Genotype interactions and intima-media thickness

Association studies between candidate gene polymorphisms and arterial phenotype are numerous, and often provide conflicting results. Isolated studies have frequently proved inconclusive and meta-analysis have been necessary to clarify the net effect of a given polymorphism on arterial phenotype. Even in these cases, the contribution of a given gene polymorphism to the variance of a specific phenotype was limited. Although often mentioned, studies combining two or more gene polymorphisms were seldom performed. This is a major interest of the study by Balkestein et al. [1] in this issue of the journal. The authors investigated, in 380 subjects, the relationship between intima –media thickening of the carotid and femoral arteries, and polymorphisms in the genes encoding the angiotensin-converting enzyme (ACE I/D), aldosterone synthase (−C344T) and the cytoskeleton protein α-adducin (Gly460Trp). The main finding was that the effect of the angiotensin-converting enzyme (ACE) genotype on intima–media thickness (IMT) was observed only at the site of the femoral artery and was confined to carriers of the α-adducin 460Trp allele or the aldosterone synthase −344T allele. Indeed, in the presence of the 460Trp allele, the mean femoral IMT, measured in II homozygotes of the ACE gene, was 59 μm lower than the population mean (575 μm), whereas in DD homozygotes it was on average 86 μm higher than the population mean. In addition, in carriers of the −344T allele, the corresponding deviations from the population mean amounted to −48 and 73 μm. The authors suggest that the femoral wall thickening may be due, in the first genotype combination, to the combined effects of higher systemic ACE levels, which probably stimulate the local generation of angiotensin II, and higher activity of the sodium pump in carriers of the α-adducin 460Trp allele, which enhances proximal tubular reabsorption of sodium, and may modulate the availability of Ca2+ ions, thus regulating smooth muscle cell growth. In the second genotype combination, femoral wall thickening may be due to the combination of increased local generation of angiotensin II and stimulation of aldosterone synthesis. This article is the fourth one of a larger study [1–4], the FLEMENGHO study, which recruited nuclear families in the framework of the European Project on genes in Hypertension, and which previously demonstrated that blood pressure, renal function and arterial stiffness were influenced by these candidate genes. A major interest of this fourth study [1] is that it quantitatively demonstrates whether interactions between genes possibly linked to cardiovascular disease are responsible for arterial wall thickening. The present article provides the first evidence of a synergism between these candidate genes and the femoral IMT. The article by Balkestein et al. [1] has several strengths. The phenotype has been carefully determined. The high-definition echotracking system used in this study has been previously validated in large subsets of patients with various diseases and its accuracy and reproducibility is well accepted. All vascular measurements have been performed by the same investigator to reduce variability. A large number of subjects (n = 380) were included in the whole study, in order to allow investigation of a sufficient number of subjects having a specific combination of genotypes. This number is necessary to take into account the influence of several factors known to influence wall thickness, such as age, gender, mean blood pressure, smoking, low-density lipoprotein cholesterol and intake of anti-hypertensive drugs. Interestingly, after adjustment for these confounding factors, the amplitude of the changes in femoral IMT between two combinations of genotypes, remained large enough to be clinically relevant. For example, the femoral IMT was 145 μm higher in subjects with the ACE/DD polymorphism than in ACE/II subjects, in the presence of the α-adducin 460Trp allele. Using a strategy focused both on gene interactions in a large population and investigation at two arterial sites, the authors observed significant results, whereas previous studies analysing only one polymorphism [5], reported no significant association. This study, although conclusive, exhibits several limits common to the candidate-gene approach. First, despite the large size of the study population, a high number of statistical comparisons is necessary to take into account all possible interactions, increasing the probability of a significant association (Bonferroni paradox). Second, no significant interaction was shown between genotype combinations and IMT at the site of the carotid artery. This arterial site is of major importance, since an increase in carotid IMT is generally considered as a marker of early atherosclerosis and an independent cardiovascular risk factor in older adults. The association of carotid IMT with cardiovascular risk factor, and its predictive value on cardiovascular events, were more often published than that of femoral IMT. Balkestein et al. [1] quote the Angina Prognosis Study [6], which showed that the risk of revascularization was correlated with femoral IMT, whereas carotid IMT failed to predict any cardiovascular event in this population. These two studies [1,6] indicate that measurement of femoral IMT could provide information that is complementary to the carotid IMT, and should be more often included in association and predictive studies dealing with arterial phenotype and cardiovascular events. Third, despite the precision of the arterial measurements allowed by the echotracking system, ultrasound imaging cannot discriminate between the intima layer and the media layer of the vessel wall to distinguish true arteriosclerosis (i.e. the adaptive response of the media layer to changes in tensile stress) from atherosclerosis (viewed as a disorder restricted to the intima layer). Whether arteriosclerosis or atherosclerosis processes are concerned at the molecular level by the genotype interactions observed in the present work remains an open question. Fourth, the pathogenic hypothesis underlying such a selective tropism of action of these genes on muscular versus elastic arteries requires further argumentation. The femoral artery is a muscular artery, whereas the carotid artery exhibits a higher proportion of extracellular matrix and elastic lamellae, and a relatively lower percentage of smooth muscle cells. The differences in the percentage and distribution of smooth muscle cells between the carotid and femoral artery walls may explain the different patterns of wall hypertrophy between these two sites in subjects having various gene polymorphism combinations. The growth-promoting effect of gene interactions may be lower in carotid smooth muscle cells than in femoral ones. Fifth, there are interesting discrepancies between the previously published results of this group on femoral artery distensibility [4] and the current results on arterial structure [1]. The femoral artery plays only a minor role in systemic arterial distensibility whereas aortic distensibility has been shown to be an independent predictors of cardiovascular events [7]. In the previous study by Balkestein et al. [4], ACE/DD subjects homozygous for α-adducin Gly460 had a lower femoral cross-sectional distensibility compared to the population mean [4], whereas a significant difference in femoral IMT was observed only in carriers of the α-adducin 460Trp allele, and not the Gly460 allele. Sixth, it would be interesting to determine, in further pharmacogenetic studies, whether the combination of genetic polymorphisms, observed in some patients, may predispose them to a higher efficacy of specific drugs. In conclusion, the findings of the present study by Balkestein et al. [1], together with previous studies by the same group [2–4], may have important clinical implications for the assessment of cardiovascular risk and the therapeutic approach to hypertension, including pharmacogenetic studies. Confirmation of these results would also strengthen mechanistic hypotheses concerning gene interactions at the molecular level and promote in-vitro studies which may lead to the discovery of new pharmacological targets. These ‘candidate gene’ studies [1–4] would benefit from confrontation with studies using an unbiased genomic approach, which may not only confirm the validity of the present analysis, but also provide new information on unexpected genes differently expressed between subjects.