Pharmacology Conference

Abstract III

SEX DIFFERENCES IN PHARMACOGENETICS: PHASE II ENZYMES

Richard M. Weinshilboum, M.D.

Almost all drugs and xenobiotics undergo biotransformation. Phase II pathways in drug and xenobiotic metabolism include conjugation reactions such as glucuronidation, sulfation, acetylation, methylation or glutathione conjugation. Virtually all phase II pathways for drug metabolism display large individual variations in activity that are due to the effects of inheritance (pharmacogenetics). In experimental animals, large sex-related differences in levels of phase II enzyme activities have been demonstrated. This presentation will compare and contrast data with regard to sex-related differences in phase II drug-metabolizing enzymes in humans and experimental animals, briefly describe our present knowledge of pharmacogenetic variations in phase II enzymes in humans and will present data -- when available -- with regard to the relationship of sex to pharmacogenetic variations in phase II enzymes in humans. Finally, possible directions for future research with regard to sex-related variation in phase II drug metabolism in humans will be discussed.

Pharmacodynamic Polymorphisms: Contribution to Variability in Drug Response?

Darrell R. Abernethy, M.D., Ph.D.

Considerable time and effort has been devoted to determining sources of interindividual variability in drug response over the past 30 years. This is with the implicit assumption that identifying and understanding these sources would lead to improved pharmacotherapy by permitting more effective individualization of therapy. To the present, virtually all of these efforts have focused on determining sources of interindividual variability in drug clearance and biotransformation. A usually unstated assumption has been that exposure of an effector to a given drug concentration will result in a given drug effect. In addition until recently there has been a belief that a significant genotypic or phenotypic alteration in receptors, signaling pathways, and other sites of specific drug effect would not be compatible with survival. This has been abundantly disproved with the maturation of the genome project and recent understanding that substantial variation in both genotype and phenotype of drug targets is quite compatible with life. Of interest, a number of the genotypic variants, for example of the _2-adrenoceptor, occur in the population rather frequently. Most research has focused on possible disease associations with these variants, however an obvious extension is the study of such variants to transduce a drug response. If indeed frequent variants which confer altered drug sensitivity occur, this becomes a major source, if not a predominant source of interindividual variability in drug response. The extension of such a hypothesis is that knowledge of genotype or phenotype of effector in an individual would offer a powerful tool to better predict individual drug responses. Preliminary data for one genotypic variant, endothelial nitric oxide synthase, and one phenotypic variant, splice variants of the L-type calcium channel will be used to develop the concept. Learning how to predict which variants will have importance to understand an individual drug response is a research challenge that is now possible to address.
Estrogen Receptors; New Insights, New Drug, New Issues

Donald P. McDonnell, Ph.D.

Estrogen containing medicines have been used successfully for the past fifty years for the treatment of conditions associated with menopause. Although initially considered a reproductive hormone, millions of years of clinical exposure to estrogen(s) have indicated that its influence extends to a variety of target tissues not generally considered to be involved in reproduction. Specifically, estrogen has positive actions in the skeleton, the cardiovascular system and possibly the central nervous system, activities which combine to effect a positive impact on mortality and morbidity. However, despite the medical benefits afforded by estrogen replacement therapy (ERT), the number of women who initiate or remain on therapy for greater than one year is relatively small. This is due in part to the fear that estrogens increase the risk of getting breast cancer. Consequently, it was anticipated several years ago that there was an unmet medical need for novel estrogen receptor modulators which would retain the beneficial effects of estrogens in most target organs but which were inactive in the breast. Although the perfect tissue-selective estrogen remains to be identified progress in this direction has been made. In the past year, for instance, we have seen Selective Estrogen Receptor Modulators (SERMs) enter into the clinic for the prevention of osteoporosis. Compounds of this class, which function as estrogens in the skeletal system but oppose estrogen action in the breast, represent the first step in developing the perfect hormone replacement therapy medicine. In this presentation, the complex pharmacology of SERMs and how they differ mechanistically from estradiol, the physiological ligand of the estrogen receptor, will be discussed.

Hormones and Drug Interactions: A CNS Perspective

Patricia D. Kroboth, Ph.D.

The objective of this presentation is to provide an overview of the pharmacodynamic interactions between specific CNS active drugs and progesterone and DHEA. In addition, a brief overview of drug interactions with estrogen will be provided. The focus of the interactions will be on agents that affect the GABA-receptor complex and the cholinergic system. The importance of these interactions is only partly due to the presence of clinical interactions. These interactions may also provide us with insight as to how sex can affect response to drugs. Ultimately, however, drugs serve as probes for receptors and the associated neurotransmitters that may be involved in the pathogenesis of psychiatric or debilitating diseases, or addictive behaviors. Sex differences in prevalence and severity of several psychiatric diseases exist. Depression and schizophrenia are two such diseases. Additionally, there appears to be a difference between the sexes in difficulty with successful smoking cessation. By briefly examining these drug interactions, intriguing, unanswered questions will be identified regarding the role of hormones and hormone manipulation in disease, prevention of cognitive decline, and the potential overlay of hormones and genetic differences in affecting the occurrence of disease or response to drugs.
Menstrual Cycle, Disease and Pharmacotherapy: Chronobiologic and Investigative Issues

Michael H. Smolensky, Ph.D.

The biology of young women varies in a predictable manner over time as utradian (high frequency and pulsatile), circadian (~24-hour), circamensual (~monthly) and circannual (~yearly) rhythms. Numerous studies document the role of circadian rhythms in day-night patterns of severe medical events and the exacerbation of symptoms of chronic diseases. Circadian rhythms also affect the response of patients to diagnostic tests. They also are known to affect the kinetics and dynamics of therapeutic interventions, sometimes in a profound manner, according to the time of their ingestion, inhalation or infusion (Hrushesky et. al, Temporal Control of Drug Delivery, NY Acad. Sci. 618, 1991; Touitou & Haus, Biologic Rhythms in Clinical and Laboratory Medicine, Spinger-Veralg, 1992; Redfern & Lemmer, Physiology & Pharmacology of Biological Rhythms, Springer-Verlag, 1997). Surprisingly, the study of menstrual rhythms in the occurrence and intensity of disease has received relatively little attention (Dalton, The premenstrual Syndrome and Progersterone Therapy. Yearbook Medical, 1977; Case & Reid, Arch. Intern. Med. 145:1405, 1998). Indeed, the symptoms of many common medical conditions, such respiratory (allergic rhinitis and bronchial asthma), digestive, metabolic and neurological ones, can exhibit profound menstrual cycle variation in their intensity. Several theories have been advanced to explain such differences; these include predictable-in-time alteration in endocrine status, immune function and perception. The role of menstrual cycle-dependent differences in the kinetics and effects medications in such patterns has yet to be explored. The dissolution characteristics as well as the distribution, metabolism and elimination of medications can all be affected by dosing time during the 24 hours due to influence of the human circadian time structure. Thus, the design of studies aimed at elucidating the role of menstrual stage on the pharmacokinetics and dynamics of medications and other chemical agents, such as ones found in the ambient and work environments, must be based on sound chronobiologic and pharmacologic criteria. These former include the standardization of subjects to a common activity-rest synchronizer schedule and the use of marker rhythms to denote the status of menstrual cycle stage and function, including the occurrence and quality of ovulation. The effect of birth control interventions on physiologic and bicohemical functions (Reinberg et. al, Chronob. Internat=l. 13:199-211,1996; Ferin et. al. Biorhythms and Human Reproduction, Wiley, 1972) and in turn the kinetics and effects of medications and other chemical agents also requires investigation. Study of menstrual cycle-dependencies of the kinetics and effects of chemical agents is complicated. It necessitates investigative protocols that take into account the effect of dosing time during the 24 hours as well as menstrual cycle. The results of studies suggest the same time of day need not be representative of the same circadian time during different stages of the menstrual cycle (See Ferin et. al, Biorhythms and Human Reproduction, Wiley, 1972). Even the season of study could have a significant influence on findings (See Reinberg and Smolensky, Biological Rhythms and Medicine, Springer-Verlag, 1983).
Sex Differences in Relationships Between Pharmacokinetics and Pharmacodynamics

William J. Jusko, Ph.D.

The role of sex or gender as a factor in the pharmacokinetics of drugs has become well appreciated. Women often exhibit modestly more rapid clearances of drugs metabolized by the CYP3A4 pathway (eg. methylprednisolone). They may also show alterations in disposition of drugs in relation to phase of the menstrual cycle (eg. theophylline), pregnancy (eg. caffeine), or after menopause (eg. verapamil). Women on oral contraceptives are likely to show more rapid clearances of conjugated drugs (eg. oxazepam), but have reduced clearances of many oxidized compounds (eg. prednisolone). Interpretation of literature data is sometimes complicated by the need to assess whether pharmacokinetic parameters are properly normalized for body weight differences and if the phase of the menstrual cycle was monitored.

The assessment of pharmacodynamic differences between men and women requires control of pharmacokinetic factors and utilization of appropriate methodology to relate response to plasma or biophase drug concentration. Some therapeutic areas have notable examples of marked sex differences in drug efficacy. For example, in the cardiovascular area, aspirin is less effective in women in prevention of stroke, perhaps related to sex-hormone-dependent differences in platelet aggregation. Tirilazad, a new agent for treatment of stroke, is also less effective in women. This was partly attributed to its more rapid clearance in women, but this factor has since been ruled out in more careful study.

Sex differences in analgesic effects have been well-studied, but with diverse findings. Opioids such as pentazocine show greater efficacy for pain relief in women, but the NSAID ibuprofen exhibits better responses in men with no differences in kinetics.

The immune system shows intricacies with sex differences in types of disorders, role of sex hormones, and drug responses. Women have a higher incidence of autoimmune diseases (MS, RA, SLE). Sex is sometimes a factor in drug therapy for organ transplantation with differences found in organ rejection rates and in response to drug therapy. This may be related either to differences in drug clearances (eg. corticosteroids, cyclosporine) or to intrinsic differences in lymphocyte sensitivity. Unraveling the underlying determinants of sex differences in the immune system and its responsiveness to drugs needs further study.

Support: Grant No. GM 24211 from the National Institute of General Medical Sciences
The Pittsburgh Cocktail: An experimental paradigm to investigate the selective regulation of individual cytochrome P450 (CYP) enzymes in humans

R. A. Branch, M.D., M. Romkes, Ph.D., R. Frye, Ph.D., J. Wilson, Ph.D.

A wide variety of probe drugs have been advocated to evaluate in vivo activity of individual drug metabolizing enzymes. The APittsburgh Cocktail@ contribution to this field was based on the following premises:

(1) Selection was made of substrates that are either entirely or predominantly metabolized by a single drug metabolizing enzyme in vitro and are suitable and safe to be given in single low doses in humans.

(2) A detailed pharmacokinetic study in normal subjects was used to quantitate the fractional metabolic clearance to metabolite identified from the in vitro information. This measure was used as a gold standard of in vivo enzyme activity and was compared to a variety of single point estimates obtained from urine, blood, or both. The selection of a phenotypic trait was assessed by the correlation to fractional metabolic clearance and ease of acquisition. This phenotypic marker provides an in vivo measure of drug metabolizing activity for the enzyme of interest.

(3) In order to develop a cocktail combination, it is important to ensure that simultaneous administration of multiple probes does not result in mutual drug interactions.

At the present time, the Pittsburgh Cocktail consists of caffeine, mephenytoin, debrisoquine, chlorzoxazone and dapsone as probe drugs for metabolism by CYP1A2, CYP2C19, CYP2D6, CYP2E1 and CYP3A4, respectively. The use of dapsone as a probe for CYP3A4 remains controversial and is under further study as it is also metabolized by CYP2E1 and CYP2C9. We are also determining whether or not flurbiprofen can be added as a probe for CYP2C9. The cocktail study involves simultaneous drug administration, with blood samples at 0, 4 and 8 hours and an 8 hour urine collection. Due to the ease of its acquisition, it can be conducted in an outpatient or clinic setting.

We are acquiring cumulative experience with this tool. We have shown that there is no evidence of mutual drug:drug interaction between the component drugs. Each measure has been remarkably stable within subjects over time in the absence of experimental perturbation, yet intersubject variance is wide. Genetic polymorphisms are confirmed for CYP2D6 and CYP2C19, but not identifiable for CYP2E1. Quantitation of mRNA in white blood cells or mucosal tissue has been shown to relate to in vivo measure of CYP2D6 and CYP2E1. The cocktail approach has proved a robust technique to demonstrate selectivity and magnitude of changes in individual enzyme activities in drug induction and drug inhibition studies.

In a preliminary comparison between genders in 85 normal volunteers who have undertaken the Pittsburgh Cocktail, we have confirmed the wide intersubject variation for each phenotypic trait measure in each gender. In an initial analysis, raw observations were plotted along with 90% bootstrap confidence intervals for the ratio of median in females to median in males. These intervals were based on the Abias-corrected@ approach using 2,000 bootstrap samples. Intervals whose end points fall outside of the 0.80 - 1.25 range would not be considered evidence for bioequivalence according to Schuirmann=s Atwo one-sided tests@ approach with alpha = 0.05.

Using these Agoalposts,@ only CYP2D6 and CYP2C19 show evidence of bioequivalence. If the criteria are relaxed to 2/3 to 3/2, then CYP2E1 and CYP1A2 would also exhibit bioequivalence between genders. However, in no instance were the 90% confidence limits either above or below 1.0 to achieve statistical difference between genders. This analysis is considered preliminary and merits an increase in sample size. It does, however, question the application of rigid a priori definitions of gender equivalence and non-equivalence in drug development.
Relation of Genotype and Phenotype to Clinical Outcomes

Maureen T. Cronin, Ph.D. and David Flockhart, M.D.

Emerging pharmacogenetic data that describe genotype-phenotype relationships are anticipated to permit more accurate prediction of complex phenotypes from composite genotypes. These data have great potential to diminish inter-individual variability in response to medications. The next step to making a good match between a patient=s genetic profile and optimal pharmacotherapy will depend on building up and refining this nascent database by incorporating more genetic analysis and phenotype characterization into clinical trial protocols. To make this possible, molecular genetic testing must become routinely available. The ideal automated system for molecular diagnosis of pharmacogenetic traits would isolate genomic DNA from an easily accessed patient sample, carry out required PCR/labeling reactions, score polymorphisms in the PCR products, provide a composite genotype, predict phenotype and evaluate the prediction using a clinically relevant measure of therapeutic outcome. In this complex and rapidly changing environment, genotyping is best supported by hybridization-based assays on DNA microarrays. However, this method is faced with the technical challenge of providing genotyping methods capable of responding to the accelerating pace of polymorphism discovery. Evolving reliable pharmacotherapeutic predictions depends on integrating increasingly complex genotype and phenotype information and correlating it with outcomes from various therapeutic alternatives. Experience with preliminary studies designed to predict drug metabolism phenotype from CYP2D6 and CYP2C19 genotypes underscores the need for rapid array design iteration capability. New strategies have been developed to address this problem and are being demonstrated using an NAT2 model.

Computational biology in cardiovascular medicine: insights into
gene function and drug action

Thomas J. Colatsky, Ph.D., J. Jeremy Rice, Ph.D.,
Adam L. Muzikant, Ph.D. and Gang Chen, Ph.D.

The computational modeling of biological systems provides an effective means of integrating the vast amounts of genetic, physiological and clinical data currently available to the medical researcher (Noble et al, 1999). By constructing biophysically detailed computer models of cells, tissues and organs, one can determine the effects of allelic variations on drug response, probe the role of specific proteins in disease progression, and define the effects of specific gene mutations on biological function. These computational tools also serve as powerful adjuncts to experimental studies, by allowing the researcher to explore relationships among complex sets of variables, identify critical gaps in knowledge, and generate new hypotheses and directions for future research. This presentation will focus on computational models of the heart, and their use in understanding the cardiac electrophysiological differences observed between men and women both at baseline and in response to drugs. It is well established women have a longer QT interval than men, and also show a greater predisposition to the adverse effects of QT-prolonging drugs such as sotalol (Lehmann et al, 1999). In single cell experiments, Liu et al (1998) measured reductions in two repolarization currents, IKr (rapid delayed rectifier) and IK1 (inward rectifier) that may contribute to these sex-based differences. We have simulated these reductions in IKr and IK1 using detailed models of single epicardial, endocardial and mid-myocardial cells, and then compared the resulting changes in repolarization time course to published results. Allelic variations in electrical substrate and physiological factors such as hypokalemia were also modeled to define possible arrhythmogenic triggers that might contribute to sex-based differences in drug-induced proarrhythmia. The simulations indicate that a relatively modest (10-15%) decrease in both IKr and IK1 is sufficient to reproduce the observed differences in male vs. female QT intervals. These results illustrate the utility of computational models in defining pharmacogenetic differences, and in testing and extending experimental hypotheses from pre-clinical to clinical situations.

References:
Lehmann MH, Hardy S, Archibald D, MacNeil DJ. (1999) JTc prolongation with d,l-sotalol in women versus men. Am J Cardiol, 83:354-9
Liu XK, Katchman A, Drici MD, Ebert SN, Ducic I, Morad M, Woolsey RL. (!998) Gender differences in the cycle-length dependent QT and potassium currents in rabbits. J Pharmacol Exp Therap. 285:672-679.
Noble D, Levin J, Scott W. (1999) Biological simulations in drug discovery. Drug Discovery Today. 4:10-16.

SEX DIFFERENCES IN AGING AND ALZHEIMER'S DISEASE
Gillian Einstein, Ph.D.

Epidemiological studies demonstrate that both the incidence and prevalence of many mental disorders are sexually dimorphic. This holds true for changes in cognition that accompany normal aging and neurodegenerative disease. In both cases, Hormone Replacement Therapy improves performance of females on verbal and spatial cognitive tasks. Thus, the decrease in gonadal steroids that accompanies aging in females may accelerate age related cognitive decline.
To explore the underlying mechanism of this sexually dimorphic decline, we carried out a combination of anatomical and behavioral experiments on female and male gonadectomized rats. Rats were ovariectomized at 2 months of age and treated with no estradiol, chronic estradiol for over one year, or acute estradiol. We found that in hippocampal dentate granule cells: (1) In females deprived of estrogens long-term spine density was decreased significantly, but not in males; (2) In females, acute estradiol replacement increased spine density significantly, but decreased it in males; (3) When estradiol was administered 72 and 24 hours before sacrifice, on a delayed-match-to-sample water-maze female rats showed improved memory performance as compared to rats that received only oil injections.
These results suggest that: (1) Female and male dentate granule cells age differently; (2) The aging female hippocampus is still responsive to estrogens; (3) Working memory performance is improved during the period when estradiol has induced increased spine density in regions CA1and the dentate gyrus of the hippocampus. These results have important implications for the mechanism of cognitive improvement in aging female rats and have tantalizing parallels for women's health.