Review of Acute Human-Toxicity Estimates for Selected Chemical-Warfare Agents (1997)
Chapter: ED50 for Severe Effects
man, or 5 mg/kg. The existing LD50 estimate for GD following percutaneous liquid exposure is the same (CDEPAT 1994).
The available animal-toxicity data are insufficient for estimating the human LD50 for percutaneous exposure to GD liquid with any degree of confidence. The most often quoted estimate is 350 mg for a 70-kg man, an estimate based on a study by Cullumbine et al. (1954) investigating the effect of percutaneous exposure of rabbits to GD liquid. The rabbit is more sensitive to GD liquid than any other nonhuman species (Henry 1989). The reported LD50s in animals range from < 1 to 14 mg/kg (< 70 to 980 mg for a 70-kg man). Given this wide range in LD50 values, the subcommittee's degree of confidence in CDEPAT's proposed estimate of 350 mg for a 70-kg man is moderately low. The subcommittee recommends that CDEPAT's proposed estimate serve as an interim value until further research is conducted to establish this estimate with a greater degree of confidence.
ED50 for Severe Effects
CDEPAT's proposed estimate for the ED50 for severe effects following percutaneous exposure to GD liquid on bare skin is 200 mg for a 70-kg man or approximately 3 mg/kg. There is no existing ED50 estimate for GD (CDEPAT 1994).
The available data are insufficient for estimating the human ED50 for severe effects (incapacitation, prostration, collapse, and convulsion) after percutaneous exposure to GD liquid. Data on exposure of pigs indicate that the slopes for lethality and severe effects are parallel and that the ID50-to-LD50 ratio is about 0.8 (Manthei et al. 1988; CDEPAT 1994). Other data suggest that the ratio is approximately 0.6 (Reutter et al. 1992). On the basis of the more conservative ratio of 0.6, CDEPAT derived the human-toxicity estimate of 3 mg/kg (5 mg/kg (LD50) × 0.6 = 3 mg/kg). In the absence of adequate human data or animal data on GD for this effect, the subcommittee accepts CDEPAT's approach of using the ID50-to-LD50 ratio of 0.6 to derive the ED50 estimate. The subcommittee's degree of confidence in CDEPAT's proposed estimate is low to moderate because of the insufficient data. The subcommittee recommends that the proposed estimate serve as an interim value until further research is conducted to establish this estimate with a greater degree of confidence.
