Nitrous oxide: time to stop laughing?

Dephlogisticated air (nitrous oxide) was discovered by Joseph Priestley in 1786 [1] when he noticed that ‘a candle burned in this air with a remarkably vigorous flame.’ He called it dephlogisticated air because of its ability to sustain combustion. Priestley assumed that the new gas had a propensity to absorb phlogiston that explained this observation (at the time phlogiston was thought to be the flammable part of any substance). The analgesic properties of this new gas were noted by Davy in his Researches, Chemical and Philosophical [2] in which he records how the inhalation of three or four breaths of the gas greatly relieved his toothache. He went on to conclude that ‘As nitrous oxide in its extensive operation appears capable of destroying physical pain, it may probably be used with advantage during surgical operations in which no great effusion of blood takes place.’ Today's anaesthetists are only too familiar with ‘great effusions of blood’ and it is this that has brought nitrous oxide once more under the spotlight, since operations involving heavy blood loss do not usually last less than two hours. Ever since its first (unsuccessful) use as an anaesthetic agent by Horace Wells in 1845, the role of nitrous oxide in contemporary anaesthetic practice has been keenly debated. For further information about the history of nitrous oxide, readers are referred to a seminal series of articles by Smith in the British Journal of Anaesthesia, of which the most recent is referenced [3]. The rest of this editorial will attempt to provide a current review of its place in modern anaesthesia. Nitrous oxide has a minimum alveolar concentration (MAC) of 105% at 1 atm in humans [4] and thus at ambient pressure cannot be used as a single anaesthetic agent. It is most widely used as a carrier gas for other more potent volatile agents, its analgesic properties reducing opioid need in the operative period. In combination with 50% oxygen (as Entonox) it is used extensively in delivery suites around the world for pain relief in labour and by field emergency staff for analgesia, usually for trauma. It has also been used in alcohol dependency withdrawal programmes. The administration of high concentrations of nitrous oxide to patients has caused concern among anaesthetists who worried about the risk of delivering a hypoxic mixture of inspired gas. It was this fear that led to the introduction of mechanisms linking oxygen and nitrous oxide flowmeters such that a hypoxic mixture could not be delivered. Should another gas be deemed suitable to replace nitrous oxide in the future then the technology for hypoxic mixture prevention will already be in place. Up until recently there has not been a rapidly acting opioid that could provide the same degree of analgesia as nitrous oxide, whose action would also terminate as quickly as that of the gas. The arrival of remifentanil means that this is now no longer the case and may stimulate some commentators to suggest that nitrous oxide can no longer justify its place in the anaesthetic armoury. The cost issue will no doubt be raised by those responsible for budgets; this debate surrounds any new anaesthetic technique and will thus not be discussed further here. The kinetics of nitrous oxide have intrigued investigators ever since its discovery, although only in the last 50 years have we understood how the gas washes in and out of the body. Stenqvist [5] provides a useful review for those wishing to read more about this topic. The relative insolubility of nitrous oxide has aroused most interest since it is this aspect of the gas that has led to much of the criticism associated with it. It increases the size of pneumothoraces and air emboli, it increases middle ear pressure and causes bowel distention, both of which may be associated with postoperative nausea and vomiting. It has been known that its presence can cause postoperative hypoxaemia since Fink described the phenomenon of diffusion anoxia in 1955 [6], but whether or not this is important clinically has been contested in the literature [7, 8]. Since the incidence of hypoxaemia in the immediate postoperative period can be significantly reduced by the administration of supplemental oxygen this point remains largely of academic interest. What is certain is that diffusion hypoxia can only occur if there is present in sufficient quantities a gas that is less soluble in blood than air. The most convincing evidence that nitrous oxide may be detrimental to those to whom it is administered has been gathered by investigators concerned with the effects of the gas on bone marrow. In 1978, Amess and co-workers demonstrated that exposure for 24 h to nitrous oxide interfered with synthesis of DNA [9]. This confirmed the earlier work of Banks et al. [10] who showed that nitrous oxide converts the cobalt in vitamin B12 from the monovalent form to the bivalent form and thus inactivates it. This leads to the megaloblastic changes seen in individuals exposed to nitrous oxide for long periods and Nunn [11] has suggested that it is probably also responsible for the fetotoxicity attributable to nitrous oxide. It is well documented that prolonged exposure to nitrous oxide can lead to subacute combined degeneration of the spinal cord, but there has recently been a report of a patient who developed this condition after a single exposure lasting 8 h [12]. Interference with DNA synthesis may be detected by demonstrating an abnormal result from a deoxyuridine (dU) suppression test and abnormal tests have been reported in patients exposed to nitrous oxide for as little as 2–4 h [13, 14]. Amos and colleagues [15] raised the possibility that compromised patients may be more susceptible to the effects of nitrous oxide, by reporting a series of patients requiring intensive care in which 18 out of 22 who developed megaloblastic changes went on to die. Thus it seems that nitrous oxide may cause bone marrow changes after as little as 2 h exposure and that particularly ill patients may fare less well subsequently. To counter this there is good evidence [16] that administration of 30 mg folinic acid twice daily can prevent development of an abnormal dU suppression test and also the megaloblastic changes seen after nitrous oxide exposure. Whether or not we should be giving one drug to prevent the side-effects of another is a matter for debate. We do this every day when we administer anticholinergic drugs before reversing the action of muscle relaxants; some may say that this is an unavoidable part of modern anaesthetic practice whereas nitrous oxide is not. In rats nitrous oxide is undoubtedly teratogenic [17–19] and this is surely a matter for concern. Although the dose required was large (70% inspired concentration) for several hours or days, this is not without the realms of possibility in modern practice. The irrefutable fact remains that nitrous oxide has been clearly shown to cause teratogenesis in a mammalian model and when alternative techniques are available serious consideration should be given to their use. Certainly drugs have been withdrawn from the market on much less convincing evidence than this. There have been suggestions in the past that chronic low levels of anaesthetic gases may impair the performance of staff working in these environments. In 1978 Smith and Shirley [20] concluded that there was no evidence that this was the case and this was also the opinion of Dale and Husum in 1994 [21]. The issue of whether or not nitrous oxide may be harmful to staff who are exposed to chronic low levels of the gas remains unresolved. There is evidence that exposure is associated with an increased incidence of spontaneous abortion [22, 23], that it is not [24] and that it is associated with increased time to conception [25]. It seems unlikely that this matter will be resolved in the near future, but what is clear is that the use of scavenging dramatically reduces the levels of nitrous oxide in the clinical environment [26]. Where nitrous oxide is in use routinely, the onus is on those responsible for service provision to ensure that adequate scavenging facilities are in place. There have also been concerns raised about the effect of nitrous oxide on the atmosphere, its contribution to the greenhouse effect and depletion of the ozone layer. While it is true that nitrous oxide does contribute to these phenomena [27], ecologically minded anaesthetists can rest easily in the knowledge that anaesthetic-related nitrous oxide represents about 1% of the total global production of the agent [28] and this accounts for 0.05% of the greenhouse effect. So what of the future? Attempts to predict future trends are notoriously inaccurate but in the case of nitrous oxide it appears probable that its use will decline. In remifentanil there is available a short-acting opioid that could replace all of the anaesthetic uses of nitrous oxide. While this drug could probably also be developed for use in delivery suites, that will probably take a little longer. There is currently a resurgence in the use of the inert gas xenon as an anaesthetic agent. It is the only one of the inert gases that has anaesthetic properties at ambient pressure, having an MAC of 71% [29] and preliminary reports of its use were favourable. Because it must be prepared from air, in which it is present at a concentration of 0.0000087%, it is very expensive; current prices are about £10 per litre although costs can be minimised with the use of a closed circle system [30]. In 1990 Nunn and Halsey [31] suggested that the time was not ripe for xenon to replace nitrous oxide in the anaesthetic arena; seven years later it may be time to reconsider whether nitrous oxide can still justify its place in our formulary.