Research Article For reprint orders, please contact: [email protected]

Pharmacogenomics

Genetic epidemiology of pharmacogenetic variations in CYP2C9, CYP4F2 and VKORC1 genes associated with warfarin dosage in the Indian population Aim: Warfarin, a widely used anticoagulant, exhibits large interindividual variability in dose requirements. CYP2C9 and VKORC1 polymorphisms in various ethnic groups have been extensively studied as genetic markers associated with variable drug response. However, allele frequencies of these variants have not been assessed in major ethnic groups in the Indian population. Materials & methods: To study the functional variants known to affect warfarin dosing, we reanalyzed genotype microarray datasets generated as a part of genome-wide association studies as well as data from the Indian Genome Variation database. We examined data from 2680 individuals across 24 ethnically diverse Indian subpopulations. Results: Allelic distribution of VKORC1 (-1639G>A) showed a greater degree of variation across Indian subpopulations, with frequencies as low as 6.5% in an out-group subpopulation to >70% in Tibeto–Burmans. Risk allele frequency of CYP4F2*3 (V433M) was higher in north Indians (0.30–0.44), as compared with other world populations, such as African–American (0.12), Caucasian (0.34) and Hispanic (0.23). TheVKORC1 variant (-1639A) was shown to be prevalent amongst Tibeto–Burmans, whereas CYP2C9 (R144C, I359L) and CYP4F2 (V433M) variants were observed in considerable variability amongst Indo–Europeans. The frequency of CYP2C9*3 (I359L) in north Indians was found to be higher than in most Asian populations. Furthermore, geographical distribution patterns of these variants in north India showed an increased trend of warfarin extensive metabolizers from the Himalayan to Gangetic region. Combined allele frequency (CYP2C9*3 and CYP4F2*3) data suggest that poor metabolizers varied in the range of 0.38–1.85% in Indo–Europeans. Conclusion: Based on genotypic distribution, the majority of the Indian subpopulation might require higher doses for stable anticoagulation, whereas careful assessment is required for Tibeto–Burmans who are expected to have intermediate dose requirement. This is the largest global genetic epidemiological study examining variants associated with warfarin that could potentially be valuable to clinicians in optimizing dosage strategies.

Anil K Giri‡,1,2, Nazir M Khan‡,1, Sandeep Grover‡,1, Ismeet Kaur1, Analabha Basu3, Nikhil Tandon4, Vinod Scaria2,5, IGV Consortium, INDICO § , Ritushree Kukreti1, Samir K Brahmachari1 & Dwaipayan Bharadwaj*,1,2 CSIR-Institute of Genomics & Integrative Biology, Delhi, 110 020, India 2 Academy of Scientific & Innovative Research (AcSIR), Anusandhan Bhavan, 2 Rafi Marg Delhi, 110 001, India 3 National Institute of BioMedical Genomics, Kalyani, 741 251, India 4 Department of Endocrinology, All India Institute of Medical Sciences, New Delhi, 110 029, India 5 GN Ramachandran Knowledge Center for Genome Informatics, CSIR-Institute of Genomics & Integrative Biology, Mathura Road, Delhi, 110 020, India *Author for correspondence: Tel.: +91 11 29879246 Fax: +91 11 2766 7471 db@ igib.res.in ‡ Authors contributed equally § A complete list of the members of the INDICO consortia can be found in the online Supplementary Material 1

Original submitted 4 April 2014; Revision submitted 23 May 2014 Keywords:  allele frequency • CYP2C9 • CYP4F2 • pharmacogenomics • VKORC1 • warfarin

Background Warfarin is one of the most commonly used oral anticoagulants and is widely prescribed for the treatment of a number of medical conditions, including deep vein thrombosis, pulmonary embolism and heart valve prosthesis to avoid thromboembolic attack [1] . Warfarin has a narrow therapeutic index and has been

10.2217/PGS.14.88 © 2014 Future Medicine Ltd

well known to be associated with marked interindividual variability in drug response. An overdose of warfarin increases the risk of unexpected bleeding, while insufficient dose results in failure to prevent thrombosis [2] . In a cross sectional study conducted on south Indians, Nekkanti et al. have demonstrated that 26% of treated patients developed warfarin-

Pharmacogenomics (2014) 15(10), 1337–1354

part of

ISSN 1462-2416

1337

Research Article  Giri, Khan, Grover et al. induced bleeding [3] . This necessitates frequent monitoring of the effect of warfarin in clinical settings as different individuals respond differently to the drug. In addition to clinical characteristics of patients, common genetic variants in VKORC1 and functional variants of CYP2C9*2 (c.C430T; p.Arg144Cys), CYP2C9*3 (c.A1075C; p.Ile359Leu) and CYP4F2*3 (c.G1297A, p.Val433Met) are known to affect the individual response to warfarin [2,4,5] . Since genetic architecture of individuals varies greatly in terms of genotype composition, different warfarin doses are required for different individuals. Individual variation in drug dosage requirement and action can be ascertained by analyzing relevant variants [6,7] . Genotyping of patients is not feasible in daily practice under resource limiting conditions, which necessitates the creation of a genotype frequency map at a population scale for relevant variants affecting drug metabolism. Creation of such maps at the subpopulation scale is necessary in large countries with ethnically stratified populations, such as in India. Previous studies have substantiated this diversity by suggesting that allele frequency between groups in India are larger than in Europe [8] . A recent study by Gan et al. showed that among the Asian populations, Indians require higher warfarin doses compared with the Chinese and Malay patients, and these variations could mainly be due to different allele frequencies and polymorphisms in CYP2C9 and VKORC1 [9] . These results were further supported by findings of others who demonstrated that the VKORC1 haplotype in Indians was significantly different from other Asian populations [10,11] . It is thus imperative to ascertain differences in genotype frequencies in smaller populations within countries such as India to aid in the understanding and prediction of population level differences in drug metabolism and action. A number of resources have information on genetic variations in the Indian population. The Indian Genome Variation (IGV) database catalogues the common patterns of genetic variations in important complex disease candidate genes and also provides allele frequencies of major and minor variants among large Indian subpopulations [12] . Furthermore, availability of genome-wide datasets generated from genome-wide association studies (GWAS) in major populations provide a new opportunity towards understanding genome epidemiology and mining relevant genotypes and allele frequencies for clinically relevant variants, which could be of immense value to practicing clinicians. Here, we use a proof of concept of such approaches to mine and analyze genetic variants associated with metabolism of warfarin in the Indian population. To the best of our knowledge, this is the first large-scale genetic epidemi-

1338

Pharmacogenomics (2014) 15(10)

ological analysis of genotypes associated with warfarin pharmacogenetics in any global population. Our data suggest that allelic and genotypic characteristics of the CYP2C9, VKORC1 and CYP4F2 variants in Indians are quite different from those in other ethnic groups across the globe. Materials & methods Study population

In the present study, a total of 2680 individuals were recruited from two previously published consortium studies carried out in Indian population: Indian Diabetes Consortium (INDICO) and IGV consortium [12,13] . The target population in the IGV study consisted of 552 individuals belonging to 24 diverse endogamous Indian subpopulations of Indo–European (IE), Dravidian, Tibeto–Burman (TB) and Austro–Asiatic ethnicities spread across several regions and states of India. The INDICO study included 2128 individuals, comprising of 1219 men and 909 women of IE ethnicity that were also participants of a previously published GWAS for Type 2 diabetes (T2D) in Indians [13,14] . T2D cases were collected at the outpatient department of the endocrinology clinic of All India Institute of Medical Sciences, New Delhi, India [14] and control samples were collected from Diabetes Awareness Camps organized in various regions of Delhi and nearby states [13] . At the time of collection, samples were de-identified and coded with unique IDs ready for use in genetic studies. The analyses were performed separately for cases and controls, as the original study was performed for T2D and other associated traits [14] . Since we did not see significant differences in distribution of allele frequencies in cases and controls, we merged the datasets for further analysis. The study was approved by the human ethics committees of all the participating institutions. Written informed consent was obtained from all the participants for genetic studies including the possibilities of genotyping for various diseases. The study was carried out in accordance with the principles of the Helsinki Declaration. Genes & variants selection

The variants rs1799853 (CYP2C9*2), rs1057910 (CYP2C9*3), rs2108622 (CYP4F2*3) and rs7294, rs9923231 and rs9934438 of the VKORC1 gene were selected for analysis. Out of these six variants, rs1799853, rs1057910 and rs9923231 belonged to the level of 1A, wherea the remaining three were classified as level 1B in PharmGKB. Furthermore, polymorphisms found in rs1799853 (C>T), rs1057910 (A>C) and rs210862 (C>T) were reported as missense variants. For IGV samples, genotype information of rs1799853 (CYP2C9*2), rs1057910 (CYP2C9*3) and

future science group

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

rs9923231 (VKORC1) were mined, whereas genotype information for the above variants for INDICO samples were extracted from Illumina (Illumina Inc., CA, USA) Human 610 Quad Bead Chip used in the GWAS of T2D. Genotyping

Peripheral blood samples were used for isolation of genomic DNA. Genotyping of selected SNPs was performed using Sequenom (Sequenom Inc., CA, USA) for IGV samples and Illumina platform for INDICO samples according to methods as described previously [12,14] . Stringent quality controls were applied to the genotyping data as described earlier [12,14] . All the polymorphisms passed the quality check. The INDICO study samples had an average call rate of 98.2% and a concordance rate of >99%. Global datasets & comparison

Allele frequencies of the variants were evaluated from other population scale databases including HapMap, 1000 Genome datasets and the Allele Frequency Database (ALFRED). We also compared the allele frequencies of the variants presently analyzed with previously published reports from India [15,16] . A bubble plot was generated to visualize the global comparison of allele frequency using ggplot2 package in R. Statistical analysis

Each polymorphism was tested for Hardy–Weinberg equilibrium (HWE) in the study population. No significant deviation from HWE was observed for all the studied variants (p > 0.05). Observed genotype frequencies were compared with those expected under HWE using the χ 2 test. Pairwise one sided proportion test (greater than) was performed to compare the difference between allele frequencies of various racial groups globally. One-sided binomial test of proportion was used to compare the allele frequency between different states of north India. All the statistical analysis were performed using R platform. Results Genotype & allele frequencies of warfarinmetabolizing genes in Indians

We analyzed the genotype and allele frequency distributions of CYP2C9*2, CYP2C9*3 and VKORC1 (-1639G>A) polymorphisms to assess the metabolizer frequency in 24 Indian subpopulations and the results are summarized in Table 1. Table 1 shows the frequency of individuals with altered metabolizing CYP2C9 genotypes across 24 Indian subpopulations and it was found that 4.5–8.7% of individuals representing several IEs from the north and west

future science group

Research Article

of India displayed intermediate metabolizer status for CYP2C9. The amino acid substitutions in the CYP2C9 protein at position 144 (Arg to Cys) and 359 (Ile to Leu) due to CYP2C9*2 and CYP2C9*3 variants, respectively, have been found to limit the metabolism of warfarin into active forms (poor metabolizer) because of reduced enzymatic activity [9] . Hence, homozygous variants confer a poor metabolizer status whereas the homozygous wildtype allele would confer an extensive warfarin metabolizer status [9] . Of all the possible metabolizers based on distribution of CYP2C9*2 and CYP2C9*3 alleles, only intermediate metabolizers with CYP2C9*1/*3 genotype were detected among IEs in the Indian population (Table 1). The -1639G>A polymorphism is a promotor polymorphism found in the upstream promoter region of VKORC1 and the variant genotype (-1639AA) has been shown to be associated with reduced warfarin requirements [17] . As compared with CYP2C9 allelic variants, the VKORC1 variant (-1639G>A or rs9923231) showed highly variable distribution with an allele frequency as low as 6.5% in an out-group (OG) subpopulation to more than 70% in TBs (Table 1 & Supplementary Figure 1; see online at: www.futuremedicine.com/doi/suppl/10.2217/ pgs.14.88). Additionally, IEs showed a great degree of variation for the VKORC1 variant that follows a gradient effect from the western (6.8–23.9%) to northern (19.6–37%) to northeastern (23.8–54.6%) parts of India (Table 1 & Supplementary Figure 1) . Previous studies on genomic variations in Indians demonstrated that TBs residing in north and northeastern parts of India showed close genetic affinity with Asian populations (Chinese and Japanese), whereas IEs from north-India showed affinities with the Utah residents with northern and western ancestry from the CEPH collection (CEU) population of the HapMap project [18] . Therefore, we examined the allele frequencies of TBs vis-a-vis IEs and HapMap populations. Results showed that TBs exhibited a similar allele frequency with Japanese and Chinese populations (Han Chinese in Beijing [CHB] and Chinese in Metropolitan Denver, Colorado [CHD]) for VKORC1 and CYP2C9 variants; however, IEs showed a different frequency from the CEU population for said variants (Supplementary Figure 2) . Furthermore, the OG subpopulation, an isolated tribal population residing in the western part of India has been shown to be a population with more recent African ancestry [18] , hence we compared the allele frequency of the OG sub­population with that of the African population. It was found that the OG subpopulation showed similar allele frequency patterns with Africans (for VKORC1) and with Yoruba in Ibadan, Nigeria (YRI), a population of African descent (for CYP2C9; Supplementary Table 1). Taken together, these results

www.futuremedicine.com

1339

1340

Pharmacogenomics (2014) 15(10)

18 18 (100.0)

22 22 (100.0)

18 18 (100.0)

16 16 (100.0)

18 18 (100.0)

22 22 (100.0)

20 20 (100.0)

17 17 (100.0)

23 21 (91.3)

18 17 (94.4)

20 19 (95.0)

22 22 (100.0)

22 22 (100.0)

21 21 (100.0)

21 21 (100.0)

22 22 (100.0)

22 22 (100.0)

21 21 (100.0)

20 20 (100.0)

20 20 (100.0)

19 19 (100.0)

18 17 (94.4)

22 21 (95.5)

21 21 (100.0)

AA-C-IP5

DR-C-IP2

IE-E-IP1

IE-E-LP2

IE-E-LP4

AA-E-IP3

IE-N-LP5

IE-N-LP1

IE-N-LP9

IE-N-IP2

IE-N-SP4

TB-N-SP1

TB-N-IP1

IE-NE-IP1

IE-NE-LP1

TB-NE-LP1

DR-S-IP4

DR-S-LP3

DR-S-LP2

IE-W-LP3

IE-W-LP4

IE-W-LP1

IE-W-LP2

OG-W-IP

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

*1/*2 

0 (0.00)

1 (4.5)

1 (5.6)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

1 (5.0)

1 (5.6)

2 (8.7)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

*1/*3 

*2/*3 

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

0 (0.00) 0 (0.00)

*2/*2 

Genotype counts (frequency)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

0 (0.00)

*3/*3 

CYP2C9 allelic variants (*2, *3)

  100.0

  95.5

  94.4

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  95.0

  94.4

  91.3

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

  100.0

EM (wt/ wt)

0.00

4.5

5.6

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

5.0

5.6

8.7

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

IM (wt/ mut)

†

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

PM (mut/ mut)

Metabolizer frequency

Population nomenclature as per Indian Genome Variation Consortium convention. EM: Extensive metabolizer; IM: Intermediate metabolizer; mut: Mutant genotype; PM: Poor metabolizers; wt: Wild-type genotype.

 

 

*1/*1 

n 

 

Population† 

  23

  22

  23

  22

  21

  23

  23

  23

  23

  21

  22

  22

  22

  23

  23

  23

  21

  23

  22

  23

  21

  23

  22

  19

 

n GA

0 (0.0)

1 (4.8)

0 (0.0)

3 (13.6)

2 (10.5)

AA

0 (0.0)

5 (21.7)

20 (87.0) 3 (13.0)

19 (86.4) 3 (13.6)

13 (56.5) 9 (39.1)

14 (63.6) 6 (27.3)

17 (81.0) 4 (19.1)

19 (82.6) 4 (17.4)

19 (82.6) 4 (17.4)

  0.76

  0.46

0 (0.0)

0 (0.0)

1 (4.4)

2 (9.1)

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

  0.94

  0.93

  0.76

  0.77

  0.91

  0.91

  0.91

  0.91

14 (60.9)   0.28

2 (9.5)

10 (45.5) 7 (31.8)

19 (82.6) 4 (17.4)

4 (17.4)

14 (63.6)   0.25

  0.80

  0.63

  0.76

  0.71

  0.80

  0.77

  0.85

  0.91

  0.91

  0.73

  0.82

G

0.06

0.07

0.24

0.23

0.1

0.09

0.09

0.09

0.72

0.24

0.54

0.70

0.75

0.20

0.37

0.24

0.29

0.20

0.28

0.15

0.09

0.09

0.27

0.18

A

Allele frequency

11 (50.0) 10 (45.5)   0.30

5 (22.7)

13 (61.9) 6 (28.6)

5 (22.7)

1 (4.6)

3 (13.6)

1 (4.4)

13 (56.5) 2 (8.7) 15 (65.2) 7 (30.4)

8 (34.8)

12 (52.2) 11 (47.8) 0 (0.0)

10 (47.6) 10 (47.6) 1 (4.8)

14 (60.9) 9 (39.1)

12 (54.6) 10 (45.5) 0 (0.0)

16 (69.6) 7 (30.4)

18 (85.7) 2 (9.5)

19 (82.6) 4 (17.4)

13 (59.1) 6 (27.3)

14 (73.7) 3 (15.8)

GG

Genotype counts (frequency)

VKORC1 rs9923231

Table 1. Distribution of CYP2C9 allelic variants and VKORC1 rs9923231 across 24 Indian Genome Variation subpopulations.

Research Article  Giri, Khan, Grover et al.

future science group

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

suggested that IEs showed a distinct genetic pattern for the variants involved in metabolism of warfarin. The dissimilar allele frequency for the variants associated with warfarin metabolism among IEs might be due to admixture with the indigenous populations over different time periods [18] . Further­more, owing to availability of limited data for these variants among IEs in the IGV study, we further examined the allelic and genotypic frequency distribution of variants in warfarin-­metabolizing genes in detail by utilizing a distinct study (INDICO) that recruits mainly IEs residing in north India [13] . Genotype & allele frequencies of warfarinmetabolizing genes in IEs

To explore the allelic distribution pattern among IEs in detail, we divided the total individuals from the INDICO study (n = 2128) into five major groups according to their state of origin in the northern part of India (Punjab, Haryana, Delhi, Uttar Pradesh [UP] and Bihar). The allele frequencies are summarized in Table 2 & Figure 1 and genotype frequencies are given in Supplementary Table 2. The allele frequencies for CYP2C9*2 and CYP2C9*3 were found to be 0.22 and 0.11, respectively, in the north

Research Article

Indian population, although it varies between various states selected for the present study (Table 2) . Allele frequency of Bihar inhabitants was significantly less than that of Punjab for rs1057910 (p = 0.01), whereas it was higher for rs1799853 (p = 0.0001). Based on their genotype frequency as shown in Supplementary Table 2, individuals living in Punjab, Haryana, Delhi, UP and Bihar were classified into the extensive (72, 54.6, 61, 60.4 and 56.3% for CYP2C9*2, respectively, and 76.9, 79.6, 77.4, 78.6 and 82.4% for CYP2C9*3, respectively) intermediate (25.8, 38.9, 33.3, 34.4 and 35.5% for CYP2C9*2, respectively, and 21.6, 19.8, 20.8, 20.2 and 16.7% for CYP2C9*3, respectively) and poor metabolizer (2.2, 6.5, 5.6, 5.2 and 8.2% for CYP2C9*2, respectively, and 1.5, 1.2, 1.4, 1.3 and 0.8 for CYP2C9*3, respectively) groups. CYP4F2*3 (V433M) was recently identified as an important genetic determinant of warfarin dosing and has been shown to be associated with reduced capacity to metabolize vitamin K1 because of amino acid substitutions in the CYP4F2 protein at position 433 (Val to Met) and hence individuals carrying this allele are predisposed to high levels of hepatic vitamin K1, necessitating a higher warfarin dose to achieve stable anticoagulation [19] . The allele frequency was most

Table 2. Allele frequencies of CYP2C9, CYP4F2 and VKORC1. Serial number   

SNP

Variants

 

 

Punjab

Haryana

Delhi

Uttar Pradesh

Bihar

Total† 

1

CYP2C9*2 (rs1799853)

 

n = 225

n = 108

n = 213

n = 1098

n = 268

n = 2128

C

0.78

0.74

0.78

0.78

0.74

0.78

T

0.22

0.26

0.22

0.22

0.26

0.22

 

n = 223

n = 108

n = 212

n = 1098

n = 266

n = 2128

A

0.88

0.89

0.88

0.89

0.91

0.89

C

0.12

0.11

0.12

0.11

0.09

0.11

 

n = 226

n = 108

n = 213

n = 1101

n = 267

n = 2127

C

0.7

0.56

0.6

0.59

0.57

0.6

T

0.3

0.44

0.4

0.41

0.43

0.4

 

n = 225

n = 108

n = 212

n = 1102

n = 267

n = 2127

T

0.7

0.69

0.73

0.73

0.77

0.73

C

0.3

0.31

0.27

0.27

0.23

0.27

 

n = 225

n = 108

n = 213

n = 1101

n = 267

n = 2128

C

0.79

0.82

0.84

0.84

0.87

0.84

T

0.21

0.18

0.16

0.16

0.13

0.16

 

n = 225

n = 108

n = 213

n = 1102

n = 267

n = 2125

G

0.79

0.82

0.84

0.84

0.87

0.84

A

0.21

0.18

0.16

0.16

0.13

0.16

2

3

4

5

6

CYP2C9*3 (rs1057910)

CYP4F2*3 (rs2108622)

VKORC1 (rs7294)

VKORC1 (rs9923231)

VKORC1 (rs9934438)

Frequency 

Total number includes additional individuals from other states of north India, in addition to above mentioned states.

†

future science group

www.futuremedicine.com

1341

Research Article  Giri, Khan, Grover et al. prevalent in Haryana (0.44) and Bihar (0.43), while less frequent in Punjab (0.30; Table 2). The variant homozygosity for CYP4F2*3 genotype was 7.96% in Punjab, whereas it was 17.6% in the inhabitants of Bihar (Supplementary Table 2) . Among the selected VKORC1 variants, individuals residing in Bihar had significantly lower VKORC1 allele frequencies 0.23 (rs7294), 0.13 (rs9923231), 0.13 (rs9934338) compared with individuals dwelling in Punjab 0.30 (rs7294), 0.21 (rs9923231), 0.21 (rs9934338; p = 1.88 × 10 -4, 3.52 × 10 -7, 3.52 × 10 -7, respectively). The overall VKORC1 allele frequency in the north Indian population was 0.27 (rs7294), 0.16 (rs9923231) and 0.16 (rs9934338; Table 2). Combinatorial effect of CYP2C9, CYP4F2 & VKORC1 gene variants in the population

The frequency of variants taken together has been shown to be a better strategy to predict the outcome of

CYP2C9*2 (rs1799853)

CYP2C9*3 (rs1057910)

CYP4F2*3 (rs2108622)

VKORC1 (rs7294)

VKORC1 (rs9923231)

VKORC1 (rs9934438)

Punjab

Haryana

T 15%

T 26%

warfarin exposure and cumulative frequencies of two or more genotypes of different genetic variants are termed as combined genotype frequencies [19] . We evaluated the combined genotype frequencies of CYP2C9*3 and CYP4F2*3 for IEs. As shown in Table 3, our data clearly show that the frequency of individuals harboring both variants was 58.7% (Punjab), 76.8% (Haryana), 69.3% (Delhi), 72.3% (UP) and 73.7% (Bihar). Warfarin-sensitive genotype frequency describes the cumulative frequency of different genotypes that were shown to decrease warfarin metabolism, for example, homozygous variant genotype of CYP2C9*3 (CC) and homozygous wild-type genotype of CYP4F2*3 (CC). Among these subjects, warfarin-sensitive (intermediate dose requirement) genotype frequencies of CYP2C9 (CC) as well as CYP4F2 (CC) were seen in Punjab (0.45%), Haryana (1.85%), UP (0.73%) and Bihar (0.38%), whereas none of the individuals residing in Delhi possessed both genotypes (Table 3) . Delhi

Uttar Pradesh

Bihar

T 22%

T 22%

T 26%

C 11%

C 12%

C 11%

C 9%

T 30%

T 44%

T 40%

C 30%

C 31%

C 85%

C 12%

A 88%

C 70%

T 70%

A 21%

A 21%

G 79%

G 79%

C 74%

A 89%

C 86%

T 69%

C 78%

A 88%

C 60%

C 27%

T 73%

A 18%

A 16%

A 18%

A 16%

G 82%

G 82%

G 84%

G 84%

C 78%

A 89%

T 41%

C 27%

A 16%

C 59%

T 73%

G 84%

A 16%

G 84%

C 74%

A 91%

T 43%

C 57%

C 23%

T 77%

A 13%

G 87%

A 13%

G 87%

Figure 1. Allele frequencies of CYP2C9, CYP4F2 and VKORC1. Allele frequencies of CYP2C9*2, CYP2C9*3, CYP4F2*3, VKORC1 (rs7294, rs9923231 and rs9934438) are shown as pie charts for different states of India. Wild-type and variant allele frequencies are represented by blue and purple, respectively. Please see color figure at www.futuremedicine.com/doi/pdf/10.2217/pgs.14.88

1342

Pharmacogenomics (2014) 15(10)

future science group

future science group

Extensive metabolizer: CYP2C9 wild-type homozygote. Intermediate metabolizer: CYP2C9 heterozygote. Poor metabolizer: CYP2C9 variant homozygote. Freq: Frequency.

0.84

0.16   3

  16 1.5

0.38 1

4  

  0.09

0.64 7

1  

  0.47

0.47   1

  1 0

0 0

0

Research Article

  0

0

1.79 4

TT

 

0.63   12 0.38 1   0.73 8   0   0 1.85 2   0.45 1

Poor CC metabolizer (CC) TC

9.12

3.25   62

  174 7.52

1.88 5

20  

  3.55

8.56 94

39  

  12.26

5.19   11

  26 9.26

2.78 3

10  

  1.79

10.76

4

24

TT

6.82   130 4.51 12   7.83 86   4.25   9 5.56 6 7.62 17

Intermediate CC metabolizer (AC) TC

 

12.6

36.8   702

  241 15.41

38.35 102

41  

  37.16

13.75 151

408  

  8.96

37.74   80

  19 15.74

40.74 44

17  

 

13 TT

5.83

68 TC

30.49

29.2   556 26.32 70   27.69 304   30.66   65 23.15 25   92 CC

n

Extensive metabolizer (AA)

41.26

  n Freq. (%)     Freq. (%)   n Freq. (%)

 

n

Freq. (%)

  Delhi (n = 212)   Haryana (n = 108)   Punjab (n = 223) CYP4F2*3

We compared the allele frequencies of north Indians with previously published data from south Indians. CYP2C9*3 allele frequency in north Indians was higher than in south Indians residing in the Tamil Nadu and Kerala provinces (Table 5) . Furthermore, allele frequency

CYP2C9*3

Comparison of allele frequencies with previous studies in Indian & other populations

Table 3. Combined CYP2C9 and CYP4F2*3 genotype frequency.

The graphical representation of the distribution of the CYP4F2*3 genetic variants in the selected IEs is shown in Figure 2, which depicts that the highest frequency of the variant was found in Bihar and Haryana whereas the lowest frequency was reported in Punjab (Figure 2) . The frequency distribution of CYP2C9*3 (rs1057910) and VKORC1 (rs9923231) followed the reverse trend as that of CYP4F2*3 (rs2108622) with the highest frequency observed in Punjab and the lowest in Bihar (Figure 3) . The variant homozygosity for CYP2C9*3 and VKORC1 (rs9923231) resulted in poor metabolism of warfarin whereas CYP4F2*3 homozygous variants behave as extensive metabolizers [19] . The geographical distribution pattern of these variants showed an increased trend of extensive metabolizers and concomitant decreased trend of poor metabolizers from the Himalayan region (Punjab) to the Gangetic region (Bihar; Figure 3). The opposite trend of metabolizer frequencies in these regions demonstrated the example of genetic homeostasis maintained through adaptation to varied environments. These results suggest that in the absence of other nongenetic factors, the standard dose of warfarin might result in excessive bleeding in patients from the Himalayan region, whereas it may be insufficient to prevent thrombosis in patients from the Gangetic region. These results further suggest that individuals residing in the Gangetic planes (Bihar) would require a higher dose of warfarin to attain a therapeutic anticoagulant response.

n

Geographical distribution of genetic variants

Freq. (%)

Uttar Pradesh (n = 1098)

 

n

Bihar (n = 266)

 

Total (n = 1907)

Since the US FDA has recommended warfarin doses (low, intermediate or high) based on the possible combinations of CYP2C9*3 and VKORC1 genotypes, we analyzed the distribution of these combinations across 24 Indian subpopulations to identify variability in dose requirement across India. Based on genotypic distribution of these loci in the Indian population, we observed that the majority of Indian subpopulation may have a high-dose requirement for attaining therapeutic international normalized ratio (Table 4, Figure 2 & Supplementary Table 3) . However, a much more careful assessment of dosing regimens is required in TBs in whom a high proportion of individuals with intermediate dose requirement (warfarin sensitive) was observed (42.9–66.7%).

Freq. (%)

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

www.futuremedicine.com

1343

Research Article  Giri, Khan, Grover et al.

Table 4. Distribution of expected proportion of subpopulations with different levels of warfarin therapeutic dosage requirement based on distribution of VKORC1 and CYP2C9 genotypes across 24 Indian subpopulations. Population_ID† 

Dose requirement

n

%

AA-C-IP5

High

15

93.80

Intermediate

1

6.3‡

Low

0

0.00

 

16

100.00

High

22

100.00

Intermediate

0

0.00

Low

0

0.00

 

22

100.00

High

18

85.70

Intermediate

3

14.30‡

Low

0

0.00

 

21

100.00

High

22

100.00

Intermediate

0

0.00

Low

0

0.00

 

22

100.00

High

20

100.00

Intermediate

0

0.00

Low

0

0.00

 

20

100.00

High

21

100.00

Intermediate

0

0.00

Low

0

0.00

 

21

100.00

High

18

100.00

Intermediate

0

0.00

Low

0

0.00

 

18

100.00

High

15

100.00

Intermediate

0

0.00

Low

0

0.00

 

15

100.00

High

18

100.00

Intermediate

0

0.00

Low

0

0.00

18

100.00

High

14

70.00

Intermediate

6

30.00‡

AA-E-IP3

DR-C-IP2

DR-S-IP4

DR-S-LP2

DR-S-LP3

IE-E-IP1

IE-E-LP2

IE-E-LP4

IE-NE-IP1 † ‡

1344

Population nomenclature as per Indian Genome Variation Consortium convention. Proportion of each subpopulation with intermediate dose requirement of warfarin, if present in a subpopulation.

Pharmacogenomics (2014) 15(10)

future science group

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

Research Article

Table 4. Distribution of expected proportion of subpopulations with different levels of warfarin therapeutic dosage requirement based on distribution of VKORC1 and CYP2C9 genotypes across 24 Indian subpopulations (cont.). Population_ID† 

Dose requirement

n

%

IE-NE-IP1 (cont.)

Low

0

0.00

20

100.00

IE-NE-LP1

High

17

89.50

Intermediate

2

10.50‡

Low

0

0.00

19

100.00

High

15

88.20

Intermediate

2

11.80‡

Low

0

0.00

17

100.00

High

16

100.00

Intermediate

0

0.00

Low

0

0.00

IE-N-IP2

IE-N-LP1

IE-N-LP5

IE-N-LP9

IE-N-SP4

IE-W-LP1

IE-W-LP2

IE-W-LP3

† ‡

16

100.00

High

20

100.00

Intermediate

0

0.00

Low

0

0.00

20

100.00

High

21

91.30

Intermediate

2

8.70‡

Low

0

0.00

23

100.00

High

18

90.00

Intermediate

2

10.00‡

Low

0

0.00

20

100.00

High

16

88.90

Intermediate

2

11.10‡

Low

0

0.00

18

100.00

High

20

95.20

Intermediate

1

4.80‡

Low

0

0.00

21

100.00

High

20

100.00

Intermediate

0

0.00

Low

0

0.00

20

100.00

Population nomenclature as per Indian Genome Variation Consortium convention. Proportion of each subpopulation with intermediate dose requirement of warfarin, if present in a subpopulation.

future science group

www.futuremedicine.com

1345

Research Article  Giri, Khan, Grover et al.

Table 4. Distribution of expected proportion of subpopulations with different levels of warfarin therapeutic dosage requirement based on distribution of VKORC1 and CYP2C9 genotypes across 24 Indian subpopulations (cont.). Population_ID† 

Dose requirement

n

%

IE-W-LP4

High

17

94.40

Intermediate

1

5.60‡

Low

0

0.00

18

100.00

High

21

100.00

Intermediate

0

0.00

Low

0

0.00

21

100.00

High

9

40.90

Intermediate

13

59.10‡

Low

0

0.00

OG-W-IP

TB-NE-LP1

TB-N-IP1

TB-N-SP1

22

100.00

High

12

57.10

Intermediate

9

42.90‡

Low

0

0.00

21

100.00

High

7

33.30

Intermediate

14

66.70‡

Low

0

0.00

21

100.00

Population nomenclature as per Indian Genome Variation Consortium convention. ‡ Proportion of each subpopulation with intermediate dose requirement of warfarin, if present in a subpopulation. †

differs significantly when compared with other ethnic groups. Similarly, CYP4F2*3 was found to be more frequent in north Indians than in Caucasians, Asians and African–Americans (Supplementary Table 4) . Comparison of allele frequencies with the global population

We further compared the allele frequency of the variants vis-a-vis global population datasets available in the public domain. We have used data from the HapMap and 1000 Genome consortium for reference. In summary, the allele frequencies of the variants have been noted to be significantly different in the global population. The population frequencies of the alleles in different datasets are summarized in Supplementary Table 5. Allele frequency distributions among various populations are shown in a bubble plot (Figure 4) . In the case of rs1057910, the frequency has been noted to be variable from 0 to 0.131 in the global population, while for rs2108622, rs7294, rs9923231 and rs9934438 it has been 0.158–0.432, 0.046–0.688, 0.022–0.954 and 0.022–0.954, respectively (Supplementary Table 5) .

1346

Pharmacogenomics (2014) 15(10)

Discussion Warfarin is the mainstay of oral anticoagulation therapy worldwide because of its low cost, excellent oral bioavailability and proven efficacy. However, its narrow therapeutic index, severe adverse effects and the high interindividual variability in dose response complicate the use of warfarin in clinical settings. There have been no large-scale studies on genetic epidemiology of polymorphisms affecting warfarin metabolism in India. Here, we constructed a pharmacogenetic map of warfarin and determined initial dosing recommendations based on US FDA approvals of genetic testing in 2680 individuals across 24 ethnically diverse Indian subpopulations. The present study is the largest epidemiological genetic study globally, where genetic data were taken from previously published GWAS and IGV studies. The GWAS included both T2D cases and healthy controls. Although T2D subjects do not truly reflect controls for the present study, they were implicated in the genetic profiling of warfarin because we did not find significant differences in the allele frequencies of studied variants among T2D cases and

future science group

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

controls. None of the studied variants were associated with diabetes status in the published studies, hence, we were not expecting any serious implications of using diabetic individuals in the pharmacogenetic study of warfarin. Our analysis showed that the VKORC1 variant (-1639G>A) was prevalent among TBs and showed a high degree of variation in IEs, whereas CYP2C9 variants were observed with substantial variability in IEs residing in the northern and western part of India (Table 1) . The detailed study among IEs that further classified patients according to the state of their origin showed an increasing trend of extensive metabolizers for the variants involved in warfarin metabolism from

Research Article

the western Himalayan region (Punjab) to the eastern Gangetic region (Bihar; Figure 3). In a recent study, Scott et al. have studied the association of the CYP4F2*3 polymorphism with warfarin dosing among Asian and Caucasian populations [19] . Nevertheless, frequency of this variation in the Indian population has not been studied before. We are the first to report the presence of the CYP4F2*3 allele in the Indian population in the context of warfarin metabolism. Interestingly, among all six variants analyzed in the present study, CYP4F2*3 was the most prevalent and the allele frequency ranged from 0.30–0.44. The predicted outcome for warfarin dosing is determined

CYP4F2*3 (rs2108622, C>T) variant allele T N Punjab 30% Uttar Pradesh

44%

Bihar Haryana

41% 43%

Warfarin requirement

High dose 0

600

Low dose

kilometers

Figure 2. Allele frequency distribution among north Indians. Distribution of CYP4F2*3 allele frequency (variant allele T) was shown in various states of India. States are shaded according to minor allele frequency.

future science group

www.futuremedicine.com

1347

Research Article  Giri, Khan, Grover et al.

B

A

0.40

0.48 VKORC1 (rs9923231; T allele) Poor metabolizer

Allele frequency

Allele frequency

0.48

0.32 0.24 0.16 0.08

CYP2C9*3 (C allele) Poor metabolizer

0.40 0.32 0.24 0.16 0.08 0

0 Punjab Haryana Delhi

UP

C

Delhi

UP

Bihar

CYP4F2*3 (T allele) Extensive metabolizer

0.48 Allele frequency

Punjab Haryana

Bihar

0.40 0.32 0.24 0.16 0.08 0 Punjab

Haryana

Delhi

UP

Bihar

Figure 3. Trends of warfarin metabolizers among north Indians. Distribution of allele frequency of poor metabolizing variants ([A] VKORC1 [-1639G>A], [B] CYP2C9*3) and extensive metabolizing variants ([C] CYP4F2*3) was shown in various states of India. VKORC1 (-1639G>A) and CYP2C9*3 behave as poor metabolizers and CYP4F2*3 acts as an extensive metabolizer. Poor metabolizer frequency decreases from Punjab to Bihar whereas extensive metabolizer frequency inecreases from Punjab to Bihar. The solid black lines denote trend for increase/decrease in the allele frequency from one state to another. UP: Uttar Pradesh.

by the cumulative genotype frequency of relevant variants. No studies, to our knowledge, have investigated the combined frequency of CYP2C9 and CYP4F2 genotypes among Indians. Here, we demonstrated that genotyping of CYP4F2 and CYP2C9 variants may be helpful for predicting adverse drug reaction in IEs. FDA recommended doses, based on possible combinations of CYP2C9*3 and VKORC1 genotypes, can be broadly classified into low (0.5–2 mg), intermediate (3–4 mg) and high (5–7 mg) dose. Based on genotypic distribution of these loci in the Indian population, we observed that the majority of Indian sub­ populations may have a high dose requirement, except for TBs where a high proportion of individuals require an intermediate dose of warfarin in order to attain therapeutic international normalized ratio (Figure 5, Table 4 & Supplementary Table 3) . Such a strategy could potentially bring down the number of adverse drug events associated with high warfarin doses in TBs. However, our conclusion regarding dose require-

1348

Pharmacogenomics (2014) 15(10)

ment among the studied subpopulations may be limited because of the small number of individuals in each group, but we can also observe a trend emerging from these small numbers. Furthermore, in clinical practice the prediction of warfarin dosing based on allele frequencies alone may be inadequate since possibilities of contributions from other polymorphisms as well as nongenetic factors typical to each ethnic group cannot be excluded. The expected dose requirements among various ethnic groups based on only genotypes cannot verified in the absence of clinical data showing actual dose requirement and associated bleeding events. The high incidence of warfarin associated bleeding events among Indians may not be strictly owing to genetic variations because genetic variation accounts for A) frequency being as low as 6.5% in the OG sub­population to more than 70% in TBs. Furthermore, combined allele frequency data in IEs showed a distinct geographical distribution pattern, demonstrating an increased trend of extensive metabolizers and decreased trend of poor metabolizers from the Himalayan to the Gangetic region, suggesting that a standard dose of warfarin might result in excessive bleeding in patients from the Himalayan region, whereas it may be

insufficient to prevent thrombosis in patients from the Gangetic region. Furthermore, global comparison datasets showed that allelic and genotypic characteristics of the CYP2C9, VKORC1 and CYP4F2 variants in Indians are quite different from those in other ethnic groups across the globe. Here, we showed how genotype microarray datasets generated as a part of GWAS or genetic variation databases could be potentially useful in mining relevant genotypes and allele frequencies for clinically relevant variants that could be potentially valuable to clinicians in optimizing dosage strategies. Future perspective The very low VKORC1 (-1639G>A) variant frequency in the OG subpopulation suggests that some additional variants may play a role in warfarin dose requirement in this population. The high degree of variability among Indian subpopulations suggests that discrete GWAS among these subpopulations will likely identify additional genes and common variants with reasonable effects on warfarin dosing. Furthermore, in-depth analyses of pharmacogenomics markers, exome sequencing and whole genome sequencing studies have the potential to identify rare variants with significant effect size in these subpopulations. Additionally, genotypic micro-

Executive summary Background • Warfarin-induced bleeding is highly prevalent in India with 26% of treated south Indian patients developing excessive bleeding. • Among Asians, Indians require higher warfarin doses compared with Chinese and Malay patients, and these variations could mainly be due to different allele frequencies and polymorphisms in CYP2C9 and VKORC1. • In the absence of data concerning the allele frequencies of these variants in major ethnic groups in the Indian population, we investigated the mapping of six genetic variants in three important genes involving 2680 individuals across 24 ethnically diverse Indian subpopulations.

Genotype & allele frequencies of warfarin-metabolizing genes in Indians • The majority of Indians across 24 Indian subpopulations showed extensive metabolizer status for CYP2C9 genotypes, whereas 4.5–8.7% of individuals representing several Indo–European populations from the north and west of India displayed intermediate metabolizer status for CYP2C9. • As compared with CYP2C9 allelic variants, the VKORC1 variant (-1639G>A or rs9923231) showed highly variable distribution with an allele frequency as low as 6.5% in an out-group subpopulation to more than 70% in Tibeto–Burmans.

Geographical distribution of genetic variants in Indo–Europeans • The geographical distribution pattern of genetic variants (CYP2C9*3, CYP4F2*3 and VKORC1) showed an increased trend of extensive metabolizers and concomitant decreased trend of poor metabolizers in patients from the Himalayan region (Punjab) to Gangetic region (Bihar). • In the absence of other nongenetic factors, the standard dose of warfarin might result in excessive bleeding in patients from the Himalayan region, whereas it may be insufficient to prevent thrombosis in patients from the Gangetic region.

Conclusion • Based on genotypic distribution of major loci in the Indian population, our data suggests that the majority of Indian subpopulations may have a high-dose requirement for attaining therapeutic international normalized ratio; however, a careful assessment of dosing regimen is required in Tibeto–Burmans in whom a high proportion of individuals with intermediate dose requirement was observed. • When compared globally, the allelic and genotypic characteristics of CYP2C9, CYP4F2 and VKORC1 variants in Indians are quite distinct from those in other ethnic groups.

1352

Pharmacogenomics (2014) 15(10)

future science group

CYP2C9, CYP4F2 & VKORC1 polymorphisms in Indians 

array datasets available from various genome variation databases and other GWAS in a stratified population would provide a valuable resource to create the genetic landscape of relevant pharmacogenetic variants that may help in deciding the warfarin dose. Acknowledgements The authors are thankful to all the participating subjects for their support and cooperation in carrying out the study. The authors also acknowledge collaborators at OpenPGx.org for help in curating pharmacogenomic datasets from the ­literature.

Financial & competing interests disclosure Funding by the Council of Scientific and Industrial Research (CSIR), Government of India through CARDIOMED

References Papers of special note have been highlighted as: • of interest; •• of considerable interest

Research Article

BSC0122-(2) and SIP0006 is greatfully acknowledged. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.

•

Recent study demonstrating the requirement of high warfarin doses in Indians compared with other Asians.

10

Lee MT, Chen CH, Chuang HP et al. VKORC1 haplotypes in five east-Asian populations and Indians. Pharmacogenomics 10(10), 1609–1616 (2009).

11

Limdi NA, Wadelius M, Cavallari L et al. Warfarin pharmacogenetics: a single VKORC1 polymorphism is predictive of dose across 3 racial groups. Blood 115(18), 3827–3834 (2010).

1

Daly AK, King BP. Pharmacogenetics of oral anticoagulants. Pharmacogenetics 13(5), 247–252 (2003).

2

Eriksson N, Wadelius M. Prediction of warfarin dose: why, when and how? Pharmacogenomics 13(4), 429–440 (2012).

3

Nekkanti H, Mateti UV, Vilakkathala R, Rajakannan T, Mallayasamy S, Padmakumar R. Predictors of warfarininduced bleeding in a south Indian cardiology unit. Perspect. Clin. Res. 3(1), 22–25 (2012).

12

•

Recent paper describing the degree to which warfarin is associated with adverse events in the Indian population.

Indian Genome Variation Consortium. The Indian Genome Variation database (IGVdb): a project overview. Hum. Genet. 118(1), 1–11 (2005). 

•

4

Takeuchi F, McGinnis R, Bourgeois S et al. A genome-wide association study confirms VKORC1, CYP2C9, and CYP4F2 as principal genetic determinants of warfarin dose. PLoS Genet. 5(3), e1000433 (2009).

A genetic database of the Indian population covering important genes involved in complex diseases. Provides a source of genotype information for analysis of pharmacogenetic variants.

13

A genome-wide association study (GWAS) describing genetic polymorphisms in three important genes as the major determinants of warfarin dosing.

INDICO: the development of a resource for epigenomic study of Indians undergoing socioeconomic transition. Hugo J. 5(1–4), 65–69 (2011).

•

Provides detailed characteristic information of study participants.

14

Tabassum R, Chauhan G, Dwivedi OP et al. Genome–wide association study for Type 2 diabetes in Indians identifies a new susceptibility locus at 2q21. Diabetes 62(3), 977–986 (2013).

•

A GWAS carried out among Indo–Europeans residing in north India, which provided the source of genotype information for the analysis of the pharmacogenetic variants.

15

Adithan C, Gerard N, Vasu S, Balakrishnan R, Shashindran CH, Krishnamoorthy R. Allele and genotype frequency of CYP2C9 in Tamilnadu population. Eur. J. Clin. Pharmacol. 59(8–9), 707–709 (2003).

16

Jose R, Chandrasekaran A, Sam SS et al. CYP2C9 and CYP2C19 genetic polymorphisms: frequencies in the south Indian population. Fundam. Clin. Pharmacol. 19(1), 101–105 (2005).

17

Wang D, Chen H, Momary KM, Cavallari LH, Johnson JA, Sadee W. Regulatory polymorphism in vitamin K epoxide reductase complex subunit 1 (VKORC1) affects gene

••

5

Cini M, Legnani C, Cosmi B et al. A new warfarin dosing algorithm including VKORC1 3730 G > A polymorphism: comparison with results obtained by other published algorithms. Eur. J. Clin. Pharmacol. 68(8), 1167–1174 (2012).

6

Gupta S, Jain S, Brahmachari SK, Kukreti R. Pharmacogenomics: a path to predictive medicine for schizophrenia. Pharmacogenomics 7(1), 31–47 (2006).

7

Gupta M, Kaur H, Grover S, Kukreti R. Pharmacogenomics and treatment for dementia induced by Alzheimer’s disease. Pharmacogenomics 9(7), 895–903 (2008).

8

Reich D, Thangara JK, Patterson N, Price AL, Singh L. Reconstructing Indian population history. Nature. 461(7263), 489–494 (2009).

9

Gan GG, Phipps ME, Lee MM et al. Contribution of VKORC1 and CYP2C9 polymorphisms in the interethnic variability of warfarin dose in Malaysian populations. Ann. Hematol. 90(6), 635–641 (2011).

future science group

www.futuremedicine.com

1353

Research Article  Giri, Khan, Grover et al. expression and warfarin dose requirement. Blood 112(4), 1013–1021 (2008). 18

Indian Genome Variation Consortium. Genetic landscape of the people of India: a canvas for disease gene exploration. J. Genet. 87(1), 3–20 (2008).

29

Halling J, Petersen MS, Damkier P et al. Polymorphism of CYP2D6, CYP2C19, CYP2C9 and CYP2C8 in the Faroese population. Eur. J. Clin. Pharmacol. 61(7), 491–497 (2005).

19

Scott SA, Khasawneh R, Peter I, Kornreich R, Desnick RJ. Combined CYP2C9, VKORC1 and CYP4F2 frequencies among racial and ethnic groups. Pharmacogenomics 11(6), 781–791 (2010).

30

Kimura M, Ieiri I, Mamiya K, Urae A, Higuchi S. Genetic polymorphism of cytochrome P450s, CYP2C19, and CYP2C9 in a Japanese population. Ther. Drug Monit. 20(3), 243–247 (1998).

••

Landmark paper describing the importance of combined allele frequencies as a determinant of warfarin response in indiviuals.

31

20

Siddiqi A, Khan DA, Khan FA, Naveed AK. Impact of CYP2C9 genetic polymorphism on warfarin dose requirements in Pakistani population. Pak. J. Pharm. Sci. 23(4), 417–422 (2010).

Mark L, Marki-Zay J, Fodor L, Hajdara I, Paragh G, Katona A. Cytochrome P450 2C9 polymorphism and acenocoumarol therapy. Kardiol. Pol. 64(4), 397–402 (2006).

32

Nasu K, Kubota T, Ishizaki T. Genetic analysis of CYP2C9 polymorphism in a Japanese population. Pharmacogenetics 7(5), 405–409 (1997).

21

Aynacioglu AS, Brockmoller J, Bauer S et al. Frequency of cytochrome P450 CYP2C9 variants in a Turkish population and functional relevance for phenytoin. Br. J. Clin. Pharmacol. 48(3), 409–415 (1999).

33

Scordo MG, Aklillu E, Yasar U, Dahl ML, Spina E, Ingelman–Sundberg M. Genetic polymorphism of cytochrome P450 2C9 in a Caucasian and a black African population. Br. J. Clin. Pharmacol. 52(4), 447–450 (2001).

22

Bozina N, Granic P, Lalic Z, Tramisak I, Lovric M, Stavljenic–Rukavina A. Genetic polymorphisms of cytochromes P450: CYP2C9, CYP2C19, and CYP2D6 in Croatian population. Croat. Med. J. 44(4), 425–428 (2003).

34

Sullivan-Klose TH, Ghanayem BI, Bell DA et al. The role of the CYP2C9–Leu359 allelic variant in the tolbutamide polymorphism. Pharmacogenetics 6(4), 341–349 (1996).

35

Taube J, Halsall D, Baglin T. Influence of cytochrome P-450 CYP2C9 polymorphisms on warfarin sensitivity and risk of over-anticoagulation in patients on long-term treatment. Blood 96(5), 1816–1819 (2000).

36

Wang SL, Huang J, Lai MD, Tsai JJ. Detection of CYP2C9 polymorphism based on the polymerase chain reaction in Chinese. Pharmacogenetics 5(1), 37–42 (1995).

37

Yasar U, Eliasson E, Dahl ML, Johansson I, Ingelman– Sundberg M, Sjoqvist F. Validation of methods for CYP2C9 genotyping: frequencies of mutant alleles in a Swedish population. Biochem. Biophys. Res. Commun. 254(3), 628–631 (1999).

38

Yoon YR, Shon JH, Kim MK et al. Frequency of cytochrome P450 2C9 mutant alleles in a Korean population. Br. J. Clin. Pharmacol. 51(3), 277–280 (2001).

39

Sana Abbas US, Bhatti S, Naqi A, Ali M, Nawaz N, Alam R. Interethnic diversity of CYP2C9 alleles on warfarin pharmacogenetics in healthy Sindhi and Baluchi population samples: correlation to global allelic frequencies. Open Access Sci. Rep. 1(11), 4 (2012).

40

Giri AK, Khan NM, INDICO Pharmacogenetic landscape of clopidogrel in north Indians suggest distinct inter-population differences in allele frequencies. Pharmacogenomics 15(5), 643–653 (2014).

23

24

Bravo-Villalta HV, Yamamoto K, Nakamura K, Baya A, Okada Y, Horiuchi R. Genetic polymorphism of CYP2C9 and CYP2C19 in a Bolivian population: an investigative and comparative study. Eur. J. Clin. Pharmacol. 61(3), 179–184 (2005). Cascorbi I AE, Sachse C, Brockmoller J, Roots I. A novel CYP2C9 intron 2T/C transition and linkage to mutations Leu359 and Cys144. Clin. Pharmacol. Ther. 63, 198 (1998).

25

Dickmann LJ, Rettie AE, Kneller MB et al. Identification and functional characterization of a new CYP2C9 variant (CYP2C9*5) expressed among African Americans. Mol. Pharmacol. 60(2), 382–387 (2001).

26

Dorado P, Berecz R, Norberto MJ, Yasar U, Dahl ML, LLerena A. CYP2C9 genotypes and diclofenac metabolism in Spanish healthy volunteers. Eur. J. Clin. Pharmacol. 59(3), 221–225 (2003).

27

28

1354

impaired CYP2C9 metabolism in a Caucasian population. Eur. J. Clin. Pharmacol. 57(1), 47–49 (2001).

Gaedigk A, Casley WL, Tyndale RF, Sellers EM, Jurima–Romet M, Leeder JS. Cytochrome P4502C9 (CYP2C9) allele frequencies in Canadian Native Indian and Inuit populations. Can. J. Physiol. Pharmacol. 79(10), 841–847 (2001). Garcia-Martin E, Martinez C, Ladero JM, Gamito FJ, Agundez JA. High frequency of mutations related to

Pharmacogenomics (2014) 15(10)

future science group