Health Consequences of Exclusion or Underrepresentation of Women in Clinical Studies (II)
Leslie Z. Benet
Of course any presentation such as this must begin with the "apple pie" de rigueur pronouncement supporting equal access for women in clinical studies. I wholly support such a position and believe that women should be included in all clinical studies in proportion to their prevalence in the population experiencing the phenomena being tested. However, such a position does not address the health consequences of exclusion or underrepresentation. One must, thus, consider separately the Committee's broad definition of the term "clinical studies" to include epidemiological studies, health services research, and outcomes research, as well as randomized clinical trials.
Negative, unacceptable consequences would, and previously did, result from the exclusion or underrepresentation of women in disease epidemiology research, such as the Framingham Study, or in health services research. Similarly, it would be a mistake to exclude women from clinical studies directed toward outcomes research. The latter may be a gray area, since the underrepresentation or exclusion in outcomes research may reflect the lack of clinical investigation where therapeutic interventions are tested. In any case, there seems to be no rationale, except perhaps convention, which could justify the exclusion of women from clinical studies, where no therapeutic intervention is being evaluated.
Therefore, the focus of this presentation will relate to the health consequences of exclusion or underrepresentation of women in randomized
clinical trials which involve a therapeutic intervention. Furthermore, let me further exclude the issue of fetal toxicity, as this has been well covered in other presentations. My focus, then, is whether exclusion or underrepresentation underrepresentation of women in clinical studies involving therapeutic interventions may lead to negative health consequences for the potential female population to be administered this therapeutic intervention. Although not all therapeutic interventions will be drugs or chemicals, I will focus my remarks in that area since that is my field.
Basically, the question can be rephrased as, "Will differences in drug disposition or drug response due to gender be clinically significant?" My answer is, "Rarely, if ever."
The basis for my negative response concerning the potential for health consequences in women relates to the intersubject and intrasubject variability which patients exhibit in both the pharmacokinetics and pharmacodynamics of drugs. Most drugs on the market today exhibit a wide therapeutic index. That is, major differences exist between the dose, blood concentration, or receptor concentration necessary to achieve a positive therapeutic response versus those measurements which elicit a toxic response. However, there are a number of drugs critical to our therapeutic armamentarium which exhibit a narrow therapeutic index. That is, small changes in dose or concentration can shift a patient from an efficacious state to a toxic condition or to a state where no efficacy is exhibited (corresponding respectively to increases and decreases in dose or concentration from the therapeutic dosage regimen). It is not well recognized that although many drugs exhibit substantial intrasubject, variability as well as marked intersubject variability, all narrow therapeutic index drugs, by definition, must exhibit low intrasubject variability. If this were not true, a patient maintained on a particular dose would experience cycles of efficacy, lack of efficacy, and toxicity during a constant dosage regimen. In fact, if a narrow therapeutic index drug does exhibit high intrasubject variability, such drugs never get past Phase II testing during the drug development Process, since it is not possible to show efficacy for a particular dose (or steady state concentration).
Narrow therapeutic index drugs, however, may and often do exhibit marked interpatient variability and thus these drugs are titrated in the patient by the clinician to the appropriate dose or concentration, which, once achieved, will be maintained owing to the low intrasubject variability of the drug. It is my hypothesis, that the health consequences of exclusion or underrepresentation of women in clinical studies for narrow therapeutic index drugs will be minimal, since it is immaterial whether women differ significantly from men in their pharmacokinetics and/or pharmacodynamics of the drug, since in each case the drug will be titrated by the clinician to the appropriate dose or concentration.
Similarly, there should be few health consequences to large gender differences for wide therapeutic index or range drugs, since the variability inherent in the male population most likely already encompasses the difference
between the male and female patient populations. For the reasons stated above, it seems obvious that we would not have noted any major gender differences in terms of drug kinetics or dynamics for drugs going on the market even though they had received inadequate testing in women. I believe this to be the case.
Although I suspect there are only minimal health consequences of exclusion or underrepresentation of women in clinical studies, I favor inclusion of women as early as possible in drug development, including Phase I studies, but definitely in studies in special populations (e.g., aged, subjects with renal or hepatic disease) and of course in Phase III studies when the treatment will be utilized in women. New techniques in clinical pharmacology, such as population pharmacokinetics, where a few systemic drug concentration measurements are made in each patient during Phase III, can be utilized to identify the presence or lack of drug-drug and drug-disease interactions, and to identify other unanticipated variabilities such as metabolic heterogeneity, as well as gender differences. Such information will be useful in understanding the drug disposition process as a function of gender, age, disease, environment, and chronopharmacologic effects.
An example of such a gender dependent finding relates to the explosion of knowledge about drug metabolizing isozymes which has occurred during the last few years. We have identified, purified, characterized, and cloned specific isozymes with emphasis on the major metabolic system—the cytochromes P450. At least seven, and maybe more, gene superfamilies have been identified in humans. Furthermore, innumerable isozymes of the cytochromes P450 and other enzyme systems are continuing to be identified, both in humans and in the animal species used in preclinical testing. Up until now, no marked gender difference in the presence of these isozymes has been identified in humans, although certain animal species do exhibit marked gender-specific differences in various isozyme levels. In one case, for a very important human metabolic isozyme, identified as P45O-3A4, a potential gender difference has been suggested. Earlier this week at the National Institutes of Health's "Workshop on Menopause," I hypothesized that clearance in postmenopausal women is lower than that in premenopausal women for many drugs owing to the decrease of a particular metabolic isozyme. I suggested that the isozyme which is decreased in menopause is P45O-3A4. I further hypothesized that progestational agents may restore P45O-3A4 to premenopausal levels. If my hypotheses are correct, then a gender difference due to menopause, which previously may have been confounded with an age effect, could be important in evaluating and optimizing new drug therapies, as well as in correctly defining the disposition of drugs in postmenopausal women not receiving progestational agents. The hypotheses which I made could only be developed as a result of our increasing understanding of the importance and specificity of the metabolic isozymes. But, the results above could not be obtained if studies were not carried out comparing pre- and postmenopausal women.
Therefore, in conclusion, although I feel that there are minimal immediate health consequences to the exclusion or underrepresentation of women in clinical studies, the advancement of therapy and our understanding of the basic processes involved can only be enhanced by inclusion of women at all levels of clinical investigation.