Interpretation of Erythropoietin and Haemoglobin Levels in Patients with Various Stages of Chronic Kidney Disease.

BACKGROUND: The production of erythrocytes is regulated by the hormone erythropoietin (EPO), which maintains the blood haemoglobin (Hb) levels constant under normal conditions. Human EPO is a glycoprotein hormone and its synthesis is controlled by the hypoxia-inducible transcription factor. The aim of this study was to establish EPO and Hb levels in patients with chronic kidney disease (CKD), as well as in control subjects, and to investigate the relationship between these parameters.
METHODS: This cross-sectional, observational study included 356 subjects with CKD divided into 4 subgroups according to their glomerular filtration rate (GFR). The control group consisted of 206 age and sex matched healthy subjects with GFR rate ≥90 mL/min/1.73 m2. EPO, Hb and serum creatinine levels were determined by using immunochemical and spectrophotometric methods. GFR was determined using the MDRD formula.
RESULTS: The CKD patients had significantly lower levels of haemoglobin (p<0.0005) and hematocrit (p<0.0005) compared to control group. Our results showed that Hb levels decreased, whereas serum creatinine increased with the increasing renal failure. The CKD patients in all four groups had significantly lower (p<0.0005) Hb levels, and significantly higher (p<0.0005) creatinine levels compared to the control group. The median EPO in group I and II were significantly higher (p=0.002; p=0.018), while median EPO in group III and IV were significantly lower (p=0.03; p=0.011) compared to the control group.
CONCLUSIONS: In patients with CKD, GFR positively correlated with Hb and EPO, while the correlation between GFR and serum creatinine was negative.

Introduction

The main humoral regulator of erythropoiesis is the hormone erythropoietin (EPO) (1, 2). EPO is a glycoprotein hormone composed of 165 amino acid residues with four complex carbohydrate chains attached to peptide in four binding positions. The kidneys are the primary source of EPO and its synthesis is controlled by hypoxia-inducible transcription factors (HIFs) (3). HIF is activated in all cells by exposure to hypoxia, as well as cobalt or iron chelators (4). However, under low-oxygen conditions (hypoxia), the number of circulating red blood cells increases to deliver oxygen efficiently into peripheral organs. Therefore, under hypoxic conditions, the kidney will produce and secrete EPO to increase the proliferation of erythrocytes by targeting the colony-forming unitery throid (CFU-E), proerythroblast and basophilic erythro blast subsets in the differentiation (2, 5). EPO binds to its receptor (EpoR) on the surfaces of immature erythroid cells and stimulates signalling cascades for proliferation, differentiation, and antiapoptosis (6).

The plasma concentration of EPO depends on the rate of its secretion in the kidney, and is inversely related to the oxygen content of the blood (2, 3). The dynamics of the secretion of EPO into circulation and, hence, its plasma concentrations are mainly determined by the transcription rate of the EPO gene, which itself is essentially under control of the cellular oxygen content. A decrease in oxygen concentrations increases the activity of the hypoxia-inducible transcription factor (HIF), which in turn triggers EPO gene transcription. Disorders of kidney function lead to abnormal EPO secretion, which may result in severe anaemia or polycythemia (7, 8). In patients with chronic kidney disease (CKD), normochromic normocytic anaemia may develop due to EPO deficiency (7–9). Although anaemia in patients with CKD is multifactorial in origin, it is primarily associated with erythropoietin production deficiency as the glomerular filtration rate (GFR) falls. Once the estimated GFR trends below 60 mL/min/1.73 m2 (Stage 3a CKD), erythropoietin production by the kidneys falls, and anaemia may develop (10). Many studies have shown that every tenth person in the world has a CKD. Renal anaemia is a consequence of CKD and occurs at an early stage, and it gets worse as the disease progresses (11, 12).

According to the WHO, anaemia is defined as the haemoglobin concentration less than 130 g/L for men and less than 120 g/L for premenopausal women (13). Anaemia is a frequent complication of CKD, occurring in over 90% of patients receiving renal replacement therapy and is associated with high risk of death. However, the process is multifactorial, with several other contributing factors: absolute and functional iron deficiency, folate and vitamin B12 deficiencies, reduced red blood cell life span, and suppression of erythropoiesis (14).

Apart from the mentioned facts, anaemia in kidney patients is not recognized and treated at the right time; therefore, the aim of this study was to evaluate EPO and Hb levels in patients with CKD and in control subjects, and to investigate the relationship between these parameters.

Materials and Methods

This cross-sectional, observational study included 562 subjects: a group of 356 patients with CKD (174 males and 182 females; age 61.7±16.1 years) and a control group consisting of 206 healthy subjects (104 males and 102 females; age 59.2±14.4 years).

CKD was defined in accordance with the Kidney Disease Initiatives and Global Outcome guidelines (15). The subjects with CKD were classified into four CKD stages according to GFR rate: Group I (n=88): patients with stage 1 CKD – GFR = 60.00–89.99 mL/min/1.73 m2, Group II (n=141): patients with stage 2 CKD – GFR = 30.00–59.99 mL/min/1.73 m2, Group III (n=73): patients with stage 3 CKD – GFR = 15.00–29.99 mL/min/1.73 m2, and Group IV (n=54): patients with stage 4 CKD – GFR ≤ 14.99 mL/min/1.73 m2.

The control group included apparently healthy subjects with GFR rate ≥90 mL/min/1.73 m2, who did not suffer from any form of anaemia, liver or renal disease, or had any bone marrow complications. The Modification of Diet in Renal Disease (MDRD) formula was used for the calculation of GFR (16). All CKD patients underwent a standard diagnostic protocol comprising a detailed anamnestic questionnaire, medical history and physical examination. Blood samples from CKD patients and healthy subjects were obtained in the morning after a 12-h fasting period. The blood samples were collected into empty tubes and immediately stored at +4 °C. Bio chemical analyses were performed on the same day.

Erythropoietin levels were determined using the enzymatic-chemiluminescent immunometric method on IMMULITE/IMMULITE 1000 systems from Siemens Healthcare Diagnostics. Haemoglobin concentration was determined by the counter of blood elements (Cell Dyn 3700 system), using a haemoglobin-hydroxylamine colorimetric method. Creatinine was determined by a modified kinetic Jaffe’s reaction reported by Larsen (17), on an autoanalyzer Dimension RxL Siemens.

The study protocol was performed in accordance with the Helsinki Declaration as revised in 2000. Upon careful explanation of the study procedure, all subjects signed written consent for participation in the study. The study protocol was approved by the Institute for Research and Development of the Clinical Center University of Sarajevo (Approval number: 0901-2-50/15).

Statistical analysis

Data distribution was determined by the Kolmogorov-Smirnov test for normality. Data was reported as median and interquartile range (IQR). Variable differences between the groups were determined by the nonparametric Mann-Whitney U-test. Correlation coefficient between variables was assessed by Spearman’s test. Statistical significance was set at p<0.05. For statistical analysis of the obtained data, we used the software package SPSS for Windows (version 19.0, SPSS Inc., Chicago, Illinois, USA).

Results

Median haemoglobin concentrations were significantly lower (p<0.0005) in Group I [136.5 g/L (128.3–147.8 g/L)], Group II [127.0 g/L (111.5– 140.0 g/L)], Group III [114.0 g/L (105.0–128.5 g/L)] and Group IV [115.0 g/L (98.0–123.0 g/L)] than in the control group [146.0 g/L (140.0–152.0 g/L)]. Median haemoglobin concentration in Group I was statistically higher (p<0.0005) than in Group II, III, and IV. A significant difference in haemoglobin concentration was also found when Group II was compared to Group III (p=0.002) and Group IV (p<0.0005). Comparison of haemoglobin concentrations between other groups showed no significant differences ().

Box-and-whisker plots of haemoglobin concentration (g/L) in the studied groups.

The solid horizontal lines denote the median value, the box represents 25% and 75% interquartile ranges and the whiskers represent minimum and maximum values.

Group I – patients with stage 1 CKD (GFR=60.00–89.99 mL/min/1.73 m2); Group II – patients with stage 2 CKD (GFR=30.00–59.99 mL/min/1.73 m2); Group III – patients with stage 3 CKD (GFR=15.00–29.99 mL/min/1.73 m2); Group IV – patients with stage 4 CKD (GFR ≤14.99 mL/ min/1.73 m2)

p – probability; NS – not significant; * – compared to Group II; ° – compared to Group III; # – compared to Group IV; ⋄ – compared to Control group

Patients in all stages of CKD had significantly higher median serum creatinine concentrations than in the control group [74.0 μmol/L (65.0–86.3 μmol/L)] (p<0.0005). Median serum creatinine con cen tra tion in Group I [104.0 μmol/L (85.3–126.8 μmol/L)] was significantly (p<0.0005) lower than in Group II [152.0 μmol/L (117.5–189.0 μmol/L)], Group III [250.0 μmol/L (172.0–293.0 μmol/L)] and Group IV [471.5 μmol/L (328.8–713.3 μmol/L)], respectively. A significant difference in serum creatinine concentration was also found when Group II was compared with Group III (p<0.0005) and Group IV (p<0.0005), and when Group III was compared with Group IV (p<0.0005) ().

Box-and-whisker plots of creatinine concentration (mmol/L) in the studied groups.

The solid horizontal lines denote the median value, the box represents 25% and 75% interquartile ranges and the whiskers represent minimum and maximum values.

Group I – patients with stage 1 CKD (GFR=60.00–89.99 mL/min/1.73 m2); Group II – patients with stage 2 CKD (GFR=30.00–59.99 mL/min/1.73 m2); Group III – patients with stage 3 CKD (GFR=15.00–29.99 mL/min/1.73 m2); Group IV – patients with stage 4 CKD (GFR 14.99 mL/ min/1.73 m2)

p – probability; NS – not significant; * – compared to Group II; ° – compared to Group III; # – compared to Group IV; ⋄ – compared to Control group

The median erythropoietin concentrations in the control group [11.0 mIU/mL (8.4–15.4 mIU/mL)] were significantly lower than in Group I [15.3 mIU/mL (10.3–18.2 mIU/mL); (p=0.002)] and Group II [13.1 mIU/mL (9.9–16.2 mIU/mL); (p=0.018)] and significantly higher than median erythro poietin concentrations in Group III [10.2 mIU/mL (7.8–12.2 mIU/mL); (p=0.03)] and Group IV [8.9 mIU/mL (7.1–12.5 mIU/mL); (p=0.011)]. Median erythropoietin concentration in Group I was sig nificantly higher than in Group III (p<0.0005) and Group IV (p<0.0005). A significant difference in erythropoietin concentration was also observed between Group II and III (p<0.0005), and between Group II and IV (p<0.0005). Comparison of the erythropoietin concentrations between other groups showed no significant differences ().

Box-and-whisker plots of erythropoietin concentration (mIU/mL) in the studied groups.

The solid horizontal lines denote the median value, the box represents 25% and 75% of interquartile ranges and the whiskers relate to minimum and maximum values.

Group I – patients with stage 1 CKD (GFR=60.00–89.99 mL/min/1.73 m2); Group II – patients with stage 2 CKD (GFR=30.00–59.99 mL/min/1.73 m2); Group III – patients with stage 3 CKD (GFR=15.00–29.99 mL/min/1.73 m2); Group IV – patients with stage 4 CKD (GFR 14.99 mL/ min/1.73 m2)

p – probability; NS – not significant; * – compared to Group II; ° – compared to Group III; # – compared to Group IV; ⋄ – compared to Control group

There was a statistically significant positive correlation between GFR and haemoglobin concentration in the CKD group (rho=0.501; p=0.0001) and control group (rho=0.155; p=0.026). Statistically significant positive correlation was found between GFR and concentration of erythropoietin in CKD group (rho=0.342; p=0.0001), while negative, although not significant correlation was observed between GFR and concentration of erythropoietin in the control group of healthy subjects (rho=-0.045; p=0.517). Statistically significant negative correlation was observed between GFR and serum creatinine concentration in both CKD group (rho=-0.789; p=0.0001) and the control group (rho=-0.532; p=0.0001) ().

Discussion

Chronic kidney disease is an important epidemic and public health problem associated with a high risk of vascular disease and cardiovascular mortality, and with kidney disease progression. The prevalence of CKD is so high that it represents a worldwide epidemic and a public health problem all over the world (18). Anaemia is a major co-morbidity of CKD and as kidney disease progresses, anaemia increases in prevalence, affecting nearly all patients with stage 5 CKD. It is associated with reduced quality of life and increased incidence of cardiovascular disease, hospitalizations, cognitive impairment, and mortality (19).

EPO deficiency is the most significant cause of anaemia in CKD and has been demonstrated to occur at each stage of kidney failure. Given that the kidney is the sole source of EPO synthesis in adults, reduction in kidney mass as occurs in progressive CKD often results in impairment of EPO production, resulting in anaemia (20). EPO is a glycoprotein hormone with a molecular mass of 30.4 kDa composed of a single 165 amino acid residues chain to which four glycans are attached (3). EPO is primarily produced by peritubular interstitial cells in the deep cortex and outer medulla of kidneys of the adult and in hepatocytes in the fetus (21). These cells are sensitive to hypoxia and once they are sensed, it leads to an increased EPO production (22). Erythropoietin is essential for terminal maturation of erythrocytes (23). In the confines of the bone marrow, erythroid progenitor cells proliferate into large, nucleated proerythroblasts, which are committed into producing cells of the erythroid series (2). Proerythroblasts are exquisitely sensitive to EPO. They proliferate and develop into erythroblasts and afterwards into reticulocytes that enter the peripheral circulation where they mature into red blood cells (22).

Normal development of erythrocytes and biosynthesis of haemoglobin depend on an optimal biochemical state and adequate supplies of the necessary building blocks including protein, vitamins, hormones, iron and copper. If these components are lacking for a prolonged time, erythropoiesis and biosynthesis of haemoglobin slows and anaemia may result (2, 24). Anaemia is a common finding in patients with CKD and is primarily due to reduced production of erythropoietin in the kidney and shortened red cell survival (21).

Our study showed that anaemia was already observed below a GFR of 90 mL/min/1.73 m2. Radtke et al. (25) in their study showed that renal anaemia is manifested in patients with creatinine clearance below 40 mL/min. Fehr et al. showed that the prevalence of anaemia with haemoglobin <110 g/L was around 30 to 40% in patients with severe renal failure and around 15% in patients with clearance of 20 to 39 mL/min (9). The National Health and Nutrition Examination Survey (NHANES) database III was used in two different studies examining the correlation between prevalent haemoglobin concentration and eGFR. Using a cut-off value of 130 g/L in men and 110 g/L in women, the prevalence of renal anaemia increased from 1% at eGFR of 60 mL/min/1.73 m2 to 90% at eGFR of 30 mL/ min/1.73 m2 and to 33–67% at eGFR of 15 mL/min/1.73 m2 (12, 26–27). However, NHANES IV using the WHO definition of anaemia showed a lower prevalence of anaemia for each CKD stage (28).

Our study showed that the concentration of haemoglobin decreased with increasing renal failure. The CKD patients in all four groups had significantly lower (p<0.0005) concentrations of haemoglobin compared to the control group. Being below 30 mL/min/1.73 m2 of GFR (CKD III and IV), the EPO levels were lower than in the control group. This can be attributed to the serious damage of the kidney tissues and the lack of normal EPO biosynthesis, secretion and regulation. Several studies examined the relationship between erythropoietin and haemoglobin levels in patients with various CKD stages. Radtke et al. (25) measured EPO levels in 135 patients. The first group consisted of 117 patients with chronic renal disease, and the second group of 18 patients suffering from end stage renal failure. They established significantly increased average levels of EPO in all groups as compared to the average level of EPO for 59 reference subjects. They also established that levels of EPO decreased with increased deterioration of excretory renal function (25).

Fehr et al. (9) attempted to establish a quantitative association between erythropoietin levels and haemoglobin at different levels of creatinine clearance. A total number of 5120 consecutive patients were investigated. Analysis of erythropoietin and haemoglobin levels showed that patients with creatinine clearance >90 mL/min (group E) and 60–90 mL/min (group A) had identical levels of haemoglobin and erythropoietin. However, EPO levels only rose mildly in patients with creatinine clearance 20–40 mL/min (group B) and 40–60 mL/min (group C) compared to group E, but were equal between groups A and E (9).

Chandra et al. (29) studied 48 children with various degrees of renal insufficiency. They showed a significant correlation of haemoglobin and creatinine clearance below a clearance of 40 mL/min.

In the present study, the serum creatinine levels were significantly higher (p<0.0005) in all stages of CKD patients than in healthy controls. Serum creatinine levels increased with deteriorating renal function, with the highest levels recorded in CKD group IV patients. The median EPO concentrations in group I and II were significantly higher (p=0.002; p=0.018), while median EPO concentrations in group III and IV were significantly lower (p=0.03; p=0.011) than in the control group. In patients with CKD, GFR positively correlated with the concentration of haemoglobin and erythropoietin, while the correlation between GFR and se rum creatinine concentration was negative.

Lack of EPO is a primary factor causing anaemia in CKD. The renal fibroblasts lose their EPO-producing capacity after injury and trans-differentiate to scar-producing myofibroblasts (29). The successful correction of CKD anaemia has resulted in reduction of associated morbidity and improvement of functionality, exercise tolerance, cognitive function and overall quality of life. Moreover, significant reduction of cardiovascular morbidity and mortality has occurred. In 2007, the National Kidney Foundation (NKF) and Kidney Disease Outcomes Quality Initiative (KDOQI) Work Group updated the recommended haemoglobin level target between 110 and 120 g/L for all patients with CKD and issued a statement warning about greater risk when targeting haemoglobin levels higher than 130 g/L (31). Also, International Clinical Practice Guideline for Anemia in Chronic Kidney Disease suggests to target haemoglobin levels ≤115 g/L, with individualized consideration of slightly higher haemoglobin targets to improve quality of life (32). Therefore, change in the global approach to CKD from the end-stage renal disease (ESRD) treatment to much more aggressive primary and secondary prevention is an imperative (33).

Conclusion

Erythropoietin biosynthesis and secretion are inversely regulated by haemoglobin levels. A low tissue pO2 leads to up-regulation of the hypoxia-inducible factor-alpha, a transcription factor that regulates the erythropoietin biosynthesis in the renal cortex. Basically, »relative erythropoietin deficiency« in the context of renal failure can be caused by two different ways: either the biosynthesis of erythropoietin is decreased due to the tissue damage by the underlying disease, or the set-point for erythropoietin secretion is lowered in relation to tissue oxygenation. Our results showed positive correlation between glomerular filtration rate and haemoglobin and erythropoietin concentration in patients with chronic kidney disease, while the correlation between glomerular filtration rate and serum creatinine concentration was negative.

Table I

Correlations between serum glomerular filtration rate (GFR) and clinical laboratory parameters in CKD group and control group.

	GFR (mL/min/1.73/m2)
Variables	CKD group (n=356)		Control group (n=206)
	rho	p	rho	P
Haemoglobin (g/L)	0.501	0.0001	0.155	0.026
Creatinine (μmol/L)	-0.789	0.0001	-0.532	0.0001
Erythropoietin (mIU/mL)	0.342	0.0001	-0.045	0.517

n – number of cases; rho – Spearman’s correlation coefficient; p – probability