Inter‐ethnic differences in pharmacokinetics—is there more that unites than divides?

Abstract Inter‐ethnic variability in pharmacokinetics (PK) has been attributed to several factors ranging from genetic to environmental. It is not clear how current teaching in higher education (HE) reflects what published literature suggests on this subject. This study aims to gain insights into current knowledge about inter‐ethnic differences in PK based on reports from published literature and current teaching practices in HE. A systematic literature search was conducted on PubMed and Scopus to identify suitable literature to be reviewed. Insights into inter‐ethnic differences in PK teaching among educators in HE and industry were determined using a questionnaire. Thirty‐one percent of the studies reviewed reported inter‐ethnic differences in PK, of these, 37% of authors suggested genetic polymorphism as possible explanation for the inter‐ethnic differences observed. Other factors authors proposed included diet and weight differences between ethnicities. Most respondents (80%) who taught inter‐ethnic difference in PK attributed inter‐ethnic differences to genetic polymorphism. While genetic polymorphism is one source of variability in PK, the teaching of genetic polymorphism is better associated with interindividual variabilities rather than inter‐ethnic differences in PK as there are no genes with PK implications specific to any one ethnic group. Nongenetic factors such as diet, weight, and environmental factors, should be highlighted as potential sources of interindividual variation in the PK of drugs.

been linked to race, ethnicity, and ancestry, and prescribing recommendations have been directed at certain races or ethnicities. 1 The terms race and ethnicity are often used interchangeably, and their meanings are not always clear despite attempts at standardization. 2,3 Ethnicity is often used to describe cultural and social factors 2 and can have varying meanings across the English-speaking world 3 with some emphasizing the cultural aspects while others incorporate aspects of ancestry and race. The categories used vary between countries and the UN Statistics Division has noted that no internationally accepted criteria for ethnicity are possible. 4 In human genetics, anthropology, and population genetics it is now accepted that ideas of race or ethnicity have no biological or genetic basis. 3 Environmental factors, such as diet which has cultural and ethnic differences, have been shown to affect the pharmacokinetics (PK) of drugs. 5 For example, many traditional Asian and African diets contain spices, vegetables, and herbs 6 while a typical European diet is more likely to contain high levels of fat, 7 though a recent study argued that the levels of fat in traditional Asian diet may be comparable to fat levels in Western diet. 8 Some individuals may maintain their traditional diet irrespective of their geographical location. A study showed that, in subjects who consumed low-fat vegetarian diet, there was a resultant slower absorption rate of paracetamol compared to their nonvegetarian counterparts. The authors explained that the difference may be due to reduced gastric emptying resulting from consumption of vegetarian diet. 9 Green tea and its extracts, popular in East Asia but consumed throughout the world, are common and regularly used beverages and supplements; however, the polyphenols in the tea may affect absorption and metabolism by altering the levels of drug metabolizing enzymes and transporters involved in the drug disposition. 10 Herbal medication, which is often used traditionally in some parts of the world, may have a broad range of effects on the PK of conventional medicines. In Western societies, St John's Wort is well known for its ability to interact with prescription medicines and there are numerous examples from other cultures. 11 In South America, Cat Claw, a medicinal plant used by the Amazonians for its immunostimulant and antiviral properties, may inhibit the CYP3A4 isozyme, 12 decreasing drug metabolism which therefore increases the plasma exposure of antiretroviral agents such as atazanavir, ritonavir, and saquinavir. 11 Furthermore, garlic may selectively inhibit certain isozymes of cytochrome P450 (CYP) enzymes and induce expression of intestinal P-glycoproteins thereby potentially influencing the absorption and elimination of drugs. 11 Interindividual variations in drug distribution can also be attributed to variations in body weight or composition. Hayes et al., (2015) demonstrated that the weight-for-age in adults from different malaria endemic countries differ significantly ranging from an average of 50 kg in the South Asians to 68 kg in the Americas; 13 this contrasts with the generally accepted average adult 72 kg weight based on healthy Caucasians. 14 This weight variation may result in differences in the distribution kinetics of lipophilic drugs across these populations since higher body weight may be indicative of a higher fat mass and higher volume of distribution of such drugs. 15 Plasma protein binding may influence the extent of drug distribution. The two main plasma proteins responsible for the binding of drug molecules in humans are α 1 acidic glycoprotein (AAG) and the human serum albumin. Differences in α 1 acidic glycoprotein (AAG) expression levels have been demonstrated between Whites and Chinese. 16 Zhou et al., (1990) showed that there was a higher unbound fraction of disopyramide, propranolol, and diphenhydramine in Chinese patients (n = 8) compared to White patients (n = 8) and they attributed this to lower levels of AAG in the Chinese. 16 Levels of AAG may be affected by a number of environmental factors 17 and there are polymorphisms that show differences in AAG allele frequencies between ethnic groups. 18,19 Genetic polymorphisms of metabolizing enzymes and transporter proteins may affect several pharmacokinetic processes. 1 One of the earliest reported and best-known examples of this is the ethnic differences in response to alcohol due to striking differences in the presence of certain alleles for alcohol dehydrogenase and aldehyde dehydrogenase in European, Asian, Sub-Saharan African, and African American groups. 20 Polymorphisms have been widely reported in the CYP450 enzyme superfamily, a group of enzymes which are responsible for the metabolism of a wide range of drugs.
For example, the functionally deficient allele, CYP2B6*6, prevalent in African Americans (46.7%) compared with Asians (17.4%) has been reported to be responsible for inter-ethnic difference in the clearance of efavirenz. A review highlighted the clinical relevance of this observation in Asians and reported that due to the higher efavirenz plasma exposure observed in them following standard doses, optimization of dosing regimen may be warranted. 21 There is substantial literature describing environmental and physiological differences that exist between ethnic groups some of which have pharmacokinetic significance and may result in interethnic variabilities in PK. Therefore, this study aims to explore literature evidence to determine the extent to which there are PK differences between ethnic groups, determine the mechanisms suggested for differences in PK in published literature, investigate how current teaching in higher education (HE) reflects the findings from literature, and recommend ways to assist educators in HE to approach the subject of inter-ethnic differences in PK so that learners have a better understanding of the potential reasons why inter-ethnic variation in PK may occur and potential impact these differences may have on treatment outcomes.

| Search strategy
To survey published literature that reported significant inter-ethnic differences in the pharmacokinetic of drugs, a scoping review of literature was conducted. This involved a systematic approach whereby the preferred reporting items for systematic reviews and meta-analyses (PRISMA) strategy was adapted to suit the purpose of this review and the process was conducted by two members of the project team such that each stage of the process completed by one member was checked separately by the other member of the team.
When there was disagreement between the two members, the criteria used at the respective stage were reviewed by both members in order to maintain consistency. In this current study, while reviewing the literature on ethnic diversity in PK, terminologies similar to those used in published literature sources such as "Caucasians" for example, were used during literature search.

| Initial search
To identify potentially relevant publications, two databases were used namely PubMed and SCOPUS, and the search term used for both databases is reported in Table S1 in the Supplementary Material. The selection of search terms used in identifying articles included terms such as "ethnic," "blacks," "African American," "Asian," "Caucasian," and "European." These terms were chosen because they are terms commonly attributed to race or ethnicity.

| Screening of initial search
Duplicates were removed from the articles identified in the initial search of the two databases. The titles and abstracts from the identified articles were screened to remove those where inter-ethnic differences were not mentioned in the context of differences in diet/ lifestyle or absorption, distribution, metabolism, and excretion. Also, articles that did not involve humans were excluded.

| Eligibility
After screening, the remaining articles were assessed based on whether they demonstrated differences in PK between ethnic groups. The articles which showed inter-ethnic differences in PK were separated from those that did not show inter-ethnic differences. Also, reviews and articles deemed not relevant to the research or already had the primary papers in the search were removed. The articles were deemed irrelevant if the authors did not directly measure pharmacokinetic parameters of drugs referenced in their study.

| Included articles
The articles included in the final analyses excluded articles which did not have a pharmacokinetic implication for the differences observed.

| Survey of teaching practices
A questionnaire that targeted academics who have either taught PK or are currently teaching PK in HE or the industry was designed and shared with contact of the project team through university email contact lists, social media platforms, and the British Pharmacological Society community forum. Responses were collected anonymously; no identifying information was required, and the questionnaire took less than 5 min to complete. The details of the questions asked in the survey are reported in Figure S1 of Supplementary Material 1 and involved tallying the number of articles where authors reported interethnic differences in PK parameters that may be due to inter-ethnic differences in one or more of the ADME processes. The percentage ADME tallies were determined to compare the number of articles where each ADME process was involved in inter-ethnic variabilities to the total number of articles included in the analysis. In the questionnaire analysis, the number of respondents who thought there were inter-ethnic differences in either one or more of the ADME processes was tallied. The percentage of those who thought any of the ADME process was involved in inter-ethnic difference was determined in comparison with the total number of respondents. The percentages obtained from the article analyses and questionnaires were weighed equally and the resultant percentages were plotted on a bar chart.
The drugs and disease class of the drugs that were studied in the final included articles were tabulated; the table included the drug class, the key PK parameter differences between ethnic groups and the ADME processes that may have been affected, the authors' explanation for inter-ethnic difference observed, and the authors' recommendation or other relevant remarks.

| Published literatures finding difference or similarities in PK between ethnicities
A comparison between studies that showed differences in PK and those that did not show differences in PK was conducted and presented as a percentage of the total number of PK publications in each decade of publication starting from papers published in 1980 till March 2021.

| Authors explanation/suggestions for interethnic differences in PK between ethnicities
The papers which reported inter-ethnic differences in PK were screened to determine authors' explanations or suggested reasons for inter-ethnic differences in PK. The authors' reasons were sorted in similar categories and represented as percentages of the total number of papers.

| Literature search, screening, eligibility, and inclusion
A total of 69 studies were included in the final analyses. This was down from 725 studies identified from the initial literature search ( Figure 1).

| Summary of literature evidence showing inter-ethnic differences seen across a wide range of drugs treatments.
From the literature, the most reported drug classes where ethnic differences were observed were those that treat cardiovascular diseases, central nervous system-related conditions, and human immunodeficiency virus (HIV) ( Table 1). Moreover, the most reported pharmacokinetic process affected for these drugs was metabolism resulting from polymorphisms of CYP enzymes and transporters.
Only two papers reported excretion to be the pharmacokinetic process affected and both the drugs studied, phenytoin and morphine are used for the treatment of central nervous system disorders.
Propranolol, nifedipine, paracetamol, rosuvastatin, midazolam, efavirenz, and cyclosporine were the most studied across multiple different papers in various ethnic groups.
Additionally, from all the statins included in this review, systemic absorption was mainly affected between the studied ethnicities. Moreover, polymorphisms of genes encoding the transporters OATP1B1, ABCG2, and SLCO1B1 were linked to the ethnic differences.
Overall, those classified as Caucasians (sometimes referred to as White) and Asians were the two most commonly studied.

| Papers showing ethnic differences in pharmacokinetics compared to those showing no significant differences.
Overall, of the research articles, the majority (69%) did not find significant ethnic differences in pharmacokinetic processes ( Figure 2 The authors explanation or mechanism for inter-ethnic difference is evidence based on data provided in the respective paper. b The authors explanation or mechanisms for inter-ethnic differences was postulated and no evidence was found in the respective paper.
c Observed results and pharmacokinetic parameters to be evidence based.
d Observed results and pharmacokinetic parameters to be speculative and no evidence was found to support this in the respective papers.  Table S3.  Table S2.

| Summary of responses from the respondents
Of the 15 respondents who taught inter-ethnic differences in PK, only one discussed examples that were not necessarily related to drug metabolism. This respondent stated, "Without detailing the processes, racial differences affect transporters (absorption, metabolism, and excretion), enzymes (metabolism) and body composition (distribution and possibly metabolism and excretion)." Notably, of all the respondents who taught inter-ethnic differences in PK, 80% attributed genetic polymorphisms to inter-ethnic differences in PK.

| Summary of PK processes associated with inter-ethnic difference in PK from published literature and questionnaires
In both the literature and current HE teaching, the process of metabolism appears to be the most reported mechanism linked to ethnic differences making up 60% of taught material and 50% of reported literature ( Figure 3). Our literature search identified higher frequencies of papers suggesting inter-ethnic differences in absorption and F I G U R E 2 Comparison between decades of published data which showed significant pharmacokinetic differences between ethnic groups (green) and which showed no significant pharmacokinetic differences (gray). Studies identified in 2021 was between January and March. Two hundred and thirty-seven studies were compared which included 163 studies that showed no significant difference in pharmacokinetics and 74 which showed significant difference in pharmacokinetics F I G U R E 3 Comparison of current teaching and the literature publications for each pharmacokinetic process. Literature percentage tally was based on the number of papers that linked PK differences to the ADME processes in comparison to the total number or papers. Respondent percentage tally was based on the numbers of respondents who thought there were ethnic differences in one or more of the ADME processes compared to the total number or respondents. distribution compared to the respondents' answers. Absorption and distribution made up 8%-12% of teaching respondents compared to about 23% in the reported studies analyzed (Figure 3). On the other hand, our literature search had lower frequencies of papers reporting inter-ethnic differences linked to excretion (2%) compared to the respondents' (19%).

| Suggested or investigated mechanisms linked to inter-ethnic PK differences in published literature
Genetic polymorphism was the predominant mechanism (37%) proposed or suggested by authors as an explanation for the inter- Inter-ethnic difference in enzyme activity was also proposed as a probable mechanism.

| DISCUSS ION
Several factors can contribute to interindividual variation in PK.
Some of the factors might be more common in certain ethnic groups than others. It is important to understand what physiological, sociocultural factors, and other underlying mechanism may lead to interindividual differences in PK. It is equally important that a proper reflection of how these factors could impact on interindividual differences in PK is included as a part of PK-related teaching in HE. This study aimed to review evidence from literature which suggest ethnic differences in PK, identify the mechanism proposed by the authors, investigate how current teaching addresses these factors, and make recommendations for teaching inter-ethnic differences in PK in HE.

| Terminology and methodology
One major observation from this review is the use of the terms "ethnicity" and "race" which are often used interchangeably and without definition although "ethnicity" is often used to indicate more social and cultural characteristics. The current United Nations recommendation is that individuals are asked to self-report their ethnicity rather than it being assumed by the investigator 4 however the majority of the papers included in this review do not specify how participants were characterized in their respective groups. The way that ethnic groups are defined varies around the world: one paper referred to an ethnic group as "Hispanic" however this is open to interpretation as individuals identifying as such may have ancestry of European, Native American, or African descent. 22 While we have searched for literature citing ethnic differences, the implied meaning of the term in these contexts is generally closer to that of race since many papers frequently used the terms "Caucasians," "Whites," or  In published literature and the current teaching practice, most ethnic sensitivity to PK is thought to involve drug metabolism (Figure 3;

| Diet and its effect on drug absorption, distribution, and metabolism
Diet can change drug disposition by altering the luminal pH, direct binding to drug, gastric emptying, intestinal transit, mucosal absorption, hepatic blood flow, and induction of enzymes. 5 It has also been shown that for certain lipophilic antimalarials like lumefantrine, there is 16-fold increase in bioavailability if the drug is administered with a high-fat meal. 29 This is thought to be because the consumption of fatty meals increases in vivo solubility of lumefantrine as a result of increased bile micelles production. 30 In this present study, many papers reported diet to be a crucial factor responsible for some inter-ethnic differences in PK. For example, both modafinil 31 and ziprasidone 32 were investigated in the five prominent ethnic groups in China and diet was implicated for the pharmacokinetic differences observed (  (Figure 3). Of all the studies identifying inter-ethnic differences in PK, 9% suggested diet as possible reason for the inter-ethnic difference observed (Figure 4).

| Body weight as a physiological factor and its effect on distribution
The distribution of a drug has many determinants including tissue binding, organ blood flow, and drug plasma protein binding. 35 Furthermore, in obese patients, tissue blood flow is decreased, and cardiac function and structure are modified. 35 In Asian populations, the average body weight is 60 kg 36 which is significantly lower than many other populations. This is reflected with the drugs efavirenz 36 and selumetinib 37 where body weight was postulated as the underlying reason for the observed distribution differences seen between the studied populations (Table 1). Furthermore, lower body weight has also been considered a risk factor for increased systemic exposure of certain drugs such as tenofovir disoproxil fumarate (TDF) which may therefore provide justification for adjusting dosing regimens. 36 Notably, no respondent mentioned inter-ethnic differences in weight as an example of what they would teach their students on inter-ethnic differences in PK.

| Metabolizing enzyme and transporter
abundance and polymorphism and effects on drug absorption, distribution, and metabolism In addition to isoforms of metabolizing enzymes, there were also references to genetic polymorphism or protein abundances relating to plasma proteins particularly α 1 acidic glycoprotein (AAG). For example, the level of AAG-bound propranolol was significantly lower in the Chinese compared to Caucasians (Table 1). 41 Genetic polymorphisms of membrane-bound efflux transporters such as P-gp may influence absorption and is one of the proposed mechanisms for anticancer drug resistance. 42 Another example of polymorphism occurs in organic anion-transporting polypeptide (OATP) transporter superfamily. Located on the basolateral membrane of hepatocytes, these proteins are involved in the uptake and clearance of many drugs therefore influencing absorption and excretion. 43 It has been reported that there may be differences in the PK of transporter substrates between Caucasians and Asians. 44 Interestingly, with the same genotype, OATP1B1, the activity in an Asian population is still half that of Caucasians suggesting that besides allele frequencies of transporter genes, there is an underlying ethnic variability, distinct from genetics, that contributes to the pharmacokinetic variations. 44 In the survey of teaching practices, respondents mainly provided examples related to genetic polymorphism of CYPs and OATP1B1 as what they would teach students on the subject of inter-ethnic variabilities in drug disposition (see Supplementary Material Table S2).

Genetic polymorphism was the predominant mechanism suggested
in the literature identified in the review as linked to inter-ethnic differences in drug disposition ( Figure 4).
The importance of polymorphisms of metabolizing enzymes and transporters as a mechanism associated with inter-ethnic differences in PK has received much greater attention in the literature (Table 2 and Figure 4) and in HE in PK (Table 2; Figure 3) although it remains to be seen whether this increased attention reflects the importance of this mechanism or whether it is the result of researchers and teachers selecting this for their attention. It is important to note that the attribution of inter-ethnic differences in PK to genetic polymorphism that some authors in this review made were postulated, and their actual studies neither confirmed nor refuted their postulation. This may, therefore, increase the possibility of unconscious bias when interpreting these studies and may lead to attributing nearly all unexplained variabilities in PK to genetic polymorphisms.
It is important to note that, while there is genetic diversity between individuals, different races or ethnicities cannot be categorized on the basis of their genetics as greater genetic diversity has been reported within an ethnic groups compared to between ethnic groups, for example, there is greater genetic diversity within Africans than between Africans and other non-African populations. 3,45

| What does this all mean for teaching PK in the future?
We recommend that educators explain the variety of factors that can lead to interindividual variation in PK rather than focussing on inter-ethnic differences. These factors include diet, weight, age, concurrent medicines, co-morbidities, isoforms of metabolizing enzymes and transporters, differing levels of expression of plasma proteins, enzymes, and transporters, and use a balanced array of examples that include both environmental and genetic factors.
The evidence upon which current PK teaching discusses variabilities in the PK of drugs between ethnicities as being predominantly linked to genetic polymorphism is at best weak. No genetic trait is unique to one race or ethnicity 46 and traits are influenced by both environmental and genetic factors. Educators should ensure that students recognize that there is greater genetic diversity within ethnic groups than between them and should reinforce the consensus view among geneticists that there is no biological or genetic basis to the idea of race. 3 Therefore, single-nucleotide polymorphisms should be discussed primarily in the context of interindividual variation in PK, rather than inter-ethnic differences in PK.
While evidence for varying prevalence of certain SNPs among ethnic groups could be discussed, this should be in the context of a critical analysis of the prevalence of the resultant PK variabilities within the ethnic group and the potential pitfalls in using race or ethnicity as a proxy for a variable(s) that could be measured directly.
This analytical perspective provides scope for discussions around the rationale behind ethnic-dependent use or dosing of certain medicines in current teaching 40  The interpretation of educators' perspective about inter-ethnic differences in PK in this present study was based on a small size and limited geographical representation of respondents, therefore, an extrapolation of our findings on general educators' teaching practices would be difficult to reach. Another limitation of this review was that no qualitative research methodologies were used to analyze the narrative text response from the respondents.

| CON CLUS ION
There is a considerable literature comparing PK between ethnic groups with more papers reporting similarities than differences Logically though, if we know, for example, that weight can have an impact on PK, then it would be better to assess a patient's weight when prescribing rather than using ethnicity as a proxy. The same is true for the other factors that can lead to interindividual variation.
In our analysis, we identified some methodological issues with some of the studies that report differences in PK between ethnic groups. One aspect of this was in the definitions of the terms used to describe ethnicity and a lack of clarity on how the term "ethnicity" was determined. Another aspect is in the experimental design, sampling, and statistical analysis in some of the papers and it will be important to pursue this further by focussing on certain classes of drugs to weigh up the strength of the evidence for and against ethnic differences. This information will provide a rich source of material in teaching of critical thinking.

ACK N OWLED G M ENTS
The authors thank the British Pharmacological Society (BPS) for providing the teaching grant to conduct the study and for the support provided by their staff during this project.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R S CO NTR I B UTI O N S
OO, NK, and JF contributed to the conception and design of the research and the acquisition, analysis and interpretation of data generated. OO, NK, JW, and JK were all involved in drafting the manuscript and revising it for important intellectual content.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.