Herbert Yu

Obesity and risk of ovarian cancer subtypes: evidence from the Ovarian Cancer Association Consortium

Whilst previous studies have reported that higher BMI increases a womans risk of developing ovarian cancer, associations for the different histological subtypes have not been well defined. As the prevalence of obesity has increased dramatically, and classification of ovarian histology has improved in the last decade, we sought to examine the association in a pooled analysis of recent studies participating in the Ovarian Cancer Association Consortium. We evaluated the association between BMI (recent, maximum and in young adulthood) and ovarian cancer risk using original data from 15 case-control studies (13 548 cases and 17 913 controls). We combined study-specific adjusted odds ratios (ORs) using a random-effects model. We further examined the associations by histological subtype, menopausal status and post-menopausal hormone use. High BMI (all time-points) was associated with increased risk. This was most pronounced for borderline serous (recent BMI: pooled OR=1.24 per 5 kg/m(2); 95% CI 1.18-1.30), invasive endometrioid (1.17; 1.11-1.23) and invasive mucinous (1.19; 1.06-1.32) tumours. There was no association with serous invasive cancer overall (0.98; 0.94-1.02), but increased risks for low-grade serous invasive tumours (1.13, 1.03-1.25) and in pre-menopausal women (1.11; 1.04-1.18). Among post-menopausal women, the associations did not differ between hormone replacement therapy users and non-users. Whilst obesity appears to increase risk of the less common histological subtypes of ovarian cancer, it does not increase risk of high-grade invasive serous cancers, and reducing BMI is therefore unlikely to prevent the majority of ovarian cancer deaths. Other modifiable factors must be identified to control this disease.

Elucidating the Landscape of Aberrant DNA Methylation in Hepatocellular Carcinoma

Background Hepatocellular carcinoma (HCC) is one of the most common cancers and frequently presents with an advanced disease at diagnosis. There is only limited knowledge of genome-scale methylation changes in HCC. Methods and Findings We performed genome-wide methylation profiling in a total of 47 samples including 27 HCC and 20 adjacent normal liver tissues using the Illumina HumanMethylation450 BeadChip. We focused on differential methylation patterns in the promoter CpG islands as well as in various less studied genomic regions such as those surrounding the CpG islands, i.e. shores and shelves. Of the 485,577 loci studied, significant differential methylation (DM) was observed between HCC and adjacent normal tissues at 62,692 loci or 13% (p<1.03e-07). Of them, 61,058 loci (97%) were hypomethylated and most of these loci were located in the intergenic regions (43%) or gene bodies (33%). Our analysis also identified 10,775 differentially methylated (DM) loci (17% out of 62,692 loci) located in or surrounding the gene promoters, 4% of which reside in known Differentially Methylated Regions (DMRs) including reprogramming specific DMRs and cancer specific DMRs, while the rest (10,315) involving 4,106 genes could be potential new HCC DMR loci. Interestingly, the promoter-related DM loci occurred twice as frequently in the shores than in the actual CpG islands. We further characterized 982 DM loci in the promoter CpG islands to evaluate their potential biological function and found that the methylation changes could have effect on the signaling networks of Cellular development, Gene expression and Cell death (p = 1.0e-38), with BMP4, CDKN2A, GSTP1, and NFATC1 on the top of the gene list. Conclusion Substantial changes of DNA methylation at a genome-wide level were observed in HCC. Understanding epigenetic changes in HCC will help to elucidate the pathogenesis and may eventually lead to identification of molecular markers for liver cancer diagnosis, treatment and prognosis.

Common variation at 2p13.3, 3q29, 7p13 and 17q25.1 associated with susceptibility to pancreatic cancer.

Pancreatic cancer is the fourth leading cause of cancer death in the developed world. Both inherited high-penetrance mutations in BRCA2 (ref. 2), ATM, PALB2 (ref. 4), BRCA1 (ref. 5), STK11 (ref. 6), CDKN2A and mismatch-repair genes and low-penetrance loci are associated with increased risk. To identify new risk loci, we performed a genome-wide association study on 9,925 pancreatic cancer cases and 11,569 controls, including 4,164 newly genotyped cases and 3,792 controls in 9 studies from North America, Central Europe and Australia. We identified three newly associated regions: 17q25.1 (LINC00673, rs11655237, odds ratio (OR) = 1.26, 95% confidence interval (CI) = 1.19–1.34, P = 1.42 × 10−14), 7p13 (SUGCT, rs17688601, OR = 0.88, 95% CI = 0.84–0.92, P = 1.41 × 10−8) and 3q29 (TP63, rs9854771, OR = 0.89, 95% CI = 0.85–0.93, P = 2.35 × 10−8). We detected significant association at 2p13.3 (ETAA1, rs1486134, OR = 1.14, 95% CI = 1.09–1.19, P = 3.36 × 10−9), a region with previous suggestive evidence in Han Chinese. We replicated previously reported associations at 9q34.2 (ABO), 13q22.1 (KLF5), 5p15.33 (TERT and CLPTM1), 13q12.2 (PDX1), 1q32.1 (NR5A2), 7q32.3 (LINC-PINT), 16q23.1 (BCAR1) and 22q12.1 (ZNRF3). Our study identifies new loci associated with pancreatic cancer risk.

Case–Control Study of Aspirin Use and Risk of Pancreatic Cancer

Background: Pancreas-cancer prognosis is dismal, with 5-year survival less than 5%. Significant relationships between aspirin use and decreased pancreas-cancer incidence and mortality have been shown in four of 13 studies. Methods: To evaluate further a possible association between aspirin use and risk of pancreatic cancer, we used data from a population-based Connecticut study conducted from January 2005 to August 2009, of 362 pancreas-cancer cases frequency matched to 690 randomly sampled controls. Results: Overall, regular use of aspirin was associated with reduced risk of pancreatic cancer [odds ratio (OR), 0.52; 95% confidence interval (CI), 0.39–0.69]. Increments of decreasing risk of pancreatic cancer were observed for each year of low-dose or regular-dose aspirin use (OR, 0.94; 95% CI, 0.91–0.98 and OR, 0.98; 95% CI, 0.96–1.01, respectively) and for increasing years in the past that low-dose or regular-dose aspirin use had started (OR, 0.95; 95% CI, 0.92–0.99 and OR, 0.98; 95% CI, 0.96–1.00, respectively). Reduced risk of pancreatic cancer was seen in most categories of calendar time period of aspirin use, for both low-dose aspirin and regular-dose aspirin use. Relative to continuing use at the time of interview, termination of aspirin use within 2 years of interview was associated with increased risk of pancreatic cancer (OR, 3.24; 95% CI, 1.58–6.65). Conclusions: Our results provide some support that a daily aspirin regimen may reduce risk of developing pancreatic cancer. Impact: Long-term aspirin use has benefits for both cardiovascular disease and cancer, but appreciable bleeding complications that necessitate risk–benefit analysis for individual applications. Cancer Epidemiol Biomarkers Prev; 23(7); 1254–63. ©2014 AACR.

Breast Cancer Risk in Metabolically Healthy but Overweight Postmenopausal Women

Adiposity is an established risk factor for postmenopausal breast cancer. Recent data suggest that high insulin levels in overweight women may play a major role in this relationship, due to insulins mitogenic/antiapoptotic activity. However, whether overweight women who are metabolically healthy (i.e., normal insulin sensitivity) have elevated risk of breast cancer is unknown. We investigated whether overweight women with normal insulin sensitivity [i.e., homeostasis model assessment of insulin resistance (HOMA-IR) index, or fasting insulin level, within the lowest quartile (q1)] have increased breast cancer risk. Subjects were incident breast cancer cases (N = 497) and a subcohort (N = 2,830) of Womens Health Initiative (WHI) participants with available fasting insulin and glucose levels. In multivariate Cox models, metabolically healthy overweight women, defined using HOMA-IR, were not at elevated risk of breast cancer compared with metabolically healthy normal weight women [HRHOMA-IR, 0.96; 95% confidence interval (CI), 0.64-1.42]. In contrast, the risk among women with high (q3-4) HOMA-IRs was elevated whether they were overweight (HRHOMA-IR, 1.76; 95% CI, 1.19-2.60) or normal weight (HRHOMA-IR, 1.80; 95% CI, 0.88-3.70). Similarly, using fasting insulin to define metabolic health, metabolically unhealthy women (insulin q3-4) were at higher risk of breast cancer regardless of whether they were normal weight (HRinsulin, 2.06; 95% CI, 1.01-4.22) or overweight (HRinsulin, 2.01; 95% CI, 1.35-2.99), whereas metabolically healthy overweight women did not have significantly increased risk of breast cancer (HRinsulin, 0.96; 95% CI, 0.64-1.42) relative to metabolically healthy normal weight women. Metabolic health (e.g., HOMA-IR or fasting insulin) may be more biologically relevant and more useful for breast cancer risk stratification than adiposity per se.

A novel model to combine clinical and pathway-based transcriptomic information for the prognosis prediction of breast cancer.

Breast cancer is the most common malignancy in women worldwide. With the increasing awareness of heterogeneity in breast cancers, better prediction of breast cancer prognosis is much needed for more personalized treatment and disease management. Towards this goal, we have developed a novel computational model for breast cancer prognosis by combining the Pathway Deregulation Score (PDS) based pathifier algorithm, Cox regression and L1-LASSO penalization method. We trained the model on a set of 236 patients with gene expression data and clinical information, and validated the performance on three diversified testing data sets of 606 patients. To evaluate the performance of the model, we conducted survival analysis of the dichotomized groups, and compared the areas under the curve based on the binary classification. The resulting prognosis genomic model is composed of fifteen pathways (e.g. P53 pathway) that had previously reported cancer relevance, and it successfully differentiated relapse in the training set (log rank p-value = 6.25e-12) and three testing data sets (log rank p-value<0.0005). Moreover, the pathway-based genomic models consistently performed better than gene-based models on all four data sets. We also find strong evidence that combining genomic information with clinical information improved the p-values of prognosis prediction by at least three orders of magnitude in comparison to using either genomic or clinical information alone. In summary, we propose a novel prognosis model that harnesses the pathway-based dysregulation as well as valuable clinical information. The selected pathways in our prognosis model are promising targets for therapeutic intervention.

Testing the proportional hazards assumption in case-cohort analysis

BackgroundCase-cohort studies have become common in epidemiological studies of rare disease, with Cox regression models the principal method used in their analysis. However, no appropriate procedures to assess the assumption of proportional hazards of case-cohort Cox models have been proposed.MethodsWe extended the correlation test based on Schoenfeld residuals, an approach used to evaluate the proportionality of hazards in standard Cox models. Specifically, pseudolikelihood functions were used to define “case-cohort Schoenfeld residuals”, and then the correlation of these residuals with each of three functions of event time (i.e., the event time itself, rank order, Kaplan-Meier estimates) was determined. The performances of the proposed tests were examined using simulation studies. We then applied these methods to data from a previously published case-cohort investigation of the insulin/IGF-axis and colorectal cancer.ResultsSimulation studies showed that each of the three correlation tests accurately detected non-proportionality. Application of the proposed tests to the example case-cohort investigation dataset showed that the Cox proportional hazards assumption was not satisfied for certain exposure variables in that study, an issue we addressed through use of available, alternative analytical approaches.ConclusionsThe proposed correlation tests provide a simple and accurate approach for testing the proportional hazards assumption of Cox models in case-cohort analysis. Evaluation of the proportional hazards assumption is essential since its violation raises questions regarding the validity of Cox model results which, if unrecognized, could result in the publication of erroneous scientific findings.

Circulating MicroRNAs in Relation to EGFR Status and Survival of Lung Adenocarcinoma in Female Non-Smokers

Objectives Lung adenocarcinoma is considered a unique disease for Asian female non-smokers. We investigated whether plasma microRNA (miRNA) expression profiles are different by the EGFR status and are associated with survival outcomes of the patients. Methods Using real-time RT-PCR, we analyzed the expression of 20 miRNAs in the plasma of 105 female patients with lung adenocarcinoma. Kaplan-Meier survival analysis and Cox proportional hazards regression were performed to determine the association between miRNA expression and overall survival. Time dependent receiver operating characteristic (ROC) analysis was also performed. Results In the 20 miRNAs, miR-122 were found differently expressed between wild and mutant EGFR carriers (P=0.018). Advanced disease stage and tumor metastasis were independently associated with poor prognosis of patients with lung adenocarcinoma (P=0.010 and 1.0×10-4). Plasma levels of miR-195 and miR-122 expression were also associated with overall survival in the patients, especially in those with advanced stage (HR=0.23, 95%CI:0.07-0.84; and HR=0.22, 95%CI:0.06-0.77) and EGFR mutation (HR=0.27, 95%CI:0.08-0.96; and HR=0.23, 95%CI=0.06-0.81). Moreover, a model including miR-195, miR-122 may predict survival outcomes of female patients with lung adenocarcinoma (AUC=0.707). Conclusions Circulating miR-195 and miR-122 may have prognostic values in predicting the overall survival as well as predicting EGFR mutation status in non-smoking female patients with lung adenocarcinoma. Measuring plasma levels of miR-195 and miR-122 may especially be useful in EGFR mutant patients with lung adenocarcinoma.

Circulating Unsaturated Fatty Acids Delineate the Metabolic Status of Obese Individuals

Background Obesity is not a homogeneous condition across individuals since about 25–40% of obese individuals can maintain healthy status with no apparent signs of metabolic complications. The simple anthropometric measure of body mass index does not always reflect the biological effects of excessive body fat on health, thus additional molecular characterizations of obese phenotypes are needed to assess the risk of developing subsequent metabolic conditions at an individual level. Methods To better understand the associations of free fatty acids (FFAs) with metabolic phenotypes of obesity, we applied a targeted metabolomics approach to measure 40 serum FFAs from 452 individuals who participated in four independent studies, using an ultra-performance liquid chromatograph coupled to a Xevo G2 quadruple time-of-flight mass spectrometer. Findings FFA levels were significantly elevated in overweight/obese subjects with diabetes compared to their healthy counterparts. We identified a group of unsaturated fatty acids (UFAs) that are closely correlated with metabolic status in two groups of obese individuals who underwent weight loss intervention and can predict the recurrence of diabetes at two years after metabolic surgery. Two UFAs, dihomo-gamma-linolenic acid and palmitoleic acid, were also able to predict the future development of metabolic syndrome (MS) in a group of obese subjects. Interpretation These findings underscore the potential role of UFAs in the MS pathogenesis and also as important markers in predicting the risk of developing diabetes in obese individuals or diabetes remission after a metabolic surgery.

Genome-wide association study of endometrial cancer in E2C2

Endometrial cancer (EC), a neoplasm of the uterine epithelial lining, is the most common gynecological malignancy in developed countries and the fourth most common cancer among US women. Women with a family history of EC have an increased risk for the disease, suggesting that inherited genetic factors play a role. We conducted a two-stage genome-wide association study of Type I EC. Stage 1 included 5,472 women (2,695 cases and 2,777 controls) of European ancestry from seven studies. We selected independent single-nucleotide polymorphisms (SNPs) that displayed the most significant associations with EC in Stage 1 for replication among 17,948 women (4,382 cases and 13,566 controls) in a multiethnic population (African America, Asian, Latina, Hawaiian and European ancestry), from nine studies. Although no novel variants reached genome-wide significance, we replicated previously identified associations with genetic markers near the HNF1B locus. Our findings suggest that larger studies with specific tumor classification are necessary to identify novel genetic polymorphisms associated with EC susceptibility.