Chronic Kidney Disease: The “Perfect Storm” of Cardiometabolic Risk Illuminates Genetic Diathesis in Cardiovascular Disease

A fact well known to the readers of the Journal: heart disease is the number one killer in our society, contributing to six hundred thousand deaths annually(1). Coronary heart disease (CHD) is responsible for approximately 2/3rds of these deaths, and 935,000 Americans suffer myocardial infarction (MI) each year(1). Hypertension, elevated LDL cholesterol, smoking and diabetes are well appreciated as risk factors. However, a particularly ominous risk for CHD arises in the setting of chronic kidney disease (CKD)(2). As renal function declines below a glomerular filtration rate of 60 cc/min/1.73 m2, the relative risk for cardiovascular mortality progressively increases by 3 -fold as compared to those without CKD(3) – and almost 10-fold in the setting of CKD with diabetes(4). Indeed, the cardiovascular risk conveyed by CKD exceeds that conveyed by diabetes(5), (6)

The relationship between mineral metabolism and cardiovascular risk has been increasingly appreciated over the past two decades, now codified in the KDIGO designation of CKD-MBD, for “Chronic Kidney Disease – Mineral and Bone Disorder(7).” The CKD-MBD designation only partially captures the clinical implications; the underlying recognition that vascular calcium deposition and skeletal bone mass are reciprocally and functionally coupled in CKD is somewhat reflected in the “MBD” – but the appellation does not fully capture the cardiovascular risk associated with perturbation in this relationship. London and colleagues have nicely demonstrated that the present and extent of arterial intimal and medial calcification is a strong predictor of cardiovascular mortality(8). The fortunate minority of those with CKD on RRT (renal replacement therapy) who do not exhibit overt arterial calcification enjoy much greater longevity(8). His group's elegant histomorphometric studies identified that dialysis patients with the most severe low-turnover bone disease had the most extensive arterial calcium load(9). While hypercholesterolemia, hypertension, and hypercoagulability in CKD all certainly contribute, CAD morbidity and mortality stubbornly persists in this high – risk population even when aggressive therapeutic regimens targeting these risk factors are implemented(10,11). Of note, the cardiovascular consequences of perturbations in mineral metabolism – viz., phosphate and calcium metabolism – appear to be emerging in patients with normal renal function as well. For example, at every level of renal function, patients with T2DM experience increased vascular calcium loads(12) – and CKD and T2DM synergistically enhance the risk for myocardial infarction. Moreover, serum calcium and phosphate levels portend all-cause cardiovascular mortality at every level of renal function(13). Thus, a role for calciotropic hormone signaling and mineral metabolic in cardiovascular disease is increasingly apparent. The prevailing metabolic and mechanical insults that most frequently cause CKD(14) also promote cardiovascular calcification(15) – disease that is worsened by declining renal function that globally perturb mineral metabolism including the bone-vascular axis(16). As such, CKD represents a metabolic “perfect storm” for cardiovascular disease.

In this issue of the Journal, Reilly, Rader and colleagues successfully explore the enhanced CAD risk the Chronic Renal Insufficiency Cohort (CRIC) to reveal novel insights into the molecular genetics of cardiovascular disease (17). They implement the working hypothesis that the high-risk milieu of CKD would amplify the negative impact of genetic diathesis – functioning as a metabolic magnifying glass to help identify genes conveying CHD risk. As such, this strategy is somewhat but imperfectly analogous to “sensitized screening” used in genetic studies of lower eukaryotes(18). Coronary artery calcification (CAC) score obtained by MSCT (multislice computed tomography) was used quantify the aberrant vascular mineral metabolism that demarcates and contributes to arterial disease in the CRIC cohort. The IBC cardiometabolic SNP array (19)was then used to interrogate genetic variability at ca. 2100 loci known to influence biologically plausible and proven risk pathways in CHD. Because of the small CRIC cohort size (N = 1509), two other sets of CAC-phenotyped patient populations – the PennCAC (n = 2560) and the Amish Family Calcification Study (AFCS); n = 784) – were used as independent “filters” to help further validate initial findings from CRIC. Importantly, these latter groups were not enriched for patients with CKD. This strategy helped identify 23 gene loci that were correlated with CAC scores in (a) the CRIC and PennCAC cohort; and/or (b) the CRIC and AFCS cohort. Importantly, chr9p21 – the premier locus identified as conveying both CAC and CHD risk in GWAS applied to the general population(20) – was identified by this methodology. Finally, to provide robust evidence for the “hard” primary clinical endpoints of MI, these loci were used to interrogate the Pakistani Risk of Myocardial Infarction (PROMIS) cohort. Following this analysis, 4 loci -- chr9p21, COL4A1, ATP2B1 and ABCA4 – were identified as contributing to MI risk. Both chr9p21 and COL4A1 had been previously identified as contributing to CAC scores in GWAS(20). At this stage, even when correcting for multiple comparisons (not done at earlier stages of discovery), ATP2B1 and COL4A1 genetics significantly portended increase risk for MI in the PROMIS cohort lacking enrichment for CKD(17).

Why is this study so significant? It is enlightening on several levels. Firstly, the impact of CKD specifically upon coronary artery calcium load prior to RRT has needed assessment in a large well-phenotyped cohort – and CRIC has provided this information(12,17). Moreover, it confirms and extends the association of chr9p21 and COL4A1 genotype with coronary calcium load(17,20). Type IV collagen is vital to basement membrane function, vascular integrity, and alterations contribute to tissue calcification risk in other venues(21,22). Indeed, type IV collagen protein accumulation is more intense around sites of calcification in patients with CKD(23). This suggests that additional research into the biological role of type IV collagen in vascular calcification is warranted. Additionally, when viewed in the context of other recent observations, this study suggests that low-cost focused genotyping may yet provide important, cost-effective risk stratification in patients with equivocal CAD risk. Coronary artery calcification arises primarily from atherosclerotic calcification vs. medial artery calcification (24), and for individuals at intermediate CHD risk by Framingham criteria, CAC scoring helps better assess the probability for MI (25,26). Thus, given these new results, genotyping for CAC risk could be envisioned in an algorithm that better guides or augments the implementation of CT imaging to stratify individual clinical CAD risk.(27). Moreover, the loci identified once again implicate mineral metabolism in the pathogenesis of cardiovascular disease. The most intriguing is ATP2B1, a.k.a. PMCA1. Via alternative splicing, this gene encodes several different protein isoforms of PMCA1, a plasma membrane ATPase that pumps cytosolic calcium across the membrane and out of cells; via this action in vascular smooth muscle cells (VSMCs) PMCA1 regulates vascular tone and blood pressure(28). However, PMCA1 also plays global “housekeeping” functions that are critical to cell viability, and it would be interesting to know whether VSMC propensity for apoptotic cell death(29)is altered by PMCA1 insufficiency(30) as relevant to acute coronary events. Compensatory upregulation of VSMC PMCA4 with PMCA1 deficiency may mitigate some of this risk. Moreover, PMCA1 also plays a critical role in regulating bone-resorbing osteoclast activation by RANKL(31). This TNF superfamily member is not only necessary for osteoclast differentiation, function, and survival (15,31) – but is also implicated in the pathobiology of vascular calcification(32). The ATP2B1 rs11105354A allele tracked both lower CAC score and blood pressure, and it may be that global calcium homeostasis is also altered in this genetic setting(17). Of note, the ATP2B1 rs11105354A allele was significantly associated with increased serum calcium levels(17). Finally, as the authors point out(17), this type of “sensitized screening” (18)may be useful in the identification of new signaling pathways that contribute to the pathogenesis of CAD through a focus on atherosclerosis phenotypes in patients with CKD. It's interesting to note that the TLL1 genetic variant identified by Cresci et al as significantly associated with the extend of CAD appears specific to the setting of type II diabetes (33); similar examples of selective interactions between genotype, metabolic milieu and CAD risk may also arise in the setting of CKD. Nevertheless, given the clinical management challenges we face with our increasingly aged and dysmetabolic patient populations -- often with concomitant reductions in renal function – the insights afforded by genomics combined with multidisciplinary phenotyping of patients with CKD will no doubt provide novel strategies for assigning and mitigating cardiovascular disease risk.