effects. However, the effects are not disabling and are transient and reversible upon cessation of exposure.

AEGL-2 is the airborne concentration (expressed as ppm or mg/m³) of a substance above which it is predicted that the general population, including susceptible individuals, could experience irreversible or other serious, long-lasting adverse health effects or an impaired ability to escape.

AEGL-3 is the airborne concentration (expressed as ppm or mg/m³) of a substance above which it is predicted that the general population, including susceptible individuals, could experience life-threatening health effects or death.

Airborne concentrations below the AEGL-1 represent exposure concentrations that could produce mild and progressively increasing but transient and nondisabling odor, taste, and sensory irritation or certain asymptomatic, nonsensory effects. With increasing airborne concentrations above each AEGL, there is a progressive increase in the likelihood of occurrence and the severity of effects described for each corresponding AEGL. Although the AEGL values represent threshold concentrations for the general public, including susceptible subpopulations, such as infants, children, the elderly, persons with asthma, and those with other illnesses, it is recognized that individuals, subject to idiosyncratic responses, could experience the effects described at concentrations below the corresponding AEGL.

SUMMARY

arrhythmia. Death that occurs within 24 h of exposure is likely due to pulmonary damage (Lindberg et al. 1997). Lewisite-2 (L-2) and lewisite-3 (L-3) are co-products concurrently formed with L-1(Trammel 1992). L-1 yield is greater than 65%, and approximate yields of L-2 and L-3 are 7-10% and 4-12%, respectively (Lindberg et al. 1997). L-2 and L-3 are present in smaller quantities and have comparatively low volatility, so those compounds will be less toxicologically significant than L-1. Furthermore, the toxicity of L-2 and L-3 is comparable to L-1 (Lindberg et al. 1997). Therefore, AEGL values were derived for “lewisite”, rather than for the individual lewisite compounds, and are considered protective for L-1, L-2, and L-3 compounds.

Appropriate data were not available for deriving AEGL-1 values for lewisite. Odor cannot be used as a warning for potential exposure. For L-1, the odor threshold is reported to be between 14-23 mg/m³ (1.7-2.7 ppm), concentrations greater than those that are highly irritating and higher than the AEGL-2 and AEGL-3 values. Therefore, AEGL-1 values are not recommended.

No inhalation studies with both concentration and duration parameters and with effects consistent with the definition of AEGL-2 end points were available. Therefore, the AEGL-2 values for lewisite, were determined by taking a threefold reduction in the AEGL-3 values; the resulting values are considered to be estimated thresholds for irreversible effects (NRC 2001). The reduction approach is considered appropriate because of the steep concentration-response curve for mortality from lewisite. In studies with mice, the 10-min LC₅₀ (lethal concentration, 50% lethality) was 200 mg/m³ [24 ppm], and the 10-min LC₁₀₀ (lethal concentration, 100% lethality) was 240 mg/m³ [28 ppm]. In dogs, no deaths occurred after a 7.5-min exposure to lewisite at 126 mg/m³ [15 ppm], and the LC₅₀ was 176 mg/m³ [21 ppm]).

AEGL-3 values for lewisite were based on lethality data for L-1 in dogs (Armstrong 1923). Points of departure were the calculated LC₀₁ values: 38.7 mg/m³ (4.6 ppm) for the 10-min value, 14.0 mg/m³ (1.7 ppm) for the 30-min value, 7.4 mg/m³ (0.87 ppm) for the 1-h value, 2.1 mg/m³ (0.25 ppm) for the 4-h value, and 1.1 mg/m³ (0.13 ppm) for the 8-h value. The LC₀₁ values are considered estimated lethality thresholds. Interspecies and intraspecies uncertainty factors of 3 each were applied. The interspecies uncertainty factor of 3 is supported by data suggesting little species variability with regard to lethality from inhalation exposure to lewisite; C × T values are relatively constant across species, except for the guinea pig, and the interspecies uncertainty factor of 3 encompasses the two- to three-fold difference in sensitivity between guinea pigs and rats, mice, rabbits, dogs, and goats. The intraspecies uncertainty factor of 3 is supported by the steep concentration-response curve with regard to lethality, which implies limited intraspecies variation. Thus, the total uncertainty factor was 10.

TABLE 5-1 AEGL Values for Lewisite

Abbreviations: LC₀₁, lethal concentration, 1% lethality; NR, not recommended.

^aAbsence of AEGL-1 values does not imply that exposure below the AEGL-2 values is without adverse effects.

1. INTRODUCTION

Lewisite is an organic arsenical with vesicant properties. It can exist as a trans- or cis-isomer; in aqueous solutions, the cis-isomer undergoes photoconversion to the trans-isomer. Pure lewisite is a colorless, odorless oily liquid; however, synthesized agent is an amber to dark brown liquid with a geraniumlike odor (Munro et al. 1999). Lewisite causes local corrosive damage and may cause systemic poisoning after absorption through skin or mucous membranes. Exposure to lewisite causes almost immediate irritation and burning sensation of the eyes, skin, upper respiratory tract, and lungs. Death may result from direct pulmonary damage or circulatory failure from fluid loss and arrhythmia. Death that occurs within 24 h of exposure is likely due to pulmonary damage (Lindberg et al. 1997).

Lewisite was developed as a chemical warfare blister agent during the latter part of World War I and was named after its inventor Captain W. Lee Lewis. When the first ship loaded with lewisite reached Europe in 1918, the war ended, and the cargo was dumped into the sea. During the period between World War I and World War II, few studies on lewisite were conducted; however, when World War II began, the research efforts intensified. Results of those studies suggested that lewisite had limited utility as a war gas because of hydrolysis to a nonvolatile and water insoluble oxide, poor penetration of protective clothing, and difficulty in attaining lethal concentrations on the battle field (Lindberg et al. 1997). Also, lewisite is so immediately irritating at low concentrations (about 6-8 mg/m³) that troops would be warned of its presence, even before detecting the geranium-like odor at 14-23 mg/m³, and take protective action by deploying gas masks or retreating from the toxic atmosphere (Gates et al. 1946).

Lewisite-1 (L-1) is formed by the reaction of acetylene with arsenic trichloride using aluminum trichloride as a catalyst. Arsenic trichloride, lewisite-2 (L-2),

and lewisite-3 (L-3) are co-products concurrently formed with L-1 (Trammel 1992). L-1 yield is greater than 65%, and approximate yields of arsenic trichloride, L-2, and L-3 are 16-21%, 7-10%, and 4-12%, respectively (Lindberg et al. 1997). Therefore, an accidental release from storage tanks of L-1 will likely be the release of a mixture of L-1, L-2, L-3, and arsenic trichloride. Exposure will be to these compounds and to potential hydrolysis products, sodium arsenite (NaAsO2) and arsenic acid (H3AsO4). Toxicologic data on arsenic trichloride are limited; however, effects are similar to those of L-1 (corrosiveness, damage to skin, eyes, and mucous membranes). With regard to lethality, arsenic trichloride appears to be approximately 2-3 times less toxic than L-1; the LCt₅₀ (lethal concentration [product of concentration and time] that will cause death in 50% of the exposed population) for arsenic trichloride is 4,000-5,000 mg-min/m³ whereas the LCt₅₀ for L-1 is 1,200-1,500 mg-min/m³ (Flury 1921). L-2 and L-3 will be less significant toxicologically because of their smaller quantities and comparatively low volatility. However, the toxicity of L-2 and L-3 is reportedly comparable to L-1 (Lindberg et al. 1997). Therefore, AEGL values were derived for “lewisite” rather than for the individual lewisite compounds, and are considered protective for L-1, L-2, and L-3. In addition, in the review of literature, L-1, L-2, and L-3 are discussed only if the primary literature makes a distinction.

TABLE 5-3 Chemical and Physical Data for Lewisite Compounds

2. HUMAN TOXICITY DATA

2.2. Nonlethal Toxicity

2.2.1. Individual Studies

Lewisite is immediately and highly irritating at concentrations of about 68 mg/m³. The geranium-like odor is reportedly detectable at 14-23 mg/m³ (Gates et al. 1946).

Inhalation of lewisite at10 mg/m³ for 30 min reportedly resulted in severe intoxication and incapacitation that lasted for several weeks, and exposure at 10 mg/m³ for 15 min caused inflammation of the eyes and swelling of the eyelids (Franke 1968, as cited in Ottinger et al. 1973). Ottinger et al. (1973) is a review article, and did not provide experimental details or information regarding the severity of effects. No further details were available.

A dermal vapor study was conducted by Eldridge (1923). To select “men of average resistance” for the study, pin-point drops of 0.1 or 2% solutions of liquid lewisite in alcohol were applied to the forearms of 52 male volunteers at Edgewood Arsenal. If a subject showed no reaction to the 2% solution, he was classified as “resistant” and not used in the dermal vapor study. If a subject showed a marked reaction to the 0.1% solution, he was classified as “sensitive” and not used in the dermal vapor study. Of the 52 men, 14 were resistant and 3 were sensitive. Further dermal tests with liquid lewisite were done on the sensitive and resistant subjects; the tests showed that the sensitive subjects had no effects when treated with a 0.01% lewisite solution, and the resistant subjects developed blistering with a 5% solution. Dermal effects included blanching or graying of the skin, followed by severe erythema within 15-30 min. Vesication, accompanied by some edema, occurred within 12 h. Within less than 24 h, a raised area of redness measuring 2 × 2.5 inches in diameter appeared, accompanied by a 1.5-inch diameter blister surrounded by hundreds of minute vesicles. Forty-eight hours later, the raised area of redness had increased to 6 × 3.5 inches in diameter, and fluid seeped from the blister. The smaller vesicles also ruptured as the severity of the burns continued to increase up to the fourth day. No change was noted from days 4 to 7, and from day 7 onward improvement was observed until the burns were completely healed by the end of week 4. The men described a stinging sensation that lasted for 2 min and occurred within 2.5 min of exposure. No further sensation was noted until approximately 20 min later, when the stinging sensation was again reported and lasted for approximately 2 h. Five hours later, a “continuous feeling of discomfort” was described; burning lasted until the blister ruptured 22.5 h after lewisite was administered. Intermittent stinging and burning followed and the area became sore to the touch. By the end of day 6, the pain was more severe and occurred at shorter time intervals. By day 9, all pain had resolved.

Groups of 1-7 men (from the 35 male volunteers of average sensitivity described above) were exposed on their arms to varying concentrations of lewisite vapor for periods ranging from 10 min to 3 h to determine the concentration nec-

2.5. Carcinogenicity

In 1940, a World War II German soldier was accidentally exposed to lewisite on his lower right leg. The blistered lesion never healed, and was diagnosed as malignant in 1948. Bowen’s disease (intraepidermal squamous cell carcinoma) was diagnosed 38 years later (Krause and Grussendorf 1978).

Wada et al. (1962) reported increased incidences of cancer mortality (14% respiratory tract; 9.6% digestive tract) in workers from the Okuno-Jima poison gas factory. When cancer rates were correlated with job classification, the frequency of respiratory and gastrointestinal tract neoplasms were highest in workers involved in the production of mustard gas or lewisite, followed by those who worked indirectly with mustard gas or lewisite, and the lowest frequency was found in those that had no direct contact with mustard or lewisite (Yamakido et al. 1985). However, this information is confounded by the fact that workers were also exposed to mustard gas in addition to lewisite, and the factory also produced hydrocyanic acid, diphenylcyanarsine, chloroacetophenone, and phosgene.

2.6. Summary

TABLE 5-5 Data on Humans Exposed to Lewisite in Air

TABLE 5-6 Skin Effects in Humans Exposed to Liquid Lewisite

3.1.1. Rats

Olajos et al. (1998) exposed groups of six male and six female Sprague-Dawley rats (head-only) to product solution (waste stream) from the chemical neutralization of Chemical Agent Identification Sets (CAIS). The CAIS waste stream contained chloroform (vehicle), t-butanol (vehicle), and lewisite. Exposures to the CAIS waste stream were at 6,000, 12,000, 18,000, or 24,000 ppm and to the chloroform-butanol solvent were at 24,000 ppm for 1 h. The concen-

tration of lewisite in the test atmospheres was 0, 0.17, 0.67, 0.96, and 0.31 mg/m³ for the vehicle controls and the 6,000, 12,000, 18,000, and 24,000 ppm groups, respectively. Toxic signs were consistent with those of chloroform and butanol, and were noted in the control (vehicle) and waste-stream exposed animals. Ocular effects (corneal opacity and erosion) and pulmonary function effects (decreased minute volume) were similar in control and waste-stream exposed groups. The authors concluded that effects were due to chloroform and butanol, not lewisite.

3.1.2. Mice

Silver and McGrath (1943) exposed groups of 20 male CF-1 mice to varying concentrations of cis- or trans-lewisite for 10 min. Animals were exposed in a 386-L continuous flow chamber. Lewisite was vaporized by passing 20-30 L of air per minute (L/min) through lewisite in a bubbler at room temperature. Chamber airflow was maintained at 250 L/min. Lewisite concentrations in the chamber were measured analytically using a wet test meter. No animals were placed in the chamber until the chamber atmosphere reached equilibrium (approximately 10 min). Ten-min mouse LC₅₀ values of 190 and 200 mg/m³ were determined for the cis- and trans-isomers, respectively. All mice exposed to lewisite for 10 min at 240 mg/m³ died.

3.1.3. Dogs

Harrison et al. (1946) exposed dogs to lewisite at 50 mg/m³ for 30 min (8 dogs), 61 mg/m³ for 30 min (9 dogs), or 121 mg/m³ for 10 min (5 dogs). Clinical signs included vomiting, urination, defecation, salivation, and respiratory distress; 80% of the dogs died 3-48 h after exposure. No other information was available.

A dermal LD₅₀ of 15 mg/kg (Cameron et al. 1946) and subcutaneous LD₅₀ of 2 mg/kg (Cameron et al. 1946) were reported for dogs.

3.1.4. Rabbits

A 7.5-min LC₅₀ of 160 mg/m³ and a 60-min LC₅₀ of 25 mg/m³ was reported for rabbits (Gates et al. 1946). A dermal LD₅₀ of 6 mg/kg (Cameron et al. 1946) and intravenous LD₅₀ of 0.5 mg/kg (Cameron et al. 1946) were also reported.

3.1.5. Guinea Pigs

A 9-min LC₅₀ of 111 mg/m³ and a 60-min LC₅₀ of 8 mg/m³ were reported for guinea pigs (Gates et al. 1946). A dermal LD₅₀ of 12 mg/kg (Cameron et al. 1946) and subcutaneous LD₅₀ of 1 mg/kg (Cameron et al. 1946) were also reported.

3.1.6. Goats

A 100-min LC₅₀ of 12.5 mg/m³ (Gates et al. 1946) and a dermal LD₅₀ of 15 mg/kg (Cameron et al. 1946) were reported for goats.

3.2. Nonlethal Toxicity

3.2.1. Rats

3.2.3. Rabbits

Ocular lesions, but no deaths, were reported in rabbits exposed to lewisite at 1 mg/m³ for 30 min (Gates et al. 1946). No additional experimental details, including severity of effects, were reported.

3.2.4. Pigs

Lindsay et al. (2004) dermally exposed three large white pigs to lewisite at 0.3 mg/cm². While under anesthesia, an area of dorsal skin (35 cm × 25 cm) was shaved. Exposures were then conducted using inverted glass chambers; lewisite (in hexane) was pipetted onto 10-cm² glass-fiber discs fitted tightly in the roof of each circular, glass chamber. The heat from the animals vaporized the lewisite so that the skin was exposed to vapor, but not lewisite liquid. Pigs were monitored in their pens for 24 h and were then killed. Full skin thickness samples from control (non-exposed) and lewisite-treated skin were excised to examine the degradative processes in connective tissue components of skin, especially glycoproteins, using immunostaining and gel electrophoresis. There was no evidence of cross linking of laminin or of type III or IV collagen in lewisite-treated pigs. There was evidence of degradation of laminin and type IV collagen only.

3.3. Developmental and Reproductive Effects

Hackett et al. (1987) administered lewisite to CD rats and New Zealand white rabbits by gastric intubation. Rats were dosed daily on days 6 through 15 of gestation with lewisite at 0, 0.5, 1.0, 2.0, or 2.5 mg/kg in a range-finding study and with 0, 0.5, 1.0, or 1.5 mg/kg in a teratology study. Rabbits were dosed on gestation days 6 through 19 with lewisite at 0, 0.5, 1.0, 1.5, and 2.0 mg/kg in a range-finding study and at 0, 0.07, 0.2, and 0.6 mg/kg in a teratology study. In rats, no maternal of fetal effects were noted at 1.5 mg/kg. At 2.0 mg/kg, maternal mortality (10%), decreased maternal and fetal body weight, and decreased numbers of viable fetuses were found. In rabbits, maternal mortality ranged from 13% in the 0.07-mg/kg group to 100% in the 2.0-mg/kg group. This mortality rate limited the sample size and made identification of other potential fetal or maternal effects difficult. However, at 0.07 mg/kg, only maternal mortality was noted. At 0.6 mg/kg (highest dose in the teratology study), effects included 86% maternal mortality, decreased maternal body weight gain, an increased incidence of fetal stunting, and a tendency toward decreased fetal body weight (Hackett et al. 1987).

In a 42-week, two-generation reproductive study in rats, parental males and females were administered lewisite in sesame oil by gastric intubation at 0, 0.10, 0.25, or 0.60 mg/kg/day for 5 days/week prior to mating, during mating, and after mating until the birth of offspring. Dams continued to be exposed to lewisite during lactation. After weaning, male and female offspring selected to

continue on the study were exposed similarly to lewisite. There were no treatment-related effects on reproductive performance, fertility, or reproductive organ weights of male or female rats through two consecutive generations. There were no treatment-related effects in offspring (Sasser et al. 1989).

3.4. Genotoxicity

Lewisite did not induce mutations in Salmonella typhimurium strains TA97, TA98, TA100, or TA102 with or without metabolic activation up to concentrations limited by toxicity (1.0 μg/plate) (Stewart et al. 1989). Lewisite was negative for mutation at the HGPRT locus in Chinese hamster ovary (CHO) cells at concentrations ranging from 0.12 to 2.0 μM (Jostes et al. 1989). However, lewisite induced chromosomal aberrations in CHO cells at concentrations of 0.50, 0.75, and 1.0 μM (Jostes et al. 1989). Lewisite was negative in the Dro-sophilla melanogaster sex-linked recessive lethal assay (Auerbach and Robson 1946, 1947) and negative in a dominant lethal assay in CD rats at concentrations of 0.375, 0.75, or 1.5 mg/kg (Bucci et al. 1993).

3.5. Carcinogenicity

No data were located regarding the carcinogenicity of lewisite in animals.

3.6. Summary

TABLE 5-8 Summary of Acute Inhalation Data from Animals Exposed to Lewisite

4.2. Mechanism of Toxicity

Dermal or intravenous exposure to lewisite leads to local skin edema and pulmonary edema due to increased capillary permeability. There is no evidence of edema or capillary permeability in any other part of the body. The increased capillary permeability results in blood plasma loss and leads to a sequence of physiological events termed “lewisite shock” which is similar to shock observed in severe burn cases. Functional changes in the lungs, kidneys, respiratory tract, cardiovascular system, and lymphatic system may be the result of a disturbance of osmotic equilibrium (Goldman and Dacre 1989).

TABLE 5-9 Summary of Acute Oral, Dermal, Subcutaneous, and Intravenous Data from Animals Exposed to Lewisite

As reviewed in Goldman and Dacre (1989) and Young (1999), the mechanism of toxicity of lewisite is the formation of stable complexes between the arsenite moiety of lewisite and sulfhydryls groups that are critical to the function of proteins and thiol cofactors (e.g., dihydrolipoic acid, keratin, alcohol dehydrogenase, pyruvate dehydrogenase, succinic dehydrogenase, succinic oxidase, hexokinase). The formation of stable complexes with protein thiols is also the primary mechanism of toxicity of arsenite. Although lewisite can hydrolyze to yield arsenite, the reaction occurs at alkaline conditions and/or high temperature and is unlikely to be important in lewisite toxicology (Goldman and Dacre 1989).

4.3. Structure-Activity Relationships

Lewisite and arsenite share a common mechanism of action in disruption of protein function by formation to complexes with protein sulfhydryls. Toxicologic data on arsenic trichloride, L-2 and L-3, co-products concurrently formed with L-1, are limited. However, effects are similar qualitatively to those of L-1 (corrosiveness, damage to skin, eyes, and mucous membranes). With regard to lethality, arsenic trichloride appears to be approximately 2-3 times less toxic than L-1; the LCt₅₀ for arsenic trichloride is 4,000-5,000 mg-min/m³ whereas the LCt₅₀ for L-1 is 1,200-1,500 mg-min/m³ (Flury 1921). The toxicity of L-2 and L-3 is reportedly comparable to L-1 (Lindberg et al. 1997). Silver and McGrath (1943) found no substantial difference in 10-min LC₅₀ values (190 and 200 mg/m³) for the cis- and trans-isomers of lewisite.

Inhalation data for sodium arsenite, a hydrolysis product of L-1, are not available. However, Inns et al. (1988) compared the acute intravenous toxicity of lewisite and sodium arsenite in New Zealand white rabbits. The LD₅₀ of lewisite was 1.8 mg/kg. Rapid panting was noted 5 min after injection, and was followed by prostration and death within 4 h. By 24 h after injection, surviving rabbits appeared normal. The LD₅₀ for sodium arsenite was 7.6 mg/kg, with hypoactivity noted 20 min after injection. On the basis of trivalent arsenic content, lewisite was 6.5 times more toxic than the inorganic sodium arsenite, and clinical signs and times of death and recovery differed between the compounds. Severe pulmonary damage (gross and histopathologic) was found in rabbits treated with lewisite, but not in animals treated with sodium arsenite. Also, arsenic levels in the liver, kidneys, brain, stomach, duodenum, spleen, and bladder were much greater in sodium arsenite-treated rabbits than in lewisite-treated rabbits. However, arsenic content in the lungs was similar. These data suggest different mechanisms of toxicity for lewisite and inorganic trivalent arsenic, and that arsenite is not an appropriate surrogate for lewisite.

4.4. Other Relevant Information

4.4.1. Species Variability

FIGURE 5-1 LC₅₀ Data for Lewisite in Dogs. Source: Armstrong 1923.

5.2. Animal Data Relevant to AEGL-1

No animal data were relevant for establishing AEGL-1 values for lewisite.

5.3. Derivation of AEGL-1

Appropriate data were not available to derive AEGL-1 values for lewisite. Odor cannot be used as a warning for potential exposure, because the odor threshold (14-23 mg/m³ for L-1) is higher than concentrations that are highly irritating and higher than the AEGL-2 and AEGL-3 values. Therefore, AEGL-1 values are not recommended. Lack of AEGL-1 values does not imply that exposure at concentrations below the AEGL-2 values is without effect.

6. DATA ANALYSIS FOR AEGL-2

6.1. Human Data Relevant to AEGL-2

No human data were available for establishing AEGL-2 values for lewisite.

6.2. Animal Data Relevant to AEGL-2

No animal data were available for establishing AEGL-2 values for lewisite. Ocular inflammation and lesions reported by Gates et al. (1946) and Ottinger et al. (1973) were considered an inappropriate basis for AEGL-2 values. The Ottinger et al. (1973) report is a review paper, which noted that “a lower concentration of 0.01 mg/L causes inflammation to the eyes and swelling of the lid after 15 minutes” (as

cited in Franke 1968); however, no experimental details were provided and attempts to obtain the Franke (1968) report were not successful. Because the primary data could not be obtained for review, the information was considered unsuitable for deriving AEGL-2 values. Gates et al. (1946) reported approximate concentrations necessary to produce ocular lesions in 30 min (0.001 mg/L in rabbits and 0.20 mg/L in dogs); however, no experimental details or descriptions of the lesions were reported. For both the Ottinger et al. (1973) and Gates et al. (1946) reports, sufficient detail is not available to determine the severity of effects and do not provide no-effect levels.

6.3. Derivation of AEGL-2

No inhalation data with both concentration and duration parameters and with effects consistent with the definition of AEGL-2 end points were available. Therefore, the AEGL-2 values for lewisite were determined by taking a three-fold reduction in the AEGL-3 values; the resulting values are considered to be estimated thresholds for irreversible effects (NRC 2001). The reduction approach is considered appropriate because of the steep concentration-response curve for mortality from lewisite. In studies with mice, the 10-min LC₅₀ was 200 mg/m³ and the 10-min LC₁₀₀ was 240 mg/m³. In dogs, no deaths occurred after a 7.5-min exposure to lewisite at 126 mg/m³, and the LC₅₀ was 176 mg/m³.

TABLE 5-10 AEGL-2 Values for Lewisite

7.3. Derivation of AEGL-3

TABLE 5-11 AEGL-3 Values for Lewisite

TABLE 5-12 AEGL Values for Lewisite

^aNR, not recommended; absence of an AEGL-1 does not imply that exposure below the AEGL-2 values is without adverse effect.

8.2. Comparisons with Other Standards and Guidelines

No exposure standards or guidelines were for L-1, L-2, or L-3 were found.

8.3. Data Adequacy and Research Needs

Human data were insufficient for deriving AEGL values for lewisite. Mouse and dog lethality studies were well conducted and were not inconsistent with the limited lethality data in other species. No data on concentration-response relationships for AEGL-2 effects were suitable for deriving AEGL-2 values. Data were available only for L-1; however, given the low volatility and small volume of L-2 and L-3 in total lewisite and the similar toxicity of L-2 and L-3 with L-1 (Lindberg et al. 1997), AEGL-values derived for lewisite should be protective for L-1, L-2, and L-3 compounds.

9. REFERENCES

Because lewisite compounds were developed as chemical warfare agents, military literature is a major source of relevant toxicity data. Consequently, many of the study reports have “limited distribution”, which is a separate issue from “classification”. For various reasons, sources may have a restricted distribution because of treaty restrictions on data access with allies, concerns regarding distribution of engineering information characterizing agent dissemination or generation in other sections of the same document, and related issues. To ensure

public access to pertinent toxicity data originating from limited-distribution materials, pertinent data from those sources have been incorporated into this chapter.

Armstrong, G.C. 1923. The toxicity of M-1 by inhalation for dogs. Chapter II in The Toxicity, Pathology, Chemistry, Mode of Action, Penetration, and Treatment for M-1 and its Mixtures with Arsenic Trichloride. Part 1. ADB954935. Edgewood Arsenal, Aberdeen Proving Ground, MD. August 13, 1923 (unclassified report/limited distribution).

Auerbach, C., and J.M. Robson. 1946. Chemical production of munitions [letter]. Nature 157(3984):302.

Auerbach, C., and J.M. Robson. 1947. Tests of chemical substances for mutagenic action. Proc. R. Soc. Edinb. Biol. 62:284-291.

Goldman, M., and J.C. Dacre. 1989. Lewisite: Its chemistry, toxicology, and biological effects. Rev. Environ. Contam. Toxicol. 110:76-115.

Olajos, E.J., E.W. Morgan, R.A. Renne, H. Salem., B. McVeety, R. Johnson, and R.L. Phelps. 1998. Acute inhalation toxicity of neutralized chemical agent identification sets (CAIS) containing agent in chloroform. J. Appl. Toxicol. 18(5): 363-371.

Ottinger, R.S., J.L. Blumenthal, D.F. Dal Porto, G.I. Gruber, M.J. Santy, and C.C. Shih. 1973. Recommended Methods of Reduction, Neutralization, Recovery, or Disposal of Hazardous Wastes. Volume VII. Propellants, Explosives, Chemical Warfare. EPA-670/2-73-053-g. U.S. Environmental Protection Agency, Washington, DC.

Prentiss, A.M. 1937. Vesicant agents. Pp. 177-300 in Chemicals in War: A Treatise on Chemical Warfare. New York: McGraw-Hill Book Company, Inc.

Reutter, S.A., D.R. Sommerville, and L.L. Miller, Jr. 2003. Review and Recommendations for Human Toxicity Estimates for FM 3-11.9. ECBC-TR-349. Edgewood Chemical Biological Center, U.S. Army Soldier and Biological Chemical Command. September, 2003 (unclassified report/limited distribution).

Young, R.A. 1999. Appendix F. Health risk assessment for lewisite. Pp. 277-294 in Review of the U.S. Army’s Health Risk Assessments for Oral Exposure to Six Chemical-Warfare Agents. Washington, DC: National Academy Press.

APPENDIX A

DERIVATION OF AEGL VALUES FOR LEWISITE

Derivation of AEGL-1 Values

The available data were insufficient to derive AEGL-1 values for lewisite.

Derivation of AEGL-2 Values

In the absence of relevant data to derive AEGL-2 values, a one-third reduction of the AEGL-3 values was used to derive AEGL-2 values (NRC 2001).

Derivation of AEGL-3 Values

APPENDIX B

ACUTE EXPOSURE GUIDELINE LEVELS FOR LEWISITE

Derivation Summary

AEGL-1 VALUES

The available data on lewisite were inadequate to derive AEGL-1 values.

AEGL-2 VALUES

Data adequacy: The available data on lewisite were inadequate to derive AEGL-2 values. When data are lacking and the concentration-response curve is steep, AEGL-2 values may be derived by dividing the AEGL-3 values by 3 (NRC 2001). A steep concentration-response curve has been demonstrated for lewisite. In studies with mice, the 10-min LC₅₀ was 200 mg/m³ and the 10 -min LC₁₀₀ was 240 mg/m³. In dogs, no deaths occurred after a 7.5-min exposure to lewisite at 126 mg/m³, and the LC₅₀ was 176 mg/m³.

AEGL-3 VALUES

Key reference: Armstrong, G.C. 1923. The toxicity of M-1 by inhalation for dogs. Chapter II in The Toxicity, Pathology, Chemistry, Mode of Action, Penetration, and Treatment for M-1 and its Mixtures with Arsenic Trichloride. Part 1. ADB954935. Edgewood Arsenal, Aberdeen Proving Ground, MD. August 13, 1923. (unclassified report/limited distribution).

Test species/Strain/Number: Dog; breed not specified; 1-17 per group.

Exposure route/Concentrations/Durations:
Inhalation; 126, 176, 231, 274, and 330 mg/m³ for 7.5 min
Inhalation; 68.7, 87.7, 96, 102, 125, and 233 mg/m³ for 15 min
Inhalation; 11.5, 24.5, 30.6, 41.5, 48, and 58.6 mg/m³ for 30 min
Inhalation; 5.8, 8, 25, 35, 43, and 53 mg/m³ for 1 h
Inhalation; 4.8, 12.5, 17.9, 24.5, and 34.5 mg/m³ for 2 h
Inhalation; 2.1, 6.2, 10, and 16.9 mg/m³ for 4 h

End point/Concentration/Rationale: Calculated LC₀₁ values, considered thresholds for lethality.

Uncertainty factors/Rationale:
Total uncertainty factor: 10
Interspecies: 3, data suggest little species variability with regard to lethality from inhalation exposure to L-1; C × T values are relatively constant across species, except for the guinea pig, and the interspecies uncertainty factor of 3 encompasses the 2- to 3-fold difference in sensitivity between guinea pigs and rats, mice, rabbits, dogs, and goats.
Intraspecies: 3, steep concentration-response curve with regard to lethality implies limited intraspecies variation. In studies with mice, the 10-min LC₅₀ was 200 mg/m³ and the 10-min LC₁₀₀ was 240 mg/m³. In dogs, no deaths occurred after a 7.5-min exposure to lewisite at 126 mg/m³, and the LC₅₀ was 176 mg/m³.

Modifying factor: Not applicable

Animal-to-human dosimetric adjustment: Not applicable

Time scaling: Points of departure were time-specific LC₀₁ values.

Data adequacy: Data are adequate to derive AEGL-3 values for lewisite.

APPENDIX C

CALCULATION OF LC₀₁ VALUE FOR DOGS

Data source: Armstrong (1923)

Filename: ten Berge Spreadsheet Data for Log Probit Model Date: 15 October 2010 Time: 09:03:51

Used Probit Equation Y = B0 + B1*X1 + B2*X2
X1 = concentration mg/m³, ln-transformed
X2 = minutes, ln-transformed

Chi-Square = 15.93
Degrees of freedom = 29
Probability Model = 9.76E-01

Ln(Likelihood) = -29.24

B 0 = -7.7323E+00
Student t = -3.1898 B 1 = 1.7999E+00
Student t = 5.3334 B 2 = 1.6615E+00
Student t = 5.2230

Variance B 0 0 = 5.8761E+00
Covariance B 0 1 = -8.1104E-01
Covariance B 0 2 = -7.6250E-01
Variance B 1 1 = 1.1390E-01
Covariance B 1 2 = 1.0355E-01
Variance B 2 2 = 1.0120E-01

Estimation ratio between regression coefficients of ln(conc) and ln(minutes)
Point estimate = 1.083
Lower limit (95% CL) = 0.976
Upper limit (95% CL) = 1.191

Estimation of concentration mg/m³ at response of 1%
Minutes = 10
Point estimate concentration mg/m³ = 3.869E+01 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 1.699E+01 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 5.741E+01 for response of 1%

Estimation of concentration mg/m³ at response of 1%
Minutes = 30
Point estimate concentration mg/m³ = 1.403E+01 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 6.185E+00 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 2.064E+01 for response of 1%

Estimation of concentration mg/m³ at response of 1%
Minutes = 60
Point estimate concentration mg/m³ = 7.400E+00 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 3.237E+00 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 1.094E+01 for response of 1%

Estimation of concentration mg/m³ at response of 1%
Minutes = 120
Point estimate concentration mg/m³ = 3.903E+00 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 1.682E+00 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 5.838E+00 for response of 1%

Estimation of concentration mg/m³ at response of 1%
Minutes = 240
Point estimate concentration mg/m³ = 2.058E+00 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 8.675E-01 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 3.138E+00 for response of 1%

Estimation of concentration mg/m³ at response of 1%
Minutes = 480
Point estimate concentration mg/m³ = 1.085E+00 for response of 1%
Lower limit (95% CL) concentration mg/m³ = 4.447E-01 for response of 1%
Upper limit (95% CL) concentration mg/m³ = 1.697E+00 for response of 1%

APPENDIX D

CATEGORY PLOT FOR LEWISITE

FIGURE D-1 Category plot of animal and human toxicity data and AEGL values for lewisite.

TABLE D-1 Data Used in the Category Plot for Lewisite

For category: 0 = no effect, 1 = discomfort, 2 = disabling, 3 = lethal; SL = some lethality.