Propensity to calcification as a pathway to renal hypoxia in chronic kidney disease and in hypertension

Arterial rigidity is a hallmark of chronic kidney disease (CKD) ^[1]. Arterial calcification recapitulates normal bone formation ^[2], and this process is considered central-to-arterial stiffness in CKD ^[3] and predicts major clinical outcomes including death, cardiovascular events and progression to kidney failure ^[4] in these patients. Renal hypoxia measured in vivo by blood oxygenation level-dependent MRI (BOLD-MRI) goes along with the severity of renal dysfunction and the risk of progression of CKD ^[5,6]. In theory arterial stiffness by vascular calcification may explain renal hypoxia in CKD and in essential hypertension, another condition typically associated with arterial stiffness.

The hypothesis has now been explored by Pruijm et al.^[7] in an article published in this issue of the Journal. Along with renal oxygenation measured by BOLD-MRI ^[8], these authors estimated the individual propensity to calcification by an in-vitro test quantifying the ability of serum samples to delay the formation of crystalline hydroxyapatite-containing calciproteins from amorphous calciproteins containing Ca and P (the longer the time for this transformation, the lesser the calcification propensity) ^[9] as well as echo-Doppler renovascular resistances and arterial stiffness (pulse wave velocity). Confirming previous reports, propensity to calcification was by 24% more pronounced in CKD patients than in healthy controls. Importantly, also essential hypertensives had a distinctly raised calcification propensity (+15%) as compared with healthy controls and in a global analysis patients with a higher propensity to calcification exhibited lower levels of renal oxygenation, higher renal resistances and a more pronounced degree of arterial stiffness and were more likely to be hypertensive, which was true both in CKD and in essential hypertensive patients. Furthermore, arterial stiffness per se was inversely related to renal oxygenation. Overall, these cross-sectional analyses are compatible with the hypothesis that calcification propensity may be the remote event in the pathway whereby arterial stiffness eventually triggers renal hypoxia and renal damage (Fig. 1). However attracting, these data are by design hypothesis generating rather than hypothesis testing. In brief, uncertainties remain on modifiable factors that can be leveraged to attenuate calcification propensity, on the validity of the ‘in-vitro’ test of calcification propensity as a prognostic factor, and on the causal nature of the calcification propensity–renal oxygenation link.

FIGURE 1: Mechanisms whereby calcification propensity may lead to renal hypoxia. Factors responsible for calcification propensity reported in the figure are discussed into detail in a review by Pasch ^[30]. ^*Systemic inflammation triggers calcification propensity but it is also a well known risk factor for chronic kidney disease progression ^[16], and it may therefore be a confounder for the calcification propensity–renal hypoxia link. ^**Hydroxyapatite-calciproteins in smooth muscle cells in the vessel wall may per se trigger local inflammation ^[31] and perhaps systemic inflammation thus representing an amplification loop in this pathway mechanism.

Calcification in the arterial system depends on an unbalance between inhibitors and activators of the calcification process in conditions of saturated calcium and phosphate extracellular fluid concentrations. This unbalance activates mechanisms similar to the endochondral and membranous ossification process ^[2]. In addition to aging and hyperphosphatemia – a risk factor peculiar to the late stages of CKD ^[10] – diabetes and hypercholesterolemia are well established procalcifying stimuli ^[2]. Molecular imaging techniques documented that inflammatory mechanisms are strong drivers of calcification since the early phases of atherosclerosis ^[11]. CKD patients are notoriously predisposed to vascular calcification and hyperphosphatemia is a well known procalcifying stimulus in this condition ^[3]. In longitudinal studies in patients with atherosclerosis combining PET by 18-flourodeoxyglucose uptake – a biomarker of inflamed areas – and computed tomography, inflammation clearly antedates arterial calcification ^[12].

Factors determining the propensity to calcification in the test adopted by Pruijm et al., largely overlap with those implicated in vascular calcification in CKD patients (Fig. 1). Fetuin-A is a protein extremely sensitive to inflammation, which reduces substantially during this process (an inverse acute phase reactant), and the first test of calcification propensity, the calciproteins fetuin test, devised by Hamano et al.^[13] in reality was an inflammation-centered test. The test devised by Pasch et al.^[9] applied in the current study is also dependent on the inflammatory background because low levels of the inverse acute phase reactants and calcification inhibitors albumin and fetuin-A are among the main factors favoring amorphous phosphate crystallization and because, in crude terms, both serum C-reactive protein and TNFα are both associated with calcification propensity as measured by the same test ^[9]. Thus, these well conceived tests of calcification propensity in most circumstances (e.g. in essential hypertension ^[14] and primarily inflammatory chronic diseases ^[15]) reflect the procalcifying effect of inflammation. CKD is an inherently inflammatory disease ^[16] incited by a host of risk factors including phosphate excess ^[17] which is also an important determinant of the in-vitro test by Pasch et al.^[9]. It was stressed that measuring the propensity to calcification is important because the results of this test predict mortality in stage 3–4 CKD patients ^[18] and in transplant patients ^[19]. However, until now the prognostic validity of this test was examined in two small studies including just 43 ^[18] and 81 ^[19] deaths and in the database of the EVOLVE study ^[20]. Furthermore, the predictive power of calcification propensity has not been compared with predictive models specific to the CKD population like the Chronic Renal Impairment in Birmingham (CRIB) cohort model ^[21]. In this model, just four factors (age, N-terminal probrain natriuretic peptide, troponin T and smoking) showed a high discrimination power for death not only in the cohort in which the model was developed [area under the receiver operating curve (ROC) curve 0.82] but also in an external, confirmation cohort (area under the ROC curve again = 0.82) which is a discrimination power substantially higher than that of calcification propensity (0.74 in the study by Smith ^[18] and just 0.65 in the EVOLVE study ^[20]). However ingenious and useful for research purposes aimed at elucidating the risk for vascular calcification, the test does not give information on the precise risk factors underlying calcification propensity that can be leveraged to reduce the risk of death and cardiovascular events by vascular calcification. In clinical practice, rather than a biochemical test measuring the predisposition to calcification, it is preferable having detailed information on individual determinants of the risk for vascular calcification in hypertension and in CKD – including serum calcium, phosphate, PTH, vitamin D and inflammation biomarkers like C reactive protein and parathormone, IL-6 and TNFα – because the level of these biomarkers may be useful to profile interventions targeting vascular risk.

Arterial stiffening is the key element of the link between calcification propensity and renal oxygenation (Fig. 1). Stiffness of major (large) arteries is determined mainly by the extracellular matrix composition, whereas smooth muscle cells have a limited role in the mechanical properties of the same arteries ^[22]. In large elastic arteries, elastin (a low turnover protein, representing the 90% of arterial elastic fibers) provides extensibility during the pulsatile flow, whereas collagen provides strength and protection of vascular integrity. Elastic fibers degradation is increased by age and disease states (like CKD and hypertension), and degradation of this protein is a very early phenomenon in the uremic mouse in which it appears a factor necessary for increased stiffness ^[23]. Elastin has a high affinity for calcium and calcium accumulation in elastic lamellae markedly increases arterial stiffness ^[24]. However, calcium excess is not a prerequisite for arterial stiffening because in a mouse model of type 2 diabetes, an osteogenic-transcription factor like Runx2 promotes aortic fibrosis and stiffness independently of calcification ^[25]. High matrix metalloproteinase activity triggered by inflammatory cytokines is a primary driver of elastin degradation in arteriosclerosis and the osteogenic transformation of macrophages favors the formation of microcalcifications in the vessel wall that in turn amplify inflammation ^[11]. Elegant longitudinal imaging studies in man ^[12] support the view that vascular calcification follows inflammation. Thus, propensity to calcification may be seen as a proxy of inflammation (the triggering event). On the other hand, the fact that systemic inflammation has per se an independent role in renal disease progression in CKD patients ^[26] suggests that this alteration may be a confounder for the interpretation of the link among calcification propensity, arterial stiffness and renal hypoxia (Fig. 1).

Arterial stiffening is a process related to systemic inflammation in essential hypertension ^[27] and in CKD ^[28]. Also independently of calcification, vascular wall rigidity has per se the potential to engender renal damage. The renal circulation is a low resistance bed that makes this organ vulnerable to pressure load and to pulsatile changes in blood flow resulting in endothelial dysfunction, ischemia and tissue injury ^[29]. The capillary rarefaction that commonly occurs in these conditions renders the kidney particularly susceptible to low ambient oxygen tension. In such a context, the superimposed hemodynamic stress by arterial stiffness and the resulting microvascular damage may critically aggravate renal hypoxia. However, the thought provoking observations by Pruijm et al. fall short of proving that the link among calcification propensity, arterial stiffness and renal hypoxia is causal in nature. The causal graph presented in Fig. 1 makes explicit the critical mediators and confounders at play in the ‘calcification propensity’ hypothesis of renal hypoxia. Full testing of the hypothesis clearly requires specific experiments in animal models and appropriate longitudinal and intervention studies in patients with CKD and in essential hypertensive patients. This intriguing study is just the start of the story rather than a scientific proof of the existence of the pathway that Pruijm et al. incriminate as responsible for renal hypoxia.

ACKNOWLEDGEMENTS

Conflicts of interest

There are no conflicts of interest.