Iron Deficiency and Anaemia in Heart Failure: Understanding the FAIR-HF Trial

Chronic heart failure (CHF) is increasingly viewed as a multi-system disease which, beyond the impairment of cardiac function, also
affects the functional capacity of other organs such as the kidneys and skeletal muscle. Anaemia and chronic kidney disease (CKD) are
the most prevalent comorbidities in CHF and they both confer an independent worse prognosis. Mechanisms unrelated to haemodynamic
dysfunction, such as inadequate tissue oxygen supply and impaired oxygen use by the skeletal muscle, may underlie impaired exercise
tolerance. Anaemia in CHF can be the consequence of reduced glomerular filtration rate, reduced plasma flow, impaired erythropoietin
(EPO) production, and haemodilution.
In up to 53% of out-patients with anaemia and CHF due to systolic dysfunction, some sort of haematinic deficiency can be found, although
up to 27% of patients with CHF and without anaemia can also have a haematinic deficiency. In another study of 148 patients with
anaemia and CHF, most patients (57%) presented anaemia of chronic disease, which was indicated by an inadequate production of
EPO relative to the degree of anaemia and/or a defective iron supply for erythropoiesis. Both these findings seem to be associated with
elevated levels of inflammatory cytokines.

Correction of Anaemia
Great early enthusiasm was generated by Silverberg et al.[5] small initial studies of correction of anaemia in patients with CHF and CKD.
He achieved an increase in haemoglobin, functional class, and left ventricular ejection fraction, and decreases in hospitalizations for heart
failure, with EPO and intravenous (IV) iron. Erythropoietin-stimulating agents (ESA) received most of the attention from the medical
community, and randomized studies were performed assessing the benefits of ESA administration in patients with anaemia and CHF.
However, iron replacement was given a secondary role and was only corrected with oral instead of IV iron, except in Silverberg's studies.
Two recent meta-analyses of these studies with ESA demonstrated a beneficial effect on heart failure hospitalizations and some signs of
symptomatic improvement with no increase in mortality or other adverse events.[7,8] However, in the larger studies, there was no
significant improvement in exercise capacity despite the increase in haemoglobin.[9,10] If we add the patients with CHF that were
included in the recent TREAT trial, that was designed to assess ESA treatment in diabetic patients with CKD and anaemia, there is a
neutral effect on mortality and non-fatal heart failure events.[11] In this situation, an adequately powered clinical trial is needed and
hopefully, the RED-HF trial will shed more light on the usefulness of ESA therapy in patients with anaemia and CHF.[12]
These findings are in contrast to the consistent clinical improvements found in recent studies using only IV iron in patients with CHF and
iron deficiency (ID), with or without anaemia.[13–17] In the FAIR-HF trial, 459 patients were randomized to receive IV iron as ferric
carboxymaltose vs. placebo. Among patients receiving IV iron, 50% reported being much or moderately improved, as compared with
28% of patients receiving placebo, according to the patient global assessment (odds ratio for improvement, 2.51; 95% CI, 1.75–3.61).
There was also a significant improvement in the New York Heart Association functional class, in the distance on the 6-minute walk test
and quality-of-life assessments. There were no differences in adverse events, but there was a trend for fewer hospitalizations for any
cardiovascular disease in the IV iron group. It is important to note that treatment with IV iron was beneficial to both patients with and
without anaemia.[16] This finding was also found in the FERRIC-HF trial and is consistent with the concept that impaired physical
performance in iron-deficient animal models is due to two facts: (i) the impaired oxidative capacity of the skeletal muscle, as myoglobin,
mitochondrial cytochrome, and iron-sulfur content, and total mitochondrial oxidative capacity decreases. (ii) The diminished oxygen
transport when anaemia develops.[18] Also, the outcomes assessed in the FAIR-HF trial tended to estimate lower intensity endurance
exercise, which correlates tightly with tissue ID. If maximal aerobic capacity had been assessed, probably no benefit would have been
found in the non-anaemic group, as this parameter depends fundamentally on the haemoglobin concentration and not on the oxidative
capacity of muscle.[19]

Iron Deficiency in Chronic Heart Failure
It therefore seems reasonable to look further into the pathogenesis of anaemia in patients with CHF. Although vitamin B12 or folic acid
deficiency must be assessed, as it may affect up to 19% of patients with CHF and anaemia, it seems to be of secondary importance.[1]
Iron deficiency plays a critical role in the anaemia of CHF, and it can contribute to EPO resistance, as the bone marrow will not respond to
EPO unless adequate iron stores are present.[20] An important point is that despite seemingly adequate iron stores assessed by serum
iron and ferritin, up to 73% of patients with anaemia, normal kidney function, and advanced CHF had ID as assessed by bone marrow
aspiration in a study by Nanas et al.[21] This study demonstrated that neither serum iron nor ferritin levels proved to be reliable markers of
ID. The reason for this higher than expected serum ferritin may have been due to inflammatory mediators that accompany the CHF
syndrome and this is why in these patients a higher value of ferritin (<100 µg/L) defines absolute ID.[16]
Iron is an essential trace element that can donate electrons in its ferrous form-Fe(II)-and accept electrons in its ferric form-Fe(III). This
capability makes it a useful component of cytochromes and oxygen-binding molecules, such as haemoglobin and myoglobin, but can also
promote the generation of free radicals and makes iron potentially toxic. For this reason iron is bound to proteins, yielding solubility in
aqueous solutions such as blood, without the risk of free radical generation.[20] Since iron is not actively excreted from the body, iron
haemostasis is mostly regulated by iron absorption in the duodenum and proximal jejunum (see Figure 1). There are two different iron
absorption pathways: one for haem-bound iron, mostly bound to porphyrins in meat-based foods, and another for non-haem iron, mostly
found in vegetables. Haem-iron constitutes only 10% of dietary iron, but because of greater bioavailability it represents 30% of the total
absorbed iron and it is absorbed via a specific membrane transporter. Non-haem iron is mostly found in the Fe(III) state and is reduced to
Fe(II) in the intestinal apical membrane by a ferrireductase, which is induced by ID. Fe(II) is then transported by a divalent metal
transporter 1 (DMT1) into the enterocyte. Vitamin C, amino acids containing cysteines, and gastric acid reduce Fe(III) non-haem iron to
the more easily absorbable Fe (II). On the other hand, tannins (tea, coffee), oxalates (spinach), phosphates (milk), antacids, and proton
pump inhibitors reduce non-haem iron absorption.[20] The iron absorbed by DMT1 can be incorporated to enterocyte ferritin, and
eventually lost when the enterocyte is replaced, or it can be exported by a transport protein called ferroportin, within the basolateral
membrane.[22] Hepcidin is a peptide synthesized by the liver in response to an increase in transferrin saturation (TSAT), microbial
infection, or inflammation. It blocks intestinal iron absorption and iron release from the liver and spleen by binding to ferroportin and
degrading it.[23] Iron released into the bloodstream is bound to apotransferrin and yields transferrin, which transports iron to all body cells.
[20] Transferrin levels correlate inversely with the amount of body iron stores. These transferrin molecules bind to transferrin receptors on
the cell surface and this complex is internalized. In erythroid cells, iron moves mainly into the mitochondria to be used in haem synthesis
for the subsequent formation of haemoglobin outside the mitochondria. In non-erythroid cells, iron is stored as ferritin and haemosiderin.
[24] The concentration of soluble transferrin receptors in plasma is increased in absolute ID,
but not by the acute phase response,
helping to differentiate between ID anaemia and anaemia of chronic disease.[25]

Diagnosis of Iron Deficiency
Diagnosing ID in CHF patients is important, as iron plays a key role in oxygen uptake, transport, and storage, in the oxidative metabolism
of the skeletal muscle and in erythropoiesis.[22] The evaluation of iron metabolism must include the determination of serum iron,
transferrin, TSAT, and ferritin (see ). In the FAIR-HF trial, true ID was defined as ferritin <100 µg/L, normally accompanied by high
transferrin and low TSAT. In these patients, overt bleeding and poor dietary intake need to be evaluated, and it seems logical that these
patients will respond to IV iron. Patients with functional ID defined as ferritin between 100 and 299 µg/L, and a TSAT <20% also benefited
from IV iron. But what does this functional ID mean? If we consider anaemia of CHF as anaemia of chronic disease, then there will be
increased uptake and retention of iron in the cells of the reticuloendothelial system. This is achieved by the expression of DMT1, which is
up-regulated by cytokines. Divalent metal transporter 1 mediates iron transport into the intestinal mucosal cells and into the activated
macrophages, but the export of iron from these cells is inhibited by down-regulation of the expression of ferroportin by means of an
increase in hepcidin. This protein also inhibits iron absorption from the gut, and hepcidin levels seem to reflect iron load and response to
EPO rather than inflammation and EPO resistance.[26] Thus, this will imply normal or increased ferritin with low serum iron, low transferrin,
low TSAT and thus poor availability of iron at the bone marrow.[25] In a recent study of 546 patients with systolic CHF, ID (absolute or
functional) was found in 37% of patients and ID, but not anaemia, was related to an increased risk of death or heart transplantation in
multivariable analysis, reinforcing its importance as an independent predictor of unfavourable outcome.[27]

Treatment of Iron Deficiency
The FAIR-HF trial demonstrated that IV iron therapy in patients with CHF and ID improves their functional status and increases
haemoglobin in those that were also anaemic.[16] One may wonder whether oral iron, which is cheaper, could have achieved similar
results. We think that most likely the answer is no, because the absorption of oral Fe(II) iron preparations in the anaemia of chronic
disease is blocked by hepcidin;[23] these preparations are poorly tolerated, mainly due to gastrointestinal side-effects; and a number of
drug interactions may occur, such as with proton pump inhibitors.[20] Also, even if oral iron was absorbed, hepcidin would impair iron
delivery to the bone marrow, which is why the anaemia of chronic disease does not respond to oral iron.[23] Intravenous iron, on the other
hand, is provided as carbohydrate complexes with iron in its Fe-III form.[20] They are cleared from the bloodstream and taken into the
reticuloendothelial system within a few hours of administration. The majority of the dose is deposited in long-term storage, but a portion of
this iron is rapidly bound to transferrin and available for transport to the bone marrow,[28] bypassing the restrictions on iron release
imposed by hepcidin. In all the clinical trials, large amounts of iron were delivered over minutes or hours as pulse therapy. This could lead
to poor utilization of this iron with tissue deposition, free radical formation, and increased risk of infection because of a decrease in
cellular immunity and promotion of bacterial growth.[23] However, there was no increase in the risk of infection or cardiovascular events in
the FAIR-HF trial.[16] Also, in contrast to dextran iron, currently used IV iron preparations, such as ferric saccharate, ferric gluconate, or
ferric carboxymaltose are well tolerated and do not require a test dose as they have fewer hypersensitivity reactions.[29] The total iron
dose required for iron repletion in the FAIR-HF trial was calculated according to Ganzoni's formula[30] and was administered as ferric
carboxymaltose in weekly doses of 200 mg as an intravenous bolus injection. This is an advantage over the administration of 200 mg of
ferric saccharate, which is normally given over a period of 1–2 h, as higher doses can cause hypotension.[29] After this correction phase,
a maintenance phase was started, in which 200 mg of iron was given monthly until Week 24. Measures of iron metabolism and
haemoglobin were performed every 2 months in the maintenance phase. If the ferritin level exceeded 800 µg/L or was between 500 and
800 µg/L with a TSAT of more than 50%, or if the haemoglobin level was higher than 160 g/L, IV iron was discontinued.[16] It is important
to monitor these parameters to avoid iron overload and avoid IV iron in the presence of a suspected infection.

After the positive findings of the FAIR-HF trial, there is a need for cardiologists to study the iron metabolism of patients with CHF, and it
seems reasonable to undertake a work-up evaluation if absolute or functional ID is found and start treatment with IV iron, although mortality
and morbidity studies would help to further define the role of IV iron in CHF and its usefulness associated with ESA agents in patients with
anaemia and CHF.