A minimal-flow system for xenon anesthesia

The Modification and Performance of a Large Animal Anesthesia Machine (Tafonius®) in Order to Deliver Xenon to a Horse

Introduction

Xenon, due to its interesting anesthetic properties, could improve the quality of anesthesia protocols in horses despite its high price. This study aimed to modify and test an anesthesia machine capable of delivering xenon to a horse.

Materials and methods

An equine anesthesia machine (Tafonius, Vetronic Services Ltd., UK) was modified by including a T-connector in the valve block to introduce xenon, so that the xenon was pushed into the machine cylinder by the expired gases. A xenon analyzer was connected to the expiratory limb of the patient circuit. The operation of the machine was modeled and experimentally tested for denitrogenation, wash-in, and maintenance phases. The system was considered to consist of two compartments, one being the horse’s lungs, the other being the machine cylinder and circuit. A 15-year-old, 514-kg, healthy gelding horse was anesthetized for 70 min using acepromazine, romifidine, morphine, diazepam, and ketamine. Anesthesia was maintained with xenon and oxygen, co-administered with lidocaine. Ventilation was controlled. Cardiorespiratory variables, expired fraction of xenon (FeXe), blood gases were measured and xenon was detected in plasma. Recovery was unassisted and recorded.

Results

FeXe remained around 65%, using a xenon total volume of 250 L. Five additional boli of ketamine were required to maintain anesthesia. PaO₂ was 45 ± 1 mmHg. The recovery was calm. Xenon was detected in blood during the entire administration time.

Conclusion

This pilot study describes how to deliver xenon to a horse. Although many technical problems were encountered, their correction could guide future endeavors to study the use of xenon in horses.

A Mass-Sensitive Approach for the Detection of Anaesthetic Xenon

Quartz crystal microbalances (QCMs) were utilized for the detection of the noble gas xenon (Xe) by combining them with different recognition layers such as permethylated calixarenes (tetramethyl-tert-butylcalix[4]arene (Cal4Me), hexamethyl-tert-butylcalix[6]arene (Cal6Me)), and polyurethanes, with covalently embedded Cal4OH (Poly4Cal), or Cal6OH (Poly6Cal). A third type of sensitive material is synthesized from polyacrylic acid (PAA) and polyvinyl alcohol (PVA) and utilized as a sensitive coating. The results demonstrate that the Cal4Me layer gives higher response towards Xe, while, by the use of a second uncoated QCM channel, the influence of ambient humidity could be nearly completely compensated by signal subtraction. Moreover, the Cal4Me sensor shows excellent reversibility and rapid response time, providing a potentially reliable way to determine Xe during anaesthesia.

A cryogenic machine for selective recovery of xenon from breathing system waste gases

BACKGROUND

Xenon has many characteristics that make it very attractive as an anesthetic and therapeutic drug. Unfortunately, the supply of xenon is fixed, and therefore reclamation and recovery from even the most efficient breathing circuits is desirable. We built and evaluated a cryogenic device to recover xenon from waste anesthetic gases.

METHODS

Xenon was selectively frozen to -139.2 degrees C from test gas mixtures at ambient pressure (STP). The machine ran on standard 240 V 13 A electrical current without refrigerants that required replenishing, e.g., liquid nitrogen. A wide range of xenon/oxygen mixtures were processed over a range of freezing chamber temperatures. Efflux gas and thawed reclaimed xenon were collected separately. Xenon purity and yield (fraction recovered) were measured and calculated on each occasion.

RESULTS

Gas was processed at 300 mL/min, and the operating temperature was -139.2 (0.096) degrees C [Mean (sd)]. Purity and yield were >90% and >70% for gas mixtures containing > or =20% xenon, increasing to >95% and >85%, respectively, with an input gas xenon fraction > or =40%. Efficiency improved linearly with reducing temperature.

CONCLUSIONS

Xenon of high purity (>90%) and yield (>70%) for such a machine was recovered from all gas mixtures containing > or =20% xenon. The operating temperature of the freezing chamber is a major influence on the efficiency of recovery.

Nitrogen accumulation during closed circuit anesthesia depends on the type of surgery

STUDY OBJECTIVE

The aim of this study is to test the hypothesis that the amount of nitrogen that accumulates within the closed breathing system would be greater during open abdominal surgery than during superficial surgery with small wounds.

DESIGN

Prospective, comparative study.

SETTING

Operating rooms of a university hospital.

PATIENTS

Fourteen American Society of Anesthesiologists physical status I and II adult patients scheduled for abdominal surgery (n = 7) or tympanoplasty (n = 7).

INTERVENTIONS

After induction of anesthesia and endotracheal intubation, the patients were denitrogenated for 30 minutes using 100% oxygen at a fresh gas flow of 10 L/min. The breathing system was then closed and patients were anesthetized using 60% xenon in oxygen, supplemented with epidural anesthesia in the abdominal surgery group and sevoflurane in the tympanoplasty group.

MEASUREMENTS

Nitrogen concentration in the breathing system was determined by gas chromatography immediately before and 2 hours after the breathing system was closed.

MAIN RESULTS

The median (range) increase in nitrogen concentration during the 2-hour period of closed circuit anesthesia was greater in the abdominal surgery patients than in the tympanoplasty patients (6.5% [4.0%-10.2%] vs 2.5% [1.4%-8.4%], P = 0.035, Mann-Whitney U test).

CONCLUSIONS

The amount of nitrogen accumulation during closed circuit anesthesia is greater during open abdominal surgery than in superficial surgery such as tympanoplasty. We postulate that during open abdominal surgery, nitrogen in the ambient air enters the body across the peritoneum and then diffuses into the alveoli to be exhaled.

Nitrogen Diffusion into Closed Anesthesia Systems

OBJECTIVE

In order to reduce losses of gases through plastic components and to reduce nitrogen accummulation during closed system anaesthesia we investigated either 10 sets of anaesthetic tubing made of silicon as used in standard clinical practice and 10 sets made of latex, which are not used anymore due to concerns about latex allergies. The results were compared to each one set made of conventional industrial rubber.

METHODS

Anaesthetic tubings were connected to ventilators with low fresh gas losses, suitable for closed system anaesthesia. For nitrogen measurements, a mass spectrometer was used. The fresh gas flow was set to exceed losses by leakages and the amount of gases, extracted from the system by the mass spectrometer.

RESULTS

Highest accumulation of nitrogen was found using tubings made of silicone.

CONCLUSION

If closed anaesthetic systems in the future will be used in intensive care therapy or in case of long lasting procedures in which closed system anaesthesia is proceeded, materials other than silicone should be investigated to avoid regular purging of system and consecutive losses of gas mixtures.

The analgesic effect of xenon on the formalin test in rats: a comparison with nitrous oxide

To investigate the analgesic effects of xenon, we performed formalin tests in rats under 0.5 minimum alveolar anesthetic concentration xenon or nitrous oxide and stained the lumbar spinal cord for c-fos (n = 18) and the phosphorylated N-methyl-D-aspartate (NMDA) receptor (n = 24) by using the avidin-biotin-peroxidase method. After 20 min of 79% xenon, 68% nitrous oxide, or 100% inhaled oxygen, 10% formalin (100 microL) was injected into the left rear paw of the animals except for a control group. Nociceptive behavior was observed for 1 h. The rats were killed 2 h after the formalin injection, and the lumbar spinal cord was stained for c-fos or the phosphorylated NMDA receptor immunohistochemically. Animals in the xenon and nitrous oxide groups showed less nociceptive behavior than did the oxygen group. Although the number of c-fos-positive cells in the lumbar spinal cord in the nitrous oxide group was not decreased, that in the xenon group decreased. The number of phosphorylated NMDA receptor-positive cells in the xenon group was significantly less than in the nitrous oxide and oxygen groups. Inhaled xenon suppressed nociceptive behaviors, c-fos expression, and activation of the NMDA receptor during the formalin test in rats. These results confirm that xenon's analgesic effects result from inhibition of the NMDA receptor.

IMPLICATIONS

Inhaled xenon suppressed nociceptive behaviors, c-fos expression, and activation of the N-methyl-D-aspartate receptor during the formalin test in rats. Xenon's analgesic effect was speculated to result from the inhibition of the N-methyl-D-aspartate receptor in vivo.

Comparative effects of xenon and nitrous oxide on diaphragmatic contractility in dogs

BACKGROUND

Xenon at two different concentrations (30%, 60%) has no effect on diaphragmatic contractility. This study was undertaken to compare the effects of xenon and nitrous oxide (N2O), a commonly used and well-established gas anesthetic, on diaphragmatic contractility in dogs.

METHODS

Twenty-one pentobarbitone-anesthetized dogs were randomly divided into three groups of seven each: group 1 received xenon 30% (0.25 MAC) in oxygen; group 2 received N2O 47% (0.25 MAC) in oxygen; and group 3 received N2O 60% (0.32 MAC) in oxygen. Diaphragmatic contractility was assessed by transdiaphragmatic pressure (Pdi) at low- (20-Hz) and high-frequency (100-Hz) stimulation, after maintaining 60 min of stable condition. The integrated electrical activity of diaphragm (Edi) to each stimulus was measured.

RESULTS

With an inhalation of xenon 30%, N2O 47%, or N2O 60%, Pdi and Edi at both stimuli did not change. No difference in Pdi or Edi was observed among the groups.

CONCLUSION

When used at clinical concentration, xenon or N2O does not affect contractility and electrical activity of the diaphragm in dogs.

Monitoring xenon in the breathing circuit with a thermal conductivity sensor. Comparison with a mass spectrometer and implications on monitoring other gases

OBJECTIVE

To test the accuracy of a thermal conductivity xenon sensor in vitro and in vivo and to test the effect of xenon on other anesthetic gas analyzers as determined by a mass spectrometry gold standard.

METHODS

The xenon concentration was measured with a prototype of a thermal conductivity sensor and a mass spectrometer in vitro and in 6 patients. Further in vitro experiments determined the impact of xenon on the measurements of oxygen, carbon dioxide and desflurane with three commercially available anesthesia gas monitors.

RESULTS

In vitro the thermal conductivity sensor and an associated computer, when calibrated against a mass spectrometer using a third order polynomial calibration curve measured the xenon concentration to a 95% confidence limit of -1.2 to +1.8 vol% compared to mass spectrometry. In vivo and under clinical conditions with a mixture of xenon, O2 and CO2 the 95% confidence limit was -2.5 to +1.6 vol% with a mean bias of -0.5 vol% over a concentration range of 20 to 70 vol%. Xenon induced a clinically relevant bias on the measurements of oxygen (up to 5 vol%), carbon dioxide and desflurane (both twofold overestimation) in a Hewlett-Packard M1025B monitor. In contrast there was only a small bias on the measurements of a Drager PM8060 and a Datex AS3 compact monitor, which was statistically significant (oxygen and desflurane) but clinically irrelevant.

CONCLUSION

Thermal conductivity is a clinically useful technique to measure xenon in the breathing circuit despite its statistically significant but clinically irrelevant error compared to mass spectrometry. Other gases of interest have to be measured with selected monitors explicitly approved or tested for use with xenon.

Exploration of xenon as a potential cardiostable sedative: a comparison with propofol after cardiac surgery

Xenon anaesthesia is thought to have minimal haemodynamic side-effects. It is, however, expensive and requires special delivery systems for economic use. In this randomised cross-over study, we: (i) investigated the haemodynamic profile and recovery characteristics of xenon compared with propofol sedation in postoperative cardiac surgery patients, and (ii) evaluated a fully closed breathing system to minimise xenon consumption. We demonstrated a significantly faster recovery from xenon (3 min 11 s) than propofol sedation (25 min 23 s). Relative to propofol, xenon sedation produced no change in heart rate or mean arterial pressure and there were significantly higher mean values for central venous pressure (10.6 vs. 8.9 mmHg), pulmonary artery occlusion pressure (11.2 vs. 9.5 mmHg), mean pulmonary artery pressure (20.1 vs. 18.3 mmHg) and systemic vascular resistance index (2170 vs. 1896 dyn.s.cm-5.m-2). The haemodynamic profile seen with propofol reflected its known vasodilator effects. This was supported by the almost identical left ventricular stroke work indexes seen with both methods of sedation.

The effects of 30% and 60% xenon inhalation on pial vessel diameter and intracranial pressure in rabbits

Xenon may increase cerebral blood flow and intracranial pressure (ICP). To evaluate the effects of xenon on brain circulation, we measured pial vessel diameter changes, CO(2) reactivity, and ICP during xenon inhalation in rabbits. Minimum alveolar anesthetic concentration (MAC) for xenon was established in rabbits (n = 6). By using a cranial window model, pial vessel diameters were measured at 30% and 60% xenon inhalation and in time control groups (n = 15). ICP, mean arterial blood pressure, and heart rate were recorded during 30% and 60% xenon inhalation (n = 5). Pial vessel diameters were measured during hypocapnia and hypercapnia conditions in 60% Xenon and Control groups (n = 14). MAC for xenon was 85%. Xenon (0.35 and 0.7 MAC) dilated the arterioles (10% and 18%, respectively) and venules (2% and 4%, respectively) (P < 0.05). Dilation of arterioles was more prominent than that of venules. ICP, mean arterial blood pressure, and heart rate did not change during xenon inhalation. No difference in CO(2) reactivity was observed between Xenon and Control groups (P = 0.79). Sixty percent xenon (0.7 MAC) dilated brain vessels, but venule changes were small. Xenon did not increase ICP and preserved CO(2) reactivity of the brain vessels.

IMPLICATIONS

Xenon might increase cerebral blood flow; however, 0.7 minimum alveolar anesthetic concentration xenon preserved both low intracranial pressure and CO(2) reactivity of the cerebral vessels in the normal rabbit.

Xenon expenditure and nitrogen accumulation in closed-circuit anaesthesia

The high price of xenon has prevented its use in routine, clinic anaesthetic practice. Xenon therefore has to be delivered by closed-circuit anaesthesia. The accumulation of nitrogen is a significant problem within the closed circuit and necessitates flushing, which in turn increases gas expenditure and costs. In previous investigations, nitrogen concentrations between 12% and 16% have been reported in closed-circuit anaesthesia. In order to avoid such nitrogen accumulation, we denitrogenised seven pigs using a non-rebreathing system and connected the animals to a system primed with a xenon/oxygen mixture. In comparison, seven pigs were anaesthetised with xenon using a standard low-flow anaesthetic procedure. Anaesthesia time was 2 h. Nitrogen concentrations in the closed system ranged from 0.08 to 7.04% and were not significantly different from those observed during low-flow anaesthesia. Closed-circuit anaesthesia reduced the xenon expenditure 10-fold compared with low-flow anaesthesia.

[Xenon anesthesia: from myth to reality]

OBJECTIVE

To analyze the current knowledge concerning xenon anaesthesia.

DATA SOURCES

References were obtained from computerized bibliographic research (Medline), recent review articles, the library of the service and personal files.

STUDY SELECTION

All categories of articles on this topic have been selected.

DATA EXTRACTION

Articles have been analysed for history, biophysics, pharmacology, toxicity and environmental effects and using prospect.

DATA SYNTHESIS

The noble gas xenon has anaesthetic properties that have been recognized 50 years ago. Xenon is receiving renewed interest because it has many characteristics of an ideal anaesthetic. In addition to its lack of effects on cardiovascular system, xenon has a low solubility enabling faster induction of and emergence from anaesthesia than with other inhalational agents. Nevertheless, at present, the cost and arety of xenon limit its widespread use in clinical practice. The developement of closed rebreathing system that allowed recycling of xenon and therefore reducing its waste has led to a recent interest in this gas. Reducing its cost will help xenon to find its place among anaesthetic agents. An European multicentric clinical trial under submission will contribute to the discussion of the opportunity for xenon introduction in anaesthesia.

Effect of xenon on diaphragmatic contractility in dogs

PURPOSE

This study was undertaken to examine the effect of xenon on diaphragmatic contractility in pentobarbitone- anesthetized, mechanically ventilated dogs.

METHODS

Twenty-one dogs were randomly allocated to three groups (n=7 of each): Group I received oxygen 100%; Group II received xenon 30% in oxygen; Group III received xenon 60% in oxygen. Diaphragmatic contractility was assessed by measuring transdiaphragmatic pressure (Pdi) generated during supramaximal stimulation of phrenic nerves at the neck at low-frequency (20-Hz) and high-frequency (100-Hz) stimulation, after maintaining 60 min of stable condition.

RESULTS

With inhalation of xenon at two different concentration (30% and 60%), no changes were observed in Pdi at either concentration. There was no difference in Pdi among the three groups.

CONCLUSION

Increasing the concentration of xenon to 60% has no effect on diaphragmatic contractility in dogs. This suggests that xenon may be used safely as an anesthetic with respect to respiratory muscle function.

Measurement of rubidium and xenon absolute polarization at high temperatures as a means of improved production of hyperpolarized (129)Xe

We report on a rubidium-xenon (Rb-Xe) polarization unit for the continuous production of large quantities of hyperpolarized (129)Xe. The unit includes two diagnostic systems which enable the absolute measurement of both the (85)Rb and (129)Xe polarization in situ and at high temperatures. The Rb diagnostic system allows the measurement of one- or two-dimensional images of the absolute Rb polarization and thus enables the experimental study of light penetration into the optical pumping cell. The equilibrium Xe polarization measured in the optical pumping cell and in the freezing unit is typically approximately 20%, under optimal flowing conditions, and this is much lower than the volume-averaged rubidium polarization.

Xenon has greater inhibitory effects on spinal dorsal horn neurons than nitrous oxide in spinal cord transected cats

Xenon (Xe) suppresses wide dynamic range neurons in cat spinal cord to a similar extent as nitrous oxide (N2O). The antinociceptive action of N2O involves the descending inhibitory system. To clarify whether the descending inhibitory system is also involved in the antinociceptive action of Xe, we compared the effects of Xe on the spinal cord dorsal horn neurons with those of N2O in spinal cord-transected cats anesthetized with alpha-chloralose and urethane. We investigated the change of wide dynamic range neuron responses to touch and pinch by both anesthetics. Seventy percent Xe significantly suppressed both touch- and pinch-evoked responses in all 12 neurons. In contrast, 70% N2O did not show significant suppression in touch- and pinch-evoked responses. These results suggest that the antinociceptive action of Xe might not be mediated by the descending inhibitory system, but instead may be produced by the direct effect on spinal dorsal horn neurons.

IMPLICATIONS

Xenon (Xe) is an inert gas with anesthetic properties. We examined the antinociceptive effects of Xe and nitrous oxide (N2O) in spinal cord-transected cats. Our studies indicate that Xe has a direct antinociceptive action on the spinal cord that is greater than that of N2O.

A comparative study of the antinociceptive action of xenon and nitrous oxide in rats

We attempted to clarify the mechanism of antinociceptive action induced by xenon and nitrous oxide. Eighty percent of nitrous oxide or 80% xenon was applied to rats inside enclosed clear plastic glass cylinders with their tails protruding for assessment of the tail-flick response to radiant heat. With repeated testing, there was a rapid reduction to nitrous oxide antinociception within 90 min, which was interpreted as development of tolerance, but not to xenon antinociception. Nitrous oxide antinociception was blocked by the intraperitoneal administration of 0.1 or 1.0 mg/kg yohimbine, but not by 1.0 or 5.0 mg/kg L659-066 or by 5.0 or 10 mg/kg naloxone. Xenon antinociception was not affected by any of these drugs. Yohimbine and L659-066 are characterized as alpha 2-adrenoceptor antagonists. Although yohimbine penetrates the blood-brain barrier after systemic administration, L659-066 does not penetrate it and act peripherally. Therefore, the results indicate that alpha 2-adrenoceptors, but not opioid receptors, may play a key role in antinociception induced by nitrous oxide in the central nervous system. Furthermore, the mechanism of xenon antinociception differs from that of nitrous oxide because it does not involve either alpha 2 or opioid receptors.

IMPLICATIONS

The precise mechanism of antinociceptive action of nitrous oxide and xenon remains unknown. It is still controversial whether an opioid system plays a role in antinociception induced by nitrous oxide. The results of the study showed that antagonism of central alpha 2-adrenoceptors, but not opioid receptors, reverses the antinociception induced by nitrous oxide but not by xenon, which indicates that alpha 2-adrenoceptors may play a key role in nitrous oxide antinociception.

Diffusion of anaesthetic gases through different polymers

BACKGROUND

Improvement of working conditions and anaesthesia with closed systems includes reduction of gas leaks during anaesthesia. One source of contamination is the permeation of gases through plastic materials. The volume of gas permeating through a polymer depends on its molecular structure, the solubility and the diffusion coefficient.

METHODS

We designed an experimental set-up to measure the permeation rates of nitrous oxide, enflurane and isoflurane through components of the anaesthetic ventilator made of silicone, latex, rubber and polyvinylchloride (PVC). Reservoir bags, ventilation tubes and endotracheal tubes were investigated.

RESULTS

The highest permeation rates of anaesthetic gases were observed with silicone materials. Permeation through silicone exceeded that of the least permeable material by more than 10.000 times. By summarizing the permeation rates of the single items, the use of silicone increases the anaesthetic system's leakage rate by 4 ml/min, which means an increase of 18% in a modern anaesthetic ventilator and of 31% in a closed system.

CONCLUSIONS

The highest permeation rates were found for nitrous oxide through silicone, although nitrous oxide has a known low solubility in plastic materials. The result demonstrates the dependency of the leakage rate on the diffusibility. The leakage of anaesthetic gases caused by silicone items does not alone lead to unacceptable pollution of operating theatres. To minimize the total leakage rates of minimal-flow-systems, however, plastic materials with low solubility and low diffusion coefficients have to be used.

MR imaging with hyperpolarized 3He gas

Magnetic resonance images of the lungs of a guinea pig have been produced using hyperpolarized helium as the source of the MR signal. The resulting images are not yet sufficiently optimized to reveal fine structural detail within the lung, but the spectacular signal from this normally signal-deficient organ system offers great promise for eventual in vivo imaging experiments. Fast 2D and 3D GRASS sequences with very small flip angles were employed to conserve the norenewable longitudinal magnetization. We discuss various unique features associated with performing MRI with hyperpolarized gases, such as the selection of the noble gas species, polarization technique, and constraints on the MR pulse sequence.

Clinical experience with minimal flow xenon anesthesia

Xenon is a more potent anesthetic than nitrous oxide, and give more profound analgesia. This investigation was performed to assess the potential of xenon for becoming an anesthetic inspite of its high manufacturing cost. Seven ASA I-II patients undergoing cholecystectomy (n = 4), hernia repair (n = 2), or mammoplasty (n = 1) were studied. Denitrogenation by 15-20 min of oxygen breathing under propofol anesthesia was followed by fentanyl-supplemented xenon anesthesia administered via an automatic minimal flow system which held the oxygen concentration at 30%. Xenon anesthesia lasted 76-228 min and 8-14 l of xenon (ATPD) was used, of which 5.6-8.1 l was expended during the first 15 min. Anesthesia appeared to be satisfactory, and the patients woke up rapidly after xenon was discontinued. The automatic system made minimal flow xenon anesthesia easy to administer, but nitrogen accumulation is still a problem. Assuming a xenon price of 10 US$ per litre, the average cost for xenon was about 65 US$ for the first 15 min and then about 25 US$ for each subsequent hour of anesthesia.

Left ventricular performance and cerebral haemodynamics during xenon anaesthesia. A transoesophageal echocardiography and transcranial Doppler sonography study

The effects of xenon anaesthesia on myocardial function and cerebral blood flow velocities were investigated with transoesophageal echocardiography and transcranial Doppler sonography. Seventeen ASA 1 patients undergoing open cholecystectomy (n = 16) or abdominal hysterectomy (n = 1) were studied. Anaesthesia with 65% xenon in oxygen was induced by ventilating the lungs through a circle system with minimal fresh gas flow. The echocardiographically obtained mean (SD) fractional area change in a short axis view of the left ventricle at the level of the papillary muscles was 65 (10)% (n = 14) before xenon. There was no significant change after 5, 10 and 15 min of xenon anaesthesia. Cerebral blood flow velocities were unchanged during the first 5 min of xenon anaesthesia, but were significantly increased in the left and right middle, and the right anterior, cerebral arteries after 15 and 30 min (n = 16) (p < 0.05). In conclusion, xenon anaesthesia had no adverse effect on myocardial function, but probably increased cerebral flood flow.

A computer-controlled closed anaesthetic breathing system

We describe the design and working of a computer-controlled, closed anaesthetic breathing system which rapidly achieves and maintains a prescribed end-tidal concentration of isoflurane in oxygen. The system is simple to set up and not expensive; the only nonstandard component is a modified glass syringe. We have demonstrated that gas analysers may contribute as much as the patient to the accumulation of nitrogen within the breathing system. Details of our clinical experience with the system are presented in an accompanying article.

An evaluation of anaesthetic loss from a closed breathing system

We present the results of a laboratory study of the loss of isoflurane from a to-and-fro system, ventilated with oxygen, using a standard ventilator connected to the system via a long corrugated tube (the trunk) in place of the reservoir bag. No conventional fresh gas supply is used and isoflurane is injected as a liquid directly into the soda lime canister. The loss of isoflurane from the system due to mixing in the trunk was generally less than that lost by absorption into rubber and plastic system components. Glass syringes used for injection of liquid isoflurane were found to be a potential source of much greater leaks. The results showed that a to-and-fro system ventilated with oxygen via an open trunk functions as a virtually completely closed anaesthetic breathing system.