Trends in iron therapy
In the pre-ESA era, most dialysis patient required multiple transfusions to maintain Hb levels in the 7-8 g/dL range and the iron from the transfused blood could not be excreted. Consequently, serum ferritin levels were typically in the 1000-1500 ng/mL range. Administration of ESAs to these patients in the clinical trials during the 1980s produced a rapid reduction in serum ferritin levels as storage iron was incorporated into newly synthesized RBCs.
To maintain adequate delivery of iron to the erythroid marrow in the setting of ongoing blood loss from HD, most patients in the post-ESA era have required IV iron supplementation as oral iron is not sufficiently bioavailable to meet iron needs. It has been estimated that ongoing blood losses in HD patients are 6.26 mL/day from the gastrointestinal (GI) tract (as opposed to 3.15 mL/day in non-dialysis patients with CKD and 0.83 mL/day in normal volunteers); this calculates to about 700 mg of GI iron losses per year.
It is also estimated that around 7 mg of iron are lost from the HD treatment itself (blood left in extracorporeal circuit and phlebotomy) and additional yearly iron losses of around 500 mg may accrue from vascular access procedures. This totals a mean of around 2200 mg of iron lost per year by HD patients; DOPPS reports that the mean IV iron dose for HD patients in the US in 2016 was around 250 mg/month or 3000 mg/year.
It is likely that this positive iron balance is due to the failure of administered IV iron to adequately support erythropoiesis, since much of the administered IV iron is locked in stores in the body (represented by escalating mean serum ferritin levels in the US HD population) so the provider gives more and more IV iron in an attempt to raise transferrin saturation and improve iron delivery to the bone marrow.
However, an analysis of the rise in mean serum ferritin levels in the US HD population (302 ng/mL in 1993 to 825 ng/mL in 2014) could not be completely explained by positive iron balance and the decrease in mean Hb levels (and the iron content of the RBC mass).
It is suspected that the IV iron itself may be inducing an inflammatory state which increases hepcidin levels and contributes to a greater proportion of the administered iron being sequestered in inaccessible stores. It is worth noting that both the 2006 KDOQI anemia guidelines and the 2012 KDIGO anemia guidelines recommend against IV iron administration if the serum ferritin exceeds 500 ng/mL. However, that does not appear to be the practice since the mean serum ferritin in US HD patients was 803 ng/mL in Oct. 2016—yet, 82.9% of patients received IV iron during the prior 3 months, according to DOPPS data.
Concerns have been raised that this positive iron balance may be toxic, as magnetic imaging studies reveal excess liver iron when serum ferritin levels exceed 300 mg/dL. It is known that iron decreases host defenses and may promote the growth of certain pathogens. IV iron doses exceeding an average of 400 mg/month have been associated with an increase in all-cause mortality, according to DOPPS. Infection is now the leading cause of hospitalization among ESRD patients in the US, eclipsing cardiovascular disease, and liberal use of IV iron has been implicated as a contributing factor.
Despite what may be an imbalance in the supply of iron to the erythroid marrow (decreased due to absolute iron deficiency or iron mobilization defects in the setting of inflammation) and demand (increased due to ESA therapy and the need for RBC output to compensate for decreased RBC life span and increasing the Hb level from below target to target range), IV iron is an unphysiologic route to address this imbalance.
Typical doses of IV iron administered during a dialysis treatment (50-125 mg) overwhelm iron control pathways that evolved to recycle 2-4 mg of iron daily from senescent RBCs and iron absorbed orally to replace GI losses. Administered iron that exceeds iron transport mechanisms must be “parked” in stores from which release may be blocked by hepcidin. Studies have suggested that small doses of IV iron given frequently are more effective in supporting erythropoiesis than larger doses of IV iron given less frequently because less iron must be stored in the former case. Unfortunately, in the US, administration of small frequent doses of IV iron products is not economically feasible because the products are supplied in non-reusable, fixed-dose vials. This has led to the exploration of alternative routes of bioavailable iron administration that allow for small frequent dosing.
Oral vs. intravenous iron
Oral iron has been long been considered ineffective in hemodialysis patients because of its poor bioavailability and the magnitude of iron requirements to overcome ongoing iron losses and functional iron deficiency. In non-dialysis CKD patients, both the 2006 KDOQI and the 2013 KDIGO anemia guidelines recommend considering a 3-6 month trial of oral iron in iron deficient patients before defaulting to IV iron. The Randomized Trial to Evaluate IV and Oral Iron in CKD (REVOKE) study assigned patients with stages 3 and 4 CKD and iron deficiency anemia to oral ferrous sulfate (325 mg 3 times daily for 8 weeks) or IV iron sucrose (200 mg every 2 weeks x 5 doses). REVOKE was prematurely terminated by an independent data safety monitoring board when it was noted there was a substantial signal for harm from IV iron and no reasonable likelihood of detecting a difference in the rate of decline in glomerular filtration rate between the two groups (the primary end-point).4
The Ferinject Assessment in Patients with Iron Deficiency and Non-dialysis Dependent CKD (FIND-CKD) study examined the ESA sparing effect of two doses of IV iron vs oral iron in anemic CKD patients receiving ESAs. Patient were randomly assigned 1:1:1 to IV ferric carboxymaltose (Ferinject in Europe, Injectafer in the US) with a target serum ferritin level of 200-400 ng/dL, a target ferritin level of 400-600 ng/dL, or oral iron with ferrous sulfate 100 mg elemental iron twice daily. Patients treated with IV ferric carboxymaltose with higher target serum ferritin quickly achieved and maintained target Hb levels with no difference in infectious or cardiovascular events compared to the other groups.5 Like the DRIVE study in HD patients, higher serum ferritin targets may be associated with increased ESA responsiveness in the short term, but these studies were likely of insufficient duration to detect adverse events that may result from positive iron balance over the long term.
Oral ferric citrate
Following the demonstration that oral ferric citrate, a phosphate binder with absorbable iron, led to increased ferritin and TSAT levels and decreased ESA requirements in dialysis patients, a study was undertaken to assess the efficacy of ferric citrate in the treatment of iron deficiency anemia in non-dialysis patients with CKD. Patients were randomized to ferric citrate or placebo for 16 weeks. The primary outcome was an increase in Hb by >1 g/dL at any time during the 16- week period, and 52% of patients treated with ferric citrate achieved the primary end-point as compared to 19.1% receiving placebo (p<0001). Rates of adverse events were similar in the two groups.6
These findings, along with those in patients on dialysis, suggest that ferric citrate may be a safer alternative to IV iron as the former remains under the tight controls that limit iron absorption from the gastrointestinal tract and it provides smaller amounts of iron more frequently that can be more effectively transported to the erythroid marrow.
Another route to deliver iron in small doses more frequently is via the dialysate in HD patients. Ferric pyrophosphate citrate (FPC, Triferic) provides 5-7 mg to the patient during each dialysis session, equivalent to the amount of iron lost during a dialysis treatment. The goal is to maintain iron balance, not to replace accumulated iron deficits or extraordinary iron losses. FPC was shown in one randomized prospective trial to be safe and effective in decreasing IV iron (48%) and ESA (35%) requirements compared to placebo.
In another randomized study where IV iron was prohibited and ESA doses remained constant (except for a rescue pathway with study drug if targets fell out of range), patients receiving FPC maintained stable Hb levels whereas those treated with placebo sustained a 0.36 g/dL decrease in Hb levels (p=0.011). In both studies FPC did not raise serum ferritin levels, most likely because the small doses of administered iron can be accommodated by circulating transferrin. FPC is supplied in a 272 mg powder packet which is added to a the bicarbonate mix of a central delivery system and in a 5 mL ampule that is added to a 1 gallon jug of liquid bicarbonate concentrate for a single patient. FPC cannot be used in dialysate delivery systems that use solid bicarbonate concentrate.
Despite its physiologic appeal and reported efficacy/safety, FPC has not made significant inroads into the dialysis market following its approval by the FDA in early 2015. Concerns remain regarding its cost, the possible growth of siderophilic microorganisms in the dialysate lines, and the unusual design of the RCTs that led to its FDA approval. As is the case with IV iron, FPC is in the dialysis payment bundle and is not separately reimbursable.
Choice of iron supplement
The choice of IV iron agent depends upon the clinical setting. For HD patients who can receive IV iron via the extracorporeal circuit as frequently as three times per week, repeated low doses of a relatively inexpensive agent such as iron sucrose (Venofer) or sodium ferric gluconate (Ferrlecit) is preferred. These two agents have more free iron per mg than the newer agents. Since many of the immediate side effects of IV iron (nausea, vomiting, hypotension) are due to free iron, single doses of these agents are generally limited to 200-250 mg.
IV iron agents such as iron dextran (INFeD), ferumoxytol (Feraheme) and ferric carboxymaltose (Injectafer® in the US) are encapsulated iron nanoparticles that provide very little free iron and therefore larger single doses can be tolerated. This is an advantage in patients with non-dialysis CKD who must come to an infusion center to receive IV iron and in whom minimizing venipunctures is important to preserve vascular access sites. The maximum single dose of ferumoxytol is 510 mg; ferric carboxymaltose is 750 mg, and iron dextran is >1000 mg.
The disadvantage of iron nanoparticles is that the carbohydrate capsules may provoke allergic or anaphylactic reactions, so patients must be closely monitored during and after each infusion. The newest IV iron nanoparticle preparation is iron isomaltoside (Monofer®) which can be given in single doses as high as 1000 mg. Iron isomaltoside is approved in Europe and is undergoing clinical trials in the US.
Summary and conclusions
The treatment of anemia in patients with CKD has come a long way. Before 1989, dialysis patients were treated with transfusions and androgens yet still had Hct levels in the mid 20s range with characteristic pale yellow skin and poor quality of life. Replacement of the missing hormone, erythropoietin, with a pharmacologic recombinant human DNA version held the promise of solving the problem like treating a type 1 diabetic patient with recombinant human insulin. Over the ensuing 2-3 decades, it has become apparent that use of ESAs and iron needed to support erythropoiesis is not simple, although it is likely that no CKD patient or professional caring for these patients would wish to return to the pre-1989 era.
One of the lessons learned is that anemia may be an adaptive response to CKD and that increasing the Hb level to that in the normal population is unnecessary from a QOL standpoint and harmful from a vascular standpoint. Although the FDA recommends the minimum dose of ESA to avoid transfusions, this is neither realistic nor helpful guidance since unpredictable intercurrent illness may decrease Hb levels and necessitate transfusions in previously stable patients. The current “unofficial” Hb target range for patients receiving ESA therapy of 9.5-11.0 g/dL appears to be a reasonable balance between risk and benefit.
Nonetheless there remains a subset of CKD patients who are unable to achieve the target Hb level or who require large doses of ESA to do so. Furthermore, the continuous rise in mean serum ferritin levels among ESRD patients in the US has been a cause for concern, as have imaging studies demonstrating increased liver iron content in patients with serum ferritin levels exceeding 300 ng/mL.
Observational studies have demonstrated an association between higher IV iron doses and adverse outcomes, and an increase in infection-related hospitalizations among dialysis patients has also raised suspicion that liberal IV iron use as an ESA-sparing strategy in the bundled payment environment may be playing a role.
To address these concerns, evolution—not revolution—is in order. No treatment, whether it is a pharmacologic version of a complex missing hormone or a simple mineral, like iron, is not without risks. For the majority of dialysis patients who have Hb levels in the target range on modest doses of ESAs and IV iron, perhaps the best strategy is to leave well enough alone unless the HIF-PHIs have a compelling cost advantage or safety advantage based on phase 3 clinical trials.
For dialysis patients who are unable to achieve target Hb levels, require large doses of ESAs, and/or have high serum ferritin levels, the HIF-PHIs may hold promise as an alternate anemia treatment strategy. Until HIF-PHIs are available, non-IV routes or smaller more frequent of IV iron administration might be considered for patients in whom high serum ferritin levels are a concern as it is becoming increasingly clear that commonly used doses of IV iron cannot be effectively utilized for erythropoiesis and may be pro-inflammatory.
For anemic patients with non-dialysis CKD, ESA use has decreased due to FDA relabeling and increasingly restrictive reimbursement policies by prescription drug plans. The mean Hb level among incident dialysis patients has fallen as a result. The fact that ESAs must be administered parenterally is another barrier as many patients are either reluctant to self-inject (including some who already take insulin) or find it inconvenient to visit an infusion center where they can receive their ESA injections by a nurse. The advent of an orally administered class of erythropoietic agents, the HIF-PHIs, appears to fill a greater unmet need in the non-HD CKD population than it does in the HD population for this reason.
- Agarwal R, Kusek JW, Pappas MK. A randomized trial of intravenous and oral iron in chronic kidney disease. Kidney Int. 2015 Oct:88(4):905-914.
- Macdougall IC, Block AH, Carrera F, Eckardt KU, Gaillard C, Van WD et al. FIND-CKD: a randomized trial of intravenous ferric carboxymaltose versus oral iron in patients with chronic kidney disease and iron deficiency anemia. Nephrol Dial Transplant 2014:29:2075-2084.
- Fishbane S, Block GA, Loram L, Neylan J, Pergola, PE, Uhlig K et al. Effects of ferric citrate in patients with nondialysis-dependent CKD and iron deficiency anemia. J Am Soc Nephrol.