Figure 5.1--Chemical Structures of Nerve Agents
Nerve agents have been used in some wars since that period (SIPRI, 1971; UN, 1984; Cordesman and Wagner, 1990, 1991); to suppress internal uprisings in Iraq (Macilwain, 1993); and more recently, in large-scale terrorist attacks (Ohtomi et al., 1996; Morita et al., 1995, Okumura et al., 1996). Nerve agents have also been the subject of much concern during and since the Gulf War. They occasioned considerable defensive efforts and, later, concerns that coalition forces might have been exposed to them during the war (Riegle and D'Amato, 1994; House, 1997; Senate, 1994). The concern was increased by the discovery that U.S. forces had unknowingly destroyed a substantial amount of nerve agents in demolitions at the Iraqi depot at Khamisiyah shortly after the end of the Gulf War, resulting in possible exposure to low concentrations of nerve agents over a large area (OSAGWI, 1997a; CIA, 1997).
There is a great deal of literature on nerve agents and organophosphate pesticides, including several recent books on chemical agents (Somani, 1992; Marrs et al., 1996; Sidell, Takafuji, and Franz, 1997), with one giving a detailed summary of human studies in the UK and United States (Marrs et al., 1996; Smart, 1997; Sidell, 1997; Dunn et al., 1997; Sidell and Hurst, 1997).
This chapter provides an overview of nerve agent effects but looks especially at information about the effects of low, or inapparent, exposures on mood, memory, thinking, strength, and behavior. It also pays particular attention to information that may provide insight on mechanisms for long-term neuropathy.
A pilot plant at Munster Lager provided enough tabun for field trials in 1939 (OSRD, 1946). Later, larger production plants were built but encountered considerable delays, with full production of tabun beginning in 1942. Sarin proved more difficult to produce, and only in 1945 were the Germans able to produce several hundred tons of it. Soman was not produced in quantity (SIPRI, 1971; SIPRI, 1973).
Several hundred accidents occurred during the production of nerve agents, and ten workers were killed. Exposure to low levels of tabun was so common that workers were given extra milk and fat rations because it was observed that larger fat consumption had a protective effect (Harris and Paxman, 1982).
The United States and the UK conducted extensive research during World War II on some related compounds, diisopropyl flurorophosphate (DFP) (also designated as agent P-3) being the best known (OSRD, 1946), but these less-toxic variants appeared most suitable as incapacitating agents because of their ocular effects. Achieving lethal concentrations was difficult.
After the war, the United States, the UK, and the former Soviet Union conducted extensive classified research and development. The German plants and technical information were in the part of Germany the Soviets occupied. That appears to have contributed to a very large postwar Soviet chemical effort (Seagrave, 1981; Harris and Paxman 1982; SIPRI, 1971).
The United States began producing sarin on a large scale in the early 1950s; occupational exposures from that period also provided useful data. No worker died, but nearly 1,000 sustained some exposure. Illnesses were generally brief, usually only a few days, sometimes a few weeks (Craig and Freeman, 1953; Gaon and Werne, 1955; Craig et al., 1959; Holmes, 1959; Marrs et al., 1996, Sidell, 1997). These workers have been subject to only limited follow-up, using small groups and controls (Metcalf and Holmes, 1969).
Defensive research into detection, decontamination, and treatment continues. The perception that soman was a key element in the Soviet arsenal, coupled with recognition of its high toxicity and resistance to therapy, resulted in research emphasis on this agent. The rapidity of action and resistance to oxime therapy lead to the development of pretreatment drugs (carbamate reversible inhibitors), as well as deployment of diazepam drugs with some NATO forces (NATO, 1973; Gall, 1981; Marrs et al., 1996; Sidell, 1997; Dunn et al., 1997).
Problems related to aging chemical munitions in stockpiles and decisions in many countries to eliminate chemical weapons have resulted in research into lower-dose exposures and the longer-term implications of exposure of nonmilitary populations (SIPRI, 1980; Watson et al., 1989; Dacre, 1989).
Meanwhile, development and use of organophosphate-based insecticides has proliferated, and they continue to be widely used in agriculture (Hayes, 1982). Although these insecticides are less toxic than the nerve agents, the illnesses they produce clinically resemble those nerve agents produce (Grob and Harvey, 1953; Hayes, 1982). The toxicity of these insecticides to humans is thus relevant (Haley and Kurt, 1997; Haley, Kurt, and Horn, 1997), and this chapter includes information from pesticide studies where it seems helpful. However, Sidell stresses the clinical differences between the organophosphate insecticides and nerve agents, noting that cholinergic crises from pesticides last much longer than those from military nerve agents (Sidell, 1997; Sidell and Hurst, 1997). On the other hand, reviews of possible long-term effects of nerve agents have regarded organophosphate pesticide experience as being informative (NAS, 1982; Karczmar, 1984; Boskovic and Kusic, 1980; Jamal, 1995b).
It was recognized early that the clinical-pharmacological effects of nerve agents and related organophosphate pesticides resembled the strong actions of the neurotransmitter acetylcholine (ACh). This chemical activates specialized receptors at the nerve synaptic junction, promoting discharge of the nerve on the other side of the synapse and stimulating the action of the nerve. ACh is rapidly destroyed by the enzyme acetylcholinesterase (AChE) (one of a family of serine esterase enzymes), which plays a regulatory role to limit the effects of ACh.
A key mechanism of action of nerve agents is their inhibition of AChE, which results in physiological-pathological overstimulation by excessive ACh (OSRD, 1946; Somani, 1992, Ch. 4). This common mechanism explains the similar effects of many nerve agents and their response to therapy with atropine and oximes.
These agents also inhibit a variety of other enzyme systems (e.g., serine esterases), and their effects impinge on other biological systems via mechanisms that the inhibition of AChE does not fully explain. Increased understanding of neurobiology and neurotransmitters has aided the understanding of these agents (O'Neill, 1981; Prioux-Guyonneau et al., 1982).
WEAPONIZATION
The earliest nerve agents, tabun and sarin, were considerably more toxic than the existing chemical gas weapons, such as phosgene, by a factor of 7 to 40 (Franke, 1967; OSRD, 1946). These agents were hard to detect; even when exposures were insufficient for rapid fatality, they injured and incapacitated soldiers. Liquid contamination of soils, clothing, and material could provide a secondary vapor hazard for variable periods. Artillery shells that detonated the same as ordinary shells could deliver these agents effectively (OSRD, 1946). During World War II, the Germans used aerial bombs and spray tanks for delivery. The vapor density allowed the agent to flow into lower terrain, trenches, bunkers etc., extending the hazard after the attack, which the Germans regarded as desirable.
Subsequently, many agents and potential agents were synthesized and tested. Toxicities turned out to be rather similar (Callaway and Blackburn, 1954). The several G agents varied in the threat they posed via the skin (sarin was not very effective), and efforts were made to mix them with other agents that might enhance skin penetration, such as mustards or lewisite (SIPRI, 1973; Krustanov, 1962). However, a variety of other factors, such as stability, ease of production, and physical properties, may have been more important than toxicity in weaponization decisions (SIPRI, 1971, 1973). Efforts were made to thicken the nerve agents with additives to increase their persistence and penetration (SIPRI, 1973). In the end, several countries adopted sarin, while the former Soviet Union produced soman and thickened soman (SIPRI, 1971, 1973).
The later development of the V agents, such as VX, provided a number of very toxic compounds. Although not very volatile, these could be disseminated in aerosols and provided a very high percutaneous hazard with an environmental persistence far greater than the G agents. Both Western and Soviet forces adopted these agents.
Nerve agents can be delivered by free rockets, guided missiles, and mines, as well as mortar and artillery shells, aerial bombs and submunitions, and spray tanks. Weaponized nerve agents are suitable for a large variety of military operations and for both tactical and strategic use.
Defensively, nerve agents can be used to disorganize forces in assembly areas and reserve formations. The more persistent agents can impede advancing forces, especially by reinforcing other obstacles. During the Gulf War, commanders were reasonably concerned that operations to breach Iraqi defenses might be subject to chemical attack (Clancy and Franks, 1997).
Because of the hazards and difficulties of deploying chemical weapons, the United States (and perhaps other countries) developed so-called binary weapons during the 1970s and 1980s. The ingredients to produce a nerve agent were stored separately in the munitions and then were combined to produce the agent shortly before impact (Rutman, 1976; Eyring, 1976). There are reports that Saddam Hussein claimed Iraq had such weapons,[1] but UNSCOM found none (UNSCOM, 1991, 1992, 1995). The United States found the development of such weapons to be challenging. A variety of ingredients were potentially involved, and some of the reaction by-products were also toxic (Rutman, 1976; McNamara et al., 1979). Sarin and VX are the most commonly discussed binary agents in the U.S. stockpile, but other theoretically highly toxic, although less stable, agents might be produced (Lohs, 1975).
There have been reports of a highly toxic Soviet binary nerve agent, called Novichok, designed to be undetectable by U.S. detectors (Smart, 1997). The information came from a migr who indicated that Iraq might have acquired agents of this family. Information about these newer Soviet agents (33 and 232) is only to be found in press reports interviews and Internet postings (Englund, 1992a, 1992b; Adams, 1996; Tucker, 1996). No detailed or peer-reviewed scientific data are available.
Iraq's Capability
Iraq's acts against the Kurds were an early indication of its chemical capability,[2] while the Iran-Iraq War showed considerable and improving Iraqi use of a variety of agents, not all of which were identified (Cordesman and Wagner, 1990). Tabun was definitely used against Iranian forces (UN, 1984). Typical nerve agent casualties were independently confirmed, and tabun was identified in a bomb, mixed with chlorobenzene (a stabilizer) in a percentage quite similar to what the Germans used in World War II (OSRD, 1946). In later fighting, Iraq appears to have used nerve agents with some success in attacks in the southern sector (Cordesman and Wagner, 1990). Although sarin and cyclosarin might have been used in these attacks, there was reason to be concerned that Iraq's large chemical program might also have produced soman. More recently, UNSCOM suspected and later documented that Iraq had produced VX. Iraq initially admitted to some research on VX, but admitted to UNSCOM late in 1996 that it had produced 3.9 tons of VX as well as 58.5 tons of precursor chemicals (Miller, 1998).
After the Gulf War, the UN became aware that Iraq had substantial stocks of tabun, sarin, and cyclosarin--with sarin and cyclosarin being present in the 122-mm rockets destroyed at Khamisiyah. A barrage of such rockets can rapidly establish a lethal concentration over a large area, representing great danger to personnel not wearing respirators. Cyclosarin can also be a more persistent threat than sarin and is a greater percutaneous hazard (U.S. Army, 1990).
It seems unlikely that potential use of such agents against coalition forces had been the reason Iraq chose them. Development of weapons takes considerable time, and the coalition formed rapidly. Tabun has some persistence and is the easiest agent to produce. It is also capable of producing incapacity for many military functions at levels well below lethal concentrations (OSRD, 1946). As with other Iraqi nerve agents, tabun is suitable for both offensive and defensive use.
The discovery that Iraq had substantial stocks of cyclosarin was interesting because, although this agent was fairly well known, no major power had adopted it. Iraq may have selected it to provide a more persistent and percutaneously effective agent than sarin, one that also has formidable inhalation toxicity. With a sarin production capability, Iraq may have found it easier to produce cyclosarin than to develop VX. However, Sidell, Takafuji, and Franz (1997) indicates that Iraq may have produced cyclosarin because precursor chemicals for sarin--but not those for cyclosarin (e.g., cyclohexyl alcohol)--had been embargoed.
Coalition forces offered many potential targets to Scud missiles: airfields, ports, assembly areas, and logistic facilities, some proximate to urban areas. The Iraqi Scuds had payloads sufficient to place considerable agent on target, although not with great accuracy. Before the air war began, Iraq had a substantial air force, which had demonstrated some ability to deliver chemical agents (Cordesman, 1990; UN, 1984; Zilinskas, 1997), which threatened both the same targets as missiles did and other tactical targets.
Writings after the war indicated that U.S. commanders were concerned about the threat of chemical agents to their forces, especially during initial efforts to breach Iraqi defenses, when friendly forces would be concentrated in identifiable locations and not be moving rapidly (Clancy and Franks, 1997). Training emphasized protective equipment and, as the attack risk increased, the use of pretreatment medications.
Potential for Exposure
Both before and after the start of the air war, there were many alarms from chemical-agent detection systems. The significance of these alarms remains controversial. They apparently resulted from other environmental contaminants, and confirmatory tests generally did not find proof of an agent, although some allied force reports continue to raise doubts (Riegle and D'Amato, 1994). Some have alleged that a nerve agent was present, perhaps from attacks on Iraqi chemical storage facilities, while the general position of DoD and other analysts has been that sarin would be unlikely to present a hazard after being dispersed over hundreds of kilometers and thus having the opportunity to disperse and to hydrolyze (OSAGWI, 1998b; PAC, 1996b).
In two separate accidental exposures during the 1950s at Dugway Proving Ground, workers developed signs, symptoms, and laboratory evidence of mild nerve agent exposure in a test area three days after a sarin test, when it was thought safe to work without protection. In both incidents, it was noted there was a lot of dust blowing at the time of exposure, but the exact locations with respect to test area were not indicated. At the time, it was suspected that sarin had survived longer than expected because it was trapped on dust particles. No environmental samples were taken. The severity resembled that seen with a vapor exposure CT of 15 mg-min/m3. This suggests that, in some circumstances, sarin trapped on dust particles may persist for a long time and represent a hazard when stirred into the air (Brody and Gammill, 1954; Craig and Freeman, 1953). Craig and Freeman (1953) described an exposure that took place 24 hours after a test: The safety officer had thought it safe to be in the immediate test area without protection because the weather was warm. Again, exposure seemed to be from dust.
Earlier military research had shown that sarin and the organophosphate pesticide paraoxon trapped on small inert particles were highly toxic to experimental animals (Asset and Finklestein, 1951). Particle delivery is a key means of distributing pesticides. No information was available about the details of the models used to estimate agent dispersion from Iraq to Saudi Arabia, or for the Khamisiyah event, or whether particle trapping is even considered relevant to such models. There are indications that sarin trapped on dust may persist and be dangerous longer than is commonly thought. Declassified reports (Defense Intelligence Agency, 1997) indicate an awareness of Iraqi "dusty mustard," but also noted that other agents might be used in dusty form.
The DoD position has been that Iraq did not use chemical weapons, and there do not appear to have been any readily recognizable casualties from nerve agent attacks. However, as discussed in detail below, there is precedent for misinterpretation of low-level exposures (Gaon and Werne, 1955), and there is some reason to think that the pyridostigmine bromide (PB) pretreatments U.S. troops received could have reduced the intensity of response to low-level challenges (Gall, 1981; Husain, Kumar, et al., 1993; Vijayaraghavan, Husain, et al., 1992).
Controversy thus remains about Iraqi use of chemical weapons,[3] with allegations that there might have been some. Even if the Iraqi higher command had explicitly instructed troops not to use chemicals, they appear to have been present in the operational area.[4] At least some of these lacked distinctive markings, making accidental release feasible. There is no proof that this occurred, however. UNSCOM, as cited in PAC (1997), indicates there were no chemical agents in Kuwait or in Iraq south of Khamisiyah. OSAGWI has also extensively investigated all suspected cases and to date has not been able to confirm chemical weapon exposure except one case for a single individual and in the case of Khamisiyah (OSAGWI, 1997d).
Khamisiyah
It is clear that U.S. forces unknowingly destroyed Iraqi chemical weapons in March 1991 at the Khamisiyah depot, thinking that these were conventional munitions (OSAGWI, 1997a). Rockets containing sarin and cyclosarin were destroyed by explosive charges, releasing some agent into the atmosphere. Several studies have attempted to model exposures from this incident (Babarsky, 1998; CIA, 1997). One study under way (Gray et al., 1998) used the plume analysis to identify troops who had been more and less exposed to the sarin. The case narrative (OSAGWI, 1997a) had indicated that there were no reports of immediate clinical effects on the nearest troops. In an effort to rule out longer-term effects from low-level exposure, the researchers are now comparing the hospitalization experiences of the 61,000 personnel potentially exposed to the plume with a group of 250,000 who had been in the region but not in the plum pattern. Low-level exposure effects are discussed later in this review.
Hypotheses
Unexplained illnesses in personnel returning from the Gulf War generally do not "fit" the pattern of readily recognized disorders associated with nerve agents. Many hypotheses are being tested. Congressional testimony (Riegle and D'Amato, 1994; House, 1997; Senate, 1994) questioned whether a combination of exposures to chemicals might have produced a new delayed-onset disease. The chemicals may have included pesticides, such as the personal repellent diethyl-m-toluamide (DEET), the anti-nerve agent prophylactic PB, and perhaps chemical warfare agents.
Veteran Reports
Studies of selected, defined small groups of ill Gulf veterans and controls in the United States (Haley, Kurt, and Horn, 1997; Haley, Horn, et al., 1997; Haley and Kurt, 1997; Hom, Haley, and Kurt, 1997) both found epidemiological indications of unusual exposures (e.g., flea collars, being outside during attacks) and identified three to six clinical syndromes. They found subtle indications of diffuse neurological injury in a smaller group of 23 veterans and suggested a variant of delayed organophosphate neuropathy, with the suggestion, based on animal research, that nerve agents could not be ruled out as being involved (Abou-Donia et al., 1996; Husain, Kumar, et al., 1993; Husain, Vijayaraghavan, et al., 1993). Jamal et al. (1996), studying ill UK Gulf veterans, found indications of subtle neurological injury, with sensory peripheral neuropathy being the most striking, although exposure studies were not reported. Where the Jamal studies correspond to those of Haley and Hom, they do not always agree; for example, Haley and Hom did not detect sensory neuropathy, while Jamal did not find the abnormal evoked responses that Haley and Hom did. There was, however, evidence of some organic neurological disorder in both groups of ill veterans. The significance of these findings is controversial; while the authors considered them statistically significant, others question the statistical techniques.
The RAND report on pesticides (Cecchine et al., 2000) will document the use of anticholinesterase pesticides in the Gulf theater, while self-reported exposure interviews (Haley and Kurt, 1997) document some unauthorized use of commercial "flea-collar" devices that contained chlorpyrifos. Pesticides are discussed here only with respect to possible interactions with nerve agents.
This review cannot determine the causes of illnesses in Gulf War veterans, but it provides a background of information about nerve agent effects to help analyze hypotheses and plan further studies. PB is discussed only in the context of interactions with nerve agents; a separate report has been issued on PB (Golomb, 1999).
Figure 5.1--Chemical Structures of Nerve Agents
Nerve Agent Chemical Structure
| Agent | X | R1 | R2 |
| Tabun (GA) | CN | N(CH3)2 | C2H5 |
| Sarin (GB) | F | CH3 | CH(CH3)2 |
| Soman (GD) | F | CH3 | CH(CH3)C(CH)3 |
| Cyclosarin (GF) | F | CH3 | Cyclohexyl |
| VX | SCH2CH2N[CH(CH3)2]2 | CH3 | C2H5 |
SOURCE:
SIPRI (1973).
NOTE: Keyed to Figure 5.1.
A carbon-phosphorous bond is common to the nerve agents but is rare in the less-toxic organophosphate pesticides (SIPRI, 1973). Thousands of organophosphate compounds have been synthesized. Because of their chemical characteristics, nerve agents slowly degrade in water, with half-lives from 5 to 40 hours, depending on pH.
The military agents are racemic mixtures of stereoisomers. There are (+) and (-) forms of tabun and sarin, while soman has four chiral forms (Benschop, Berends, and de Long, 1981). The different isomers and mixtures thereof have important toxicological and kinetic differences; for example, the (-) isomer of sarin is more toxic than the racemic mixture (SIPRI, 1973; Boter and Dijk, 1969).
The M8A1 alarm system, which U.S. forces use widely, is designed to detect nerve agents as vapors or aerosols. It responds within less than 2 minutes to G agents in the range of 0.1 to 0.2 mg/m3 and to VX at 0.4 mg/m3. The bias is toward sensitivity, not specificity, and a large number of interfering chemicals can produce false alarms (smokes, fuels, insecticides, paint fumes, cologne) (OSAGWI, 1997f).
The M256A1 system is used not as an alarm but for confirmatory testing. This is a slower response system than the M8A1 alarm, taking 15 minutes for nerve agent analyses, but it is able to detect nerve agent vapors at 0.005 mg/m3 of G agents and 0.02 mg/m3 of V agents. This system is less influenced by the interferants that affect the M8A1 (DSB, 1994).
The UK's CAM also uses ion mobility spectrometry but responds to nerve agents at or below 0.1 mg/m3 within less than a minute. This device also can detect mustard agent vapors (DSB, 1994).
The Fox nuclear, biological, and chemical reconnaissance vehicle is optimized to detect and mark surface contamination by chemical weapons. The system uses a mass spectrometer that is configured to determine suspected threat chemicals promptly, then determine the spectra of specific agents more definitively. This system is less sensitive than the alarms--requiring levels of about 62 mg/m3 and 45 seconds to respond to nerve agents. Events during the Gulf War showed that oil fires and oil vapors could interfere with the system and cause some false alarms. The vehicle was also equipped with the M8A1 alarm (OSAGWI, 1997c).
There are also detector papers used to detect droplets, and according to DSB (1994), M8 paper responds to G or V droplets of 0.02 ml with a color change within 20 seconds or less. M9 paper responds with a color change but to smaller 100 µm drops.
Regarding the sensitivity of detectors in comparison with human thresholds for eye effects, the Subcommittee on Toxicity Values for Selected Nerve and Vesicant Agents (NAS, 1997) estimates miosis levels as follows:
Urinary metabolites of sarin were followed in a patient from the Matsumoto, Japan, incident. Two metabolites--methylphosphonic acid (MPA) and isopropylmethylphosphonic acid (IMPA)--were identified on the first day after the attack. By day 3, MPA was barely detectable, but IMPA was measured for one week. Total excretion was 2.1 mg for IMPA and 0.45 mg for MPA. Estimated total sarin exposure was 0.05 mg/kg (Nakajima, Sasaki, et al., 1998).
Analysis of urine samples from four Tokyo victims not only documented sarin and sarin metabolites, such as IMPA, but also detected chemicals associated with sarin products, such as ethyl-sarin and its metabolite, ethyl methylphosphonic acid. Other contaminants associated with sarin production were found in substantial amounts: ethyl alcohol, isopropyl alcohol, isopropyl methyl phosphonate, and diethylphosphonate (Minami et al., 1998).
Although congressional testimony (Riegle and D'Amato, 1994) emphasized agent alarms after the start of the air war, there were alarms before then, probably false positives. To conserve battery power, fewer detectors were turned on before the air war and therefore one would expect fewer alarms before the air war (OSAGWI, 1999).
Persistence of Nerve Agents
| Agent | Persistence |
| Tabun | Heavily splashed liquid lasts one to two days, depending on weather. Takes 20 times as long as water to evaporate. Persists in water one day at 20°C and six days at 5°C. |
| Sarin | Little persistence. Evaporates as fast as water or kerosene. |
| Soman | Heavily splashed liquid lasts one to two days (depending on weather). Takes four times as long as water to evaporate. Thickeners can extend the duration of persistence. |
| Cyclosarin | Heavily splashed liquid lasts one to two days (depending on weather). Takes 20 times as long as water to evaporate. |
| Thiosarin | Unknown |
| VXa | Splashed liquid can persist for weeks to months. Calculated to evaporate 1,500 times slower than sarin. |
SOURCE:
U.S. Army (1990).
aOther V agents are similar.
There are differences in opinion as to how pertinent studies of organophosphate pesticides are to understanding nerve agents. Clinical differences between organophosphate pesticides and nerve agents should be kept in mind, as will be discussed later. Sidell (1997) emphasized the rapid onset of nerve agent effects compared with those of organophosphate pesticides and noted the longer and more-difficult-to treat course of serious organophosphate pesticide poisoning. Likewise, no seriously poisoned nerve agent casualty has been reported to have the intermediate syndrome that Senanayake and Karalliedde (1987) described as arising from pesticide exposure.
Cholinesterase inhibitors have been used therapeutically in the past for glaucoma and are used currently to treat myasthenia gravis (Harrison, 1997). Cholinesterase inhibitors are currently the most established treatment strategy in Alzheimer's disease--several drugs, including tacrine, donazepil, and rivastigmine are in use, and many others are under study (Nordberg and Svensson, 1998).
At the clinical level, the early signs and symptoms of nerve agents and those of organophosphate pesticides are identical, so there is no ready distinction between them. Early investigators such as Grob and Harvey (1953, 1958) showed equal interest in other organophosphate pesticides and at times included them in studies using nerve agents. Organophosphate pesticides are sometimes used in laboratory models to understand nerve agent effects. Since follow-up information on nerve agents is limited, the somewhat greater human experience with organophosphate pesticides may be instructive in looking for possible effects. Those interested in organophosphate pesticides have also noted common features of longer-term effects (Korsak and Sato, 1977). It is of course impossible to discuss delayed neuropathy without considerable reference to non-nerve agent organophosphates; likewise, experience with organophosphate pesticides illuminates understanding of the common mechanisms of tolerance both classes share. Interactions between environmental exposures to organophosphate pesticides and the effects of nerve agents are also possible.
The model is that these agents inhibit AChE, resulting in excessive ACh effects within the nervous system. The peripheral cholinergic systems are best understood, while the central nervous system picture continues to evolve. Central cholinergic systems are important in protective systems, locomotion, alertness, and memory and in the regulation of a number of cyclic and periodic behaviors (Petras, 1984).
It has also been recognized that nerve agents can also inhibit enzymes outside of the cholinergic system, chiefly serine esterases. O'Neill (1981) reviewed data that indicated a role for anticholinesterase compounds, such as DFP or nerve agents, in altering the metabolism and persistence of important neuropeptides, such as endorphins, enkephalins, and substance P, which are degraded by serine esterases--producing some symptoms of agent exposure that do not respond to atropine. Experimental support of this concept of non-ChE effects on the brain is provided by Clement and Copeman (1984), who found a long-lasting analgesia in mice following exposure to soman and sarin, which was reversed by the opiate antagonist nalaxone. They also infer that inhibition of proteases may increase the effects of endogenous opioids. No information was available on changes in opioid receptors following exposure to anticholinesterase drugs.
Before the discovery of nerve agents, it was known that some organophosphorus compounds (e.g., TOCP) could cause a delayed neuropathy occurring weeks after exposure, in people and animals, the effect being quite separate from any AChE effects. This disorder, organophosphate-induced delayed neuropathy (OPIDN), has been the subject of much study (Abou-Donia, 1981; Johnson, 1975; Johnson, 1992). The disorder was complex, with sensitivities varying according to age, species, and chemical. The hen came to be the standard screening model (Johnson, 1975).
An enzyme, neuropathy target esterase (NTE), was recognized as playing a role in the disorder. Naturally, the nerve agents came under scrutiny for their ability to induce neuropathy (see later discussions). Different agents have been found to vary considerably in their ability to inhibit NTE and to produce neuropathy. It was shown (in hens) that repeated subneurotoxic doses of DFP or sarin, in animals protected by atropine and the oxime P2S, could develop delayed neuropathy even at dosing intervals of 16 days (Davies and Holland, 1972).
Prior to the Gulf War, the prevailing view was that some military agents could acutely produce delayed neuropathy only at very high levels, many times the lethal doses of the agents (and requiring "heroic" treatment efforts for the cholinergic problems) (Marrs, Maynard, and Sidell, 1996; Sidell, Takafuji, and Franz, 1997, Ch. 8; Bucci, Parker, and Gosnell, 1992b, 1992c; Gordon et al., 1983). The reviewer did not find discussions of hazards from repeated low-dose exposures, although neuropathy from repeated low doses of other neurotoxic chemicals was known.
But the concern with illnesses after the Gulf War has turned attention toward that possibility. Haley et al. have suggested that an atypical delayed neuropathy might be involved in some Gulf War illnesses (Haley, Horn, et al., 1997; Haley and Kurt, 1997; Haley, Kurt, and Hom, 1997; Hom, Haley, and Kurt, 1997; and Somani, 1997). There is a report that inhaled sarin in mice at daily doses that do not produce cholinergic signs of illness can produce typical delayed neuropathy after 10 days of exposure (the does would be very symptomatic in humans) (Husain, Vijayaraghavan, et al., 1993); see later discussions.
Regardless of the proximate mechanism of action, a complex cascade of effects can follow once the toxic effect of the nerve agent is initiated: seizures and hypoxia, with the excitotoxins the seizures release producing neuronal injury (Lipton and Rosenberg, 1994).
Figure 5.2--Chemical Structure of ACh
Figure 5.3--Cholinergic Action and Effects of Drugs and Toxins
Figure 5.4--AChE Active Site, with ACh
Figure 5.5--Reaction of Nerve Agents with AChE
Nerve agents and organophosphate pesticides bind to the enzyme, first in a reversible way. Many agents then "age" the enzyme, producing a very difficult-to-reverse bond (Figure 5.5). The aging rate varies with the agent and is an important therapeutic consideration. VX ages slowly (many hours), while soman ages rapidly (6 minutes) (SIPRI, 1976; De Jong, 1987). The carbamate inhibitors, such as eserine and PB, also inhibit the enzyme, but reversibly.
As a consequence of nerve agent inhibition of AChE, ACh accumulates at synapses, giving rise to uncoordinated bursts of signals, initially stimulating function and then paralyzing it. This brings about the characteristic signs and symptoms, which are usually grouped as follows (Klaassen, 1996; Goodman and Gilman, 1990; Marrs et al., 1996; Koelle, 1994):
Alkylating agents used in chemotherapy and for immune suppression, such as cyclophosphamide, inhibit AChE and serum cholinesterase, requiring modifications of anesthesia given proximate to their use (PDR, 1998). Sulfur mustard also inhibits cholinesterase (Krustanov, 1962; Dacre and Goldman, 1996).
Anticholinesterase Potency of Organophosphates
| Agent | pI50 |
| Tabun | 8.6 |
| Sarin | 8.9 |
| Soman | 9.2 |
| Thiosoman | 8.9 |
| Cyclosarin | 10.1 |
| VX | 8.8 |
| DFP | 6.5 |
| Parathion | 4.9 |
NOTE:
See text for definition of pI50.
SOURCES: (Dacre, 1984; SIPRI, 1973).
Toluene, an organic solvent, at levels of 2,000 parts per million, has been shown in vivo and in vitro to inhibit AChE bound to red blood cell and synaptic membranes (Korpela and Tahti, 1988). Other aromatic and chlorinated hydrocarbons inhibited red cell AChE in vitro (Korpela and Tahti, 1986a, 1986b). Another study showed in vitro red cell inhibition from a variety of hydrocarbons (including benzene, xylene, and trichloroethylene); ethanol also slightly inhibited the enzyme (Korpela and Tahti, 1986a, 1986b). Lower levels of toluene (300 parts per million) in vitro also inhibited red cell AChE.
A comparison of rates of inhibition of eel AChE by oxono (P = O) inhibitors and thiono analogs (P = S) of several agents (paraoxon, fonofos, sarin, and soman) showed that the oxono compounds had substantially higher phosphorylation rate constants than their thiono analogs. This was thought to be due to differences in hydrophobicity of the analogs (Maxwell and Brecht, 1992). Aging of the enzyme was not reported.[6] Note that AChE is found in a number of tissues that lack neural connections, such as erythrocytes, lymphocytes, basophils, spermatozoa, and placenta (Sastry and Sadavongvivad, 1979). The functional role of the enzyme in such tissues is not known, and although it is inhibited there by nerve agents, the biological consequences are little understood (Sastry and Sadavongvivad, 1979; Meier et al., 1985).
The main current approach is to use a drug like PB that binds to cholinesterase in a reversible manner, in an amount that leaves many functional sites untouched, avoiding toxicity. When a soldier encounters a nerve agent, such as soman, that binds irreversibly to the enzyme, the sites occupied by the drug are protected from attack. After the dose of nerve agent has reacted elsewhere, the reversible drug PB leaves the enzyme site, restoring function. This approach is needed for agents, such as soman, that "age" and bind rapidly to the enzyme in a way that oxime drug treatment cannot reverse.
Koster (1946) used the carbamate physostigmine to protect cats from DFP. The very rapid aging of the enzyme bound by soman (about 6 minutes) caused great interest in NATO countries in using carbamates to pretreat for this Soviet threat agent. Several countries, including the United States, chose PB (used to treat myasthenia gravis) (Sidell, Takafuji, and Franz, 1997). This drug, the subject of a separate RAND report (Golomb, 1999), ordinarily does not cross the blood-brain barrier because of its polar nature (Goodman and Gilman, 1990).
No discussion was found on the situation of continued PB use after unrecognized exposure to nerve agent. One concludes that an unstated assumption within the research and medical community about PB use was that attacks would be obvious and that PB would not be used after intoxications.
Adverse Health Effects
Considerable effort went into developing administration regimens for PB that would be safe and effective and that would not impair military performance (Gall, 1981). But the actual use of PB during the war has produced unexpected responses and concern. These issues have been exhaustively investigated in a separate volume of this series (Golomb, 1999).
Treatments
Atropine, the mainstay of treatment agents (Marrs, Maynard, and Sidell, 1998; Grob and Harvey, 1953), antagonizes the effects of ACh on muscarinic receptors. It does not act on nicotinic receptors. Oximes, a second class of treatment drugs, are used to reactivate AChE by displacing the agent from the enzyme, but only before aging has occurred. As noted, the treatment of soman poisoning is difficult, and it was for this agent primarily that pyridostigmine was used as a pretreatment. U.S. forces use injectable forms of atropine, pralidoxime, and diazepam. Diazepam is used for its anticonvulsant effect (U.S. Army Medical Research Institute of Chemical Defense, 1995).
It is known that cyclosarin is somewhat resistant to treatment with some common oximes, including pralidoxime, based on animal and in vitro studies (Coleman et al., 1966; Clement, 1992; Worek et al., 1998; Kassa and Bajgar, 1995). However, rhesus monkeys pretreated with pyridostigmine and challenged with intramuscular doses of 5 LD50 of cyclosarin, followed by treatment with atropine and 2-PAM (pralidoxime) had 100 percent survival (Koplovitz et al., 1992). This suggests that the treatment means available to U.S. forces during the Gulf War would have been effective against cyclosarin. Oximes have a number of other effects on allosteric sites and receptors and block overstimulated ganglia.
As mentioned above, however, AChE inhibition does not explain all aspects of nerve agent toxicity (Van Meter, Karczmar, and Fiscus, 1978; Kaufer et al., 1998; O'Neil, 1981). Several nerve agents appear to be weak direct agonists of receptors, with VX acting strongly on nicotinic receptor ion channel sites (Albuquerque et al., 1983, 1985; Eldefrawdi et al., 1985). There are also indications of direct agent binding to synaptic membranes (Anderson and Chamberlain, 1988), and soman has also been shown to act directly on the receptor (Hoskins, 1982). Cholinergic stimulation by inhibition of cholinesterase in effect stimulates expression of the proto-oncogene c-fos in the brain. The long-term significance of this is uncertain, as will be discussed later (Kaufer et al., 1998; Friedman et al., 1996).
Nerve agents can also inhibit other serine esterases (e.g., trypsin, chymotrypsin, and thrombin) (Meier et al., 1985; O'Neill, 1981; Walday, Aas, and Fonnum, 1991; Pasternack and Eisen, 1985) and serine proteases involved in regulating neuropeptides (e.g., substance P and met-enkephalin (O'Neill, 1981; Clement and Copeman, 1984). The functional significance of these findings is unclear, although O'Neil (1981) suggests that some signs and symptoms of nerve agent poisoning might be mediated by enkephalins, whose persistence in the brain is prolonged by exposure to DFP, for example.
Delayed Neuropathy and Neuropathy Target Esterase (NTE)
Of particular interest is NTE, which has been implicated in a form of delayed neuropathy known as OPIDN (Abou-Donia, 1981; Johnson, 1975; Johnson, 1992). It has been hypothesized that some of the neurological findings in Gulf War illness patients arise because of combined chemicals including nerve agents that may have produced this type of neuropathy (discussed further under "Clinical Findings," below).
In general it has been very difficult to produce delayed neuropathy in animals using nerve agents. Doses vary in excess of lethal levels, requiring pretreatment and treatment (Gordon et al., 1983). The natural substrate of NTE and the detailed mechanism of toxicity are not known. Toxicity only occurs when a sufficiently large amount of the enzyme is inhibited. There is no known treatment. The enzyme is widely distributed in the nervous system and has also been demonstrated in lymphocytes and platelets, which have been used in screening and toxicity studies (Bertoncin et al., 1985; Lotti, 1991). The hen has become the standard research animal for NTE studies (e.g., Olajos, DeCaprio, and Rosenblum, 1978).
Johnson (1972) noted that only a tiny amount of toxin was required to produce an effect, the rest being dissipated via nonspecific reactions and degradation mechanisms. He raised concern that if another compound overloaded or blocked such pathways, the threshold dose of the neurotoxic compound would decrease.
Phenylmethylsulfonyl fluoride has been used as a pretreatment protective of NTE but increases toxicity if given after a NTE inhibitory agent, such as TOCP or mipafox (Pope and Padilla, 1990). There has been speculation that administration of PB after a toxic exposure might have the same effect (Haley, Kurt, and Horn, 1997; Halley, Horn, et al., 1997; Halley and Kurt, 1997).
The temporal properties of delayed neuropathy are complex. Classically, after an acute exposure to TOCP or DFP at sufficient doses, there is a 10- to 14-day delay before onset of signs and symptoms. In human TOCP cases, there is weakness and then paralysis, chiefly involving the lower extremities. There is degeneration of axons both peripherally and in the spinal cord. Recovery is rare but does occur (Hayes, 1982). Animal studies with organophosphate chemicals have shown that cumulative effects can produce such lesions; in some studies, doses six weeks apart were cumulatively able to produce the neuropathy (Lotti, 1991). Although no reports of the effects of combinations of different chemicals that inhibit NTE emerged, there seems to be some potential for complex interactions.
The classical findings of delayed neurotoxicity have generally been found from the medulla to the periphery (Abou Donia, 1981). Some possibility remains that higher brain centers might be affected. Prendergast, Terry, and Buccafusco (1997, p. 116), point out in the review section of their paper that impairment of AChE does not predict cognitive impairment well in animals and suggest that NTE, which has also been associated with cognitive impairment, might be involved. They did not measure this enzyme in their studies
The lipophilic nature of the nerve agents indicates that, as a group, they can readily penetrate the skin, lung, and gastrointestinal tract and, after entering the circulation, can be widely distributed, largely according to regional blood flow. The considerable species variation in sensitivity to these agents appears to reflect differences in the amount and distribution of nonspecific esterases that can bind the agents (Somani, 1992), although there are species differences in AChE affinity for organophosphate agents (Wang and Murphy, 1982).
Dermal
The skin does provide some degree of environmental protection, particularly against vapors. Military agents vary in the threat they pose. Tabun and sarin are rather volatile, and high concentrations (vastly higher than toxic respiratory doses) are required to produce toxicity through the skin by vapor exposure. Humans exposed to CTs of 1,000 to 1,300 mg-min/m3 of tabun showed only a decline in serum and red cell cholinesterase. Even at a CT of 2,000 mg-min/m3, subjects had no symptoms (Krakow and Firth, 1949). The NRC's Committee on Toxicology (NRC, 1997) cited 1951 work by McGrath in which humans were exposed to sarin vapor at CTs of 190 to 1,010 mg-min/m3 without lowering blood enzyme levels. Exposure to levels of 1,225 to 1,850 mg-min/m3 resulted in declines of ChE from 31 to 90 percent. No illness occurred, but two of nine subjects showed sweating. The committee considered 1,200mg-min/m3 to be a threshold effect ECT50 exposure.
Liquid agents applied to the skin are readily absorbed (Blank et al., 1957), but for volatile agents, such as sarin, evaporation reduces the amount of agent available for absorption (Grob et al., 1953). Applying 5 mg/kg of tabun to the skin of volunteers produced no illness but resulted in a 30-percent fall in AChE and notable local sweating (Freeman et al., 1954).
Chlorpyrifos, a pesticide of some Gulf War interest, is very poorly absorbed through the skin (Nolan et al., 1984). Some agents of intermittent volatility (e.g., soman and cyclosarin) present a greater hazard, and VX presents the greatest percutaneous hazard (Sim, 1962).
People show distinct regional differences in skin absorption; for VX, the highest rates are on the head and neck (Sim, 1962). Moisture, heat, and abrasions can increase agent transfer, while the total area exposed is important (Blank et al., 1957). The dermal toxic dose for many agents is considerably higher than for parenteral or respiratory exposure, reflecting not only evaporative and mechanical losses but also the ability of the skin and underlying tissues to bind agents and to inactivate them enzymatically (Fredriksson, 1969). For example, the skin dose of VX required to reduce red-cell AChE by 50 percent in humans is 32 µg/kg, while the intravenous effect is attained with 1 µg/kg. VX, on the other hand, is not hydrolyzed efficiently in the skin.
Ocular
Agents can be readily absorbed from the conjunctival sac and the eye (Grob and Harvey, 1958). Theoretically, dangerous amounts of agent could be absorbed as droplets by this route, but no data suggesting this is likely. The marked local effects of miosis, dim vision, impaired night vision, headache, lethargy, and impaired accommodation are predominant features of eye toxicity.
Respiratory
Vapors and aerosols are well absorbed from the lung. Oberst et al. (1959, 1968) demonstrated that humans exposed at rest to doses of sarin retained 89 percent of the inspired agent, less (79.5 percent) if exercising. The authors noted that CTs were not highly reliable indicators of toxicity. CTs ranging from 7 to 9.7 mg-min/m3 (exercising men) and 33 to 42.6 mg-min/m3 of sarin all produced similar absorbed doses. Particles with adsorbed agent can also be dangerous by this route (Asset and Finklestein, 1951).
Gastrointestinal
Gastrointestinal exposure of animals has been extensively studied for organophosphate pesticides, but little has been done with nerve agents. One study suggests that gastrointestinal exposure to a nerve agent produces rapid and serious intoxication (Karakchiev, 1973). Human studies (Sidell and Groff, 1974) with 4 µg/kg of VX taken orally showed a rapid response with fall of red-cell AChE. Maximum inhibition was 70 percent at two to three hours, with mild gastrointestinal disturbances (colic, nausea, vomiting, and diarrhea). Systemic symptoms were rapid at 20 minutes in other studies of sarin (Grob and Harvey, 1953) and VX (Sim et al., 1971).
Metabolism
Intra-arterial administration of sarin (Grob and Harvey, 1953) at 6 µg/kg showed the agent would pass the capillary bed with immediate symptoms and gradual decline of red-cell AChE to 28 percent over one hour. Smaller doses (3 to 4 µg/kg) produced no symptoms and only a minimal decline in red-cell AChE, suggesting that such lower levels may be detoxified.
Animal data show that rapidly acting agents at high dosage levels are not cleared effectively but that there are detoxification systems capable of dealing with lower levels of challenge (Somani, 1992; Fonnum and Sterri, 1981). Guinea pigs metabolize sarin at a rate of 0.013 µg/kg/min and soman at a rate of 0.009 µg/kg/min (Somani, 1992, p. 89). The rates of metabolism probably vary, with the isomers involved as in the case of soman (Benschop et al., 1981, 1984; De Bisschop et al., 1985). Carboxyesterases are important in metabolizing sarin, soman, and tabun (Walday, Aas, and Fonnum, 1991; Maxwell, 1992), but these enzymes are quantitatively much more important in rodents than humans. Nonspecific enzymes in serum and liver (aliesterases) metabolize all agents (Somani, 1992, p. 91).
Female mice are known to have plasma butyrlcholinesterase levels about twice that of matched males, with carboxyesterase levels 1.3 times those of males. The detoxifying or protective effects of these enzymes were not detectable in comparisons of the brain AChE levels in males and females three hours after intraperitoneal injection of 4 mg/kg of DFP or 0.3 mg/kg of sarin (Tuovinen et al., 1997).[8]
Some forms of paroxonase (PON1) in the serum hydrolyze sarin and soman at a high rate, breaking the P-F bond (Davies, 1996). Serum anhydride hydrolase (parathionase) is also active at least against soman (Broomfield, 1992). Variations in the abundance of these enzymes in human populations may produce variations in sensitivity to agents (Mutch et al., 1992).
The serum and some tissues contain an enzyme butyrylcholinesterase (EC 3.1.1.8), whose main biological purpose is unknown. This enzyme can hydrolyze ACh. Nerve agents and carbamates bind to this enzyme. Studies of an Israeli soldier who was hypersensitive to PB showed a variant of this enzyme that had low affinity (1/20th normal) for PB and other cholinesterases (Lowenstein-Lichtenstein et al., 1995; Schwarz et al., 1995). This suggests a role for the normal enzyme in decreasing the effects of low doses of anticholinesterase agents.
Elimination Excretion
In the case of soman, there appears to be a deposit site, probably in muscle, that does not inactivate the agent but rather stores and later releases it in toxic form. No such phenomenon has been reported for other agents, but the possibility apparently has not been evaluated. The effect lasts hours not days (Van Helden and Wolthuis, 1983; Wolthuis, Benschop, and Berends, 1981; Van Helden, Berends, and Wolthuis, 1984).
In general, the agents disappear rapidly from the blood, with rapid formation of hydrolysis products. The main metabolic product of sarin, IMPA, remains in tissues, although a great deal is eliminated rapidly in the urine, with a half-life of 3.7 hours (Fleisher et al., 1969). Cyclosarin studies show it is hydrolyzed to a similar analog with a half-life of 9.9 hours. Soman metabolism is more complex and biphasic, with half-lives of 3.6 hours and 18 hours, and soman accumulates in the lung (Shih, McMonagle, et al., 1994). Pinacolyl phosphonic acid is a major soman metabolite. IMPA was detected in the urine of Japanese sarin casualties--in one case for a week (Nakajima, Sasaki, et al., 1998; Minami et al., 1998).
Distribution
Sublethal amounts of soman injected intravenously in mice yield only trace amounts in tissues after 1 minute, with most converted to pinacolyl phosphonic acid. Studies show distribution to blood, choroid plexus, and spinal fluid at 2 minutes, with a distinct concentration in hypoglossal and vestibular nuclei, and later in the thalamus and caudate nucleus (Traub et al., 1985).
Blood flow is a key determinant in distribution of all agents (Somani, 1992, p. 79), and the higher percentage of the cardiac output going to the brain increases the risk there. Sites at which the brain is active show increased metabolic activity associated with vasodilation and increased blood flow, raising the possibility that distribution of agent might vary according to brain activity at the time of exposure (Scremin, Shih, and Corcoran, 1991; Scremin and Jenden, 1996).
Regional inhibition of AChE has been used to examine distribution effects, as has alteration of receptor properties. For example, Churchill et al. (1984) report such alteration in the olfactory bulb, hippocampus, and cortex, and such inhibition seems to be long lasting (Shipley et al., 1985). The latter regions are important in memory, while limbic system involvement influences mood and activity. The effects show age differences in the distribution of inhibition (Shih, Penetar, et al., 1990), but the correlation with inhibited AChE and clinical findings is uncertain.
Although the carbamates, such as PB, do not bind to aliesterase (Somani, 1992), there is concern that PB might occupy other binding sites, rendering them unavailable to bind nerve agents and thus increasing the amount of agent available at sites where toxicity is manifested. However, in animal studies with VX and soman, the agent was not increased in the brains with PB pretreatment (Anderson et al., 1992).
Acute Exposures, Acute Effects
There is no information on thiosarin, and the information on cyclosarin is sparse. Table 5.4 gives lethality and incapacitation estimates for the others.[9] There has also been a no-effect estimate of a CT of 1.6 mg-min/m3 for VX (McNamara et al., 1973). However, a study cited by the Committee on Toxicology (NRC, 1997) could not identify an adverse level (no-observed-adverse-effect level) based on VX exposures ranging from 0.7 to 25 mg-min/m3 (cited report was Bramwell et al., 1963). Lethality and incapacity estimates for cyclosarin are now available based on the extensive review of the Subcommittee on Toxicity Values for Selected Nerve and Vesicant Agents (NAS, 1997), although these are sometimes based on analogies with better-known agents. Some of their findings are included in Table 5.4. Generally, the toxicity of cyclosarin falls between that of soman and sarin (Cresthull et al., 1957), although some authors credit it as being twice as toxic as sarin (Karakchiev, 1973). Appendix B gives effect estimates for other species and modes of exposure. It is of some note that female animals show greater sensitivity to nerve agents (Callaway and Blackburn, 1954, for example), and the rate of recovery of AChE is slower (Woodard et al., 1994). A follow-up study of some Tokyo subway sarin cases found subtle neurological deficiencies in female cases (compared to control) but not in the male cases (Yokoyama, Araki, et al., 1998a).
Estimates of Nerve Agent Lethality or Incapacitation to Humans
| Skin LD50 (mg liquid,70 kg man) | Respiratory LCT50(mg-min/m3) | Respiratory ICT50/Severe Effects(mg-min/m3) | |||||
| Agent | NAS | Somani | NAS | Somani | Other | NAS | Other |
| Tabun | <1,500 | 200-1,000 | <70 | 100-200 | 150-400 | <50 | 100-300 |
| Sarin | <1,700 | 100-500 | <35 | 50-100 | 70-100 | <25 | 15-75 |
| Soman | 350 | 50-300 | <35 | 25-50 | 50-80 | <25 | 5-25 |
| Cyclosarin | 350 | <35 | <25 | ||||
| VX | <<5 | 5-15 | <15 | 5-15 | 30-100 | 10 | 5-50 |
SOURCES: Somani (1992), p. 77; OSRD (1946), U.S. Army (1990), Karakchiev (1973), McNamara et al. (1973), Trask et al. (1959) (NRC, 1997).
Also of interest are fairly large-order variations in sensitivity to nerve agents at different times of the circadian cycle, as Elsmore (1981) showed in LD50s of rats given soman at intervals around the clock. Agents also disrupt circadian rhythms (Mougey et al., 1985). This might mean that nerve agents or pesticides might be more toxic to troops at night than in the day, when most human studies have been done.
The U.S. Army Surgeon General has established exposure limits to a number of nerve agents for workers (eight-hour exposures) and the civilian population (MMWR, 1988). Concentration limits were established for exposures (see Table 5.5).
Nerve Agent Exposure Limits Established by the U.S. Army Surgeon General
| General Population (72 hours, mg/m3) | Workers (8 hours, mg/m3) | |
| Tabun | 3 x 10-6 | 1 x 10-4 |
| Sarin | 3 x 10-6 | 1 x 10-4 |
| VX | 3 x 10-6 | 1 x 10-5 |
SOURCE:
MMWR (1988).
NOTE: The MMWR summary did not include soman and cyclosarin. The document
from the Surgeon General (DAMD17-85R0072, p. 49) on which the MMWR report was
based also shows an 8-hour time-weighted average soman of 3x10-5
mg-min/m3. The 8-hour time-weighted average for cyclosarin was the
same as for sarin.
Such exposure limits are selected both to avoid any clinical signs and generally to provide at least an order-of-magnitude safety margin. Estimated remote cumulative doses for the Khamisiyah release appear to be higher: 0.01296 mg-min/m3 (CIA, 1997).
Behavioral Effects
A study of 29 troops in the mid-1940s (Marrs et al., 1996) found that humans exposed to a CT of 28 mg-min/m3 of tabun had definite symptoms; all 29 had miosis and vomiting; 26 were depressed; 22 were fatigued; and some night performance was impaired.[10] Recovery took one week, so it seems unlikely a similar outbreak in the Gulf would have escaped medical notice. Low inhaled-dose exposure to sarin of 5 µg/kg (calculated from respiratory exposure) did not impair a variety of complex tasks.[11]
An investigation of dermally applied EA1701 (an early designator for VX) using a micrometer syringe, at several levels of exposure, found mood, thinking, and behavioral changes in 93 human volunteers exposed to VX at levels that did not produce gastrointestinal, respiratory, or muscle symptoms, although some experienced nausea. In that study, decreases in red cell AChE correlated with anxiety and decreased mental performance; the exposures were of several levels and some had no fall in AChE (Bowers et al., 1964). These volunteers are presumed to be included in the long-term follow-up study that found no long-term effects (NAS, 1985). The authors did not give the actual doses (perhaps for security reasons) but drew attention to the considerable mental effects without peripheral signs of cholinesterase inhibition.
Another study (Sidell, 1967), using intravenous VX at three levels (three subjects at the lowest, four next, and 18 at the highest), with a placebo control group (four subjects), found no significant blood pressure or heart rate changes. AChE showed 70 percent inhibition. Doses were 1.3 µg/kg, 1.4 µg/kg, and 1.5 µg/kg. There were few peripheral symptoms. Although eight were nauseated and four vomited, these symptoms took an hour to develop. Twelve subjects were dizzy or lightheaded, and nervousness was common. A number facility tests showed a significant decrease only in the 1.5-µg/kg group. Presumably, the volunteers noted above were included in the long-term follow-up study, which did not find long-term effects in volunteers (NAS, 1985), but no short-term follow-up was included in the reports, so the duration of the effects is uncertain.
Low doses of soman and sarin (1/40th to 1/9 LD50) alter rodent behavioral performance, although the animals appear well. They seemed anxious (i.e., they hesitated on some tasks and were less inquisitive), but only sarin impaired coordination and balance (Sirkka et al., 1990). Behavioral changes were also observed in rats (open field locomotion) given sign-free doses of soman (4 µg/kg) and sarin (20 µg/kg) intraperitoneally lasting over 12 hours (Nieminen et al., 1990).
Marmoset studies (Wolthuis, Groen, et al., 1995) of cholinesterase inhibitors showed little physiological response at low levels, although blood AChE was decreased. However, there was definite disruption in a number of tests (e.g., visual discrimination, eye-hand coordination, and choice-time). There were increases in no-attempt behavior (i.e., failures to respond to rewards). PB had more behavioral effects than had been expected, given that it does not usually pass the blood-brain barrier. The performance decrements took place at levels of agent without overt clinical signs. There have been many studies of animal performance in response to nerve agents, with an emphasis on soman (Hartgraves and Murphy, 1992). At 0.5 LD50, soman and VX were more disruptive of performance than tabun and sarin (Mays, 1985).
Ocular Effects
The eye is sensitive to vapor or aerosols, with clinically and operationally important effects occurring at low levels (OSRD, 1946; NAS, 1997; Sim, 1956). These effects should have appeared if there were significant low-level exposures in the Gulf. Miosis refers to constriction of the pupils but is usually associated with a constellation of other problems: dim vision, pain, impaired night vision, difficulty focusing, and appearance of eye inflammation. CTs for miosis (mg-min/m3) are 20 for tabun, 2 for sarin, 0.1 for soman, and 0.09 for VX (OSRD, 1946; Karakchiev, 1973; McNamara et al., 1973; Sim, 1956; Johns 1952). Human night vision is impaired via a retinal effect at 5 CT of sarin (Rubin and Goldberg, 1957b). The no-effect level for VX is 0.02 mg-min/m3 (McNamara et al., 1973). The recent report of the Subcommittee on Toxicity Values for Selected Nerve and Vesicant Agents (NAS, 1997), which used multiple sources that may have been different from the above citations, determined the CT for miosis of sarin to be 0.5 mg-min/m3, a much lower figure than that noted above, although they noted that there were also reports of no effects at this level of exposure.
Dermal Effects
Dermal exposures generally require higher doses to generate the effects seen through other routes (see Appendix B). They are associated with slower onset of symptoms, fewer eye and respiratory symptoms, more cardiovascular symptoms, and nervous system symptoms. Reducing blood AChE by 50 percent required total doses of 400 mg of sarin, 65 mg of soman, and 30 mg of cyclosarin (Marrs et al., 1996; note the difference in dermal effect between sarin and cyclosarin). Parenteral (e.g., intravenous) and gastrointestinal routes of exposure were reviewed but do not seem relevant to Gulf War situations. Sweating is the common marker of dermal exposure and can be rather persistent.
Combined Effects
Tabun and mustard show a marked increase in toxicity and lethality when animals are exposed to both, and serum cholinesterase recovers more slowly than when the agents are used singly (Krustanov, 1962). The recognition of possible mixed use of sarin and cyclosarin prompted study of their combined toxicity; animals did not show unique toxicity, and therapy with standard measures was satisfactory (Clement, 1994). (See the earlier discussion on treatment mentioning the resistance of cyclosarin to oxime.)
Stress and Steroid Effects
How stress influences the effects of agents has not been studied extensively. Adrenalectomy did not alter the toxicity of sarin and soman in Wistar rats. Pretreatment with ACTH, adrenal cortical extract, cortisone, prednisolone, or corticosterone did not decrease soman toxicity. Soman toxicity was significantly decreased by pretreatment with prednisolone or cortisone plus atropine, compared to atropine alone (Stabile, 1967).
Modulation by Pretreatment
Humans exposed to a CT of 5 mg-min/m3 (30 min) of sarin after PB pretreatment had an altered miosis course with less conjunctival irritation and a shorter course of symptoms (i.e., two to three days versus seven to ten days) (Gall, 1981). This demonstrates that the clinical response of pretreated persons to low doses of agents may be modified by the pretreatment, possibly decreasing or preventing some of the signs and symptoms.
There has been concern, however, that pretreatment medications might enhance the toxicity of some agents. Physostigmine after DFP did not protect but rather enhanced toxicity (Koster, 1946).
There are limited indications that PB, not followed by treatment (e.g., with atropine or oxime), may decrease the duration and severity of symptoms and perhaps their occurrence in humans and animals exposed to low doses of an agent (e.g., sarin). Husain, Vijayaraghavan, et al. (1993) showed that sign-free PB and physostigmine pretreatments, also not followed by any treatment, provided a definite, favorable modification of the pulmonary function decreases in rats exposed to a CT of 51 mg-min/m3 of sarin. Rats are less sensitive to sarin than humans; the LCT50 for the rat is about 220 (Callaway and Blackburn, 1954), and the estimated LCT50 for humans is about 75 mg-min/m3 (NAS, 1997). Related studies by the same group (Vijayaraghavan et al., 1992) showed that, in rats exposed to sarin at a CT of 51 mg/m3 aerosol, pretreatment with carbamates, PB (0.075 mg/kg intramuscular), or a "symptom free" dose of physostigmine 20 minutes before sarin exposure protected lung AChE and increased survival. Physostigmine afforded better results. No treatment was given.
Longer Exposures and Tolerance
There is less information about chronic exposures, especially with measured doses. There are no reported exposure levels in the accidental occupational exposures reviewed, some of which may have reflected low-level exposure. Hartgraves and Murphy (1992) provide a substantial review of the behavioral effects of low-dose exposures to agents, some of which were subchronic or chronic. Chronic low doses of soman impaired primate responses, but the responses were not exacerbated by physostigmine pretreatment (Blick et al., 1993).
Animals and humans exposed repeatedly to sublethal levels of anticholinesterase compounds (inhibitors, such as nerve agents, organophosphate pesticides, drugs, carbamates, and carbamate pesticides) over time (days to a week) develop a condition known as tolerance, in which further administration of the inhibitor does not produce further signs and symptoms of exposure. Animal models have also shown behavioral tolerance to sustained sublethal exposures to DFP (Costa et al., 1982; Modrow and McDonough, 1986; Russell et al., 1975; Wolthuis, Philippens, and Vanwersch, 1991; Chippendale et al., 1972). Behavioral tolerance to soman in rats was seen, although performance decrements were noted on days of soman administration (Russell, et al., 1986). Doses of 35 µg/kg were given subcutaneously three times a week for four weeks (Modrow and McDonough, 1986). Dogs exposed to 25 µg/kg/day of sarin vapor for five days were symptomatic but showed signs of developing tolerance (Cresthull et al., 1960). A large, long-term study designed to simulate occupational exposures used beagles, exposing them daily to 10 mg-min/m3 of sarin for six months (Jacobson et al., 1959), which resulted in some illness, no direct mortality, signs of tolerance, and full recovery after the end of study.
Russell et al. (1986), showed that prolonged administration of soman (11 doses over 22 days, 35 µg/kg--0.3 log of the LD50) produced few signs of toxicity, although body temperature fell initially and then showed tolerance. Hypoalgesia continued, but tolerance was shown after initial decrements for a variety of temporal and performance activities. Brain AChE levels stabilized during the study despite continued administration of soman, implying some compensating regulatory activity (Russell, et al., 1986).
In contrast a single sublethal dose of soman in rats (100 to 150 µg/kg, intramuscular) did not produce seizures immediately but greatly altered spontaneous motor activity and test performance lasting for over 21 days. Some animals were very excitable and developed seizures when handled (Haggerty et al., 1986).
Nonspecificity
Tolerant organisms show decreased response not only to the inducing chemical but also to other anticholinesterases and cholinergic compounds such as carbachol and oxytremorine. They also show increased sensitivity to the effects of antagonists, such as atropine (Costa et al., 1982; Modrow and McDonough, 1986).
It has been shown that tolerance to the organophosphate pesticide disulfoton and the agent DFP can be induced by administering small doses that do not produce any overt signs of toxicity (Schwab and Murphy, 1981). A similar finding has been observed in humans taking the inhibitor echothiophate for glaucoma (DeRoetth et al., 1965). Tolerance has been seen in pesticide workers (Hayes, 1982) and is the probable explanation for the production and laboratory workers in the U.S. nerve agent program who had very low levels of cholinesterase but who reported no symptoms (Freeman et al., 1956; Holmes, 1959).
Extensive research has excluded increased metabolic clearance of the inhibitors as an explanation of tolerance. The uptake of choline and synthesis of ACh in the presynaptic tissues is not impaired (Costa et al., 1982).
Receptors
There are two main classes of receptors for ACh: muscarinic (with three subgroups of differing affinities) and nicotinic. Peripheral tissues, such as the gastrointestinal and pulmonary systems, are muscarinic, while skeletal muscles are nicotinic. The central nervous system contains both types of receptors, but their role is less well understood than in peripheral tissues and the autonomic nervous system. The receptors are primarily found on postsynaptic membranes, although there are some presynaptic muscarinic receptors (Costa et al., 1982).
Decreases in the abundance of both muscarinic and nicotinic receptors in response to sustained exposure to anticholinesterases has been demonstrated and seems to be the paramount mechanism of tolerance (Costa et al., 1982; Schwab et al., 1983; Bartholomew et al., 1985). Receptor decrease has been seen in tissue cultures, as well as in vivo. There may be additional mechanisms distal to the synapse involved in tolerance (Schwab et al., 1983).
Muscarinic receptors have been the most studied. Their abundance is decreased by chronic exposure to anticholinesterases or direct-acting cholinergic compounds (downregulation), while the binding affinity of the receptors is not altered (Costa et al., 1982). Indications are that receptors are internalized within the cell much as ligand-bound insulin receptors are. De novo synthesis of new receptors is required for recovery of normal abundance of receptors. In addition to decreases in receptor abundance, the function of the remaining receptors is altered, with decreased binding of agonists and antagonists in animals tolerant of organophosphate pesticides (Schwab et al., 1983; Costa et al., 1982; Schwab et al., 1981).
Nicotinic receptors appear to be more stable, although desensitization of nicotinic cholinergic motor end plates is fairly rapid. Other nicotinic receptors are slower to decrease than muscarinic receptors, although downregulation does occur. (Buccafusco et al., 1997).
The extent of downregulation in different parts of the central nervous system varies considerably--e.g., the brain stem showed much less downregulation in muscarinic receptors of rodents than did the striatum and cortex (Bartholomew et al., 1985). A single sublethal dose of soman in rats produced a reversible decline in muscarinic receptors of the telencephalon but an irreversible decline in the pyriform cortex (Pazdernik et al., 1986). Downregulation from low levels of anticholinesterases has also been demonstrated in vitro (isolated synaptic membranes of the bovine caudate nuclei) (Volpe, Biagioni, and Marquis, 1985).
Carbamates also induce tolerance, although their binding to AChE is reversible. Short-acting carbamates, such as physostigmine, require sustained infusions to induce tolerance. Tolerance to neostigmine has been shown in people and animals. The mechanisms of tolerance with carbamates may be more complex than with organophosphate agents, but downregulation of muscarinic receptors has been shown with them.
Duration
For indirectly acting cholinergics (anticholinesterases), the duration of cholinesterase inhibition is the critical factor, since anticholinesterase level appears to be the ultimate regulator of sensitivity to these chemicals. Prolonged exposure to anticholinesterases produces a decline in receptor abundance (Schwab et al., 1983).
The separate RAND report on PB (Golomb, 1999) also considers receptor effects. Most studies of this compound as a pretreatment have been fairly short, many of three to five days of exposure (Gall, 1981). Prolonged use of this drug, which is fairly long acting (the oral half-life is about four hours), might induce tolerance, at least in peripheral tissues, thus decreasing the effects of nerve agents by a mechanism additional to its reversible binding to AChE.
As noted previously, there is now reason to suspect that, under severe stress conditions, PB can pass the blood brain barrier and can act centrally as well as peripherally in downregulating receptors (Friedman et al., 1996).
Tolerance May Not Be Beneficial
The decrease of muscarinic and nicotinic receptors in the brains of animals tolerant to organophosphates raises the possibility that the balance of neuronal connections might be modified, with effects on higher brain functions. Such effects, rather than being protective, might represent a pathological process (Taylor et al., 1979). Animal studies have shown correlations of reduced memory and decreased abundance of brain nicotinic receptors (Gattu and Buccafusco, 1997).
Research inspired by illnesses in Gulf War veterans (Buccafusco et al., 1997; Wickelgren, 1997) has demonstrated decreased (over 50 percent) abundance of nicotinic receptors in cortical striatal and hippocampal neurons of rats exposed to sign-free doses of DFP (0.25 mg/kg/day for 14 days). Three weeks after withdrawal of DFP, treated animals showed impaired learning of a water maze, although previously trained animals retained their maze memories. There was no recovery of hippocampal nicotinic receptors three weeks after stopping DFP. A report in Science (Wickelgren, 1997, p. 1404) indicates that DFP-treated rats given nicotine before the water maze test learned adequately. Related studies in nonhuman primates given DFP 0.01 mg/kg/day for 25 days did not show altered performance in a delayed-matching-to-sample task, although red-cell AChE fell to 76 percent of control. Similar results occurred with 0.015 mg/kg/day for 15 days. Impaired performance was encountered at levels of 0.02 mg/kg/day, but these animals showed mild overt toxicity (Prendergast, 1998).
Induction of C-fos
The emerging picture of how cholinesterase inhibitors rapidly induce the expression of the transcription factor for c-fos points the way for possible long-term effects and added mechanisms of tolerance. It also appears that severe stress-induced release of ACh in animal models can also induce c-fos expression. The changes in gene expression initially enhance and later inhibit neuronal excitability mediated by muscarinic receptors. C-fos, an early immediate transcription factor, mediates selective regulatory effects on long-lasting activities of genes involved in ACh metabolism. This appears to create a situation in which the effects of cholinesterase inhibitors might persist long after the agents are no longer present (Kaufer et al., 1998). The role of c-fos and other intermediate early genes (IEGs), such as c-Jun, that seem to play an important role in translating stimuli into longer-term adaptive responses of cells is vast and complex and eventually may explain the longer-term effects of brief chemical exposures. A brief summary of recent information on IEGs and c-fos can be found in Appendix C.
The finding that cholinergic stimulation or stress can induce the expression of immediate early genes, such as c-fos, is not surprising in view of the variety of stimuli that activate this transforming factor. The possibility of this proto-oncogene playing a role in producing long-term effects from exposure to agents that produce cholinergic activity seems great, but the details remain to be demonstrated. This might be the mechanism by which short-term exposures produce long-term effects without killing large numbers of cells.
Although stress of various kinds increases c-fos, the regions involved vary with the stress model employed. It remains to be demonstrated which, if any, cholinergic stimuli produce effects convergent with stress responses.
Delayed Effects
The clinical manifestations of typical OPIDN begin about two weeks after exposure, with a progressive peripheral neuropathy, which can also involve the central nervous system, with axon degeneration and later demyelination. Sustained lower doses are as toxic as single exposures to larger doses, provided some threshold is crossed, with chronic dermal exposure being suspect (Cherniack et al., 1986; Hayes, 1982). Additive effects with long intervals between exposures (up to six weeks) have been demonstrated (Hayes, 1982; Davies and Holland, 1972; Abou-Donia, 1981).[12] No human case of typical delayed neurotoxicity arising from nerve agents has been reported. Sarin in repeated sublethal exposures did produce a typical neuropathy in mice (Husain, Vijayaraghavan, et al., 1993).
Sarin can produce delayed neurotoxicity in animals. However, very high levels of acute exposure (30 to 60 times LD50) are required to produce the effect in hens protected from cholinesterase toxicity by treatments (PB, oximes, atropine) (Gordon et al., 1983). It was recognized that some humans with analogously high acute doses might survive as battle casualty treatment improved, possibly resulting in delayed neurotoxicity. But after examining sustained exposure of rodents to sarin aerosol, Husain (Husain, Vijayaraghavan, et al., 1993; Husain and Pant, 1994) questioned the impression that it took very high levels of repeated challenge with sarin to produce delayed toxicity. Husain et al. did not produce delayed neuropathy in Wistar albino rats with daily 20-minute sarin exposures for ten days (250 mg-min/m3 exposures). However, white albino mice given 5 mg/m3 of sarin for 20 minutes (i.e., 100 CT) daily for ten days developed classic delayed neurotoxicity (weakness, ataxia, and twitching) beginning at day 14 and confirmed by tissue pathology. The mice were not initially ill from sarin exposure and manifested no symptoms of anticholinesterase intoxication. The doses would be lethal in the range for humans (NRC, 1997). Since the hen has become the standard animal for OPIDN studies, there is less information about the effects in mice.
AChE levels in the brain were reduced by only 19 percent. Platelets and the spinal cord showed marked decreases in NTE levels, although less than in mipafox controls. This report has not been replicated. Rodents have generally been considered resistant to delayed neuropathy and have been used in research on the subject. The authors did not discuss why the species differed. The hen has become the standard animal for OPIDN toxicity studies with less information from mice.
Studies of the effects of the isomers of soman and sarin hinted at the possibility of nerve agents producing NTE effects at lower levels of exposure. A trend of increased inhibition of lymphocyte NTE in hens exposed to Sarin II (an isomer of sarin) suggested that longer exposures at lower levels might cause cumulative toxicity (Crowell et al., 1989). The P+ isomer of soman is a potent inhibitor of NTE, suggesting that this isomer alone could produce neuropathy at unprotected LD50 levels (Johnson, Read, and Benschop, 1985).
There has been considerable comment that this "low-level" exposure raises the possibility that sustained exposures in humans might result in neuropathy. The exposures in the study are at the upper range of LCT50 for rodents. Rodents, however, are more resistant to nerve agents than humans, and the mice in question did not require any treatment. However, the results are not congruent with earlier studies with sarin in hens or with the experience in dogs (which are not a standard animal for NTE research). Dogs do develop delayed neuropathy from DFP (Johnson, 1975). However, chronically exposing dogs to sarin vapor did not produce any neuropathy (Jacobson et al., 1959). The main weight of information makes it difficult to attribute delayed neuropathy to sarin or cyclosarin, given the very low levels calculated for the Khamisiyah release.
The studies of Gordon et al. (1983) demonstrated that soman and tabun did not produce delayed neuropathy at doses 38 times the LD50 of soman and 82 times the LD50 of tabun. In these studies, animals were provided the appropriate chemical therapy to enhance survivability following supra doses of agent. The same studies looked at the molar concentrations of agents required to inhibit in vitro 50 percent of the two enzymes, NTE and AChE, and calculated the ratios of the two (Table 5.6). The presumption was that the larger the number, the greater the likelihood of encountering delayed neurotoxicity from the agent in question. The results of other studies are summarized in Table 5.7.
A measure of the complexity and difficulty of this field is demonstrated by the Lenz et al. (1996) finding that sustained infusion of high daily doses (57 µg/kg/day) of VX in rats not provided chemical therapy reduced brain NTE by 90 percent at 14 days. No study of pathology or clinical response was reported. VX was not previously thought to be capable of significant NTE affects.
Severity-Sequelae Relationships
It is uncertain whether sequelae always correlate with severity at onset, although it seems intuitively obvious. Holmes (1959) reviewed the experiences of a group of workers exposed to sarin at various levels (although none so severely as to suffer seizures). He found that the more seriously exposed were ill longer. However, Stephens et al. (1996), in a study of groups exposed to organophosphate pesticides, found no correlation between acute exposure effects and the severity of performance shortfalls in later neuropsychological testing. In the reports of accidental cases, there are instances of patients with mild initial symptoms who had rather protracted later symptoms (Gaon and Werne, 1955; Brody and Gammill, 1954; Craig and Freeman, 1953).
Molar Concentration of Agent Required to Inhibit Half of Enzyme Activity
| NTE | AChE | AChE/NTE | |
| DFP | 9.3x10-7 | 1.05x10-6 | 1.1300 |
| Sarin | 3.38x10-7 | 1.9x10-9 | 0.0056 |
| Soman | 3.77x10-7 | 4.6x10-10 | 0.0012 |
| Tabun | 6.65x10-6 | 3.5x10-9 | 0.0005 |
| VX | 2.5x10-4 | 3.6x10-10 | 10-6 |
SOURCE:
Reprinted by permission from Archives of Taxicology, Gordon et al.
(1983), pp. 71-82. ©1983 Springer-Verlag, Berlin, Germany.
NOTE: Cyclosarin was not studied.
Results of Other Delayed-Neuropathy Studies
| Tabun (GA) | A 90-day study in hens at maximum tolerated dose (plus atropine) did not demonstrate delayed neuropathy (Willems et al. 1984). |
| Soman (GD) | To date, only repeated doses on the order of 150 times the LD50 have produced delayed neuropathy (Gordon et al., 1983; Willems, Nicaise, and De Bisschop, 1984). In a 90-day subchronic study at daily doses of GD insufficient to produce clinical signs, no delayed clinical or histological neuropathy resulted (Hayward et al., 1990). |
| Cyclosarin (GF) | GF has not been studied as extensively as the other agents. Vranken, De Bisschop, and Willems (1982) demonstrated that GF in vitro is a very potent inhibitor of NTE, but at doses where some lethality was encountered, no neuropathy occurred (Willems, et al., 1983). |
| VX | Most in vitro and in vivo studies fail to suggest that VX has any delayed neuropathic potential (Gordon et al., 1983; Vranken, De Bisschop, and Willems, 1982; Willems, Nicaise, and De Bisschop, 1983). However, Lenz, Maxwell, and Austin, (1996) raises some doubts about this conclusion. |
| Thiosarin | There is no information about this agent. |
This section concentrates on ocular, dermal, and respiratory exposure routes, emphasizing lower levels of exposure and clinical severity and information about long-term consequences and any patterns of illness similar to illnesses in Gulf War veterans. Earlier case reports suggest that clinicians did not expect long-term symptoms to arise from very mild exposures and sometimes considered alternate explanations for such patients when encountered (e.g., chronic anxiety). There is not much information about repeated or sustained exposures.
The amount of information about specific agents is uneven. There is much information about human exposures to sarin, but much less for tabun, soman, and VX. The only information available on human experience with cyclosarin is from secondary sources. The 1982 NAS review indicated that 27 volunteers were exposed to cyclosarin and that there were apparently both some sensory and oxime treatment studies. No information about thiosarin is available.
For nerve agents and pesticides, it is not always easy for the clinician to determine if an exposure has occurred, and AChE levels correlate poorly with the clinical findings. Holmes (1959) stated, with respect to a sarin production facility, that
The examiner frequently asks himself the question "Is this a true exposure?" If so how serious is it? Except possibly for miosis, there seems to be no single symptom which occurs in every exposure. Only in the more severe exposures was miosis present in every instance. It is apparent that in milder exposures a single symptom related to a system occurs. When this happens it raises a question as to whether a particular symptom is a result of exposure or represents a symptom related to some other medical problem, such as a cold etc. When several symptoms related to a system occur, there is little doubt in the examiners mind both that this is a true exposure and in all probability a fairly severe exposure. When there is a scattering of symptoms related to different systems then the question arises if an exposure has occurred. Correlation with acetylcholinesterase is unreliable--in many instances a person is judged to have a mild exposure when the red cell acetylcholinesterase shows a greater drop than expected.Acute Effects
Table 5.8 summarizes signs and symptoms arising from nerve agent exposure by several routes. No clinical differences are expected between various nerve agents. The clinical signs and symptoms from other organophosphate pesticide exposure are also quite similar (Hayes 1982; Namba et al., 1971). Table 5.9 classifies the severity of poisoning and is the basis of classifications used through this report. The emergency department of a Japanese hospital in Tokyo (Okumura et al., 1996) has summarized its experience with 640 victims of the subway attack. The department treated and released 82 percent (528) of the cases. The detailed findings in this group have not been reported, but these cases were considered to have been full recoveries. Of the 111 patients (17.3 percent of those in emergency department) admitted, 107 were considered moderate cases (16.7 percent) and four were considered severe. One of these died. Table 5.10 summarizes the clinical findings in the moderate and severe cases. There were two deaths, one in the emergency department and one later and 638 patients were considered to have made a full recovery. The moderate cases were hospitalized for 2.4 days (mean).
Signs and Symptoms Following Short-Term Nerve Agent Exposure
| Site of Action | Signs and Symptoms |
| Ciliary body | Frontal headache, eye pain on focusing, blurred vision |
| Conjunctivae | Hyperemia |
| Nasal mucous membranes | Rhinorrhea, hyperemia, but this may also be present after systemic absorption |
| Following systemic absorption of liquid and prolonged vapor exposure | |
| Bronchial tree | Tightness in chest sometimes with prolonged wheezing, expiration suggestive of bronchoconstriction or increased secretion, dyspnea, slight pain in chest, increased bronchial secretion, cough, pulmonary edema, cyanosis |
| Gastrointestinal | Anorexia, nausea, vomiting, abdominal cramps, epigastric and substernal tightness (cardiospasm) with "heartburn" and eructation, diarrhea, tenesmus, involuntary defecation |
| Sweat glands | Increased sweating |
| Salivary glands | Increased salivation |
| Lachrymal glands | Increased lachrymation |
| Heart | Bradycardia |
| Pupils | Slight miosis, sometimes unequal, later maximal miosis (pinpoint pupils); sometimes mydriasis is observed |
| Bladder | Frequent, involuntary microurination |
| Striated muscle | Easy fatigue, mild weakness, muscular twitching, fasciculation cramps, generalized weakness including muscles of respiration with dyspnea and cyanosis |
| Sympathetic ganglia | Pallor, occasional elevation of blood pressure |
| Central nervous system | Ataxia, generalized weakness, coma with absence of reflexes, Cheyne-Stokes respiration, convulsions, depression of respiratory and circulatory centers resulting in dyspnea and fall in blood pressure; emotional effects very often occur |
SOURCE: NATO (1973).
Severity Classifications for Organophosphate Pesticide Poisoning
| Term | Description |
| Latent | No clinical manifestations |
| Serum cholinesterase activity is 50 to 90 percent of normal | |
| Mild | Fatigue, headache, dizziness, nausea and vomiting, increased salivation and sweating, chest tightness, abdominal cramps or diarrhea, can still walk, numbness of extremities |
| Serum cholinesterase activity is 20 to 50 percent of normal | |
| Moderate | Unable to walk, generalized weakness, difficulty talking and fasciculations, in addition to the symptoms but more miosis associated with mild poisoning |
| Serum cholinesterase activity is 10 to 20 percent of normal | |
| Severe | Unconsciousness, loss of pupillary light reflex, fasciculations, flaccid paralysis, moist rales in lungs, seizures, respiratory difficulties and cyanosis, secretion from mouth and nose |
| Serum cholinesterase activity is less than 10 percent of normal |
SOURCE: Modified from Namba et al. (1971); AD Little (1986), Ch. 5.
NOTE: Namba's classification, developed from experience with parathion and methyl parathion poisoning, has been adopted for classification of nerve agent exposures.
Signs and Symptoms in Patients with Moderate to Severe Sarin Exposure
| Patients | ||||
| Sign or Symptom | Number | %(n=111) | ||
| Eye | Miosis | 110 | 99.0 | |
| Eye pain | 50 | 45.0 | ||
| Blurred vision | 44 | 39.6 | ||
| Dim vision | 42 | 37.8 | ||
| Conjunctival injection | 30 | 27.0 | ||
| Tearing | 10 | 9.0 | ||
| Chest | Dyspnea | 70 | 63.1 | |
| Cough | 38 | 34.2 | ||
| Chest oppression | 29 | 26.1 | ||
| Wheezing | 7 | 6.3 | ||
| Tachypnea | 28 | 31.8a | ||
| Gastrointestinal tract | Nausea | 67 | 60.4 | |
| Vomiting | 41 | 36.9 | ||
| Diarrhea | 6 | 5.4 | ||
| Neurologic | Headache | 83 | 74.8 | |
| Weakness | 41 | 36.9 | ||
| Fasciculations | 26 | 23.4 | ||
| Numbness of extremities | 21 | 18.9 | ||
| Decrease of consciousness level | 19 | 17.1 | ||
| Vertigo and dizziness | 9 | 8.1 | ||
| Convulsion | 3 | 2.7 | ||
| Ear, nose and throat | Running nose | 28 | 25.2 | |
| Sneezing | 5 | 9.0 | ||
| Psychological | Agitation | 37 | 33.3 | |
SOURCE:
Okumura et al. (1996). ©1996 American College of Emergency Physicians
(ACEP). Reprinted by permission.
an=88.
Table 5.11 summarizes the mild cases encountered in the workers accidentally exposed to "G" agents (tabun and sarin) studied by Craig and Freeman (1953). These cases show a lesser prevalence of eye, gastrointestinal, and nervous system findings but more rhinorrhea than in the Japanese cases, but the Japanese cases were a single event, while Craig and Freeman summarized multiple events. Their table is somewhat misleading in that, as in the Japanese cases, miosis was the most consistent finding (48 of their 53 cases). Most of their cases recovered rapidly, 78 percent within two days, most of the remaining ten cases within one week, with one case still symptomatic at 20 days.
Incidence of Symptoms in Workers Accidentally Exposed to Tabun and Sarin (mild cases)
| Symptoms | Number of Workers | % Exposed Workers |
| Respiratory symptoms | 41 | 77 |
| Pressure sensation | 35 | |
| Localization not recorded | 17 | |
| Throat | 2 | |
| High sternum | 3 | |
| mid sternum | 9 | |
| low sternum | 3 | |
| not localized | 1 | |
| Cough | 20 | |
| Unproductive | 15 | |
| Productive | 5 | |
| Wheezing | 9 | |
| Inability to obtain a satisfactorily full inspiration | 8 | |
| Increased exertional dyspnea | 7 | |
| Dyspnea at rest | 1 | |
| Rhinorrhea | 31 | 58 |
| Eye Symptoms | 29 | 55 |
| Dim vision | 24 | |
| Impaired accommodation | 13 | |
| Pain on accommodation | 6 | |
| Central Nervous System Symptoms | 27 | 51 |
| Headache | 17 | |
| Headache as only CNS symptom | 7 | |
| Disturbed sleep | 13 | |
| Mood change | 12 | |
| Easily fatigued | 10 | |
| Increased perspiration | 3 | |
| Dizziness | 2 | |
| Gastrointestinal symptoms | 14 | 26 |
| Anorexia and/or nausea | 9 | |
| Increased GI activity | 6 | |
| Diarrheal stool | 6 | |
| Vomiting | 2 | |
| Miscellaneous | ||
| Unpleasant taste to tobacco | 10 | |
| Poor driving | 6 |
SOURCE: Craig and Freeman (1953).
Commonly, cases are stratified clinically according to severity, as follows:
Severe Intoxications. Several severe intoxications will be described, from accidents and the incidents in Japan to give a clinical picture of the problems that confronted military planners in the Gulf and to describe some of the sequelae of severe intoxications. Sidell (1973, 1974) reported two severe intoxications requiring hospitalization. The first was a 33-year-old man who sustained an accidental combined respiratory, cutaneous, and mucosal exposure to soman (less than 1 ml). He immediately decontaminated himself, was asymptomatic on arrival at the emergency room about five minutes later, but then collapsed. He was cyanotic with labored breathing, had a blood pressure of 180/80 mm Hg, and had a heart rate of 150/min. His conjunctivae were very inflamed; hehad marked oral and nasal secretions and widespread fasciculations. He was given intravenous atropine and 2-pyridine aldoxime methiodide (2-PAM). Cyanosis worsened, but he became conscious about 30 minutes later. Fasciculations continued, and he was restless, with nausea and vomiting, and electrocardiogram changes showed sinus tachycardia and then atrial fibrillation for 20 hours. His physical condition improved more rapidly than his psychiatric condition. He was observed to be depressed and withdrawn and had bad dreams that improved with scopolamine treatment. He had difficulty calculating. AChE remained low, but other laboratory tests were normal. By six weeks, he was back to his premorbid level and was doing well at six months.
The second was a 52-year-old man whose mask malfunctioned in a sarin-filled room. He noted respiratory difficulty; on arrival at the emergency room 5 to 10 minutes later, he was cyanotic and having seizures. He was given intravenous atropine and oxime and required respiratory assistance for apnea. Fasciculations were prominent, and he had marked wheezing, developed an S4 gallop, and later had electrocardiogram changes typical of ischemia. He resumed breathing one hour later. This patient was more alert at three hours and was able to ambulate at nine hours. He was rehospitalized four months later because of fatigue and dyspnea on exertion and was diagnosed with depression at six months. He died of myocardial infarction 18 months after exposure (Sidell, 1973, 1974).
There were two major sarin events in Japan, the first at Matsumoto in June 1994 (seven deaths and about 600 persons poisoned). The second was in Tokyo in March 1995 when sarin was released into the subway (11 died and 5,000 were poisoned) (Morita et al., 1995; Okumura et al., 1996). A third event was a man sprayed with VX (Nozaki and Aikawa, 1995).
The Japanese cases were well-documented. Some patients were comatose on admission, with miosis, seizures, fasciculations, flushing, tachycardia, hypotension, and respiratory distress; hypoxia was common. Many required intubation. Creatine kinase and glucose levels were elevated; many patients were acidotic at pH 6.8. Reports on secretions varied (Suzuki et al., 1995; Nozaki and Aikawa, 1995). Many recovered well, but some reported dysesthesias. One person had retrograde amnesia for 70 days (Hatta et al., 1996), and another was delirious and had hallucinations for over one week (Inoue, 1995). There were few deaths following hospitalization, but one man remained in a vegetative state in Matsumoto six months later.
The VX patient initially presented with blurred vision; seizures, fasciculation, and sweating followed. There was no miosis. The patient became cyanotic and was on a respirator for several days. He required atropine drip and intravenous diazepam. He was released after 15 days with brachial plexus neuropathy and antegrade and retrograde amnesia. Unlike the sarin cases, bradycardia had been prominent (Nozaki and Aikawa, 1995).
Moderate Exposures. At the hospital receiving the largest number of patients from the Japanese subway attack, 111 were categorized as severely or moderately injured (4 severe and 107 moderate) on admission (see Table 5.10). In these patients, miosis (99 percent) and headache (75 percent) were the most frequent symptoms, followed by dyspnea (63 percent) and nausea (60 percent); bradycardia was uncommon. Even at discharge, headache and eye pain were common. All but one made full recoveries (one severe case died), although 37 (33 percent of the 111) had acute stress disorders and four were diagnosed with PTSD at three months (Okumura et al., 1996). Of 213 patients seen initially at another Tokyo hospital, none had complaints at three-month follow-up (Yokoyama et al., 1996), similar to other Japanese reports.
In a three-week follow-up of some 117 mild and moderate cases in Matsumoto (Morita et al., 1995), the initial symptoms included rhinorrhea (78), headache (53), dark vision (52), sneezing (24), fatigability (18), dizziness (17), and nausea (14). Others reported diplopia, dysesthesia, vomiting, dysphagia, increased tearing, or gait disturbances. Most of these symptoms cleared by three weeks. At six months, five people visited hospitals regularly, with diverse complaints. One man with no history of lung disease was mildly hypoxic, and another had low-grade fevers.
Grob et al. (1953) reported a moderately severe reaction to percutaneous sarin in a volunteer. Illness was delayed several hours after exposure, lacked immediate eye or respiratory findings, and ran a protracted course with waves of recurring symptoms over four to five days. The dose was about 0.18 mg/kg through abraded skin. After 2-3/4 hours, there was local sweating; at 5-3/4 hours, the patient experienced general sweating, giddiness, and abdominal cramps. Blood pressure rose, and he was given atropine. One hour later, nausea, sweating, generalized weakness, and fatigability existed. He was short of breath with abdominal cramps but no wheezing. Maximum symptoms occurred at 10 to 11 hours, with dilated pupils and decreased vital capacity. He had mild symptoms with atropine at 13 hours, was fatigued and weak at 21 hours, and had insomnia and nightmares at 40 hours, finally recovering over the ensuing days.
Mild Exposures. Respiratory. Review of the inhalation exposures of sarin, tabun and some V-agent accidents supports the description in Wilson (1954) of mild intoxication, of symptoms that develop rapidly and then evolve over time:
The chief effects consist of a feeling of constriction in the throat and chest, a tendency to cough, and eyes that quickly become red and painful with contracted pupils such that the subject finds it painful to focus on near objects. A severe and persistent frontal headache usually follows, and he becomes dejected and not inclined to bother to do anything unless he must. At night he is restless and has difficulty getting to sleep, and when he does has vivid dreams and nightmares; these symptoms may last 3-5 days. With larger doses, there is anorexia, nausea, vomiting, abdominal pain, salivation and diarrhea. There is sweating and generalized muscle weakness and fasciculations. Psychological changes include restlessness, irritability and insomnia. The appearance of the symptoms bears no relation to the plasma cholinesterase activity.
The occupational experience related to sarin production and testing was extensive. Holmes (1959) analyzed 991 cases in two groups, stratified by four levels of red-cell cholinesterase levels, with initial and follow-up examinations. All but a few cases were mild, and there was little treatment. These cases did not have long-term follow-up of the whole group, but cases were followed enough to determine that 10 percent had symptoms lasting two or more weeks. Appendix B includes information from this extensive report.
Several other reports also shed light on this issue: Brody and Gammill (1954), 75 cases; Gaon and Werne (1955), 244 persons; Craig and Freeman (1953), 53 cases; and Finesinger et al. (1950), 40 cases. Sarin was the primary agent to which people were exposed, but Freeman also reported four tabun exposures and two combined tabun-and-sarin exposures. The reports arose from medical examinations of exposed persons and later follow-up exams. The symptoms are those shown in Table 5.11 (Brody and Gammill, 1954; Craig and Freeman, 1953, had similar findings). Data from Holmes (1959) suggest a tendency for persons with higher percentages of cholinesterase inhibition to have more symptoms and longer periods of illness. Gaon and Werne (1955) could not make such an assertion; their cases, which lasted over three weeks, showed 47 percent with no significant reduction in enzyme level. In a series of 182 cases, they reported that 106 (58.2 percent) had recovered in three days, 34 (18.6 percent) in one week, 19 (10.4 percent) in two weeks, but 23 (12.6 percent) took three weeks or more. They did not describe their follow-up process but documented two cases with persisting symptoms 10 to 11 months after exposure (recurring headaches, dizzy spells, fatigue, syncope, and weakness for one; memory problems, inattention, and fatigue in the second).
Table 5.11 gives an idea of the prevalence of symptoms, with eye problems being acutely the most prevalent, as was noted previously. Note the considerable percentages of symptoms related to the central nervous system, including mood changes, thinking problems, and fatigue.
Although most cases resolved within three days, some 12 percent of Gaon's cases persisted for three weeks or more (Gaon and Werne, 1955). Many e