Detecting familial hypercholesterolemia by serum lipid profile ...

0 downloads 0 Views 1MB Size Report
18 Jul 2017 - c Occupational Medicine Division, Garibaldi Hospital of Catania, Catania ...... Update and analysis of the University College London low density.
Accepted Manuscript Detecting familial hypercholesterolemia by serum lipid profile screening in a hospital setting: clinical, genetic and atherosclerotic burden profile Roberto Scicali, MD, Antonino Di Pino, MD, Roberta Platania, MD, Giacomo Purrazzo, MD, Viviana Ferrara, MD, Alberto Giannone, MD, Francesca Urbano, PhD, Agnese Filippello, PhD, Venerando Rapisarda, MD, Emanuele Farruggia, MD, Salvatore Piro, MD PhD, Agata Maria Rabuazzo, MD, Francesco Purrello, MD PII:

S0939-4753(17)30157-6

DOI:

10.1016/j.numecd.2017.07.003

Reference:

NUMECD 1754

To appear in:

Nutrition, Metabolism and Cardiovascular Diseases

Received Date: 18 May 2017 Revised Date:

4 July 2017

Accepted Date: 10 July 2017

Please cite this article as: Scicali R, Di Pino A, Platania R, Purrazzo G, Ferrara V, Giannone A, Urbano F, Filippello A, Rapisarda V, Farruggia E, Piro S, Rabuazzo AM, Purrello F, Detecting familial hypercholesterolemia by serum lipid profile screening in a hospital setting: clinical, genetic and atherosclerotic burden profile, Nutrition, Metabolism and Cardiovascular Diseases (2017), doi: 10.1016/ j.numecd.2017.07.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Detecting familial hypercholesterolemia by serum lipid profile screening in a hospital setting: clinical, genetic and atherosclerotic burden profile Roberto Scicali MD1, Antonino Di Pino MD1, Roberta Platania MD1, Giacomo Purrazzo

RI PT

MD1, Viviana Ferrara MD1, Alberto Giannone MD1, Francesca Urbano PhD1, Agnese Filippello PhD1, Venerando Rapisarda MD2, Emanuele Farruggia MD3, Salvatore Piro MD PhD1, Agata Maria Rabuazzo MD1, Francesco Purrello MD1.

2

Department of Clinical and Experimental Medicine, University of Catania, Italy.

Department of Clinical and Experimental Medicine, Section of Occupational Medicine,

SC

1

University of Catania, Italy.

Occupational Medicine Division, Garibaldi Hospital of Catania, Catania, Italy.

M AN U

3

Corresponding author:

Francesco Purrello, MD, Department of Clinical and Experimental Medicine, University of

TE D

Catania, Internal Medicine, Garibaldi Hospital, Via Palermo, 636 95122 Catania, Italy, Fax number: +39-0957598421, Phone number: +39-0957598401

AC C

EP

E-mail: [email protected]

1

ACCEPTED MANUSCRIPT

Abstract Background and Aims: Familial hypercholesterolemia (FH) is underdiagnosed and public cholesterol screening may be useful to find new subjects. In this study, we aim to investigate the

burden using intima-media thickness (IMT).

RI PT

prevalence of FH patients in a hospital screening program and evaluate their atherosclerotic

Methods and results: We screened 1575 lipid profiles and included for genetic analysis adults with a low-density lipoprotein (LDL) cholesterol > 190 mg/dL and triglycerides < 200 mg/dL and first-degree child relatives with LDL cholesterol > 160 mg/dL and triglycerides < 200 mg/dL.

SC

The diagnosis of FH was presumed by Dutch Lipid Clinic Network (DLCN) criteria and confirmed by the presence of the genetic variant. Mean common carotid intima media thickness

M AN U

(IMT) was assessed using consensus criteria. After confirming LDL cholesterol value and excluding secondary hypercholesterolemia, 56 subjects with a DLCN ≥ 4 performed genetic analysis. Of these, 26 had an FH genetic variant. The proportion of patients with a mutation having a DLCN score of 6–8 was 75 %; in individuals with a DLCN score > 8 it was 100 %. Mean IMT was higher in FH patients compared to non FH (0.73 [0.61–0.83] vs 0.71 [0.60–0.75] mm, p < 0.01). Moreover, we detected two mutations not previously described. Finally, simple

TE D

regression analysis showed a correlation of IMT with LDL cholesterol > 190 mg/dL and corneal arcus (p < 0.01 and p < 0.001, respectively).

Conclusions: A hospital screening was useful to detect FH subjects with increased

AC C

EP

atherosclerosis. Also, next-generation sequencing was able to detect new FH mutations.

2

ACCEPTED MANUSCRIPT

Keywords:

Low-density

Lipoprotein

Cholesterol,

Screening

Program,

Familial

AC C

EP

TE D

M AN U

SC

RI PT

Hypercholesterolemia, Cardiovascular Risk Assessment, Intima Media Thickness

3

ACCEPTED MANUSCRIPT

Introduction Familial hypercholesterolemia (FH) is the most frequent monogenetic disease characterized by autosomal inheritance disorders in genes related to low-density lipoprotein (LDL) cholesterol

RI PT

metabolism that results in lifelong elevation of LDL cholesterol and premature atherosclerotic cardiovascular disease (CVD)[1]. Different mutations in genes encoding key proteins in the LDL receptor pathways are involved, leading to decreased cellular uptake of LDL and increased plasma LDL cholesterol concentrations[2]. FH is defined both as an autosomal and recessive/dominant disease. Autosomal dominant hypercholesterolemia (ADH) is caused by

SC

mutations in low-density lipoprotein receptors (LDLR), apolipoprotein B (ApoB) and the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene[3]. Autosomal recessive

M AN U

hypercholesterolemia (ARH) is caused by mutations in the LDL receptor adapter protein 1 (LDLRAP1) gene[4]. Traditionally, the prevalence of the heterozygous (He) FH phenotype in the general population has been reported as being 1 in 500 and the prevalence of the homozygous (Ho) FH form is estimated to be 1 in 1 million[5]. However, a recent screening of 98,098 individuals from the Copenhagen General Population Study evaluated the prevalence of individuals classified with probable or definite FH as 1 in 217 using the Dutch Lipid Clinic

TE D

Network (DLCN) score (DLCN score > 5 points) or an LDL cholesterol above 4.4 mmol/L (170 mg/dL) was ~1/200[6]. Based on these data, in Italy there should be ~ 300–350.000 subjects with heterozygous FH[7] with most of them without a diagnosis and treatment[8]. Identification of these subjects may have relevant clinical implications because if left untreated, the risk of

EP

premature coronary artery disease (CAD) is significantly higher in heterozygous patients than unaffected individuals[9].

AC C

According to these considerations, public cholesterol screening is essential to find individuals with FH who should be included in cardiovascular risk prevention programs[10]. General practitioners (GPs) request over 90% of LDL cholesterol measurements in the general population and might also have a critical role in detecting patients with FH[11]. Furthermore, workplace cholesterol screening programs may be useful to detect subjects at high risk for cardiovascular disease; indeed, occupational medicine services are periodically in contact with all employees, including persons not visiting GPs regularly. Moreover, a workplace cholesterol screening program would be useful to detect subjects at high risk for cardiovascular disease[12].

4

ACCEPTED MANUSCRIPT

In this study, we aim to investigate the prevalence of patients with FH in a hospital cholesterol screening program and evaluate their atherosclerotic burden using intima-media

AC C

EP

TE D

M AN U

SC

RI PT

thickness (IMT) as a macroangiopathic biomarker.

5

ACCEPTED MANUSCRIPT

Methods Study design and Population. We evaluated serum lipid profiles from GP screened subjects undergoing lipid management at the Day Service of the Internal Medicine Division and in health care workers referred to the

RI PT

Occupational Medicine Division at the Hospital Garibaldi in Catania, Italy from 12th January 2015 to 27th January 2017. A total of 1575 lipid profiles were screened. Inclusion criteria for genetic evaluation were: adults with an LDL cholesterol > 190 mg/dL and triglycerides < 200 mg/dL and first-degree child relatives with an LDL cholesterol > 160 mg/dL and triglycerides
325 mg/dL (8 points). For child probands, LDL cholesterol value was modified

AC C

as an LDL ≤ 160 mg/dL (1 point), 161–250 mg/dL (3 points), 251–325 mg/dL (5 points) and > 325 mg/dL (8 points). Using the DLCN criteria, a diagnosis of FH was presumed definite if the total score was > 8, probable if the score was 6–8, possible if the score was 3–5. The diagnosis of FH was confirmed by the presence of a genetic variant. Mononuclear cell isolation was performed for genetic analysis. The study protocol was approved by the Institutional Review Board of the Department and conformed to the ethical guidelines of the 1975 Helsinki Declaration. Informed written consent was obtained from each patient before entering the study, and they did not receive any monetary payment. IMT measurements were performed in genetically evaluated patients without history of CVD. Arterial

6

ACCEPTED MANUSCRIPT

hypertension was considered present when the measurement of brachial blood pressure (BP) exceeded 140 mm Hg (systolic) and/or 90 mm Hg (diastolic) on at least two different occasions, or if the patient was on antihypertensive medication. Statin therapy was defined as a daily intake of statins. For the diagnostic classification, an LDL cholesterol correction factor table was used to

RI PT

estimate the pretreated LDL cholesterol levels of patients on cholesterol lowering medication as previously reported[14]. Current smoking habits were divided into either current smoking (defined as any number of cigarettes in the previous month), or not current smoking. CVD was considered as documented by previous myocardial infarction (MI), acute coronary syndrome

SC

(ACS), coronary revascularization (percutaneous coronary intervention (PCI), coronary artery bypass graft surgery (CABG) or other arterial revascularization procedures, stroke or transient

M AN U

ischemic attack (TIA), or peripheral arterial disease (PAD).

Patients were classified into 2 groups according to molecular genetic testing results: 1) non FH group (individuals without genetic variant, 30 subjects), 2) FH group (individuals with genetic variant found, 26 subjects). Measurements

Serum total cholesterol, triglycerides and high-density lipoprotein (HDL) cholesterol were

TE D

measured by currently available enzymatic methods as previously described[15]. LDL cholesterol was calculated using the Friedewald formula. During physical examination, brachial BP measurements were performed under standardized conditions. Mononuclear cell (MNC) isolation

EP

MNCs were isolated from blood samples (10–12 ml) using lympholyte medium (LymphoprepTM, Stemcell Technologies) according to the manufacturer’s instructions.

AC C

Carotid ultrasound examinations

Ultrasound scans were performed using a high-resolution B-mode ultrasound system (MyLab 50 Xvision; Esaote Biomedica SpA, Florence, Italy) equipped with a 7.5-MHz linear array transducer. Scans were performed by a single physician, as previously described[16]. Mean IMT was defined for each individual as the average of the right and the left common carotid IMT.

Genetic analysis

7

ACCEPTED MANUSCRIPT

Genetic analysis was performed at the Center for the Study of Atherosclerosis at Bassini Hospital (Milan), as part of the “LIPIGEN” project of the Italian Society for the Study of Atherosclerosis (SISA). Genomic DNA was extracted from MNCs. Progenika’s Familial Hypercholesterolemia

RI PT

Genetic Analysis (SEQPRO LIPO) detected variants associated with Autosomal Dominant Hypercholesterolemia (ADH) and Autosomal Recessive Hypercholesterolemia (ARH).

Specifically, the assay detected substitutions and indels in exons and intron-exon boundaries of the following genes: LDLR (18 exons), ApoB (regions of exons 26 and 29

SC

involved in LDLR binding), PCSK9 (12 exons) and LDLRAP1 (9 exons). The assay also detected Copy Number Variations (CNV) in the LDLR gene associated with FH. Elaborate

M AN U

proprietary algorithms were used to detect the presence of genomic deletions and duplications in all 18 exons and the promoter region in LDLR. A full analysis of the genes by next generation sequencing (NGS) was performed to detect index cases. A specific genetic Capillary Electrophoresis sequencing (for point mutations detection) or Multiplex Ligation-dependent Probe Amplification (MLPA) analysis (for CNV variants) was performed to detect familial cases. Fifty-eight regions of interest (20Kb) were amplified from genomic DNA, isolated from MNCs,

TE D

in five different types of Polymerase Chain Reaction (PCR). A second amplification was performed in order to include an 8nt index sequence used for sample identification, as well as the adaptors used for sequencing on MiSeq (Illumina) equipment[17]. The PCR master mixes included highly purified primers for the specific amplification of regions in which mutations

EP

causing ADH and ARH can be found, as well as internal controls on chromosome 21. DNA variants were identified and classified by comparison with reference sequence and Progenika’s extensive database. When a variant related to ADH or ARH was identified, the report indicated

AC C

the specific mutation found; the gene, exon, nucleotide, and amino acid position; and pathogenicity status (pathogenic, probably pathogenic, or possibly pathogenic). To confirm probably or possibly pathogenic status of the novel genetic variants, the in silico prediction of the LDLR, APOB and PCSK9 gene missense mutation effect was performed using PolyPhen-2 (http://genetics.bwh.harvard.edu/pph2/) and SIFT Human Protein (http://sift.jcvi.org/) refined SIFT[18]. The SEQPRO LIPO report indicated whether the specific mutation found was homozygous or heterozygous. Heterozygous FH is considered a heterozygous mutation in LDLR, APOB or PCSK9 genes. Homozygous FH results from homozygous mutations in LDLR or

8

ACCEPTED MANUSCRIPT

LDLRAP1 genes. Compound heterozygous FH is considered as compound mutations in either the LDLR or ARH genes. Double heterozygous FH results from the presence of a mutation in two of the four assayed genes. In cases where hypocholesterolemic variants were identified, they were also reported.

RI PT

Heterozygous pathogenic, possibly pathogenic or probably pathogenic mutations found in

AC C

EP

TE D

M AN U

SC

LDLRAP1 were reported to give information about the carrier status of the patient.

9

ACCEPTED MANUSCRIPT

Statistical Analysis The distributional characteristics of each variable, including normality, were assessed by the Kolmogorov-Smirnov test. Descriptive analysis of the general population was performed and

RI PT

subject characteristics were expressed as means ± standard deviation (SD) or frequencies. Descriptive analysis of the FH subgroup was performed and subject characteristics were expressed as means ± standard deviation (SD) or frequencies. When necessary, numerical variables were logarithmically transformed for statistical analysis to reduce skewness (triglycerides and IMT mean) and values were expressed as median and interquartile range. To

SC

compare qualitative variables, the χ2 for qualitative variables were used. Then, to test differences in clinical and biochemical characteristics among the two groups we used Student’s t test. Mean

M AN U

IMT was further studied for its association with DCLN criteria (family history, physical examination, LDL cholesterol value, molecular genetic testing). For this purpose, family history was analyzed as a dicotomic variable (yes or no). Moreover, the physical examination was evaluated as two separate dicotomic variables: corneal arcus (presence or not) and tendon xanthomas (presence or not). Finally, also the LDL cholesterol value was analyzed as a dicotomic variable with an LDL cut-off value > 190 mg/dL (yes or no). Prior to multivariate analysis,

TE D

variance inflation due to covariates was verified by estimating a variance inflation factor < 2. All statistical analyses were performed using IBM SPSS Statistics for Windows version 23. For all

AC C

EP

tests, p < 0.05 was considered significant.

10

ACCEPTED MANUSCRIPT

Results A total of 1575 lipid profiles were screened; of these, 1450 were excluded because they did not meet inclusion criteria. A total of 125 patients were considered. In forty-one subjects it was

RI PT

not possible to obtain a second blood sample to confirm the LDL cholesterol value and evaluate hypercholesterolemia. In eighty-four patients we obtained a second blood sample and also a clinical and physical examination. Twenty-eight subjects with secondary hypercholesterolemia were excluded. Finally, 56 patients with a DCLN score ≥ 4 were selected to perform genetic analysis (Figure 1).

SC

Among the 56 subjects, 26 (15 index cases and 11 relatives) had a genetic variant establishing diagnosis of FH. The genetic and phenotypic characteristics of FH subjects are

M AN U

presented in Table 1. The most frequent genetic variants were LDLR mutations (87.1 %) and the majority of these were pathogenic. Moreover, 1/3 of LDLR genetic variants were possibly pathogenic. The in silico analysis showed that all the possibly pathogenic mutations had a pathogenic effect. No possibly pathogenic mutations were found in APOB and PCSK9 gene. Twenty-one (80.8 %) subjects were heterozygous FH, 3 (11.5 %) were double heterozygous FH and 2 were compound heterozygous FH.

TE D

The descriptiption of genetic and proteic variants is reported in Table 2. While the most frequent LDLR pathogenic variant was c.352G>T (p.Asp118Tyr ), c.1118G>A was the most present LDLR possibly pathogenic mutation. Also, two possibly pathogenic variants c.788A>G (p.Asp263Gly) and c.2120A>C (p.Asp707Ala) were not previously reported. The first novel

EP

variant was detected in one index case and relative. The second new mutation was found in one subject and the cascade screening is in progress.

AC C

Figure 2 shows the proportions of patients with a genetic mutation according to their DLCN score. All subjects had a DLCN ≥ 4. The proportion of patients with a detected mutation having a DLCN score of 4–5 was 17.6 %; in subjects with a DLCN score of 6–8 it was 75 %. Moreover, the percentage of genetically confirmed FH individuals was 100 % in subjects with a DLCN score > 8.

The general characteristics of the study subjects are presented in Table 3. FH patients were younger than non FH subjects (42.7 ± 16.85 vs 50.75 ± 12.18, p = 0.04). Also, body mass index was lower in the FH group than in the non FH group (23.96 ± 4.24 vs 26.82 ± 3.63, p = 0.02). The percentage of men was similar between the two groups. While subjects without genetic

11

ACCEPTED MANUSCRIPT

variant did not exhibit corneal arcus, 13 FH patients (50 %) had corneal arcus. None of the patients exhibited tendon xanthomas. FH patients had higher total cholesterol, triglycerides, LDL-C and non-HDL cholesterol compared to not FH subjects. While the proportion of LDL-C < 251 mg/dL was higher in the non FH group, the FH group exhibited a higher percentage of LDL

RI PT

> 250 mg/dL. Concerning risk factors, the FH group exhibited lower systolic and diastolic blood pressure than the non FH group, but the difference was not statistically significant. Only 10 (38.5 %) FH patients were on statin therapy. However, the percentage of patients on statin and antihypertensive therapy was similar between the two groups.

SC

To evaluate the atherosclerotic burden of the study population, IMT measurement was performed in 48 individuals without history of CVD: 25 non FH subjects and 23 FH patients.

M AN U

Mean IMT (Figure 3) was significantly higher in FH patients compared to non FH subjects (0.73 [0.61–0.83] vs 0.71 [0.60–0.75] mm, p < 0.01)

In the simple regression analysis between mean IMT and DLCN determinants, mean IMT was correlated to LDL cholesterol > 190 mg/dL (r = 0.377, p < 0.01) and corneal arcus (r =

AC C

EP

TE D

0.480, p < 0.001).

12

ACCEPTED MANUSCRIPT

Discussion In this study, we investigated the prevalence of patients with FH in a hospital cholesterol screening program. Furthermore, we evaluated their atherosclerotic burden using intima-media

RI PT

thickness (IMT) as a macroangiopathic biomarker. We found 26 (46.4% of subjects underwent genetic evaluation) subjects with FH defined by clinical diagnosis and genetic analysis. Also, we found two novel LDLR mutations not previously described

in literature. Moreover, in a

subgroup of individuals without CVD, we found that FH subjects exhibited higher mean IMT compared to non FH individuals. Finally, we demonstrated that LDL > 190 mg/dL and corneal

SC

arcus were correlated with IMT.

Primary prevention is the most effective strategy for reducing the burden of cardiovascular

M AN U

disease in FH individuals[19]. Thus, sequential and combined screening programs for detecting new FH cases are recommended in clinical practice[20]. In our study, we showed a useful multidisciplinary approach including GPs, internal and preventive medicine specialists to find new FH patients. Vickery et al. have already reported that preventive specialists and GPs are central to improve diagnosis of FH in primary care settings[21].

In our study, we showed that DLCN score was a simple but very sensible method to select

TE D

probably FH patients undergoing to genetic analysis in a large population. Also, we confirmed that a DLCN ≥ 6 had high sensibility in our population. We found that 90.5% of subjects with a DLCN score ≥ 6 had a genetic mutations. Our reports are comparable to preliminar findings by

was 92%[22].

EP

Casula et al where the proportion of subjects with a DLCN score ≥ 6 having a genetic variant

In our reports, we found that a diagnostic algorithm by using DLCN score and genetic

AC C

testing was useful to detect new FH subjects. Previous studies evaluated this combined approach in a hospital setting. Maglio et al found that the percentage of individuals with an FH genetic mutation was 65%[23]. In our population, 46.4% of subjects had a detected FH causing mutation. The disagreement between the two findings may be explained by the different prevalence of subjects with a probable or definite FH according to DLCN score. While in the study by Maglio et al 74% of the population was classified with probable or definite FH, in our reports the percentage of subjects with probable or definite FH was only 37.5%. In our study population, the percentage of index cases of FH among subjects undergoing genetic analysis was 26.8 %. This is not the first study exploring the systematic detection of familiar hypercholesterolemia in primary

13

ACCEPTED MANUSCRIPT

health care. Kirke et al. reported a lower percentage of FH index cases (18.6%) in a study testing three methods of case detection (pathology laboratory database search, workplace health checks and general practice database search)[24]. The discordance between the two studies may be found in the different recruitment modalities. Kirke et al. performed genetic evaluation only in

RI PT

individuals with a DLCN score > 5; in contrast, in this study, we considered as candidates for genetic analysis also subjects with a DLCN score ≥ 4. This consideration is supported by the observation that, in our study, the proportion of patients with a detected FH causing mutation was 17.6 % among subjects with a DLCN score of 4–5. Our reports are comparable to findings by

SC

Grenkowitz et al. where the proportion of subjects with an FH genetic mutation was 22.7 % among individuals with a DLCN score of 3–5[25]. Furthermore, we showed that 87.1 % of the

M AN U

FH causing mutations were found in the LDLR gene; these results are in line with previous reports showing alterations of LDLR as about 90 % of the overall mutations detected[26]. In this study, a full analysis of the genes by NGS was able to detect two novel LDLR mutations not previously described. Previously, Alves et al reported by NGS two novel genetic variants in APOB gene[27]. In this context, the combination of diagnostic criteria and nextgeneration sequencing technology could be used for FH diagnosis in clinical setting[28].

TE D

FH is a common cause of premature CAD[29]. If left untreated, heterozygous FH patients typically develop CAD before age 55. In our population, the percentage of FH patients on statin therapy was only 38.5 %. Other studies underlined the very high risk of CAD due to undertreatment of FH in the community and primary care. Benn et al. have shown at least half of

EP

FH affected subjects are not receiving cholesterol-lowering medication. Furthermore, they have reported a very high risk for CAD in patients without cholesterol-lowering medication [odds ratio13.2 (10.0-17.4)] in definite/probable FH compared with non-FH subjects, after adjusting for

AC C

CV risk factors. The corresponding adjusted odds ratio for coronary artery disease was significantly lower in FH subjects on cholesterol-lowering medication [odds ratio 10.3 (7.813.8)][30].

Concerning carotid IMT, our results are in line with earlier reports showing an increased IMT in FH subjects without prior history of CVD[31,32]. Previous studies have reported similar findings: Junyent M et al. have shown advanced carotid atherosclerosis in FH patients in relation to LDLR mutational class supporting the utility of genetic testing in FH beyond providing a secure

diagnosis.

Furthermore,

a

retrospective

study

in

patients

with

familial

14

ACCEPTED MANUSCRIPT

hypercholesterolemia mutations showed that carotid IMT is the simpler and cheaper technique for detection of atheroma in younger patients. It can also be performed in primary care[33,34]. In our study, we showed a significant association between IMT and both LDL cholesterol > 190 mg/dL and corneal arcus. Corneal arcus is a lipid-rich deposit at the corneoscleral limbus

RI PT

that shares some similarities with the lipid deposition of atherosclerosis[35]. However, epidemiologic studies examining the association between corneal arcus and coronary heart disease (CHD) have yielded mixed results. Rouhiainen et al. and other authors found a significant association between IMT and corneal arcus in hypercholesterolemic patients[36,37]. Other

SC

authors did not support this hypothesis and reported that corneal arcus predicts CVD and CHD in the community-based Framingham Heart Study cohort due to the strong association of corneal

M AN U

arcus with increasing age[38]. A possible explanation to support the association of IMT and corneal arcus may be that the last one could reflect a continuous exposure to high LDL cholesterol levels. So, corneal arcus may represent a sign of cholesterol burden. Recently, Gallo et al showed that early coronary calcifications were correlated to cholesterol burden in HeFH subjects without history of CVD[39].

There are several limitations to our study. First, population size was relatively small but we

TE D

were still able to show significant differences in the two groups and independent association of IMT and corneal arcus. Also, we did not know the percentages of patients on statin therapy among screened lipid profiles with an LDL-C < 191 mg/dL. Finally, multivariate models used in this study for evaluating the variability of IMT may have been better explained by the fact that

EP

other confounding factors such as diet or physical activity were not taken into consideration. In conclusion, our study showed that an integrated approach including GPs, internal and preventive medicine specialists was useful to detect subjects with FH in a hospital setting.

AC C

Furthermore, these patients seem to exhibit an increased carotid atherosclerosis compared to non FH subjects. Moreover , the combination of diagnostic criteria and genetic analysis could be used for FH diagnosis in clinical setting. Finally, next-generation sequencing technology was useful to detect new FH mutations. The clinical impact of these novel mutations may be explored in future studies.

15

ACCEPTED MANUSCRIPT

Acknowledgments We wish to thank the Lipid TransPort Disorders Italian Genetic Network (LIPIGEN) study for financial support to carry out genetic analysis.

RI PT

We wish to thank the Scientific Bureau of the University of Catania for language support. Funding

or not-for-profit sectors.

Conflict of interest

AC C

EP

TE D

M AN U

The authors have no conflicts of interest to disclose.

SC

This research did not receive any specific grant from funding agencies in the public, commercial,

16

ACCEPTED MANUSCRIPT

References [1]

Hovingh GK, Davidson MH, Kastelein JJP, O’Connor AM. Diagnosis and treatment of familial hypercholesterolaemia. Eur Heart J 2013;34:962–71.

[2]

Goldstein JL, Brown MS. The LDL receptor. Arterioscler Thromb Vasc Biol 2009;29:431–8. doi:10.1161/ATVBAHA.108.179564.

[3]

RI PT

doi:10.1093/eurheartj/eht015.

Austin MA, Hutter CM, Zimmern RL, Humphries SE. Genetic causes of monogenic

2004;160:407–20. doi:10.1093/aje/kwh236. [4]

SC

heterozygous familial hypercholesterolemia: a HuGE prevalence review. Am J Epidemiol

Garcia CK, Wilund K, Arca M, Zuliani G, Fellin R, Maioli M, et al. Autosomal recessive

M AN U

hypercholesterolemia caused by mutations in a putative LDL receptor adaptor protein. Science 2001;292:1394–8. doi:10.1126/science.1060458. [5]

Liyanage KE, Burnett JR, Hooper AJ, van Bockxmeer FM. Familial hypercholesterolemia: epidemiology, Neolithic origins and modern geographic distribution. Crit Rev Clin Lab Sci 2011;48:1–18. doi:10.3109/10408363.2011.565585.

[6]

Benn M, Watts GF, Tybjærg-Hansen A, Nordestgaard BG. Mutations causative of familial

TE D

hypercholesterolaemia: screening of 98 098 individuals from the Copenhagen General Population Study estimated a prevalence of 1 in 217. Eur Heart J 2016;37:1384–94. doi:10.1093/eurheartj/ehw028. [7]

Cefalù AB, Emmanuele G, Marino G, Fiore B, Caldarella R, Vivona N, et al.

EP

Effectiveness of screening for known mutations in Sicilian patients with "probable" familial hypercholesterolemia. Nutr Metab Cardiovasc Dis

[8]

AC C

2001;11:394–400.

Nordestgaard BG, Chapman MJ, Humphries SE, Ginsberg HN, Masana L, Descamps OS, et al. Familial hypercholesterolaemia is underdiagnosed and undertreated in the general population: guidance for clinicians to prevent coronary heart disease: Consensus Statement of the European Atherosclerosis Society. Eur Heart J 2013;34:3478–90. doi:10.1093/eurheartj/eht273.

[9]

De Backer G, Besseling J, Chapman J, Hovingh GK, Kastelein JJP, Kotseva K, et al. Prevalence and management of familial hypercholesterolaemia in coronary patients: An analysis of EUROASPIRE IV, a study of the European Society of Cardiology. 17

ACCEPTED MANUSCRIPT

Atherosclerosis 2015;241:169–75. doi:10.1016/j.atherosclerosis.2015.04.809. [10]

Mata P, Alonso R, Pérez-Jiménez F. Screening for Familial Hypercholesterolemia: a Model for Preventive Medicine. Rev Española Cardiol (English Ed 2014;67:685–8. doi:10.1016/j.rec.2014.01.015. Bell DA, Hooper AJ, Bender R, McMahon J, Edwards G, van Bockxmeer FM, et al.

RI PT

[11]

Opportunistic screening for familial hypercholesterolaemia via a community laboratory. Ann Clin Biochem 2012;49:534–7. doi:10.1258/acb.2012.012002. [12]

Prior JO, van Melle G, Crisinel A, Burnand B, Cornuz J, Darioli R. Evaluation of a

SC

multicomponent worksite health promotion program for cardiovascular risk factors-

correcting for the regression towards the mean effect. Prev Med (Baltim) 2005;40:259–67.

[13]

M AN U

doi:10.1016/j.ypmed.2004.05.032.

Haase A, Goldberg AC. Identification of people with heterozygous familial hypercholesterolemia. Curr Opin Lipidol 2012;23:282–9. doi:10.1097/MOL.0b013e3283556c33.

[14]

Haralambos K, Whatley SD, Edwards R, Gingell R, Townsend D, Ashfield-Watt P, et al. Clinical experience of scoring criteria for Familial Hypercholesterolaemia (FH) genetic

TE D

testing in Wales. Atherosclerosis 2015;240:190–6. doi:10.1016/j.atherosclerosis.2015.03.003. [15]

Calanna S, Scicali R, Di Pino A, Knop FK, Piro S, Rabuazzo AM, et al. Lipid and liver abnormalities in haemoglobin A1c-defined prediabetes and type 2 diabetes. Nutr Metab

[16]

EP

Cardiovasc Dis 2014;24:670–6. doi:10.1016/j.numecd.2014.01.013. Di Pino A, Scicali R, Calanna S, Urbano F, Mantegna C, Rabuazzo AM, et al. Cardiovascular risk profile in subjects with prediabetes and new-onset type 2 diabetes

AC C

identified by HbA1c according to American Diabetes Association criteria. Diabetes Care 2014;37:1447–53. doi:10.2337/dc13-2357. [17]

Futema M, Plagnol V, Whittall RA, Neil HAW, Humphries SE, Humphries SE, et al. Use of targeted exome sequencing as a diagnostic tool for Familial Hypercholesterolaemia. J Med Genet 2012;49:644–9. doi:10.1136/jmedgenet-2012-101189.

[18]

Leigh SEA, Foster AH, Whittall RA, Hubbart CS, Humphries SE. Update and Analysis of the University College London Low Density Lipoprotein Receptor Familial Hypercholesterolemia Database. Ann Hum Genet 2008;72:485–98. doi:10.1111/j.1469-

18

ACCEPTED MANUSCRIPT

1809.2008.00436.x. [19]

Vallejo-Vaz AJ, Kondapally Seshasai SR, Cole D, Hovingh GK, Kastelein JJP, Mata P, et al. Familial hypercholesterolaemia: A global call to arms. Atherosclerosis 2015;243:257– 9. doi:10.1016/j.atherosclerosis.2015.09.021. Kindt I, Mata P, Knowles JW. The role of registries and genetic databases in familial hypercholesterolemia. Curr Opin Lipidol 2017;28:152–60. doi:10.1097/MOL.0000000000000398.

[21]

RI PT

[20]

Vickery AW, Bell D, Garton-Smith J, Kirke AB, Pang J, Watts GF. Optimising the

SC

detection and management of familial hypercholesterolaemia: central role of primary care and its integration with specialist services. Heart Lung Circ 2014;23:1158–64.

[22]

M AN U

doi:10.1016/j.hlc.2014.07.062.

Casula M, Averna M, Arca M, Bertolini S, Tarugi P, Calandra S, et al. The LIPIGEN (LIPid transport disorders Italian Genetic Network) study. G Ital dell’Arteriosclerosi 2017;8:71–80.

[23]

Maglio C, Mancina RM, Motta BM, Stef M, Pirazzi C, Palacios L, et al. Genetic diagnosis of familial hypercholesterolaemia by targeted next-generation sequencing n.d.

[24]

TE D

doi:10.1111/joim.12263.

Kirke AB, Barbour RA, Burrows S, Bell DA, Vickery AW, Emery J, et al. Systematic detection of familial hypercholesterolaemia in primary health care: a community based prospective study of three methods. Heart Lung Circ 2015;24:250–6.

[25]

EP

doi:10.1016/j.hlc.2014.09.011.

Grenkowitz T, Kassner U, Wühle-Demuth M, Salewsky B, Rosada A, Zemojtel T, et al. Clinical characterization and mutation spectrum of German patients with familial

AC C

hypercholesterolemia. Atherosclerosis 2016;253:88–93. doi:10.1016/j.atherosclerosis.2016.08.037. [26]

Bertolini S, Pisciotta L, Rabacchi C, Cefalù AB, Noto D, Fasano T, et al. Spectrum of mutations and phenotypic expression in patients with autosomal dominant hypercholesterolemia identified in Italy. Atherosclerosis 2013;227:342–8. doi:10.1016/j.atherosclerosis.2013.01.007.

[27]

Alves AC, Etxebarria A, Soutar AK, Martin C, Bourbon M. Novel functional APOB mutations outside LDL-binding region causing familial hypercholesterolaemia. Hum Mol

19

ACCEPTED MANUSCRIPT

Genet 2014;23:1817–28. doi:10.1093/hmg/ddt573. [28]

Vandrovcova J, Thomas ERA, Atanur SS, Norsworthy PJ, Neuwirth C, Tan Y, et al. The use of next-generation sequencing in clinical diagnosis of familial hypercholesterolemia. Genet Med 2013;15:948–57. doi:10.1038/gim.2013.55. Pang J, Poulter EB, Bell DA, Bates TR, Jefferson V-L, Hillis GS, et al. Frequency of

RI PT

[29]

familial hypercholesterolemia in patients with early-onset coronary artery disease admitted to a coronary care unit. J Clin Lipidol 2015;9:703–8. doi:10.1016/j.jacl.2015.07.005. [30]

Benn M, Watts GF, Tybjaerg-Hansen A, Nordestgaard BG. Familial hypercholesterolemia

SC

in the danish general population: prevalence, coronary artery disease, and cholesterollowering medication. J Clin Endocrinol Metab 2012;97:3956–64. doi:10.1210/jc.2012-

[31]

M AN U

1563.

Hjuler Nielsen M, Irvine H, Vedel S, Raungaard B, Beck-Nielsen H, Handberg A. Elevated atherosclerosis-related gene expression, monocyte activation and microparticlerelease are related to increased lipoprotein-associated oxidative stress in familial hypercholesterolemia. PLoS One 2015;10:e0121516. doi:10.1371/journal.pone.0121516.

[32]

Plana N, Ferré R, Merino J, Aragonès G, Girona J, Heras M, et al. Heterozygous familial

TE D

hypercholesterolaemic patients have increased arterial stiffness, as determined using the augmentation index. J Atheroscler Thromb 2011;18:1110–6. [33]

Junyent M, Gilabert R, Jarauta E, Núñez I, Cofán M, Civeira F, et al. Impact of lowdensity lipoprotein receptor mutational class on carotid atherosclerosis in patients with

EP

familial hypercholesterolemia. Atherosclerosis 2010;208:437–41. doi:10.1016/j.atherosclerosis.2009.07.058. [34]

Clarke REJ, Padayachee ST, Preston R, McMahon Z, Gordon M, Graham C, et al.

AC C

Effectiveness of alternative strategies to define index case phenotypes to aid genetic diagnosis of familial hypercholesterolaemia. Heart 2013;99:175–80. doi:10.1136/heartjnl2012-302917. [35]

Zech LA, Hoeg JM. Correlating corneal arcus with atherosclerosis in familial hypercholesterolemia. Lipids Health Dis 2008;7:7. doi:10.1186/1476-511X-7-7.

[36]

Rouhiainen P, Salonen R, Rouhiainen H, Salonen JT. Association of corneal arcus with ultrasonographically assessed arterial wall thickness and serum lipids. Cornea 1993;12:142–5.

20

ACCEPTED MANUSCRIPT

[37]

Ogura M, Hori M, Harada-Shiba M. Association Between Cholesterol Efflux Capacity and Atherosclerotic Cardiovascular Disease in Patients With Familial Hypercholesterolemia. Arterioscler Thromb Vasc Biol 2016;36:181–8. doi:10.1161/ATVBAHA.115.306665.

[38]

Fernandez AB, Keyes MJ, Pencina M, D’Agostino R, O’Donnell CJ, Thompson PD.

RI PT

Relation of corneal arcus to cardiovascular disease (from the Framingham Heart Study data set). Am J Cardiol 2009;103:64–6. doi:10.1016/j.amjcard.2008.08.030.

Gallo A, Giral P, Carrié A, Carreau V, Béliard S, Bittar R, et al. Early coronary calcifications are related to cholesterol burden in heterozygous familial

EP

TE D

M AN U

SC

hypercholesterolemia. J Clin Lipidol 2017;11:704–711.e2. doi:10.1016/j.jacl.2017.03.016.

AC C

[39]

21

ACCEPTED MANUSCRIPT

Table 1. Genetic and Phenotypic Characteristics of FH Subjects Total (n = 26)

RI PT

31 (100) 21 (67.7) 8 (25.8) 2 (6.5)

SC

19 (61.3) 8 (25.8) 1 (3.2) 1 (3.2) 2 (6.5)

15 (57.7) 11 (42.3) 21 (80.8) 3 (11.5) 2 (7.7)

M AN U

Molecular Genetic Testing Genetic variant, n (%) Pathogenic genetic variant, n (%) Possibly pathogenic genetic variant, n (%) Hypocholesterolemic genetic variant, n (%) Genetic variant pathogenicity LDLR pathogenic variant, n (%) LDLR possibly pathogenic variant, n (%) APOB pathogenic variant, n (%) PCSK9 pathogenic variant, n (%) PCSK9 hypocholesterolemic variant, n (%) FH case Index, n (%) Relative, n (%) FH phenotype Heterozygous FH, n (%) Double heterozygous FH, n (%) Compound heterozygous FH, n (%)

Table 1. Genetic and Phenotypic Characteristics of FH Subjects. Data are presented as percentages.

FH = familial hypercholesterolemia, LDLR = low-density lipoprotein receptor, APOB =

AC C

EP

TE D

apolipoprotein B, PCSK9 = proprotein convertase subtilisin/kexin type 9.

22

ACCEPTED MANUSCRIPT

Table 2. Genetic Analysis Description of the FH Subjects. Proteic Variant

Mutation Class

Pathogenicity

31 (100) 9 (29.0) 3 (9.7) 3 (9.7) 4 (12.9) 2 (6.5) 2 (6.5) 1 (3.2) 1 (3.2) 1 (3.2) 1 (3.2) 1 (3.2) 2 (6.5) 1 (3.2)

LDLR LDLR LDLR LDLR LDLR LDLR LDLR LDLR LDLR LDLR APOB PCSK9 PCSK9

4 8 4 8 5 1 1 11 9 14 26 1 -

p.Asp118Tyr p.Tyr375Trpfs*7 p.Glu140* p.Gly373Asp p.Asp263Gly p.Gly20Arg p.Ala540Thr p.Thr404Ile p.Asp707Ala p.Arg3527Gln p.Arg46Leu -

Amino acid change Null allele Null allele Amino acid change Amino acid change Null Allele Amino acid change Amino acid change Amino acid change Amino acid change Amino acid change Amino acid change Promoter

Pathogenic Pathogenic Pathogenic Possibly Pathogenic Possibly Pathogenic Pathogenic Pathogenic Pathogenic Possibly Pathogenic Possibly Pathogenic Pathogenic Hypocholesterolemic Pathogenic

RI PT

Exon

SC

c.352G>T c.1120_1123dupGGCT c.418G>T c.1118G>A c.788A>G c.1-?_67+?del# c.58G>A c.1618G>A c.1211C>T c.2120A>C c.10580G>A c.137G>T c.(-331)C>A

Gene

M AN U

Genetic variant

Number (%)

Table 2. Genetic Analysis Description of the FH Subjects. Data are presented as percentages. #

Deletion of promoter and exon 1.

TE D

FH = familial hypercholesterolemia, LDLR = low-density lipoprotein receptor, APOB =

AC C

EP

apolipoprotein B, PCSK9 = proprotein convertase subtilisin/kexin type 9.

23

ACCEPTED MANUSCRIPT

Table 3. General Characteristics of the Study Population FH group (genetic variant)

56 47.01 ± 14.96 27 (48.2) 25.5 ± 4.14

30 50.75 ± 12.18 16 (53.3) 26.82 ± 3.63

26 42.7 ± 16.85 11 (42.3) 23.96 ± 4.24

13 (23.2) -

-

277.14 ± 59.16 51.96 ± 10.8 120 (91.75 - 163.5) 200.8 ± 59.9 225.18 ± 59

260.13 ± 37.44 52.3 ± 11.23 125 (101 - 181.75) 181.28 ± 37.87 207.83 ± 38.38

1 (1.8) 36 (64.3) 13 (23.2) 6 (10.7)

28 (93.3) 2 (6.7) -

p Value between two groups

RI PT

Non FH group (no genetic variant)

0.04 0.44 0.02

13 (50) -

296.77 ± 72.98 51.58 ± 10.49 97.5 (71.75 - 150) 223.32 ± 72.43 245.19 ± 71.92

0.03 0.81 0.02 < 0.01 0.02

1 (3.8) 8 (30.8) 11 (42.3) 6 (23.1)

< 0.001 < 0.001 -

SC

M AN U

-

123.66 ± 11.93 77.41 ± 9.49 15 (26.8)

126.33 ± 11.52 79.5 ± 8.65 6 (20.0)

120.58 ± 11.86 75 ± 10 9 (34.6)

0.07 0.08 0.24

23 (41.1) 12 (21.4)

13 (43.3) 8 (26.7)

10 (38.5) 4 (15.4)

0.79 0.35

TE D

Demographic Characteristics N Age, years Men, n (%) Body mass index, kg/m2 Physical examination Corneal arcus, n (%) Tendon xanthomas, n (%) Lipid Values Total cholesterol, mg/dL HDL cholesterol, mg/dL Triglycerides, mg/dL LDL cholesterol, mg/dL Non-HDL cholesterol, mg/dL LDL cholesterol Values, mg/dL LDL < 191, n (%) LDL 191 - 250, n (%) LDL 251 - 325, n (%) LDL > 325, n (%) Risk Factors Systolic blood pressure, mm Hg Diastolic blood pressure, mmHg Smoking, n (%) Treatment Statin therapy, n (%) Antihypertensive therapy, n (%)

All patients

Table 3. General Characteristics of the Study Population. Data are presented as mean ± standard deviation, percentages, or median (interquartile range).

lipoprotein.

EP

FH = familial hypercholesterolemia, HDL = high-density lipoprotein, LDL = low-density

AC C

The estimated pretreated LDL cholesterol was included in LDL cholesterol values profile.

24

ACCEPTED MANUSCRIPT

Figure Legends

Figure 1. Flow diagram of enrollment of the study subjects.

RI PT

LDL = low-density lipoprotein, CVD = cardiovascular disease, IMT = intima-media thickness.

Figure 2. Mutation Detection in the Study Population stratified by DLCN score. Percentages of subjects with a DLCN score of 4 - 5, 6 - 8 and >8

SC

DLCN = Dutch Lipid Clinic Network. † = p value < 0.001 versus non FH

M AN U

To compare the percentages of patients in the DLCN score groups, the χ2 for qualitative variables was used.

Figure 3. Difference in IMT values of the non FH and FH subjects.

Values are adjusted for age.

TE D

Data are presented as mean ± standard deviation.

IMT = intima-media thickness

FH = familial hypercholesterolemia

AC C

EP

* = p value < 0.01

25

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Highlights

We performed a hospital cholesterol screening program to detect patients with FH. DLCN score ≥ 6 had high sensibility to detect probably FH subjects. Genetic analysis was useful to detect new FH mutations.

AC C

EP

TE D

M AN U

SC

Patients with FH had higher intima-media thickness.

RI PT

Only 38.5% of FH subjects were on statin therapy.