liquid chromatography gradient and data-dependent acquisition MS options followed a previously published method [18, 19]. In addition, the acquired MS spectra were searched using Sequest HT on Proteome Discoverer (version 2.3, Thermo Scientific, USA) against the Swiss-Prot human proteome sequence database (May 2017) [18, 19].
Measurement of biomarkers
The BAL samples obtained from the 116 patients were used to measure the candidate biomarker levels. The Human B-lymphocyte antigen CD20 ELISA kit (LS Bio Inc., Seattle, WA, USA), the Human chloride intracellular channel 4 (CLIC4) ELISA kit (MyBioSource, San Diego, CA, USA), the Human Sec1 family domain-containing protein 1 (SCFD1) ELISA Kit (MyBioSource) and the ELISA kit for transporter associated with antigen processing 1 (TAP1) (MyBioSource) were used for the quantitative detection of CD20, CLIC4, SCFD1 and TAP1, respectively, in each sample, according to the manufacturers' instructions.
Statistical analysis
All the values were expressed as the mean±sd for the continuous variables or as percentages for the categorical variables. The Mann–Whitney U-test for numerical data and the Pearson's chi-squared test for categorical data were used to compare the variables between the groups. The receiver operating characteristic (ROC) curves were constructed to illustrate the best cut-off values for the biomarkers for predicting infectious causes or mortality. The C-statistic data between the combination of factors for predicting non-infectious causes were compared using the method proposed by Kang et al. [20]. Then a regression analysis was used to identify the independent risk factors for the non-infectious causes. Variables with p-values of <0.1 in the univariate analysis were entered into the multivariate models. ROC curves were constructed to illustrate the various cut-off values of CD20, neutrophil count and alveolar macrophage count in BAL fluid. Statistical significance was tested by comparing the area under the curve (AUC) with 0.5. We calculated p-values using a Z-statistic to determine if the AUC was significantly different from 0.5. To find the optimal threshold for classification, we used Youden's J statistic, which maximises the sum of sensitivity and specificity. A two-tailed p-value of <0.05 was considered to indicate statistical significance. The data were analysed using the SPSS Statistics 21 statistical analysis software (IBM Corporation, Armonk, NY, USA). The protein abundance plot was drawn using the R (version 3.6.0) software package ggplot2. The Venn diagram was drawn using jvenn [21]. The pathway enrichment analysis was performed using the FunRich 3.1.3 software [22].
Results
Selection of candidate protein biomarkers
The BALF exosome samples were collected from 20 patients divided into four groups, consisting of different combinations of differential diagnoses of infection or mortality (supplementary table 1). The proteomes of these 20 samples were analysed using duplicated liquid chromatography–tandem mass spectrometry. A total of 1776 proteins were quantified. Among them, 47 proteins were evaluated in the lung tissue samples using the Human Protein Atlas (https://www.proteinatlas.org/), and the top 10 exosomal proteins in the Vesiclepedia [23] were included (figure 1a). Most of these proteins (∼87.88%) were identified at least once in the Vesiclepedia, and about half of them (∼52.7%) were depicted as the exosome in the cellular component category of the Gene Ontology annotation (figure 1b). As a result of the proteomic profiling of the exosomes in the BALF from the 20 patients in the four groups, CD20, CLIC4, SCFD1 and TAP1 were mutually exclusive in each group and selected as candidate biomarkers for predicting the presence of infection or survival outcomes (figure 1c). The MS/MS spectra of these four proteins are shown in supplementary figure 1.
FIGURE 1.
Proteomes of bronchoalveolar fluid exomes a) Distribution of the protein abundance based on label-free quantification. The lung-elevated proteins are highlighted in red and the top 10 exosomal proteins are highlighted in black. b) Functional gene enrichment analysis of all the identified proteins using the FunRich 3.1.3 software for the cellular components. The blue bars represent the percentage of protein genes that are assigned to the indicated term; the yellow line (axis on right side) represents the value of log10(1/p-value). If the p-value is 0.05, the right-axis value is 1.3. The red bar indicates the p-value of 0.05 as a cut-off reference. c) Venn diagram of the number of proteins in the four groups (infectious/survival, non-infectious/survival, infectious/non-survival and non-infectious/non-survival) and the result table of the mass spectrometry peaks in the four groups: chloride intracellular channel 4 (CLIC4), transporter associated with antigen processing 1 (TAP1), Sec1 family domain-containing protein 1 (SCFD1) and B-lymphocyte antigen CD20 (CD20).
Patients’ clinical characteristics
A total of 116 patients participated in the measurement of the candidate biomarkers (CD20, CLIC4, TAP1 and SCFD1) in the BALF. The mean age of the subjects was 63.0 years, and 86% were men (table 1). The mortality rate, Sequential Organ Failure Assessment score, serum CRP and PCT levels did not differ between the patients with infectious or non-infectious causes. The most common microbial aetiology of the infectious causes was methicillin-resistant Staphylococcus aureus in 21 patients (table 2), whereas acute exacerbation of interstitial lung disease (n=13) was the most common non-infectious cause.
TABLE 1.
Comparison of the baseline characteristics of the non-infectious and the infectious causes of bilateral lung infiltrates
| Characteristics | All patients (n=116) | Non-infectious (n=58) | Infectious (n=58) | p-value |
|---|---|---|---|---|
| Age, years | 63.0±14.2 | 59.3±15.7 | 66.7±11.5 | 0.004 |
| Male | 86 (74.1) | 43 (74.1) | 43 (74.1) | 1.000 |
| Ever smoker | 55 (47.4) | 27 (46.6) | 28 (48.3) | 0.852 |
| Comorbidities | ||||
| Cardiovascular | 41 (35.3) | 17 (29.3) | 24 (41.4) | 0.174 |
| Diabetes | 13 (11.2) | 4 (6.9) | 9 (15.5) | 0.141 |
| Chronic kidney diseases | 11 (9.5) | 5 (8.6) | 6 (10.3) | 0.751 |
| Chronic respiratory diseases | 48 (41.4) | 24 (41.4) | 24 (41.4) | 1.000 |
| Gastrointestinal diseases | 21 (18.1) | 8 (13.8) | 13 (22.4) | 0.228 |
| Neurological diseases | 12 (10.3) | 2 (3.4) | 10 (17.2) | 0.015 |
| Malignancy | 41 (35.3) | 20 (34.5) | 21 (36.2) | 0.846 |
| SOFA score | 6.5±2.9 | 6.5±3.2 | 6.6±2.7 | 0.900 |
| P/F ratio | 182.8±88.5 | 192.7±94.3 | 172.8±81.8 | 0.228 |
| WBCs, µL−1 | 13 385±9352 | 11 405±6559 | 15 364±11 201 | 0.022 |
| CRP, mg·dL−1 | 13.6±9.8 | 13.0±8.7 | 14.3±10.9 | 0.457 |
| PCT, ng·mL−1 | 3.16±12.4 | 1.91±5.50 | 4.2±16.1 | 0.425 |
| BAL | ||||
| Neutrophil, % | 49.2±32.5 | 40.5±28.5 | 58.4±34.2 | 0.003 |
| Lymphocyte, % | 9.2±10.7 | 9.8±9.5 | 8.5±12.0 | 0.506 |
| Alveolar macrophage, % | 33.8±27.5 | 39.5±27.2 | 27.8±26.8 | 0.023 |
| CD20, ng·mL−1 | 213.4±156.2 | 248.6±154.5 | 177.6±150.9 | 0.014 |
| CLIC4, pg·mL−1 | 161.8±233.0 | 144.7±142.1 | 178.9±297.9 | 0.431 |
| TAP1, ng·mL−1 | 1.4±3.1 | 1.0±0.9 | 2.0±4.3 | 0.102 |
| SCFD1, ng·mL−1 | 7.0±5.2 | 5.8±5.1 | 8.1±5.2 | 0.194 |
Data are presented as mean±sd or n (%). SOFA: Sequential Organ Failure Assessment; P/F ratio: ratio of the partial pressure of arterial oxygen to the fraction of inspired oxygen; WBC: white blood cell; CRP: C-reactive protein; PCT: procalcitonin; BAL: bronchoalveolar lavage; CD20: B-lymphocyte antigen CD20; CLIC4: chloride intracellular channel 4; SCFD1: Sec1 family domain-containing protein 1; TAP1: transporter associated with antigen processing 1.
TABLE 2.
Infectious and non-infectious causes of bilateral lung infiltrates
| Infectious cause | n=58 |
| MRSA | 21 (36.2) |
| CRAB | 13 (22.4) |
| Corynebacterium | 6 (10.3) |
| Stenotrophomonas maltophilia | 5 (8.6) |
| Pseudomonas aeruginosa | 4 (6.9) |
| Gram-negative species# | 9 (15.5) |
| Gram-positive species¶ | 3 (5.2) |
| Influenza A | 1 (1.7) |
| Non-infectious cause | n=58 |
| AE-ILD | 13 (22.4) |
| Pulmonary congestion | 10 (17.2) |
| DAH | 9 (15.5) |
| Drug-induced lung injury | 7 (12.1) |
| AIP | 5 (8.6) |
| Radiation pneumonitis | 3 (5.2) |
| Organising pneumonia | 3 (5.2) |
| Malignancy | 3 (5.2) |
| Other+ | 5 (8.6) |
Data are presented as n (%). MRSA: methicillin-resistant Staphylococcus aureus; CRAB: carbapenem-resistant Acinetobacter baumannii; AE-ILD: acute exacerbation of interstitial lung disease; DAH: diffuse alveolar haemorrhage; AIP: acute interstitial pneumonia. #: The Gram-negative species were Elizabethkingia meningoseptica (n=3), Klebsiella pneumoniae (n=2), Serratia marcescens (n=1), Haemophilus influenzae (n=1), Enterobacter cloacae (n=1) and Escherichia coli (n=1). ¶: The Gram-positive species were Streptococcus pneumoniae (n=1), Enterococcus faecium (n=1) and Burkholderia cepacia (n=1). +: The other causes were postoperative ventilator-induced lung injury (n=2), hepatic failure-induced acute respiratory distress syndrome (n=1), trauma (n=1) and acute eosinophilic pneumonia (n=1).
Predicting values of non-infectious biomarkers
Among the candidate biomarkers, the level of CD20 was significantly elevated in patients with non-infectious causes, compared with those with infectious causes (248.6 versus 177.6 ng·mL−1; p=0.014; figure 2). The levels of CLIC4, TAP1 and SCFD1 did not differ between the two groups.
FIGURE 2.
The level of the biomarkers in the bronchoalveolar lavage fluid according to the infectious and non-infectious causes in patients with bilateral lung infiltrates. CD20: B-lymphocyte antigen CD20; CLIC4: chloride intracellular channel 4; TAP1: transporter associated with antigen processing 1.
In the ROC analysis, CD20 in the BALF was a significant predictor for non-infectious causes (figure 3; AUC=0.668, 95% confidence interval (CI) 0.567–0.769; p<0.001; cut-off value, 167.6 ng·mL−1; sensitivity 74.1%; specificity 63.2%). In a multivariate logistic regression analysis, a high CD20 level (≥167.6 ng·mL−1) in the BALF was independently associated with non-infectious causes (OR 4.414, 95% CI 1.804–10.798; p=0.001) (table 3). Blood white blood cell count and neutrophil percentage in BALF did not significantly differentiate between infectious and non-infectious causes.
FIGURE 3.
Receiver operating characteristic (ROC) curve for discriminating the non-infectious causes from the infectious causes in patients with bilateral lung infiltrates. CD20: B-lymphocyte antigen CD20; BAL: bronchoalveolar lavage; AM: alveolar macrophage; AUC: area under the curve; CI: confidential interval.
TABLE 3.
Risk factors for non-infectious causes of bilateral lung infiltrates assessed using a logistic regression analysis
| Risk factors | OR (95% CI) | p-value |
|---|---|---|
| Univariate analysis | ||
| Sex | 1.000 (0.436–2.296) | 1.000 |
| Age, years | 1.042 (1.011–1.073) | 0.007 |
| CD20 ≥167.6 ng·mL−1 | 4.914 (2.215–10.903) | 0.000 |
| WBCs, µL−1 | 1.000 (1.000–1.000) | 0.031 |
| BAL WBCs, µL−1 | 1.000 (0.999–1.000) | 0.059 |
| BAL neutrophils ≤16.5% | 0.847 (0.380–1.885) | 0.683 |
| BAL alveolar macrophages ≥22.0% | 4.362 (1.982–9.600) | 0.000 |
| Multivariate analysis | ||
| Age | 1.041 (1.007–1.076) | 0.019 |
| CD20 ≥167.6 ng·mL−1 | 4.414 (1.804–10.798) | 0.001 |
| WBCs, µL−1 | 1.000 (1.000–1.000) | 0.217 |
| BAL alveolar macrophages ≥22.0% | 2.230 (0.908–5.480) | 0.080 |
CD20: B-lymphocyte antigen CD20; WBC: white blood cell; BAL: bronchoalveolar lavage.
The level of CD20 correlated with the percentage of neutrophils (r=−0.190, p=0.042) and alveolar macrophages in the BALF (r=0.232, p=0.013). There were no significant differences in the levels of CD20 (207.8±150.8 versus 222.2±165.5; p=0.632), CLIC4 (153.3±237.0 versus 175.3±228.5; p=0.623), TAP1 (1.6±3.9 versus 1.1±0.9; p=0.414) and SCFD1 (3.5±3.6 versus 5.6±5.6; p=0.342) between the survivors and the non-survivors.
Discussion
In this study, we found several candidate protein biomarkers in the exosomes of the BALF using proteomic profiling for the diagnosis of infection and the prediction of mortality in patients with acute lung injury.
Our results showed that CD20, CLIC4, TAP1 and SCFD1 were exclusively quantified in each sample group. Among the candidate biomarkers, the CD20 levels in the BALF were significantly higher in patients with non-infectious causes than in those with infectious causes. However, these candidate biomarkers were not associated with the mortality of patients with bilateral lung infiltrates.
Confirming the existence of lung infections is critical for appropriate treatment selection, better outcomes, and predicting the prognosis early and accurately in patients with bilateral lung infiltrates. The clinical symptoms and the radiologic findings for the non-infectious disease mimic pneumonia, and many clinicians focus on pneumonia at the initial diagnosis and provide antibiotic treatment. However, acute lung injury associated with non-infectious causes requires specific management such as immune-modulating drugs, and often has poor outcomes without proper treatment [24]. Furthermore, the culturing of microorganisms to confirm the infectious aetiology will delay diagnosis and treatment [25], and the identification rate of pathogens is low. This uncertainty leads to the use of unnecessary antibiotics. Furthermore, broad-spectrum antibiotics in patients without infection may be harmful [26], and can be associated with superinfection with multi-resistant bacteria [27] and an increased duration of hospitalisation [28]. Consequently, a clinical tool is needed to improve the precise diagnosis and the decision-making process to avoid over-diagnosis or late diagnosis, both of which may result in worse outcomes [29].
Several biomarkers have been studied for the diagnostic stratification of infectious and non-infectious ICU patients. Pneumonia is the leading infectious cause of mortality worldwide and contributes significantly to the burden of antibiotic consumption. PCT and CRP remain the most widely used biomarkers. Meta-analyses suggest similar sensitivities for both markers in the diagnosis of infection (75% for CRP versus 77% for PCT), but PCT has a slightly higher specificity (67% for CRP versus 79% for PCT) [7]. However, PCT is elevated in various non-infectious conditions, such as cirrhosis, pancreatitis, mesenteric infarction and aspiration pneumonitis [9], as well as in conditions such as systemic inflammation response syndrome, multi-organ failure and previous infections. The CRP level is determined by its rate of synthesis in the liver, which determines the response to the inflammatory intensity. CRP levels are elevated in various pathologies, such as trauma, surgery and immunological-mediated inflammatory diseases [30]. A single elevated plasma CRP concentration is not very informative for diagnosing nosocomial pneumonia, and CRP tends not to increase in immunosuppressed patients [29, 31]. A meta-analysis by Shi et al. suggested that soluble triggering receptor expressed on myeloid cells-1 is a useful biomarker in ICU patients suffering from bacterial lung infections, but that further studies were required to confirm the ideal cut-off value [32]. Hellyer et al. showed that interleukin-1 beta and interleukin-8 effectively exclude ventilator-associated pneumonia (VAP) [33]. Such studies had heterogeneity in the clinical criteria for group stratification and showed a lack of specificity.
Biomarkers for pneumonia may indicate inflammation or may be released specifically after lung injury due to infection. Previous studies have compared the BAL findings of patients with bacterial pneumonia with those of patients with non-infectious diseases. Stolz et al. reported that the percentage of neutrophils in BALF and serum PCT levels were predictors of bacterial infection [34]. Sternberg et al. investigated the usefulness of BAL in assessing pneumonia in renal transplant patients and showed that the percentage of neutrophils in the BALF could predict bacterial pneumonia [35]. In our study, the differential cell count (percentage of neutrophils or macrophages) in BALF was not useful for distinguishing between infectious and non-infectious causes. Our study aimed to identify candidate biomarkers in BALF that directly reflect the inflammatory response to either infectious or non-infectious stimuli. However, the measured levels of these biomarkers should be interpreted cautiously and always correlated with clinical findings, as many confounding factors need to be considered for accurate interpretation [30]. Lu et al. described the BAL proteome from patients with VAP by identifying 206 proteins and creating a proteome map. Four of these 206 proteins – gelsolin, serum amyloid p-component, vitamin D-binding protein and pyruvate kinase – were significantly higher in the BAL of patients with VAP [36]. In addition, Nguyen et al. identified a BAL protein signature (S100A8, lactotransferrin and actinin-1) that identifies VAP patients with acute lung injury [37].
When independently identifying proteins in each group (infection versus non-infection, survival versus death), we screened for the available proteins associated with an immune reaction, intracellular transport or cell death. Among the four candidate proteins, only CD20 significantly increased in the non-infectious group. The biomarker CD20 is a B-lymphocyte antigen encoded by a membrane-spanning 4A family member (membrane-spanning 4-domains A1). It plays a role in B-cell development, in differentiation into plasma cells and in T-cell-independent antibody responses [38]. B-cells produce immunoglobulins, present antigens to T-cells and play additional key roles in the immune system. The function of CD20 is not fully elucidated. However, its structure suggests major hydrophobic regions, and it has been described as having features of a calcium channel with possible roles in B-cell activation and differentiation. Owing to its suitability for monoclonal antibody targeting, CD20-depleting agents are increasingly being used both on- and off-label for haematologic malignancies and autoimmune diseases [39, 40]. Nevertheless, the mechanism of action of CD20 in BALF requires further investigation.
Our study has some potential limitations that should be mentioned. First, this study is a single-centre study, and the number of enrolled patients is small. Although we used 20 samples for proteome profiling, dividing these into four groups further reduced the sample size per group. However, the biomarker detection group, comprising 116 patients, provided improved statistical power. The diagnostic performance of the candidate biomarkers was improved by considering disease with culture-proven bacterial pneumonia as having an infectious cause. However, the non-infectious causes were diverse because the prevalence of non-infectious causes was relatively low. For the validation of CD20 as a new biomarker, external validation in a multicentre study is needed. Second, to be precise, our result for CD20 means that this biomarker may be useful for distinguishing between bacterial pneumonia and non-infectious lung injury. Although bacterial pneumonia is a common cause of acute lung injury, compared with viral pneumonia or fungal pneumonia, further evaluation for viral or fungal infections is still needed. Third, CD20 and other biomarker candidates did not correlate with mortality outcomes in our study. However, this lack of correlation may be due to several factors. In predicting mortality among critically ill patients, a biomarker measured at a single initial time point may have limited sensitivity. The effects of initial treatment should also be considered. For instance, in sepsis cases, the change in lactate levels measured 6 h after the initial measurement is crucial for predicting mortality [41]. In our study, it was challenging to obtain serial samples because repeated BAL procedures are difficult to perform, especially in critically ill patients. This limitation restricted our ability to track biomarker changes over time, which may have affected our capacity to establish correlations with mortality outcomes. Fourth, the imbalance in comorbidities such as cardiovascular diseases, neurologic diseases and diabetes between infectious and non-infectious cases could potentially introduce confounding effects. Finally, omics data are a good tool for evaluating the markers of the disease process, but this analysis is limited to correlation and mostly reflects reactive processes rather than causative ones.
Conclusions
This study shows that the level of CD20 in BALF would be a useful biomarker to assist clinicians in excluding bacterial infectious causes in patients with bilateral lung injury. However, external validation is needed in further larger-scale studies to prove the clinical usefulness of this biomarker.
Supplementary material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material 00762-2024.SUPPLEMENT (244.7KB, pdf)
Footnotes
Provenance: Submitted article, peer reviewed.
Ethics approval: This study was approved by the Asan Medical Center Institutional Review Board (number 2011-0001).
Conflict of interest: The authors declare no conflict of interest.
Support statement: This study was supported by grant 2014-595 from the Asan Institute for Life Sciences, Asan Medical Center, and grant NRF-2017R1D1A1B03035034 from the National Research Foundation of Korea. Funding information for this article has been deposited with the Crossref Funder Registry.
References
- 1.Rubenfeld GD, Caldwell E, Peabody E, et al. Incidence and outcomes of acute lung injury. N Engl J Med 2005; 353: 1685–1693. doi: 10.1056/NEJMoa050333 [DOI] [PubMed] [Google Scholar]
- 2.Ong T, McClintock DE, Kallet RH, et al. Ratio of angiopoietin-2 to angiopoietin-1 as a predictor of mortality in acute lung injury patients. Crit Care Med 2010; 38: 1845–1851. doi: 10.1097/CCM.0b013e3181eaa5bf [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.National Heart, Lung, Blood Institute Acute Respiratory Distress Syndrome Clinical Trials Network , Parsons PE, Matthay MA, Ware LB, et al. Elevated plasma levels of soluble TNF receptors are associated with morbidity and mortality in patients with acute lung injury. Am J Physiol Lung Cell Mol Physiol 2005; 288: L426–L431. doi: 10.1152/ajplung.00302.2004 [DOI] [PubMed] [Google Scholar]
- 4.Pugin J, Ricou B, Steinberg KP, et al. Proinflammatory activity in bronchoalveolar lavage fluids from patients with ARDS, a prominent role for interleukin-1. Am J Respir Crit Care Med 1996; 153: 1850–1856. doi: 10.1164/ajrccm.153.6.8665045 [DOI] [PubMed] [Google Scholar]
- 5.Gibot S, Cravoisy A, Levy B, et al. Soluble triggering receptor expressed on myeloid cells and the diagnosis of pneumonia. N Engl J Med 2004; 350: 451–458. doi: 10.1056/NEJMoa031544 [DOI] [PubMed] [Google Scholar]
- 6.Simon L, Gauvin F, Amre DK, et al. Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis. Clin Infect Dis 2004; 39: 206–217. doi: 10.1086/421997 [DOI] [PubMed] [Google Scholar]
- 7.Wacker C, Prkno A, Brunkhorst FM, et al. Procalcitonin as a diagnostic marker for sepsis: a systematic review and meta-analysis. Lancet Infect Dis 2013; 13: 426–435. doi: 10.1016/S1473-3099(12)70323-7 [DOI] [PubMed] [Google Scholar]
- 8.Ware LB. Prognostic determinants of acute respiratory distress syndrome in adults: impact on clinical trial design. Crit Care Med 2005; 33: Suppl, S217–S222. doi: 10.1097/01.CCM.0000155788.39101.7E [DOI] [PubMed] [Google Scholar]
- 9.Karakioulaki M, Stolz D. Biomarkers in pneumonia-beyond procalcitonin. Int J Mol Sci 2019; 20: 2004. doi: 10.3390/ijms20082004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.De Toro J, Herschlik L, Waldner C, et al. Emerging roles of exosomes in normal and pathological conditions: new insights for diagnosis and therapeutic applications. Front Immunol 2015; 6: 203. doi: 10.3389/fimmu.2015.00203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lo Cicero A, Stahl PD, Raposo G. Extracellular vesicles shuffling intercellular messages: for good or for bad. Curr Opin Cell Biol 2015; 35: 69–77. doi: 10.1016/j.ceb.2015.04.013 [DOI] [PubMed] [Google Scholar]
- 12.Kadota T, Fujita Y, Yoshioka Y, et al. Extracellular vesicles in chronic obstructive pulmonary disease. Int J Mol Sci 2016; 17: 1801. doi: 10.3390/ijms17111801 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Monsel A, Zhu YG, Gudapati V, et al. Mesenchymal stem cell derived secretome and extracellular vesicles for acute lung injury and other inflammatory lung diseases. Expert Opin Biol Ther 2016; 16: 859–871. doi: 10.1517/14712598.2016.1170804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Makiguchi T, Yamada M, Yoshioka Y, et al. Serum extracellular vesicular miR-21-5p is a predictor of the prognosis in idiopathic pulmonary fibrosis. Respir Res 2016; 17: 110. doi: 10.1186/s12931-016-0427-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kim TH, Hong SB, Lim CM, et al. The role of exosomes in bronchoalveolar lavage from patients with acute respiratory distress syndrome. J Clin Med 2019; 8: 1148. doi: 10.3390/jcm8081148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Selevsek N, Chang CY, Gillet LC, et al. Reproducible and consistent quantification of the Saccharomyces cerevisiae proteome by SWATH-mass spectrometry. Mol Cell Proteomics 2015; 14: 739–749. doi: 10.1074/mcp.M113.035550 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hasin Y, Seldin M, Lusis A. Multi-omics approaches to disease. Genome Biol 2017; 18: 83. doi: 10.1186/s13059-017-1215-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ahn HS, Kim JH, Jeong H, et al. Differential urinary proteome analysis for predicting prognosis in type 2 diabetes patients with and without renal dysfunction. Int J Mol Sci 2020; 21: 4236. doi: 10.3390/ijms21124236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kim EY, Ahn HS, Lee MY, et al. An exploratory pilot study with plasma protein signatures associated with response of patients with depression to antidepressant treatment for 10 weeks. Biomedicines 2020; 8: 455. doi: 10.3390/biomedicines8110455 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kang L, Chen W, Petrick NA, et al. Comparing two correlated C indices with right-censored survival outcome: a one-shot nonparametric approach. Stat Med 2015; 34: 685–703. doi: 10.1002/sim.6370 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bardou P, Mariette J, Escudie F, et al. jvenn: an interactive Venn diagram viewer. BMC Bioinformatics 2014; 15: 293. doi: 10.1186/1471-2105-15-293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Pathan M, Keerthikumar S, Chisanga D, et al. A novel community driven software for functional enrichment analysis of extracellular vesicles data. J Extracell Vesicles 2017; 6: 1321455. doi: 10.1080/20013078.2017.1321455 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pathan M, Fonseka P, Chitti SV, et al. Vesiclepedia 2019: a compendium of RNA, proteins, lipids and metabolites in extracellular vesicles. Nucleic Acids Res 2019; 47: D516–D519. doi: 10.1093/nar/gky1029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Black AD. Non-infectious mimics of community-acquired pneumonia. Pneumonia (Nathan) 2016; 8: 2. doi: 10.1186/s41479-016-0002-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Iregui M, Ward S, Sherman G, et al. Clinical importance of delays in the initiation of appropriate antibiotic treatment for ventilator-associated pneumonia. Chest 2002; 122: 262–268. doi: 10.1378/chest.122.1.262 [DOI] [PubMed] [Google Scholar]
- 26.Improving Medicine through Pathway Assessment of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators , Kett DH, Cano E, Quartin AA, et al. Implementation of guidelines for management of possible multidrug-resistant pneumonia in intensive care: an observational, multicentre cohort study. Lancet Infect Dis 2011; 11: 181–189. doi: 10.1016/S1473-3099(10)70314-5 [DOI] [PubMed] [Google Scholar]
- 27.Amyes SG. Strategies and options for minimizing resistance emergence in pulmonary infections. Chest 1998; 113: Suppl, 228S–232S. doi: 10.1378/chest.113.3_Supplement.228S [DOI] [PubMed] [Google Scholar]
- 28.Birmingham MC, Hassett JM, Schentag JJ, et al. Assessing antibacterial pharmacoeconomics in the intensive care unit. Pharmacoeconomics 1997; 12: 637–647. doi: 10.2165/00019053-199712060-00004 [DOI] [PubMed] [Google Scholar]
- 29.Salluh JIF, Souza-Dantas VC, Povoa P. The current status of biomarkers for the diagnosis of nosocomial pneumonias. Curr Opin Crit Care 2017; 23: 391–397. doi: 10.1097/MCC.0000000000000442 [DOI] [PubMed] [Google Scholar]
- 30.Morley D, Torres A, Cilloniz C, et al. Predictors of treatment failure and clinical stability in patients with community acquired pneumonia. Ann Transl Med 2017; 5: 443. doi: 10.21037/atm.2017.06.54 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Youssef J, Novosad SA, Winthrop KL. Infection risk and safety of corticosteroid use. Rheum Dis Clin North Am 2016; 42: 157–176, ix–x. doi: 10.1016/j.rdc.2015.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shi JX, Li JS, Hu R, et al. Diagnostic value of sTREM-1 in bronchoalveolar lavage fluid in ICU patients with bacterial lung infections: a bivariate meta-analysis. PLoS One 2013; 8: e65436. doi: 10.1371/journal.pone.0065436 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hellyer TP, Morris AC, McAuley DF, et al. Diagnostic accuracy of pulmonary host inflammatory mediators in the exclusion of ventilator-acquired pneumonia. Thorax 2015; 70: 41–47. doi: 10.1136/thoraxjnl-2014-205766 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Stolz D, Stulz A, Muller B, et al. BAL neutrophils, serum procalcitonin, and C-reactive protein to predict bacterial infection in the immunocompromised host. Chest 2007; 132: 504–514. doi: 10.1378/chest.07-0175 [DOI] [PubMed] [Google Scholar]
- 35.Sternberg RI, Baughman RP, Dohn MN, et al. Utility of bronchoalveolar lavage in assessing pneumonia in immunosuppressed renal transplant recipients. Am J Med 1993; 95: 358–364. doi: 10.1016/0002-9343(93)90303-7 [DOI] [PubMed] [Google Scholar]
- 36.Lu YT, Han CL, Wu CL, et al. Proteomic profiles of bronchoalveolar lavage fluid from patients with ventilator-associated pneumonia by gel-assisted digestion and 2-D-LC/MS/MS. Proteomics Clin Appl 2008; 2: 1208–1222. doi: 10.1002/prca.200800069 [DOI] [PubMed] [Google Scholar]
- 37.Nguyen EV, Gharib SA, Palazzo SJ, et al. Proteomic profiling of bronchoalveolar lavage fluid in critically ill patients with ventilator-associated pneumonia. PLoS One 2013; 8: e58782. doi: 10.1371/journal.pone.0058782 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kuijpers TW, Bende RJ, Baars PA, et al. CD20 deficiency in humans results in impaired T cell-independent antibody responses. J Clin Invest 2010; 120: 214–222. doi: 10.1172/JCI40231 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Du FH, Mills EA, Mao-Draayer Y. Next-generation anti-CD20 monoclonal antibodies in autoimmune disease treatment. Auto Immun Highlights 2017; 8: 12. doi: 10.1007/s13317-017-0100-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Barun B, Bar-Or A. Treatment of multiple sclerosis with anti-CD20 antibodies. Clin Immunol 2012; 142: 31–37. doi: 10.1016/j.clim.2011.04.005 [DOI] [PubMed] [Google Scholar]
- 41.Ryoo SM, Lee J, Lee YS, et al. Lactate level versus lactate clearance for predicting mortality in patients with septic shock defined by sepsis-3. Crit Care Med 2018; 46: e489–e495. doi: 10.1097/CCM.0000000000003030 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material 00762-2024.SUPPLEMENT (244.7KB, pdf)