Role of Hepcidin in Anemia of Chronic Disease in Rheumatoid Arthritis

Introduction

Rheumatoid arthritis (RA) is a systemic autoimmune inflammatory disorder that affects multiple joints and is characterized by a number of extra-articular manifestations such as rheumatoid nodules, vasculitis, lung or heart disease, anemia of chronic disease (ACD), and peripheral neuropathy. In fact, ACD has been identified as a predictor of radiographic progression as well as an indicator of active clinical or subclinical inflammatory state.^[1,2] In addition, anemia in RA may contribute to secondary manifestations such as fatigue, reduced mobility, cardiovascular conditions, and increased hospitalization.^[3,4] Thus, in order to provide an index of the patient's level activity and progression, the DAS28 composite score has been proposed.^[5] However, although hemoglobin levels are considered an index of system inflammation, they are not included in the disease activity scores (DAS), while more “traditional” markers of inflammation such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are included.^[6,7] ACD in RA has a multifacet pathogenesis that implicates changes in iron metabolism, impairment of erythropoiesis, downregulation of response to erythropoietin (EPO), or most likely a combination of all the above.^[8] Thus, an index of anemia implicated in iron metabolism, erythropoiesis, and inflammation could provide more global information regarding the complex interaction of homeostatic markers of erythropoiesis/iron metabolism and mediators of inflammation.^[9-13] In this regard, hepcidin (HEP) has received increased attention.^[9,14] In this complex interplay of anemia, inflammation, and iron metabolism, HEP (along with EPO) is considered a mediator of hematopoiesis, but its production is also influenced by markers of inflammation such as interleukin-6 (IL-6) and tumor necrosis factor alpha.^[15] Typically, HEP binds to cell membrane-localized ferroportin and initiates the internalization and degradation of ferroportin that results in rapid iron sequestration into specialized storage cells and iron retention in the duodenal epithelium.^[16]

Thus, the aim of the present cross-sectional observational study was to assess the relationships of HEP with markers of anemia, iron metabolism, inflammation, and erythropoiesis.

Methods

Results

HEP had statistically significant mostly moderate negative relationships with markers of anemia with the exception of RDW that was positive (►Table 1). In addition, HEP correlated negatively with all markers of iron metabolism with the exception of a positive correlation with sTFr (►Table 1). Finally, HEP had small-to-moderate positive correlations with all markers of inflammation and erythropoiesis (►Table 1).

All markers of anemia loaded significantly on factor 1 of the PCA. All markers of inflammation loaded on factor 2 (►Table 1). Some markers of iron metabolism loaded significantly on factor 1 (IRON, TFsat, sTFr), while the rest (TIBC, FER) loaded significantly on factor 2 (►Table 1). All markers of erythropoiesis (HEP, EPO) loaded on factor 1; HEP had also a significant loading on factor 2 (►Table 1). ►Fig. 1 represents the loadings of every single variable on both factors in rotated space as well as the unique combination of markers of anemia, iron metabolism, inflammation, and erythropoiesis that “reduces” to the corresponding factors. HEP is the only variable that contributes to both factors.

Close

Fig. 1

Component plot of the extracted factors in rotated space. CRP, C-reactive protein; DAS28, disease activity score 28; EPO, erythropoietin; ESR, erythrocyte sedimentation rate; FER, ferritin; Hb, hemoglobin; HEP, hepcidin; IL-6, interleukin 6; IRON, iron metabolism serum iron; MCH, mean corpuscular hemoglobin; MCV, mean corpuscular volume; MCHC, mean corpuscular hemoglobin concentration; RDW, red cells distribution width; RETHb, reticulocytes hemoglobin; TIBC, total iron binding capacity; sTFr, soluble transferrin receptor; TFsat, transferrin saturation.

Export to PPT

Discussion

The present cross-sectional observational study aimed to clarify relationships between HEP and a comprehensive set of variables covering markers of anemia, iron metabolism, inflammation, and erythropoiesis. Our results revealed that in a group of anemic and nonanemic RA patients, HEP is correlated (to varying magnitudes) to markers anemia, iron metabolism, inflammation, and erythropoiesis (►Table 1). This is in accordance with previous studies demonstrating similar associations between HEP and anemia markers^[21-25] or iron metabolism^[25-29] however, others have reported no correlation.^[23,30-32]

In order to better understand the dynamics of this complex correlation matrix, we performed data reduction using PCA where linear combinations of the original variables were “reduced” to single factors.^[19] Our PCA analysis revealed that regarding the factor 1, explaining approximately 1/3 of total variance, variables that had significant loadings were all markers of anemia, markers of iron metabolism (IRON, TFsat, and sTFr), and erythropoiesis (EPO and HEP). Anemic state in chronic disease is influenced by the complex interaction between inflammation, erythropoiesis, and iron metabolism^[15,33-35] however, it is not clear to what extend anemia is merely a marker of disease severity and progression or a causative factor that specifically impacts on the underlying inflammatory process by the progression of joint destruction.^[15,34] Inflammation-induced cytokines activate HEP production (IL-6) and increased levels of HEP induce a state of “functional iron deficiency”^[33] however as the disease progresses low, IRON levels may actually exert inhibitory effects on HEP production.^[15] On the other end of the spectrum low IRON availability further activates erythropoiesis, while EPO may act regulatory effects on HEP.^[15] This loop of inflammation-induced erythropoiesis-iron availability-anemia is depicted in factor 1 of our PCA analysis.

We propose that the longitudinal monitoring of these markers may be useful in patient evaluation as the disease progresses through time. For example, MCH and MCHC directly measure Hb content per red blood cell (RBC) and RETHb provides direct information regarding the iron supply to red-cell precursors.^[36] In addition, it has been demonstrated that, at least in the early stages of RA, the upregulation of IL-6 suppresses the maturation of erythrocytes in bone marrow, which leads to the immature erythrocytes entering the circulation, resulting in the changes of RDW level.^[37,38] Furthermore, it has been proposed that RDW can be influenced not only by anemia but also from inflammation.^[39] Similarly, sTFr has been proposed as a marker reflecting iron status, namely to identify functional iron deficiency in RA.^[40] HEP had the lowest loading on factor 1, which is not unexpected given the cascade of events that span from the level of stimulation of these markers to the level of “disturbed” iron status at the single RBC.^[15] Therefore, the combination of the above variables is considered to measure different aspects of some common underlying dimension just as suggested by our PCA.^[19] Given that most variables that had significant loading on factor 1 are associated with the individual erythrocyte level (MCH, MCHC, RETHb, sTFr), it can be concluded that factor 1 represents a global index of “RBC health status.”

Our PCA analysis further demonstrated the existence of second (independent) factor explaining approximately 16% of total variance. Variables having substantial loadings were markers of inflammation (CRP and ESR), clinical activity (DAS28), another subset of iron metabolism (TIBC and FER), and finally HEP. Given that TIBC actually reflects transferrin levels,^[41,42] both TIBC and FER are also considered acute phase protein along with CRP.^[43,44] In fact TIBC as a surrogate of transferrin had a strong negative loading on factor 2, while FER and CRP had a positive loading that demonstrates their role as “negative” and “positive” acute phase protein, respectively.^[45] Albeit weak, HEP did have a meaningful loading on factor 2 also that may underscore its function in inflammation^[14] and also justify the small correlations between HEP and inflammation markers (►Table 1). In this regard, our results are in agreement with previous studies demonstrating that HEP represents the link between the “crossroad” of anemia and inflammation,^{[23,25,35,46,47]} although not clinical disease activity^[28,48] however, given that neither ESR nor CRP is included in the composite activity analysis,^[7] it is not surprising that DAS28 had also substantial loading on factor 2. Thus, it can be concluded that factor 2 represents a global index of “acute phase reactants.”

The pattern that emerged from our analysis is that the majority of variables (MCH, sTFr, MCHC, RETHb, RDW, Hb, EPO, MCV, TFsat, IRON) loaded highly on “RBC health status” and considerably less on “acute phase reactants,” while the opposite was true for the remaining set of variables (CRP, ESR, DAS28, TIBC, FER) (►Table 2). The exception was HEP that presented similar loadings on both factors. On one hand from a physiological/clinical perspective, this may represent the pivotal role of HEP in the interaction between anemia and inflammation outlined above.^{[14,23,35,47]} On the other hand from a “statistical” perspective, if a variable has similar-sized loadings across two or more factors this could be due to the factors representing similar constructs (that is being correlated) or that is not a good index distinguishing those constructs.^[19] Given that the two rotations produced similar solutions, “RBC health status” and “acute phase reactants” are likely independent dimensions; this is further supported by the fact that no other variable exhibited substantial loadings (> 0.4) on both factors. Thus, the possibility remains that HEP cannot distinguish well either “RBC health status” or “acute phase reactants.” However, given that both loadings of HEP were above the cutoff value of 0.4, it provides some evidence in favor of its diagnostic potential.^[14,49] Thus, it can be concluded that HEP can be viewed as an index of “RBC health status” in the presence of inflammation.^[50]

The present findings should be interpreted in view of potential limitations. The cross-sectional design limits any causal inferences and is inherent in all observational studies. In addition, PCA is used as an exploratory technique and the corresponding loadings do not provide biological associations. Furthermore, PCA analysis limits conclusions to the sample collected and generalization of results requires cross-validation of the factor structure in a different sample. In this regard, PCA was used to generate future hypotheses, mainly regarding the diagnostic potential of HEP in a complex state of various degrees of inflammation, anemia, and iron availability.

In conclusion, we demonstrated that HEP shows significant relationships with markers of anemia, iron metabolism, inflammation, and erythropoiesis in a population of anemic and nonanemic RA patients. Furthermore, reducing the combination of these markers onto two “latent” factors describing “RBC health status” and “acute phase reactants,” HEP demonstrated significant loading on both factors. However, despite the pivotal role of HEP in these relationships approximately half of total data variability remained unexplained. Larger studies are needed to identify more “latent” variables than can potentially explain more variability and/or ascribe more physiological roles on HEP.

Abbreviations: CI, confidence interval; CRP, C-reactive protein; DAS28, disease activity score 28; EPO, erythropoietin; ESR, erythrocyte sedimentation rate; FER, ferritin; Hb, hemoglobin; HEP, hepcidin; IL-6, interleukin 6; IRON, iron metabolism serum iron; MCH, mean corpuscular hemoglobin; MCV, mean corpuscular volume; MCHC, mean corpuscular hemoglobin concentration; RDW, red cells distribution width; RETHb, reticulocytes hemoglobin; TIBC, total iron binding capacity; sTFr, soluble transferrin receptor; TFsat, transferrin saturation.