The Pharmacogenomics Journal (2013) 13, 417 -- 422 & 2013 Macmillan Publishers Limited All rights reserved 1470-269X/13 www.nature.com/tpj

ORIGINAL ARTICLE A genome-wide association study for irinotecan-related severe toxicities in patients with advanced non-small-cell lung cancer

J-Y Han1, ES Shin2, Y-S Lee3, HY Ghang4, S-Y Kim3, J-A Hwang3, JY Kim1 and JS Lee1

The identification of patients who are at high risk for irinotecan-related severe diarrhea and neutropenia is clinically important. We conducted the first genome-wide association study (GWAS) to search for novel susceptibility for irinotecan-related severe toxicities, such as diarrhea and neutropenia, in non-small-cell lung cancer (NSCLC) patients treated with irinotecan chemotherapy. The GWAS putatively identified 49 single-nucleotide polymorphisms (SNPs) associated with grade 3 diarrhea (G3D) and 32 SNPs associated with grade 4 neutropenia (G4N). In the replication series, the SNPs rs1517114 (C8orf34), rs1661167 (FLJ41856) and rs2745761 (PLCB1) were confirmed as being associated with G3D, whereas rs11128347 (PDZRN3) and rs11979430 and rs7779029 (SEMAC3) were confirmed as being associated with G4N. The final imputation analysis of our GWAS and replication study showed significant overlaps of association signals within these novel variants. This GWAS screen, along with subsequent validation and imputation analysis, identified novel SNPs associated with irinotecan-related severe toxicities.

The Pharmacogenomics Journal (2013) 13, 417--422; doi:10.1038/tpj.2012.24; published online 5 June 2012 Keywords: C8orf34; FLJ41856; PLCB1; PDZRN3; SEMA3C; irinotecan

INTRODUCTION technologies has paved the way for a genome-wide association 13 Irinotecan is one of the most active chemotherapeutic agents for study (GWAS) to identify novel susceptibility genes. Therefore, patients with colorectal and lung cancers. However, some patients we conducted the first GWAS to search for novel genes involved suffer from irinotecan-related severe toxicities, such as diarrhea in susceptibility to irinotecan-related severe toxicity, such as and neutropenia.1 Although comprehensive data supporting the diarrhea and neutropenia, in NSCLC patients treated with association between homozygous UGT1A1*28 and grade 4 irinotecan chemotherapy. neutropenia exist, this association does not account for all 2--4 toxicities observed with irinotecan chemotherapy. Furthermore, MATERIALS AND METHODS genetic variations in other UGT1A genes and drug transporters Study population have been suggested to contribute to the variability of irinotecan To construct a relatively homogenous treatment group, we restricted pharmacokinetics and toxicities.5--8 Previously, we investigated the patients who participated in clinical trials and received irinotecan-based associations of variants in other UGT1A genes and transporters chemotherapy as first-line therapy for their advanced NSCLC. In addition, with irinotecan pharmacokinetics and toxicity in Korean non- we collected toxicity data prospectively through clinical trials. For the initial small-cell lung cancer (NSCLC) patients treated with irinotecan- 9--11 GWAS discovery stage, cases and controls were obtained from two phase II based chemotherapy. We found that the UGT1A1*6/*6 and studies of irinotecan and cisplatin chemotherapy, which enrolled chemo- SLCO1B1 521TC or CC genotypes were associated with a naive patients with advanced NSCLC. A total of 81 patients received significantly higher area under the curve (AUC) of SN-38, an irinotecan 80 mg mÀ2 intravenously (i.v.) on days 1 and 8 plus cisplatin active metabolite of irinotecan, and were independently pre- 60 mg mÀ2 i.v. on day 1 every 3 weeks, whereas 22 patients were treated dictive for grade 4 neutropenia (G4N). Although no significant with irinotecan 65 mg mÀ2 and cisplatin 30 mg mÀ2 i.v. on days 1 and 8 pharmacokinetic parameters for severe diarrhea were found, the every 3 weeks. Full details of the studies have been reported UGT1A9*22 9/9, ABCC2 3972CC and ABCG2 34GA or AA genotypes previously.14,15 The significance of the single-nucleotide polymorphisms 12 were independently predictive for grade 3 diarrhea (G3D). (SNPs) identified at the discovery stage was subsequently confirmed Nevertheless, much of the variation in irinotecan pharmaco- through a replication study in independent subjects enrolled in another kinetics and toxicity remains unexplained. randomized phase II study of irinotecan-based chemotherapy, which also Owing to the severity of irinotecan-related toxicity, identifica- recruited chemo-naive patients with advanced NSCLC. In this study, 73 tion of patients who are at high risk for severe diarrhea and patients received irinotecan 80 mg mÀ2 i.v. on days 1 and 8 plus cisplatin neutropenia is clinically essential. To date, irinotecan pharmaco- 60 mg mÀ2 i.v. on day 1 every 3 weeks, and 73 patients received irinotecan genetic studies have mainly consisted of association analyses of 90 mg mÀ2 i.v. on days 1 and 8 plus oral capecitabine 1000 mg mÀ2 twice candidate genes. However, the toxicity phenotypes of cancer daily on days 1 through 14 every 3 weeks. These studies were conducted patients are complicated by many factors; thus, a more compre- under the approval of the ethical review boards and the guidelines hensive and systemic approach is needed. The recent develop- for good clinical practice. All subjects gave informed consent for the ment of new-generation and cost-effective high-throughput genetic testing.

1Lung Cancer Branch, Research Institute and Hospital, National Cancer Center, Goyang, Gyeonggi, Republic of Korea; 2DNALink, Seoul, Republic of Korea; 3Functional Genomic Branch, Research Institute and Hospital, National Cancer Center, Goyang, Gyeonggi, Republic of Korea and 4Korea Research Institute of Bioscience and Biotechnology, Daejeon, Korea. Correspondence: Dr J-Y Han, Center for Lung Cancer, Research Institute and Hospital, National Cancer Center, 111 Jungbalsan-ro, Ilsandong-gu, Goyang-si, Gyeonggi-do, 410-769, Republic of Korea. E-mail: [email protected] Received 23 February 2012; revised 26 April 2012; accepted 30 April 2012; published online 5 June 2012 GWAS for irinotecan toxicities J-Y Han et al 418 Definition of case and control for irinotecan-related severe Table 1. Characteristics of study subjects toxicities Our goal was to find novel genetic variants associated with irinotecan- GWAS sample (%) Replication sample (%) related severe diarrhea and neutropenia through a GWAS. Cases were required to have G3D or G4N during the course of irinotecan-based Number 101 146 chemotherapy. Controls had not experienced G3D or G4N during Age, years irinotecan-based chemotherapy. Toxicity was assessed throughout treat- Median (range) 60 (29--76) 58 (31--77) ment and was graded using the National Cancer Institute’s Common Terminology Criteria (NCI-CTC version 2.0). Gender Male 77 (76) 134 (92) Female 24 (24) 12 (8) Genotyping and quality control Peripheral blood samples were obtained at baseline, and genomic DNA Grade 4 neutropenia was extracted using a QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, Yes 24 (24) 21 (14) No 77 (76) 125 (86) USA) in accordance with the manufacturer’s instructions. A genome-wide scan was performed using an Affymetrix Genome-Wide SNP array Grade 3 diarrhea 5.0 (Affymetrix, Santa Clara, CA, USA) comprising 440 094 genome-wide Yes 11 (11) 13 (9) SNPs. Genotype calls were determined from the fluorescence intensities No 90 (89) 133 (91) using the DM algorithm with a 0.33 P-value setting, as well as the B-RLMM algorithm. Abbreviation: GWAS, genome-wide association study. Quality control procedures included the following steps. First, only samples with at least 95% call rate were included. SNPs with a call rate o95% in each panel of case or control were discarded. Second, SNPs deviating from Hardy--Weinberg equilibrium in controls (Po0.001) were in the cases (G3D or G4N) was o0.15, we identified 48 SNPs for not included in the w2 test. Finally, those with a minor allele frequency of G3D and 31 SNPs for G4N. Next, we proceeded to replicate these 40.05 were included for analysis. associations in an independent NSCLC patient population receiv- Based on the initial GWAS results, we selected the most promising SNPs ing an irinotecan-based chemotherapy. The replication analysis for subsequent genotyping in the replication sample according to the revealed the three most significantly associated SNPs for each following inclusion criteria: (1) Po1 Â 10À4 in the GWAS allelic association toxicity. The significant associations observed in the GWAS and analyses and (2) minor allele frequency in the case population (G3D or the replication study are shown in Table 2. rs1517114 in intron 3 G4N) of X0.15. Genotyping was performed using an iPLEX Gold assay on of C8orf34 ( 8 open reading frame 34), rs1661167 in 0 the MassARRAY platform (Sequenom, San Diego, CA, USA) based on the 3 untranslated region of FLJ41856 (CEACAM22P (carcinoem- matrix-assisted laser desorption/ionization time-of-flight spectrometry, bryonic antigen-related cell adhesion molecule 2, )) and according to the manufacturer’s instructions.16 For quality control of the rs2745761 in intron 2 of PLCB1 (phospholipase C, b1) showed genotyping assay, 10% of the samples were run as duplicates and tested strong associations with G3D. Regarding G4N, rs11128347 in for concordance. Additionally, genotype clusters were examined manually intron 3 of PDZRN3 (PDZ domain--containing ring finger 3) and two to determine their fitness. The resultant genotype data were collected SNPs (rs11979430 and rs7779029) located close to SEMA3C (sema using Typer software (Sequenom, version 4.0). domain, immunoglobulin domain (Ig), short basic domain, secreted, (semaphorin) 3C) on showed strong associations. A statistically significant LD was observed between rs11979430 Statistical analysis and rs7779029 (r2 ¼ 0.911). Upon combining the GWAS and The association of a specific SNP with G3D or G4N was analyzed by replication results, the statistical significances were significantly comparing the minor allele frequencies of the cases and controls, with improved for all of these SNPs. The combined P-values for these significance determined by P-values of Jonckheere--Terpstra tests. The SNPs are summarized in Table 2. odds ratio of cases with a selected SNP compared with the controls and its relevant 95% confidence intervals were determined. As the sample size Validation and missing data imputation at the genotyped SNPs was small, the thresholds for declaring significance were Po1.0 Â 10À4 for the GWAS and Po0.05 for the replication study. All association tests were To provide independent estimates of genotypes associated with based on the comparison of alleles. PLINK 1.07 and SAS 9.1.3 (SAS Institute, G3D or G4N, we imputed all of the known SNPs in the 50-kb Cary, NC, USA) were used for statistical analysis. Linkage disequilibrium region of interest based on the genotyped SNPs. The imputed (LD) analysis was quantified with Haploview, version 4.2 (http://www. association signals in the combined samples (GWAS þ replication) broadinstitute.org/haploview/haploview).17 For the genomic regions of are plotted in Supplementary Figures 1--4. According to the interest, the IMPUTE program version 1 (https://mathgen.stats.ox.ac.uk/ imputation results in the combined GWAS and replication data set, impute/impute_v1.html) was utilized to impute the genotypes of all SNPs 47 SNP markers for G3D and 26 SNP markers for G4N achieved À4 located in the regions based on Asian HapMap data.18 P-values of o1 Â 10 . Among them, 22 SNPs showed a stronger association with G3D, which achieved a combined P-value of o1 Â 10À5 and were located within a 24.5-kb LD block of RESULTS C8orf23 intron 3 and an 11.9-kb LD block of PLCB1 intron 2. GWAS discovery and identification of novel SNPs Regarding G4N, 16 SNPs achieved a combined P-value of À5 The study design included an initial GWAS discovery stage and a o1 Â 10 ; these were located within a 42.6-kb LD block of follow-up replication and validation stage. The characteristics of PDZRN3 intron 3. Detailed information for the plotted SNPs is the respective samples are detailed in Table 1. In the GWAS presented in Table 3. discovery stage, 312 230 SNPs passed the quality control criteria for GWAS analysis for G4D and G3D. We found 137 SNPs for G3D Previously described genetic associations and 42 SNPs for G4N with Po1.0 Â 10À4. The association P-values Previously, we found that UGT1A1*6 and SLCO1B1 521T4C were were very modest, with only one reaching the established significantly associated with irinotecan-related G4N. Moreover, genome-wide significance level at 10À7 (best P ¼ 6.9 Â 10À8). UGT1A9*22, ABCC2 3972C4T and ABCG2 34G4A were associated The distribution of P-values across the genome is illustrated in with irinotecan-related G3D. Thus, we investigated their effects on Figure 1. After discarding the SNPs whose minor allele frequency irinotecan-related G4N or G3D in the present study. Table 4 shows

The Pharmacogenomics Journal (2013), 417 -- 422 & 2013 Macmillan Publishers Limited GWAS for irinotecan toxicities J-Y Han et al 419 a 7 6 5 4 3

-log10 (P) 2 1 0 Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7Chr8 Chr9 Chr10 Chr11 Chr12 Chr13 Chr14 Chr15 Chr16 Chr17 Chr18 Chr19 Chr20 Chr21 Chr22 ChrXY b 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5

-log10(P) 2.0 1.5 1.0 0.5 0.0 Chr1 Chr2 Chr3 Chr4 Chr5 Chr6 Chr7Chr8 Chr9 Chr10 Chr11 Chr12 Chr13 Chr14 Chr15 Chr16 Chr17 Chr18 Chr19 Chr20 Chr21 Chr22 ChrXY Figure 1. Manhattan plot of genome-wide association study (GWAS) results for testing the associations with grade 3 diarrhea (a) and grade 4 neutropenia (b).

Table 2. Novel SNPs identified by GWAS and replication study

Toxicity Ch Nearest SNP Position Location Allelea Stage MAF case MAF control P-value OR (95% CI)

G3D 8 C8orf34 rs1517114 69551771 Intron 3 C/G GWAS 0.455 0.112 2.5 Â 10À5 6.6 (2.5--17.4) Replication 0.292 0.133 0.035 2.7 (1.04--7.00) Combined 0.370 0.124 8.5 Â 10À6 4.1 (2.1--8.0) 19 FLJ41856 rs1661167 49743908 30 UTR G/A GWAS 0.409 0.088 1.9 Â 10À5 7.2 (2.6--19.5) Replication 0.292 0.136 0.041 2.6 (1.01--6.7) Combined 0.348 0.118 1.9 Â 10À5 4.0 (2.0--7.9) 20 PLCB1 rs2745761 8225943 Intron 2 C/G GWAS 0.318 0.047 8.7 Â 10À6 9.5 (3.0--29.7) Replication 0.231 0.089 0.022 3.1 (1.1--8.5) Combined 0.271 0.072 6.4 Â 10À6 4.8 (2.3--10.0) G4N 3 PDZRN3 rs11128347 73702251 Intron 3 C/G GWAS 0.229 0.042 9.2 Â 10À5 6.7 (2.3--19.4) Replication 0.214 0.088 0.014 2.8 (1.2--6.7) Combined 0.222 0.071 1.6 Â 10À5 3.7 (2.0--7.0) 7 SEMA3C rs11979430 80347192 Intron 2 T/C GWAS 0.313 0.085 9.2 Â 10À5 4.9 (2.1--11.5) Replication 0.250 0.129 0.044 2.3 (1.01--5.0) Combined 0.284 0.113 3.6 Â 10À5 3.1 (1.8--5.5) 7 SEMA3C rs7779029 80370048 Intron 2 C/T GWAS 0.326 0.089 7.1 Â 10À5 5.0 (2.1--11.5) Replication 0.262 0.136 0.037 2.3 (1.04--4.9) Combined 0.296 0.119 2.8 Â 10À5 3.1 (1.8--5.4) Abbreviations: Ch, chromosome; CI, confidence interval; GWAS, genome-wide association study; G3D, grade 3 diarrhea; G4N, grade 4 neutropenia; MAF, minor allele frequency; OR, odds ratio; SNP, single-nucleotide polymorphism; UTR, untranslated region. aThe former allele represents the minor allele. ORs for minor allele (allele model). All P-values listed are two sided. the analysis of SNPs in the previously mentioned genes. Whereas potentially associated with irinotecan-related severe diarrhea UGT1A1 was present in our GWAS panel, the remaining four SNPs and neutropenia in Korean patients with NSCLC. The follow-up were not present. Thus, we analyzed these four SNPs in the stage confirmed the associations of SNPs with irinotecan- replication samples (Table 4). Although UGT1A1*6 failed to reach related severe diarrhea and neutropenia, as the associations genome-wide significance, the combined analysis showed that were replicated in an independent NSCLC patient sample. homozygous UGT1A1*6 was significantly associated with G4 Furthermore, imputation analysis of our GWAS and replication neutropenia. Additionally, the ABCC2 3972 C4T and SLCO1B1 study showed significant overlaps of the association signals. 521T4C variants were significantly associated with G3D and G4N, We demonstrated that 47 and 26 SNPs showed strong asso- respectively, in the replication stage. ciations (Po1 Â 10À4) with G3D and G4N, respectively. In particular, a narrow 11.9-kb window of intron 2 of PLCB1 contained 7 SNPs that achieved P-values of o1 Â 10À5 and DISCUSSION showed complete LD with rs274561. Additionally, 13 SNPs This study represents the first GWAS for severe diarrhea in intron 3 of C8orf34 achieved P-values of o1 Â 10À5 and and neutropenia, which are clinically important and potentially showed high LD with rs1517114. According to our imputa- fatal toxicities for cancer patients treated with irinotecan tion analysis for G4N, 16 SNPs in intron 3 of PDZRN3 achieved chemotherapy. Through the GWAS, we identified new SNPs P-values of o1 Â 10À5 and showed complete LD with rs11128347.

& 2013 Macmillan Publishers Limited The Pharmacogenomics Journal (2013), 417 -- 422 GWAS for irinotecan toxicities J-Y Han et al 420 Table 3. Major imputation results

LD information SNP ID Ch Position Region OR (95% CI) P-value Sig SNP ID SNP ID D0 r2

Grade 3 diarrhea rs6133571 20 8228341 Intron 4.972 (2.387--10.357) 3.34926EÀ06 rs2745761 rs6133571 1 1 rs6140592 20 8228417 Intron 4.972 (2.387--10.357) 3.35EÀ06 rs2745761 rs6140592 1 1 rs6140593 20 8231850 Intron 4.972 (2.387--10.357) 3.34926EÀ06 rs2745761 rs6140593 1 1 rs6140594 20 8232033 Intron 4.972 (2.387--10.357) 3.34926EÀ06 rs2745761 rs6140594 1 1 rs6140595 20 8234291 Intron 4.972 (2.387--10.357) 3.34926EÀ06 rs2745761 rs6140595 1 1 rs16994609 20 8236976 Intron 4.972 (2.387--10.357) 3.35EÀ06 rs2745761 rs16994609 1 1 rs6039133 20 8237823 Intron 4.972 (2.387--10.357) 3.34926EÀ06 rs2745761 rs6039133 1 1 rs2890430 8 69565334 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs2890430 1 0.952 rs4489312 8 69565808 Intron 4.179 (2.186--7.988) 4.29E-06 rs1517114 rs4489312 1 0.952 rs1867634 8 69570970 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs1867634 1 0.952 rs4236960 8 69571065 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs4236960 1 0.952 rs4588839 8 69571337 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs4588839 1 0.952 rs4590428 8 69572083 Intron 4.179 (2.186--7.988) 4.29EÀ06 rs1517114 rs4590428 1 0.952 rs4590429 8 69572292 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs4590429 1 0.952 rs4527853 8 69573888 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs4527853 1 0.952 rs4737262 8 69574270 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs4737262 1 0.952 rs2380465 8 69575664 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs2380465 1 0.952 rs2380467 8 69575900 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs2380467 1 0.952 rs10091735 8 69576256 Intron 4.179 (2.186--7.988) 4.29453EÀ06 rs1517114 rs10091735 1 0.952 rs4596639 8 69572345 Intron 4.113 (2.149--7.872) 5.93533EÀ06 rs1517114 rs4596639 1 0.952 rs2745761 20 8225943 Intron 4.781 (2.295--9.96) 6.40316EÀ06 rs2745761 rs2745761 1 1 rs1517114 8 69551771 Intron 4.125 (2.126--8.006) 8.64575EÀ06 rs1517114 rs1517114 1 1 rs17383779 8 69549434 Intron 3.981 (2.065--7.674) 1.25EÀ05 rs1517114 rs17383779 1 1 rs16934643 8 69549514 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs16934643 1 1 rs8180911 8 69549871 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs8180911 1 1 rs1850386 8 69552657 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs1850386 1 1 rs10089645 8 69554699 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs10089645 1 1 rs1400550 8 69557536 Intron 3.981 (2.065--7.674) 1.25EÀ05 rs1517114 rs1400550 1 1 rs10105022 8 69558175 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs10105022 1 1 rs2380461 8 69560751 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs2380461 1 1 rs4466385 8 69561364 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs4466385 1 1 rs4481578 8 69561534 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs4481578 1 1 rs4602871 8 69561723 Intron 3.981 (2.065--7.674) 1.24503EÀ05 rs1517114 rs4602871 1 1 rs6982746 8 69553999 Intron 3.879 (2.012--7.48) 1.90156EÀ05 rs1517114 rs6982746 1 1 rs4330678 8 69571996 Intron 4.878 (2.196--10.835) 2.47888EÀ05 rs1517114 rs4330678 1 0.976 rs1727747 19 49739960 30 UTR 3.888 (1.99--7.596) 2.65439EÀ05 rs1661167 rs1727747 1 1 rs11670955 19 49714537 Intron 13.158 (2.839--60.99) 2.81795EÀ05 rs1661167 rs11670955 0.678 0.064 rs8111044 19 49741978 30 UTR 3.696 (1.896--7.205) 5.30781EÀ05 rs1661167 rs8111044 1 1 rs1661167 19 49743908 30 UTR 4.498 (2.047--9.885) 5.71687EÀ05 rs1661167 rs1661167 1 1 rs1727748 19 49740163 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1727748 1 1 rs1661176 19 49740219 30 UTR 3.63 (1.868--7.052) 6.27EÀ05 rs1661167 rs1661176 1 1 rs1661175 19 49740469 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1661175 1 1 rs1661174 19 49740498 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1661174 1 1 rs1661173 19 49740619 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1661173 1 1 rs1727751 19 49740943 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1727751 1 1 rs1626475 19 49742330 30 UTR 3.63 (1.868--7.052) 6.26512EÀ05 rs1661167 rs1626475 1 1 rs2662996 20 8245604 Intron 3.6 (1.828--7.088) 9.68EÀ05 rs2745761 rs2662996 0.972 0.609

Grade 4 neutropenia rs9866965 3 73707335 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9866965 1 1 rs9811288 3 73707427 Intron 3.837 (2.048--7.189) 9.65106E-06 rs11128347 rs9811288 1 1 rs9843996 3 73708244 Intron 3.837 (2.048--7.189) 9.65EÀ06 rs11128347 rs9843996 1 1 rs9850092 3 73709717 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9850092 1 1 rs9816865 3 73710931 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9816865 1 1 rs12629627 3 73712238 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs12629627 1 1 rs6771982 3 73716017 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs6771982 1 1 rs9830956 3 73718710 Intron 3.837 (2.048--7.189) 9.65EÀ06 rs11128347 rs9830956 1 1 rs4677310 3 73725380 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs4677310 1 1 rs4440067 3 73729398 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs4440067 1 1 rs9858993 3 73731771 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9858993 1 1 rs9879918 3 73735523 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9879918 1 1 rs9866634 3 73737242 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs9866634 1 1 rs4676938 3 73741067 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs4676938 1 1 rs4676939 3 73741175 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs4676939 1 1 rs12630477 3 73749909 Intron 3.837 (2.048--7.189) 9.65106EÀ06 rs11128347 rs12630477 1 1 rs11128347 3 73702251 Intron 3.714 (1.982--6.962) 1.68EÀ05 rs11128347 rs11128347 1 1 rs7779029 7 80370048 Intron 3.114 (1.797--5.397) 2.82905EÀ05 rs11979430 rs7779029 1 0.966

The Pharmacogenomics Journal (2013), 417 -- 422 & 2013 Macmillan Publishers Limited GWAS for irinotecan toxicities J-Y Han et al 421 Table 3 (Continued )

LD information SNP ID Ch Position Region OR (95% CI) P-value Sig SNP ID SNP ID D0 r2

rs10487876 7 80312439 Intron 3.086 (1.784--5.338) 3.08244EÀ05 rs11979430 rs10487876 0.966 0.917 rs2189959 7 80312878 Intron 3.086 (1.784--5.338) 3.08244EÀ05 rs11979430 rs2189959 0.966 0.917 rs17154550 7 80319318 Intron 3.086 (1.784--5.338) 3.08EÀ05 rs11979430 rs17154550 0.966 0.917 rs17154553 7 80323277 Intron 3.086 (1.784--5.338) 3.08244EÀ05 rs11979430 rs17154553 0.966 0.917 rs7795137 7 80328382 Intron 3.086 (1.784--5.338) 3.08EÀ05 rs11979430 rs7795137 0.966 0.917 rs7784762 7 80334149 Intron 3.086 (1.784--5.338) 3.08244EÀ05 rs11979430 rs7784762 0.966 0.917 rs7793465 7 80370531 Intron 3.054 (1.775--5.254) 3.14869EÀ05 rs11979430 rs7793465 1 0.952 rs11979430 7 80347192 Intron 3.12 (1.783--5.46) 3.71EÀ05 rs11979430 rs11979430 1 1 Abbreviations: Ch, chromosome; CI, confidence interval; LD, linkage disequilibrium; OR, odds ratio; SNP, single-nucleotide polymorphism; UTR, untranslated region.

Table 4. Previously described genetic associations

Toxicity SNP Ch Position Gene Allelea Stage MAF case MAF control P-value OR (95% CI)

G3D rs3740066 10 101594197 ABCC2 T/C Replication 0.462 0.242 0.016 2.7 (1.2--6.1) rs2231137 4 89061114 ABCG2 A/G Replication 0.269 0.246 0.795 1.1 (0.5--2.8) rs3832043 2 234580463 UGT1A9 T10/T9 Replication 0.583 0.648 0.650 1.5 (0.3--7.2) G4N rs4148323 2 234333883 UGT1A1 A/G GWAS 0.283 0.190 0.013 7.3 (1.2--42.7) Replication 0.250 0.141 0.0004 NA Combined 0.267 0.159 2.2 Â 10À5 5.6 (3.0--80.6) rs4149056 12 21222816 SLCO1B1 C/T Replication 0.275 0.138 0.028 2.4 (1.1--5.2) Abbreviations: Ch, chromosome; CI, confidence interval; GWAS, genome-wide association study; G3D, grade 3 diarrhea; G4N, grade 4 neutropenia; MAF, minor allele frequency; NA, not assessable; OR, odds ratio; SNP, single-nucleotide polymorphism. aThe former allele represents the minor allele. ORs for minor allele (allele model except UGT1A1, recessive model). All P-values listed are two sided.

Nine additional SNPs in intron 2 of SEMA3C achieved P-values Moreover, a higher baseline total bilirubin level was correlated of o1 Â 10À4. with a significantly lower absolute neutrophil count nadir during Among the candidate genes identified in our study for G3D, irinotecan chemotherapy.12 Thus, lower bilirubin glucuronidation PLCB1 has been implicated in gastrointestinal mucositis after activity contributes to a higher SN-38 AUC, which results in severe irinotecan therapy in an animal study. neutropenia. A microarray investigation found more than 500 genes, The PDZRN3 gene is expressed in a variety of organs and tissues, including PLCB1, to be significantly altered by irinotecan in the including the heart, skeletal muscle and liver. In skeletal muscle, rat gut.19 PLCB1 has also been implicated in severe ulcerative the expression of PDZRN3 is developmentally regulated. This colitis.20 The encoded by PLCB1 catalyzes the formation of protein is essential for differentiation of myoblasts into myo- inositol 1,4,5-trisphosphate and diacylglycerol from phosphatidy- tubes.22 Depletion of PDZRN3 by RNA interference revealed that linositol 4,5-bisphosphate. This reaction uses calcium as a cofactor PDZRN3 inhibits the upregulation of alkaline phosphatase activity and plays an important role in the intracellular transduction of in vitro.23 Furthermore, PDZRN3 is frequently downregulated or many extracellular signals. PLCB1 is activated by two G-protein a deleted in various cancers, including biliary tract cancer.24 These subunits, a-q and a-11. Two transcript variants encoding different findings suggest that the genetic variants of PDZRN3 may affect isoforms have been found for PLCB1 (provided by RefSeq. http:// bilirubin glucuronidation, which may lead to a decrease in SN-38 www.ncbi.nlm.nih.gov/gene/23236). glucuronidation and ultimately the development of severe C8orf34 encodes a protein that is related to cyclic AMP (cAMP)-- neutropenia. dependent protein kinase regulators. The cAMP-dependent SEMA3C is a secreted protein that may be involved in cell protein kinase pathway mediates the intestinal secretary process, survival mechanisms. It is expressed intensely in the heart, skeletal which leads to severe diarrhea.21 This relationship suggests that muscle, colon, small intestine, ovary, testis and prostate. It shows genetic variants of C8orf34 may be associated with treatment- faint expression ubiquitously among other organs, including the related severe diarrhea. brain. It binds with high affinity to the receptor NRP2 (http:// There is accumulating evidence that plasma exposure to SN-38 www.phosphosite.org/proteinAction.do?id ¼ 2380717&showAllSi- is associated with an increased risk of irinotecan-related severe tes ¼ true). A genome-wide association meta-analysis revealed toxicity, primarily neutropenia. The genetic variants of UGT1A1, that variants in SEMA3C are associated with total serum bilirubin which is responsible for bilirubin glucuronidation, have been level, as are UGT1A1 and SLCO1B1.25 Because the plasma associated with reduced glucuronidation of SN-38 and increased concentration of SN-38 is strongly correlated with total serum severe neutropenia in patients treated with irinotecan.2--4 Pre- bilirubin level, SEMA3C variants may lead to severe neutropenia viously, we found that the absolute neutrophil count was after irinotecan treatment. negatively correlated with the SN-38 AUC and that patients Recently, other group reported several candidate SNPs that might who experienced G4N had a significantly higher SN-38 AUC.9,12 help to predict the outcome of platinum-based chemotherapy

& 2013 Macmillan Publishers Limited The Pharmacogenomics Journal (2013), 417 -- 422 GWAS for irinotecan toxicities J-Y Han et al 422 26 in NSCLC. However, the efficacy of chemotherapy or patients’ 7 Schellens JH, Maliepaard M, Scheper RJ, Scheffer GL, Jonker JW, Smit JW et al. survival can be affected by many factors such as the drug Transport of topoisomerase I inhibitors by the breast cancer resistance protein. combination and the subsequent salvage chemotherapies given Potential clinical implications. Ann NY Acad Sci 2000; 922: 188--194. to patients. Thus, we analyzed the effects of GWAS on toxicity only 8 Nozawa T, Minami H, Sugiura S, Tsuji A, Tamai I. Role of organic anion transporter in this study. Despite the small number of patients, we collected OATP1B1 (OATP-C) in hepatic uptake of irinotecan and its active metabolite, the toxicity data prospectively throughout the clinical trials 7-ethyl-10-hydroxycamptothecin: in vitro evidence and effect of single nucleotide using the same toxicity criteria. However, this study has some polymorphisms. Drug Metab Dispos 2005; 33: 434--439. considerable limitations. First, patients received different doses 9 Han JY, Lim HS, Shin ES, Yoo YK, Park YH, Lee JE et al. Comprehensive analysis of UGT1A polymorphisms predictive for pharmacokinetics and treatment outcome of irinotecan. However, previously we had demonstrated that in patients with non-small-cell lung cancer treated with irinotecan and cisplatin. the incidences of G3 and G4 were not significantly different J Clin Oncol 2006; 24: 2237--2244. 10 according to the irinotecan dose schedules. Second, different 10 Han JY, Lim HS, Yoo YK, Shin ES, Park YH, Lee SY et al. Associations of ABCB1, drugs were used in combination with irinotecan. As cisplatin- ABCC2,andABCG2 polymorphisms with irinotecan-pharmacokinetics and clinical containing regimens are more likely to develop neutropenia outcome in patients with advanced non-small cell lung cancer. Cancer 2007; 110: compared with non-platinum regimens, patients in the GWAS 138--147. stage showed a trend toward higher incidence of G4N than 11 Han JY, Lim HS, Lee SY, Kim HT, Lee JS. Influence of the organic anion transporting those in the replication stage (24% vs 14%, P ¼ 0.06). This finding polypeptide 1B1 (OATP1B1) polymorphisms on irinotecan-pharmacokinetics suggests that cisplatin combination in the GWAS stage may and clinical outcome of patients with advanced non-small cell lung cancer. affect discovering SNPs for G4N. However, fewer patients received Lung Cancer 2008; 59: 69--75. 12 Han JY, Lim HS, Park YH, Lee SY, Lee JS. Integrated pharmacogenetic prediction of cisplatin in the replication stage. Furthermore, there was no irinotecan pharmacokinetics and toxicity in patients with advanced non-small cell difference in G3D between two groups (11% vs 9%, P ¼ 0.604). lung cancer. Lung Cancer 2009; 63: 115--120. These findings suggest that the replication samples are more likely 13 Savas S, Liu G. Genetic variations as cancer prognostic markers: review and to represent irinotecan-related toxicities rather than cisplatin. update. Hum Mutat 2009; 30: 1369--1377. Thus, validation though the replication stage may contribute to 14 Han JY, Lee DH, Song JE, Lee SY, Kim HY, Kim HT et al. Randomized Phase II study select SNPs associated with irinotecan-related G4N. Third, the of two opposite administration sequences of irinotecan and cisplatin in patients power to detect genome-wide significance was low because of with advanced nonsmall cell lung carcinoma. Cancer 2006; 106: 873--880. the small sample size. Finally, the biologic mechanisms underlying 15 Han JY, Lim HS, Lee DH, Ju SY, Lee SY, Kim HY et al. Randomized phase 2 study of these associations are still unknown. irinotecan plus cisplatin versus gemcitabine plus vinorelbine as first-line chemotherapy with second-line crossover in patients with advanced nonsmall In summary, this GWAS in Korean patients with NSCLC cell lung cancer. Cancer 2008; 113: 388--395. identified potential candidate gene regions that are associated 16 So¨derlund-Strand A, Dillner J, Carlson J. High-throughput genotyping of with irinotecan-related severe diarrhea and neutropenia. This is oncogenic human papilloma viruses with MALDI-TOF mass spectrometry. Clin the first study to explore the genetic variants in irinotecan-related Chem 2008; 54: 86--92. severe toxicities on a genome-wide scale. Follow-up functional 17 Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and validation and replication studies in the larger cohort are needed haplotype maps. Bioinformatics 2005; 21: 263--265. to confirm our findings. 18 Marchini J, Howie B, Myers S, McVean G, Donnelly P. A new multipoint method for genome-wide association studies via imputation of genotypes. Nat Genet 2007; 39: 906--913. CONFLICT OF INTEREST 19 Bowen JM, Gibson RJ, Tsykin A, Stringer AM, Logan RM, Keefe DM. analysis of multiple gastrointestinal regions reveals activation of The authors declare no conflict of interest. common cell regulatory pathways following cytotoxic chemotherapy. Int J Cancer 2007; 121: 1847--1856. 20 Haritunians T, Taylor KD, Targan SR, Dubinsky M, Ippoliti A, Kwon S et al. Genetic ACKNOWLEDGEMENTS predictors of medically refractory ulcerative colitis. Inflamm Bowel Dis 2010; 16: This study was supported by a grant from the National Cancer Center (no. 0810130). 1830--1840. 21 Blumenstein I, Keseru¨ B, Wolter F, Stein J. The chemopreventive agent resveratrol stimulates cyclic AMP-dependent chloride secretion in vitro. Clin Cancer Res 2005; REFERENCES 11: 5651--5656. 1 Van Cutsem E, Douillard JY, Kohne CH. Toxicity of irinotecan in patients with 22 Ko JA, Kimura Y, Matsuura K, Yamamoto H, Gondo T, Inui M. PDZRN3 (LNX3, colorectal cancer. N Engl J Med 2001; 345: 1351--1352. SEMCAP3) is required for the differentiation of C2C12 myoblasts into myotubes. 2 Ando Y, Saka H, Ando M, Sawa T, Muro K, Ueoka H et al. Polymorphisms of UDP- J Cell Sci 2006; 119: 5106--5113. glucuronosyltransferase gene and irinotecan toxicity: a pharmacogenetic analysis. 23 Honda T, Yamamoto H, Ishii A, Inui M. PDZRN3 negatively regulates BMP-2- Cancer Res 2000; 60: 6921--6926. induced osteoblast differentiation through inhibition of Wnt signaling. Mol Biol 3 Iyer L, Das S, Janisch L, Wen M, Ram´ırez J, Karrison T et al. UGT1A1*28 Cell 2010; 21: 3269--3277. polymorphism as a determinant of irinotecan disposition and toxicity. Pharmaco- 24 Miller G, Socci ND, Dhall D, D0Angelica M, DeMatteo RP, Allen PJ et al. genomics J 2002; 2: 43--47. Genome wide analysis and clinical correlation of chromosomal and transcrip- 4 Innocenti F, Undevia SD, Iyer L, Chen PX, Das S, Kocherginsky M et al. Genetic tional mutations in cancers of the biliary tract. J Exp Clin Cancer Res 2009; variants in the UDP-glucuronosyltransferase 1A1 gene predict the risk of severe 28:62. neutropenia of irinotecan. J Clin Oncol 2004; 22: 1382--1388. 25 Johnson AD, Kavousi M, Smith AV, Chen MH, Dehghan A, Aspelund T et al. 5 de Jong FA, Scott-Horton TJ, Kroetz DL, McLeod HL, Friberg LE, Mathijssen RH Genome-wide association meta-analysis for total serum bilirubin levels. Hum Mol et al. Irinotecan-induced diarrhea: functional significance of the polymorphic Genet 2009; 18: 2700--2710. ABCC2 transporter protein. Clin Pharmacol Ther 2007; 81: 42--49. 26 Tan XL, Moyer AM, Fridley BL, Schaid DJ, Niu N, Batzler AJ et al. Genetic 6 Iyer L, Ram´ırez J, Shepard DR, Bingham CM, Hossfeld DK, Ratain MJ et al. Biliary variation predicting cisplatin cytotoxicity associated with overall survival in lung transport of irinotecan and metabolites in normal and P-glycoprotein-deficient cancer patients receiving platinum-based chemotherapy. Clin Cancer Res 2011; mice. Cancer Chemother Pharmacol 2002; 49: 336--341. 17: 5801--5811.

Supplementary Information accompanies the paper on the The Pharmacogenomics Journal website (http://www.nature.com/tpj)

The Pharmacogenomics Journal (2013), 417 -- 422 & 2013 Macmillan Publishers Limited