5
Assumptions and Limitations
TEST PROTOCOL CONSIDERATIONS
An important aspect of the technical review of the MIST methodology is identifying all assumptions in the BRHA model. These assumptions must be closely examined to establish their impact on the design and implementation of the model. The resulting limitations must also be identified and evaluated to establish their impact on the applicability of the model to real-life situations. The committee's analysis of the test protocol raised the following questions.
The basic measurements of MIST/BRHA are protection factors for test suits for various anatomical sites against a 12,000 mg/m3-min (concentration x time [Ct] factor) exposure to MeS (methyl salicylate). The protection factors are used to derive effective Cts against VX and HD. The MIST/BRHA can be used to rank the relative protective value of test suits against VX and HD but should not be used to predict physiological effects. For example, it would be wrong to suggest that wearing a given protective suit in a VX exposure of 3,225 mg/m3-min (see Table 2-7) would result in symptoms of VX poisoning associated with a 70 percent depression in red blood cell cholinesterase. The data do not support a calculation of agent percutaneous absorption from the mass of MeS collected in passive samplers attached to the skin.
Given an effective Ct of 25 mg/m3-min for VX exposure to produce a 70 percent depression in cholinesterase in an unprotected person, and 1,000 mg/m3-min for HD to produce severe burns on an unprotected forearm, a Ct of 12,000 mg/m3-min in the MIST would be considered a massive challenge, especially for VX (if 1 mg of MeS is taken as the equivalent to 1 mg of chemical agent). Because all exposures were at a Ct of 12,000 mg/m3-min, suit rankings might be different at lower Ct challenges.
The selection of MeS as a simulant for both VX and HD was based on its historical use as an agent simulant and on its safety for human use (see Chapter 3). One must ask, however, whether MeS is a reliable simulant of chemical agents. How does the diffusivity, solubility, surface tension, wetability, etc., of MeS compare with chemical agents? The physicochemical properties of VX, HD, and MeS are very different. MeS has a vapor pressure approximately 2 orders of magnitude higher than VX. Also, we do not know the relative distributions of these chemicals in system-level tests of chemical and biological protective suits.
The BRHA is based on the assumptions that the regional variation in VX skin toxicity is the same for liquid and vapor exposures and that the relative regional variation in skin toxicity to HD is the same as for VX. These assumptions have not been tested. For some compounds, some studies have shown that the regional variation in skin permeability appears to be compound dependent (Table 5-1); other studies have shown that permeability coefficients for liquid and vapor exposures are different (Barry et al., 1984).
Based on the statistical analyses of local and systemic effective Cts (Tables 2-5 and 2-7, respectively), approximately 75 percent of the
TABLE 5-1 Regional Variations in Human Skin Permeability as a Function of Test Substance
|
|
Relative Permeability |
|
|
Anatomic Site |
Hydrocortisone |
Parathion |
Malathion |
Benzoic Acid |
Forearm (ventral) |
1.0 |
1.0 |
1.0 |
1.0 |
Forearm (dorsal) |
1.1 |
- |
- |
- |
Foot arch (plantar) |
0.1 |
- |
- |
- |
Ankle (lateral) |
0.4 |
- |
- |
- |
Palm |
0.8 |
1.4 |
0.9 |
- |
Back |
1.7 |
- |
- |
0.8 |
Abdomen |
- |
2.2 |
1.4 |
1.6 |
Scalp |
3.5 |
3.7 |
- |
- |
Axilla |
3.6 |
7.4 |
4.2 |
- |
Forehead |
6.0 |
4.1 |
3.4 |
3.0 |
Jaw angle |
13.0 |
3.9 |
10.0 |
- |
Scrotum |
42.0 |
12.0 |
- |
- |
test suits have overlapping CIs (confidence intervals) (at 95 percent significance level) with the battle dress overgarment. The CIs may be accurate reflections of the performance of the test suits; however, because the suits were not all tested the same number of times, conclusions cannot be drawn about the relative merits of the suits or about the discriminatory power of the MIST/BRHA. One must question then whether the tests have been replicated adequately to draw statistically reliable conclusions.
Because the MIST procedure is expensive, the natural tendency is to minimize the number of replications. It might be more appropriate to screen suits more rigorously prior to the MIST and to subject only the most promising candidates to the MIST/BRHA with more replications. Data from the MIST/BRHA have undergone ANOVA (analysis of variance). Additional tests (e.g., Dunnett, Neuman-Kuels test) should also be employed for multiple comparisons of protective ensembles. It may be of value to analyze data from individual anatomic sites as well as data from different test subjects.
The passive detectors are intended to measure skin deposition of MeS. However, no data have been established to compare MeS skin deposition directly with passive detector deposition.
The MIST is an accelerated test. In other words, higher concentrations of simulant are used for shorter periods of exposure (two hours) than are specified and required for suit protection (24 hours). Is this trade-off of time/concentration justified?
Have the temperature and relative humidity in the test chamber been appropriately chosen and controlled? One might suspect that the barrier and permeation properties of materials are temperature and humidity dependent; thus, performance must be evaluated under appropriate, relevant conditions. In the MIST procedure, is MeS introduced in the appropriate concentration (vapor challenge) to simulate a realistic chemical and biological threat? Are the convective dynamics (airflow rates) realistically reproduced and adequately controlled in the test chamber?
HUMAN FACTORS CONSIDERATIONS
The committee has two concerns about human factors associated with the test operations procedure for the MIST that should be addressed. The first deals with the closures of the protective garments
and the second with the physical exercise routine to simulate field conditions under which the soldier would be expected to function.
The test operations procedure ensures that garments are properly closed at the beginning of each test. However, a complete, reliable interpretation of test results requires knowing the degree to which the closures remain closed during the test. The closures could be checked by the test supervisors at the end of the 120-minute exposure period when they check the positions of the passive samplers. Information about the closures would be helpful for interpreting differences in absorption levels at different anatomical sites. A related issue is the probability that soldiers would keep the garments closed under real combat conditions. The test procedure also specifies that an interview be conducted and a human factors questionnaire be completed by the test subject at the conclusion of the test. The interview and questionnaire could be critical to determining whether the suit would be worn as intended, with full closure, to evaluate test results from a practical standpoint.
The physical exercise routine in the MIST protocol may not adequately simulate field conditions, if for no other reason than that field conditions cover a wide range of variables and are almost impossible to specify. Nevertheless, the physical exercise must be rigorous and as reflective of anticipated field conditions as possible. The extent to which the movement disturbs the suit closures could be very important. Furthermore, body heat generated during exercise would increase the likelihood that a soldier might loosen the garment. The perspiration level during exercise is also important because perspiration changes the distribution of chemicals on the skin surface. It would be advantageous to review relevant literature comparing the types of exercises used in the MIST with actual field conditions to ensure that field conditions are adequately simulated.
ACETYLCHOLINESTERASE INHIBITION AS A BIOLOGICAL MARKER
Acetylcholine is neurotransmitter released at many autonomic nerve endings that binds to neurons and causes them to fire. Acetylcholinesterase is the enzyme that breaks the acetylcholine bond and returns the neuron to the resting state. Certain nerve toxins have long been believed to inhibit neural acetylcholinesterase enzyme activity based on the demonstrated ability of these
agents to decrease acetylcholinesterase activity in vivo and the observation of the continued firing of neurons that could be explained by the action of these toxins. Understanding how known cholinesterase inhibitors work has proven to be very useful for developing neurotoxic pesticides and diagnostic tests for humans affected by them. These diagnostic tests are based on the presence in human red blood cell membranes of readily measurable levels of acetylcholinesterase activity, as well as the presence in plasma of a related enzyme known as pseudocholinesterase. However, in recent years a debate has developed about the usefulness of blood cholinesterase activity as a biological marker to predict neurotoxic effects.
Biological markers can be divided into markers that indicate the amount of exposure and markers that indicate presumed susceptibility. There is a continuum between markers of exposure and markers of susceptibility, and some markers can be classified as both. Interpreting biological markers depends on understanding the toxicology of the chemical, including its absorption, distribution, metabolism, excretion, and toxicity to the target organ.
Blood cholinesterase activity, either red blood cell or serum, should be considered both as an exposure marker and an effect marker. An example of an ideal marker of both exposure and effect is carboxyhemoglobin. Carbon monoxide bound to hemoglobin is both an integrated measure of carbon monoxide exposure in the past 8 to 12 hour period, and, through our understanding of the mechanism of CO toxicity, a predictor of adverse consequences.
Blood acetylcholinesterase activity is also, to some extent, a marker of both exposure and effect. The major limitation is the obvious fact that the target organ of concern is the brain, but the measurements are of enzyme activity in the blood. Subtle variations in enzyme structure between different tissues must be taken into account. Another limitation is that blood cholinesterase levels may vary due to genetic and disease factors. For most individuals, however, blood cholinesterase activity is a suitable biological marker of exposure. It is also a very useful biological marker of effect, as long as it is recognized that observable effects, such as fasciculations, do not begin until there is perhaps a 40 to 50 percent decrease in cholinesterase. Thus a 20 percent decrease in enzyme activity can be a useful marker of exposure but is not a definitive marker of effect.
The Army has used the data from a study by Sim (1962), which detailed the amount of liquid VX required to cause a 70 percent depression in red blood cell cholinesterase following application to
different body regions as a quantitative indicator of regional sensitivity to HD or VX. Because cholinesterase measurement is quite variable, however, the data alone cannot be used to assign regional differences in agent sensitivity. The data may also have been compromised by prior, incidental, exposures, such as those found at Rocky Mountain Arsenal where low red blood cell cholinesterase levels could not be correlated with test exposure but were associated with carelessness in putting suits on and eating food placed on surfaces where used protective gloves had been placed.
BIOLOGICAL INTERPRETATION OF THE MIST/BRHA
The MIST generates an ensemble protection factor based on the ratio of simulant concentration outside the suit to the concentration inside the suit at each of many locations around the body (see Chapter 2). A protection factor for each specified location or region of the ensemble is calculated as the ratio of simulant detected in the absence and presence of the ensemble. Because component or swatch tests have been used to eliminate ensembles constructed from unacceptable materials, the MIST is particularly useful for detecting leaks around seams and closures.
The BRHA, combined with the MIST, simply weights the mass of simulant collected at a particular anatomic site by the surface area of a given skin region and the estimated regional variation in human skin permeability to chemical agent vapor. The results of the MIST/BRHA are still based on protection factors and require knowledge of the regional variations in skin penetration by the agent vapor. Currently, BRHA estimates of regional variations in VX and mustard vapor penetration are based on the data from Sim (1962), who studied droplets of liquid VX in humans.
The only way to validate the BRHA is through direct measurements of VX and mustard vapor penetration on excised human skin from different anatomic sites. An apparatus used at Dugway Proving Ground for generating stable vapor concentrations in swatch tests could be adapted for studies of excised skin. Regional variations in skin penetration, based on the small amount of data currently available, depends on the compound in question (Wester and Maibach, 1989) and may not even be relevant to vapor exposures (Barry et al., 1984) because most of the studies used solvents with the test chemicals. Sufficient data are not available to support the use of regional
variations in transepidermal water loss in humans or regional variations in pesticide absorption in animals as measures of regional variations in skin absorption of agent vapor in humans.
Translating data from the MIST/BRHA into the biological effect of a simulated agent exposure (physiologic endpoint) will probably require developing a simulant for each chemical agent of concern because VX, mustard, and soman, for example, have different physical properties. Passive detectors may need to be modified or abandoned altogether because no artificial membrane has yet been shown to simulate the differential permeability of the skin and its response to changing temperatures and humidity. Noninvasive measurement of a simulant in the stratum corneum or the measurement of simulant and metabolite in urine or saliva may be more practical. The experience gained from monitoring civilians exposed to pesticides and other chemicals should be used to advantage. For example, scientists at the National Institute for Occupational Safety and Health have outlined a protocol for validating diffusive sampling techniques in the laboratory and in the field (Cassineili et al., 1987).