Copyrig
Eryth m
Marti
Introd
The ery
poietin
hemoglobin (Hb) values and reduce transfusion require-
ments in cancer patients receiving chemotherapy.
Initially approved for this indication in 1993 (erythro-
poietin) and 2002 (darbepoetin), more recent studies
evaluated the role of ESAs in anemic cancer patients
not receiving chemotherapy. Several placebo-controlled
studies also examined the quality of life (QoL) of anemic
cancer patients who received ESAs to achieve normal or
near normal Hb values. Using ESAs in patients not
receiving chemotherapy and targeting normal Hb values
are not approved uses for ESAs in cancer patients, and
several of these trials raised safety concerns. A decrease in
survival following ESA treatment was first observed in
two trials in 2003. In the ENHANCE trial, patients with
advanced head and neck cancer treated with radiotherapy
had a higher risk for locoregional progression when also
receiving epoetin beta. The target Hb values were at
least 14 g/dl (women) and at least 15 g/dl (men). The
BEST trial with metastatic breast cancer patients receiv-
ing chemotherapy (target Hb
14 g/dl) was terminated
prematu
survival
thrombo
ESAs w
in 2006
addition
survival
ts i
o a
me
the American Society of Clinical Oncology/American
Society of Hematology (ASCO/ASH) clinical practice
guidelines [5], and changes in the product labeling
for ESAs.
Transfusion requirements and quality of life
ESAs are indicated to reduce the transfusion require-
ments of anemic cancer patients receiving chemotherapy.
Patients treated with ESAs on average receive one red
blood cell (RBC) unit less than controls. There is no
significant difference in transfusions rates when starting
ESA treatment at an Hb less than 10 g/dl compared with
starting ESA treatment at higher Hb values [6]. In a
recent study with anemic cancer patients not receiving
chemotherapy [7], darbepoetin treatment failed to sig-
nificantly reduce the incidence of transfusions compared
with placebo.
Treatment with darbepoetin 200mg every 2 weeks
decreased the symptom burden using theM.D. Anderson
Departmen
Stem Cell T
Hannover,
Correspond
of Hemato
Transplanta
Neuberg-S
Tel: +49 5
e-mail: fenn
Current O
en
em
ent
has
lism
tim
w
er
gu
fo
in,
pinc
1040-874as documented in two meta-analyses published withrely in 2003 because of a significantly reduced
in women receiving ESAs. An increased risk of
embolic events for cancer patients receiving
Symptom Inventory (MDASI) in 2422 cancer patients
with nonmyeloid malignancies [8]. Improvements in
QoL scores are not consistently seen in other studies
cancer patients treated with ESAs. The study by
ropoietin in cancer-related ane
n H. Fenner and Arnold Ganser
uction
thropoiesis-stimulating agents (ESAs), erythro-
and darbepoetin, are widely used to increase
even
led t
treat
t of Hematology, Hemostasis, Oncology and
ransplantation, Hannover Medical School,
Germany
ence to Martin H. Fenner, MD, Department
logy, Hemostasis, Oncology and Stem Cell
tion, Hannover Medical School, Carl-
tr 1, 30625 Hannover, Germany
11 532 4077; fax: +49 511 532 8077;
er.martin@mh-hannover.de
pinion in Oncology 2008, 20:685–689
Purpose of review
Erythropoiesis-stimulating ag
cancer patients receiving ch
erythropoiesis-stimulating ag
Recent findings
Several recently published p
risk of venous thromboembo
treated with erythropoiesis-s
Summary
To minimize risks associated
patients, the most recent Am
Hematology clinical practice
recommendations should be
Keywords
anemia, cancer, erythropoiet
Curr Opin Oncol 20:685–689

2008 Wolters Kluwer Health | Lip
1040-8746ht © Lippincott Williams & Wilkins. Unauthorized r
. During the last 18 months, the results of six
al phase III clinical trials showed a decreased
and/or an increased risk of thromboembolic
Smith e
scores. Q
or trans
6
2008 Wolters Kluwer Health | Lippincott Williams & Wilkinsn cancer patients receiving ESAs. These findings
renewed discussion on the role of ESAs in the
nt of cancer-related anemia [1–4], an update of
ott Williams & Wilkinsia
ts reduce the transfusion requirements of anemic
otherapy. Risks associated with the use of
s in cancer patients have more recently been identified.
e III trials and a meta-analysis have shown an increased
and a decreased survival in anemic cancer patients
ulating agents.
ith erythropoiesis-stimulating agent use in cancer
ican Society of Clinical Oncology/American Society of
idelines and Food and Drug Administration
llowed.
survival, venous thromboembolismeproduction of this article is prohibited.
t al. [7 ] failed to detect a difference in QoL
oL is more difficult to measure than Hb values
fusion requirements and sufficient QoL data are
DOI:10.1097/CCO.0b013e3283136971

Copyrigh
missing
Older tr
meta-an
anemic
Beutel a
Hb valu
effects
significa
by ESA
pretreat
improve
[5].
Iron su
Patients
tored w
label ra
examine
patients
187 can
to place
ESA tr
(sodium
significa
as an H
differen
placebo
Similar
with ly
higher r
ally rece
In 149 c
who we
iron sig
darbepo
analysis
with no
11 g/dl
iron (so
placebo in addition to darbepoetin 500mg every 3 weeks
[12].
serious
gastroin
rate (he

12 g/d
intraven
ing intr
sions (9
intraven
for all c
differen
high-mo
because
lar-weig
and iron
ou
ng
se
bo
H
na
f t
nts
g/d
-an
d
de
vin
in
arc
on
ase
tly
fica
nts
era
nts
rel
: 1
nts
ic
the
en
[7
dar
nts
18
de
of
an
bo
y i
1.5
ad
guidelin
s
ase
iv
cre
me
a
res
r p
sin
r p
w
686 LeuThe treatment was usually well tolerated with
adverse events occurring in 3%, mostly related to
testinal symptoms. The hematologic response
re defined as increase in Hb
2 g/dl or Hb
l) was significantly higher in patients receiving
ous iron (86 vs. 73%, P¼ 0.0011). Patients receiv-
avenous iron also received fewer RBC transfu-
vs. 20%, P¼ 0.005). Based on these studies,
ous iron supplementation should be considered
ancer patients receiving ESAs [13]. Of the four
t available iron preparations for intravenous use,
lecular-weight iron dextrane should not be used
ESA
incre
Surv
A de
treat
(head
The
cance
able
cance
mentfor up to 40% of patients in these ESA trials.
ials frequently did not report the data needed for
alysis. The effect of ESA treatment on QoL in
cancer patients was summarized in a review by
nd Ganser [6]. Only studies with a mean baseline
e of 10 g/dl or less reported statistically significant
on QoL. For those trials reporting a statistically
nt difference in QoL, the improvement in QoL
treatment was small [5]. For patients with
ment Hb more than 10 g/dl, the data on the
ment in QoL by ESA treatment are inconclusive
pplementation
receiving ESAs should have iron stores moni-
ith iron repletion when indicated. Four open-
ndomized studies published in 2007 and 2008
d the role of iron supplementation in cancer
receiving ESAs. Henry et al. [9] randomized
cer patients planned to receive chemotherapy
bo, oral iron, or intravenous iron in addition to
eatment. Patients receiving intravenous iron
ferric gluconate 125mg once weekly) had a
ntly higher rate of hematologic responses defined
b increase by at least 2 g/dl and there was no
ce in Hb response between the oral iron and
groups (73 vs. 45 vs. 41%, analysis as treated).
results were reported in a study with 67 patients
mphoproliferative disorders [10] with again a
ate of hematologic responses in patients addition-
iving intravenous iron (87 vs. 53%, P¼ 0.0014).
ancer patients with chemotherapy-related anemia
re not iron-deficient, treatment with intravenous
nificantly increased the hematologic response to
etin (77 vs. 62%, P¼ 0.0495, intention-to-treat
) [11]. In the largest trial to date, 396 patients
nmyeloid malignancies and an Hb value less than
were randomized to receive either intravenous
dium ferric gluconate 200mg every 3 weeks) or
Ven
Duri
analy
boem
the C
in Re
risk o
patie
13.5
meta
strate
confi
recei
fore,
Rese
lines
incre
recen
signi
patie
moth
patie
matu
VTE
patie
anem
radio
frequ
trols
the
patie
apy [
inclu
most
trials
place
cantl
(RR
was
kemiat © Lippincott Williams & Wilkins. Unauthorized re
of more frequent side effects [14]. Low-molecu-
ht iron dextrane or the iron salts, ferric gluconate
saccharate, are the preferred formulations.
(hazard
Anemic
based cs thromboembolism
the last few years, a number of trials and meta-
s reported an increased risk of venous throm-
lism (VTE) following ESA treatment. In 2006
OIR (Correction of Hemoglobin and Outcomes
l Insufficiency) trial demonstrated an increased
hrombovascular events in chronic kidney disease
when ESAs were used to achieve a target Hb of
l compared with 11.3 g/dl. In the same year, a
alysis published as Cochrane review demon-
a relative risk (RR) for VTE of 1.67 [95%
nce interval (CI) 1.35–2.06] in cancer patients
g ESAs. The increased risk of VTE was, there-
cluded in the 2006 European Organisation for
h and Treatment of Cancer (EORTC) guide-
the use of ESAs in cancer patients [15]. This
in VTE was also demonstrated in several more
published trials. The BRAVE trial [16] found a
ntly increased risk of thromboembolic events in
with metastatic breast cancer treated with che-
py. The GOG-191 trial with cervical cancer
receiving radiochemotherapy was closed pre-
y because of concerns of an increased rate of
1/57 patients in the treatment arm and 4/52
in the control arm experienced VTE [17]. In
cancer patients not receiving chemotherapy or
rapy, patients receiving darbepoetin had more
t thromboembolic events compared with con-
]. An increased risk of VTE was also observed in
bepoetin group of 600 small-cell lung cancer
receiving first-line platinum-based chemother-
]. An updated meta-analysis published in 2008
d the data from the 2006 Cochrane review and
these more recent trials, totaling 38 phase III
d 8172 cancer patients treated with ESAs or
[19]. Patients treated with ESAs had a signifi-
ncreased risk of VTE compared with controls
7 95% CI 1.31–1.87). The increased risk of VTE
ded as a new topic to the 2007 ASCO/ASH
es [5], and the guidelines recommend using
cautiously in situations associated with an
d risk of VTE.
al
ased survival in cancer patients receiving ESA
nt was first reported in 2003 for the ENHANCE
nd neck cancer) and BEST (breast cancer) trials.
ults of five additional trials studying survival in
atients receiving ESA treatment became avail-
ce 2007. In the Amgen 103 trial enrolling anemic
atients not receiving chemotherapy, ESA treat-
as associated with an increased mortality riskproduction of this article is prohibited.
ratio 1.22; 95% CI 1.03–1.45; P¼ 0.022, [7]).
lung cancer patients not receiving platinum-
hemotherapy also had an increased mortality risk

Copyrig
(hazard
and the
The D
(DAHA
regional
advance
Breast
PREPA
receivin
the add
cervical
due to p
bolic ev
recruite
75% in
between
600 pa
platinum
CI 0.78
The m
patients
51 phas
from th
recent s
mortalit
ment w
4062 pa
related
ratio 1.2
increase
treatme
0.99–1.
The he
all trial
value fo
higher
ASCO/A
lines [5
and str
target
doses sh
that is
should
The 20
recomm
anemic
(with th
drome).
these re
on the
include
not the
trials [1
hr
dec
in
f V
p
Rs
atio
ata
mp
ss
of t
of
al
(R
ren
tum
pec
od
orn
sur
po
e s
ral
ces
an
nce
E
t
d
sur
uro
e [
si
Ep
as
.01
sta
nts
oR
. B
a
chi
e-m
-m
fica
r p
ba
g
en
ce
ropratio 1.84) in the Canadian EPO-CAN-20 trial,
trial was stopped prematurely [20].
anish Head and Neck Cancer Group
NCA) 10 trial reported an increased risk of loco-
failure (RR 1.44, P¼ 0.03) in patients with locally
d head and neck cancer receiving radiotherapy.
cancer patients receiving chemotherapy in the
RE trial had a shortened survival when also
g darbepoetin. The GOG-191 trial studying
ition of darbepoetin to radiochemotherapy in
cancer patients was terminated prematurely
otential concerns for an increase in thromboem-
ents [17]. At this point, overall survival in 114
d patients was 61% in the darbepoetin group and
the control group. No difference in survival
the darbepoetin and placebo arm was seen in
tients with small-cell lung cancer receiving
-based chemotherapy (hazard ratio 0.93; 95%
–1.11; P¼ 0.431) [18].
ost recent meta-analysis on survival of cancer
receiving ESAs by Bennett et al. [19] included
e III trials with 13 611 patients, including 38 trials
e 2006 Cochrane analysis and most of the more
tudies mentioned above. Significantly increased
y risks and faster tumor growth after ESA treat-
ere identified in eight of these trials involving
tients. In the six trials evaluating ESAs in cancer-
anemia, the mortality risk was increased (hazard
9; 95% CI 1.00–1.67). The mortality risk was also
d in 45 trials with 11 522 cancer patients with
nt-related anemia (hazard ratio 1.09; 95% CI
19).
moglobin stopping value was at least 13 g/dl in
s that reported an increased mortality risk, a
r which ESAs are not approved and which is
than the 12 g/dl recommended in the 2002
SH guidelines. The 2007 ASCO/ASH guide-
] include information until November 2007
ess the importance of adhering to the 12 g/dl
level, including the recommendations that
ould be reduced when the Hb reaches a level
needed to avoid transfusions and that ESAs
be withheld when the Hb exceeds 12 g/dl.
07 ASCO/ASH guidelines also added a stronger
endation against the use of erythropoietin in
cancer patients not receiving chemotherapy
e exception of low-risk myelodysplastic syn-
The 2007 ASCO/ASH guidelines differ in
commendations from the EORTC guidelines
use of ESAs in anemic cancer patients that
Eryt
The
seen
risk o
could
(Epo
form
the d
is co
expre
tion
level
norm
PCR
diffe
and
nons
antib
Calif
cell
125I-E
on th
Seve
indu
hum
enha
[24].
roden
looke
and
in ne
tissu
had a
high
decre
P¼ 0
158
Patie
of Ep
vival
EpoR
in ar
Abov
below
signi
cance
The
nalin
Kuyk
scien
Erythht © Lippincott Williams & Wilkins. Unauthorized r
data only until October 2006, and therefore
results from the Amgen 103 and EPO-CAN-20
5].
tumor c
stream e
and melopoietin receptors on tumor cells
reased survival of cancer patients receiving ESAs
several trials cannot be explained by the increased
TE alone. Sinclair et al. [21] argued that ESAs
romote tumor growth via erythropoietin receptors
) on tumor cells and/or ESAs could promote the
n of tumor blood vessels. The interpretation of
on the expression of functional EpoR on tumors
licated. Most tumor cell lines and tumor tissues
EpoR mRNA and protein. Genomic amplifica-
he EpoR locus is a rare event in tumors and the
EpoR mRNA expression is often comparable to
tissue. Using quantitative reverse transcriptase-
T-PCR), many studies failed to demonstrate a
ce in EpoR mRNA expression between normal
or tissue. Agarwal et al. [22] pointed out the
ific binding of the commonly used commercial
y C-20 (Santa Cruz Biotechnology, Santa Cruz,
ia, USA), which can also not distinguish
face expression from intracellular expression.
binding studies failed to detect functional EpoR
urface of tumor cells [23].
studies have shown that erythropoietin treatment
downstream effects in cancer cells. Treatment of
ovarian cancer cells with erythropoietin results in
d signaling and a paclitaxel-resistant phenotype
rythropoietin induced tumor angiogenesis in a
breast cancer model [25]. Two recent studies
at the correlation between EpoR expression
vival. EpoR expression was significantly higher
blastoma tumor samples compared with control
26] and patients with a higher EpoR expression
gnificantly better survival (P¼ 0.03). In contrast,
oR expression was associated with a significantly
ed 5-year disease-specific survival (61 vs. 81%,
) in a multivariate analysis of tumor samples from
ge I nonsmall cell lung cancer patients [27].
in the ENHANCE trial that showed expression
by immunhistochemistry had a decreased sur-
lau et al. [28] studied the mRNA expression for
s well as the downstream targets Jak2 and Hsp70
ved tumor samples from the ENHANCE trial.
edian mRNA expression for EpoR and Jak2 and
edian expression of Hsp70 correlated with a
ntly poorer outcome in those head and neck
atients with unresected tumors.
sic science studies on EpoR expression and sig-
in cancer cells were recently summarized by
dal et al. [29]. The authors identified 52 basic
studies that had studied EpoR expression in
oietin in cancer-related anemia Fenner and Ganser 687eproduction of this article is prohibited.
ells. Strong evidence for the presence and down-
ffects of EpoR was seen in head and neck cancer
anoma; in the other tumor cell lines the evidence

Copyrigh
was mix
lines stu
downstr
lation. T
stimulat
convinc
seen in
Regula
The dec
in 2007
mittee (
(FDA)
2008. T
Novem
ings sta
shorten
patients
and non
a Hb m
survival
when E
ESAs s
patient
The las
that ES
cancers
that the
forms [3
cines A
product
stating
method
July 30
(FDA)
informa
not ind
therapy
therapy

10g/d
hemoglo
[33].
Conclu
ESAs c
chemot
of blood
an incre
when ta
ASCO/A
box wa
More b
underst
cre
ts f
oc
re
of p
ighl
 of spe
 of out
nal
Lite
uri F
nce
ythr
08;
crifi
elos
e ca
zzo J
tien
ciet
:13
rren
s.
ute
nce
ith
emi
erap
ntro
ndo
sur
abril
e M
uce
tien
eks
nry
can
emi
:23
den
oeti
quire
ndom
dra
pple
ficie
stit
al co
ery 3
uce
gest
Intra
ce o
erb
J Clin
erb
aem
kem
ythro
ancer 2007; 43:258–270.
pro
sur
d/o
e Va
oma
688 Leusion
an be used in anemic cancer patients receiving
herapy to raise Hb values and reduce the number
transfusions. The use of ESAs is associated with
ased risk of VTE and it can decrease survival
rgeting Hb values of at least 12 g/dl. The 2007
SH guidelines and the most recent FDA black
14 Au
an
15 Bo
er
C
16 Aa
on
an
th
17 Thtory changes
reased survival reported in several phase III trials
prompted the Oncologic Drug Advisory Com-
ODAC) of the US Food and Drug Administration
to hold two meetings in May 2007 and March
he FDA box warnings were revised in March and
ber 2007. The November 2007 black box warn-
te that clinical studies with ESAs have shown
ed overall survival or time-to-tumor progression in
with advanced breast, head and neck, lymphoid
small cell lungmalignancies when dosed to target
ore than 12 g/dl, that the risks of shortened
and tumor progression have not been excluded
SAs are dosed to target a Hb <12 g/dl, and that
hould only be used to treat anemia while the
is also receivingmyelosuppressive chemotherapy.
t ODAC meeting in March 2008 recommended
As should not be given to patients with curable
or metastatic breast or head and neck cancer and
FDA should mandate the provision of consent
0,31]. On June 26, 2008, the European Medi-
gency (EMEA) recommended updating the
information for epoetin-containing medicines
that blood transfusions should be the preferred
of correcting anemia in cancer patients [32]. On
, 2008, the US Food and Drug Administration
issued a letter ordering changes in the product
tion of epoetin-containing medicines: ESAs are
icated for patients receiving myelosuppressive
when the anticipated outcome is cure, ESA
should not be initiated at hemoglobin levels
L, and ESA therapy should be withheld when the
bin exceeds a level needed to avoid transfusions
Additio
World
1 Kh
ca
2 Er
20
3 Sa
4 M
tiv
5

Ri
pa
So
26
The cu
patient
6 Be
ca
7

Sm
an
th
co
This ra
shorter
8 G
th
ind
pa
we
9 He
nifi
an
12
10 He
ep
re
ra
11 Pe
su
de
12

Ba
tri
ev
ind
The lar
ESAs.
inciden
13 Aued or there were insufficient data. In all of the cell
died, one or more studies showed an increase in
eam protein phosphorylation with EpoR stimu-
he basic science data are still incomplete, but
ion of tumor growth via EpoR remains the most
ing hypothesis to explain the decreased survival
some clinical trials with ESAs.
a de
resul
the C
Refe
Papers
been h
kemiat © Lippincott Williams & Wilkins. Unauthorized re
rnings should, therefore, be followed closely.
asic and clinical research is needed to better
and why several phase III studies have reported
mainta
10.0 g
radiati
317–M, Leonard RC, Barnadas A, et al. Effect of once-weekly epoetin beta
vival in patients with metastatic breast cancer receiving anthracycline-
r taxane-based chemotherapy: results of the Breast Cancer-Anemia and
lue of Erythropoietin (BRAVE) study. J Clin Oncol 2008; 26:592–598.
s G, Ali S, Hoebers FJ, et al. Phase III trial to evaluate the efficacy ofased survival following ESA treatment. First
rom an individual patient data meta-analysis by
hrane collaboration are expected later this year.
nces and recommended reading
articular interest, published within the annual period of review, have
ighted as:
cial interest
standing interest
references related to this topic can also be found in the Current
rature section in this issue (pp. 730–731).
R. Weighing the hazards of erythropoiesis stimulation in patients with
r. N Engl J Med 2007; 356:2445–2448.
opoietin analogues: an unnecessary class of drugs. Lancet Oncol
9:81.
cing the cash cow. Nat Biotechnol 2007; 25:363.
ky BL. Erythropoiesis-stimulating agents: benefits and risks in suppor-
re of cancer. Curr Oncol 2008; 15:S10–S15.
D, Somerfield MR, Hagerty KL, et al. Use of epoetin and darbepoetin in
ts with cancer: 2007 American Society of Clinical Oncology/American
y of Hematology clinical practice guideline update. J Clin Oncol 2008;
2–149.
t standard clinical practice guidelines on the use of ESAs in cancer
l G, Ganser A. Risks and benefits of erythropoiesis-stimulating agents in
r management. Semin Hematol 2007; 44:157–165.
RE, Aapro MS, Ludwig H, et al. Darbepoetin alpha for the treatment of
a in patients with active cancer not receiving chemotherapy or radio-
y: results of a phase III, multicenter, randomized, double-blind, placebo-
lled study. J Clin Oncol 2008; 26:1040–1050.
mized trial demonstrated an increased risk of thromboembolism and
vival in anemic cancer patients receiving ESA treatment.
ove JL, Perez EA, Tomita DK, et al. Assessing symptom burden using
. D. Anderson symptom inventory in patients with chemotherapy-
d anemia: results of a multicenter, open-label study (SURPASS) of
ts treated with darbepoetin-alpha at a dose of 200 microg every 2
. Cancer 2007; 110:1629–1640.
DH, Dahl NV, Auerbach M, et al. Intravenous ferric gluconate sig-
tly improves response to epoetin alfa versus oral iron or no iron in
c patients with cancer receiving chemotherapy. Oncologist 2007;
1–242.
us M, Birgega˚rd G, Na¨sman P, et al. Addition of intravenous iron to
n beta increases hemoglobin response and decreases epoetin dose
ment in anemic patients with lymphoproliferative malignancies: a
ized multicenter study. Leukemia 2007; 21:627–632.
zzoli P, Farris A, Del Prete S, et al. Randomized trial of intravenous iron
mentation in patients with chemotherapy-related anemia without iron
ncy treatedwith darbepoetin alpha. J Clin Oncol 2008; 26:1619–1625.
L, Vandebroek A, Altintas S, et al. Randomized, multicenter, controlled
mparing the efficacy and safety of darbepoetin alpha administered
weeks with or without intravenous iron in patients with chemotherapy-
d anemia. J Clin Oncol 2008; 26:1611–1618.
trial comparing intravenous iron to placebo in cancer patients receiving
venous iron improved the hematologic response rate and reduced the
f transfusions.
ach M. Should intravenous iron be the standard of care in oncology?
Oncol 2008; 26:1579–1581.
ach M, Ballard H, Glaspy J. Clinical update: intravenous iron for
ia. Lancet 2007; 369:1502–1504.
eyer C, Aapro MS, Courdi A, et al. EORTC guidelines for the use of
poietic proteins in anaemic patients with cancer: 2006 update. Eur Jproduction of this article is prohibited.
ining hemoglobin levels above 12.0 g/dL with erythropoietin vs above
/dL without erythropoietin in anemic patients receiving concurrent
on and cisplatin for cervical cancer. Gynecol Oncol 2008; 108:
325.