We found no differences between alfacalcidol and paricalcitol and their influence on NT-proBNP, OPG or inflammatory markers during 16 weeks of treatment in hemodialysis patients. Fetuin-A increased significantly in the alfacalcidol-treated group compared with the paricalcitol-treated group during the first treatment period. Owing to a significant period effect we could not analyze crossover data for differences in changes in fetuin-A.
Fetuin-A was shown in experimental studies to be an inhibitor of vascular calcification
[21, 22] and low levels of fetuin-A have been associated with increased cardiovascular morbidity and mortality in dialysis patients
[23–25]. It is unknown whether an increase in fetuin-A level, as found in the present study, protects against cardiovascular disease.
Fetuin-A is known to be a negative acute phase reactant
, and in accordance with former cross-sectional studies in dialysis patients demonstrating a negative association between levels of Crp and fetuin-A
[25, 27], we even found a negative correlation between changes in these parameters. This points to fetuin-A being part of the malnutrition-inflammation-atherosclerosis syndrome in dialysis patients.
We found an increase in fetuin-A during the first treatment period in both groups. This increase was only statistically significantly higher in the alfacalcidol-treated group compared with the paricalcitol-treated group. However, a significant increase in fetuin-A during paricalcitol treatment has been demonstrated after eight weeks of treatment in an uncontrolled study in hemodialysis patients
. There was a tendency towards a higher prevalence of patients with diabetes at baseline in the paricalcitol-treated group. This may modulate changes in fetuin-A, as diabetic patients had higher fetuin-A levels and may respond differently than non-diabetic patients
[29, 30]. The discrepancy between the effects in period 1 and period 2 could be due to a carryover effect from the first treatment period, or the significant difference during period 1 may simply be due to chance. The actual effect of vitamin D analogs on fetuin-A remains to be demonstrated in a placebo-controlled trial.
Alfacalcidol and paricalcitol appeared to increase OPG. This increase may be due to a direct effect of vitamin D analogs or to suppression of parathyroid hormone, which was found to correlate with changes in OPG. Because PTH decreased least in the paricalcitol-treated patients during the second treatment period, this may explain why OPG did not show any significant changes during this period. The changes in OPG during vitamin D analog treatment have, as far as we know, only been studied in an uncontrolled trial of maxacalcitol, where the opposite effects were found, namely that this led to a decrease in OPG
OPG is an important factor in bone metabolism where OPG acts as a decoy receptor for receptor activator of nuclear factor кB (RANK) ligand (RANKL) and inhibits activation of RANK, inhibits maturation of osteoclasts and prevents bone resorption
. In animal models, OPG deficiency causes vascular calcification
 and OPG treatment has been shown to block vascular calcification
. However, in dialysis patients, OPG has been associated with calcification and the progression of calcification
[35–37]; both in the general population and in the population of chronic kidney disease (CKD) patients, OPG was associated with increased cardiovascular morbidity and mortality
[18, 38, 39]. Whether an increase in OPG during treatment with vitamin D analogs increases the risk of cardiovascular disease or actually reflects a vascular protective mechanism, remains to be further explored.
Left ventricular hypertrophy and cardiac dysfunction are risk factors for cardiovascular mortality in patients with CKD
. NT-proBNP predicts cardiovascular and total mortality in hemodialysis patients
[41, 42]. In experimental models, vitamin D analogs inhibit left ventricular hypertrophy
[12, 13]. In the PRIMO trial, paricalcitol attenuated the increase in BNP compared with placebo in patients with CKD stages 3–4
. However, in the PRIMO trial and the very similar OPERA trial
, paricalcitol did not change the left ventricular mass index. In post hoc analysis, a reduction in left atrial volume was detectable after 48 weeks of paricalcitol treatment, which could be an early marker of an increase in left ventricular mass
. We explored changes in NT-proBNP during alfacalcidol and paricalcitol treatment in hemodialysis patients and found, as in predialysis patients, a steady increase in NT-proBNP during the intervention period in both treatment groups. The increase in Nt-proBNP may have been attenuated by the vitamin D analogs. Unfortunately, no untreated group was present.
High levels of parathyroid hormone have been associated with left ventricular hypertrophy and high levels of NT-proBNP in patients with CKD and end-stage renal-disease
. This may be due to a direct effect of parathyroid hormone on cardiac myocytes
. We explored the relation between changes in PTH during treatment with vitamin D analogs and changes in NT-proBNP, and found that the degree of suppression of hyperparathyroidism during 16 weeks of treatment did not influence changes in NT-proBNP.
FGF23 levels are increased as kidney function declines
 and in dialysis patients treatment with vitamin D analogs increases FGF23
. High FGF23 levels are associated with increased mortality and left ventricular hypertrophy in dialysis patients
[48, 49]. Experimental studies support that FGF23 induces left ventricular hypertrophy
. We found no relation between changes in FGF23 during treatment with vitamin D analogs and changes in NT-proBNP.
The reason why changes in parathyroid hormone or FGF23 do not relate to changes in NT-proBNP may be due to the short intervention period or because these factors are only biomarkers of other mechanisms involved in the pathophysiology of left ventricular hypertrophy and heart failure in patients with CKD.
The malnutrition-inflammation complex are common in dialysis patients and a risk factor for morbidity and mortality
. Experimental studies have demonstrated an anti-inflammatory effect of vitamin D
[9, 52]. We did not find any changes in inflammatory markers during treatment with alfacalcidol or paricalcitol. These findings are in conflict with some previous reports. In a placebo-controlled oral study in CKD stages 1–3 patients
 and in two uncontrolled (intravenous and oral) studies in dialysis patients, a decrease in hs-Crp, IL-6 and TNF-α was found during paricalcitol treatment
[54, 55]. However, Moe et al. did not find any changes in TNF-α or IL-6 after 12 weeks of treatment with intravenous paricalcitol in a placebo-controlled study in hemodialysis patients with low PTH
. Furthermore, 8 weeks of high-dose cholecalciferol did not influence inflammatory markers in predialysis and dialysis patients
. Whether an anti-inflammatory effect of vitamin D and its analogs depends on the administration route, dose or patient population including level of hyperparathyroidism, remains to be explored. Furthermore, the present study may be insufficiently powered to detect minor changes in inflammatory parameters.
The present study has several limitations. The study size was small and minor differences may not be detected. Patients participating in the present study were comparable with participants in the main study concerning baseline values; still the missing data may lead to bias, especially in the analysis of uncrossed data. These prevalent patients may differ from incident dialysis patients. Other markers, especially ones involved in calcification, such as matrix-GLA-protein and osteopontin, and direct measurement of vascular calcification using cardiac-CT or abdominal x-ray, may add further information to the influence of vitamin D analogs on calcification in dialysis patients.