Search

Epidemiologic studies linking high serum iron with cancer risks are limited and inconclusive, despite evidence implicating body iron in human carcinogenesis. A cohort of 309,443 adults in Taiwan who had no history of cancer had serum iron levels tested at the time of recruitment (1997–2008). Initially measured iron levels were associated with subsequent cancer risk by linking individuals with the National Cancer Registry and National Death File. HRs were calculated by the Cox model. One third of males (35%) and one fifth of females (18%) had high serum iron (≥120 μg/dL), which was associated with a 25% increase in risk for incidence of all cancers [HR, 1.25; 95% confidence interval (CI), 1.16–1.35] and with a 39% increase in risk for mortality from all cancers (HR, 1.39; 95% CI, 1.23–1.57). The relationship between serum iron and cancer risk was a J-shaped one, with higher cancer risk at both ends, either at lower than 60 μg/dL or higher than 120 μg/dL. At the higher end, cancer risk increased by 4% for every 10 μg/dL increment above 80 μg/dL, showing a dose–response relationship, with 60 to 79 μg/dL as a reference level. In a sensitivity analysis, the increases in risk were still observed after the first 5 years of cancer cases were excluded. Liver cancer risk was increased in HBV (−) non-hepatitis B carrier (3-fold) and HBV (+) hepatitis B carrier (24-fold). Lifestyle risks such as smoking, drinking, or inactivity interacted synergistically with high serum iron and significantly increased the cancer risks. The liver (HR, 2.49; 95% CI, 1.97–3.16) and the breast (HR, 1.31; 95% CI, 1.01–1.70) were the two major cancer sites where significant cancer risks were observed for serum iron either ≥120 μg/dL or ≥140 μg/dL, respectively. This study reveals that high serum iron is both a common disorder and a marker of increased risk for several cancers. Cancer Res; 74(22); 6589–97. ©2014 AACR.

Background: Previous lung cancer models were mostly based on case control studies and used limited epidemiological risk factors. The clinical utility was hampered due to lack of added value beyond smoking history. In this study, we developed a series of lung cancer risk prediction models using extensive data, including lab testing data, from a large prospective cohort in Taiwan. In particular, we developed a risk prediction model within heavy smokers with ≧30 pack years that could further stratify this group and improve the screening efficiency of Low Dose Computed Tomography (LDCT). Methods: The cohort participants were divided into equal training and validation sets, by using variables from personal history and standard lab tests. Cox proportional hazards regression and goodness of fit were used to generate area under the curve (AUC). Results: A total of 1,774 incident lung cancer cases were identified among 505,030 participants recruited in a 14-year period during 1994-2007. We developed a “History only” model that included smoking status, body mass index (BMI), family history, and personal history; a “lab only” model that included blood in sputum, lung function test and serum biomarkers like carcinoembryonic antigen (CEA), bilirubin, and alpha fetoprotein (AFP); and an integrative model with all these variables. The AUC for “history only” model was 0.837, “lab only” model, 0.839, and integrative model, 0.862. We also developed integrative models for ever-smokers and never-smokers with AUCs of 0.872 and 0.801, respectively. Furthermore, since LDCT is recommended for heavy smokers with ≥ 30 smoking pack-years but with a high false positive rate, we developed a risk model in this group to further stratify this group. A total of 454 incident lung cancer cases were identified among 23,582 heavy smokers. Six variables are significantly associated with lung cancer risk: age, smoking status (current or former), smoking intensity (number of packs smoked in a day), BMI, lung function test and CEA. Current smokers, who smoked >1 pack per day, with BMI Conclusion: Routinely collected lab data can significantly improve lung cancer risk prediction models. An integrative lung cancer risk prediction model among heavy smokers with ≧30 pack years can stratify subjects with a 3-4 fold difference in lung cancer risk for LDCT screening. The model found decreased lung function and high CEA values, along with detailed smoking history and low BMI, offered a unique predictive ability to improve the efficiency of LDCT screening. Citation Format: Xifeng Wu, Chi Pang Wen, Yuanqing Ye, MinKwang Tsai, Xia Pu, Wong-Ho Chow, Chad Huff, Sonia Cunningham, Maosheng Huang, Shiuan Bei Wu, Chwen Keng Tsao, Jian Gu. The use of lung cancer risk prediction model to stratify heavy smokers: The power of routinely collected lab data. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr LB-298. doi:10.1158/1538-7445.AM2014-LB-298

Background. Lung cancer is the leading cause of cancer death in the United States. Cost-effective screening and early detection is essential for reducing lung cancer mortality. Serum metabolites are emerging as promising biomarkers for improving risk prediction and early detection of lung cancer. Methods. In the discovery phase, we performed global, unbiased metabolomic profiling in serum samples from 20 healthy controls, 20 early stage, and 20 late stage lung cancer patients who were matched on age and gender. We then selected three differentially expressed metabolites for validation in two additional case control populations, followed by final validation of one metabolite in a large Taiwanese prospective cohort of 435,985 individuals. Results. A total of 403 named metabolites were identified by metabolomic profiling in the discovery phase, with 306 metabolites remaining after initial quality assurance. Of these, 29 metabolites exhibited a significant trend when comparing normal controls, early stage and late stage lung cancer cases. Three of the top differentially expressed metabolites, gamma-glutamylalanine, bilirubin and allantoin, were selected and validated in two additional sets of subjects, one set comprised of 50 controls, 50 early stage and 50 late stage lung cancer cases, and the second comprised of 123 controls, 123 early stage and 123 late stage lung cancer cases. The most promising metabolite, bilirubin was further validated in a large prospective cohort. In this cohort, the incidence rate of lung cancer per 10,000 person-years in men was 3.73 (95% confidence interval [CI], 3.13-4.43) in the highest quartile of bilirubin level compared to 7.02 (95% CI, 6.16-7.99) in the lowest quartile, which translates into a 38% increase in lung cancer incidence for the low bilirubin group (P=0.003). The corresponding lung cancer specific mortality rate in men was 2.46 (95% CI, 2.02-3.00) in the highest quartile compared to 4.84 (95% CI, 4.19-5.60) in the lowest quartile, a 53% increase in lung cancer specific mortality for the low bilirubin group (P Conclusions. Smokers with low bilirubin levels had significantly increased risks for both lung cancer incidence and mortality. Serum bilirubin is a valuable biomarker for lung cancer risk prediction in ever-smokers. Citation Format: Fanmao Zhang, Chi Pang Wen, Dong Liang, Heath Skinner, Jian Gu, Wong-Ho Chow, Yuanqing Ye, Michelle A.T. Hildebrandt, Xia Pu, Maosheng Huang, Min Kuang Tsai, Chwen Keng Tsao, Xifeng Wu. Metabolomic profiling identifies bilirubin as a novel serum marker for lung cancer. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr LB-27. doi:10.1158/1538-7445.AM2013-LB-27

Background: Attention has been focused on the increased risk of high cholesterol, but the effect of low cholesterol remains unclear and receives little attention. Method: The study cohort, 537,430 adults in Taiwan who went through a medical screening program between 1996 and 2008, was followed up for an average of 8.5 years for mortality and cancer incidence, with 18,745 deaths and 17,714 cancer cases identified. Low and very low cholesterol or LDL were defined as < 180 and Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-324. doi:1538-7445.AM2012-LB-324

Models predicting liver cancer are only available for the high risk population with hepatitis carriers but not for the general population. we developed risk prediction models for liver cancer using data collected from a large health screening program in Taiwan with long term follow up. The cohort included 444,023 subjects at baseline. During a median follow up time of 7.4 years, 1,668 subjects developed liver cancer. Epidemiologic data, medical history and routine blood panel including serum transaminases data were collected with optional additional testing for HBV or HCV status. Liver cancer cases were ascertained by computerized record linkage with both the National Cancer Registry and the National Death Certification profiles in Taiwan. Stepwise Cox regression analysis was performed to identify significant predictors in the multivariate models. Individualized risk of developing liver cancer in 10 years was calculated from baseline probability and relative risk profile estimated from the Cox regression model. Models were developed separately to provide risk prediction for subjects who chose to have HCV tested (130,533 subjects at baseline and 416 liver cancer cases) and who chose otherwise (313,490 subjects and 1,252 liver cancer cases). Since the results for both sub-cohorts are comparable and we only reported results for the sub-cohort with HCV tested. In the sub-cohort with HCV tested, the model with only questionnaire data identified significant main effects for male gender, older age, prior history of diabetes, pack year of smoking, alcohol use, and physical inactivity. This epidemiologic model had an AUC of 0.798 (95% CI=0.772-0.815) for 10-year risk prediction. With only data from measures of Transaminases, the model achieved an AUC of 0.927 (95% CI=0.911-0.938), a significant increase as compared to the model with only epidemiologic variables. The addition of HBV improved the AUC to 0.936 (95% CI=0.913-0.958). Similar results were obtained for sub-cohort without HCV testing. The addition of HCV status in combination with HBV status, the AUC improve to 0.941 (95% CI=0.918-0.967). Among models using key history and blood panels, the use of serum Transaminases only, available in routine health check-ups, was able to provide most information needed in predicting liver cancer in the general public. Additional testing of HBV and /or HCV further increased the prediction power. This simple transaminase-based model for the general public can be valuable in clinical practice for identifying high risk individuals. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 5509. doi:1538-7445.AM2012-5509