Effect of the Physical Properties of Isoflurane, Sevoflurane, and Desflurane on Pulmonary Resistance in a Laboratory Lung Model

The setup of the experiment is presented in figure 1. A two-chamber test lung with fixed resistance (5600i Dual Adult System; Michigan Instruments, Grand Rapids, MI) was connected to an anesthesia machine (Julian; Draëger Medical, Lübeck, Germany) using the volume control mode of ventilation. The test lung served as a quantitative, calibrated “test load” simulating human pulmonary physiology. The elastance and resistance of the experimental chambers were set at 20 cm H²O · l⁻¹and 15 cm H²O · l⁻¹· s⁻¹, respectively.5 

A screen pneumotachograph, which used a differential pressure–based flow sensor and was used for the measurement of flow rate (RSS100-HR; Hans Rudolph, Kansas City, MO), and a pressure transducer were inserted between the endotracheal tube and the Y piece of the respiratory circuit for the continuous recordings of measured flow, tidal volume, and inspiratory pressures (P^peakand P^plateau). This particular pneumotachograph provides correction for gas density, viscosity, temperature, and barometric pressure. Moreover, to confirm accuracy, the flows measured at baseline and after the addition of anesthetic gases were compared. Because the lung model was ventilated with volume control mode, it was mandatory that flow remain constant, and only minor breath-to-breath variations should have been recorded. The pulmonary resistance (R^plm) was calculated by the inspiratory pressure method,6where the driving pressure for airflow is the difference between end-inspiratory peak airway pressure and plateau pressure (P^peak– P^plateau) during occlusion of the expiratory port. The P^peakand P^plateaupressures were measured from the waveform of the inspiratory pressure. Flow was measured from the waveform of the pneumotachograph (fig. 2). The overall pulmonary resistance (R^plm) was calculated from the equation: pulmonary resistance R^plm= (P^peak– P^plateau)/(flow) at baseline, when a mixture of 25% oxygen in air was used (without volatile agent) and also at mixtures of 1.0, 1.5, and 2.0 MAC of isoflurane, sevoflurane, and desflurane with 25% oxygen in air. Finally, to discriminate between the effects of the concentration of the anesthetics or the effects of their physical properties, we also measured the resistance at the same low (1.2%) and high (4%) concentrations of isoflurane sevoflurane, and desflurane with 25% oxygen in air. These particular concentrations were chosen because 1.2% was 1 MAC of isoflurane, and 4% is the highest concentration that could be delivered from the isoflurane vaporizer. For each concentration of the volatile agent, five measurements were recorded.

The parameters of the volume control mode of ventilation were the same for all measurements: tidal volume = 600 ml, respiratory rate = 8 breaths/min, positive end-expiratory pressure = 0 cm H²O, T^I:T^E= 1:1.5, T^I:T^plateau= 50%, and fresh gas flow = 8 l/min. The MAC values of each anesthetic agent were calculated by the Agent Analyser (IRIA; Draëger Medical, Lübeck, Germany) of the anesthetic apparatus using the volatile agent end-expiratory concentrations and corrected for age (at a set age of 40 yr), altitude, and gas mixture. IRIA measures the concentration of the anesthetics based on infrared light absorption. The absorption of infrared light at 3 μm is used for the measurement of carbon dioxide–nitrous oxide and at 8 μm for the measurement of end-expiratory volatile anesthetics. Age-corrected MAC values were then calculated according to the Mapleson formula for patients older than 1 yr: MAC^age= MAC⁴⁰× 10 [−0.00269 × (age − 40)].7 

All values represent the means from a total of five consecutive respiratory cycles. The increase of pulmonary resistance caused by the volatile agents was also expressed as a percentage of the baseline value. Data were analyzed by analysis of variance for repeated measurements, and post hoc  Tukey and least significant difference tests were  used. All values are expressed as mean ± SD.

The measured end-expiratory concentrations for isoflurane, sevoflurane, and desflurane at the three studied MAC values are shown in table 1. Desflurane was used at greater end-expiratory concentrations at all MAC values (6.6, 9.7, and 13.3% representing 1, 1.5, and 2 MAC, respectively), because it is the least potent of the three volatile agents.

The flow measurements for all gas compositions are shown in table 2. None of the differences recorded were significant. This finding confirmed that the pneumotachograph measures were valid because, with the use of volume control ventilation mode, inspiratory flow was mandatory to remain constant, and only minor breath-to-breath variations were acceptable. Otherwise, the pneumotachograph gave false values because of the altered physical properties of the gas mixtures.

The baseline value of pulmonary resistance of the lung model at 25% oxygen in air was 15.64 ± 0.15 cm H²O · l⁻¹· s⁻¹. The effects of the different concentrations of isoflurane, sevoflurane, and desflurane in 25% oxygen in air on pulmonary resistance are shown in table 3and figure 3.

Compared with the baseline value, isoflurane significantly increased pulmonary resistance only at 2 MAC (P = 0.005). Sevoflurane increased pulmonary resistance at 1.5 and at 2 MAC by 17.8% and 23.3%, respectively (P < 0.001 for both comparisons). Desflurane increased the calculated resistance of the test lung model at all concentrations used (P < 0.001 for all concentrations). At 1 MAC, where the end-expiratory concentration of desflurane was 6.6%, the increase of pulmonary resistance from the baseline value was 44.5%. At 1.5 MAC desflurane, an end-expiratory concentration of 9.7% was used, and the increase of pulmonary resistance was 63.7% from the baseline value. The greatest increase of pulmonary resistance, by 87%, was observed at 2 MAC when an end-expiratory concentration of 13.3% was used.

When comparing the effect of the three volatile agents on the pulmonary resistance at equivalent concentrations, it was shown that at 1 MAC isoflurane and sevoflurane had similar effects on pulmonary resistance, whereas desflurane caused a significant increase compared with the two other agents (P < 0.001; fig. 4). At 1.5 MAC, sevoflurane caused a significant increase of pulmonary resistance compared with isoflurane (P < 0.001), and desflurane caused the greatest increase compared with the two other agents (P < 0.001 for all comparisons). Sevoflurane and desflurane did increase overall pulmonary resistance by 17.8% and 63.7%, respectively. At 2 MAC, a statistically significant increase of pulmonary resistance was observed with all of the three volatile agents. Sevoflurane had a significantly greater effect compared with isoflurane (P < 0.001), and desflurane caused the greatest increase compared with the two other agents (P < 0.001 for both comparisons; table 3and fig. 4).

There was no significant increase of pulmonary resistance from the baseline when the three anesthetics were delivered at the same low concentration of 1.2%, whereas at 4%, all three produced a significant increase that ranged from 25.8% for desflurane to 28% for isoflurane and 28.9% for sevoflurane (P < 0.001 for the three agents; table 3). The differences between the three anesthetics at the aforementioned concentrations were not significant.

Pulmonary resistance  is defined as the opposition to the flow of gases caused by frictional forces within the respiratory system, and it is calculated according to the equation: Resistance = driving pressure/flow rate.1These frictional pressure losses (ΔP) produced by flow in tubes are a function of flow rate, tube geometry, and fluid physical properties.2 

The inspiratory pressure method6was used to calculate the overall pulmonary resistance according to the equation R = (P^peak− P^plateau)/flow. During passive ventilation, rapid airway occlusion at the end of inspiration produces an immediate decrease in transpulmonary pressure from its peak (P^peak) to a lower value (P¹), followed by a gradual decrease in pressure to an apparent plateau (P^plateau), which represents the static end-expiratory elastic recoil of the lung. This permits pulmonary flow resistance to be partitioned between the true intrinsic resistance of the airways (R^min= (P^peak– P¹)/flow) and an additional effective resistance due to time-constant inequalities and tissue stress adaptation (R^L= (P¹− P^plateau)/flow).8 

This method provides comparable values of resistance with other techniques but does not permit discrimination between patients with airway obstruction and those with infiltrative lung disease. However, the limitations of this method are not considered to affect the measurements during the experiment of the current study.9 

In the current study, the use of a laboratory lung model that simulates the respiratory tract eliminated the effects of lung volume and bronchial tone on airway resistance that largely affect airway resistance in vivo . Therefore, any change to the calculated resistance resulted from changes in the gas mixture physical properties, and changes in the flow pattern that influences the pressure decrease determined by the difference between peak pressure and plateau pressure (ΔP = P^peak− P^plateau). The test lung served as a quantitative, calibrated “test load” simulating human pulmonary physiology for use in testing respiratory equipment and pulmonary research. It is designed to realistically simulate the mechanics of the adult respiratory system from the upper airway. The lung model is not a detailed model of actual human anatomy, and its use helped in determining whether the physical properties of a gas mixture alter the measurements of a fixed resistance. The adult airway is constructed by using a hose assembly, and the airway resistance exhibits parabolic characteristics in regard to pressure change as a function of flow. This nonlinear parabolic characteristic is similar to that seen in standard endotracheal tubes. Similar models have been previously used for studies examining lung mechanics during different modes of ventilation.5,10Compliance can be independently set for each lung using a precision-calibrated spring, and plug-in resistors are provided to simulate pulmonary resistance. The baseline resistance selected (15 cm H²O · l⁻¹· s⁻¹) has previously been reported to occur during general anesthesia.11The lung model mostly represents the central airways and more specifically the trachea and the main bronchi, whereas it cannot simulate alveoli. Therefore, time constant inequalities did not occur, P¹pressure was not observed, and only P^peakand P^plateaupressures were recorded (fig. 2).

The lung model was ventilated with constant flow inflations, and the inspiratory flow was recorded with screen pneumotachograph that provided correction for gas density, viscosity, temperature, and barometric pressure. This type of pneumotachograph contains a fine-mesh screen that provides a small fixed resistance to airflow. As gas flows through the pneumotachograph, a microprocessor-based system converts the measured differential pressure to volumetric flow rate. The screen is actively heated to prevent condensation, which leads to alterations of the resistance and inaccurate results. Furthermore, it was easy to confirm its accuracy by comparing the flow measured at baseline and after the addition of anesthetic gases. Because the lung model was ventilated with volume control mode, it was mandatory that flow remain constant, and only minor breath-to-breath variations should have been recorded.

The fact that flow measurements remained constant for all gas mixtures delivered confirmed the accuracy of the pneumotachograph.

Volatile anesthetic agents produce dose-dependent decreases in airway resistance after antigen-induced bronchoconstriction in animal models.12Halothane is considered to have the most potent bronchodilatory properties due to decreased vagal tone,13but all volatile anesthetics possess bronchodilatory properties.14,15However, there are no studies regarding the effects of the physical properties of volatile anesthetics on airway resistance. Habre et al.  4estimated the viscosity and density values of the pure component of volatile anesthetics. (isoflurane: viscosity 0.892 Pa s × 10⁻⁵, density 5.19 kg/m³; sevoflurane: viscosity 1.276 Pa s × 10⁻⁵, density 6.12 kg/m³; desflurane: viscosity 1.452 Pa s × 10⁻⁵, density 5.44 kg/m³). They also calculated the viscosity and density of their mixtures with air, 100% oxygen, and 50% oxygen. They concluded that volatile agents in the clinically applied concentrations produced a relatively small decrease of the viscosity values of gas mixtures with maximal decrease in viscosity of 3.3–3.5% at 2 MAC desflurane, whereas they significantly increased the density of the gas mixture with maximal increase of density to 47.7% at 2 MAC desflurane in air.4It should be mentioned that either increased density or decreased viscosity increase pulmonary resistance. The current study clearly demonstrated that isoflurane, sevoflurane, and desflurane at high concentrations increased the calculated overall pulmonary resistance in a test lung model, and the greatest increase was observed at high desflurane concentrations (2 MAC). The significant increase of pulmonary resistance observed with desflurane was considered to be the result of the high concentrations required to achieve 1, 1.5, and 2 MAC values (6.6, 9.7, and 13.3%, respectively; table 1) and not from unusually deviant values of viscosity and density of the pure component of the anesthetic.4This observation was supported by the finding that the three anesthetics produced similar effects on airway resistance when delivered at the same low or high concentration, suggesting that the increase of the resistance is due to the concentration delivered rather than differences of the physical properties of the agents used. Similar effects of increased resistance measurements are reported in previous studies regarding the effects of xenon on respiratory mechanics,16,17which showed that the high density of xenon increased airway pressure and resistance in animal models.

The findings of the current study may explain the discrepancies between in vivo  and in vitro  studies regarding the bronchodilatory properties of desflurane. Mercier et al.  18found that desflurane relaxed proximal isolated human bronchi in a dose-dependent manner. On the contrary, Goff et al.  19found that desflurane exerted no bronchodilation in patients undergoing elective surgery during general anesthesia. In the former study, the relaxation of the bronchi was measured directly, whereas in the latter, the isovolume method was used to assess total airway resistance, which is a product of bronchial tone, lung volume, flow rate, and the physical properties of the inspired gas. The current study may also explain the results of Dikmen et al. ,11who found that desflurane produced bronchodilation at 1 MAC but increased airway resistance at 2 MAC. At 2 MAC, the increased density of the gas mixture may have offset the bronchodilatory effect of the anesthetic, and the overall resistance measured was increased. Theoretically, the increased pulmonary resistance caused by the physical properties of anesthetics might have deleterious effects on patients with chronic obstructive pulmonary disease or asthma. However, desflurane, which had the most pronounced effects of pulmonary resistance, has been widely used without adverse effects even in patients breathing spontaneously,20and there are no studies in humans to support that it should be used with caution in patients with airway hyperreactivity.

In conclusion, the high density of volatile anesthetics significantly increases airway resistance. This phenomenon is more pronounced with less potent agents that are delivered in high concentrations. Studies of the effects of these agents on airway resistance should take into account that a percent of these effects may result from the altered density of the inspired gas mixture.