FGF-23, α-Klotho Gene Polymorphism and Their Relationship with the Markers of Bone Metabolism in Chronic Peritoneal Dialysis Patients

Abstract

Objective:

The aim of the present study was to evaluate the variations of some major bone metabolism markers with reference to klotho gene polymorphism (KGP) and bone mineral density (BMD) values in patients on chronic peritoneal dialysis (CPD).

Materials and Methods:

In 51 CPD patients and 40 healthy persons, assays for intact parathormone (iPTH), fibroblast growth factor 23 (FGF-23), osteoprotegerin (OPG), osteocalcin (OC), procollagen type-1 N terminal propeptide (PINP), beta- crosslaps (beta CTx), tartrate resistant acid phosphatase (TRAP5b), bone alkaline phosphatase (BAP), 1,25(OH)D3, and 25(OH)D3 and α-klotho gene mutations were performed.

Results:

In CPD patients, 1,25(OH)D3 and 25(OH)D3 deficiency rates were 96% and 94% respectively. iPTH (249 pg/mL vs 39 pg/mL) and FGF-23 (1089 RU/mL vs 153 RU/mL), OPG, OC, PINP, beta CTx, TRAP5b levels were significantly higher in patients. iPTH levels and whole-body BMD values were negatively correlated in patients. The rate of KGP was similar in all groups.

Conclusion:

In CPD patients, besides vitamin D deficiency, high levels of OPG, OC, PINP, beta CTx, TRAP5b were evident. Positive correlation between iPTH levels and BAP and PINP levels suggested a diagnostic value for those markers during the management of CKD MBD. On the other hand, high serum TRAP5b concentrations did not seem to be affected by neither calcitriol treatment nor the severity of hyperparathyroidism. iPTH and FGF-23 levels and whole-body BMD values showed a significant negative correlation. We were unable to show any correlation between KGP and any of the CKD-MBD parameters measured in this study.

Introduction

Chronic kidney disease mineral and bone disorder (CKD-MBD) is a complex and multifactorial entity. Although parathormone is assumed to be the most prominent metabolic marker during the course of CKD-MBD, now it is well known that several newly recognized other endocrine, paracrine and autocrine factors do contribute in the CKD-MBD scene. One of the major physiologic contributors, fibroblast growth factor-23 (FGF-23) which is a hormone produced by osteocytes in response to hypophosphatemia, increases the urinary excretion of inorganic phosphate (Pi) and inhibits the renal production of 1,25(OH)D3, helping to mitigate hypophosphatemia in patients with chronic kidney disease and also in normal persons [1]. It has been shown that in patients with chronic kidney disease (CKD), circulating FGF-23 levels are markedly elevated and independently associated with mortality [2]. FGF-23 induced receptor activation requires klotho as a co-receptor. Klotho gene is expressed in kidneys, parathyroid glands and brain. Klotho strongly attaches to FGF-23 membranous receptors located in the cellular membrane and modulates the functions of several cellular ion and transporter channels [3]. Klotho gene is also known to have some anti-aging and anti-atherosclerosis properties [4, 5]. Klotho gene polymorphism has been shown to be associated with the severity of renal failure and aging parameters such as common carotid artery intima media thickness, low density lipoprotein and uric acid levels in CKD patients [6–8]. Klotho gene mutations may cause early dysfunction and thrombosis of vascular accesses in haemodialysis patients [9]. Therefore, klotho gene polymorphism may potentially affect all FGF-23 related CKD-MBD parameters as well as mortality and morbidity in CKD.

Another important factor in regulating bone metabolism in uremic patients is RANKL-RANK pathway. At the bone level, 1,25(OH)D3 increases the expression of the receptor activator of nuclear factor-κB ligand (RANKL) in osteoblasts and induces osteoprotegerin (OPG) synthesis, an inhibitor of osteoclast activity and osteoclastogenesis [10].

Dual energy X-ray absorptiometry (DEXA) has been widely used for the assessment of bone mineral deficiency status in non-ESRD population and also in ESRD patients. In ESRD patients, however, interpretation of the results obtained by DEXA technique is conflicting. In some DEXA based studies in peritoneal dialysis patients, assessment of BMD resulted in a similar prevalence rate of osteoporosis compared to the healthy controls and an increased prevalence of osteoporosis in some others [11–13]. In studies specifically performed in peritoneal dialysis patient population, DEXA based assessment of BMD has shown an osteoporosis rate of 19 % and 36% of osteopenia, indicating the presence of subnormal amount of bone mass in 55% of patients [14].

In this study, along with klotho gene polymorphism and bone mineral density, we have evaluated the levels of some other major important parameters of bone turnover and metabolism namely iPTH, FGF-23, osteoprotegerin, osteocalcin (OC), procollagen type-1 N terminal propeptide (PINP), beta crosslaps (beta CTx), tartrate resistant acid phosphatase (TRAP5b), bone alkaline phosphatase (BAP), 1–25(OH)D3, and 25(OH)D3, in chronic peritoneal dialysis patients.

Materials and Methods

This cross sectional, controlled study included 51 patients on chronic peritoneal dialysis either on CAPD or APD (Group 1; G1), and 40 healthy persons (Group 2; G2) (Table 1). Patients further divided into two subgroups regarding their iPTH levels as Group 3 (G3; iPTH<300 pg/mL; n=31) and Group 4 (G4; iPTH >300 pg/mL; n=20). Mean age of the patients in G1 and G2 were similar (51±16 years and 47±11 years respectively; p=0.171). Gender distributions in groups 1 and 2 were also similar (F/M; 19/32-22/18 respectively, p=0.091). Mean duration of the patients on peritoneal dialysis treatment was 43±40 months. Demographic data of G3 and G4 are shown in Table 2. All patients who have been on chronic peritoneal dialysis treatment more than 3 months, older than 20 years, with no history of more than one month of corticosteroid therapy, organ transplantation, parathyroidectomy or any acute infection within last two months have been considered to be eligible for the study. Patients were on their regular ESRD medications including phosphate binders, antihypertensives, vitamin D and its analogues indicated by their clinical status and laboratory values without any restriction. No patient was on any specific treatment for osteoporosis including calcitonin or bisphosphonates. 23 patients have been on calcitriol treatment and 28 patients were not on any kind of vitamin D receptor activation therapy. In all patients and control subjects, bone mineral density was measured by DEXA. In all participants of the study, serum fibroblast growth factor-23, intact parathormone, calcium, inorganic phosphorus, bone alkaline phosphatase, osteoprotegerin, osteocalcin, tartrate resistant acid phosphatase, beta CTx, procollagen type-1 N-terminal propeptide, 1,25(OH) D3 and 25(OH)D3, Pi, Ca assays were performed. For all biochemical assays, venous blood samples were stored at −80 °C until the laboratory assays were performed. ELISA based techniques used for TRAP 5b (Immunodiagnostic Systems, IDS Ltd., Boldon, Tyne & Wear, NE35, 9PD); bone alkaline phosphatase (Immunodiagnostic Systems, IDS Ltd., Boldon, Tyne & Wear, NE35, 9PD); osteoprotegerin (Bender MedSystems GmbH, Vienna, Austria); C-Terminal FGF; (ALPCO Diagnostic, Salem, NH 03079, USA) assays. 1,25(OH)D3 assays performed through radioimmunoassay technique (Biosource Europe S.A, Niveles, Belgium). For N-mid osteocalcin, beta CTx, 25(OH)D3, PINP, electrochemiluminescence immunoassay (ECLIA) technique was used (Roche Modular Analytics E170 Immunoassay Analyser, Roche Diagnostics GmbH, Mannheim, Germany). Bone mineral density was measured by a dual energy X-ray absorptiometer (GE Healthcare, Lunar DPX Bone Densitometer).

Table 1.

Comparison of bone turnover and other bone metabolism indicators between chronic peritoneal dialysis patients and healthy subjects

Parameters	Group 1 (PD patients) n=51	Group 2 (Healthy control group) n=40	p value
Gender (F/M)	19/32	22/18	0.091
Age	51±16	47±11	0.171
Body mass index (kg/m2)	26.3±5.1	24.5±2.7	0.034
i-PTH (pg/mL)	249 (3–1635)	39 (15–107)	0.000
FGF-23 c-terminal (RU/mL)	1089 (51–1626)	153 (24–991)	0.000
OPG (pg/mL)	56.6(2.3–1338)	19.9 (1.2–98)	0.000
OC (ng/mL)	110.5(20–1477)	23.7 (13–300)	0.000
PINP (ng/mL)	422.5±412.5	49.5±19.3	0.000
Beta CTx (ng/mL)	2.19±2.17	0.24±0.15	0.000
TRAP5b (U/L)	1.9±1.3	1±0.5	0.000
1,25(OH)D3 (pg/mL)	8±5.2	28.4±11	0.000
25(OH)D3 (ng/mL)	9.4±5.9	20.6±9.2	0.000
Total ALP (U/L)	339±204	203±56	0.000
BAP (U/L)	32.7±27	12.2±5.2	0.000
Ca (mg/dL)	9.3±0.6	9.6±0.4	0.021
Pi (mg/dL)	4.7±1.2	3.7±0.5	0.000
CaXPi (mg2/dL2)	44.3±10	36.3±5.7	0.000
Klotho gene polymorphism (Wild/heterozygous/mutated) (n)	38/13/0	30/9/1	0.507
Klotho gene polymorphism (Wild/heterozygous/mutated) (%)	74.5/25.5/0	75/22.5/2.5	0.507
Whole-Body BMD (g/cm2)	1.12±0.13	1.16±0.1	0.276

i-PTH: intact parathormon; FGF-23: fibroblast growth factor 23; RU/mL: reference unit/mL; OPG: osteoprotogerin; OC: osteocalcin; PINP: procollagen type-1 N terminal propeptide; Beta CTx: beta-crosslaps; TRAP5b: tartarate resistant acid phosphatase; ALP: alkaline phosphatase; Ca: calcium; Pi: inorganic phosphorus

Table 2.

Comparison of various CKD-MBD parameters in chronic peritoneal dialysis patients with lower (Group 3: PTH<300 pg/mL) and higher (Group 4: PTH>300 pg/mL) iPTH levels

	Group 3 (PTH<300 pg/mL) n=31	Group 4 (PTH>300 pg/mL) n=20	P
Age (Years)	58±14	43±15	0.007
Gender (F/M)	9/22	10/10	0.112
Duration of PD (Months)	28.8±22	65.9±50	0.004
i-PTH (pg/mL)	160±86	718±416	0.000
Calcitriol use (+/−)	11/20	12/8	0.089
FGF-23 (RU/mL)	688±518	1158±406	0.004
BAP (U/L)	21±9	52±37	0.000
OPG (pg/mL)	55.5 (2.3–338)	55 (2.8–136)	0.692
OC (ng/mL)	95.6 (20–879)	380 (20–1500)	0.000
PINP (ng/mL)	270 (7.4–1200)	700 (97–1829)	0.001
Beta CTx (ng/mL)	1.2±0.87	3.8±2.6	0.000
TRAP5b (U/L)	1.7±1	2.2±1.6	0.390
1,25(OH)D3 (pg/mL)	7.6±44	7.7±5.7	0.877
25(OH)D3 (ng/mL)	11±6.4	8.7±8.7	0.028
Ca (mg/dL)	9.3±0.5	9.2±0.7	0.862
Pi (mg/dL)	4.4±0.4	5.4±1.3	0.018
CaXPi (mg2/dL2)	40.6±8.4	51±10	0.009
Whole body BMD (g/cm2)	1.17±0.1	1±0.12	0.000
Body mass index (kg/m2)	27.4±5	25.4±5	0.356
Klotho gene polymorphism (n) (Wild/heterozygous/mutated)	24/7/0	14/6/0	0.553
Klotho gene polymorphism (%)(Wild/heterozygous/mutated)	63.2%/53.8%/0	36.8%/46.2%/0	0.553

i-PTH: intact parathormon; FGF-23: fibroblast growth factor 23; RU/mL: reference unit/mL; OPG: osteoprotogerin; OC: osteocalcin; PINP: procollagen type-1 N terminal propeptide; Beta CTx: beta-crosslaps; TRAP5b: tartarate resistant acid phosphatase; ALP: alkaline phosphatase; Ca: calcium; Pi: inorganic phosphorus

Klotho Gene F352V (rs9536314) Polymorphism Analysis

DNA extracts were prepared from whole blood samples containing EDTA using Qiagen EZ1 DNA Blood 200 mL kits (Catalogue No: 951034, QIAGEN GmbH, Hilden, Germany) and BioRobot® EZ1 (QIAGEN GmbH, Hilden, Germany) apparatus and kept at −80 °C until the day of gene mutation studies. DNA content was measured in all DNA extracts and was diluted to obtain a concentration of 2 ng/mL. Sample DNA was amplified using real-time polymerase chain reaction (PCR) (Fw Primer: ACTATCCCGAGAGCATGAAGAATAAC, Rw Primer: AAAGTCAGCAGTTCCTTTGATGAAC (anti: GTTCATCAAAGGAACTGCTGACTTT), raZor Prob: TTTTCTCAGATTCAGTAAAATCAGGCAGAATAGATGA (anti: CATCTATTCTGCCTGATTTTACTGAATCTGAGAAAA) with Rotor-Gene® Q) (QIAGEN GmbH, Hilden, Germany) under the conditions shown in Table 3.

Table 3.

Polymerase chain reaction procedure for Klotho gene F352V (rs9536314) polymorphism analysis

	Step	Time	Temperature
	Enzyme activation	8 min	95°C
Touchdown Cycles X 6	Denaturation	15 s	95°C
	Reduce by 1.0 °C/cycle	45 s	66 – 61°C
Cycling X 50	Denaturation	15 s	95°C
	Data Collection	45 s	60°C

After PCR, products were heated to denaturation at 95 °C for 45 s which was followed by cooling down to 55 °C for 2 min to facilitate the heteroduplex formation and thereafter melting slowly at 0.1°C/s from 55°C to 95°C (Amplikon Tm: (+/−2°C) 77.7°C). High resolution melting (HRM)-curve analysis and genotyping were performed by fluorescent signal detection (The emission signal was assessed using the fluorescent dye SYBR Green) with Rotor-Gene® Q 1.7 software (Figure 1). Positive and negative controls (homozygous wild type, heterozygous, and homozygous mutant) were always included in each assay. The study protocol was approved by the medical ethics committee of Akdeniz University Faculty of Medicine, Antalya.

Figure 1.

Derivative (dF/dT) plot of melting curve consisting of two melting regions. Samples with the V allele (blue) had a lower Tm of 63.9°C, while samples harboring the F allele (green) showed a Tm of 67.2°C. The V/F heterozygous samples (red) manifested both melting peaks of these alleles. Each genotype had a unique path distinguishing itself from the others.

Statistical Analysis

An SPSS for Windows 13:0 statistical package (Statistical Package for Social Sciences; Chicago, IL, USA) was used for statistical analysis. Continuous variables were expressed as means±SD and compared using Student t test. If not normally distributed, continuous variables were compared using Mann Whitney U test. Categorical variables were expressed as per cents and compared using chi square test. For assessing normalization, we used Shapiro-Wilk test. All statistical tests were performed at p<0.05 level.

Results

As shown in Table 1, age and gender distribution was similar in patient and control groups. PD patients had a higher average body mass index than the controls (26.3±5 vs 24.5±2.7 respectively, p=0.034). iPTH (249 pg/mL vs 39 pg/mL) and FGF-23 (1089 RU/mL vs 153 RU/mL) levels were significantly higher in patients. Similarly, OPG, OC, PINP, beta CTx, TRAP5b, BAP, Pi levels and CaXPi were found to be higher in patient group compared to the healthy control subjects. On the other hand, both 1.25(OH)D3 and 25(OH)D3 levels were significantly lower in patient group. Out of 51 patients, only 2 patients for 1.25(OH)D3 (>22 pg/mL) and 3 for 25(OH) D3 (>15 ng/mL) were not deficient indicating a deficiency rate of 96% and 94% respectively. No significant difference was found between patient and control groups regarding klotho gene polymorphism. Whole body bone mineral density was significantly lower in patient group (Table 1).

As shown in Table 2, in order to increase the relevance of our FGF-23 related findings with current clinical practice, we analysed the relationship between serum FGF-23 concentrations with some major CKD-MBD related clinical parameters and frequently used laboratory cut off values including gender, duration of PD, calcitriol usage, serum concentrations of iPTH (300 pg/mL), Ca (9.5 mg/dL), Pi (5.5 mg/dL), CaXPi (55 (mg2/dL2) and also for T score values describing normal, osteopenic and osteoporotic bone mineral density status. FGF-23 levels were higher in women, patients not using calcitriol, and also in patients with higher measured serum iPTH, CaXPi, longer time on peritoneal dialysis treatment and lower serum calcium. When patients were stratified based on their T-scores, osteoporotic subjects had significantly higher FGF-23 levels compared to the patients with osteopenia and also normal BMD values (Table 4).

Table 4.

C-terminal FGF-23 levels in various stratifications based on some major CKD-MBD related parameters and frequently used clinical reference limits

Parameter	Group characteristics	FGF-23 level	p
Gender	Woman (n:19)	1069±527	0.017
	Man (n:22)	756±498
PD duration (Months)	≤24 (n:21)	543±484	0.000
	>24 (n:30)	1103±427
Calcitriol use	Users (n:23)	1133±418	0.002
	Non-users (n:28)	659±516
iPTH (pg/mL)	≤300	688±518	0.004
	>300	1158±406
Ca (mg/dL)	<9.5 (n:28)	808±552	0.369
	≥9.5 (n:23)	952±494
Pi (mg/dL)	<5.5 (n:38)	729±510	0.001
	≥5.5 (n:13)	1291±316
CaXPi (mg2/dL2)	<55 (n:40)	761±519	0.003
	≥55 (n:11)	1278±323
Whole-Body T score	>−1 (n:29) → 1st group	730±510	1–2:0.293
	−1/−2.5 (n:13) → 2nd group	888±532	1–3:0.001
	>−2.5 (n:9) → 3rd group	1311±331	2–3:0.043

When patients compared based on their oral calcitriol use, among various parameters of CKD-MBD including iPTH, FGF-23, OPG, OC, PINP, beta CTx, TRAP5b, Ca, Pi, CaXPi, BAP, 1.25(OH)D3, 25(OH)D3, and whole body BMD, only FGF-23 and iPTH levels and PD duration have shown a statistically significant correlation with calcitriol use. In calcitriol users, C-Terminal FGF-23 levels were significantly higher (p=0.002). Calcitriol users had higher serum iPTH concentrations compared to non-users probably due to higher rate of adynamic bone disease in patients not using calcitriol treatment. Longer time on peritoneal dialysis was associated with higher incidence of calcitriol use (Table 5).

Table 5.

Calcitriol treatment and its association with some CKD-MBD parameters

	Calcitriol users (n=23)	Calcitriol non-users (n=28)	p
Gender (F/M)	11 (47.8%)/12 (52.2%)	8 (28.6%)/20 (71.4%)	0.157
Age (years)	46±15	54±16	0.182
PD duration (Months)	42 (6–173)	22.5 (6–131)	0.020
i-PTH (pg/mL)	330 (2.9–1635)	198 (10.8–1514)	0.039
Ca (mg/dL)	9.3±0.5	9.2±0.6	0.271
CaxPi (mg2/dL2)	45.7±8.9	43.5±11.8	0.307
Pi (mg/dL)	4.8±0.9	4.7±1.4	0.526
BAP (U/L)	40±31	27.5±24	0.176
1,25(OH)D3 (pg/mL)	9.3±6.7	6.9±3.4	0.508
25(OH)D3 (ng/mL)	9.9±8.6	10.3±6.1	0.391
OPG (pg/mL)	45.8 (2.8–133)	57.8 (2.3–1338)	0.147
OC (ng/mL)	285 (20–1500)	98.5 (31–895)	0.056
PINP (ng/mL)	603±523	294±263	0.135
Beta CTx (ng/mL)	2.9±2.6	1.72±1.6	0.291
TRAP5b (U/L)	1.7±1.2	2±1.4	0.670
FGF-23 (RU/mL)	1133±418	659±516	0.002
Whole-Body BMD (g/cm2)	1±0.1	1.13±0.1	0.335

i-PTH: intact parathormon; FGF-23: fibroblast growth factor 23; RU/mL: reference unit/mL; OPG: osteoprotogerin; OC: osteocalcin; PINP: procollagen type-1 N terminal propeptide; Beta CTx: beta-crosslaps; TRAP5b: tartarate resistant acid phosphatase; ALP: alkaline phosphatase; Ca: calcium; Pi: inorganic phosphorus

As shown in Table 4, 51 chronic peritoneal dialysis patients participated in the study were also divided into two groups depending on their iPTH levels as Group 3 (G3, iPTH<300 pg/mL, n=31) and Group 4 (G4, iPTH>300 pg/mL, n:20). Calcitriol use was similar in two groups (G3: 11/20 vs G4: 12/8, p=0.089). Serum Ca, 1.25(OH)D3, OPG and TRAP5b levels were also similar in two groups. 25(OH)D3 level was significantly higher in patients with lower serum levels of iPTH (11±6.4 vs 8.7±8.7; G3 vs G4, p=0.028). Out of 51 patients, only 2 patients for 1,25(OH)D3 (>22 pg/mL) and 3 for 25(OH)D3 (>15 ng/mL) were not deficient for vitamin D indicating a deficiency rate of 96% and 94% respectively. FGF-23, OC, PINP, beta CTx, BAP as well as 25(OH)D3, Pi and CaXPi values were significantly higher in Group 4 (Table 2). Whole body BMD values were also found to be significantly lower in Group 4. No significant difference was observed regarding the prevalence of their klotho gene polymorphism between groups 3 and 4 (Table 2). Bone mineral density was lower in high PTH group (G4 vs G3) when measured both in whole body measurement technique (Table 6) and also specifically at femoral site (Table 7).

Table 6.

Bone mineral density values measured using whole body technique for whole body and body regions in patient groups G3 and G4

	Group 3 (PTH<300 pg/mL) n=31	Group 4 (PTH>300 pg/mL) n=20	p value
Whole body	1.17±0.1	1±0.12	0.000
Whole body BMD measurement expressed in various body regions
Cranium	2.2±0.3	1.8±0.3	0.002
Arm	0.9±0.1	0.8±0.1	0.001
Leg	1.2±0.1	1±0.1	0.000
Trunk	0.9±0.1	0.8±0.1	0.001
Ribs	0.7±0.1	0.6±0.07	0.013
Pelvis	1±0.1	0.9±0.1	0.003
Vertebra	1±0.1	0.9±0.1	0.000

Table 7.

Bone mineral density values as specifically measured at femoral site in patient groups G3 and G4

	Group 3 (PTH<300 pg/mL) n=31	Group 4 (PTH>300 pg/mL) n=20	p
Femur neck	0.86±0.1	0.73±0.1	0.014
Femur upper neck	0.74±0.17	0.63±0.14	0.020
Femur wards	0.71±0.17	0.57±0.18	0.012
Femur trochanter	0.76±0.14	0.64±0.13	0.003
Femur trunk	1.1±0.16	0.93±0.21	0.003
Femur total	0.92±9.14	0.78±0.15	0.003

Discussion

Calcium-phosphate disturbances in chronic renal failure and dialysis patients affect the quality of life due to bone and joint pain, low-trauma fractures and clinical problems associated with metastatic calcifications. It is also firmly established that CKD-MBD is responsible for the increased cardiovascular mortality and morbidity rate in patients with chronic renal failure. Hence, calcium and phosphate disturbances are more than being solely a bone problem in dialysis population and in order to increase our success in designing and maintaining our treatment options in patients with chronic renal failure, we need to elucidate the relationship between the determinants of CKD-MBD. Although peritoneal dialysis is an established mode of treatment as a chronic dialysis option worldwide, most of our knowledge on CKD-MBD parameters in dialysis patients has been derived from the studies carried out in haemodialysis patients. However, the course, clinical presentation and treatment of CKD-BMD in patients on peritoneal dialysis, is not necessarily identical with the haemodialysis patients. In peritoneal dialysis patients, factors such as steady-state biochemical control, relatively higher phosphorus intake and clearance, additional peritoneal loss of 25(OH) D3 and higher glucose load may affect the course of CKD-MBD making its nature different than haemodialysis patients [15, 16]. In this study, we aimed to expand our knowledge on major metabolic parameters of bone mineral metabolism and their interactions in chronic peritoneal dialysis patients. Although we have several new bone metabolic markers to be used during the clinical course of CKD-MBD, serum iPTH concentration still remains as the main parameter in the diagnosis and management of metabolic bone problems in dialysis patients [17]. We believe, the results of our study support that opinion.

Our results show that, serum iPTH levels were significantly higher in peritoneal dialysis patients compared to the control group (Table 1). Although the upper and lower thresholds for the definition of high and low iPTH levels in dialysis population is controversial, in our study, we have preferred to stratify our patient group into two subgroups as G3 and G4, 300 pg/mL serum concentration of iPTH taken as a cut-off point [18]. A significant statistical difference between low and high iPTH groups was evident regarding osteoblastic bone formation activity (OC, PINP), osteoclastic bone resorption (Beta CTx), bone turnover (BAP) markers and also 25(OH)D3, Pi, CaXPi and BMD values. Serum concentration of osteocalcin, a marker of osteoblastic bone formation activity, was higher in patient group compared to the control subjects (G1 vs G2) and patients with high PTH levels, (G4 vs G3) suggesting a causative relationship between osteocalcin and CKD BMD rather than sole accumulation due to renal dysfunction (Tables 1, 2, 4, 5). In patients maintaining calcitriol treatment, mean osteocalcin level was also higher compared to the non-users, suggesting that an increased osteoblastic activity may possibly induced by calcitriol.

Among several known metabolic bone markers, serum levels of procollagen I amino-terminal propeptide, bone alkaline phosphatase and TRAP5b were reported not to be affected by renal dysfunction and accumulation due to impaired excretion [19]. Our results have shown a strong correlation among iPTH levels and serum concentrations of BAP and PINP and hence confirmed this opinion with the exception of TRAP5b. Although, mean serum TRAP5b concentration was significantly higher in patients compared to control subjects (Table 1), unlike some former reports in haemodialysis patients, it did not seem to be affected by neither the severity of hyperparathyroidism (Table 5) nor the use of calcitriol (Table 2), questioning the diagnostic value of serum TRAP5b measurements in chronic peritoneal dialysis patients. This may be due to different clearance characteristics between two dialysis modalities or some other reasons which are not known at the moment. Receptor activator of nuclear factor-κB ligand (RANKL) is a protein expressed by osteoblastic cells which has been shown to bind to receptor activator of nuclear factor-κB (RANK) and induces primary mediator of osteoclast differentiation, activation and survival. Therefore, RANKL is responsible for the osteoclast-mediated bone resorption. Osteoprotegerin, a soluble RANKL decoy receptor that binds RANKL, is the key endogenous regulator of the RANKL-RANK pathway and may have a very important role in the pathogenesis of osteoporosis [20]. Glucocorticoid treatment has been shown to decrease the circulating osteoprotegerin levels [21]. Overexpression of osteoprotegerin gene in transgenic mice resulted in osteoporosis, while OPG-deficient mice developed severe osteoporosis [22, 23]. Our results show that mean serum osteoprotegerin concentration is significantly higher in chronic peritoneal dialysis patients compared to the control subjects (Table 1) and it did not seem to be affected neither by calcitriol treatment (Table 5) nor the severity of hyperparathyroidism (Table 2). Therefore, in peritoneal dialysis patients, any diagnostic value for osteoprotegerin levels during the course of CKD-MBD may seem questionable. Beta CTx, an osteoclastic bone resorption marker, has been reported to increase during the course of secondary hyperparathyroidism, but also reported to accumulate due to impaired excretion into the urine [19]. Our results have shown that, PD patients had significantly higher serum concentrations of beta CTx compared to normal subjects. Therefore, it was hard to rule out any accumulation caused by decreased excretion rate of beta CTx. However, in our patients with higher mean PTH levels, serum beta CTx concentrations were even higher than patients with lower mean PTH concentrations, reminding some contribution of metabolic events in its increase rather than being caused only by accumulation due to renal insufficiency (Tables 1, 2, 4, 5). On the other hand, no difference was observed regarding their serum beta CTx concentrations in PD patients using or not using calcitriol participated in our study (Table 5). Another important point to be investigated in this study was the predictive roles of Ca and Pi levels over the PTH induced bone changes as evidenced with the changes by the metabolic markers measured in this study. Despite a mean iPTH concentration difference of 558 pg/mL between our high and low PTH subgroups G3 and G4, Ca concentrations remained similar while Pi levels were found to be significantly higher in high PTH group. Therefore, phosphorus concentration may be assumed to be more prominent in predicting bone related metabolic events compared to the calcium levels in peritoneal dialysis patients. Recent findings suggest that FGF-23, a potent phosphatonin secreted by osteocytes and osteoblasts with the ability for supressing active vitamin D synthesis and PTH release from parathyroid cells, requires klotho for its physiologic effects as a co-receptor [24]. During the clinical course of chronic renal failure, as a physiologic response for phosphate retention and vitamin D administration, FGF levels increase progressively. In our study, possibly due to significantly higher serum phosphorus levels in chronic peritoneal dialysis, patients were shown to have significantly higher FGF-23 levels compared to the control subjects. Furthermore, in patients with higher mean iPTH levels, again probably due to their higher serum phosphorus levels, we have found significantly higher FGF-23 levels than the chronic PD patients with lower mean serum concentrations of iPTH, G4 vs G3 respectively (Tables 2, 4–6). It has been suggested that high FGF-23 levels in advanced renal failure remain ineffective in supressing PTH release due to underexpression of klotho/FGF-23 complex in hypertrophic parathyroid cells [25]. Indeed, serum pre-treatment FGF-23 levels were found to be lower in haemodialysis patients with good response to IV calcitriol administration indicating that pre-treatment serum FGF-23 levels might have a important value in predicting the clinical response for calcitriol treatment [26]. Several studies demonstrated that intravenous calcitriol therapy may increase serum concentrations of FGF-23 in haemodialysis patients with secondary hyperparathyroidism [27]. In our chronic peritoneal dialysis patients receiving oral calcitriol treatment, FGF-23 levels were also significantly higher compared to the non-users suggesting a calcitriol induced increase in FGF-23 synthesis, supporting the findings of those studies above.

As we have stated before, FGF-23 requires klotho as a co-receptor for its physiologic actions. Klotho gene polymorphism has been shown to be associated with the severity of renal failure in CKD patients [6]. Because of its anti-atherosclerotic properties and critical role in PTH and phosphorus metabolism, along with its cofactor FGF-23, mutations in klotho gene may potentially have some important impact on the course and the concentrations of major parameters of CKD-MBD and possibly on bone mineral density values [28]. It is suggested that in uremic rats with seconder hyperparathyroidism, along with FGF-23 gene, klotho gene expression is increased compared to non-uremic rats [29]. In our study group of 51 chronic peritoneal dialysis patients and 40 healthy control subjects, we have detected only one full klotho gene mutation which was not enough for any evaluation based on our findings. Heterogeneous genes consisted of about ¼ of both groups and distribution between patients and control groups was not statistically different. Patients representing higher serum iPTH levels (G4; 718±416 pg/mL) and as low as near adynamic bone disease levels (G3; 160±86 pg/mL) seemed to be similar regarding the distribution of heterogeneous klotho genes (Tables 2, 4–6).

On the other hand, mean iPTH level was found to be lower in patients not using calcitriol treatment (Table 4). This can be explained by the selection of high PTH patients as candidates for oral calcitriol treatment and higher incidence of adynamic bone disease in non-users. On the other hand, both 1,25(OH) D3 and 25(OH)D3 levels were significantly lower in patient group compared to the control subjects. Out of 51 patients, only 2 patients for 1.25(OH)D3 (>22 pg/mL) and 3 for 25(OH) D3 (>15 ng/mL) were not deficient for vitamin D, indicating deficiency rates of 96% and 94% respectively. Therefore, almost all patients participated in the study, seemed to be calcitriol deficient, including 23 patients maintaining oral calcitriol replacement therapy at the study day (Table 1). That was an interesting finding because in our peritoneal dialysis outpatient clinic, indication for routine use and dose adjustment of calcitriol is titrated depending on the results of regular (1–3 months depending on the clinical need of patients) iPTH concentration measurements as indicated by KDIGO guidelines [17]. In haemodialysis patients, similarly, it has been demonstrated that some patients may remain vitamin D deficient despite even IV vitamin D replacement therapy [30]. Similar findings have been published for peritoneal dialysis patients remaining in calcitriol and/or calcidiol deficient despite replacement therapy [31]. These findings may be due to increased rate of consumption or insufficient dosage but we believe that further studies certainly are needed to clarify the underlying mechanism. As indicated in Table 5, none of the bone metabolism markers other than iPTH and FGF-23 have shown any correlation with calcitriol treatment in our study population.

One of the major aims of this study was to investigate the possible effects of renal osteodystrophy on bone mineral density through evaluating the levels of bone metabolic markers in peritoneal dialysis patients. The diagnostic value of bone mineral density measurements in the evaluation of bone mineral deficiency during the course of CKD-MBD is not well established [11, 14]. In our study, when measured as whole body DEXA technique, which has been reported to be more precise with a higher reproducibility compared to single site measurements, we have found a negative correlation between iPTH levels and bone mineral density measurements in the patient group [32]. Additionally, when patients stratified based on their T-scores, osteoporotics had significantly higher FGF-23 levels compared to the patients with normal measurements of BMD and osteopenia (p=0.001 and p=0.043 respectively) suggesting a strong negative correlation between FGF-23 levels and whole body BMD in patients treated with chronic peritoneal dialysis (Table 4). When used whole-body technique, the statistical difference regarding bone mineral density between high and low iPTH groups was evident in the whole body scale, and also whole body measurement was expressed in various body parts as shown in Table 6. When bone mineral density was measured specifically at the femoral site (Table 7), although we have observed a similar correlation between iPTH levels and BMD, measurements over different areas of femur were inconsistent. Strongest statistical significance was achieved in the whole body measurement and it was not surpassed by any measurement performed at femoral site. Therefore, our results suggest that whole-body BMD measurement technique, because of its higher reproducibility and accuracy, may be a valuable choice in the bone mineral density measurements in chronic peritoneal dialysis patients [32]. However, considering the unavailability of any bone histomorphic evaluation in our study, it is hard to tell whether our results correlate well with clinical and histologic osteoporosis or not. On the other hand, several factors have been reported to interfere with BMD measurements carried out by DEXA technique. In obese people, some degree of over measurement error has been observed and the magnitude of that error was shown to rise with incremental increases in the body fat in the examination site. Erroneously high measurements due to aortic calcifications especially in antero-posterior projections is another known problem in lumbar spine DEXA [33, 34]. Same argument may be true for measurement errors caused by tissue calcifications which may be expected to increase during the course of chronic renal failure. As shown in Table 2, calculated CaXPi was as high as 51±10 mg²/dL² in G4 although the mean value was still below the recommended limit of 55 mg²/dL². However, in the same group, BMD was significantly lower than our low PTH group which excludes any major measurement error caused by increased CaXPi. On the other hand, it is still hard to exclude some positive measurement error in low iPTH group caused by tissue calcifications because of the higher expected prevalence of adynamic bone disease in those patients with lower iPTH levels which is a well-known risk factor for increased metastatic calcifications.

In conclusion, based on our findings derived from this study, bone mineral density significantly decreases in chronic PD patients, especially in those with iPTH levels exceeding 300 pg/mL and higher serum FGF-23 levels. Therefore those patients may require special attention regarding their bone health. In chronic peritoneal dialysis patients, vitamin D deficiency expressed by subnormal serum levels of 1,25(OH)D3 and 25(OH)D3 is very common and almost a rule. A strong positive correlation between iPTH levels and BAP and PINP levels may suggest a diagnostic value for those markers during the management of CKD-MBD. iPTH and FGF-23 levels have shown a significant negative correlation with whole-body BMD values suggesting a diagnostic value for whole body BMD measurement technique in chronic peritoneal dialysis patients. We were unable to show any correlation between klotho gene polymorphism and any of the CKD-MBD parameters measured in this study.