Nosocomial infections are significant causes of morbidity, mortality, and added costs in the health care setting (1). The normal systemic response to the injury that leads to intensive care unit (ICU) admission can be immunosuppressive and seems capable of increasing the risk of nosocomial infections (2). Various strategies for prevention of nosocomial infection have been tested. For instance, immunonutrition has been shown to reduce ventilator-associated pneumonia (VAP) after surgery, but such a demonstration has failed in medical ICUs (3). It may be related to the variety of medical ICU patients, suggesting heterogeneity of their immune status. Based on this assumption, two strategies for the assessment of immunonutrition usefulness can be discussed, the first one consists of enrolling large numbers of patients in clinical trials, the second one consists of the development of biomarkers predicting the occurrence of nosocomial infection.
We hypothesized that the NO fraction present in the airways (upper and lower) of critically ill patients under mechanical ventilation could constitute such a biomarker. NO is a broad-spectrum messenger involved in all biological processes including immunity (4) and also has direct antimicrobial activity (5). Some arguments suggest its usefulness for predicting immunosuppression and consequently the risk of nosocomial infection: (a) its fraction has been shown to be decreased in upper and lower respiratory tracts of patients with cystic fibrosis, and its deficiency has experimentally been shown to favor Pseudomonas aeruginosa infection (6, 7); (b) a beneficial effect of NO inhalation on bacterial clearance of P. aeruginosa has been shown in animals (8); (c) immunonutrition using nutriments with l-arginine, the precursor of NO synthesis, has been associated with a reduction of nosocomial infections in surgical patients (3); and (d) we demonstrated that a decrease in exhaled NO constitutes a marker of postaggressive immunosuppression in an experimental model of sepsis (9). Besides this biological plausibility, NO measurement in the respiratory tract can be easily and immediately obtained at the bedside and is already used to assess various respiratory diseases (10). All these arguments emphasize that NO fraction in the respiratory tract may constitute a relevant marker of immunosuppression in critically ill patients. Consequently, the aim of this proof-of-concept study was to evaluate whether the NO fraction in both upper (nasal) and lower (exhaled) respiratory tracts constitutes a predictive marker for nosocomial infection in medical ICU patients under mechanical ventilation.
MATERIALS AND METHODS
Design of the study
In preliminary experiments (Dr Jerome Aboab, unpublished data; 10 patients), we found a positive correlation between ex vivo endotoxin-stimulated IL-6 production by whole blood and nasal NO of the patient, whereas exhaled NO (mean value in a sampling bag) did not correlate to cytokines. This prompted us to make partitioning of exhaled NO in its bronchial and alveolar origins.
This prospective study was conducted in a 20-bed medical ICU during a 6-month period. Our aim was to include patients at high risk for nosocomial infection but who will also survive until occurrence of nosocomial infection.
Inclusion criteria were ages older than 18 years, a recent acute organ injury (<3 days), and mechanical ventilation on admission with an expected duration of mechanical ventilation longer than 3 days; noninclusion criteria were previous immunosuppression (human immunodeficiency virus infection, systemic corticosteroid treatment, history of cancer with chemotherapy or radiotherapy within the past year) and recent surgery or trauma; and exclusion criterion was a follow-up duration (either caused by weaning from mechanical ventilation or death) shorter than 3 days.
Exhaled NO, which was partitioned into its bronchial and alveolar origins, nasal NO, NO end products (nitrite plus nitrate) in urine, and plasma concentrations of IL-6 and IL-10 were measured on day 1 (0-24 h since admission) and on day 3 (48-72 h since admission) because only early predictive biomarkers would be clinically relevant.
Nosocomial infections were recorded for 15 days. We hypothesized that NO fractions measured at both days 1 and 3 could not serve as a risk factor for nosocomial infection acquisition at later time points. The investigators assessing the outcome of interest (i.e., nosocomial infections) were masked to the results of airway NO levels.
The study protocol was approved by the local ethics committee, and written informed consent was given by the patient's next of kin.
The method has previously been described in detail in spontaneously breathing patients (11), and the two-compartment model has also previously been used in mechanically ventilated patients (12). Briefly, a pneumotachometer was connected serially between the Y-piece from the ventilatory circuit and the humidifying antibacterial filter leading to the tracheal tube. A suction catheter was inserted into the tracheal tube via the catheter mount to sample gas for 3 min from the distal orifice of the tracheal tube to the NO-analyzing system during mechanical ventilation. NO was detected with a chemiluminescent analyzer (ENDONO 8000, Seres, Aix en Provence, France). Inhaled air was NO free. The analog signals of NO and flow were digitized and stored for further analysis. The dependence of exhaled NO on exhalation flow rate can be explained by a simple two-compartment model of the lung. Theoretical studies have provided convincing proof that this flow dependence can be attributed to a double origin of NO from the alveoli and from the bronchial tree (11, 13, 14). In this model, NO concentration (ppb), which is stable in alveolar space, is increased by bronchial excretion (nanoliters per minute) during the passage of exhalate through airways and allows the calculation of two flow-independent NO exchange parameters (independent of ventilatory settings): alveolar NO concentration and maximum airway wall NO flux. For instance, in 40 mechanically ventilated preoperative patients, the median alveolar concentration and airway flux were 2.25 ppb (1.15-5.06) and 24 nL/min (19-38), respectively (manuscript submitted).
To measure nasal NO fraction, air was sampled through a nasal prong introduced approximately 2 cm from the aperture of one nostril (10). The nasal NO fraction (ppb) was obtained at a 1.3-L/min sampling flow rate (1 min sampling); nasal NO fraction was computed as the nasal plateau value minus ambient NO. The intrasubject variability was checked in preliminary experiments and was good according to measurements made in intubated and nonintubated patients (10, 15). In 10 mechanically ventilated preoperative patients, the median nasal concentration was 324 ppb (246-511) (manuscript submitted).
Urine NO end products
Urine samples were heated at 90°C in an acidic environment (HCl 6N) in the presence of vanadium (III) chloride to catalyze the reduction of NO3− to NO2− and then to NO. The NO fraction was measured with the chemiluminescent analyzer, as previously described (16).
Cytokine concentration in plasma
IL-6 and IL-10 concentrations were measured in plasma using specific enzyme-linked immunosorbent assays according to the manufacturer's instructions (Immunotech, Paris, France).
Criteria for nosocomial infections (at least 48 h after admission)
Nosocomial pneumonia criteria were new and persistent lung infiltrates on chest roentgenograms, macroscopically purulent tracheal secretions, and positivity of a quantitative protected specimen brush culture, defined as at least one microorganism recovered at a significant concentration (>103 colony forming units [cfu]/mL), and/or positivity of bronchoalveolar lavage fluid with a significant threshold of >104 cfu/mL (17). Early-onset pneumonia and late-onset pneumonia were defined according to the day of occurrence since admission (early, second - fourth day; late, longer than fifth day). Ventilator-associated pneumonia was defined by a nosocomial pneumonia acquired while mechanically ventilated. In patients with bacteremia, diagnosis of central venous catheter-related infection was confirmed by a positive quantitative tip culture with a significant threshold of 103 cfu/mL. Diagnosis of primary bacteremia was confirmed by at least one positive blood culture (two or more blood cultures when coagulase-negative staphylococci were isolated) without another site simultaneously infected with the same microorganism. A urinary tract infection was defined by association of the two following criteria: leukocyturia (>10 white blood cells/μL) and a urine culture of 105 cfu/mL in patients with clinical signs of infection. Sinusitis criteria were fever and/or purulent nasal secretions, radiological opacification of the maxillary sinuses, and a positive microbiological culture (sinus puncture).
On admission to the ICU, the following parameters were recorded: age, sex, body mass index, indication for mechanical ventilation, and severity of illness evaluated by the Simplified Acute Physiology Score (SAPS) II (18). The Sequential Organ Failure Assessment (SOFA) was calculated on day 1 and day 3 (19). Length of mechanical ventilation, length of ICU stay, and ICU outcome were recorded.
An independent investigator (L.T.) was responsible for data collection and analysis in the clinical research unit.
The aim of the study was to compare NO fractions in patients who will subsequently acquire a nosocomial infection with those who will not; we decided to enroll at least 12 patients with nosocomial infection (based on preliminary experiments in 10 patients). To this end, 40 patients were needed with a nosocomial infection rate of 30%.
Quantitative variables were expressed as median (25th-75th percentiles) because NO parameters are not normally distributed. Comparison of quantitative variables used Mann-Whitney U test, and comparison of categorical variables used chi-square test (Fisher correction when necessary). Nasal and bronchial NO fractions at days 1 and 3 were compared using Wilcoxon test. Relationships between quantitative variables were evaluated using Spearman rank correlation coefficient. The sensitivity and specificity of possible cutoff points for nasal NO in discriminating between subjects with nosocomial infection were determined with receiver-operator characteristic curves. A two-tailed value of P < 0.05 was considered statistically significant. Statistical analyses were performed using MedCalc software (Mariakerke, Belgium).
Sixty consecutive patients were prospectively included, of whom 45 were followed up for more than 3 days and constituted the studied population (Fig. 1). Indications for admission were cardiovascular (cardiac arrest, n = 5; cardiac insufficiency, n = 8; septic shock, n = 3), respiratory (acute exacerbation of chronic obstructive pulmonary disease [COPD], n = 6; pneumonia, n = 5), neurological (stroke, n = 10; epilepsy, n = 5), and metabolic (acute renal failure, n = 1; miscellaneous, n = 2). Fourteen patients were infected on admission. Their other characteristics are described in Table 1, accordingly to the occurrence of a nosocomial infection in the first 15 days. NO parameters were similar whatever the indication for admission or the presence of infection on admission; specifically, the patients admitted for a respiratory reason (all with suspected infection) did not have higher values for both nasal and bronchial NO.
A significant increase in both nasal NO (P = 0.003) and bronchial NO (P = 0.005) was evidenced between days 1 and 3. Nasal and bronchial NO did not significantly correlate at both days 1 and 3.
Among the 45 patients, 15 acquired at least one nosocomial infection within 15 days of ICU stay (16 infections, one patient had two episodes of VAP); the median time to occurrence of nosocomial infection was 4.0 days (3.0-7.7), all but one (an episode of bacteremia at day 3) were VAP (7 early-onset VAP).
Markers for nosocomial infection acquisition
Nasal NO was the only marker significantly different between these two subgroups. The predictive values of the different markers for nosocomial infection acquisition are shown in Table 2.
Bronchial NO was not different between the 31 patients who did not acquire VAP as compared with the 14 patients who had VAP whatever the time point (day 1, P = 0.91; day 3, P = 0.32).
Markers for ICU survival
Patients who did not survive had higher SAPS II scores (P = 0.030), higher SOFA scores on day 3 (P = 0.006), and higher IL-6 concentrations on day 3 (P = 0.014), whereas their nasal NO concentrations were not significantly different both on days 1 and 3. Nonsurvivors tended to acquire more nosocomial infections (P = 0.057).
Urine NO end products were not significantly different between survivors and nonsurvivors or between patients with or without nosocomial infection. Nevertheless, urine NO end products at day 1 correlated to both SAPS II and SOFA scores (ρ = −0.35, P = 0.027 and ρ = −0.46, P = 0.004, respectively).
The main result of this preliminary study is to suggest that nasal NO is a quite sensitive and specific biomarker of subsequent (within 15 days) nosocomial infection (mainly VAP) acquisition. The aim of this study was to find a relevant biomarker, available at bedside, which may further help us select medical ICU patients who would benefit from piloted immunonutrition in subsequent trials.
Risk factors for nosocomial infection can be classified as endogenous (caused by acquired immunosuppression) and exogenous (caused by ICU environmental). Endogenous factors bear most of the statistical weight (21), suggesting that ICU-related immunosuppression would constitute a relevant end point for nosocomial infection prevention. Several lines of evidence suggest that NO may constitute a biomarker for nosocomial infection prediction, and we previously established the link between its decrease in exhaled gas and markers of immunosuppression in an animal model (9). This preliminary study was designed to assess whether exhaled and/or nasal NO may constitute relevant biomarkers. The fact that nasal NO was a better predictive marker than bronchial NO is an advantage for future developments of its measurement in the ICU because it is easily measured at the bedside as compared with bronchial NO. We also show that nasal NO decrease is not a risk factor for mortality (specificity for infection).
NO fraction in airways is commonly viewed as an "inflammatory" marker, which would be increased in critically ill patients. Besides its frank increase in eosinophilic (atopic) inflammation (22), its increase in bacterial or viral infections of the respiratory tract is modest and even controversial, for instance in COPD patients (23, 24). Our patients who were hospitalized for a respiratory reason (exacerbation of COPD and pneumonia, with infection on admission) had similar levels of both nasal and bronchial NO, which could be related to the small size of the sample. Nevertheless, Adrie and colleagues (15) demonstrated that mechanically ventilated patients with pneumonia (mainly caused by aspiration) had higher levels of both exhaled and nasal NO; but they also stated that the global baseline values of exhaled NO were markedly lower than that in some recent reports. Other investigators found a decrease in exhaled NO in settings of pulmonary inflammation such as ARDS or cardiopulmonary bypass (25, 26). Nasal NO has also been shown to be reduced in sinus epithelium of critically ill patients with radiological maxillary sinusitis and sepsis (27). Even urine NO end products have been shown to decrease in the most severe patients with ARDS (nonsurvivors) (28). Overall, a generalized impairment of NO production can be suspected, which may contribute to depressed immunity because NO has a broad-spectrum antimicrobial activity (5). There is no definite explanation for the decreased NO production; we previously suggested that it could result from an increased arginase activity, which uses the same precursor, l-arginine (9, 16). Deja and colleagues (27) showed that in critically ill patients with sinusitis and sepsis, maxillary NO production is almost completely suppressed by downregulation of iNOS mRNA. The specific involvement of one NO synthase may explain why we do not show a parallel involvement of NO production of the different sites (nasal, bronchial, alveolar, and urine) because these anatomical sites can be differently affected by diseases and treatments (7, 29, 30).
This study has limitations. These results are preliminary, but this trial was designed as a proof-of-concept study, aimed at defining whether nasal or exhaled NO would be selected in a larger trial. Whether NO levels could be increased in the respiratory tract at the early phase of sepsis was beyond the scope of our study. A larger trial is needed to assess whether the predictive value of nasal NO is similar in infected versus noninfected patients on admission. This study mainly highlights the link between nasal NO decrease and VAP acquisition; it remains to evaluate whether the decrease in nasal NO also predicts other nosocomial infections. Chemiluminescence analyzers are expensive, but more recent analyzers have been developed (electrochemical devices), which are less expensive and available at the bedside. Finally, one may ask whether nosocomial sinusitis may constitute a bias for nasal NO measurement. The early time points of nasal NO measurements (days 1 and 3) make unlikely the presence of nosocomial sinusitis.
In conclusion, this preliminary proof-of-concept study suggests that nasal NO measurement may constitute a relevant biomarker for prediction of nosocomial infection (at least VAP), which warrants confirmation in a larger multicenter trial.
The authors thank Dr Jerome Aboab for having performed preliminary experiments in the medical ICU of Henri Mondor Hospital and Mr Xavier Gautier for cytokine determinations.
1. Safdar N, Dezfulian C, Collard HR, Saint S: Clinical and economic consequences of ventilator-associated pneumonia: a systematic review. Crit Care Med
2. Munford RS, Pugin J: Normal responses to injury prevent systemic inflammation and can be immunosuppressive. Am J Respir Crit Care Med
3. Heyland DK, Novak F, Drover JW, Jain M, Su X, Suchner U: Should immunonutrition become routine in critically ill patients? A systematic review of the evidence. JAMA
4. Bogdan C: Nitric oxide and the immune response. Nat Immunol
5. Fang FC: Perspectives series: host/pathogen interactions. Mechanisms of nitric oxide-related antimicrobial activity. J Clin Invest
6. Kelley TJ, Drumm ML: Inducible nitric oxide synthase expression is reduced in cystic fibrosis murine and human airway epithelial cells. J Clin Invest
7. Hubert D, Aubourg F, Fauroux B, Trinquart L, Sermet I, Lenoir G, Clement A, Dinh-Xuan AT, Louis B, Mahut B, et al: Exhaled nitric oxide
in cystic fibrosis: relationships with airway and lung vascular impairments. Eur Respir J
8. Jean D, Maitre B, Tankovic J, Meignan M, Adnot S, Brun-Buisson C, Harf A, Delclaux C: Beneficial effects of nitric oxide inhalation on pulmonary bacterial clearance. Crit Care Med
9. Attalah HL, Honore S, Eddahibi S, Marcos E, Soussy CJ, Adnot S, Delclaux C: Decreased exhaled nitric oxide
as a marker of postinsult immune paralysis. J Appl Physiol
10. American Thoracic Society, European Respiratory Society: ATS/ERS recommendations for standardized procedures for the online and offline measurement of exhaled lower respiratory nitric oxide and nasal nitric oxide, 2005. Am J Respir Crit Care Med
11. Mahut B, Louis B, Zerah-Lancner F, Delclaux C: Validity criteria and comparison of analytical methods of flow-independent exhaled NO parameters. Respir Physiol Neurobiol
12. Tornberg DC, Bjorne H, Lundberg JO, Weitzberg E: Multiple single-breath measurements of nitric oxide in the intubated patient. Am J Respir Crit Care Med
13. Delclaux C, Mahut B, Zerah-Lancner F, Delacourt C, Laoud S, Cherqui D, Duvoux C, Mallat A, Harf A: Increased nitric oxide output from alveolar origin during liver cirrhosis versus bronchial source during asthma. Am J Respir Crit Care Med
14. George SC, Hogman M, Permutt S, Silkoff PE: Modeling pulmonary nitric oxide exchange. J Appl Physiol
15. Adrie C, Monchi M, Dinh-Xuan AT, Dall'Ava-Santucci J, Dhainaut JF, Pinsky MR: Exhaled and nasal nitric oxide as a marker of pneumonia in ventilated patients. Am J Respir Crit Care Med
16. Bartolucci P, Ngo MT, Beuzard Y, Galacteros F, Saber G, Rideau D, Eddahibi S, Maitre B, Adnot S, Delclaux C: Decrease in lung nitric oxide production after peritonitis in mice with sickle cell disease. Crit Care Med
17. Chastre J, Fagon JY: Ventilator-associated pneumonia. Am J Respir Crit Care Med
18. Le Gall JR, Neumann A, Hemery F, Bleriot JP, Fulgencio JP, Garrigues B, Gouzes C, Lepage E, Moine P, Villers D: Mortality prediction using SAPS II: an update for French intensive care units. Crit Care
19. Ferreira FL, Bota DP, Bross A, Melot C, Vincent JL: Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA
20. DeLong ER, DeLong DM, Clarke-Pearson DL: Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics
21. Kollef MH: Ventilator-associated pneumonia. A multivariate analysis. JAMA
22. Mahut B, Peiffer C, Thibaudon M, Chevalier-Bidaud B, Defrance-Hutinet MF, Trinquart L, Delclaux C: What does a single exhaled nitric oxide
measurement tell us in asthmatic children? J Asthma
23. Taylor DR, Pijnenburg MW, Smith AD, De Jongste JC: Exhaled nitric oxide
measurements: clinical application and interpretation. Thorax
24. Bhowmik A, Seemungal TA, Donaldson GC, Wedzicha JA: Effects of exacerbations and seasonality on exhaled nitric oxide
in COPD. Eur Respir J
25. Beghetti M, Silkoff PE, Caramori M, Holtby HM, Slutsky AS, Adatia I: Decreased exhaled nitric oxide
may be a marker of cardiopulmonary bypass-induced injury. Ann Thorac Surg
26. Brett SJ, Evans TW: Measurement of endogenous nitric oxide in the lungs of patients with the acute respiratory distress syndrome. Am J Respir Crit Care Med
27. Deja M, Busch T, Bachmann S, Riskowski K, Campean V, Wiedmann B, Schwabe M, Hell B, Pfeilschifter J, Falke KJ, et al: Reduced nitric oxide in sinus epithelium of patients with radiologic maxillary sinusitis and sepsis. Am J Respir Crit Care Med
28. McClintock DE, Ware LB, Eisner MD, Wickersham N, Thompson BT, Matthay MA: Higher urine nitric oxide is associated with improved outcomes in patients with acute lung injury. Am J Respir Crit Care Med
29. Degano B, Genestal M, Serrano E, Rami J, Arnal JF: Effect of treatment on maxillary sinus and nasal nitric oxide concentrations in patients with nosocomial maxillary sinusitis. Chest
30. Mahut B, Escudier E, de Blic J, Zerah-Lancner F, Coste A, Harf A, Delclaux C: Impairment of nitric oxide output of conducting airways in primary ciliary dyskinesia. Pediatr Pulmonol
Exhaled nitric oxide; mechanical ventilation; critically ill patient