Abstract
Interindividual variability in analgesic effects of nonsteroidal anti‐inflammatory drugs prescribed for sickle cell disease (SCD) pain is attributed to polymorphisms in the CYP2C8 and CYP2C9 enzymes. We described CYP2C8 and CYP2C9 genotype/phenotype profiles and frequency of emergency department (ED) visits for pain management in an African American SCD patient cohort. DNA from 165 unrelated patients was genotyped for seven CYP2C8 and 15 CYP2C9 alleles using the iPLEX ADME PGx multiplexed panel. CYP2C8*1 (0.806),*2 (0.164), *3 (0.018), and *4 (0.012) alleles were identified. Genotype frequencies were distributed as homozygous wild type (66.7%), heterozygous (27.8%), and homozygous variant/compound heterozygous (5.4%), respectively. CYP2C9*1 (0.824), *2 (0.027), *3 (0.012), *5 (0.009), *6 (0.009), *8 (0.042), *9 (0.061), and *11(0.015) were observed with extensive (68.5%), intermediate (18.1%) and poor predicted metabolizers (0.6%), respectively. Fifty‐two and 55 subjects, respectively had at least one variant CYP2C8 or CYP2C9 allele. Although the distribution of the CYP2C9 (p = 0.0515) phenotypes was marginally significantly in high and low ED users; some CYP2C8 and CYP2C9 allelic combinations observed in 15.2% (25) of the cohort are associated with higher risks for analgesic failure. CYP2C8 and CYP2C9 preemptive genotyping could potentially enable clinicians to identify patients with impaired metabolic phenotypes.
Keywords: NSAIDs, sickle cell disease, pharmacogenetics, CYP2C9, CYP2C8
Introduction
Sickle cell disease (SCD) is one of the most common genetic blood disorders worldwide that affects predominantly people of African ancestry.1 The hallmark of SCD is the occurrence of painful vaso‐occlusive episodes (VOCs) that can start as early as 6 months of age in pediatric patients and continue to occur unpredictably throughout adult life. Severe VOC pain is associated with acute chest syndrome, organ failure and frequent visits to the emergency department (ED) for parenteral opioid treatment.2 Prodromal signs of VOC pain are treated with a weak opioid such as codeine but more commonly with nonsteroidal anti‐inflammatory drugs (NSAIDs) because of their anti‐inflammatory, analgesic, and antipyretic effects.3, 4 However, many individuals with SCD display variable response to NSAID treatment and some even fail to achieve adequate analgesia with standard doses of NSAIDs.2, 3
NSAIDs (e.g., ibuprofen, diclofenac, ketoprofen, naproxen, flurbiprofen, meloxicam, piroxicam, and tenoxicam) are metabolized by two enzymes of the cytochrome P450 superfamily, mainly the CYP2C8 and CYP2C9.5, 6 Polymorphisms in these two genes have been associated with decreased enzyme activity and alteration of NSAIDs pharmacokinetic parameters.6 Both the CYP2C8 and CYP2C9 enzymes are highly polymorphic and various allelic variants reported. More than 16 alleles and over 60 variants have been characterized for the CYP2C8 and CYP2C9 enzymes, respectively (http://www.cypalleles.ki.se/). Allelic variants impacts the metabolic activity of the CYP450 enzymes; and previous determinations of enzymatic activity and expression of most CYP450 drug metabolizing enzymes revealed four distinct metabolic phenotypes: ultrarapid metabolizers (UMs), extensive metabolizers (EMs), intermediate metabolizers (IMs) and poor metabolizers (PMs).5, 6 PMs are compound heterozygous for different inactivating alleles or homozygous for an inactivating variant and may display variation in the severity of functional enzyme deficiencies. IMs carry one functional allele and one nonfunctional allele but may demonstrate a wide range of enzymatic activity. EMs have two functional alleles. UMs carry multiple copies of functional alleles. Current NSAIDs dosing strategy in patients with SCD is based on the assumption that the individual patient is an EM. However, accumulated evidence indicates association between decreased or loss of function CYP2C8 and CYP2C9 alleles with suboptimal therapeutic response and adverse effects of NSAIDs.5, 6, 7, 8 For SCD patients, suboptimal therapeutic may possibly be linked with higher likelihood of being admitted to hospital for either analgesic drug failure. To date however, relatively few studies have attempted to bridge the concept of pharmacogenetic variability as a determinant of interindividual response to NSAID therapy in SCD patients.9, 10, 11, 12 In this study, we determined the frequency of pharmacologically relevant allelic variants of the CYP2C8 and CYP2C9 enzymes in a SCD patient cohort and correlate metabolic phenotypes with frequency of ED visits.
Methods
Human subjects
The study participants were randomly selected patients with SCD receiving care at the Georgia Regents University Comprehensive Sickle Cell Center clinics. The clinics are located in six towns in southeastern Georgia. The study was approved by the Georgia Regents University Institutional Review Board. Written informed consent or assent was obtained from each patient prior to inclusion into the study. Study participants were recruited between January 2011 and January 2013. Medical records of the study participants were reviewed to abstract SCD genotypes, NSAID prescriptions, clinical, and acute care utilization data.
CYP2C8 and CYP2C9 genotyping
Whole blood samples (10 mL in tubes containing EDTA) were collected from the study participants in steady state. Genomic DNA was extracted using the Puregene DNA Purification Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. We used the iPLEX ADME PGx multiplex panel (Sequenom, Inc, San Diego, CA, USA) to genotyped seven CYP2C8 alleles (*1, *2, *3, *4, *5, *7, and *8) and 15 CYP2C9 alleles (*1, *2, *3, *4, *5, *6, *8, *9, *10, *11, *12, *13, *15, *25, and *27) across all study participants as previously described.13 Briefly, the iPLEX® ADME PGx multiplexed panel uses Sequenom Bioscience's iPLEX biochemistry with specific ADME oligo multiplex mixes on the MassARRAY system to simultaneously interrogate 192 biologically relevant polymorphisms in 36 pharmacogenes. After running the reactions, mutations were detected, quantified, and genotype reports automatically created using Sequenom TYPER software (http://bioscience.sequenom.com/iplex‐adme‐pgx‐panel). TYPER software assigns the wild‐type (*1) CYP2C8 and CYP2C9 alleles in the absence of other detectable variant alleles The CYP allele designations refer to those defined by the Cytochrome P450 Allele Nomenclature Committee.14
Statistical analysis
The primary outcome measure was genotype frequencies. The secondary end point compares CYP2C8 and CYP2C9 genotypes with the number of ED visits for VOC pain. Descriptive analyses were used for baseline demographic and clinical data and to compare allele frequencies between the study participants and published data of other populations. The CYP2C8 and CYP2C9 allele and genotype frequencies were presented as percentage of the study cohort with 95% confidence interval. The observed genotype frequencies were compared with those expected for concordance with Hardy‐Weinberg equilibrium using the X 2 test. A p value of less than 0.05 was deemed to represent statistical significance. All statistical analyses were performed using SPSS statistical package version 19.0.
Results
Demographics
This study elucidates the allelic variants of the CYP2C8 and CYP2D9 in a SCD cohort. A total of 165 SCD patients (82 males) were recruited. The study participants were all African Americans. Race was self‐reported by the subjects. The study participants’ demographic features and clinical characteristics are summarized in Table 1. The subjects ranged in age from 16 to 61 years and their body mass index ranged from 15.3 to 38.4. SCD genotype frequencies were distributed as SS (97.5%), SB Thal° (1.8%), and S‐Los Angeles (0.6%), respectively. Ibuprofen, aspirin and naproxen were the NSAIDs routinely prescribed. One hundred ten subjects (67%) were prescribed hydroxyurea. In terms of treatment of SCD, hydroxyurea is the only FDA approved drug and has been associated with decreased frequency of VOC and morbidity.2, 4
Table 1.
Cohort demographic and clinical characteristics.
| Sex | Male Female |
No (%) of subjects 82 (49.6%) 83 (50.4%) |
| Age | ||
| Median | ||
| Male | 30.5 yrs. (16 – 57) | |
| Female | 31.7 yrs. (18 – 61) | |
| Ethnic origin | African‐American | 165 (100%) |
| SCD genotype | SS SB Thal° S‐Los Angeles |
161 (97.5%) 3 (1.8%) 1 (0.6%) |
| BMI | ||
| Median | ||
| Male | 23.1 (15.3 – 35.2) | |
| Female | 23.7 (16.5 – 38.4) | |
| NSAIDs | Ibuprofen (Motrin, Advil) Naproxen (Aleve) Aspirin Acetaminophen (Tylenol) |
50 (30.3%) 10 (6.1%) 5 (3%) 20 (12.1%) |
| Hydroxyurea therapy | Yes | 110 (67%) |
CYP2C8 alleles and genotype frequencies
Table 2 showed our cohort CYP2C8 allele and genotype frequencies. The CYP2C8*1 is considered the wild type with normal enzyme activity. The abnormal CYP2C8*2 and *3 alleles are the most prevalent alleles associated with decreased enzyme activity, but are unevenly distributed in racial and geographic populations.15 The CYP2C8*4 through CYP2C8*14 variants alleles are rare and found in less than 1% of racial populations.16, 17 In our cohort, we identified four CYP2C8 alleles. The CYP2C8*1 wild‐type allele frequency was 0.806. Of the three variant alleles identified, the *2 occurred in the highest frequency (0.164). This was followed by CYP2C8*3 (0.018) and *4 (0.012), respectively. The CYP2C8*5, *7, and *8 rare variants found in less than 1% of populations, mainly Asians, were not detected in our cohort.18, 19
Table 2.
Allelic, genotypic and phenotype frequencies of CYP2C8 gene polymorphisms
| Allele | Genetic mutation | Enzyme activity | Number of alleles | Allelic frequency | 95% CI (±) | 95% CI range |
|---|---|---|---|---|---|---|
| *1 | Wild‐type | Normal | 266 | 0.806 | 0.043 | 0.763–0.849 |
| *2 | Missense | Decreased | 54 | 0.164 | 0.040 | 0.124–0.204 |
| *3 | Missense | Decreased | 6 | 0.018 | 0.014 | 0.004–0.033 |
| *4 | Missense | Decreased | 4 | 0.012 | 0.012 | 0.000–0.024 |
| *5 | Frame shift | None | 0 | 0.000 | 0.000 | |
| *7 | Missense | None | 0 | 0.000 | 0.000 | |
| *8 | Missense | Decreased | 0 | 0.000 | 0.000 | |
| Genotype | Number of subjects | Observed genotype frequency (%) | Expected genotype frequency (%) | Predicted phenotype | ||
| Wild‐type | ||||||
| *1/*1 | 110 | 66.7 | 65.0 | EM | ||
| Heterozygous | ||||||
| *1/*2 | 40 | 24.2 | 26.4 | IM | ||
| *1/*3 | 4 | 2.4 | 2.9 | IM | ||
| *1/*4 | 2 | 1.2 | 2.0 | IM | ||
| Subtotal | 46 | 27.8 | ||||
| Homozygous variant/compound heterozygous | ||||||
| *2/*2 | 5 | 3.0 | 2.7 | PM | ||
| *2/*3 | 2 | 1.2 | 0.6 | PM | ||
| *2/*4 | 2 | 1.2 | 0.4 | PM | ||
| *3/*3 | 0 | 0.0 | 0.0 | PM | ||
| *3/*4 | 0 | 0.0 | 0.0 | PM | ||
| Subtotal | 9 | 5.4 | ||||
| Total | 165 | 100 | 99.9 | |||
CI = confidence interval.
The CYP2C8*1, *2, *3, and *4 frequencies were concordant with Hardy‐Weinberg equilibrium. Because of substrate‐dependent functional activity of the CYP2C8 alleles and discrepancies between in vitro and in vivo data the genotype frequencies were distributed as homozygous wild type, heterozygous and homozygous variant/compound heterozygous, respectively.17 The CYP2C8 1/*1 was the most frequent genotype related to NSAID metabolism and it corresponds to an predicted EM phenotype (66.7%); the CYP2C8 *1/*2, *1/*3, and *1/*4 genotypes correspond to an IM phenotype and accounted for 27.8% of the study cohort; whereas the PMs (*2/*2, *3/*3, and *2/*4 genotypes) accounted for 5.4% of cohort.
CYP2C9 allele, genotypes, and phenotype frequencies
The CYP2C9 allele, genotype and predicted metabolic phenotype frequencies are summarized in Table 3. We surveyed 15 CYP2C9 alleles (*1, *2, *3, *4, *5, *6, *8, *9, *10, *11, *12, *13, *15, *25, and *27) and identified eight alleles (*1, *2, *3, *5, *6, *8, *9, *11) with the wild‐type *1 occurred in the highest frequency (0.824) in our cohort. The CYP2C9*8 (0.042) and *9 (0.061) were the most common variant alleles. The combined frequency for the reduced activity CYP2C9*2, *3, *5, *8, *9, *11 and the null *6 variants was 0.176. The CYP2C9 *4, *12, *13, *15, *25, and *27 alleles were not detected in our cohort. The high frequency (0.061) of the CYP2C9*9 allele made *1/*9 the most common genotype with allelic variant in the cohort (9.7%). The observed frequencies for the overall cohort were concordant with Hardy‐Weinberg equilibrium. CYP2C9 phenotypes have been designated extensive (two functional alleles), intermediate (one functional allele/one dysfunctional), and PMs (two nonfunctional alleles). Based on the observed genotypes and published criteria, we assigned predicted phenotype frequencies for our study participants as follows: extensive (68.5%), intermediate (18.1%), and PMs (0.6%), respectively.20, 21, 22, 23, 24 Because some of the variant CYP2C9 alleles do not have clear phenotypic consequences, the predicted metabolic phenotype for four CYP2C9 genotypes (*5/*9, *6/*8, *8/*9, and *9/*11) were indeterminate.
Table 3.
Allelic, genotypic, and phenotypic frequencies of CYP2C9 gene polymorphisms.
| Allele | Genetic mutation | Enzyme activity | Number of alleles | Allelic frequency | 95% CI (±) | 95% CI range |
|---|---|---|---|---|---|---|
| *1 | Wild‐type | Normal | 272 | 0.824 | 0.041 | 0.783–0.865 |
| *2 | Missense | Decreased | 9 | 0.027 | 0.018 | 0.010–0.045 |
| *3 | Missense | Decreased | 4 | 0.012 | 0.012 | 0.000–0.024 |
| *4 | Missense | NA | 0 | 0.000 | 0.000 | |
| *5 | Missense | Decreased | 3 | 0.009 | 0.010 | −0.001–0.019 |
| *6 | Frame shift | None | 3 | 0.009 | 0.010 | −0.001–0.019 |
| *8 | Missense | Decreased | 14 | 0.042 | 0.022 | 0.021–0.064 |
| *9 | Missense | NA | 20 | 0.061 | 0.026 | 0.035–0.086 |
| *10 | Missense | NA | 0 | 0.000 | 0.000 | |
| *11 | Missense | Decreased | 5 | 0.015 | 0.013 | 0.002–0.028 |
| *12 | Missense | Decreased | 0 | 0.000 | 0.000 | |
| *13 | Missense | Decreased | 0 | 0.000 | 0.000 | |
| *15 | Missense | NA | 0 | 0.000 | 0.000 | |
| *25 | Frame shift | None | 0 | 0.000 | 0.000 | |
| *27 | NA | NA | 0 | 0.000 | 0.000 | |
| Predicted metabolizer phenotype | Number of subjects | Observed genotype frequency (%) | Expected genotype frequency (%) | |||
| Extensive metabolizer | ||||||
| *1/*1 | 113 | 68.5 | 67.9 | |||
| Intermediate metabolizer | ||||||
| *1/*2 | 8 | 4.8 | 4.5 | |||
| *1/*3 | 3 | 1.8 | 2.0 | |||
| *1/*5 | 2 | 1.2 | 1.5 | |||
| *1/*6 | 2 | 1.2 | 1.5 | |||
| *1/*8 | 11 | 6.7 | 7.0 | |||
| *1/*9 | 16 | 9.7 | 10.0 | |||
| *1/*11 | 4 | 2.4 | 2.5 | |||
| Subtotal | 46 | 27.8 | ||||
| Poor metabolizer | ||||||
| *2/*3 | 1 | 0.6 | 0.1 | |||
| Unknown metabolizer status | ||||||
| *5/*9 | 1 | 0.6 | 0.1 | |||
| *6/*8 | 1 | 0.6 | 0.1 | |||
| *8/*9 | 2 | 1.2 | 0.5 | |||
| *9/*11 | 1 | 0.6 | 0.2 | |||
| Subtotal | 5 | 3.0 | ||||
CI = confidence interval.
Correlation between study participants’ genomic and clinical data
Table 4 describes the study participants’ CYP2C8 and CYP2C9 genomic and ED clinical data. Out of a total of 152 participants with ED visit clinical records, we had 39 high ED users (≥3 ED visits per year) and 113 low ED users (<3 ED visits per year). The distribution for the CYP2C8 predicted phenotypes was not significantly different in high and low ED users (p = 0.1668). However, the distribution of predicted phenotypes was marginally significantly in high and low ED users (p = 0.0515) for the CYP2C9. Table 5 presents detailed genomic and clinical data, including NSAIDs prescriptions, number of VOC days and pain score ranges during ED visits or hospital admissions for selected participants with deficient genomic metabolic and clinical risk profiles for NSAIDs therapeutic failure. Table 6 compares our CYP2C8 and CYP2C9 data to allelic data previously reported in other African American and African ethnic populations.19, 20, 21, 22, 23, 24, 25 There is limited published data available from other populations for several minor frequency CYP2C8 and CYP2C9 alleles making it somewhat difficult to compare our data with some African ethnic groups with high incidence of SCD.
Table 4.
Distribution of CYP2C8 & CYP2C9 phenotypes and ED visits.
| CYP2C8 | *High | +Low | CYP2C9 | High | Low |
|---|---|---|---|---|---|
| EM | 23 (0.61) | 83 (0.74) | EM | 27 (0.71) | 91 (0.81) |
| IM | 15 (0.39) | 29 (0.26) | IM | 9 (0.24) | 10 (0.09) |
EM = extensive metabolizer; IM = intermediate metabolizer; ED = emergency department; UNK = Unknown.
*(≥3 ED visits per year); +(<3 ED visits per year).
Table 5.
Genomic and clinical data for selected SCD participants with compromised metabolic genotypes.
| # of subject | CYP2C8 genotype | CYP2C8 phenotype | CYP2C9 genotype | CYP2C9 phenotype | Pain range* | # of VOC days | ED visit | Hosp. Adm. | NSAID, number of times prescribed and place | Hu therapy |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | *1/*1 | EM | *1/*8 | IM | 0–8 | 3 | 1 | 2 | Ibuprofen 800 mg x14, tylenol 500 mg x10, ibuprofen 800 x17 | Y |
| 2 | *1/*3 | IM | *1/*2 | IM | 0–7 | 1 | 0 | 1 | Naproxen 500 x39 | N |
| 3 | *1/*3 | IM | *1/*2 | IM | 0–9 | 67 | 0 | 5 | N/A | Y |
| 4 | *1/*1 | EM | *1/*8 | IM | 0–9 | 8 | 2 | 1 | Tylenol 250 x10, ibuprofen 800 mg x 8 (hosp), toradol 30 mg IV x5 (hosp) | Y |
| 5 | *1/*1 | EM | *1/*8 | IM | N/A | N/A | 3 | 4 | ASA unknown dose x 26 | Y |
| 6 | *1/*3 | IM | *1/*2 | IM | 0–5 | 1 | 2 | 5 | Toradol 75 mg iv (hosp), acetaminophen 875 (hosp), goody pm x64 | Y |
| 7 | *2/*3 | PM | *1/*2 | IM | 0–9 | 1 | 1 | 3 | Motrin 800 x 2 | Y |
| 8 | *1/*1 | EM | *6/*8 | UNK (PM?) | 16 | 6 | 2 | Motrin 800 x 4 (hosp), toradol 60 IM (hosp), theraflu unknown dose x 45, motrin 800 x 72, tylenol 500 x 14 | Y | |
| 9 | *1/*2 | IM | *1/*8 | IM | 0–9 | 10 | 0 | 0 | Ibuprofen 800 mg x 24, tylenol 500 mg x4 | N |
| 10 | *1/*1 | EM | *1/*8 | IM | N/A | 1 | 0 | 0 | Tylenol 500 mg x1 | N |
| 11 | *2/*3 | PM | *1/*2 | IM | 0–7 | 37 | 3 | 1 | Ibuprofen 200 mg x 85 | Y |
| 12 | *1/*1 | EM | *8/*9 | UNK | 0–9 | 0 | 0 | 1 | N/A | N |
| 13 | *1/*2 | IM | *1/*6 | IM | 0–4 | 4 | 1 | 2 | N/A | N |
| 14 | *1/*1 | EM | *1/*8 | IM | 0–8 | 0 | 0 | 1 | N/A | Y |
| 15 | *1/*1 | EM | *1/*8 | IM | 0–1 | 4 | 0 | 0 | Excedrin unknown dose x 17, advil unknown dose x 1 | N |
| 16 | *1/*2 | IM | *1/*3 | IM | 0–9 | N/A | 1 | 0 | N/A | N |
| 17 | *1/*2 | IM | *1/*8 | IM | N/A | N/A | 1 | 0 | N/A | Y |
| 18 | *1/*3 | IM | *1/*2 | IM | N/A | N/A | 0 | 2 | N/A | N |
| 19 | *1/*1 | IM | *8/*9 | UNK | N/A | N/A | 0 | 0 | N/A | Y |
| 20 | *1/*2 | IM | *1/*8 | IM | N/A | N/A | 4 | 0 | N/A | Y |
| 21 | *1/*1 | IM | *1/*3 | IM | N/A | 0 | 4 | 10 | N/A | Y |
| 22 | *1/*2 | IM | *1/*8 | IM | N/A | N/A | 0 | 0 | N/A | Y |
| 23 | *1/*1 | EM | *1/*8 | IM | N/A | N/A | 0 | 15 | Ibuprofen 800 x 11 | N |
| 24 | *1/*4 | IM | *1/*2 | IM | 0–9 | 3 | 8 | 12 | N/A | Y |
| 25 | *1/*1 | EM | *2/*3 | PM | N/A | N/A | 3 | 3 | N/A | Y |
ADM = admission; EM = extensive metabolizer; IM = intermediate metabolizer; PM = poor metabolizer; Hosp = hospital; ED = emergency department; HU = hydroxyurea; N/A = not available; UNK = unknown; Y = yes; N = no; *Pain range during ED or Hospital admission.
Table 6.
CYP2C9 & CYP2C8 frequencies in previously studied populations.
| CYP2C9 alleles frequency percentage | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Racial and ethnic group | *1 | *2 | *3 | *4 | *5 | *6 | *8 | *9 | *11 | *12 | *13 | Ref. |
| African Americans (n = 165) | 0.824 | 0.027 | 0.012 | 0.000 | 0.009 | 0.009 | 0.042 | 0.061 | 0.015 | 0.000 | 0.000 | This study |
| African American (n = 300) | 0.867 | 0.028 | 0.020 | 0.000 | 0.015 | 0.010 | 0.047 | – | 0.013 | – | 0.000 | 19 |
| Ghanaian (n = 204) | – | 0.000 | 0.000 | 0.000 | 0.000 | – | – | – | 0.020 | – | – | 21 |
| Beninese (n = 111) | 0.955 | 0.000 | 0.000 | 0.000 | 0.018 | – | – | – | 0.027 | – | – | 22 |
| Mozambican (n = 206) | – | 0.000 | 0.010 | – | 0.019 | 0.000 | 0.146 | – | 0.024 | – | – | 23 |
| Africans (n = 250) | 0.810 | 0.120 | 0.000 | – | 0.010 | 0.012 | 0.040 | 0.095 | 0.022 | 0.020 | – | 20 |
| CYP2C8 alleles frequency percentage | ||||||||||||
| *1 | *2 | *3 | *4 | *5 | *7 | *8 | *9 | *10 | *11 | *12 | ||
| African American (n = 165) | 0.806 | 0.164 | 0.018 | 0.012 | 0.000 | 0.000 | 0.000 | – | – | – | – | This study |
| African American (n = 500) | 0.878 | 0.100 | 0.010 | 0.012 | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 | – | 0.000 | 24 |
| Mozambican (n = 360) | 0.160 | 0.048 | 0.000 | 0.005 | 0.000 | – | – | – | – | – | – | 25 |
| Ghanaian (n = 204) | – | 0.170 | 0.000 | 0.000 | – | – | – | – | – | – | – | 21 |
(–) Allele not screened in study
Discussion
To the best of our knowledge, this study describes the combined CYP2C8 and CYP2C9 allelic frequencies, genotypes and predicted phenotypes for the first time in an African American SCD patient cohort. Our study identified twenty five individuals with combined impaired CYP2C8 and CYP2C9 genotypes characterized mainly as intermediate and PMs. Interindividual variation in drug response is greatest in IM phenotypic group where it is difficult to determine unequivocally the quantitative/percentage value of altered functionality of allelic variants, except for null alleles.26, 27, 28 Perhaps more significantly, included in this phenotypic group are the six individuals with the delirious CYP2C8*3 and CYP2C9*2 allelic combination associated with major ibuprofen clearance impairment and analgesic failure.28 To the best of our knowledge, no prior studies have identified specific African American subjects with these two genes allelic combination or assess their implications for NSAID analgesic failure.
We did not observe great variations in the frequency of null and reduced function alleles in the CYP2C8 and CYP2C9 genes to ranges reported in other African American populations. Interestingly, two previous genotyping studies reported slightly higher frequencies of mutant gene deletion alleles of the CYP3A5 and CYP2D6 in SCD patients compared to healthy African Americans.9, 12 These deficient alleles are associated with the PM phenotype and impaired functionality and have also been associated with failure of codeine treatment for VOC in children on hydroxyurea; and linked with higher likelihood of pediatric and adult SCD patients being admitted to the ED for parenteral opioid management.10, 11 Although the distribution of CYP2C9 predicted phenotypes was marginally significantly in high and low ED users in our study, it was unclear in the above referenced studies whether the individuals at risk for frequent ED visits also had deficient CYP2C8 and CYP2C9 genotypes. Nonetheless, in our cohort, ten out of twelve of the combined alleles of the CYP2C8 and CYP2C9 enzymes identified contribute to deficient metabolic genotypes. Fifty‐two subjects had at least one variant CYP2C9 allele (*2, *3,*5, *6, *8, *9, and *11) associated with either the IM, poor, or indeterminate phenotype. Fifty‐five subjects had at least one CYP2C8 variant allele (*2, *3, and *4) that contributes to impaired metabolic genotypes, and nine subjects were homozygous or compound heterozygous for deficient metabolic genotypes as determined in previous pharmacokinetic and pharmacodynamics studies.26, 27, 28, 29, 30, 31, 32, 33, 34
Pharmacokinetics and pharmacodynamics effects of CYP2C8 and CYP2C9 alleles
The influence of some CYP2C8 and CYP2C9 genotypes on analgesic response and therapeutic outcomes of a number of NSAIDs including ibuprofen and naproxen have been established in previous pharmacokinetics studies.26, 27, 28 Ibuprofen pharmacokinetics data is strongly related to variant CYP2C8 and CYP2C9 genotypes: heterozygous and homozygous carriers of the CYP2C8*3 allele display ibuprofen metabolic clearance reduction of approximately 62% and 10%, respectively when compared to individuals homozygous for the CYP2C8*1 and CYP2C9*1 genotype.25 Metabolic clearance values in subjects heterozygous and homozygous for CYP2C9*2 but not carrying any other allelic mutations were 96% and 84%, respectively when compared to individuals with CYP2C8*1 and CYP2C9*1 wild‐type alleles.27 The CYP2C9*2 allele when linked with the CYP2C8*3, translates into a major impairment on ibuprofen clearance as indicated above.28 Individuals carrying CYP2C9*3 variant alleles display a mean reduction of clearance of approximately 65% and 17% for heterozygous and homozygous carriers, respectively.27 In a recent study of 130 healthy individuals who received a single oral dose of 400 mg ibuprofen, the oral clearance of ibuprofen was 4.43, 3.26, 2.91, 2.05, 1.83, and 1.13 1/hour for individuals with CYP2C9*1/*1, *1/*2, *1/*3, *2/*2, *2/*3, and *3/*3 genotypes, respectively.28 The effects of the variant CYP2C9 alleles are dissimilar for all NSAIDs.29 Studies with tenoxicam, however, indicated that oral clearance among carriers of CYP2C9*2 and CYP2C9*3 decreases to approximately 70% and 55%30; while oral clearance of meloxicam in individuals with the CYP2C9*1/*13 genotype was significantly decreased by 62% compared to individuals with the CYP2C9*1/*1 genotype.31
Pharmacodynamics studies have also implicated CYP2C8 and CYP2C9 genotypes in gastrointestinal toxicity of NSAIDs. Martinez and colleagues found the CYP2C9*2 and *3 associated with a two‐and‐a‐half fold increased risk of gastric bleeding episode after dosing with NSAIDs such as celecoxib, diclofenac, ibuprofen, indomethacin, lornoxicam, piroxicam, or naproxen. The increased risk was attributed to the *2 allele which was detected in 23.4% of the study subjects with gastric bleeding episode compared with 13.7% of the control subjects.32 In another study of gastrointestinal bleeding in NSAIDs users, the frequencies of the CYP2C8*3 and CYP2C9*2 alleles were higher in NSAIDs user who experienced a bleed versus those who did not experience a bleed (CYP2C8*3, odds ratio: 2.4, p < 0.002; CYP2C9*2, odds ratio: 2.7, p < 0.013).33 Pilotto et al. found that a significantly higher frequency of CYP2C9*1/*3 and *1/*2 genotypes were identified in patients with endoscopically documented NSAID‐related gastroduodenal bleeding lesions compared to a matched control group. In the study described, the presence of the CYP2C9*3 allelic variant was associated with a significant high risk of bleeding (OR: 7:3).34
Clinical utility of CYP2C8 and CYP2C9 preemptive genotyping
Currently, data on the association of NSAIDs treatment with severe drug side effects or analgesic failure are limited in SCD patient population.4 Interestingly, recent epidemiologic data associates higher doses of some traditional and nontraditional NSAIDs with double risk of congestive heart failure, increased risk of peptic ulcer complications, gastrointestinal bleeding, and increased risk of major vascular events (nonfatal myocardial infarction, nonfatal stroke, or vascular death.8, 35, 36, 37, 38 Though these risks fall quickly after drug cessation, nonetheless, with chronic NSAID use, these risks do not wane over years of use.37, 38, 39, 40 For SCD patients, however, due to the ubiquity and chronicity of pain experience, NSAIDs therapy is initiated very early in life and throughout the lifespan. Indeed, as Table 5 illustrates, NSAIDs are routinely prescribed to SCD with compromised metabolic profiles. Preemptive genotyping to identify patients with CYP2C8 and CYP2C9 intermediate and poor predicted metabolic phenotypes could potentially facilitate early prediction of NSAID treatment outcomes both in terms of efficacy and possible development of adverse events in SCD patients. Preemptive genotyping anticipates current and future medication prescription needs of patients as opposed to current practice whereby genotyping is performed only when clinically indicated.41 With the preemptive genotyping model, a single blood sample is used to genotype specific patient populations for polymorphisms of pharmacogenetic significance in relevant drug metabolizing enzymes and transporters with regards to specific medication classes.42 CYP2C8 and CYP2C9 preemptive genotyping data could potentially facilitate quantification and clinical assessment of pharmacogenetic risk and longitudinal accumulation of NSAIDs risk burden in SCD patients. As depicted in Table 5, the identification of an impaired drug metabolic profile in a patient would alert the clinician that special attention should be given to NSAIDs drug response of the patient and would provide explanation for those individuals with unsatisfactory drug response or side effects profiles enabling clinicians to make distinctions between a medication compliance problem and a metabolic defect. More importantly, since the genetic makeup of an individual is virtually invariable, the determination of SCD patients’ drug metabolic genotypes provides lifelong applicable information for analgesic pharmacotherapy.40
Ultimately, pharmacogenetic screening for SCD patients’ CYP2C8 and CYP2C9 genotypes would be useful only if it will facilitate development of NSAIDs dosing algorithms similar to those developed for warfarin and tamoxifen dosing with the CYP2C9/VKORC1 and CYP2D6 genes, respectively.43, 44, 45 However, this would require additional functionality studies of novel CYP2C8 and CYP2C9 variant alleles, as well as the effect of SCD on CYP2C8 and CYP2C9 enzyme expressions and the role of environment factors. Perhaps more immediate is the need for appropriate PK and PD studies to determine the effects of allelic variants present in the SCD population. As shown in Tables 2 and 3, the CYP2C8*2 and CYP2C9*5,*6,*8, *9, and *11 variant alleles are common in populations with genetic susceptibility for SCD, mostly populations of African ancestry. However, while the relationships between the CYP2C8*3, CYP2C9*2, and *3 alleles and metabolic indexes of several NSAIDs are well delineated, to the best of our knowledge, the pharmacokinetic effects of the CYP2C8*2 and CYP2C9*5,*6,*8, *9, and *11 alleles on NSAIDs metabolism has not been evaluated in African American populations. This lack of pharmacokinetics data not only precluded us from assigning predicted metabolic phenotype for four genotypes (i.e., *5/*9, *6/*8, *8/*9, *9/*11), but also represents a crucial knowledge gap in the use of genetic genotype to inform pharmacogenetic prescribing practice for NSAIDs dosing or for other drugs metabolized by the CYP2C8 (i.e., amiodarone, fluvastatin, simvastatin, verapamil, montelukast, amodiaquine and chloroquine, morphine and methadone) and the CYP2C9 (i.e., losartan, tolbutamide and torsemide) enzymes often prescribed for SCD patients in various parts of the world.
Our study has limitations. Though we used a multiplex genotyping panel to determine CYP2C8 and CYP2C9 genotypes, the assignment of predicted metabolic profiles in our cohort was based on allelic combinations and associated activity levels reported in the literature. Ultimately, pharmacokinetics study remains the gold standard for discerning SCD patients’ metabolic phenotypes and for reporting individuals’ analgesic drug response profile for specific substrates. Another study limitation is that with the iPLEX ADME PGx panel, the CYP2C9*3 and *18 are indistinguishable haplotypes. Consequently, we arbitrarily reported the CYP2C9*3 allele frequency in our cohort. Additionally, the cross‐sectional nature of our study precluded reporting of daily or annual NSAIDs usage by genotypes, pain scores and analgesic response which we reserve for another project. These limitations notwithstanding, to the best of our knowledge, no study attempts to bridge the concept of pharmacogenetic variability as a determinant of interindividual response to NSAID therapy in SCD patients.
Conclusions
In summary, our study determined allele frequencies and genotypes in the CYP2C8 and CYP2C9 enzymes in a SCD patient cohort. Ten of 12 of the combined alleles of the CYP2C8 and CYP2C9 enzymes identified in our cohort contribute to deficient metabolic genotypes with fifty‐two subjects having at least one variant CYP2C9 allele associated with either the IM, poor, or indeterminate phenotype; while 55 subjects had at least one CYP2C8 variant allele that contributes to impaired metabolic genotypes. Several of the impaired function variant alleles are being reported for the first time among SCD patients. The CYP2C8 and CYP2C9 variant alleles play a significant role in the analgesic effects and toxicity of NSAIDs. These drugs used to treat VOC pain at the prodromal stage in SCD patients are associated with vascular risks, adverse effects on the gastrointestinal tract, and possibly frequent ED visits for analgesic failure.7, 8 Our study highlights the concept of pharmacogenetic variability as a determinant of interindividual response to NSAIDs therapy in SCD patients. CYP2C8 and CYP2C9 preemptive genotyping could potentially enable clinicians to identify patients with impaired metabolic phenotypes. In tandem with preemptive genotyping, additional and appropriate pharmacokinetics and pharmacodynamics studies are required to potentially enable clinicians to tailor NSAIDs dosing accordingly to achieve optimal analgesic response.
Conflict of Interest
The authors have no financial disclosures or conflicts of interest to declare.
Acknowledgment
The authors wish to thank the nursing staff at the Georgia Regents University Adult Sickle Cell Clinics. This work was accepted for presentation in abstract form at the 33rd Annual Scientific Meeting of the American Pain Society, April 30‐May 3, 2014, at the Tampa Convention Center, Tampa, Florida. This study was supported in part by a grant from the National Institute of Nursing Research, NIH, grant # 5K01NR012465 and the National Institute on Minority Health and Health Disparities, NIH, grant # 5P20MD003383. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute for Nursing Research or the National Institute on Minority Health and Health Disparities.
References
- 1. Field JJ, Knight‐Perry JE, DeBaum MR. Acute pain in children and adult with sickle cell disease: management in the absence of guidelines. Curr Opin Hematol 2009; 16: 173–178. [DOI] [PubMed] [Google Scholar]
- 2. Ballas SK, Gupta K, Adams‐Graves P. Sickle cell pain: a critical reappraisal. Blood 2012; 120: 3647–3656. [DOI] [PubMed] [Google Scholar]
- 3. Niscola P, Sorrentino F, Scaramucci L, de Fabritiis P, Cianciulli P. Pain syndromes in sickle cell disease: an update. Pain Med 2009; 10: 470–480. [DOI] [PubMed] [Google Scholar]
- 4. Jerrell JM, Tripathi A, Stallworth JR. Pain management in children and adolescents with sickle cell disease. Am J Hematol 2011; 86: 82–84. [DOI] [PubMed] [Google Scholar]
- 5. Wang B, Wang J, Huang SQ, Su HH, Zhou SF. Genetic polymorphism of the human cytochrome P450 2C9 gene and its clinical significance. Curr Drug Metab 2009; 10: 781–834. [DOI] [PubMed] [Google Scholar]
- 6. Agúndez JA, García‐Martín E, Martínez C. Genetically based impairment in CYP2C8‐ and CYP2C9‐dependent NSAID metabolism as a risk factor for gastrointestinal bleeding: is a combination of pharmacogenomics and metabolomics required to improve personalized medicine? Expert Opin Drug Metab Toxicol 2009; 5: 607–620 [DOI] [PubMed] [Google Scholar]
- 7. Bjarnason I. Gastrointestinal safety of NSAIDs and over‐the‐counter analgesics. Int J Clin Pract Suppl 2013; 178: 37–42. [DOI] [PubMed] [Google Scholar]
- 8. Coxib and traditional NSAID Trialists? (CNT) Collaboration , Bhala N, Emberson J, Merhi A, Abramson S, Arber N, Baron JA, Bombardier C, Cannon C, Farkouh ME, FitzGerald GA et al. Vascular and upper gastrointestinal effects of non‐steroidal anti‐inflammatory drugs: meta‐analyses of individual participant data from randomised trials. Lancet 2013; 382: 769–779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Yee MM, Josephson C, Hill CE, Harrington, R . Castillejo, MI, Ramjit, Osunkwo I. Cytochrome P450 2D6 polymorphism and predicted metabolism in African American children with sickle cell disease. J Pediatr Hematol Oncol 2013; 7: e301–e305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Brousseau DC, McCarver DG, Drendel, AL , Divakaran K, Panepinto JA. The effect of CYP2D6 polymorphisms on the response to pain treatment for pediatric sickle cell pain crisis. J Pediatr 2007; 150: 623–626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Shord SS, Cavallari LH, Gao W, Jeong HY, Deyo K, Patel SR, Camp JR, Labott SM, Molokie RE. The pharmacokinetics of codeine and its metabolites in blacks with sickle cell disease. Eur J Clin Pharmacol 2009; 65: 651–658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Joly P, Gagnieu MC, Bardel C, Francina A, Pondarre C, Martin C. Genotypic screening of the main opiate‐related polymorphisms in a cohort of 139 sickle cell disease patients. Am J Hematol 2012; 87: 534–536. [DOI] [PubMed] [Google Scholar]
- 13. Shetkar SS, Ramakrishnan S, Seth S, Chandna P, Verma SK, Bhargava B, Bahl VK. CYP 450 2C19 polymorphisms in Indian patients with coronary artery disease. Indian Heart J. 2014; 66: 16–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Sim SC, Ingelman‐Sundberg M. The human cytochrome P450 Allele Nomenclature Committee Web site: submission criteria, procedures, and objectives. Methods Mol Biol 2006; 320: 183–191. [DOI] [PubMed] [Google Scholar]
- 15. Xiaoping L, Zhong F, Tan X. Cytochrome P450 2C8 and drug metabolism. Curr Top Med Chem 2013; 13: 2241–2253. [DOI] [PubMed] [Google Scholar]
- 16. Daily EB, Aquilante CL. Cytochrome P450 2C8 pharmacogenetics: a review of clinical studies. Pharmacogenomics 2009; 10: 1489–1510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Aquilante CL, Niemi M, Gong L, Altman RB, Klein TE. PharmGKB summary: very important pharmacogene information for cytochrome P450, family 2, subfamily C, polypeptide 8. Pharmacogenet Genomics 2013; 23: 721–728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Dai D, Zeldin DC, Blaisdell JA, Chanas B, Coulter SJ, Ghanayem BI, Goldstein JA. Polymorphisms in human CYP2C8 decrease metabolism of the anticancer drug paclitaxel and arachidonic acid. Pharmacogenetics 2001; 11: 597–607. [DOI] [PubMed] [Google Scholar]
- 19. Scott SA, Jaremko M, Lubitz SA, Kornreich R, Halperin JL, Desnick RJ. CYP2C9*8 is prevalent among African Americans: implications for pharmacogenetic dosing. Pharmacogenomics 2009; 10: 1243–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Man N, Farmen M, Dumaual C, Teng CH, Moser B, Irie S, Noh GJ, Njau R, Close S, Wise S, Hockett R. Genetic variation in metabolizing enzymes and transporter genes: comprehensive assessment in 3 major East Asian subpopulations with comparison to Caucasians and African. J Clin Pharmacol 2010; 50: 929–940. [DOI] [PubMed] [Google Scholar]
- 21. Kudzi W, Dodoo A, Mills JJ. Characterisation of CYP2C8, CYP2C9 and CYP2C19 in a Ghanaian population. BMC Medical Genetics 2009; 10: 124. doi:10.1186/1471-2350-10-124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Allabi AC, Gala JL, Desager JP, Heusterspreute M, Horsmans Y. Genetic polymorphisms of CYP2C9 and CYP2C19 in the Beninese and Belgian populations. Br J Clin Pharmacol 2003; 56: 653–657. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Vargens DD, Damasceno A, Petzl‐Erler M, Suarez‐Kurtz G. Combined CYP2C9, VKORC1 and CYP4F2 frequencies among Amerindians, Mozambicans and Brazilians. Pharmacogenomics 2011; 12: 769–772. [DOI] [PubMed] [Google Scholar]
- 24. Martis S, Peter I, Hulot J, Kornreich R, Desnick RJ, Scott SA. Multi‐ethnic distribution of clinically relevant CYP2C genotypes and haplotypes. Pharmacogenomics J 2012; 13: 369–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Arnaldo P, Thompson RE, Lopes MQ, Suffys PN, Santos AR. Frequencies of cytochrome P450 2B6 and 2C8 allelic variants in the Mozambican population. Malays J Med Sci 2013; 20: 12–23. [PMC free article] [PubMed] [Google Scholar]
- 26. Martínez C, García‐Martín E, Blanco G, Gamito FJG, Ladero JM, Agúndez JAG. The effect of the cytochrome P450 CYP2C8 polymorphism on the disposition of (R)‐ibuprofen enantiomer in healthy subjects. Br J Clin Pharmacol 2005; 59: 62–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yasar U, Eliasson E, Forslund‐Bergengren C, Tybring G, Gadd M, Sjögvist F, Dahl ML. The role of CYP2C9 genotype in the metabolism of diclofenac in vivo and in vitro. Eur J Clin Pharmacol 2001; 57: 729–735. [DOI] [PubMed] [Google Scholar]
- 28. García‐Martín E, Martínez C, Tabarés B, Frías J, Agúndez JA. Interindividual variability in ibuprofen pharmacokinetics is related to interaction of cytochrome P450 2C8 and 2C9 amino acid polymorphisms. Clin Pharmacol Ther 2004; 76: 119–127. [DOI] [PubMed] [Google Scholar]
- 29. Bae JW, Kim JH, Choi CI, Kim MJ, Kim HJ, Byun SA, Chang YS, Jang CG, Park YS, Lee SY. Effect of CYP2C9*3 allele on the pharmacokinetics of naproxen in Korean subjects. Arch Pharm Res 2009; 32: 269–273. [DOI] [PubMed] [Google Scholar]
- 30. Vianna‐Jorge R, Perini JA, Rondinelli E, Suarez‐Kurtz G. CYP2C9 genotypes and the pharmacokinetics of tenoxicam in Brazilians. Clin Pharmacol Ther 2004; 76: 18–26. [DOI] [PubMed] [Google Scholar]
- 31. Bae JW, Choi CI, Jang CG, Lee SY. Effects of CYP2C9*1/*13 on the pharmacokinetics and pharmacodynamics of meloxicam. Br J Clin Pharmacol 2011; 71: 550–555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Martínez C, Blanco G, Ladero JM, García‐Martin E, Toxonera C, Gamito, FG , Diaz‐Rubio M, Agúndez JA. Genetic predisposition to acute gastrointestinal bleeding after NSAIDs use. Br J Pharmacol 2004; 141: 205–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Blanco G, Martínez C, Ladero JM, García‐Martin E, Toxonera C, Gamito, FG Diaz‐Rubio M, Agundez JA. Interaction of CYP2C8 and CYP2C9 genotypes modifies the risk for nonsteroidal anti‐inflammatory drugs‐related acute gastrointestinal bleeding. Pharmacogenet Genomics 2008; 18: 37–43. [DOI] [PubMed] [Google Scholar]
- 34. Pilotto A, Seripa D, Franceschi M, Scarcelli C, Colaizzo D, Grandone E, Niro V, Andriulli A, Leandro G, Di Mario F, Dallapiccola B. Genetic susceptibility to nonsteroidal anti‐inflammatory drug‐related gastroduodenal bleeding: role of cytochrome P450 2C9 polymorphisms. Gastroenterology 2007; 133: 465–471. [DOI] [PubMed] [Google Scholar]
- 35. He YJ, McLeod HL. Ready when you are: easing into preemptive pharmacogenetics. Clin Pharmacol Ther 2012; 92: 412–414. [DOI] [PubMed] [Google Scholar]
- 36. Schildcrout JS, Denny JC, Bowton E, Gregg W, Pulley JM, Basford MA, Cowan JD, Xu H, Ramirez AH, Crawford DC, Ritchie MD, Peterson JF, Masys DR, Wilke RA, Roden DM. Optimizing drug outcomes through pharmacogenetics: a case for preemptive genotyping. Clin Pharmacol Ther 2012; 92: 235–242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Ray WA, Stein CM, Hall K, Daugherty JR, Griffin MR. Non‐steroidal anti‐inflammatory drugs and risk of serious coronary heart disease: an observational cohort study. Lancet 2002; 359: 118–123. [DOI] [PubMed] [Google Scholar]
- 38. McGettigan P, Henry D. Cardiovascular risk with non‐steroidal anti‐inflammatory drugs: systematic review of population‐based controlled observational studies. PLoS Med 2011; 8: e1001098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Garcia Rodriguez LA, Tacconelli S, Patrignani P. Role of dose potency in the prediction of risk of myocardial infarction associated with nonsteroidal anti‐ inflammatory drugs in the general population. J Am Coll Cardiol 2008; 52: 1628–1636. [DOI] [PubMed] [Google Scholar]
- 40. Hernandez‐Diaz S, Rodriguez LA. Association between nonsteroidal anti‐inflammatory drugs and upper gastrointestinal tract bleeding/perforation: an overview of epidemiologic studies published in the 1990s. Arch Intern Med 2000; 160: 2093–2099. [DOI] [PubMed] [Google Scholar]
- 41. Richy F, Bruyere O, Ethgen O, Rabenda V, Bouvenot G, Audran M, Herrero‐Beaumont G, Moore A, Eliakim R, Haim M, Reginster JY. Time dependent risk of gastrointestinal complications induced by non‐steroidal antiinflammatory drug use: a consensus statement using a metaanalytic approach. Ann Rheum Dis 2004; 63: 759–766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Loovers HM, Van der Weide J. Implementation of CYP2D6 genotyping in psychiatry. Expert Opin Drug Metab 2009; 5: 1–13. [DOI] [PubMed] [Google Scholar]
- 43. Pirmohamed M, Burnside G, Eriksson N, Jorgensen AL, Toh CH, Nicholson T, Kesteven P, Christersson C, Wahlström B, Stafberg C et al. A randomized trial of genotype‐guided dosing of warfarin. N Engl J Med. 2013; 369: 2294–2303. [DOI] [PubMed] [Google Scholar]
- 44. Grossi E, Podda GM, Pugliano M, Gabba S, Verri A, Carpani G, Buscema M, Casazza G, Cattaneo M. Prediction of optimal warfarin maintenance dose using advanced artificial neural networks. Pharmacogenomics 2014; 15: 29–37. [DOI] [PubMed] [Google Scholar]
- 45. Ruddy KJ, Desantis SD, Gelman RS, Wu AH, Punglia RS, Mayer EL, Tolaney SM, Winer EP, Partridge AH, Burstein HJ. Personalized medicine in breast cancer: tamoxifen, endoxifen, and CYP2D6 in clinical practice. Breast Cancer Res Treat 2013; 141: 421–427. [DOI] [PubMed] [Google Scholar]