Frequently, pharmacomechanisms are not fully elucidated. Therefore, drug use is linked to an elevated interindividual diversity of effects, whether therapeutic or adverse, and the role of biological sex has as yet unrecognized and underestimated consequences. A pharmacogenomic approach could contribute towards the development of an adapted therapy for each male and female patient, considering also other fundamental features, such as age and ethnicity. This would represent a crucial step towards precision medicine and could be translated into clinical routine. In the present review, we consider recent results from pharmacogenomics and the role of sex in studies that are relevant to cardiovascular therapy. We focus on genome-wide analyses, because they have obvious advantages compared with targeted single-candidate gene studies. For instance, genome-wide approaches do not necessarily depend on prior knowledge of precise molecular mechanisms of drug action. Such studies can lead to findings that can be classified into three categories: first, effects occurring in the pharmacokinetic properties of the drug, e.g. through metabolic and transporter differences; second, a pharmacodynamic or drug target-related effect; and last diverse adverse effects. We conclude that the interaction of sex with genetic determinants of drug response has barely been tested in large, unbiased, pharmacogenomic studies. We put forward the theory that, to contribute towards the realization of precision medicine, it will be necessary to incorporate sex into pharmacogenomics.
In the blockbuster one-dose-fits-all drug model, trials have been large enough to show statistically significant responses in the complete sample, without considering which participants will benefit from the drug or not. In many instances, however, a given drug may typically be efficacious in only 30–60% of the patient population . Importantly, there are major differences in how we all react to drugs as individuals, with some patients having side effects and others not. In fact, adverse drug reactions have been estimated to account for up to two-thirds of drug-related hospital admissions, with an overall incidence of 6.7% having severe consequences; they are among the leading causes of death in hospitalized patients [2–5].
Several factors may influence drug effects, including age, sex, lifestyle, nutritional state and environmental factors. In addition to disease susceptibility [6–24], classic pharmacogenetic approaches designed to investigate pharmacological consequences of single-gene mutations have shown that multiple genetic variations also affect drug responses [25–44]. Previously, a female-specific over-representation of rare genetic variants in genes coding for drug-metabolizing enzymes has been reported, which was associated with a modest increase in risk of acute myocardial infarction in women . Based on this, it is to be expected that sex-specific occurrence of genetic variation would also lead to sex-specific drug therapeutic and adverse effects (Figure 1). However, this has been explored little and the role of genetic variants as sex-dependent susceptibility factors to drug responses is not clearly understood. Targeted studies considering the impact of sex on individual genetic polymorphisms that may affect pharmacokinetics and pharmacodynamics have been reviewed elsewhere [45,46].
Sex-specific (epi)genetic regulation of drug responses
In contrast to targeted pharmacogenetics of single candidates, a systems approach employing pharmacogenomics to investigate the interaction between sex and genetic variation in drug therapeutic and adverse effects is innovative and powerful. This is mainly because it facilitates the simultaneous assessment of the effects of sex on numerous genetic factors and gene networks in drug efficacy and toxicity. This is of great importance, as large-scale, genome-wide investigations of disease susceptibility have revealed loci with sex-specific association [47–55]. In addition, most interindividual differences are not caused by a single genetic variant, but rather by variations in several genes, and the influence of other factors gives rise to the multifactorial nature of the drug response [56,57]. Furthermore, such genome-wide strategies are necessary when we consider that genetic variants are mostly found in regulatory regions and not in protein-coding regions . The integration of clinical biomarkers and pharmacogenomics with sex-related variables and genetic variables may contribute towards the realization of precision medicine. In the present review, we focus on the influence of sex on the pharmacogenomics of drugs used in cardiovascular therapy.
DEFINITION AND BRIEF HISTORICAL OUTLINE OF PHARMACOGENOMICS
The recognition of interindividual differences in drug efficacy and side effects gave rise to pharmacogenetics, which is the science that investigates single-gene variation affecting drug responses . Major improvements in tools and technologies in the 1990s and early 2000s facilitated the transition to pharmacogenomics from pharmacogenetics. Pharmacogenomics studies the interaction between an individual's genetic makeup and the response to a drug at the whole-genome level . In essence, pharmacogenomics does not consider variation in just a single gene, but rather the effect of variation in the function and interaction of all genes in the genome on drug responses. This enables the simultaneous assessment of pharmacokinetics and pharmacodynamics, or the combination of both. Consequently, pharmacogenomics incorporates genomics and proteomics in the assessment of the effects of multiple genes on drug response . The ultimate goal of pharmacogenomics is the development of appropriate therapies on the basis of genetic information for each individual patient, thereby contributing towards the realization of precision medicine. The term ‘pharmacogenomics’ first began appearing in the 1990s and the initial appearance of publications in PubMed using this term was in 1998 .
SEX DIFFERENCES IN CARDIOVASCULAR PHARMACOTHERAPIES
The biological differences between men and women in anatomical and physiological features, such as body and organ weight, body composition and fat, gastrointestinal tract factors, liver metabolism and renal function, affect the therapeutic effects of drugs and may do this in a sex-specific manner. Similarly, the manifestation of adverse drug reactions may also differ between men and women. Several cardiovascular drugs exert sex-specific therapeutic and toxic effects. These have been extensively reviewed elsewhere [45,46,62–64]. In the present review, we provide a brief overview summarizing selected pronounced effects. Compared with men, women more often develop torsade-de-pointes ventricular tachycardia during the administration of drugs that prolong cardiac repolarization [65–68] or dry cough when treated with angiotensin-converting enzyme (ACE) inhibitors , and they experience greater toxicity due to β blockers [70,71]. On the other hand, men may experience painful gynaecomastia while being treated with angiotensin receptor blockers . Sex differences in pharmacokinetic and pharmacodynamic properties of drugs are expected to be the principal factors leading to sex-specific drug therapeutic and deleterious effects. On the basis of this, it would be clinically relevant to identify the contributing genetic factors and to understand the influence of sex. This, in turn, would lead to pharmacotherapies applied in a sex-specific manner (Figure 2).
Improved application of pharmacotherapies through sex-specific pharmacogenomics
ROLE OF SEX IN PHARMACOGENOMICS OF CARDIOVASCULAR DRUGS
To date, the interaction of sex with genetic variation that affects the pharmacokinetics and pharmacodynamics of cardiovascular drugs, as well as adverse drug reactions, has scarcely been investigated in unbiased pharmacogenomic approaches. In the present review, we discuss the factors that may have contributed to the paucity of relevant studies summarizing available data.
Early on, a considerable amount of work focused on pharmacogenetic approaches for the study of the drug-metabolizing cytochrome P450 (CYP) enzymes, because of their leading role in biotransformation. The levels of several cardiovascular drugs in the blood depend on the intrinsic activity of the liver biotransformation enzymes, due to their interaction which powers activation or elimination of these substances. In this process, some drugs, such as clopidogrel, losartan and simvastatin, may be metabolized to a more active form, mostly through CYP2C9, CYP2C19 and CYP3A4 . Furthermore, the elimination of these drugs and others, such as torasemide, is also driven by biotransformation. Consequently, targeted pharmacogenetic studies of single-candidate genes have been performed on CYP members in order to assess the influence of genetic variation and other factors, such as ethnicity and environmental factors, on drug kinetics. However, the interaction between sex and CYP genotype has not been investigated systematically. Nevertheless, findings of single-gene studies investigating, for example, genetic polymorphisms of CYP2C9 and CYP3A4/5 have revealed a significant influence of sex on CYP enzyme activity [74–79].
In a genome-wide study analysing the association between single nucleotide polymorphisms (SNPs) and ADP-stimulated platelet aggregation in response to clopidogrel administration, sex was included as a covariate, and there was even a strong tendency for women to respond less well than men . Several SNPs within the CYP2C18–CYP2C19–CYP2C9–CYP2C8 cluster were associated with diminished clopidogrel response . However, presentation of any findings on the potential interaction between genetic variants and sex is missing. Nevertheless, in another genome-wide association analysis, applying multivariate regression models adjusting for known genetic and non-genetic predictors of warfarin dose, such as age and sex, an SNP in the CYP4F2 gene was identified .
Another drug-metabolizing enzyme with an important role in pharmacokinetic parameters is the organic anion transporting polypeptide (OATP) 1B1, encoded by the SLCO1B1 gene, which is implicated in statin metabolism. It must be noted that, in this case, sex significantly affects pharmacokinetic characteristics of specific SLCO1B1 polymorphisms. In particular, two different, targeted, pharmacogenetic studies on statins, i.e. atorvastatin, simvastatin and pitavastatin, revealed sex-specific effects of genetic variants within the SLCO1B1 gene on statin treatment efficacy [82,83].
The approach of targeted pharmacogenetic studies of single-candidate genes is mainly based on the hypothesis that different polymorphisms have an influence on the drug-metabolizing enzyme activity, thereby leading to interindividual variation. However, several other factors might play a key role in overall enzyme function, indicating that not only are structural differences important, but also the overall quantity and availability of these enzymes may be crucial. Therefore, the influence of non-genetic factors, such as hormones, or other factors regulating protein synthesis, activation or degradation, should be taken into account.
Hormonal fluctuations due to menstruation, pregnancy and menopause may have profound drug effects in women . In fact, oral contraceptives inhibit CYP enzyme activity . As a result, fluctuations in the levels of sex hormones, such as oestrogens, particularly in premenopausal women, could potentially lead to fluctuating gene expression, which would differ from that in men [86,87]. To this extent, sex differences in the expression of several members of the CYP enzymes have been reported [88–93]. Furthermore, sex hormones and their receptors directly regulate gene expression [94–97], including genes coding for drug-metabolizing enzymes, such as CYP1A2 and CYP4A12 [98–101]. Of note, we have shown that oestradiol-mediated gene regulation may be sex specific [94,95,102]. Furthermore, polymorphisms in the genes coding for oestrogen receptors ESR1 and ESR2 may affect their function as transcription factors, thereby altering gene expression, but they may also affect circulating oestrogen levels . It is interesting that polymorphisms in the ESR1 gene have been associated with coronary heart disease risk in men [104–106].
Epigenetic changes may also lead to heritable variability regulating gene expression, thereby altering the function of downstream effectors . Some DNA methylases and demethylases responsible for epigenetic regulation, such as members of the Jarid family, are sex chromosome-linked and under the control of oestrogens, thereby contributing to sex-specific regulation of gene expression. In addition, the expression, as well as the function, of a gene may be changed by environmental factors, such as sleep and diet, that exert major effects [108,109]. In fact, environmental factors may influence epigenetic regulation in a sex-specific manner . Therefore, due to these complex interactions, there is a need for a transition to next-generation pharmacogenomic approaches that explore the relationship between sex and genome organization and function and incorporate epigenetic regulation combined with other factors and environmental exposures.
In addition, although several drug-metabolizing enzymes, particularly members of the CYP family, have been thoroughly studied and are known to be involved in interindividual variation, others with a previously unknown role in drug metabolism could contribute to pharmacokinetic differences, which could exert their effects in a sex-specific manner. Whole-genome studies would facilitate much more visibility, through the analysis of many involved genes and polymorphisms. Besides, sex should be systematically analysed as a covariate in order to identify whether it is a significant predictor of genetic variation implicated in pharmacokinetic variability. This would lead to knowledge that could enable a clinically oriented and relatively easy way to adapt medication to patients based on sex.
Together, drug responses are influenced by sex. Although the patient's sex is known to the treating physician, the impact of sex on the functional status of the potentially relevant, drug-metabolizing enzyme is not known. Besides, a drug may be metabolized by alternative pathways [111,112]. Therefore, genome-wide, sex-specific, pharmacokinetic testing through pharmacogenomics will address sex differences in the occurrence of adverse drug reactions or the lack of efficacy.
In the last few years, genome-wide association studies have been increasingly performed to identify novel genetic variants and their contribution to the pharmacodynamics of some common drugs. These studies are of great interest in the domain of cardiovascular drugs, where interindividual variability is high and pharmacomechanisms are still unclear. In fact, due to the multifactorial nature of cardiovascular diseases underlain by complex genetic causes, similar pathologies may be caused by a different set of genes between two people. Therefore, considering the use of a drug approach targeting the underlying genetic cause, different pharmacotherapies would be effective for these two people. Two of the most studied systems in pharmacodynamics are the lipid-lowering statins and the hypertension-lowering drugs. Yet again, the role of sex has not been fully explored and its effects are currently obscure.
Notably, sex has previously been shown to influence the therapeutic efficiency of statins and, to a certain extent, this even occurs independently of other factors, such as baseline lipid levels, socioeconomic status, cardiovascular comorbidities and associated risk factors . Although statin therapy is an effective intervention in the secondary prevention of cardiovascular events in both sexes, a meta-analysis found no benefit to stroke and all-cause mortality in women . However, the statistical approaches used and the degree of women's participation in the clinical trials may mask any beneficial effects [115,116]. It is, therefore, clear that it is necessary to have a better understanding of any sex-specific response to statin therapy.
At a single-candidate gene level, previous studies have revealed a significant interaction between sex and genetic variation in the response to statin therapy. These include common polymorphisms in the oestrogen receptor 1 (ESR1) and apolipoprotein A1 (APOA1) genes, which exert a sex-specific effect on high-density lipoprotein- and low-density lipoprotein (LDL)-cholesterol, total cholesterol and triacylglycerol responses [117,118]. However, in another study, apolipoprotein E (APOE) genotypes affected baseline concentrations of LDL-cholesterol and its reduction by statin therapy, but no significant gene–environment interactions were observed with sex . The lack of a positive association with sex may be the result of an inadequately powered sample, because approximately 20% of the individuals analysed in each genotype were women.
Among genome-wide association analyses on statins apparently considering sex, a study, aimed at the identification of genetic variants associated with reduction of a differential coronary heart disease event by pravastatin therapy, found that event reduction by therapy differed according to genotype for an SNP in the DnaJ heat-shock protein family member C5 beta (DNAJC5B) gene . This analysis included individuals from three randomized, placebo-controlled trials of pravastatin, i.e. Cholesterol and Recurrent Events (CARE), West of Scotland Coronary Prevention Study (WOSCOPS) and PROspective Study of Pravastatin in the Elderly at Risk/PHArmacogenomic study of Statins in the Elderly at risk for cardiovascular disease (PROSPER/PHASE). It is interesting that, although sex was included as a covariate in the description of the statistical methods, the results of this analysis were not presented or discussed. Besides, only in the last study, 40% (with events) and 53% (no events) of the participants were women, whereas in the first and second studies, 13% and none of the participants, respectively, were women.
Another genome-wide association study of the LDL-cholesterol response to statins identified significant associations at ATP-binding cassette subfamily G member 2 (ABCG2), lipoprotein a (LPA) and APOE genes . This study included sex as a covariate and it even addressed whether genetic interactions were contributing to the variance in LDL-cholesterol reduction with rosuvastatin, but no SNP in the study met the genome-wide criterion of P<5×10−8 for a sex interaction . It has to be noted that female participation in the study was 31.6%. In a similar manner, another whole-genome study of LDL-cholesterol response to statins, using data from participants in the Collaborative Atorvastatin Diabetes Study (CARDS), the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) and the observational phase of ASCOT, did not find any gene–sex interaction for all the top SNPs reported . However, only the first trial had a 47% female participation, whereas, in the other two trials, 11% and 13% of the participants, respectively, were women. An additional study identified genome-wide associations for the activity of lipoprotein-associated phospholipase A2 at baseline and 12 months after statin therapy . However, none of the genome-wide associations showed evidence of interaction with sex . Considering that the female participation in the study was 32%, the null finding could be due to power constraints. Therefore, replication of all these findings with sufficient numbers of women would be necessary before any conclusions could be reached.
In a further genome-wide approach analysing for association with statin response, although sex was included as a covariate and actually had a significant effect on response combined with age, its interaction with genetic variation was not explored, most probably because merely 18% of the participants in the subset chosen for whole-genome analysis were women . Similarly, no gene–sex interactions were tested in another genome-wide association examining predictors of response to statins combining individuals from three randomized trials . Two additional studies reporting no or very small effects lacked any sex-specific analysis [126–127].
The second pharmacodynamic field, in which clinical data show significant differences between men and women, is antihypertensive therapy . A previous single-candidate gene approach to polymorphisms on the angiotensin I-converting enzyme 2 (ACE2) gene revealed a female-specific risk for hypertension, as well as an implication in blood pressure regulation in response to captopril therapy . However, the results of recent, unbiased, genome-wide association studies are not very encouraging.
Testing the hypothesis that SNPs associated with blood pressure or hypertension would also be associated with blood pressure response to antihypertensive drugs, a study that adjusted for sex, among other factors, did not identify any polymorphisms that would have a significant effect alone . Nevertheless, gene score analyses combining several alleles were more strongly associated with the blood pressure response to these drugs than the individual polymorphisms, thereby demonstrating the importance of whole-genome analyses for phenotypes with multiple genetic contributors . Similarly, a large genome-wide study of 21 267 individuals, with a predominant female participation of up to 68%, also did not find any individual loci of interest for potential clinical routine . In contrast, suggestive evidence for genetic variants influencing antihypertensive drug responses was reported in a recent study . However, the effect of sex could not be investigated, because only male participants were included.
Overall, even though most studies mention correction for sex through matching or baseline adjustments, the predictive value of sex and its interaction with genetic variation have not been properly assessed. Future pharmacogenomic studies should therefore address these issues, because such approaches will contribute to the realization of precision medicine and may provide mechanistic explanations for the observed clinical sex-specific variation.
Adverse drug reactions
It is important to note that, although non-responders do not experience any benefit, they are still at risk from drug side effects. The biological differences between men and women may lead to sex-specific adverse drug reactions. Large-scale pharmacogenomic studies for the identification of loci that affect susceptibility to adverse drug reactions, incorporating sex-specific analyses, are rather limited.
An important adverse effect is myopathy occurring in association with statin therapy. Particularly at older ages, women are at a relatively higher risk of developing statin-associated myopathy than men . A meta-analysis of nine studies concluded that a polymorphism in the solute carrier organic anion transporter family member 1B1 (SLCO1B1) gene is associated with an increased risk for statin-related myopathy, especially in individuals receiving simvastatin . However, female participants were not included in each of these nine studies, thereby impeding the assessment of sex effects in this meta-analysis. Nevertheless, different odds ratios between men and women were reported for myopathy associated with a polymorphism in the SLCO1B1 gene, in a subgroup analysis in one of those nine studies . Another study  included sex as a covariate but did not replicate a previously reported  polymorphism in the glycine amidinotransferase (GATM) gene, associated with the risk of statin-induced myopathy.
A side effect of angiotensin-converting enzyme (ACE) inhibitors is persistent cough, which occurs more often in women than in men . Even though the underlying mechanisms remain largely unclear, different genetic polymorphisms have been associated with cough related to this drug class . Of great importance, the reported effect of the polymorphism in the ACE gene is sex specific, having a protective effect in men and increasing the risk in women . A whole-genome study identified a polymorphism in the potassium, voltage-gated, channel-interacting protein 4 (KCNIP4) gene . However, in analyses stratified by sex, the odds ratios were similar for men and women . Another side effect of ACE inhibitors is angio-oedema. Although a genome-wide association analysis did not find any genetic variant with a large effect size , a previous targeted candidate gene study reported a sex-dependent association of a polymorphism in the X-prolyl aminopeptidase P 2 (XPNPEP2) gene with ACE inhibitor-related angio-oedema .
One of the most serious adverse drug reactions is an increase in QT duration, which can lead to torsades de pointes [142–144]. Women are at higher risk than men of developing torsades de pointes during administration of cardiovascular drugs [65–68]. Targeted candidate gene studies have identified SNPs associated with drug-induced QT prolongation [145–147]. In these studies, sex was included as a covariate. However, in only one of them was a sex-stratified analysis performed, but no major effect modification by sex was identified for a statistically significant polymorphism in the potassium, voltage-gated channel, subfamily E regulatory subunit 1 (KCNE1) gene . On the other hand, a large-scale, genome-wide analysis composed of approximately 35 000 participants from 10 cohorts did not identify any common genetic variant that significantly influenced the association between QT and drugs in four therapeutic classes previously associated with QT prolongation . Although there was a predominant participation of women, with sex included as a covariate, no analysis stratified by sex was reported. In another genome-wide association analysis for common predisposing genetic variants of drug-induced torsades de pointes, in which sex was included as a covariate, no common variants with large effect sizes were identified, but there were limitations with numbers and selection of patients, and the study was not powered to detect interactions .
Collectively, there are considerable controversies in the available studies and only a relatively small fraction of the variance in drug-related side effects may be explained by the identified polymorphisms. Consequently, unbiased, whole-genome, pharmacogenomic studies incorporating gene–sex interaction analyses combined with other interactions may contribute to a better understanding of the ‘missing heritability’ of adverse drug reactions. It should be noted that there is an overall lack of understanding of several drug-related disorders. Furthermore, many major journals have minimal requirements for publishing adverse event reports and some do not have any . Therefore, following the guidelines for reporting adverse events properly will be crucial for a better understanding of the safety of medicinal products .
RELEVANCE OF EXPERIMENTAL ANIMAL MODELS
The importance of animal models in the drug development process is indubitable. Similarly, there is great value in the integration of interindividual variability of drug responses into basic research. In fact, animal models can assist in elucidating whether genetic variation is part of a drug–response mechanism or a biomarker of drug response.
Along this line, several studies have employed animal models to identify genetic polymorphisms that may underlie susceptibility to adverse drug reactions. For example, a panel of 36 inbred mouse strains was used to model genetic diversity and identified the gene CD44 as a modulator of susceptibility to paracetamol-induced toxicity . However, only male mice were used in the study. The diversity-outbred mouse model, which is another model of genetic diversity with each mouse having a unique genotype, was employed to measure peripheral blood and bone marrow micronucleated reticulocytes, after 4 weeks of exposure to benzene inhalation, displaying interindividual variation in toxicity response . In this study also only male mice were included.
In addition to the studies with mutant mouse lines and those using highly diverse animal populations to explore mediators within causal pathways, several studies employing experimental animal models have shown sex differences in genes and pathways of drug metabolism and elimination. In fact, the expression of more than 1000 hepatic genes is sex-specific in mice and rats [153–155]. These include the male-specific hepatic Bcl6, the levels of which show an unusually high interindividual variability . Moreover, the activity of different members of the CYP enzymes can also be sex specific [157,158]. As already discussed, sex steroid hormones, such as 17β-oestradiol, influence the levels of drug-metabolizing enzymes. However, we recently reported that the effects of 17β-oestradiol may differ dramatically as a function of genetic variation associated with the C57BL/6J and C57BL/6N mouse strains . Importantly, toxicity response differed significantly between these two strains, which are widely used in mouse genetic studies, with more pronounced effects actually occurring in females than in males .
Collectively, there needs to be a transition from the traditional testing paradigm for rodent studies based on a single sex and strain to a new testing paradigm for population-based rodent studies incorporating multiple strains and both sexes. Such approaches will contribute to the identification of the genetic basis for a sex-specific response.
CHALLENGES AND OPPORTUNITIES FOR CLINICAL IMPLEMENTATION OF SEX-SPECIFIC PHARMACOGENOMICS
Barriers may be inherent to pharmacogenomics, sex/gender-based medicine or a combination of both. In the first instance, obstacles have been related to the availability and costs of tools and technologies. The SNP-based chips that have been used cover only a certain number of genes; rare variants are poorly represented on such genotyping arrays, and all this has been paralleled with high costs. However, whole-genome sequencing has the benefit of identifying rare or novel variants, and the rapidly decreasing costs of sequencing the human genome can facilitate implementation in the clinic. On this basis, it would be of great importance to employ pharmacogenomics to investigate whether genetic variation occurs and contributes to drug responses in a sex-specific manner. In fact, this should be investigated not only in drug-metabolizing enzymes and drug transporters, but also in drug targets, e.g. target receptors, and associated gene expression in target organs, e.g. the heart, thereby leading to cell- and tissue-specific data. This is of great relevance, as we have previously reported sex-specific gene and protein regulation in the heart [102,161,162]. To achieve these goals, next-generation tools and methods are needed to explore the relationship between sex and genome organization and function. In fact, the investigation of sex chromosomes in genome-wide association studies has been a challenge. However, novel methods have enabled the incorporation of sex chromosomes and the study of sex differences into genome-wide association analyses [50,52,163–165], which have led to the discovery of sex- and sex chromosome-specific associations [47–55]. Another problem has recently emerged due to sample mislabelling for sex, which obviously affects the validity of sex-specific comparisons and the conclusions based on these . Therefore, it is important to keep this in mind, particularly in a meta-analysis of publicly available datasets.
In addition, it will be necessary for future physicians to be better trained in pharmacogenomics and sex/gender-based medicine. On one hand, pharmacogenomic approaches generate a large amount of data, the analysis and, ultimately, the clinical usage of which may be hindered due to lack of training and specific skills. University programmes of study and advanced courses for pharmacogenomics are limited. On the other hand, the discipline of sex/gender-based medicine has emerged relatively recently, and there is still a misconception that it deals with only women's health . This, among other things, has hampered the inclusion of sex/gender-specific aspects into medical and interprofessional education. However, the integration of sex/gender into medical education will improve the education of future caregivers . In turn, this will lead to better medical care of both men and women and contribute to the realization of precision medicine.
A change in regulatory, funding and industrial policies will also be crucial for the implementation of sex-specific pharmacogenomics in clinical medicine. Incorporation of sex-specific data into genetic databases, to understand sex-specific variability in pharmacokinetics and pharmacodynamics, is necessary. In addition, drug labelling can include sex-specific information on pharmacogenomic biomarkers. The U.S. Food and Drug Administration (FDA) has provided a list of approved drugs with pharmacogenomic information in their labelling. Analysing these data, we were interested to discover that so far such information has been included for 12 pharmacogenomic biomarkers in the labelling of cardiology-related drugs (Figure 3). Although some sex-specific information is available, inclusion of sex-specific response variability and risk for adverse events in, among others, drug labelling for pharmacogenomic biomarkers would provide a framework for sex-specific and more appropriate medical care. In the European Union, integration of sex into research and innovation is an objective of Horizon 2020, which is the biggest European research and innovation programme so far. This should also concern the framework for clinical trials, where the under-representation of women is a critical bottleneck for the understanding of sex-specific effects in pharmacogenomics.
Pharmacogenomic biomarkers in drug labelling
Together, a multilevel approach for the integration of sex into pharmacogenomics, to explore the relationship between sex and genome organization and function, will be crucial for the implementation of sex-specific pharmacogenomics in clinical routine (Figure 4).
Roadmap for sex-specific pharmacogenomic approaches
A better understanding of the interaction of sex, genetic variation and drugs will provide clinically useful information on the evaluation of the susceptibility of an individual to drug toxicity or of the therapeutic response. This, in turn, will offer prognostic information on potential sex-specific effects for any drug. Consequently, such newly generated knowledge will contribute towards a more appropriate and individualized medical therapy, ultimately paving the way for precision medicine. Investigating the interaction of sex with genetic variation in drug responses at an unbiased pharmacogenomic level will be crucial to achieve this.
We thank Dr Franziska Schwarz (scivisto.com) for the scientific illustrations.
We acknowledge support from the DZHK (German Centre for Cardiovascular Research) and the German Heart Foundation (Deutsche Herzstiftung).