ANEMIA OF CHRONIC DISEASE IN RHEUMATOID ARTHRITIS PATIENTS: POSSIBLE ROLE OF HEPCIDIN AND TNF-α

Background: Rheumatoid arthritis (RA) is a chronic disease of undetermined cause that is associated with significant disability. Anemia of chronic disease (ACD) is a well recognized extra-articular feature. Hepcidin is a recently discovered hormone that appears to be a key regulator of systemic iron homeostasis. However, the role of this hormone in ACD in RA patients had not been investigated. Cytokines play an important role in rheumatoid arthritis, of these cytokines tumor necrosis factor-α (TNF-α) seems to play a role in pathogenesis of anemia of chronic disease (ACD) in patients with RA.

Aim of the study: To evaluate the role of hepcidin and TNF-α in anemia of chronic disease associated with rheumatoid arthritis.

Subjects and methods: We evaluated serum prohepcidin (as an indicative to hepcidin level) and TNF-α in a group of 30 patients with rheumatoid arthritis suffering from ACD, in addition to 10 healthy subjects as a control group.

Results: In patients with rheumatoid arthritis, a significant increase in serum pro-hepcidin and TNF-α levels had been observed (292.23±103.89 ng/ml and 78.56± 35.87 pg/ml) when compared to control group (223.60±76.80 ng/ml and 26.40±16.06 pg/ml) (p = 0.038 and p=0.001 respectively).

Also, both serum prohepcidin levels and TNF-α correlated negatively with hemoglobin level (p=0.019 and p=0.003 respectively). A trend towards positive correlation had been observed between serum prohepcidin and TNF-α (p= 0.06).

Conclusion: This study provides evidence that both hepcidin and TNF-α are involved in the development of ACD in patients with rheumatoid arthritis.

       

INTRODUCTION

Rheumatoid arthritis (RA) is a prevalent systemic inflammatory disease that mainly affects the joints. This disease affects about 1% of the human population. Although the etiology and pathogenesis of this disease are not yet fullyunderstood it seems that an autoimmune-mechanism plays a crucial role in RA (Swaak, 2006).

 

The most frequent extraar-ticular manifestation in RA is anemia, and although rarely acknowledged as such, it can affect 60% of all patients with RA at least once during their lifelong disease course(Wolfe and Michaud, 2006).

 

Anemia not only contributes to fatigue and reduced quality of life in RA, but longstanding anemia can have deleterious cardiovascular effects and contribute to increased mortality(Swaak, 2006).

 

Anemia in RA patients can be accounted for by iron deficiency and/or less often by reduced vitamin B-12 or folic acid levels. However, the most common form of anemia in this patients group is anemia of chronic disease (ACD). Although many theories have been proposed to explain possible mechanisms underlying ACD in RA patients, its pathogenesis is still unclear (Nissenson et al., 2003).

 

One of the main underlying mechanisms of anemia of chronic disease (also termed as anemia of inflammation) is the disturbance of iron homeostasis, with increased uptake andretention of iron within cells of the reticuloendothelial system.This leads to a diversion of iron from the circulation intostorage sites of the reticuloendothelial system, subsequentlimitation of the availability of iron for erythroid progenitorcells and iron-restricted erythropoiesis (Nikolaisen et al., 2008).

 

Hepcidin is a recently discovered mediator of innate immunity, which had been suggested to be a key regulator of iron homeostasis (Park et al., 2001). Hepcidin, previously reported as liver-expressed antimicrobial-peptide, is a circulating hormone mainly synthesized in the liver by hepatocytes. Studies had reported that hepcidin regulates intestinal iron absorption and affects the release of iron from hepatic stores and from macrophages involved in the recycling of iron from hemoglobin. Furthermore, hepcidin is an acute phase peptide and its pro-duction is increased in inflammation. It has been proposed that hepcidin may be a contributing factor in the pathogenesis of anemia of chronic disease (Fleming, 2008).

However the exact role of this hormone is the pathogenesis of ACD in RA patients is not yet clear.

 

ACD is not just a matter of disturbed iron metabolism; in addition it is associated with impaired erythro-poietinproduction, impaired response of the erythroid marrowto erythro-poietin and a diminished pool of erythropoietin responsivecells (Nicolas et al., 2002).

 

Substantial evidence indicates that inflammatory cytokines subserve a crucial role in joint destruction and disease propagationin RA patients. Among these cytokines, tumor necrosis factor α (TNF-α), which has been considered as the pivotal factorin inducing and sustaining tissue damage. Apartfrom its detection in the inflamed synovial fluid, TNF-α is alsofound in elevated levels in patient sera. Moreover it had been postulated to correlate with disease activity. Furthermore,circumstantial evidence suggests that increased local TNF-α productionin the bone marrow may be implicated in the pathogenesisof anemia of chronic disease in RA patients(Papadaki et al., 2002).

 

In view of these findings, we investigated levels of serum pro-hepcidin, the pro-hormone form of hepcidin and TNF-α in a cohort of patients with RA with ACD to identify if either or both play role in the pathogenesis of ACD in RA patients.

Foot Pad

SUBJECTS AND METHODS:

A. Subjects:

This study was conducted at El-Minia University Hospital, and included 30 RA patients with ACD. In addition 10 healthy subjects, who were age and sex matched to the patients group, were included as a control group.

 

All patients satisfied 4 or more of the revised American College of Rheumatology (ACR) criteria for classifycation of Rheumatoid arthritis (Arnett  et al., 1988).

 

Anemia was defined in accordance with the World Health Organization (WHO) criteria, as presence of hemoglobin levels < 13.0 g/dl for men and < 12.0 g/dl for women, in a manner similar to Wolfe and Michaud (2006).

 

ACD was defined as anemia in associated with serum ferritin level > 50 ng/ml. (Intragumtornchai et al., 1998 and Vreugdenhil et al., 1990).

 

Exclusion criteria included: 1) Patients with iron deficiency anemia, as diagnosed by a low mean corpuscular volume (< 80 fl) with serum ferritin level < 50 ng/ml 2) Presence of any other acute or chronic medical illness 3) bedridden or post operative state 4) blood transfusion within the last 3 months 5) patients who received erythropoietin and/or iron treatment or who had recent history of bleeding were also excluded.

 

Methods:

Blood Sampling and Laboratory Procedures

Blood samples were collected from all participants in 3 tubes (Plain, citrated for ESR and K3 EDTA for CBC). After centrifugation, serum was collected and stored at -70^oC till further analysis of serum prohepcidin and TNF-α.

 

The following laboratory tests were done for all participants แผ่นแปะฝ่าเท้า

 

Routine laboratory tests:

Chemistry (ALT, AST, bilirubin, urea and creatinine) were done by fully automated machine (Dimension ES, USA). CRP was done by turbidmetric assay, Kits was supplied by Spinreact, Spain. ESR was done by Westergrenn method. CBC was performed on automated hematology analyzer (Sys-mex-K 800, Japan). Serum iron was done by colorimetric method (Biocon, Germany). Serum ferritin was done by ELISA (TECO Diagnostics USA).

 

Pro-hepcidin serum assessment:

Determination of serum prohe-pcidin concentration in serum was carriedout by commercially available solid phase enzyme-linked immuno-sorbent assay (ELISA) kits, obtained from DRG (Heidelberg,Germany, Ref: RE 54051). The detection limit was 4.0 ng/ml.

 

Serum TNF-a:

TNF-a was quantified in serum using ELISA method (Endogen, USA). The sensitivity of which is 4.0 pg/ml. The absorbance of each well was read at 450 nm and a standard curve was constructed to quantitate the TNF- a concentration in the assay samples.

 

Statistical Analysis:

Statistical analyses were performed using SPSS statistical package. Values are presented as mean ±SD, unless otherwise mentioned. Comparison between groups was by unpaired t-test for normally distributed date and Mann-Whitney U-test for data that were not normally distributed. Evaluation of the correlation was performed using Pearson and Spear-man single correlation coefficient analysis. For categorical data CHI square test was used. Multiple linear logistic regression was done for
determination of independent variable for hemoglobin level. Values of P<0.05 or less were considered statisticallysignificant.

 

RESULTS:

            Some of the baseline characteristics and clinical and laboratory data of both studied groups are demonstrated in table (1). As shown they were 30 patients with RA with ACD (11 males and 19 females), with a mean±SD age of (47.1±12.6 years). Regarding the control group, they were 10 health subjects (3 males and 7 females) with mean±SD age of (50.7±12.2 years). As expected, both ESR and CRP were significantly higher in RA group (p=0.0007 and p=0.001 respectively).

 

 

Table (1): Comparison between some baseline characteristics of RA group with ACD

                 and control group

 

Variable	RA group with ACD (n=30)	Control Group (n=10)	P-value
Age (years)	47.1±12.6	50.7±12.2	0.420
Sex (Male/Female)	11/19	3/7	1.00
Disease Duration(months)	71.00±33.63	–	NA
Morning stiffness(minutes)	54.9±4.5	–	NA
No of swollen joints	2.9±1.3	–	NA
ESR (mm/h)	50.28±24.91	17.43±7.16	0.0007
CRP (mg/dL)	48.52±55.72	4.2±0.9	0.001

ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein.

 

 

 

Comparison between some studied hematological parameters in RA group and control group is demonstrated in table (2). A significantly higher hemoglobin level (p=0.003) and iron level (p=0.002) had been observed in the control group compared to RA group with ACD. On the other hand a statistically significant higher ferritin had been observed in RA group patients with ACD (p=0.011).

 

Table (2): Comparison between some hematological parameters of RA group with

                 ACD and control group

Variable	RA group with ACD (n=30)	Control Group (n=10)	P-value
Hemoglobin(gm/dl)	10.10±1.68	12.89±1.09	0.003
RBCs (m/uL)	4.18±0.63	5.25±0.79	0.001
MCV(fl)	79.7±4.76	88.01±6.95	0.004
MCH (pg/cell)	26.1±2.1	28.2±0.66	0.001
MCHC (%)	32.43±1.25	32.9±1.10	0.278
WBC ( x109/L)	6.8±1.66	8.6±2.50	0.059
Platelets (x 109/L)	277.76±83.27	259.60±61.77	0.471
Serum iron (ug/dl)	38.6±9.69	60.8±24.66	0.002
Serum ferritin (ng/ml)	95.70±61.08	42.50±19.48	0.011

 

MCV: Mean corpuscular volume; MCH: Mean corpuscular hemoglobin; MCHC: Mean corpuscular hemoglobin concentration.

 

When comparing serum prohepcidin level and TNF-αbetween both groups (table 3), a statistically significant higher levels of both had been found in RA patients with ACD compared to control group. Prohep-cidin level in RA group with

 

ACD had been (292.23±103.89 ng/ml) compared to (223.60±76.80 ng/ml) in control group (p=0.038). Serum TNF-α in RA group with ACD was (78.56±35.87 pg/ml) compared to (26.40±16.06 pg/ml) in control group (p=0.001) (Figure 1).

 

Table (3): Comparison between levels of TNF-α and Pro-hepcidin between both groups

Variable	RA group with ACD (n=30)	Control Group (n=10)	P-value
Pro-Hepcidin (ng/mL)	292.23±103.89	223.60±76.80	0.038
TNF-α (pg/ml)	78.56±35.87	26.40±16.06	0.001

 

 

Table (4) demonstrates results of simple linear regression analysis between different studied parameters. As demonstrated, Serum prohepcidin level correlated negatively with hemoglobin level (p=0.019) (Figure 2). Likewise, TNF-α correlated negatively with hemoglobin level (p=0.003) (Figure 3). A trend towards positive correlation between TNF-α and prohepcidin level had been also observed (p=0.06) (Figure 4). Prohep-cidin and TNF-α showed no statistically significant correlation with ESR, CRP or ferritin. No statistically significant correlation had been demonstrated between hemoglobin level and CRP or ESR.

 

Table (4): Correlation between TNF-α & prohepcidin and some studied parameters in

                  RA patients with ACD

 

Variables	R	P-value
Prohepcidin Vs Hemoglobin	-0.452	0.019
TNF-α Vs Hemoglobin	-0.522	0.003
TNF-α Vs Prohepcidin	0.350	0.06
Prohepcidin Vs ESR	0.352	0.06
Prohepcidin Vs CRP	0.316	0.09
Prohepcidin Vs ferritin	0.041	0.81
TNF-α Vs ESR	0.044	0.82
TNF-α  Vs CRP	0.291	0.12
TNF-α Vs ferritin	0.033	0.68
Hemoglobin  Vs ESR	-0.041	0.74
Hemoglobin  Vs CRP	-0.038	0.69

 

Multivariate regression analysis of predictors for hemoglobin in RA patients:

On multivariate linear regression analysis, when hemoglobin had been

 

introduced in the model as the dependent variable and serum prohepcidin, TNF-α and ferritin ESR and CRP was an independent factors, (Table 5).

 

 

Table (5): Multivariate regression analysis of predictors of hemoglobin in RA

                  patients with ACD

Variable	Coefficient	Std Error	F-test	P
TNF-α	-0.009	0.003	7.00	0.01
Prohepcidin	-0.001	0.001	1.88	0.18

 

DISCUSSION:

Although anemia is not considered to be a major problem in rheumatoid arthritis, it is likely that this is not the case and that this statement is based on the fact that studies of anemia in rheumatoid arthritis are sparse with few systematic reviews (Swaak, 2006).

 

Anemia of chronic disease (ACD) is one of the most commonclinical syndromes encountered in the practice of medicine. The pathogenesis of ACD and the regulationof iron absorption and distribution rank among the major unsolvedproblems in classical hematology. In the last few years rapidprogress has been made on both problems by elucidationof the central role of hepcidin, an iron-regulatory hormoneand a mediator of innate immunity (Weiss and Goodnough, 2005).

 

In this study when investigating serum prohepcidin concentration as a marker of endogenous hepcidin levels, we found it to be increasedin patients with RA and ACD compared to control group. Moreover, a statistically signi-ficant negative correlation had been found between hemoglobin level and serum prohepcidin.

 

To our knowledge, this is one of the earliest reports of assessing serum hepcidin level in patients with RA.

 

Our finding adds to the already sparse knowledge of the role of hepcidin in anemia of chronic disease and very nicely resembles several studies addressing the problem of anemia of chronic disease in other conditions(Fleming, 2008). The same finding of an increased serum prohe-pcidin in a group of patients with RA had been verified in a study carried out by Koca et al., 2008. In their series they had documented elevated serum prohepcidin in a cohort of RA patients and ACD.

 

It had been shown in several studies that hepcidin is a key regulator for iron homeostasis in different clinical settings.

 

Since its discovery in 2000, hepcidin and its role in iron metabolism and inflammation, has been an integral part of experimental and clinical studies. Thereby, clinical research has been focused on its role on pathogenesis of anemia of chronic disease and diseases associated with dysregulation of iron absorption and iron overload like hemochromatosis (Christiansen et al., 2007).

The role of hepcidin in iron metabolism is increasingly being defined (Fleming 2008; Ganz and Nemeth, 2006 Kemna et al., 2008; Rivera et al., 2005 and Silvestri et al., 2008). Moreover, the role of hepcidin had been demonstrated in a variety of conditions associated with ACD in infectious and inflammatorydiseases (Kemna et al., 2005; Nemeth et al., 2003 and Theur et al 2006) and chronic renal failure (Kato et al., 2008) Hepcidin role had been also demonstrated in conditions associated with dysregulated iron metabolism as hemaochromatosis, (Kemna et al., 2008) and  hemoglobinopathies (Kearney et al., 2008 and Origa et al., 2007). Moreover, the role of hepcidin in ACD in childhood had been also verified in a study by Cherian et al., in (2008). In their series they were able to document that increased hepcidin production is responsible for iron accumulation in tissue macrophages and is thought to be responsible for the development of ACD in childhood anemia of chronic disease in different clinical conditions.

 

Anemia of chronic disease typically manifests itself as a hypoproliferativeanemia accompanied by a low serum iron concentration despiteadequate reticuloendothelial iron stores as reflected by increased serum ferritin (Nemeth et al., 2004). It seems that hepcidin is the long-anticipated hormone which could explain much of the characteristics of ACD. Findings from other studies demonstrated that hepcidin leads to internalization and degradation of the ironexporter ferroportin, which is present on the cell surface ofenterocytes attenuating iron uptakein the gut (Nicolas et al., 2002) and inhibits therelease of iron by macrophages (Pigeon et al., 2001). In addition to these effects on body iron distribution,hepcidin might also directly inhibit erythroid-progenitor proliferationand survival (Howard et al., 2007).

 

In our study, no statistically significant correlation had been found between serum porhepcidin and ferritin level.

 

Conflicting reports had been published about correlation between pro-hepcidin and ferritin levels. So while hepcidin has been closely associated and positively correlated with ferritin in some series (Dallalio et al., 2003) and a positive correlation was demons-trated between serum pro-hepcidin and ferritin levels in chronic renal failure patients (Malyszko et al., 2005). On the other hand Nagashima et al., in (2006) reported that serum pro-hepcidin levels negatively correlated with ferritin levels in patients with viral hepatitis C, while this correlation was positive in patients with viral hepatitis B and healthy controls. Other studies demonstrated that levels of pro-hepcidin were unrelated with ferritin or other iron parameters (Roe et al., 2007 and Taes et al., 2004). These contra-dictory findings may support the notion that measurementof pro-hepcidin with the currently available commercial assayis not very reliable as the substitute for bioactive hepcidin analysis (Kemna et al., 2005 and Tomosugi et al., 2006).

 

In this series, no statistically significant correlation had been demonstrated between hemoglobin and either ESR or CRP. A similar finding had been found in other study (Peeters et al., 1999). On the other hand, the work by Vreugdenhil et al., in (1992) demonstrated an association between ACD and increased CRP and ESR. The association between ACD and ESR levels is more complex, as ESR increases independently with falling hemoglobin levels and presence of autoantibodies (Swaak, 2006).

 

The pathogenesis of ACD is not purely a matter of deranged iron metabolism, but also includes shortened red cell survival, impaired erythropoietin production and impaired response of erythroid progenitors to erythropoietin (Adamson, 2008). Although hepcidin role in iron metabolism had been verified, it is more likely that other factors are also operative.

 

In this series a statistically significant higher levels of TNF-α had been documented in patients with ACD compared to control group.

Although serum level of TNF-α had been extensively studied in patients with rheumatoid arthritis, this study is additionally demonstrating a negative correlation between this cytokine and hemoglobin levels. Furthermore, serum prohepcidin level positively correlated with TNF-α level.

 

Previous studies have shown that patients with rheumatoid arthritis have increased serum concentrations of variety of pro-inflammatory cytokines, including TNF-a (Papadaki et al., 2002). Moreover, the role of this cytokines in the development of ACD has been well recognized. Increased levels of the cytokine have been reported in anemic patients with cancer (Lastiri et al., 2002), parasitic and bacterial infections (Kern et al., 1989) and AIDS and AIDS-related complex (Dallalio et al., 1999) in addition to rheumatoid arthritis.

 

The role of TNF in pathogenesis of ACD in RA patients was verified in a series reported by Papadiki.et al., 2002. In their series they suggested that patients with RA exhibit low frequency and increased apoptosis of bone marrow erythroid progenitorand precursor cells due to increased local production of TNF-α.Moreover, they provided in vitro and ex vivo evidence that TNF-α -inducesaccelerated apoptosis of bone marrow erythroid cells largely contributingto the pathogenesis of ACD in RA.

 

Impaired erythropoietin produ-ction and impaired response of erythroid progenitors to erythropoietin are two major pathogenic mechanisms underlying ACD. TNF-α seems to be an ideal candidate to be a mediator in ACD. An elegant experiment was carried out by Faquin et al., in (1992). In their study they had demonstrated that when TNF-α. was added to a human hepatoma cell line (these cells have the useful property of upregulating erythropoietin gene expression in response to hypoxia) erythropoietin production in response to hypoxia was dramatically reduced. The suppressive effect of the cytokine was dose-dependent. Similar finding had been verified in a study by Zhai et al., (2004). In their study they found that in vitro recombinant TNF-α inhibited the expression of erythro-poietin mRNA in hypoxic conditions and that the inhibitory effects became stronger with the increase of recom-binant TNF-α.

 

Direct evidencefor the role of TNF- in the pathogenesis of ACD in RA has becoame available from clinical trials using in vivo TNF- blockade. Papadaki et al., in (2002) were able to verify a beneficial effect of administration of a TNF-α blocker on anemia in patients with RA. In their series, approximately60% were anemic and among them 65% had ACD. After 6 doses of TNF-α blocker, a significant improvementof hemo-globin levels was observed in the total study group compared to their baseline values. Moreover, the mostprominent increase was obtained in the group of ACD patients.

 

It seems that TNF-α may also play a role in the disturbance of iron metabolism in patients with ACD. Laftah et al., (2006) demonstrated that TNF-α stimulation results in up-regulation of cellular iron import protein DMT1 (divalent metal transporter 1) and reduced the iron exporter IREG1 (iron-regulated protein 1) in a human monocyte cell line. These actions were found to be independent of hepcidin.

 

In the present study, serum TNF-α plasma levels correlated positively with serum prohepcidin, also both correlated negatively with hemoglobin level. This may imply a cause effect relationship between TNF-α and hepcidin production.

 

The mechanism leading to increased hepcidin in patients with rheumatoid arthritis can be attributed to chronic inflammation. RA is best described as a chronic inflammatory condition. RA is associated with deranged cytokine including, among others, IL-6, IL-1 and TNF-α.

 

Although, Nemeth et al., in 2003 had found that hepcidinmRNA was induced by interleukin-6 in vitro,but not by TNF-, a more recent report found that TNF-α, which is produced mainly by activated macrophages in patients with RA is capable of inducing the production of other cytokines including IL-6 which is a major

inducer for hepcidin release. In this series the author implicated TNF- also in direct induction of hepcidin expression (Brennan and McInnes, 2008).

 

This study has some limitations. First, it has never been proven that pro-hepcidin reflects level of active-mature hepcidin. Thus, if we measure bioactive hepcidin in serum or urine, our results would be more accurate, a finding that had been supported by other investigators (Kemna et al., 2005 and Tomosugi et al., 2006).

 

In inflammatory diseases, iron deficiency anemia can coexist with ACD due to poor intake and/or absorption and increased loss of iron, and so, to differentiate between ACD and iron deficiency anemia may be difficult. The most reliable tool for detecting iron deficiency is stainable iron content in bone marrow aspirate. Recently, serum transferrin receptor level was proposed as a sensitive characteristic for detection of iron deficiency(Swaak, 2006). Thus, our failure to use more sensitive indicators such as transferrin receptor to exclude iron deficiency anemia may be another limitation of the present study.

 

In summary, we have shown that both hepcidin and TNF- α play an important role in the development of ACD in patients with RA. Moreover, it seems that there is some sort of synergism between both. However, these findings require confirmation in larger cohort of patients. Data from this study may have implications in the understandingthe mechanisms of ACD associated not only with RA but also withother chronic inflammatorydiseases.

 

Foot Patch