Fridley, Brooke L.

Abstract Purpose: To identify molecular predictors of grade 3/4 neutropenic or leukopenic events (NLE) after chemotherapy using a genome-wide association study (GWAS). Experimental Design: A GWAS was performed on patients in the phase III chemotherapy study SUCCESS-A (n = 3,322). Genotyping was done using the Illumina HumanOmniExpress-12v1 array. Findings were functionally validated with cell culture models and the genotypes and gene expression of possible causative genes were correlated with clinical treatment response and prognostic outcomes. Results: One locus on chromosome 16 (rs4784750; NLRC5; P = 1.56E-8) and another locus on chromosome 13 (rs16972207; TNFSF13B; P = 3.42E-8) were identified at a genome-wide significance level. Functional validation revealed that expression of these two genes is altered by genotype-dependent and chemotherapy-dependent activity of two transcription factors. Genotypes also showed an association with disease-free survival in patients with an NLE. Conclusions: Two loci in NLRC5 and TNFSF13B are associated with NLEs. The involvement of the MHC I regulator NLRC5 implies the possible involvement of immuno-oncological pathways.

Thiopurine drugs such as 6‐mercaptopurine (6‐MP) and 6‐thioguanine (6‐TG) are effective immunosuppressants and antineoplastic agents. We used a genome‐wide approach to identify genes that might contribute to individual variation in thiopurine response. Basal gene expression profiles, genome‐wide SNP data, and 6‐MP and 6‐TG Cell growth inhibition (GI50) values were generated for 200 Coriell Institute “Human Variation Panel” lymphoblastoid cell lines. Gene expression and genome‐wide SNPs were correlated with GI50 values to identify candidate genes that might contribute to thiopurine response. We identified 104 expression probes and 161 SNPs, as well as 315 gene probes and 579 SNPs that correlated significantly with GI50 values for 6‐MP and 6‐TG, respectively (P<10 −5 and 10 −6 , respectively). Three genes, CD83 , TXNDC13 and SNAP23 , contained SNPs that contributed significantly to both 6‐MP and 6‐TG cytotoxicity and were also significantly associated with expression levels (P<10 −4 ), suggesting a “three‐way model” involving genotype, gene expression and cytotoxicity. GBP5 , a gene with an expression level that correlated negatively with 6‐MP and 6‐TG GI50 values, sensitized leukemia cells when overexpressed. Functional studies of the remainder of the candidate genes are being performed. These results may enhance our ability to predict thiopurine response phenotypes. NIH GM28157.

Objectives Betaine‐homocysteine methyltransferase (BHMT) catalyzes the remethylation of homocysteine. BHMT in the human liver is responsible for up to 25% of homocysteine metabolism. Knowledge of the relationship between genetic polymorphisms and BHMT phenotypes could enhance our understanding of the role of this reaction in drug metabolism. Methods Levels of BHMT enzyme activity and immunoreactive protein were measured in 268 human liver biopsies. Twelve tag BHMT SNPs from our own gene resequencing and HapMap data were used to genotype DNA from these samples. Genotype‐phenotype association analysis was then performed. BHMT enzyme activity averaged 0.077 ± 0.028 units/μg, with a 12 fold variation. Results BHMT protein levels averaged 6.9 ± 3.3 units/μg, with a 50‐fold variation. Three SNPs (rs41272270, rs16876512, and rs6875201), located 28kb upstream, in the 5′‐UTR and in intron 1 of BHMT , respectively, were significantly correlated with both phenotypes (p<10 −8 for enzyme activity and p<10 −5 for protein level). rs41272270 was associated with altered protein binding during electrophoresis mobility shift assay. Conclusions Three tag SNPs were significantly associated with human hepatic BHMT enzyme activity and immunoreactive protein. These observations may help us understand the regulation of BHMT in the human liver and its role in drug metabolism. Supported by NIH U01 GM61388, R01 GM28157, R01 CA132780.

Abstract Blood lipids have been associated with the development of a range of cancers, including breast, lung and colorectal cancer. For endometrial cancer, observational studies have reported inconsistent associations between blood lipids and cancer risk. To reduce biases from unmeasured confounding, we performed a bidirectional, two‐sample Mendelian randomization analysis to investigate the relationship between levels of three blood lipids (low‐density lipoprotein [LDL] and high‐density lipoprotein [HDL] cholesterol, and triglycerides) and endometrial cancer risk. Genetic variants associated with each of these blood lipid levels ( P < 5 × 10 −8 ) were identified as instrumental variables, and assessed using genome‐wide association study data from the Endometrial Cancer Association Consortium (12 906 cases and 108 979 controls) and the Global Lipids Genetic Consortium (n = 188 578). Mendelian randomization analyses found genetically raised LDL cholesterol levels to be associated with lower risks of endometrial cancer of all histologies combined, and of endometrioid and non‐endometrioid subtypes. Conversely, higher genetically predicted HDL cholesterol levels were associated with increased risk of non‐endometrioid endometrial cancer. After accounting for the potential confounding role of obesity (as measured by genetic variants associated with body mass index), the association between genetically predicted increased LDL cholesterol levels and lower endometrial cancer risk remained significant, especially for non‐endometrioid endometrial cancer. There was no evidence to support a role for triglycerides in endometrial cancer development. Our study supports a role for LDL and HDL cholesterol in the development of non‐endometrioid endometrial cancer. Further studies are required to understand the mechanisms underlying these findings.

Abstract Background: Constitutive activation of the JAK-STAT signaling pathway through acquisition of the JAK2V617F mutation plays a key role in the pathogenesis of related myeloproliferative disorders (MPD) including polycythemia vera (PV), agnogenic myeloid metaplasia (AMM), and essential thrombocythemia (ET). Hypothesis: Genetic variation in JAK2 and/or the cytokine receptor genes for erythropoietin (EPOR), thrombopoietin (MPL), and granulocyte colony stimulating factor (GCSFR), influences the expression of a specific disease phenotype (PV, AMM, or ET). Methods: We studied 179 Caucasian MPD patients (PV=84, AMM=58, ET=37) for genetic variation in 4 candidate genes (i.e. JAK2, EPOR, MPL, GCSFR) through single nucleotide polymorphism (SNP) and haplotype analyses. A total of 32 LD (linkage disequilibrium) tag-SNPs with a minimum allele frequency of at least 5% were selected from the HapMap CEU database (JAK2=13, EPOR=4, MPL=5, GCSFR=10). Genotyping was performed using archived DNA from enriched neutrophils. Results: Seventy six (94%), 26 (45%), and 14 (38%) patients with PV, AMM, and ET, respectively, carried the JAK2V617F mutation. Significant differences in allele frequencies was observed at 6 SNP loci (rs10758669, rs3808850, rs7849191, rs7046736, rs10815148, and rs12342421, p-values &lt; 0.0005), all within the JAK2 gene, in comparing our overall study population with the founder Caucasian population in the HapMap database. In the additive genotype-phenotype association analysis adjusted for gender and age at diagnosis, 3 SNP loci in JAK2 (rs7046736, rs10815148, and rs12342421) were found to be significantly, but reciprocally associated with PV (p-values &lt; 0.00006, odds ratio=0.37, 2.82, and 2.39, respectively) and ET (p-values &lt; 0.002, odds ratio=2.50, 0.36, and 0.41, respectively), but not AMM, in terms of the minor allele being ’protective’. These three SNPs were all in high LD, with the ’r2’ measures of LD between 0.59 and 0.71. Furthermore, 2 additional JAK2 SNP loci (rs10758669, p-value = 0.0024 and rs10974947, p-value = 0.0046) were significantly associated with PV (odds ratio=1.88 and 0.47, respectively), but not ET or AMM. Similarly, presence of the minor allele at a single SNP locus in EPOR (rs318699, p-value = 0.0012) was significantly associated with PV only (odds ratio=2.16). For the phenotype-genotype intragene haplotype analyses, the EPOR intragene haplotypes GAAA and GGAA were significantly associated with PV (simulated global p-value = 0.058, simulated individual p-values 0.0013 and 0.0068, haplotype frequency of 35% and 56%, respectively). In addition to EPOR, 6 intragene haplotypes within JAK2 where significantly associated with PV (simulated global p-value &lt; 0.0001, individual simulated p-values &lt; 0.03). Conclusion: The current study demonstrates a genotype-phenotype association that involves JAK2 and EPOR in the setting of PV, but not that of AMM. Some JAK2 SNPs were found to be associated with both PV and ET but in opposite direction. Therefore, genetic variation among MPDs might contribute to phenotypic diversity in the presence of specific mutations.

Abstract Abstract 873 Introduction: Clonal chromosomal anomalies have been well-characterized in the blood cells of patients with Chronic Lymphocytic Leukemia (CLL) and other hematological malignancies. However, the frequencies and types of chromosomal anomalies found in “normal” people (i.e. no history of hematological cancer) are much less well known. Here we compare chromosomal anomalies in untreated CLL patients with those found in a large sample of people (n>50,000) recruited for genome-wide association studies in the Gene Environment Association Studies (GENEVA) consortium. This comparison provides information about the extent to which people without a prior history of hematological cancer have chromosomal anomalies that are characteristic of CLL. Methods and Results: The anomaly detection method focused on acquired, rather than inherited, chromosomal anomalies identified in whole blood samples. The method and results for the GENEVA subjects were published recently (Laurie et al. 2012, Nat. Gen. 44:642–650). The same method was applied to pre-treatment samples from 220 CLL patients from US Intergroup Trial E2997 of the Eastern Cooperative Oncology Group (ECOG). Acquired anomalies were detected as mosaic mixtures of normal disomic cells along with cells containing autosomal anomalies consisting of large (generally > 1 Mb) deletions, duplications and/or acquired uniparental disomy (UPD). Detection requires a large fraction of cells with the same abnormal karyotype (>5–10%), which is presumably of clonal origin. The subjects (of all ages) came from 12 different GENEVA studies, which addressed a variety of health conditions not directly related to hematological cancer (e.g. addiction and glaucoma). DNA was extracted mainly from blood samples (or, in 8% of subjects, from buccal/saliva samples) and genotyped on Illumina SNP mircroarrays. Clonal mosaic anomalies were detected from change points in intensity and genotype data (B Allele Frequency and Log R Ratio). We found 514 anomalies in 404 GENEVA subjects out of 50,222 examined (0.8% of subjects), although the prevalence increased with age. Clinical data from about 75% of subjects with a clonal chromosomal abnormality revealed that only a small fraction (2.8%, 95% CI=1.0 – 4.7%) had a record of prior hematological cancer. In the ECOG study, blood samples from 220 CLL patients enrolled in protocol E2997 were genotyped on the Illumina HumanOmni1-Quad_v1 array and clonal mosaic chromosome anomalies were detected using the GENEVA method. Chromosomal anomalies typical of CLL also were identified using interphase FISH (i.e. 6q23.3-, 11q22-, 12q15+, 13q14-, 17p13-) as previously reported (Grever, JCO 25:799-804). We found 489 anomalies in 148 patients of the 220 examined (67% of patients). The anomalies detected by FISH and SNP microarray were highly concordant (>98%). Although, the most common recurrent anomalies occurred on chromosomes 6, 11, 12, 13 and 17 in regions that span the FISH probes (except for chromosome 6), a majority of the anomalies detected by SNP data (74% of 489) were not detected by the CLL FISH panel. The frequency of large chromosomal anomalies was much higher in CLL patients than in GENEVA (67% versus 0.8% of subjects) and the overall composition of anomaly types found was also different. The fraction of anomalies consisting of acquired UPD was substantially higher in GENEVA subjects than in ECOG (34.0% versus 10.6%), while deletions were less common in GENEVA (50.1% versus 72.4%) and duplications were similar (15.6% versus 17.0%) (Chi-squared test p-value=3 × 10−18). We compared the most commonly deleted regions (CDR) on each chromosome across studies. In the ECOG CLL patients, four chromosomes each had 10 or more overlapping deletions in their top CDR (6, 11, 13 and 17), which span the FISH probes (except for chromosome 6). The GENEVA subjects also had CDR on chromosomes 11, 13 and 17, which overlap those in ECOG. In addition, GENEVA had CDR spanning genes or regions implicated in myeloid malignancies, including DNMT3a on 2p, TET2 on 4q, PRAME on 22q, and a region on 20q13. Conclusions: These genotyping results of >50,000 subjects suggest clonal chromosomal anomalies may represent precursor states, or perhaps undiagnosed or unrecorded cases, of a variety of hematological malignancies. Further studies to validate these findings and explore the natural history of these genetic defects in individuals without hematologic malignancy are warranted. Disclosures: Shanafelt: Genentech: Research Funding; Glaxo-Smith-Kline: Research Funding; Cephalon: Research Funding; Hospira: Research Funding; Celgene: Research Funding; Polyphenon E International: Research Funding.

AbstractBackground:Germline DNA copy number variation (CNV) is a ubiquitous source of genetic variation and remains largely unexplored in association with epithelial ovarian cancer (EOC) risk.Methods:CNV was quantified in the DNA of approximately 3,500 cases and controls genotyped with the Illumina 610k and HumanOmni2.5M arrays. We performed a genome-wide association study of common (&gt;1%) CNV regions (CNVRs) with EOC and high-grade serous (HGSOC) risk and, using The Cancer Genome Atlas (TCGA), performed in silico analyses of tumor-gene expression.Results:Three CNVRs were associated (P &lt; 0.01) with EOC risk: two large (∼100 kb) regions within the 610k set and one small (&lt;5 kb) region with the higher resolution 2.5M data. Large CNVRs included a duplication at LILRA6 (OR = 2.57; P = 0.001) and a deletion at CYP2A7 (OR = 1.90; P = 0.007) that were strongly associated with HGSOC risk (OR = 3.02; P = 8.98 × 10−5). Somatic CYP2A7 alterations correlated with EGLN2 expression in tumors (P = 2.94 × 10−47). An intronic ERBB4/HER4 deletion was associated with reduced EOC risk (OR = 0.33; P = 9.5 × 10−2), and somatic deletions correlated with ERBB4 downregulation (P = 7.05 × 10−5). Five CNVRs were associated with HGSOC, including two reduced-risk deletions: one at 1p36.33 (OR = 0.28; P = 0.001) that correlated with lower CDKIIA expression in TCGA tumors (P = 2.7 × 10−7), and another at 8p21.2 (OR = 0.52; P = 0.002) that was present somatically where it correlated with lower GNRH1 expression (P = 5.9 × 10−5).Conclusions:Though CNV appears to not contribute largely to EOC susceptibility, a number of low-to-common frequency variants may influence the risk of EOC and tumor-gene expression.Impact:Further research on CNV and EOC susceptibility is warranted, particularly with CNVs estimated from high-density arrays.

AbstractX chromosome inactivation (XCI) is a key epigenetic gene expression regulatory process, which may play a role in women’s cancer. In particular tissues, some genes are known to escape XCI, yet patterns of XCI in ovarian cancer (OC) and their clinical associations are largely unknown. To examine XCI in OC, we integrated germline genotype with tumor copy number, gene expression and DNA methylation information from 99 OC patients. Approximately 10% of genes showed different XCI status (either escaping or being subject to XCI) compared with the studies of other tissues. Many of these genes are known oncogenes or tumor suppressors (e.g. DDX3X, TRAPPC2 and TCEANC). We also observed strong association between cis promoter DNA methylation and allele-specific expression imbalance (P = 2.0 × 10−10). Cluster analyses of the integrated data identified two molecular subgroups of OC patients representing those with regulated (N = 47) and dysregulated (N = 52) XCI. This XCI cluster membership was associated with expression of X inactive specific transcript (P = 0.002), a known driver of XCI, as well as age, grade, stage, tumor histology and extent of rl disease following surgical debulking. Patients with dysregulated XCI (N = 52) had shorter time to recurrence (HR = 2.34, P = 0.001) and overall survival time (HR = 1.87, P = 0.02) than those with regulated XCI, although results were attenuated after covariate adjustment. Similar findings were observed when restricted to high-grade serous tumors. We found evidence of a unique OC XCI profile, suggesting that XCI may play an important role in OC biology. Additional studies to examine somatic changes with paired tumor-normal tissue are needed.

Thymidylate synthase (TYMS) is a crucial enzyme for DNA synthesis. TYMS expression is regulated by its antisense mRNA, ENOSF1. Disrupted regulation may promote uncontrolled DNA synthesis and tumor growth. We sought to replicate our previously reported association between rs495139 in the TYMS-ENOSF1 3′ gene region and increased risk of mucinous ovarian carcinoma (MOC) in an independent sample. Genotypes from 24,351 controls to 15,000 women with invasive OC, including 665 MOC, were available. We estimated per-allele odds ratios (OR) and 95% confidence intervals (CI) using unconditional logistic regression, and meta-analysis when combining these data with our previous report. The association between rs495139 and MOC was not significant in the independent sample (OR = 1.09; 95% CI = 0.97–1.22; p = 0.15; N = 665 cases). Meta-analysis suggested a weak association (OR = 1.13; 95% CI = 1.03–1.24; p = 0.01; N = 1019 cases). No significant association with risk of other OC histologic types was observed (p = 0.05 for tumor heterogeneity). In expression quantitative trait locus (eQTL) analysis, the rs495139 allele was positively associated with ENOSF1 mRNA expression in normal tissues of the gastrointestinal system, particularly esophageal mucosa (r = 0.51, p = 1.7 × 10−28), and nonsignificantly in five MOC tumors. The association results, along with inconclusive tumor eQTL findings, suggest that a true effect of rs495139 might be small.