2013
http://informahealthcare.com/bmkISSN: 1354-750X (print), 1366-5804 (electronic)
Biomarkers, 2013; 18(5): 425–435! 2013 Informa UK Ltd. DOI: 10.3109/1354750X.2013.808263
RESEARCH ARTICLE
MDR1 mRNA expression and MDR1 gene variants as predictors ofresponse to chemotherapy in patients with acute myeloid leukaemia:a meta-analysis
Chrysoula Doxani1,2, Michael Voulgarelis2, and Elias Zintzaras1,3
1Department of Biomathematics, University of Thessaly School of Medicine, Larissa, Greece, 2Department of Pathophysiology, Division of
Haematology, University of Athens School of Medicine, Athens, Greece, and 3Center for Clinical Evidence Synthesis, The Institute for Clinical
Research and Health Policy Studies, Tufts Medical Center, Tufts University School of Medicine, Boston, MA, USA
Abstract
Data from 30 pharmacogenomic studies that investigated MDR1 mRNA expression or genevariants (C3435T, G2677TA, C1236T) and response to therapy in acute myeloid leukaemia (AML)were synthesized. Anthracycline-based regimens were mainly used. MDR1 mRNA over-expression was associated with poor response to therapy [odds ratio (OR)¼ 2.49 95%confidence interval (CI) 1.38–4.50]. The gene variants were not associated with response totreatment; the generalized ORs, a genetic model-free approach, for the variants C3435T,G2677TA and C1236T were ORG¼ 0.86 (95% CI 0.55–1.37), ORG¼ 0.97 (95% CI 0.58–1.64) andORG¼ 1.17 (95% CI 0.75–1.83), respectively. There is indication that MDR1 mRNA expressionmay be considered as a potential marker for response to chemotherapy in AML patients.
Keywords
AML, chemotherapy, expression, gene,leukemia, MDR1, meta-analysis, response
History
Received 30 January 2013Revised 21 May 2013Accepted 21 May 2013Published online 20 June 2013
Introduction
Acute myeloid leukaemia (AML) is a heterogeneous clonal
disorder of haemopoietic progenitor cells (‘‘blasts’’) which
lose the ability to differentiate normally and to respond to
normal regulators of proliferation (Estey & Dohner, 2006).
This leads to neutropenia, anaemia and thrombocytopenia
with fatal consequences to the patients, either due to an
infection or bleeding. AML is the most common myeloid
leukaemia with an incidence of 3.69 cases per 100 000 (16.1
for age above 65 years) and the median age at onset is about
70 years (Howlader et. al., 2011).
The standard therapy for AML consists of a combination
of an anthracycline, such as daunorubicin or idarubicin and
cytarabine (ara-C), with the anthracycline often administered
for 3 d and ara-C for 7 d (‘‘3þ 7’’ regimens) (Tallman et al.,
2005). Therapy aims to a prolonged complete remission (CR),
which is defined as a marrow with 55% blasts, a neutrophil
count 41000 and a platelet count 4100 000 (Cheson et al.,
2003). In patients aged less than 60 years, 80% of them
achieve complete remission with a 50% likelihood of relapse
(Burnett et al., 2011). When the remission lasts43 years then,
the likelihood of relapse declines sharply to510% (de Lima
et al., 1997). Response to therapy can vary depending on both
the biological characteristics of the disease and the clinical
characteristics of the patient (Estey, 2002; Tallman et al.,
2005).
MDR1 gene overexpression is considered to be a major
cause of multi-drug resistance and it is implicated in the
response to chemotherapy in AML patients. MDR1 is a
member of the ATP-binding cassette superfamily of trans-
porter proteins and encodes for P-glycoprotein 1 (P-gp),
which is a cellular drug efflux pump transporting toxic
endogenous substances, drugs and xenobiotics out of cells
(Hartmann et al., 2001). The action of P-gp protein in AML
blasts leads to reduced intracellular concentration of cytotoxic
drugs (Leith et al., 1999). As anthracyclines are substrates of
the protein, the failure of AML therapy may be related to
MDR1 (Illmer et al., 2002). MDR1 expression has been
suggested as a disease-related unfavourable prognostic factor
with significant impact on complete remission in AML
(Verhaak & Valk, 2010).
Increased expression of MDR1 has been attributed to
various mechanisms, including altered activity of transcrip-
tional factors, gene rearrangement or change of the methylation
status of the promoter (Kim et al., 2006). It has also been
suggested that MDR1 variants have probably an impact on P-gp
expression and function and in this way, on the effect of drugs
used for treating AML (Gottesman et al., 2002; Illmer et al.,
2002). Two single nucleotide variants, G2677TA in exon 21
and G2995A in exon 24, leading in amino acid changes of
A893S and A999T, were identified to result in changes in P-gp
(Huang, 2007). In addition, two synonymous variants, C3435T
in exon 26 and C1236T in exon 12, have been suggested
Address for correspondence: Prof. Elias Zintzaras, Head, Departmentof Biomathematics, University of Thessaly School of Medicine, 2Panepistimiou St, 41100 Biopolis, Larissa, Greece. E-mail: [email protected]
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
(Huang, 2007) to alter MDR1 mRNA by affecting secondary
structure and turnover. Three of these variants (C3435T,
G2677TA and C1236T) have been mainly studied.
The relationship between the MDR1 mRNA expression
levels or MDR1 gene variants and response to chemotherapy
in patients with AML has been investigated in several
studies producing non-conclusive or contradictory results
(Dahl et al., 2000; Fujimaki et al., 2002; Gsur et al., 1993;
Gruber et al., 1992; Guenova et al., 2010; Huh et al., 2006;
Hunault et al., 1997; Hur et al., 2008; Ho et al., 2008; Illmer
et al., 2002; Kim et al., 2006; Lin et al., 1995; Monzo et al.,
2006; Petti et al., 2003; Sato et al., 1990; Schaich et al., 2002;
Tan et al., 2003; Trnkova et al., 2007; Van den Heuvel-
Eibrink et al., 2007; Van der Holt et al., 2006; Venditti et al.,
2004; Wells et al., 1994). This might be due to limited sample
sizes, design strategies and differences in populations
(Zintzaras & Lau, 2008a). Aim of this study was to decrease
the uncertainty of the pharmacogenomic effect of MDR1
(including mRNA expression levels, both binary and con-
tinuous, and gene variants) on response to AML therapy and
to provide more conclusive evidence regarding the clinical
relevance of this pharmacogenomic association. In order to
achieve this aim, we systematically searched for all relevant
published pharmacogenomic studies and then, the available
data were synthesized using meta-analytic techniques
(Trikalinos et al., 2008; Zintzaras & Lau, 2008a).
Materials and methods
Selection of studies
A comprehensive search of the PubMed was conducted from
its inception through May 2012 to identify all articles that
investigated MDR1 gene expression and response to chemo-
therapy in AML patients. The following keywords were used
as search terms: ‘‘MDR1’’, ‘‘ABCB1’’, ‘‘gene’’, ‘‘gene
polymorphisms’’, ‘‘gene variants’’, ‘‘single nucleotide poly-
morphisms’’, ‘‘SNP’’, ‘‘acute myeloid leukemia’’, ‘‘AML’’,
‘‘idarubicin’’, ‘‘daunorubicin’’, ‘‘cytarabine’’, ‘‘anthracy-
cline’’, ‘‘chemotherapy’’ and ‘‘response’’. The abstracts
were retrieved and screened to assess their appropriateness
for inclusion in the meta-analysis. After the abstract screen-
ing, the retrieved studies were read in their entirety to assess
their eligibility for the analysis. The search was restricted to
English language articles and to articles of studies conducted
in humans. All references cited in the studies were also
reviewed to identify additional published articles not indexed
in the database.
The articles that fulfilled the following criteria were
included in the meta-analysis: (i) provide cases diagnosed
with any type of AML, (ii) use any form of chemotherapy
used as treatment, (iii) define as response the complete
remission of patients and (iv) provide the number of
responders and non-responders with positive and negative
MDR1 mRNA expression (binary outcome) or provide
summary statistics (mean, standard deviation and sample
size) of mRNA expression for responders and non-responders
or provide the genotype distribution of MDR1 variants for
responders and non-responders. Studies that investigated
disease susceptibility, progression, severity or survival were
excluded.
Data extraction
Two investigators independently extracted data. From each
study, the following information was extracted: first author,
journal and year of publication, ethnicity of study population,
demographics, mRNA expression or genotype distribution,
number of responders and non-responders, disease character-
istics and type of chemotherapy.
Data synthesis
Binary MDR1 mRNA expression
The pooled odds ratios (ORs) with the corresponding 95%
confidence interval (CI) of responders (i.e. CR) versus non-
responders (i.e. not CR) to chemotherapy for the negative
mRNA expression relative to positive expression was
estimated using the random effects (RE) model
(DerSimonian & Laird, 1986) because RE is more conserva-
tive than the fixed effects OR (Zintzaras & Lau, 2008a). The
RE model incorporates in the estimates of the between study
variability.
Continuous MDR1 mRNA expression
In synthesizing the difference between CR and not CR in
terms of mRNA expression, the pooled difference in mean
values of mRNA expression (Dm) (standardized effect size)
using the inverse of the variance of Dm as a weighting factor
(Hedges & Olkin, 1985; Whitehead, 2002) was estimated.
Then, the 95% CI of Dm was calculated.
MDR1 gene variants
In synthesizing the studies, the RE pooled generalized OR
(ORG) with the corresponding 95% CI was estimated
(Zintzaras, 2010, 2012). The ORG is a genetic model-free
approach for investigating the pharmacogenomic effect and
provides an estimate of the overall pharmacogenomic risk
effect by utilizing the complete genotype distribution of the
gene variant. The ORG is defined as follows: for any two
subjects, one responder and one non-responder to chemother-
apy, the ORG estimates the odds of being responder relative to
the odds of being non-responder when the responder has
higher mutational load than the non-responder, i.e. the risk of
response is proportional to the increased genetic exposure.
Alternatively, the ORG shows how many responders/non-
responders pairs exist in the study for which the responders
have larger mutational load relative to the number of pairs for
which the non-responders have the larger mutational load
(Zintzaras, 2010, 2012). In addition, the RE pooled ORs of
responders versus non-responders to chemotherapy for reces-
sive and dominant models of each gene variant were
estimated (DerSimonian & Laird, 1986). The OR for the
recessive model expresses the odds of response for homozy-
gotes of the mutant-type allele relative to the odds of response
for carriers of the wild-type allele and the OR for the
dominant model expresses the odds of response for carriers of
the mutant-type allele relative to the odds of response for
homozygotes of the wild-type allele. Thus, the OR for the
recessive and dominant models is estimated after merging
genotypes.
426 C. Doxani et al. Biomarkers, 2013; 18(5): 425–435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
Statistical heterogeneity across the various studies was
tested with the use of Q-statistic (Cochran, 1954; Zintzaras &
Lau, 2008a). A pQ-value 50.10 indicated a significant
heterogeneity across studies (Zintzaras & Lau, 2008a).
Heterogeneity was quantified with the I2 metric (Higgins &
Thompson, 2002; Zintzaras & Ioannidis, 2005). I2 takes
values between 0% and 100% with higher values denoting
greater degree of heterogeneity. When there is no heterogen-
eity the RE model coincides with the fixed effects model
(Zintzaras & Lau, 2008a). In the present meta-analysis, the
RE model was used as it is more conservative.
The meta-analysis consisted of the main (overall) analysis
which includes all available data and subgroup analyses by
ethnicity, type of chemotherapy and age (adults/children)
(Zintzaras & Lau, 2008a,b). A sensitivity analysis (i.e. analysis
after excluding specific studies) was performed for the studies
involving patients that received MDR1 reversal agents. The
differential magnitude of effect in large versus small studies
was examined using the Harbord’s test (Harbord et al., 2006).
Analyses were performed using Meta-Analyst V.3 (Evidence-
Based Practice Centers, Tufts Medical School, Boston, MA)
and CUMAGAS (Larissa, Greece, http://biomath.med.uth.gr)
(Zintzaras & Kitsios, 2009; Zintzaras et al., 2011). The ORG
was estimated using ORGGASMA (Larissa, Greece, http://
biomath.med.uth.gr) (Zintzaras, 2010, 2012).
Results
Eligible studies
The literature search identified 94 articles in PubMed that met
the search criteria. All articles identified through the literature
search were independently screened by two investigators
(C.D. and E.Z.). Figure 1 presents a flow chart of the retrieved
and excluded studies with specification of reasons. Data from
24 articles met the inclusion criteria. These articles involved
21 studies for mRNA expression (17 with binary outcome and
four with continuous outcome), four studies for the variant
C3435T, three studies for the variant G2677TA and two
studies for the variant C1236T. The characteristics of the
individual studies included in the meta-analysis are provided
in Table 1. The studies were published between 1990
and 2009.
Overall, the meta-analysis included 1279/172 patients for
the mRNA expression in binary/continuous scale and 816/
628/520 patients for the variants C3435T/G2677TA/C1236T,
respectively. Nineteen of the articles included patients with de
novo AML. Thirteen articles included chemotherapy-naıve
patients and two studies included only recurrent or refractory
AML patients. Eighteen articles reported the use of
anthracycline-based (daunorubicin or idarubicin) chemother-
apy, two articles used other type of chemotherapy and four
articles did not report the type of chemotherapy. All articles
that investigated gene variants used anthracycline-based
chemotherapy. In three studies (Van der Holt et al., 2006)
that investigated the variants C3435T, G2677TA and C1236T,
more than half of the patients received a P-gp inhibitor (PSC-
833) (58, 54 and 55%, respectively) and in one study that
investigated mRNA expression (in binary scale), 57% of
patients received PSC-833 (Van den Heuvel et al., 2007), The
average age of the patients varied from 8.15 to 67 years.
The articles involved patients from different ethnicities: 17
with Whites, six with East Asians and one with Indians.
Binary MDR1 mRNA expression
Figure 2 shows the results of the main (overall) analysis for
the association between MDR1 mRNA expression levels
(negative/positive) and response to chemotherapy (CR/not
CR). The meta-analysis showed that, overall, the mRNA
expression level is significantly associated with response to
chemotherapy: OR¼ 2.49 95% CI 1.38–4.50, i.e. there is 2.5-
fold higher chance for CR for negative expression than for
positive one. However, the heterogeneity between studies is
large (pQ50.01 with I2¼ 82%). The differential magnitude of
effect in large versus small studies was non-significant
(pH¼ 0.57).
In adults, the pattern of results remained the same as in the
overall analysis: the OR was significant with a similar effect
size (OR¼ 2.17 95% CI 1.10–4.25) and the heterogeneity was
also large (pQ50.01 with I2¼ 83%). When the analysis was
restricted to anthracycline-based chemotherapy, the mRNA
expression level revealed a highly significant association with
response to therapy: OR¼ 2.91, 95% CI 1.38–6.16 with the
heterogeneity being significant (pQ50.01 with I2¼ 85%). In
the subgroup analysis by ethnicity, Whites derived non-
significant association (OR¼ 1.86, 95% CI 0.96–3.61) and
large heterogeneity (pQ50.01 with I2¼ 73%). On the con-
trary, East Asians derived a significant result: OR¼ 8.96, 95%
CI 4.01–20.02, indicating that there is 9-fold higher chance
for CR when the mRNA expression is negative compared to
the positive one. The sensitivity analysis (exclusion of the
study involving patients treated with MDR1 reversal agents)
improved the effect size [OR¼ 2.94, 95% CI 1.82–4.75);
though, the heterogeneity remained significant (pQ¼ 0.01
with I2¼ 51%).
94 titles were found in Pubmed
41 were not relevant (e.g. concerned different diseases or genes or data for protein expression)
5 provided results for other outcomes or no results
21 articles were evaluated 3 articles were retrieved from references
24 articles were finally included
12 articles were reviews
5 articles were not in English
7 were not conducted in humans
Figure 1. Flow chart of studies retrieved and studies excluded, withspecification of reasons.
DOI: 10.3109/1354750X.2013.808263 MDR1 and response to chemotherapy in AML 427
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
Tab
le1
.C
har
acte
rist
ics
of
stu
die
su
sed
.
Auth
or
(yea
r)E
thnic
ity
Subje
cts
(per
cent
mal
es)
Age
inyea
rsM
edia
n(m
in–m
ax)
mea
n(�
SD
)M
DR
1var
iant
or
expre
ssio
nC
linic
alch
arac
teri
stic
sof
pat
ients
Def
init
ion
of
resp
onder
sD
efin
itio
nof
non-r
esponder
sT
ype
of
chem
oth
erap
y
Hur
etal
.(2
008)
Eas
tA
sian
200
(63.5
)44
(�14)
C3435T
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
NR
TR
elap
seor
dea
thId
arubic
in12
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e100
mg/m
2/d
IV(D
1–D
7)
Kim
etal
.(2
006)
Eas
tA
sian
81
(66.6
)39
(15–72)
C3435T
,G
2677T
Ade
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
and
AN
C4
1�
10
9/L
&P
LT
4100�
10
9/L
rela
pse
aspre
sence
of4
5%
bla
stce
lls
inbone
mar
row
inpre
-vio
usl
ynorm
albone
aspir
ates
or
evid
ence
of
extr
amed
ull
ary
leukae
mia
Idar
ubic
in12
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e100
mg/m
2/d
IV(D
1–D
7)
van
der
Holt
etal
.(2
006)
Whit
es130
(57)
67
(60–85)
C1236T
,G
2677T
A,
C3435T
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
,55%
of
pat
ients
rece
ived
aP
-gp
inhib
itor
(58%
for
C3435T
,54%
for
G2677T
Aan
d55%
for
C1236T
)
norm
oce
llula
rbone
mar
row
wit
h5
5%
bla
sts,
no
Auer
rods
and
no
evid
ence
of
extr
amed
ull
ary
involv
emen
t
Pat
ients
inw
hom
AM
Lre
lapse
dor
who
die
dw
ithin
28
day
saf
ter
CR
wer
eco
nsi
der
edas
not
hav
ing
achie
ved
CR
dau
noru
bic
in45
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
IV(D
1–D
7)
(Arm
A)
or
dau
noru
bic
in35
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
ayIV
(D1–D
7)
and
PS
C-8
33
(Arm
B)
Illm
eret
al.
(2002)
NR
T405
(NR
T)
53
(17–78)
C1236T
,G
2677T
A,
C3435T
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
afte
rth
ese
cond
cours
eof
induct
ion
ther
apy
NR
Tdau
noru
bic
in45
mg/m
2(D
1–D
3)
and
cyto
sine
arab
inosi
de
(ara
-C)
100
mg/m
2(D
1–D
7)
Gsu
ret
al.
(1993)
Whit
es75
(48)
56
(18–80)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
NR
TE
D(e
arly
dea
thw
ithin
4w
eeks
afte
rbeg
innin
gof
trea
tmen
t)þ
RD
resi
stan
tdis
-ea
se(a
fter
atle
ast
two
cycl
es
of
chem
oth
erap
y)
dau
noru
bic
in45
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
ayIV
(D1–D
7)
Gru
ber
etal
.(1
992)
Whit
es34
(52)
65
(15–87)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
atesþ
nor-
mal
cell
ula
rity
and
norm
alper
ipher
alblo
od
val
ues
no
CR
afte
rtw
oin
duct
ion
cours
esdau
noru
bic
in-v
incr
isti
ne-
ara-
C,
mit
oxan
trone-
etoposi
de-
ara-
Cor
dau
noru
bic
in-a
ra-C
-thio
guan
ine
Guen
ova
etal
.(2
010)
Whit
es17
(64)
54
(�16.8
)m
RN
Aex
pre
ssio
n(b
inar
y)
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
NR
TN
RT
NR
T
Tan
etal
.(2
003)
Eas
tA
sian
65
(46)
35.4
(6–76)
mR
NA
expre
ssio
n
(bin
ary)
de
novo
and
seco
ndar
yA
ML
,
chem
oth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
edpat
ients
NR
TN
RT
har
ringto
nin
e4
mg/m
2/d
IV(D
1–D
7)-
dau
noru
bic
in45
mg/m
2/d
IV(D
1–
D3)�
cyta
rabin
e200
mg/m
2(D
1–
D7)
or
dau
noru
bic
inþ
cyta
rabin
eor
etoposi
deþ
cyta
rabin
eor
idar
-ubic
inþ
cyta
rabin
eor
mit
oxan
troneþ
cyta
rabin
e
van
den
Heu
vel
-Eib
rink
etal
.(2
007)
Whit
es154
(56)
67
(60–85)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
/57%
of
the
pat
ients
rece
ived
aP
-gp
inhib
itor
norm
oce
llula
rbone
mar
row
wit
h5
5%
bla
sts,
no
Auer
rods
and
no
evid
ence
of
extr
amed
ull
ary
involv
emen
t
Pat
ients
inw
hom
AM
Lre
lapse
dor
who
die
dw
ithin
28
day
saf
ter
whom
AM
Lre
lapse
dor
who
die
dw
ithin
28
day
saf
ter
CR
wer
eco
nsi
der
edas
not
hav
ing
achie
ved
CR
dau
noru
bic
in45
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
IV(D
1–D
7)
(Arm
A)
or
dau
noru
bic
in35
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
ayIV
(D1–D
7)
and
PS
C-8
33
(Arm
B)
Fuji
mak
iet
al.
(2002)
Eas
tA
sian
14
(NR
T)
NR
Tm
RN
Aex
pre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
edpat
ients
CR
:m
ainta
inin
gco
mple
tere
mis
-si
on
acco
rdin
gto
esta
bli
shed
crit
eria
for4
6m
onth
s:ce
llu-
lar
mar
row
wit
h5%
bla
st
cell
s,A
NC�
1.5�
10
9/L
,P
LT
�100�
10
9/L
and
no
evi-
den
ceof
leukae
mia
inoth
ersi
tes,
ER
:re
lapse
wit
hin
6m
onth
sfr
om
rem
issi
on
NR
:ce
llula
rm
arro
ww
ith4
5%
bla
stce
lls
or
evid
ence
of
leu-
kae
mia
inoth
ersi
tes,
afte
rat
leas
ttw
oco
urs
esof
chem
oth
erap
y
NR
T
428 C. Doxani et al. Biomarkers, 2013; 18(5): 425–435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
Hunau
ltet
al.
(1997)
Whit
es98
(57)
54
(�16)
(18–90)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
pre
sence
of5
5%
bla
stce
lls
pro
-m
yel
ocy
tes,
monocy
toid
cell
sor
oth
erel
emen
tsw
hic
hca
nnot
be
clas
sifi
ed
asm
ore
mat
ure
norm
alel
emen
tsin
bone
mar
row
aspir
ates
&A
NC4
1*10
9/L
&P
LT4
100*10
9/L
and
Hgb4
100
g/L
invi
vodru
gre
sist
ance
(45%
bone
mar
row
bla
stce
lls)
and
dea
thduri
ng
apla
sia
Idar
ubic
in12
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e100
mg/m
2/d
IV(D
1–D
7)
Huh
etal
.(2
006)
Eas
tA
sian
34
(NR
T)
30
(3m
onth
sto
5yea
rs)
mR
NA
expre
ssio
n(b
inar
y)
NR
Tce
llula
rity
of4
20%
wit
h5
5%
bla
sts
inbone
mar
row
aspir
-at
esaf
ter
induct
ion
chem
oth
erap
y
cell
ula
rity
of4
20%
wit
h4
5%
bla
sts
inbone
mar
row
aspir
ates
afte
rin
duct
ion
chem
oth
erap
y
adult
AM
L:
cyta
rabin
eþ
idar
ubic
in(A
Ire
gim
en),
chil
dhood
AM
L:
Cycl
ophosp
ham
ideþ
pre
dnis
one
Sch
aich
etal
.(2
002)
Whit
es331
(NR
T)
56
(15–78)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
afte
rth
ese
cond
cours
eof
induct
ion
ther
apy
NR
Tpts�
60
yea
rs:
one
cours
eof
MA
V[m
itoxan
trone
10
mg/m
2(D
4–D
8),
cyto
sine
arab
inosi
de
100
mg/m
2
(D1–D
8),
VP
-16
100
mg/m
2(D
4–
D8)]
and
ase
cond
cours
eof
MA
MA
C[c
yto
sine
arab
inosi
de
2�
1000
mg/m
2(D
1–D
5),
m-
amsa
crin
e100
mg/m
2(D
1–D
5)]
,
pts4
60
yea
rs:
two
cours
esof
DA
[dau
noru
bic
in45
mg/m
2(D
3–D
5),
cyto
sine
arab
inosi
de
100
mg/m
2
(D1–D
7)]
Ven
dit
tiet
al.
(2004)
Whit
es240
(55)
62.5
(18–84)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
Pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
atesþ
nor-
mal
cell
ula
rity
and
norm
alper
ipher
alblo
od
val
ues
Pre
sence
of4
5%
bla
stce
lls
inbone
mar
row
inpre
vio
usl
ynorm
albone
aspir
ates
or
evi-
den
ceof
extr
amed
ull
ary
leukae
mia
regim
ens
bas
edon
cyta
rabin
e,et
opo-
side
and
anth
racy
clin
e(i
dar
ubic
in,
dau
noru
bic
in)
or
mit
oxan
trone
Dah
let
al.
(2000)
Whit
es50
(51)
(1.4
month
sto
20
yea
rs)
mR
NA
expre
ssio
n(b
inar
y)
rela
pse
dor
refr
acto
ryA
ML
,ch
e-m
oth
erap
y-t
reat
edpat
ients
Morp
holo
gic
ally
norm
albone
mar
row
,A
NC�
1500/m
Lan
dP
LT�
100
000/m
L.
PR
:5.1
–25.0
%of
bone
mar
row
bla
stce
lls
wit
hre
cover
ing
counts
and
no
circ
ula
ting
bla
sts.
Rel
apse4
25%
bla
sts
etoposi
de
and
azac
yti
din
ew
ith
itra
thec
alA
ra-Cþ
anth
racy
clin
esan
doth
eran
thra
cycl
ine-
bas
edre
gim
ens
Wel
lset
al.
(1994)
Whit
es11
(91%
)8.1
5(0
.5–15)
mR
NA
expre
ssio
n(b
inar
y)
rela
pse
dor
refr
acto
ryA
ML
,ch
e-m
oth
erap
y-t
reat
edpat
ients
anM
1bone
mar
row
(55%
bla
sts)
,ce
llula
rity
rangin
gfr
om
mod-
erat
ely
hypoce
llula
rto
hyper
-ce
llula
r,A
NC4
1000/m
m3
and
PL
T4
100
000/m
m3
PR
:M
2bone
mar
row
level
s(5
–40%
for
AM
L)
or
ifth
em
arro
wre
ached
M1,
wit
hout
reco
ver
yof
the
per
ipher
alblo
od
counts
.A
lloth
er
resp
onse
sw
ere
induct
ion
fail
ure
s.
arab
inosi
de1
g/m
2(D
1–
D4)þ
mit
oxan
trone
12
mg/m
2
(D3–D
7)
Trn
kova
etal
.(2
007)
Whit
es57
(47)
43.8
(0.4
5–78)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
AM
LC
R:
NR
Tno
CR
var
ious
pro
toco
ls(n
odet
ails
)
Lin
etal
.(1
995)
Eas
tA
sian
20
(75%
)46.5
(18–75)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
ed
pat
ients
CR
:N
RT
earl
ydea
thor
resi
stan
tdis
ease
(1)
Dau
nom
yci
n45
mg/m
2/d
(D1–
D3),
cyta
rabin
100
mg/m
2/d
(D1–
D7),
etoposi
de
100
mg/m
2/d
(D1–
D5),
(2)
dau
nom
yci
n45
mg/m
2/d
(D1–D
3),
cyta
rabin
e200
mg/m
2/d
(D1–D
7),
(3)
cyta
rabin
20
mg/m
2/
d(D
1–D
14)
Ho
etal
.(2
008)
Whit
es31
(57)
41.5
(17–69)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
and
AN
C4
1�
10
9/L
and
PL
T4
100�
10
9/L
and
Hgb4
100
g/L
NR
:no
resp
onse
toin
duct
ion
ther
apy
dau
noru
bic
in45
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
IV(D
1–D
7)
Sat
oet
al.
(1990)
Whit
es36
(40.7
)53.9
(11–82)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
edpat
ients
CR
:re
turn
tonorm
alhae
mopoie
sis
RD
dau
noru
bic
in45
mg/m
2/d
IV(D
1–D
3)
and
Cyta
rabin
e200
mg/m
2/d
IV(D
1–D
7)
(co
nti
nu
ed)
DOI: 10.3109/1354750X.2013.808263 MDR1 and response to chemotherapy in AML 429
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
Tab
le1
.C
on
tin
ued
Auth
or
(yea
r)E
thnic
ity
Subje
cts
(per
cent
mal
es)
Age
inyea
rsM
edia
n
(min
–m
ax)
mea
n(�
SD
)M
DR
1var
iant
or
expre
ssio
nC
linic
alch
arac
teri
stic
sof
pat
ients
Def
init
ion
of
resp
onder
sD
efin
itio
nof
non-r
esponder
sT
ype
of
chem
oth
erap
y
Pet
tiet
al.
(2003)
Whit
es12
(50)
54.8
(31–64)
mR
NA
expre
ssio
n(b
inar
y)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
edpat
ients
CR
:pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
and
norm
alper
ipher
alblo
od
val
ues
and
norm
alcl
inic
alfi
ndin
gs.
PR
:515%
bla
sts
innorm
alhyper
cell
ula
rm
arro
wan
dnorm
alphysi
cal
findin
gs
RD
Hig
h-d
ose
hydro
xyure
a
Xu
etal
.(1
999)
Whit
es52
(NR
T)
66
(22–91)
mR
NA
expre
ssio
n(c
onti
nuous)
de
novo
and
seco
ndar
yA
ML
,ch
emoth
erap
y-n
aıve
pat
ients
and
chem
oth
erap
y-t
reat
edpat
ients
CR
:N
RT
AM
Lth
atpro
gre
ssed
afte
rin
ten-
sive
chem
oth
erap
y,ei
ther
atdia
gnosi
sor
atre
lapse
.
NR
T
Har
tet
al.
(1997)
Whit
es33
(NR
T)
NR
Tm
RN
Aex
pre
ssio
n(c
onti
nuous)
de
novo
and
seco
ndar
yA
ML
CR
:pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
and
norm
alper
ipher
alblo
od
val
ues
stan
dar
dre
gim
ens
conta
inin
gan
thra
-cy
clin
es,
cyto
sine
arab
inosi
de,
VP
16
Chau
han
etal
.(2
012)
India
ns
30
(44)
35
(19–85)
mR
NA
expre
ssio
n(c
onti
nuous)
de
novo
AM
L,
chem
oth
erap
y-
naı
ve
pat
ients
CR
:pre
sence
of5
5%
bla
stce
lls
inbone
mar
row
aspir
ates
and
norm
alper
ipher
alblo
od
val
ues
NR
:no
CR
dau
noru
bic
in45
mg/m
2(D
1–D
3)
and
cyto
sine
arab
inosi
de
(ara
-C)
100
mg/m
2(D
1–D
7)
Di:
Day
iP
R:
par
tial
resp
on
seN
R:
no
resp
on
seR
D:
resi
stan
td
isea
seN
RT
:n
ot
rep
ort
ed
430 C. Doxani et al. Biomarkers, 2013; 18(5): 425–435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
Quantitative MDR1 mRNA expression
Figure 3 shows the pooled estimate of the difference between
CR and not CR in terms of MDR1 mRNA expression. The
pooled mean difference for comparing CRs versus non-CRs in
terms of mRNA expression yielded non-significant result
[Dm¼�0.069 (�0.48, 0.34)] and the heterogeneity was also
non-significant, but a bit large (pQ¼ 0.20 with I2¼ 36%). The
differential magnitude of effect in large versus small studies
was non-significant (pH¼ 0.68).
MDR1 gene variants
The results of the meta-analysis are shown in Table 2 and
Figure 4(a) and (b). Overall, both the model-free approach
and the genetic models (recessive and dominant) yielded no
significant association between the examined variants
(C3435T, G2677TA and C1236T) and response to chemo-
therapy. Also, the subgroup analysis by ethnicity and age
group produced no significant results.
For variants C3435T and G2677TA, the heterogeneity
varied from none (pQ¼ 0.61) to high (pQ¼ 0.01), whereas for
variant C1236T, the heterogeneity was not significant
(pQ40.15). The sensitivity analysis (exclusion of studies
involving patients treated with MDR1 reversal agents) did not
alter the pattern of results. There was a significant differential
magnitude of effect in large versus small studies (pH¼ 0.04).
Discussion
The aim of this study was to determine whether the
knowledge of MDR1 mRNA expression or gene variants
could predict the response to treatment in patients with AML.
The results suggested that MDR1 mRNA overexpression may
be considered as a predictor for poor response to anthracy-
cline-based chemotherapy. Subgroup analysis showed that in
East Asians, mRNA-negative patients had 9-fold higher
chance for CR, whereas in Whites, the effect was not
significant; this discrepancy might be due to the differences in
mutant allele frequencies among ethnicities that affect gene’s
expression (Cascorbi, 2011). The sensitivity analysis for the
study involving patients treated with MDR1 reversal agents
(Van den Heuvel et al., 2007; Van der Holt et al., 2006;)
improved the effect size, indicating that adding a P-gp
inhibitor, such as PSC-833, to standard treatment, may cause
an increase in response to treatment for the patients express-
ing the gene protein.
0.001 0.01 0.1 0.2 0.5 1 2 5 10 100 1000
Guenova, 2010
Ho, 2007
Trnkova, 2007
Huh, 2006
Venditti, 2004
Schaich, 2004
Petti, 2003
Tan, 2003
Fujimaki, 2002
Dahl, 2000
Hunault, 1997
Lin, 1995
Wells, 1994
Gsur, 1993
Gruber, 1992
Sato, 1990
pooled effect size
OR (95% CI)
Van den Heuvel, 2007
Figure 2. RE OR estimates with the corresponding 95% CI for the association between MDR1 mRNA expression and response to chemotherapy. TheOR estimate of each study is marked with a solid black square. The size of the square represents the weight that the corresponding study exerts in themeta-analysis. The CIs of pooled estimates are displayed as a horizontal line through the diamond; this line might be contained within the diamond ifthe CI is narrow. The horizontal axis is plotted on a log scale.
DOI: 10.3109/1354750X.2013.808263 MDR1 and response to chemotherapy in AML 431
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
However, the available evidence showed that the MDR1
gene variants may not be considered as predictors of response
to chemotherapy based on the current evidence. A consist-
ency between the results of mRNA expression analysis and
gene variants analysis would also be expected, because gene
variants probably have an impact on MDR1 gene expression
and function either by changing the amino acid sequence
(G2677TA) or by changing the secondary structure of the
mRNA (C3435T and C1236T). The inconsistency derived
could be attributed to the large heterogeneity between studies,
-2 -1 1
Chauhan, 2012
Hart , 1997
Trnkova, 2007
Xu, 1999
0
Δµ (95% CI)
pooled effect size
Figure 3. REs pooled estimate of the difference (Dm), with the respective 95% CI, in MDR1 mRNA expression between complete remission and not-complete remission. The Dm estimate of each study is marked with a solid black square. The size of the square represents the weight that thecorresponding study exerts in the meta-analysis. The CIs of pooled estimates are displayed as a horizontal line through the diamond; this line might becontained within the diamond if the CI is narrow. The horizontal axis is plotted on a log scale.
Table 2. Random effects odds ratios (ORs) with the corresponding 95% confidence intervals (CI) and heterogeneity results for genetic contrasts ofgenotypes of MDR1 variants C3435T, G2677TA and C1236T and response to chemotherapy.
Genetic model Population No. of studies Random effects OR (95% CI) I2 (%) P Q-test
C3435TModel-free All 4 0.86 (0.55–1.37) 64 0.04
East Asians 2 0.63 (0.26–1.53) na 0.08Adults 3 1.08 (0.80–1.47) 20 0.29
All without PSC-833 3 0.86 (0.45–1.62) 73 0.02Recessive model All 4 1.02 (0.73–1.44) 0 0.51
East Asians 2 0.85 (0.40–1.81) na 0.61Adults 3 1.06 (0.75–1.50) 0 0.41
All without PSC-833 3 1.14 (0.77–1.68) 0 0.59Dominant model All 4 0.86 (0.47–1.56) 63 0.04
East Asians 2 0.54 (0.16–1.81) na 0.07Adults 3 1.20 (0.85–1.68) 0 0.41
All without PSC-833 3 0.81 (0.35–1.88) 75 0.02G2677TAModel-free All 3 0.97 (0.58–1.64) 66 0.05
All without PSC-833 2 0.90 (0.34–2.36) na 0.02Recessive model All 3 1.30 (0.87–1.94) 0 0.55
All without PSC-833 2 1.35 (0.82–2.36) na 0.30Dominant model All 3 0.80 (0.35–1.84) 73 0.03
All without PSC-833 2 0.56 (0.07–4.40) na 0.01C1236TModel-free All 2 1.17 (0.75–1.83) na 0.18Recessive model All 2 1.04 (0.52–2.06) na 0.15Dominant model All 2 1.33 (0.93–1.90) na 0.36
na: non-applicable.
432 C. Doxani et al. Biomarkers, 2013; 18(5): 425–435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
the limited number of studies investigated the effect of
variants and/or the existence of other unknown variants with a
pharmacogenomic effect that are in linkage disequilibrium
with the examined ones. A possible pharmacogenomic
convergence of the different data sources (mRNA expression
and gene variant studies) might be necessary in order to draw
definite conclusion on the effect of MDR1 in response to
chemotherapy in AML and to proceed with standard
pharmacogenomic testing (Kitsios & Zintzaras, 2009).
The associations presented in these meta-analyses resulted
from pooling a relatively small number of studies and patients
with large heterogeneity between studies. In addition, the
impact of effect modifiers such as age, the pre-treatment
cytogenetic and molecular genetic findings in AML blasts,
de novo or therapy-related AML and response to chemother-
apy was not considered as the individual studies did not
provide the relevant data. The present report included studies
which varied in terms of study design and methodology,
sample sizes, inclusion criteria and definition of cut-off points
for over- and under-expression, Thus, the results should be
interpreted with caution. Future analyses with more studies of
homogeneous groups, with strict inclusion criteria and large
sample sizes will provide more reliable results. Thus, lack of
significant association in the gene variants analysis does not
exclude the possibility of an association.
In conclusion, there is evidence that MDR1 mRNA
overexpression in AML patients is related to poor response
to chemotherapy. On the contrary, the variants C3435T,
G2677TA and C1236T are not associated with response to
treatment. However, the minor contributing role of these
variants cannot be totally excluded. The current evidence does
not justify the implementation of a pharmacogenetic test
Figure 4. RE ORG estimates with the cor-responding 95% CI for (a) the MDR1C3435T (T versus C) variant and (b) theMDR1 G2677TA (TA versus G) and responseto chemotherapy. The horizontal axis isplotted on a log scale.
Illmer, 2002
Kim, 2006
van der Holt, 2006
Hur, 2008
pooled effect size
ORG (95% CI)
21.9.8.7.6.5.4.3.2.1
Illmer, 2002
Kim, 2006
van der Holt, 2006
pooled effect size
ORG (95% CI)
21.9.8.7.6.5.4.3.2
(a)
(b)
DOI: 10.3109/1354750X.2013.808263 MDR1 and response to chemotherapy in AML 433
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
before initiation of chemotherapy treatment in AML patients.
Thus, large and rigorous pharmacogenomic studies are
needed in order to provide more conclusive evidence on the
role of this genetic marker to respond to chemotherapy.
Declaration of interest
The authors declare no conflicts of interest. The authors alone
are responsible for the content and writing of this article.
References
Burnett A, Wetzler M, Lowenberg B. (2011). Therapeutic advances inacute myeloid leukemia. J Clin Oncol 29:487–94.
Cascorbi I. (2011). P-glycoprotein: tissue distribution, substrates, andfunctional consequences of genetic variations. Handb Exp Pharmacol201:261–83.
Chauhan PS, Bhushan B, Singh LC, et al. (2012). Expression of genesrelated to multiple drug resistance and apoptosis in acute leukemia:response to induction chemotherapy. Exp Mol Pathol 92:44–9.
Cheson B, Bennett J, Kopecky K, et al. (2003). Revised recommenda-tions of the International Working Group for diagnosis, standardiza-tion of response criteria, treatment outcomes, and reporting standardsfor therapeutic trials in acute myeloid leukemia. J Clin Oncol 21:4642–9.
Cochran WG. (1954). Some methods for strengthening the common �2tests. Biometrics 10:417–51.
Dahl GV, Lacayo NJ, Brophy N, et al. (2000). Mitoxantrone, etoposide,and cyclosporine therapy in pediatric patients with recurrent orrefractory acute myeloid leukemia. J Clin Oncol 18:1867–75.
de Lima M, Strom S, Keating M, et al. (1997). Implications of potentialcure in acute myelogenous leukemia: development of subsequentcancer and return to work. Blood 90:4719–24.
DerSimonian R, Laird N. (1986). Meta-analysis in clinical trials. ControlClin Trials 7:177–88.
Estey EH. (2002). Treatment of acute myelogenous leukemia. Oncology16:343–52.
Estey E, Dohner H. (2006). Acute myeloid leukaemia. Lancet 368:1894–907.
Fujimaki S, Funato T, Harigae H, et al. (2002). Quantitative analysis of aMDR1 transcript for prediction of drug resistance in acute leukemia.Clin Chem 48:811–17.
Gottesman MM, Fojo T, Bates SE. (2002). Multidrug resistance incancer: Role of ATP-dependent transporters. Nat Rev Cancer 2:48–58.
Gruber A, Vitols S, Norgren S, et al. (1992). Quantitative determinationof mdr1 gene expression in leukaemic cells from patients with acuteleukaemia. Br J Cancer 66:266–72.
Gsur A, Zochbauer S, Gotzl M, et al. (1993). MDR1 RNA expression asa prognostic factor in acute myeloid leukemia: an update. LeukLymphoma 12:91–4.
Guenova ML, Balatzenko GN, Nikolova VR, et al. (2010). An anti-apoptotic pattern correlates with multidrug resistance in acute myeloidleukemia patients: a comparative study of active caspase-3, cleavedPARPs, Bcl-2, Survivin and MDR1 gene. Hematology 15:135–43.
Harbord RM, Egger M, Sterne JA. (2006). A modified test for small-study effects in meta-analyses of controlled trials with binaryendpoints. Stat Med 25:3443–57.
Hart SM, Ganeshaguru K, Scheper RJ, et al. (1997). Expression of thehuman major vault protein LRP in acute myeloid leukemia. ExpHematol 25:1227–32.
Hartmann G, Kim H, Piquette-Miller M. (2001). Regulation of thehepatic multidrug resistance gene expression by endotoxin andinflammatory cytokines inmice. Int Immunopharmacol 1:189–99.
Hedges LV, Olkin I. (1985). Statistical methods for meta-analysis.Orlando: Academic Press.
Higgins JP, Thompson SG. (2002). Quantifying heterogeneity in a meta-analysis. Stat Med 21:1539–58.
Ho MM, Hogge DE, Ling V. (2008). MDR1 and BCRP1 expression inleukemic progenitors correlates with chemotherapy response in acutemyeloid leukemia. Exp Hematol 36:433–42.
Howlader N, Noone AM, Krapcho M, et al. (2011). SEER cancerstatistics review, 1975–2008. Bethesda (MD): National CancerInstitute.
Huang Y. (2007). Pharmacogenetics/genomics of membrane transportersin cancer chemotherapy. Cancer Metastasis Rev 26:183–201.
Huh HJ, Park CJ, Jang S, et al. (2006). Prognostic significance ofmultidrug resistance gene 1 (MDR1), multidrug resistance-relatedprotein (MRP) and lung resistance protein (LRP) mRNA expression inacute leukemia. J Korean Med Sci 21:253–8.
Hunault M, Zhou D, Delmer A, et al. (1997). Multidrug resistance geneexpression in acute myeloid leukemia: major prognosis significancefor in vivo drug resistance to induction treatment. Ann Hematol 74:65–71.
Hur EH, Lee JH, Lee MJ, et al. (2008). C3435T polymorphism of theMDR1 gene is not associated with P-glycoprotein function ofleukemic blasts and clinical outcome in patients with acute myeloidleukemia. Leuk Res 32:1601–4.
Illmer T, Schuler US, Thiede C, et al. (2002). MDR1 gene polymorph-isms affect therapy outcome in acute myeloid leukemia patients.Cancer Res 62:4955–62.
Kim DH, Park JY, Sohn SK,et al. (2006). Multidrug resistance-1 genepolymorphisms associated with treatment outcomes in de novo acutemyeloid leukemia. Int J Cancer 118:2195–201.
Kitsios GD, Zintzaras E. (2009). Genomic convergence of genome-wideinvestigations for complex traits. Ann Hum Genet 73:514–19.
Leith CP, Kopecky KJ, Chen IM, et al. (1999). Frequency andclinical significance of the expression of the multidrug resist-ance proteins MDR1/P-glycoprotein, MRP1, and LRP in acutemyeloid leukemia: a Southwest Oncology Group Study. Blood94:1086–99.
Lin SF, Huang SM, Chen TP, et al. (1995). MDR1 gene expression inacute myeloid leukemia: clinical correlation. J Formos Med Assoc 94:111–16.
Monzo M, Brunet S, Urbano-Ispizua A, et al. (2006). Genomicpolymorphisms provide prognostic information in intermediate-riskacute myeloblastic leukemia. Blood 107:4871–9.
Petti MC, Tafuri A, Latagliata R, et al. (2003). High-dose hydroxyurea inthe treatment of poor-risk myeloid leukemias. Ann Hematol 82:476–80.
Sato H, Preisler H, Day R, et al. (1990). MDR1 transcript levels as anindication of resistant disease in acute myelogenous leukaemia. Br JHaematol 75:340–5.
Schaich M, Harbich-Brutscher E, Pascheberg U, et al. (2002).Association of specific cytogenetic aberrations with mdr1 geneexpression in adult myeloid leukemia and its implication in treatmentoutcome. Haematologica 87:455–64.
Tallman MS, Gilliland DG, Rowe JM. (2005). Drug therapy of acutemyeloid leukemia. Blood 106:1154–63.
Tan Y, Li G, Zhao C, et al. (2003). Expression of sorcin predicts pooroutcome in acute myeloid leukemia. Leuk Res 27:125–31.
Trikalinos TA, Salanti G, Zintzaras E, Ioannidis JP. (2008). Meta-analysis methods. Adv Genet 60:311–34.
Trnkova Z, Bedrlıkova R, Markova J, et al. (2007). SemiquantitativeRT-PCR evaluation of the MDR1 gene expression in patients withacute myeloid leukemia. Neoplasma 54:383–90.
Van den Heuvel-Eibrink MM, van der Holt B, Burnett AK, et al. (2007).CD34-related coexpression of MDR1 and BCRP indicates a clinicallyresistant phenotype in patients with acute myeloid leukemia (AML) ofolder age. Ann Hematol 86:329–37.
Van der Holt B, Van den Heuvel-Eibrink MM, Van Schaik RH, et al.(2006). ABCB1 gene polymorphisms are not associated with treat-ment outcome in elderly acute myeloid leukemia patients.Clin Pharmacol Ther 80:427–39.
Venditti A, Del Poeta G, Maurillo L, et al. (2004). Combined analysis ofbcl-2 and MDR1 proteins in 256 cases of acute myeloid leukemia.Haematologica 89:934–9.
Verhaak RG, Valk PJ. (2010). Genes predictive of outcome and novelmolecular classification schemes in adult acute myeloid leukemia.Cancer Treat Res 145:67–83.
Wells RJ, Odom LF, Gold SH, et al. (1994). Cytosine arabinoside andmitoxantrone treatment of relapsed or refractory childhood leukemia:initial response and relationship to multidrug resistance gene 1. MedPediatr Oncol 22:244–9.
Whitehead A. (2002). Meta-analysis of controlled clinical trials.Chichester, UK: John Wiley & Sons, Ltd.
Xu D, Arestrom I, Virtala R, et al. (1999). High levels of lung resistancerelated protein mRNA in leukaemic cells from patients with acutemyelogenous leukaemia are associated with inferior response to
434 C. Doxani et al. Biomarkers, 2013; 18(5): 425–435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.
chemotherapy and prior treatment with mitoxantrone. Br J Haematol106:627–33.
Zintzaras E. (2010). The generalized odds ratio as a measure of geneticrisk effect in the analysis and meta-analysis of association studies. StatAppl Genet Mol Biol 9: Article 21.
Zintzaras E. (2012). The power of generalized odds ratio in assessingassociation in genetic studies with known mode of inheritance. J ApplStat 39:2569–81.
Zintzaras E, Ioannidis JP. (2005). Heterogeneity testing in meta-analysisof genome searches. Genet Epidemiol 28:123–37.
Zintzaras E, Lau J. (2008a). Synthesis of genetic association studiesfor pertinent gene-disease associations requires appropriate
methodological and statistical approaches. J Clin Epidemiol 61:634–45.
Zintzaras E, Lau J. (2008b). Trends in meta-analysis of geneticassociation studies. J Hum Genet 53:1–9.
Zintzaras E, Kitsios GD. (2009). Synopsis and synthesis ofcandidate-gene association studies in chronic lymphocytic leukemia:the CUMAGAS-CLL information system. Am J Epidemiol 170:671–8.
Zintzaras E, Doxani C, Ziogas DC, et al. (2011). Bone mineral densityand genetic markers involved in three connected pathways (focaladhesion, actin cytoskeleton regulation and cell cycle): theCUMAGAS-BMD information system. Biomarkers 16:698–708.
DOI: 10.3109/1354750X.2013.808263 MDR1 and response to chemotherapy in AML 435
Bio
mar
kers
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f Q
ueen
slan
d on
10/
08/1
3Fo
r pe
rson
al u
se o
nly.