The U.S. Army's Chemical Defense Equipment Process Action Team (CDEPAT) recently conducted an extensive review of the scientific basis for toxicity estimates in use by the Army for several chemical-warfare (CW) agents: GA, GB, GD, GF, VX, and HD. Following a detailed analysis of the toxicity of these agents and using contemporary methods of analysis, CDEPAT concluded that many of the human-toxicity estimates in use would not protect the soldier adequately (CDEPAT 1994). Recalculations of the potencies of several of the CW agents indicate that their potencies are greater than previously determined. As a result, lower exposure levels of CW agents are expected to elicit adverse effects.

Before deciding whether to implement CDEPAT's recommendations, the U.S. Department of the Army requested that the National Research Council (NRC) independently review the CDEPAT report entitled Review of Existing Toxicity Data and Human Estimates for Selected Chemical Agents and Recommended Human Toxicity Estimates for Defending the Soldier. The NRC assigned the project to the Committee on Toxicology (COT) of the Board on Environmental Studies and Toxicology. The COT convened the Subcommittee on Toxicity Values for Selected Nerve and Vesicant Agents, which conducted the study and prepared this report. Subcommittee members were chosen for their expertise in several specialties, including

toxicology, medicine, pathology, biostatistics, and risk assessment. The subcommittee was charged with determining the scientific validity of CDEPAT's proposed human-toxicity estimates for CW agents for various routes of exposure (that is, percutaneous vapor exposures, vapor inhalation exposures, and percutaneous liquid exposures). The report considers only acute¹ exposures and acute effects. It should be noted that the human-toxicity estimates for the CW agents were proposed for healthy adult male soldiers only. They must not be used for the general population. Specifically, the subcommittee was charged with the following tasks:

1. Review the scientific protocols and the quality of the toxicity data used in revising the human-toxicity estimates for acute exposures.
2. Review the toxicity estimates for mild and nonsevere effects and for severe and lethal effects.
3. Review the procedures used in deriving the human-toxicity estimates for acute exposures.
4. Determine the appropriateness of the assumptions made in deriving the human-toxicity estimates for acute exposures.

In reviewing the toxicity data and the proposed human-toxicity estimates for acute exposures, the subcommittee evaluated the quality of the data, the appropriateness of the procedures used in obtaining the estimates, and the assumptions made in deriving them. The subcommittee also determined whether the supporting documentation justified the proposed recommendations and whether the studies and toxicity end points were appropriate for deriving the toxicity estimates. In reviewing the proposed human-toxicity estimates, the subcommittee reviewed only the toxicity information presented in the CDEPAT report. It did not perform an independent literature search, nor did it review any data other than those presented in the report. In addition, the subcommittee was not asked to recommend new estimates or to address the policy or operational consequences of the proposed lower human-toxicity estimates.

The exposures used in the estimates are defined as follows:

• LCt₅₀ is the exposure to a vapor causing lethality in 50% of a given population and is expressed as the product of air concentration (c), in

• milligrams per cubic meter, and exposure duration (t), usually in minutes.² Ct refers to the product of concentration and exposure time. LCt₅₀s were derived for either inhalation or percutaneous vapor exposures. It is important to note that the product of concentration and exposure duration (c × t) is not necessarily a constant.
• ECt₅₀ is the exposure to a vapor causing a defined effect in 50% of a given population and is expressed as the product of c × t in mg-min/m³, where c × t is not necessarily a constant. The effects include those classified as threshold (minimal), mild (miosis and/or rhinorrhea), or severe (incapacitation, prostration, collapse, and convulsions). ECt₅₀s were derived for either inhalation or percutaneous vapor exposures.
• ICt₅₀ is the exposure to a vapor causing incapacitation in 50% of a given population and is expressed as the product of c × t in mg-min/m³, where c × t is not necessarily a constant.
• LD₅₀ is the acute dose of a liquid agent causing lethality in 50% of a given population and is expressed in milligrams per kilogram of body weight (mg/kg). In this report, all LD₅₀s are for percutaneous liquid exposure of bare skin.
• ID₅₀ is the acute dose of a liquid agent causing incapacitation in 50% of a given population and is expressed in mg/kg. In this report, all ID₅₀s are for percutaneous liquid exposure of bare skin.
• ED₅₀ is the acute dose of a liquid agent causing a defined effect in 50% of a given population and is expressed in mg/kg. In this report, all ED₅₀s are for percutaneous liquid exposure of bare skin. Effects include those in the threshold, mild, or severe (including incapacitation) categories.

The results of the subcommittee's deliberations are presented in Chapters 2 through 8 of this report. Chapters 2 through 7 review CDEPAT's proposed human-toxicity estimates for agents GA, GB, GD, GF, VX, and HD, respectively. Those chapters provide conclusions on the scientific validity of the proposed acute human-toxicity estimates and recommendations for research efforts. Chapter 8 evaluates the risk-estimation procedures used in the CDEPAT report. The Appendix discusses the offensive versus the defensive use of human-toxicity estimates for CW agents.