Relationships between blood bone metabolic biomarkers and anemia in patients with chronic kidney disease

Abstract

Introduction

Blood bone metabolic biomarkers are noninvasive indices for evaluating metabolic bone diseases. We investigated the relationships between blood bone metabolic biomarkers and anemia in chronic kidney disease (CKD) patients and analyzed the effects of parathyroidectomy (PTX) on the above indices.

Methods

In this cross-sectional study, 100 healthy controls and 239 CKD patients, including 46 secondary hyperparathyroidism (SHPT) patients with PTX, were enrolled. Moreover, a prospective study was conducted in which 28 PTX patients were followed up. The degree of anemia was classified as mild, moderate, or severe based on the tertiles of hemoglobin (Hb) levels of the anemic CKD patients, with cutoff values of 83 g/L and 102 g/L. Bone metabolic biomarkers, including calcium (Ca), phosphorus (P), intact parathyroid hormone (iPTH), fibroblast growth factor 23 (FGF23), and α-klotho, were tested.

Results

The mean estimated glomerular filtration rate (eGFR) in CKD patients was 25.7 ± 36.0 ml/min/1.73 m², and 84.10% of CKD patients had anemia. The baseline Hb levels in the mild, moderate, and severe anemia subgroups were 110.86 ± 5.99 g/L, 92.71 ± 5.96 g/L, and 67.38 ± 10.56 g/L, respectively. CKD patients had higher adjusted Ca, P, alkaline phosphatase (ALP), iPTH, and FGF23 levels and lower α-klotho levels than controls. Baseline adjusted Ca, P, iPTH, and α-klotho levels were associated with Hb levels in CKD patients. Blood adjusted Ca, P, and iPTH levels were correlated with anemia severity. After PTX (median interval: 6.88 months), anemia and high blood adjusted Ca, P, iPTH, and FGF23 levels were ameliorated, while α-klotho levels were increased.

Conclusions

Blood adjusted Ca, P, iPTH, and α-klotho levels were correlated with Hb levels in CKD patients. Correction of bone metabolic disorders may be a therapeutic strategy for anemia treatment.

Keywords:

• Anemia
• chronic kidney disease–mineral and bone disorder
• bone metabolic biomarkers
• parathyroidectomy
• secondary hyperparathyroidism

Introduction

Renal anemia is a common complication in patients with chronic kidney disease (CKD). The incidence of anemia in non-dialysis patients exceeds 50% [1,2]. Anemia is associated not only with increased morbidity rates in patients with cardiovascular disease, but also with an increased risk of hospitalization and mortality [1]. Traditional causes of renal anemia include relative erythropoietin (EPO) deficiency, iron deficiency, and nutritional deficiency [3]. Correction of anemia is an important therapeutic strategy to improve the survival rate and quality of life for patients with CKD. Iron and parenteral erythropoiesis-stimulating agents (ESAs) are the cornerstones of anemia management [4]. New therapeutic approaches, including hypoxia-inducible factor prolyl hydroxylase inhibitors (HIF-PHIs), attracts more attention for anemia correction in recent years [4,5].

Chronic kidney disease-mineral and bone disorder (CKD-MBD) is a common complication and an important cause of death in patients with CKD, characterized by dysregulation of mineral and bone metabolism, bone abnormalities, and vascular calcification [6,7]. Bone biopsy is the golden standard in the assessment of metabolic bone diseases, but it is invasive. Therefore, bone-related biomarkers have been used clinically to determine the bone turnover status in CKD and hemodialysis patients [8]. As the common manifestation of CKD-MBD, secondary hyperparathyroidism (SHPT) contributes to increased bone turnover, risk of fractures, vascular calcification, cardiovascular and all-cause mortality [9]. Parathyroidectomy (PTX) is effective in treating severe SHPT patients [10].

We have demonstrated that stage 5 CKD patients with severe SHPT have more significant anemia, which is remarkably improved after PTX [11]. Cross-sectional studies suggest that there may be a link between biomarkers of mineral metabolism and hemoglobin (Hb) levels [12–14]. However, the correlations between systemic bone biomarkers and anemia in CKD patients remain unclear.

Serum calcium (Ca) and phosphorus (P) are related to bone metabolism. Alkaline phosphatase (ALP) plays an important role in skeletal mineralization and is the most widely recognized biochemical marker for osteoblast activity [15]. Intact parathyroid hormone (iPTH) is a major mediator of bone remodeling and an essential regulator of Ca and P homeostasis [16]. Fibroblast growth factor 23 (FGF23), a hormone secreted by osteocytes and osteoblasts, is a potent regulator of vitamin D metabolism and phosphate homeostasis [13]. Finally, α-klotho is a membrane protein that is highly expressed in the kidney, especially in the distal tubular epithelial cells [17]. Patients with advanced CKD have a significant reduction in α-klotho levels and progressively lose the ability to prevent phosphate retention [18].

The aim of this study is to investigate the relationships between bloodbone metabolic biomarkers and anemia in CKD patients and analyze the effects of PTX on the above indices. This research provides new insights into the pathogenesis and potential therapeutic targets for anemia in CKD patients from the perspective of CKD-MBD.

Materials and methods

Study population

In this study, 239 CKD patients were enrolled from October 2011 to October 2018 from the Department of Nephrology, the First Affiliated Hospital of Nanjing Medical University. A total of 100 healthy age- and sex-matched volunteers were recruited as controls. Among the 239 CKD patients, 46 of all had severe SHPT, including 44 cases receiving hemodialysis and 2 cases receiving peritoneal dialysis. They underwent total PTX with forearm auto-transplantation. The PTX patients were derived from a wide geographical area. Some of them dropped out after the operation because of transference to other dialysis units, inability to contact, or poor compliance. A total of 28 PTX patients were followed up with a median interval of 6.88 months. This research included a cross-sectional study enrolled normal controls, CKD patients with or without severe SHPT, and a prospective follow-up study of 28 PTX patients. A study flow diagram is shown in Figure 1.

Figure 1. Flow diagram of the study.

All patients and controls provided written informed consent. The study protocol was approved by the Research Ethics Committee of the First Affiliated Hospital of Nanjing Medical University, Nanjing, China (2015-SR-146).

Inclusion criteria

The inclusion criteria of the patients included: (1) aged between 18–75 years old; (2) signed the written consent; (3) the diagnostic criteria of CKD from ‘KDIGO 2017 Clinical Practice Guideline Update for the Diagnosis, Evaluation, Prevention, and Treatment of Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD) [19]’ were met.

Exclusion criteria

Patients with a history of bleeding or blood transfusion, a history of malignancy, serious infections [20,21], iron strengthening, a lack of folic acid and/or vitamin B12, severe heart failure, or active systemic disease 3 months prior to the observation and/or during follow-up were excluded.

PTX indications

Patients with the following indications were considered for PTX: (1) persistent serum iPTH > 800 pg/mL; (2) hypercalcemia and/or hyperphosphatemia that could not be controlled by medical therapy; (3) obvious clinical manifestations such as bone pain, pruritus, ectopic calcification, and/or fracture; and (4) at least one enlarged parathyroid gland discovered by ultrasound or a radiopharmaceutical technetium-99m-methoxyisobutylisonitrile (99mTc-MIBI) scan. PTX was considered successful when the mean percentage reduction of serum iPTH levels at 20 min after PTX (io-iPTH20%) > 88.9% (sensitivity 78.6%, specificity 88.5%), or serum iPTH levels on the fourth day after PTX (D4-iPTH) ≤ 147.4 pg/mL (sensitivity 100%, specificity 99.5%) [22].

Main observation indicators

During enrollment, basic information was collected, such as demographic information, dialysis mode, medical history, and causes of CKD. In addition, the use of antihypertensive drugs was documented, including angiotensin-converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), β-adrenergic receptor blockers, and dihydropyridine Ca channel blockers (CCB). Information regarding healthy individuals was obtained through questionnaires.

Fasting venous blood was collected in the morning for routine blood tests, blood biochemistry tests, and measurement of bone metabolism-related indices, including iPTH, FGF23, and α-klotho levels. Routine blood tests were performed using an LH-750 Hematology Analyzer (Beckman Coulter, Inc., Fullerton, CA, USA). Biochemical indices were measured with an automatic biochemical analyzer (AU5400, Olympus Corporation, Japan). Serum iPTH was measured using a UniCel DxI 800 Access Immunoassay System (Beckman Coulter, Inc.). The bone metabolism-related biomarkers were measured with the following human ELISA kits: plasma cFGF23 (Immutopics, San Clemente, CA, USA) and human soluble α-klotho (Immuno-Biologic Laboratories Co., Ltd., Japan). For hemodialysis patients, blood samples were collected before dialysis. For PTX patients, these parameters were obtained no more than 2 weeks before surgery. Patients were subsequently followed up after PTX, and examinations of laboratory values and bone metabolism indices were repeated on the date of the last follow-up.

Estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease (MDRD) formula [23]. Serum Ca levels were corrected for serum albumin (Alb): adjusted Ca (mmol/L) = total Ca (mmol/L) + 0.02 × (40 − Alb [g/L]). Anemia was defined based on the KDIGO clinical practice guideline for anemia in CKD when Hb < 13.0 g/dL in men and Hb < 12.0 g/dL in women [3]. In this study, the degree of anemia was classified as mild, moderate, or severe based on the tertiles of Hb levels of the anemic CKD patients, with cutoff values for the tertiles of 83 g/L and 102 g/L.

Statistical analysis

Baseline characteristics of CKD patients were compared with controls using t-test (normally distributed continuous variables) or Mann-Whitney U tests (skewed continuous variables) for continuous variables and chi-square tests for categorical variables. Categorical variables were presented as numbers (proportions), and the continuous variables were presented as mean ± standard deviations (SD) or median (interquartile range). The Pearson correlation coefficient was used to analyze the correlation between various variables.

To adjust for confounding factors, a single-variable regression analysis was first conducted to select the statistically significant variables (p < 0.05) as independent variables. Subsequently, stepwise multivariate linear regression analysis was conducted to determine the factors independently associated with Hb levels, with adjustments for age, sex, body mass index (BMI), systolic blood pressure (SBP), and diastolic blood pressure (DBP), and medications, including CCB, ACEI/ARB, β-adrenergic receptor blocker, calcimimetics, and active vitamin D sterols, and clinical biochemistry indicators, including glucose (Glu), Alb, total cholesterol (TC), and triglycerides (TG).

The degree of anemia was classified as mild, moderate, and severe based on the tertiles of Hb levels of the anemic CKD patients, with cutoff values of 83 g/L and 102 g/L. For the comparison of bone metabolism indices among the four subgroups, a one-way ANOVA was performed. For the ordered dependent variables, ordered logistic regression analysis was performed to explore the effects of factors influencing the severity of anemia among CKD patients. A paired sample t-test was used to assess the differences between the values recorded before and after the PTX. We investigated the correlations between the changes of bone metabolism indices and Hb levels in PTX patients before and after operation by univariate linear regression analysis. p < 0.05 was considered statistically significant. All the statistical analyses were performed using the statistical package for the social sciences (SPSS), version 26.0 (SPSS Inc., Chicago, IL, USA).

Results

Characteristics of the study population

A total of 239 CKD patients (130 males and 109 females) and 100 age- and sex-matched healthy controls were enrolled. The mean age and eGFR in 239 CKD patients were 49.25 ± 13.22 years old and 25.7 ± 36.0 ml/min/1.73 m², respectively. The most common cause of CKD was chronic glomerulonephritis (77.41%). Among the CKD patients, 58.58% did not receive dialysis, 32.22% underwent hemodialysis, and 9.21% underwent peritoneal dialysis. The dialysis vintage was 60.00 (12.00–84.00) months. Of 99 dialysis patients, hemodialysis was the main dialysis mode. In our research, 30.54% patients used ACEI/ARB, 30.96% patients used active vitamin D sterols, 52.72% patients took CCB, and only 3.35% patients were treated with calcimimetics.

Compared with controls, CKD patients had lower Hb and Alb levels, lower hematocrit (Hct) values, higher TG levels, and higher SBP and DBP. Anemia was observed in most CKD patients (84.10%). The baseline Hb level was 144.40 ± 15.52 g/L in controls and 98.76 ± 25.62 g/L in CKD patients.

Meanwhile, mineral and bone metabolism disorders were observed in CKD patients, with higher P, ALP, iPTH, and FGF23 levels compared to controls. In addition, circulating Ca and α-klotho levels were significantly lower than in controls. To eliminate spurious hypocalcemia, the serum Ca concentration was adjusted for serum Alb. Adjusted Ca levels were 2.20 ± 0.11 mmol/L in controls and obviously increased in CKD patients at 2.31 ± 0.27 mmol/L. For a detailed comparison of controls and CKD patients, please see Table 1.

Table 1. Demographic data and laboratory values of controls and CKD patients.

Variables	Controls (n = 100)	CKD patients (n = 239)	p Value^a	CKD patients
				No anemia (n = 38)	Mild anemia (n = 70)	Moderate anemia (n = 68)	Severe anemia (n = 63)	p Value^b
Demographics
Age (years)	48.97 ± 13.54	49.25 ± 13.22	0.859	45.24 ± 13.04	51.07 ± 12.31	49.91 ± 12.90	48.94 ± 14.39	0.169
Male/female	51/49	130/109	0.568	22/16	41/29	34/34	33/30	0.726
Body mass index (kg/m^2)	23.56 ± 2.75	23.19 ± 3.63	0.314	24.60 ± 3.79	23.84 ± 3.60	22.47 ± 3.55	22.41 ± 3.32	0.003
Systolic pressure (mmHg)	122.63 ± 16.36	146.00 ± 25.92	<0.001	132.24 ± 16.92	144.96 ± 25.16	149.87 ± 22.44	153.51 ± 30.99	<0.001
Diastolic pressure (mmHg)	77.88 ± 10.91	86.83 ± 13.94	<0.001	85.71 ± 12.24	85.79 ± 14.81	86.85 ± 11.73	88.65 ± 16.09	0.637
Cause of CKD, n (%)
Glomerulonephritis	/	185(77.41)	/	33(86.84)	51(72.86)	52(76.47)	49(77.78)	0.423
Diabetic nephropathy	/	24(10.04)	/	2(5.26)	9(12.86)	7(10.29)	6(9.52)	0.629
Hypertensive nephropathy	/	9(3.77)	/	1(2.63)	3(4.29)	3(4.41)	2(3.17)	0.967
Polycystic kidney disease	/	9(3.77)	/	0(0)	2(2.86)	3(4.41)	4(6.35)	/
Other	/	12(5.02)	/	2(5.26)	5(7.14)	3(4.41)	2(3.17)	0.419
Dialysis mode, n (%)
Non-dialysis	/	140(58.58)	/	34(89.47)	45(64.29)	29(42.65)	32(50.79)	<0.001
Hemodialysis	/	77(32.22)	/	4(10.53)	19(27.14)	32(47.06)	34(47.06)	0.001
Peritoneal dialysis	/	22(9.21)	/	0(0)	6(8.57)	7(10.29)	9(14.29)	0.027
Dialysis vintage(m)	/	60.00(12.00–84.00)	/	102.00(30.00–168.00)	60.00(17.00–108.50)	60.00(24.00–96.00)	36.00(5.00–72.00)	0.066
Medication history, n (%)
CCB	/	126(52.72)	/	13(34.21)	34(48.57)	42(61.76)	37(58.73)	0.031
ACEI/ARB	/	73(30.54)	/	16(42.11)	22(31.43)	22(32.35)	13(20.63)	0.142
β-receptor blocker	/	53(22.18)	/	5(13.16)	16(22.86)	17(25.00)	15(23.81)	0.528
Active vitamin D sterols	/	74(30.96)	/	4(10.53)	18(25.71)	33(48.53)	19(30.16)	<0.001
Calcimimetics	/	8(3.35)	/	1(2.63)	4(5.71)	3(4.41)	0(0.00)	0.264
Laboratory values
Hemoglobin (g/L)	144.40 ± 15.52	98.76 ± 25.62	<0.001	139.39 ± 13.63	110.86 ± 5.99	92.71 ± 5.96	67.38 ± 10.56	<0.001
Hematocrit (%)	43.26 ± 4.24	30.19 ± 7.66	<0.001	41.93 ± 3.98	33.96 ± 2.12	28.41 ± 2.18	20.83 ± 3.56	<0.001
Glucose (mmol/L)	5.41 ± 0.77	5.28 ± 2.33	0.442	5.58 ± 2.82	5.64 ± 2.92	4.87 ± 1.30	5.14 ± 2.10	0.202
Albumin (g/L)	47.76 ± 2.87	35.74 ± 6.63	<0.001	37.14 ± 7.59	36.36 ± 6.92	34.57 ± 7.13	35.50 ± 4.78	0.212
Creatinine (µmol/L)	71.75 ± 15.61	625.65 ± 462.33	<0.001	206.51 ± 257.33	471.02 ± 400.79	705.20 ± 375.35	964.40 ± 437.92	<0.001
Urea (mmol/L)	5.56 ± 1.42	20.16 ± 11.90	<0.001	9.14 ± 7.15	15.68 ± 8.07	21.39 ± 8.55	30.45 ± 12.47	<0.001
Total cholesterol (mmol/L)	5.01 ± 0.83	4.98 ± 1.89	0.816	5.92 ± 2.90	5.33 ± 1.77	4.76 ± 1.51	4.27 ± 1.25	<0.001
Triglyceride (mmol/L)	1.44 ± 1.45	1.92 ± 1.33	0.003	2.02 ± 1.02	2.22 ± 1.49	1.90 ± 1.39	1.53 ± 1.16	0.024
eGFR (ml/min/1.73 m^2)	97.8 ± 18.3	25.7 ± 36.0	<0.001	65.5 ± 45.1	34.7 ± 40.5	12.5 ± 15.6	6.2 ± 4.1	<0.001
Calcium (mmol/L)	2.36 ± 0.11	2.23 ± 0.28	<0.001	2.28 ± 0.17	2.25 ± 0.30	2.26 ± 0.27	2.13 ± 0.32	0.021
Adjusted Ca (mmol/L)	2.20 ± 0.11	2.31 ± 0.27	<0.001	2.34 ± 0.15	2.32 ± 0.26	2.37 ± 0.23	2.22 ± 0.31	0.013
Phosphorus (mmol/L)	1.17 ± 0.15	1.72 ± 0.60	<0.001	1.26 ± 0.38	1.55 ± 0.48	1.86 ± 0.54	2.04 ± 0.66	<0.001
lnALP	4.31 ± 0.26	4.63 ± 0.81	<0.001	4.40 ± 0.69	4.56 ± 0.72	4.81 ± 0.94	4.64 ± 0.80	0.076
lniPTH	3.58 ± 0.39	5.18 ± 1.61	<0.001	4.04 ± 1.36	4.59 ± 1.58	5.72 ± 1.56	5.94 ± 1.14	<0.001
lnFGF23	4.15 ± 0.31	7.29 ± 2.36	<0.001	5.29 ± 1.72	6.79 ± 2.29	8.20 ± 2.40	8.07 ± 1.82	<0.001
lnKlotho	6.50 ± 0.27	5.82 ± 0.49	<0.001	6.05 ± 0.44	5.78 ± 0.50	5.80 ± 0.52	5.74 ± 0.43	0.012

Data were presented as means ± SD or median (interquartile range) for continuous variables and as n (%) for categorical variables. Bold values were statistically significant at p < 0.05.

Abbreviations: CKD: chronic kidney disease; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; CCB: dihydropyridine Ca channel blocker; ACEI/ARB: angiotensin converting enzyme inhibitors/angiotensin receptor blocker; Ca: calcium; TC: total cholesterol; TG: triglyceride; eGFR: estimated glomerular filtration rate; ALP: alkaline phosphatase; iPTH: intact parathyroid hormone; FGF23: fibroblast growth factor 23.

^aControls vs. CKD patients.

^bAnalysis of variance of 4 subgroups in CKD patients.

No correlation was observed between adjusted Ca and P levels in CKD patients (r = −0.013, p = 0.840). In 239 patients with different stages of renal function, serum-adjusted Ca levels presented little fluctuations. When eGFR was less than 60 ml/min/1.73 m², serum P levels showed an obvious upward trend, please see Supplementary Figure 1.

Correlations between blood bone biomarkers and anemia in CKD patients

The CKD patients were divided into four subgroups according to the Hb levels: no anemia (male: Hb ≥ 130 g/L, female: Hb ≥ 120 g/L, n = 38), mild anemia (male: 102 ≤ Hb < 130 g/L, female: 102 ≤ Hb < 120 g/L, n = 70), moderate anemia (83 ≤ Hb < 102 g/L, n = 68), and severe anemia (30 ≤ Hb < 83 g/L, n = 63). The baseline Hb levels in the no anemia and mild, moderate, and severe anemia subgroups were 139.39 ± 13.63 g/L, 110.86 ± 5.99 g/L, 92.71 ± 5.96 g/L, and 67.38 ± 10.56g/L, respectively. Patients with more severe anemia exhibited higher levels of P, lnALP, lniPTH, and lnFGF23 levels. Contrarily, serum-adjusted Ca and α-klotho levels showed a decreasing trend. Lower BMI and Alb levels and higher SBP were observed in CKD patients with more severe anemia (Table 1, Figure 2).

Figure 2. Baseline blood bone metabolism indices in different subgroups of CKD patients. The CKD patients were divided into four subgroups according to the tertiles of hemoglobin levels: 1: no anemia(male: Hb ≥ 130g/L; female: Hb ≥ 120g/L, n = 38), 2: mild anemia(male: 102 ≤ Hb < 130g/L; female: 102 ≤ Hb < 120g/L, n = 70), 3: moderate anemia(83 ≤ Hb < 102g/L, n = 68), 4: severe anemia(30 ≤ Hb < 83g/L, n = 63). P: analysis of variance of 4 subgroups in CKD patients. Abbreviations: Ca: calcium; P: phosphorus; ALP: alkaline phosphatase; iPTH: intact parathyroid hormone; FGF23: fibroblast growth factor 23.

Correlations between blood bone metabolism indices and Hb in 239 CKD patients are illustrated in Figure 3. At baseline, Hb levels showed significant negative correlations with lniPTH, lnFGF23 and P levels (p < 0.001). Meanwhile, a positive correlation between ln(α-klotho) and Hb was revealed (r = 0.171, p = 0.010). Hb was shown to be correlated with serum-adjusted Ca levels (r = 0.130, p = 0.044). Furthermore, sex, BMI, SBP, DBP, TC, and TG were correlated with Hb (data not shown). No correlation was found between lnALP and Hb levels.

Figure 3. Correlations between blood bone metabolism indices and Hb levels in CKD patients. Abbreviations: Hb: hemoglobin; Ca: calcium; P: phosphorus; ALP: alkaline phosphatase; iPTH: intact parathyroid hormone; FGF23: fibroblast growth factor 23.

EGFR also presented a correlation with Hb levels in CKD patients (r = 0.559, p < 0.001). To access the potential confounding effect of eGFR, linear regression analyses between each blood bone metabolism index and Hb levels were established in stages 1–4 and non-dialysis stage 5 CKD patients (n = 140). Please see Supplementary Table 1.

Independent influencing factors of Hb levels and anemia severity in CKD patients

To eliminate the influence of potential confounding factors, a multivariable stepwise linear regression analysis with Hb as the dependent variable was conducted. Assessments of covariates were described in Supplementary Table 2. In CKD patients, the Hb levels were independently inversely correlated with blood P levels (β = −0.264, p < 0.001) and lniPTH (β = −0.157, p = 0.026). Serum-adjusted Ca and ln(α-klotho) showed independent positive associations with Hb levels. There were no correlations between circulating FGF23 and Hb levels. Sex, BMI, SBP, Alb and TC were independently correlated with Hb in CKD patients (Table 2).

Table 2. Influencing factors of Hb levels in CKD patients analyzed by multiple linear stepwise regression.

Independent Variable	Unstandardized Coefficient	Standardized Coefficient	95% CI		p Value
			Lower	Upper
Female	−11.690	−0.224	−17.117	−6.263	<0.001
BMI	0.955	0.130	0.200	1.710	0.013
SBP	−0.170	−0.170	−0.275	−0.064	0.002
Alb	1.400	0.355	0.957	1.844	<0.001
TC	3.990	0.296	2.423	5.558	<0.001
Adjusted Ca	12.903	0.134	3.019	22.786	0.011
P	−11.505	−0.264	−17.216	−5.794	<0.001
lniPTH	−2.574	−0.157	−4.840	−0.307	0.026
lnklotho	5.622	0.105	0.151	11.093	0.044

Multiple stepwise regression model with adjustment for age, sex, BMI, SBP, DBP, and Glu, Alb, TC, TG, adjusted Ca, P, lniPTH, lnFGF23, lnklotho and CCB, ACEI/ARB, β-receptor blocker, calcimimetics and active vitamin D sterols.

Abbreviations: Hb: hemoglobin; CKD: chronic kidney disease; BMI: body mass index; SBP: systolic blood pressure; Alb: albumin; TC: total cholesterol; Ca: calcium; P: phosphorus; iPTH: intact parathyroid hormone; CI: confidence interval.

These findings were further strengthened in a multivariate ordered logistic regression analysis. The anemia level was analyzed based on mild, moderate, and severe anemia subgroups. After rigorous adjustment of confounders, including age, sex, BMI, SBP, DBP, medication history, and biochemical indicators, the following significant factors influencing the severity of anemia in CKD patients were identified: serum adjusted Ca (odds ratio [OR] = 0.216, 95% confidence interval [CI] = 0.068–0.684; p = 0.009), P (OR = 3.019, 95% CI = 1.621–5.624; p = 0.001), and lniPTH (OR = 1.366, 95% CI = 1.005–1.857; p = 0.046). Similarly, sex, SBP, Alb, and TC were also independent factors affecting the severity of anemia (Figure 4).

Figure 4. Multiple ordered logistic regression analysis of factors associated with anemia in CKD patients. Abbreviations: BMI: body mass index; SBP: systolic blood pressure; Alb: albumin; TC: total cholesterol; Ca: calcium; P: phosphorus; iPTH: intact parathyroid hormone; OR: odds ratio; CI: confidence interval.

Effects of PTX on laboratory indicators and bone metabolism indices in patients with severe SHPT

CKD patients were categorized into stages 1-5 based on eGFR, with 17, 21, 21, 26, and 154 patients, respectively. In the stage 5 CKD subgroup, 46 patients with severe SHPT underwent PTX; these patients were younger and hemodialysis mode was more prevalent than among stage 5 CKD patients who did not undergo PTX. Among the 46 PTX patients in the stage 5 CKD subgroup, 2 patients had no anemia, 9 patients had mild anemia, 24 patients had moderate anemia, and 11 patients had severe anemia.

A total of 28 PTX patients were followed up (median time: 6.88 months). At baseline, the average Hb level of 46 PTX patients was 93.93 g/L, and the average Hb level of 28 follow-up patients was 96.68 g/L (p > 0.05). Although the duration of follow-up differed in PTX patients, blood–bone metabolism disorders were alleviated, and no time trends were observed (data not shown). After PTX, anemia, hypoalbuminemia and bone metabolism disorders were improved significantly. Compared with baseline data, the levels of adjusted Ca, P, iPTH and FGF23 were decreased, while the levels of α-klotho were increased in the patients after PTX (Figure 5). Before and after PTX, the levels of lnFGF23 were 10.85 ± 1.18 and 8.29 ± 1.22, respectively, and the levels of ln(α-klotho) were 5.84 ± 0.36 and 6.09 ± 0.41, respectively. We further investigated the correlations between the longitudinal improvements of blood bone biomarkers and Hb levels in PTX patients, however, no correlation was found (Table 3).

Figure 5. Comparisons of pre- and post-operative laboratory values in severe SHPT patients with PTX. Circulating levels of Hb (A), Hct (B), Alb (C), adjusted Ca (D), P (E), lnALP (F), lniPTH(G), lnFGF23(H), and lnklotho(I) (pre-PTX versus post-PTX), Data were expressed as 95% CIs. Abbreviations: PTX: parathyroidectomy; SHPT: secondary hyperparathyroidism; Hb: hemoglobin; Hct: hematocrit; Alb: albumin; Ca: calcium; P, phosphorus; ALP: alkaline phosphatase; iPTH: intact parathyroid hormone; FGF23: fibroblast growth factor 23.

Table 3. Associations between the changes of bone metabolism indices and Hb levels in PTX patients before and after operation analyzed by univariate linear regression.

Independent Variable	Unstandardized Coefficient	Standardized Coefficient	95% CI		p Value
			Lower	Upper
ΔAdjusted Ca	7.992	0.166	−11.194	27.177	0.400
ΔP	1.134	0.048	−8.282	10.549	0.806
lnΔALP	1.000	0.074	−4.846	6.846	0.727
lnΔiPTH	−5.908	−0.199	−20.791	8.974	0.414
lnΔFGF23	0.074	0.005	−5.736	5.883	0.979
ln(|Δklotho|)	−2.425	−0.123	−12.44	7.589	0.616

Abbreviations: Hb: hemoglobin; Ca, calcium; P: phosphorus; ALP: alkaline phosphatase; iPTH: intact parathyroid hormone; FGF23: fibroblast growth factor 23; Δ: subtracting the values of bone markers after PTX from baseline values; |Δ|: absolute value.

Discussion

The present investigation was designed (1) to analyze the relationships between blood bone biomarkers and anemia in CKD patients and (2) to assess how PTX affected these variables in the severe SHPT subgroup. The inflammatory state has been confirmed to influence the development of renal anemia [24]; so, we excluded the CKD patients with obvious inflammation at enrollment. Hypoalbuminemia is recognized as a strong predictor of mortality in the CKD population [25]. In our research, Alb was an independent influencing factor of Hb levels in CKD patients, and Hb was also affected by sex, BMI, blood pressure, and blood lipids [26,27].

We showed that both serum Ca and P were independently associated with Hb levels in CKD patients. Mauro et al. [27] reported that a decrease in serum Ca and an increase of serum P levels were closely related to anemia in patients with advanced non-dialysis CKD. Ca was required for the in vitro proliferation and differentiation of erythroid progenitor cells induced by EPO [28]. As an EPO-activated pathway, store-operated Ca channel (SOC) related genetic polymorphisms have been proven to be correlated with the risk of EPO resistance in dialysis patients [29].

Correlations between hyperphosphatemia and anemia have also been reported [14,30]. EPO deficiency, inflammation, and oxidative stress have been implicated as potential factors linking hyperphosphatemia and anemia. Hyperphosphatemia may increase the synthesis of polyamines, which can act as toxins inhibiting erythropoiesis [30]. Hyperphosphatemia in CKD can decrease vitamin D synthesis, resulting in hypocalcemia and elevated PTH levels, which further inhibit erythropoiesis [31,32]. FGF-23 and α-klotho may play important roles in the link between hyperphosphatemia and anemia [14,32]. High serum P can increase FGF-23 levels, which in turn suppresses the activity of α-klotho and aggravates vitamin D deficiency [32,33]. Klotho deficiency increases P reabsorption in the kidneys, which results in hyperphosphatemia [34]. Low levels of α-klotho and vitamin D are associated with an increased risk of anemia [35,36].

A positive association between serum ALP levels and ESA-resistant anemia has been demonstrated in a study involving 38,328 end-stage kidney disease (ESKD) patients receiving hemodialysis [15]. Serum ALP levels are associated with the severity of resistance to ESA in CKD patients [37]. ALP is associated with mortality in ESKD patients receiving dialysis [38] and pre-dialysis patients with CKD [39,40]. However, in this study, no significant correlation was identified between ALP and Hb levels in CKD patients, which may be related to the lower number of cases and the lack of information on the use of ESA.

Increased iPTH levels can inhibit EPO synthesis, shorten the erythrocyte survival time, and lead to bone marrow fibrosis, thus reducing hematopoietic function [41]. However, the molecular mechanism underlying the inhibitory effects of high PTH levels on the synthesis of endogenous EPO and myeloid progenitor cells remains unclear. In our study, baseline iPTH levels were independently associated with Hb levels and severity of anemia in CKD patients. SHPT is a common complication of advanced CKD and a recognized cause of ESA resistance. Kamalas et al. [42] conducted a retrospective study on 43 anemic CKD patients who received ESA treatment and found that in patients with normal bone marrow examination, higher iPTH levels were associated with lower Hb levels, and these patients required higher doses of ESA, suggesting that the low reactivity of ESA in CKD patients is related to high levels of iPTH.

We confirmed that blood FGF23 levels were negatively correlated with the Hb levels in CKD patients. A recent report from the Chronic Renal Insufficiency Cohort (CRIC) study showed that increased FGF23 levels were significantly correlated with decreased Hb levels and overall rates of anemia [13]. Moreover, elevated serum FGF23 levels in CKD patients are negatively correlated with transferrin saturation, serum iron levels, and erythropoiesis [43,44]. Coe et al. [45] observed a significant increase in the number of hematopoietic stem cells (HSCs) and erythroid progenitor cells in the bone marrow and peripheral blood in FGF23 deficient mice. Elevated serum FGF23 levels can stimulate the synthesis and secretion of various pro-inflammatory cytokines (such as IL-6 and IL-1β). Cytokines upregulate the production of ferritin in the liver, increasing the serum ferritin level and leading to functional iron deficiency [46].

The anti-aging gene α-klotho is an important marker of CKD, and its corresponding protein is mainly expressed in the kidney [17]. Previous research revealed that Hct was significantly decreased in klotho gene mutant (KL^-/-) mice, indicating that klotho deficiency causes anemia by suppressing the expression of S-formylglutathione hydrolase (FGH) ^[47]. Klotho gene deficiency in kidneys leads to decreased plasma Klotho levels, which suppresses FGH expression in red blood cells (RBCs). FGH is a key enzyme in the generation of the major cellular antioxidant glutathione (GSH), which is involved in the maintenance of normal homeostasis of RBCs and kidneys. A decrease in GSH levels could damage RBCs and lead to anemia [47–49]. Moreover, glomerular collapse and interstitial fibrosis were observed in KL^-/- mice, indicating that haplodeficiency of the klotho gene (K) impaired kidney function, which in turn contributed to klotho deficiency-related anemia. In addition, full-length α-klotho is reported to have an inhibitory effect on oxidative stress, which may help to alleviate anemia [30,34]. Sangeetha et al. [50] demonstrated that loss of klotho disrupted erythropoiesis and HSCs development. These hematopoietic changes in bone marrow could be caused by the dramatic reduction in osteoblast and osteoclast numbers and bone mineral density in Klotho^−/− bones, which may be responsible for the reduced expression of hypoxia-inducible factor (HIF) and EPO in bone.

Klotho deficiency is a characteristic feature of CKD; however, the role of α-klotho in renal anemia remains unclear. Our study demonstrated that blood α-klotho levels were positively associated with Hb levels in CKD patients. Treatment of anemia with EPO has been demonstrated to enhance renal and extrarenal production of α-klotho in CKD patients [35]. However, several studies provided contradictory results. Yang Xu et al. [36] suggested that there may be a negative feedback mechanism between EPO and α-klotho, such that increasing α-klotho suppresses the production of EPO in CKD; so, low expression of α-klotho may be a compensatory mechanism to attenuate the effects of anemia. Another study showed that the expression of HIF-1 α and HIF-2 α is significantly upregulated in α-klotho-/- bone tissues, resulting in local overexpression of EPO [51]. The relationship between α-klotho and renal anemia is also affected by iron metabolism and vitamin D, but more research is required to investigate this connection [36]. Although the specific mechanism of α-klotho in renal anemia remains unclear, it is certain that α-klotho can serve as a therapeutic target for renal anemia treatment.

Clinical studies have confirmed the importance of treating SHPT in improving renal anemia in hemodialysis patients [11,52]. Zingraff et al. [53] have demonstrated that successful PTX can significantly improve anemia in SHPT patients, consistent with the conclusions in our previous reports [11,54]. In this study, PTX patients (n = 28) were followed up for at least 3 months after surgery (median follow-up: 6.88 months). Elevated serum P, iPTH, ALP, and FGF23 levels and decreased α-klotho levels were alleviated after PTX; however, we did not find correlations between corrected bone metabolism disorders and improvement of anemia after PTX.

The study has several limitations. First, this is a cross-sectional study, causal effects cannot be conclusively established between the blood bone metabolic biomarkers and anemia in CKD patients. Second, we did not collect the patient’s blood transfusion data, iron metabolism-related indices, or history of taking HIF-PHIs or ESAs. Third, bone histology analysis was not performed. Finally, the sample size of PTX cohort was small and the follow-up time was short.

Conclusions

Anemia and CKD-MBD are both common complications in CKD patients, which are strongly correlated with poor prognosis. Here, we showed that in CKD patients, increased serum P and iPTH levels, and decreased adjusted Ca and α-klotho levels were independently associated with reduced Hb levels. Moreover, adjusted Ca, P, and iPTH levels were independent factors affecting the severity of renal anemia. PTX could correct the anemia and bone metabolic disorders in patients with severe SHPT, but no correlation was found between the increased Hb levels and different corrected bone metabolism biomarkers. This study provides a reference for in-depth analysis of the pathogenesis and therapeutic targets for renal anemia, from the perspective of CKD-MBD.

Acknowledgement

The authors sincerely thank the patients and their families for participation in our study. The authors thank all medical staffs and students involved in the clinical management and analysis of samples for the patients. The study was supported by the International Society of Nephrology (ISN) Mentorship Program and the authors thank Professor Marcello Tonelli (University of Calgary, Canada) for his helpful comments on the draft of the manuscript.

Disclosure statement

The authors declare no potential conflict of interest.

Additional information

Funding

This work was funded by the National Natural Science Foundation of China [81270408, 81570666]; International Society of Nephrology (ISN) Clinical Research Program [1801-0247]; CKD Anemia Research Foundation from China International Medical Foundation [Z-2017-24-2037]; Construction Program of Jiangsu Provincial Clinical Research Center Support System [BL2014084]; Chinese Society of Nephrology [13030300415]; Jiangsu Province Key Medical Personnel Project [RC201162, ZDRCA2016002]; Outstanding Young and Middle-aged Talents Support Program of the First Affiliated Hospital of Nanjing Medical University (Jiangsu Province Hospital).