Wrong Place, Wrong Time

Oximes work by reactivating an initial inhibited organophosphate-AChE adduct before an irreversible secondary “ageing” dealkylation reaction occurs. Critically, if pralidoxime is administered after AChE has “aged,” it has zero “reactivating” efficacy. ^{Reference Bevan17}

The ageing half-lives of organophosphates vary substantially: VR has an ageing half-life of 139 h, tabun 19 h, parathion-methyl 4 h, sarin 3 h, but soman just 6 min, ^{Reference Worek, Thiermann and Szinicz18} and cyanobacterial guanitoxin theoretically instantaneously. ^{Reference Fiore, de Lima and Carmichael19,Reference Hyde and Carmichael20} Furthermore, the half-life of novichok agents (used in the Navalny and Skripal poisonings ^{Reference Eddleston and Chowdhury13} ) is not reliably published, but is believed to occur rapidly. ^{Reference Jeong and Choi21} Therefore, unless pralidoxime is administered within minutes of Soman exposure, or hours of sarin exposure, it is unlikely to have any clinical benefit.

This poses a major problem for the United Kingdom, as pralidoxime is a “Category C” antitoxin—meaning it is held “supra-regionally”. Indeed, approximately 2/3 of acute hospitals do not stock pralidoxime, and the mean estimated delivery time from holding centers to these hospitals was recently estimated to be 114 min. ^{Reference Harnett, Vithlani and Sobhdam22} Therefore, in scenario 1, there is a possible near-2-h delay until drug is administered to victims. In scenario 2—the nature of the chemical would be initially unknown, likely leading to a further delay in diagnosis of organophosphate exposure. This happened in Matsumoto 1994, where it took 3 d to diagnose sarin release by terrorists. ²³ Alternatively, with few victims, scenario 3 may be even more challenging to identify—as organophosphate poisoning presents similarly to opioid overdose—which may be presumed first to the rare nature of organophosphate exposure. ^{Reference Eddleston and Chowdhury13} Indeed, in the Skripal poisoning, pralidoxime was not administered until the following day. ^{Reference Eddleston and Chowdhury13,Reference Hulse, Haslam and Emmett24}

Therefore, as illustrated on Figure 1, in scenario 1, pralidoxime is likely to have no benefit for soman poisoning, but possible efficacy for medium/slow ageing organophosphates. In scenario 2, a day-long delay is likely to reduce the efficacy of pralidoxime to most organophosphates, whereas in scenario 3, only the slowest-ageing organophosphates such as VR/VX will have reversal from pralidoxime administration.

Only with near instantaneous oxime administration can the fastest-ageing organophosphates such as soman be reversed. This was achieved through supply of 6.7m atropine-oxime auto-injectors to civilians in Israel during the Gulf War. ^{Reference Bentur, Layish and Krivoy25} However, the much lower risk of organophosphate attack in the United Kingdom makes this economically unjustifiable here. Alternatively, whereas these delays might be inevitable for UK civilian use, in military contexts provision of auto-injectors, and a far higher index of suspicion of organophosphate exposure, may collectively enable faster recognition of attack, and administration of drugs—meaning these criticisms do not necessarily apply to military contexts.

One Size Does Not Fit All

Since the 1950s, it has been known that oximes may act particularly well (and better than pralidoxime) at reversing the poisoning of specific organophosphates. ^{Reference Kozer, Mordel and Haim26} Additionally, K-series agents have been shown to have efficacy against GA—which is impressive as GA has a lone amide electron pair—making nucleophilic attack impossible for pralidoxime. ^{Reference Kuca, Cabal and Musilek27,Reference Kassa, Jun and Karasova28} More recently, research on donor blood has allowed direct comparison of particular organophosphates with specific oximes. For every single organophosphate tested, an oxime better than pralidoxime has been found—and obidoxime, which is widely used globally, was found to be at least 3× better than pralidoxime for most G- and V-series agents. ^{Reference Worek, Thiermann and Szinicz18} Further oximes are in development globally—although to date, pralidoxime, obidoxime, and trimedoxime have remained the key oximes licensed for use in humans.

In this context, even in scenarios 1 + 2, where pralidoxime might have efficacy, there likely exists an even better oxime—of which the UK stocks none for civilian use.

Ideally, multiple oximes would be stocked—and the best oxime for a specific organophosphate administered. Although there would inevitably be further delay in the context of time-critical ageing, in the 1995 Tokyo attack, sarin was identified by gas chromatography/mass spectrometry within 2 h of the initial release ²³ —so this is not impossible.

Too Weak, or Too Dangerous?

There is some indication that the WHO recommended dosage of pralidoxime may be too low. ^{Reference Bevan17} The target plasma oxime concentration appears to be based off early experiments in cats, ^{Reference Bevan17,Reference Eddleston, Eyer and Worek29,Reference Sundwall30} which found anesthetized cats to be protected from organophosphates at this concentration—with further work showing this could be achieved in humans in 10 min with intramuscular injection of 20-30 μg/kg pralidoxime. ^{Reference Sundwall31} A fundamental problem with this extrapolation is that major differences in oxime efficacy exist between species, ^{Reference Davies, Green and Willey32} making application to humans very difficult.

Subsequent analysis in humans indicates this may not be appropriate.

First, an RCT in 2009 compared the WHO regimen with placebo in patients with pesticide poisoning and found no reduction in mortality with pralidoxime use. ^{Reference Eddleston, Eyer and Worek29} However, mortality was significantly lower in the patients where AChE activity successfully recovered—which may indicate that reactivating AChE does reduce mortality, but the WHO dosage is too low to have this effect in most patients. This is supported by a 2006 RCT, which compared high dose (above the WHO regimen) with low dose (below the WHO regimen) pralidoxime—with mortality 87% lower in the high dose group. Further research is needed, but there is, therefore, some indication that pralidoxime may work well as an oxime, but only at concentrations higher than currently recommended.

There is an alternative risk that pralidoxime might be harmful in some cases—with a nonsignificant harmful hazard ratio found in numerous meta-analyses. ^{Reference Buckley, Eddleston and Li15,Reference Kharel, Pokhrel and Ghimire16} Pralidoxime is toxic in excess, and early studies found that in rodents, pralidoxime has a minimum lethal dose of 100 μg/kg—at least 3× above that used in humans—causing death in 20 min through respiratory depression. ^{Reference Kewitz, Nachmansohn and Wilson4} Although the mechanism is uncertain, as early as 1959, ^{Reference Berry, Davies and Green33} and again more recently, ^{Reference Kuca, Cabal and Musilek27} it has been shown that at high doses, oximes may paradoxically inhibit AChE. Additionally, there is some evidence that oximes are more toxic in patients exposed to carbamate pesticides (who may present clinically almost identically, as carbamates also inhibit AChE) rather than organophosphates, as oximes may augment inhibition of AChE in these scenarios. ^{Reference Faff and Bak34,Reference Wille, Kaltenbach and Thiermann35}