Cardiac Valve Calcifications and Left Ventricular Hypertrophy in Hemodialysis Patients

Abstract

Cardiac valve calcification (VC) is a common finding in end-stage renal disease patients. It was shown recently that VC is an independent predictor for all-cause and cardiovascular mortality in peritoneal dialysis patients. In hemodialysis (HD) patients, VC was associated with all-cause and cardiovascular mortality, but after adjusting for other cardiovascular risk factors and complications, as well as left ventricular mass index (LVMI), it lost significance. The aim of the study was to assess the relationship between VC and left ventricular hypertrophy in hemodialysis patients. Echocardiographic examination with mitral and aortic valves assessment and LVMI calculation was performed in 65 HD patients ages 49 ± 12, with duration of HD therapy 38 ± 32 months. VC were found in 32 of 65 patients (49%)—Group VC(+), mitral valve calcifications (MVC) in 10, aortic valve calcifications (AVC) in 9, and both valves calcifications (MVC + AVC) in 13 patients. Patients with VC were older, on HD therapy were longer, had higher systolic and pulse pressure, and had higher LVMI. Patients with both VCs had the highest LVMI. No significant differences were found with respect to Ca, P, PTH, and mean Ca × P product, but the incidence of Ca × P product above 4.43 mmol²/L² was higher in VC(+) compared with those without VCs. VC coexists with left ventricular hypertrophy, particularly when both valves are calcified. Even short-lasting incidents of increased Ca × P product may lead to cardiac VC.

• valve calcification
• echocardiography
• left ventricular hypertrophy

Introduction

Cardiovascular (CV) disease accounts for about 50% of deaths among dialysis patients.[1] Soft-tissue calcifications, including the heart and arteries, are common findings in end-stage renal disease (ESRD) patients, even in young adults.[2&3] Using echocardiography, cardiac valve calcifications (VCs) were found in about half of hemodialysis (HD) patients.[3&4] VC is potentially dangerous due to association with valve dysfunction, myocardial ischemia, conduction defects, infective endocarditis, and heart failure.[5] It was also shown that VC is an independent predictor for all-cause and cardiovascular mortality in peritoneal dialysis patients, with extremely high all-cause mortality in patients with both: mitral and aortic valves calcified.[6] In HD patients, VC was associated with all-cause and cardiovascular mortality in unadjusted analysis.[7] But after adjusting for other CV risk factors, CV complications, and left ventricular mass index (LVMI), it lost significance. LVMI—a parameter of left ventricular hypertrophy (LVH)—remains one of the most important cardiovascular prognostic factors in ESRD patients.

Hyperphosphatemia and high calcium × phosphorus (Ca × P) product are independent risk factors of all-cause and cardiovascular mortality in dialysis patients.[8&9] Despite many studies, the pathogenesis of VC is not fully understood. In several studies, there was no relationship between calcium and phosphorus concentration, Ca × P product, and severity of VC. According to more recent studies, the process of VCs is an active one, similar to bone formation with the important roles of vascular smooth muscle cells and calcification inhibitors, such as fetuin-A and matrix Gla protein.[10]

The aim of our study was to assess the relationship between cardiac VCs and LVH and disturbances of calcium-phosphate metabolism in HD patients.

Patients and Methods

We studied 65 HD patients: F = 35, M = 30, aged 49 ± 12 (range, 23–74) years, with mean duration of HD therapy 38 ± 32 (range, 3–136) months. The underlying renal diseases were glomerulonephritis in 38 patients, hypertensive nephropathy in 8, polycystic kidney disease in 8, interstitial nephritis in 6, and unknown in 5 patients. Patients were dialyzed three times a week, with duration of HD session 4 to 5 h. Bicarbonate dialysate fluid was used and the fluid calcium concentration was 1.25 mmol/L. Patients with the history of parathyroidectomy and patients with diabetes were excluded from the study. The protocol of the study was accepted by local Ethics Committee. Informed consent was obtained from all patients.

Laboratory data from 6 consecutive months preceding echocardiography were averaged.[2] Blood was sampled twice monthly before a middle-week HD session for total calcium (Ca) and inorganic phosphate (P). Intact parathormone (PTH), albumin, and blood urea (expressed as blood urea nitrogen) were measured every 3 months. Kt/V (delivered dose of dialysis) was calculated according to The National Kidney Foundation-Dialysis Outcomes Quality Initiative (NKF-DOQI) recommendation.[11] Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured before HD session with mercury sphygmomanometer in a sitting position after 15 min rest. Blood pressure records from 2 months preceding echocardiography (26 HD charts) were averaged. The incidence of Ca × P product over 4.43 mmol²/L² (equivalent of 55 mg²/dL²) was analyzed. That value (55 mg²/dL²) was recently recommended as the upper limit of the Ca × P product to prevent cardiac and VC in dialyzed patients.[12]

Echocardiography with American Society of Echocardiography standard M-mode measurement[13] was performed after an HD session using Sonos 2000 (Hewlett-Packard) equipment with a 2.5/2.0-MHz transducer. Diagnosis of VC was based on echocardiographic criteria of dense echoes in mitral vascular calcification (MVC) or aortal vascular calcification (AVC) valve leaflets or annulus.[4-7] Interventricular septum thickness (IVS), left ventricular internal diameter (LVID), and posterior wall thickness (PW) were measured in end diastole. Left ventricular mass (LVM) was calculated using the Devereux formula: LVM = 0.8[1.04 (IVS + LVID + PW)³− LVID³] + 0.6 g.[14] End-diastolic volume (EDV) and ejection fraction (EF) were calculated from a 2D apical four-chamber view with the area-length method. Echocardiographic data from three consecutive cycles were averaged. LVM and EDV were indexed to body surface area: LVMI and EDVI, respectively. The definition of LVH was gender related: LVMI > 100 g/m² in women, and LVMI > 131 g/m² in men.[15]

All data were expressed as mean ± SD. Statistical analysis was performed using Student's t test. If variable was not normally distributed, the Mann-Whitney U test was used. Qualitative data were compared with the χ²-test; p < 0.05 was considered to be statistically significant.

Results

VC was found in 32 of 65 (49%) HD patients—VC(+) group. MVC was found in 10 patients (15%), AVC in 9 (14%), and both VCs (MVC + AVC) were found in 13 patients (20% of the study population). The clinical characteristic of patients without VCs, VC(−), and with VCs, VC(+), is presented in Table 1. Patients with VC were older, and the duration of HD therapy was longer compared with those without VC. SBP and pulse pressure were higher in VC(+) group. No significant differences were found with respect to calcium, phosphorus, mean Ca × P product, and PTH concentrations between VC(−) and VC (+) groups. But the incidence of Ca × P product above 4.43 mmol²/L² was higher in patients with VCs compared with those without—53.4% versus 38.1%, χ² = 18.28; p < 0,001 (Table 1).

Table 1. Characteristics of clinical parameters in patients without valve calcifications, VC(−), and with valve calcifications, VC(+)

Parameter	VC(−) (n = 33)	VC(+) (n = 32)	P
Age (years)	43 ± 11	51 ± 12	< 0.001
Gender (F/M)	17/16	18/14	NS
Duration of HD (months)	35 ± 40	41 ± 23	< 0.02
SBP (mmHg)	133.9 ± 15.0	142.9 ± 9.3	< 0.01
DPB (mmHg)	79.8 ± 8.4	79.8 ± 4.6	NS
PP (mmHg)	54.1 ± 8.6	63.2 ± 7.2	< 0.001
Calcium (mmol/L)	2.24 ± 0.20	2.22 ± 0.20	NS
Phosphorus (mmol/L)	2.11 ± 0.47	2.27 ± 0.41	NS
Ca × P (mmol^2/L^2)	4.71 ± 1.01	5.02 ± 0.92	NS
Ca × P > 4.43 (mmol^2/L^2)	151 of 396 (38.1%)	205 of 384 (53.4%)	< 0.001
PTH (pg/mL)	372.7 ± 315.9	522.7 ± 541.9	NS
Kt/V	1.24 ± 0.19	1.24 ± 0.15	NS
Albumin (g/dL)	4.30 ± 0.45	4.29 ± 0.33	NS

On echocardiography (Table 2), higher IVS and PW thickness, higher LVMI, and higher prevalence of LVH were found in patients with VC, despite that there were no patients with aortic stenosis and left ventricle-aortic pressure gradient above 20 mmHg (Table 2). LVMI was 133.4 ± 34.6 g/m² in patients without VC, 151.0 ± 43.3 g/m² in patients with AVC, 161.5 ± 56.2 g/m² in patients with MVC, and 198.1 ± 54.3 g/m² in patients with both VCs (AVC + MVC) (p < 0.05 vs. AVC patients and p < 0.001 vs. patients without VC) (Figure 1).

Table 2. Echocardiographic parameters of the patients without valve calcifications, VC(−), and with valve calcifications, VC(+)

	VC(−) (n = 33)	VC(+) (n = 32)	P
IVS (mm)	13.1 ± 2.9	15.6 ± 2.7	< 0.001
LVID (mm)	48.3 ± 5.9	47.2 ± 7.6	NS
PW (mm)	11.3 ± 1.8	13.3 ± 2.6	< 0.001
EDVI (mL/m^2)	47.6 ± 12.9	50.5 ± 14.1	NS
LVMI (g/m^2)	133.4 ± 34.6	173.4 ± 54.7	< 0.001
LVH	21 (63%)	28 (88%)	< 0.05
EF (%)	65.7 ± 7.2	66.7 ± 8.7	NS

Figure 1. Left ventricular mass index in patients without valve calcifications (VC[−]), with aortic valve calcifications (AVC), mitral valve calcifications (MVC), and both valves calcifications (AVC + MVC) (p < 0.05 for AVC + MVC vs. AVC patients and p < 0.001 for AVC + MVC vs. VC [−] patients).

To further investigate the relationship between calcium-phosphate metabolism and VC, patients with VC were divided into two groups (Table 3): patients with single VCs—Group 1VC(+) n = 19 (mitral or aortic)—and patients with both VCs—Group 2VC(+) n = 13 (mitral and aortic).

Table 3. Characteristics of calcium-phosphate metabolism parameters in patients with single valve calcification and with both valve calcifications

	1 VC(+) (n = 19)	2 VC(+) (n = 13)	P
Ca (mmol/L)	2.17 ± 0.19	2.31 ± 0.19	p < 0.05
P (mmol/L)	2.22 ± 0.25	2.34 ± 0.58	NS
Ca × P (mmol^2/L^2)	4.79 ± 0.54	5.36 ± 1.23	NS
Ca × P > 4.43 (mmol^2/L^2)	113 of 228 (49.6%)	92 of 156 (59.0%)	NS
PTH (pg/mL)	302.9 ± 224.8	843.9 ± 703.2	p < 0.05

Calcium concentration was significantly higher in patients with both VCs compared with patients with single VCs. Also, PTH concentration was significantly higher in patients with both VCs than in patients with single VCs and patients without VCs, (p < 0.05 for both). Patients with both VCs did not differ from those with single VCs with respect to phosphate concentration and mean calcium-phosphorus product. We found very high incidence of Ca × P product above 4.43 mmol²/L², but difference between patients with both VCs than with single VC was not significant: 59.0% versus 49.6%, χ² = 3.29, NS.

Discussion

Our study confirms that valvular heart calcification is a common finding in ESRD patients. VC was found in 49% of HD patients. In previous studies with echocardiography, VCs were found in 23% to 60% of dialysis patients.[6&7], [16], [17], [18] Mitral VCs were found in 33% to 46.3% and aortic VCs in 28% to 52%,[4&5], [18], [19] whereas the prevalence of MVC and AVC in the nondialysis control group (ages 59 ± 17 years) was 10% and 4.3%, respectively.[4] Using electron-beam computed tomography (EBCT) MVC and AVC was detected in 59% and 55% of HD patients, respectively.[3]

In agreement with data from literature we found association between VC and age and duration of HD therapy. Aging is the most important risk factor for VC in HD patients.[3-5], [18] Duration of dialysis therapy seems to be another important risk factor for VC.[5], [18&19] Such relation was also found for MVC, but not for AVC, in the study by Ribeiro et al.[4] However, in the study of Covic et al., age and duration of HD therapy did not differ between patients with and without MVC.[16]

Elevated Ca × P product was predictive parameter for VC in studies of Maher et al. and Salguiera et al.[14], [19] It is of interest that increased Ca × P product was the risk factor for MVC but not AVC in the study of Ribeiro et al.[4] Also, Ca × P product was correlated with MVC score in this study. In contrast, Pannucio et al. did not found association between Ca × P product and VC.[7]

In our study, incidence of elevated Ca × P, but not mean Ca × P product, was associated with VC. It cannot be exclude that even short-lasting incidents of increased calcium-phosphorus product or hypercalcemia, difficult to detect with routine twice monthly measurements, may lead to metastatic soft-tissue calcifications in ESRD patients.

In the study of Braun et al., AVCs were correlated with coronary artery calcium score. In the multiple regression model, age and hypertension, but not Ca, P, and PTH, were correlated with degree of cardiac artery calcifications, but direct relation with VC was not evaluated in this study.[3]

Despite many clinical studies the role of PTH in cardiac VCs remains unclear. Our study indicates the potential role of advanced secondary hyperparathyroidism in cardiac VC. Mazzaferro et al. found in their study that serum calcium concentration and PTH level, but not phosphorus concentration, were higher in patients with MVC compared with those without.[5] These relations were found in HD patients but not in predialysis patients. Other authors did not confirm any relation between PTH and cardiac VCs.[4],[16], [18]

In the opposite, Tsuchihashi et al. found in their study, that hypoparathyroidism with PTH < 60 pg/mL was the risk factor of calcified cardiovascular complications.[20] In this study, patients with hypoparathyroidism had also higher calcium concentration, and the proportion of patients receiving active vitamin D₃ tended to be higher than in those with hyperparathyroidism (PTH > 200 pg/mL). The prevalence of secondary hyperparathyroidism was 27.1% in this study, whereas in our study 44 of 65 patients (68%) had PTH > 200 pg/mL. Also, in the study of Pannucio et al., PTH concentration was lower in HD patients with VC compared with patients without VC.

Soft-tissue calcifications may be correlated both with very high and low PTH concentration.[21&22] Adynamic bone disease with low PTH concentration may predispose to calcium deposition in soft tissue as a result of decreased bone buffer capacity.[22] In contrast, severe hyperparathyroidism with parathyroid adenoma often leads to difficult to manage hypercalcemia and hyperphosphatemia. We can speculate that relationship between PTH and VCs (or soft-tissue calcifications as a complication of ESRD in general) may represent U-curve shape, with the highest risk with very low and very high PTH levels.

Left ventricular hypertrophy is an independent risk factor of cardiovascular mortality in ESRD patients.[23] VC is also an independent predictor for all-cause and cardiovascular mortality in peritoneal dialysis patients, with extremely high all-cause mortality in patients with both: mitral and aortic valves calcified.[6] In HD patients, VC was associated with all-cause and cardiovascular mortality in unadjusted analysis.[7] But after adjusting for other CV risk factors, CV complications and LVMI, it lost significance. In our study, we found very high LVMI in patients with both VCs, with LVH criteria met in all patients in this group. This finding may at least partly explain high all-cause and cardiovascular mortality in patients with both VCs in the study of Wang et al. In this study, both VCs were associated with 57% 1-year all-cause mortality, but no adjustment for LVMI and other cardiovascular risk factors was done.[6]

In our study, patients with cardiac VCs had higher systolic and pulse pressure but not diastolic pressure than patients without VCs. Hypertension is a well-known pathogenic factor of LVH due to increased afterload.[24] It is also suggested that hypertension acts as pathogenic factor of VC due to increased mechanical stress.[19],[22] In the study of Ribeiro et al., mitral VC was associated with longer duration of predialysis hypertension, but there was no correlation between actual blood pressure and VC.[4] In another study, Salgueira et al. also did not find a relationship between hypertension and VC.[14]

In contrast, a relationship between AVC and peripheral VC was found in clinical studies.[4],[19] Peripheral VC contributes to increased arterial stiffness,[25] which acts as pathogenic factor of LVH.[26&27] In our study, VC coexists with higher LVM and higher systolic and pulse pressure. Increase in pulse pressure reflects increased arterial stiffness. We assume cardiac VC is associated with increased arterial stiffness, but this issue needs further investigation.

Conclusion

Cardiac VCs are associated with LVH, particularly in patients with both VCs. Calcium-phosphate product is an important target in HD patients. Even short-lasting incidents of elevated Ca × P product should be avoided to protect against cardiac VCs.