저 시-비 리- 경 지 2.0 한민

는 아래 조건 르는 경 에 한하여 게

l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.

다 과 같 조건 라야 합니다:

l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.

l 저 터 허가를 면 러한 조건들 적 되지 않습니다.

저 에 른 리는 내 에 하여 향 지 않습니다.

것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.

Disclaimer

저 시. 하는 원저 를 시하여야 합니다.

비 리. 하는 저 물 리 목적 할 수 없습니다.

경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.

약학 사 학 논

골수성백 병 치료에서

약물반응 인자 분석

Factors influencing treatments outcomes

in myeloid leukemia

2012 8월

울 학 학원

약학과 ·임상약학 공

경 임

학 논문 원문제공 서비스에 한 동의서

본인의 학 논문에 하여 서울 학교가 아래와 같이 학 논문 작물을

제공하는 것에 동의합니다.

1.동의사항

① 본인의 논문을 보존이나 인터넷 등을 통한 온라인 서비스 목 으로

복제할 경우 작물의 내용을 변경하지 않는 범 내에서의 복제를

허용합니다.

② 본인의 논문을 디지털화하여 인터넷 등 정보통신망을 통한 논문의

일부 는 부의 복제․배포 송 시 무료로 제공하는 것에

동의합니다.

2.개인( 작자)의 의무

본 논문의 작권을 타인에게 양도하거나 는 출판을 허락하는 등

동의 내용을 변경하고자 할 때는 소속 학(원)에 공개의 유보 는

해지를 즉시 통보하겠습니다.

3.서울 학교의 의무

① 서울 학교는 본 논문을 외부에 제공할 경우 작권

보호장치(DRM)를 사용하여야 합니다.

② 서울 학교는 본 논문에 한 공개의 유보나 해지 신청 시 즉시

처리해야 합니다.

논문제목 :골수성백 병 치료에서 약물반응 인자 분석

학 구분 :박사

학 과 :약학과

학 번 :2007­30944

연 락 처 :

작 자 :김 경 임 (인)

제 출 일 :2012년 8월 1일

서울 학교총장 귀하

-i-

Abstract

Factors influencing treatments outcomes

in myeloid leukemia

Kyung Im Kim

Clinical Pharmacy, Department of Pharmacy

The Graduate School

Seoul National University

Acute myeloid leukemia (AML) is a rapidly proliferating clonal

disorder of hematopoietic stem cells. Since AML is a clinical and

biological heterogenous disease, AML patients are divided into

three cytogenetically defined risk groups with significant

differences in overall survival (OS). However, large

inter‑individual differences in treatment response and

development of resistance are still major drawbacks in AML.

Cytarabine arabinoside (ara‑C) is the key agent for treating

AML, but there is also considerable heterogeneity in the

outcomes for individual patients in same risk group. In addition,

up to 50% of AML patients show no abnormalities by

conventional cytogenetics at diagnosis. These normal karyotype

AML (NK‑AML) patients are prognostically heterogeneous

-ii-

although categorized in the intermediate‑risk group. There is also

an inter‑ethnic difference in treatment outcomes among AML

patients.

Genetic or genomic alterations may affect the expression

and/or function of specific drug protein targets and explain, at

least in part, the inter‑individual variations in the response to

specific treatments. The genetic variations, such as single

nucleotide polymorphisms (SNPs), in the genes encoding the

ara‑C transport and metabolizing pathways may play an important

role in the clinical outcomes in AML patients. Since the drug

response is the result of an interaction of numerous genetic

combinations, the combined effects of SNPs via gene‑gene

interactions as well as the effect of individual SNP may explain

the different clinical outcomes between patients. Copy number

variation (CNV) is a common type of genomic structure

variation. CNVs also have recently attracted considerable interest

as a source of genomic variation because they may play an

important role in the etiology of complex diseases and in

evolution. CNVs, depending on their size and location, are as

important as SNPs for producing variations in treatment efficacy

and/or adverse responses to chemotherapy.

To identify the susceptible genetic or genomic alterations

affecting the clinical outcomes of AML patients receiving ara‑C

based chemotherapy, we genotyped 139 SNPs of 10 candidate

genes within the ara‑C transport and metabolic pathway using

-iii-

the Illumina GoldenGate Genotyping Assay (Illumina Inc., San

Diego, CA, USA) in 97 patients with previously non‑treated de

novo AML other than M3. For 30 NK‑AML patients, we

determined the frequency of genome‑wide cytogenetic CNV

aberrations using HelixTreeⓇ software version 5.2.0 (Golden

Helix Inc., Bozeman, MT, USA). Bone marrow aspirates and

blood from AML patients were provided at the time of diagnosis

for genotyping and copy number analysis.

For SNP analysis, both effect of single SNP and SNP‑SNP

interaction on treatment outcomes were tested. And we tested

three different genetic models, including dominant, recessive, and

additive model. In multivariate anlaysis, SNP rs4694362 (CC

genotype) in DCK gene, individually, was a significant poor

prognostic factor for OS (HR, 33.202 [95% CI, 4.937-223.273],

P < 0.0001, PBonferroni = 0.017). In addition to the single SNP

effect on treatment outcomes, multivariate analysis revealed that

the presence of the SLC29A1 rs3734703 (AA or AC genotype)

in combination with TYMS rs2612100 (AA genotype) was

significantly associated with shorter relapse free survival (RFS)

compared to the combination with wild type (HR, 17.630 [95%

CI, 4.829-64.369], P < 0.0001, PBonferroni = 0.021). The effect of

these SNP‑SNP interaction also decreased the survival time,

although not statistically significant after the multiple test

adjustment (HR, 23.523 [95% CI, 4.616-119.873], P = 0.0001).

In addition, CDA rs10916827 (GG genotype) in combination with

-iv-

DCTD rs17331744 (TC or CC genotype) was associated with

less survival time (HR, 31.680, [95% CI, 6.152-162.905], P <

0.0001, PBonferroni = 0.052). These results suggest that a single

SNP and SNP‑SNP interactions may help to predict the drug

response and provide a guide in developing individualized

chemotherapy for AML patients receiving ara‑C based

chemotherapy.

For 30 NK‑AML patients, possible associations between

cytogenetic aberrations and clinical parameters were analyzed.

CNVs were identified in 23 (76.7%) of the 30 cases tested.

Multivariate analyses controlled for other clinical covariates

showed that patients having copy number loss had a decreased

probability of complete remission (OR, 0.015 [95% CI, 0-0.737],

P = 0.035). And patients who had a copy number gain of more

than four regions tended to have shorter RFS (P = 0.083) with

multivariate analysis showing that CNV increase is an

independent predictive factor for increased risk of relapse (HR,

22.104 [95% CI, 1.644-297.157], P = 0.020). In addition, we

identified genes in recurrent CNV regions utilizing data from the

University of Canada database

(http://projects.tcag.ca/variation/project.html) and the PharmGKB

(http://www.pharmgkb.org/index.jsp). It involved nineteen

previously reported AML-related genes, including HES5,

PRDM16, TNFRSF25, MTX2, TERT, ABCB8, PTP4A3, PBX3,

VENTX, AKT1, KIAA0284, ABCA3, CBFA2T3, FANCA, MLLT6,

-v-

CD7, PRTN3, CEBPA, and TYMP. Among drug-related genes,

NOS3, ERCC1, ERCC2, ATP5I, ATP5D, CYBA, NDUFS7,

SLC19A1, and P2RX1 are known to be related to ara‑C or

anthracycline response. These results suggest that CNVs may

affect the success of ara‑C and anthracycline based

chemotherapy in Korean NK‑AML patients.

In addition, the population specificity in allele frequencies

of the 139 SNPs through inter‑ethnic comparisons was assessed

in this study. For this analysis, the International HapMap

(http://hapmap.ncbi.nlm.nih.gov/) and the 1000 genomes database

(http://www.1000genomes.org/) was used. FST statistic are

calculated between Korean and other populations. Overall, there

were large differences in allele frequencies between Korean and

Caucasian or African, whereas Chinese and Japanese populations

were extremely similar to Korean. The SNPs which showed a

significant relationship with the response to ara‑C based

chemotherapy in this study represented a large divergence for

the comparisons with other populations; DCK rs4694362

(comparison with African, FST = 0.519), SLC29A1 rs3734703

(comparison of Caucasian, FST = 0.136), and TYMS rs2612100

(comparison of Caucasian, FST = 0.195).

In conclusion, the results of this study have important

implications in providing fundamental and useful information for

predicting the treatment outcomes in Korean AML patients, and

may help the development of more appropriate therapeutic

-vi-

modalities.

Keywords : Acute myeloid leukemia, Cytarabine, Resistance,

Single nucleotide polymorphism, Copy number variation, Ethnicity

Student number : 2007­30944

-vii-

Contents

Abstract ······························································································ i

1. Introduction ················································································· 1

1.1 Acute myeloid leukemia (AML)　and ara‑C ············ 1

1.2 Genetic/genomic alterations in drug response ······· 3

1.3 Aims and scopes ································································ 4

2. Materials and Methods ···························································· 6

2.1 Study population and their treatments ······················ 6

2.2 SNP genotyping ································································· 6

2.3 Copy number analysis and gene identification ······· 8

2.4 SNP database ······································································· 8

2.5 Evaluation of clinical response and toxicity ············ 9

2.6 Statistical analysis ······························································ 10

3. Results ························································································· 12

3.1 Patients’characteristics and treatment outcomes

····································································································· 12

3.2 Effect of　 single SNP or SNP‑SNP interactions on

ara‑C based treatment outcomes ································· 14

3.3 CNVs and gene identification ········································ 15

3.4 Correlation of CNVs with ara‑C based treatment

outcomes ················································································ 16

3.5 Comparison of the allele frequencies of the SNPs

between populations ·························································· 17

-viii-

4. Discussion and conclusion ····················································· 19

5. References ················································································· 26

Figure legends ··············································································· 40

Tables ································································································ 41

Figures ······························································································ 68

국 ··························································································· 73

감사 ··························································································· 78

-ix-

List of Tables

Table 1. The 139 candidate SNPs in this study ·············· 41

Table 2. AML patients characteristics and treatment

outcomes ········································································· 43

Table 3. Comparison between patients with and without

CNVs ················································································· 45

Table 4. Combined effect of SNPs on survival in AML

patients ············································································· 47

Table 5. Genes in recurrently altered CNV regions and

their characteristics ···················································· 48

Table 6. Comparison of CNVs of NK‑AML patients with CR

vs. non‑CR ······································································ 58

Table 7. Comparison of RFS by clinical and molecular

characteristics of NK‑AML patients ····················· 60

Table 8. FST values for pair‑wise comparisons between

Korean and the HapMap populations in descending

rank order ······································································· 62

-x-

List of Figures

Figure 1. Ara‑C transport and metabolic pathway ··········· 68

Figure 2. Single SNP effect of DCK rs4694362 on OS

··························································································· 69

Figure 3. Combined effects of SNPs on RFS and OS ···· 70

Figure 4. Difference in allele frequency of SNPs between

Korean and other populations ······························· 71

Figure 5. Combined effect of SLC29A1 and TYMS on ara‑C

metabolism in blast cells ········································· 72

-1-

1. Introduction

1.1 Acute myeloid leukemia (AML) and ara‑C

Acute myeloid leukemia (AML) is a rapidly proliferating clonal

disorder of hematopoietic stem cells that lose the ability to

differentiate normally.[1] AML is the most common myeloid

leukemia, with a prevalence of 3.0-4.3 cases per 100,000 rising

to 20.1-23.3 cases per 100,000 adults aged 65 years and

older.[2] Since AML is a clinical and biological heterogenous

disease, previous studies have focused on defining risk

stratification, with patients divided into three cytogenetically

defined risk groups with significant differences in overall survival

(OS).[3] However, there is considerable heterogeneity in the

outcomes for individual patients in each risk group. In addition,

up to 50% of AML patients show no abnormalities by

conventional cytogenetics.[4] These karyotypically normal AMLs

(NK‑AMLs) are prognostically heterogeneous and show various

molecular alterations, although NK‑AML is currently categorized

in the intermediate‑risk group.[5] There is also an inter‑ethnic

difference in treatment outcomes among AML patients.[6-8] In a

study by the Children’s Oncology Group of pediatric patients

with AML treated on 2 consecutive multi-institutional trials,

African patients had significantly worse survival compared with

Caucasian.[6] A similar study in patients treated at St. Jude

-2-

Children’s Research Hospital on 5 consecutive trials showed a

trend of worse outcome in African patients treated on the most

recent trial.[7]

Cytarabine arabinoside, 1‑b‑D‑arabinofuranosylcytosine

(ara‑C) is the key agent for treating AML. The combination

regimen of regular‑dose ara‑C given for 7 days with an

anthracycline for 3 days has been a standard induction therapy

for AML, achieving complete remission (CR) rates of 60-80% in

young adult AML patients.[9] However, Only 20% to 30% of

patients enjoy long‑term disease‑free survival (DFS) and the

majority of patients die primarily because of persistent or

relapsed AML.[10] Resistance to chemotherapy, including ara‑C,

is a major reason for treatment failure among patients with

AML.[11-14] Like other nucleoside analogues, ara‑C is a

prodrug that requires extensive intracellular phophorylation for

activation to its active metabolite ara‑C triphosphate

(ara‑CTP).[15] As the mechanism of action, ara‑C is transported

into leukemic cells by membrane transporters including the

solute carrier family 29 (nucleoside transporters) member 1

(SLC29A1)[Figure 1].[16] And inside the cell, ara‑C is

phosphorylated into ara‑C monophosphate (ara‑CMP) by the

deoxycytidine kinase (DCK) and eventually to ara‑CTP, which

then competes with deoxycytidine triphosphate (dCTP) for

incorporation into DNA and subsequent block of DNA synthesis

causing death of leukemic cell.[17] Other important enzymes are

-3-

cytidine deaminase (CDA) and deoxycytidylate deaminases

(DCTD) that regulate ara‑C degradation.[18]

1.2Genetic/genomic alterations in drug response

Genetic or genomic alterations may affect the expression and/or

function of specific drug protein targets and explain, at least in

part, the inter‑individual variations in the response to specific

treatments. Several previous studies have been examined that

the genetic variations, such as single nucleotide polymorphisms

(SNPs), in the genes encoding the drug transporters and drug

metabolizing pathways relevant for ara‑C activity may play an

important role in the clinical outcomes in AML

patients.[17,19-21] However, it remains to be elucidated

whether SNPs influence the prognosis of leukemia after

chemotherapy. Individual genetic variants may show no

association with the clinical outcomes of interest, but analysis of

the combined effects of SNPs via gene‑gene interactions may

provide evidence of disease susceptibility or drug

response.[22,23] These complex gene‑gene interactions have

been reported as the norm rather than the exception as a risk of

common multifactorial human diseases or drug response.[24]

Therefore, given the interconnected nature of the drug response,

delineating the combined effects of multiple genes acting

collectively in its response is an important aspect in explaining

why clinical outcomes vary so much between patients.[23,25]

-4-

Copy number variation (CNV) is a common type of

genomic structure variation. CNV is loosely defined as a deletion,

duplication or inversion of a DNA sequence of more than one

kilobase. CNVs have recently attracted considerable interest as a

source of genomic variation because they may play an important

role in the etiology of complex diseases and in evolution.[26]

CNVs are also widespread and highly polymorphic within and

between populations and influence gene expression, phenotypic

variation, and adaptation by altering gene copy number, which

can in turn cause disease or contribute to risks of various

complex trait diseases.[27,28] Although these cytogenetic

aberrations are common in healthy individuals, they occur more

frequently in cancers, including AML.[29-32] Recently, CNVs

were reported to be associated with chemotherapy response,

which could affect disease prognosis.[33,34] As such, genomic

variations like CNVs, depending on their size and location, are as

important as SNPs for producing variations in treatment efficacy

and/or adverse responses to chemotherapy. However, despite the

many cytogenetic aberrations that may be relevant for AML

pathogenesis identified in adult AML patients[35], to date no

studies have evaluated copy number changes that are correlated

with response to ara‑C based chemotherapy in NK‑AML patients.

1.3 Aims and scopes

In this study, we determined whether single SNP or SNP‑SNP

-5-

interaction in ara‑C transport and metabolic pathway contribute

to differences in ara‑C based chemotherapy responses in AML

and assessed whether there are population specificity in allele

frequencies of these SNPs through inter‑ethnic comparisons. For

NK‑AML patients, we determined the frequency of genome‑wide

cytogenetic CNV aberrations, and to test whether these genomic

variations contribute to variations in ara‑C based chemotherapy

responses.

-6-

2. Materials and Methods

2.1 Study population and their treatments

Ninety‑seven patients diagnosed with AML other than M3 and

241 Korean normal controls were included in genetic or genomic

analysis. Subjects who were diagnosed with any other cancer or

hematological malignancies or previously administered cytotoxic

drugs or radiation were excluded. Bone marrow or peripheral

blood samples were provided from AML patients at diagnosis and

normal controls. All AML patients received an induction regimen

consisting of ara‑C and idarubicin. A standard dose of ara‑C 100

mg/m2 for 7 days and idarubicin 12 mg/m2 for 3 days were

administered to 79 patients, while 18 patients were treated with

the modified dose regimen based on their general condition at a

physician’s discretion. Once patients achieved CR, the patients

received sequential consolidation therapy consisting of ara‑C and

anthracyclines or hematopoietic stem cell transplantation (HSCT).

All subjects enrolled in this study provided informed consent for

genetic analysis. This study was approved by the Institutional

Review Board of Seoul National University Hospital.

2.2 SNP genotyping

Based on literature search on PubMed

(http://www.ncbi.nlm.nih.gov/pubmed/) and PharmGKB

-7-

(http://www.pharmgkb.org/index.jsp), we selected the 10 genes

which could affect the treatment outcomes of ara‑C based

chemotherapy as follows: SLC29A1, DCK, NT5C3, CDA, DCTD,

CYP1A1, GSTM1, NQO1, MTHFR, and TYMS. The 139 candidate

SNPs in these genes were selected using the database from

NCBI (http://www.ncbi.nlm.nih.gov/) and International HapMap

project (http://hapmap.ncbi.nlm.nih.gov/)[Table1]. SNPs selection

were focused on nonsynonymous coding SNPs, that have been

reported to potentially influence protein structure, activity,

stability or localization and SNPs in the promoter region that are

known to influence the gene expression levels. Known, validated

SNPs from the literature were also included.

SNP genotyping was performed at a multiplex level using

the Illumina GoldenGate Genotyping Assay (Illumina Inc., San

Diego, CA, USA). The genotype quality score for retaining data

was set to 0.25. The deviation from the Hardy‑Weinberg

Equilibrium for the SNPs with significant patient numbers was

tested using the chi‑square test. For pairwise linkage

disequilibrium (LD) between the genetic markers, three

estimators, D, D’ and r were computed. Fifty‑five tagging SNPs

were finally selected with thresholds of r2 > 0.8 for the analysis.

These analyses were carried out using the Haploview 4.2

(Cambridge, MA, USA).

-8-

2.3 Copy number analysis and gene identification

For 30 NK‑AML patients, copy number analysis was performed

using HelixTreeⓇ software version 5.2.0 (Golden Helix Inc.,

Bozeman, MT, USA). To identify individual CNVs, we

incorporated multiple factors including log R ratio, B allele

frequency, marker distance, and population frequency of the B

allele.[36,37] The signal intensity (LRR) and allelic intensity

(BAF) ratios of all samples were exported using the Illumina

BeadStudio software. CNVs were defined using the genomic DNA

of a single reference individual, a European‑American male

(NA10851) from the HapMap study and a pooled data set from

50 randomly selected healthy Korean females. The samples were

removed from further analysis when the call rate was less than

99.0%, the number of identified CNVs exceeded 100, and LRR

standard deviation was above 0.24.[38] All samples had a call

rate greater than 99.2%.

Genes involved in copy number‑altered regions and their

relationship with cancer or drug response were identified utilizing

data from the University of Canada database

(http://projects.tcag.ca/variation/project.html) and the PharmGKB.

2.4 SNP database

To determine the ethnic differences in allele frequencies of the

139 SNPs between Korean and other populations, the

International HapMap and the 1000 genomes database

-9-

(http://www.1000genomes.org/) was used. For the HapMap

(phase III) data, four populations were selected: U.S. residents

of northern and western European ancestry (CEU), Beijing in

China (CHB), Tokyo in Japan (JPT) and Ibadan in Nigeria (YRI).

For the SNPs unshared in the International HapMap, the 1000

genomes database was used to compare the allele frequency.

2.5 Evaluation of clinical response and toxicity

The clinical and pathological information of the AML patients was

obtained by chart review from the clinical database at the study

institution. CR was defined as follows: blast cell counts in the

bone marrow < 5%; absence of extramedullary disease; absolute

neutrophil count > 1.0 X 109/L (1,000/μL); platelet count > 100

X 109/L (100,000/μL).[9] Relapse was defined as the presence

of more than 5% of blast in the bone marrow or reappearance of

blasts in the blood or development of extramedullary disease.

Relapse‑free survival (RFS) was measured from the date of

achievement of a remission until the date of relapse or death

form any cause. OS was defined for all patients of this study

and was measured from the date of entry into the study to the

date of death from any cause. Patients lost to follow‑up or

underwent HSCT after their CR were censored at their date of

last known contact or the date of HSCT, respectively.

Hematologic toxicity during the induction chemotherapy was

graded according to the National Cancer Institute Common

-10-

Terminology Criteria for Adverse Events (NCI CTCAE) version

4.0.[39]

2.6 Statistical analysis

Demographic data and patient characteristics were analyzed and

are reported as frequencies and percentages. Continuous

variables are reported as medians with ranges. Possible

associations between genetic/genomic aberrations and clinical

parameters were analyzed with the chi‑square test, Fisher’s

exact test, or Mann‑Whitney test. Survival probabilities were

estimated by the Kaplan‑Meier method, and differences in the

distributions between the genetic/genomic aberrations were

evaluated using the log‑rank test. For multiple regression

analysis, a Cox proportional hazard model was constructed for

RFS and OS, adjusting for potential confounding covariates. A

stepwise selection method was carried out to find potential

confounding covariates which explain responses very well. For

SNP analysis, both effect of single SNP and SNP‑SNP interaction

were tested. And we tested three different genetic models,

including dominant, recessive, and additive model. The

best‑fitting model was the one with the lowest p‑value among the

three models. Statistical analyses were carried out with the IBM

SPSS software (ver. 19.0; IBM SPSS, Inc., Chicago, IL, USA)

and the free statistical computing environment R (ver. 2.3.1). All

statistical tests were two‑sided, and an a priori level of

-11-

significance of 0.05 was set for all analyses. Multiple test

adjustment was additionally performed using the Bonferroni

correction.

For inter‑ethnic comparison of allele frequency, FST and

chi‑square test statistic are calculated between Korean and other

populations on biallelic genotype data, especially for the minor

allele frequency (MAF).[40] For each SNP, when minor alleles

are not accordant between two populations, an allele of one

population is fixed as reference, and the frequency population is

computed accordingly. We first compared the MAF of the 139

SNPs between Korean AML patients and normal controls, and

there were no significant differences between the two groups.

Thus, we compared the MAF of all Koreans (n = 338) with

those of other ethnic groups. Based on Wright’s qualitative

guidelines, values of FST less than 0.05 at an individual locus

represents low genetic divergence, values of between 0.05-0.15

are considered to represent moderate divergence, FST of 0.15 to

0.25 indicates large divergence, and FST greater than 0.25

represents very large divergence.[41] And a P value of 0.005

was considered as the significant level of chi‑square test to

judge population differentiation.

-12-

3. Results

3.1 Patients’ characteristics and treatment

outcomes

Baseline characteristics and treatment results of the 97

previously untreated de novo AML patients were analyzed in this

study as summarized in Table 2. The median age of patients

was 50.0 years (range, 16.0-76.0 years) and the male/female

proportion was 61/36. The most frequent

French‑American‑British (FAB) subtype was M2 with 48 patients

(49.5%) followed by M4 with seven patients (27.8%). A total of

48 patients (51.6%) were NK‑AMLs. Among the patients who

were available for their cytogenetic or molecular information, 21

patients had t(8;21)(q22;q22) or AML1/ETO, five patients had

inv(16)(p13q22) or CBFB/MYH11, and seven patients had MLL

rearrangement. Of seventy nine patients whom the NPM1 and

FLT3 internal tandem duplication (ITD) mutation information was

identified, 71 patients (91.8%) were at high risk status. Overall,

78 patients (88.6%) achieved overall remission after ara‑C based

induction chemotherapy. Sixty nine patients (78.4%) achieved the

CR after their first course of ara‑C based induction therapy, and

other nine patients achieved the CR after reinduction therapy.

Among the 78 patients, 46 patients (59.0%) were relapsed

during the follow‑up period. Median and mean follow‑up period

-13-

for the 97 patients was 10.3 and 22.0 months, respectively

(range, 0.6-108.5 months). Among all the patients, 34 patients

(25.1%) received allogenic HSCT during follow‑up period. And

58 patients (59.8%) of the 97 patients had died of their disease

progression or disease related complication by the end of

follow‑up period.

Among them, 30 previously untreated de novo NK‑AML

patients were enrolled in CNV analysis. Overall, the median

patient age was 55.5 years (range, 19-76 years) and the

male/female ratio was 15/15. The most frequent FAB subtype

was M2 (13 patients, 43.3%) and the next was M4 (7 patients,

23.3%). The median and mean follow‑up period for the 30

patients was 14.2 and 32.2 months, respectively (0.8-107.7

months). Nineteen patients (63.3%) achieved CR after their first

course of ara‑C based induction therapy, and four patients

achieved CR after reinduction therapy. Among the 23 patients

who achieved overall CR, 13 patients relapsed. Seventeen

(56.7%) of the 30 patients died due to disease progression or

disease‑related complications during the follow‑up period.

Treatment outcomes, molecular markers, and clinical parameters

such as gender, age, baseline complete blood count (CBC), and

bone marrow blast percentage at diagnosis were compared

between the patients with and without CNVs, and there were no

significant differences between the two groups [Table 3].

-14-

3.2 Effect of　 single SNP or SNP‑SNP interaction

on ara‑C based treatment outcomes

In multivariate analysis, SNP rs4694362 in DCK gene,

individually, was a significant prognostic factor for OS [Figure

2]. The presence of CC genotype was significantly associated

with less survival time compared to CT or TT genotypes (HR,

33.202 [95% CI, 4.937-223.273], P < 0.0001, PBonferroni =

0.017). However, none of the 55 SNPs, individually, had any

associations with the first CR, overall remission, or RFS after

adjusting for multiple testing.

In addition to the single SNP effect on treatment

outcomes, combined effects of SNPs were additionally analyzed

in this study [Table 4, Figure 3]. SLC29A1 rs3734703 and

TYMS rs2612100 were the multi-locus genotype combination

that best explained the RFS of AML patients receiving ara‑C

based chemotherapy. Multivariate analysis of RFS revealed that

the presence of the SLC29A1 rs3734703 (AA or AC genotype)

in combination with TYMS rs2612100 (AA genotype) was

significantly associated with shorter RFS compared to the

combination with wild type (HR, 17.630 [95% CI,

4.829-64.369], P < 0.0001, PBonferroni = 0.021). The effect of

these SNP‑SNP interactions also decreased the survival time,

although not statistically significant after the multiple test

adjustment (HR, 23.523 [95% CI, 4.616-119.873], P = 0.0001).

In addition, CDA rs10916827 (GG genotype) in combination with

-15-

DCTD rs17331744 (TC or CC genotype) tended to be associated

with less survival time (HR, 31.680, [95% CI, 6.152-162.905],

P < 0.0001, PBonferroni = 0.052). However, regarding CR or

overall remission, there were no significant combined effects

from the SNPs after the multiple test adjustment.

None of the 55 SNPs individually or in combination were

significantly associated with experience of febrile neutropenia

greater than grade 3, severe neutropenia of grade 4 (neutrophil

count < 500/mm3), or duration of severe neutropenia after ara‑C

based induction chemotherapy. None of the clinical factors was

predictive of hematologic toxicity related to ara‑C based

induction chemotherapy.

3.3 CNVs and gene identification

CNVs were identified in 23 patients (76.7%) in this study. In

total, 384 CNVs with a median of 3 CNVs per patient were

observed and affected every autosomal chromosome at least

once. Sequence losses were more common than gains, with 278

losses (size range, 45.814-43496.575 kbp) and 106 gains (size

range, 51.287-14853.0 kbp) detected. The loss and gain CNV

frequencies per sample were 9.3 and 3.5, respectively, with two

high copy gain regions (log2 ratio > 0.5). Among them, fourteen

copy number‑altered regions contained both gains and losses:

1p36.33-32, 7q32.1, 8q24.3, 9q34.2-3, 10q26.3, 11p15.5-4,

16p13.3, 16q24.2-3, 17p13.1, 17q25.3, 19p13.3, 19p13.11,

-16-

22q11.21-22, and 22q13.1.

Among the observed alterations, 71 recurrently altered

regions were found with 56 having losses (size range,

51.8-3258.3 kbp) and 15 with gains (size range, 25.2-2292.4

kbp). The genes involved in each recurrently imbalanced region

found in more than three patients are reported in Table 5, with

genes known to be related to AML, other hematologic/solid

cancers, or drugs indicated. Nineteen previously reported

AML‑related genes, including HES5, PRDM16, TNFRSF25, MTX2,

TERT, ABCB8, PTP4A3, PBX3, VENTX, AKT1, KIAA0284,

ABCA3, CBFA2T3, FANCA, MLLT6, CD7, PRTN3, CEBPA, and

TYMP were observed in copy number loss regions of this study.

Among drug‑related genes that were located in copy number loss

regions, NOS3, ERCC1, ERCC2, ATP5I, ATP5D, CYBA, NDUFS7,

SLC19A1, and P2RX1 are known to be related to ara‑C or

anthracycline response.

3.4 Correlation of CNVs with ara‑C based

treatment outcomes

The characteristics of CNV, molecular markers, and clinical

parameters such as gender, age, baseline CBC, and bone marrow

blast percentage at diagnosis were compared between the

patients with and without CR after the first induction

chemotherapy, and were also correlated with RFS and OS. There

were no significant different loci of CNV loss or gain between

-17-

the CR and non‑CR group. Except for the bone marrow blast

percentage at diagnosis (P = 0.048) and molecular prognostic

parameters such as FLT3 ITD, and NPM1 mutations, no other

clinical characteristics showed significant association with

attaining CR after the first chemotherapy [Table 6]. The number

of patients with copy number loss significantly differed between

the CR and non‑CR group (P = 0.017). Multivariate analyses

controlled for other clinical covariates showed that the presence

of copy number loss was the only independent factor that

decreased the possibility of CR (OR, 0.015 [95% CI, 0-0.737],

P = 0.035). In addition, the number of copy number‑altered

regions was observed to be a statistically significant prognostic

factor for RFS in patients having copy number gain with a gain

of more than four regions tending to be associated with shorter

RFS (P = 0.083)[Table 7]. In multivariate analysis, the high

number of gain of regions persisted as an independent predictive

factor for shorter RFS (HR, 22.104 [95% CI, 1.644-297.157], P

= 0.020). However, clinical characteristics, molecular prognostic

factors, and genomic instability showed no significant association

with OS in this study.

3.5 Comparison of the allele frequencies of the

SNPs between populations

Overall, allele frequencies of the 139 SNPs in Chinese and

Japanese populations were extremely similar to Korean, with

-18-

most of the SNPs clustering at low FST values. However, allele

frequencies in the Caucasian or African populations showed large

divergence from those of the Korean [Table 8, Figure 4].

Among all the SNPs, DCK rs4694362 which was a significant

prognostic factor for OS in this study ranked the highest in FST

for the population comparisons with African (FST = 0.519). Other

genes that have FST values greater than 0.25 were MTHFR

rs4846052 (comparison with African, FST = 0.492), DCK

rs12648166 (comparison with African, FST = 0.328), and DCTD

rs9542 (comparison with African, FST = 0.252). Among the

SNPs which showed a significant relationship with the response

to ara‑C based chemotherapy in this study, TYMS rs2612100

represented large divergence for the population comparisons with

Caucasian (FST = 0.195). SLC29A1 rs3734703 also showed

moderate difference of allele frequency compared to Caucasian

(FST = 0.136) or African (FST = 0.136). However, there were

very low divergences of allele frequency between Korean and

other populations for CDA rs10916827 and DCTD rs17331744.

-19-

4. Discussion and conclusion

Differing chemotherapy responses result from many factors,

including gender, race, environmental influences and DNA

sequence variations such as SNP and CNV.[34] Therefore,

understanding the contribution of pharmacogenetics/genomic

alteration to the differences in response to ara‑C based

chemotherapy could help individualize chemotherapy and

potentially improve outcomes in AML patients. While the

contribution of SNPs or CNVs to AML chemotherapy has been

described for Caucasian populations, their effects are less

well‑characterized for Asian populations. Here, we investigated

the association of genetics/genomic alteration and treatment

outcomes in Korean AML patients treated with ara‑C based

chemotherapy.

For SNP analysis, SNP rs4694362 in DCK gene had a

significant association with OS. The presence of CC genotype

was significantly associated with less survival time compared to

CT or TT genotypes (HR, 33.202 [95% CI, 4.937-223.273], P

< 0.0001, PBonferroni = 0.017). DCK located on chromosome

4q13.3-12.1 is a rate‑limiting enzyme which is involved in the

activation of ara‑C to ara‑CTP.[15] It seemed to play a distinct

role in development of resistance to ara‑C, since activity of DCK

could determine the intracellular ara‑CTP concentration and

-20-

therefore changed cellular sensitivity.[42] Lamba et al identified

a few SNPs in DCK that associated with its protein activity or

kinetics in the lymphoblastoid cell lines as well as a SNP, in

3′UTR region, which are significantly associated with DCK

mRNA expression and blast ara‑CTP concentrations in patients

administered ara‑C. Although meaningful SNP in our study

located in intron region of DCK, it was possible that the clinical

effect was probably due to other nonsynonymous polymorphisms

within the same LD block. Also, it cannot be ruled out that

intronic SNPs of DCK may directly regulate transcription by

alteration of RNA elongation, splicing, or maturation.[43,44]

Since the drug response is the result of an interaction of

numerous genetic combinations, our results also suggested that a

strong combined effect of SNPs via gene‑gene interaction may

help to predict the outcomes in ara‑C based chemotherapy. We

specifically found that the SLC29A1 rs3734703 AA or AC

genotypes in combination with TYMS rs2612100 AA genotype is

significantly associated with shorter RFS (HR, 17.630 [95% CI,

4.829-64.369], P < 0.0001, PBonferroni = 0.021). This combination

was also associated with less survival time. As shown in Figure

5, thymidine triphosphate (TTP) generated by TYMS is known

to enhance ara‑C toxicity by decreasing dCTP pools.[45] A

decrease in dCTP pools should lead to relative increase in the

amount of ara‑C incorporated into DNA since reduction in dCTP

levels increases the DCK activity, subsequently enhances

-21-

ara‑CTP formation. TTP also inhibits DCTD enzyme which

catalyses the conversion of ara‑CMP into inactive ara‑UMP.

Experimental studies have confirmed that synergy between ara‑C

and thymidine occurs in some tumor cell lines[46,47] and

experimental chemotherapy settings.[48,49] Another SNP‑SNP

interaction identified in our analysis includes CDA rs10916827

(GG genotype) in combination with DCTD rs17331744 (TC or

CC genotype), although this interaction showed a marginal

significance after the multiple test adjustment. Both CDA and

DCTD are the key enzymes that regulate ara‑C degradation.[17]

In pharmacogenomics studies, effects of combined SNPs via

gene‑gene interaction play an important role in characterizing a

trait that involves complex pharmacokinetic and pharmacodynamic

mechanisms, particularly when each involved feature only

demonstrates a minor effect.[50] The data in this study showed

that the combination of certain genotype in ara‑C transport and

metabolic pathway may be one kind of efficient way in predicting

treatment outcomes in ara‑C based chemotherapy.

For copy number analysis in 30 NK‑AML patients, CNVs

were identified in 23 patients (76.7%), which is a rate slightly

higher than that previously reported for Caucasian NK‑AML

patients who had CNV frequencies ranging from

23-60%.[51-53] This variation may arise due to ethnic

differences. Indeed, one study that examined CNVs among the

Korean population reported CNV differences according to the

-22-

ethnic reference set used, with a set of 90 Korean subjects

exhibiting 123 CNV regions.[54] In contrast, more CNV regions

(n = 643) were detected when compared to a reference set that

included multiple ethnic groups, which reflects the ethnic

diversity of structural variations between Korean and other

populations. Recent reports also suggest that different ethnic

groups may represent different CNV profiles that are stratified in

the human population.[55,56]

The characteristics of patients’clinical and chromosomal

factors were compared between patients with and without CR

after the first induction chemotherapy, and showed that loss of

copy number and high bone marrow blast percentage at diagnosis

are predictive markers for lower probability for achieving CR. As

shown in Table 5, many cancer and drug‑related genes were

found in recurrent copy number loss regions. Overall, a few

genes from the recurrent loss regions have been reported to be

related to the response to ara‑C or anthracycline drugs. Among

these, NOS3 (Nitric Oxide Synthase 3) may play a role in

anthracycline treatment outcomes because of its activity in

oxidative stress and quinone detoxification.[57] Several studies

demonstrated that nitric oxide (NO) produced by NOS3 interacts

with anthracycline‑induced radicals to form peroxynitrite, which

can damage lipids, DNA, and proteins via direct oxidative

reactions or indirect, radical‑mediated mechanisms.[58,59]

Enhanced tumor vasculature with higher NO levels resulting from

-23-

NOS3 activity may result in better drug delivery to cancer cells,

which is supported by a recent study showing that lower levels

of NOS3 were associated with increased risk of recurrence and

poorer survival in breast cancer.[60] In addition, DNA repair

genes such as ERCC1 and ERCC2, the ATP synthase genes

ATP5I and ATP5D, as well as the CYBA and NDUFS7 genes are

also known to be related with response to anthracycline

chemotherapy and anthracycline‑induced cardiotoxicity.[61-65]

With regard to ara‑C, in our study we found that the P2RX1

gene lies in a recurrent loss region. P2RX1, found on the p arm

of chromosome 17, encodes P2X purinoceptor 1, which is a

ligand‑gated ion channel expressed in smooth muscle and

platelets. A recent study using a whole genome approach

identified one SNP in the P2RX1 gene that is involved in the

genetic signature for susceptibility to ara‑C in the central

European population.[66] However, the precise contribution and

mechanism of genetic variations in these genes to ara‑C and

anthracycline susceptibility remains unclear. Therefore, further

in‑depth studies are needed to define the relationship between

CNVs of these candidate genes and leukemia progression or

various responses to chemotherapy in NK‑AML patients.

Another interesting finding in the present study was the

correlation of genomic instability with poor RFS rate for

NK‑AML patients. The presence of more than four copy number

gain regions was an important poor prognostic factor and was

-24-

independent of other clinical and genetic characteristics. In

contrast, we found no correlation between genetic instability and

OS. Among 30 patients, eight received allogenic HSCT, and OS

might be affected by HSCT‑related factors such as donor

selection, cytotoxic conditioning regimen, patient performance

status, and complications. In this study, FLT3 ITD and NPM1

mutations were not correlated with the response to

chemotherapy, which is in contrast to results of some previous

reports. It may be due to the small number of patients assessed

in this study.

In addition, we compared the MAF of the 139 SNPs in

ara‑C transport and metabolic pathway between Korean and other

populations. The last few years, there has been great concern

that the ethnic difference in the frequency of SNPs involved in

the pharmacology are one potential explanation for the

differences in treatment outcomes.[67] The present study

showed that the MAF of the 139 SNPs in the Korean population

differed greatly from those in Caucasians and Africans but were

similar among Asian populations. Especially, DCK rs4694362,

SLC29A1 rs3734703, and TYMS rs2612100, which were

significantly associated with a response to ara‑C based

chemotherapy in this study, showed a meaningful divergence

between Korean and African or Caucasian populations.

Our study had some limitations due to the relatively small

sample size. In an attempt to examine ara‑C toxicities, almost

-25-

patients (n = 88) experienced febrile neutropenia during the

induction chemotherapy and no differences were observed by

single or combined genotype analysis. Also, despite ara‑C being

the most important drug in AML therapy, patients receive a

multiagent therapy including anthracyclines in combination with

ara‑C. Thus, anthracyclines could have had some influence on

treatment response independent of the examined ara‑C

metabolism related SNPs.

In conclusion, our study strengthens the importance of

analyzing genetic/genomic alterations to better understand AML

progression and its response to ara‑C based chemotherapy in

AML patients. Although AML is a very heterogeneous disease

with different subtypes that are of prognostic significance, the

results of our study could help in better understanding of the

effect of single SNP, SNP‑SNP interaction or CNVs on drug

responsiveness and guide us to develop individualized

chemotherapy in AML patients receiving ara‑C based

chemotherapy. Our data also identified several candidate genes in

the context of ara‑C and anthracycline drug response in NK‑AML

patients. Further studies involving larger groups of AML patients

and functional studies to explore the underlying mechanisms of

the candidate genes and SNPs in the chemotherapeutic response

are needed to confirm our results.

-26-

5. References

1) Estey E, Dohner H.Acute myeloid leukaemia. Lancet

2006;368: 1894-907.

2) National Cancer Institute. Acute myeloid leukemia.

http://seer.cancer.gov/statfacts/html/amyl.html (accessed

April 20, 2012).

3) Slovak ML, Kopecky KJ, Cassileth PA, Harrington DH, Theil

KS, Mohamed A, Paietta E, Willman CL, Head DR, Rowe

JM, Forman SJ, Appelbaum FR. Karyotypic analysis predicts

outcome of preremission and postremission therapy in adult

acute myeloid leukemia: a Southwest Oncology

Group/Eastern Cooperative Oncology Group Study. Blood

2000;96:4075-83.

4) Mitelman F. Catalogue of Chromosome Aberrations in

Cancer. Cytogenet Cell Genet. 1983;36:1-515.

5) Bienz M, Ludwig M, Leibundgut EO, Mueller BU, Ratschiller

D, Solenthaler M, Fey MF, Pabst T. Risk assessment in

patients with acute myeloid leukemia and a normal

karyotype. Clin Cancer Res 2005;11:1416-24.

6) Aplenc R, Alonzo TA, Gerbing RB, Smith FO, Meshinchi S,

Ross JA, Perentesis J, Woods WG, Lange BJ, Davies SM.

Ethnicity and survival in childhood acute myeloid leukemia:

a report from the Children's Oncology Group. Blood

2006;108:74-80.

-27-

7) Rubnitz JE, Lensing S, Razzouk BI, Pounds S, Pui CH,

Ribeiro RC. Effect of race on outcome of white and black

children with acute myeloid leukemia: the St. Jude

experience. Pediatr Blood Cancer 2007;48:10-15.

8) Sekeres MA, Peterson B, Dodge RK, Mayer RJ, Moore JO,

Lee EJ, Kolitz J, Baer MR, Schiffer CA, Carroll AJ,

Vardiman JW, Davey FR, Bloomfield CD, Larson RA, Stone

RM. Differences in prognostic factors and outcomes in

African Americans and whites with acute myeloid leukemia.

Blood 2004;103:4036-4042.

10) Dohner H, Estey EH, Amadori S, Appelbaum FR, Buchner

T, Burnett AK, Dombret H, Fenaux P, Grimwade D, Larson

RA, Lo-Coco F, Naoe T, Niederwieser D, Ossenkoppele GJ,

Sanz MA, Sierra J, Tallman MS, Lowenberg B, Bloomfield

CD. Diagnosis and management of acute myeloid leukemia in

adults: recommendations from an international expert panel,

on behalf of the European LeukemiaNet. Blood

2010;115:453-74.

11) Tallman MS, Gilliland DG, Rowe JM. Drug therapy for acute

myeloid leukemia. Blood 2005;106:1154-63.

12) Galmarini CM, Thomas X, Calvo F, Rousselot P, El Jafaari

A, Cros E, Dumontet C. Potential mechanisms of resistance

to cytarabine in AML patients. Leuk Res 2002;26:621-9.

13) Galmarini CM, Thomas X, Calvo F, Rousselot P, Rabilloud

M, El Jaffari A, Cros E, Dumontet C. In vivo mechanisms of

-28-

resistance to cytarabine in acute myeloid leukaemia. Br J

Haematol 2002;117:860-8.

14) Cros E, Jordheim L, Dumontet C, Galmarini CM. Problems

related to resistance to cytarabine in acute myeloid

leukemia. Leuk Lymphoma 2004;45:1123-32.

15) Styczynski J. Drug resistance in childhood acute myeloid

leukemia. Curr Pharm Biotechnol 2007;8:59-75.

16) Lamba JK, Crews K, Pounds S, Schuetz EG, Gresham J,

Gandhi V, Plunkett W, Rubnitz J, Ribeiro R.

Pharmacogenetics of deoxy cytidine kinase: identification

and characterization of novel genetic variants. J Pharmacol

Exp Ther 2007;323:935-45.

17) Yamauchi T, Negoro E, Kishi S, Takagi K, Yoshida A,

Urasaki Y, Iwasaki H, Ueda T. Intracellular cytarabine

triphosphate production correlates to deoxycytidine kinase

/cytosolic 5'­nucleotidase II expression ratio in primary

acute myeloid leukemia cells. Biochem Pharmacol

2009;77:1780-6.

18) Mahlknecht U, Dransfeld CL, Bulut N, Kramer M, Thiede C,

Ehninger G, Schaich M. SNP analyses in cytarabine

metabolizing enzymes in AML patients and their impact on

treatment response and patient survival: identification of

CDA SNP C­451T as an independent prognostic parameter

for survival. Leukemia 2009;23:1929-32.

19) Stam RW, den Boer ML, Meijerink JP, Ebus ME, Peters GJ,

-29-

Noordhuis P, Janka-Schaub GE, Armstrong SA, Korsmeyer

SJ, Pieters R. Differential mRNA expression of

ara­C­metabolizing enzymes explains ara­C sensitivity in

MLL gene­rearranged infant acute lymphoblastic leukemia.

Blood 2003;101:1270-6.

20) Galmarini CM, Thomas X, Graham K, El Jafaari A, Cros E,

Jordheim L, Mackey JR, Dumontet C. Deoxycytidine kinase

and cN­II nucleotidase expression in blast cells predict

survival in acute myeloid leukaemia patients treated with

cytarabine. Br J Haematol 2003;122:53-60.

21) Yue L, Saikawa Y, Ota K, Tanaka M, Nishimura R, Uehara

T, Maeba H, Ito T, Sasaki T, Koizumi S. A functional

single­nucleotide polymorphism in the human cytidine

deaminase gene contributing to ara­C sensitivity.

Pharmacogenetics 2003;13:29-38.

22) Bhatla D, Gerbing RB, Alonzo TA, Conner H, Ross JA,

Meshinchi S, Zhai X, Zamzow T, Mehta PA, Geiger H,

Perentesis J, Davies SM. Cytidine deaminase genotype and

toxicity of cytosine arabinoside therapy in children with

acute myeloid leukemia. Br J Haematol 2009;144:388-394.

23) Han W, Kim KY, Yang SJ, Noh DY, Kang D, Kwack K.

SNP­SNP interactions between DNA repair genes were

associated with breast cancer risk in a Korean population.

Cancer 2012;118:594-602.

24) Strom SS, Estey E, Outschoorn UM, Garcia-Manero G.

-30-

Acute myeloid leukemia outcome: role of nucleotide excision

repair polymorphisms in intermediate risk patients. Leuk

Lymphoma 2010;51:598-605.

25) Moore JH, Williams SM. New strategies for identifying

gene­gene interactions in hypertension. Ann Med

2002;34:88-95.

26) Moore JH. The ubiquitous nature of epistasis in determining

susceptibility to common human diseases. Hum Hered

2003;56:73-82.

27) Zhang D, Qian Y, Akula N, Alliey-Rodriguez N, Tang J,

Gershon ES, Liu C. Accuracy of CNV Detection from GWAS

Data. PLoS One 2011;6:e14511.

28) Feuk L, Carson AR, Scherer SW. Structural variation in the

human genome. Nat Rev Genet 2006;7:85-97.

29) Fredman D, White SJ, Potter S, Eichler EE, Den Dunnen

JT, Brookes AJ. Complex SNP­related sequence variation

in segmental genome duplications. Nat Genet

2004;36:861-6.

30) Burkhardt B, Bruch J, Zimmermann M, Strauch K,

Parwaresch R, Ludwig WD, Harder L, Schlegelberger B,

Mueller F, Harbott J, Reiter A. Loss of heterozygosity on

chromosome 6q14-q24 is associated with poor outcome in

children and adolescents with T­cell lymphoblastic

lymphoma. Leukemia 2006;20:1422-9.

31) Huse K, Taudien S, Groth M, Rosenstiel P, Szafranski K,

-31-

Hiller M, Hampe J, Junker K, Schubert J, Schreiber S,

Birkenmeier G, Krawczak M, Platzer M. Genetic variants of

the copy number polymorphic beta­defensin locus are

associated with sporadic prostate cancer. Tumour Biol

2008;29:83-92.

32) Lin LJ, Asaoka Y, Tada M, Sanada M, Nannya Y, Tanaka Y,

Tateishi K, Ohta M, Seto M, Sasahira N, Kawabe T, Zheng

CQ, Kanai F, Ogawa S, Omata M. Integrated analysis of

copy number alterations and loss of heterozygosity in

human pancreatic cancer using a high­resolution, single

nucleotide polymorphism array. Oncology 2008;75:102-12.

33) Wang ZC, Lin M, Wei LJ, Li C, Miron A, Lodeiro G, Harris

L, Ramaswamy S, Tanenbaum DM, Meyerson M, Iglehart

JD, Richardson A. Loss of heterozygosity and its correlation

with expression profiles in subclasses of invasive breast

cancers. Cancer Res 2004;64:64-71.

34) Joha S, Dauphin V, Lepretre F, Corm S, Nicolini FE,

Roumier C, Nibourel O, Grardel N, Maguer-Satta V,

Idziorek T, Figeac M, Lai JL, Quesnel B, Etienne G, Guilhot

F, Lippert E, Preudhomme C, Roche-Lestienne C. Genomic

characterization of Imatinib resistance in CD34+ cell

populations from chronic myeloid leukaemia patients. Leuk

Res 2011;35:448-58.

35) Kalari KR, Hebbring SJ, Chai HS, Li L, Kocher JP, Wang L,

Weinshilboum RM. Copy number variation and cytidine

-32-

analogue cytotoxicity: a genome­wide association approach.

BMC Genomics 2010;11:357.

36) Walter MJ, Payton JE, Ries RE, Shannon WD, Deshmukh H,

Zhao Y, Baty J, Heath S, Westervelt P, Watson MA,

Tomasson MH, Nagarajan R, O'Gara BP, Bloomfield CD,

Mrozek K, Selzer RR, Richmond TA, Kitzman J, Geoghegan

J, Eis PS, Maupin R, Fulton RS, McLellan M, Wilson RK,

Mardis ER, Link DC, Graubert TA, DiPersio JF, Ley TJ.

Acquired copy number alterations in adult acute myeloid

leukemia genomes. PNAS 2009;106:12950-5.

37) Wang K, Chen Z, Tadesse MG, Glessner J, Grant SF,

Hakonarson H, Bucan M, Li M. Modeling genetic inheritance

of copy number variations. Nucleic Acids Res

2008;36:e138.

38) Wang K, Li M, Hadley D, Liu R, Glessner J, Grant SF,

Hakonarson H, Bucan M. PennCNV: an integrated hidden

Markov model designed for high­resolution copy number

variation detection in whole­genome SNP genotyping data.

Genome Res 2007;17:1665-74.

39) Kathiresan S, Voight BF, Purcell S, Musunuru K, Ardissino

D, Mannucci PM, Anand S, Engert JC, Samani NJ, Schunkert

H, Erdmann J, Reilly MP, Rader DJ, Morgan T, Spertus JA,

Stoll M, Girelli D, McKeown PP, Patterson CC, Siscovick

DS, O'Donnell CJ, Elosua R, Peltonen L, Salomaa V,

Schwartz SM, Melander O, Altshuler D, Merlini PA, Berzuini

-33-

C, Bernardinelli L, Peyvandi F, Tubaro M, Celli P, Ferrario

M, Fetiveau R, Marziliano N, Casari G, Galli M, Ribichini F,

Rossi M, Bernardi F, Zonzin P, Piazza A, Yee J, Friedlander

Y, Marrugat J, Lucas G, Subirana I, Sala J, Ramos R, Meigs

JB, Williams G, Nathan DM, MacRae CA, Havulinna AS,

Berglund G, Hirschhorn JN, Asselta R, Duga S, Spreafico M,

Daly MJ, Nemesh J, Korn JM, McCarroll SA, Surti A,

Guiducci C, Gianniny L, Mirel D, Parkin M, Burtt N, Gabriel

SB, Thompson JR, Braund PS, Wright BJ, Balmforth AJ, Ball

SG, Hall AS, Linsel-Nitschke P, Lieb W, Ziegler A, Konig I,

Hengstenberg C, Fischer M, Stark K, Grosshennig A,

Preuss M, Wichmann HE, Schreiber S, Ouwehand W,

Deloukas P, Scholz M, Cambien F, Li M, Chen Z, Wilensky

R, Matthai W, Qasim A, Hakonarson HH, Devaney J, Burnett

MS, Pichard AD, Kent KM, Satler L, Lindsay JM, Waksman

R, Knouff CW, Waterworth DM, Walker MC, Mooser V,

Epstein SE, Scheffold T, Berger K, Huge A, Martinelli N,

Olivieri O, Corrocher R, McKeown P, Erdmann E, Konig IR,

Holm H, Thorleifsson G, Thorsteinsdottir U, Stefansson K,

Do R, Xie C, Siscovick D. Genome­wide association of

early­onset myocardial infarction with single nucleotide

polymorphisms and copy number variants. Nat Genet

2009;41:334-41.

40) Cancer Therapy Evaluation Program, Common Terminology

Criteria for Adverse Events, 1. Version 4.0 (2009), DCTD,

-34-

NCI, NIH, DHHS. http://ctep.cancer.gov. Accessed 2

February 2012.

41) Han K, Kim K, Oh JM, Kim IW, Kim KI, Park TS.

Unbalanced sample size effect on genome­wide population

differentiation studies. Int J Data Mining and Bioinformatics

(in press)

42) Wright S: Variability Within and Among Natural Populations.

University of Chicago, Chicago, IL, USA, 4 (1978).

43) Kufe DW, Munroe D, Herrick D, Egan E, Spriggs D. Effects

of 1­beta­D­arabinofuranosylcytosine incorporation on

eukaryotic DNA template function. Mol Pharmacol

1984;26:128-34.

44) Baralle M, Pastor T, Bussani E, Pagani F. Influence of

Friedreich ataxia GAA noncoding repeat expansions on

pre­mRNA processing. Am J Hum Genet 2008;83:77-88.

45) Murani E, Ponsuksili S, Seyfert HM, Shi X, Wimmers K.

Dual effect of a single nucleotide polymorphism in the first

intron of the porcine secreted phosphoprotein 1 gene:

allele­specific binding of C/EBP beta and activation of

aberrant splicing. BMC Mol Biol 2009;10:96.

46) Ryan DP, Garcia-Carbonero R, Chabner BA. Cytidine

Analogues. In: Cancer Chemotherapy & biotherapy: Priciple

& Practice, Vol. Fourth edition 183-211 (Lippincott

Williams & Wilkins, 2005).

47) Harris AW, Reynolds EC, Finch LR. Effect of thymidine on

-35-

the sensitivity of cultured mouse tumor cells to

1­beta­D­arabinofuranosylcytosine. Cancer Res

1979;39:538-41.

48) Grant S, Lehman C, Cadman E. Enhancement of

1­beta­D­arabinofuranosylcytosine accumulation within

L1210 cells and increased cytotoxicity following thymidine

exposure. Cancer Res 1980;40:1525-31.

49) Danhauser LL, Rustum YM. Effect of thymidine on the

toxicity, antitumor activity, and metabolism of

1­beta­D­arabinofuranosylcytosine in rats bearing a

chemically induced colon carcinoma. Cancer Res

1980;40:1274-80.

50) Fram R, Major P, Egan E, Beardsley P, Rosenthal D, Kufe

D. A phase I­II study of combination therapy with

thymidine and cytosine arabinoside. Cancer Chemother

Pharmacol 1983;11:43-7.

51) Lane HY, Tsai GE, Lin E. Assessing gene­gene

interactions in pharmacogenomics. Mol Diagn Ther

2012;16:15-27.

52) Bullinger L, Kronke J, Schon C, Radtke I, Urlbauer K,

Botzenhardt U, Gaidzik V, Cario A, Senger C, Schlenk RF,

Downing JR, Holzmann K, Dohner K, Dohner H.

Identification of acquired copy number alterations and

uniparental disomies in cytogenetically normal acute myeloid

leukemia using high­resolution single­nucleotide

-36-

polymorphism analysis. Leukemia 2010;24:438-49.

53) Suela J, Alvarez S, Cifuentes F, Largo C, Ferreira BI, Blesa

D, Ardanaz M, Garcia R, Marquez JA, Odero MD, Calasanz

MJ, Cigudosa JC. DNA profiling analysis of 100 consecutive

de novo acute myeloid leukemia cases reveals patterns of

genomic instability that affect all cytogenetic risk groups.

Leukemia 2007;21:1224-31.

54) Tyybakinoja A, Elonen E, Piippo K, Porkka K, Knuutila S.

Oligonucleotide array­CGH reveals cryptic gene copy

number alterations in karyotypically normal acute myeloid

leukemia. Leukemia 2007;21:571-4.

55) Jeon JP, Shim SM, Jung JS, Nam HY, Lee HJ, Oh BS, Kim

K, Kim HL, Han BG. A comprehensive profile of DNA copy

number variations in a Korean population: identification of

copy number invariant regions among Koreans. Exp Mol

Med 2009;41:618-28.

56) Kidd JM, Newman TL, Tuzun E, Kaul R, Eichler EE.

Population stratification of a common APOBEC gene deletion

polymorphism. PLoS Genet 2007;3:e63.

57) Redon R, Ishikawa S, Fitch KR, Feuk L, Perry GH, Andrews

TD, Fiegler H, Shapero MH, Carson AR, Chen W, Cho EK,

Dallaire S, Freeman JL, Gonzalez JR, Gratacos M, Huang J,

Kalaitzopoulos D, Komura D, MacDonald JR, Marshall CR,

Mei R, Montgomery L, Nishimura K, Okamura K, Shen F,

Somerville MJ, Tchinda J, Valsesia A, Woodwark C, Yang F,

-37-

Zhang J, Zerjal T, Armengol L, Conrad DF, Estivill X,

Tyler-Smith C, Carter NP, Aburatani H, Lee C, Jones KW,

Scherer SW, Hurles ME. Global variation in copy number in

the human genome. Nature 2006;444:444-54.

58) Choi JY, Nowell SA, Blanco JG, Ambrosone CB. The role of

genetic variability in drug metabolism pathways in breast

cancer prognosis. Pharmacogenomics 2006;7:613-24.

59) Ambs S, Hussain SP, Harris CC. Interactive effects of nitric

oxide and the p53 tumor suppressor gene in carcinogenesis

and tumor progression. FASEB J 1997;11:443-8.

60) Felley­Bosco E. Role of nitric oxide in genotoxicity:

implication for carcinogenesis. Cancer Metastasis Rev

1998;17:25-37.

61) Choi JY, Barlow WE, Albain KS, Hong CC, Blanco JG,

Livingston RB, Davis W, Rae JM, Yeh IT, Hutchins LF,

Ravdin PM, Martino S, Lyss AP, Osborne CK, Abeloff MD,

Hayes DF, Ambrosone CB. Nitric oxide synthase variants

and disease­free survival among treated and untreated

breast cancer patients in a Southwest Oncology Group

clinical trial. Clin Cancer Res 2009;15:5258-66.

62) Biason P, Hattinger CM, Innocenti F, Talamini R, Alberghini

M, Scotlandi K, Zanusso C, Serra M, Toffoli G. Nucleotide

excision repair gene variants and association with survival

in osteosarcoma patients treated with neoadjuvant

chemotherapy. Pharmacogenomics J 2011 Aug 9. Doi:

-38-

10.1038/tpj.2011.33

63) Mordente A, Meucci E, Silvestrini A, Martorana GE,

Giardina B. New developments in anthracycline­induced

cardiotoxicity. Curr Med Chem 2009;16:1656-72.

64) Nohl H, Gille L, Staniek K. The exogenous NADH

dehydrogenase of heart mitochondria is the key enzyme

responsible for selective cardiotoxicity of anthracyclines. Z

Naturforsch C 1998;53:279-85.

65) Pawlowska J, Tarasiuk J, Wolf CR, Paine MJ, Borowski E.

Differential ability of cytostatics from anthraquinone group

to generate free radicals in three enzymatic systems:

NADH dehydrogenase, NADPH cytochrome P450 reductase,

and xanthine oxidase. Oncol Res 2003;13:245-52

66) Wojnowski L, Kulle B, Schirmer M, Schluter G, Schmidt A,

Rosenberger A, Vonhof S, Bickeboller H, Toliat MR, Suk

EK, Tzvetkov M, Kruger A, Seifert S, Kloess M, Hahn H,

Loeffler M, Nurnberg P, Pfreundschuh M, Trumper L,

Brockmoller J, Hasenfuss G. NAD(P)H oxidase and

multidrug resistance protein genetic polymorphisms are

associated with doxorubicin­induced cardiotoxicity.

Circulation 2005;112:3754-62.

67) Hartford CM, Duan S, Delaney SM, Mi S, Kistner EO,

Lamba JK, Huang RS, Dolan ME. Population­specific

genetic variants important in susceptibility to cytarabine

arabinoside cytotoxicity. Blood 2009;113:2145-53.

-39-

68) O'Donnell PH, Dolan ME. Cancer pharmacoethnicity: ethnic

differences in susceptibility to the effects of chemotherapy.

Clin Cancer Res 2009;15:4806-14.

-40-

Figure legends

Figure 1. Ara‑C transport and metabolic pathway

Figure 2. Single SNP effect of DCK rs4694362 on OS. (CT vs.

TT, HR = 33.202 [95% CI, 4.937-223.273], P < 0.0001,

PBonferroni = 0.017).

Figure 3. Combined effects of SNPs on RFS and OS. (A)

Combined effect of SLC29A1 rs3734703 and TYMS rs2612100

on RFS (CC x AA vs. AA/AC x AA, HR = 17.630 [95% CI,

4.829-64.369], P < 0.0001, PBonferroni = 0.021). (B) Combined

effect of CDA rs10926817 and DCTD rs17331744 on OS (GG x

TT vs. GG x TC/CC, HR = 31.680 [95% CI, 6.152-162.905], P

< 0.0001, PBonferroni = 0.052).

Figure 4. Difference in allele frequency of SNPs between Korean

and other populations. Abbreviations: CEU, Central European;

CHB, Chinese; JPT, Japanese; KOR, Korean; YRI, African.

Figure 5. Combined effect of SLC29A1 and TYMS on ara‑C

metabolism in blast cells

-41-

Gene No. of SNPs SNPs

SLC29A1 10 rs7753792, rs1057985, rs507964, rs3778504, rs693955, rs747199, rs9394992, rs324148, rs760370, rs3734703

DCK 2 rs12648166, rs4694362

NT5C3 42 rs7792057, rs6462445, rs6462446, rs6942974, rs12155477, rs11532669, rs2392209, rs16879126, rs17170153,

rs3750117, rs3750118, rs3750119, rs6955792, rs17170180, rs7793793, rs4562213, rs6462449, rs6462450,

rs2049758, rs17170218, rs4720097, rs12668520, rs17170228, rs6956397, rs7806813, rs6462453, rs4720098,

rs10085768, rs6948212, rs4316067, rs6946062, rs7801986, rs10486512, rs7776847, rs10231011, rs10951370,

rs4338000, rs4723242, rs10251079, rs10281012, rs10256717, rs6954923

CDA 19 rs1253904, rs532545, rs603412, rs602946, rs2072671, rs471760, rs10916824, rs818199, rs818196, rs577042,

rs4655226, rs10799647, rs818194, rs10916827, rs580032, rs527912, rs1689924, rs477155, rs12404655

DCTD 28 rs11132158, rs11132159, rs6835318, rs7277, rs1130902, rs3811810, rs3190314, rs9542, rs2464974,

rs1960207, rs12499918, rs17331744, rs7663494, rs3886768, rs13148414, rs6552621, rs17331968,

rs13114435, rs13139377, rs10520543, rs10009825, rs13147196, rs13101260, rs9990999, rs7688234,

rs13116494, rs17272827, rs13116598

CYP1A1 1 rs3809585

GSTM1 1 rs412543

[Table 1] The 139 candidate SNPs in this study

-42-

NQO1 6 rs10517, rs1437135, rs1800566, rs2917669, rs4986998, rs689452

MTHFR 15 rs11121832, rs13306556, rs1476413, rs1537516, rs17367504, rs17421511, rs1801131, rs1801133, rs1994798,

rs3737965, rs4846048, rs4846049, rs4846052, rs6541003, rs9651118

TYMS 15 rs1001761, rs1004474, rs1051527, rs2244500, rs2612095, rs2612100, rs2847149, rs2847150, rs2847153,

rs2853532, rs2853533, rs2853741, rs3786362, rs502396, rs699517

-43-

Characteristics Σn n % Median (range)

Gender

Male/Female

97

61/36

62.9/37.1

Age (years) 97 50.0 (16.0-76.0)

FAB classification

M0

M1

M2

M4

M5

M6

M7

97

2

13

48

27

5

1

1

2.1

13.4

49.5

27.8

5.2

1.0

1.0

Bone marrow blast (%) 89 60.9 (6.8-98.7)

WBC (x103/㎕) 94 18.0 (1.1-314 5)

Hb (g/dL) 94 8.2 (4.1-19.4)

PLT (x10³/㎕) 94 38.5 (3.0-278.0)

Karyotype

Normal/Abnormal

97

48/45

51.6/48.4

Inv(16) or CBFB

Positive/Negative

45

6/39

13.3/86.7

t(8;21) or ETO

Positive/Negative

95

21/74

22.1/77.9

MLL rearrangement

Positive/Negative

85

7/78

8.2/91.8

NPM1/FLT3 risk statusa

Low/High

79

8/71

10.1/89.9

[Table 2] AML patients characteristics and treatment outcomes

-44-

Overall remission

Yes/ No

88

78/10

88.6/11.4

Relapse before HSCT

Yes/ No

78

46/32

59.0/41.0

HSCT during the F/U period

Yes/ No

97

34/63

25.1/64.9

Abbreviations: F/U, follow‑up; HSCT, hematopoietic stem cell

transplantation.

a. NPM1/FLT3 status: High risk group, NPM1 wild/FLT3 ITD(-),

NPM1 wild/FLT3 ITD(+), NPM1 mutated/FLT3 ITD(+); Low risk

group, NPM1 mutated/FLT3 ITD(-)

-45-

Characteristics

Patients

with CNVs

(n = 23)

Patients

without CNVs

(n = 7)

P

Gender, No. (%)

Male

Female

14 (60.9)

9 (39.1)

1 (14.3)

6 (85.7)

0.080

Age (years)

Mean (SD)

52.4 (15.9)

54.6 (5.8)

0.595

FAB classification, No. (%)

M0

M1

M2

M4

M5

1 (4.3)

5 (21.7)

8 (34.8)

6 (26.1)

3 (13.0)

0 (0)

0 (0)

5 (71.4)

1 (14.3)

1 (14.3)

0.432

Bone marrow blast (%)

Mean (SD)

63.8 (24.1)

49.4 (26.0)

0.187

WBC (x103/㎕)

Median (range)

18.7 (2.2-189.7)

19.7 (5.1-173.1)

0.774

Hb (g/dL)

Median (range)

7.4 (4.5-19.4)

9.3 (4.6-11.5)

0.631

PLT (x10³/㎕)

Median (range)

37.0 (6.0-181.0)

54.0 (8.0-179.0)

1.000

FLT3 ITD, No. (%)

Positive

Negative

5 (27.8)

13 (72.2)

1 (33.3)

2 (66.7)

1.000

[Table 3] Comparison between patients with and without CNVs

-46-

NPM1 mutation, No. (%)

Positive

Negative

8 (42.1)

11 (57.9)

2 (66.7)

1 (33.3)

0.571

CR after induction CTx., No.(%)

Positive

Negative

13 (56.5)

10 (43.5)

6 (85.7)

1 (14.3)

0.215

Relapse after CR, No. (%)

Positive

Negative

11 (64.7)

6 (35.3)

2 (33.3)

4 (66.7)

0.341

HSCT during F/U period, No. (%)

Positive

Negative

7 (30.4)

16 (69.6)

1 (14.3)

6 (85.7)

0.398

Death during F/U period, No. (%)

Positive

Negative

14 (60.9)

9 (39.1)

3 (42.9)

4 (57.1)

0.666

Abbreviations: CR, complete remission; CTx, chemotherapy; FLT3 ITD,

FMS‑like tyrosine kinase 3 internal tandem duplication; F/U, follow‑up;

HSCT, hematopoietic stem cell transplantation; NPM1, nucleophosmin1.

-47-

Endpoint SNP‑SNP interaction Genotype HRb (95% CI) Pc Pd

RFSa

SLC29A1

rs3734703

TYMS

rs2612100

CC x AA (n = 34)

CC x AG/GG (n = 30)

AA/AC x AA (n = 11)

AA/AC x GG/AG (n = 21)

1

3.969 (1.689-9.329)

17.630 (4.829-64.369)

0.999 (0.351-2.842)

0.002

< 0.0001

0.999

0.021

OSa

CDA

rs10916827

DCTD

rs17331744

GG x TT (n = 25)

GG x TC/CC (n = 15)

GA/AA x TT (n = 40)

GA/AA x TC/CC (n = 17)

1

31.680 (6.152-162.905)

6.374 (1.624-25.021)

6.322 (1.056-37.853)

< 0.0001

0.008

0.043

0.052

[Table 4] Combined effect of SNPs on survival in AML patients

Abbreviations: CI, confidence interval; HR, hazard ratio; OS, overall survival; RFS, relapse free survival; SNP,

single nucleotide polymorphism.

a. Dominant model

b. Adjusted covariates are subtype, normal karyotype, t(8;21), WBC, bone marrow blast count, and each SNP

for RFS and subtype, WBC, PLT, bone marrow blast count, and each SNP for OS.

c. P value for each genotype combination; Cox proportional hazard model

d. Statistically significant after the Bonferroni correction

-48-

Chr. Start, nt End. nt Length, kbp Cytoband Genes involveda

Gains

4 162209025 162287870 78.8 4q32.2 Unknown

8 43680397 43910848 230.5 8p11.1 Unknown

17 43924123 43949296 25.2 17q21.32 Unknown

Losses

1 1045729 3767779 2722.1 1p.36.33-

1p.36.32

MIR200A, MIR429, TNFRSF18, TNFRSF4, TTLL10, SDF4,

FAM132A, UBE2J2, B3GALT6, ACAP3, CPSF3L, SCNN1D,

DVL1, PUSL1, GLTPD1, MXRA8, TAS1R3, AURKAIP1,

CCNL2, MRPL20, TMEM88B, VWA1, ATAD3A, ATAD3B,

ATAD3C, SSU72, CDK11A, CDK11B, MIB2, MMP23B,

SLC35E2, SLC35E2B, NADK, GNB1, TMEM52, CALML6,

KIAA1751, GABRD, PRKCZ*, SKI, MORN1, PEX10, PANK4,

TNFRSF14*, REP1, PLCH2, HES5, MMEL1, ACTRT2,

FLJ42875, PRDM16, MIR4251, ARHGEF16, MEGF6,

[Table 5] Genes in recurrently altered CNV regions and their characteristics

-49-

MIR551A, WDR8, TPRG1L, TP73, KIAA0495, LRRC47,

CCDC27, KIAA0562

1 5957785 6630412 672.6 1p.36.31 NPHP4, KCNAB2, CHD5, RNF207, RPL22, ICMT, GPR153,

HES3, ACOT7, HES2, MIR4252, ESPN, TNFRSF25,

PLEKHG5, NOL9, TAS1R1, ZBTB48, KLHL21, PHF13,

THAP3, DNAJC11

2 176755210 176879235 124.0 2q31.1 MTX2, HOXD1

4 638987 1851365 1212.4 4p16.3 PDE6B, ATP5I*, MFSD7, MYL5, PCGF3, CPLX1, GAK,

IDUA, TMEM175, DGKQ, SLC26A1, FGFRL1, RNF212,

SPON2, KIAA1530, TMED11P, MAEA, CRIPAK, CTBP1,

FAM53A, SLBP, WHSC1, TMEM129, TACC3, FGFR3,

LETM1

5 1 1988770 1988.8 5p15.33 PLEKHG4B, EXOC3, ZDHHC11, SLC6A19, SDHAP3, IRX4,

CCDC127, CEP72, TRIP13, SLC6A18, LRRC14B, TPPP,

TERT, MIR4277, SDHA, BRD9, CLPTM1L, SLC6A3*,

MRPL36, NDUFS6*, PDCD6, AHRR, NKD2, LPCAT1,

SLC12A7, SLC9A3*

7 1 2526797 2526.8 7p22.3 FAM20C, HEATR2, MICALL2, MAD1L1, NUDT1, PDGFA*,

-50-

SUN1, UNCX, INTS1, TFAMP1, FTSJ2, SNX8, FLJ44511,

GPER, TMEM184A, PRKARIB, MIR339, COX19, MAFK,

PSMG3, KIAA1908, ADAP1, GET4, ELFN1, EIF3B, LFNG,

CHST12

7 26885779 27072503 186.7 7p15.2 Unknown

7 150086395 150591510 505.1 7p36.1 TMEM176B, ABP1*, KCNH2*, ATG9B, ABCB8, ACCN3,

NOS3*, TMUB1, FASTK, CDK5*, SLC4A2, AGAP3, ASB10,

GBX1, CHPF2, MIR671, ABCF2, SMARCD3

8 21954537 22086997 132.5 8p21.3 FGF17, EPB49, FAM160B2, HR, NUDT18, REEP4, LGI3,

SFTPC, BMP1*

8 144408549 145707981 1299.4 8q24.3 ZFP41, GLI4, ZNF696, TOP1MT, RHPN1, MAFA, ZC3H3,

PYCRL, TIGD5, GSDMD, NAPRT1, EEF1D, TSTA3, ZNF623,

ZNF707, BREA2, FAM83H, SCRIB, MIR937, PUF60,

MAPK15, NRBP2, EPPK1, PLEC, MIR661, PARP10, GRINA,

SPATC1, GPAA1, OPLAH, EXOSC4, SHARPIN, CYC1,

MAF1, KIAA1875, HEATR7A, BOP1, HSF1, DGAT1, SCRT1,

FBXL6, GPR172A, ADCK5, MIR939, MIR1234, SLC39A4,

VPS28, NFKBIL2, FOXH1, KIFC2, MFSD3, CYHR1, CPSF1,

-51-

PPP1R1, VPS28, CYHR1, RECQ, GPT*

8 142259486 142632471 373.0 8q24.3 DENND3, SLC45A4, GPR20, PTP4A3, FLJ43860

9 127318586 129617946 2299.4 9q33.3 MAPKAP1, PBX3, FAM125B, LMX1B, ZBTB43, ZBTB34,

RALGPS1, GARNL3, SLC2A8, RPL12, SNORA65, ZNF79,

LRSAM1, FAM129B, STXBP1, PTRH1, SH2D3C, TTC16,

ENG

9 135816553 138284852 2468.3 9q34.2-

9q34.3

VAV2, BRD3, NCRNA00094, RNU6ATAC, WDR5, RXRA*,

COL5A1, FCN1, FCN2, OLFM1, KIAA0649, MRPS2, LCN1,

OBP2A, PAEP, GLT6D1, LCN9, SOHLH1, KCNT1,

CAMSAP1, NACC2, UBAC1, LHX3, QSOX2

10 133752383 135374737 1622.4 10q26.3 JAKMIP3, DPYSL4, STK32C, LRRC27, PWWP2B, INPP5A,

NKX6­2, GPR123, MIR202, KNDC1, UTF1, VENTX,

ADAM8, TUBGCP2, CALY, ZNF511, ECHS1, PRAP1, PAOX,

MTG1, SPRN, DUX4(L2,3,5,6,7), CYP2E1*, FRG2B, SYCE1,

SPRNP1

11 1 1557881 1557.9 11p15.5 BET1L, SCGB1C1, ODF3, SIRT3, RIC8A, PSMD13, NLRP6,

IFITM1, IFITM2, IFITM3, IFITM5, ATHL1, B4GALNT4,

SIGIRR, ANO9, PTDSS2, PKP3, RNH1, LRRC56, HRAS,

-52-

IRF7, SCT, MIR210, RASSF7, CDHR5, DEAF1, DRD4*,

TMEM80, PHRF1, EPS8L2, TALDO1, PDDC1, SLC25A22,

LRDD, RPLP2, SNORA52, POLR2L, EFCAB4A, CD151,

TSPAN4, CHID1, PNPLA2, CEND1, AP2A2, MUC2, MUC6,

MUC5B, TOLLIP, BRSK2, MOB2, DUSP8

12 130948875 132087336 1138.5 12q24.33 ULK1, PUS1, EP400, SNORA49, DDX51, NOC4L,

EP400NL, GALNT9, FBRSL1, PXMP2, P2RX2, POLE*,

CHFR, GOLGA3, ZNF26, ZNF605, PGAM5, ANKLE2

14 103604906 105175621 1570.7 14q32.33 ASPG, MIR203, KIF26A, TMEM179, INF2, ADSSL1,

AKT1*, SIVA1, ZBTB42, KIAA0284, MGC23270,

AHNAK2, PLD4, CDCA4, GPR132, JAG2, NUDT14, BRF1,

BTBD6, PACS2, MTA1, CRIP1, CRIP2, TMEM121

16 1 3134387 3134.4 16p13.3 HBA1*, HBA2, AXIN1, LMF1, GFER, TSC2, SSTR5*,

CANCNA1H, GNPTG, CLCN7, IGFALS, HAGH, PKD1,

ABCA3

16 83676307 84402785 726.5 16q24.1 KIAA0513, FAM92B, KIAA0182, MIR1910, COX4NB, GINS2,

COX4I1

16 85079868 85170125 90.3 16q24.1 FOXF1, MTHFSD, FLJ30679, FOXC2

-53-

16 86202175 88827254 2625.1 16q24.2-

16q24.3

JPH3, KLHDC4, SLC7A5, CA5A, BANP, ZNF469, ZFPM1,

ZC3H18, CYBA*, MVD, IL17C, SNAI3, MGC23284, RNF166,

FAM38A, CTU2, APRT1, GALNS, CDT1*, CBFA2T3,

TRAPPC2L, PABPN1L, APRT, ACSF3, CDH15, ZNF778,

ANKRD11, SPG7*, RPL13*, SNORD68*, CPNE7, DPEP1,

CHMP1A, SPATA2L, CDK10, FANCA, ZNF276, SPIRE2,

MC1R*, TCF25, TUBB3, CENPBD1, DEF8, AFG3L1, GAS8,

PRDM7, DBNDD1

17 3696555 3835227 138.7 17p13.2 CAMKK1, P2RX1*, ATP2A3

17 33821353 34160276 338.9 17q12 ARHGAP23, SRCIN1, CISD3, MLLT6, PCGF2

17 76554231 78774742 2220.5 17q25.3 RPTOR, CHMP6, FLJ90757, BAIAP2, AATK, MIR338,

MIR657, MIR1250, MIR3065, MIR3186, AZI1, SLC38A10,

TMEM105, BAHCC1, ACTG1, FSCN2, NPLOC4, TSPAN10,

ARL16, CCDC137, HGS, MRPL12, SLC25A10, PDE6G*,

DYSFIP1, P4HB, ARHGDIA, SIRT7, ANAPC11, PCYT2*,

THOC4, MAFG, NPB, PYCR1, MYADML2, NOTUM, DUS1L,

GPS1, FASN*, ASPSCR1, STRA13, CCDC57, CD7, TEX19,

HEXDC, CSNK1D, SLC16A3, UTS2R, SECTM1, NARF*,

-54-

FOXK2, WDR45L, RAB40B, FN3KRP, FN3K, TBCD, ZNF750,

B3GNTL1, METRNL

18 75222831 76117153 894.3 18q23 ATP9B, NFATC1*, CTDP1, KCNG2, PQLC1, TXNL4A,

HSBP1L1, ADNP2, PARD6G

19 1 2261981 2262.0 19p13.3 WASH5P, FAM138A, FAM138F, OR4F17, FLJ45445,

PPAP2C, MIER2, THEG, C2CD4C, SHC2, ODF3L2,

MADCAM1, CDC34*, BSG, HCN2*, POLRMT, RNF126,

FGF22, GZMM, PRSSL1, PALM, PTBP1, FSTL3, LPPR3,

MIR3187, AZU1*, PRTN3, MED16, ELANE, KISSIR,

ARID3A, CFD, GRIN3B, CNN2, ABCA7, SBNO2, WDR18,

GPX4, POLR2E, HMHA1, CIRBP, STK11*, ATP5D*, MIDN,

EFNA2, RPS15, APC2, NDUFS7*, GAMT, MUM1, DAZAP1,

PCSK4, REEP6, MEX3D, MBD3, UQCR11, ADAMTSL5,

TCF3, PLK5P, ONECUT3, FAM108A1, ATP8B3, ADAT3,

REXO1, MIR1, 9, SCAMP4, KLF16, CSNK1G2, BTBD2,

MKNK2, MOBKL2A, AP3D1, IZUMO4, DOT1L, 6, CS27,

PLEKHJ1, JSRP1, SF3A2, AMH, MIR4321, OAZ1, LINGO3

19 17157595 18921318 1763.7 19p13.11 MYO9B, NR2F6, USE1, OCEL1, USHBP1, ABHD8, ANKLE1,

-55-

ANO8, MRPL34, DDA1, PLVAP, GTPBP3, BST2, NXNL1,

SLC27A1, TMEM221, PGLS, FAM125A, FAM129C, UNC13A,

GLT25D1, MAP1S, FCHO1, JAK3, B3GNT3, SLC5A5*,

INSL3, RPL18A, SNORA68, CCDC124, KCNN1, ARRDC2,

IL12RB1, MAST3, PIK3R2, IFI30, RAB3A, PDE4C*,

MPV17L2, KIAA1683, JUND, LSM4, PGPEP1, LRRC25,

GDP15, MIR3189, ISYNA1, SSBP4, ELL, UBA52, KLHL26,

TMEM59L, CRTC1, FKBP8, CRLF1, LASS1, COPE, UPF1,

COMP, GDF1, HOMER3, DDX49

19 38358874 38597603 238.7 19q13.11 LRP3, SLC7A10, CEBPA, CEBPG, PEPD

19 49895805 51221731 1325.9 19q13.31-

19q13.32

CEACAM16, BCL3*, CBLC, BCAM*, PVRL2, TOMM40*,

APOC1*, APOC1P1, CLPTM1, AOPE, RELB, CLASRP,

GEMIN7, NKPD1, ZNF296, BLOC1S3, TRAPPC6A,

EXOC3L2, ERCC1*, ERCC2*, PPP1R13L*, CKM, KLC3*,

CD3EAP*, RTN2, VASP, OPA3, FOSB, PPM1N, GIPR*,

MIR330, GPR4, EML2, MIR642A, QPCTL, SNRPD2, FBX046,

DMPK, DMWD, RSPH6A, SIX5, NOVA2, FOXA3, IRF2BP1,

CCDC61, MYPOP, NANOS2, SYMPK

-56-

19 60266554 60883384 616.8 19q13.42 RDH13, EPS8L1, PPP1R12C, TNNT1, TNNI3, SYT5,

PTPRH, TMEM86B, PPP6R1, HSPBP1, BRSK1, TMEM150B,

IL11, COX6B2, SUV420H2, FAM71E2, TMEM190, RPL28,

UBE2S, ISOC2, SSC5D, ZNF628, SHISA7, SBK2, NAT14,

ZNF579, FIZ1, ZNF524, ZNF865, ZNF784, ZNF580,

ZNF581, EPN1, CCDC106, U2AF2

19 63540987 63811651 270.7 19q13.43 ZSCAN22, A1BG, ZNF497, A1BGAS, ZNF837, RPS5,

ZNF584, SLC27A5, ZBTB45, TRIM28, CHMP2A, UBE2M,

MZF1, MGC2752

20 60123448 62435964 2312.5 20q13.33 PSMA7, LSM14B, SS18L1, GTPBP5, HRH3*, OSBPL2,

LAMA5, ADRM1, CABLES2, RPS21, GATA5, MIR1­1,

MIR133A2, NTSR1, OGFR, COL9A3, TCFL5, SLCO4A1*,

DPH3P1, DIDO1, SLC17A9, NCRNA00029, HAR1B, YTHDF1,

MIR124­3, NKAIN4, HARIA, BHLHE23, BIRC7, MIR3196,

FLJ16779, ARFGAP1, MIR4326, COL20A1, CHRNA4*,

KCNQ2, EEF1A2, PRIC285, PTK6, SRMS, PPDPF, GMEB2,

STMN3, RTEL1, ARFRP1, ZGPAT, ZBTB46, LIME1,

SLC2A4RG, TPD52L2, TNFRSF6B, UCKL1, NCRNA00176,

PCMTD2, RGS19, ZNF512B, DNAJC5, SAMD10, UCKL1AS,

-57-

TCEA2, SOX18, NPBWR2, OPRL1, MYT1, PRPF6

21 43430725 46689054 3258.3 21q22.3 CRYAA, SIK1*, HSF2BP, RRP1B, RRP1, PDXK, CSTB,

AGPAT3, PWP2, ICOSLG, TRAPPC10, DNMT3L, PFKL,

AIRE, TRPM2, KRTAP10­1, KRTAP12­1, UBE2G2,

ITGB2, SUMO3, PTTG1IP, ADARB1, NCRNA00163,

POFUT2, NCRNA00175, COL18A1, SLC19A1*, PCBP3,

COL6A1, COL6A2, FTCD, LSS, MCM3AP, PCNT

22 15936976 16048532 111.6 22q11.1 IL17RA, CECR1, CECR4, CECR5, CECR6

22 44570586 45471982 901.4 22q13.31 ATXN10, WNT7B, MIRLET7A3, MIRLET7B, PPARA*,

PKDREJ, TTC38, GTSE1, TRMU, CELSR1, GRAMD4, CERK

22 48896256 49691432 795.2 22q13.33 MOV10L1, PANX2, TRABD, SELO, TUBGCP6, HDAC10*,

PLXNB2, MAPK11*, MAPK12*, FAM116B, PPP6R2, SBF1,

ADM2, MIOX, LMF2, NCAPH2, SCO2, TYMP*, ODF3B,

KLHDC7B, MAPK8IP2, ARSA, CPT1B, ACR, RPL23AP82,

RABL2B

Abbreviations: Chr, chromosome; nt, nucleotide.

a. Genes known to be involved in AML are marked in bold, and genes involved in other hematologic or solid

cancer are underlined, and drug related genes are marked with asterisk.

-58-

CharacteristicsCR

(n = 19)

Non‑CR

(n = 11)P

Gender, No. (%)

Male/Female

9 (47.4)/ 10 (52.6)

6 (54.5)/ 5 (45.5)

0.705

Age (years)

Mean (SD)

53.5 (15.8)

51.9 (11.3)

0.768

FAB classification, No. (%)

M0

M1

M2

M4

M5

1 (5.3)

3 (15.8)

8 (42.1)

5 (26.3)

2 (10.5)

0 (0)

2 (18.2)

5 (45.4)

2 (18.2)

2 (18.2)

0.890

Bone marrow blast (%)

Median (range) 59.0 (4.0-96.0) 70.0 (49.8-93.2)

0.048

WBC (x103/㎕)

Median (range)

14.8 (2.2-189.7) 19.7 (3.7-118.5)

0.747

Hb (g/dL)

Median (range) 7.8 (4.8-11.5) 7.3 (4.5-19.4)

0.561

PLT (x10³/㎕)

Median (range)

56.0 (11.0-181.0)

33.0 (6.0-144.0)

0.254

FLT3 ITD, No. (%)

Positive/Negative

4 (33.3)/ 8 (66.7)

2 (22.2)/ 7 (77.8)

0.659

NPM1 mutation, No. (%)

Positive/Negative

6 (66.7)/ 3 (33.3)

4 (30.8)/ 9 (69.2)

0.192

[Table 6] Comparison of CNVs of NK‑AML patients with CR vs.

non‑CR

-59-

Loss of copy number

Patients, No. (%)

No. of loss region/patient

Median (range)

7 (36.8)

12 (1-44)

9 (81.8)

2 (1-79)

0.017

0.782

Gain of copy number

Patients, No. (%)

No. of gain region/patient

Median (range)

12 (63.2)

3 (1-22)

8 (72.7)

3 (1-16)

0.592

1.000

Abbreviations: FAB, French‑American‑British; FLT3 ITD, FMS‑like

tyrosine kinase 3 internal tandem duplication; NPM1, nucleophosmin1.

-60-

Characteristics No. (%)RFS ± SE

(median, months)P

Gender

Male

Female

12 (52.2)

11 (47.8)

8.2 ± 1.472

21.4

0.022

Age (years)

≦ 55.5

> 55.5

11 (47.8)

12 (52.2)

-

9.7 ± 2.165

0.110

FAB classification

M0

M1

M2

M4

M5

1 (4.3)

3 (13.1)

11 (47.8)

5 (21.7)

3 (13.1)

5.5

13.8 ± 4.572

17.7 ± 7.985

6.7 ± 2.300

9.7 ± 5.715

0.122

Bone marrow blast (%)

≦ 66.7

> 66.7

13 (59.1)

9 (40.9)

11.3 ± 1.359

13.8 ± 12.373

0.545

WBC (x103/㎕)

≦ 18.825

> 18.825

11 (47.8)

12 (52.2)

11.3 ± 8.092

9.7 ± 4.386

0.384

Hb (g/dL)

≦ 7.5

> 7.5

11 (47.8)

12 (52.2)

10.9 ± 2.058

16.5 ± 6.928

0.374

PLT (x10³/㎕)

≦ 39

> 39

10 (43.5)

13 (56.5)

10.9 ± 4.071

11.3 ± 4.853

0.625

[Table 7] Comparison of RFS by clinical and molecular

characteristics of NK‑AML patients

-61-

FLT3 ITD

wild type

ITD

12 (75.0)

4 (25.0)

11.3 ± 4.592

6.7 ± 3.050

0.662

NPM1 mutation

wild type

mutation

11 (64.7)

6 (35.3)

10.9 ± 2.293

8.2 ± 8.451

0.506

Loss of copy number

< 3

≧ 3

6 (54.5)

5 (45.5)

8.4 ± 3.087

11.3 ± 0.438

0.896

Gain of copy number

< 4

≧ 4

9 (60.0)

6 (40.0)

17.7 ± 1.789

10.9 ± 3.552

0.083

Abbreviations: FAB, French‑American‑British; FLT3 ITD, FMS‑like

tyrosine kinase 3 internal tandem duplication; NPM1, nucleophosmin1.

-62-

KOR‑CEU KOR‑YRI KOR‑JPT KOR‑CHB

SNP FST SNP FSTa SNP FST SNP FST

rs2853533 0.197 rs4694362 0.519 rs818199 0.083 rs818199 0.096

rs2612100 0.195 rs4846052 0.492 rs17170228 0.016 rs11132158 0.017

rs699517 0.192 rs12648166 0.328 rs3786362 0.016 rs471760 0.015

rs2847150 0.187 rs9542 0.252 rs16879126 0.016 rs2072671 0.010

rs2853532 0.185 rs507964 0.214 rs17170153 0.016 rs4846052 0.010

rs1051527 0.183 rs9651118 0.200 rs6956397 0.015 rs1801131 0.010

rs3734703 0.136 rs4846048 0.198 rs1801133 0.013 rs1476413 0.010

rs2917669 0.132 rs6541003 0.196 rs689452 0.010 rs6541003 0.009

rs689452 0.122 rs6835318 0.188 rs3190314 0.009 rs4694362 0.008

rs10517 0.115 rs693955 0.186 rs1960207 0.009 rs12404655 0.008

rs4846052 0.113 rs11132158 0.181 rs1130902 0.009 rs11532669 0.007

rs7776847 0.112 rs1801133 0.177 rs2049758 0.008 rs11121832 0.007

rs477155 0.109 rs3750117 0.176 rs6835318 0.008 rs1004474 0.007

rs6541003 0.109 rs2392209 0.174 rs7277 0.008 rs532545 0.006

rs3786362 0.107 rs7277 0.174 rs747199 0.008 rs7776847 0.006

rs2847153 0.098 rs1130902 0.174 rs1537516 0.008 rs17421511 0.006

rs1994798 0.097 rs3190314 0.174 rs13306556 0.008 rs4846049 0.005

rs580032 0.091 rs1994798 0.172 rs6462445 0.008 rs9542 0.005

rs3750117 0.086 rs6552621 0.166 rs13139377 0.008 rs1800566 0.005

rs6552621 0.084 rs7663494 0.161 rs7792057 0.008 rs12668520 0.005

[Table 8] FST values for pair‑wise comparisons between Korean

and the HapMap populations in descending rank order

-63-

rs10251079 0.083 rs6462453 0.49 rs10085768 0.008 rs1994798 0.005

rs6462446 0.082 rs12668520 0.149 rs1800566 0.007 rs17170218 0.005

rs4846048 0.080 rs6956397 0.147 rs4338000 0.007 rs1437135 0.005

rs502396 0.077 rs7793793 0.146 rs2392209 0.007 rs4846048 0.005

rs2847149 0.074 rs2049758 0.146 rs4720097 0.007 rs693955 0.004

rs4694362 0.073 rs6954923 0.139 rs4723242 0.007 rs4655226 0.004

rs2244500 0.073 rs3734703 0.136 rs1437135 0.007 rs3809585 0.004

rs2612095 0.072 rs7688234 0.134 rs7793793 0.007 rs2847153 0.004

rs2612095 0.072 rs6462450 0.133 rs6462453 0.007 rs1537516 0.003

rs2612095 0.072 rs6462446 0.130 rs7801986 0.007 rs13306556 0.003

rs2612095 0.071 rs10281012 0.124 rs3734703 0.007 rs3750117 0.003

rs603412 0.069 rs17170180 0.120 rs2847149 0.007 rs603412 0.003

rs4846049 0.069 rs4846049 0.117 rs9651118 0.007 rs9651118 0.003

rs1800566 0.068 rs6948212 0.117 rs1004474 0.007 rs6835318 0.002

rs2853741 0.068 rs6955792 0.114 rs17367504 0.006 rs10916824 0.002

rs1437135 0.068 rs818194 0.112 rs3778504 0.006 rs324148 0.002

rs1801131 0.052 rs10486512 0.109 rs2853533 0.006 rs1057985 0.002

rs10009825 0.049 rs3786362 0.107 rs1476413 0.006 rs10916827 0.002

rs1057985 0.046 rs747199 0.107 rs4720098 0.006 rs4986998 0.002

rs7688234 0.042 rs477155 0.100 rs12668520 0.005 rs7688234 0.002

rs6956397 0.041 rs689452 0.098 rs2244500 0.005 rs9394992 0.002

rs17170218 0.041 rs10517 0.090 rs2612095 0.005 rs3190314 0.002

rs507964 0.041 rs10085768 0.083 rs760370 0.005 rs12648166 0.002

rs13139377 0.039 rs11121832 0.082 rs2464974 0.005 rs6946062 0.002

-64-

rs7663494 0.039 rs603412 0.081 rs2847153 0.005 rs7277 0.002

rs1476413 0.039 rs4720098 0.078 rs1001761 0.005 rs3737965 0.002

rs17170228 0.039 rs1057985 0.075 rs10251079 0.005 rs502396 0.002

rs16879126 0.038 rs6462445 0.074 rs324148 0.005 rs3811810 0.002

rs17170153 0.038 rs7792057 0.074 rs6462446 0.005 rs17367504 0.001

rs10799647 0.036 rs10256717 0.073 rs10256717 0.004 rs760370 0.001

rs10520543 0.035 rs4720097 0.073 rs2847150 0.004 rs1130902 0.001

rs13114435 0.035 rs4723242 0.073 rs4846052 0.004 rs6956397 0.001

rs4720098 0.034 rs7801986 0.071 rs10916824 0.004 rs2847150 0.001

rs2072671 0.034 rs2917669 0.071 rs2612100 0.004 rs747199 0.001

rs6462453 0.032 rs4338000 0.070 rs2853532 0.004 rs2464974 0.001

rs532545 0.032 rs13101260 0.066 rs818196 0.004 rs507964 0.001

rs2392209 0.032 rs412543 0.065 rs577042 0.004 rs10009825 0.001

rs6948212 0.031 rs1800566 0.065 rs4986998 0.004 rs13139377 0.001

rs7793793 0.031 rs324148 0.062 rs1051527 0.004 rs3786362 0.001

rs2049758 0.031 rs17170228 0.060 rs471760 0.004 rs6954923 0.001

rs6955792 0.031 rs16879126 0.059 rs699517 0.003 rs10486512 0.001

rs3811810 0.031 rs17170153 0.059 rs1801131 0.003 rs13101260 0.001

rs13147196 0.031 rs2847153 0.086 rs6946062 0.003 rs17170180 0.001

rs10486512 0.030 rs10916824 0.048 rs10916827 0.003 rs12499918 0.001

rs10281012 0.030 rs2072671 0.046 rs3750117 0.003 rs1689924 0.001

rs1689924 0.029 rs471760 0.043 rs6541003 0.003 rs2853532 0.001

rs6954923 0.029 rs10251079 0.041 rs2853741 0.003 rs6948212 0.001

rs17170180 0.028 rs532545 0.034 rs7776847 0.003 rs1051527 0.001

-65-

rs1801133 0.028 rs11532669 0.033 rs10281012 0.003 rs818196 0.001

rs10085768 0.027 rs1004474 0.032 rs12648166 0.003 rs7663494 0.001

rs9990999 0.027 rs3811810 0.031 rs11532669 0.003 rs2847149 0.001

rs6462445 0.026 rs17421511 0.030 rs532545 0.002 rs2853533 0.001

rs7792057 0.026 rs6946062 0.028 rs603412 0.002 rs2612100 < 0.001

rs4338000 0.025 rs17170218 0.028 rs6948212 0.002 rs2244500 < 0.001

rs4720097 0.025 rs2464974 0.027 rs4655226 0.002 rs2612095 < 0.001

rs4723242 0.025 rs818196 0.027 rs507964 0.002 rs17331744 < 0.001

rs17421511 0.024 rs12499918 0.026 rs9542 0.002 rs17272827 < 0.001

rs7801986 0.024 rs12404655 0.024 rs4846049 0.002 rs10517 < 0.001

rs471760 0.024 rs13147196 0.023 rs6955792 0.002 rs1001761 < 0.001

rs10256717 0.024 rs1437135 0.022 rs2917669 0.002 rs10281012 < 0.001

rs17367504 0.021 rs580032 0.021 rs10486512 0.002 rs477155 < 0.001

rs17272827 0.019 rs1689924 0.020 rs580032 0.002 rs13116494 < 0.001

rs9651118 0.018 rs2853741 0.019 rs1057985 0.002 rs699517 < 0.001

rs1960207 0.017 rs577042 0.018 rs11132158 0.002 rs6955792 < 0.001

rs12668520 0.016 rs7776847 0.017 rs10009825 0.001 rs3778504 < 0.001

rs693955 0.016 rs13148414 0.016 rs3811810 0.001 rs13147196 < 0.001

rs17331744 0.014 rs3809585 0.016 rs1994798 0.001 rs3734703 < 0.001

rs3886768 0.014 rs2853533 0.016 rs6552621 0.001 rs10799647 < 0.001

rs3778504 0.014 rs2853532 0.016 rs6954923 0.001 rs2917669 < 0.001

rs13148414 0.013 rs17331744 0.016 rs2072671 0.001 rs17170228 < 0.001

rs11132158 0.013 rs3886768 0.016 rs502396 0.001 rs6462450 < 0.001

rs12404655 0.012 rs13139377 0.016 rs10517 0.001 rs689452 < 0.001

-66-

rs1537516 0.010 rs699517 0.015 rs17421511 0.001 rs6462453 < 0.001

rs13306556 0.010 rs2244500 0.014 rs17272827 0.001 rs1960207 < 0.001

rs577042 0.009 rs10520543 0.014 rs13147196 0.001 rs2049758 < 0.001

rs6462450 0.009 rs2847149 0.014 rs6462450 0.001 rs2392209 < 0.001

rs1004474 0.008 rs13114435 0.014 rs12499918 0.001 rs13116598 < 0.001

rs760370 0.007 rs4655226 0.014 rs9394992 < 0.001 rs3886768 < 0.001

rs4655226 0.007 rs1001761 0.013 rs3737965 < 0.001 rs10520543 < 0.001

rs13116494 0.006 rs2612095 0.013 rs3809585 < 0.001 rs17170153 < 0.001

rs4986998 0.006 rs2847150 0.012 rs693955 < 0.001 rs16879126 < 0.001

rs527912 0.005 rs3778504 0.011 rs10799647 < 0.001 rs7793793 < 0.001

rs324148 0.005 rs1051527 0.010 rs4846048 < 0.001 rs6462445 < 0.001

rs13116598 0.004 rs10799647 0.009 rs7688234 < 0.001 rs6552621 < 0.001

rs7277 0.004 rs10916827 0.005 rs477155 < 0.001 rs13148414 < 0.001

rs10916824 0.004 rs502396 0.004 rs13101260 < 0.001 rs7792057 < 0.001

rs3190314 0.004 rs3737965 0.003 rs13114435 < 0.001 rs10085768 < 0.001

rs1130902 0.004 rs760370 0.003 rs3886768 < 0.001 rs1801133 < 0.001

rs3809585 0.004 rs1476413 0.002 rs17331744 < 0.001 rs527912 < 0.001

rs6835318 0.004 rs1801131 0.002 rs10520543 < 0.001 rs818194 < 0.001

rs818196 0.002 rs13306556 0.002 rs11121832 < 0.001 rs4338000 < 0.001

rs12499918 0.002 rs9990999 0.002 rs12404655 < 0.001 rs6462446 < 0.001

rs9542 0.002 rs4986998 0.002 rs17170218 < 0.001 rs580032 < 0.001

rs3737965 0.001 rs2612100 0.001 rs13148414 < 0.001 rs4720098 < 0.001

rs412543 0.001 rs17272827 0.001 rs7663494 < 0.001 rs4723242 < 0.001

rs13101260 0.001 rs13116494 0.001 rs527912 < 0.001 rs4720097 < 0.001

-67-

rs747199 0.001 rs527912 0.001 rs13116598 < 0.001 rs10251079 < 0.001

rs6946062 < 0.001 rs13116598 0.001 rs17170180 < 0.001 rs10256717 < 0.001

rs2464974 < 0.001 rs17367504 < 0.001 rs13116494 < 0.001 rs2853741 < 0.001

rs12648166 < 0.001 rs10009825 < 0.001 rs818194 < 0.001 rs577042 < 0.001

rs10916827 < 0.001 rs9394992 < 0.001 rs1689924 < 0.001 rs13114435 < 0.001

rs818194 < 0.001 rs1537516 < 0.001 rs9990999 < 0.001 rs9990999 < 0.001

rs9394992 < 0.001 rs4694362 < 0.001 rs7801986 < 0.001

Abbreviations: CEU, Central European; CHB, Chinese; JPT, Japanese;

KOR, Korean; YRI, African.

-68-

[Figure 1] Ara‑C transport and metabolic pathway

-69-

[Figure 2] Single SNP effect of DCK rs4694362 on OS

-70-

[Figure 3] Combined effects of SNPs on RFS and OS

-71-

[Figure 4] Difference in allele frequency of SNPs between

Korean and other populations

-72-

[Figure 5] Combined effect of SLC29A1 and TYMS on ara‑C

metabolism in blast cells

-73-

국

골 병 료에

약 인자 분

경 임

울 학 학원

약학과 ·임상약학 공

주요어 : 골 병, 시타라 , 내 , 단일염 다 ,

단 복 변이, 인종

학번 : 2007­30944

골 병(acute myeloid leukemia, AML) 림프구 계통

구 구 포 암 증식에 한 질 , 자 포 학

이상에 라 후 양 군, 간군 불량군 분 한다. Cytarabine

arabinoside (ara‑C)는 AML 료 해 도요법과 공고요법 모 에

사용 는 핵심약 이나, 체 AML 자 약 20-30%만이 장

병생존에 도달하는 등 같 후군에 속한 자들에 도 그

료 이 매우 다양한 이 주 난 이다. 또한 AML 자 약

50%는 포 학 이상 가지지 않는 상핵 (normal karyotype,

NK‑AML) 간 후군 분 다. 그러나 이들 자들에 도

-74-

그 료 과에 매우 큰 차이를 보이고 있어, NK‑AML 자에

염색체상 변이 외에 료 할 있는 또 다른 인자 규명이

실하다.

Genetic/genomic alteration 약 특 단 질 능

또는 에 향 끼침 써 개인 간 약 료 과 차이에

여하는 것 주목 고 있다. 특히 앞 여러 연구는 ara‑C 송

사 과 에 속한 자 단일염 변이(single nucleotide

polymorphism, SNP) 같 다양 과 ara‑C 료효과 간

상 계에 주목해 다. 그러나 부분 연구가 복잡한 자

상 작용 통한 SNP들 간 병합효과를 고 하지 않고 개별

SNP 써 향 평가에 그쳐 아직 자간 료 다양 에 한

원인규명에 있어 그 결 에 이르지 못하고 있다. 또한 체

구조 변이인 단 복 변이(copy number variation, CNV)는 암 또는

여타 질 직 원인 또는 감 인자 작용함이 알 있 며

근에는 질 생 외에도 항암 료 시 료 과 또는 이상 에

여함이 보고 있다.

AML 자에 ara‑C 료 개인 간 차이를 명하

하여, 본 연구에 는 M3를 외한 AML 자 97명 상

그들 진단 시 골 또는 말 액샘플 이용하여 ara‑C 송

사에 여하는 139개 후보 SNP Illumina GoldenGate

Genotyping Assay (Illumina Inc., San Diego, CA, USA) 이용하여

분 하 다. 또한 NK‑AML 자 30명에 해 는 HelixTreeⓇ

software version 5.2.0 (Golden Helix Inc., Bozeman, MT, USA)

이용하여 추가 CNV 분 시행하 다.

개별 SNP 또는 SNP‑SNP 조합과 료 과간 상

-75-

분 하 며, dominant, recessive additive SNP model 모

용하여 그 가장 합한 model 이용하 다. Ara‑C 송 사

각 SNP과 AML 료 과에 향 미 는 것 알 진 임상 ,

포 학 요인 함께 고 한 다변량 분 에 DCK 자 SNP

rs4694362 일하게 체생존 간에 통계 있는 향

끼쳤다. DCK SNP rs4694362 CC 가지고 있는 자들

CT 또는 TT 가지고 있는 자들에 해 통계

있게 체 생존 간 감소를 보 다 (HR, 33.202 [95% CI,

4.937-223.273], P < 0.0001, PBonferroni = 0.017). 또한 SNP 조합과

AML 료 과에 향 미 는 것 알 진 임상 , 포 학

요인 함께 고 한 다변량 분 에 SLC29A1 자 SNP

rs3734703 AA 또는 AC 과 TYMS rs2612100 AA

조합 wild type 조합에 해 통계 있게

재 험도를 증가시 다 (HR, 17.630 [95% CI, 4.829-64.369], P

< 0.0001, PBonferroni = 0.021). 이 SNP 간 조합 체 생존 간

감소에도 여하 다 (HR, 23.523 [95% CI, 4.616-119.873], P =

0.0001). 이외에 CDA rs10913827 GG 과 DCTD rs17331744

TC 또는 CC 간 조합 wild type 조합에 해

체 생존 간 감소 경향 보 다 (HR, 31.680, [95% CI,

6.152-162.905], P < 0.0001, PBonferroni = 0.052). 약 체

연구에 자‑ 자 상 작용 고 한 SNP간 조합 , 단독 SNP

향만 는 명하지 못하는 복잡한 약 동태학 약 역학

명하는데 요한 역할 할 있다. 라 본 연구결과는 ara‑C

약 송 사과 에 참여하는 다양한 SNP들 상 작용이 ara‑C

학요법 는 인 AML 자 약 결 에 여할 있

-76-

인하고 그 근거를 마 하 다는 에 를 가진다.

30명 NK‑AML 자 증 23명(76.7%)에 CNV가

견 었 며, 이들 자군 료 과 진단 시 주요 임상지

후인자 등 특 CNV가 견 지 않 자군과 통계

있는 차이를 보이지 않았다. 다변량 분 시행한 결과, 복 감소가

있는 자들에 통계 있는 낮 해 (complete

remission, CR) 도달 도를 인할 있었다 (OR, 0.015 [95% CI,

0-0.737], P = 0.035). 또한 복 증가를 가진 자 , 4개

이상 증가구역 가진 자들 그 지 않 자들보다 통계

있게 증가 재 험 가지는 것 다변량 분 에 인할

있었다 (HR, 22.104 [95% CI, 1.644-297.157], P = 0.020).

또한 본 연구에 는 University of Canada database

(http://projects.tcag.ca/variation/project.html) PharmGKB

database (http://www.pharmgkb.org/index.jsp)를 이용하여 CNV가

생한 에 존재하는 자를 도출하고 해당 자 암 또는

약 과 상 여부를 인하 다. 그 결과, 19개 자(HES5,

PRDM16, TNFRSF25, MTX2, TERT, ABCB8, PTP4A3, PBX3,

VENTX, AKT1, KIAA0284, ABCA3, CBFA2T3, FANCA, MLLT6,

CD7, PRTN3, CEBPA TYMP)가 AML과 가짐이 보고

있었다. 또한 복 감소구역에 포함 NOS3, ERCC1, ERCC2,

ATP5I, ATP5D, CYBA, NDUFS7, SLC19A1 P2RX1 9개

자 경우 ara‑C 또는 anthracycline계 항암 효과

이상 과 있 이 인 하 다.

이 함께 본 연구에 는 ara‑C 항암 학요법에 인종간

료 다양 명하 한 일 상 139개 SNP에 해

-77-

인구집단 거리 분 통 인 법인 FST 분 법

이용하여 인종 간 minor allele frequency (MAF) 차이를 분 하 다.

집단 International HapMap database (phase III,

http://hapmap.ncbi.nlm.nih.gov/) 내 Caucasian, Chinese, Japanese

African 4개 인종이었다. 139개 SNP International HapMap

database에 포함 지 않 SNP에 해 는 1000 genomes database

(http://www.1000genomes.org/)를 이용하 다. 139개

SNP MAF는 아시아 인종 내에 매우 사하 며, Caucasian과

African 경우 한국인과 큰 차이를 보 다. 한국인과 타인종과

, DCK rs4694362가 African과 시에 가장 큰 FST값 가 다

(FST = 0.519). 이 외에 본 연구에 ara‑C 료 과 한

상 보인 TYMS rs2612100과 SLC29A1 rs3734703 역시

Caucasian 또는 African과 시에 한 인종간 MAF 차이를

보 다. 이는 ara‑C 료 과가 인종 간 차이를 보임에 있어,

약 사 에 여하는 자 SNP 도 인종 간 차이가 여할

것이라는 근 연구가 뒷 침하는 결과이다.

결과 본 연구 결과는 한국인 AML 자 료 과를

할 있는 근거 용 있 며, 이는 이들 자 료 과

향상 한 근 를 마 하 다는 를 가진다.