The noble gas xenon is an almost ideal anaesthetic and offers several potential advantages over classical inhalational agents, such as stable haemodynamics, potent cardioprotective and neuroprotective properties, minimal side effects and fast onset/emergence from anaesthesia.1–7 However, to date these experimentally obtained data have been difficult to verify in clinical settings. The high price of xenon due to its low availability as well as its high minimum alveolar concentration (MAC) value of 63% results in very high costs of a combined xenon/opioid anaesthesia, which limits the use of xenon mainly to clinical trials.8–10 One important reason for the lack of clinical data is a technical one: nowadays, only a small number of anaesthesia circuits are suited for xenon anaesthesia. The closed anaesthesia circuit (Physioflex, Draeger, Luebeck, Germany) was favoured; however, this company does not produce this device any longer. Semi-closed circuits with manually controlled xenon and oxygen fresh gas flow achieve comparable xenon consumptions but lack patient safety and usability.11,12 Owing to cessation of production of the Physioflex, other anaesthesia machines such as semi-closed circuits have to be modified for xenon application to meet an increasing demand after the admittance of xenon as an anaesthetic agent by the European Medicines Agency (EMAEA) in 2007.
The specific fast initial and slow long-term uptake of xenon requires a time-dependent adjustment of xenon and oxygen gas flow, which was ideally carried out in an automatic way. Thus, optimization of the initial xenon flow during the wash-in phase and adjustment of the fresh gas flow to maintain a constant system volume during maintenance should reduce overall consumption. In addition to body weight and fresh gas flow, the filling volume of the anaesthesia circuit, leakage and accumulation of nitrogen during maintenance will influence xenon consumption.
This experimental animal study was designed to optimize a semi-closed anaesthesia apparatus (Tangens 2C Xe; EKU Electronics, Leiningen, Germany) with an electronically regulated fresh gas flow with respect to a closed circuit system (Physioflex). The different nonsystematically chosen wash-in procedures were adapted step by step to the inertia of the semi-closed circuit, which could not be mathematically predicted. The main economization is expected to be reached by a stepwise variation of xenon and oxygen flow during the wash-in procedure and maintenance. Optimal fast and economic protocols should be provided as automated processes for the user.
Tangens 2C Xe
The Tangens 2C Xe resembles a classical nonrebreathing bellows-in-a-bottle system (semi-closed system) consisting of an electronic mixed gas control unit (EGAMIX) and an out-of-circle vaporizer to deliver the volatile anaesthetic (Fig. 1). The EGAMIX system is newly developed by EKU. Fresh gas is delivered to the circle system during expiration. The precision of the xenon concentration measurement averages about 5% using thermal conductivity, whereas the oxygen concentrations are measured using a paramagnetic sensor with an error of ±1%. The total filling volume of this anaesthesia circuit is about 3.5 l, including a soda lime canister, as recommended by the manufacturer. The inhaled anaesthesia concentration follows the fresh gas flow of oxygen, air and xenon, which are controlled individually and time dependently by the EGAMIX system following a specific programmed protocol.
Physioflex is a closed circuit anaesthesia system. In place of the ‘bellows-in-a-bottle’ principle, the Physioflex uses a number of movable membrane chambers. The stroke of the movable membranes is used for volume measurement to control and adapt the system's volume to the requirements of mechanical ventilation by a central processing unit. The closed circuit Physioflex systems control the fresh gas flow based on the patient's oxygen consumption. Therefore, the Physioflex uses three different control circuits: one for the oxygen consumption, one for the volatile anaesthetic concentration and the third for controlling the circuit volume, which is turned by a blower at 60 l min−1. The Physioflex uses continuous paramagnetic oxygen analysis. As the Physioflex is a closed anaesthesia circuit, the oxygen dosage equals the patient's oxygen uptake. The xenon concentration is determined within the circuit by thermal conductivity. Xenon wash-in is started using the flush option twice, which refreshes the contents of the system with 3 l of the desired oxygen–xenon mixture (30: 70) within 30 s.13 The total respiratory system volume of the Physioflex including the soda lime canister is about 4 l.
All experimental procedures and protocols used in this investigation were reviewed and approved by the local animal care committee as well as the governmental animal care office (no. 50.203–2-AC 38 4/05; Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Recklinghausen, Germany). Furthermore, all procedures and protocols were in accordance with the Guide for Care and Use of Laboratory Animals.14
Eleven pigs weighing 30–35 kg received an intramuscular injection of 4 mg kg−1 azaperone. Anaesthesia was induced by the intravenous injection of 3 mg kg−1 propofol, enabling tracheal intubation. The animals were ventilated with 100% oxygen using a tidal volume of 10 ml kg−1 and a respiratory frequency of 20–24 min−1 to keep the end-tidal partial carbon dioxide tension (pCO2) between 36 and 42 mmHg. Anaesthesia was maintained by continuous infusion of 15 mg kg−1 h−1 thiopental. Ringer's solution (10 ml kg−1 h−1) was administered continuously for fluid maintenance. Basic monitoring included ECG, tail pulse oximetry, temperature determination and invasive femoral artery and central venous pressure measurement using a standard anaesthesia monitor (AS/3, Datex Ohmeda, Helsinki, Finland). After completion of the instrumentation, the expiratory oxygen concentration was set to 30% and the animals were allowed to recover for 1 h with a continued thiopental infusion to achieve normothermia (38.5°C) and stable blood pressure. Before each wash-in of xenon, an almost complete denitrogenation of the animals took place by ventilating with 100% oxygen for 10 min. As pigs are highly susceptible to actelectasis, repeated arterial blood gas samples were taken and these did not reveal a statistically significant difference compared with tail pulse oximetry and capnometry (ABL500; Radiometer, Copenhagen, Denmark).
Wash-in protocols in a semi-closed circuit
To optimize the xenon consumption, we established a stepwise (1–3 steps) electronic adjustment of xenon and oxygen flows during the wash-in phase with an inspiratory oxygen target of 30%. End-tidal oxygen concentrations were used as control quantities to switch between consecutive wash-in steps for patient safety reasons, as the electronic control unit tends to overshoot during the initial gas application. Therefore, we wanted to prevent any hypoxic gas application.
The number of steps, flow rates and oxygen thresholds between the steps were changed according to the protocols presented in Table 1. The number of steps (1–3) was increased in order to increase the initial xenon flow and to reduce it both earlier and to a greater extent. Repeated wash-in phases were performed for each animal, with at least 30 min allowed between protocols in order to reach a xenon concentration within the circuit of below 1%.
Maintenance protocol in a semi-closed circuit
After the wash-in phase, the system switches at a FEO2 level of 0.32 for the maintenance phase of xenon anaesthesia. The fresh gas oxygen flow was manually set between 0.2 and 0.4 l min−1 to keep the bellows in abeyance. Xenon was infused automatically at a rate of 0.3–0.5 l min−1 if the inspiratory oxygen concentration exceeded 0.32.
The wash-in phase of the closed system was executed by using the ‘flush option’ twice with a target oxygen concentration within the circuit of 30%. During maintenance, oxygen was automatically added in 5 ml portions to keep the concentration constant. Xenon was also automatically added to keep the system volume constant. During maintenance, the flush option was not used for strange gas reduction.
The time and xenon consumption needed to reach an end-tidal xenon concentration of 60% (FXe = 0.6) were defined as the wash-in phase and compared between the different wash-in protocols (Fig. 2). The resulting xenon flow was plotted as a function of time (Fig. 2). Total xenon consumption during maintenance was calculated for an additional 2 h and for 30 min intervals (Fig. 3). Xenon flow was a parameter of the electronic gas control unit and its consumption was measured as the integral of flow over time. The parameters were displayed on the screen and stored in exportable data tables in both machines for post-hoc analysis. Volumes and flows were corrected for body weight. Concentrations of xenon and oxygen were given at the end of maintenance, whereas the strange gas portion was calculated as the difference between them (Figs 4 and 5). Differences between protocols for xenon consumption, time variables and gas concentrations were assessed using one-way analysis of variance (ANOVA) with Bonferroni's post-hoc test for multiple comparisons (Prism 5.0; GraphPad Software, San Diego, California, USA). Differences during the maintenance phase for 30 min values were analysed using a two-way ANOVA for repeated measurements (Prism 5.0). A P value less than 0.05 was regarded as significant. Results are presented in bar charts and XY graphs with mean ± SD (Prism 5.0).
Whereas the closed system consumed 0.164 ± 0.016 l kg−1 to reach an FXe of 0.6, the consumption with the semi-closed circuit depended on the protocol. Of the multistep protocols, only protocol 3D consumed significantly less xenon at 0.116 ± 0.012 l kg−1 (Fig. 2a). The lowest values were achieved with protocol 1, wherein only the consumed oxygen was replaced by xenon (0.065 ± 0.019 l kg−1). However, this optimal value was reached at the expense of time, requiring 10.2 ± 1.2 min to reach the target xenon concentration. The multilevel protocols (2, 3B–D) demonstrated comparable time intervals (1.5–2.5 min) to the closed system (2.8 ± 1.2 min; Fig. 2b). The best relation between time and consumption was achieved with protocol 3D and was related to a short but high initial xenon flow with a fast decrease (Fig. 3). In other words, the conservation of a higher initial flow will increase xenon consumption (protocol 2–3C), whereas lower initial flows will increase the duration to reach the target concentration (protocol 1, 3A).
No differences were observed in the total consumption during the maintenance phase of 2 h between the two machines (closed = 0.303 ± 0.02 l kg−1, n = 6; semi = 0.292 ± 0.05 l kg−1, n = 4) and they showed comparable time-dependent effects (Fig. 4). Both machines conserved acceptable xenon concentrations at the end of this phase (closed = 0.62 ± 0.02 l kg−1, semi = 0.62 ± 0.01 l kg−1). Whereas a slightly lower FIO2 was observed within the semi-closed circuit (0.29) compared with the closed system (0.31), the amount of strange gas was similar in both circuits (0.09 l kg−1 ± 0.005 vs. 0.07 l kg−1 ± 0.01; Fig. 5).
The present study demonstrated that, by using an electronic control algorithm for the xenon wash-in procedure, a relevant xenon-sparing effect could be achieved for the tested semi-closed anaesthesia circuit. Time-dependent changes in xenon flow and possible wash-in time will clearly influence overall xenon consumption. The slow wash-in of protocol 1 was the most xenon-sparing, but it was also the most time-consuming routine. This procedure, in which consumed oxygen was replaced by xenon to maintain the circuit volume, reflects the lowest possible value, because xenon was not added in excess and thus wasted through the ‘pop-off’ valve. In clinical settings, this period would need to be bridged by the continuous intravenous application of a different hypnotic drug (e.g. propofol). The introduction of more steps, with a short and high initial flow followed by a quick decrease (protocol 3D) provides the optimal relationship between duration and consumption during the wash-in phase, with better results compared with the closed system. The relative xenon delivery over time was closer to the relative decrease in xenon uptake modified from Taheri and Eger15 when compared with the closed system with a lower area under the curve, thus wasting less xenon and reflecting a higher efficiency of the system (Fig. 6). Therefore, an exponential decay of xenon flow instead of a stepped decrease might lead to a further approximation of xenon consumption to the body's uptake. Xenon uptake resembles an exponential function wherein the sharp increase at the beginning is mainly due to xenon's low blood/gas partition coefficient (λblood/gas = 0.14). Therefore, the equilibrium between the alveolar partial pressure and the inspired xenon concentration is achieved much faster than with other inhalational agents such as isoflurane or desflurane.16 An oversupply of xenon after reaching this equilibrium means a waste of xenon. During the wash-in phase, the closed system could not demonstrate its advantages because it behaved like a semi-closed system due to high fresh gas flow. The 15% higher system volume additionally helped to explain the 40% higher xenon consumption.
No differences in xenon consumption and gas concentrations were found for the maintenance phase because, in the semi-closed system, fresh gas flow was manually adjusted to be in the range of oxygen consumption with a visual control of the system volume as judged by the position of the bellows. As expected, xenon consumption decreased over time due to the filling up of the slow body compartments. Nalos et al.17 calculated that, after 2 h, 0.22 l kg−1 was dissolved in the pig. If 3 l of xenon is necessary to exchange the system volume and lung volume in this setting, the combined use of protocol 1 (slow wash-in) followed by the maintenance phase will consume 0.265 l kg−1 for 2 h of anaesthesia. Thus, approximately 1.5 l xenon would be wasted. Using the faster protocol (3D) in combination with the maintenance phase, 3 l of xenon or 24% of the total consumption would be wasted. Lower xenon consumptions using semi-closed circuits, as described by Burov et al.11 and Luttropp et al.,12 during maintenance (0.2 and 0.133 l kg−1) will be related to the presence of a higher strange gas content when compared with the current study and thus only xenon concentrations below 0.6 can be achieved. The higher fat/blood coefficient (λfat/blood = 15–20) in combination with a lesser blood supply to the fatty tissue leads to a prolonged uptake of xenon in the so-called slow body compartments. Theoretically, a 75 kg patient could take up to 16.5 l of xenon during the first 2 h of anaesthesia, reflecting the minimal possible consumption.17
The use of xenon as an anaesthetic agent requires additional measuring technologies and the calibration of flow sensors for different xenon concentrations. Correct measurement of gas concentrations and flow is fundamental to the implemented control technique and will influence xenon consumption. Most respiratory systems use thermal conductivity to delineate the measurement of the xenon concentration in the inspiratory arm of the anaesthesia circuit. The measurement error of the thermal conductivity for xenon averages about 5% when compared with laser refractometry.18 The use of low flow anaesthesia enhances this error by causing a larger accumulation of humidity in the tubing, which can be avoided by water traps and a small chamber of silica-gel desiccant placed in the sampling line.18 The correct measurement of the xenon concentration is a prerequisite to judgements regarding the accumulation of strange gas, which often serves as a trigger for flushing the system, a procedure that will then dramatically increase the total xenon consumption.
Apart from optimization of xenon flow, additional measures for saving xenon include the choice of correct tubing19,20 and extensive preoxygenation.21 These techniques will reduce strange gas accumulation over time and thus limit the necessity to flush the system. Recycling of xenon is not an option for a reduction in overall consumption, because only 1–2 l will be exhaled during the short recovery from anaesthesia (3.6 min to extubation)22 due to the long half-life of 97 min.23
Whereas initial xenon consumption depends on the system volume, lung volume and thus body weight and fresh gas flow, any further waste is influenced by body weight and the fresh gas flow related to oxygen consumption, strange gas accumulation and leakage. Thus, a semi-closed circuit for xenon application could be additionally optimized using an automatic volume control system, nonlinear xenon fresh gas flow and measurement of inspiratory and expiratory xenon concentration to calculate individual uptake. Additional measurement of xenon behind the pop-off valve will help to avoid xenon waste. Although precise xenon sensors are not available, the fresh gas flow of xenon and oxygen should be strictly controlled by the system volume and the oxygen concentration to avoid hypoxic gas mixtures and low xenon concentrations.18 Automated protocols using the expiratory oxygen concentration as the control value are suitable for preventing low inspiratory oxygen concentrations and thus increasing safety.
The major limitations of this study are the use of animals and the nonsystematic modification of the wash-in protocols. Despite these limitations, the semi-closed circuit could be optimized for xenon usage with lower xenon consumption than the compared standard closed system. Two different automated protocols could be provided to the user to choose between an economic but slow version and a fast procedure with a minimized but comparably higher consumption.
The present work was supported by EKU-Electronics (Leiningen, Germany) and Air Liquide Medical (Duesseldorf, Germany). Special thanks are extended to Thaddäus Stopinski (Institute of Laboratory Animal Sciences), Renate Nadenau and Christian Bleilevens (Department of Anesthesiology) for their help in the laboratory.
1 Dingley J, Mason RS. A cryogenic machine for selective recovery of xenon
from breathing system waste gases. Anesth Analg 2007; 105:1312–1318.
2 Hein M, Roehl AB, Baumert JH, et al
. Establishment of a porcine right ventricular infarction model for cardioprotective actions of xenon
and isoflurane. Acta Anaesthesiol Scand 2008; 52:1194–1203.
3 Hein M, Baumert JH, Roehl AB, et al
alters right ventricular function. Acta Anaesthesiol Scand 2008; 52:1056–1063.
4 Baumert J, Falter F, Eletr D, et al
anaesthesia may preserve cardiovascular function in patients with heart failure. Acta Anaesthesiol Scand 2005; 49:743–749.
5 Baumert J, Hein M, Gerets C, et al
. The effect of xenon
anesthesia on the size of experimental
myocardial infarction. Anesth Analg 2007; 105:1200–1206.
6 Weber NC, Toma O, Wolter JI, et al
. The noble gas xenon
induces pharmacological preconditioning in the rat heart in vivo via induction of PKC-epsilon and p38 MAPK. Br J Pharmacol 2005; 144:123–132.
7 Weber NC, Frassdorf J, Ratajczak C, et al
induces late cardiac preconditioning in vivo: a role for cyclooxygenase 2? Anesth Analg 2008; 107:1807–1813.
8 Coburn M, Kunitz O, Baumert J, et al
. Randomized controlled trial of the haemodynamic and recovery effects of xenon
or propofol anaesthesia. Br J Anaesth 2005; 94:198–202.
9 Kunitz O, Baumert J, Hecker K, et al
does not modify mivacurium induced neuromuscular block. Can J Anaesth 2005; 52:940–943.
10 Rossaint R, Reyle-Hahn M, Schulte Am Esch J, et al
. Multicenter randomized comparison of the efficacy and safety of xenon
and isoflurane in patients undergoing elective surgery. Anesthesiology 2003; 98:6–13.
11 Burov NE, Molchanov IV, Potapov VN, et al
. Perspective in the development and supply of equipment for xenon
anesthesia. Med Tekh 2005:13–18.
12 Luttropp HH, Thomasson R, Dahm S, et al
. Clinical experience with minimal flow xenon
anesthesia. Acta Anaesthesiol Scand 1994; 38:121–125.
13 Tenbrinck R, Hahn MR, Gultuna I, et al
. The first clinical experiences with xenon
. Int Anesthesiol Clin 2001; 39:29–42.
14 Guide for the care and use of laboratory animals. Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council. 7th ed. Washington, DC: National Academy Press; 1996.
15 Taheri S, Eger EI. A demonstration of the concentration and second gas effects in humans anesthetized with nitrous oxide and desflurane. Anesth Analg 1999; 89:774–780.
16 Hanne P, Marx T, Musati S, et al
: uptake and costs. Int Anesthesiol Clin 2001; 39:43–61.
17 Nalos M, Wachter U, Pittner A, et al
. Arterial and mixed venous xenon
blood concentrations in pigs during wash-in of inhalational anaesthesia. Br J Anaesth 2001; 87:497–498.
18 King R, Bretland M, Wilkes A, et al
measurement in breathing systems: a comparison of ultrasonic and thermal conductivity methods. Anaesthesia 2005; 60:1226–1230.
19 Luttropp HH, Rydgren G, Thomasson R, et al
. A minimal-flow system for xenon
anesthesia. Anesthesiology 1991; 75:896–902.
20 Marx T, Froba G, Bader S, et al
. Diffusion of anaesthetic gases through different polymers. Acta Anaesthesiol Scand 1996; 40:275–281.
21 Reinelt H, Marx T, Schirmer U, et al
expenditure and nitrogen accumulation in closed-circuit anaesthesia. Anaesthesia 2001; 56:309–311.
22 Goto T, Saito H, Nakata Y, et al
. Emergence times from xenon
anaesthesia are independent of the duration of anaesthesia. Br J Anaesth 1997; 79:595–599.
23 Marx T, Kotzerke J, Musati S, et al
. Time constants of xenon
elimination after anesthesia. Appl Cardiopulm Pathophysiol 2000; 9:91–96.
Keywords:© 2010 European Society of Anaesthesiology
experimental; semi-closed anaesthesia circuit; xenon; xenon consumption