The CYP2C19*2 and CYP2C19*17 Polymorphisms Influence Responses to Clozapine for the Treatment of Schizophrenia
Received 22 August 2019
Accepted for publication 21 December 2019
Published 11 February 2020 Volume 2020:16 Pages 427—432
Checked for plagiarism Yes
Review by Single anonymous peer review
Peer reviewer comments 3
Editor who approved publication: Dr Roger Pinder
Christielly Rodrigues-Silva,1 Agostinho Tavares Semedo,1 Hiasmin Franciely da Silva Neri,1 Rosana Pereira Vianello,2 Carlos Galaviz-Hernández,3 Martha Sosa-Macías,3 Rodrigo Bernini de Brito,1,4 Paulo César Ghedini1
1Department of Pharmacology, Institute of Biological Sciences, Federal University of Goiás, Goiânia, GO, Brazil; 2Empresa Brasileira de Pesquisa Agropecuária - Embrapa, Santo Antônio de Goiás, GO, Brazil; 3Instituto Politécnico Nacional, Academia de Genómica, CIIDIR-Durango, Durango, México; 4Brain Institute Medical Clinic, Bueno Medical Center Building, Goiânia, GO, Brazil
Correspondence: Paulo César Ghedini
Laboratório de Farmacologia Bioquímica e Molecular, Instituto de Ciências Biológicas, UFG, Cep 74690-900, Goiânia, GO, Brazil
Tel +55 62 3521-1725
Email [email protected]
Introduction: Clozapine (CLZ) is the gold standard drug for treatment-refractory schizophrenia (TRS). However, approximately 30% of patients partially respond to CLZ, defining this subset with super refractory schizophrenia (SRS). Alterations in enzyme activity may affect CLZ responses; the CYP3A4, CYP1A2 and CYP2C19 genes are primarily responsible for CLZ metabolism.
Objective: The aim of this study was to assess if CYP2C19 variants were associated with TRS or SRS.
Methods: CYP2C19*2 loss-of-function and CYP2C19*17 gain-of-function polymorphism genotype testing were performed in 108 individuals undergoing pharmacological treatment for TRS or SRS. DNA was extracted and polymorphisms were analyzed by polymerase chain reaction (PCR) and sequencing.
Results: CYP2C19*17 had positive correlations with SRS and lower Brief Psychiatric Rating Scale (BPRS) scores for TRS. In addition, CYP2C19*2 was associated with lower CLZ dosages for TRS.
Conclusion: These results show that CYP2C19*2 and CYP2C19*17 polymorphisms influence CLZ responses during schizophrenia treatment.
Keywords: schizophrenia, CYP2C19*2, CYP2C19*17, treatment response, clozapine
Schizophrenia is a chronic and severe psychiatric disorder that exhibits variability in response to many antipsychotic drugs. Although pharmacotherapy can be used to treat this disorder, approximately one-third to one half of patients are unresponsive to current treatments, and are therefore deemed to exhibit treatment-refractory-schizophrenia (TRS).1,2
Refractoriness is defined as the lack of satisfactory clinical response to treatments with at least two antipsychotic drugs, of the first or second generation, given at therapeutic doses and for at least six weeks during treatment.3 When refractoriness occurs, CLZ is the drug of choice for treatment.4–6 Nevertheless, approximately 30% of TRS patients do not fully respond to CLZ as a monotherapy, meaning these patients are as known as CLZ non-responders or super-refractory schizophrenics (SRS).7–9 For this condition, treatments are combined with CLZ and other antipsychotics, antidepressants or mood stabilizers.10,11
Factors such as ethnicity, co-medication, age, gender, diet, as well as genetic variability to cellular receptors and drug metabolism, have been shown to affect CLZ responses.12–14 With regards to TRS and SRS genetic components, variability in cytochrome P450 (CYP) enzyme activities have been described as significant in influencing CLZ bioavailability.15–17 CYP3A4 and CYP1A2 are primarily responsible for the formation of N-desmethylclozapine,18,19 with CYP2C19 playing a role and CYP2C9 and 2D6 playing more modest roles.20 In light of these functions, our group previously observed CYP1A2*1F polymorphism associations with SRS,9 thereby shedding some light on the relevance of CYP-genetic polymorphisms in CLZ responses. Although involvement of the CYP1A2*1F polymorphism in response to CLZ treatment has been shown, there are little data evaluating associations between CYP2C19 polymorphisms and TRS or SRS.
Thirty-five polymorphic variants have been identified in the CYP2C19 gene (https://www.pharmgkb.org/page/cyp2c19), of which three are clinically relevant (CYP2C19*2, *3 and *17); CYP2C19*3 is significantly frequent in Asian populations.16 CYP2C19*2 (c.681G>A, rs4244285) is an allelic variant that encodes for a non-functional protein21–23 and CYP2C19*17 (−806C>T, rs12248560) affects promoter responsiveness, increasing CYP2C19 expression.13,23–25 Phenotypically, CYP2C19*2 and CYP2C19*17 variants are associated with poor metabolizers (PM) and ultra-rapid metabolizers (UM), respectively, while extensive metabolizers (EM) are homozygous for the wild-type allele, *1.25 Subjects with *2/*2 genotype are PM, those heterozygous for *1/*2 or *2/*17 are intermediate metabolizers (IM), and *1/*17 and *17/*17 are UM.21,23,25
We recently identified a relationship between the CYP1A2*1F polymorphism and CLZ therapeutic outcomes;9 however, there is a dearth of research investigating associations of CYP2C19 polymorphisms with refractoriness to CLZ responses. Therefore, this study sought to evaluate pharmacogenetic associations of CYP2C19*2 and CYP2C19*17 polymorphisms with TRS and SRS.
Materials and Methods
One hundred and eight schizophrenia patients (108) from Goiás state, Brazil, were included. Inpatients and outpatients were recruited from the Brain Institute – Bueno Medical Centre or the Distribution Centre of High-Cost Drugs of the Secretary of Health. For allele and genotype frequency comparisons, 137 healthy individuals (control group; both sexes, 28 ± 11 years old) were also included in the study. All selected individuals were classified as pardos, according to classifications used by the Brazilian Institute of Geography and Statistics (IBGE).26 Pardos consider themselves as a mixture of native Brazilian, European, West African and/or South Asian.
Schizophrenia diagnosis was defined according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-V). Psychopathological and clinical data were acquired by interviews with patients or accessing their medical records. All participants provided informed written consent, and all study protocol were approved by the research ethics committee from the Federal University of Goiás and Goiás State Secretary of Health (protocols 1,483734 and 1,537538, respectively), in accordance with the World Medical Association Declaration of Helsinki.
Only individuals receiving CLZ for at least six months were included in the study. Patients were classified as TRS (n = 63) or SRS (n = 45) following criteria described by Kane et al (1988)4 and de Brito et al (2015).9 Characteristics of the study population are shown (Table 1).
Table 1 Characteristics of Patients
For the 45 SRS patients for whom another drug was added, 52% received a second-generation antipsychotic, with risperidone being the most frequent (30%), followed by quetiapine (11%), olanzapine (3.7%), aripiprazole (3.7%), and ziprasidone (3.7%). 48% received a first-generation agent. Five patients received lamotrigine in combination with CLZ and other antipsychotics. None of the CLZ-associated drugs were metabolized primarily by 2C19 enzyme.
Genomic DNA was obtained from whole-blood using the PureLink™ Genomic DNA Mini Kit (Invitrogen™, Carlsbad, CA, USA). We then used PCR to amplify the regions of interest. Reactions consisted of 50 µL final volume, containing PCR buffer 10X, 2 mM MgCl2, 0.1 mM dNTPs, 100 ng genomic DNA, 1 U Taq DNA polymerase (Invitrogen™) and 0.5 µM of each oligonucleotide. CYP2C19*2 was amplified using the following primers; Forward; 5ʹ-CAACCAGAGCTTGGCATATTGTATC–3ʹ and Reverse; 5ʹ–GCCCTTAGCACCAAATTCTC–3ʹ; and CYP2C19*17 was amplified using the following primers; Forward; 5ʹ–TAAAGTCCCGAGGGTTGTTG–3ʹ and Reverse 5ʹ–ATTTAACCCCCTAAAAAAAACG–3ʹ. Cycling conditions for the *2 allele were 40 cycles at 95ºC/45 s; 56ºC/30 s; and 72ºC/30 s. Cycling conditions for the *17 polymorphism were 35 cycles at 94ºC/30 s; 52ºC/30 s; and 72ºC/30 s. PCR products were electrophoresed in a 1% agarose gel and excised fragments were purified using the GFX™ PCR DNA and Gel Band Purification Kit (GE Healthcare, Chicago, Illinois, USA), followed by sequencing on an ABI3500 Genetic Analyzer (Applied Biosystems, Foster City, California, USA) using BigDye Terminator Mix v. 3.1 chemistry (Applied Biosystems).
Statistical analyses were performed using GraphPad Prism (version 6.0, GraphPad Prism Software Inc., San Diego, CA, USA), and P < 0.05 was the threshold for statistical significance. Allelic and genotypic frequencies, sex, smoking and coffee status were assessed using the chi-square or Fisher´s exact tests. Age, BMI (body mass index), CLZ dosage and BPRS scores were analyzed using “t” tests or ANOVA for two or more groups, respectively. Genotype frequencies were obtained by direct count, and Hardy–Weinberg Equilibrium (HWE) was calculated using the χ2 goodness-of–fit statistic. Differences in allele and genotype frequencies were evaluated using the χ2 test (and Fisher’s exact). Haplotype frequency estimations (analysis of multiple genotype associations of A (CYP2C19*2) and T (CYP2C19*17) polymorphisms) and associations between polymorphisms and TRS, SRS or schizophrenia risk were evaluated using multivariate logistic regression analyses, using three models (co-dominant, dominant and recessive) on SNPStats software.27 Additionally, we assessed if BPRS or different oral CLZ dosages were independently associated with each single nucleotide polymorphism (SNP) (*2 or *17) in TRS or SRS, adjusting for potential confounding effects and dichotomising patients using observed global median of BPRS scores (<38 or ≥38) and CLZ doses (<400 or ≥400 mg/kg) as reference. Odds ratios (ORs) and 95% confidence intervals (CIs) were also calculated.
Demographic data showed significantly higher BPRS scores and CLZ dosages in the SRS group when compared with the TRS group. Factors such as age, gender, smoking, and coffee drinking were not significantly different between SRS and TRS groups (Table 1). The genotype distribution in the study sample did not deviate from HWE (P = 0.54 for TRS, and P= 0.06 for SRS). The distributions of alleles and genotypes showed a significantly increased frequency of CYP2C19*17 in the SRS group when compared with the control (P = 0.007 and P = 0.0002, respectively) and TRS (P = 0.028 and P = 0.003, respectively) groups (Table 2). The CYP2C19*2 allele showed a similar frequency between SRS, TRS and control groups (Table 3). Genotype frequencies showed that the heterozygous genotype, CYP2C19*1/*17 was significantly associated with SRS, when compared with the control (OR = 4.24, 95% CI, 2.08–8.65; P < 0.0001) and the TRS group (OR = 4.20, 95% CI, 1.86–9.47; P < 0.00001). Haplotype estimations revealed a significant association of GT with SRS (OR = 2.85, 95% CI, 1.32–6.18; P < 0.0091). The presence of *2 and *17 polymorphisms in the same individual (CYP2C19*2/*17) meant these individuals were classified as intermediate metabolizers.21,23,25 Frequencies of CYP2C19*2/*17 were similar for the TRS (5%), SRS (9%) and control group (6%) (P > 0.05).
Table 2 Allelic and Genotype Frequencies of CYP2C19*17 in Control (Healthy Individuals), TRS and SRS Individuals. Statistical Analysis Values (χ2, P-value), Found for Frequencies Comparisons Between Three Groups, are Presented Too
Table 3 Allelic and Genotype Frequencies of CYP2C19*2 in Control (Healthy Individuals), TRS and SRS Individuals. Statistical Analysis Values (χ2, P-value), Found for Frequencies Comparisons Between Three Groups, are Presented Too
We asked if each SNP was associated with BPRS scores and oral CLZ doses in TRS or SRS independently of each other, and found that the lowest oral CLZ dose (<400 mg/day) in the TRS group was associated with the CYP2C19*2 allele (OR = 0.29, 95% CI, 0.09–0.90; P < 0.031). However, the CYP2C19*17 allele showed a significant association with lower BPRS scores (<38) in the TRS group (OR = 0.13, 95% CI, 0.03–0.61; P < 0.0025). In the SRS group, oral CLZ doses did not show associations with SNPs; however, the lowest BPRS scores (<38) revealed a significant association with the CYP2C19*2 allele (OR = 0.12, 95% CI, 0.02–0.86; P < 0.034).
Pharmacogenetic studies investigate genetic sources of inter-individual variation that can impact on clinical outcomes.1,25,27 Previous reports have highlighted the clinic influence of CYP2C19 polymorphisms in therapeutic responses to proton pump inhibitors, anti-platelet and antidepressant drugs, as well as associations with breast cancer risk.28–32 In terms of antipsychotic responses, specifically for CLZ treatments, a previous study found no differences in the distribution of the CYP2C19*17 polymorphism between TRS and SRS patients and healthy controls.33 Interestingly, a more recent study suggested that CYP2C19*17 is protective against the development of diabetes during CLZ treatment, and increases the likelihood of improving schizophrenia.34 Our findings revealed an association between CYP2C19*17 and SRS, suggesting this allele is related to therapeutic response impairments for CLZ treatment.
A previous study found that 74% of SRS patients possessed a gain-of-function polymorphism from CYP1A2 (CYP1A2*1F).9 Similar to CYP2C19*17, CYP1A2*1F is a variant that encodes increased enzymatic activity.21,35 In this work, we detected the CYP2C19*17 polymorphism in 64% of patients from our SRS subgroup, of which 55.5% had the UM genotype (*1/*17). These two enzymes contribute more than 50% to in vitro CLZ biotransformation20 which may explain the findings of major frequencies of the two polymorphic variants in patients with incomplete response to CLZ – SRS.
As expected, CLZ doses were higher in SRS patients, when compared with TRS patients. Interestingly, those TRS patients carrying the CYP2C19*2 allele had lower average CLZ doses, suggesting that these doses, in carriers of this polymorphism, are adequate to maintain tissue levels for an effective therapeutic response. In contrast to our data, Piatkov et al (2017)34 observed that carriers of CYP2C19*17 required lower average CLZ doses, which was associated with better clinical responses. However, these authors did not evaluate the CYP2C19*2 allele, nor the association of CYP2C19*17 with SRS. On the other hand, we observed that TRS patients carrying the CYP2C19*17 variant generated lower BPRS scores which may be associated with better clinical responses, corroborating elements from Piatkov et al (2017).34
Van de Bilt et al (2015)33 evaluated if refractory patients would have ended up in this condition because they would be UM and the hypothesis was not confirmed, because the CYP2C19*17 was equally distributed between refractory and non-refractory patients. Our study showed that this polymorphism was more frequent in SRS patients; however, this observation is not related to disease risk, since CYP2C19*17 is equally distributed between refractory patients and healthy controls. Finally, we observed a CYP2C19*2 allele association with lower BPRS scores in the SRS group, suggesting protective effects in these patients. This interesting finding will require more studies to delineate this association.
Our study had some limitations; we did not have access to CLZ and norclozapine metabolite plasma levels measures, and we did not determine CYP2C19 expression or levels in patient samples.
Our results indicate that the CYP2C19*17 allele is associated with SRS, whereas the CYP2C19*2 allele is associated with better clinical responses to CLZ in TRS patients. These findings should be validated and replicated in larger scale studies assessing CLZ plasma levels measurements, before any clinical implications should be considered.
The authors report no conflicts of interest in this work.
1. Lally J, MacCabe JH. Personalized approaches to pharmacotherapy for schizophrenia. BJ Psych Adv. 2016;22:78–86. doi:10.1192/apt.bp.114.013433
2. Krivoy A, Gaughran F, Weizman A, Breen G, MacCabe JH. Gene polymorphisms potentially related to the pharmacokinetics of clozapine: a systematic review. Int Clin Psychopharmacol. 2016;31(4):179–184. doi:10.1097/YIC.0000000000000065
3. Soroka E, Inga-Ujma Lesiczka IU, Tracz M, Bereza B, Olajossy M. Treatment-refractory and super refractory schizophrenia – a case study. Curr Probl Psychiatry. 2017;18(4):292–299. doi:10.1515/cpp-2017-0022
4. Kane J, Honigfeld G, Singer J, Meltzer H. Clozapine for the treatment-resistant schizophrenic: a double-blind comparison with chlorpromazine. Arch Gen Psychiatry. 1988;45(9):789–796. doi:10.1001/archpsyc.1988.01800330013001
5. Martin AK, Mowry B. Increased rare duplication burden genomewide in patients with treatment-resistant schizophrenia. Psychol Med. 2016;46(3):469–476. doi:10.1017/S0033291715001701
6. Gillespie AL, Samanaite R, Mill J, Egerton A, MacCabe JH. Is treatment-resistant schizophrenia categorically distinct from treatment responsive schizophrenia? A systematic review. BMC Psychiatry. 2017;17(1):12. doi:10.1186/s12888-016-1177-y
7. Remington G, Saha A, Chong SA, Shammi C. Augmentation strategies in clozapine-resistant schizophrenia. CNS Drugs. 2005;19(10):843–872. doi:10.2165/00023210-200519100-00004
8. Chetty M, Murray M. CYP-mediated clozapine interactions: how predictable are they? Curr Drug Metab. 2007;8(4):307–313. doi:10.2174/138920007780655469
9. de Brito RB, de Carvalho Araújo L, Diniz MJA, et al. The CYP1A2-163C > a polymorphism is associated with super-refractory schizophrenia. Schizophr Res. 2015;169(1–3):502–503. doi:10.1016/j.schres.2015.10.018
10. Toto S, Grohmann R, Bleich S, et al. Psychopharmacological treatment of schizophrenia over time in 30 908 inpatients: data from the AMSP study. Int J Neuropsychopharmacol. 2019;22(9):560–573. doi:10.1093/ijnp/pyz037
11. Henna Neto J, Elkis H. Clinical aspects of super-refractory schizophrenia: a 6-month cohort observational study. Braz J Psychiatry. 2007;29(3):228–232. doi:10.1590/S1516-44462007000300007
12. Bersani FS, Capra E, Minichino A, et al. Factors affecting interindividual differences in clozapine response: a review and case report. Hum Psychopharmacol. 2011;26(3):177–187. doi:10.1002/hup.1191
13. Zanger UM, Schwab M. Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther. 2013;138(1):103–141. doi:10.1016/j.pharmthera.2012.12.007
14. Arranz MJ, Gallego C, Salazar J, Arias B. Pharmacogenetic studies of drug response in schizophrenia. Expert Rev Precis Med Drug Dev. 2016;1:79–91. doi:10.1080/23808993.2016.1140554
15. Prior TI, Baker GB. Interactions between the cytochrome P450 system and the second-generation antipsychotics. J Psychiatry Neurosci. 2003;28(2):99–112.
16. Ingelman-Sundberg M, Sim SC, Gomez A, Rodriguez-Antona C. Influence of cytochrome P450 polymorphisms on drug therapies: pharmacogenetic, pharmacoepigenetic and clinical aspects. Pharmacol Ther. 2007;116(3):496–526. doi:10.1016/j.pharmthera.2007.09.004
17. Moore TR, Hill AM, Panguluri SK. Pharmacogenomics in psychiatry: implications for practice. Recent Pat Biotechnol. 2014;8(2):152–159. doi:10.2174/1872208309666140904113615
18. Eiermann B, Engel G, Johansson I, Zanger UM, Bertilsson L. The involvement of CYP1A2 and CYP3A4 in the metabolism of clozapine. Br J Clin Pharmacol. 1997;44(5):439–446. doi:10.1046/j.1365-2125.1997.t01-1-00605.x
19. Fang J, Coutts RT, McKenna KF, Baker GB. Elucidation of individual cytochrome P450 enzymes involved in the metabolism of clozapine. Naunyn Schmiedebergs Arch Pharmacol. 1998;358(5):592–599. doi:10.1007/PL00005298
20. Olesen OV, Linnet K. Contributions of five human cytochrome P450 isoforms to the N-demethylation of clozapine in vitro at low and high concentrations. J Clin Pharmacol. 2001;41(8):823–832. doi:10.1177/00912700122010717
21. Scott SA, Sangkuhl K, Shuldiner AR, et al. PharmGKB summary: very important pharmacogene information for cytochrome P450, family 2, subfamily C, polypeptide 19. Pharmacogenet Genomics. 2012;22(2):159–165. doi:10.1097/FPC.0b013e32834d4962
22. Flaten HK, Kim HS, Campbell J, Hamilton L, Monte AA. CYP2C19 drug-drug and drug-gene interactions in ED patients. Am J Emerg Med. 2016;34(2):245–249. doi:10.1016/j.ajem.2015.10.055
23. Hicks JK, Sangkuhl K, Swen JJ, et al. Clinical Pharmacogenetics Implementation Consortium guideline (CPIC) for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants: 2016 update. Clin Pharmacol Ther. 2017;102:37–44. doi:10.1002/cpt.597
24. Arici M, Özhan G. CYP2C9, CYPC19 and CYP2D6 gene profiles and gene susceptibility to drug response and toxicity in Turkish population. Saudi Pharm J. 2017;25(3):376–380. doi:10.1016/j.jsps.2016.09.003
25. Ragia G, Arvanitidis KI, Tavridou A, Manolopoulos VG. Need for reassessment of reported CYP2C19 allele frequencies in various populations in view of CYP2C19*17 discovery: the case of Greece. Pharmacogenomics. 2009;10(1):43–49. doi:10.2217/146224188.8.131.52
26. Petruccelli JL, Saboia AL. Características Étnico-Raciais Da População: Classificações e Identidades. Número 02. Rio de Janeiro: IBGE; 2013:208.
27. Solé X, Guinó E, Valls J, Iniesta R, Moreno V. SNPStats: a web tool for the analysis of association studies. Bioinformatics. 2006;22(15):1928–1929. doi:10.1093/bioinformatics/btl268
28. Justenhoven C, Hamann U, Pierl CB, et al. CYP2C19*17 is associated with decreased breast cancer risk. Breast Cancer Res Treat. 2009;115(2):391–396. doi:10.1007/s10549-008-0076-4
29. Zabalza M, Subirana I, Sala J, et al. Meta-analyses of the association between cytochrome CYP2C19 loss- and gain-of-function polymorphisms and cardiovascular outcomes in patients with coronary artery disease treated with clopidogrel. Heart. 2012;98(2):100–108. doi:10.1136/hrt.2011.227652
30. McDonough CW, McClure LA, Mitchell BD, et al. CYP2C19 metabolizer status and clopidogrel efficacy in the Secondary Prevention of Small subcortical strokes (SPS3) study. J Am Heart Assoc. 2015;4(6):e001652. doi:10.1161/JAHA.114.001652
31. Ichikawa H, Sugimoto M, Sugimoto K, Andoh A, Furuta T. Rapid metabolizer genotype of CYP2C19 is a risk factor of being refractory to proton pump inhibitor therapy for reflux esophagitis. J Gastroenterol Hepatol. 2016;31(4):716–726. doi:10.1111/jgh.2016.31.issue-4
32. Jukić MM, Haslemo T, Molden E, Ingelman-Sundberg M. Impact of CYP2C19 genotype on escitalopram exposure and therapeutic failure: a retrospective study based on 2087 patients. Am J Psychiatry. 2018;175(5):463–470. doi:10.1176/appi.ajp.2017.17050550
33. van de Bilt MT, Prado CM, Ojopi EP, et al. Cytochrome P450 genotypes are not associated with refractoriness to antipsychotic treatment. Schizophr Res. 2015;168(1–2):587–588. doi:10.1016/j.schres.2015.08.002
34. Piatkov I, Caetano D, Assur Y, et al. CYP2C19*17 protects against metabolic complications of clozapine treatment. World J Biol Psychiatry. 2017;18(7):521–527. doi:10.1080/15622975.2017.1347712
35. Thorn CF, Aklillu E, Klein TE, Altman RB. PharmGKB summary: very important pharmacogene information for CYP1A2. Pharmacogenet Genomics. 2012;22(2):159–165. doi:10.1097/FPC.0b013e32834d4962
This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.Download Article [PDF]