Polymorphisms of genes involved in the hypoxia signaling pathway and the development of abdominal aortic aneurysms or large-artery atherosclerosis

Polymorphisms of genes involved in the hypoxia signaling pathway and the development of abdominal aortic aneurysms or large-artery atherosclerosis Ewa Strauss, PhD,a Krzysztof Waliszewski, MD, PhD,b Grzegorz Oszkinis, MD, PhD,b and Ryszard Staniszewski, MD, PhD,b Poznan, Poland Background: The pathogenesis of aortic diseases, both aneurysmal and occlusive, is associated with the occurrence of local ischemic/hypoxic conditions, but the genetic factors that differentiate the predisposition to specific types of aortic diseases are largely unknown. In this study, the functional variants in genes involved in the hypoxia signaling pathway, hypoxia-inducible factor-1a (HIF1A) 1772C>T, 1790G>A, and vascular endothelial growth factor (VEGFA) L634G>C, were analyzed in search of the associations specific to abdominal aortic aneurysm (AAA) development. Methods: The study encompassed a series of 518 patients with AAA, 354 patients with aortoiliac occlusive disease, and 541 controls. In AAA patients, the occurrence of peripheral arterial disease (PAD) was examined with duplex arterial scanning. Genotypes were determined by the polymerase chain reaction/restriction fragment length polymorphism method or with TaqMan probes. Results: In univariate analysis, a significantly increased risk for development of AAA without coexisting PAD was found in VEGFA L634C allele carriers (effect of allele dose: odds ratio [OR], 1.38; P [ .012). In VEGFA L634CC homozygotes, the risk was enhanced by the interaction with HIF1A 1772CC-1790GG genotype (OR, 2.41; P [ .008). This joint effect of homozygous genotypes also influenced the AAA risk independently of PAD coexistence (OR, 1.87; P [ .036). In contrast, the minor allele of the HIF1A 1772C>T polymorphism (1772T and 1772T-1790G haplotype) was significantly associated with the occurrence of AAA with concomitant PAD (OR, 2.02; P [ .009 for the dominant model). This effect was enhanced in the VEGF L634GG homozygotes (OR, 2.86; P [ .005) and among smokers (OR, 3.10; P [ .001). The individual effects of the HIF1A 1772 and VEGFA L634 polymorphisms on AAA risk remained significant in multivariable analysis after adjustment for the traditional vascular risk factors and analyzed polymorphisms. None of the studied variants influenced the risk of aortoiliac occlusive disease. Conclusions: This study identifies polymorphisms in the HIF1A and VEGF genes as potential genetic markers that indicate the predisposition to either AAA coexisting with peripheral atherosclerosis or AAA without such lesions, suggesting the genetic heterogeneity of this disease. The HIF1A 1772T allele also seems to be a genetic risk factor that determines sensitivity to cigarette smoke exposure. Further work is needed to confirm the findings in an independent samples set and to study the functional role of studied variants in AAA. (J Vasc Surg 2014;-:1-9.) Clinical Relevance: Abdominal aortic aneurysm (AAA) is a life-threatening disease with poorly understood genetic background. We hypothesized that the genes involved in the vascular response to hypoxia could harbor changes that differentiate the predisposition to aneurysmal and occlusive/atherosclerotic types of aortic diseases. This study reveals the significant differences in the distribution of functional variants in the genes encoding the hypoxia-inducible factor-1a (HIF1A) and the vascular endothelial growth factor (VEGFA) between AAA patients with and without concomitant peripheral atherosclerosis. This genetic variation should be considered in future studies concerning the development of gene-based prognostic scores and antiangiogenic medical therapy for AAAs.

From the Institute of Human Genetics of the Polish Academy of Sciencesa; and the Department of General and Vascular Surgery, Poznan University of Medical Sciences.b The work has been supported by the National Science Centre in Poland under grant No. NN403250440. Author conflict of interest: none. Additional material for this article may be found online at www.jvascsurg.org. Reprint requests: Ewa Strauss, PhD, Institute of Human Genetics of the Polish Academy of Sciences, Strzeszynska 32, 60-479 Poznan, Poland (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214/$36.00 Copyright Ó 2014 by the Society for Vascular Surgery. http://dx.doi.org/10.1016/j.jvs.2014.02.007

Abdominal aortic aneurysm (AAA) is an abnormal progressive dilation of the infrarenal aorta associated with a risk of aortic rupture at advanced stages.1 The disease is related to high mortality rate because of its mostly asymptomatic course of development. It occurs mainly in elderly men and affects w9% of adults older than 65 years. Because the causal mechanisms of AAA initiation and rupture are still not fully elucidated,2 surgical repair is the only available method of treatment. One of the proposed mechanisms of AAA development is the hypoxia-mediated weakening of the aortic wall.3 Studies have demonstrated that hypoxia exists in vivo in aneurysms and may accelerate the production of matrix 1

JOURNAL OF VASCULAR SURGERY --- 2014

2 Strauss et al

metalloproteinases4 and increase inflammation and excessive neovascularization.5 AAA tissue has been reported to display increased immunopositivity for the hypoxiainducible factor-1a (HIF-1a) and the endothelial Ets-1 transcription factor in comparison to normal aortic tissue.4 HIF-1a is an oxygen-sensitive subunit of the HIF-1 transcription factor, which, in reduced oxygen concentration, induces the expression of hypoxia-inducible genes,6 including Ets-1.4 In AAA tissue, HIF-1a and Ets-1 localize to vascular smooth muscle cells and inflammatory infiltrate, with a predominance of expression observed in the medial layer of the aneurysm wall. Matrix metalloproteinase 2 (MMP-2) localizes to the same regions, and the highest HIF-1a, Ets-1, and MMP-2 expression levels have been reported in the tunica media distant from the vasa vasorum and in the most dilated part of the AAA sac. The AAA tissue immediately flanking the vasa vasorum constituted the only region found to be negative for Ets-1, which indicates that oxygen diffusion from these vessels may be insufficient to protect aortic tissue undergoing extensive matrix remodeling from hypoxia. Concomitant atherosclerosis is present in most AAA cases. Atherosclerosis is also related to localized hypoxia.7 However, there are important differences in terms of molecular features, epidemiology, and heritability between these two vascular pathologic processes, confirming that they constitute completely distinct disease entities.8 Interindividual differences in vascular response to hypoxia are partially subject to genetic control; thus, the inherited factors may differentiate the predisposition to aneurysmal and occlusive/atherosclerotic types of arterial diseases. The candidate genes are those encoding HIF-1a (HIF1A) and the vascular endothelial growth factor (VEGFA), the key initiator of hypoxia-induced angiogenesis. The HIF1A gene, located at chromosome 14q21-24, is composed of 15 exons.9 The two variants in exon 12, 1772C>T (rs11549465) and 1790G>A (rs11549467), result in amino acid changes (P582S and A588T, respectively) within the oxygen-dependent degradation domain.10 Protein isoforms encoded by the 1772T or 1790A allele display enhanced transactivation capacity under either normoxic or hypoxic conditions.11,12 The VEGFA gene is located at chromosome 6p21.3 and consists of eight exons exhibiting alternative splicing.13 The
634G>C (rs2010963) variant results in decreased basal promoter activity and reduced protein production.14 Our preliminary study found an association between the HIF1A 1772T allele, smoking, and the development of an AAA phenotype with concomitant peripheral arterial disease (PAD).15 However, that study was too small to assess the individual role of a single variant in casecontrol analysis. In this work, we extended our pilot study to include the HIF1A 1790 and VEGFA
634 singlenucleotide polymorphisms (SNPs), enlarged the group of AAA patients characterized in terms of concomitant PAD, and included an additional reference group consisting of patients with atherosclerosis in the location partially shared with AAAs (aortoiliac occlusive disease [AIOD])16

to provide a more extensive verification of the hypothesis that the functional variants of hypoxia-related genes are the inherited factors differentiating the predisposition to aneurysmal or to occlusive/atherosclerotic types of arterial diseases. METHODS Study population. The study enrolled 872 patients scheduled for surgery because of either AAA (518 patients) or AIOD (354 patients), referred to the Department of General and Vascular Surgery of the Poznan University of Medical Sciences in the years 1999 to 2011. The preliminary diagnosis was based on ultrasound scanning. The presence of AAA was evaluated by computed tomography angiography or magnetic resonance angiography; the presence of AIOD was evaluated by computed tomography angiography alone. The control group consisted of 541 subjects selected during the same time from the Poznan district. The exclusion criteria included known aneurysms, PAD, and age younger than 37 years for the controls and the presence of AAA for the AIOD group. Intervieweradministered questionnaires were used to obtain information concerning cigarette smoking history (former and current smokers were examined together). In 337 AAA patients, the presence of PAD was determined on the basis of physical examinations of pulse in the superficial arteries supplemented with ultrasound duplex color scanning. In 33.2% of the AAA patients, obstruction of the dorsalis pedis or posterior tibial arteries was revealed; in 27.6%, the examinations found obstructions of the femoropopliteal or aortoiliac arteries. As the obstruction of the dorsalis pedis or tibial artery in AAAs may reflect embolism from the aneurysm and the studied patients did not undergo additional hemodynamic studies, only the femoropopliteal or aortoiliac obstructions were considered clinically confirmed PAD. The AAA patients with coexisting PAD had intermittent claudication (met the criteria of stage II PAD as defined by Fontaine), and they had not been previously treated surgically for their PAD. All patients were treated pharmacologically with statins, antiplatelet drugs, and other drugs (antihypertensive or antidiabetic), depending on their clinical condition. The study protocol was approved by the local Bioethics Committee, and all subjects provided their informed consent. Genotyping. Genomic DNA was extracted from circulating blood lymphocytes by use of the chemical method. Polymorphisms were ascertained by the polymerase chain reaction/restriction fragment length polymorphism method in accordance with previously described procedures.17,18 The genotypes of the HIF1A 1790 variant were validated by a predesigned TaqMan SNP genotyping assay employing the ABI 7900HT Fast RealTime PCR System (Life Technologies, Carlsbad, Calif). The allelic discrimination test gave a 98% agreement with the polymerase chain reaction/restriction fragment length polymorphism method, and these results were accepted. Data analysis. Genotype frequencies were tested for Hardy-Weinberg equilibrium by the c2 test

JOURNAL OF VASCULAR SURGERY Volume -, Number -

Strauss et al 3

Table I. Clinical characteristics of patients with abdominal aortic aneurysm (AAA), patients with aortoiliac occlusive disease (AIOD), and control subjects P value for comparison of groups (if significant) Variable

Controls (n ¼ 541)

Median age, years 60 (53, 68) Age range, years 37-94 Male gender 69.9 Smoking 36.6 Hypertension 42.5 Diabetes 12.8 Obesity 25.7 Aortic diameter d Lipids and lipoproteins profile, mmol/L TC 5.6 (4.7, 6.6) HDLC 1.4 (1.1, 1.7) Decreased HDLC level 13.3 LDLC 3.5 (2.7, 4.4) TG 1.4 (1.1, 2.0)

AAA (n ¼ 518)

AIOD (n ¼ 354)

69 (62, 75) 40-94 86.3 80.7 73.0 16.2 17.4 60.0 (54, 71)

59 (54, 66) 43-86 69.3 91.3 63.8 20.3 11.5 d

5.2 (4.3, 6.1) 1.1 (0.9, 1.4) 37.0 3.2 (2.4, 4.1) 1.5 (1.1, 2.0)

5.3 (4.5, 6.5) 1.2 (1.0, 1.4) 34.2 3.3 (2.5, 4.3) 1.6 (1.2, 2.2)

AAA vs controls

AIOD vs controls

<.0001 <.0001 <.0001 <.0001 .030

<.0001 <.0001

AAA vs AIOD <.0001

<.0001 <.0001 <.0001 <.0001

<.0001 <.001 .030 .001

<.0001 <.0001

HDLC, high-density lipoprotein cholesterol; LDLC, low-density lipoprotein cholesterol; TC, total plasma cholesterol; TG, triglyceride. Decreased HDLC level: <1.0 mmol/L for men and <1.2 mmol/L for women. Variables are expressed as median (interquartile range) or as percentages. The differences in the distribution of potentially modifiable vascular risk factors were evaluated with adjustment for age and gender.

(http://ihg.gsf.de/cgi-bin/hw/hwa1.pl). Haploview v4.2 was used for HIF1A haplotype analysis, and QUANTO software (version 1.2.4; http://hydra.usc.edu/GxE/) was employed for sample size calculations. The statistical power of the study to detect the alleles associated with an odds ratio (OR) of 1.4 to 2.0 for the SNPs with a frequency of 0.100 to 0.300 (in the additive or dominant model) was 80%. This means that the study had sufficient power to detect an association of the HIF1A 1772 and VEGFA
634 SNPs with AAA phenotypes of different advancement of atherosclerosis in the case-control analysis. Univariate analyses were used to compare the demographic parameters of the studied groups: the MannWhitney U test for age and the c2 test for sex ratio. Differences in the distributions of genotype and modifiable risk factors were assessed with adjustment for age (in 10-year strata) and gender by logistic (for qualitative variables) or conditional (for quantitative variables) regression analysis. Continuous variables were analyzed after logarithmic transformation because of the deviation of distribution from the normal curve. The relations between the traditional vascular risk factors (age, gender, smoking status, arterial hypertension, diabetes, obesity, decreased highdensity lipoprotein cholesterol level) and genotype distribution were also investigated. These analyses were performed with Statistica v8.0 software, and the observed differences were considered significant at P < .05. The Bonferroni correction for multiple comparisons was applied to the two-loci interaction analysis. RESULTS Traditional risk factors distribution. The development of both aortic diseases was significantly associated with smoking, hypertension, diabetes, and high-density lipoprotein cholesterol level (Table I). The AAA cases were

older (required surgical intervention on average a decade later), and a higher percentage of them were men (6.3-fold male predominance was revealed, compared with 2.5-fold in the AIOD group). The profiles of the modifiable risk factors did not differ between AAA and AIOD in such categories as hypertension and blood lipid parameters; in turn, smoking and diabetes were more frequently represented in AIOD, whereas obesity was more frequent in the AAA group. The AAA patients with concomitant PAD, compared with those without, had a significantly higher frequency of smokers (88.2% vs 77.3%; P ¼ .02) and smaller aortic diameters (median [interquartile range], 57.5 [48, 69] mm vs 62.5 [55, 72] mm; P < .01). Genotype distribution. The genotype distributions in the three studied SNPs (Table II) showed no deviations from the Hardy-Weinberg equilibrium in either the patients or the controls. Because of the small distance between the two HIF1A variants, only the three common 1772-1790 haplotypes were observed in the studied population: C-G, T-G, and C-A. As a consequence, the effects of the 1772T and 1790A alleles correspond to the impacts of the 1772T-1790G and 1772C-1790A haplotypes, respectively. The main results of the one-locus analysis are presented in Table III (see also Supplementary Table I, online only for details of the statistical analysis), whereas the statistically significant two-loci interactions are shown in Table IV. We observed that without the stratification of AAA patients by PAD coexistence, only the joint effect of the HIF1A 1772CC-1790GG and VEGFA
634CC genotypes significantly influenced the AAA risk (OR, 1.78; P ¼ .036; Table IV), but there were no significant effects of the individual genotypes (heterozygotes, homozygotes) or alleles (in dominant, recessive, and additive models) on AAA or

JOURNAL OF VASCULAR SURGERY --- 2014

4 Strauss et al

Table II. Distribution of genotypes, alleles, and haplotypes of the hypoxia-inducible factor-1a (HIF1A) and vascular endothelial growth factor (VEGFA) polymorphisms in patients with abdominal aortic aneurysms (AAAs) stratified according to peripheral arterial disease (PAD) coexistence, patients with aortoiliac occlusive disease (AIOD), and controls AAA Genotype

Controls (n ¼ 541)

All (N ¼ 518)

Without PAD (n ¼ 244)

HIF1A 1772 (rs11549465) CC 463 (86.5) 448 (86.7) 66 (12.7) CT 69 (12.9)a TT MAF 3 (0.6) .070 3 (0.6) .070 HIF1A 1790 (rs11549467) GG 509 (94.1) 494 (95.4) GA 31 (5.7) 23 (4.4) AA MAF 1 (0.2) .030 1 (0.2) .024 HIF1A 1772-1790 (rs11549465-rs11549467) CC-GG 434 (81.1) 425 (82.2) CC-GA 28 (5.2) 22 (4.3) CC-AA 1 (0.2) 1 (0.2) CT-GG 66 (12.3) 66 (12.8) CT-GA 3 (0.6) 0 (0.0) TT-GG 3 (0.6) 3 (0.6) TT-GA 0 (0.0) 0 (0.0) d 69 (13.4) (CTþTT)-GG 69 (12.9) Haplotype frequency C-G .899 .907 C-A .031 .023 T-G .070 .070 T-A .000 .000 VEGFA
634 (rs2010963) GG 280 (52.5) 253 (49.2) 213 (41.4) GC 218 (40.9)g CC MAF 35 (6.6) .270 48 (9.3) .301 HIF1A 1772-1790/VEGFA
634 (rs11549465-rs11549467/rs2010963) CC-GG/GG 227 (43.1) 201 (39.2) CC-GG/GC 174 (33.0) 177 (34.5) i,j 43 (8.4)i CC-GG/CC 26 (4.9) l CT-GG/GG 34 (6.5) 35 (6.8) CT-GG/GC 25 (4.7) 27 (5.3) CT-GG/CC 6 (1.1) 4 (0.8) CC-GA/GG 13 (2.5) 14 (2.7) CC-GA/GC 14 (2.7) 7 (1.4) CC-GA/CC 1 (0.2) 1 (0.2) TT-GG/GG 1 (0.2) 2 (0.4) Other rare <1% 6 (1.1) 2 (0.4) 37 (7.2) (CTþTT)-GG/GG 35 (6.7)o

With PAD (n ¼ 93)

AIOD (n ¼ 354)

215 (88.1) 28 (11.5)b 1 (0.4) .06

70 (76.1) 21 (22.8)a-c 1 (1.1) .125

303 (85.6) 49 (13.8)c 2 (0.6) .075

231 (94.7) 12 (4.9) 1 (0.4) .029

91 (97.9) 2 (2.15) 0 (0.0) .011

335 (94.6) 19 (5.4) 0 (0.0) .027

202 12 1 28 0 1 0 29

69 1 0 21 0 1 0 22

285 18 0 49 0 1 1 50

(82.8) (4.9) (0.4) (11.5) (0.0) (0.4) (0.0) (11.9)e

(75.0) (1.1) (0.0) (22.8) (0.0) (1.1) (0.0) (23.9)d-f

(80.5) (5.1) (0.0) (13.8) (0.0) (0.3) (0.3) (14.1)f

.910 .029 .061 .000

.870 .005 .125 .000

.900 .025 .073 .001

108 (44.8) 106 (44.0)g,h 27 (11.2) .332

50 (54.4) 36 (39.1) 6 (6.5) .261

182 (51.7) 147 (41.8)h 23 (6.5) .274

34 29 5 14 6 1 1 0 0 1 0 15

145 119 19 21 25 3 15 2 1 0 2 21

87 88 24 11 15 2 8 3 1 1 1 12

(36.1) (36.5) (10.0)j,k (4.6)m (6.2) (0.8) (3.3) (1.2) (0.4) (0.4) (0.4) (5.0)p

(37.4) (31.9) (5.5) (15.4)l-n (6.6) (1.1) (1.1) (0.0) (0.0) (1.1) (0.0) (16.4)o-r

(41.2) (33.8) (5.4)k (6.0)n (7.1) (0.9) (4.3) (0.6) (0.3) (0.0) (0.6) (6.0)r

MAF, Minor allele frequency. Major association between the genotypes and the studied diseases. a-c HIF1A 1772CTþTT genotypes: a, b, P < .01; c, P < .05. d-f HIF1A 1772(CTþTT)-1790 GG genotypes: d, e, P < .01; f, P < .05. g,h Dose of the VEGFA
634C allele: g, h, P < .05 (see Supplementary Table I, online only, for details). i-r Differences in the frequency of the HIF1A/VEGFA genotype combinations: i, k, P < .05; j, l, n, o, P < .01; m, p, r, P # .001.

AIOD risk (Supplementary Table I, online only). The studied polymorphisms were also not correlated with the aortic diameter or age of AAA patients, even when we excluded the cases with the largest, accidentally detected, or ruptured aneurysms (above median diameter) from analysis. Genotype distribution in relation to the coexistence of PAD in AAA. In the AAA group, the genotype frequencies did not differ significantly between patients with femoropopliteal and aortoiliac occlusions or between patients without arterial obstructions and those with obstructions of the dorsalis pedis or tibial arteries. Therefore,

in further studies, these patients were analyzed together (as subgroups of AAA patients with and without clinically confirmed PAD, respectively). Compared with AAA patients without concomitant PAD, the subgroup with coexisting PAD had a significantly higher frequency of HIF1A 1772T allele carriers (23.9% vs 11.9%; P < .01; Table II and Supplementary Table I, online only) and an increased percentage of subjects with HIF1A 1772(CTþTT)1790GG/VEGF
634GG genotypes (16.4% vs 5.0%; P < .001; Tables II and IV). In the AIOD group, there was a lower frequency of HIF1A 1772T allele carriers (14.4%)

JOURNAL OF VASCULAR SURGERY Volume -, Number -

Strauss et al 5

Table III. The crude and adjusted odds ratios (ORs) with 95% confidence intervals (CIs) for the associations between the hypoxia-inducible factor-1a (HIF1A) rs11549465 and vascular endothelial growth factor (VEGFA) rs2010963 polymorphisms and development of abdominal aortic aneurysms (AAAs) with or without concomitant peripheral arterial disease (PAD) Multivariate analysisa

Univariate analysis Effect AAA without PAD VEGFA
634G>C (rs2010963) GCþCC CC Dose of C allele AAA with PAD HIF1A 1772C>T (rs11549465) CTþTT Dose of T allele HIF1A 1772C>T (rs11549465) in CTþTT Dose of T allele

OR (95% CI)

P

OR (95% CI)

P

1.36 (1.00-1.85) 1.80 (1.06-3.04) 1.35 (1.07-1.71)

.047 .029 .013

1.50 (1.01-2.22) 1.99 (1.00-3.96) 1.45 (1.07-1.97)

.042 .048 .016

2.02 (1.18-3.47) 1.90 (1.15-3.13) smokers 3.12 (1.55-6.27) 3.12 (1.58-6.18)

.009 .012

2.15 (1.05-4.43) 2.01 (1.03-3.92)

.036 .041

.001 .001

3.63 (1.51-8.71) 3.60 (1.52-8.53)

.004 .003

a Adjusted for the traditional vascular risk factors: age (in 10-year strata), gender, smoking status, arterial hypertension, diabetes, obesity, decreased high-density lipoprotein cholesterol level (<1.0 mmol/L for men and <1.2 mmol/L for women), and HIF1A or VEGFA genotypes.

Table IV. Two-loci interactions between the hypoxia-inducible factor-1a (HIF1A) 1772C>T-1790G>A (rs11549465rs11549467) haplotypes and the vascular endothelial growth factor (VEGFA)
634G>C (rs2010963) polymorphism in relation to abdominal aortic aneurysm (AAA) risk Compared groups AAA AAA AAA AAA AAA

vs controls without PAD vs controls with PAD vs controls without PAD vs AIOD with PAD vs AIOD

AAA with PAD vs AAA without PAD

Genotype combination HIF1A 1772-1790/VEGFA
634 CC-GG/CC CC-GG/CC (CTþTT)-GG/GG CC-GG/CC CT-GG/GG (CTþTT)-GG/GG CT-GG/GG (CTþTT)-GG/GG

OR (95% CI) 1.87 2.41 2.86 2.11 2.87 3.11 3.26 3.20

(1.11-3.15) (1.31-4.42) (1.42-5.79) (1.09-4.07) (1.39-5.89) (1.53-6.32) (1.35-7.88) (1.36-7.53)

P value crude

P value corrected

.018 .004 .003 .028 .003 .001 .007 .006

.036 .008 .005 .056 .009 .003 .014 .012

AIOD, Aortoiliac occlusive disease; CI, confidence interval; OR, odds ratio; PAD, peripheral arterial disease. Only significant interactions between two loci, as identified by logistic regression, are presented. ORs are presented with reference to the homozygous genotype for the most frequent allele (CC-GG/GG). P values (crude and corrected for the two-loci analysis) for comparison of studied and reference groups are shown.

and HIF1A 1772(CTþTT)-1790GG/VEGF
634GG genotypes (6.0%) compared with AAA subjects with concomitant PAD (23.9%, P ¼ .03 and 16.4%, P ¼ .001, respectively) as well as a lower proportion of VEGFA
634C allele carriers (48.3%) in comparison to that among AAA patients without coexisting PAD (55.2%, significant effect of allele dose, P ¼ .03; Tables II and IV and Supplementary Table I, online only). The association of the HIF1A and VEGFA alleles with the two studied AAA subphenotypes reflected the observed differences in the genotype distribution between the patient subgroups. Univariate case-control analysis revealed a significantly increased risk for development of AAA with coexisting PAD in HIF1A 1772T allele (1772T-1790G haplotype) carriers (dominant effect of the allele: OR, 2.02; P ¼ .009; Table III) and in the combinations of genotypes HIF1A 1772(CTþTT)-1790GG/VEGFA
634GG (OR, 2.86; P ¼ .005; Table IV), whereas a significantly increased

risk for development of AAA without coexisting PAD was noted in VEGFA
634C allele carriers (effect of allele dose: OR, 1.38; P ¼ .012; Table III) and in the HIF1A 1772CC-1790GG/VEGFA
634CC homozygotes (OR, 2.41; P ¼ .008; Table IV). In a multivariable logistic regression controlling for conventional vascular risk factors (age, gender, smoking status, arterial hypertension, diabetes, obesity, decreased high-density lipoprotein cholesterol level) and the studied SNPs, the HIF1A 1772C>T polymorphism was a significant independent predictor of AAA with coexisting PAD in dominant (OR, 2.15; P ¼ .036) and additive (OR, 2.01; P ¼ .040) models, whereas the VEGFA
634G>C polymorphism was an independent predictor of AAA without concomitant PAD in all models tested: dominant (OR, 1.5; P ¼ .042), recessive (OR, 1.99; P ¼ .048), and additive (OR, 1.45; P ¼ .016; Table III). Effect of cigarette smoking. The influence of smoking on the association between the HIF1A 1772C>T

6 Strauss et al

SNP and the development of AAA with concomitant PAD was found. The effect of a 2.01-fold increase in risk for 1772T allele carriers, observed in the whole group, was further enhanced among smokers (a 3.10-fold increase in risk; P ¼ .001; Table III; see also Supplementary Table II, online only for information on genotype frequencies in smokers), and this effect remained significant after adjustment for conventional vascular risk factors. In nonsmokers, the impact of this allele was nonsignificant, but there were few nonsmokers among AAA patients with PAD (12.0%), which may affect the results of this analysis. DISCUSSION The data presented in this study provide information addressing the effects of functional polymorphisms in the HIF1A and VEGFA genes on the susceptibility to AAA and AIOD (the results are summarized in the Supplementary Fig, online only). The main outcome was that the HIF1A 1772C>T and VEGFA
634G>C polymorphisms influenced the risk of AAA, but the effect of opposite alleles of a particular SNP was distinct, depending on PAD coexistence. Namely, the VEGFA
634C allele was an independent predictor for development of AAA without concomitant PAD, and this effect was further enhanced by the interaction with the HIF1A 1772C allele (the 1772C-1790G haplotype). Moreover, the joint effect of the HIF1A 1772CC-1790GG and VEGF
634CC genotypes affected the AAA risk independently from PAD coexistence. On the other hand, the HIF1A 1772T allele (the 177T-1790G haplotype) was an independent predictor of the development of AAA with concomitant PAD, and this effect was enhanced by the interaction with the VEGFA
634G allele and in smokers. However, none of studied variants affected the AAA size, the age at surgical treatment, or the risk of AIOD. On the basis of these observations, two mechanisms by which hypoxia-regulated genes promote distinct AAA phenotype formation may be postulated. The first is the less developed vasa vasorum in the HIF1A 1772CC and VEGFA
634CC double homozygotes contributing to aortic tissue degeneration and AAA formation. The second is the upregulation of the inflammatory pathways in smokers with the HIF1A 1772(CTþTT) and VEGFA
634GG genotypes, promoting atherosclerosis development in patients with undefined AAA genetic background. The HIF1A and VEGFA SNPs and the development of AAA and AIOD. A growing number of studies support the thesis that arterial wall hypoxia has a causal role in aneurysm3 and atherosclerosis19 formation. The anatomic structure of the human abdominal aorta results in the predisposition to tissue hypoxia because of the poorly developed vasa vasorum vessels below the renal arteries.20 From this point, the aorta is markedly thinner. This anatomy results in an increased likelihood of AAA formation, especially in the presence of factors that increase aortic wall hypoxia, such as the occurrence of atherosclerotic plaques and intraluminal thrombus21 as well as the stenosis of adventitial vasa vasorum.22

JOURNAL OF VASCULAR SURGERY --- 2014

HIF-1a stimulates protective angiogenic response to hypoxia by enhancing VEGFA gene expression. However, the encoding protein, which also acts as an inflammatory factor,7 may promote the development of inflammation in AAA sacs.23 Both the HIF1A and VEGFA genes are overexpressed in the aortic walls of human and experimental AAAs, contributing to pathogenesis,4,5 but the relation between their expression levels and the coexistence of atherosclerosis has not yet been determined. The genetic differences observed in our study suggest the existence of a constitutive variation between the AAA patients with and without concomitant PAD, which may affect the response of an organism to proangiogenic and antiangiogenic factors.24 The HIF1A 1772T and VEGFA
634G alleles that were associated with the presence of PAD have been previously found to determine increased HIF-1a11,12 and VEGFA14,25 protein levels, thus increasing the stimulation of the hypoxia signaling pathway in response to inducing agents. The opposite 1772C and
634C alleles, associated with AAA without concomitant PAD, appear to be related to insufficient local vascular response to hypoxia and may also result in the limited development of vasa vasorum during ontogenesis. Presently, the use of antiangiogenic therapy is one of the most promising strategies for the medical treatment of AAAs.3 Because the variation in hypoxia-related genes may substantially affect the response to this therapy,24 pharmacogenomic approaches should be applied for the development of such medical treatment. The phenotypic effects of the studied SNPs were initially analyzed in malignant diseases, and there are repeated reports showing the correlations of the HIF1A 1772T and 1790A alleles with increased tumor size and disease aggressiveness, most probably resulting from increased microvessel density in the tumors of the carriers of these variants.11 These alleles are also overrepresented in highperformance athletes and in populations living in highaltitude environments, which reveals that they determine better adaptation to tissue hypoxia.24 In some groups of patients with vascular diseases, the biologic effects of these SNPs seem to be opposite to those observed in patients with malignant diseases and in healthy subjects. For example, it has been demonstrated that in patients with coronary artery disease, the HIF1A 1772T and VEGFA
634G alleles are associated with a reduced collateral vessel density, which suggests decreased angiogenic response to myocardial ischemia.26,27 HIF1A 1772T was also correlated with the development of stable angina rather than acute myocardial infarction as the initial clinical presentation of this disease, most probably as a result of the lower neovascularization within atherosclerotic plaques.28 In contrast, in hemodialysis patients, the HIF1A polymorphisms were associated with acute myocardial infarction and hypotension.29 There were no differences in the incidence of myocardial infarction between the subgroups of AAA patients analyzed in our study, whereas coronary artery disease was slightly more common in those with concomitant PAD (62.0% vs 51.2%; P ¼ .08; data not

JOURNAL OF VASCULAR SURGERY Volume -, Number -

presented), indicating the link between the HIF1A 1772T and VEGFA
634G alleles and the co-occurrence of systemic atherosclerosis in AAAs. Contrary to our observations, the effect of the polymorphisms of genes involved in extracellular matrix remodeling, such as MMP2, MMP3, MMP-13, and TIMP1, on AAA risk seems to be independent of the coexistence of atherosclerosis in other locations30; it may, however, depend on positive family history of AAA.31 The lack of influence of the HIF1A alleles on AIOD risk corresponds to the results of previous studies on PAD.32 In a study including 917 patients and 969 controls, no association was found between both SNPs in exon 12 and the development of PAD, its Fontaine stage, the ankle-brachial index (ABI) of patients, and their age at the onset of the disease. These results indicate that the genetic variants influencing the response to hypoxia in AAAs may be different from those in AIOD. Genetic-environmental interaction. In accordance with other observations concerning the predominant role of cigarette smoking in the etiopathogenesis of largeartery diseases,33,34 smoking was the most common modifiable risk factor for both AAAs and AIOD in our study. In AAA patients, the proportion of smokers was higher among those with coexisting PAD, and smoking enhanced the risk for development of this phenotype determined by the HIF1A 1772T allele, which confirmed the results of our previous study.15 The present results suggest that HIF1A and VEGFA SNPs are independent inherited risk factors for AAA, but the manifestation of vascular phenotype in HIF1A 1772T allele carriers may be also enhanced by cigarette smoking, indicating the role of epigenetic factors in pathogenesis. Oxidative stress and tobacco smoke are known to change the epigenetic modification of chromosomes, such as histone acetylation and inflammatory gene promoter methylation, which contribute to the overexpression of inflammatory genes35 and thus promote causal disease pathways in the development of atherosclerosis. The corresponding genetic-environmental interaction has been previously described with respect to hepatocellular carcinoma occurrence; the risk for the HIF1A 1790A allele carriers was reported as fourfold higher for the nonselected cases and 11-fold higher for smokers.36 The HIF1A 1772T allele saw a similar trend. Cigarette smoking has been associated with increased HIF1-a levels in malignant tissues; therefore, the authors speculated that the interaction effect contributes to pathogenesis because both risk allele and exposure promote HIF1A gene expression, increasing the risk. The same mechanism may be postulated in the case of AAAs. However, the effects of tobacco smoke exposure appear to depend on tissue type as well as on the duration and intensity of hypoxia. For example, studies of vascular cell lines and animal models of acute vascular ischemia have shown that cigarette smoke may also act as an antiangiogenic factor by minimizing HIF-1 and VEGFA production in response to hypoxia.37 Smoking was also correlated with lacking or poor

Strauss et al 7

development of collateral blood vessels in chronic coronary artery disease.38 Despite the observed phenotype variability, the HIF1A 1772T allele, which was found in one in 10 smoking controls and in one in four smoking AAA patients with concomitant PAD, is a promising candidate gene variant for studies concerning individual sensitivity to cigarette smoke exposure. The identification of the groups that are at high risk for development of smoking-dependent lifethreatening diseases allows implementation of primary prevention in these groups to reduce the incidence of premature deaths from vascular causes. It has been observed that the reduction of cigarette smoking in European countries has resulted in a significant diminution of AAA incidence in Sweden and a decrease in mortality related to aneurysm rupture in the United Kingdom.39 Limitations. Our study was primarily limited by the absence of PAD severity measurements, such as the ABI, making it difficult to compare our findings directly with the results of other studies. We used duplex scanning as a more precise method40 because, in many cases, ABIs could not be reliably obtained (accurate ankle pressure measurements were not possible because of incompressible calcified arteries or nonaudible pulses). In addition, all asymptomatic patients who had not undergone the ultrasound control of the lower limb arteries were excluded from the AAA subphenotype analysis. However, when we excluded patients with known PAD from the whole AAA group and reanalyzed the data (including 404 cases and 541 controls in the analysis), the association of the VEGFA
634C allele dose with the AAA without concomitant PAD was still significant (OR, 1.32; P ¼ .012). The second limitation is that this was only a clinical association study, and thus we have no experimental data supporting the idea that the phenotypic effects of the studied alleles observed in patients with other complex diseases are also expressed and functional in the abdominal aorta. However, that functionality may be expected because the HIF1A and VEGFA genes are expressed in AAAs,4,5 and they are not located in imprinted genomic domains (both alleles of particular SNPs are transcribed). Further studies correlating HIF1A and VEGFA genotypes with the density of the vasa vasorum in the aorta as well as with the serum VEGF levels in patients may be helpful in elucidating the potential causal mechanisms of AAA development. CONCLUSIONS In this study, we provide preliminary genetic evidence that the HIF1A 1772 and 1790 and VEGFA
634 SNPs in white populations constitute potential risk factors for the susceptibility to AAA but not to AIOD. These polymorphisms, with the exception of the low-frequency HIF1A 1790 variant, are potential genetic markers that indicate predisposition to either AAA with concomitant PAD or AAA without such lesions. This genetic variation suggests the genetic heterogeneity of the disease. Smoking was a factor that potentiated the genetic risk of atherosclerosis development

JOURNAL OF VASCULAR SURGERY --- 2014

8 Strauss et al

in AAAs. These genetic and environmental factors should be taken into account in the development of prognostic scores and antiangiogenic medical therapy in the treatment of AAAs. The results of this study require replication in other populations for generalization.

16.

17.

AUTHOR CONTRIBUTIONS Conception and design: ES Analysis and interpretation: ES Data collection: ES, KW, GO Writing the article: ES Critical revision of the article: KW, GO, RS Final approval of the article: ES, KW, GO, RS Statistical analysis: ES Obtained funding: ES, GO, RS Overall responsibility: ES

18.

19.

20. 21.

REFERENCES 1. Wassef M, Upchurch GR, Kuivaniemi H, Thompson RW, Tilson MD. Challenges and opportunities in abdominal aortic aneurysm research. J Vasc Surg 2007;45:192-8. 2. Kuivaniemi H, Elmore JR. Opportunities in abdominal aortic aneurysm research: epidemiology, genetics, and pathophysiology. Ann Vasc Surg 2012;26:862-70. 3. Vijaynagar B, Bown MJ, Sayers RD, Choke E. Potential role for antiangiogenic therapy in abdominal aortic aneurysms. Eur J Clin Invest 2013;43:758-65. 4. Erdozain OJ, Pegrum S, Winrow VR, Horrocks M, Stevens CR. Hypoxia in abdominal aortic aneurysm supports a role for HIF-1a and Ets-1 as drivers of matrix metalloproteinase upregulation in human aortic smooth muscle cells. J Vasc Res 2011;48:163-70. 5. Kaneko H, Anzai T, Takahashi T, Kohno T, Shimoda M, Sasaki A, et al. Role of vascular endothelial growth factor-A in development of abdominal aortic aneurysm. Cardiovasc Res 2011;91:358-67. 6. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc Natl Acad Sci U S A 1995;92:5510-4. 7. Holm PW, Riemer HJAS, Clark JZ, Jan LH, René AT. Atherosclerotic plaque development and instability: a dual role for VEGF. Ann Med 2009;41:257-64. 8. Harrison SC, Holmes MV, Agu O, Humphries SE. Genome wide association studies of abdominal aortic aneurysmsdbiological insights and potential translation applications. Atherosclerosis 2011;217:47-56. 9. Semenza GL, Rue EA, Iyer NV, Pang MG, Kearns WG. Assignment of the hypoxia-inducible factor 1a gene to a region of conserved synteny on mouse chromosome 12 and human chromosome 14q. Genomics 1996;34:437-9. 10. Hon WC, Wilson MI, Harlos K, Claridge TDW, Schofield CJ, Pugh CW, et al. Structural basis for the recognition of hydroxyproline in HIF-1a by pVHL. Nature 2002;417:975-8. 11. Tanimoto K, Yoshiga K, Eguchi H, Kaneyasu M, Ukon K, Kumazaki T, et al. Hypoxia-inducible factor-1a polymorphisms associated with enhanced transactivation capacity, implying clinical significance. Carcinogenesis 2003;24:1779-83. 12. Fu XS, Choi E, Bubley GJ, Balk SP. Identification of hypoxia-inducible factor-1a (HIF-1a) polymorphism as a mutation in prostate cancer that prevents normoxia-induced degradation. Prostate 2005;63:215-21. 13. Wei MH, Popescu NC, Lerman MI, Merrill MJ, Zimonjic DB. Localization of the human vascular endothelial growth factor gene, VEGF, at chromosome 6p12. Hum Genet 1996;97:794-7. 14. Stevens A, Soden J, Brenchley PE, Ralph S, Ray DW. Haplotype analysis of the polymorphic human vascular endothelial growth factor gene promoter. Cancer Res 2003;63:812-6. 15. Strauss E, Waliszewski K, Oszkinis G, Staniszewski R. Gene-environment interaction for the HIF1-A 1772C>T polymorphisms and

22.

23.

24. 25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

cigarette smoking increase susceptibility to abdominal aortic aneurysm [in Polish]. Przeg Lek 2012;69:744-9. Kashyap VS, Pavkov ML, Bena JF, Sarac TP, O’Hara PJ, Lyden SP, et al. The management of severe aortoiliac occlusive disease: endovascular therapy rivals open reconstruction. J Vasc Surg 2008;48: 1451-7. Fransen K, Fenech M, Fredrikson M, Dabrosin C, Soderkvist P. Association between ulcerative growth and hypoxia inducible factor-1a polymorphisms in colorectal cancer patients. Mol Carcinog 2006;45: 833-40. Kariyazono H, Ohno T, Khajoee V, Ihara K, Kusuhara K, Kinukawa N, et al. Association of vascular endothelial growth factor (VEGF) and VEGF receptor gene polymorphisms with coronary artery lesions of Kawasaki disease. Pediatr Res 2004;56:953-9. Bjornheden T, Levin M, Evaldsson M, Wiklund O. Evidence of hypoxic areas within the arterial wall in vivo. Arterioscler Thromb Vasc Biol 1999;19:870-6. Wolinsky H, Glagov S. Comparison of abdominal and thoracic aortic medial structure in mammals. Circ Res 1969;25:677-86. Vorp DA, Lee PC, Wang DH, Makaroun MS, Nemoto EM, Ogawa S, et al. Association of intraluminal thrombus in abdominal aortic aneurysm with local hypoxia and wall weakening. J Vasc Surg 2001;34: 291-9. Tanaka H, Zaima N, Sasaki T, Hayasaka T, Goto-Inoue N, Onoue K, et al. Adventitial vasa vasorum arteriosclerosis in abdominal aortic aneurysm. PLoS One 2013;8:e57398. Scott DJ, Allen CJ, Honstvet CA, Hanby AM, Hammond C, Johnson AB, et al. Lymphangiogenesis in abdominal aortic aneurysm. Br J Surg 2013;100:895-903. Rogers MS, D’Amato RJ. Common polymorphisms in angiogenesis. Cold Spring Harb Perspect Med 2012;2. Watson CJ, Webb NJ, Bottomley MJ, Brenchley PE. Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: correlation with variation in VEGF protein production. Cytokine 2000;12:1232-5. Resar JR, Roguin A, Voner J, Nasir K, Hennebry TA, Miller JM, et al. Hypoxia-inducible factor 1a polymorphism and coronary collaterals in patients with ischemic heart disease. Chest 2005;128:787-91. Lin TH, Wang CL, Su HM, Hsu PC, Juo SH, Voon WC, et al. Functional vascular endothelial growth factor gene polymorphisms and diabetes: effect on coronary collaterals in patients with significant coronary artery disease. Clin Chim Acta 2010;411:1688-93. Hlatky MA, Quertermous T, Boothroyd DB, Priest JR, Glassford AJ, Myers RM, et al. Polymorphisms in hypoxia inducible factor 1 and the initial clinical presentation of coronary disease. Am Heart J 2007;154: 1035-42. Zheng ZL, Hwang YH, Kim SK, Kim S, Son MJ, Ro H, et al. Genetic polymorphisms of hypoxia-inducible factor-1 alpha and cardiovascular disease in hemodialysis patients. Nephron Clin Pract 2009;113: c104-11. Saracini C, Bolli P, Sticchi E, Pratesi G, Pulli R, Sofi F, et al. Polymorphisms of genes involved in extracellular matrix remodeling and abdominal aortic aneurysm. J Vasc Surg 2012;55:171-9. Ogata T, Shibamura H, Tromp G, Sinha M, Goddard KAB, Sakalihasan N, et al. Genetic analysis of polymorphisms in biologically relevant candidate genes in patients with abdominal aortic aneurysms. J Vasc Surg 2005;41:1036-42. Bahadori B, Uitz E, Mayer A, Harauer J, Dam K, TruschnigWilders M, et al. Polymorphisms of the hypoxia-inducible factor 1 gene and peripheral artery disease. Vasc Med 2010;15:371-4. Vardulaki KA, Walker NM, Day NE, Duffy SW, Ashton HA, Scott RA. Quantifying the risks of hypertension, age, sex and smoking in patients with abdominal aortic aneurysm. Br J Surg 2000;87:195-200. Lu JT, Creager MA. The relationship of cigarette smoking to peripheral arterial disease. Rev Cardiovasc Med 2004;5:189-93. Moodie FM, Marwick JA, Anderson CS, Szulakowski P, Biswas SK, Bauter MR, et al. Oxidative stress and cigarette smoke alter chromatin remodeling but differentially regulate NF-kB activation and proinflammatory cytokine release in alveolar epithelial cells. FASEB J 2004;18:1897-9.

JOURNAL OF VASCULAR SURGERY Volume -, Number -

36. Hsiao PC, Chen MK, Su SC, Ueng KC, Chen YC, Hsieh YH, et al. Hypoxia inducible factor-1a gene polymorphism G1790A and its interaction with tobacco and alcohol consumptions increase susceptibility to hepatocellular carcinoma. J Surg Oncol 2010;102: 163-9. 37. Michaud SÉ, Ménard C, Guy LG, Gennaro G, Rivard A. Inhibition of hypoxia-induced angiogenesis by cigarette smoke exposure: impairment of the HIF-1a/VEGF pathway. FASEB J 2003;17:1150-2. 38. Koerselman J, de Jaegere PPT, Verhaar MC, Grobbee DE, van der Graaf Y. Coronary collateral circulation: the effects of smoking and alcohol. Atherosclerosis 2007;191:191-8. 39. Thompson SG, Ashton HA, Gao L, Buxton MJ, Scott RA. Final follow-up of the Multicentre Aneurysm Screening Study (MASS)

Strauss et al 9

randomized trial of abdominal aortic aneurysm screening. Br J Surg 2012;99:1649-56. 40. Ascher E, Marks NA, Hingorani AP, Schutzer RW, Nahata S. Duplexguided balloon angioplasty and subintimal dissection of infrapopliteal arteries: early results with a new approach to avoid radiation exposure and contrast material. J Vasc Surg 2005;42:1114-21.

Submitted Nov 28, 2013; accepted Feb 6, 2014.

Additional material for this article may be found online at www.jvascsurg.org.

9.e1 Strauss et al

JOURNAL OF VASCULAR SURGERY --- 2014

Supplementary Fig (online only). Forest plots of the individual and joint effect of the hypoxia-inducible factor-1a (HIF1A) 1772C>T (rs11549465), 1790G>A (rs11549467), and vascular endothelial growth factor (VEGFA) þ405G>C (rs2010963) polymorphisms on abdominal aortic aneurysm (AAA), with and without coexisting peripheral arterial disease (PAD), and aortoiliac occlusive disease (AIOD) risk. For each individual polymorphism, the estimation of the dominant (HIF1A) or additive (VEGF) effects of the minor allele is shown. On the left, investigated effects are listed, whereas the right-hand part shows the odds ratios (ORs) for the studied diseases (black squares) with the corresponding 95% confidence intervals (CIs, horizontal lines). The vertical line represents no effect, and the overlap of the CIs with these lines indicates no statistically significant effect.

JOURNAL OF VASCULAR SURGERY Volume -, Number -

Strauss et al 9.e2

Supplementary Table I (online only). The univariate statistical analysis of the associations between single hypoxiainducible factor-1a (HIF1A) and vascular endothelial growth factor (VEGFA) polymorphisms, as well as HIF1A haplotypes, and abdominal aortic aneurysm (AAA), with or without concomitant peripheral arterial disease (PAD), or aortoiliac occlusive disease (AIOD) occurrence Compared groups OR and P value

Genotype or effect

AAA vs controls

AAA without PAD vs controls

AAA with PAD vs controls

HIF1A 1772C>T (rs11549465) Heterozygous 0.99 0.87 .950 .572 Homozygous 1.03 0.72 .968 .774 Dominant effect 0.99 0.87 .958 .544 Recessive effect 1.04 0.73 .999 .999 Additive effect 1.00 0.87 .967 .531 Difference in MAF 0.99 0.87 .997 .529 HIF1A 1790 G>A (rs11549467) Heterozygous 0.76 0.85 .340 .648 Homozygous 1.03 2.20 .983 .567 Dominant effect 0.77 0.90 .351 .743 Recessive effect 1.04 2.22 .999 .525 Additive effect 0.81 1.00 .379 .850 Difference in MAF 0.79 0.94 .369 .846 VEGFA
634G>C (rs2010963) Heterozygous 1.08 1.26 .546 .157 Homozygous 1.52 2.00 .079 .012 Dominant effect 1.14 1.36 .284 .047 Recessive effect 0.88 1.07 .636 .787 Additive effect 1.19 1.38 .121 .012 Difference in MAF 1.16 1.34 .123 .013 HIF1A 1772-1790 (rs11549467-rs11549467) CC-GG 1.08 1.12 .650 .578 CC-GA 0.81 0.94 .456 .853 CT-GG 1.04 0.92 .834 .732 (CTþTT)-GG 1.04 0.91 .855 .728 Rare <1% 0.59 0.62 .728 .728 Haplotype frequencies C-G 1.10 1.13 .531 .506 C-A 0.75 0.93 .281 .818 T-G 0.99 0.87 .965 .587 T-A Not estimable Not estimable Not MAF, Minor allele frequency; OR, odds ratio. The crude ORs and P values are shown.

AAA with AIOD vs AAA vs AAA without AAA with PAD vs AAA controls AIOD PAD vs AIOD PAD vs AIOD without PAD

2.01 .011 2.21 .485 2.02 .009 1.95 .471 1.84 .011 1.90 .010

1.09 .682 1.02 .984 1.08 .688 1.01 .999 1.07 .704 1.07 .703

0.91 .646 1.02 .987 0.92 .656 1.03 .999 0.93 .679 0.93 .678

0.81 .392 0.71 .775 0.80 .373 0.72 .999 0.81 .371 0.81 .371

1.86 .033 2.17 .521 1.87 .028 1.93 .501 1.74 .029 1.77 .030

2.30 .008 3.07 .406 2.33 .006 2.67 .473 2.14 .006 2.18 .006

0.36 .150 1.86 .672 0.35 .137 1.93 .999 0.36 .134 0.35 .129

0.93 .812 0.51 .417 0.90 .730 0.51 .999 0.90 .653 0.88 .652

0.82 .534 2.04 .410 0.86 .623 2.06 .999 1.01 .726 0.90 .724

0.92 .816 4.35 .229 0.99 .983 4.37 .408 1.32 .850 1.07 .848

0.39 .192 3.67 .999 0.39 .192 Not estimable 0.39 .192 0.39 .280

0.42 .253 0.84 .531 0.39 .206 0.87 .999 0.41 .193 0.37 .259

0.93 .741 0.96 .930 0.93 .747 0.59 .320 0.96 .789 0.95 .792

1.04 .797 1.01 .969 1.03 .809 0.60 .074 1.02 .851 1.02 .853

1.04 .774 1.50 .133 1.10 .473 1.47 .165 1.17 .231 1.14 .234

1.22 .268 1.98 .025 1.32 .099 1.80 .051 1.36 .031 1.32 .032

0.89 .639 0.95 .915 0.90 .650 1.00 .999 0.95 .713 0.93 .718

0.73 .229 0.480 .122 0.68 .119 0.55 .226 0.71 .077 0.71 .078

0.70 .173 0.20 .104 2.10 .007 2.12 .010 0.83 .999

0.96 .820 0.97 .922 1.14 .513 1.11 .616 0.43 .330

1.12 .526 0.83 .566 0.91 .645 0.94 .764 1.37 .999

1.16 .481 0.97 .927 0.81 .396 0.82 .463 1.45 .999

0.73 .245 0.21 .143 1.84 .035 1.91 .027 1.93 .501

0.62 .107 0.21 .124 2.28 .009 2.33 .010 1.33 .999

0.75 .228 0.17 .049 1.90 .016 estimable

1.01 .964 0.82 .563 1.05 .779 4.54 .398

1.09 .605 0.91 .754 0.94 .777 0.23 .406

1.12 .560 1.13 .720 0.83 .485 0.48 .999

0.74 .238 0.21 .147 1.80 .036 1.28 .999

0.66 .151 0.19 .081 2.18 .010 Not estimable

JOURNAL OF VASCULAR SURGERY --- 2014

9.e3 Strauss et al

Supplementary Table II (online only). Distribution of genotypes and alleles of the hypoxia-inducible factor-1a (HIF1A) and vascular endothelial growth factor (VEGFA) polymorphisms in smokers: patients with abdominal aortic aneurysm (AAA) stratified according to peripheral arterial disease (PAD) coexistence and controls AAA Genotype distribution HIF1A 1772 (rs11549465) CC CT TT MAF HIF1A 1790 (rs11549467) GG GA AA MAF VEGFA
634 (rs2010963) GG GC CC MAF

Controls (n ¼ 178)

Without PAD (n ¼ 184)

With PAD (n ¼ 82)

160 (89.9) 18 (10.1) 0 (0.0) 0.051a

159 (86.4) 25 (13.6) 0 (0.0) 0.068b

60 (74.1) 20 (24.7) 1 (1.2) 0.136a,b

173 (95.1) 9 (4.9) 0 (0.0) 0.025

176 (95.7) 8 (4.3) 0 (0.0) 0.022

80 (97.6) 2 (2.4) 0 (0.0) 0.012

96 (53.3) 70 (38.9) 14 (7.8) 0.272

MAF, Minor allele frequency. Association between the genotypes and the studied diseases. a Dominant effect of the 1772T allele: OR, 3.1 (1.6-6.2); P ¼ .001. b Difference in the frequency of 1772T allele carriers; P ¼ .015.

83 (45.9) 76 (42.0) 22 (12.2) 0.331

47 (58.0) 29 (35.8) 5 (6.2) 0.241