Pharmacogenetic profile of a South Portuguese population: results from the pilot study of the European Health Examination Survey in Portugal.

Original Paper Received: September 30, 2014 Accepted after revision: January 8, 2015 Published online: March 10, 2015

Public Health Genomics DOI: 10.1159/000373920

Pharmacogenetic Profile of a South Portuguese Population: Results from the Pilot Study of the European Health Examination Survey in Portugal Vânia Gaio a Isabel Picanço b Baltazar Nunes a Aida Fernandes c Francisco Mendonça d Filomena Horta Correia d Álvaro Beleza c Ana Paula Gil a Mafalda Bourbon b Astrid Vicente b Carlos Matias Dias a Marta Barreto da Silva a

Departamentos de a Epidemiologia and b Promoção da Saúde e Prevenção das Doenças Não Transmissíveis, Instituto Nacional de Saúde Doutor Ricardo Jorge, Lisbon, and c Laboratório de Saúde Pública Dra. Laura Ayres, and d Administração Regional de Saúde do Algarve, Faro, Portugal

Abstract Background: The genetic inter-individual variability of drug response can lead to therapeutic failure or adverse drug reactions (ADRs). The aims of this study were to assess the pharmacogenetic profile of a South Portuguese population according to established dosing guidelines for commonly prescribed drugs and to compare it with that of previously genotyped populations. Methods: A cross-sectional study was developed in the context of the Portuguese Component of the European Health Examination Survey (EHES). A total of 47 pharmacogenetically relevant variants in 23 different genes were genotyped in 208 participants. Allelic and genotypic frequencies were calculated, and the pharmacogenetic profile of the participants was defined. A comparative analysis was conducted through electronic database search. Pairwise Fst calculations were performed to assess the genetic distance between populations. Results: We found a significant small differentiation between the Portuguese regional populations regarding CYP2C9 rs1057910, CYP2D6

© 2015 S. Karger AG, Basel 1662–4246/15/0000–0000$39.50/0 E-Mail [email protected] www.karger.com/phg

rs3892097, MTHFR rs1801133 and F5 rs6025. When considering 4 HapMap populations, ADH1B rs2066702, ADH1B rs1229984, NAT2 rs1799931 and VKORC1 rs9923231 displayed a significant population differentiation. We found that 18.9% of the participants are intermediate or poor metabolizers for at least 3 drugs simultaneously and that 84.6% of the participants have at least one therapeutic failure or ADR risk allele for the considered drugs. Conclusions: There is a high prevalence of risk alleles associated with an altered drug metabolism regarding drugs largely used by the South Portuguese population. This knowledge contributes to the prediction of their clinical efficacy and/or toxicity, optimizing therapeutic response while improving cost-effectiveness. © 2015 S. Karger AG, Basel

Background

Drug response is a complex trait influenced by several genetic and environmental factors. Although it is widely accepted that a strong genetic component is involved in the disposition (absorption, distribution, metabolism and excretion) of a given drug, age, gender, drug interactions, underlying diseases, smoking and alcohol conMarta Barreto da Silva Departamento de Epidemiologia Instituto Nacional de Saúde Doutor Ricardo Jorge Avenida Padre Cruz, PT–1649-016 Lisbon (Portugal) E-Mail marta.barreto @ insa.min-saude.pt

Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/27/2015 4:37:16 PM

Key Words Pharmacogenetics · Genetic distance · Genetic variants · Population-based study

2


meaning of the information because genetic variations occur at very different frequencies among subpopulations [16], and the major groups are not representative of the population diversity. It is therefore important to perform genetic population-based studies in the different populations to assess the different variants’ frequency distribution and to identify groups at risk for therapeutic failure or ADRs. Understanding the relationship between genotype variants and drug response could have several public health benefits, including the selection of therapeutic options at the national level. If patients have genetic variations that prevent drugs from being properly metabolized, the prescription of these drugs is likely to be ineffective or harmful and health-care resources would be wasted. Consequently, it would be possible to decrease the overall health costs by reducing the prescription of drugs that are ineffective or harmful to particular patients and at the same time provide these patients with alternative medicines that would be more effective for their particular case [17]. Moreover, advances in DNA sequencing technologies have led to a successive cost reduction in the pharmacogenetic profile acquirement, which could facilitate the routinely implementation of these genetic tests in the near future [18]. In Portugal, pharmacogenetic studies are scarce, and the pharmacogenetic profile of the Portuguese population is still largely unknown. Consequently, the main goals of the present study were to characterize pharmacogenetically relevant variants in a South Portuguese population and to compare these with the frequencies of genetic variants obtained for populations in other Portuguese regions. The analyzed variants were those located in genes codifying drug-metabolizing enzymes, drug transporters and drug targets. We have also determined the pharmacogenetic susceptibility profile of the studied population specifically regarding the genotyped variants and compared the obtained frequencies with those from 4 HapMap populations: residents of Utah, USA, of northern and western Europe ancestry (CEU), Han Chinese people in Beijing, China (HCB), Japanese people in Tokyo, Japan (JPT), and Yoruba people in Ibadan, Nigeria (YRI). Methods Study Design and Participants We have used a cross-sectional study designed for the pilot study of the Portuguese Component of the European Health Examination Survey (EHES) [19, 20]. This pilot study was conducted

Gaio et al.


sumption also modulate the drug response [1]. Given that genetic variants in genes coding for drug-metabolizing enzymes, receptors, transporters and other drug targets are strongly associated with variability in the efficacy and toxicity of many drugs [2], patients’ genomes can account for 20–95% of the variation in the drug response [3]. This inter-individual variability can lead to therapeutic failure or adverse drug reactions (ADRs), which represent a major health-care problem, being responsible for significant morbidity, mortality and costs in health-care systems [4]. A recent meta-analysis [5] has indicated that ADRs may occur in 16.88% of hospitalized patients. In Portugal, 16,720 ADRs were identified among 9,271,122 hospitalizations between 2000 and 2009, representing 1.26% of the total hospitalized patients [6]. By improving the predictability of drug response in individual patients through the application of genetic information in guiding drug therapy decisions, it will be possible to avoid ADRs and to improve drug efficacy and safety [7]. Today, dosing guidelines for at least 27 different drugs provided by the Clinical Pharmacogenetics Implementation Consortium (CPIC) to support the clinical implementation of genetic information in drug prescription are already available, and many others are being developed [8]. There are 7 pharmacogenes with well-established dosing guidelines: TPMT, which allows to define the metabolic profile of thiopurine drugs used in immunologic disorders and cancer treatments [9]; CYP2C19, whose alleles can be predictors of patient response to both clopidogrel, used as a blood clot inhibitor in cardiovascular disease conditions, and tricyclics, frequently used in psychiatric disorders and pain management [10]; CYP2C9 and VKORC1, which influence the metabolic profile of warfarin, the most commonly used oral anticoagulant for the treatment and prevention of thromboembolic events [11]; DPYD, whose variants are related to the detoxifying metabolism of fluoropyrimidines, which are chemotherapeutic agents associated with fatal adverse reactions [12]; UGT1A1, which allows to define the metabolic profile of irinotecan, another chemotherapeutic agent [13], and CYP2D6, whose alleles can act as predictors of the metabolic profile of several drugs, including codeine, an analgesic opioid used as a pain relief agent, and tricyclic antidepressants [14]. Typically, drug effectiveness is tested taking into consideration the pharmacogenetic profiles of the major human populations (European, African and Asian ancestries), which are more extensively characterized [15]. However, the use of a self-defined ancestry of a patient as a clinical guide to prescribe a drug can compromise the

SNP Selection and Genotyping Genomic DNA was isolated from whole blood containing EDTA according to standard procedures [22]. We selected variants based on their involvement in drug metabolism using the PharmGKB database information [23]. The criteria for the selection of the variants included available VIP annotation (Very Important Pharmacogene summaries) and/or previously established dosing guidelines. A total of 47 variants in 23 different genes were selected: 35 variants were located in drug-metabolizing enzymes (ADH1B rs2066702, ADH1B rs1229984, ADHIC rs698, COMT rs4680, CYP2C19 rs4244285, CYP2C19 rs4986893, CYP2C8 rs11572080, CYP2C8 rs10509681, CYP2C9 rs1799853, CYP2C9 rs1057910, CYP2D6 rs3892097, CYP2D6 rs5030655, CYP2D6 rs5030656, CYP2D6 rs16947, CYP2D6 rs1065852, CYP2D6 rs35742686, CYP3A4 rs2740574, CYP3A5 rs776746, DPYD rs67376798, DPYD rs1801265, DPYD rs55886062, DPYD rs3918290, GSTP1 rs1695, GSTT1 gene deletion, MTHFR rs1801131, MTHFR rs1801133, NAT2 rs1799929, NAT2 rs1799930, NAT2 rs1799931, NAT2 rs1801279, TPMT rs1800462, TPMT rs1800460, TPMT rs1142345, TPMT rs1800584 and UGT1A1 rs8175847), 3 in drug transporters (ABCB1 rs1045642, ABCB1 rs1128503 and ABCB1 rs2032582), 7 in drug targets (VKORC1 rs9923231, ADRB1 rs1801252, ADRB2 rs1042714, ADRB2 rs1042713, TYMS rs34743033, TYMS rs2853542 and TYMS rs34489327) and 2 genetic variations indirectly affecting drug response (F5 rs6025 and KCNJ11 rs5219). Twenty-two variants were carried out by Sequenom. The remaining 25 variants were not Sequenom compatible because they had several SNPs in their vicinity or were included in iPlexing options with very few SNPs. Therefore, 24 variants were genotyped by restriction fragment length polymorphism (RFLP) and 1 by sequencing methods (suppl. table 1; see www.karger.com/doi/10.1159/000373920 for all suppl. material). An electronic database search was conducted on PubMed to select previous genetic studies in the Portuguese population that were used in the comparative analysis of the obtained allele frequencies. The criteria used to select other studies on pharmacogenetic variant frequencies include the following: population-based studies, sample size ≥100 participants and sampling region information. Searches were performed in April 2014. In addition to our

The European Health Examination Survey in Portugal

sample from São Brás de Alportel representing the South Portuguese region, 6 other Portuguese regions were considered: North, Centre, Lisbon and Tagus Valley (LVT), Alentejo, Azores and Madeira. Statistical Analysis Allele frequencies were obtained, and 95% confidence intervals (CI) were calculated for all allele frequencies observed. All variants were tested for Hardy-Weinberg equilibrium using the HardyWeinberg R package [24]. The level of significance was set at 5% for all statistical tests. Monozygotic SNPs were excluded from the analysis. Pairwise Fst calculations were performed using the Arlequin 3.1 software package [25]. The statistical significance of the Fst values was estimated by permutation analysis (10,000 permutations, p < 0.05). Pairwise Fst values ranged from 0 to 1, where 1 means that the two populations were completely separated and 0 means that no divergence between the populations existed. The interpretation of the Fst values was performed as follows: small population differentiation (0–0.05), moderate population differentiation (0.05–0.15), large population differentiation (0.15–0.25) and very large population differentiation (values >0.25) [26].

Results

Population Characteristics Of the 221 participants recruited, we genotyped those with available DNA samples (n = 208). Clinical and demographic characteristics of the participants influencing drug response are described in table 1. Of all participants, 42.2% (n = 88) were men and 57.8% (n = 120) were women. The participants’ age ranged from 26 to 91 years (mean ± standard deviation: 56.3 ± 16.0) (table 1). All participants were Caucasian. More than half of the participants had at least one chronic disease (63.1%), and the majority (54.4%) received medication for this disease. Moreover, 39.0% of the participants were smokers, and 7.8% were excessive alcohol consumers. Genotyping Data The genotype and allele frequencies of the analyzed variants with the respective value of the Hardy-Weinberg equilibrium test are described in supplementary table 2. All SNPs were in Hardy-Weinberg equilibrium except for CYP2D6 rs1065852 (p = 0.033) and VKORC rs9923231 (p = 0.026). Four monozygotic SNPs were detected (CYP2C19 rs4986893, TPMT rs1800584, DPYD rs55886062 and DPYD rs3918290), and the genotyping of 4 SNPs (CYP3A5 rs776746, NAT2 rs1799929, NAT2 rs1801279 and CYP2D6 rs35742686) failed. A total of 17 studies examining the pharmacogenetic variants studied were identified, but only 9 studies, whose Public Health Genomics DOI: 10.1159/000373920

3


between 2010 and 2011 in the population served by the São Brás de Alportel Health Center in the south of Portugal (Algarve). The study cohort consisted of 11,089 individuals (2.6% of the total South Portuguese population) aged 25 years or older. The study involved an observational and descriptive epidemiological examination with data collected through a detailed questionnaire (including health status, health determinants and medication use) and a physical examination complying with the recommendations proposed by the Feasibility of a European Health Examination Survey (FEHES) [7]. Blood samples were also collected. Participants were selected using a simple random sampling scheme from the National Health System number database, which covers over 99% of the total population from the São Brás de Alportel Health Center, as previously reported [21]. All participants received a brief description of the objectives of the study, after which they signed an informed consent form. The study protocol was approved by the Ethics Committee of the National Health Institute Doutor Ricardo Jorge and by the Comissão Nacional de Proteção de Dados (authorization 199/2011).

participants (n = 208) Sex Men Women Age, years Body mass index Chronic diseases1 Medications2 Smoking3 Excessive alcohol consumption4

42.2 (35.5 – 49.0) 57.8 (51.0 – 64.5) 56.3 ± 16.0 27.9 ± 4.7 63.1 (56.5 – 69.7) 54.4 (47.6 – 61.2) 39.0 (32.4 – 45.7) 7.8 (4.1 – 11.4)

Data are presented as mean ± standard deviation for continuous variables and percentages for proportions with the respective 95% confidence intervals in parentheses. 1 Chronic diseases are defined as illness or health problems which have lasted, or are expected to last, for 6 months or more. 2 Medications for chronic diseases were considered. 3 Daily and occasional smokers were considered. 4 Excessive alcohol consumption was defined as more than 3 drinks per day for men and as more than 2 drinks per day for women [27].

references are listed in table 2, met the inclusion criteria: 4 of the articles were from the LVT region, 2 from the Centre, 2 from the Azores and 1 was from North Portugal. The studies from the LVT region included variants in the ABCB1, CYP2C19, CYP2D6, MTHFR and NAT2 genes. The 2 studies from the Centre referred to variants in the ABCB1, CYP2C19, CYP2C9, CYP2D6, CYP3A4, GSTP1, NAT2, TYMS and UGT1A1 genes. The 2 studies from the Azores included variants in the F5, MTHFR and UGT1A genes. Finally, the study from the North included variants in the ABCB1 gene. The comparison between the obtained minor allele frequencies (MAFs) in the South Portuguese population and the MAFs of the previously described variants for the other Portuguese populations, when available, is shown in table 2. When analyzing linkage disequilibrium (LD) patterns among SNPs located in the same gene, strong evidence of LD was observed for the variants genotyped in the ABCB1 and ADRB2 genes and between rs1799929 and rs1799930 of the NAT2 gene. Uninformative evidence of LD was detected between the variants analyzed in the TPMT and DPYD genes and between rs1799931 of the NAT2 gene and other variants of the same gene. It was possible to perform at least one comparison between two different Portuguese regions for 17 different variants (fig. 1). A significant small differentiation between the Portuguese populations regarding 4


4 variants was detected: CYP2C9 rs1057910 (pairwise FstSouth versus Centre = 0.020, p < 0.05), CYP2D6 rs3892097 (pairwise FstSouth versus Centre = 0.043, p < 0.05), MTHFR rs1801133 (pairwise FstSouth versus Azores = 0.014, p < 0.05) and F5 rs6025 (pairwise FstSouth versus Azores = 0.014, p < 0.05). Moreover, in these 4 variants, we observed significantly lower MAFs in the South than in the other Portuguese regions (table 2). The remaining analyzed variants did not show significant differentiation among the different Portuguese populations. No information was available for the allele frequencies in the populations from the autonomous regions Madeira and Alentejo, and no previous pharmacogenetic studies on a representative sample of the Portuguese population were found. Comparison with HapMap Populations When comparing the allele frequencies of the South Portuguese population with those of the main 4 HapMap populations, the obtained allele frequencies are similar to the frequencies observed in the CEU population for the majority of the variants. Exceptions are ABCB1 rs1045642, ADH1B rs1229984, ADHIC rs698 and COMT rs4680, whose 95% CI do not overlap between the South Portuguese and CEU populations (table 3). In fact, in the pairwise Fst analysis used as a differentiation measure between populations, we found that the South Portuguese population displays a higher genetic similarity to the CEU population (blue diamonds in fig. 2) than to any other of the considered HapMap populations, with pairwise Fst values A, rs9923231

Normal (*1*1GG, *1*1GA, *1*2GG) Intermediate (*1*1AA, *1*2AA, *2*2GG or GA, *1*3GG or GA, *2*3 GG) Deficient

66.8 (73.2 – 60.4) 33.2 (39.6 – 26.8)

CYP2C9 VKORC1

0.00

Fluoropyrimidines Toxic effects

DPYD

*2A, rs3918290 *13, rs55886062 rs67376798

Normal (*1*1) Intermediate (*1/*rs67376798) Deficient

98.5 (100 – 96.9) 1.49 (0 – 3.1) 0.00

Irinotecan

Toxic effects

UGT1A1

*28, rs8175347

Normal (*1/*1) Intermediate (*1/*28) Deficient (*28/*28)

44.1 (50.9 – 37.3) 44.6 (51.4 – 37.8) 11.3 (15.6 – 6.9)

Codeine

Respiratory depression

CYP2D6

*6, rs5030655 *9, rs5030656 *4, rs389209 + rs1065852 *10, rs1065852

Normal (*1/*1, *1/*6, *1*9, *1*4, *1*10) Intermediate (*4/*10) Deficient

77.8 (83.4 – 72.1) 22.2 (27.9 – 16.6) 0.00

Tricyclics

Central nervous system effects Cardiac effects

CYP2D6

*6, rs5030655 *9, rs5030656 *4, rs389209 + rs1065852 *10, rs1065852

Normal (*1*1 + *1*1/*1*4/*1*6/*1*9/*1*10) Intermediate (*4*10 + *1*2/*1*1 + *1*2/*4*10 + *1*1) Deficient (*1*1 + *2*2)

59.4 (66.1 – 52.7)

CYP2C19

*2, rs4244285

39.1 (45.8 – 32.5) 1.45 (0 – 3.1)



9


1 Normal metabolizers are homozygotic for the wild-type allele; intermediate metabolizers carry 2 decreased-function alleles or are heterozygous for 1 deficient allele and 1 decreased-function allele; deficient metabolizers carry 2 non-functional alleles.

The Pharmacogenetic Profile of the South Portuguese Population We found that 84.6% of the South Portuguese population has at least 1 risk allele to potentially develop ADRs or therapeutic failure if treatment for cancer, cardiovascular and/or psychiatric disorders is started, which could be avoided if individual genetic information was taken into consideration. Unfortunately, the current Portuguese health-care system is not properly aligned to ensure an efficient use of these pharmacogenetic information advances, and costs related to ADRs continue to be a major economic and health-care problem. With the recognition that genetic information can improve the drug prescribing process, along with the availability of expert consensus guidelines on how to apply genotype results to drug therapy decisions, implementing pharmacogenetics would be a major step forward in health-care quality. The Potential of Applying Pharmacogenetic Profiles in the Portuguese Population: Irinotecan, Clopidogrel and Warfarin We observed that more than 50% of the participants have a deficient or intermediate pharmacogenetic profile regarding irinotecan, similar to the percentage already reported among Caucasians [37]. Irinotecan is an anticancer agent commonly used in colorectal cancer treatment, which is one of the most frequent cancer types in the Portuguese population [38]. Despite the controversy associated with the cost-effectiveness analysis performed for 10


Table 5. Description of the pharmacogenetic biomarkers with

potential to prioritize medication selection and possibly reduce health-care expenditure in the Portuguese population Gene

Variant ID Drug

Frequency in the South Portuguese population

CYP2C19

rs4244285 Clopidogrel

12.8 (9.6 – 16.0)

CYP2C9

rs1799853 Warfarin rs1057910

13.7 (10.4 – 17) 3.4 (1.6 – 5.1)

VKORC1

rs9923231 Warfarin

42.3 (37.6 – 47.1)

UGT1A1

rs8175347 Irinotecan

33.6 (29 – 38.2)

this drug [39], there is evidence that UGT1A1 genotyping potentially offers a less expensive alternative than current practice, and its application in cancer treatment could be especially important due to the small therapeutic window between efficacy and severe toxicity of this drug [40]. Another drug with a considerable percentage of intermediate/deficient metabolizers (24.3%) is clopidogrel. According to the National Authority of Medicines and Health Products (INFARMED), clopidogrel is the second-most active substance contributing to expenditure growth. In 2010, the proportion of the Portuguese population that received the defined daily dose of clopidogrel was 10.6 per 1,000 inhabitants, while in 2003 it was only 1 per 1,000 inhabitants. Consequently, the total cost associated with clopidogrel in 2010 was EUR 53,993,831 compared to EUR 7,410,401 in 2003 [41]. Given these values, the pharmacogenetic profile information associated with this drug could have particular impact as there is probably a higher percentage of patients taking the drug with reduced therapeutic effect. These patients are at risk of developing adverse cardiovascular events despite taking the drug [42, 43]. If the individuals’ pharmacogenetic information is considered, it will be possible to decrease the overall health costs associated with ineffective clopidogrel prescription and, at the same time, provide these patients with other medications that will be more effective in their particular case, thus improving health-care quality. Regarding warfarin therapy, a small cost-benefit study of warfarin therapy has already been performed by the Portuguese National Health Institute Doutor Ricardo Jorge [44]. The predicted cost involved in warfarin diagnosis in the Portuguese population was EUR 1.2 million/ year, whereas the benefits were expected to be EUR 1.6 million/year, avoiding 391 hospitalizations caused by serious bleeding episodes per year. In addition, a recent meGaio et al.


Comparison between the South Portuguese Population and the 4 HapMap Populations As expected, we observed a high genetic proximity of the South Portuguese population to the CEU population. Nevertheless, among the other HapMap populations, there are 4 variants with very large population differentiation and 6 variants with large population differentiation. Differentiation between HapMap populations is particularly large for genetic variants located in the alcohol dehydrogenase (ADH) genes, for example ADH1B rs1229984 which shows the highest Fst values, reflecting a large genetic distance between the Eastern-Asian population and any other population in the world. This is likely due to a random genetic drift, as previously described [36]. These findings reinforce the need to analyze genetic variants prior to prescribing drugs whose metabolization and/or toxicity is influenced by these variants in order to maximize the treatment, avoid ADRs and have implications at the decision-making and planning levels in different countries.

ta-analysis has concluded that there is evidence of improved clinical effectiveness associated with genotypeguided warfarin dosing [45]. Based on the issues described for irinotecan, clopidogrel and warfarin and the findings regarding the allele frequencies of their associated variants, we consider that CYP2C19 rs4244285, CYP2C9 rs1799853, CYP2C9 rs1057910, VKORC1 rs9923231 and UGT1A1 rs8175347 have the potential to be used as pharmacogenetic biomarkers to prioritize medication selection and possibly reduce health-care expenditure in the Portuguese population (table 5). Study Limitations Because of the small study size, we only genotyped a few genetic variants in each gene, and, consequently, the genetic susceptibility of the South Portuguese population to ADRs was probably underestimated. For instance, the CYP2C9, CYP2D6, CYP2C19, DPYD and TPMT genes have already a wide list of identified loss-of-function alleles and reduced-activity alleles that are included in the CPIC dosing guidelines but not in the present study. A better approach to obtain the pharmacogenetic profile of this population would be the total sequencing of these genes, which will be economically possible in a short time given the advances in the next-generation sequencing technologies. With this approach, higher frequencies of intermediate and deficient metabolizers will probably be identified, and, consequently, a higher impact of the pharmacogenetic information application could be achieved. Several cost-effectiveness pharmacogenetic analyses have already been performed, and even though these strategies will only be cost-effective for certain combinations of diseases, drugs and genetic tests, they have great potential to improve the effectiveness and safety of pharmaceutical care, particularly in cancer treatment [46].

Conclusion

Pharmacogenetics has the potential to increase drugs’ efficacy and safety and, consequently, health, decreasing the cost of drug treatment and improving health-care quality. The cost-effectiveness analysis of pharmacogenetics has already been performed, and, even though some strategies will only be cost-effective for certain combinations of diseases, drugs and genetic tests, it has great potential to improve the effectiveness and safety of pharmaceutical care. In particular, CYP2C19 rs4244285, CYP2C9 rs1799853, CYP2C9 rs1057910 VKORC1 rs9923231 and UGT1A1 rs8175347 have the potential to be used as pharmacogenetic biomarkers to prioritize medication selection and possibly reduce health-care expenditure in the Portuguese population.

Acknowledgment The authors are grateful to the Association for Research and Development of the Faculty of Medicine (AIDFM), Lisbon. The authors are also grateful to J. Costa for assistance with genotyping (Instituto Gulbenkian de Ciência/Genomics Unit). The pilot study of the Portuguese Component of the European Health Examination Survey (EHES) project has received funding from the European Commission/DG Sanco (agreement No.: 20092301-EHES JA-EAHC). This study has also received funding from the Portuguese Foundation for Science and Technology (FCT) (project reference: PTDC/SAU-ESA/101743/2008).

Disclosure Statement The authors declare that they have no competing interests.

References


5 Miguel A, Azevedo LF, Araújo M, Pereira AC: Frequency of adverse drug reactions in hospitalized patients: a systematic review and meta-analysis. Pharmacoepidemiol Drug Saf 2012;21:1139–1154. 6 Miguel A, Marques B, Freitas A, Lopes F, Azevedo L, Pereira AC: Detection of adverse drugs reactions using hospital databases – a nationwide study in Portugal. Pharmacoepidemiol Drug Saf 2013;22:907–913. 7 FEHES – Feasibility of a European Health Examination Survey. http://www.thl.fi/fehes/index.html.


8 The Pharmacogenomics Knowledge Base. http://www.pharmgkb.org. 9 Relling MV, Gardner EE, Sandborn WJ, Schmiegelow K, Pui CH, Yee SW, Stein CM, Carrillo M, Evans WE, Klein TE; Clinical Pharmacogenetics Implementation Consortium: Clinical Pharmacogenetics Implementation Consortium guidelines for thiopurine methyltransferase genotype and thiopurine dosing. Clin Pharmacol Ther 2011; 89: 387– 391.

11


1 Weinshilboum RM, Wang L: Pharmacogenetics and pharmacogenomics: development, science, and translation. Annu Rev Genomics Hum Genet 2006;7:223–245. 2 Meyer UA: Pharmacogenetics and adverse drugs reactions. Lancet 2000;356:1667–1671. 3 Crews KR, Hicks JK, Pui C-H, Relling MV, Evans WE: Pharmacogenomics and individualized medicine: translating science into practice. Clin Pharmacol Ther 2012;92:467–475. 4 Davies EC, Green CF, Mottram DR, Pirmohamed M: Adverse drugs reactions in hospitals: a narrative review. Curr Drug Saf 2007;2:79–87.

12


21 Gaio V, Nunes B, Fernandes A, Mendonça F, Correia FH, Beleza Á, Gil AP, Bourbon M, Vicente A, Dias CM, Barreto da Silva M: Genetic variation at the CYP2C19 gene associated with metabolic syndrome susceptibility in a south Portuguese population: results from the pilot study of the European Health Examination Survey in Portugal. Diabetol Metab Syndr 2014;6:23. 22 Lahiri DK, Nurnberger JI: A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res 1991;19:5444. 23 Ford ES: Risks for all-cause mortality, cardiovascular disease, and diabetes associated with the metabolic syndrome: a summary of the evidence. Diabetes Care 2005;28:1769–1778. 24 de Ferranti S, Gauvreau K, Ludwig D, Neufeld E, Newburger J, Rifai N: Prevalence of the metabolic syndrome in American adolescents: findings from the Third National Health and Nutrition Examination Survey. Circulation 2004;110:2494–2497. 25 Scheinder S, Roessli D, Excoffier L: Arlequin ver.2.000: a software for population genetics data analysis. Geneva, Genetics and Biometry Laboratory, University of Geneva, 2000. 26 Wright S: Genetical structure of populations. Nature 1950;166:247–249. 27 Matias Dias C, Briz T: Distribuição conjunta de determinantes de saúde relacionados com comportamentos na população portuguesa. Observações_Boletim Epidemiológico, vol 1, No 1, July–September. Lisboa, Instituto Nacional de Saúde Doutor Ricardo Jorge, 2012. http://hdl.handle.net/10400.18/878. 28 Correia C, Santos P, Coutinho A, Vicente A: Characterization of pharmacogenetically relevant CYP2D6 and ABCB1 gene polymorphisms in a Portuguese population sample. Cell Biochem Funct 2009;27:251–255. 29 Santos M, Carvalho S, Lima L, Nogueira A, Assis J, Mota-Pereira J, Pimentel P, Maia D, Correia D, Gomes S: Common genetic polymorphisms in the ABCB1 gene are associated with risk of major depressive disorder in male Portuguese individuals. Genet Test Mol Biomarkers 2014;18:12–19. 30 Oliveira E, Marsh S, van Booven DJ, Amorim A, Prata MJ, McLeod HL: Pharmacogenetically relevant polymorphisms in Portugal. Pharmacogenomics 2007;8:703–712. 31 Ruas JL, Lechner MC: Allele frequency of CYP2C19 in a Portuguese population. Pharmacogenetics 1997;7:333–335. 32 Castro R, Rivera I, Ravasco P, Jakobs C, Blom H, Camilo M, De Almeida I: 5,10-Methylenetetrahydrofolate reductase 677C→T and 1298A→C mutations are genetic determinants of elevated homocysteine. QJM 2003; 96:297–303. 33 Gil JP, Lechner MC: Increased frequency of wild-type arylamine-N-acetyltransferase allele NAT2*4 homozygotes in Portuguese patients with colorectal cancer. Carcinogenesis 1998;19:37–41.

34 Pacheco PR, Brilhante MJ, Ballart C, Sigalat F, Polena H, Cabral R, Branco CC, Mota-Vieira L: UGT1A1, UGT1A6 and UGT1A7 genetic analysis: repercussion for irinotecan pharmacogenetics in the São Miguel Island Population (Azores, Portugal). Mol Diagn Ther 2009;13:261–268. 35 Cross DS, Ivacic LC, Stefanski EL, McCarty CA: Population based allele frequencies of disease associated polymorphisms in the Personalized Medicine Research Project. BMC Genet 2010;11:51. 36 Osier MV, Pakstis AJ, Soodyall H, Comas D, Goldman D, Odunsi A, Okonofua F, Parnas J, Schulz LO, Bertranpetit J: A global perspective on genetic variation at the ADH genes reveals unusual patterns of linkage disequilibrium and diversity. Am J Hum Biol 2002; 71: 84–99. 37 Shimoyama S: Pharmacogenetics of irinotecan: an ethnicity-based prediction of irinotecan adverse events. World J Gastrointest Surg 2010;2:14. 38 Direção Geral da Saúde: Direção de Serviços de Epidemiologia e Estatísticas de Saúde. Lisboa, Divisão de Epidemiologia, 2006. http://www. dgsaude.min-saude.pt/pns/vol1311html. 39 Maitland ML, Vasisht K, Ratain MJ: TPMT, UGT1A1 and DPYD: genotyping to ensure safer cancer therapy? Trends Pharmacol Sci 2006;27:432–437. 40 Houtsma D, Guchelaar H, Gelderblom H: Pharmacogenetics in oncology: a promising field. Curr Pharm Des 2010;16:155–163. 41 Furtado C, Oliveira R: Análise da evolução do mercado total de medicamentos entre 2003 e 2010. Infarmed, 2012. http://www.infarmed.pt/ portal/page/portal/infarmed/monitorizacao_ do_mercado/observatorio/introducao_de_ficheiros/monit-mt-2003–2010_20110513_.pdf. 42 Rossi JS, Cammarata M, Dharmavaram J, Weck K, Walko C, Gabriel D, Kuritzky J, Muldrew K, Stouffer GA: Clopidogrel dose adjustment after outpatient screening for cyp2c19 variant alleles: a pilot study. Pharmacogenomics 2014;15:915–923. 43 Yasmina A, de Boer A, Klungel OH, Deneer VH: Pharmacogenomics of oral antiplatelet drugs. Pharmacogenomics 2014;15:509–528. 44 Raimundo A, Picanço I, Barreto da Silva M, Vicente AM: Análise de custo-benefício da farmacogenética na terapêutica com varfarina. Observações_Boletim Epidemiológico, vol 2, No 3, January–March. Lisboa, Instituto Nacional de Saúde Doutor Ricardo Jorge, 2013. 45 Goulding R, Dawes D, Price M, Wilkie S, Dawes M: Genotype-guided drug prescribing: a systematic review and meta-analysis of randomized control trials. Br J Clin Pharmacol 2014, Epub ahead of print. 46 Veenstra DL, Higashi MK, Phillips KA: Assessing the cost-effectiveness of pharmacogenomics. AAPS PharmSci 2000;2:80–90.

Gaio et al.


10 Scott SA, Sangkuhl K, Gardner E, Stein C, Hulot J, Johnson J, Roden D, Klein T, Shuldiner A; Clinical Pharmacogenetics Implementation Consortium: Clinical Pharmacogenetics Implementation Consortium guidelines for cytochrome P450-2C19 (CYP2C19) genotype and clopidogrel therapy. Clin Pharmacol Ther 2011;90:328–332. 11 Johnson J, Gong L, Whirl-Carrillo M, Gage B, Scott S, Stein C, Anderson J, Kimmel S, Lee M, Pirmohamed M, Wadelius M, Klein TE; Altman RB; Clinical Pharmacogenetics Implementation Consortium: Clinical Pharmacogenetics Implementation Consortium guidelines for CYP2C9 and VKORC1 genotypes and warfarin dosing. Clin Pharmacol Ther 2011;90:625–629. 12 Caudle K, Thorn C, Klein T, Swen J, McLeod H, Diasio R, Schwab M: Clinical Pharmacogenetics Implementation Consortium guidelines for dihydropyrimidine dehydrogenase genotype and fluoropyrimidine dosing. Clin Pharmacol Ther 2013;94:640–645. 13 Kweekel D, Gelderblom H, Van der Straaten T, Antonini N, Punt C, Guchelaar H: UGT1A1* 28 genotype and irinotecan dosage in patients with metastatic colorectal cancer: a Dutch Colorectal Cancer Group study. Br J Cancer 2008;99:275–282. 14 Hicks J, Swen J, Thorn C, Sangkuhl K, Kharasch E, Ellingrod V, Skaar T, Müller D, Gaedigk A, Stingl J; Clinical Pharmacogenetics Implementation Consortium: Clinical Pharmacogenetics Implementation Consortium guideline for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants. Clin Pharmacol Ther 2013;93:402–408. 15 International HapMap Project. http://HapMap.ncbi.nlm.nih.gov/index.html.en. 16 Ioannidis JP, Ntzani EE, Trikalinos TA: ‘Racial’ differences in genetic effects for complex diseases. Nat Genet 2004;36:1312–1318. 17 Robertson JA, Brody B, Buchanan A, Kahn J, McPherson E: Pharmacogenetic challenges for the health care system. Health Aff (Millwood) 2002;21:155–167. 18 Personalized Medicine Coalition: The case for personalized medicine. 2014. http://www. personalizedmedicinecoalition.org/userfiles/ pmc-corporate/file/pmc_case_for_personalized_medicine.pdf. 19 Lemos M, Cabrita F, Silva H, Vivan M, Placido F, Regateiro F: Genetic polymorphism of CYP2D6, GSTM1 and NAT2 and susceptibility to haematological neoplasias. Carcinogenesis 1999;20:1225–1229. 20 Branco CC, Pereirinha T, Cabral R, Pacheco PR, Mota-Vieira L: Thrombotic genetic risk factors and warfarin pharmacogenetic variants in São Miguel’s healthy population (Azores). Thromb J 2009;7:9.

[Physical activity in adult working population: results from the European National Health Survey for Spain (2009)].

Prevalence and correlates of hypertension in a transitional southeastern European population: results from the Albanian Demographic and Health Survey.

Oral Health and Risk of Arthritis in the Scottish Population: Results from the Scottish Health Survey.

Pilot study on the European Portuguese version of the Confusion Assessment Method.

Are chronic diseases related to height? Results from the Portuguese National Health Interview Survey.

Differences in participation rates and lessons learned about recruitment of participants--the European Health Examination Survey Pilot Project.

The extent and determinants of diabetes and cardiovascular disease comorbidity in South Africa - results from the South African National Health and Nutrition Examination Survey (SANHANES-1).

Prevalence and cardiovascular risk profile of chronic kidney disease in Italy: results of the 2008-12 National Health Examination Survey.

Prevalence of age-related macular degeneration in a large European cohort: results from the population-based Gutenberg Health Study.

Comorbidities of rheumatoid arthritis: Results from the Korean National Health and Nutrition Examination Survey.

Correction: Comorbidities of rheumatoid arthritis: Results from the Korean National Health and Nutrition Examination Survey.

The European Health and Behaviour Survey: rationale, methods and initial results from the United Kingdom.

Is Hypozincemia Related to Tinnitus?: A Population Study Using Data From the Korea National Health and Nutrition Examination Survey.

Health profile of centenarians in Portugal: a census-based approach.

The Mental Health in Austrian Teenagers (MHAT)-Study: preliminary results from a pilot study.

The family role in children׳s sleep disturbances: Results from a cross-sectional study in a Portuguese Urban pediatric population.

Generational differences in fast food intake among South-Asian Americans: results from a population-based survey.

The current condition of the workers' general health examination in South Korea: a retrospective study.

Cross-cultural adaptation and validation of the Portuguese version of the "European Health and Behaviour Survey-section B".

Sex and age differences in the achievement of control targets in patients with type 2 diabetes: results from a population-based study in a South European region.

Demographic characteristics and health behaviours of consenters to medical examination. Results from the Welsh Heart Health Survey.

Medical education in pharmacogenomics-results from a survey on pharmacogenetic knowledge in healthcare professionals within the European pharmacogenomics clinical implementation project Ubiquitous Pharmacogenomics (U-PGx).

A European Profile of Games for Health.

Health examination utilization in the visually disabled population in Taiwan: a nationwide population-based study.