1
Introduction
BACKGROUND
The U.S. Army Medical Research and Materiel Command is addressing the increasingly contentious issue of the effects of exposures to drugs, biologics, and chemicals and their possible interactions. U.S. troops routinely receive a number of vaccines, and when they are deployed they may be given antimalarial drugs, anti-biological warfare drugs and vaccines, and additional vaccines to protect against indigenous diseases. They may be further exposed to sundry chemicals, such as permethrin or N,N-diethyl-m-toluamide (DEET), environmental contaminants, and warfare byproducts; some personnel will also be using individually prescribed and over-the-counter medications. Although the adverse effects of most single products have been relatively well studied (for example, in the data submitted to the Food and Drug Administration [FDA] for approval of a new drug), it is largely unknown whether and how their combined use may provoke unanticipated interactions. For the purposes of this report, agents are said to interact if the presence of one agent affects an exposed individual's reactivity to other agents.
The U.S. Army contracted with the Institute of Medicine to address the issue of the interactions of drugs, chemicals, and biologics. The Institute assembled a committee of experts in pharmacology, drug safety assessment, immunology, vaccinology, epidemiology, biostatistics, occupational health, environmental health, toxicology, and biomedical administration. The names of potential committee members were sought from a variety of sources, and the final committee roster was approved by the chairman of the National Research Council.
Using a broad range of sources, the committee informed itself on this topic and then deliberated to reach consensus recommendations. The committee reviewed the available scientific literature and heard personal testimony from officials of the U.S. Army, the FDA, the Centers for Disease Control and Prevention (CDC), and the British Ministry of Defence (a list of presenters can be found in the Appendix).
THE COMMITTEE'S CHARGE
The committee was asked to address the following questions:
-
What are the drugs, biologics, and chemicals that U.S. military forces currently receive as prophylactic-preventive agents, and what additional prophylactic-preventive agents are planned?
-
What does the published scientific literature tell us about the health effects of combinations of these prophylactic-preventive agents in the U.S. military, in other human populations, or in model systems (e.g., animal, in vitro, and computer)?
-
Do the experiences of the militaries of other nations shed any light on issues #1 and #2?
-
If there are important gaps in our knowledge, where are they and how would they best be filled?
-
Should, and in what way, the Army modify or expand its development and utilization strategies to ascertain possible interactions of prophylactic-preventive agents?
-
Should the Army undertake any new programs to provide information on drug, biologic, and chemical interactions?
SCOPE OF THE PROBLEM
The number of potential drugs, chemicals, and biologics to which military personnel may be exposed is quite large (see Chapter 2), and a complete study of their interactions would by necessity involve examination of all their possible combinations. For example, in the case of 25 agents, there are 225 – 1 (or 33,554,431) combinations. Even if it were possible to study all combinations of agents in epidemiologic or animal model systems, it is unlikely that such a strategy would work. Many confounding factors would be encountered in epidemiologic studies; for example, host susceptibility factors such as age, race, sex, comorbid conditions, and other effect modifiers could affect the results.
With exposure to drugs, biologics, and chemicals being such common occurrence in the military, it is critical to know whether coexposure to two or
more of these agents yields a different health effect than what would be expected by adding each separate effect. Reports of drug interactions began to appear in the medical literature in the 1940s when para-aminobenzoic acid was found to reduce the renal tubular excretion of salicylates (Gilman et al., 1990). Since then, many drug interactions have been discovered. There have also been major advances in the understanding of the mechanisms and pharmacokinetic principles of drug interactions.
However, little or no information exists on interactions among drugs, biologics, and chemicals. The committee's own Medline search did not turn up any notable articles related to its charge other than the few noted below. Additionally, discussions with liaisons from the Canadian and British militaries provided no additional insight regarding interactions.
Nonetheless, this does not mean that interactions have not been the subject of separate study in the disciplines of pharmacology, toxicology, statistics, and epidemiology. The committee has, indeed, drawn briefly on material from each of these disciplines as necessary. As an aid to the reader who would like to pursue further discipline-specific findings, the committee offers the following short discussion.
There is a vast literature on drug-drug interactions; for example, standard compendia can be used to identify potentially interacting drug combinations (Hansten and Horn, 1993; Tarto, 1995). However, the bulk of the more than 10,000 publications on drug interactions in the medical literature primarily comprises pharmacokinetic or pharmacodynamic studies with humans and animals, case reports, review articles, animal studies, and in vitro studies. Thus, the existing literature provides very little information about how often drug interactions actually occur in humans or how often they produce clinically meaningful adverse effects. Although epidemiologic methods have demonstrated strong utility in the detection and quantitation of drug reactions in general, they have not been applied often in the study of interactions. There are also no meaningful estimates of the importance of drug interactions as a clinical problem (Jankel and Speedie, 1990; Jankel and Fitterman, 1993).
The existing literature also provides some information on the potential for interactions among certain biologics to which military personnel may be exposed. With regard to vaccine interactions, Grabenstein (1990) reviewed interactions between vaccines, vaccines and immunoglobulins, and vaccines and other drugs. Similar to drug interactions, vaccine interactions can lead to potentiated or diminished effects, enhanced or impaired elimination, or other metabolic or pharmacologic effects. Vaccines, for example, have been shown to affect the metabolism of other drugs, possibly by interfering with human liver cytochrome P450 isozymes (Kramer and McClain, 1981). More recently, Gizurarson (1996) summarized the known and suspected vaccine-vaccine and vaccine-drug interactions.
The toxicology literature on interactions is considerable, although such work is more apt to be indexed under “complex mixtures” than “interactions,” per se. One recent review article is by Mauderly (1993), who cites a more detailed treatise, the NRC report on complex mixtures (NRC, 1988). It is worth noting that major research efforts directed toward the study of complex mixtures date back to the mid-1970s. Studies of the effects of tobacco smoke and of engine exhaust provide examples of studies of complex mixtures. Toxicologists generally have a much broader range of study tools than investigators who are limited to studying humans, but this advantage is offset by the disadvantage that data not derived from human sources must be extrapolated to humans, with varying degrees of uncertainty. Nevertheless, the considerable advantages of toxicological studies led the committee to recommend their use in addition to studies of humans (see Chapter 5 and Chpater 6).
Statistics and epidemiology each have made their contributions to the study of interactions. Over the years, statisticians have developed a considerable body of work on the subject of interactions. The statistical term interaction denotes a departure from additivity in some (linear) mathematical model of exposure and response, and thus much of the statistical literature on departure from additivity in models is pertinent. In Chapter 5, the committee has made particular mention of newly developed statistical techniques for the study of interaction.
Epidemiology provides at least two definitions of the term interaction, one the more traditional “statistical” definition above and the other arising from the counterfactual model of effects (see Greenland, 1993). The committee has operated using the more traditional definition of interaction in this report. Epidemiologists have rightly emphasized the practical difficulties in assessing interactions. For example, the power to detect interactions in epidemiologic settings is typically much less than the power to detect single-factor effects, and the impact of measurement errors on interactions is not nearly as well studied as is the impact of measurement errors on main effects. Given the difficulty in studying interactions using epidemiologic data, some have recommended that “design and analysis is best focused on accurate estimation of the entire dose-response surface relating incidence to covariates, rather than on isolated aspects of this surface, such as statistical interaction.” (Greenland, 1993, p. 64)
REPORT FOCUS
To reduce the problem to a more manageable level and as an organizing framework for this report, the committee advocates a categorical approach to the study of interactions. This approach would categorize interactions into three classes—known, potential, and unknown—so that different strategies may be applied to each class.
In the remainder of the report, Chapter 2 discusses the current prophylactic agents given to military personnel, Chapter 3 briefly describes the surveillance tools currently available to the military, Chapter 4 presents the strategy for identifying and dealing with the various types of interactions, Chapter 5 addresses in greater detail the study and management of the various types of interactions, and Chapter 6 contains the committee's findings and recommendations. A glossary of technical terms and acronyms is also provided.