Published ahead of print on August 19, 2003, doi:10.1164/rccm.200306-784OC
© 2003 American Thoracic Society Multiple Single-breath Measurements of Nitric Oxide in the Intubated PatientDepartment of Anesthesiology and Intensive Care, Karolinska Hospital, and Department of Physiology and Pharmacology, Karolinska Institute, Stockholm, Sweden Correspondence and requests for reprints should be addressed to Daniel C. Törnberg, M.D., Department of Anesthesiology and Intensive Care, Karolinska Hospital, S-171 76 Stockholm, Sweden. E-mail: danieltornberg{at}hotmail.com
Multiple flow rate measurements of exhaled nitric oxide (NO) have been advocated to fractionate NO from alveolar and bronchial sources. The aim of this study was to develop a method by which multiple single-breath exhalations at various flow rates could be performed in intubated, mechanically ventilated patients. Nine patients without lung disease were studied awake and after intubation, during general anesthesia. A suction ejection system connected to a restrictor valve was used to control the exhalation flow rate. From these measurements the fraction of alveolar NO (FANO), the fraction of airway wall NO (FawNO), and the airway wall transfer rate (DNO) were calculated. The fraction of exhaled NO was reduced by 50% after intubation. DNO was also reduced by intubation (from 10 ± 1.3 to 6.4 ± 2.1 nl second-1 ppb-1 x 10-3) whereas neither FawNO nor FANO was affected. The peak NO concentration after 20 seconds of apnea during general anesthesia was similar to calculated FawNO. The vacuum aspiration method used in this study allowed for reproducible multiple single-breath measurements and calculation of alveolar and bronchial NO parameters. Further studies will reveal whether this methodology will improve the value of exhaled NO analysis in intubated, mechanically ventilated patients with pulmonary disease.
Key Words: exhaled • lung function • mechanical ventilation • pulmonary Exhaled nitric oxide (NO) is gaining considerable scientific interest as a marker of airway inflammatory disease in patients with, for example, asthma, chronic obstructive pulmonary disease, and cystic fibrosis (1–3). Recommendations for single-breath analysis of fraction of exhaled NO (FENO) have been published by task force groups within both the European Respiratory Society (4) and the American Thoracic Society (5) and the first diagnostic instrument was approved by the U.S. Food and Drug Administration for clinical use in individuals with asthma. FENO is highly flow dependent and knowledge of flow during NO sampling is therefore crucial for comparison of data (6). For this reason, single-breath exhalations with controlled flow represent the most widely used method in awake, cooperative patients. To further extend NO analysis a theoretical two-compartment model of the airway, with one distal alveolar component and one additive connecting bronchial component, has been suggested (7, 8). In this model exhaled NO is the sum of two origins: NO of alveolar origin (fraction alveolar NO [FANO]) and NO from the airway wall epithelium (fraction of airway wall NO [FawNO]). The addition of airway wall NO to alveolar NO is affected by the airway wall transfer rate [DNO]. DNO and NOflux (the product of FawNO and DNO) are dependent on the NO-releasing area within the airways. By the use of multiple single-breath exhalations at various flow rates it has been possible to calculate these flow-independent parameters (FANO, FawNO, DNO, and NOflux) in awake, cooperating patients (9–11). The increase in FENO in asthma consists of increased DNO and increased airway wall NO, where the latter is reduced by steroid treatment (9). Increased FANO has been described in patients with alveolitis (11). From studies of tracheotomized awake patients it is known that oral FENO is higher than tracheal FENO (12, 13) and that this difference can be reduced by an oral antibacterial mouthwash (14). Whether the addition of NO during oral exhalation will affect calculation of the flow-independent parameters has not been elucidated. Few studies have been performed in mechanically ventilated patients and none with multiple flow rate analysis. The preferred method has been to measure NO concentration on-line during tidal breathing. FENO is increased in ventilator-associated pneumonia in the intensive care unit (15), in lung transplant recipients affected by bacterial airway infection or bronchiolitis obliterans syndrome (16, 17), but is decreased during acute respiratory distress syndrome (ARDS) (18). In patients subjected to cardiopulmonary bypass unchanged or decreased levels of FENO have been described (19–21). In accordance with studies of awake individuals we wanted to use multiple single-breath measurements to calculate flow-independent parameters in the intubated, noncooperative patient. Therefore, we developed a method using preset constant expiratory flow rates, which permitted measurements both awake preoperatively and after intubation during general anesthesia in the same subject. Also, NO concentration and output during tidal breathing with mechanical ventilation were compared with the flow-independent parameters and analyzed with different levels of positive end-expiratory pressure (PEEP).
The study was approved by the local ethics committee at the Karolinska Institute (Stockholm, Sweden). Nine nonsmoking female patients scheduled for laparoscopic gynecologic surgery (mean age, 39 years; range, 21–56 years), without pulmonary disease, were included after informed consent had been obtained. In each patient measurements were made immediately before and during general anesthesia. NO analysis was made by the chemiluminescence technique. A fraction of the exhaled air was sampled into an NO analyzer at a flow rate of 6 ml second–1 (model 77 AM; Eco Physics, Dürnten, Switzerland). The analyzer was calibrated with NO-free air and NO gas, which was diluted 15 times in the analyzer (10 ppm; AGA Linde, Lidingö, Sweden) before each experiment.
NO Measurements in the Awake Patient
NO Measurements during General Anesthesia Total intravenous anesthesia was induced with propofol at 2–3 mg kg-1, alfentanil at 1 mg, and rocuronium at 0.6 mg kg-1 and then maintained with propofol at 8–10 mg kg-1 hour-1.
Single-breath Measurements The measured expiratory flow levels during vacuum aspiration were 6 ± 1, 24 ± 2, 57 ± 2, 109 ± 2, 195 ± 5, and 312 ± 7 ml second–1. A mean value of two or three consecutive single-breath exhalations at each flow rate was used.
Apnea
Tidal Breathing During these measurements the NO scrubber was connected between the hospital compressed air system and the ventilator (S5 anesthesia delivery unit [S5/ADU]; Instrumentarium, Datex-Ohmeda Division, Bromma, Sweden). The pneumotachometer was connected to the Y-piece of the respiratory tubings. NO was continuously measured via the catheter at the distal end of the endotracheal tube. NO mean concentration (ppb) over 30 seconds, NO peak concentration (ppb, mean of peaks during 30 seconds), and NO output (nl second-1) were calculated. NO output was measured at PEEP levels of 3 and 10 cm H2O, 30 seconds after alterations in PEEP level.
Nasal NO Measurements
Calculations
 is flow.
Nonlinear Regression
Slope–Intercept
>> DNO (/DNO > 10) the linear approximation of the exponential function is 1 - DNO/ and from Equations 1 and 2:
To calculate FANO we used the flows equal to or higher than 100 ml second-1. The slope is FANO and the intercept with the y axis is NOflux, which equals (FawNO - FANO) x DNO.
NO Output during Tidal Breathing
Statistics
Single-breath Measurements There was no significant difference between FENO50awake (15.6 ± 1.3 ppb, vacuum aspiration system) and conventional FENO50 (16.2 ± 2.2 ppb, biofeedback method). At five of six flow rates during single-breath exhalations, FENO were significantly lower after intubation (FENO awake: 74 ± 9, 34 ± 5, 16 ± 2, 9 ± 1, 5 ± 0.5, and 4 ± 0.5 ppb versus FENO under general anesthesia: 42 ± 7, 16 ± 4, 7 ± 2, 4 ± 1, 2 ± 0.4, and 2 ± 0.4 ppb for 6, 20, 50, 100, 200, and 300 ml second-1, respectively) (p = 0.07 for 20 ml second-1; see Figure 2) . The relative differences (FENO under general anesthesia/FENO awake) were 0.43, 0.47, 0.55, 0.55, 0.43, and 0.54, respectively.
Flow-independent Parameters There were no significant differences between the data obtained by the two different methods for calculation of the flow-independent parameters (Table 1) .
Flow-independent parameters calculated by nonlinear regression and least-squares fit are shown in Table 1 and Figure 3 . There was no significant difference in FawNO between measurements before (98 ± 14 ppb) and after intubation (128 ± 45 ppb) and the correlation was significant (r2 = 0.53, p = 0.02).
There was no significant difference in alveolar NO concentration during general anesthesia compared with the awake state (FANOga: 1.05 ± 0.27 ppb versus FANOawake: 1.58 ± 0.35 ppb; p = 0.09). DNOga was significantly reduced compared with DNOawake (DNOga: 6.4 ± 2.1 nl second-1 ppb-1 x 10-3 versus DNOawake: 10 ± 1.3 nl second-1 ppb-1 x 10-3; p < 0.05). Also, a decrease in NOflux during general anesthesia was observed (NOfluxga: 0.378 ± 0.061 nl second-1 versus NOfluxawake: 0.874 ± 0.084 nl second-1) (p = 0.012; see Figure 3). The calculated slope–intercept values revealed no difference in FANO between awake and intubated states but a significantly higher NOflux in the awake state (Table 1). After 20 seconds of apnea the peak NO concentration was 93 ± 12 ppb. These values did not differ significantly from FawNO but no correlation between apnea peak NO levels and the least-squares fit-calculated FawNO was evident.
Tidal Breathing Peak NO concentration did not correlate to any of the flow-independent parameters. However, NO output correlated significantly with two of the flow-independent parameters from nonlinear regression: NOfluxga (r2 = 0.76, p = 0.005) and FANOga (r2 = 0.65, p = 0.016); whereas no correlation was observed to FawNOga (p > 0.05; r2 = 0.11, p = 0.41).
Effects of PEEP
Nasal Measurements
We present here a novel method for multiple single-breath measurements in mechanically ventilated patients by which we were able to calculate NO from different compartments in the human airways. Furthermore, by using the same method both in the awake state and during general anesthesia, we were able to compare the influence of intubation on the flow-independent parameters. For every flow rate the absolute levels of FENO were about 50% lower during anesthesia than in the awake state. Among the flow-independent parameters DNO and NOflux decreased on intubation whereas FawNO and FANO were unchanged. NO output during tidal breathing correlated to both NOflux and FANO. NO output increased with increasing PEEP. The rationale for calculating flow-independent parameters is that these would extend the diagnostic value of exhaled NO as an inflammatory marker. Indeed, results from different studies show some efficiency of these parameters in discrimination between different pulmonary disorders (9–11). It is reasonable to believe that this extension of FENO analysis would be of use also in diseases common in the intensive care unit such as pneumonia, acute lung injury, and ARDS. However, to our knowledge, multiple single-breath exhalations to calculate flow-independent parameters in mechanically ventilated patients have not been studied previously. The aspiration method used here to control expiration was relatively easy to perform and gave reproducible exhalation flows within each patient. A slight increase in flow rate from the preset level was noted on exhalation, and was probably due to the increased pressure difference across the restrictor valve in the vacuum aspiration system. However, the actual flow rate was measured and used in the calculations. Compared with the conventional biofeedback method often used in cooperating patients, the two methods gave similar FENO50 values but the variation between exhalations within each patient was actually smaller with the vacuum aspiration system (data not shown). At the higher flows (200 and 300 ml second–1), during mechanical ventilation, the exhalation time was short but long enough to obtain a detectable NO level. Taken together, this method is feasible to evaluate flow-independent parameters in mechanically ventilated patients as well as to study the same patient when awake. However, in spontaneously breathing, intubated patients this method may be more difficult to perform, because an undisturbed exhalation could be difficult to achieve. In the present study all patients were given a muscle relaxant. Furthermore, in patients with severe pulmonary dysfunction, for example, ARDS, it is questionable whether disconnection of the respiratory tubings to perform single-breath measurements may be performed without risking alveolar collapse and deteriorating oxygenation. In most cases, however, this method could be used and should preferably be followed by a lung recruitment maneuver. The flow-independent parameters (FANO, FawNO, DNO, and NOflux) were calculated by two different methods according to suggestions in the literature (7, 9). When using nonlinear regression there were difficulties in reaching convergence when all six flow rates were used. Therefore only three (and in two patients four) flows were used for calculations. From a practical point of view this is desirable in the clinical setting but it may weaken parameter estimation. However, we found similar values for FANO and NOflux with the slope–intercept method. Furthermore, the flow-independent parameters we obtained were in the same range as presented previously by others (9, 10). To further evaluate our data, all six flow rates were used in a nonlinear regression, two-parameter model, fixing FANO at zero. The FawNO and DNO values obtained with this model were similar to those obtained with three or four flows (data not shown). Taken together, three or four flow rates seem to be sufficient to estimate the flow independent parameters in the awake and intubated subject. Theoretically a long apnea would lead to an equilibration between FawNO and airway lumen NO. Interestingly, 20 seconds of apnea gave similar NO values as calculated FawNO. However, there was no correlation between NO peak values after 20 seconds of apnea and FawNO. The lack of correlation could be due to insufficient time to reach equilibrium. We chose 20 seconds of apnea based on the findings of Dweik and co-workers, who obtained a plateau in intratracheal NO levels during breath-hold within this period of time (23). However, it is possible that a longer apnea would have produced values even more in match with FawNO. The twice-as-high FENO values found in the awake state compared with general anesthesia are in line with previous findings from our group, when tracheotomized awake patients were exhaling through the mouth or the tracheal cannula (13). This speaks against the anesthetic drugs being responsible for the reduction in FENO after intubation. Furthermore, in vitro results show that propofol may actually stimulate the NO pathway in rat tracheal epithelial cells (24). Moreover, nasal NO did not decrease during anesthesia. The exact source of the NO adding to the exhaled levels in the awake state is not known. There is a substantial contribution of NO from the oral cavity, where a stepwise reduction of salivary nitrate to nitric oxide occurs. These reactions are dependent on bacterial nitrate reductases and the acidic environment in the crypts of the tongue (25, 26). The oral addition of NO can be reduced by antibacterial mouthwash (14). Other sources of NO may also contribute, including inducible NO synthase, which seems to be present in the oral mucosa (K. Alving, personal communication). In the present study the entire NO-producing area above the cuff of the endotracheal tube was blocked from contributing to the exhaled air. Judging from the addition of NO in the awake state, this area corresponds to about 50% of the total NO-producing area in the airways. This was evident by a reduction in DNO of a similar magnitude. However, because DNO is proportional to both the lumenal surface area and a mucosal transfer coefficient (9) we cannot exclude that the intubation per se affected mucosal transfer of NO. FawNO did not differ between the awake state and after intubation. This finding suggests that the factors determining FawNO are similar in the upper trachea and oral compartments as in the rest of the bronchial system, which supports the use of a two-compartment model in awake, orally exhaling subjects. A consequence of the unchanged FawNO and reduced DNO was a reduction of NOflux. According to the two-compartment model FANO should not be affected by reducing the NO-producing area. We found no significant reduction in FANO after intubation. Measuring exhaled NO online during tidal breathing has been the preferred method in the relatively few studies performed on mechanically ventilated patients. Although this method has several practical advantages it may not be the optimal method to reveal differences in FENO in various diseases. Tidal breathing FENO is measured during unstable expiratory flow conditions and is therefore sensitive to changes in tidal volume and respiratory frequency. In the present study NO output during tidal breathing correlated both to FANO and FawNO, indicating that tidal breath NO analysis could not identify the source of NO. NO gradually increases during exhalation and finally reaches a peak before next inhalation. The peak NO level in each exhalation during tidal breathing coincides with the lowest flow. This fact would suggest that peak NO should, at least vaguely, reflect or correlate to FawNO. We could not find such a correlation in our study. This could be due to the different sampling response times for NO and flow measurements and we cannot exclude that the true NO peak remained undetected because of the response time of the analyzer. Taken together, NO measurements during tidal breathing, although practical, may miss important information regarding NO release from the lungs. Nevertheless, some studies have shown altered FENO levels during tidal breathing measurements in ventilator-associated pneumonia (15), ARDS (18), and after cardiopulmonary bypass during coronary artery bypass surgery (20). Alterations in FENO levels during tidal breathing may also be detected after administration of nitroglycerin (21) and endothelium-dependent vasoactive substances such as acetylcholine, substance P, bradykinin, and endothelin (27). In this study, NO output increased with increasing PEEP. This has previously been shown in rabbits and stretch-dependent mechanisms and reduced cardiac output have been postulated to cause these effects (28, 29). Pulmonary recruitment effects with reduced atelectasis or altered resistance and compliance may also explain the increase in NO output by PEEP. These facts again emphasize the importance of ventilator settings when analyzing FENO during tidal breathing. We measured nasal NO in this study to evaluate any effect of anesthesia on airway NO production in general. Nasal NO is dependent on NO synthase activity in the nasal cavity and paranasal sinuses (30). Interestingly, nasal NO actually increased after intubation during general anesthesia, which does not support reduced NO synthase activity caused by the administered anesthetics. On the other hand, the unchanged FawNO does not suggest a general increase in airway NO production during anesthesia. The lower levels of nasal NO measured before anesthesia are probably due to phonation and ventilation of the nasal cavity, which will also enhance paranasal sinus ventilation (31, 32). General anesthesia with endotracheal intubation excludes phonation and voluntary nasal ventilation, which might lead to a greater accumulation of NO in the nose and sinuses. The possible use of extended NO analysis to evaluate where in the respiratory system NO is generated is intriguing. We conclude that flow-independent parameters can be calculated from controlled single-breath exhalations in mechanically ventilated patients. Compared with the awake situation, some of these parameters are affected by intubation, mainly because of a reduction in NO-producing area in connection with expiratory flow. NO concentration obtained during tidal breathing does not correlate to any of the flow-independent parameters and must be considered a blunter tool in evaluating exhaled nitric oxide in mechanically ventilated patients. Further studies will reveal whether the flow-independent parameters will improve NO analysis in the intensive care unit.
The authors thank Prof. Kjell Alving for valuable expert advice, and Kerstin Pregner, R.N. and Kia Rosberg, R.N. for technical assistance.
Supported by grants from the Swedish Heart Lung Foundation, the Swedish Research Council, and the AGA-Linde Research Fund, and by funds from the Karolinska Institute. Conflict of Interest Statement: D.C.T. has no declared conflict of interest; H.B. has no declared conflict of interest; J.O.L. owns shares in Aericrine AB, a Swedish biotech company, and in 2002 received a grant from AstraZeneca of $30,000 for research related to intestinal inflammation; E.W. has patents and applications regarding the use of NO measurements on diagnosis, and is co-founder and member of the board of Aerocrine AB in Sweden, owns 85,000 shares, and is not employed by the company but participates in R&D discussions. Received in original form June 13, 2003; accepted in final form August 18, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
