Atherosclerosis 188 (2006) 398–405
Autosomal recessive hypercholesterolemia (ARH) and homozygous familial hypercholesterolemia (FH): A phenotypic comparison Livia Pisciotta a , Claudio Priore Oliva b , Giovanni Mario Pes c , Lilla Di Scala d , Antonella Bellocchio a , Raffaele Fresa a , Alfredo Cantafora e , Marcello Arca f , Sebastiano Calandra b,∗∗ , Stefano Bertolini a,∗ a
b
Department of Internal Medicine, University of Genoa, Vaile Benedetto XV 6, I-16132 Genoa, Italy Department of Biomedical Sciences, University of Modena and Reggio Emilia, Via Campi 287, I-41100 Modena, Italy c Department of Biomedical Sciences, University of Sassari, Italy d Department of Mathematics, University of Pavia, Italy e National Institute of Health, Rome, Italy f Department of Clinical and Applied Medical Therapy, University of Rome “La Sapienza”, Italy Received 22 September 2005; accepted 4 November 2005 Available online 15 December 2005
Abstract Autosomal recessive hypercholesterolemia (ARH) is a rare disorder, due to complete loss of function of an adaptor protein (ARH protein) required for receptor-mediated hepatic uptake of LDL. ARH is a phenocopy of homozygous familial hypercholesterolemia (HoFH) due to mutations in LDL receptor (LDLR) gene; however, previous studies suggested that ARH phenotype is less severe than that of HoFH. To test this hypothesis we compared 42 HoFH and 42 ARH patients. LDLR and ARH genes were analysed by Southern blotting and sequencing. LDLR activity was measured in cultured fibroblasts. In ARH plasma LDL cholestrol (LDL-C) level (14.25 ± 2.29 mmol/L) was lower than in receptor-negative HoFH (21.38 ± 3.56 mmol/L) but similar to that found in receptor-defective HoFH (15.52 ± 2.39 mmol/L). The risk of coronary artery disease (CAD) was 9-fold lower in ARH patients. No ARH patients ≤20 years of age were found to have CAD as opposed to 43% of HoFH. The CAD prevalence was or tended to be lower in ARH also in the 21–40 (45% versus 86%) and 41–60 (78% versus 100%) age groups. Heterozygous ARH carriers showed higher level of LDL-C (+17%) than non-carrier family members. In conclusion the clinical phenotype of ARH is milder than that of receptor-negative HoFH and resembles that observed in receptor-defective HoFH. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Co-dominant and recessive hypercholesterolemia; Clinical phenotype; Coronary artery disease
1. Introduction Familial hypercholesterolemia (FH) (OMIM 143890) is one of the most common inherited metabolic diseases with a frequency of 1:500 for heterozygotes and 1:1,000,000 for homozygotes [1]. The key feature of FH, which has a co-
∗
Corresponding author. Tel.: +39 010 3537992; fax: +39 010 3537797. Corresponding author. Tel.: +39 059 2055423; fax: +39 059 2055426. E-mail addresses:
[email protected] (S. Calandra),
[email protected] (S. Bertolini). ∗∗
0021-9150/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2005.11.016
dominant transmission, is the high plasma level of LDL cholesterol (LDL-C), frequently associated with tendon xanthomas and premature coronary artery disease (pCAD). The primary cause of LDL-C elevation in FH is a genetic defect of the LDL-receptor (LDLR) [1]. The elevation of LDLC and the FH phenotype is gene-dose dependent; in the rare homozygous or compound heterozygous patients (collectively indicated as FH homozygotes (HoFH)), who have inherited two mutant LDLR genes, plasma LDL-C increases 4–5-fold and severe coronary artery disease (CAD) occurs in the first two decades of life in most of the cases [1,2]. FH homozygotes with different types of LDLR mutations show
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
a residual LDLR activity in cultured fibroblasts varying from 0 to 30% [1,3]. This variability to some extent explains the phenotypic variation in these patients; those with receptornegative mutations (≤5% residual LDL receptor activity) [4] tend to show a more severe phenotype, as compared to patients with receptor-defective mutations [1,3]. A syndrome clinically similar to FH is familial defective apolipoprotein B-100 (FDB) (OMIM 144010), which is due to a few mutations in the gene encoding apolipoprotein B100. In its heterozygous form FDB occurs with a frequency of about 1:1000 in central Europe but is much less common in other populations [5]. The few FDB homozygotes reported so far have a much milder phenotype than HoFH [5]. A third co-dominant form of hypercholesterolemia (designated autosomal dominant hypercholesterolemia type 3, HCHOLA3) (OMIM 603776) has recently been described [6]. It is due to mutations in a gene named pro-protein convertase subtilisin kexin 9 (PCSK9), encoding a protease which is a member of the subtilisin-like protein convertase family. Some missense mutations of this protein have been associated with autosomal dominant hypercholesterolemia [6–9]. No homozygous patients with PCSK9 mutations have been reported so far. Since the first report by Khachadurian and Uthman of some Lebanese families with an apparent recessive transmission of FH [10], evidence has been accumulating of the existence of a recessive form of severe hypercholesterolemia, phenotypically similar to homozygous FH but not linked to LDLR gene. This disorder, designated autosomal recessive hypercholesterolemia (ARH) (OMIM 603813), is caused by mutations in the ARH gene, which encodes an adaptor protein required for normal LDL receptor-mediated endocytosis in hepatocytes [11–13]. ARH is considered an exceedingly rare disorder in most countries with the exception of the island of Sardinia (Italy), where the disease is not uncommon (frequency 1:40,000 for homozygotes and compound heterozygotes), probably as the result of founder effect and inbreeding [14]. Only two mutations of ARH gene (W22X and c.432insA) account for all reported ARH in the Sardinian population [14]. ARH has been considered a phenocopy of HoFH due to LDLR mutations. However, the careful observation of some patients in highly specialised clinical settings and the survey of the cases reported in literature suggest that the clinical phenotype of ARH is somewhat less severe and more variable within the same family than HoFH [15,16]. Moreover, ARH patients seem to be more responsive to lipid-lowering drug therapy, as compared to HoFH [14,17,18]; however, there are ARH cases that showed poor response to statin treatment [19]. The comparison between ARH and HoFH patients is no easy task for the following reasons: (i) the small number of patients available for study (especially in the case of ARH); (ii) the great phenotypic variability in both HoFH and ARH patients; (iii) the large number of mutations (especially in the case of HoFH) which have a different effect on LDLR activity. To
399
overcome these problems we have taken advantage of two favourable circumstances: (i) the previous phenotypic and genetic characterization of a large number of Italian HoFH patients carrying mutations of LDLR gene and (ii) the recent identification and characterization of several ARH patients from Sardinia.
2. Methods 2.1. Recruitment of patients 2.1.1. FH patients In this comparative analysis we included 29 homozygotes and 7 compound heterozygotes (in whom both allelic mutations were detected), who had been previously reported by our group [3], and six patients recently characterized: 5 homozygotes (2 with deletion of exons 1–12, 1 with p.D221G, 1 with p.N370T and 1 with p.A399T) and 1 compound heterozygote (p.D90N/C222X) (Table 1). From the evaluation of LDL receptor activity in cultured skin fibroblasts [3] we were able to classify 14 of them as receptor-negative (≤5% residual LDL receptor activity) and 26 as receptor-defective (8–30% residual LDL receptor activity) [4]. Cultured skin fibroblasts from two patients with p.G373D mutation were not available; these subjects were defined as functionally “unclassified” (see Table 1). 2.1.2. ARH patients We collected clinical and biochemical data of 42 molecularly characterized Sardinian ARH patients from 24 unrelated families (Table 2). Twenty-eight of them had been previously reported [14]. We also recruited non-affected members of ARH families. Informed consent was obtained from the patients and their family members or, in the case of children, from their parents. The study protocol was approved by the institutional human investigation committee of each participating institution. 2.2. Biochemical analysis Fasting plasma lipid concentrations were measured before any hypolipidemic treatment. Total cholesterol (T-C), triglycerides (TG) and high density lipoprotein cholesterol (HDLC) levels were measured enzymatically (Roche Diagnostics GmbH, Mannheim, Germany), using an automated analyzer (Hitachi model 912, Hitachi, Ltd., Tokyo); low density lipoprotein cholesterol (LDL-C) was calculated by Friedewald’s formula. 2.3. LDL receptor activity The assay of 125 I-labeled LDL binding, internalisation and degradation by cultured skin fibroblasts was performed as previously reported [3].
Values are mean ± S.D.; values adjusted for gender and age are reported in italics. Receptor-defective alleles: dupl. exons 16–17, p.D221G, p.C352W, p.N370T, p.C379R, p.A399T, p.V523M, p.P685L, IVS12 −1G > A, IVS15 +1G > A, p.D118Y/p.V523M, p.C379R/p.V523M, p.G592E/p.P685L. Receptor-negative alleles: del exons 2–12, del exons 13–18, p.Q33X, p.Q125X, p.G549D, IVS10 +1G > A, del. exons 13–15/p.G549D, p.Q35X/p.H485R, p.C89R/p.G592E, p.D90N/p.C222X. Unclassified allele: p.G373D. * P < 0.002 vs. receptor-defective (Mann–Whitney test).
0.82 ± 0.21, 0.79 ± 0.21 0.69 ± 0.20, 0.75 ± 0.17 0.78 ± 0.05, 0.73 ± 0.03 15.52 ± 2.39, 16.08 ± 2.82 21.38 ± 3.56* , 20.12 ± 3.24* 12.14 ± 1.73, 13.68 ± 2.19 16.95 ± 2.44, 17.48 ± 2.89 22.78 ± 3.52* , 21.58 ± 3.22* 13.52 ± 1.68, 14.99 ± 2.11 15/11 6/8 2/0 Receptor-defective Receptor-negative Unclassified
21 12 1
HDL-C (mmol/L) LDL-C (mmol/L) T-C (mmol/L) No. of families No. of subjects M/F LDLR mutations
Table 1 Plasma lipid profile in homozygous FH patients
1.29 ± 0.52, 1.28 ± 0.52 1.43 ± 0.42, 1.45 ± 0.42 1.29 ± 0.25, 1.28 ± 0.26
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
TG (mmol/L)
400
2.4. Patient genotyping FH patients were genotyped by Southern blot analysis, SSCP analysis, direct sequencing of genomic DNA, Northern blot analysis and sequencing of RT-PCR products [3]. ARH patients were genotyped by sequencing all nine exons, as well as the splice junctions, of ARH gene using intronic primers (the sequence of these primers is reported in Supplementary Table 1). Exon 1 was amplified using AdvantageTM -GC2 Polymerase Mix (Clontech, Palo Alto, CA) under the following conditions: 94 ◦ C for 5 min; 30 cycles at 94 ◦ C for 30 s/65 ◦ C for 30 s/72 ◦ C for 1 min, followed by a final extension at 72 ◦ C for 7 min. Exons 2–9 were amplified using Expand High Fidelity PCR System (Roche Diagnostics GmbH, Mannheim, Germany), under the following conditions: 94 ◦ C for 3 min; 30 cycles 94 ◦ C for 15 s/55 ◦ C for 30 s/72 ◦ C for 30 s, followed by a final extension at 72 ◦ C for 7 min. The amplification products were purified by ExoSAP-ITM (Amersham Pharmacia Biotech Italia, Cologno Monzese, Italy) and sequenced by automatic sequencer CEQ2000 DNA Analysis System (Beckman Coulter, Fullerton, CA). The mutations were designated according to the Human Genome Variation Society (http://www.hgvs.org/ mutnomen). The numerical series of codons includes the sequence of the signal peptide. 2.5. Clinical criteria for CAD assessment FH and ARH patients were considered positive for CAD (CAD+), if they: (a) had died from coronary disease, confirmed by autopsy; (b) had a history of documented myocardial infarction, coronary artery bypass graft (CABG) or percutaneous transluminal coronary angioplasty (PTCA); (c) suffered from angina pectoris with positive exercise ECG and thallium test; (d) had silent ischemia detected during bicycle exercise testing and confirmed by thallium test or stress echocardiography; (e) had angiographically documented coronary atherosclerosis with stenosis >50% in the main arteries. FH and ARH patients were considered free of coronary artery disease (CAD−) if they had a negative bicycle exercise test, confirmed by thallium test or stress echocardiography. Some asymptomatic children only had undergone resting ECG and echocardiography. 2.6. Statistical analysis All statistical analyses were performed using the SAS System (version 8.02, SAS Institute Inc., Cary, NC). Differences in the distribution of categorical variables were assessed by Chi-square or Fisher’s exact tests. Differences between groups for continuous variables were assessed by Mann–Whitney test. Triglyceride levels were logarithmically transformed before analysis. The probability of having CAD in FH patients with respect to ARH patients was evaluated by logistic regression analysis, carried out using PROC
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
401
Table 2 Plasma lipid profile in ARH patients ARH mutations
No. of subjects M/F
No. of families
T-C (mmol/L)
LDL-C (mmol/L)
HDL-C (mmol/L)
TG (mmol/L)
A/A B/B A/B A + B/B A + B/A + B
7/5 9/6 5/7 1/0 2/0
5 10 7 1 1
15.38 ± 1.64, 15.46 ± 1.67 16.35 ± 3.00, 16.33 ± 2.90 15.85 ± 1.94, 15.70 ± 1.76 19.73, 20.10 14.42 ± 0.91, 14.83 ± 0.95
13.63 ± 1.72, 13.74 ± 1.76 14.69 ± 2.89, 14.63 ± 2.78 14.18 ± 1.88, 14.05 ± 1.73 18.25, 18.54 13.03 ± 0.87, 13.37 ± 0.94
1.25 ± 0.31, 1.23 ± 0.30 1.16 ± 0.15, 1.18 ± 0.17 1.13 ± 0.16, 1.11 ± 0.15 1.03, 1.10 0.89 ± 0.02, 0.94 ± 0.03
1.10 ± 0.47, 1.06 ± 0.47 1.08 ± 0.22, 1.11 ± 0.26 1.18 ± 0.52, 1.18 ± 0.45 0.94, 0.97 1.10 ± 0.12, 1.11 ± 0.10
Values are mean ± S.D.; values adjusted for gender and age are reported in italics. Mutant alleles: A = p.W22X; B = c.432insA > Fs144 > X170.
GENMOD in SAS. The explanatory variables in the models were age (used as a categorical variable with the following assigned values: 1 = ≤20; 2 = 21–40; 3 = 41–60) and the disease (ARH = 1; FH = 2). All the logistic models include a random effect term, which is the familial relationship among observations from patients belonging to the same family. Likelihood ratio statistics was used to compare models with different combinations of predictors with an inclusion level of 0.05 and an exclusion level of 0.1. Odds ratios and 95% confidence intervals are reported.
3. Results 3.1. Plasma lipid profile in HoFH patients Table 1 shows the pre-treatment plasma lipids in 42 patients who had the clinical diagnosis of homozygous FH and were homozygous or compound heterozygous for LDLR gene mutations. Patients with receptor-negative mutations have higher T-C and LDL-C levels than patients with receptor-defective mutations, as previously reported by our group [4].
heterozygotes for the two ARH gene mutations found in the Sardinian population. 3.3. Comparison between HoFH and ARH patients Table 3 shows the comparison between ARH and HoFH patients. ARH patients were older than FH patients; their mean plasma levels of T-C, LDL-C and TG were lower, and that of HDL-C was higher than in HoFH patients. These differences were most striking when the comparison was made with HoFH with receptor-negative mutations. By contrast, when we compared ARH patients with HoFH patients carrying receptor-defective mutations, we found no significant differences in age and gender adjusted plasma lipids—with the exception of HDL-C (whose level was higher in ARH patients). We found no differences between HoFH (receptornegative and receptor-defective) and ARH patients with regard to tendon xanthomatosis, planar and tuberous xanthomatosis. However, the prevalence of overt coronary heart disease was lower in ARH patients than in HoFH patients taken as a whole. 3.4. Assessment of cardiovascular risk
3.2. Plasma lipid profile in ARH patients Table 2 shows the pre-treated plasma lipids in 42 Sardinian ARH patients identified in 24 families. The plasma T-C and LDL-C levels were similar in homozygotes and compound
Since it is well established that the prevalence of CAD in FH increases with age, we stratified our patients in three age groups (≤20, 21–40 and 41–60 years of age) (Table 4). No ARH patients ≤20 years of age were found to have CAD, in
Table 3 Comparison between homozygous FH and ARH patients
N Age (year) T-C (mmol/L) LDL-C (mmol/L) HDL-C (mmol/L) TG (mmol/L) Tx (%) Px (%) Tux (%) CAD+ (%)
All homozygous FH patients
Receptor-negative FH patients
Receptor-defective FH patients
ARH patients
42 23.2 ± 15.9* 18.73 ± 4.07‡ , 18.38 ± 3.61† 17.31 ± 4.07‡ , 16.93 ± 3.56† 0.77 ± 0.21‡ , 0.81 ± 0.20‡ 1.34 ± 0.48* , 1.34 ± 0.48* 92.9 57.1 71.4 66.7*
14 12.8 ± 11.8¶ 22.78 ± 3.52¶ , 21.37 ± 3.24¶ 21.38 ± 3.56¶ , 19.86 ± 3.24¶ 0.69 ± 0.20¶ , 0.81 ± 0.17¶ 1.43 ± 0.42• , 1.46 ± 0.42• 85.7 78.6 78.6 71.4§
26 28.1 ± 15.8 16.95 ± 2.44# , 17.06 ± 2.78 15.52 ± 2.39# , 15.63 ± 2.73 0.82 ± 0.21** , 0.81 ± 0.22** 1.29 ± 0.52, 1.29 ± 0.52 96.2 50 65.4 61.5
42 31.1 ± 13.3 15.92 ± 2.33, 16.27 ± 2.45 14.25 ± 2.29, 14.63 ± 2.42 1.16 ± 0.22, 1.13 ± 0.21 1.11 ± 0.39, 1.10 ± 0.38 92.9 50.0 65.0 40.5
Values are mean ± S.D.; values adjusted for gender and age are reported in italics. All FH vs. ARH: * P < 0.05; † P < 0.01; ‡ P < 0.001; receptor-negative FH vs. ARH: § P < 0.05; • P < 0.01; ¶ P < 0.001; ARH vs. receptor-defective FH: # P < 0.05; ** P < 0.001 (Mann–Whitney test and Fisher test). Tx: tendon xanthomas; Px: planar xanthomas; Tux: tuberous xanthomas; CAD: coronary artery disease.
402
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
Table 4 Prevalence of CAD in homozygous FH and ARH patients Homozygous FH patients
Receptor-negative FH patients
Receptor-defective FH patients
ARH patients
Age group ≤20 N Age (range) LDL-C HDL-C TG CAD (%)
21 9.3 ± 5.2 (1–20)* 19.33 ± 4.16† 0.68 ± 0.20‡ 1.27 ± 0.42 9/21 (42.8%)†
12 8.7 ± 5.2 (1–19)§ 21.8 ± 3.64¶ 0.64 ± 0.16¶ 1.47 ± 0.44• 8/12 (66.6%)•
9 10.1 ± 5.4 (1–20) 15.96 ± 1.65 0.74 ± 0.24** 1.00 ± 0.17 1/9 (11.1%)
11 14.2 ± 4.5 (6–19) 15.36 ± 2.64 1.11 ± 0.17 1.03 ± 0.21 0/11 (0%)
Age group 21–40 N Age (range) LDL-C HDL-C TG CAD (%)
14 31.5 ± 5.4 (22–39) 15.2 ± 2.60 0.86 ± 0.20‡ 1.43 ± 0.58 12/14 (85.7%)*
1 31 19.11 1.08 1.05 1/1 (100.0%)
11 31.4 ± 6.0 (22–39) 15.44 ± 2.24# 0.85 ± 0.22** 1.49 ± 0.64 9/11 (81.8%)
22 32.4 ± 6.1 (21–40) 13.66 ± 2.03 1.16 ± 0.18 1.06 ± 0.31 10/22 (45.4%)
Age group 41–60 N Age (range) LDL-C HDL-C TG CAD (%)
7 48.1 ± 3.7 (43–54) 15.4 ± 3.53 0.89 ± 0.13* 1.35 ± 0.48 7/7 (100.0%)
1 44 18.02 0.98 1.31 1/1 (100.0%)
6 48.8 ± 3.5 (43–54) 15.00 ± 3.66 0.88 ± 0.13# 1.36 ± 0.52 6/6 (100.0%)
9 48.6 ± 5.9 (42–58) 14.33 ± 2.19 1.24 ± 0.33 1.33 ± 0.64 7/9 (77.7%)
Values are mean ± S.D. Lipid values: mmol/L. All FH vs. ARH: * P < 0.05; † P < 0.01; ‡ P < 0.001; receptor-negative FH vs. ARH: § P < 0.05; • P < 0.01; ¶ P < 0.001; ARH vs. receptor-defective FH: # P < 0.05; ** P < 0.01 (Mann–Whitney test and Fisher test).
striking contrast with the percentage of CAD patients found in HoFH (especially in receptor-negative patients). In this age group the plasma lipid profile of ARH patients was similar to that of receptor-defective HoFH patients, with the exception of HDL whose level was higher in ARH. A significant difference in CAD prevalence between ARH and HoFH (taken as a whole) was also observed in the 21–40 age group. The odds ratios (95% confidence interval) for CAD, adjusted for age and for familial relationship, in FH patients with respect to ARH patients were the following: all FH versus ARH 9.1 (4.4–19.1); receptor-defective FH versus ARH 5.4 (1.5–19.8); receptor-negative FH versus ARH 25.9 (3.6–186.5). Table 5 shows the clinical features of ARH and HoFH patients with CAD. ARH patients with CAD were much older, had lower plasma T-C and LDL-C and higher HDL lev-
els than receptor-negative HoFH patients and show clinical features similar to those of receptor-defective HoFH patients with the exception of plasma HDL level. The comparison of ARH patients with or without CAD showed that there was no difference in plasma lipid profile and the prevalence of xanthomatosis. However, patients with CAD were 15 years older than patients without CAD (P < 0.0001) (data not shown). 3.5. Plasma lipid profile in ARH heterozygotes Plasma lipid levels in ARH heterozygotes are shown in Table 6. Plasma T-C tended to be higher in ARH carriers as compared to non-carrier family members, but this difference was not significant. However, the mean level of plasma LDLC was significantly higher (+17%) in carriers (especially in
Table 5 Comparison between homozygous FH and ARH patients with CAD
N Age (year) T-C (mmol/L) LDL-C (mmol/L) HDL-C (mmol/L) TG (mmol/L) Tx (%) Px (%) Tux (%)
All homozygous FH patients
Receptor-negative FH patients
Receptor-defective FH patients
ARH patients
28 28.9 ± 15.1* 18.82 ± 4.25† , 19.04 ± 3.44* 17.36 ± 4.32† , 17.60 ± 3.44* 0.77 ± 0.20‡ , 0.76 ± 0.16‡ 1.45 ± 0.47, 1.44 ± 0.48 100 50.0 71.4
10 16.4 ± 12.1• 22.92 ± 3.54• , 21.92 ± 2.94• 21.59 ± 3.62• , 20.50 ± 2.96• 0.68 ± 0.24• , 0.76 ± 0.17§ 1.42 ± 0.30, 1.44 ± 0.31 100 80.0 90.0
16 36.3 ± 12.5 16.91 ± 2.43, 17.79 ± 2.46 15.37 ± 2.40, 16.33 ± 2.44 0.82 ± 0.18# , 0.75 ± 0.16# 1.48 ± 0.58, 1.47 ± 0.58 100 37.5 56.3
17 40.2 ± 10.1 15.80 ± 1.97, 16.95 ± 2.32 14.06 ± 1.95, 15.32 ± 2.36 1.14 ± 0.18, 1.03 ± 0.19 1.31 ± 0.49, 1.29 ± 0.49 100 46.7 86.7
Values are mean ± S.D.; values adjusted for gender and age are reported in italics. All FH vs. ARH: * P < 0.05; † P < 0.01; ‡ P < 0.001; receptor-negative FH vs. ARH: § P < 0.01; • P < 0.001; ARH vs. receptor-defective FH: # P < 0.001 (Mann–Whitney test and Fisher test).
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
403
Table 6 Plasma lipid profile in ARH heterozygotes
M/F Age (year) T-C (mmol/L) LDL-C (mmol/L) HDL-C (mmol/L) TG (mmol/L)
Non-carrier family members
All ARH mutation carriers
W22X carriers
c.432insA carriers
36/24 41.6 ± 20.0 5.24 ± 1.21, 5.29 ± 1.07 3.07 ± 1.05, 3.09 ± 1.01 1.57 ± 0.53, 1.58 ± 0.50 1.33 ± 1.02, 1.35 ± 0.95
32/28 45.0 ± 20.3 5.61 ± 1.16, 5.57 ± 1.13 3.58 ± 1.10* , 3.56 ± 1.07* 1.48 ± 0.42, 1.47 ± 0.41 1.24 ± 0.71, 1.22 ± 0.67
11/13 40.4 ± 20.8 5.86 ± 1.35, 5.92 ± 1.27 3.87 ± 1.33* , 3.92 ± 1.25* 1.48 ± 0.39, 1.46 ± 0.41 1.12 ± 0.39, 1.19 ± 0.38
21/15 48.1 ± 19.7 5.45 ± 1.02, 5.34 ± 0.97 3.39 ± 0.88, 3.32 ± 0.88 1.47 ± 0.44, 1.47 ± 0.41 1.31 ± 0.85, 1.25 ± 0.82
Values are mean ± S.D.; values adjusted for gender and age are reported in italics. ARH mutation carriers vs. non-carriers: * P < 0.02 (Mann–Whitney test).
those with the W22X mutation, +26%) while plasma HDL-C and TG levels were comparable to those found in non-carrier family members.
4. Discussion In this study there has been a comparison between the plasma lipid profile and the clinical manifestations in patients with autosomal recessive hypercholesterolemia and in patients with homozygous familial hypercholesterolemia due to mutations in LDLR gene. This comparison led us to the following conclusions: (i) the plasma level of LDL-C in ARH patients was much lower than that observed in HoFH with receptor-negative mutations but close to that seen in patients with milder forms of HoFH due to receptor-defective mutations; (ii) the plasma HDLC level was higher in ARH patients than in HoFH patients, irrespective of the residual LDLR activity; (iii) the risk of coronary artery disease was 9-fold lower in ARH patients than in HoFH taken as a whole. Altogether, the results of this comparison consolidate, by means of stronger statistical bases, the suggestions derived from the description of ARH patients reported over the last few years. It can be argued, however, that our conclusions apply only to Sardinian ARH patients, who belong to a population with a unique genetic background [14]. We think that our conclusions can be extended to all ARH patients for the reasons illustrated below. (A) The ARH gene mutations found in Sardinian patients (p.W22X and c.432insA > Fs144 > X170), like those found in all other ARH cases reported so far, are predicted to result in the absence of ARH protein (because of either premature termination of translation or failure to produce mRNA) [15]; in this respect the basic defect in ARH function is expected to be comparable among ARH patients throughout the world – in striking contrast to the position of HoFH – where heterogeneity of mutant alleles results in a variable effect on LDLR function. (B) The plasma LDL-C level found in our ARH patients (14.25 ± 2.29 mmol/L) is fairly similar to the mean LDLC level found in non-Sardinian ARH patients reported so far in literature (14.02 ± 3.19 mmol/L) [11,17,18,20–24]. (C) The mean plasma LDL-C levels of genotyped heterozygous FH patients from Sardinia is similar to that observed in heterozygous FH of continental Italy (see Supplementary Table
2), thus indicating that the specific Sardinian genetic background has no substantial effect on the expression of defects of receptor-mediated LDL metabolism. The differences in the level of plasma T-C and LDLC between our ARH and HoFH patients resemble those observed in mice with targeted inactivation of ARH and LDLR genes, respectively. Under basal dietary conditions (chow diet containing only 0.02% cholesterol) the mean plasma level of cholesterol in the arh−/− mice was intermediate (83 mg/dL) to the wild-type (68 mg/dL) and ldlr−/− mice (196 mg/dL). This difference was greatly reduced or abolished when mice were fed with a chow containing 0.1–0.2% cholesterol [25]. If the result of these observations can be applied to humans, it is expected that dietary habits as well as variations in intestinal absorption of cholesterol may have a substantial effect on plasma LDL-C in ARH patients. Like HoFH, ARH patients are prone to develop premature CAD. We found that 40% of our ARH patients had premature CAD. However, as opposed to patients with HoFH, our ARH patients develop clinically overt CAD later in life—after 20 years of age. This indicates that coronary atherosclerosis as well as aortic valve stenosis develop more slowly in ARH patients than in HoFH (especially in subjects with receptornegative mutations), probably as the result of lower plasma level of LDL-C and, possibly, higher level of HDL-C. We calculated that the risk of developing CAD in ARH patients is about 5-fold and 25-fold lower than in receptordefective and in receptor-negative HoFH patients, respectively. However, the large 95% confidence intervals of the calculated odds, due to the relatively small number of subjects investigated, indicate that these estimates must be considered as approximate values. Again it can be argued that this reduced CAD risk in ARH patients as opposed to HoFH, is due to some “protective” genetic background present in Sardinian population only. That does not seem to be the case as the prevalence of CAD in genotyped heterozygous FH patients from Sardinia was similar to that observed in heterozygous FH patients from continental Italy (see Supplementary Table 2). ARH appears to be a fully recessive disorder, as in the majority of cases parents of affected individuals have normal plasma cholesterol. However, it has been reported that in almost 30% of the cases parents had cholesterol levels >6 mmol/L [14,17,19,26]. Our comparison of the plasma lipid profile in 60 ARH carriers and in 60 non-affected family
404
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405
members clearly showed that there was no significant difference in plasma cholesterol (Table 6). However, we found a small but significant difference in the level of LDL-C, which was slightly higher in ARH carriers (+17%) than in nonaffected family members. This difference, however, does not allow the discrimination of ARH carriers from non-carriers during population screening. This is in keeping with the observation in mice where no significant differences were seen in plasma lipid levels between the arh+/− mice and arh+/+ mice, irrespective of dietary treatment [25,27]. The milder phenotype of ARH patients with respect to HoFH suggests that ARH patients might respond to lipidlowering therapy (LDL-apheresis, statins, cholestyramine, ezetimibe) much better than HoFH. Indeed, such appears to be the case in the few reports of long-term follow up treatment of ARH patients [18,28]. Although a previous report had indicated a 30–60% reduction of plasma total cholesterol following statin therapy in Sardinian ARH patients [14], we have not been able to collect sufficiently structured data on therapy (i.e. type of drug and dosage, compliance of the patients, schedule of LDL-apheresis, use of bile acid sequestrant, etc.) to provide, at this stage, a reliable and homogeneous overview of the result of treatment. However, the results of long-term
Fig. 1. Long-term treatment of three patients: (A) FH homozygote with receptor-negative mutation (p.G549D); (B) FH compound heterozygote with receptor-defective mutations (p.C379R/p.V522M); (C) ARH homozygote (c.432insA). Aph: LDL-apheresis every 15–30 days (the values reported are the mean ± S.D. of LDL-C levels observed before Aph treatment); S: simvastatin 40, 60 mg/day; A: atorvastatin 10, 20, 40, 60, 80 mg/day; C: cholestyramine 8, 16 g/day; E: ezetimibe 10 mg/day.
treatment of three of our patients (Fig. 1) suggest that the response to lipid-lowering therapy might be satisfactory in ARH patients, intermediate in HoFH patients with receptordefective mutations but poor in HoFH with receptor-negative mutations.
Acknowledgement This work was supported by grants made by Telethon (GGP02149) to S.B. and to S.C.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis. 2005.11.016.
References [1] Goldstein JL, Hobbs H, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. 8th ed. New York: McGraw-Hill; 2001. p. 2863–913. [2] Rader DJ, Cohen J, Hobbs HH. Monogenic hypercholesterolemia: new insights in pathogenesis and treatment. J Clin Invest 2003;111:1795–803. [3] Bertolini S, Cassanelli S, Garuti R, et al. Analysis of LDL receptor gene mutations in Italian patients with homozygous familial hypercholesterolemia. Arterioscler Thromb Vasc Biol 1999;19:408– 18. [4] Bertolini S, Cantafora A, Averna M, et al. Clinical expression of familial hypercholesterolemia in clusters of mutations of LDL-receptor gene causing receptor-defective or receptor-negative phenotype. Arterioscler Thromb Vasc Biol 2000;20:e41–52. [5] Myant NB. Familial defective apolipoprotein B-100: a review including some comparison with familial hypercholesterolemia. Atherosclerosis 1993;104:1–18. [6] Abifadel M, Varret M, Rabes J-P, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 2003; 34:154–8. [7] Timms KM, Wagner S, Samuels ME, et al. A mutation in PCSK9 causing autosomal-dominant hypercholesterolemia in a Utah pedigree. Hum Genet 2004;114:349–53. [8] Leren TP. Mutation in the PCSK9 gene in Norwegian subjects with autosomal dominant hypercholesterolemia. Clin Genet 2004;65:419–22. [9] Maxwell KN, Breslow JL. Proprotein convertase subtilisin kexin 9: the third locus implicated in autosomal dominant hypercholesterolemia. Curr Opin Lipidol 2005;16:167–72. [10] Khachadurian AK, Uthman SM. Experiences with the homozygous cases of familial hypercholesterolemia. A report of 52 patients. Nutr Metab 1973;15:132–40. [11] Garcia CK, Wilund K, Arca M, et al. Autosomal recessive hypercholesterolemia caused by mutations in a putative LDL receptor adaptor protein. Science 2001;292:1394–8. [12] Wilund KR, Yi M, Campagna F, et al. Molecular mechanisms of autosomal recessive hypercholesterolemia. Hum Mol Genet 2002;11:3019–30. [13] Cohen JC, Kimmel M, Polanski A, Hobbs HH. Molecular mechanisms of autosomal recessive hypercholesterolemia. Curr Opin Lipidol 2003;14:121–7.
L. Pisciotta et al. / Atherosclerosis 188 (2006) 398–405 [14] Arca M, Zuliani G, Wilund K, et al. Autosomal recessive hypercholesterolemia in Sardinia, Italy, and mutations in ARH: a clinical and molecular genetic analysis. Lancet 2002;359: 841–7. [15] Soutar AK, Naoumova RP, Traub L. Genetics, clinical phenotype, and molecular cell biology of autosomal recessive hypercholesterolemia. Arterioscler Thromb Vasc Biol 2003;23: 1963–70. [16] Fellin R, Zuliani G, Arca M, et al. Clinical and biochemical characterization of patients with autosomal recessive hypercholesterolemia (ARH). Nutr Metab Cardiovasc Dis 2003;13:278–86. [17] Naumova RP, Neuwirth C, Lee P, et al. Autosomal recessive hypercholesterolemia: long-term follow-up and response to treatment. Atherosclerosis 2004;174:165–72. [18] Lind S, Olsson AG, Eriksson M, et al. Autosomal recessive hypercholesterolemia: normalization of plasma LDL cholesterol by ezetimibe in combination with statin treatment. J Intern Med 2004;256:406–12. [19] Barbagallo CM, Emmanuele G, Cefal`u AB, et al. Autosomal recessive hypercholesterolemia in a Sicilian kindred harbouring the 432insA mutation of the ARH gene. Atherosclerosis 2003;166:395– 400. [20] Al-Kateb H, B¨ahring S, Hoffmann K, et al. Mutation in the ARH gene and a chromosome 13q locus influence cholesterol levels in a new form of digenic-recessive familial hypercholesterolemia. Circ Res 2002;90:951–8.
405
[21] Harada-Shiba M, Takagi A, Miyamoto Y, et al. Clinical features and genetic analysis of autosomal recessive hypercholesterolemia. J Clin Endocrinol Metab 2003;88:2541–7. [22] Rodenburg J, Wiegman A, Vissero MN, et al. A boy with autosomal recessive hypercholesterolemia. Neth J Med 2004;62:89–93. [23] Thomas H-P, Vogt A, Wilund KR, et al. Autosomal recessive hypercholesterolemia in three sisters with phenotypic homozygous familial hypercholesterolemia: diagnosis and therapeutic procedures. Ther Apher Dial 2004;8:275–80. [24] Canizales-Quinteros S, Aguilar-Salinas CA, Huertas-Vazquez A, et al. A novel ARH splice site mutation in a Mexican kindred with autosomal recessive hypercholesterolemia. Hum Genet 2005;116:114–20. [25] Jones C, Hammer RE, Li W-P, et al. Normal sorting but defective endocytosis of the low-density lipoprotein receptor in mice with autosomal recessive hypercholesterolemia. J Biol Chem 2003;278:29024–30. [26] Eden ER, Patel DD, Sun X-M, et al. Restoration of LDL receptor function in cells from patients with autosomal recessive hypercholesterolemia by retroviral expression of ARH1. J Clin Invest 2002;110:1695–702. [27] Harada-Shiba M, Takagi A, Marutsuka K, et al. Disruption of autosomal recessive hypercholesterolemia gene shows different phenotype in vitro and in vivo. Circ Res 2004;95:945–52. [28] Naumova RP, Neuwirth C, Lee P, et al. Autosomal recessive hypercholesterolemia: long-term follow-up and response to treatment. Atherosclerosis 2004;174:165–72.
