Journal of Cardiology 62 (2013) 249256

Contents lists available at ScienceDirect

Journal of Cardiology
journal homepage: www.elsevier.com/locate/jjcc

Original article

Effect of anemia correction to the modestly high hemoglobin level in

patients with chronic kidney disease on left ventricular hypertrophy
Makoto Akaishi (MD, PhD, FJCC) a, , Michiaki Hiroe (MD, PhD, FJCC) b ,
Yoshiyuki Hada (MD, PhD, FJCC) c , Makoto Suzuki (MD, FJCC) d ,
Yoshiharu Tsubakihara (MD, PhD) e , Tadao Akizawa (MD, PhD) f , KRN321 Study Group
a
Department of Cardiology, Kitasato University Kitasato Institute Hospital, Tokyo, Japan
b
Cardiology Division, National Center for Global Health and Medicine, Tokyo, Japan
c
Department of Cardiology, Sakakibara Memorial Clinic, Tokyo, Japan
d
Department of Clinical Laboratory, Toho University Medical Center Ohashi Hospital, Tokyo, Japan
e
Department of Comprehensive Kidney Disease Research, Osaka University Graduate School of Medicine, Osaka, Japan
f
Division of Nephrology, Department of Medicine, Showa University School of Medicine, Tokyo, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Background and purpose: To assess effects of long-term anemia management on left ventricular hyper-
Received 4 October 2012 trophy in patients with chronic kidney disease (CKD) not on dialysis, we performed secondary outcome
Received in revised form 15 April 2013 analyses of a randomized controlled study that evaluated effects of anemia management with erythro-
Accepted 22 April 2013
poiesis stimulating agents in this population.
Available online 18 June 2013
Methods and subjects: Subjects [hemoglobin (Hb) < 10.0 g/dL, 2.0 serum creatinine < 6.0 mg/dL] were
randomized either to high Hb (11.0 target Hb 13.0 g/dL with darbepoetin alfa), or to low Hb group
Keywords:
(9.0 target Hb 11.0 g/dL with recombinant human erythropoietin), and followed up to 48 weeks. Data
Cardiovascular disease
Hypertrophy
from echocardiographic evaluation and values of neurohumoral factors associated with heart failure were
Drug therapy assessed in subjects whose data were evaluable both at the baseline and at the end point.
Renal function Results: The high Hb group achieved target range Hb levels (12.1 1.1 g/dL, at 32 weeks, N = 111), which
Natriuretic peptide was signicantly higher (p < 0.001) than the low Hb group (N = 95). Though blood pressure and renal
function changes were similar between the groups, left ventricular diastolic dimension was signicantly
decreased only in the high Hb group (p < 0.001), and the change in left ventricular mass index (LVMI)
correlated coarsely but signicantly with the achieved Hb levels (r = 0.147, p = 0.032). The higher Hb
levels were associated with greater reduction in LVMI and left ventricular wall thickness, and the lower
Hb levels with the greater increase in human arterial- or brain natriuretic polypeptide levels.
Conclusions: Anemia correction targeting modestly higher Hb levels better preserves cardiac function in
CKD patients not on dialysis.
2013 Japanese College of Cardiology. Published by Elsevier Ltd. All rights reserved.

Introduction
Chronic kidney disease (CKD) is an independent risk factor for
cardiovascular disease (CVD), and the risk of CVD increases as
glomerular ltration rate (GFR) decreases. A prospective analy-
sis showed that CKD is a signicant risk factor for CVD also in
the general Japanese population [13]. Furthermore, CKD patients
with anemia are at a greater risk of developing CVD. Observa-
tional studies revealed that anemia or left ventricular hypertrophy
Corresponding author at: Department of Cardiology, Kitasato University Kitasato
Institute Hospital, 5-9-1 Shirokane, Minato-ku, Tokyo 108-8642,
Japan. Tel.: +81 03 3444 6161; fax: +81 03 3448 0553.
E-mail address: akaishim@insti.kitasato-u.ac.jp (M. Akaishi).
(LVH) associated with CKD is an additional risk for CKD and
total mortality [4,5]. Cardio-renal anemia syndrome (CRA syn-
drome) is another complication, in which CKD, chronic heart
failure (CHF), and anemia can cause or be caused by the oth-
ers, or act as risk factors for each other [6,7]. As CKD progresses,
the patients may develop anemia. If appropriate anemia manage-
ment breaks the vicious cycle of CRA syndrome, the progression
of each disease can be prevented. Approximately 75% of CKD
patients are complicated with LVH at the initiation of renal replace-
ment therapy, and have various complications of CVD such as
CHF, coronary artery disease, or peripheral vascular disease [810].
The importance of anemia management from the early stages of
CKD has been emphasized and several controlled studies have
been conducted to determine the reasonable target levels in cor-
recting anemia [1115]. However, the optimal target levels for

0914-5087/$ see front matter 2013 Japanese College of Cardiology. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jjcc.2013.04.008
250 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256
anemia correction remain controversial in regard to reducing
the risk of developing CVD and improving the prognosis in CKD
patients.
In Japan, recombinant human erythropoietin (rHuEPO) has been
used in clinical settings since 1990. Because periodic erythro-
cyte transfusion was the only therapeutic choice for CKD patients
with anemia before the introduction of rHuEPO, anemia correc-
tion with rHuEPO dramatically decreased the number of CKD
patients requiring erythrocyte transfusion and greatly improved
patients quality of life, as well as their prognosis. The hemoglobin
(Hb) levels for anemia correction with rHuEPO at around 10 g/dL
have been recommended. However, in a late phase II clinical
study of darbepoetin alpha, LVH and myocardial stress were
signicantly reduced in CKD patients not on dialysis whose
Hb levels were increased to greater than 11.0 g/dL in16 weeks
[16,17]. In the present study, we further analyzed the secondary
outcomes of a 48-week, randomized, controlled study [17] to deter-
mine whether maintaining target hemoglobin (Hb) levels higher
than those achieved by conventional anemia treatment levels
(11.0 Hb 13.0 g/dL and 9.0 Hb 11.0 g/dL, respectively) bene-
cially affects LVH and cardiac overload in patients with CKD not on
dialysis.
Methods
Study design and patients
The present study is a sub-analysis focused on cardiac func-
tions in study subjects of a multicenter open-label randomized
controlled study involving 79 medical centers conducted between
November 2005 and April 2007 [17]. In subsets of the study popu-
lation that were followed up for 48 weeks, data regarding cardiac
functions were assessed in the present study comparing values at
baseline and at 32 weeks after starting erythropoiesis stimulating
agent (ESA: darbepoietin alfa or rHuEPO) treatment.
The protocol of the main study was approved by institutional
review boards of all the participating medical centers. All the par-
ticipants in the study gave written informed consent.
The detailed protocol of the study was described elsewhere [17],
and a brief summary of the protocol is presented here.
The study included patients with CKD not on dialysis, whose
Hb levels were less than 10.0 g/dL, and whose serum creatinine
(Cr) levels were 2.0 mg/dL and <6.0 mg/dL or creatinine clearance
(Ccr) levels 30.0 mL/min.
The enrolled patients were randomized to either of the two
treatment groups: the high Hb group, in which darbepoetin alfa
was administered to achieve a target Hb level of 11.0 g/dL and
13.0 g/dL, or the low Hb group, in which rHuEPO was administered
to reach a target Hb level of 9.0 g/dL and 11.0 g/dL. Randomiza-
tion was performed using the minimization method with Hb levels
(<9.0 g/dL or 9.0 g/dL), serum Cr levels (<4.0 mg/dL or 4.0 mg/dL),
and the presence or absence of diabetes mellitus, and participating
medical centers as adjustment factors.
Subjects in the high Hb group started with subcutaneous dar-
bepoetin alfa 60 g every two weeks. After reaching the target
Hb level (11.0 Hb 13.0 g/dL), the Hb level was maintained with
subcutaneous darbepoetin alpha every 2 or 4 weeks. Subjects in
the low Hb group started to receive subcutaneous rHuEPO at a
dose of 6000 IU once a week, in accordance with the dosage and
administration approved in Japan. After reaching the target Hb
level (9.0 Hb 11.0 g/dL), the Hb level was maintained with sub-
cutaneous rHuEPO once a week or every 2 weeks for 48 weeks.
During the study, a combination of supplemental iron as neces-
sary, blood pressure control, and dietary control for salt and protein
intake were also administered as described in detail elsewhere
[17].
Measurement of cardiac function and other examinations
Echocardiography, assay for neurohumoral factors [serum
human brain natriuretic peptide (BNP) and human atrial natri-
uretic peptide (hANP)], and chest X-ray were performed to evaluate
cardiac function at baseline and 32 weeks or at discontinuation
of the treatment. Blood pressure, heart rate, complete blood cell
counts, and blood chemistry were assessed every 2 or 4 weeks
depending on the injection schedule of darbepoetin alpha or
rHuEPO.
Since M-mode imaging was necessary for echocardiographic
measurements, the following patients with factors that might
hamper appropriate image acquisition and/or its evaluation were
excluded.
Those who discontinued the study earlier than 16 weeks after
the start of treatment were also excluded from the evaluation.
(1) Patients with pericardial disease or myocarditis.
(2) Patients with multiple ventricular premature contractions.
(3) Patients with atrial brillation.
(4) Patients with myocardial infarction.
(5) Patients with organic valvular disease or with moderate to
severe valvular regurgitation.
(6) Patients with hypertrophic cardiomyopathy.
(7) Patients with arteriovenous stula.
At each medical center, the following ve types of echocardio-
grams were obtained: left ventricular M-mode echocardiogram and
the end-diastolic/end-systolic long and short axis cross-sectional
images of the left ventricle, and all the images were collected
and evaluated by an external evaluation committee consisting of
3 cardiologists. Two cardiologists rst evaluated the quality of
echocardiograms, and then measured left ventricular end diastolic
dimension (LVDd), left ventricular end systolic dimension (LVDs),
interventricular septal thickness (IVST), and posterior wall thick-
ness (PWT) by left ventricular M-mode echocardiography using
the ASE method (American Society of Echocardiography leading
edge recommendations) [18]. The mean value of the measurement
results by two cardiologists was used as the nal measurement
value. When the two measurements disagree, the results were dis-
cussed with a third cardiologist and the nal decision reected
the consensus of all three cardiologists. In order to avoid bias, all
evaluations and measurements were conducted after masking the
treatment groups, the timing of examinations, and patient charac-
teristics.
From the measured left ventricular parameters, left ven-
tricular mass was calculated using Devereuxs formula [19]:
0.8 {1.04 [(LVDd + IVST + PWT)3 LVDd3 ]} + 0.6, which was
divided by body surface area to obtain LVMI. To assess left ventri-
cular systolic function, left ventricular ejection fraction (LVEF) was
calculated using left ventricular volume by the Teichholz method.
Analyses
Patient characteristics
Summary statistics were obtained and compared by two-sample
t-test for continuous variables and by 2 test for frequencies of
nominal variables.
Comparison of cardiac function indices between the high- and
low-Hb groups
Cardiac function was analyzed in patients whose echocardio-
grams were obtained and evaluable both before and after the
 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256
treatment in this study. Parameters chosen for the study were
obtained at the start of treatment (baseline) and when nal
assessment of LVMI was performed (end point). Mean changes
and 95% condence intervals were calculated for each group
between the baseline and end point values, and differences
between the two groups were analyzed with a two-sample t-test.
Within-group comparison of changes in LVMI and other param-
eters from baseline to end point was conducted by one-sample
t-test.
Correlation between achieved Hb levels and changes in cardiac
function indices (LVMI and other parameters)
Data from the two groups were combined irrespective of the
treatment received and divided into ve categories based on
Hb levels at the end point (Hb < 10.0 g/dL, 10.0 Hb < 11.0 g/dL,
11.0 Hb < 12.0 g/dL, 12.0 Hb < 13.0 g/dL, Hb 13.0 g/dL) to ana-
lyze the relationship between the changes in cardiac function
parameters and Hb levels. Correlation between the change in
the LVMI and the Hb levels at the endpoint was also assessed.
Summary statistics and 95% condence intervals of the various
parameters studied in each Hb category were calculated. An anal-
ysis of covariance was performed with LVMI at the baseline as
a covariate, and the adjusted mean and 95% condence inter-
val were calculated. Similar analyses were performed for other
parameters. In addition, to evaluate correlation between the end
point Hb levels and cardiac function indices such as changes in
LVMI, contribution of the changes in Hb levels to each param-
eter of cardiac functions was determined by a linear regression
model using cardiac function indices at baseline as a covariate.
Data were presented as mean standard deviation. If interval esti-
mates were required, two-sided 95% condence intervals were
calculated. Two-sided p-values of 0.05 were considered as signif-
icant.
All analyses were performed with SAS (Cary, NC, USA).
Results
Patients and baseline characteristics
A total of 322 patients identied as being eligible for the present
study were randomized in a ratio of 1:1 to either the high or low
Hb group. Of the 322 patients (161 in the high Hb group and 161
in the low Hb group), echocardiograms at the start of study were
available in 244 patients, excluding 78 patients who discontinued
the study earlier than 16 weeks after starting the treatment, who
had a disease that might prevent appropriate M-mode echocar-
diographic evaluation, or those who did not have all the required
measurements at the time of baseline evaluation. Thereafter,
echocardiography was performed at 32 weeks or at study dis-
continuation. Excluding 38 patients whose cardiac measurements
were not available, the nal analysis for cardiac function in the
present study included 206 patients (111 in the high Hb group and
95 in the low Hb group), whose echocardiography at both 0 week
(baseline) and post-treatment were evaluable (Fig. 1). Echocar-
diograms were not evaluated in a total of 116 (36.0%) patients,
due either to unclear images, inappropriate echo beam direction,
pericardial effusion, discontinuation earlier than 16 weeks after
starting the treatment, or no examination performed. Thirty six
cases were discontinued after 16 weeks, including 12 cases in the
high Hb group and 24 cases in the low Hb group due to introduction
of hemodialysis (15 cases), adverse events (7 cases), inadequate
response to darbepoetin alpha or rHuEPO (5 cases), discontinu-
ation of drug because of excess response (7 cases), or voluntary
251
withdrawal (2 cases). These 36 discontinued cases after 16 weeks,
whose observation period was 21.7 4.6 weeks, were included for
analysis using the last measurement at the discontinuation.
No signicant difference between the groups was observed in all
the baseline patient characteristics other than gender and transfer-
rin saturation (Table 1). All the other baseline parameters including
Hb level, the serum Cr level, and estimated glomerular ltration rate
(eGFR) were not different between the two groups. No imbalance
was detected between the groups in past and present history of
CVD, including angina, atrial brillation, congestive heart failure,
myocardial infarction, hypertension, and cerebral infarction.
Time course of Hb levels, blood pressure, and kidney function
Fig. 2 shows the changes in Hb levels, blood pressure, and eGFR
during the study. As shown in Fig. 2, subjects in the high Hb
group achieved the lower limit of the target Hb ranges, 11.0 g/dL,
by 8 weeks, and thereafter remained within the target Hb range
(11.013.0 g/dL) for 32 weeks. In the low Hb group, the Hb level
remained within the target Hb range (9.011.0 g/dL) from the start
to the end of the treatment. The baseline Hb levels of the two groups
were mostly equivalent (9.3 0.8 g/dL), while the Hb levels at the
echocardiogram measurements (32 weeks) of high Hb group were
higher (12.0 1.2 g/dL) than in the low Hb group (10.1 1.0 g/dL,
Fig. 2A). Blood pressure did not change signicantly, and eGFR
showed a tendency to decrease during the study, and showed sim-
ilar time course of changes in both groups (Fig. 2B and C).
Changes in cardiac function indices
Echocardiographic parameters
Echocardiographic parameters before and after the study treat-
ment are summarized in Table 2. LVMI and left ventricular wall
thickness (IVST + PWT) were signicantly decreased only in the
high Hb group. LVMI did not change from baseline in the low Hb
group, whereas it decreased by 7.8 g/m2 in the high Hb group,
and the changes in LVMI were statistically signicant between the
groups (p = 0.009, Table 2B). As shown in Table 2B, the decrease
in left ventricular wall thickness (IVST + PWT) as well as that of
PWT was greater in the high Hb group, and it was also statistically
signicant between the groups (p = 0.023 and 0.016, respectively).
There was no difference in the changes of other echocardiographic
parameters (Table 2B).
Cardiothoracic ratio (CTR) did not show any difference between
the two groups, either before or after the treatment. Also, there
was no signicant difference in changes of heart rates between the
two groups (0.9 11.1 and 1.5 9.5/min, high and low Hb groups,
respectively).
Neurohumoral factors
BNP levels increased by 23.3 pg/mL only in the low Hb group
(p = 0.03), whereas h-ANP was unchanged in both groups (Table 2).
Change in levels of BNP from baseline to the 32nd week signicantly
correlated with the changes in LVMI (p = 0.014). Change in levels of
h-ANP also showed a weak trend to associate with the changes in
LVMI, but was not found to be statistically signicant (p = 0.092).
Correlation between the achieved Hb levels and changes in
cardiac function indices (LVMI and other parameters)
Hb levels achieved after the 32-week treatment correlated
coarsely but signicantly with the change in LVMI (r = 0.147,
p = 0.034). Fig. 3 shows that the changes in LVDd, hANP, or BNP
decreased as Hb levels increased (p = 0.008, p < 0.001 and p = 0.045).
However, no signicant relationships between the changes in EF
with Hb levels were observed.
252 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256

Fig. 1. Collection of echocardiograms. Of 322 eligible patients, a total of 206 patients whose echocardiographic measurements for both baseline and at 32 weeks (or at
discontinuation) were available, were included in the present analyses. A total of 116 patients were excluded for either a disease history that could inuence M-mode
echocardiographic evaluation or missing M-mode echocardiographic measurements. Hb, hemoglobin; LVMI, left ventricular mass index.

Relationship between baseline LVMI and improvement in LVMI by
anemia correction
Fig. 4 shows the change in LVMI in each group that was
categorized based on the baseline LVMI (LVMI < 100 g/m2 , 100
LVMI < 130 g/m2 , 130 LVMI < 160 g/m2 , and LVMI 160 g/m2 ).
Decrease in LVMI was greater in the more severe LVH category.
Relationship between changes in LVMI and kidney function
Decrease in LVMI after the 32-week treatment also showed sig-
nicant correlation with changes in kidney functions determined
as serum creatinine levels or eGFR (p = 0.007, p = 0.049 between the
changes of LVMI and serum creatinine levels or eGER, respectively).
Discussion
The present study showed, in patients with CKD not on dialy-
sis, that correcting anemia with target Hb levels at 11.013.0 g/dL
decreases LVMI more signicantly compared to the conventional
treatment level of 9.011.0 g/dL. The present results also demon-
strated that, regardless of the treatment method, higher Hb level
achieved was associated with greater reduction of LVH and lesser
production of neurohumoral factors that reect ventricular over-
load. Patients with more severe baseline LVH responded more
signicantly with decreases in the LVMI. Together, the present
results suggest that anemia correction to modestly higher Hb levels
in CKD patients not on dialysis could reduce incidence of LVH.
Recently, four large-scale randomized controlled studies
(ACORD, CHOIR, CREATE, and TREAT) were conducted to deter-
mine the effects of Hb level normalization in patients with CKD
not on dialysis in association with the risk of developing LVH
or CVD [1315,20]. All four studies used higher target levels
than the present study (more than 13.0 g/dL in the TREAT study,
13.015.0 g/dL in the ACORD study, 13.5 g/dL in the CHOIR study,
and 13.015.0 g/dL in the CREATE study, as target Hb levels). In any
of these studies, no risk reduction for the development of LVH or
CVD was observed in high Hb groups compared to low Hb groups
(Hb level of approximately 11.0 g/dL). Furthermore the TREAT study
 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256 253

Table 1
Subject characteristics.

Characteristic High Hb group Low Hb group p-Value

(n = 111) (n = 95)

Age (yr) 63.6 12.4 61.4 11.0 0.173

Female sex no. of patients (%) 51 (45.9) 57 (60.0) 0.044
Cause of chronic kidney disease 0.717
no. of patients (%)
Glomerulonephritis 55 (49.5) 46 (48.4)
Diabetic nephropathy 22 (19.8) 23 (24.2)
Pyelonephritis 0 (0.0) 1 (1.1)
Polycystic kidney disease 5 (4.5) 2 (2.1)
Nephrosclerosis 15 (13.5) 10 (10.5)
Other 14 (12.6) 13 (13.7)
Cardiovascular history no. of 106 (95.5) 92 (96.8) 0.618
patients (%)
Weight (kg) 56.6 10.9 56.2 11.2 0.820
Height (cm) 159.0 9.3 157.6 7.9 0.279
Transferrin saturation (%) 32.9 12.4 29.3 11.2 0.030
Ferritin (ng/mL) 119.3 122.2 107.4 100.8 0.449
Creatinine (mg/dL) 3.52 1.08 3.49 1.08 0.845
Creatinine clearance (mL/min) 19.6 7.2 19.0 6.6 0.541
eGFR (mL/min/1.73 m2 ) 12.9 4.9 12.7 4.9 0.715
hANP (pg/mL) 47.8 29.7 53.9 37.9 0.197
BNP (pg/mL) 116.3 162.6 97.5 138.5 0.376
CTR (%) 49.4 5.2 49.1 6.2 0.724

Hb, hemoglobin; eGFR, estimated glomerular ltration rate; hANP, human atrial
natriuretic peptide; BNP, brain natriuretic peptide; CTR, cardiothoracic ratio.
eGFR was estimated according to the Modication of Diet in Renal Disease formula.

demonstrated signicant increase in stroke in the high Hb group.

A meta-analysis of nine clinical studies in patients with CKD at
various stages [21] also reported that patients with a target Hb
level within the normal range (13.0 g/dL in seven studies and
12.0 g/dL in two studies) had higher risk of total death, shunt
thrombosis, and poor blood pressure control. These recent reports
suggested that normalization of the Hb level in patients with CKD
not on dialysis is not necessarily benecial but rather could be detri-
mental. However, since characteristics of patients with CKD not on
dialysis and the frequency of CVD are distinct in Japanese compared
to those in Westerners, results from non-Japanese clinical stud-
ies may not be applicable to Japanese patients. In addition, the Hb
level after achieving the target Hb in the present study was approx-
imately 12.0 g/dL in the high Hb group (not more than 13.0 g/dL),
which was lower than the target Hb levels in most previous studies.
The modestly high Hb target level was achieved with the protocol
that included the rule for discontinuation of the drug.
Although several reports described that anemia correction
improves LVMI [22], few studies investigated LVMI or LVH in a long-
term and randomized condition. There have been a few studies
conducted in Japan, which found that anemia correction favorably
affects cardiac function in CKD patients not on dialysis [23,24].
However, those studies were small in sample size or retrospec-
tive assessment not a controlled comparative study with certain
target Hb levels. The present study was originally conducted as
a long-term randomized controlled study to evaluate the effect
of the different levels of anemia correction on renal function, in
patients with CKD not on dialysis. Here, we further investigated Fig. 2. Time course of hemoglobin (Hb) levels, blood pressure, and estimated
the long-term cardio-protective effect of anemia correction levels glomerular ltration rates (eGFR). (A) Time course of mean Hb levels. The target
in the same study population. Hb level was 11.013.0 g/dL in the high Hb group and 9.011.0 g/dL in the low Hb
group. (B) Time course of mean blood pressure. Diast-BP, diastolic blood pressure;
In patients whose Hb levels were targeted between 11.0 and Syst-BP, systolic blood pressure. (C) Time course of mean eGFR.
13.0 g/dL, LVMI was signicantly improved. Analysis combining the
high and low Hb groups showed a signicant relationship between a decrease in left ventricular overload. As anemia contributes to
the achieved Hb levels and LVMI improvement, suggesting that left ventricular overload through high output state, stimulation
anemia correction has a profound effect on the regression of LVH. of sympathetic nervous system or relative myocardial ischemia,
The higher achieved Hb level, irrespective of the treatment groups, achieving near-normal Hb levels helps to improve the left ven-
was also associated with the lower hANP or BNP levels. hANP tricular loads, as shown in the present study. The signicant
and BNP levels reect left ventricular overload [2528]. Thus, correlation detected in the present study between the levels of
the observed improvement in neurohumoral factors represents Hb achieved and reduction of LVH further supports the benecial
254 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256

Table 2
Summary of echocardiography data and humoral factors of cardiac function.

High Hb group Low Hb group

Baseline End point One sample t-test Baseline End point One sample t-test
Mean SD Mean SD Mean SD (n = 95) Mean SD (n = 95)
(n = 111) (n = 111)

(A) Baseline and end point values of markers for cardiac functions in the high- and low-Hb groups
LVDd (mm) 49.35 5.49 49.42 5.70 0.835 49.31 5.65 49.54 5.37 0.455
LVDs (mm) 29.81 6.00 29.08 5.61 0.108 29.12 5.81 29.02 5.69 0.759
IVST + PWT (mm) 21.41 3.35 20.45 2.64 <0.001 21.16 3.42 20.95 3.63 0.317
IVST (mm) 11.03 1.92 10.61 1.67 0.004 10.97 1.92 10.85 2.18 0.42
PWT (mm) 10.39 1.64 9.84 1.23 <0.001 10.19 1.73 10.10 1.65 0.408
LVMI (g/m2 ) 127.7 37.2 119.9 30.3 <0.001 126.1 35.1 126.0 36.2 0.955
LVEF (%) 64.28 9.68 66.05 9.36 0.059 65.72 9.84 66.51 8.83 0.276
BNP (pg/mL) 116.3 162.6 136.2 201.7 0.119 97.5 138.5 120.8 174.6 0.03
hANP (pg/mL) 47.8 29.7 48.9 36.0 0.684 53.9 37.9 56.9 49.7 0.342

High Hb group Low Hb group Two sample t-test

Mean SD (n = 111) Mean SD (n = 95)

(B) Changes of cardiac function markers in the high- and low-Hb groups
LVDd (mm) 0.08 3.91 0.22 2.92 0.763
LVDs (mm) 0.76 4.70 0.10 3.31 0.253
IVST + PWT (mm) 0.96 2.52 0.21 2.08 0.023
IVST (mm) 0.42 1.50 0.12 1.44 0.152
PWT (mm) 0.54 1.48 0.10 1.12 0.016
LVMI (g/m )
2
7.8 22.4 0.1 19.3 0.009
LVEF (%) 1.89 9.82 0.79 7.06 0.365
BNP (pg/mL) 19.9 133.5 23.3 103.4 0.838
hANP (pg/mL) 1.1 27.3 3.0 31.1 0.625

LVDd was measured at the peak of the R wave on the ECG, and LVDs were measured at the time when the interventricular septum was located most posteriorly. IVST and
PWT were measured at end diastole. LVMI was calculated using the measurements by Devereuxs formula. LVEF was calculated using the Teichholz method.
Measurements of markers in individual patients at the baseline were subtracted from those at the end points, and the values were compared between the two groups by the
two-sample t-test. Mean of the subtracted values are also shown.
Hb, hemoglobin; LVDd, left ventricular end diastolic dimension; LVDs, left ventricular end systolic dimension; IVST, interventricular septal thickness; PWT, posterior wall
thickness; LVMI, left ventricular mass index; LVEF, left ventricular ejection fraction; BNP, brain natriuretic peptide; hANP, human atrial natriuretic peptide.

effects of anemia correction on improving LVH. Patients with
higher baseline LVMI showed more signicant decrease in the
LVMI, suggesting that aggressive anemia correction would be more
effective in patients with more advanced LVH.
LVH, preceding overt cardiac failure, is a classic risk factor for
developing CVD, as is anemia. Increasing the Hb levels to between
11.0 and 13.0 g/dL may prevent increase in LVMI, and regressing
LVH can eliminate the risk factors for developing CVD, implying that
anemia correction may improve prognosis via two related mech-
anisms: (1) by ameliorating anemia as an independent CVD risk
factor; and (2) by suppressing LVH as another independent risk
factor for CVD.
The present study also revealed a correlation between dete-
rioration of kidney function and increases in LVMI. Although
baseline and changes in kidney function determined by creati-
nine or eGFR were not statistically different between both groups,

Fig. 3. Changes in cardiac function parameters classied into ve categories according to Hb levels. (A) Changes in LVMI from baseline to end point by Hb levels. (B) Changes
in hANP from baseline to end point by Hb levels. (C) Changes in BNP from baseline to end point by Hb levels. (D) Changes in LVDd from baseline to end point by Hb levels.
Hb, hemoglobin; LVMI, left ventricular mass index; hANP, human atrial natriuretic peptide; BNP, brain natriuretic peptide; LVDd, left ventricular end diastolic dimension.
 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256
Fig. 4. The more baseline LVMI, the more improvement of LVMI was observed in High Hb group. The groups were classied into four categories for LVMI changes according
to baseline LVMI. LVMI, left ventricular mass index.
after long-term follow up of this study, the renal composite
outcomes determined by time to occurrence of end-stage renal
disease (ESRD), doubling of serum creatinine or death was sig-
nicantly lower in the high Hb group [29]. This may suggest a
kidney-mediated relationship between the correction of anemia
and amelioration of cardiac function or these three functions in
triad may interact with one another in this patient population.
Although this is a randomized prospective study, bias due to
the open-label controlled study design may be present. To mini-
mize the bias, LVMI was evaluated by three cardiologists blinded
to treatment group, patient characteristics, and other information
for all the patients. The two groups used different drugs to achieve
the targeted Hb levels. At the time the present study was conducted,
rHuEPO administration was limited not to reach an Hb level at
approximately 12 g/dL by the Japanese health insurance system,
while darbepoetin alfa was under investigation for its safety and
efcacy and was accessible to aim higher Hb levels. Because dar-
bepoetin alfa and rHuEPO share similar pharmacological proles,
except that the former has a longer half-life, the effects on car-
diac functions observed in the present study could reasonably be
interpreted as pure outcomes from the different Hb levels achieved.
However, we cannot fully exclude the possibility that an unknown
effect distinct to either of the drugs might have caused differen-
tial outcome observed in the present study, since, by the study
design, there were substantial differences, including the Hb levels
achieved between the two groups. Use of two study drugs, though
similar in action, is a limitation of the present study. We hope that
our ndings that suggest benecial effects of anemia correction
with modestly high target Hb levels would further encourage larger
studies to justify use of either ESA to treat modestly anemic patients
with CKD for better outcomes.
Conclusions
These results from the present study support our previous
observation in a late phase II clinical study that increasing the
Hb level to at least 11.0 g/dL would reduce LVH and be cardio-
protective in patients with CKD not on dialysis.
255
A long-term clinical study using harder end points such as death
or development of overt CVD is warranted to investigate the effects
of anemia correction up to the modestly high Hb level (more than
11 g/dL and less than 13 g/dL) on the prognosis of CKD patients.
Acknowledgments
We would like to express our deepest appreciation to the fol-
lowing advisors in this study. Fumitake Gejyo (Niigata University),
Shinichi Nishi (Kobe University Graduate School of Medicine),
Masashi Suzuki (Shinrakuen Hospital), Takashi Akiba (Tokyo
Womens Medhical University), Yasuhiko Iino (Nippon Medical
University Hospital), Akira Saito (Yokohama Dai-ichi Hospital),
Yuzou Watanabe (Kasugai Municipal Hospital), Hideki Hirakata
(Fukuoka Red Cross Hospital).
This randomized controlled study was sponsored by Kyowa
Hakko Kirin Co., Ltd. (Tokyo, Japan). Registration of Clinical
Trials: The Cochrane Renal Group registry: autoid 121, crg id
CRG030600049. The authors participated in this study as advisers
and received consultation fees from Kyowa Hakko Kirin Co., Ltd.
References
[1] Ninomiya T, Kiyohara Y, Tokuda Y, Doi Y, Arima H, Harada A, Ohashi Y,
Ueshima H, Japan Asteriosclerosis Longitudinal Study Group. Impact of kid-
ney disease and blood pressure on the development of cardiovascular disease:
an overview from the Japan Arteriosclerosis Longitudinal Study. Circulation
2008;118:2694701.
[2] Ueshima K, Yasuno S, Oba K, Fujimoto A, Mukoyama M, Ogihara T, Saruta T,
Nakao K. Impact of left ventricular hypertrophy on the time-course of renal
function in hypertensive patients a subanalysis of the CASE-J trial. Circ J
2010;74:21328.
[3] Kiyosue A, Hirata Y, Ando J, Fujita H, Morita T, Takahashi M, Nagata D, Kohro T,
ImaiNagai R. Relationship between renal dysfunction and severity of coronary
artery disease in Japanese patients. Circ J 2010;74:78691.
[4] Vlagopoulos PT, Tighiouart H, Weiner DE, Grifth J, Pettitt D, Salem DN, Levey
AS, Sarnak MJ. Anemia as a risk factor for cardiovascular disease and all-cause
mortality in diabetes: the impact of chronic kidney disease. J Am Soc Nephrol
2005;16:340310.
[5] Weiner DE, Tighiouart H, Vlagopoulos PT, Grifth JL, Salem DN, Levey AS, Sarnak
MJ. Effects of anemia and left ventricular hypertrophy on cardiovascular disease
in patients with chronic kidney disease. J Am Soc Nephrol 2005;16:180310.
256 M. Akaishi et al. / Journal of Cardiology 62 (2013) 249256
[6] Silverberg D, Wexler D, Blum M, Wollman Y, Iaina A. The cardio-renal anaemia
syndrome: does it exist? Nephrol Dial Transplant 2003;18(Suppl. 8):viii712.
[7] Iwanaga Y, Miyazaki S. Heart failure, chronic kidney disease, and biomarkers
an integrated viewpoint. Circ J 2010;74:127482.
[8] Harnett JD, Foley RN, Kent GM, Barre PE, Murray D, Parfrey PS. Congestive heart
failure in dialysis patients: prevalence, incidence, prognosis and risk factors.
Kidney Int 1995;47:88490.
[9] Parfrey PS, Harnett JD, Foley RN. Heart failure and ischemic heart disease in
chronic uremia. Curr Opin Nephrol Hypertens 1995;4:10510.
[10] Shiba N, Sjimokawa H. Chronic kidney disease and heart failure bidi-
rectional close link and common therapeutic goal. J Cardiol 2011;57:
817.
[11] Levin A, Djurdjev O, Thompson C, Barrett B, Ethier J, Carlisle E, Barre P, Magner
P, Muirhead N, Tobe S, Tam P, Wadgymar JA, Kappel J, Holland D, Pichette
V, et al. Canadian randomized trial of hemoglobin maintenance to prevent
or delay left ventricular mass growth in patients with CKD. Am J Kidney Dis
2005;46:799811.
[12] Roger SD, McMahon LP, Clarkson A, Disney A, Harris D, Hawley C, Kerr P, Lynn
K, Parnham A, Pascoe R, Voss D, Walker R, Levin A. Effects of early and late
intervention with epoetin alpha on left ventricular mass among patients with
chronic kidney disease (stage 3 or 4): results of a randomized clinical trial. J Am
Soc Nephrol 2004;15:14856.
[13] Ritz E, Laville M, Bilous RW, ODonoghue D, Scherhag A, Burger U, de Alvaro
F. Target level for hemoglobin correction in patients with diabetes and CKD:
primary results of the Anemia Correction in Diabetes (ACORD) Study. Am J
Kidney Dis 2007;49:194207.
[14] Singh AK, Szczech L, Tang KL, Barnhart H, Sapp S, Wolfson M, Reddan D, CHOIR
Investigators. Correction of anemia with epoetin alfa in chronic kidney disease.
N Engl J Med 2006;355:208598.
[15] Dreke TB, Locatelli F, Clyne N, Eckardt KU, Macdougall IC, Tsakiris D, Burger
HU, Scherhag A. Normalization of hemoglobin level in patients with chronic
kidney disease and anemia. N Engl J Med 2006;355:207184.
[16] Akaishi M, Suzuki M, Hada Y, Hiroe M, Aonuma K, Tsubakihara Y, Akizawa T.
The favorable effect of continuous adequate treatment of anemia on left ven-
tricular mass in patients with chronic kidney disease in Japan. Eur Soc Cardiol
2007:P1369.
[17] Akizawa T, Gejyo F, Nishi S, Iino Y, Watanabe Y, Suzuki S, Saito A, Akiba T,
Hirakata S, Fukuhara S, Morita S, Horoe M, Hada Y, Suzuki M, Akaishi M,
et al. Positive outcomes of high hemoglobin target in patients with chronic
kidney disease not on dialysis: a randomized controlled study. Ther Apher Dial
2011;15:43140.
[18] Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quanti-
tation in M-mode echocardiography: results of a survey of echocardiographic
measurements. Circulation 1978;58:107283.
[19] Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N.
Echocardiographic assessment of left ventricular hypertrophy: comparison to
necropsy ndings. Am J Cardiol 1986;57:4508.
[20] Pfeffer MA, Burdmann EA, Chen CY, Copper ME, de Zeeuw D, Eckardt KU, Feyzi
JM, Ivanovich P, Kewalramani R, Levey AS, Lewis EF, McGill JB, McMurray JJ,
Parfrey P, Parving HH, et al. A trial of darbepoetin alfa in type 2 diabetes and
chronic kidney disease. New Engl J Med 2009;361:201932.
[21] Phrommintikul A, Haas SJ, Elsik M, Krum H. Mortality and target haemoglobin
concentrations in anaemic patients with chronic kidney disease treated with
erythropoietin: a meta-analysis. Lancet 2007;369:3818.
[22] Parfrey P, Lauve M, Latremouille-Viau D, Lefebvre P. Erythropoietin therapy and
left ventricular mass index in CKD and ESRD patients: a meta-analysis. Clin J
Am Soc Nephrol 2009;4:75562.
[23] Hayashi T, Suzuki A, Shoji T, Togawa M, Okada N, Tsubakihara Y, Imai E, Hori
M. Cardiovascular effect of normalizing the hematocrit level during erythro-
poietin therapy in predialysis patients with chronic renal failure. Am J Kidney
Dis 2000;35:2506.
[24] Suzuki M, Hada Y, Akaishi M, Hirose M, Aonuma K, Tsubakihara Y, Akizawa
T. Effects of anemia correction by erythropoiesis-stimulating agents on car-
diovascular function in non-dialysis patients with chronic kidney disease. Int
Heart J 2012;53:23843.
[25] Anand IS, Kuskowski MA, Rector TS, Florea VG, Glazer RD, Hester A, Chi-
ang YT, Akny N, Maggioni AP, Opasich C, Latini R, Cohn JN. Anemia and
change in hemoglobin over time related to mortality and morbidity in patients
with chronic heart failure: results from Val-HeFT. Circulation 2005;112:
11217.
[26] Anand IS. Anemia and chronic heart failure. J Am Coll Cardiol 2008;52:50111.
[27] Sato Y, Fujiwara H, Takatsu Y. Biochemical markers in heart failure. J Cardiol
2012;59:17.
[28] Tsutamoto T, Wada A, Sakai H, Ishikawa C, Tanaka T, Hayashi M, Fujii M,
Yamamoto T, Dohke T, Ohnishi M, Takashima H, Kinoshita M, Hiroe M. Relation-
ship between renal function and plasma brain natriuretic peptide in patients
with heart failure. J Am Coll Cardiol 2006;47:5826.
[29] Tsubakihara Y, Gejyo F, Nishi S, Iino Y, Watanabe Y, Suzuki M, Saito
A, Akiba T, Hirakata H, Akizawa T. High target hemoglobin with
erythropoiesis-stimulating agents has advantages in the renal function of
non-dialysis chronic kidney disease patients. Ther Apher Dial 2012;16:
52940.