lable at ScienceDirect

Atherosclerosis 233 (2014) 206e210

Contents lists avai

Atherosclerosis

journal homepage: www.elsevier .com/locate/atherosclerosis

Presence and type of low density lipoprotein receptor (LDLR) mutationinfluences the lipid profile and response to lipid-lowering therapy inBrazilian patients with heterozygous familial hypercholesterolemia

Paulo Caleb Junior Lima Santos a,*, Aline Cruz Morgan a, Cintia Elin Jannes a,Luciana Turolla a, Jose Eduardo Krieger a, Raul D. Santos b, Alexandre Costa Pereira a,*

a Laboratory of Genetics and Molecular Cardiology, Heart Institute (InCor), University of Sao Paulo Medical School, SP, Brazilb Lipid Clinic, Heart Institute (InCor), University of Sao Paulo Medical School, SP, Brazil

a r t i c l e i n f o

Article history:Received 25 August 2013Received in revised form4 December 2013Accepted 5 December 2013Available online 4 January 2014

Keywords:LDLR geneLipid-lowering therapyFamilial hypercholesterolemiaType of mutation

* Corresponding authors. Laboratory of Genetics andInstitute, University of Sao Paulo Medical School, Av.44 Cerqueira César, São Paulo, SP CEP 05403-000, Brfax: þ55 11 2661 5022.

E-mail addresses: pacaleb@usp.br (P.C.JuniorL. Sanusp.br (A.C. Pereira).

0021-9150/$ e see front matter � 2014 Published byhttp://dx.doi.org/10.1016/j.atherosclerosis.2013.12.028

a b s t r a c t

Objectives: Familial hypercholesterolemia (FH) is an autosomal dominant disease caused mainly by LDLRmutations. This study assessed the influence of the presence and type of LDLR mutation on lipid profileand the response to lipid-lowering therapy in Brazilian patients with heterozygous FH.Methods: For 14 � 3 months, 156 patients with heterozygous FH receiving atorvastatin were followed.Coding sequences of the LDLR gene were bidirectionally sequenced, and the type of LDLR mutations wereclassified according to their probable functional class.Results: The frequencies of the types of LDLR mutations were: null-mutation (n ¼ 40, 25.6%), defective-mutation (n ¼ 59, 37.8%), and without an identified mutation (n ¼ 57, 36.6%). Baseline total cholesterol(TC) and low-density lipoprotein cholesterol (LDL-C) were higher in patients carrying a null mutation(9.9 � 1.9 mmol/L, 7.9 � 1.7 mmol/L), compared to those with a defective (8.9 � 2.2 mmol/L,7.0 � 2.0 mmol/L), or no mutation (7.9 � 1.9 mmol/L, 5.8 � 1.9 mmol/L) (p < 0.001). After treatment, theproportion of patients attaining an LDL-C<3.4 mmol/L was significantly different among groups: null(22.5%), defective (27.1%), and without mutations (47.4%) (p ¼ 0.02). The presence of LDLR mutations wasindependently associated with higher odds of not achieving the LDL-C cut-off (OR 9.07, 95% CI 1.41e58.16, p ¼ 0.02).Conclusions: Our findings indicate that the presence and type of LDLR mutations influence lipid profileand response to lipid-lowering therapy in Brazilian patients with heterozygous FH. Thus, more intensivecare with pharmacological therapeutics should be performed in patients who have a molecular analysisindicating the presence of a LDLR mutation.

� 2014 Published by Elsevier Ireland Ltd.

1. Introduction

Familial hypercholesterolemia (FH; OMIM 143890) is an auto-somal dominant disease mainly caused by mutations in the low-density lipoprotein receptor (LDLR) gene. This common disorderis characterized by severely elevated LDL-C (low-density lipopro-tein cholesterol), xanthomas, and the development of prematurecardiovascular disease (CVD). To date, more than 1000mutations in

Molecular Cardiology, HeartDr. Enéas de Carvalho Aguiar,azil. Tel.: þ55 11 2661 5929;

tos), alexandre.pereira@incor.

Elsevier Ireland Ltd.

the gene that encodes LDL receptors have been described world-wide [1e3].

The LDLR gene (606945) consists of 18 exons, with correlationbetween exons and the functional domains of the LDL receptorprotein. Regarding the type of LDLRmutations, five classes based onphenotypic effects have been proposed. Class I mutations usuallyare nonsense mutations, large deletions and promoter mutationsresulting in no detectable LDL receptor protein. Class II mutations(transport defective alleles) cause either complete (class II a) orpartial (class II b) block of the transport of the LDL receptor from theendoplasmic reticulum to the Golgi apparatus. Class III mutationsresult in defective LDL binding; class IV mutations cause a defi-ciency in the internalization of LDL and class V mutations fail torecycle effectively [3e6].

Table 1Baseline clinical and laboratory characteristics of FH patients receiving lipid-lowering therapy according to gender.

Variables Females Males p value

106 50

Age, years 54.3 � 15.3 48.8 � 12.0 0.03BMI, Kg/m2 27.5 � 5.3 26.7 � 5.5 0.46Race/color, %White 73.6 74.0Intermediate 19.8 14.0 0.40Black 6.6 12.0

Smoking, %No 73.6 40.0Ex-smoker 13.2 44.0 <0.01Yes 7.6 16.0

Total cholesterol, mmol/L 8.8 � 2.3 8.7 � 1.8 0.88LDL-C, mmol/L 6.8 � 2.2 6.8 � 1.9 0.85HDL-C, mmol/L 1.1 � 0.3 1.2 � 0.3 0.14Triglycerides, mmol/L 1.6 � 0.8 2.0 � 1.4 0.06Creatine kinase, U/L 67 � 73 55 � 43 0.59Type of the LDLR gene mutation,

% (null/defective)(17.0/42.5) (44.0/28.0) 0.01

Patient without identifiedmutation, %

40.5 28.0

FH: familial hypercholesterolemia; BMI: body mass index; LDL-C: low-density li-poprotein cholesterol; HDL-C: high-density lipoprotein cholesterol.Race/color was categorized in White, Intermediate (“Mulatto or Pardo” in Portu-guese, person with admixture between White and Black) and Black.Biochemical data were adjusted for age and race.

P.C.JuniorL. Santos et al. / Atherosclerosis 233 (2014) 206e210 207

Statins are the first choice of treatment to reduce cardiovascularmorbidity and mortality in FH, a condition that is considerablyundertreated [7e10]. Furthermore, the influence of the type of LDLRmutations on response to treatment is not completely clear, andfew studies have been performed with controversial findings[4,6,11e18]. The aim of this study was to assess the influence of thetype of LDLR mutation on lipid profile and the response to lipid-lowering therapy in Brazilian patients with heterozygous FH.

2. Materials and methods

2.1. Patients clinical and laboratory evaluation

This study included 156 index Brazilian patients with a het-erozygous FH phenotype from the Lipid Clinic, Heart Institute(InCor), University of São Paulo Medical School, São Paulo, Brazil.These patients were diagnosed using biochemical and clinical data[19,20] andwere followed for at least 12months from January 2006to December 2012 while receiving lipid-lowering therapy. Treat-ment was administered without any knowledge of the kind of ge-netic defect causing the FH phenotype by the attending physician.Atorvastatin dosage and the association of ezetimibe, and adjuvantcholesterol lowering medication, were set according to thediscretion of the physician, to attain the greatest LDL-C loweringeffect possible. All patients were assessed at least 3 times duringfollow-up. The study protocol was approved by the InstitutionalEthics Committee (CAPPesq numbers 022/11 and 191/04), andwritten informed consent was obtained from all participants priorto entering the study.

Plasma total cholesterol and LDL-C values were measured byenzymatic methods obtained at least in the following periods: atbaseline (first value), at the initiation of atorvastatin use (imme-diately before), and on average after 1-year of treatment onset (thetwo first measured values were not necessarily the same). Forcholesterol values, to convert from the SI unit (mmol/L) to theconventional unit (mg/dL), divide by the conversion factor 0.0259.Laboratory assessment of plasma creatine kinase (CK) concentra-tions was performed at least 3 times during follow-up by a kineticmethod. Musculoskeletal treatment side effects considered were:myalgia (atorvastatin-induced muscle pain, irrespective of CKvalues, at onset of treatment or in dose up-titration until the firstyear of follow-up), CK elevations of more than 3 times the upperlimit of the normal range (irrespective of symptoms), and rhab-domyolysis, as previously described [21].

2.2. Sequencing

Genomic DNA was extracted from peripheral blood following astandard salting-out procedure. Coding sequences of the LDLR gene(18 exons) were amplified by polymerase chain reaction (PCR) us-ing primers described in the Supplementary Table 1. PCR productswere purified using ExoSAP-IT� reagent (GE Healthcare, NJ, USA)and were bidirectionally sequenced using the ABI TerminatorSequencing Kit and the ABI 3500XL Sequencer� according to themanufacturer’s instructions (Applied Biosystems, Foster City, CA,USA). In addition, sequencing for the APOB and PCSK9 genes andmultiplex ligation-dependent probe amplification (MLPA) tech-nique for the identification of LDLR gene deletions/insertions wereperformed.

LDLR gene mutations were classified according to their probablefunctional class as previously reported in the literature [5,6,22e26]and through the use of a curated database (www.jojogenetics.nl/wp/database/). Null-mutations included classes 1 and 2A, andlarge rearrangements or deletions leading to a frameshift and apremature stop codon. Defective-mutation included classes 2B, 3, 4,

and 5, and non-frameshift small deletions or insertions. There arealso patients without an FH-related mutation who were classifiedin the “without identified mutation”.

2.3. Statistical analysis

Continuous variables data are presented as mean and standarddeviation and categorical variables as frequencies. Chi-square orFisher exact tests were performed for comparative analysis of thecategorical variables according to gender or the type of LDLR mu-tation. The Student t test or analysis of variance (ANOVA) was usedfor comparing age, body mass index, CK values, cholesterol data,and variation of LDL-C means according to the type of the LDLRmutation. Variation data were calculated as 1-year treatment valueminus baseline value divided by baseline value, expressed as apercentage. Biochemical data were adjusted for age, gender, andself-referred ethnicity. Variation data and 1-year treatment datawere also adjusted for baseline values. Atorvastatin doses and useof ezetimibe, were also inserted in the model as covariates. Tukey’spost hoc test was performed for identifying statistically differentgroups. In addition, logistic regression univariate and multivariateanalyses were performed to evaluate predictors for LDL-C<3.4 mmol/L after 1 year. All statistical analyses were carried outusing SPSS software (version 16.0, IBM, New York, NY), with thelevel of significance set at p � 0.05.

3. Results

3.1. General characteristics of FH patients receiving lipid-loweringtherapy

Baseline characteristics of study subjects according to genderare shown in Table 1. Of the 156 patients (mean age 52.5 � 14.5years), 106 (67.9%) were female and 50 (32.1%) were male. All pa-tients received atorvastatin for at least 12 months (mean andstandard deviation 14 � 3 months of follow-up of therapy). Initialdaily doses were 20 mg (n ¼ 14, 9.0%), 40 mg (n ¼ 51, 32.7%), 60 mg(n¼ 2, 1.3%) and 80 mg (n¼ 89, 57.0%); and only 10 (6.4%) had their

Table 3Logistic regression multivariate analysis of factors associated with LDL-C>3.4 mmol/L.

Variables OR 95% IC p value

Gender (male) 4.01 0.63e25.30 0.14Age 1.02 0.95e1.08 0.63Self-declared ethnicity 1.21 0.51e2.87 0.66Body mass index 1.03 0.89e1.21 0.68Smoking 1.46 0.38e5.58 0.58Atorvastatin dose 0.97 0.95e0.99 0.02Ezetimibe use 0.75 0.11e5.25 0.75Baseline LDL-C 1.00 0.99e1.01 0.86Presence of LDLR mutations 9.07 1.41e58.16 0.02

LDL-C: low-density lipoprotein cholesterol.

P.C.JuniorL. Santos et al. / Atherosclerosis 233 (2014) 206e210208

doses modified during follow-up. None of study subjects dis-continued the drug during follow-up. In addition to atorvastatintreatment, 89 (57.1%) patients also received ezetimibe 10 mg/day.The most frequently used drugs were: acetylsalicylic acid (37.8%),angiotensin converting enzyme inhibitors (33.3%), beta-blockers(30.8%), hydrochlorothiazide (25.0%), and metformin (13.5%).Frequent comorbidities were: hypertension (58.3%), diabetes mel-litus (14.7%), and acute myocardial infarction (9.6%).

During the one-year follow-up period, 17 patients (11.6%)developed a myalgia clinical phenotype, and no patient developedrhabdomyolysis. No difference regarding the frequency of myalgia(p ¼ 0.45) or CK values (p ¼ 0.70) according to the type of the LDLRmutation was observed (Table 2).

The frequencies of the type of LDLR mutations were null-mutation (n ¼ 40, 25.6%), defective-mutation (n ¼ 59, 37.8%), andwithout identified mutation (n ¼ 57, 36.6%). Patients carrying thep.(Glu317Glyfs*15), p.(His388Profs*53), p.(Phe629Tyrfs*16),p.(Ser791Cysfs*70), and exons 15 þ 16 deletion new mutationswere classified in the null-mutation group. The p.Ala391Thr andc.1706-10G > A genetic alterations were considered non-pathogenic based on in silico programs. Thus, we included pa-tients carrying both alterations (n ¼ 2) in the without FH-relatedmutation group. Neither patient presented APOB or PCSK9 muta-tions. Supplementary Table 2 shows the LDLR gene mutations ac-cording to their probable functional class.

The frequencies of ezetimibe use and atorvastatin doses weredifferently distributed according to the type of the mutation(p ¼ 0.002, p ¼ 0.04, respectively). The frequency of ezetimibe usewas 67.5%, 67.8% and 38.6% respectively in thosewith null, defectiveand no identified mutations. The frequency of atorvastatin dose of80 mg/day was 70% in those with null-mutation, 61% in those withdefective-mutation, and 45.6% in those without identifiedmutation.

Table 2Biochemical data and response to lipid-lowering therapy of FH patients according totype of LDLR mutation.

Variables Type of the LDLR mutation p value

Null Defective Without

40 59 57

Baseline TC, mmol/L 9.9 � 1.9a 8.9 � 2.2b 7.9 � 1.9c <0.001Baseline LDL-C, mmol/L 7.9 � 1.7a 7.0 � 2.0b 5.8 � 1.9c <0.0011-year treatment TC,

mmol/L6.7 � 2.1a 6.4 � 1.9a 5.5 � 1.6b 0.02

1-year treatment LDL-C,mmol/L

4.9 � 1.8a 4.6 � 1.5a,b 3.6 � 1.4b 0.003

Variation of TC/Baseline

TC, %

�37 � 19a �26 � 23b �27 � 27b 0.03

Variation of LDL-C/BaselineLDL-C, %

�44 � 19a �31 � 29b �32 � 35b 0.04

Variation of TC, mmol/L �3.9 � 2.3a �2.4 � 2.4b �2.3 � 2.4b 0.02Variation of LDL-C, mmol/L �3.6 � 2.0a �2.4 � 2.3b �2.2 � 2.2b 0.021-year treatment status, %

(LDL-C<3.4 mmol/L)22.5 27.1 47.4 0.02

Myalgia, % 12.5 15.3 7.0 0.45Creatine kinase, U/L 142 � 68 133 � 92 148 � 90 0.70

FH: familial hypercholesterolemia; TC: total cholesterol; LDL-C: low-density lipo-protein cholesterol.Biochemical data were adjusted for age, gender, and race. Variation data and 1-yeartreatment data were also adjusted for baseline data.Values with different superscript letters are significantly different (Tukey’s post hoctest).Variation data/baseline were calculated as 1-year treatment value minus baselinevalue, divided by baseline value, expressed as a percentage. Variation data were alsoexpressed as mmol/L (final value minus initial value). For cholesterol values, toconvert from the SI unit (mmol/L) to the conventional unit (mg/dL), divide by theconversion factor 0.0259.

3.2. Association of type of the LDLR mutation with lipid profile andresponse to lipid-lowering therapy

Table 2 shows baseline and 1-year treatment lipid levels. Base-line total cholesterol (TC) and low-density lipoprotein cholesterol(LDL-C) were higher in patients carrying null compared to thedefective and without a mutation group (p < 0.001, for both pa-rameters). In addition, LDL-C values were significantly differentamong groups in the initial time of lipid-lowering therapy and inthe 1-year treatment period (Supplementary Fig. 1).

After treatment, patients carrying null-mutations had higher TCand LDL-C values, while those without identified mutations hadlower TC and LDL-C values (p ¼ 0.02, p ¼ 0.003, respectively). Thevariations of TC and LDL-C, expressed in percentage consideringbaseline values or expressed in mmol/L, were higher in the null-mutation group compared to other groups, adjusted for cova-riates (Table 2). Even adjusting for ezetimibe use and atorvastatindose as additional covariates, 1-year treatment and variation dataremained significantly different. Supplementary Table 3 showsstratified analysis for biochemical data and response to lipid-lowering therapy according to type of LDLR mutation.

The proportion of patients with LDL-C<3.4 mmol/L, after 1 yeartreatment was significantly different among groups: null-mutation(22.5%), defective-mutation (27.1%), and without mutation (47.4%),p¼ 0.02. The analysis of predictors for achieving LDL-C<3.4mmol/Lis shown in Table 3 and Supplementary Table 4. The presence ofLDLR mutation was independently associated with higher odds ofnot achieving the LDL-C cutoff in an adjustedmodel (OR 9.07, 95% CI1.41e58.16, p ¼ 0.02) (Table 3).

4. Discussion

The main findings of this study indicate the influence of thepresence and type of LDLR mutation found in Brazilian patientswith heterozygous FH phenotype on the lipid profile and responseto lipid-lowering treatment. Thus, focusing on reducing CVDmorbidity and mortality in FH, more intensive care with pharma-cological therapeutics should be performed in patients who have amolecular analysis indicating the presence of a LDLR mutation.

Corroborating our findings, some studies have previously shownpositive associations between the type of LDLR mutation and somephenotypes: higher TC and LDL-C levels, different response tostatins, and higher frequencies of premature CVD (myocardialinfarction, angina pectoris, percutaneous coronary intervention orother invasive procedures and coronary artery bypass grafting)[4,6,11,14e17,22,27e29]. The present study extends these analysesto the 1-year follow-up period in patients receiving lipid-loweringtreatment. Thus, the higher TC and LDL-C values identified in thebaseline and 1-year periods in the null-mutation patient group

P.C.JuniorL. Santos et al. / Atherosclerosis 233 (2014) 206e210 209

support the idea that the relative reduction in plasma cholesterolconcentrations can be more apparent in null-mutation carriersthan in othermutation groups. On the other hand, the proportion ofpatients achieving the therapeutic target is lower in null-mutationcarriers. Consequently, more aggressive treatment and attentionshould be practiced on these patients.

Regarding this issue some studies have reported controversialresults [4,6,12,13,16,17,27]. Sun et al. examined 42 patients withheterozygous FH receiving simvastatin and concluded that therewere no differences in the response to treatment in terms of eitherrelative reduction or absolute decrease in LDL-C concentrationbetween patients with different LDL-receptor defects [12]. Sij-brands et al. compared the response to simvastatin, 20 mg dailyfor 9 weeks, in 27 FH patients. They observed that the percentageLDL lowering response to simvastatin treatment was similar inpatients with negative or positive mutations [13]. Chaves et al.observed that 22 FH patients with null-mutations showed apoorer response to simvastatin treatment than in 20 patients withdefective-mutations [17]. On the other hand, Miltiadous et al.studied 49 patients receiving atorvastatin 20 mg/day and followedfor 12 weeks. They showed that patients with a class V mutationexhibited higher percentage decrease in LDL-C after statin therapycompared to patients carrying class II mutations [6]. Couture et al.evaluated 63 children and adolescents with heterozygous FH andreceiving simvastatin 20 mg/day for 6 weeks. They concluded thatthe nature of LDLR gene mutations and other genetic and consti-tutional factors play a significant role in predicting the efficacy[16]. However, some aspects should explain part of these contro-versial results, such as small sample size and power of analysis,different types of statins or drug combinations, the use of differentend-points (such as decrease of LDL-C or achievement of LDL-Cgoal), mutation heterogeneity within the same predicted class,presence of other pharmacogenetic markers, ethnicity, anddifferent follow-up times. Thus, a robust meta-analysis, performedwith previous studies but with similar characteristics, mightindicate a resolution to the conflicting conclusions currentlyavailable in the literature.

Here, our casuistic was formed with over two-fold more femalethan male patients. In addition, the male group was younger andpresented a higher proportion of null mutations. A hypothesis forthese differences is that females, as previously described, arecharacteristically more aware of their health problems and aremore able to seek specialized care. It is known that women aremore prevalent in population-wide screening and tend to be moreadherent to primary care programs. Men with null mutations havemore severe dyslipidemia and this may have triggered an overallincreased recall in this specific substrata. The gender variable wassignificant in a multivariate analysis of factors associated with LDL-C<3.4 mmol/L. However, this finding on FH treatment scenarioneeds further investigation. Also, even though the null-mutationpatient group had higher frequencies of ezetimibe and atorvasta-tin 80 mg/day use, they had a lower frequency of attaining LDL-Cvalues <3.4 mmol/L and higher risk of not achieving the thera-peutic target. Our data are similar to previous studies that reportthe necessity of improving FH management [2,22,28].

In this context, and for a genetic disease with more than 1000known mutations, a randomized prospective study plus a largefollow-up of patients receiving pharmacological treatment willallow better assessment of the effects consequent to the type ofLDLRmutation on the outcomes of lipid-lowering treatment and tounderstand what is the best way to prevent CVD in this particularpopulation of high-risk individuals. It is possible that patients car-rying different functional LDLR mutations may respond differentlyto drugs, because HMG-CoA reductase may respond differentlyaccording to pathological status.

There are limitations in our study. First, our sample size isrelatively small. Second, it is not possible to completely exclude theinfluence of ezetimibe use and atorvastatin dose according tomutation, but even with this difference, null-mutation carriers hadworse LDL-C status. Third, we supported FH patients in whom nomutation was found by biochemical and clinical phenotypes, butwe were not able to sequence LDLR non-coding regions or toconduct functional tests. Finally, other variables could still beevaluated in terms of management of FH according to the type ofLDLR mutation, for examples: cost-effectiveness, other markersassociated with the pharmacogenomics of statins, use of concom-itant medications, comorbidities, age, gender, and ethnicity [30e33].

In conclusion, our findings indicate that the presence and typeof LDLR mutation influences the lipid profile and the response tolipid-lowering therapy in Brazilian patients with FH. Patients car-rying null-mutations have higher LDL-C values at baseline and after1-year and, although having a higher nominal decrease in LDL-C,the frequency of patients achieving a therapeutic target is lower.Thus, more intensive care with pharmacological therapeuticsshould be performed in patients who have a molecular analysisindicating the presence of a LDLR mutation.

Conflict of interest

RDS has the following conflicts of interest to declare: honorariaor consulting fees from Astra Zeneca, Aegerion, Amgen, Biolab,Bristol Myers Squibb, Biolab, Merck, Pfizer, Sanofi, Regeneron,Genzyme, Novo-Nordisk, Lilly, Boehringer Ingelheim, Novartis and,Nestle.

The other authors declare no conflict of interest.

Acknowledgments

PCJL Santos and AC Morgan are recipients of fellowships fromFAPESP, Proc. 2010-17465-8, Proc. 2013-09295-3, Proc. 2013-20614-3, and Proc. 2012-08118-8, respectively, Brazil. We also thank thepatients who participated in the study. The technical assistance ofthe Laboratory of Genetics and Molecular Cardiology group, HeartInstitute (InCor), and Sociedade Hospital Samaritano e Ministérioda Saúde (PROADI-SUS; SIPAR: 25000.180.672/2011-81) is grate-fully acknowledged.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.atherosclerosis.2013.12.028.

References

[1] Huijgen R, Vissers MN, Defesche JC, Lansberg PJ, Kastelein JJ, Hutten BA. Fa-milial hypercholesterolemia: current treatment and advances in management.Expert Rev Cardiovasc Ther 2008;6:567e81.

[2] Pijlman AH, Huijgen R, Verhagen SN, et al. Evaluation of cholesterol loweringtreatment of patients with familial hypercholesterolemia: a large cross-sectional study in The Netherlands. Atherosclerosis 2010;209:189e94.

[3] Schaefer JR, Kurt B, Sattler A, Klaus G, Soufi M. Pharmacogenetic aspects infamilial hypercholesterolemia with the special focus on FHMarburg (FHp.W556R). Clin Res Cardiol Suppl 2012;7:2e6.

[4] Choumerianou DM, Dedoussis GV. Familial hypercholesterolemia andresponse to statin therapy according to LDLR genetic background. Clin ChemLab Med 2005;43:793e801.

[5] Hobbs HH, Russell DW, Brown MS, Goldstein JL. The LDL receptor locus infamilial hypercholesterolemia: mutational analysis of a membrane protein.Annu Rev Genet 1990;24:133e70.

[6] Miltiadous G, Xenophontos S, Bairaktari E, Ganotakis M, Cariolou M, Elisaf M.Genetic and environmental factors affecting the response to statin therapy inpatients with molecularly defined familial hypercholesterolaemia. Pharma-cogenet Genomics 2005;15:219e25.

P.C.JuniorL. Santos et al. / Atherosclerosis 233 (2014) 206e210210

[7] Baigent C, Keech A, Kearney PM, et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 partici-pants in 14 randomised trials of statins. Lancet 2005;366:1267e78.

[8] Ghatak A, Faheem O, Thompson PD. The genetics of statin-induced myopathy.Atherosclerosis 2010;210:337e43.

[9] Santos RD, Gagliardi AC, Xavier HT, et al. Arq Bras Cardiol 2013;99:1e28.[10] Nordestgaard BG, Chapman MJ, Humphries SE, et al. Familial hyper-

cholesterolaemia is underdiagnosed and undertreated in the general popu-lation: guidance for clinicians to prevent coronary heart disease: consensusstatement of the European atherosclerosis society. Eur Heart J 2013, Aug 15[Epub ahead of print].

[11] Torres AL, Moorjani S, Vohl MC, et al. Heterozygous familial hypercholester-olemia in children: low-density lipoprotein receptor mutational analysis andvariation in the expression of plasma lipoprotein-lipid concentrations.Atherosclerosis 1996;126:163e71.

[12] Sun XM, Patel DD, Knight BL, Soutar AK. Influence of genotype at the lowdensity lipoprotein (LDL) receptor gene locus on the clinical phenotype andresponse to lipid-lowering drug therapy in heterozygous familial hyper-cholesterolaemia. The Familial Hypercholesterolaemia Regression StudyGroup. Atherosclerosis 1998;136:175e85.

[13] Sijbrands EJ, Lombardi MP, Westendorp RG, et al. Similar response to sim-vastatin in patients heterozygous for familial hypercholesterolemia withmRNA negative and mRNA positive mutations. Atherosclerosis 1998;136:247e54.

[14] Jeenah M, September W, Graadt van Roggen F, de Villiers W, Seftel H,Marais D. Influence of specific mutations at the LDL-receptor gene locus onthe response to simvastatin therapy in Afrikaner patients with heterozygousfamilial hypercholesterolaemia. Atherosclerosis 1993;98:51e8.

[15] Heath KE, Gudnason V, Humphries SE, Seed M. The type of mutation inthe low density lipoprotein receptor gene influences the cholesterol-lowering response of the HMG-CoA reductase inhibitor simvastatin inpatients with heterozygous familial hypercholesterolaemia. Atherosclerosis1999;143:41e54.

[16] Couture P, Brun LD, Szots F, et al. Association of specific LDL receptor genemutations with differential plasma lipoprotein response to simvastatin inyoung French Canadians with heterozygous familial hypercholesterolemia.Arterioscler Thromb Vasc Biol 1998;18:1007e12.

[17] Chaves FJ, Real JT, Garcia-Garcia AB, et al. Genetic diagnosis of familial hy-percholesterolemia in a South European outbreed population: influence oflow-density lipoprotein (LDL) receptor gene mutations on treatment responseto simvastatin in total, LDL, and high-density lipoprotein cholesterol. J ClinEndocrinol Metab 2001;86:4926e32.

[18] Bertolini S, Cantafora A, Averna M, et al. Clinical expression of familial hy-percholesterolemia in clusters of mutations of the LDL receptor gene thatcause a receptor-defective or receptor-negative phenotype. ArteriosclerThromb Vasc Biol 2000;20:E41e52.

[19] Santos PC, Gagliardi AC, Miname MH, et al. SLCO1B1 haplotypes are notassociated with atorvastatin-induced myalgia in Brazilian patients with fa-milial hypercholesterolemia. Eur J Clin Pharmacol 2012 Mar;68(3):273e9.

[20] Williams RR, Hunt SC, Schumacher MC, et al. Diagnosing heterozygous fa-milial hypercholesterolemia using new practical criteria validated by molec-ular genetics. Am J Cardiol 1993;72:171e6.

[21] Link E, Parish S, Armitage J, et al. SLCO1B1 variants and statin-inducedmyopathyea genomewide study. N Engl J Med 2008;359:789e99.

[22] Alonso R, Mata N, Castillo S, et al. Cardiovascular disease in familial hyper-cholesterolaemia: influence of low-density lipoprotein receptor mutationtype and classic risk factors. Atherosclerosis 2008;200:315e21.

[23] Bertolini S, Cassanelli S, Garuti R, et al. Analysis of LDL receptor gene muta-tions in Italian patients with homozygous familial hypercholesterolemia.Arterioscler Thromb Vasc Biol 1999;19:408e18.

[24] Hobbs HH, Brown MS, Goldstein JL. Molecular genetics of the LDL receptorgene in familial hypercholesterolemia. Hum Mutat 1992;1:445e66.

[25] Leigh SE, Foster AH, Whittall RA, Hubbart CS, Humphries SE. Update andanalysis of the University College London low density lipoprotein receptorfamilial hypercholesterolemia database. Ann Hum Genet 2008;72:485e98.

[26] Mozas P, Castillo S, Tejedor D, et al. Molecular characterization of familialhypercholesterolemia in Spain: identification of 39 novel and 77 recurrentmutations in LDLR. Hum Mutat 2004;24:187.

[27] Leitersdorf E, Eisenberg S, Eliav O, et al. Genetic determinants of respon-siveness to the HMG-CoA reductase inhibitor fluvastatin in patients withmolecularly defined heterozygous familial hypercholesterolemia. Circulation1993;87:III35e44.

[28] Mata N, Alonso R, Badimon L, et al. Clinical characteristics and evaluation ofLDL-cholesterol treatment of the Spanish Familial HypercholesterolemiaLongitudinal Cohort Study (SAFEHEART). Lipids Health Dis 2011;10:94.

[29] Vohl MC, Szots F, Lelievre M, et al. Influence of LDL receptor gene mutationand apo E polymorphism on lipoprotein response to simvastatin treatmentamong adolescents with heterozygous familial hypercholesterolemia.Atherosclerosis 2002;160:361e8.

[30] Alvim RO, Freitas SR, Ferreira NE, et al. APOE polymorphism is associated withlipid profile, but not with arterial stiffness in the general population. LipidsHealth Dis 2010;9:128.

[31] Niemi M, Pasanen MK, Neuvonen PJ. Organic anion transporting polypeptide1B1: a genetically polymorphic transporter of major importance for hepaticdrug uptake. Pharmacol Rev 2011;63:157e81.

[32] Santos PC, Alvim Rde O, Ferreira NE, et al. Ethnicity and arterial stiffness inBrazil. Am J Hypertens 2011;24:278e84.

[33] Santos PC, Soares RA, Nascimento RM, et al. SLCO1B1 rs4149056 poly-morphism associated with statin-induced myopathy is differently distributedaccording to ethnicity in the Brazilian general population: Amerindians as ahigh risk ethnic group. BMC Med Genet 2011;12:136.