the pharmacogenetics of type 2 diabetes: a systematic …...in 2013, there existed multiple...

The Pharmacogenetics of Type 2Diabetes: A Systematic Review

OBJECTIVE

We performed a systematic review to identify which genetic variants predict re-sponse to diabetes medications.

RESEARCH DESIGN AND METHODS

We performed a search of electronic databases (PubMed, EMBASE, and CochraneDatabase) and a manual search to identify original, longitudinal studies of theeffect of diabetes medications on incident diabetes, HbA1c, fasting glucose, andpostprandial glucose in prediabetes or type 2 diabetes by genetic variation. Twoinvestigators reviewed titles, abstracts, and articles independently. Two investi-gators abstracted data sequentially and evaluated study quality independently.Quality evaluations were based on the Strengthening the Reporting of GeneticAssociation Studies guidelines and Human Genome Epidemiology Networkguidance.

RESULTS

Of 7,279 citations, we included 34 articles (N = 10,407) evaluating metformin (n =14), sulfonylureas (n = 4), repaglinide (n = 8), pioglitazone (n = 3), rosiglitazone (n =4), and acarbose (n = 4). Studies were not standalone randomized controlled trials,and most evaluated patients with diabetes. Significant medication–gene inter-actions for glycemic outcomes included 1) metformin and the SLC22A1, SLC22A2,SLC47A1, PRKAB2, PRKAA2, PRKAA1, and STK11 loci; 2) sulfonylureas and theCYP2C9 and TCF7L2 loci; 3) repaglinide and the KCNJ11, SLC30A8, NEUROD1/BETA2, UCP2, and PAX4 loci; 4) pioglitazone and the PPARG2 and PTPRD loci;5) rosiglitazone and the KCNQ1 and RBP4 loci; and 5) acarbose and the PPARA,HNF4A, LIPC, and PPARGC1A loci. Data were insufficient for meta-analysis.

CONCLUSIONS

We found evidence of pharmacogenetic interactions for metformin, sulfonyl-ureas, repaglinide, thiazolidinediones, and acarbose consistent with their phar-macokinetics and pharmacodynamics. While high-quality controlled studies withprespecified analyses are still lacking, our results bring the promise of personal-ized medicine in diabetes one step closer to fruition.Diabetes Care 2014;37:876–886 | DOI: 10.2337/dc13-1276

1Division of General InternalMedicine, The JohnsHopkins University School of Medicine,Baltimore, MD2Welch Center for Prevention, Epidemiology, andClinical Research, Baltimore, MD3Department of Epidemiology, The JohnsHopkins Bloomberg School of Public Health,Baltimore, MD4Center for Health Care Research and Policy, TheMetroHealth System, Cleveland, OH5Division of General Internal Medicine, TheMetroHealth System/Case Western ReserveUniversity, Cleveland, OH6Department of Biostatistics and Epidemiology,Case Western Reserve University, Cleveland, OH7Department of Health Policy andManagement,The Johns Hopkins Bloomberg School of PublicHealth, Baltimore, MD8University of Maryland School of Medicine,Baltimore, MD

Corresponding author: Nisa M. Maruthur,[email protected].

Received 31 May 2013 and accepted 19November 2013.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1.

© 2014 by the American Diabetes Association.See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.

Nisa M. Maruthur,1,2,3

Matthew O. Gribble,2,3

Wendy L. Bennett,1,2 Shari Bolen,4,5,6

Lisa M. Wilson,7 Poojitha Balakrishnan,3

Anita Sahu,8 Eric Bass,1,7

W.H. Linda Kao,1,2,3 and

Jeanne M. Clark1,2,3

876 Diabetes Care Volume 37, March 2014

SYSTEM

ATICREV

IEW

mailto:[email protected]://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://creativecommons.org/licenses/by-nc-nd/3.0/http://creativecommons.org/licenses/by-nc-nd/3.0/

In 2013, there existed multiplepharmacologic interventions for theprevention and treatment of type 2diabetes (1). However, all evaluations ofknown efficacious interventions revealthat some patients respond totreatment while others do not. Asrecognized by the American DiabetesAssociation in its 2012 statement on themanagement of hyperglycemia, care intype 2 diabetes must become morepatient-centered (2), and theindividualization of diabetes preventionand treatment based on geneticvariation has great potential.

Narrative reviews have commented onthe promise of pharmacogenomics oftype 2 diabetes (3–6), and prominentindividual studies have found statisticallysignificant pharmacogenetic interactionsassociated with diabetes risk andglycemic outcomes (7–11). However,prior reviews have not systematicallyevaluated this literature to inform futureresearch questions, and these reviews donot address the quality issues that affectthe existing literature on diabetespharmacogenetics. The clinical utility ofgenetic variation for tailoring diabetesmedications rests on the identification ofsubstantial and statistically significantpharmacogenetic interactions frominternally valid studies and confirmationof their findings in varied populationsbased on race/ethnicity.

We conducted a systematic review ofobservational and experimental studiesto determine if the effect of diabetesmedications on diabetes incidence,HbA1c, fasting glucose (FG), andpostprandial glucose (PPG) varies byindependent genetic variation inpatients with impaired FG, impairedglucose tolerance, or type 2 diabetes.We hypothesized that 1) geneticvariation associated with drugtransporters, metabolizers, targets, andmechanisms of action would modify theeffect of specific drugs and 2) theexisting evidence would be insufficientto recommend clinical use ofpharmacogenetic interactions becauseof a lack of well-conducted studiesacross diverse populations.

RESEARCH DESIGN AND METHODS

Senior members of the study werediabetes and obesity researchers with

training in clinical epidemiology, clinicaltrials, and systematic reviewmethodology (E.B., W.L.B., J.M.C., S.B.,W.H.L.K., N.M.M., and M.O.G.) andgenetic epidemiology (P.B., W.H.L.K.,N.M.M., and M.O.G.). The team alsoincluded an experienced projectmanager with expertise in the conductof systematic reviews (L.M.W.).

We searched the PubMed, EMBASE,Cochrane electronic databases and alsomanually searched key review articles,key journals’ tables of contents, and thereferences of included articles. Keyjournals were selected based on contentarea and ones that commonly publishedthe included articles. The PubMedsearch and list of key journals areprovided in Supplementary Tables 1 and2. The electronic search included datesof database inception through 13March2013, and themanual search of tables ofcontents included January to March2013. The search was limited to studiespublished in English.

We included original articles on theeffect of Food and Drug Administration(FDA)-approved diabetes medications(Supplementary Table 3) on diabetesincidence, HbA1c, FG, and PPG in adultswith either type 2 diabetes or increaseddiabetes risk because of impaired FG (FG5.55–6.94 mmol/L) or impaired glucosetolerance (2-h postload [75 g] glucose7.77–11.04 mmol/L) by commongenetic variation. We considered anyindependent genetic variation (e.g.,single nucleotide polymorphisms[SNPs], copy number variants) eligibleand excluded variation such ashaplotypes. Eligible study designs were1) controlled studies evaluating theeffect of a drug for one allele/genotypeversus another over time and2) uncontrolled studies evaluating theeffect (change in outcome or incidenceof outcome) of a drug comparing onegenotype/allele to another. Weexcluded studies of less than 24-hduration and did not include results forHbA1c in studies shorter than 3 months.

We excluded case reports, case series,and cross-sectional studies; studies notwritten in English (due to lack ofavailability of resources to interpretthese articles); and studies that includedparticipants on more than one diabetes

medication. We did not contact authorsto obtain additional results fromincluded studies.

Two investigators reviewed each title,abstract, and full-text articleindependently. A citation was advancedto abstract review if a single investigatorincluded it. Abstracts and full-textarticles were reviewed using astandardized and piloted eligibilitycriteria form, and disagreements wereresolved through consensus.

We developed data abstraction formsbased on included abstracts and articles.Data abstraction forms were pilotedextensively and included information onstudy design, study populationcharacteristics, genetic variation understudy, and study results on outcomes ofinterest. Abstraction forms werecompleted using DistillerSR onlinesystematic review managementsoftware. Two investigators abstracteddata sequentially using the finalizedstandardized forms.

We developed quality abstraction formsbased on the Strengthening theReporting of Genetic AssociationStudies guidelines for reporting ofgenetic association studies (12). Inthe absence of guidelines forpharmacogenetic studies, we alsoincorporated recommendations fromthe HuGENet (Human GenomeEpidemiology Network) HuGE ReviewHandbook (13) and priormethodological papers (14). Weconsidered a study to be randomized if itrandomized participants for thepharmacogenetic study and was notsimply based on a prior randomizedstudy. Forms captured elements ofquality control of genotyping, includingmethod of genotyping and genotypingcall rate, and we considered a call rate$95% to be acceptable. We calculatedgenotype call rates when possible. Wealso recorded genotyping concordanceas a genotyping quality metric. Weconsidered selective reporting ofinteractions based on positive resultsand selection bias related to availabilityof genotyping (Supplementary Data).Two investigators evaluated the qualityof each study independently, anddisagreements were resolved throughconsensus.

care.diabetesjournals.org Maruthur and Associates 877

http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org

We performed a qualitative synthesis ofincluded studies’ results. We wereunable to perform quantitativesyntheses with meta-analyses becausetoo few studies contained the sameSNP–drug interactions with commonoutcomes.

Funding sources had no role in thedesign, conduct, analysis, orinterpretation of the study.

RESULTS

Of 7,279 citations, we included 34articles from 21 studies (7,8,10,15–45)comprised of 10,407 subjects (Fig. 1).The included articles used one of threestudy designs: 1) subanalysis of prior

randomized controlled trials (RCTs; n =13); 2) analysis of observational data(n = 8); and 3) nonrandomized,experimental, pre–post design withouta control group (n = 13) (Table 1). Noneof the studies were de novo RCTsspecifically designed to evaluatepharmacogenetic interactions. With theexception of the Stop Non-InsulinDependent Diabetes Mellitus (STOP-NIDDM) trial (20–23) and the DiabetesPrevention Program (DPP) (7,8,17–19,36,37,45), all studies evaluatedpharmacogenetic interactions inpatients with diabetes. In the DPP, arandomized trial ofmetformin, a lifestyleintervention, and placebo for diabetes

prevention, a broad candidate geneapproach (more than 1,590 candidategene loci) was taken to evaluateassociations of SNPs with diabetes andinteractions between genetic variantsand the trial’s interventions (1 to 3.2years of follow-up) (7,8,17–19,36,37,45).Genetics of Diabetes Audit and ResearchTayside (GoDARTS) investigatorsperformed retrospective analyses ofobservational data from patients withdiabetes who had 12 to 18 months offollow-up using both a genome-wide(10) and candidate gene approach(15,16). The ethnic composition of eachstudy is provided in Table 1. No studyevaluated interactions for the other

Figure 1—Selection of included studies.

878 Pharmacogenetics of Type 2 Diabetes Diabetes Care Volume 37, March 2014

included FDA-approved medications ofinterest.

MetforminGenetic interactions with metforminwere reported in 14 articles (Table 2;Supplementary Table 4) (7,8,10,15–19,34–38,45).

Genes encoding the metformintransporters, SLC22A1, SLC22A2, andSLC47A1, were each studied in fourarticles that evaluated differentoutcomes. For the SLC22A1 locus,rs683369 was associated with responseto metformin with respect to diabetesrisk in the DPP over 3.2 years (7), and

three SNPs were not associated withHbA1c in people with diabetes in twoother studies (16,38). For SLC22A2,rs662301 was associated with risk ofdiabetes at 3.2 years in the metforminarm versus placebo in the DPP (7), andtwo SNPs were not associated withresponse to metformin with FG and

Table 1—Description of included studies

First author Parent studyStudydesign

Diabetesstatus Comparator Gene N Ethnicity*

MetforminChoi, 2011 (34) SOPHIE Observational Diabetes None MATE2-K† 253 MultiethnicDong, 2011 (35) NA Experimental Diabetes None SRR 44 ChineseFlorez, 2006 (17) DPP Experimental Prediabetes Placebo TCF7L2 3,548 MultiethnicFlorez, 2008 (19) DPP Experimental Prediabetes Placebo WFS1 3,548 MultiethnicFlorez, 2007 (18) DPP Experimental Prediabetes Placebo KCNJ11 3,547 MultiethnicFlorez, 2012 (36) DPP Experimental Prediabetes Placebo ‡ 2,890 MultiethnicFlorez, 2012 (37) DPP Experimental Prediabetes Placebo ATM 2,984 MultiethnicMoore, 2008 (9) DPP Experimental Prediabetes Placebo § 3,548 MultiethnicMoore, 2009 (8) DPP Experimental Prediabetes Placebo ENPP1 3,548 MultiethnicJablonski, 2010 (7) DPP Experimental Prediabetes Placebo ** 2,294 MultiethnicPearson, 2007 (15) GoDARTS Observational Diabetes None TCF7L2 945 NR

Tkáč, 2013 (38) NA Experimental Diabetes NoneSLC47A1, SLC22A1,

SLC22A2 148 CaucasianZhou, 2009 (16) GoDARTS Observational Diabetes None SLC22A1 1,014 NRZhou, 2011 (10) GoDARTS Observational Diabetes None †† 3,540 NR

SulfonylureasSuzuki, 2006 (25) NA Experimental Diabetes None CYP2C9 134 AsianGloyn, 2001 (24) UKPDS Experimental Diabetes None KCNJ11 364 WhitePearson, 2007 (15) GoDARTS Observational Diabetes None TCF7L2 901 NRBecker, 2008 (26) Rotterdam Study Observational Diabetes None CYP2C9 475 White

RepaglinideGong, 2012 (39) NA Experimental Diabetes None PAX4, NEUROD1/BETA2 43 ChineseHe, 2008 (27) NA‡‡ Experimental Diabetes None KCNJ11 100 Asian§§Huang, 2010 (28) NA Experimental Diabetes None SLC30A8 48 ChineseJiang, 2012 (40) NA Experimental Diabetes None SLC30A8 209 ChineseSheng, 2011 (41) NA Experimental Diabetes None NAMPT 35 ChineseQin, 2010 (29) NA Observational Diabetes None NOS1AP 100 AsianWang, 2012 (44) NA Experimental Diabetes None UCP2 41 ChineseYu, 2011 (30) NA‡‡ Experimental Diabetes None KCNQ1 91 Asian

PioglitazoneBlüher, 2003 (32) NA Experimental Diabetes None PPARG2 131 NRPei, 2013 (42) NA Experimental Diabetes None PPARG2, PTPRD 67 ChineseSaitou, 2010 (33) NA Observational Diabetes Diet ACE 222 Asian**

RosiglitazoneJiang, 2012 (40) NA Experimental Diabetes None SLC30A8 209 ChineseWang, 2008 (31) NA Experimental Diabetes None ABCA1 93 Asian**Yu, 2011 (30) NA‡‡ Experimental Diabetes None KCNQ1 91 AsianZhou, 2011 (43) Unclear Experimental Diabetes None RBP4 42 Chinese

AcarboseAndrulionyte, 2004 (22) STOP-NIDDM Experimental Prediabetes Placebo PPARG2 770 White***Andrulionyte, 2007 (20) STOP-NIDDM Experimental Prediabetes Placebo PPARA 767 White***Andrulionyte, 2006 (21) STOP-NIDDM Experimental Prediabetes Placebo HNF4A 769 White***Zacharova, 2005 (23) STOP-NIDDM Experimental Prediabetes Placebo LIPC 770 98% White

NA, not applicable; NR, not reported; SOPHIE, Study Of Pharmacogenetics In Ethnically diverse populations; UKPDS, UK Prospective Diabetes Study.*If only one ethnicity reported, the prevalence of that ethnicity is 100% in the study population. †This study also evaluated SNPs in OCT1, OCT2, andMATE1. ‡The study evaluated SNPs from the following loci: G6PC2, MTNR1B, GCK, DGKB, GCKR, ADCY5, MADD, CRY2, ADRA2A, FADS1, PROX1,SLC2A2, GLIS3, C2CD4B, IGF1, and IRS1. §The study evaluated SNPs from the following loci: CDKN2A/B, EXT2, CDKAL1, IGF2BP2, HHEX, LOC387761,and SLC30A8. **The study evaluated 1,590 SNPs. ††The study conducted a genome-wide association study and genotyped 705,125 SNPs from theAffymetrix 6.0 microarray. ‡‡He et al. (27) and Yu et al. (30) are based on the same study population. §§Assumed based on location of study (China,Korea, Japan). ***Assumed based on reported results in Zacharova et al. (23) since from same study.


http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org

Table 2—Interaction between metformin and selected SNPs for glycemic outcomes

Putative function Gene SNPs*

SNPs with significant interactions/SNPsstudied for outcome†

Diabetes risk HbA1c FG

Metformin transporters

SLC22A1rs4646281 (16), rs12208357 (16),rs683369 (7), rs622342 (38) 1/1 0/3 NA

SLC22A2rs662301 (7), rs11920090 (36),rs316019 (34,38) 1/1 0/1‡ 0/1

SLC47A1 rs8065082 (7), rs2289669 (34,38) 1/1 0/1§ NA

AMP-activated protein kinasepathway/gluconeogenesis

PRKAA1 rs249429 (7) 1/1 NA NAPRKAB2 rs6690158 (7) 1/1 NA NAPRKAA2 rs9803799 (7) 1/1 NA NAATM rs11212617 (10)**(37)†† 1/1 0/1‡‡ 0/1STK11 rs741765 (7) 1/1 NA NAPPARA rs4253652 (7) 1/1 NA NAPPARGC1A rs10213440 (7) 1/1 NA NAPCK1 rs4810083 (7) 1/1 NA NA

Insulin secretionKCNJ11 rs5219 (E23K) (18), rs7124355 (7) 1/2 NA NAABCC8 rs4148609 (7) 1/1 NA NA

TCF7L2rs12255372 (15,17),rs7903146 (15,17) 0/2 0/2 NA

WFS1rs734312, rs10010131,rs752854 (19) 0/3 NA NA

CDKN2A/B rs10811661 (45)§§ 1/1 NA NAHNF4A rs11086926 (7) 1/1 NA NAHNF1B rs11868513 (7) 1/1 NA NAGLIS3 rs7034200 (36) NA NA 0/1G6PC2 rs573225 (36) NA NA 0/1MADD rs7944584 (36) NA NA 1/1MTNR1B rs10830963 (36) NA NA 0/1ADCY5 rs11708067 (36) NA NA 0/1

Insulin sensitivityADIPOR2 rs758027 (7) 1/1 NA NAENPP1 rs1044498 (8) 1/1 NA NACAPN10 rs3792269 (7) 1/1 NA NAGCK rs2908289 (7), rs917793 (36) 1/1 NA 0/1IRS1 rs4675095 (36) NA NA 0/1IGF1 rs855228 (36) NA NA 0/1GCKR rs780094 (36) NA NA 0/1

Energy metabolismMEF2A rs424892 (7) 1/1 NA NAMEF2D rs6666307 (7) 1/1 NA NACRY2 rs11605924 (36) NA NA 0/1

OtherITLN2 rs6701920 (7) 1/1 NA NAGCG rs6733736 (7) 1/1 NA NAPKLR rs17367421 (7) 1/1 NA NAPPARGC1B rs741579 (7) 1/1 NA NASRR rs391300 (35) NA 0/1 1/1***PROX1 rs340874 (36) NA NA 0/1DGKB rs2191349 (36) NA NA 0/1ADRA2A rs10885122 (36) NA NA 0/1FADS1 rs174550 (36) NA NA 0/1C2CD4B rs11071657 (36) NA NA 1/1

NA, not applicable. *Jablonski et al. (7) explored a total of 1,590 candidate SNPs. Results for 24 loci for which the P for interaction was , 0.05 arepresented. In total, 91 SNPs demonstrated a significant interaction with metformin, and the 24 SNPs reported here represent the loci for these 91SNPs. †Number of SNPs with significant interaction (P , 0.05) out of the total number of SNPs studied at the locus. ‡rs316019 was evaluated forinteraction effect on HbA1c in two studies. §rs2289669was evaluated for interaction effect on HbA1c in two studies. **This studywas a genome-wideassociation study and genotyped 705,125 SNPs using the Affymetrix 6.0 microarray. ††rs11212617 was evaluated in two studies. ‡‡Two studiesevaluated the interaction between rs11212617 and metformin with HbA1c as an outcome. §§This study explored 10 other candidate SNPs for whichthe interaction between the genetic variant and treatment was not significant in the following loci: EXT2, CDKAL1, IGF2BP2, HHEX, LOC387761, andSLC30A8. ***P = 0.048 for 2-h PPG.


HbA1c as outcomes in three otherarticles (34,36,38). For the SLC47A1locus, rs8065082 was associated withresponse to metformin for diabetes risk(7), and rs2289669 was not associatedwith metformin response in patientswith diabetes (HbA1c as the outcome)(34,38).

rs11212617 at the ATM locus predictedresponse to metformin for diabetesrisk at 1 year in the DPP (37) andattainment of HbA1c ,7% (53 mmol/mol) at 18 months in GoDARTS (10).However, results for HbA1c and FG at1 year were not significant in the DPP(37). Neither GoDARTS or the DPP

found a significant interaction forTCF7L2 SNPs for attainment of HbA1c,7% (53 mmol/mol) (15) or diabetesincidence (17).

Statistically significant interactionsbetween metformin and geneticvariants were also reported for genesencoding additional proteins associatedwith AMP-activated protein kinase–dependent inhibition ofgluconeogenesis [PRKAB2, PRKAA2,PRKAA1, STK11 (46,47), PCK1, PPARA(48), and PPARGC1A (7,49)], insulinsecretion [KCNJ11 (7,18), ABCC8,CDKN2A/B (45), HNF4A, and HNF1B (7)];and insulin sensitivity [ADIPOR2, ENPP1

(8), CAPN10, and GCK (7)] (Table 2;Supplementary Table 4).

SulfonylureasFour studies evaluated the interactionbetween sulfonylureas and SNPs (Table3; Supplementary Table 5) (15,24–26).Of two studies evaluating SNPs inCYP2C9 (25,26), the gene encoding theprimary hepatic cytochrome P450enzyme, which metabolizessulfonylureas, one small study found agreater mean change from baseline inHbA1c at 6 months by diplotype ofrs1057910 (25). Notably, the sample sizefor the variant diplotype was very small(n = 2) (25).

Table 3—Interaction between sulfonylureas, repaglinide, thiazolidinediones, and acarbose and selected SNPsfor glycemic outcomes

Gene SNPs

SNPs with significant interactions/SNPs studied for outcome*

Diabetes HbA1c FG PPG†

SulfonylureasTCF7L2 rs79031462, rs12255372 (15) NA 2/2 NA NAKCNJ11 rs5219 (E23 K), rs1800467 (24) NA NA 0/2 NACYP2C9‡ rs1057910 (25,26), rs1799853 (26) NA 1/1 0/2 NA

RepaglinideKCNJ11 rs5219 (27) NA 1/1 0/1 1/1ABCC8 rs1799854 (27) NA 0/1 0/1 0/1KCNQ1 rs2237892, rs2237895, rs2237897 (30) NA 0/3 0/3 0/3SLC30A8 rs13266634 (28,40),§ rs16889462 (28) NA 1/2 1/2 1/2NOS1AP rs10494366 (29) NA 0/1 0/1 0/1NEUROD1/BETA2 A45T (39) NA 0/1 1/1 1/1PAX4 R121 W (39) NA 0/1 0/1 1/1NAMPT 23186C/T (41) NA 0/1 0/1 0/1UCP2 rs659366 (44) NA 1/1 1/1 0/1

PioglitazonePPARG2 rs1801282 (Pro12Ala) (32,42)** NA 0/1 1/1 0/1ACE rs1799752 (33) NA 0/1 NA NAMTHFR rs1801133 (33) NA 0/1 NA NAPTPRD rs17584499 (42) NA 0/1 0/1 1/1

RosiglitazoneABCA1 rs2230806, rs4149313, rs2230808 (31) NA 0/3 0/3 0/3KCNQ1 rs2237892, rs2237895, rs2237897 (30) NA 0/3 0/3 1/3SLC30A8 rs13266634 (40) NA 0/1 0/1 0/1RBP4 rs3758539, rs10882283 (43) NA 1/2 1/2 0/2

AcarbosePPARA rs1800206, rs4253776, rs4253623, rs135547,

rs135542, rs135539, rs4259701, rs8138102,rs4253728, rs11090819, rs4253778 (20) 2/11 NA NA NA

HNF4Ars2425637, rs3818247, rs4810424,

rs2071197, rs736824, rs1885088 (21) 2/6 NA NA NALIPC rs2070895 (23) 1/1 NA NA NAPPARG2 rs1801282 (22) 0/1 NA NA NAPPARGC1A rs8192673 (22) 1/1 NA NA NA

NA, not applicable. *Number of SNPs with significant interaction (P , 0.05) out of the total number of SNPs studied at the locus. †PPG is the 2-hglucose result from oral glucose tolerance test. ‡By convention, for CYP2C9, numerals (e.g., 1, 2, and 3) are used to identify haplotypes rather thanbase or amino acid changes, and the “1” allele is the wild-type or ancestral haplotype (technically, “1A”; The Human Cytochrome P450 AlleleNomenclature Committee; CYP2C9 allele nomenclature; http://www.cypalleles.ki.se/cyp2c9.htm, accessed 5 November 2013). §Two studiesevaluated rs13266634 for an interaction with HbA1c, FG, and PPG, with one of these finding a significant interaction for each outcome. **Two studiesevaluated the Pro12Ala variant for an interaction with HbA1, FG, and HbA1c; a single study found a significant interaction for FG.


http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://www.cypalleles.ki.se/cyp2c9.htmhttp://care.diabetesjournals.org

GoDARTS evaluated the interactionbetween two TCF7L2 SNPs andsulfonylureas and reported asignificant association with responseto medication (15). Another studyevaluating the interaction betweentwo KCNJ11 SNPs and sulfonylureasdid not find any differences in thechange in FG across genotypes at 12months (24).

RepaglinideEight articles reported on geneticinteractions with repaglinide (Table 3;Supplementary Table 6) (27–30,39–41,44). Of two SNPs evaluated in theSLC30A8 gene (28,40), rs13266634 wasassociated with response to repaglinideusing HbA1c, FG, and PPG at 8 weeksas outcomes (28). Similar, butnonsignificant, results were observedwith the other SNP evaluated (D9 = 0.928for rs168889462 and rs13266634) (28).Notably, rs13266634 has been one ofthe most replicated genetic risk variantsin type 2 diabetes (50).

A single study reported a significantlydifferent change in HbA1c and PPG at6 months (and similar, nonsignificantresults for FG) by E23K genotype ofKCNJ11, which encodes the potassiumchannel inhibited by binding ofrepaglinide to its receptor on the b-cell(27). SNPs in NEUROD1/BETA2, PAX4(39), and UPC2 (44) predicted responseto repaglinide for some glycemicoutcomes (Table 3).

PioglitazoneThree studies reported on interactionsbetween pioglitazone and geneticvariation (32,33,42) (Table 3;Supplementary Table 7). The Pro12Alavariant was associated withpioglitazone response in one (42) oftwo studies evaluating this SNP(32,42). A single study reported asignificant effect of PTPRD rs17584499genotype on PPG at 12 weeks but noton HbA1c or FG (42).

RosiglitazoneFour studies reported on response torosiglitazone by genetic variation(Table 3; Supplementary Table 8)(30,31,40,43). Individual studiesreported significant interactionsbetween the KCNQ1 (30) and RBP4 (43)loci and rosiglitazone for some, but notall, glycemic measures.

AcarboseInteractions between acarbose and thePPARA, HNF4A, LIPC, PPARG2, andPPARGC1A loci were evaluated in theSTOP-NIDDM trial with 3.3 years offollow-up for diabetes risk (Table 3;Supplementary Table 9) (20–23). Two of11 SNPs from the PPARA locus wereassociated with response to acarbose(20). Of six SNPs from the HNF4A locus,two were associated with response toacarbose (21). Single SNPs at the LIPCand the PPARGC1A loci were alsoassociated with response to acarbose(22,23).

Quality of Included StudiesWe provide detailed results on thequality of included studies inSupplementary Table 10. None of theincluded studies was a prospectiveRCT designed to evaluate apharmacogenetic interaction, and only13 of 34 (38%) had a control group.Twenty-six of 34 (76%) studies did notreport on losses to follow-up.Pharmacogenetic analyses wereprespecified in 24 of 34 (71%) studiesand were either not reported or notprespecified in the remainder of thestudies. Sixteen of 34 (47%) studiesaddressed the issue of multiplecomparisons (or only looked at a singleSNP). Thirty-one of 34 studies (91%)addressed population stratification byadjusting for admixture or self-reportedrace/ethnicity or only included onerace/ethnicity. All studies providedsome information on method ofgenotyping. Only 14 of 34 (41%)reported on genotyping or SNP-specificcall rate. Most studies (24 of 34 [71%])did not report on genotypingconcordance. Twenty-seven of 34 (79%)reported on testing for Hardy–Weinbergproportions. Studies did not report onmasking of genotyping personnel, and41% declared some form of industrysupport.

CONCLUSIONS

In this systematic review, we identified34 articles on the pharmacogenetics ofdiabetes medications, with severalreporting statistically significantinteractions between genetic variantsandmedications for glycemic outcomes.Most pharmacogenetic interactionswere only evaluated in a single study,

did not use a control group, and/or didnot report enough information to judgeinternal validity. However, our resultsdo suggest specific, biologicallyplausible, gene–medicationinteractions, and we recommendconfirmation of the biologicallyplausible interactions as a priority,including those for drug transporters,metabolizers, and targets of action. Inparticular, we recommend follow-up ofthe 1) SLC22A1, SLC22A2, SLC47A1,PRKAB2, PRKAA2, PRKAA1, and STK11loci for metformin; 2) CYP2C9 andTCF7L2 loci for sulfonylureas; 3) KCNJ11,SLC30A8, NEUROD1/BETA2, UCP2, andPAX4 loci for repaglinide; 4) PPARG2 andPTPRD for pioglitazone; 5) KCNQ1 andRBP4 loci for rosiglitazone; and 6)PPARA,HNF4A, LIPC, and PPARGC1A locifor acarbose.

Given the number of comparisonsreported in the included studies and thelack of accounting for multiplecomparisons in approximately 53% ofstudies, many of the reported findingsmay be false positives. However, weexpect interactions between responseto medications and genes encodingtheir transporters, metabolizers,targets, and components of theirpathways for action as observed in theincluded studies. The DPP reportedsignificant interactions betweenmetformin and loci for its transporters(SLC22A1, SLC22A2, and SLC47A1) (7). Itdeserves mention that positive findingswere not replicated in other studiesevaluating these loci (16,34,38), butoutcomes (mean change in quantitativetraits, achievement of HbA1c ,7% (53mmol/mol) versus diabetes risk in theDPP) and follow-up time (6 to 18monthsvs. 3 years in the DPP) differed in theother studies as well as did study design.The DPP also reported on significantinteractions for loci associated withmetformin pharmacodynamics(PRKAB2, PRKAA2, PRKAA1, and STK11)(7,46). The primary action of metforminis the inhibition of hepatic glucoseproduction through inhibition ofgluconeogenesis, and interactions withloci associated within this pathway(PCK1, PPARA, and PPARGC1A) werereported (7,48,49). Sulfonylureas andrepaglinide bind to the sulfonylureareceptor (encoded by ABCC8), which


http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1

then inhibits the function of thepotassium channel encoded by KCNJ11and causes b-cell depolarization andeventual insulin secretion. While we dididentify interactions betweenrepaglinide and KCNJ11 (27), this locuswas not associated with sulfonylureaaction in a single study that evaluatedFG (24). Variation in CYP2C9, whichencodes an enzyme that metabolizessulfonylureas, was associated withresponse to sulfonylureas in one (25) oftwo studies (25,26). Finally, thethiazolidinediones activate peroxisomeproliferator-activated receptor greceptors, which regulate expression ofgenes important for sensitivity toinsulin. Thus, variation in PPARG wouldlikely affect response to this class ofmedications, and this was suggested inone (42) of the two studies (32,42)evaluating this for pioglitazone and wasnot evaluated for rosiglitazone.

Many putative loci were not evaluatedin the included studies. Variation in thehepatic cytochrome P450 enzymes,which metabolize diabetes medications,would be expected to impact theireffects, including variation in CYP3A4and CYP2C8 for repaglinide (51), CYP2C8and CYP2C9 for rosiglitazone (52), andCYP2C8 and CYP3A4 for pioglitazone(53). The transporter encoded bySLCOB1 transports repaglinide intohepatocytes for metabolism, andvariation in this gene could affect theresponse to this medication (51).Acarbose primarily decreases intestinalglucose absorption by inhibiting brushborder enzymes that hydrolyzecarbohydrates and is mainly excretedfecally and does not seem to haveobvious pharmacokinetic orpharmacodynamic targets.

Generally, we would also expect geneticvariants that impact b-cell function toaffect the response to insulinsecretagogues and genetic variantsthat impact insulin sensitivity to affectresponse to insulin sensitizingmedications. Also, because of itsprimary effect on PPG, geneticvariation impacting glucose-stimulated insulin secretion wouldlikely impact the response toacarbose. This rationale may explainother observed significantpharmacogenetic interactions (e.g.,

rosiglitazone–KCNQ1, repaglinide–SLC30A8, sulfonylurea–TCF7L2, andacarbose–HNF4A).

Prior work in this area has consisted ofmainly narrative reviews, many of whichhave included the studies that weidentified (3–6). We add to thisliterature by using a thorough andsystematic approachwith double reviewat all levels to identify as many studiesas possible that have reported someinteraction between individual diabetesmedications of interest and diabetes riskand glycemic outcomes. Thus wepresent the state of the literature on thepharmacogenetics of type 2 diabetes,which lays the groundwork for directingfuture research efforts. Another novelcontribution of our systematic review isthe collection of detailed qualityinformation from included studies,which aids in the interpretation of priorstudies and illuminates areas forimprovement and standardization.

The major limitation of the literature onthe pharmacogenomics of type 2diabetes is the lack of high-qualitystudies to identify and confirm findingsfor specific drug–SNP–outcomecombinations: 1) The rationale forselection of loci and interactions studiedwas often not clear, which raises theconcern of selective reporting of resultsand publication bias; in particular, wewould be less suspicious of false-positiveresults in the setting of prespecifiedanalyses based on prior evidence and/orbiologic plausibility with adjustment formultiple comparisons. Therefore, it islikely that positive results were reportedand that null results were not. 2) Thesmall size of many of the studies doesnot exclude the possibility thatinteractions exist but could not beidentified because of lack of power; wereported study results as significantbased on a P value less than 0.05 buthave noted results when P values were,0.20 when possible. Meta-analysiscould help to address this issue, but theheterogeneity of studies with specificSNP–drug interactions, outcomes, andfollow-up times differing across studiesprecluded quantitative synthesis withmeta-analyses. While our qualitativesynthesis summarizes the literature andsuggests the existence of specific gene–drug interactions, we could not

complete meta-analyses to quantifythese observations. 3) Most studies didnot have a placebo or other controlgroup. Therefore, our inferences oftenrelied on the results of a singlemedication intervention on change in orincidence of outcomes by genotype;these types of studies do not exclude thepossibility that we are simply observingthe effect of genotype and notspecifically modification of the responseto the medication. 4) Studies did notgenerally provide information todetermine the potential for selectionbias based on availability of genotypinginformation, on losses to follow up, or onthe amount and handling of missingdata. Regarding selection bias due toavailability of genotyping, participantbehaviors (e.g., adherence tointervention, follow-up) and outcomes(diabetes, death) may have differedbetween those with and withoutgenotyping information; these kinds ofdifferentiating characteristics inparticipants included in genetic analysescould impact the observed gene–druginteractions. 5)While studies did provideinformation on methods for genotyping,information on SNP-specific call rate wasoften not reported, and studies did notreport on masking of personnelperforming genotyping. 6)While none ofthe included studies were actually denovo RCTs, which would limit selectionbias and confounding most completely,several articles were based on data fromprior, well-conducted randomized trials(7,8,17–23,36,37,45). In the case ofthese trials, we would not expect thatparticipant characteristics correlatedwith genotype would be related toassigned intervention and thus can feelmore confident about the robustness ofthese studies. To address these issueswith quality, we did tailor our inclusioncriteria and abstraction tools to limit theinclusion of poor-quality studies and tounderstand the important potentialsources of bias.

One limitation of our systematic reviewmethodology is the exclusion of studiesof patients on more than one diabetesmedication. We sought to identifypharmacogenetic interactions that werebased on a single drug and single geneticvariant and wanted to avoid drug–drug–SNP interactions. Future work could


http://care.diabetesjournals.org

address these types of interactions.Because of this exclusion, we did notinclude additional articles on thepharmacogenetics of diabetes in thisreview (11,54–56). These studies didreport positive findings regarding gene–drug interactions based onpharmacodynamics [e.g., PPARG–rosiglitazone (54), SLC22A1–metformin(11), IRS1–sulfonylurea (55), and ATM–metformin (56)]. These studies confirmour hypotheses regarding the possibilityof gene–drug interactions based onpharmacodynamics but are stillindividual studies different enough fromthe existing literature to preclude meta-analysis. We also excluded studies ofnon-FDA-approved medications andtherefore did not include the article byFeng et al. evaluating interactions withgliclazide; this study did demonstrate asignificant interaction betweengliclazide and ABCC8 and KCNJ11 SNPsconsistent with the pharmacodynamicsof sulfonylureas (57). We limited ouranalyses to diabetes and glycemicoutcomes (HbA1c, FG, PPG) becausethese are more commonly andconsistently measured and are alsostrongly associated with improvementsin long-term complications andmortality (58). Future studies shouldevaluate other important efficacy andsafety outcomes. Finally, because ofstudy resource limitations, we excludednon-English language studies from ourinitial search and cannot estimate thenumber of otherwise-eligible studiesthat we excluded based on this.

Guidance for the Development ofFuture Evidence in DiabetesPharmacogeneticsWe recommend that guidelines for thedesign, analysis, and reporting ofpharmacogenetic studies of diabetesmedications be developed to improvestudy quality and enhancecomparability among studies; we haveprovided a prioritized list of quality andreporting items in Supplementary Table11. The incorporation of response tomedications based on genetic variationinto clinical practice cannot occurwithout well-designed studiesconfirming significant pharmacogeneticinteractions. Based on the limitationsof the current literature, we recommendthe following for future studies: 1) a priori

specification of the SNPs andmedications to be studied, 2) the use ofexperimental designs, 3) inclusion of aconcurrent comparison group whenpossible, 4) agreement in the diabetespharmacogenetics communityregarding standardized outcomes andfollow-up (e.g., HbA1c at 3 months),5) sufficient power for the primaryoutcome, 6) adjustment for multiplecomparisons if multiple SNPs areexamined, and 7) controlling forpopulation stratification andrelatedness. In addition, independentreplication is important. Werecommend that diabetespharmacogenetics studies use currentguidelines for reporting of geneticassociation studies (12) and that theseguidelines be extended to emphasizeinformation relevant topharmacogenetic studies, includingprespecified reporting of analyses withrationale, estimates of type 2 error,standardized reporting of medicationinterventions, and reporting ofdifferences between genotyped andnongenotyped subjects when possible.

In conclusion, for all known efficaciousdiabetes preventive and therapeuticpharmacologic agents, some patientsbenefit or experience harm while othersdo not. In this systematic review,we findevidence of biologically plausiblepharmacogenetic interactions formetformin, sulfonylureas, repaglinide,pioglitazone, rosiglitazone, andacarbose, but these results requireconfirmation in future studies todetermine if an individual’s geneticinformation can be used to individualizethe choice of prediabetes and diabetespharmacologic management.Importantly, our results should guidethe development of guidelines for thedesign, conduct, and reporting ofstudies of the pharmacogenetics of type2 diabetes and other chronic conditions.These promising results show thepotential of using genetic variation totailor therapy for type 2 diabetesprevention and management.

Acknowledgments. The authors thank thefollowing for their contribution to thisarticle: Elisabeth Haberl (Johns HopkinsUniversity, Baltimore,MD), Padmini Ranasinghe(Johns Hopkins University, Baltimore, MD), and

Leslie Jackson (University of Indiana,Indianapolis, IN).

Funding. N.M.M. was supported by theNational Institutes of Health/National Centerfor Research Resources (1KL2RR025006-01).M.O.G. was supported by the National Heart,Lung, and Blood Institute (5T32HL007024) andthe National Institute of Diabetes and Digestiveand Kidney Diseases (5T32DK-062707). P.B. wassupported by the National Heart, Lung, andBlood Institute (5T32HL007024). S.B. wassupported by the National Center for ResearchResources (KL2 RR024990) and the NationalCenter for Advancing Translational Sciences(KL2TR000440), National Institutes of Health.The Baltimore Diabetes Research Center alsoprovided support for this study (P60 DK-079637).

Duality of Interest. No potential conflicts ofinterest relevant to this article were reported.

Author Contributions. N.M.M. conceived ofthe study, refined the study question anddesign, obtained data, interpreted results, anddrafted and revised the manuscript. M.O.G.,W.L.B., S.B., L.M.W., W.H.L.K., and J.M.C.refined the study design, obtained data,interpreted results, and contributed to therevision of the manuscript. P.B. and A.S.obtained data and contributed to the revision ofthe manuscript. E.B. refined the study designand contributed to the revision of themanuscript.

References1. American Diabetes Association. Standards

of medical care in diabetesd2013.Diabetes Care 2013;36(Suppl. 1):S11–S66

2. Inzucchi SE, Bergenstal RM, Buse JB, et al.;American Diabetes Association (ADA);European Association for the Study ofDiabetes (EASD). Management ofhyperglycemia in type 2 diabetes: a patient-centered approach: position statement ofthe American Diabetes Association (ADA)and the European Association for the Studyof Diabetes (EASD). Diabetes Care 2012;35:1364–1379

3. Franks PW, Pearson E, Florez JC. Gene-environment and gene-treatmentinteractions in type 2 diabetes: progress,pitfalls, and prospects. Diabetes Care 2013;36:1413–1421

4. Mannino GC, Sesti G. Individualized therapyfor type 2 diabetes: clinical implications ofpharmacogenetic data. Mol Diagn Ther2012;16:285–302

5. Pearson ER. Pharmacogenetics in diabetes.Curr Diab Rep 2009;9:172–181

6. Manolopoulos VG, Ragia G, Tavridou A.Pharmacogenomics of oral antidiabeticmedications: current data andpharmacoepigenomic perspective.Pharmacogenomics 2011;12:1161–1191

7. Jablonski KA, McAteer JB, de Bakker PIW,et al.; Diabetes Prevention ProgramResearch Group. Common variants in 40genes assessed for diabetes incidence and


http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc13-1276/-/DC1

response to metformin and lifestyleintervention in the diabetes preventionprogram. Diabetes 2010;59:2672–2681

8. Moore AF, Jablonski KA, Mason CC, et al.;Diabetes Prevention Program ResearchGroup. The association of ENPP1 K121Qwith diabetes incidence is abolished bylifestyle modification in the diabetesprevention program. J Clin EndocrinolMetab 2009;94:449–455

9. Moore AF, Florez JC. Genetic susceptibilityto type 2 diabetes and implications forantidiabetic therapy. Annu Rev Med 2008;59:95–111

10. Zhou K, Bellenguez C, Spencer CC, et al.;GoDARTS and UKPDS DiabetesPharmacogenetics Study Group; WellcomeTrust Case Control Consortium 2; MAGICinvestigators. Common variants near ATMare associated with glycemic response tometformin in type 2 diabetes. Nat Genet2011;43:117–120

11. Becker ML, Visser LE, van Schaik RH,Hofman A, Uitterlinden AG, Stricker BH.Genetic variation in the organic cationtransporter 1 is associated with metforminresponse in patients with diabetes mellitus.Pharmacogenomics J 2009;9:242–247

12. Little J, Higgins JP, Ioannidis JP, et al.STrengthening the REporting of GeneticAssociation studies (STREGA)danextension of the STROBE statement. Eur JClin Invest 2009;39:247–266

13. Little J, Higgins JPT, Eds. The HuGENet HuGEReview Handbook, version 1.0. Ottawa,HuGENet, 2006

14. Wacholder S, Chatterjee N, Hartge P. Jointeffect of genes and environment distortedby selection biases: implications forhospital-based case-control studies. CancerEpidemiol Biomarkers Prev 2002;11:885–889

15. Pearson ER, Donnelly LA, Kimber C, et al.Variation in TCF7L2 influences therapeuticresponse to sulfonylureas: a GoDARTsstudy. Diabetes 2007;56:2178–2182

16. Zhou K, Donnelly LA, Kimber CH, et al.Reduced-function SLC22A1 polymorphismsencoding organic cation transporter 1 andglycemic response to metformin:a GoDARTS study. Diabetes 2009;58:1434–1439

17. Florez JC, Jablonski KA, Bayley N, et al.;Diabetes Prevention Program ResearchGroup. TCF7L2 polymorphisms andprogression to diabetes in the DiabetesPrevention Program. N Engl J Med 2006;355:241–250

18. Florez JC, Jablonski KA, Kahn SE, et al.Type 2 diabetes-associated missensepolymorphisms KCNJ11 E23K and ABCC8A1369S influence progression to diabetesand response to interventions in theDiabetes Prevention Program. Diabetes2007;56:531–536

19. Florez JC, Jablonski KA, McAteer J, et al.;Diabetes Prevention Program ResearchGroup. Testing of diabetes-associatedWFS1 polymorphisms in the DiabetesPrevention Program. Diabetologia 2008;51:451–457

20. Andrulionyte L, Kuulasmaa T, Chiasson JL,Laakso M; STOP-NIDDM Study Group.Single nucleotide polymorphisms of theperoxisome proliferator-activatedreceptor-alpha gene (PPARA) influencethe conversion from impaired glucosetolerance to type 2 diabetes: the STOP-NIDDM trial. Diabetes 2007;56:1181–1186

21. Andrulionyte L, Laukkanen O, Chiasson JL,Laakso M; STOP-NIDDM Study Group.Single nucleotide polymorphisms of theHNF4alpha gene are associated with theconversion to type 2 diabetes mellitus: theSTOP-NIDDM trial. J Mol Med (Berl) 2006;84:701–708

22. Andrulionyte L, Zacharova J, Chiasson JL,Laakso M; STOP-NIDDM Study Group.Common polymorphisms of the PPAR-gamma2 (Pro12Ala) and PGC-1alpha(Gly482Ser) genes are associated with theconversion from impaired glucosetolerance to type 2 diabetes in the STOP-NIDDM trial. Diabetologia 2004;47:2176–2184

23. Zacharova J, Todorova BR, Chiasson JL,Laakso M; STOP-NIDDM Study Group. TheG-250A substitution in the promoter regionof the hepatic lipase gene is associated withthe conversion from impaired glucosetolerance to type 2 diabetes: the STOP-NIDDM trial. J Intern Med 2005;257:185–193

24. Gloyn AL, Hashim Y, Ashcroft SJ, Ashfield R,Wiltshire S, Turner RC; UK ProspectiveDiabetes Study (UKPDS 53). Associationstudies of variants in promoter and codingregions of beta-cell ATP-sensitive K-channelgenes SUR1 and Kir6.2with Type 2 diabetesmellitus (UKPDS 53). Diabet Med 2001;18:206–212

25. Suzuki K, Yanagawa T, Shibasaki T, KaniwaN, Hasegawa R, Tohkin M. Effect of CYP2C9genetic polymorphisms on the efficacy andpharmacokinetics of glimepiride in subjectswith type 2 diabetes. Diabetes Res ClinPract 2006;72:148–154

26. Becker ML, Visser LE, Trienekens PH,Hofman A, van Schaik RH, Stricker BH.Cytochrome P450 2C9 *2 and *3polymorphisms and the dose and effect ofsulfonylurea in type II diabetes mellitus.Clin Pharmacol Ther 2008;83:288–292

27. He YY, Zhang R, Shao XY, et al. Associationof KCNJ11 and ABCC8 geneticpolymorphisms with response torepaglinide in Chinese diabetic patients.Acta Pharmacol Sin 2008;29:983–989

28. Huang Q , Yin JY, Dai XP, et al. Associationanalysis of SLC30A8 rs13266634 andrs16889462 polymorphisms with type 2

diabetes mellitus and repaglinide responsein Chinese patients. Eur J Clin Pharmacol2010;66:1207–1215

29. Qin W, Zhang R, Hu C, et al.; InternationalType 2 Diabetes 1q Consortium. A variationin NOS1AP gene is associated withrepaglinide efficacy on insulin resistance intype 2 diabetes of Chinese. Acta PharmacolSin 2010;31:450–454

30. YuW, Hu C, Zhang R, et al. Effects of KCNQ1polymorphisms on the therapeutic efficacyof oral antidiabetic drugs in Chinesepatients with type 2 diabetes. ClinPharmacol Ther 2011;89:437–442

31. Wang J, Bao YQ, Hu C, et al. Effects ofABCA1 variants on rosiglitazonemonotherapy in newly diagnosed type 2diabetes patients. Acta Pharmacol Sin 2008;29:252–258

32. Blüher M, Lübben G, Paschke R. Analysis ofthe relationship between the Pro12Alavariant in the PPAR-gamma2 gene and theresponse rate to therapy with pioglitazonein patients with type 2 diabetes. DiabetesCare 2003;26:825–831

33. Saitou M, Osonoi T, Kawamori R, et al.Genetic risk factors and the anti-atherosclerotic effect of pioglitazone oncarotid atherosclerosis of subjects withtype 2 diabetesda retrospective study.J Atheroscler Thromb 2010;17:386–394

34. Choi JH, Yee SW, Ramirez AH, et al.A common 59-UTR variant in MATE2-Kis associated with poor response tometformin. Clin Pharmacol Ther 2011;90:674–684

35. Dong M, Gong ZC, Dai XP, et al. Serineracemase rs391300 G/A polymorphisminfluences the therapeutic efficacy ofmetformin in Chinese patients withdiabetes mellitus type 2. Clin ExpPharmacol Physiol 2011;38:824–829

36. Florez JC, Jablonski KA, McAteer JB, et al.;Diabetes Prevention Program ResearchGroup. Effects of genetic variantspreviously associated with fasting glucoseand insulin in the Diabetes PreventionProgram. PLoS ONE 2012;7:e44424

37. Florez JC, Jablonski KA, Taylor A, et al.;Diabetes Prevention Program ResearchGroup. The C allele of ATM rs11212617does not associate with metforminresponse in the Diabetes PreventionProgram. Diabetes Care 2012;35:1864–1867

38. Tkáč I, Klimčáková L, Javorský M, et al.Pharmacogenomic association between avariant in SLC47A1 gene and therapeuticresponse to metformin in type 2 diabetes.Diabetes Obes Metab 2013;15:189–191

39. Gong ZC, Huang Q, Dai XP, et al. NeuroD1A45T and PAX4 R121W polymorphisms areassociated with plasma glucose level ofrepaglinide monotherapy in Chinesepatients with type 2 diabetes. Br J ClinPharmacol 2012;74:501–509


http://care.diabetesjournals.org

40. Jiang F, Li Q, Hu C, et al. Association of aSLC30A8 genetic variant with monotherapyof repaglinide and rosiglitazone effect innewly diagnosed type 2 diabetes patients inChina. Biomed Environ Sci 2012;25:23–29

41. Sheng FF, Dai XP, Qu J, et al. NAMPT-3186C/T polymorphism affects repaglinideresponse in Chinese patients with Type 2diabetes mellitus. Clin Exp PharmacolPhysiol 2011;38:550–554

42. Pei Q, Huang Q, Yang GP, et al. PPAR-g2 andPTPRD gene polymorphisms influence type2 diabetes patients’ response topioglitazone in China. Acta Pharmacol Sin2013;34:255–261

43. Zhou F, Huang Q, Dai X, Yin J, Wu J, Zhou H,Gong Z, Liu Z. Impact of retinol bindingprotein 4 polymorphism on rosiglitazoneresponse in Chinese type 2 diabeticpatients. Zhong Nan Da Xue Xue Bao Yi XueBan 2011;36:949–957

44. Wang S, Se YM, Liu ZQ, et al. Effect ofgenetic polymorphism of UCP2-866 G/A onrepaglinide response in Chinese patientswith type 2 diabetes. Pharmazie 2012;67:74–79

45. Moore AF, Jablonski KA, McAteer JB, et al.;Diabetes Prevention Program ResearchGroup. Extension of type 2 diabetesgenome-wide association scan results inthe diabetes prevention program. Diabetes2008;57:2503–2510

46. Viollet B, Guigas B, Sanz Garcia N, Leclerc J,Foretz M, Andreelli F. Cellular andmolecular mechanisms of metformin: anoverview. Clin Sci (Lond) 2012;122:253–270

47. Gong L, Goswami S, Giacomini KM, AltmanRB, Klein TE. Metformin pathways:pharmacokinetics and pharmacodynamics.Pharmacogenet Genomics 2012;22:820–827

48. Rakhshandehroo M, Knoch B, Muller M,Kersten S. Peroxisome proliferator-activated receptor alpha target genes.PPAR Res 2010;2010:612089

49. Liu C, Lin JD. PGC-1 coactivators in the controlof energy metabolism. Acta Biochim BiophysSin (Shanghai) 2011;43:248–257

50. Cauchi S, Del Guerra S, Choquet H, et al.Meta-analysis and functional effects of theSLC30A8 rs13266634 polymorphism onisolated human pancreatic islets. MolGenet Metab 2010;100:77–82

51. Whirl-Carrillo M, McDonagh EM, HebertJM, et al. Pharmacogenomics knowledgefor personalized medicine. Clin PharmacolTher 2012;92:414–417

52. Baldwin SJ, Clarke SE, Chenery RJ.Characterization of the cytochrome P450enzymes involved in the in vitrometabolism of rosiglitazone. Br J ClinPharmacol 1999;48:424–432

53. ACTOS oral tablet, pioglitazonehydrochloride oral tablet [package insert].

Indianapolis, IN, Eli Lilly & Co/TakedaPharmaceutical Co Ltd, 2004

54. Kang ES, Park SY, Kim HJ, et al. Effects ofPro12Ala polymorphism of peroxisomeproliferator-activated receptor gamma2gene on rosiglitazone response in type 2diabetes. Clin Pharmacol Ther 2005;78:202–208

55. Seeringer A, Parmar S, Fischer A, et al.Genetic variants of the insulin receptorsubstrate-1 are influencing the therapeuticefficacy of oral antidiabetics. Diabetes ObesMetab 2010;12:1106–1112

56. van Leeuwen N, Nijpels G, Becker ML,et al. A gene variant near ATM issignificantly associated with metformintreatment response in type 2 diabetes:a replication and meta-analysis of fivecohorts. Diabetologia 2012;55:1971–1977

57. Feng Y, Mao G, Ren X, et al. Ser1369Alavariant in sulfonylurea receptor geneABCC8 is associated with antidiabeticefficacy of gliclazide in Chinese type 2diabetic patients. Diabetes Care 2008;31:1939–1944

58. Stratton IM, Adler AI, Neil HA, et al.Association of glycaemia withmacrovascular and microvascularcomplications of type 2 diabetes (UKPDS35): prospective observational study. BMJ2000;321:405–412


the pharmacogenetics of type 2 diabetes: a systematic …...in 2013, there existed multiple...

Documents