Xenon anaesthesia for all, or only a select few?

Xenon has attracted great interest from clinicians as a novel anaesthetic gas as it possesses a number of desirable properties, including low solubility, lack of metabolism, and hemodynamic stability 1-7. Xenon is relatively expensive, with a current market price of approximately £7 per litre ($10USD, €9) at standard temperature and pressure (STP). However, with advanced scavenging and low-flow techniques, it is conceivable that a xenon anaesthetic could be only slightly more expensive than one using commonly available volatile anesthetics 1. Depending on the anticipated clinical benefits, which are currently the subject of much research, this additional expenditure could be worthwhile. Something similar to the preceding paragraph could have been written about desflurane at the time of its introduction, or isoflurane, or ondansetron, or virtually any pharmaceutical; these agents have now displaced their less expensive but inferior alternatives and become widespread. Xenon, however, is fundamentally different. The economics of xenon extraction are totally unlike most drugs used in healthcare, for which patent protection, or lack thereof, is of paramount concern. It is therefore important not only to consider the current price of xenon but the likely price elasticity of xenon should it be widely adopted as an anaesthetic gas. For reasons outlined below, above a certain level of consumption, the supply of xenon is likely to be extremely price inelastic; increases in consumption beyond a certain fixed capacity will lead to large increases in price. This is often termed xenon's ‘supply ceiling.’ As shown below, even relatively modest consumption of xenon as an anaesthetic could easily exceed the maximum available supply of low-cost xenon. We therefore believe that research into xenon is best focused on the narrow, but high clinical-impact applications of xenon. Specifically, we feel that xenon's potential neuroprotective effects warrant further investigation. Furthermore, in the future it may be necessary for national societies or other large institutions to delineate acceptable and unacceptable uses of xenon to avoid shortages or price spikes. We have attempted to estimate the likely ceiling on the availability of xenon for use as an anaesthetic, given current technology, as well as the implications regarding what proportion of worldwide anaesthetics could be performed using xenon. Currently used volatile anaesthetics are composed of the relatively abundant elements of fluorine, carbon, oxygen and hydrogen; as such their price is minimally influenced by the cost of their feedstock. Instead, it is driven by capital costs, energy costs, regulatory compliance, and other factors, many of which can be reduced by economies of scale and clever engineering. Furthermore, increased use of a particular volatile anaesthetic is unlikely to affect the price of fluorine or other feedstocks, as production of volatile anaesthetics represent only a tiny fraction of annual consumption of these elements. Therefore, volatile anaesthetics are very price elastic—increasing consumption minimally increases unit cost, and in the long term may actually decrease price as new manufacturers are induced to enter the market. This is typical of most pharmaceuticals, particularly off-patent drugs. In contrast, xenon is a relatively rare element and must be purified from the atmosphere. It is manufactured as a byproduct of the production of liquid oxygen and nitrogen 1, 2, and an annual production of approximately 5–10 million litres 8-10. Air separation units designed primarily for production of liquid oxygen and nitrogen through distillation can be fitted with additional equipment to separate out xenon and krypton as byproducts. The only current source of xenon is from the atmosphere, where it constitutes a mere 87 parts per billion of dry air. Air separation units have a high capital cost, and consume large amounts of energy to refrigerate and liquefy air. Much of this expense is relatively fixed, regardless of whether the unit separates and sells all of the gases present in the atmosphere or only a few. Oxygen is generally the gas in highest demand via air distillation 11, although there are large markets for liquid nitrogen and argon, as well as smaller markets for less common gases such as neon, krypton, and xenon. Although the market price of oxygen varies depending on local market conditions and quantity supplied, in bulk it is generally estimated to cost £0.07-0.14 ($0.10-$0.20, €0.09-0.18) per kg; recent large volume sales to government agencies in the United States have disclosed prices of $0.12 per kg 12. Assuming these market prices, 100 cubic meters of dry air contains £2.41 [$3.43, €3.01] of oxygen and £0.06 [$0.09, €0.08] of xenon. The nitrogen, argon, etc. contained therein may also be co-produced and sold for a price comparable to that of the oxygen, if a local market exists. The production of cryogenic gas products is a relatively mature, stable industry, so we can assume this total revenue is approximately the amount required to pay for the capital and operating costs of an air separation unit, with a reasonable rate of return on investment. Therefore, the vast majority of revenue, and profitability, of an air separation unit depends on the sale of more common atmospheric gases. Given current market prices, the decision to construct an air separation unit will not depend in any substantial way on the price of xenon, because, at its current price point, it is such a tiny fraction of overall revenue. However, the decision to add xenon purification ability to an existing facility, which represents a smaller additional investment and ongoing cost, will be influenced by xenon's price. If an air separation unit has already been constructed (to supply demand for liquid oxygen, nitrogen, etc.) these large upfront costs are sunk. The decision to make the smaller additional investment to add xenon purification ability will be influenced by xenon's price. At xenon's current price, not all air separation units have xenon purification ability, so supply could conceivably be expanded to meet increased demand with only modest increases in price. The salient question, then, is whether there is enough ‘idle’ xenon capacity, in the form of oxygen production units that do not attempt to capture xenon, to meet the projected demand for xenon as an anaesthetic. If xenon demand exceeds this capacity, additional increases in supply will be impossible using current methods and despite large increases in the price per litre, shortages will result. Estimates of worldwide oxygen production vary 11, 13, 14. Estimates that include on-site production and consumption by end-users estimate total production at up to 440 million tons 13, although other sources estimate values of approximately 280 million tons 11 in 2013. The relative abundances of oxygen and xenon in the atmosphere are 20% and 87 parts per billion. Therefore, air that has been liquefied to produce the 440 million tons of oxygen contained approximately 120 million litres (at STP) of xenon; this represents a ceiling on easily recovered xenon. Realistically, not all facilities that produce liquid oxygen can be retro-fitted to produce xenon. Many may be too low in capacity 5, or too remote, or have other reasons that xenon production is not feasible. Furthermore, recovery of xenon is generally less than 100%, with approximately 70% being more typical. Therefore, the true ceiling on xenon production is likely much lower than 120 million litres. Regarding the amount of xenon required for anaesthetic purposes, Nakata et al. 1 found that approximately 36 l of xenon were used per case, and another recent article 2 suggested 15 l could be used per case. Even in the case of perfect scavenging from the anaesthesia circuit, xenon will be lost due to evaporation from the surgical site, and some residual xenon will be present in the patient's lung and tissues on emergence 15-17. Furthermore, the occasional need to disconnect the circuit, as well as the inevitability of circuit leaks, will cause some xenon to be lost to the atmosphere. Volumes as low as 4-6 l have been reported for initial equilibration 15-17, although further xenon was consumed during maintenance of anaesthesia. Perhaps it could be possible that only 5 l of xenon are used per anaesthetic, a quantity only slightly larger than the xenon present in the lung alone at steady state during 70% xenon anaesthesia. At current prices, this would represent an expenditure of £35 [$50, €44] per case, a very reasonable amount. Global volume of major surgeries requiring general or regional anaesthesia is estimated as 234 million cases per year 18. Optimistically assuming only 5 l are used per case, and optimistically assuming that our ceiling on xenon production is 120 million l, only 10% of all anaesthetics could be performed using xenon before supply is exceeded. Making more realistic assumptions, that 36 l are used per case on average, and that only 1/3 of worldwide oxygen production capacity can be used to produce xenon, only 0.5% of all anaesthetics could be performed using xenon before the supply ceiling is exceeded. If all available xenon produced as byproduct was exceeded, further xenon could only be obtained by producing xenon as a primary product. Unfortunately, the price of xenon would need to be orders of magnitude higher to justify this production. As noted above, 100 cubic meters of dry air contains approximately £2.40 of oxygen, a comparable value in other gases such as nitrogen and argon, and approximately £0.06 of xenon. If the common gases were unsalable due to a glutted market, xenon would need to increase in price by a factor of 30-40 (making the xenon present in 100 cubic meters worth the value of oxygen in that volume) for plant revenue to be unchanged; this corresponds to a xenon cost of £210 [$300, €263] per l. Even if excess supply only caused oxygen to decrease in price by 20%, xenon would still need to increase in price by a factor of six to make up the difference. This would correspond to a xenon cost of £42 [$60, €52] per l, and a cost per anaesthetic (assuming 36 l were used) of £1517 [$2160, €1895]. Although xenon production as a primary product at current prices is not feasible using current technology, we must assess the likelihood of technological advances providing new and novel ways of purifying xenon. For example, a substance that adsorbs xenon avidly could be discovered, or a membrane permeable only to xenon could be developed. Unfortunately, while incremental improvements to current processes are possible, order-of-magnitude improvements in process economics are unlikely. The dilute concentration of xenon in our atmosphere has much greater entropy compared with purified xenon; as such, there is a thermodynamic limit to the amount of energy required to extract xenon from the atmosphere. Specifically, in the above examples pumps are still required to push vast volumes of air across a membrane or past an absorber, requiring both capital and operating expenditures. Air separation is a mature technology that has existed for over a century, making order of magnitude improvements in process economics unlikely. The study of the economics of purifying dilute substances is quite well developed 19-21 and can be used to inform the above estimates of xenon cost if xenon is produced as a primary product. The cost of separating dilute substances generally increases at a rate inversely proportional to the concentration, regardless of whether the substance is a dilute, gas phase pollutant (like sulfur dioxide) being removed from a waste stream, or a dilute concentration of a valuable mineral (like platinum) being concentrated from ore, or a biological product. This is frequently illustrated using what are generally termed ‘Sherwood plots.’ A Sherwood plot showing extraction cost versus concentration (in 1998 US dollars) for a variety of products is shown in Fig. 1. As shown, gases are generally less expensive to extract than other substances; however, given the atmospheric concentration of xenon is 87 parts per billion, or approximately 10−7, a predicted extraction cost based on Fig. 1 would be approximately $10,000-100,000.kg−1, which corresponds to a price of $60-600.l−1 (in 1998 US dollars), which corresponds to £42-420.l−1. This estimate is very similar to our previous estimate of the cost of xenon by primary production, but was obtained via a different methodology, and bolsters the claim that as soon as xenon must be produced as a primary product, (and not a by-product of oxygen production), its cost will markedly increase. The above discussion implies two possible scenarios for xenon use in anaesthesia: First, at the current level of oxygen and xenon production, xenon may find applications as an anaesthetic, provided that demand is kept well below the current supply ceiling. Although the proportion of worldwide anaesthetics that could be performed using xenon is relatively small, this still represents a very large absolute number of anaesthetics. Nonetheless, consumption per patient must be minimised, and xenon must be reserved for a limited number of indications where it has the highest clinical impact; further research is required to delineate these applications. Furthermore, should compelling clinical indications for xenon anaesthesia be discovered, coordination of providers and suppliers, and agreement on appropriate allocation of the limited amount of xenon may be necessary to avoid causing unexpected shortages or price spikes. Second, xenon could become available for more routine use, but this would be dependent on technological advances outside anaesthesia. Specifically, if worldwide liquid oxygen demand increases markedly, this would in turn greatly expand the availability of xenon. There are several scenarios in which this could occur. For example, combustion of hydrocarbons with pure oxygen instead of air has been proposed to allow subsequent carbon dioxide sequestration 22. Coal-to-liquid fuel conversion may also increase demand for liquid oxygen 13. However, these applications are at the present time not widely deployed. The second scenario is rather speculative. Assuming the first scenario comes to pass, it is worth asking what applications might combine high clinical impact and low surgical volume, and focusing research energy on these applications. Paediatric anaesthesia, particularly for neonates or infants, might be a particularly compelling use of xenon. For example, the incidence of cyanotic congenital heart disease requiring surgical repair is estimated at 1.3/1000 births 23, performing all of these cases using xenon would not appreciably impact supply. Many other congenital malformations have similar incidences and could likewise conceivably be done with xenon without risking shortages. Furthermore, xenon has been shown to have some neuroprotective effects in rat pups 24. Although human data is lacking 25, if a clinically relevant effect was found in human neonates it would be a compelling indication for xenon. Xenon has also been shown to have neuroprotective effects in models of the adult brain after traumatic brain injury 26, and possibly beneficial effects after out-of-hospital cardiac arrest 27, implying it could find use in adults as well. On the other hand, characteristics of xenon anesthesia such as rapid emergence are less compelling. High-volume, low-risk surgeries, in which short emergence time is an important characteristic of an inhaled anaesthetic, would consume too much xenon in aggregate, bringing us closer to the supply ceiling with marginal clinical benefit. Using xenon for rapid induction is also problematic, as any mask induction is likely to waste significant amounts of xenon to the atmosphere. Lastly, the cardiovascular stability associated with xenon can be achieved using other techniques, such as careful selection and dosing of induction and maintenance agents, as well as administration of vasopressors. The existence of these alternative techniques make xenon less compelling given its limited supply. In conclusion, while xenon holds significant promise as an anaesthetic, its successful deployment will require responsible management of this finite resource. No external funding or competing interests declared.