In the chronic setting, AI predominantly results from the body’s inability to increase erythrocyte production to compensate for relatively small decrements in erythrocyte survival (reviewed in Ref. 1). In the steady state, erythrocyte production is sufficiently high so that the resulting anemia is mild to moderate. The anemia associated with acute critical illness has the same pathogenesis as other forms of AI but it develops more rapidly perhaps because of the more extensive erythrocyte destruction and intensive diagnostic phlebotomy common in this setting. The key questions about the pathogenesis of AI, still only partially answered, are: (1) What accounts for the inability of the AI marrow to increase erythropoiesis? and (2) How is this deficit connected to the characteristic hypoferremia and sequestration of iron in macrophages and hepatocytes? Anemia of CKD is similar to AI but the underlying renal pathology also impairs the ability of the kidneys to produce enough EPO leading to insufficient compensatory erythropoiesis.
Human studies indicate that transfused AI erythrocytes have a normal life span in normal recipients but transfused normal erythrocytes have a decreased life span in AI recipients.1 This finding suggests that increased erythrocyte destruction is caused by the activation of hosts factors such as macrophages that prematurely remove aging erythrocytes from the bloodstream. The explanation is consistent with the predominance of young erythrocytes in AI. Whether extrinsic factors, such as bacterial toxins and medications, or host-derived antibodies or complement contribute to this process is unknown.
SUPPRESSIVE EFFECTS OF INFLAMMATION ON ERYTHROPOIETIC PRECURSORS
Some cytokines, chiefly tumor necrosis factor (TNF)-α, IL-1, and the interferons, exert a suppressive effect on erythroid colony formation.12 Interferon-γ overproduction suppresses erythropoiesis in a mouse model13 by reducing erythrocyte life span and decreasing erythropoiesis without any evidence of iron restriction. It is not known to what extent and under what conditions these mechanisms contribute to human AI.
INADEQUATE ERYTHROPOIETIN SECRETION AND RESISTANCE TO ERYTHROPOIETIN
The normal response to increased destruction of erythrocytes is transient anemia followed by an increase in EPO production and subsequent compensatory increase in erythropoiesis. One proposed explanation for the inadequate marrow response in AI is less EPO production than expected based on other types of anemia. Studies of patients with rheumatoid arthritis and AI indicated that EPO levels are increased but less so than in IDA.14,15,16,17,18,19 The findings were similar in patients with anemias associated with solid tumors or hematologic malignancies.20,21 However, these comparisons did not take into account the potentiating effect of iron deficiency on hypoxia sensing (Chaps. 32 and 42).22 This effect could increase EPO production in IDA above that in other types of anemia, and make EPO production in AI appear low in comparison. In support of the EPO suppression hypothesis are experiments with EPO-producing cell lines indicate that production of the hormone is inhibited by inflammatory cytokines including TNF-α and IL-1. The inhibition is mediated by the effects of the transcription factor GATA-1 on the EPO gene promoter, and the suppression of EPO production can be reversed by a GATA inhibitor.23 Moreover, both baseline and hypoxia-induced EPO gene expression is suppressed in rats treated with bacterial lipopolysaccharide or IL-1β to mimic a septic state.24 However, suppression of EPO production is not the major mechanism of AI. If it were, administration of relatively small amounts of EPO should be sufficient to reverse the AI.
In contrast, relative EPO deficiency is often a major contributor to anemia of CKD. Most destructive diseases affecting the kidneys also decrease the release of EPO.25,26 In the kidney, interstitial fibroblasts of neural crest origin26,27 are probably the main source of EPO, but the identity of EPO-producing cells in the kidney remains controversial, mostly because the basal production of EPO is very low and ultrasensitive methods are required to detect the source of the hormone. In response to anemia or hypoxia, the number of renal cells producing EPO increases. In advanced CKD, the kidneys undergo end-stage fibrosis, during which these fibroblasts may transdifferentiate into myofibroblasts and lose their ability to produce appropriate amounts of EPO in response to hypoxia.26,27 However, these or other renal cells can be activated to increase their EPO output by the administration of therapeutic prolyl-hydroxylase inhibitors28 (Chap. 32), as indicated by the lower stimulated EPO production by anephric patients compared to those with end-stage renal disease and retained kidneys. Studies in animal models indicate that the impairment of EPO production in end-stage kidneys may be reversible and could be therapeutically restored.26,27
Inflammation is also a strong contributor to the pathogenesis of anemia of CKD. Patients who had renal disease with inflammation, as measured by increased serum CRP greater than 20 mg/L, required on the average 80 percent higher doses of EPO than patients with simple primary EPO deficiency from renal disease.29 In another study, patients with CRP greater than 50 mg/L reached lower concentrations of Hgb than patients with CRP less than 50 mg/L, despite higher doses of erythropoiesis-stimulating agents.30 Inflammation thus induces a state of relative resistance to EPO, contributing to the pathogenesis of anemia of CKD.
ERYTHROPOIESIS RESTRICTION AS A RESULT OF IRON UNAVAILABILITY
Interleukin-6, Hepcidin, and Hypoferremia
Hypoferremia, one of the defining features of AI, develops within hours of the onset of inflammation.1 Although previous studies of cytokine mediators of hypoferremia of inflammation were inconclusive, subsequent work31 indicates that the response is dependent on IL-6 which induces the iron-regulatory hormone, hepcidin.32 Unlike wild-type mice, mice deficient in either hepcidin33 or IL-634 do not become hypoferremic during turpentine-induced inflammation. In human hepatocyte cell cultures, IL-6 is a potent and direct inducer of hepcidin and neither IL-1 nor TNF-α share this activity. The central role of IL-6 is further indicated by the observation that IL-6-deficient mice do not acutely induce hepcidin in response to turpentine inflammation. Infusion of IL-6 into human volunteers induces hepcidin release within hours and causes concomitant hypoferremia.35 The IL-6–hepcidin axis now appears to be responsible for the induction of hypoferremia during inflammation. However, these studies do not exclude the potential contribution of other cytokines, including activin B and interferon-γ,13,36 to AI in human diseases or more complex mouse models. In support of multiple pathways of AI in a mouse model of inflammation, either the ablation of hepcidin or the ablation of IL-6 ameliorated the anemia, but neither restored normal Hgb concentration.37,38
Serum Iron Concentration Is Dependent on Iron Released from Macrophages and Hepatocytes
In the steady state, almost all of the approximately 20 to 25 mg of iron that daily enters the plasma iron/transferrin pool comes from macrophage recycling of senescent erythrocytes and from hepatocyte iron stores; only approximately 1 to 2 mg come from dietary iron. Only approximately 2 to 4 mg of iron is bound to transferrin but the entire daily iron flow transits through this compartment; thus, the iron in this pool turns over every few hours. During inflammation the release of iron from macrophages and probably also from liver stores is markedly inhibited.39,40,41,42,43,44,45 Studies in transgenic mice lacking hepcidin and mice overexpressing hepcidin indicate that the peptide is a negative regulator of iron release from macrophages and of intestinal iron uptake.46,47 During inflammation, IL-6 induces hepcidin production, which in turn inhibits iron release from macrophages (and probably from hepatocytes), leading to hypoferremia (Fig. 37–2). Hepcidin acts by binding to cell membrane-associated ferroportin molecules that are the only conduits for iron export, and inducing ferroportin internalization and degradation.48 As hepcidin concentrations increase, less and less ferroportin is available for iron export and the iron release into plasma from macrophages, hepatocytes, and enterocytes decreases.
Diagram of the effect of inflammation on iron concentrations in plasma. Arrows labeled “Hepcidin” indicate control points where hepcidin inhibits iron flow into the plasma transferrin compartment.
Erythropoiesis in Anemia of Inflammation Is Limited by Iron
As an intermediate step during the synthesis of heme, iron becomes incorporated into protoporphyrin IX. Zinc is an alternative protoporphyrin ligand. In iron deficiency, increased amounts of zinc are incorporated into protoporphyrin. In AI, zinc protoporphyrin is also increased.49 Insufficient iron reaches the sites of heme synthesis in developing erythrocytes, leading to the substitution of zinc. Moreover, the number of sideroblasts, nucleated erythrocyte precursors that stain for iron with Prussian blue, is decreased in AI.1 A further indication of the limiting role of iron in patients with AI but no evidence of iron deficiency is that coadministration of parenteral iron can resolve the resistance of AI to EPO.50,51 Attempts to treat AI with iron alone generally have been less successful, as iron became rapidly trapped in the macrophage compartment.1,52,53
In the context of anemia of CKD, increased zinc protoporphyrin and decreased reticulocyte Hgb is also characteristic of functional iron deficiency during intense bursts of erythropoiesis stimulated by pharmacologic doses of EPO derivatives.54
Inhibition of Intestinal Absorption of Iron and Other Factors Leading to Systemic Iron Deficiency
In longstanding AI, erythrocytes can become hypochromic and microcytic, partly because progressive depletion of iron stores worsens the iron restriction. Intestinal absorption of iron is inhibited55,56,57 during inflammation by an IL-6 and hepcidin-mediated mechanism.58,59,60,61,62 Only 1 to 2 mg of the daily iron needed for erythropoiesis comes from the diet and most adults have 400 to 2000 mg of iron stores (Chap. 42); therefore, a considerable amount of time is needed to deplete the stored iron. True iron deficiency can eventually develop in chronic inflammatory diseases, especially in children who have smaller iron stores and an additional requirement for iron because of body growth, or in conditions where IL-6 levels are particularly high, such as systemic-onset juvenile chronic arthritis.63 The anemia in these children was accompanied by an appropriate EPO increase, but was unresponsive to oral iron replacement. The anemia was corrected, at least partially, by parenteral iron.
In anemia of CKD, several additional factors may contribute to true iron deficiency, including the blockade of intestinal iron absorption by higher hepcidin concentrations from its decreased renal clearance and the blood losses from hemodialysis, phlebotomy for laboratory studies, and occult gastrointestinal bleeding.
AI is primarily the result of slightly decreased red cell survival and of macrophage iron sequestration leading to iron-restricted erythropoiesis. Depending on the underlying disease, the condition is compounded by inadequate EPO production, suppressive effect of inflammation on erythropoietic precursors, or depletion of iron stores. Anemia of CKD is dominated by the effects of relative EPO insufficiency but inflammation and blood loss also contribute to its pathogenesis.