Core Curriculum

Update on Anemia in ESRD and Earlier

Stages of CKD: Core Curriculum 2018
Steven Fishbane and Bruce Spinowitz

Anemia is a frequent complication during the later stages of chronic kidney disease. When present, it Complete author and article
may cause symptoms such as fatigue and shortness of breath. The pathogenesis of anemia in chronic information provided before
references.
kidney disease is complex, but a central feature is a relative deficit of erythropoietin. New information
has elucidated the critical role of the hypoxia-sensing system in mediating erythropoietin synthesis and Am J Kidney Dis. 71(3):
release. Iron deficiency is a second important factor in the anemia of chronic kidney disease. New 423-435. Published online
January 11, 2018.
insights into the dynamics of iron metabolism have clarified the role of chronic inflammation and
hepcidin as key mediators of impaired iron utilization. In this article, we review the epidemiology, doi: 10.1053/
pathobiology, clinical evaluation, and treatment of anemia in chronic kidney disease. j.ajkd.2017.09.026
© 2017 by the National
Kidney Foundation, Inc.

Anemia remains an important complication FEATURE EDITOR:

Case: A 54-year-old man with diabetes melli- experienced by patients with kidney disease, Asghar Rastegar
tus, hypertension, and coronary artery disease is although one that is treatable. The prevalence of
being treated for chronic kidney disease (CKD).
anemia depends on its definition, but generally ADVISORY BOARD:
His estimated glomerular filtration rate has
increases in frequency and severity in the more Ursula C. Brewster
declined over the past 2 years from 40 to 14 mL/ Michael Choi
min/1.73 m2. The patient reports increased advanced stages of CKD. Studying adult patients
Ann O’Hare
fatigue and asks about the causes of his anemia. at Boston health clinics, Hsu et al published in Manoocher Soleimani
Red blood cell indexes are normal, and iron 2001 the fact that mean hematocrit (Hct) values
test results and serum folate and vitamin B12 decreased with creatinine clearance < 60 mL/ The Core Curriculum
concentrations are found to be normal. min in men and <40 mL/min in women. More aims to give trainees
severe anemia (Hct < 33%) was common among in nephrology a
Question 1: What is the most likely cause patients with estimated glomerular filtration strong knowledge
base in core topics in
or causes of the patient’s anemia? rates < 30 mL/min/1.73 m2 in women the specialty by
a) Diabetes mellitus and <20 mL/min/1.73 m2 in men. Hsu et al providing an over-
b) Relative erythropoietin deficiency published a separate study the next year that used view of the topic and
c) Iron deficiency citing key references,
the third National Health and Nutrition Exami-
d) Multiple myeloma including the founda-
nation Survey (NHANES III [1988-1994]).
tional literature that
For answer, see Appendix. Among 15,971 adults, anemia as defined by led to current clinical
hemoglobin (Hb) concentration < 12 g/dL in approaches.
Introduction men and <11 g/dL in women was more com-
mon with creatinine clearances < 70 mL/min
Kidney failure produces numerous changes that and <50 mL/min in men and women, respec-
destabilize homeostasis. An important example tively. A more significant mean decrease in Hb
is diminished erythropoiesis, with anemia being concentration of 1.0 g/dL was found for patients
a common complication of kidney disease. The with creatinine clearances < 30 mL/min.
anemia that accompanies kidney failure was Among patients with diabetes mellitus and
recognized by Sir Robert Christison in 1839, who CKD, anemia tends to be more severe and to
observed that “by far the most remarkable char- develop at an earlier point in CKD. El-Achkar
acter of the blood in the advanced stage of the et al studied 5,380 individuals who were sur-
Bright’s disease is a gradual and rapid reduction veyed as part of the National Kidney Founda-
of its colouring” and “no other natural disease tion’s Kidney Early Evaluation Program (KEEP),
came as close to hemorrhage for impoverishing a community-based screening program for pa-
the red particles of the blood.” Similarly, Richard tients at higher risk for kidney disease. Using a
Bright had noticed that patients with kidney definition of anemia as Hb concentration <
disease had paleness of the skin: “after a time, 12 g/dL in men and in women older than 50
the healthy colour of the countenance fades.”* years and <11 g/dL in women 50 years and
younger, the prevalence was greater among
*Christison R. On Granular Degeneration of the
patients with CKD with diabetes (Fig 1). In pa-
Kidnies and Its connexions With Dropsy Inflammations tients with stage 3 CKD, 22.2% of patients with
and Other Diseases. Black, Edinburgh: 1839, pp. 63-74. diabetes were anemic, increasing to 52.4% in

AJKD Vol 71 | Iss 3 | March 2018 423

 Core Curriculum

stage 4 CKD. The difference in prevalence between those with

Question 2: Which factor is most responsible for
and without diabetes was greatest in stage 3 CKD, for which
sensing cellular hypoxia?
the rate of anemia was 3 times as high among the former. a) Erythropoietin
Erythropoietin deficiency is the most common cause of b) Hepcidin
anemia in CKD, and the deficiency may be more severe in c) Hypoxia-inducible factor (HIF)-prolyl hydroxylase
patients with diabetes. In a study of 694 anemic patients, d) Fibroblast growth factor 23
Symeonidis et al observed that serum erythropoietin con- e) Ferroportin
centrations, in relation to anemia severity, were lower in
For answer, see Appendix.
patients with diabetes.

Additional Readings
► El-Achkar TM, Ohmit SE, McCullough PA, et al. Higher prevalence Physiology/Pathophysiology
of anemia with diabetes mellitus in moderate kidney insufficiency:
the Kidney Early Evaluation Program. Kidney Int. 2005;67:1483-
Background
1488. The erythropoietic system maintains homeostasis in the
► Hsu C, Bates D, Kuperman G, Curhan G. Relationship between red blood cell supply to achieve adequate tissue oxygen
hematocrit and renal function in men and women. Kidney Int. delivery. Balance is achieved by replacing erythrocytes lost
2001;59:725-731. due to senescence and bleeding (if the blood loss is not
► Hsu C, McCulloch C, Curhan G. Epidemiology of anemia asso- severe). In addition, it has long been known that hypoxia
ciated with chronic renal insufficiency among adults in the United
States: results from the Third National Health and Nutrition Ex-
stimulates new erythrocyte production. Hypoxia could be
amination Survey. J Am Soc Nephrol. 2002;13:504-510. due to pulmonary disease, reduced tissue perfusion, or
+ ESSENTIAL READING living at high altitude. The expectation that a circulating
► Symeonidis A, Kouraklis-Symeonidis A, Psiroyiannis A, et al. factor governed the erythropoietic response was followed
Inappropriately low erythropoietin response for the degree of by the discovery of erythropoietin and the cloning of its
anemia in patients with noninsulin-dependent diabetes mellitus. gene in 1985. Soon thereafter, the mechanism by which
Ann Hematol. 2006;85(2):79-85. cells sensed hypoxia and the central role of the transcrip-
tional factor, hypoxia inducible factor 1 (HIF-1), was
identified.
Tissue oxygen availability is sensed continually at the
cellular level. If hypoxia is detected, a multifaceted response
is triggered. An important component of the response is
increased production of the glycoprotein hormone eryth-
ropoietin. This 30.4-kDa molecule is the key stimulus for
erythrocyte production in mammals. It acts as a true hor-
mone in that it is produced in the kidneys and circulates and
acts at tissue receptors throughout the body, most impor-
tantly in bone marrow. Erythropoietin binds to its marrow
cell-surface receptors to stimulate erythropoiesis.

Hypoxia Sensing: The HIF System

The body’s sensing of tissue hypoxia, and thereby recog-
nition of anemia, occurs by the HIF system (Fig 2). Central
to this function are 2 proteins, HIF-α and HIF-β. HIF-α is
continually produced, but when sufficient oxygen is pre-
sent, it is rapidly “marked” (hydroxylated) for degradation
by enzymes, the HIF-prolyl hydroxylases. The prolyl
hydroxylases work as oxygen sensors because they require
oxygen as a co-substrate. After hydroxylation, HIF-α is
recognized by the von Hippel-Lindau protein, poly-
ubiquitinated, and destroyed. HIF-β is constitutively
expressed, but is not sensitive to hypoxic degradation.
When tissue hypoxia occurs, HIF-α accumulates, trans-
Figure 1. Prevalence of anemia (hemoglobin < 12 g/dL in men locates to the nucleus, forms a heterodimer with HIF-β, and
and <11 g/dL in women) in patients with or without diabetes. binds to hypoxia response elements of a large number of
Abbreviation: GFR, glomerular filtration rate. Reproduced from oxygen-sensitive genes. One of these is the erythropoietin
El-Achkar et al (Kidney Int. 2015;67:1483-1488) with permission
gene, leading to increased erythropoietin production.
of the copyright holder (International Society of Nephrology).
Numerous other genes, including those coding for enzymes

424 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

Figure 2. Under normoxic conditions, hypoxia inducible factor (HIF)-α is hydroxylated by prolyl hydroxylase domain proteins and then
undergoes proteosomal degradation. Under hypoxic conditions, HIF-α does not undergo degradation, translocates to the nucleus,
binds with HIF-β, and activates the hypoxia response element, initiating gene transcription of erythropoietin. Reproduced from West
(NEJM. 2017;376:1965-1971) with permission of the copyright holder (Massachusetts Medical Society).

and transporters involved with iron metabolism, angio- erythropoietin production and activation of iron meta-
genesis, and mitochondrial genesis, are also stimulated. bolism genes. This differential effect provides a potential
There are currently 3 forms of HIF-α that have been therapeutic opportunity to target one or the other HIF
identified, HIF-1α, HIF-2α, and HIF-3α. It is unclear how molecules. Small-molecule inhibitors of prolyl hydroxy-
diverse the differential transcriptional response to these lases, which in effect stabilize HIF-α, are currently being
forms might be. There are many shared targets of HIF-1 studied for the potential treatment of anemia in patients
and HIF-2, but certain genes are regulated more by one with CKD. By stimulating the production of erythropoietin
or the other. Most relevant to the current discussion, HIF-2 and iron-regulatory proteins, a concerted approach to
appears to play a greater role in the regulation of anemia treatment may be achievable.

AJKD Vol 71 | Iss 3 | March 2018 425

 Core Curriculum

Erythropoietin iron deficiency include occult blood loss, infection, sys-

Erythropoietin is a highly glycosylated molecule of 165 temic inflammatory conditions, surgical procedures,
amino acids. The amino acid sequence is important for venipuncture, impaired absorption secondary to elevated
receptor binding, while the 4 carbohydrate chains affect hepcidin concentrations, and in dialysis, retention of blood
the molecule’s metabolism. Based on detection of by the dialysis apparatus. It is estimated that hemodialysis
messenger RNA transcripts coding for erythropoietin, patients may lose ≥2,000 mg of iron per year as a result of
production occurs primarily in renal cortical interstitial these factors. The magnitude of iron loss in non–dialysis-
cells in adults. A secondary minor source of erythropoietin dependent CKD is unclear.
production is the liver. Interestingly, in the absence of the The deficiency of iron available for erythropoiesis in
kidneys, hepatic erythropoietin production can increase CKD is frequently compounded by a relative block in iron
significantly. absorption from the intestines and reduced iron release
Following synthesis, erythropoietin is secreted directly from storage in macrophages and the liver. This iron
into the bloodstream; there is no storage within the cells in blockade phenomenon is mediated by the master regulator
which it is produced. When in circulation, the volume of of iron homeostasis, the liver-produced circulating protein
distribution approximates that of the plasma volume space. hepcidin. Elevated hepcidin concentrations cause the
Recombinant erythropoietin, and presumably native cellular iron transporter ferroportin to be internalized into
erythropoietin, circulates with a half-life of approximately cells. The result is that iron does not enter the circulation
5 to 12 hours. through enterocytes or from storage tissues. Hepcidin
Erythropoietin quantities are traditionally expressed as concentrations increase in response to increased iron
units, with 1 unit representing the erythropoietic effect of storage in the body and decrease when iron deficiency is
stimulation with 5 μmol of cobalt chloride. Under basal present. A secondary, but important, cause of increased
conditions, serum erythropoietin concentration is steady hepcidin concentration is inflammation. This response
and present at low levels (10-30 U/L). In the presence of limits the availability of iron to microorganisms during
anemia, serum erythropoietin can increase sharply to infection. In patients with CKD, it may often be a mal-
concentrations that can exceed 1,000 U/L. It should be adaptive response. An occult state of inflammation in-
noted that erythropoietin concentrations vary according to creases the hepcidin concentration and blocks iron
the assay used. Older assays had a normal range of 10 to available for erythropoiesis. This can sometimes be
30 U/L. More recent assays have been reported to vary recognized by the fairly common finding in hemodialysis
from about 2 to 4 U/L for the lower limits to 12 to 18 U/L patients of reduced transferrin saturation (TSAT; decreased
for upper limits. circulating iron) occurring at the same time that the serum
The action of erythropoietin occurs after its interaction ferritin concentration is discordantly increased (elevated
with the erythropoietin receptor, present in highest iron storage). The frequent presence of inflammation in
quantities on the cell membranes of bone marrow CKD results in iron-restricted erythropoiesis due to hep-
erythroid precursor cells. The receptors appear to be di- cidin effects, but often superimposed on underlying true
mers, but after binding of the ligand erythropoietin, there iron deficiency.
is a conformational change in the receptor’s homodimeric
structure. This activated erythropoietin-receptor interac- Erythropoiesis in CKD
tion results in a critical cascade of signal transduction (this Among patients with CKD, there is a relative deficiency in
pathway is reviewed in depth by Richmond et al). The erythropoietin production, and this is the main reason that
result is increased cell division and enhanced survival of anemia develops. Other factors that may contribute include
red blood cell precursors. This is particularly true for iron deficiency, blood loss, inflammation, hemolysis, and
erythroid progenitors known as BFU-E and CFU-E (burst nutritional deficits (Box 1). However, the central role of
forming unit- and colony forming unit-erythroid, erythropoietin deficiency is well documented and best
respectively), which express high cell-surface concentra- demonstrated by the consistent and robust improvement in
tions of erythropoietin receptors. Remarkably, after bind- Hb concentrations after treatment with recombinant human
ing to its receptor, erythropoietin rapidly disappears, erythropoietin (rHuEPO). Surprisingly, serum erythropoi-
indicating that cellular internalization may be one mech- etin concentration is often not reduced in CKD, except in
anism of erythropoietin (and erythropoiesis) relation to the degree of anemia present. However, as kidney
downregulation. failure progresses, erythropoietin deficiency becomes more
pronounced and the decline in serum concentrations par-
Iron and Hepcidin allels kidney excretory functional loss.
Effective erythropoiesis is dependent on adequate avail- In addition to experiencing diminished erythrocyte
ability of both erythropoietin and iron. Iron deficiency is production, patients with CKD may also have shortened
common in patients with CKD, occurring in ≥50% of red blood cell survival. Ma et al recently studied 54 he-
patients with non–dialysis-dependent CKD and a greater modialysis patients and found mean red blood cell survival
percentage of patients receiving dialysis. The reasons for to be only 73.2 ± 17.8 days. Similarly, Sato et al found

426 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

► Semenza GL. Regulation of tissue perfusion in mammals by

Box 1. Common Causes of Anemia in CKD hypoxia-inducible factor 1. Exp Physiol. 2007;92(6):988-991.
+ ESSENTIAL READING
• Relative erythropoietin deficiency ► Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible
• Iron deficiency factor 1 is a basic-helix-loop-helix-PAS heterodimer
• Blood loss regulated by cellular O2 tension. Proc Natl Acad Sci U S A.
• Reduced erythrocyte survival duration 1995;92(12):5510-5514.
• Inflammation ► Wang GL, Semenza GL. Purification and characterization of
• Infection hypoxia-inducible factor 1. J Biol Chem. 1995;270:1230-1237.
• Underlying hematologic disease ► Watowich SS. The erythropoietin receptor: molecular structure
• Hyperparathyroidism (dialysis patients) and hematopoietic signaling pathways. J Investig Med.
• Hemolysis 2011;59(7):1067-1072.
• Nutritional deficits ► Wenger RH. Mammalian oxygen sensing, signalling and gene
Abbreviation: CKD, chronic kidney disease. regulation. J Exp Biol. 2000;203(pt 8):1253-1263.
► Zhu H, Bunn HF. Oxygen sensing and signaling: impact on the
regulation of physiologically important genes. Respir Physiol.
erythrocyte lifespan to be 89 ± 28 days. The cause of 1999;115(2):239-247.
shortened red blood cell survival in CKD remains incom-
pletely understood, but blood loss from the dialysis pro-
cedure is one contributing factor. Symptoms and Outcomes Related to Anemia
in CKD
Additional Readings Impact of Anemia on Symptoms/Quality of Life
► Eschbach JW, Mladenovic J, Garcia JF, Wahl PW, Adamson JW. The Anemia results in reduced oxygen delivery to the body’s
anemia of chronic renal failure in sheep. Response to erythropoietin- organs and tissues, making symptoms such as fatigue,
rich plasma in vivo. J Clin Invest. 1984;74(2):434-441. shortness of breath, insomnia, headaches, and reduced
► Fandrey J. Oxygen-dependent and tissue-specific regulation of mental acuity common. It should be noted that these
erythropoietin gene expression. Am J Physiol Regul Integr Comp
Physiol. 2004;286(6):R977-R988.
nonspecific symptoms could also be due to uremia and
► Ganz T. Hepcidin, a key regulator of iron metabolism and mediator other causes. Fatigue may develop slowly as anemia pro-
of anemia of inflammation. Blood. 2003;102(3):783-788. gresses, and patients may not be fully aware of how their
+ ESSENTIAL READING lives have been affected. Focused questioning often reveals
► Jacobs K, Shoemaker C, Rudersdorf R, et al. Isolation and char- restricted activity in compensation for reduced functional
acterization of genomic and cDNA clones of human erythropoi- capacity.
etin. Nature. 1985;313:806-810. An extension of anemia-related symptoms is the overall
► Jelkmann W. Molecular biology of erythropoietin. Intern Med.
2004;43(8):649-659. + ESSENTIAL READING
impact on patients’ quality of life. Not unexpectedly, pa-
► Kaelin WG Jr. The VHL tumor suppressor gene: insights into tients receiving dialysis often have significantly diminished
oxygen sensing and cancer. Trans Am Clin Climatol Assoc. quality of life due to existing comorbid conditions and/or
2017;128:298-307. the dialysis procedure itself. Before the availability of
► Koury ST, Koury MJ. Erythropoietin production by the kidney. rHuEPO treatment, severe anemia was common
Semin Nephrol. 1993;13(1):78-86. among patients receiving dialysis, with Hb often in the
► Lin FK, Suggs S, Lin CH, et al. Cloning and expression of the range of 6 to 9 g/dL. Nephrologists who treated patients in
human erythropoietin gene. Proc Natl Acad Sci USA.
1985;82:7580-7584.
the pre-rHuEPO era, when severe anemia was common,
► Ma J, Dou Y, Zhang H, et al. Correlation between inflammatory observed that anemia significantly degraded many patients’
biomarkers and red blood cell life span in chronic hemodialysis quality of life. However, in the pre-rHuEPO era, there
patients. Blood Purif. 2017;43(1-3):200-205. were few studies of the actual impact of anemia. One
► Nalbant D, Saleh M, Goldman FD, Widness JA, Veng-Pedersen P. exception was a study of 1,013 hemodialysis patients re-
Evidence of receptor-mediated elimination of erythropoietin by ported by Moreno et al, who used the Karnofsky scale and
analysis of erythropoietin receptor mRNA expression in Sickness Impact Profile to assess quality of life and func-
bone marrow and erythropoietin clearance during anemia.
J Pharmacol Exp Ther. 2010;333(2):528-532.
tion. The most striking finding was that 26% to 31% of
► Radtke HW, Claussner A, Erbes PM, Scheuermann EH, patients had severely diminished scores on the Karnofsky
Schoeppe W, Koch KM. Serum erythropoietin concentration in scale. Aspects of impaired function included work, recre-
chronic renal failure: relationship to degree of anemia and ation, home management, and sleep. Notably, lower Hb
excretory renal function. Blood. 1979;54(4):877-884. concentration was strongly associated with diminished
► Richmond TD, Chohan M, Barber DL. Turning cells red: signal quality of life. Like all associations, this does not prove that
transduction mediated by erythropoietin. Trends Cell Biol. anemia is the cause of diminished quality of life.
2005;15(3):146-155.
► Sato Y, Mizuguchi T, Shigenaga S, et al: Shortened red blood cell
lifespan is related to the dose of erythropoiesis-stimulating agents
Impact of Anemia on Mortality Risk
requirement in patients on hemodialysis. Ther Apher Dial. Among patients with CKD, for whom cardiovascular co-
2012;16:522-528. morbid conditions are common, it is a reasonable

AJKD Vol 71 | Iss 3 | March 2018 427

 Core Curriculum

hypothesis that reduced carriage of oxygen to tissues could Other Effects of Anemia in CKD
contribute to mortality risk. Observational studies have There has been interest in whether anemia might hasten
consistently demonstrated a strong association between the progression of kidney disease by depriving the kidneys
lower Hb concentrations in dialysis patients and increased of needed oxygen. An analysis of the RENAAL (Reduction
risk for death. Ma et al performed a study of 96,369 he- of Endpoints in NIDDM With the Angiotensin II Antago-
modialysis patients, finding that lower baseline Hct was nist Losartan) trial sought to address the issue. The primary
associated with higher mortality risk. For Hct < 27%, the purpose was to evaluate losartan compared to conventional
relative risk for death was 1.33 compared with patients antihypertensive treatment for renal and other outcomes in
with Hct of 30% to 33%. Sandgren et al studied mortality patients with type 2 diabetes. Mohanram et al, in a post
risk in Medicare beneficiaries and found that patients with hoc analysis, found that among 1,513 participants, initial
CKD and anemia have a 270% increase in 2-year mortality Hb concentration was a significant predictor of time to
compared with patients without these diagnoses. Several dialysis therapy and doubling of serum creatinine con-
other studies also concluded that lower Hb concentrations centration. For every 1-g/dL decrease Hb concentration,
in CKD are associated with increased risk for mortality. It is there was 11% greater chance of these dual renal out-
important to note that in contrast to these observational comes. However, because anemia becomes more prevalent
studies, interventional trials have failed to find that as kidney function declines, it is unclear which is the cause
rHuEPO treatment, generally to near-normal Hb concen- and which is the effect: does anemia hasten kidney func-
trations, improves mortality risk. Rather, the trials tion decline or does anemia simply become more frequent
designed to increase Hb to near-normal concentrations with diminished kidney function? Although Mohanram
(>13 g/dL) have found increased risk for mortality and and colleagues adjusted for a number of potential con-
cardiovascular complications. This leaves the question of founding variables, there remained a high risk for residual
causality in the relationship of anemia to increased mor- confounding. In contrast to the RENAAL study, TREAT
tality risk unresolved. (Trial of Darbepoetin Alfa in Type 2 Diabetes and Chronic
Kidney Disease) found different results. More than 4,000
Impact of Anemia on Left Ventricular Hypertrophy patients with type 2 diabetes were randomly assigned to
There is a great burden of cardiac disease in patients with treatment with darbepoetin alfa or placebo. Patients in the
CKD. It is highly plausible that anemia, by increasing the darbepoetin group were treated to a target Hb concentra-
work of the heart and reducing tissue oxygen delivery, tion of 13 g/dL; in the placebo group, treatment was
could exacerbate cardiac injury. The most definitive rela- reserved for patients with worsening anemia. More than
tionship between anemia and heart disease is the devel- 600 patients in the study progressed to end-stage renal
opment of left ventricular hypertrophy (LVH). For disease. Importantly, there was no significant reduction in
example, Foley et al found LVH to be present in 73.9% of the rate of kidney failure in the darbepoetin alfa group
patients new to dialysis therapy. This is important because compared to placebo. A meta-analysis of erythropoiesis-
LVH is strongly and independently associated with stimulating agent (ESA) intervention trials in CKD was
increased mortality risk. In addition, the stiffened left reported by Elliott et al; these investigators found no
ventricle is functionally sensitive to blood volume changes. benefit to ESA treatment with respect to kidney disease
This results in heightened risk in dialysis patients for both progression.
pulmonary edema and intradialytic hypotension, depend- Although there are other adverse effects of anemia, a
ing on volume status. more detailed discussion is beyond the scope of the current
The relationship between anemia and LVH in CKD has article. We conclude our discussion of the impact of
been studied by several investigators. Silberberg et al anemia in kidney disease by reminding the reader that in
analyzed a cohort of 78 hemodialysis patients. A strong our view, the clearest detrimental effect is the diminution
association was demonstrated between worsening anemia of life experience due to anemia-related symptoms such as
and LVH. Among patients in the lowest quartile of Hb fatigue.
concentrations, mean left ventricular mass index was
w30% higher than in the highest quartile. Levin et al Additional Readings
studied 175 patients with non–dialysis-dependent CKD. ► Association between recombinant human erythropoietin and
LVH was found in 38.9% of patients. There was an as- quality of life and exercise capacity of patients receiving haemo-
sociation with anemia; each 1-g/dL lower Hb concen- dialysis. Canadian Erythropoietin Study Group. BMJ.
tration was associated with 6% greater LVH. Early 1990;300(6724):573-578.
rHuEPO intervention studies, often underpowered, hin- ► Besarab A, Bolton WK, Browne JK, et al. The effects of normal as
compared with low hematocrit values in patients with cardiac
ted at the potential of treatment to induce regression of
disease who are receiving hemodialysis and epoetin. N Engl J
LVH. In contrast, larger well-powered randomized Med. 1998;339(9):584-590.
controlled trials have generally found no benefit to ane- ► Delano BG. Improvements in quality of life following treatment
mia correction with rHuEPO for improvement or slowing with r-HuEPO in anemic hemodialysis patients. Am J Kidney Dis.
of LVH. 1989;14(2)(suppl 1):14-18.

428 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

► Drüeke TB, Locatelli F, Clyne N, et al; CREATE Investigators. The initial evaluation of anemia in CKD involves a
Normalization of hemoglobin level in patients with chronic targeted clinical assessment. Although relative erythro-
kidney disease and anemia. N Engl J Med. 2006;355(20): poietin deficiency will usually be the most important
2071-2084.
causal element, other contributing factors should be
► Elliott S, Tomita D, Endre Z. Erythropoiesis stimulating agents and
reno-protection: a meta-analysis. BMC Nephrol. 2017;18(1):14.
considered. A focused history, physical examination,
► Foley RN, Parfrey PS, Morgan J, et al. Effect of hemoglobin levels and laboratory review should be conducted (Box 2). The
in hemodialysis patients with asymptomatic cardiomyopathy. goal is similar to anemia evaluation in patients without
Kidney Int. 2000;58(3):1325-1335. kidney disease: a broad search for signs of reduced
► Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalent erythrocyte production, hemolysis, sequestration, and
left ventricular hypertrophy in the predialysis population: identifying bleeding. In patients with CKD, special emphasis should
opportunities for intervention. Am J Kidney Dis. 1996;27(3):347-354.
be placed on testing for iron deficiency and occult blood
► Ma JZ, Ebben J, Xia H, Collins AJ. Hematocrit level and associated
mortality in hemodialysis patients. J Am Soc Nephrol.
loss because of their frequent occurrence in this
1999;10(3):610-619. population.
► Moreno F, Lopez Gomez JM, Sanz-Guajardo D, Jofre R, Valder- We would suggest that blood testing include a
rabano F. Quality of life in dialysis patients. A Spanish multicentre complete blood cell count with red blood cell indexes.
study. Spanish Cooperative Renal Patients Quality of Life Study Abnormal leukocyte or platelet counts should suggest
Group. Nephrol Dial Transplant. 1996;11(suppl 2):125-129. the possibility of a more generalized bone marrow
► Parfrey PS, Lauve M, Latremouille-Viau D, Lefebvre P. Erythropoietin
disorder. Erythrocytic microcytosis or macrocytosis
therapy and left ventricular mass index in CKD and ESRD patients: a
meta-analysis. Clin J Am Soc Nephrol. 2009;4:755-762.
helps point to factors other than erythropoietin defi-
► Pfeffer MA, Burdmann EA, Chen CY, et al; TREAT Investigators. A ciency that may be contributing to anemia. A reticulo-
trial of darbepoetin alfa in type 2 diabetes and chronic kidney cyte count may help to better understand the
disease. N Engl J Med. 2009;361(21):2019-2032. + ESSENTIAL appropriateness of the bone marrow’s response to ane-
READING mia. One test that is not advised in the anemia evalua-
► Sarnak MJ, Levey AS. Cardiovascular disease and chronic tion is measurement of serum erythropoietin. Because
renal disease: a new paradigm. Am J Kidney Dis.
the deficiency is relative, test results rarely help in the
2000;35(4)(suppl 1):S117-S131.
► Singh AK, Szczech L, Tang KL, et al; CHOIR Investigators.
evaluation or subsequent management.
Correction of anemia with epoetin alfa in chronic kidney disease. Other causes for anemia may become apparent during
N Engl J Med. 2006;355(20):2085-2098. evaluation. Symptoms of bone pain and a low anion gap
may suggest the need to evaluate for a dysproteinemic
state. Heavy menstrual bleeding may suggest gynecologic
Case: A 76-year-old woman with diabetes mellitus and
pathology such as fibroid tumors of the uterus. A diet low
stage 4 CKD is evaluated for progressive anemia. Blood test
results are notable for the following values: serum potas-
in leafy vegetables or high mean erythrocyte cell volume
sium, 5.5 mEq/L; serum creatinine, 3.2 mg/dL; and serum could suggest folic acid deficiency. Progressive weight loss
calcium, 12.6 mg/dL. Hb concentration is 7.5 g/dL with might indicate a need to evaluate for malignancy. An
normal erythrocyte indexes. Serum ferritin concentration is enlarged spleen found on physical examination may sug-
358 ng/mL and TSAT is 20.2%. gest a primary hematologic disease.
Because of the frequency of iron deficiency in this
Question 3: In addition to erythropoietin deficiency, population, iron indexes should be assessed in all patients
what other cause of anemia is important to exclude in with CKD and anemia. The most commonly used iron tests
this case? are measurement of serum ferritin and TSAT. The former
a) Iron deficiency reflects primarily on the body’s stores of iron. In patients
b) Hyperparathyroidism with CKD, serum ferritin concentrations are often
c) Malignancy
increased independent of iron status by the presence of
d) Hypothyroidism
e) Endocarditis
For answer, see Appendix. Box 2. Evaluation of Anemia in CKD

• Focused history and physical examination

• Blood testing
Diagnostic Evaluation > Chemistries
The World Health Organization defines anemia as Hb > Complete blood cell count (including red blood cell

concentration < 12.0 g/dL in women and <13.0 g/dL in indexes)

men. In our opinion, in non–dialysis-dependent CKD, the > Reticulocyte count
> Serum ferritin
nephrologist should take an active role in management
> Transferrin saturation
when Hb concentration decreases to <12 g/dL. The diag-
> Folic acid
nosis of anemia does not imply that rHuEPO treatment > Vitamin B12
should necessarily be initiated, but rather that evaluation Abbreviation: CKD, chronic kidney disease.
and ongoing monitoring are necessary.

AJKD Vol 71 | Iss 3 | March 2018 429

 Core Curriculum

inflammation, requiring a more holistic approach to

Case: A 28-year-old woman with systemic lupus initiates
diagnosis. TSAT is a measure of circulating iron. Plasma
hemodialysis treatment after a long course of various
total iron bound to transferrin varies from 2 to 4 mg, immunoregulatory treatments. As she begins dialysis ther-
whereas the daily need for iron in bone marrow is much apy, epoetin alfa treatment is started as well, with an Hb
higher, necessitating rapid turnover of iron from tissue concentration of 7.1 g/dL. The Hb concentration increases
stores and transport to the bone marrow. over 2 months to 9.8 g/dL but fails to increase further
In the general population, either serum ferritin concen- despite subsequent increases in her epoetin dose. She
tration < 30 ng/mL or TSAT < 15% are strongly suggestive reports continued fatigue. Her lupus is inactive by symp-
of iron deficiency. However, both iron tests tend to be toms and serologic test results. Erythrocyte indexes are
inaccurate for the diagnosis of iron deficiency in patients normal, serum ferritin concentration is 26 ng/mL, and TSAT
with CKD. The results of both tests reflect not only iron is 13.7%.
status, but also other factors that limit the ability to gauge
iron status. In hemodialysis patients, there is no level of Question 4: What would be the next step in anemia
treatment?
either test that optimizes both sensitivity and specificity. For
a) Increase lupus treatment drugs
example, TSAT < 21% comes closest to reasonable accuracy, b) Increase dialysis time
with sensitivity of 81% for predicting response to intrave- c) Increase the epoetin dose further
nous (IV) iron, but with specificity of only 63%. Using the d) Change to peritoneal dialysis
tests together and review of trends may help improve e) Treat with IV iron
diagnostic utility. Among hemodialysis patients, iron defi-
ciency is likely with serum ferritin concentration < 300 ng/ For answer, see Appendix.
ml or TSAT < 20%. In non–dialysis-dependent CKD, iron
deficiency is probably present with serum ferritin concen-
trations < 100 ng/mL or TSAT < 15%. However, it must be Treatment of Anemia in CKD
recognized that iron deficiency may still be present in both Introduction
populations with considerably higher values of either test. The cornerstone of anemia treatment in CKD is rHuEPO
Other tests have been used in the assessment of iron therapy. Up to this point in the article, we have used the
status. Reticulocyte Hb content (CHr) is a test that leverages term rHuEPO as an indication of treatment with erythro-
the short circulating life of reticulocytes. Because the cells poietin analogues. At the time of writing, new drugs that are
are detectable in the bloodstream for only approximately 24 not erythropoietin analogues are in development to stimu-
hours, their Hb/iron content is a “snapshot” reflection of late erythropoiesis. The broader term that encompasses all
very recent iron status. Studies have found good diagnostic these drugs is ESAs. In this section on anemia treatment, we
utility, but the lack of widespread availability has limited its use the broader term ESA as the term for all drugs, including
use. Another test, the percentage of hypochromic red blood erythropoietin analogues that stimulate erythropoiesis.
cells, has been demonstrated to have good utility as long as Treatment with ESAs should ideally be withheld until a
blood is processed fairly soon after sampling. This test, like preliminary anemia evaluation has been completed. There
CHr, has failed to enter mainstream clinical use. is little to be gained from treatment with these drugs in a
patient who has iron deficiency or occult bleeding. If
Additional Readings emergency anemia treatment is required in a patient who
► Fishbane S, Galgano C, Langley RC Jr, Canfield W, Maesaka JK. is actively bleeding or has a rapid decrease in Hb con-
Reticulocyte hemoglobin content in the evaluation of iron status of centration, blood transfusion should be considered.
hemodialysis patients. Kidney Int. 1997;52(1):217-222.
► Fishbane S, Kowalski EA, Imbriano LJ, Maesaka JK. The evaluation Iron Treatment
of iron status in hemodialysis patients. J Am Soc Nephrol. Before initiating ESA treatment, iron status should be
1996;7(12):2654-2657. assessed, as discussed. Iron treatment may be required
► Fishbane S, Pollack S, Feldman HI, Joffe MM. Iron indices in
chronic kidney disease in the National Health and Nutritional Ex-
before, during, or at times instead of ESA therapy. KDIGO
amination Survey 1988-2004. Clin J Am Soc Nephrol. (Kidney Disease: Improving Global Outcomes) recom-
2009;4(1):57-61. mendations for when to initiate iron treatment in CKD are
► Macdougall IC. What is the most appropriate strategy to monitor listed in Box 3. There is a wide choice of both oral and IV
functional iron deficiency in the dialysed patient on rhEPO ther- iron agents available for treatment. Oral agents are
apy? Merits of percentage hypochromic red cells as a marker of generally ineffective for hemodialysis patients and only
functional iron deficiency. Nephrol Dial Transplant. modestly effective in non–dialysis-dependent CKD. One
1998;13(4):847-849.
► Stancu S, Stanciu A, Zugravu A, et al. Bone marrow iron, iron
exception is the oral iron phosphate binder ferric citrate,
indices, and the response to intravenous iron in patients which is highly efficacious as an iron supplement in both
with non-dialysis-dependent CKD. Am J Kidney Dis. patient populations. It is unclear why this oral iron agent is
2010;55(4):639-647. able to deliver iron to a greater extent than other forms of

430 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

Box 3. KDIGO Guideline Recommendations for When to Treat With Iron in CKD
2.1.2: For adult CKD patients with anemia not on iron or ESA therapy we suggest a trial of IV iron (or in CKD ND patients
alternatively a 1–3 month trial of oral iron therapy) if (2C):
• an increase in Hb concentration without starting ESA treatment is desired* and
• TSAT is ≤30% and ferritin is ≤500 ng/ml (500 μg/l)
2.1.3: For adult CKD patients on ESA therapy who are not receiving iron supplementation, we suggest a trial of IV iron (or in CKD
ND patients alternatively a 1–3 month trial of oral iron therapy) if (2C):
• an increase in Hb concentration** or a decrease in ESA dose is desired*** and
• TSAT is ≤30% and ferritin is ≤500 ng/ml (500 μg/l)
*Based on patient symptoms and overall clinical goals, including avoidance of transfusion, improvement in anemia-related symp-
toms, and after exclusion of active infection.
**Consistent with Recommendations #3.4.2 and 3.4.3.
***Based on patient symptoms and overall clinical goals including avoidance of transfusion and improvement in anemia-related
symptoms, and after exclusion of active infection and other causes of ESA hyporesponsiveness.
Abbreviations: CKD, chronic kidney disease; ESA, erythropoiesis-stimulating agent; IV, intravenous; Hb, hemoglobin; KDIGO, Kidney Disease: Improving Global Out-
comes; ND, nondialysis; TSAT, transferrin saturation.
Guideline recommendations are ©KDIGO. Reproduced with permission of KDIGO from Kidney Disease: Improving Global Outcomes (KDIGO) Anemia Work Group.
KDIGO Clinical Practice Guideline for Anemia in Chronic Kidney Disease. Kidney inter., Suppl. 2012;2:279–335.

oral iron and without an apparently greater burden of side may be initiated on oral iron agents instead of or before
effects. In dialysis patients, when used as a phosphate treatment with IV iron. When IV iron is used in these
binder, the drug will result in significant increases in iron patients, there should be appropriate observation for the
parameters and potentially improves ESA response. development of hypotension or hypersensitivity reactions.
In contrast to oral forms of iron, IV iron is generally In addition, in both these patient populations, care should
highly efficacious. It is a standard-of-care treatment for be taken to preserve veins that may later be needed to
hemodialysis patients. Among these patients, the most create vascular access for hemodialysis.
commonly used agent in the United States is iron sucrose, IV iron treatment is highly efficacious, but at the time of
usually administered at 50 to 100 mg/wk as needed. Other this writing, its safety has not been fully evaluated. As a
forms of IV iron are available, including iron dextran, result, clinicians are not able to fully consider the balance
ferric gluconate, ferumoxytol, ferric carboxymaltose, and of benefit and risk when making treatment decisions. The
iron isomaltoside. The great preponderance of iron sucrose KDIGO recommendations help ensure effective treatment
use in the United States probably reflects market forces while avoiding potential risks (Box 3). One safety concern
more than any specific drug properties. All these drugs are has been the possibility that IV iron could make iron more
effective, and all probably carry some minor risk for hy- available to bacteria and other microorganisms. Because of
potension or hypersensitivity reactions. The major differ- this, we would suggest that IV iron not be administered
ence is that a larger amount of iron can be administered at during acute infections, especially when bacteremia is
a single administration with iron dextran, ferumoxytol, present. Treatment of iron deficiency can always be safely
ferric carboxymaltose, and iron isomaltoside compared postponed until after an infection is fully treated.
with iron sucrose and ferric gluconate.
In hemodialysis, access to the circulation is simple, by ESA Treatment
patients’ dialysis lines, and the available drugs are well toler- Treatment with ESAs is the hallmark of anemia therapy
ated. There are 3 strategies for IV iron administration in these in CKD. At the time of writing, all available ESAs are
patients. The first strategy is a repletion approach. Testing is analogues of erythropoietin (Table 1). The first avail-
performed for iron deficiency every 1 to 3 months. If iron able was epoetin alfa, approved by the US Food and
deficiency is detected, a short course of IV iron is adminis- Drug Administration (FDA) in 1989 (only 5 years after
tered. A typical course would be 1,000 mg of iron sucrose or cloning of the erythropoietin gene). Epoetin alfa is
ferric gluconate over 10 to 12 dialysis treatments. A second similar to native erythropoietin and is produced by
approach might be termed maintenance therapy. In antici- recombinant DNA technology in massive cell cultures.
pation of blood loss, a weekly dose of IV iron is administered. The second ESA developed was darbepoetin alfa, which
A third method, administering large amounts of iron in a differs from native erythropoietin by 5 amino acids and
single dose, is not widely used in hemodialysis. There is no by additional carbohydrate content that changes the
great evidence to support one IV iron method over the other. pharmacokinetics, resulting in an extended serum half-
Given the marked difference between the mentioned uses of life (approximately 2-3 times longer than epoetin alfa).
IV iron, other strategies could be used as well. A third ESA is methoxy polyethylene glycol-epoetin
In patients with non–dialysis-dependent CKD or those beta, which has a significantly increased serum half-
treated with peritoneal dialysis, IV iron use is complicated life (Table 1). Less-frequent dosing is the expected
by the need to establish IV access. Because of this, patients advantage of an ESA with a longer half-life. Although

AJKD Vol 71 | Iss 3 | March 2018 431

 Core Curriculum

Table 1. Plasma Half-Lives of Erythropoietin Analogues with transfusion. Moreover, immune sensitization may
Half-Life, h limit the ability of patients to undergo eventual kidney
transplantation. The ability of ESA treatment to reduce
Intravenous Subcutaneous
Administration Administration transfusions has been well demonstrated by placebo-
Epoetin alfa 6.8 19.4 controlled randomized trials. From the initial clinical
Darbepoetin alfa 25.3 48.8 study of rHuEPO in 1987, Eschbach et al noted, “Of 18
Methoxy polyethylene 130 133 patients receiving effective doses of recombinant human
glycol-epoetin beta erythropoietin, 12 who had required transfusions no
Note: Epoetin alfa, US trademarks, Epogen, Procrit; darbepoetin alfa, longer needed them.”(p. 73) Many subsequent studies have
US trademark, Aranesp; and methoxy polyethylene glycol-epoetin beta, US trade- demonstrated a transfusion-sparing effect. Interestingly,
mark, Mircera.
during a more recent period when mean Hb concentra-
tions decreased in dialysis patients (2010-2012), the rate
this benefit is most evident in patients with non– of monthly blood transfusions increased from 9% to 11%.
dialysis-dependent CKD with anemia, it could have As discussed in an earlier section, the symptoms of anemia
advantages in the dialysis facility, specifically with can improve markedly with ESA treatment. This was easiest to
respect to freeing nurses to dedicate more time to observe in the initial clinical trials of epoetin, in which patients
patient assessment and education. However, the serum entered studies with very low Hb concentrations,
half-life of ESAs may not always correlate with required usually <8 g/dL. For example, Delano et al studied 37 he-
dose frequency. A recent Cochrane review of 14 ran- modialysis patients treated with ESAs, with the mean Hct
domized controlled trials concluded that short-acting increasing from 19.8% to 31.5%. Eighty-four percent of pa-
ESAs for patients with non–dialysis-dependent CKD tients experienced an improved sense of well-being. In
given at higher doses for extended intervals (2 or 4 addition, improvements were found in appetite (81%), sexual
weeks) were noninferior to more frequent dosing in- function (62%), socializing (70%), and sleep (68%). In
tervals in achieving and maintaining Hb concentrations. another study, The Canadian Erythropoietin Study Group
Recommendations for initiation and maintenance randomly assigned 118 patients to treatment with placebo,
therapy with ESAs from the KDIGO guideline are provided low-dose ESA, or high-dose ESA. After 6 months of follow-up,
in Box 4. Initial suggested starting doses are listed in mean Hb concentration in the placebo group was 7.4 g/dL. In
Table 2. These doses may not correlate exactly with FDA ESA-treated groups, there was an improvement of 3 to 4 g/dL.
prescribing instructions; they reflect the opinions of the ESA-treated patients experienced significant reduction in fa-
authors. Before starting treatment, iron status should be tigue and improvement in several facets of quality of life. The
optimized and blood pressure optimally controlled. Cooperative Multicenter EPO Clinical Trial Group studied 333
After initiating treatment with an ESA, Hb concentration hemodialysis patients with baseline Hcts of 22.2%. The first
should be measured weekly until Hb stability and goals are follow-up for symptoms and functional capacity was assessed
achieved. A reasonable goal is an increase of 1 g/dL in Hb at a mean of 4.4 months, at which time mean Hct had
concentration within the first month of treatment. If the increased to 35%. The number of patients free of physical
increase in Hb concentration is excessive (>1 g/dL over 2 limitations increased from 27% to 47.6%. The number of
weeks), the ESA dose should be reduced by 25% to 50%. patients reporting that they were very active increased from
As Hb concentration increases, blood pressure response 19.8% to 35.5%. At baseline, 26.7% of patients felt full of
should be monitored because blood pressure will increase energy, which increased to 52.7% at follow-up.
during treatment in some patients. We recommend In contrast to partial correction of severe anemia, there
checking iron status monthly during initial ESA treatment. is less clarity as to whether ESA treatment to full Hb
As the Hb concentration increases, a large amount of iron concentration normalization results in further increments
is transferred from storage tissues to the developing in quality of life. More importantly, both cardiovascular
erythron, and iron deficiency is frequently induced. This and thromboembolic safety risks have become apparent.
may limit the effectiveness of ESA treatment. There are 4 major studies that have helped definitively
The target Hb concentration during ESA treatment is demonstrate the general lack of benefit and increased risk
controversial. The target should reflect a balancing of benefits with full Hb concentration normalization.
and risks as applied to the individual patient. Some consid- Besarab et al performed a randomized controlled trial of
erations and the supporting key studies are discussed next. 1,233 hemodialysis patients. Epoetin alfa was used in both
The benefits of ESA treatment are clear; avoidance of groups to target an Hct of 30% or 42% for 29 months. At
blood transfusions and improvement in anemia-related the end of the study, there was no clear benefit of the
symptoms. Before the availability of ESAs, blood trans- higher Hct target. However, there was a strong trend
fusions were frequently required by dialysis patients. toward increased mortality risk.
Before the availability of ESAs, Hb concentrations of he- Drueke et al randomly assigned 603 patients with
modialysis patients were often <8 g/dL. Although the non–dialysis-dependent CKD to ESA treatment to an Hb goal
blood supply in 2017 is generally considered safe, there of 10.5 to 11.5 g/dL or 13.0 to 15.0 g/dL with 3 years of
are still certain risks and resource considerations associated follow-up. In this study, there was a finding of a positive

432 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

Box 4. KDIGO Guideline Recommendations for Initiation and Maintenance of ESAs

3.1: Address all correctable causes of anemia (including iron deficiency and inflammatory states) prior to initiation of ESA therapy.
(Not Graded)
3.2: In initiating and maintaining ESA therapy, we recommend balancing the potential benefits of reducing blood transfusions and
anemia-related symptoms against the risks of harm in individual patients (e.g., stroke, vascular access loss, hypertension). (1B)
3.3: We recommend using ESA therapy with great caution, if at all, in CKD patients with active malignancy—in particular when cure
is the anticipated outcome—(1B), a history of stroke (1B), or a history of malignancy (2C).
3.4.1: For adult CKD ND patients with Hb concentration ≥ 10.0 g/dl (≥100 g/l), we suggest that ESA therapy not be initiated. (2D)
3.4.2: For adult CKD ND patients with Hb concentration <10.0 g/dl (<100 g/l) we suggest that the decision whether to initiate ESA
therapy be individualized based on the rate of fall of Hb concentration, prior response to iron therapy, the risk of needing a
transfusion, the risks related to ESA therapy and the presence of symptoms attributable to anemia. (2C)
3.4.3: For adult CKD 5D patients, we suggest that ESA therapy be used to avoid having the Hb concentration fall below 9.0 g/dl
(90 g/l) by starting ESA therapy when the hemoglobin is between 9.0–10.0 g/dl (90–100 g/l). (2B)
3.4.4: Individualization of therapy is reasonable as some patients may have improvements in quality of life at higher Hb concentration
and ESA therapy may be started above 10.0 g/dl (100 g/l). (Not Graded)
3.4.5: For all pediatric CKD patients, we suggest that the selection of Hb concentration at which ESA therapy is initiated in the
individual patient includes consideration of potential benefits (e.g., improvement in quality of life, school attendance/performance,
and avoidance of transfusion) and potential harms. (2D)
ESA MAINTENANCE THERAPY
3.5.1: In general, we suggest that ESAs not be used to maintain Hb concentration above 11.5 g/dl (115 g/l) in adult patients with
CKD. (2C)
3.5.2: Individualization of therapy will be necessary as some patients may have improvements in quality of life at Hb concentration
above 11.5 g/dl (115 g/l) and will be prepared to accept the risks. (Not Graded)
3.6: In all adult patients, we recommend that ESAs not be used to intentionally increase the Hb concentration above 13 g/dl
(130 g/l). (1A)
3.7: In all pediatric CKD patients receiving ESA therapy, we suggest that the selected Hb concentration be in the range of 11.0 to
12.0 g/dl (110 to 120 g/l). (2D)
Abbreviations: CKD, chronic kidney disease; ESA, erythropoiesis-stimulating agent; Hb, hemoglobin; KDIGO, Kidney Disease: Improving Global Outcomes; ND,
nondialysis.
Guideline recommendations are ©KDIGO. Reproduced with permission of KDIGO from Kidney Disease: Improving Global Outcomes (KDIGO) Anemia Work Group.
KDIGO Clinical Practice Guideline for Anemia in Chronic Kidney Disease. Kidney inter., Suppl. 2012;2:279–335.

quality-of-life benefit. Patients randomly assigned to the Pfeffer et al reported on a study of 4,038 patients with
higher Hb target experienced improved general health and type 2 diabetes mellitus and non–dialysis-dependent CKD.
physical function scores. However, as in the Besarab et al Participants were randomly assigned to darbepoetin alfa
study, there was a non–statistically significant trend to treatment to an Hb target of 13 g/dL (actual achieved
increased risk for death in the higher Hb target group. median Hb was 12.5 g/dL) or placebo, with rescue dar-
Singh et al studied 1,432 patients with non–dialysis- bepoetin treatment only as required. With a mean of 29.1
dependent CKD, treated with epoetin alfa to a target Hb months of follow-up, no substantial benefit of the treated
concentration of 13.5 g/dL or 11.3 g/dL. Median patient group was found for any end point. The major safety risk
exposure was 16 months. No clinical benefits were identified was increased risk for stroke in the darbepoetin-
demonstrable for the higher Hb group. However, there treated group (hazard ratio, 1.92; 95% confidence interval,
were significant safety signals, including increased risk for 1.38-2.68). In addition, as in many of these studies, there
cardiovascular events in a composite end point in the were increases in thromboembolic complications.
higher Hb group. The adverse result was driven by The reasons for increased risk with ESA treatment to normal
increased risk for death and increased risk for hospitali- Hb targets are unclear. Greater blood viscosity at higher Hb
zations for congestive heart failure. concentrations may contribute by increasing vascular endo-
thelial wall stress. In contrast, it may be that risk is not due to
Table 2. Suggested Starting Dose for ESAs
the higher achieved Hb concentration itself, but instead to the
Hemodialysis NDD-CKD very high ESAs doses required to normalize Hb concentra-
Epoetin alfa 50-100 U/kg, 50-100 U/kg tions. It is possible that these supraphysiologic erythropoietin
3×/wk every 1-2 wk concentrations might have detrimental off-target effects.
Darbepoetin 0.45 μg/kg 0.45 μg/kg Along these lines, analyses of the higher Hb target studies have
alfa every wk every 2- 4 wk
Methoxy polyethylene 0.6 μg/kg, 0.6 μg/kg
consistently had an intriguing finding. Although there was
glycol-epoetin beta every 2 wk every 2-4 wk increased risk in the higher Hb target groups, counterintui-
Note: Epoetin alfa, US trademarks, Epogen, Procrit; darbepoetin alfa, US trade- tively, patients who achieved higher Hb concentrations in the
mark, Aranesp; and methoxy polyethylene glycol-epoetin beta, US trademark, studies had better outcomes. It was higher ESA doses that
Mircera.
Abbreviations: CKD, chronic kidney disease; ESA, erythropoiesis-stimulating correlated most strongly with adverse outcomes. This sug-
agent; NDD, non–dialysis-dependent. gests, but does not prove, that high ESA doses may be toxic.

AJKD Vol 71 | Iss 3 | March 2018 433

 Core Curriculum

► Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson

JW. Correction of the anemia of end-stage renal disease
with recombinant human erythropoietin. Results of a com-
bined phase I and II clinical trial. N Engl J Med.
1987;316(2):73-78.
► Evans RW, Rader B, Manninen DL. The quality of life of hemodi-
alysis recipients treated with recombinant human erythropoietin.
Cooperative Multicenter EPO Clinical Trial Group. JAMA.
1990;263(6):825-830.
► Fishbane S, Block GA, Loram L, et al. Effects of ferric citrate
in patients with nondialysis-dependent CKD and iron
Figure 3. Target hemoglobin (Hb) concentration during deficiency anemia. J Am Soc Nephrol. 2017;28(6):1851-
erythropoiesis-stimulating agent treatment is an important deter- 1858.
minant of the balance of benefits and risks. At lower targets, ► Hahn D, Esezobor CI, Elserafy N, Webster AC, Hodson EM.
there are clear benefits with little evidence of risk; at higher Hb Short-acting erythropoiesis-stimulating agents for anemia in pre-
targets there is increased risk with no accompanying benefit. dialysis patients. Cochrane Database Syst Rev.
2017;1:CD011690.
► Macdougall IC, Gray SJ, Elston O, et al. Pharmacokinetics of
The FDA recommends using the lowest ESA dose required to
novel erythropoiesis stimulating protein compared with epoetin
achieve therapeutic benefit. alfa in dialysis patients. J Am Soc Nephrol. 1999;10(11):2392-
Taken together, all acquired knowledge on ESA treatment 2395.
in patients with CKD indicates clear benefits for patients ► Szczech LA, Barnhart HX, Inrig JK, et al. Secondary analysis of
with baseline Hb concentrations < 10 g/dL and moderate the CHOIR trial epoetin-alpha dose and achieved hemoglobin
treatment goals. In contrast, risks are present with extended outcomes. Kidney Int. 2008;74(6):791-798.
treatment to Hb targets > 13 g/dL. It is less clear regarding
the relative balance of benefit and risk in patients treated to Article Information
Hb targets between 10 and 13 g/dL (Fig 3). It is probably
Authors’ Full Names and Academic Degrees: Steven Fishbane,
true that as Hb concentration exceeds 11 g/dL and ap- MD, and Bruce Spinowitz, MD.
proaches 13 g/dL, the potential benefits of treatment
Authors’ Affiliations: Division of Nephrology, Department of
diminish and risks increase. Our recommendation is to Medicine, Hofstra Northwell School of Medicine, Great Neck (SF);
target an Hb concentration of 10 to 11.5 g/dL, seeking to and Division of Nephrology, New York-Presbyterian/Queens,
provide patients with the benefits of therapy while dimin- Flushing, NY (BS).
ishing potential risks. As in all medical treatment, clinical Address for Correspondence: Steven Fishbane, MD, 100
judgment is needed. Individualization of the Hb target may Community Dr, Great Neck, NY 11021. E-mail: sfishbane@
be considered based on patient characteristics. For example, northwell.edu
if a patient is asymptomatic with an active life and ESA- Support: Dr Fishbane receives support from a Centers for Medicare
& Medicaid Innovation Health Care Innovations Award.
treated Hb concentration is 10 g/dL, there is little reason
to increase the Hb concentration further. In contrast, for a Financial Disclosure: Dr Fishbane has received research support
from Astra Zeneca, Corvidia, Fibrogen, ZS Pharma, Amgen, and
younger active patient experiencing continued fatigue with Akebia and consulting fees from Astra Zeneca, Pfizer, Keryx, and
an Hb concentration of 10.5 g/dL, treatment to a higher Hb Akebia. Dr Spinowitz has received research support from AMAG,
concentration may be tried. Akebia, FibroGen, Corvidia, and ZS Pharma and consulting fees
from Akebia, ZS Pharma, and Fresenius.

Additional Readings Peer Review: Received June 14, 2017 in response to an invitation
from the journal. Evaluated by 2 external peer reviewers and a
► Auerbach M, Macdougall I. The available intravenous iron member of the Feature Advisory Board, with direct editorial input
formulations: history, efficacy, and toxicology. Hemodial Int. from the Feature Editor, the Education Editor, and a Deputy Editor.
2017;21(suppl 1):S83-S92. + ESSENTIAL READING Accepted in revised form September 23, 2017.

APPENDIX
Answer to Question 1: (b) In stage 5 CKD, anemia is hydroxylases (PHDs) “tag” HIF with ubiquitin, triggering
common, especially among patients with diabetes. Dia- degradation. When hypoxia is present, HIF is stabilized
betes is not the cause of the anemia. Rather, relative and interacts with and promotes transcription of many
erythropoietin deficiency is the primary causal factor. genes responsible for cellular protection against hypoxia,
including erythropoietin.
Answer to Question 2: (c) Hypoxia-inducible factor
(HIF)-prolyl hydroxylase plays the central role in oxygen Answer to Question 3: (c) The patient’s level of kidney
sensing. In the presence of sufficient oxygen, prolyl function is sufficiently diminished to indicate that relative

434 AJKD Vol 71 | Iss 3 | March 2018

Core Curriculum

erythropoietin deficiency is probably present. However, as Answer to Question 4: (e) During the fairly large
part of a complete evaluation of anemia, hypercalcemia increase in hemoglobin concentration from 7.1 to
was identified. When present in CKD, hypercalcemia 9.8 g/dL, large quantities of iron were transferred from
should always suggest the possibility of a malignancy (in the body’s storage pools to the developing erythron.
CKD, patients generally have a normal or low serum cal- Iron deficiency frequently develops during the first
cium concentration). The triad of anemia, reduced kidney months of recombinant human erythropoietin therapy
function, and hypercalcemia might place special emphasis for this reason. The best answer is to add treatment with
on the possibility of multiple myeloma. intravenous iron.

AJKD Vol 71 | Iss 3 | March 2018 435