Clinical Cancer Research Prevention Award Frontiers in Basic Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Clinical Cancer Research 14, 1788, March 15, 2008. doi: 10.1158/1078-0432.CCR-07-1472
© 2008 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hoskins, J. M.
Right arrow Articles by Baiget, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hoskins, J. M.
Right arrow Articles by Baiget, M.

Cancer Therapy: Clinical

Irinotecan Pharmacogenetics: Influence of Pharmacodynamic Genes

Janelle M. Hoskins1, Eugenio Marcuello3, Albert Altes5, Sharon Marsh1, Taylor Maxwell2, Derek J. Van Booven1, Laia Paré4, Robert Culverhouse1, Howard L. McLeod1 and Montserrat Baiget4

Authors' Affiliations: Departments of 1 Internal Medicine and 2 Biology, Washington University School of Medicine, St. Louis, Missouri; 3 Departments of Medical Oncology and 4 Genetics, Hospital de la Santa Creu i Sant Pau, CIBERER, Barcelona, Spain; and 5 Department of Hematology, Hospital Esperit Sant, Sta. Coloma de Gramenet, Spain

Requests for reprints: Howard L. McLeod, UNC Institute for Pharmacogenomics and Individualized Therapy, University of North Carolina, Chapel Hill, Campus Box 7360, 3202 Kerr Hall, Chapel Hill, NC 27599-7360. Phone: 919-966-0512; Fax: 919-962-0644; E-mail: hmcleod{at}unc.edu.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose: Irinotecan is an important drug for the treatment of solid tumors. Although genes involved in irinotecan pharmacokinetics have been shown to influence toxicity, there are no data on pharmacodynamic genes. CDC45L, NFKB1, PARP1, TDP1, and XRCC1 have been shown to influence the cytotoxic action of camptothecins, including irinotecan. Polymorphisms in the drug target of camptothecins, topoisomerase I (TOP1), and downstream effectors may influence patient outcomes to irinotecan therapy. We undertook a retrospective candidate gene haplotype association study to investigate this hypothesis.

Experimental Design: Haplotype compositions of six candidate genes were constructed in European (n = 93), East Asian (n = 94), and West African (n = 95) populations. Haplotype-tagging single nucleotide polymorphisms (htSNP) were selected based on genealogic relationships between haplotypes. DNA samples from 107 European, advanced colorectal cancer patients treated with irinotecan-based regimens were genotyped for htSNPs as well as three coding region SNPs. Associations between genetic variants and toxicity (grade 3/4 diarrhea and neutropenia) or efficacy (objective response) were assessed.

Results: TOP1 and TDP1 htSNPs were related to grade 3/4 neutropenia (P = 0.04) and response (P = 0.04), respectively. Patients homozygous for an XRCC1 haplotype (GGCC-G) were more likely to show an objective response to therapy than other patients (83% versus 30%; P = 0.02). This effect was also seen in a multivariate analysis (odds ratio, 11.9; P = 0.04). No genetic variants were associated with diarrhea.

Conclusions: This is the first comprehensive pharmacogenetic investigation of irinotecan pharmacodynamic factors, and our findings suggest that genetic variation in the pharmacodynamic genes may influence the efficacy of irinotecan-containing therapies in advanced colorectal cancer patients.

Irinotecan (CPT-11) is used for second-line treatment of metastatic colorectal cancer and as salvage therapy in 5-fluorouracil refractory disease. The major dose-limiting toxicities of CPT-11–based therapies are delayed diarrhea and severe or fatal myelosuppression (1). CPT-11 itself is a relatively weak inhibitor of topoisomerase I (Topo I), and its clinical activity is dependent on in vivo conversion to the more potent metabolite, SN-38, a reaction catalyzed by carboxylesterase enzymes (CES1 and CES2; ref. 2).

The pharmacokinetics of CPT-11 and its active metabolite, SN-38, are determined by numerous drug transporters and metabolizing enzymes (3). Many pharmacogenetic studies have investigated the influence of genetic variation in these pathways on patient-to-patient variation in CPT-11 pharmacokinetics and toxicity (4, 5). The strongest and most consistent association has been shown for a tandem repeat TA polymorphism [–53(TA)6>7TAA] in the TATA box of the promoter region of the UDP-glucuronosyltransferases 1A1 gene (UGT1A1; refs. 4, 5). The deactivation to SN-38 glucuronide is principally catalyzed by UGT1A1 (6). Patients homozygous for the seven dinucleotide repeat –53(TA)7 (UGT1A1*28) are exposed to more SN-38 and have a higher incidence of grade 3/4 neutropenia than patients with the common allele (7, 8). Recent evidence indicates the UGT1A1*28/*28 genotype explains a considerable portion of the risk of severe neutropenia at moderate to high doses of irinotecan but not at low doses, suggesting that additional genetic or nongenetic factors explain the risk, especially when low doses of irinotecan are administered (9).

Proteins that mediate the cellular response to CPT-11 have been largely ignored in previous pharmacogenetic investigations. There is substantially less knowledge about the CPT-11 pharmacodynamics, including DNA damage repair or cell death pathways, following the formation of camptothecin-Topo I-DNA complexes. Genetic variation in the drug target of SN-38, Topo I (gene symbol TOP1), and cellular downstream effectors that lead to DNA repair or cell death are potential sources of patient-to-patient variation in CPT-11–induced toxicity and clinical response (10). This variation might improve prediction of patient outcomes to CPT-11–based therapies. Numerous proteins are involved in mediating tumoral resistance to camptothecins (10, 11), and the Pharmacogenomics and Pharmacogenetics Knowledge Base Web site (PharmGKB)6 has proposed that a handful of these proteins, including the drug target, Topo I, and cell cycle division 45–like protein (CDC45L), nuclear factor-{kappa}B (p50 subunit; NFKB1), poly(ADP-ribose) polymerase I (PARP1), tyrosyl DNA phosphodiesterase (TDP1), and X-ray cross-complementation factor (XRCC1), are involved in the pharmacodynamic pathway of irinotecan. Findings from studies using yeast or mammalian cancer cells suggest that these proteins influence the cytotoxic action of camptothecins (12–16).

Given the important roles these proteins play in determining cellular response following exposure to camptothecins, we hypothesized that genetic variation in the PharmGKB irinotecan pharmacodynamic genes, TOP1, CDC45L, NFKB1, PARP1, TDP1, and XRCC1, influences patient outcomes to CPT-11 therapy. To evaluate this hypothesis, we undertook an exploratory retrospective candidate gene haplotype association test in a cohort of European, advanced colorectal cancer patients treated with CPT-11–based regimens. As the number of events of severe neutropenia (grade 3/4) was low, the haplotype-neutropenia analysis should be considered preliminary. The influence of UGT1A1*28 on treatment outcomes has previously been investigated in this patient cohort (17).


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Patients. Study protocols were approved by the Institutional Review Boards of Hospital de la Santa Cruz y San Pablo (Barcelona, Spain) and Washington University School of Medicine (St. Louis, MO). One hundred and seven European patients with advanced colorectal cancer treated with CPT-11–containing regimens at Hospital de la Santa Creu i Sant Pau were included in the study. The study was approved by Washington University Medical Center Human Studies Committee (06-0262) and Hospital de la Santa Cruz y San Pablo Comité Etico de Investigación Clínica (120705). Patients gave written informed consent before being enrolled in the study. The initial objective of the study was to evaluate the influence of UGT1A1*28 genotype on CPT-11 treatment outcomes, and the findings were published elsewhere for 95 of the 107 patients (17). Patient pathologic characteristics are outlined in Table 1 . Patients were treated with chemotherapy regimens (Table 1) and underwent chemotherapy cycles until severe toxicity or disease progression appeared.


View this table:
[in this window]
[in a new window]

  		Table 1. Baseline characteristics of 107 advanced colorectal cancer patients, chemotherapy, and response to treatment

 
DNA was collected for genetic analysis from all 107 patients. Presence and grade of neutropenia and diarrhea were recorded for all 107 patients in accordance with the criteria of WHO. Response to treatment was analyzed in 89 of the 107 patients. Complete remission was defined as the disappearance of tumor masses and disease-related symptoms as well as the normalization of the initially abnormal tests and/or biopsies lasting for at least 1 mo. Partial remission was considered when measurable lesions decreased by at least 50%. Clinical response was assumed when a complete or partial remission was obtained. Patients without criteria of clinical response but without progression were considered patients with stable disease; disease progression during or after treatment was also considered.

Samples and DNA extraction. DNA was isolated from the whole blood of 107 patients with advanced colorectal cancer receiving CPT-11–based regimens, as previously described (16). DNA from 93 European Americans and 94 Han Chinese was obtained from Coriell.7 DNA was isolated from 95 Ghanaian healthy blood donors as previously described (18).

Polymorphism selection. Human PARP1, CDC45L, NFKB1, TDP1, TOP1, and XRCC1 genomic sequences were annotated to confirm splice site junctions using the University of California at Santa Cruz Golden Path Human Genome Browser.8 Once the gene was annotated, Polymorphism Mining and Annotation Program (PolyMAPr) was used to map the polymorphisms identified by gene resequencing and reported in public databases (dbSNP,9 CGAP-GAI,10 and JSNP11) to the gene sequences (19, 20).

Alleles, single nucleotide polymorphisms (SNP), and insertion/deletions (indels) were preferentially selected for construction of haplotype identities in the European, East Asian, and West African samples if they were in coding regions and in regulatory regions affecting gene expression if there was epidemiologic evidence to suggest that they altered the risk of developing a disease or they were predicted to affect function by PolyPhen (20). For PARP1, CDC45L, NFKB1, and XRCC1, alleles that distinguished between common haplotypes (frequency of ≥5%) were selected from data published on 90 unrelated individuals, representative of a U.S. population (see the National Institute of Environmental Health Sciences Web site12 for a description of the population; ref. 21). NFKB1 –794delATTG (previously denoted –94 ins/del ATTG) and XRCC1 –77T>C, c.580C>T (R194W), and c.839G>A (R280H) were selected from the literature (22–24). At the time of study commencement, genetic variation data for TOP1 and TDP1 were not available on the National Institute of Environmental Health Sciences Web site. Alleles with estimated frequencies of ≥5% in at least one ethnic group by resequencing were selected for haplotype construction of these two genes (19).

Genotyping. Genomic DNA samples from 93 healthy, unrelated European Americans, 94 Han Chinese, and 95 Ghanaians were typed for 41 SNPs and 2 indels (see Table 2 for a list of alleles, PCR primers, and conditions available on request) across the candidate genes to construct haplotype identities for these genes. Haplotype-tagging SNPs (htSNP) were selected as described below. Putative causal SNPs in PARP1 and XRCC1 were selected from the literature (23–27). Genomic DNA samples of 107 advanced colorectal cancer patients were genotyped for 10 htSNPs across the candidate genes and for 4 putative causal SNPs (see Table 3 ).


View this table:
[in this window]
[in a new window]

  		Table 2. Minor allelic frequencies determined by pyrosequencing the candidate genes

 

View this table:
[in this window]
[in a new window]

  		Table 3. Allele and genotype frequencies for htSNPs and possible causative SNPs of the candidate genes in 107 advanced colorectal cancer patients treated with CPT-11–based regimens

 
With the exception of XRCC1 –77T>C, genotyping was done using PCR and pyrosequencing as previously described (28). Primer sets for the htSNPs and putative causal SNPs are presented in Supplementary Table S1. For XRCC1 –77T>C, DNA was amplified using Thermo-Start PCR Master Mix (ABgene13), an annealing temperature of 55°C, and primers presented in Supplementary Table S1, yielding a 139-bp product. PCR amplicons were incubated with BsrBI at 37°C for 2 h. Fragments were visualized using 4% low-melt agarose and genotypes were scored based on the following fragment patterns: C/C = 139 bp; C/T = 139 + 79 + 60 bp; and T/T = 79 + 60 bp.

Selection of haplotype markers. Haplotype networks were estimated using TCS software package to examine the genealogic relationships between haplotypes (29, 30). htSNPs were selected from the haplotype networks and were polymorphisms that distinguished two or three major haplotypes in the European sample.

Statistics. Hardy-Weinberg equilibrium was assessed by HWSIM,14 a QuickBasic program, which uses a Monte Carlo permutation procedure. The {chi}2 test was used to compare the observed genotype frequencies with those expected under Hardy-Weinberg equilibrium. Polymorphism and Haplotype Analysis Suite (PHASE version 0.9) was used to estimate pairwise linkage disequilibrium (D') and haplotypes and to assign diplotypes (31). Analyses of diarrhea and neutropenia used a dichotomized outcome (grade 3/4 versus grade 0/1/2). Possible associations between binary outcomes (toxicities or response to treatment) and genetic factors (genotypes or diplotypes) were tested using {chi}2 tests or Fisher exact tests depending on the data distribution. In a post hoc calculation of power, based on the known sample size, number of clinical response events, and a significance threshold of P < 0.05, we had 80% power to detect an odds ratio of ≥3.6 for variants with a minor allele frequency of 0.33 (the mean minor allele frequency in this study).

Initially, univariate tests were used to assess associations between the outcomes (neutropenia and objective response to treatment) and a variety of genetic, demographic, and clinical factors, including UGT1A1*28 genotype (TA6/TA6 versus TA6/TA7 versus TA7/TA7), age (<65 versus ≥65 y), sex (male versus female), line of chemotherapy (first versus ≥second), performance status (treated as a class variable: 0 versus 1 versus 2), previous treatment (surgery versus radiotherapy), total bilirubin (≤1.0 versus >1.0 mg/dL), and chemotherapeutic regimen (CPT-11 alone versus CPT-11 + raltitrexed versus CPT-11 + 5-fluorouracil versus CPT-11 + 5-fluorouracil/leucovorin). Factors that seemed to be associated with outcomes (as determined by univariate tests; P < 0.10) were included in a stepwise logistic regression analysis in an attempt to identify independent predictors of grade 3/4 neutropenia and objective response. No P value adjustments were made for multiple hypotheses testing.


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Analysis of candidate gene sequence variation
The chromosomal locations, number of exons, and sizes of the candidate genes are presented in Supplementary Table S2. On average, 248 unique polymorphisms reported in the public databases dbSNP, CGAG-GAI, and JSNP and identified by gene sequencing were mapped to each gene using PolyMAPr. At the time of analyses, sequencing identified 1 to 18 novel SNPs per gene (19). We determined the allelic frequencies of 41 SNPs and 2 indels by pyrosequencing germ-line DNA from European Americans, East Asians (Han Chinese), and West Africans (Ghanaians; Table 2). Genotypes were tested for deviations from Hardy-Weinberg equilibrium. With the exception of TDP1 IVS12+79 in East Asians (P = 0.03, two tailed), genotype frequencies in subgroups agreed with the Hardy-Weinberg equilibrium.

Estimation of haplotypes of the candidate genes
Genotype data for 43 alleles among the candidate genes from unrelated, healthy individuals of European (n = 93), East Asian (n = 94), and West African descent (n = 95) were used to estimate haplotype identities using PHASE (Supplementary Tables S3-S8). For each gene, some individuals were homozygous for all loci and others were heterozygous at a single site, which provided unambiguous information about the specific combinations of alleles. Other haplotypes were inferred by PHASE; some of these were observed in another subgroup, which provided evidence that they were true haplotypes. Haplotypes with a frequency of ≥5% were considered common and those with frequencies of <5% were considered rare. A summary of the number of common haplotypes observed in each ethnic group is presented in Supplementary Table S2. Haplotype structures and frequencies for each gene in the three populations are listed in Supplementary Tables S3 to S8.

Selection of haplotype markers
Cladograms of haplotype identities were generated using haplotypes estimated in all ethnic groups combined to gain an understanding of the relationships between the haplotypes (see Supplementary Figs. S1-S6 for cladograms of the candidate genes). Alleles that distinguished common haplotypes in the European sample (htSNPs) were selected from the cladograms for pharmacogenetic assessment (Table 3). Two htSNPs (included SNPs and indels) that divided PARP1, CDC45L, NFKB1, and XRCC1 haplotype compositions into three haplotype groups were selected. Owing to the limited haplotypic diversity of TOP1 and TDP1 in Europeans, a marker that distinguished two major haplotype groups was chosen for each gene.

Genotypes/haplotypes of the advanced colorectal cancer patients treated with CPT-11–based regimens
Allele and genotype frequencies of the 10 htSNPs (8 SNPs and 2 indels) and 4 putative causative SNPs in the patient cohort are presented in Table 3. Observed genotype frequencies did not differ significantly from those expected from the Hardy-Weinberg equilibrium. Haplotypes were estimated for CDC45L, NFKB1, PARP1, and XRCC1. Three haplotype identities for CDC45L T-C (IVS3+87T>C-IVS8-24C>T), C-C, and C-T, were estimated. The frequencies of the haplotypes were 0.51, 0.25, and 0.24, respectively. For NFKB1, haplotypes ATTG-G (–794delATTG-IVS5+1995A>G), del-A, and ATTG-A with frequencies of 0.37, 0.34, and 0.29, respectively, were estimated. For PARP1, haplotypes of T-C (c.852T>C-IVS19-297C>T), C-C, and T-T were estimated with frequencies of 0.50, 0.35, and 0.15, respectively. Four haplotypes were estimated for XRCC1, del-G (–1449delGGCC-c.1196G>A), GGCC-A, GGCC-G, and del-A, with frequencies of 0.43, 0.36, 0.20, and 0.01, respectively.

Analysis of toxicity to CPT-11–based therapies
The incidences of grade 3/4 toxicities in the patient cohort are presented in Table 1. The incidence of diarrhea was not associated with any candidate gene genotypes or diplotypes. Associations between candidate gene htSNP genotypes and diplotypes, and neutropenia are presented in Table 4 . A gene-dose effect of TOP1 IVS4+61 genotype on the incidence of grade 3/4 neutropenia was observed, with G/G patients experiencing the lowest (6 of 73 patients, 8%) and A/A the highest incidence (2 of 4 patients, 50%; G/A, 6 of 30, 20%; P = 0.035). PARP1, CDC45L, NFKB1, TDP1, and XRCC1 genotypes and diplotypes were not significantly associated with severe neutropenia, although a trend toward PARP1 C-C/C-C (c.852-IVS19-297) patients experiencing a 2.6-fold higher incidence of neutropenia (33%, 4 of 12 patients) than patients with other PARP1 diplotypes (13%, 12 of 94 patients; P = 0.08) was observed (Fig. 1A ). None of the seven XRCC1 GGCC-G/GGCC-G (–1449-c.1196) patients experienced neutropenia compared with 16 of 100 patients with other diplotypes who did (0% versus 16%; P > 0.10). The study, however, lacked adequate power to show a significant difference.


View this table:
[in this window]
[in a new window]

  		Table 4. Univariate analyses of the relationships between responses to CPT-11–based therapies and candidate gene alleles in patients with advanced colorectal cancer

 

Figure 1
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Objective response and grade 3/4 neutropenia rates in a cohort of 107 patients with advanced colorectal cancer treated with CPT-11–based regimens by candidate gene genotypes: PARP1 (c.852T>C-IVS19-297C>T) C-C/C-C diplotype and neutropenia (n = 106; P = 0.08; A), TDP1 IVS12+79 and rate of response (n = 89; P = 0.044; B), and XRCC1 (–1449delGGCC-c.1196G>A) GGCC-G/GGCC-G diplotype and rate of response (n = 89; P = 0.016; C). Genotypes were compared using the {chi}2 trend test and diplotype groups using the Fisher's exact test.

 
Analysis of efficacy to CPT-11–based therapies
Clinical response data for the patients are presented in Table 1. Associations of candidate gene htSNP genotypes and diplotypes with response are presented in Table 4. A gene-dose effect of TDP1 IVS12+79G>T on objective response was observed, with G/G patients (64%, 7 of 11 patients) having the highest proportion of responders compared with G/T (32%, 16 of 50 patients) and T/T (25%, 7 of 28 patients; P = 0.044; Fig. 1B). Most patients with the XRCC1 GGCC-G/GGCC-G diplotype (83%, 5 of 6 patients) had an objective response compared with 30% of patients with other XRCC1 diplotypes (25 of 83 patients; P = 0.016; Fig. 1C). No significant relationships between CDC45L and NFKB1 genotypes and diplotypes and tumor response were observed.

Analysis of putative causal alleles
Epidemiologic studies have found that SNPs in PARP1 [c.2285T>C (V762A)] and XRCC1 [–77T>C, c.580C>T (R194W), and c.839G>A (R280H)] alter the risk of developing some cancers, and the SNPs have been shown in vitro to alter function by affecting either gene transcription or protein function (23–27). It is therefore possible that these SNPs could be causative alleles and influence patient response to CPT-11 treatment. Patients were genotyped for the four SNPs and associations with clinical variables were assessed. The allele and genotype frequencies are presented in Table 3. Genotype distributions for the SNPs agreed with those predicted by the Hardy-Weinberg equilibrium. Owing to the expected small number of patients in the diplotype groups, only genotype-outcome analyses were conducted. The PARP1 and XRCC1 SNPs were not associated with toxicities or response rate, suggesting that these SNPs are not causative and do not explain the associations between PARP1 and XRCC1 htSNPs and haplotypes, and clinical outcomes observed in our patient population.

Multivariate analyses
Neutropenia. Univariate analyses suggest that TOP1 IVS4+61 (G/G versus A/G versus A/A), PARP1 diplotype (C-C/C-C versus other diplotypes), age, bilirubin, and performance status are associated with grade 3/4 neutropenia (P < 0.1), whereas other baseline clinical pathologic characteristics and UGT1A1*28 genotype were not associated (P > 0.1; ref. 17). A logistic regression model, using the associated factors (P < 0.1) as predictor variables, was used to identify independent factors that predict CPT-11–induced neutropenia. When these factors were considered jointly, only total bilirubin significantly predicted neutropenia (odds ratio, 1.15; 95% confidence interval, 1.0-1.3; P = 0.01). High total bilirubin (>1 mg/dL) predicted grade 3/4 neutropenia.

Clinical response
The line of chemotherapy was significantly associated with clinical response; 27 of 62 patients who received first-line therapy before participating in the present study achieved a clinical response with CPT-11–based treatment compared with 3 of 45 patients (44% versus 6%) that had received second-line therapy or more (P = 0.001, Fisher's exact test). The other baseline pathologic characteristics and UGT1A1*28 were not associated with tumor response. Because TDP1 IVS12+79 genotype (G/G versus T/G versus T/T) and XRCC1 diplotype (GGCC-G/GGCC-G versus other diplotypes) showed association to clinical response in the univariate analyses, they and line of chemotherapy were used as predictors in a stepwise logistic regression predicting CPT-11 response. The resulting model indicated that patients who at baseline had received two or more previous chemotherapy treatments were less likely to respond to CPT-11 therapy (odds ratio, 0.13; 95% confidence interval, 0.03-0.5; P = 0.003), whereas patients homozygous for the XRCC1 GGCC-G haplotype (odds ratio, 11.9; 95% confidence interval, 1.1-128; P = 0.04) were most likely to respond to therapy. TDP1 genotype (P = 0.14) was not predictive.


   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
There is considerable clinical interest in identifying factors that predict which patients are likely to respond to CPT-11 or are at high risk of excessive and potentially life-threatening toxicity. Such factors could help physicians individualize therapies such that patients with a low chance of achieving a clinical response to CPT-11 could be offered alternative chemotherapies, whereas those likely to respond to CPT-11 but at high risk of excessive toxicity could be administered a lower starting dose and monitored carefully for adverse events. Based on preclinical and clinical evidence, resistance to CPT-11 has been proposed to result from three different mechanisms: (a) inadequate accumulation of SN-38 in the tumor (determined by pharmacokinetic factors), (b) alterations in Topo I activity that result in reduced levels of the SN-38-Topo I-DNA complex (pharmacodynamic factors), and (c) alterations in the events downstream from the ternary complex, for example, apoptosis, cell cycle regulation, checkpoints, and DNA repair (pharmacodynamic factors; refs. 11, 32). In the present study, we investigated the potential of six candidate genes involved in mediating the pharmacodynamics of camptothecins to affect treatment outcomes of CPT-11 therapy in a retrospective candidate gene haplotype association study6 (12–16). The present exploratory study assessed associations of htSNPs and haplotypes of the genes with treatment outcomes in a cohort of advanced colorectal cancer patients who received CPT-11–based regimens. We found that genetic variation in XRCC1 was associated with the efficacy of CPT-11–based therapies but found no independent genetic factors that predicted toxicity. Given the hypothesis generating nature of our study and the limited number of observations of clinical response, the association between XRCC1 and clinical response requires validation in an independent, larger patient cohort. This is the first comprehensive investigation of the influence of genetic variation of pharmacodynamic genes on CPT-11 treatment outcomes.

XRCC1 plays a critical role in base excision repair by bringing together a complex of DNA repair proteins, including PARP1 and DNA polymerase β (33). We assessed whether genetic variation in XRCC1 was associated with CPT-11 treatment outcomes. We found that more patients with the XRCC1 c.1196G>A (R399Q) G/G genotype achieved a clinical response (46% versus 26%; P = 0.10) than patients with A alleles. This is the first study to evaluate an association between XRCC1 genetic variation and tumor response to CPT-11–based therapies in patients. Codon 399 is located within the BRCT-I interaction domain (amino acids 301-402) of XRCC1, a region that is critical for the role of XRCC1 in single-strand break repair and cell survival (34) and a poly(ADP-ribose)–binding motif (amino acids 379-400) that interacts with ribosylated PARP1 (33, 35, 36). The trend we observed for XRCC1 c.1196 agrees with the results of two previous pharmacogenetic studies (37, 38). The majority of patients with metastatic colorectal cancer who responded to oxaliplatin and 5-fluorouracil therapy were G/G patients (8 of 11 patients, 73%; 3 were G/A), whereas 66% of nonresponders (33 of 50 patients) carried an A allele (P < 0.05; ref. 37). In a study of Korean metastatic colorectal cancer patients, XRCC1 c.1196 G/Gs survived longer (P < 0.05) and had a better response rate (P < 0.05) to FOLFOX therapy than patients with A/G and A/A genotypes (38). Taken together, the findings suggest that wild-type XRCC1, encoded by the G allele with arginine at position 399, has poor DNA repair capacity compared with the variant glutamine-containing form. However, these in vivo observations are in conflict with the results of several in vitro studies that have consistently shown that the glutamine residue at codon 399 is associated with reduced DNA repair capacity as assessed by the persistence of DNA adducts, elevated levels of sister chromatid exchanges, increased RBC glycophorin A, TP53 mutations, and prolonged cell cycle delay (23). Based on the results of in vitro studies, tumors carrying the c.1196 variant allele (399Gln) would be expected to have impaired XRCC1 activity resulting in reduced DNA repair efficiency following exposure to DNA-damaging agents and leading to increased cell death and a positive tumor response to CPT-11 therapy rather than the opposite situation that has been shown in the present and previous patient studies. The reason for the discrepancy between the in vivo and in vitro genotype-outcome associations is unclear but may be explained by cell-specific effects of c.1196 (R399Q) on XRCC1 activity (23).

XRCC1 haplotype was a better predictor of clinical response than XRCC1 c.1196. All but one of the patients homozygous for the XRCC1 GGCC-G haplotype (83%, five of six patients) responded to therapy (Fig. 1C). In multivariate analysis, the haplotype still predicted response. The finding suggests the haplotype encodes an XRCC1 with reduced DNA repair activity compared with other haplotypes so that tumors homozygous for the haplotype are more sensitive to the cytotoxic effects of CPT-11 and consequently respond better to therapy than tumors with other haplotype pairs. –1449delGGCC did not affect responses to CPT-11 therapy, suggesting that the indel in combination with c.1196 did not directly influence XRCC1 activity. However, haplotype GGCC-G may have indirectly affected XRCC1 activity by tagging a causative allele. We assessed whether three XRCC1 SNPs, –77T>C, c.580 (R194W), and c.839 (R280H), for which there is in vitro and in vivo evidence to suggest they affect XRCC1 activity, were responsible for the haplotype-outcome association (23, 24, 27, 39). However, none of the SNPs predicted CPT-11 outcomes, suggesting that they do not on their own explain the relationship between the GGCC-G haplotype and response to CPT-11 therapy and additional polymorphisms carried on the haplotype may explain the association.

This is the first study to identify a genetic marker that independently predicts clinical response to CPT-11. Although our findings suggest that the XRCC1 GGCC-G haplotype might be a good candidate for tumor response to CPT-11, the number of patients studied with this haplotype pair was small; therefore, the association needs to be confirmed in a larger cohort. Interestingly, although not significant, none of the patients homozygous for XRCC1 GGCC-G experienced severe neutropenia. The lack of significant association might be explained by the low incidence of neutropenia in the cohort and the small number of patients with the diplotype. Together with the association with response, the findings suggest that patients with the GGCC-G/GGCC-G diplotype have a good tumor response to CPT-11 and are also tolerant to the chemotherapy, making them excellent candidates for CPT-11–based therapies. The frequency of the haplotype is higher in East Asians and West Africans (~50%) than in Europeans (~20%), suggesting that there may be ethnic differences in the tolerance of and response to CPT-11.

Despite the strong associations between UGT1A1*28 genotype and neutropenia that have been observed, many studies, including the present study, have failed to show the association (7, 8, 17, 40–43). The CPT-11 dose seems to be an important variable, with UGT1A1 genotype being most predictive at higher doses (9). This suggests that other factors, genetic or nongenetic, contribute to heterogeneity in toxicity to CPT-11. We investigated whether six genes involved in the pharmacodynamics of camptothecins influenced toxicity and therapeutic outcomes to CPT-11 therapy using a haplotype approach in a cohort of European, advanced colorectal cancer patients treated with CPT-11–based regimens. We found that the more TOP1 IVS4+61 variant alleles a patient carried, the greater their chance of experiencing excessive neutropenia. This association might be explained by the association of the variant with elevated SN-38-Top I-DNA complex formation in bone marrow cells, resulting in increased bone marrow sensitivity to CPT-11. Additionally, more of the patients homozygous for PARP1 C-C haplotype (c.852T>C-IVS19-297C>T) tended to experience the toxicity than patients with other haplotype pairs, suggesting that the C-C haplotype might be associated with reduced PARP1 DNA repair capacity, which renders the bone marrow cells less able to combat the cytotoxic effects of camptothecins. However, neither TOP1 htSNP nor the PARP1 haplotype was independent predictive markers of neutropenia by logistic regression, suggesting that these genetic markers might not influence a patient's risk of neutropenia. However, due to the low incidence of this adverse event, the study was underpowered to show an association. In our study, baseline bilirubin was the only factor that independently predicted neutropenia.

We also studied the influence of pharmacodynamic candidate gene diversity on diarrhea. None of the candidate genes was associated with the toxicity, suggesting that they do not influence a patient's risk of CPT-11–induced diarrhea. This is to be expected because UGT1A1*28 genotype explains a large proportion of the diarrhea cases in our cohort and may overshadow the contributions of the other genes that play smaller roles (17). UGT1A1*28 genotype is not associated with risk of diarrhea in most published studies, and therefore, a role for the pharmacodynamic genes in diarrhea should not be ruled out and requires testing in other cohorts treated with CPT-11 (9).

The objective of the study was to investigate whether genetic variation in genes involved in the pharmacodynamics of camptothecins influenced the toxicity to and efficacy of CPT-11. All patients studied were European with advanced colorectal cancer, suggesting that the sample had reduced allelic heterogeneity and population structure. The study was exploratory and hypothesis generating; therefore, the results should be interpreted with caution. In addition, several limitations of the study deserve mention. The population investigated was small (n = 107), suggesting that the study may have had reduced power to detect genotype-phenotype associations when allele and haplotype frequencies were low and the effect sizes of variants were small. The patients were treated with one of four CPT-11–containing therapies, which may have contributed to the interpatient variation in drug response. However, we tested whether chemotherapy regimen influenced treatment outcomes and found that it did not predict toxicity and efficacy by univariate analyses, suggesting that it may not be a confounding variable. It is possible that several of the associations we observed between genetic variants and patient outcomes, which were of marginal statistical significance (ranging from P < 0.05 to P < 0.005), were due to chance (i.e., false positives) and would not have sustained significance if corrected for multiple comparisons.

In conclusion, we have conducted the first comprehensive pharmacogenetic investigation of camptothecin pharmacodynamic factors using candidate genes that were selected from the literature and used a haplotype-outcome association study approach. Results of indirect gene-outcome associations using alleles and haplotypes suggest that an XRCC1 haplotype (GGCC-G) predicts clinical response to CPT-11–based therapies. None of the genetic variants independently predicted diarrhea or neutropenia. In combination with a patient's UGT1A1*28 genotype status and XRCC1 haplotype could enable individualization of treatment of patients with advanced colorectal cancer by identifying, before treatment, patients who might be at increased risk of CPT-11–induced toxicity and therefore require dose adjustment and/or monitoring for severe toxicities and those unlikely to benefit from CPT-11–based chemotherapy who could be offered alternate therapies, respectively. This study was hypothesis generating and further studies in large patient cohorts homogeneous for ethnicity, type of cancer, and chemotherapy regimen are required to confirm whether XRCC1 predicts the therapeutic efficacy of CPT-11–based therapies. In addition, studies are required to determine whether the genetic variants are prognostic or predictive factors.


   Footnotes
 
Grant support: NIH Pharmacogenetics Research Network grant U01 GM63340 and Fondo de Investigación Sanitaria grant FIS 05-1218.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

Current address for J.M. Hoskins and H.L. McLeod: UNC Institute for Pharmacogenomics and Individualized Therapy, University of North Carolina, Chapel Hill, North Carolina.

6 http://www.pharmgkb.org Back

7 http://coriell.umdnj.edu/ccr/ccrsumm.html Back

8 http://genome.ucsc.edu/cgi-gin/hgGateway Back

9 http://www.ncbi.nlm.nih.gov/SNP/index.html Back

10 http://lpgws.nci.nih.gov/perl/snpbr Back

11 http://gdb.jst.go.jp/HOWDY Back

12 http://egp.gs.washington.edu Back

13 http://www.abgene.com Back

14 http://krunch.med.yale.edu/hwsim/hwsim.txt Back

Received 6/14/07; revised 9/24/07; accepted 1/ 9/08.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Chabot GG. Clinical pharmacokinetics of irinotecan. Clin Pharmacokinet 1997;33:245–59.[Medline]
  2. Xu G, Zhang W, Ma MK, McLeod HL. Human carboxylesterase 2 is commonly expressed in tumor tissue and is correlated with activation of irinotecan. Clin Cancer Res 2002;8:2605–11.[Abstract/Free Full Text]
  3. Marsh S, Yu J, Hoskins JM. Colorectal cancer pharmacogenomics. Curr Colorectal Cancer Rep 2006;2:217–24.[CrossRef]
  4. Smith NF, Figg WD, Sparreboom A. Pharmacogenetics of irinotecan metabolism and transport: an update. Toxicol In Vitro 2006;20:163–75.[CrossRef][Medline]
  5. Marsh S, McLeod HL. Pharmacogenetics of irinotecan toxicity. Pharmacogenomics 2004;5:835–43.[CrossRef][Medline]
  6. Gagne JF, Montminy V, Belanger P, Journault K, Gaucher G, Guillemette C. Common human UGT1A polymorphisms and the altered metabolism of irinotecan active metabolite 7-ethyl-10-hydroxycamptothecin (SN-38). Mol Pharmacol 2002;62:608–17.[Abstract/Free Full Text]
  7. Rouits E, Boisdron-Celle M, Dumont A, Guerin O, Morel A, Gamelin E. Relevance of different UGT1A1 polymorphisms in irinotecan-induced toxicity: a molecular and clinical study of 75 patients. Clin Cancer Res 2004;10:5151–9.[Abstract/Free Full Text]
  8. Innocenti F, Undevia SD, Iyer L, et al. Genetic variants in the UDP-glucuronosyltransferase 1A1 gene predict the risk of severe neutropenia of irinotecan. J Clin Oncol 2004;22:1382–8.[Abstract/Free Full Text]
  9. Hoskins JM, Goldberg RM, Qu P, Ibrahim JG, McLeod HL. UGT1A1*28 genotype and irinotecan-induced neutropenia: dose matters. J Natl Cancer Inst 2007;99:1290–5.[Abstract/Free Full Text]
  10. Pommier Y. Camptothecins and topoisomerase I: a foot in the door. Targeting the genome beyond topoisomerase I with camptothecins and novel anticancer drugs: importance of DNA replication, repair and cell cycle checkpoints. Curr Med Chem Anti-Canc Agents 2004;4:429–34.[CrossRef]
  11. Rasheed ZA, Rubin EH. Mechanisms of resistance to topoisomerase I-targeting drugs. Oncogene 2003;22:7296–304.[CrossRef][Medline]
  12. Reid RJ, Fiorani P, Sugawara M, Bjornsti MA. CDC45 and DPB11 are required for processive DNA replication and resistance to DNA topoisomerase I-mediated DNA damage. Proc Natl Acad Sci U S A 1999;96:11440–5.[Abstract/Free Full Text]
  13. Malanga M, Althaus FR. Poly(ADP-ribose) reactivates stalled DNA topoisomerase I and induces DNA strand break resealing. J Biol Chem 2004;279:5244–8.[Abstract/Free Full Text]
  14. Barthelmes HU, Habermeyer M, Christensen MO, et al. TDP1 overexpression in human cells counteracts DNA damage mediated by topoisomerases I and II. J Biol Chem 2004;279:55618–25.[Abstract/Free Full Text]
  15. Park SY, Lam W, Cheng YC. X-ray repair cross-complementing gene I protein plays an important role in camptothecin resistance. Cancer Res 2002;62:459–65.[Abstract/Free Full Text]
  16. Cusack JC, Jr., Liu R, Houston M, et al. Enhanced chemosensitivity to CPT-11 with proteasome inhibitor PS-341: implications for systemic nuclear factor-{kappa}B inhibition. Cancer Res 2001;61:3535–40.[Abstract/Free Full Text]
  17. Marcuello E, Altes A, Menoyo A, Del Rio E, Gomez-Pardo M, Baiget M. UGT1A1 gene variations and irinotecan treatment in patients with metastatic colorectal cancer. Br J Cancer 2004;91:678–82.[CrossRef][Medline]
  18. Ameyaw MM, Collie-Duguid ES, Powrie RH, Ofori-Adjei D, McLeod HL. Thiopurine methyltransferase alleles in British and Ghanaian populations. Hum Mol Genet 1999;8:367–70.[Abstract/Free Full Text]
  19. Freimuth RR, Xiao M, Marsh S, et al. Polymorphism discovery in 51 chemotherapy pathway genes. Hum Mol Genet 2005;14:3595–603.[Abstract/Free Full Text]
  20. Freimuth RR, Stormo GD, McLeod HL. PolyMAPr: programs for polymorphism database mining, annotation, and functional analysis. Hum Mutat 2005;25:110–7.[CrossRef][Medline]
  21. Collins FS, Brooks LD, Chakravarti A. A DNA polymorphism discovery resource for research on human genetic variation. Genome Res 1998;8:1229–31.[Free Full Text]
  22. Karban AS, Okazaki T, Panhuysen CI, et al. Functional annotation of a novel NFKB1 promoter polymorphism that increases risk for ulcerative colitis. Hum Mol Genet 2004;13:35–45.[Abstract/Free Full Text]
  23. Hu Z, Ma H, Chen F, Wei Q, Shen H. XRCC1 polymorphisms and cancer risk: a meta-analysis of 38 case-control studies. Cancer Epidemiol Biomarkers Prev 2005;14:1810–8.[Abstract/Free Full Text]
  24. Hu Z, Ma H, Lu D, et al. A promoter polymorphism (–77T>C) of DNA repair gene XRCC1 is associated with risk of lung cancer in relation to tobacco smoking. Pharmacogenet Genomics 2005;15:457–63.[Medline]
  25. Lockett KL, Hall MC, Xu J, et al. The ADPRT V762A genetic variant contributes to prostate cancer susceptibility and deficient enzyme function. Cancer Res 2004;64:6344–8.[Abstract/Free Full Text]
  26. Zhang X, Miao X, Liang G, et al. Polymorphisms in DNA base excision repair genes ADPRT and XRCC1 and risk of lung cancer. Cancer Res 2005;65:722–6.[Abstract/Free Full Text]
  27. Hao B, Wang H, Zhou K, et al. Identification of genetic variants in base excision repair pathway and their associations with risk of esophageal squamous cell carcinoma. Cancer Res 2004;64:4378–84.[Abstract/Free Full Text]
  28. Marsh S, King CR, Garsa AA, McLeod HL. Pyrosequencing of clinically relevant polymorphisms. Methods Mol Biol 2005;311:97–114.[Medline]
  29. Clement M, Posada D, Crandall KA. TCS: a computer program to estimate gene genealogies. Mol Ecol 2000;9:1657–9.[CrossRef][Medline]
  30. Templeton AR, Crandall KA, Sing CF. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 1992;132:619–33.[Abstract]
  31. Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet 2002;70:425–34.[CrossRef][Medline]
  32. Pommier Y, Pourquier P, Urasaki Y, Wu J, Laco GS. Topoisomerase I inhibitors: selectivity and cellular resistance. Drug Resist Updat 1999;2:307–18.[CrossRef][Medline]
  33. El-Khamisy SF, Masutani M, Suzuki H, Caldecott KW. A requirement for PARP-1 for the assembly or stability of XRCC1 nuclear foci at sites of oxidative DNA damage. Nucleic Acids Res 2003;31:5526–33.[Abstract/Free Full Text]
  34. Taylor RM, Thistlethwaite A, Caldecott KW. Central role for the XRCC1 BRCT I domain in mammalian DNA single-strand break repair. Mol Cell Biol 2002;22:2556–63.[Abstract/Free Full Text]
  35. Masson M, Niedergang C, Schreiber V, Muller S, Menissier-de Murcia J, de Murcia G. XRCC1 is specifically associated with poly(ADP-ribose) polymerase and negatively regulates its activity following DNA damage. Mol Cell Biol 1998;18:3563–71.[Abstract/Free Full Text]
  36. Pleschke JM, Kleczkowska HE, Strohm M, Althaus FR. Poly(ADP-ribose) binds to specific domains in DNA damage checkpoint proteins. J Biol Chem 2000;275:40974–80.[Abstract/Free Full Text]
  37. Stoehlmacher J, Ghaderi V, Iobal S, et al. A polymorphism of the XRCC1 gene predicts for response to platinum based treatment in advanced colorectal cancer. Anticancer Res 2001;21:3075–9.[Medline]
  38. Suh KW, Kim JH, Kim DY, Kim YB, Lee C, Choi S. Which gene is a dominant predictor of response during FOLFOX chemotherapy for the treatment of metastatic colorectal cancer, the MTHFR or XRCC1 gene? Ann Surg Oncol 2006;13:1379–85.[Abstract/Free Full Text]
  39. Fan J, Otterlei M, Wong HK, Tomkinson AE, Wilson DM III. XRCC1 co-localizes and physically interacts with PCNA. Nucleic Acids Res 2004;32:2193–201.[Abstract/Free Full Text]
  40. Ando Y, Saka H, Ando M, et al. Polymorphisms of UDP-glucuronosyltransferase gene and irinotecan toxicity: a pharmacogenetic analysis. Cancer Res 2000;60:6921–6.[Abstract/Free Full Text]
  41. Iyer L, Das S, Janisch L, et al. UGT1A1*28 polymorphism as a determinant of irinotecan disposition and toxicity. Pharmacogenomics J 2002;2:43–7.[CrossRef][Medline]
  42. Carlini LE, Meropol NJ, Bever J, et al. UGT1A7 and UGT1A9 polymorphisms predict response and toxicity in colorectal cancer patients treated with capecitabine/irinotecan. Clin Cancer Res 2005;11:1226–36.[Abstract/Free Full Text]
  43. Massacesi C, Terrazzino S, Marcucci F, et al. Uridine diphosphate glucuronosyl transferase 1A1 promoter polymorphism predicts the risk of gastrointestinal toxicity and fatigue induced by irinotecan-based chemotherapy. Cancer 2006;106:1007–16.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 JCOHome page
P. N. Lara Jr, R. Natale, J. Crowley, H. J. Lenz, M. W. Redman, J. E. Carleton, J. Jett, C. J. Langer, J. P. Kuebler, S. R. Dakhil, et al.
Phase III Trial of Irinotecan/Cisplatin Compared With Etoposide/Cisplatin in Extensive-Stage Small-Cell Lung Cancer: Clinical and Pharmacogenomic Results From SWOG S0124
J. Clin. Oncol., May 20, 2009; 27(15): 2530 - 2535.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
J. M. Hoskins and H. L. McLeod
UGT1A and Irinotecan Toxicity: Keeping It in the Family
J. Clin. Oncol., May 20, 2009; 27(15): 2419 - 2421.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hoskins, J. M.
Right arrow Articles by Baiget, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hoskins, J. M.
Right arrow Articles by Baiget, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online