We investigated the feasibility of an intramuscular cobinamide sulfite injection for reversing the physiologic effects of cyanide poisoning in an animal model (11
). Cobinamide has many characteristics that suggest it may be effective in this role, including high binding affinity for cyanide, rapid binding kinetics, and high solubility in aqueous media (8
Cyanide can be lethal within minutes, therefore, mobilization of an intramuscular drug must be very rapid and capable of neutralizing cyanide's effects before irreversible damage occurs. In preliminary experiments performed prior to initiating the present study, we injected cobinamide (without addition of sodium sulfite) dissolved in phosphate-buffered saline intramuscularly into cyanide toxic rabbits (at and above doses that were known to be effective when administered intravenously). However, DOS and CWNIRS data, as well as red blood cell cyanide concentrations, and blood-gas evaluations, all showed cobinamide was not effective by this intramuscular route, even at high doses. A relatively slow rate of transfer of cobinamide into the blood was seen, and dissection of intramuscular injection sites revealed high amounts of cobinamide remaining in the muscle following sacrifice. While cobinamide has an extraordinarily high binding affinity for cyanide, it also reacts with and binds to nitric oxide (30
). We hypothesized that the injected cobinamide induced localized vasoconstriction around the site of injection from binding nitric oxide, resulting in sequestration in the muscle bed. To overcome this problem, we generated cobinamide sulfite by adding sodium sulfite in slight molar excess (1.25 fold higher than the cobinamide dose) to the cobinamide solution prior to injection. Sulfite forms a co-ordinate bond with cobalamin (31
), and, we, therefore, presume that sulfite forms a similar co-ordinate bond with cobinamide. Sulfite binds to cobalamin with moderate affinity (KA
), and is likely to bind to cobinamide with at least as great an affinity. We further hypothesized that sulfite bound to cobinamide would prevent cobinamide from scavenging nitric oxide, thereby allowing cobinamide to be mobilized from the muscle bed. Because cobinamide binds cyanide with a much higher affinity than it binds sulfite, sulfite does not interfere with cyanide binding to cobinamide, as we showed in in vitro
studies. Sulfite exhibits relatively low levels of toxicity to animals, and is used as a food preservative (32
). Injecting cobinamide sulfite at a dose calculated to neutralize the administered cyanide rapidly reversed optical evidence of cyanide toxicity within minutes, with cobinamide appearing in the plasma 2-3 minutes after injection. Dissection of the muscle at the injection site demonstrated good mobilization of the injected solution. The calculated cobinamide sulfite dose required to neutralize a lethal human dose of cyanide could be administered as a 3-4 mL injection.
Rapid diagnosis and continual assessment of treatment response is essential to optimize cyanide countermeasures, particularly in the mass casualty setting. Having an antidote such as cobinamide that is stable, and can be administered intramuscularly in small volumes would allow for stockpiling and distribution of antidote kits to areas of potential risk including industrial sites, military installations, and to rapid response disaster and antiterrorist teams. However, sodium sulfite in liquid form is rapidly oxidized, and in the oxidized state does not prevent nitric oxide (NO) binding to cobinamide. Therefore, the sodium sulfite must be dissolved and mixed with cobinamide shortly before IM injection. In practice, this can be accomplished using a dual-chambered syringe with solid sodium sulfite in one chamber and a cobinamide solution in the other chamber; the two mix immediately before injection.
As a vitamin B12
analog, cobinamide appears to be relatively non-toxic (35
), and we have administered cobinamide sulfite to mice and rabbits at doses 15-40 times higher than we used in this study without evidence of toxicity. A high safety index would allow cobinamide sulfite to be used for potential victims in whom cyanide poisoning could not be confirmed with certainty prior to antidote administration. Furthermore, because cobinamide is a direct binding agent, and does not induce methemoglobin formation, it has the potential to be used in patients with combined carbon monoxide and cyanide poisoning, commonly seen in smoke inhalation injury, in contrast to the concerns of nitrite treatment in such cases (39
Cyanide is not easily detected by rapid assay, either in blood samples, or through noninvasive monitoring. Furthermore, blood cyanide levels do not correlate closely with the degree of toxicity for a number of well described reasons (13
). While diffuse optical spectroscopy does not measure cyanide levels directly, it can assess tissue oxy- and deoxyhemoglobin concentrations, as an indirect, but quantitative, noninvasive measure of the impact of cyanide on physiologic functions (11
). It is of interest that blood cyanide levels did not decrease as quickly as would be expected by the rapid recovery seen in optical oxy- and deoxyhemoglobin measurements. This is consistent with reversing cytochrome-c oxidase inhibition after only partial unblocking of cyanide binding as demonstrated by Leavesley et al. (44
In control animals, deoxyhemoglobin recovery time constants were very different between muscle and CNS (), with the CNS recovering much more slowly. In contrast, in cobinamide sulfite-treated animals, the CNS deoxyhemoglobin recovery rate was much faster than in muscle. These results suggest that cyanide detoxification by cobinamide is very effective in recovering CNS oxy- and deoxyhemoglobin extraction function compared to peripheral muscle. We speculate that the high blood flow rate in the CNS may account for this effect.
In addition to binding cyanide, cobinamide can bind nitric oxide (9
), but the binding constant is several orders of magnitude lower for NO than for cyanide (9
). Thus, cobinamide preferentially binds cyanide, but cobinamide can bind NO when given in excess of available cyanide (9
). NO binding by cobinamide could result in vasoconstriction and increased blood pressure. In preliminary dose ranging studies, we observed oxy- and deoxyhemoglobin overshoot, accompanied by significant hypertension, when cobinamide (without sulfite) was administered intravenously in substantial excess of cyanide.
There are a number of limitations to this study. Utilizing DOS and CWNIRS based hemoglobin oxygenation as the major outcome indicator for cyanide toxicity reversal must be interpreted with caution, since this is not a lethal model, and we have not demonstrated the ability to increase survival in cyanide poisoning with IM Cobinamide sulfite injection. In addition, we do not demonstrate any evidence of CNS function recovery or improvements. Therefore, we cannot draw conclusions regarding effectiveness of cobinamide in treating lethal cyanide poisoning from this study.
With regard to specific limitations, animals were anesthetized for comfort and safety in compliance with animal welfare regulations, but this caused no adverse hemodynamic or other detectable events. Second, the number of animals studied was limited, as was the duration of follow-up, preventing subtle toxicities from cyanide poisoning or cobinamide from being detected. As this is a sub-lethal model, determining whether similar beneficial effects and complete toxicity reversal will be seen in a higher dose lethal model will require separate investigations.
While an intramuscular injection may be ideal for mass casualty settings, intravenous administration of a cyanide antidote may be preferable for individual exposures because of potentially more rapid systemic distribution. However, the time required to establish intravenous access could offset the advantage of faster distribution of intravenous versus intramuscular drug administration, and hydroxocobalamin and other cyanide antidotes must be infused over 15 minutes (45
). Further studies will be needed to address these issues as well. Specific DOS and CWNIRS limitations for real time assessment of cyanide toxicity and reversal are that DOS and CWNIRS measure average tissue constituents to a depth of 4-8 mm (at the source detector separations used in these studies). Deeper tissue effect and organ specific toxicities cannot be assessed using the current study design. Furthermore, as with all near-infrared optical absorption technologies, other potential optically interfering agents, or medical conditions that might alter oxyhemoglobin levels could affect the ability to diagnose and monitor cyanide toxicity, including commonly encountered combined carbon monoxide and cyanide toxicity states. DOS capabilities in these scenarios will need to be investigated as well.
We used DOS to monitor oxy- and deoxyhemoglobin in peripheral muscles and CWNIRS to monitor oxy- and deoxyhemoglobin in the CNS because DOS is limited in maximal source detector separations and signal intensity. Therefore the maximal DOS depth of penetration is shallower, with more limited capabilities for CNS measurements. The two methods yield similar results when simultaneously monitoring over two separate muscle beds, validating their use.
In conclusion, this study demonstrates the feasibility of an intramuscular injection of cobinamide sulfite for rapidly reversing the physiologic effects of cyanide toxicity on oxy- and deoxyhemoglobin as an indicator of cytochrome oxidase blockage and tissue oxygen extraction capabilities. In addition, the study provides further evidence of the value of DOS and CWNIRS as tools for real-time, quantitative assessment of cyanide toxicity effects and comparisons between various potential cyanide treatment regimens. If additional studies confirm the efficacy and safety of intramuscular cobinamide sulfite in lethal cyanide exposures, this approach may be paradigm shifting in treating mass exposures to cyanide.