Association of OAZ1 Gene Polymorphisms With Subclinical and Clinical Vascular Events
Objective— Proliferation and migration of vascular smooth muscle cells (VSMCs) are striking features shared by vascular ageing, atherosclerosis, and in-stent restenosis. VSMC biology depends in part on polyamines whose metabolism is closely regulated by ornithine decarboxylase antizyme 1 (OAZ1). Therefore, we sought for association between OAZ1 gene polymorphisms and various outcomes involving VSMC proliferation.
Methods and Results— Systematic screening of the OAZ1 gene enabled to detect 21 variants. The impact of 4 selected tag polymorphisms (+849C/T, +851G/T, +1804G/A, and +2222A/G) was evaluated in 3 independent association studies. In a sample of 205 patients, the +2222G allele was associated with an increased risk of 6-month coronary in-stent restenosis (OR [95%CI]=2.1 [1.2 to 3.6]; P=0.0071). In a sample of 1001 subjects participating to the EVA study, the +2222G allele was longitudinally associated with a 4-year increase in common carotid intima-media thickness (P=0.047). In a case-control study (466 cases versus 466 controls), the risk of coronary heart disease associated with the +2222G allele was 1.3 (95%CI=[1.1 to 1.6]; P=0.026). No other significant association was consistently detected.
Conclusions— We identified the OAZ1+2222A/G polymorphism as a potential genetic marker of vascular events. Our findings strengthen the hypothesis that the polyamine metabolism plays a role in vascular diseases.
- ornithine decarboxylase antizyme 1
- gene mutations
- coronary in-stent restenosis
- common carotid intima-media thickness
- coronary heart disease
Abnormal proliferation and migration of vascular smooth muscle cells (VSMCs) within the arterial wall is a common feature of vascular ageing, atherosclerosis, and in-stent restenosis.1–3 In many cell types, the cellular proliferation is a complex mechanism requiring polyamines, which are highly involved in the progression of the cell cycle.4 Interestingly, convergent data obtained from animal models or cell cultures underlined the implication of polyamines in VSMC proliferation and migration.5–8 Some experiments have provided evidence for polyamine-dependent restriction points during the G0–G1 transition and G1 phase in various cell types, including VSMCs.9,10 Specific inhibition of polyamine production leads to the inhibition of VSMC proliferation in rat aorta as a result of the activation of the p42/p44 MAPK pathway.10 These data lead us to assume that alterations of the polyamine pathway may be involved in the VSMC-related neointimal development observed in various pathophysiological processes.
Polyamine pathway is an elaborately orchestrated metabolism in which ornithine decarboxylase antizyme 1 (OAZ1) plays a central role.11 OAZ1, whose gene (OAZ1) is located on chromosome 19p13.3, negatively regulates the intracellular pool of polyamines by inhibiting their production and importation. The production of OAZ1 itself is induced by high levels of polyamines.11 Based on its function, OAZ1 may play an important role in cell cycle progression and cell proliferation. Some experiments demonstrated that overproduction of OAZ1 in various cell types coincides with growth inhibition.12,13 Therefore, a reduction in OAZ1 activity may increase intracellular polyamine concentrations, promoting cell proliferation. Consistently, the reduction or the lack of expression of the OAZ1 gene was observed in certain late stage cancers, both in animals and humans.14,15 Other data suggest that antizyme 1 may also influence the cell proliferation through a direct interaction with cell cycle regulators. For example, functional OAZ1 can act as a tumor suppressor by targeting the cyclin D1 for degradation, this cyclin also participating in the progression through the G1 phase of the VSMC cycle.16
In this context, we hypothesized a possible implication of OAZ1 in vascular disorders related to VSMC proliferation and neointimal formation. We analyzed the impact of OAZ1 gene polymorphisms in 3 independent association studies with various outcomes: coronary in-stent restenosis, common carotid intima-media thickening, and coronary heart disease risk.
The main subjects characteristics are described in the supplemental Table I (available online at http://atvb.ahajournals.org).
The In-Stent Restenosis Study
The sample was composed of 205 consecutive White males and females who underwent a successful implantation of a Palmaz-Schatz stent at the Cardiology Hospital (Lille, Northern France) during the period 1994–1997, and had a 6-month angiographic follow-up.17 Coronary stenting and quantitative computer-assisted angiographic measurements were performed by standard techniques, as previously described.18 About 35% of subjects had a coronary stenting for unstable angina and 46.8% had previously suffered a myocardial infarction. Coronary stents were implanted as a bailout procedure after failed balloon angioplasty (43.6%), or because there was a suboptimal result after balloon angioplasty (36.1%), or electively (20.3%). In 19% of subjects, a single stent did not entirely cover the lesion and 2 or 3 overlapping stents were implanted. The degree of stenosis before PTCA was 64.6±15%, the stenosis length was 8.8±6.1 mm, the post-dilatation diameter was 2.55±0.41 mm, and the 6-month reference diameter was 2.91±0.43 mm. Information about medical history and drug intake were obtained using a standardized questionnaire. In-stent restenosis was defined as a >50% diameter stenosis at the 6-month follow-up.19
The EVA (Etude du Vieillissement Arteriel) Study
Details of the EVA Study have been reported previously.20 The EVA Study is a longitudinal study on cognitive and vascular aging composed of White males and females aged 59 to 71 years who were recruited from the electoral rolls of the city of Nantes (Western France) between June 1991 and July 1993. Reliable high-resolution ultrasound examinations of the carotid arteries were performed at baseline and 4-year follow-up in 1104 subjects.21 Our analyses will focus on the B-mode ultrasound common carotid intima-media thickness (CCA-IMT) measurements. All measurements of CCA-IMT were made at a site free of any discrete plaques, and details of the protocol have been described elsewhere.20 Longitudinal changes in CCA-IMT were computed as the difference between the 4-year follow-up and baseline values. The mean baseline CCA-IMT was 0.665±0.120 mm. The study sample was composed of 1001 subjects with fully available phenotypic and genetic data.
The CHD Case-Control Study
A sample of 466 White males with CHD and fully available phenotypic and genetic data were drawn from the EUROASPIRE study (EUROpean Action on Secondary Prevention by Intervention to Reduce Events) which has been described elsewhere.22 The present report focuses on patients enrolled from hospitals of the Urban Community of Lille (Northern France) during the first and the second EUROASPIRE surveys conducted in 1995 to 1996 and in 1999 to 2000, respectively. Consecutive patients with an established CHD were retrospectively enrolled from hospital admission lists, with the following diagnosis: CHD treated with coronary bypass grafting or percutaneous transluminal coronary angioplasty, acute myocardial infarction, and acute myocardial ischemia. The selected patients were interviewed and examined at least 6 months after their initial admission. The major cardiovascular risk factors and their management were collected from hospital records.
The control group was composed of White subjects recruited within the framework of the WHO-MONICA (Multinational mONItoring of trends and determinants of CArdiovascular disease) project, and randomly sampled from the electoral rolls of the Urban Community of Lille (Northern France) between 1995 and 1997.23 From this population-based sample, we selected 466 males without personal history of CHD, matched by age to the CHD cases.
Search for OAZ1 Gene Polymorphisms
In 24 healthy subjects, a systematic screening of the OAZ1 gene polymorphisms was performed by dHPLC24 or sequencing of the whole gene sequence, from 866 bp upstream of the transcription initiation site to 167 bp downstream of the 3′UTR. The sequences of oligonucleotides used for the polymerase chain reaction (PCR) amplifications as well as the detected polymorphisms (n=21) are listed in the supplemental Table II. All the mutations identified by dHPLC (denaturing High-Performance Liquid Chromatography) were confirmed by sequencing.
Genomic DNA was extracted from white blood cells and genotypes were mainly characterized by restriction enzyme digestion following PCR amplification (Supplemental Table III).
Genotype distributions and allele frequencies between groups were compared using the χ2 test. When necessary, the Fisher’s exact test was applied. Quantitative variables were compared between genotypes by ANOVA, or ANCOVA after adjusting for covariates (General Linear Model procedure). Logistic regressions were performed assuming an allele-dose effect after the goodness-of-fit of the model had been evaluated by the likelihood ratio test. In the in-stent restenosis study, multiple logistic regression tests assessing the risk associated with the polymorphisms were adjusted for age, gender, diabetes, reference diameter, postdilatation diameter, stenosis length, number of stents, and percentage of stenosis before PTCA. Quantification of the impact of OAZ1 polymorphisms on longitudinal CCA-IMT change and CHD risk was adjusted for age, BMI, hypercholesterolemia, hypertension, diabetes, and smoking habits. For the 4-year CCA-IMT change, models were further adjusted for gender and baseline CCA-IMT. Analyses were carried out using the SAS release 8.2 software (SAS Statistical Institute). The pairwise linkage disequilibrium (LD) between SNPs was calculated using the standard definition of r2. Then, a set of tag SNPs that captured all the variants identified in the OAZ1 gene was selected using the program Tagger implemented in Haploview (http://www.broad.mit.edu/mpg/tagger/). The haplotype frequencies were estimated by using a stochastic-EM (SEM) algorithm implemented in the THESIAS software.25 Haplotype effects were assessed by a likelihood ratio test. Throughout, probability values <0.05 were interpreted as significant.
Genetic Data and Selection of Relevant Polymorphisms for Association Studies
Systematic dHPLC-screening and sequencing of the whole OAZ1 sequence in 24 healthy subjects allowed to identify 21 single nucleotide polymorphisms (SNPs), with an average spacing of approximately 200 bp. All the SNPs were named according to their position from the transcription initiation site.
The allele frequencies were determined in a sample of 148 French CHD-free subjects (supplemental Table III). Power calculations indicated that for variants with a minor allele frequency less than 10%, the power to detect an odds ratio (OR) ≥3 did not exceed 75%, especially in the in-stent restenosis study which included the lowest number of patients. We therefore took forward only the 15 SNPs with minor allele frequencies above 10% for further analyses.
A linkage disequilibrium map (Figure) was generated, and the program Haploview/Tagger was then used to determine tag SNPs from this map. The results enabled to select a final set of 4 tag SNPs which captured all of the common SNPs identified in OAZ1 with a mean pairwise r2>0.8. The +1804G/A (rs2523175) SNP was a tag SNP for the −378A/G (rs2854110), +1662C/T (rs2523174), +1763C/T (rs2074457), and +2061insT (rs28384674) SNPs. The +2222A/G (rs2074458) tag SNP captured both the +580T/C (rs2250625) and the +1115T/C (rs2523173) SNPs whereas the +849 C/T (rs2523172) tag SNP captured the +2790C/T (rs2523176) SNP. Finally, the +310C/T (rs4806836), +867C/T (rs12981166), +1391G/A (rs12461851), and +2992C/T (rs12983632) SNPs were tagged by the +851G/T (rs12979409) SNP. On the basis of their minor allele frequencies and the samples size, the 4 +849C/T, +851G/T, +1804G/A, and +2222A/G tag SNPs would allow detecting an OR of 2.6, 2.9, 2.7, and 2.5, respectively, in the in stent-restenosis study, and an OR of 1.5, 1.7, 1.6, and 1.5, respectively, in the CHD case–control study, with a power of at least 75% at the 0.05 significance level. Hence, the +849C/T, +851G/T, +1804G/A, and +2222A/G tag SNPs were selected to test the impact of OAZ1 genetic variation on clinical and subclinical manifestations of VSMC proliferation.
Association of OAZ1 SNPs With In-Stent Restenosis Risk
The 4 selected OAZ1 tag SNPs were genotyped in 205 patients who underwent a Palmaz-Schatz stent implantation (Table 1). The genotype distributions fulfilled the Hardy-Weinberg equilibrium. The subjects were classified in 2 groups defined by the presence (>50% diameter stenosis) or the absence (≤50% diameter stenosis) of restenosis at the 6-month follow-up, and the genotype distributions were compared between the 2 groups. No significant difference in genotype distribution could be detected for the +1804G/A tag SNP. Under the hypothesis of a codominant effect, the +849C/T, +851G/T, and +2222A/G tag SNPs were significantly associated with the risk of in-stent restenosis (P for trend of 0.027, 0.025, and 0.0013, respectively). The presence of the +849T allele was associated with a decreased risk of restenosis (OR=0.5; 95% CI=[0.2 to 0.9]; P=0.030) which did not persist after adjustment for covariates. The presence of the +851T allele was associated with an increased risk of restenosis (OR=2.1; 95% CI=[1.0 to 3.9]; P=0.043, after adjustment). Concerning the +2222A/G SNP, the rare +2222G allele was associated with an increased risk of restenosis (OR=2.2; 95% CI=[1.3 to 3.5]; P=0.0017) which remained significant after adjustment for covariates (OR=2.1; 95% CI=[1.2 to 3.6]; P=0.0071). Compared with the +2222AA frequent homozygotes, the adjusted OR was 2.4 (95% CI=]1.0 to 5.7;]; P=0.037) for the +2222AG heterozygotes, and 4.2 (95%CI=[1.4 to 13.0]; P=0.012) for the +2222GG homozygotes.
Association of OAZ1 SNPs With Longitudinal CCA-IMT Change
In the EVA study, the genotype distributions of the 4 OAZ1 polymorphisms were compatible with the Hardy–Weinberg distribution (Table 2). Overall, the mean baseline CCA-IMT did not differ between subgroups of genotypes whatever the SNP considered. The mean 4-year longitudinal CCA-IMT change (ΔCCA-IMT) was compared between subgroups of genotypes. No significant difference could be detected between subgroups of +849C/T, +851G/T, and +1804G/A genotypes. Conversely, subjects carrying at least one +2222G allele consistently showed an increase in ΔCCA-IMT compared with frequent +2222AA homozygotes (adjusted P trend=0.015).
Association of OAZ1 SNPs With Coronary Heart Disease
The OAZ1 genotypes were obtained in 466 male patients with CHD and 466 age-matched controls (Table 3), and the genotype distributions were in accordance with the Hardy-Weinberg equilibrium. The distribution of the +849C/T, +851G/T, and +1804G/A genotypes did not differ between CHD cases and controls. Conversely, the +2222G allele was more frequent in cases than in controls (P trend=0.044). Assuming an allele-dose effect, the risk of CHD associated with the +2222G allele was estimated to 1.2 (95% CI=]1.0 to 1.5]; P=0.044), and 1.3 (95% CI=[1.1 to 1.6]; P=0.026) after adjustment for covariates. Compared with the +2222AA frequent homozygotes, the adjusted odds-ratio was 1.3 (95%CI=[0.9 to 1.8]; P=0.095) for the +2222 AG heterozygotes, and 1.6 (95%CI=[1.0 to 2.6]; P=0.050) for the +2222GG homozygotes.
We reported associations between genetic polymorphisms of OAZ1, an enzyme that regulates polyamines metabolism, and clinical and subclinical outcomes involving VSMCs. In vitro experimental evidences assign a crucial role to polyamines in the growth and the proliferation of VSMCs. OAZ1 is very closely implicated in the complex regulation of the intracellular pool of polyamines and, as a consequence, in cellular proliferation.
Because most of the OAZ1 variants described in the NCBI/SNP database were not validated and only one was represented in the HapMap database, we searched for OAZ1 SNPs by systematic d-high-performance liquid chromatography (d-HPLC) screening and sequencing. Among the 21 identified SNPs, 4 were selected for association studies: the +849C/T, +851G/T, +1804G/A, and +2222A/G tag SNPs which captured all of the common SNPs.
Among these tag SNPs, only the +2222A/G polymorphism was consistently associated with all the phenotypes. The highly significant association of this polymorphism with in-stent restenosis was of particular interest because restenosis after coronary stenting is primarily the consequence of a neointimal hyperplasia within the stent.2 Conversely, we did not find any association between the +2222A/G polymorphism and restenosis after a conventional balloon angioplasty where the contribution of neointimal hyperplasia is limited (data not shown).26 Consequently, our study strongly argues in favor of an impact of OAZ1 on VSMC biology, at least in the in-stent restenosis process. Albeit significant, the associations of the +2222A/G SNP with the 4-year ΔCCA-IMT and the CHD risk were a little weaker. This discrepancy may be explained, at least in part, by the different involvement of VSMC according to the disease phenotype. In-stent restenosis, which is initiated by mechanical, biochemical, or immunologic injury to the vessel wall, is known to involve VSMC proliferation as the primary pathophysiologic mechanism.2,27 By contrast, the exact function of VSMCs in intima-media thickening and atherosclerosis lesion formation is still debated. Mechanisms for IMT progression likely include macrophage infiltration, extracellular matrix reorganization, as well as a perturbation in the balance between VSMC proliferation and cell death.3,28 In early atherosclerosis, VSMCs may contribute to the development of the atheroma by accelerating lipid accumulation or macrophage chemotaxis whereas VSMCs may also be important in maintaining the stability of the plaque through the formation of a fibrous cap.29,30 In addition to these differences, some discrepancies exist in the disease progress. In-stent restenosis occurs within the first 6 months after stent implantation, thus providing a model in which VSMC proliferation and neointimal hyperplasia are particularly accelerated. By contrast, intima-media thickening and atherosclerotic lesion formation are more progressive processes. Nevertheless, our results for the +2222A/G SNP were noteworthy consistent in the 3 association studies.
Significant and consistent associations of the +2222A/G polymorphism with all the phenotypes could not be extended to the 3 other OAZ1 tag SNPs. A haplotype analysis was performed in an attempt to clarify this discrepancy (supplemental Table IV). Because of the high linkage disequilibrium, only 4 haplotypes were frequent enough to be explored (frequency >10%), and 2 of the 3 mutated haplotypes carried the rare +2222G allele. Despite a trend toward an increased risk of in stent restenosis and a higher ΔCCA-IMT in patients carrying haplotypes with the rare +2222G allele, the haplotype analysis failed to consistently associate a unique mutated haplotype to the various phenotypes.
The +2222A/G SNP was located in intron 2, 40 bp upstream of the 3′ splice site, that is potentially in the lariat branchpoint sequence. Interestingly, expressed sequence tags (EST) described in the UCSC genome browser database suggested the existence of an alternative OAZ1 mRNA characterized by the retention of the second intron. Predictive sequence of this alternative OAZ1 transcript showed a premature stop codon which may lead to either the synthesis of a truncated OAZ1 protein or a non-sense–mediated message decay. Consistently, using mRNA isolated from human VSMC lines, we were able to confirm the coexistence of both normal and alternatively spliced mRNA isoforms by RT-PCR (data not shown). Further experiments are needed to evaluate the impact of the OAZ1 +2222A/G polymorphism on alternative OAZ1 mRNA synthesis and to determine the biological activities associated with the alternative OAZ1 mRNA isoform. However, some data showed that mutations in branchpoint sequence may cause intron retention resulting in human inherited disorders.31 Altogether, these data reinforce the hypothesis of a functional impact of the +2222 tag SNP, this hypothesis being also supported by the fact that no relevant functional arguments could be found for the 2 other SNPs captured by the +2222A/G polymorphism. However, we cannot exclude a lack of statistical power to detect a significant but lower effect of another OAZ1 SNP.
Our results suggested an association of the OAZ1 +2222A/G polymorphism per se with various subclinical or clinical vascular events. This association should be clarified by assessing associations of the SNP with accurate phenotypes, particularly the intracellular polyamine content. Unfortunately, measurements of cellular polyamines were not available in our population samples. Moreover, because of the possibility of a type I error (false-positive) resulting from multiple testing, it may be argued that the Bonferroni correction should be applied. When using this correction with the single test significance level established as α=0.0125 (α=0.05 divided by 4, ie, the total number of SNPs tested), the association between the OAZ1+2222A/G SNP and in-stent restenosis risk remained significant (P trend=0.0013). Conversely, the associations of the OAZ1+2222A/G SNP with 4-year ΔCCA-IMT and CHD risk were only at borderline significance (P trend=0.015 and 0.044, respectively). However, because some of the 4 SNPs tested are in moderate LD resulting in a low degree of independence between markers, the Bonferroni correction method may be too conservative and may introduce the risk of type II errors (false-negative). Finally, the consistency of our results in the 3 independent association studies supports the hypothesis of a significant impact of OAZ1 gene polymorphisms on vascular pathophysiology.
In conclusion, we showed evidence for a potential influence of the OAZ1 gene on the risk of subclinical and clinical vascular events in 3 independent association studies. Our findings reinforce the hypothesis of a link between the polyamine pathway and the vascular pathophysiology, and could provide a potential genetic target for improving the detection of patients at increased risk of vascular diseases.
Sources of Funding
Julie Dumont was supported by the Conseil Régional Nord – Pas-de-Calais and the Institut National de la Santé et de la Recherche Médicale (Inserm). mRNA isolated from human VSMC were a kind gift from Dr Brigitte Jude and Dr Florence Pinet. The restenosis study was supported by a grant from the Direction de la Recherche et des Etudes Doctorales, Inserm, and the Institut Pasteur de Lille. The EVA study was supported by Inserm, the Merck, Sharp and Dohme-Chibret Co and EISAI Co (France). The EUROASPIRE study was supported by an educational grant made to the European Society of Cardiology, Merck, Sharp and Dohme-Chibret Co. The WHO-MONICA population study (Lille) was supported by the Conseil Régional du Nord-Pas de Calais, the Fondation pour la Recherche Médicale, ONIVINS, the Parke-Davis Laboratory, the Mutuelle Générale de l’Education Nationale, the Réseau National de Santé Publique, the Direction Générale de La Santé, Inserm, the Institut Pasteur de Lille, the Université de Lille 2, and the Unité d’Evaluation du Centre Hospitalier et Universitaire de Lille.
Original received September 8, 2006; final version accepted August 2, 2007.
Hoffmann R, Mintz GS. Coronary in-stent restenosis - predictors, treatment and prevention. Eur Heart J. 2000; 21: 1739–1749.
Vazquez-Padron RI, Lasko D, Li S, Louis L, Pestana IA, Pang M, Liotta C, Fornoni A, Aitouche A, Pham SM. Aging exacerbates neointimal formation, and increases proliferation and reduces susceptibility to apoptosis of vascular smooth muscle cells in mice. J Vasc Surg. 2004; 40: 1199–1207.
Durante W, Liao L, Peyton KJ, Schafer AI. Thrombin stimulates vascular smooth muscle cell polyamine synthesis by inducing cationic amino acid transporter and ornithine decarboxylase gene expression. Circ Res. 1998; 83: 217–223.
Bauer PM, Buga GM, Ignarro LJ. Role of p42/p44 mitogen-activated-protein kinase and p21waf1/cip1 in the regulation of vascular smooth muscle cell proliferation by nitric oxide. Proc Natl Acad Sci U S A. 2001; 98: 12802–12807.
Murakami Y, Matsufuji S, Miyazaki Y, Hayashi S. Forced expression of antizyme abolishes ornithine decarboxylase activity, suppresses cellular levels of polyamines and inhibits cell growth. Biochem J. 1994; 304: 183–187.
Newman RM, Mobascher A, Mangold U, Koike C, Diah S, Schmidt M, Finley D, Zetter BR. Antizyme targets cyclin D1 for degradation. A novel mechanism for cell growth repression. J Biol Chem. 2004; 279: 41504–41511.
Humphries S, Bauters C, Meirhaeghe A, Luong L, Bertrand M, Amouyel P. The 5A6A polymorphism in the promoter of the stromelysin-1 (MMP3) gene as a risk factor for restenosis. Eur Heart J. 2002; 23: 721–725.
Amant C, Bauters C, Bodart JC, Lablanche JM, Grollier G, Danchin N, Hamon M, Richard F, Helbecque N, McFadden EP, Amouyel P, Bertrand ME. D allele of the angiotensin I-converting enzyme is a major risk factor for restenosis after coronary stenting. Circulation. 1997; 96: 56–60.
Bauters C, Banos JL, Van Belle E, Mc Fadden EP, Lablanche JM, Bertrand ME. Six-month angiographic outcome after successful repeat percutaneous intervention for in-stent restenosis. Circulation. 1998; 97: 318–321.
Bonithon-Kopp C, Touboul PJ, Berr C, Leroux C, Mainard F, Courbon D, Ducimetiere P. Relation of intima-media thickness to atherosclerotic plaques in carotid arteries. The Vascular Aging (EVA) Study. Arterioscler Thromb Vasc Biol. 1996; 16: 310–316.
Bonithon-Kopp C, Touboul PJ, Berr C, Magne C, Ducimetiere P. Factors of carotid arterial enlargement in a population aged 59 to 71 years: the EVA study. Stroke. 1996; 27: 654–660.
EUROASPIRE. A European Society of Cardiology survey of secondary prevention of coronary heart disease: principal results. EUROASPIRE Study Group. European Action on Secondary Prevention through Intervention to Reduce Events. Eur Heart J. 1997; 18: 1569–1582.
Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994; 90: 583–612.
Mintz GS, Popma JJ, Pichard AD, Kent KM, Satler LF, Wong C, Hong MK, Kovach JA, Leon MB. Arterial remodeling after coronary angioplasty: a serial intravascular ultrasound study. Circulation. 1996; 94: 35–43.
Braun-Dullaeus RC, Mann MJ, Dzau VJ. Cell cycle progression: new therapeutic target for vascular proliferative disease. Circulation. 1998; 98: 82–89.
Korshunov VA, Berk BC. Flow-induced vascular remodeling in the mouse: a model for carotid intima-media thickening. Arterioscler Thromb Vasc Biol. 2003; 23: 2185–2191.
Schwartz SM, deBlois D, O’Brien ER. The intima. Soil for atherosclerosis and restenosis. Circ Res. 1995; 77: 445–465.