Figures
Abstract
Background
The electronic nose (e-nose) detects volatile organic compounds (VOCs) in exhaled air. We hypothesized that the exhaled VOCs print is different in stable vs. exacerbated patients with chronic obstructive pulmonary disease (COPD), particularly if the latter is associated with airway bacterial infection, and that the e-nose can distinguish them.
Methods
Smell-prints of the bacteria most commonly involved in exacerbations of COPD (ECOPD) were identified in vitro. Subsequently, we tested our hypothesis in 93 patients with ECOPD, 19 of them with pneumonia, 50 with stable COPD and 30 healthy controls in a cross-sectional case-controlled study. Secondly, ECOPD patients were re-studied after 2 months if clinically stable. Exhaled air was collected within a Tedlar bag and processed by a Cynarose 320 e-nose. Breath-prints were analyzed by Linear Discriminant Analysis (LDA) with “One Out” technique and Sensor logic Relations (SLR). Sputum samples were collected for culture.
Results
ECOPD with evidence of infection were significantly distinguishable from non-infected ECOPD (p = 0.018), with better accuracy when ECOPD was associated to pneumonia. The same patients with ECOPD were significantly distinguishable from stable COPD during follow-up (p = 0.018), unless the patient was colonized. Additionally, breath-prints from COPD patients were significantly distinguished from healthy controls. Various bacteria species were identified in culture but the e-nose was unable to identify accurately the bacteria smell-print in infected patients.
Citation: Shafiek H, Fiorentino F, Merino JL, López C, Oliver A, Segura J, et al. (2015) Using the Electronic Nose to Identify Airway Infection during COPD Exacerbations. PLoS ONE 10(9): e0135199. https://doi.org/10.1371/journal.pone.0135199
Editor: Konstantinos Kostikas, University of Athens Medical School, GREECE
Received: May 26, 2015; Accepted: July 18, 2015; Published: September 9, 2015
Copyright: © 2015 Shafiek et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This project was supported by the Instituto de Salud Carlos III (FIS PI12/02628) and the Spanish Respiratory Society. The funding sources were used to cover fungible expenses and had no influence on the results.
Competing interests: The authors declare that no competing interests exist related with the technology or therapy used in this manuscript.
Introduction
Exacerbations of Chronic Obstructive Pulmonary Disease (ECOPD) are important events in the natural history of the disease because they influence disease progression, morbidity [1] and mortality [2]. The diagnosis of ECOPD relies on the increase of respiratory symptoms reported by the patient with or without an increase in sputum volume or purulence [3, 4]. Hence, despite the high sanitary burden of ECOPD, their diagnosis and treatment are currently empirical [3, 5].
ECOPD are characterized by a burst of pulmonary and systemic inflammation [6], generally believed to be the result of viral and/or bacterial airway infections [7, 8]. However, potential pathogenic micro-organisms (PPM) can be recovered during ECOPD in only a proportion of patients (30% of sputum cultures and 50% of bronchial secretion cultures) [9, 10]. These percentages increase slightly (59%) if quantitative PCR is used [11]. Besides, PPB can also be recovered in a proportion of clinically stable COPD patients [8]. Therefore, novel methods capable of identifying more precisely the role of infection during ECOPD are needed. They can contribute to improve the quality of care provided to these patients since they may foster a more appropriate use of antibiotic therapy under these circumstances.
Over the past decade there has been increasing interest on the potential diagnostic value of volatile organic compounds (VOCs) exhaled in human breath [12]. The electronic nose (e-nose) constitutes an emerging non-invasive technique capable of detecting and differentiating VOCs patterns (“smell prints”) in humans (“breatheomics”) [13, 14]. Recent studies have shown that the e-nose can reliably identify patients with bronchial asthma [15], lung cancer [16], bacterial pneumonia [17] and bacterial sinusitis [18], as well as to identify and classify various bacterial species cultured in vitro [19, 20]. In a pilot study, our group previously showed that the e-nose can detect bacterial colonization in clinically stable patients with COPD [16, 21].
Considering all these previous observations we hypothesized that the VOCs smell prints will be different during ECOPD vs. clinical stability, will discriminate infectious vs. non-infectious ECOPD or pneumonia, and will identify the specific bacterial species present in the airways during ECOPD. Accordingly, in this study we sought to: (1) determine the sensitivity, specificity, positive predictive value and negative predictive value of the e-nose in the bacteriological diagnosis of ECOPD with or without pneumonia which is our primary objective; (2) investigate whether the VOCs smell print changes in the transition from ECOPD to clinical stability; (3) explore the feasibility of developing a system capable of identifying and establish the presence of bacterial infection in an exhaled air sample by analysing libraries of bacteria previously identified and characterized in vitro.
Methods
Study design and Ethics
The current study was divided into 2 parts. The first part is a cross-sectional descriptive, case-controlled study that included COPD patients hospitalized because of ECOPD (with and without pneumonia) and clinically stable COPD patients as well as healthy control subjects. The second part is a longitudinal study that compared ECOPD patients (1st visit) and the same ones at clinical stability two months later (2nd visit); Fig 1. The sample size was calculated using a two sided Fisher exact’s test for independent case-control study with a significance level (α) of 0.05 and power of test of 80% considering that PPM growth in sputum could be detected in 30% of the patients with ECOPD [22, 23] and based on our pilot study published by Sibila et al [21]. All the patients were recruited from the ward (ECOPD), out-patient clinic (stable COPD) and lung function laboratory (controls) at Son Espases University Hospital in Palma de Mallorca (Spain). All of the participants signed their informed consent, and the study protocol was approved by the local Ethics Committee of Balearic Islands.
Study population
COPD was defined by a post-bronchodilator FEV1/FVC<70% and history of tobacco consumption >10 pack-year following GOLD recommendations [5]. We included in the study patients with a diagnosis of ECOPD according to GOLD guidelines [5] defined by acute worsening of daily respiratory symptoms-beyond the normal day to day variations- that required a change in medication; ECOPD patients that had associated pneumonia proved by the presence of new infiltrates on chest radiography [24] were also included. These patients were compared with clinically stable COPD defined by absence of an exacerbation requiring antibiotic or steroid therapy within 30 days prior to inclusion and control healthy subjects with normal lung function recruited from our pulmonary function unit (Fig 1). Infective episode of ECOPD was defined by the presence of PPM growth in sputum culture (see below) in patients fulfilling the previous definition of ECOPD. Colonization was defined by the presence of PPM growth in stable COPD patients. Patients with no PPM growth in sputum culture were considered non-infected for the study purpose. All ECOPD were followed up and studied again 2 months after exacerbation, when clinically stable.
Patients with a history of asthma, primary bronchiectasis, lung cancer, active tuberculosis, interstitial lung diseases, active infection other than those of pulmonary origin including upper airways infection, and those who had taken antibiotics in the last seven days were excluded from the study.
Characterization of participants.
We obtained in all patients a detailed clinical history (including age, gender, smoking history, relevant comorbid conditions, number of previous exacerbations, and drug history), a chest X-ray, pulmonary function measurements (including forced spirometry, carbon monoxide diffusing capacity (DLCO) and static lung volumes) following international recommendations [25] and using reference values of a Mediterranean population [26]. Chest Computed Tomography (CT) was performed if indicated clinically. Spontaneous sputum and exhaled air samples were collected consecutively in all patients.
E-nose measurements
Exhaled breath gas was collected for e-nose assessment as previously described [15, 27]. Briefly, the exhaled gas from all subjects −as well as clear air collected from the same environment as baseline reference− were collected in a 5 Liters Tedlar bag after 3 minutes of tidal breathing through a Hans-Rudolph valve with an expiratory silica reservoir exposed to dry air and an inspiratory filter. The e-nose device (Cyranose 320; Smith Detections, Pasadena, CA), with 32 organic polymeric nano-composite sensor arrays, was then connected to the Tedlar bag as five to six measurements of the each sample were obtained over 5 minutes duration. Changes in the nanosensor’s electrical resistance between clear air and subjects’ exhaled breath generate a breath-print VOCs profile. As a difference from previous publications, the resistance value was monitored during the whole measuring time using several parameters (as slope values) to generate the breath-print profile. Transcutaneous oxygen saturation (SaO2) was continuously monitored during the procedure.
Sputum culture
Spontaneous sputum samples were collected within 24 hours of presentation of ECOPD symptoms as well as from clinically stable COPD cases. All the sputum samples were collected in sterile containers and transferred to the microbiological laboratory within 2 hours of collection for Gram staining and bacterial culture. All the sputum samples were classified according to Murray-Washington criteria, thus grade IV (10–25 epithelial cells and > 25 leucocytes per field) and grade V (< 10 epithelial cells and > 25 leucocytes per field) were cultured quantitatively and a bacterial load ≥106 colony forming units (CFU)/ml was considered as significant. Culture was performed in non-selective (sheep blood agar, chocolate agar) and selective media (MacConkey agar) at 37°C for 48 hours in 5% CO2, and bacterial species identification was conducted biochemically (Microscan, Siemens). The identified microorganisms were classified as potentially pathogenic microorganisms (PPM) including Streptococcus pneumoniae, Haemophilus influenzae, Pseudomonas aeruginosa, Staphylococcus aureus methicillin susceptible or resistant, Moraxella catarrhalis, and Gram negative-bacilli, and non-PPM included Candida spp., Streptococcus viridans, Neisseria spp., Staphylococcus epidermidis and Corynebacterium spp.
In vitro e-nose learning
To let the e-nose learn the breath print of specific bacteria, at least one representative clinical isolate for each species included in the potentially pathogenic microorganism group was analyzed by using the e-nose, specifically 3 S. pneumoniae, 3 H. influenzae, 3 P. aeruginosa, 1 S. aureus, 1 M. catarrhalis, and 1 E.coli isolate were included (Table A in S1 File). These bacteria were cultured on appropriate non-selective media (sheep blood agar and/or chocolate agar) at 37°C for 18–24 hours in 5% CO2 on Petri plates. Then, plates were exposed to e-nose inhalation to build up a library of VOCs pattern for each bacteria following previously described methodology [19]. Briefly, filtered air from a Tedlar bag was introduced into the close plates through a hole located in one extreme of the plate cover; then the VOC were extracted and delivered to the e-nose through a second hole located in the other side of the plate cover (Fig 2). Several exposition times were tried and finally an exposition time of about 6 minutes was selected as it coincides with no changes in the slope of the sensors. The e-nose sensors registered continuously the relative changes in resistance due to volatile organic compounds in the headspace (i.e., the space over the agar media in the Petri plates) using mathematical algorithms for each bacterium independently on the culture media used. Each measure was compared with a reference measure to obtain the sensor readout, being the reference a non-inoculated Petri plate. For each plate of the different bacterial strains, the measurement was repeated eight times with a total duration of 20 minutes.
Data analysis
Results are presented as mean ± standard deviation (SD) for quantitative variables or absolute numbers and relative frequencies for categorical variables. Comparisons between different groups were analysed with the Chi Square, Mann-Whitney or Kruskal-Wallis tests as appropriate. MedCalc (version 9.2.1.0, Acacialaan 22, B-8400 Ostend, Belgium) was used for analysis.
Breath-print data from ECOPD patients (infected and non-infected), ECOPD with associated pneumonia, stable COPD as well as healthy controls were analysed using two methods: Linear Discriminant Analysis (LDA) for comparison between the different groups [15] and Sensor Logic Relations (SLR), a new data management approach to identify different microorganisms in exhaled breath. Briefly, instead of considering the values of the relative resistance to generate the breath-print for each patient, breath-print is built with Logic Relations between the Resistor values as “Resistor 2 value is greater than Resistor 5”. This technique is adopted to consider the relative sensitivity of two sensors to a certain VOC profile (additional information provided in S1 File). The results were expressed as success classification ratio [28] as well as sensitivity and specificity.
Results
Population characteristics
We included in the study 185 subjects, 93 of whom had a diagnosis of ECOPD (19 of them with associated pneumonia), fifty patients with clinically stable COPD and 30 control healthy subjects (Fig 1).
Table 1 presents the main demographic and clinical characteristics of the population studied. COPD patients-either ECOPD or clinically stable- were mostly males, with a similar age and proportion of co-morbidities. Cumulative smoking exposure (pack-years) was higher in COPD patients than healthy controls; however the ex-smokers constituted the highest proportion among all groups (Table 1). COPD patients had moderate to severe airflow limitation and reduced DLCO whereas controls had normal lung function (Table 1). Previous admissions were higher in the ECOPD group compared to stable COPD patients and healthy controls. Chest CT scanners were available for 108 participants. Emphysematous changes were detected in > 50% of all COPD patients and in 13.3% of smoker controls. Bronchiectasis was detected only among COPD groups being more prevalent in ECOPD either with or without associated pneumonia (20.5% and 15.7% respectively; Table 1) than stable COPD patients.
Microbiological cultures
Forty-two patients (45%) with ECOPD and seventeen (34%) of stable COPD had PPM in the baseline characterization (Table 1 and S1 Fig A in S1 File). Sputum samples did not show growth of PPM in 19% of exacerbation episodes and were considered as non-infective episodes. Pseudomonas aeruginosa was the most commonly isolated microorganism (34%) followed by Haemophilus influenzae (13.6%), Streptococcus Pneumoniae (10.2%), Moraxella catarrhalis (8.5%) then Staphylococcus aureus methicillin susceptible (6.8%). Other species of bacteria were isolated in 23.7% and Candida spp. was isolated in 10.2%.
Breath-print analysis
Exhaled-breath air collection during the ECOPD was safe without evidence of significant desaturation (p = 0.773, Table 1). The breath-prints from different COPD groups, either exacerbated or stable, were statistically discriminated from those obtained from healthy control subjects (p < 0.05, Table 2).
ECOPD compared to stable COPD.
In the cross-sectional analysis, breath-prints from ECOPD were significantly distinguishable from stable COPD in case of absence of PPM (p < 0.05; Table 2). Breath-prints of stable COPD patients with evidence of PPM, indicating the presence of airway colonization, were not statistically different from infective ECOPD (p = 0.074; Table 2).
In the longitudinal analysis, when the same patient was compared between exacerbation and stability, the e-nose was able to distinguish breath-prints during ECOPD (1st visit) from clinical stability (2nd visit) with a classification success ratio of 70%, sensitivity of 72% and specificity of 67% (Table 3). Stable COPD patients during the 2nd visit with evidence of PPM in sputum were excluded from the analysis.
Pneumonia vs. non-pneumonia patients.
Breath-prints from pneumonic ECOPD were significantly distinguishable from stable COPD (classification success ratio: 86%, with sensitivity of 88% and specificity of 75%, p <0.001; Table 2). Further, the e-nose was able to significantly discriminate the breath-prints of ECOPD associated with pneumonia from non-pneumonic ECOPD regardless the presence of PPM (classification success ratio: 88%, p = 0.005 with a sensitivity of 91% and specificity of 75%, Table 2, Fig 3A) or absence of PPM (classification success ratio: 86%, p = 0.021; sensitivity = 85%, specificity = 86%; Table 2, Fig 3B).
Infected vs. non-infected ECOPD.
Among the ECOPD, the e-nose was able to discriminate breath-prints of infected patients versus non-infected patients with a classification success ratio of 75% (p = 0.018), sensitivity of 88% and specificity of 60% (Table 3, Fig 3C). Moreover, the percentage of success ratio was further improved in the pneumonic subgroup of ECOPD (classification success ratio: 100%, p = 0.014; sensitivity and specificity of 100%; Table 3, Fig 3D).
Bacterial identification in vitro.
When comparing different species of bacteria in vitro, the e-nose successfully discriminated the smell-print of two single species of bacteria (p <0.01, Fig 4A–4D and Fig B in S1 File) using the LDA method. Upon analysing the same smell-prints using SLR, one single species of bacteria (namely Pseudomonas aeruginosa, Streptococcus pneumoniae, Haemophilus influenzae, and Escherichia coli) was successfully discriminated versus all other species with success ratios varying between 97 and 100%. Results upon single comparisons are provided in the on-line data supplement (Tables A, B, C, D and E in S1 File).
The graphs shows two-dimensional principal component (PC) analysis plots showing smell-prints discrimination between 2 species of bacteria.
Bacterial identification in vivo.
Despite the successful discrimination of bacteria species in vitro, and the effective detection of infective ECOPD, the e-nose failed to identify a breath-print of an individual bacterium in the infected patient when compared to a non-infected, except for Haemophilus influenzae. Where different sensitive sensors obtained during in vitro analysis (Tables B, C and D in S1 File) did not show similar sensitivity for analysing the same bacterial species in vivo, this was not the case for Hemophilus influenza in which sensors 14, 15 and 19 (Table 4 and Table D in S1 File) were linked to the presence of Hemophilus influenza both in vitro and in vivo. Moreover, two different species of bacteria could be differentiated with a success classification ratio of 82%, 95% and 100% using SLR analysis for Pseudomonas aeruginosa vs Hemophilus influenzae, Pseudomonas aeruginosa vs Escherichia coli and Hemophilus influenza vs. Escherichia coli respectively (Table 4).
Discussion
In the current study, we described for the first time the validity of the e-nose technology in the discrimination of infective episodes for ECOPD. Using simple mathematical algorithms the e-nose distinguished specific breath-prints for infected ECOPD, and was able to detect when the exacerbation was associated with pneumonia in comparison to non-infected ECOPD. Further, it was able to identify single species of bacteria in vitro and still could differentiate between two different groups of bacteria such as Pseudomonas aeruginosa and Hemophilus influenzae in vivo.
Previous studies
Sputum cultures for microorganism detection are considered to be of limited value in investigating the etiology of ECOPD [9, 10]. Stolz et al [29] identified pathogenic bacteria in the sputum samples in 36% to 38% in their randomized groups of ECOPD. Recently, Boixeda et al [30] identified bacterial agent in only 24.1% of their cohort as an etiology of ECOPD, being Pseudomonas aeruginosa and Haemophilus influenza the most common detected microorganisms. Shykhon et al [31] and Lai et al [32] identified infection from the upper respiratory tract using e-nose with an accuracy of 88.2%. Thaler et al [18] reported correct e-nose diagnosis of bacterial sinusitis in 72% of their cohort. Further, Hanson et al [17] and Hockstein et al [33] succeeded to detect bacterial infection in ICU patients with ventilator associated pneumonia using e-nose with a prediction accuracy of more than 90%. Much more recently, Sibila et al [21] reported 88% accuracy of e-nose in the identification of airway bacterial colonization in stable COPD patients. Our results agree with these previous observations, as 45% of our sputum samples during ECOPD showed PPM and we were able to differentiate them from non-infective episodes with e-nose technology.
Interpretation of results
ECOPD is generally associated with a burst of airway inflammation [6, 34, 35]. This could be the cause for a specific VOCs breath pattern in this condition, which is independent of the presence of infection. Our results confirmed this concept as we demonstrated that a specific breath pattern of ECOPD distinguishable from stable COPD by classification success ratio of 76% in the absence of PPM. Additionally, we demonstrated a significant distinct breath-print pattern of ECOPD using e-nose and clearly distinguishable from healthy controls. This also can be explained on the basis that the e-nose can clearly differentiate the COPD as a chronic inflammatory airway disease [5, 21] with special breath pattern from others. Also, Fens et al [27] and Dragonieri et al [36] discriminated COPD from asthma and from non-small cell cancer with accuracy of 96% and 85% respectively.
Interestingly, we demonstrated a specific breath pattern for infected ECOPD versus non-infective episodes with a classification success ratio of 75%. This pattern can be due to firstly the intensive airway inflammatory response and oxidative stress associated with bacterial infection in ECOPD [37, 38] that could be a direct cause of specific VOCs breath pattern easily detectable by the e-nose. Secondly, the e-nose technology is able to identify bacterial pathogens [19, 20]. This is supported by the recent results demonstrated by our previous pilot study [21] as breath-print for bacterial colonized COPD was differentiated from non-colonized with accuracy of 89%. Also, it may explain the inability of the e-nose to differentiate colonized stable patients from acutely infected during exacerbation in our analysis.
Further, the breath pattern of ECOPD was distinct from those presented with associated pneumonia either in case of positive PPM (classification success ratio of 88%) or not (classification success ratio of 86%) indicating two different acute events with higher inflammatory response in case of pneumonia [39].
According to our results, different bacterial species were clearly discriminated in vitro which was previously demonstrated by Dutta et al [19] and Pavlou et al [20]. Moreover, in COPD patients, the e-nose technology could differentiate between two different groups of bacteria in the exhaled breath samples using SLR; as Pseudomonas aeruginosa versus Hemophilus influenzae could be discriminated with excellent success ratio. Joensen et al [40] could differentiate cystic fibrosis patients colonized with Pseudomonas aeruginosa and failed to differentiate patients colonized with other pathogens. However, there was no clear link between the in vitro sensitive sensors used for detection of bacterial species and in vivo detection except partially for Hemophilus influenza. This could be due to high complexity of the exhaled breath which could be affected by various factors as diet, medications, endogenous metabolism, metabolism by concomitant resident and pathogen microorganisms as well as varies airway inflammatory process accompanied the ECOPD [6, 34, 35].
Clinical implications
Our results showed that the e-nose is able to discriminate a specific exhaled breath pattern for severe ECOPD-compared to healthy control, stable COPD and COPD with pneumonia- using a noninvasive, handheld, well tolerated and rapid novel technique. Although it is not specific enough to guide antibiotic therapy, it has the potential to become a diagnostic tool for ECOPD associated to bacterial infection in daily clinical practice rather than being dependent on symptomatic diagnosis [3, 4]. Further, our study shows the ability of the e-nose to identified the infective ECOPD as well as COPD with pneumonia which help in the decision of starting antibiotics rather than being empirical or depending on the severity of exacerbation [41]. This will guard against the emergence of multidrug resistance and modulate the economic burden of antibiotic therapy on one hand and different biomarkers for diagnosing bacterial infection as a cause for ECOPD or guidance of antibiotic therapy [29, 42] on the other hand.
Further studies are needed to determine whether stratification of patients by severity or by clinical phenotype (i.e. emphysematous/non-emphysematous, frequent exacerbator, COPD/asthma overlap, etc) are able to reduce the variance of measures.
Study limitations
Our study has some limitations; firstly, some patients were unable to produce sputum samples and they were considered as non-infective episodes, which could not be the case, and we assume that this fact could be responsible for the low specificity found. Further, we did not consider the variation of the exacerbations’ duration at assessment. Secondly, we did not search for viral aetiology as a cause of ECOPD [43] or pneumonia in our cohort. However, bacteria still constitute the commonest cause of ECOPD [7]. Thirdly, we used only Cyranose e-nose as a diagnostic tool in all our analysis. However, Lai et al [32] and Dutta et al [19] used a similar device to identify bacterial pathogens in upper respiratory infection. More recently, Joensen et al [40] and Sibila et al [21] used Cyranose to identify infection in exhaled breath in cystic fibrosis and colonized stable COPD respectively. Further, technical reproducibility and repeatability of e-nose measurements was tested previously for Cyranose e-nose by Fens et al [27] and McWilliams et al [44] with excellent correlation coefficients indicating similar sensor responses. Finally, our study detected VOC mixtures rather than identifying a specific VOC. Further, we did not measure inflammation in the sputum which is highly accentuated during ECOPD [6, 34, 35]. In order to know which compound of VOCs is associate with exacerbations and responsible to exhaled breath discrimination as well as better identification of inflammation, the application of other techniques as gas chromatography and mass spectrometry is required.
Conclusions
The e-nose is a well-tolerated, non-invasive, and rapid technique that could be a useful tool to identify infective ECOPD episodes of bacterial aetiology. Furthermore, it is able to successfully distinguish ECOPD associated with pneumonia even if microorganism is not cultured. However, further studies are needed to improve the accuracy in detecting specific micro-organisms in vivo despite its consistent results in vitro.
Supporting Information
S1 File. Contains the following files: Table A. Database of the bacteria analyzed in the culture media; Table B. LDA results in the comparison between two different species of bacteria in the culture media; Table C. LDA results in the comparison between one species of bacteria and the other species in the culture media; Table D. SLR results in the comparison between one species of bacteria and the other species in the culture media; Table E. LDA results in the comparison between two groups of bacteria in the breath-print of the patients; Fig A. The prevalence of infection, mixed flora and invalid sputum for culture among different studied groups; Fig B. Comparison between e-nose smell-prints among different species of bacteria.
The graphs shows two-dimensional principal component (PC) analysis plots showing smell-prints discrimination between 2 species of bacteria in Fig B-A and between one species and all the others in Fig B-B and C.
https://doi.org/10.1371/journal.pone.0135199.s001
(DOCX)
Acknowledgments
Hanaa Shafiek was granted by the University of Alexandria and Ministry of Higher Education of Egypt as member of ParOwn (the Partnership and Ownership initiative). Authors are grateful to all the Lung Function Unit team and research nurses Rocio Cordova and Magdalena Pan for their contribution in the logistics of the study.
This study was funded by the Spanish Institute of Health Carlos III (FIS 12/02628) with FEDER Funds from the European Union and a grant from the Spanish Respiratory Society (SEPAR).
Author Contributions
Conceived and designed the experiments: BGC CL AO JLM. Performed the experiments: HS FF OS CL JLM IdP. Analyzed the data: HS JLM BGC. Contributed reagents/materials/analysis tools: JLM JS IdP CL. Wrote the paper: HS BGC AA AO.
References
- 1. Donaldson GC, Seemungal TA, Bhowmik A, Wedzicha JA. Relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax. 2002 Oct;57(10):847–52. pmid:12324669
- 2. Spencer S, Jones PW. Time course of recovery of health status following an infective exacerbation of chronic bronchitis. Thorax. 2003 Jul;58(7):589–93. pmid:12832673
- 3. Anthonisen NR, Manfreda J, Warren CP, Hershfield ES, Harding GK, Nelson NA. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med. 1987 Feb;106(2):196–204. pmid:3492164
- 4. Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004 Jun;23(6):932–46. pmid:15219010
- 5. Vestbo J, Hurd SS, Agusti AG, Jones PW, Vogelmeier C, Anzueto A, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med. 2013 Feb 15;187(4):347–65. pmid:22878278
- 6. Hurst JR, Perera WR, Wilkinson TM, Donaldson GC, Wedzicha JA. Systemic and upper and lower airway inflammation at exacerbation of chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2006 Jan 1;173(1):71–8. pmid:16179639
- 7. Papi A, Bellettato CM, Braccioni F, Romagnoli M, Casolari P, Caramori G, et.al. Infections and airway inflammation in chronic obstructive pulmonary disease severe exacerbations. Am J Respir Crit Care Med 2006;173:1114–21. pmid:16484677
- 8. Smith CB, Golden C, Klauber MR, Kanner R, Renzetti A. Interactions between viruses and bacteria in patients with chronic bronchitis. J Infect Dis. 1976 Dec;134(6):552–61. pmid:12235
- 9. Monso E, Ruiz J, Rosell A, Manterola J, Fiz J, Morera J, et al. Bacterial infection in chronic obstructive pulmonary disease. A study of stable and exacerbated outpatients using the protected specimen brush. Am J Respir Crit Care Med. 1995 Oct;152(4 Pt 1):1316–20. pmid:7551388
- 10. Fagon JY, Chastre J, Trouillet JL, Domart Y, Dombret MC, Bornet M, et al. Characterization of distal bronchial microflora during acute exacerbation of chronic bronchitis. Use of the protected specimen brush technique in 54 mechanically ventilated patients. Am Rev Respir Dis. 1990 Nov;142(5):1004–8. pmid:2240819
- 11. Garcha DS, Thurston SJ, Patel AR, Mackay AJ, Goldring JJ, Donaldson GC, et al. Changes in prevalence and load of airway bacteria using quantitative PCR in stable and exacerbated COPD. Thorax. 2012 Dec;67(12):1075–80. pmid:22863758
- 12. Fens N, van der Schee MP, Brinkman P, Sterk PJ. Exhaled breath analysis by electronic nose in airways disease. Established issues and key questions. Clin Exp Allergy. 2013 Jul;43(7):705–15. pmid:23786277
- 13. Turner AP, Magan N. Electronic noses and disease diagnostics. Nat Rev Microbiol. 2004 Feb;2(2):161–6. pmid:15040263
- 14. Moser B, Bodrogi F, Eibl G, Lechner M, Rieder J, Lirk P. Mass spectrometric profile of exhaled breath—field study by PTR-MS. Respir Physiol Neurobiol. 2005 Feb 15;145(2–3):295–300. pmid:15705543
- 15. Dragonieri S, Schot R, Mertens BJ, Le Cessie S, Gauw SA, Spanevello A, et al. An electronic nose in the discrimination of patients with asthma and controls. J Allergy Clin Immunol. 2007 Oct;120(4):856–62. pmid:17658592
- 16. Machado RF, Laskowski D, Deffenderfer O, Burch T, Zheng S, Mazzone PJ, et al. Detection of lung cancer by sensor array analyses of exhaled breath. Am J Respir Crit Care Med. 2005 Jun 1;171(11):1286–91. pmid:15750044
- 17. Hanson CW 3rd, Thaler ER. Electronic nose prediction of a clinical pneumonia score: biosensors and microbes. Anesthesiology. 2005 Jan;102(1):63–8. pmid:15618788
- 18. Thaler ER, Hanson CW. Use of an electronic nose to diagnose bacterial sinusitis. Am J Rhinol. 2006 Mar-Apr;20(2):170–2. pmid:16686381
- 19. Dutta R, Hines EL, Gardner JW, Boilot P. Bacteria classification using Cyranose 320 electronic nose. Biomed Eng Online. 2002 Oct 16;1:4. pmid:12437783
- 20. Pavlou A, Turner AP, Magan N. Recognition of anaerobic bacterial isolates in vitro using electronic nose technology. Lett Appl Microbiol. 2002;35(5):366–9. pmid:12390482
- 21. Sibila O, Garcia-Bellmunt L, Giner J, Merino JL, Suarez-Cuartin G, Torrego A, et al. Identification of airway bacterial colonization by an electronic nose in Chronic Obstructive Pulmonary Disease. Respir Med. 2014 Sep 19.
- 22. Rodriguez-Roisin R. Toward a consensus definition for COPD exacerbations. Chest. 2000 May;117(5 Suppl 2):398S–401S. pmid:10843984
- 23. Voelkel NF, Tuder R. COPD: exacerbation. Chest. 2000 May;117(5 Suppl 2):376S–9S. pmid:10843980
- 24. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis. 2007 Mar 1;44 Suppl 2:S27–72. pmid:17278083
- 25. Laszlo G. Standardisation of lung function testing: helpful guidance from the ATS/ERS Task Force. Thorax. 2006 Sep;61(9):744–6. pmid:16936234
- 26. Roca J, Sanchis J, Agusti-Vidal A, Segarra F, Navajas D, Rodriguez-Roisin R, et al. Spirometric reference values from a Mediterranean population. Bull Eur Physiopathol Respir. 1986 May-Jun;22(3):217–24. pmid:3730638
- 27. Fens N, Zwinderman AH, van der Schee MP, de Nijs SB, Dijkers E, Roldaan AC, et al. Exhaled breath profiling enables discrimination of chronic obstructive pulmonary disease and asthma. Am J Respir Crit Care Med. 2009 Dec 1;180(11):1076–82. pmid:19713445
- 28. Zhang H, Balaban OM, Principe JC. Improving pattern recognition of electronic nose data with time-delay neural networks. Sensors and Actuators B. 2003;96:385–9.
- 29. Stolz D, Christ-Crain M, Bingisser R, Leuppi J, Miedinger D, Muller C, et al. Antibiotic treatment of exacerbations of COPD: a randomized, controlled trial comparing procalcitonin-guidance with standard therapy. Chest. 2007 Jan;131(1):9–19. pmid:17218551
- 30. Boixeda R, Rabella N, Sauca G, Delgado M, Martinez-Costa X, Mauri M, et al. Microbiological study of patients hospitalized for acute exacerbation of chronic obstructive pulmonary disease (AE-COPD) and the usefulness of analytical and clinical parameters in its identification (VIRAE study). Int J Chron Obstruct Pulmon Dis. 2012;7:327–35. pmid:22745532
- 31. Shykhon ME, Morgan DW, Dutta R, Hines EL, Gardner JW. Clinical evaluation of the electronic nose in the diagnosis of ear, nose and throat infection: a preliminary study. J Laryngol Otol. 2004 Sep;118(9):706–9. pmid:15509368
- 32. Lai SY, Deffenderfer OF, Hanson W, Phillips MP, Thaler ER. Identification of upper respiratory bacterial pathogens with the electronic nose. Laryngoscope. 2002 Jun;112(6):975–9. pmid:12160294
- 33. Hockstein NG, Thaler ER, Torigian D, Miller WT Jr, Deffenderfer O, Hanson CW. Diagnosis of pneumonia with an electronic nose: correlation of vapor signature with chest computed tomography scan findings. Laryngoscope. 2004 Oct;114(10):1701–5. pmid:15454757
- 34. Corhay JL, Moermans C, Henket M, Nguyen Dang D, Duysinx B, Louis R. Increased of exhaled breath condensate neutrophil chemotaxis in acute exacerbation of COPD. Respir Res. 2014;15:115. pmid:25260953
- 35. Gessner C, Scheibe R, Wotzel M, Hammerschmidt S, Kuhn H, Engelmann L, et al. Exhaled breath condensate cytokine patterns in chronic obstructive pulmonary disease. Respir Med. 2005 Oct;99(10):1229–40. pmid:16140223
- 36. Dragonieri S, Annema JT, Schot R, van der Schee MP, Spanevello A, Carratu P, et al. An electronic nose in the discrimination of patients with non-small cell lung cancer and COPD. Lung Cancer. 2009 May;64(2):166–70. pmid:18834643
- 37. Gompertz S, O'Brien C, Bayley DL, Hill SL, Stockley RA. Changes in bronchial inflammation during acute exacerbations of chronic bronchitis. Eur Respir J. 2001 Jun;17(6):1112–9. pmid:11491152
- 38. White AJ, Gompertz S, Bayley DL, Hill SL, O'Brien C, Unsal I, et al. Resolution of bronchial inflammation is related to bacterial eradication following treatment of exacerbations of chronic bronchitis. Thorax. 2003 Aug;58(8):680–5. pmid:12885984
- 39. Huerta A, Crisafulli E, Menendez R, Martinez R, Soler N, Guerrero M, et al. Pneumonic and nonpneumonic exacerbations of COPD: inflammatory response and clinical characteristics. Chest. 2013 Oct;144(4):1134–42. pmid:23828375
- 40. Joensen O, Paff T, Haarman EG, Skovgaard IM, Jensen PO, Bjarnsholt T, et al. Exhaled breath analysis using electronic nose in cystic fibrosis and primary ciliary dyskinesia patients with chronic pulmonary infections. PLoS One. 2014;9(12):e115584. pmid:25542036
- 41. Ram FS, Rodriguez-Roisin R, Granados-Navarrete A, Garcia-Aymerich J, Barnes NC. Antibiotics for exacerbations of chronic obstructive pulmonary disease. Cochrane Database Syst Rev. 2006(2):CD004403. pmid:16625602
- 42. Bathoorn E, Liesker JJ, Postma DS, Koeter GH, van der Toorn M, van der Heide S, et al. Change in inflammation in out-patient COPD patients from stable phase to a subsequent exacerbation. Int J Chron Obstruct Pulmon Dis. 2009;4:101–9. pmid:19436694
- 43. Connors AF Jr, Dawson NV, Thomas C, Harrell FE Jr, Desbiens N, Fulkerson WJ, et al. Outcomes following acute exacerbation of severe chronic obstructive lung disease. The SUPPORT investigators (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments). Am J Respir Crit Care Med. 1996 Oct;154(4 Pt 1):959–67. pmid:8887592
- 44. McWilliams A, Beigi P, Srinidhi A, Lam S, MacAulay C. Sex and Smoking Status Effects on the Early Detection of Early Lung Cancer in High-Risk Smokers using an Electronic Nose. IEEE Trans Biomed Eng. 2015 Mar 11.