Search

Supplementary methods. A document with supplementary materials, including the following: (1) subjects: cohort details; lifestyle, anthropometrics, and biochemical measurements; and outcome definition; (2) laboratory methods: EPICOR sample preparation; discovery phase: Illumina Human450K Methylation Assay; replication phase on EPIC-NL sample: Sequenom MassARRAY; and (3) supplementary statistical methods: case-control differential methylation; removal of technical biases; DNA methylation and MI risk; DNA methylation and time to disease (TTD); supplementary references. (DOCX 73 kb)

Supplemental Tables S1, S2, S3, and S4. Table S1. top 6 genic DMRs in EPICOR MI overall cases vs controls. Table S2A. details of the ZBTB12-DMR CpGs in EPICOR subjects. Table S2B. details of the ZBTB12-DMR CpGs in EPIC-NL subjects. Table S3A. EPIC-NL MI risk, adjusted models. Table S3B. EPICOR, adjusted models. Table S4A. EPICOR case-control differential methylation analysis: comparison of models with and without lipids adjustment. Table S4B. EPIC-NL case-control differential methylation analysis: comparison of models with and without lipids, batch, and WBCs adjustments. (DOCX 55 kb)

INTRODUCTION: population-based survival analyses are fundamental to assess the impact of public health interventions and new therapies in cancer control. This monograph updates previous reports on cancer patient survival in Italy up to the year 2007. MATERIAL AND METHODS: we extracted from the Network of Italian Cancer Registries (AIRTUM) database over 1,490,000 records of tumours diagnosed during 1990-2007 and followed up to the end of 2008, including all multiple tumours. We used the Ederer II method to estimate relative survival (RS) for 29 different types of neoplasm. Five-year relative survival rates were analysed by gender and macroarea. Trends in 5-, 10- and 15-year RS were studied by gender over six 3-year diagnostic periods, from 1990 to 2007. Conditional 5-year RS was also computed by gender and macroarea. Hybrid approaches were applied to exploit the recent survival experiences of cases diagnosed up to 2007. Adjustment for age was performed using EUROCARE weights. Additional sections describe cancer patient survival in childhood and in elderly patients and provide a comparison of cancer patient survival rates in Italy with those of other countries. RESULTS: Standardized 5-year RS for all tumours but skin in 52% for men and 61% for women. Patient survival has improved for almost all types of cancer: from 1990 to 2007 5-year RS has increased by 15% for all cancers but skin; the exceptions are some cancers with poor prognosis, where patient survival has remained basically unchanged. In males, RS was usually lower than in females, but trend analysis shows that the gap is narrowing. We also report persisting lower RS in southern Italy: 5-year RS in the South is usually from 4% to 10% lower than in the North and Centre. CONCLUSION: this study provides valuable information for all stakeholders in cancer control, both in Italy and elsewhere. Increasing survival reflects improvements in various areas of cancer control. On the other hand, delayed diagnosis and suboptimal management are consistent with the reported differences in survival within the country.

This is an update of incidence and mortality cancer data provided by the Italian Network of Cancer Registry (AIRTUM) relative to the period 2003-2005.AIRTUM is a network of general and specialized population-based cancer registries that covers about 1/3 of the Italian resident population (www.registri-tumori.it). Incidence and mortality data for the period 2003-2005 are based on 20 Registries. The five most frequently diagnosed cancers were: - prostate (18.5%), non melanoma skin (15.8%), lung (13.1%), colorectal (12.0%), bladder (5.7%) among males; - breast (24.9%), non melanoma skin (15.1%), colorectal (11.9%), lung (5.0%) and stomach (4.1%) among females. In the same period the most frequent causes of cancer death were: - cancer of the lung (27.6%), colorectal (10.7%), prostate (8.5%), stomach (7.3%) and liver (6.1%) among males; - breast cancer (16.3%), colorectal (11.9%), lung (10.3%), stomach (7.2%) and pancreas (6.5%) among females. According to the age-specific incidence rates one man and one woman every two will receive a cancer diagnosis during his/hers life (from birth to the age of 84 years). From 1993-1995 to 2003-2005, overall crude cancer incidence rate (males and females together) increased from 555.4 to 654.8 x 100,000. Standardization showed that 63% of this increase was due to ageing of the population. Moreover, most of the residual increase was among those cancer sites (breast, prostate, colorectal, thyroid and melanoma) for which early detection may have played a relevant role in anticipating (and therefore increasing) the number of diagnoses. Due to population ageing also overall cancer mortality did not show any decrease when crude rates were compared. On the contrary, standardized mortality rates (all cancers together) showed a strong decrease (311.4 vs. 266.5 x 100.000). The risk of receiving a diagnosis or dying because of cancer is still lower in residents in the regions of the South of Italy than in those of Central and Northern Italy, but they are becoming more and more similar. In Italy cancer incidence and mortality rates are similar to those in northern European countries and in USA among males, but they are still lower for women.

This collaborative study, based on data collected by the network of Italian association of cancer registries (AIRTUM), provides updated estimates on the incidence risk of multiple primary cancer (MP). The objective is to highlight and quantify the bidirectional associations between different oncological diseases. The quantification of the excess or decreased risk of further cancers in cancer patients, in comparison with the general population, may contribute to understand the aetiology of cancer and to address clinical follow-up.Data herein presented were provided by AIRTUM population-based cancer registries, which cover nowadays 48% of the Italian population. This monograph utilizes the AIRTUM database (December 2012), considering all malignant cancer cases diagnosed between 1976 and 2010. All cases are coded according to ICD-O-3. Non-melanoma skin cancer cases, cases based on death certificate only, cases based on autopsy only, and cases with follow-up time equal to zero were excluded. To define multiple primaries, IARC-IACR rules were adopted (http://www.iacr.com.fr/MPrules_july2004.pdf). Data were subjected to standard quality control procedures (described in the AIRTUM data management protocol) and specific quality control checks defined for the present study. A cohort of cancer patients was followed over time from first cancer diagnosis until the date of second cancer diagnosis, death, or the end of follow-up, to evaluate whether the number of observed second cancer cases was greater than expected. Person years at risk (PY) were computed by first cancer site, geographic area (North, Centre, South and Islands), attained age, and attained calendar-year group. All second cancers diagnosed in the cohort's patients were included in the observed numbers of cases. The expected number of cancer cases was computed multiplying the accumulated PY by the expected rates, calculated from the AIRTUM database stratified by cancer site, geographic area, age, and calendar-year group. The Standardized Incidence Ratio (SIR) was calculated as the ratio of observed to expected cancer cases. The Excess Absolute Risk (EAR) beyond the expected amount were calculated subtracting the expected number of subsequent cancers from the observed number of cancer cases; the difference was then divided by the PY and the number of cancer cases in excess (or deficit) was expressed per 1,000 PY. Confidence intervals were stated at 95%. The two months (60 days) after first cancer diagnosis were defined as "synchronicity period", and in the main analysis observed and expected cases during this period were excluded. It was estimated the excess risk in the period after first diagnosis (≥ 0 months), excluding the synchronicity period (≥ 2 months), and during the following periods: 2-11, 12-59, 60-119 and 120 months after diagnosis. First-cancer-site-and-gender-specific sheets are presented, reporting both SIRs and EARs.For 5,979,338 person-years a cohort of 1,635,060 cancer patients (880,361 males and 754,699 females) diagnosed between 1976 and 2010 was followed. The mean follow-up length was 14 years. Overall, 85,399 metachronous (latency ≥2 months) cancers were observed, while 77,813 were expected during the study period: SIR: 1.10 (95%CI 1.09-1.10), EAR: 1.32 x 1,000 person-years (95%CI 1.19 - 1.46). The SIR was 1.08 (95%CI 1.08-1.09) for men (54,518 observed and 50,260 expected) and 1.12 (95%CI 1.11-1.13) for women (30,881/27,553), and the EAR 1.61 (95%CI 1.37-1.84) and 1.08 x 1,000 person-years (95%CI 0.93-1.24), respectively.Moreover, during the first two months after first cancer diagnosis (synchronous period) 14,807 cancers were observed while 3,536 were expected (SIR: 4.16; 95%CI 4.09-4.22); the SIR was 4.08 (95%CI 4.00-4.16) for men and 4.32 (95%CI 4.20-4.45) for women.The mean age of patients at first cancer diagnosis was 67.0 years among males and 65.8 among females.The risk of MP was related to age being higher for younger patients and lower for older ones. In relation to the time of first cancer diagnosis, the SIR was very high at the beginning and then decreased, although remaining constantly over 1, and then rose over time. No strong differences were evident across the different incidence periods, which all showed an increased MP risk.Women had higher SIRs than expected for 18 cancer sites, men for 12. The statistically significantly SIRs lower than 1 were 2 and 8, respectively. Increased overall MP risk was observed for patients of both sexes with a first primary in the oral cavity (SIR men: 1.93; SIR women: 1.48), pharynx (SIR men: 2.13; SIR women: 1.99), larynx (SIR men: 1.57; SIR women: 1.79), oesophagus (SIR men: 1.45; SIR women: 1.41), lung (SIR men: 1.09; SIR women: 1.13), kidney (SIR men: 1.14; SIR women: 1.15), urinary bladder (SIR men: 1.29; SIR women: 1.22), thyroid (SIR: 1.22 in both sexes), Hodgkin lymphoma (SIR men: 1.59; SIR women: 1.94), and non-Hodgkin lymphoma (SIR men: 1.13; SIR women: 1.12), and for the heterogeneous group "other sites" (SIR men: 1.09; SIR women: 1.07). Moreover, men had a higher MP risk if the first cancer was in the testis (SIR: 1.24), while the same was true for women with gallbladder (SIR: 1.21), skin melanoma (SIR: 1.17), bone (SIR: 1.41), breast (SIR: 1.12), cervix uteri (SIR: 1.23) and corpus uteri (SIR: 1.23), and ovarian cancer (SIR: 1.18). On the contrary, a first liver or pancreas cancer were associated with a decreased MP risk in both sexes (liver SIR: 0.86 and 0.81 for men and women, respectively; pancreas SIR: 0.70 and 0.78 for men and women, respectively), as were those of colon (SIR: 0.93), rectum (SIR: 0.83), gallbladder (SIR: 0.80), prostate (SIR: 0.93), mesothelioma (SIR: 0.65), and central nervous system (SIR: 0.82) among men. Among the cancers for which the EAR is statistically significant, those with higher Excess Absolute Risk of MP were those of the oral cavity (EAR: 16.0 x 1,000 person-years in men and 5.4 in women), pharynx (17.6 and 9.1), larynx (11.4 and 8.8), and oesophagus (8.5 and 4.8).This descriptive study provides quantitative information on the risk of developing a second cancer in an Italian population-based cohort of approximately 1.65 million cancer patients, compared to the risk of the general population. During the follow-up time (on average 14 years) cancer patients had an MP risk that was 10% higher in comparison to the general population and an Excess Absolute Risk of 1.32 x 1,000 person-years. Study of MPs and their risk measures are dependent on methods used in the calculation. The definition of MP is not univocal and using different rules can greatly change the number of cancers in a patient with MPs. However, the AIRTUM cancer registries adopt the same recommendations for MP definition. This monograph was therefore made possible by the shared rules and standards used by AIRTUM registries. The cancer site-specific sheets, which represent the core of the monograph, can be useful to highlight and quantify the bidirectional associations among different diseases and therefore provide indications for clinical follow-up. Lifestyle changes in more healthful directions can have a positive effect in the cancer patient population and should always be recommended.

OBJECTIVES: the aim of the present monograph is to update the estimation of the number of people living with cancer in Italy, to describe geographic variability, and estimate the number of long-term survivors, i.e., people living five years or more after a cancer diagnosis. MATERIALS AND METHODS: the study included the data of the AIRTUMdatabase. Twenty-four Cancer Registries (CRs) (covering 27% of the Italian population) collected information on the incidence and vital status of 1,275,353 cases diagnosed between 1978 and 2005. For each CR, the observed prevalence was calculated up to the maximum observable duration. To estimate the complete prevalence (all living patients, independently from time since diagnosis) and the prevalence for lengths of time exceeding the CR maximum duration of registration, the observed prevalence was corrected through a completeness index. Completeness indices, gender, age and site specific, were estimated by means of statistical regression models using cancer incidence and survival data available from CRs with more than 15 years of observation. As of 1 January 2006, the prevalence was estimated (as absolute numbers and as a proportion per 100,000 inhabitants) for 46 cancer sites, by gender, age class, years since diagnosis and geographic areas. RESULTS: as of 2006, 2,244,000 persons (4%of the Italian population) were alive with a cancer diagnosis. A relevant geographic variability emerged, with proportions between 4%-5% among CRs in the Centre and North of Italy, and proportions between 2%-3% in the South. Forty-four percent of prevalent subjects (988,000) were males and 56% (1,256,000) females. Fifty-seven percent (1,285,680 people, 2.2% of total population) of these patients was represented by long-term survivors. In patients aged 75 years or more, the proportions of prevalent cases were 19%in males and 13%in females, and 10%between 60 and 75 years of age in both genders.More than half a million Italian women were alive with a breast cancer diagnosis (42%of women with a neoplasm), followed by women with cancers of the colonrectum (12%), corpus uteri (7%), thyroid (5%), and cervix uteri (4%). In men, 22%of prevalent cases (216,716) included patients with prostate cancer, 18% with bladder cancer, and 15%with colon-rectum cancer. Percentages of long-term survivors higher than 70% were reported for cancers of the cervix uteri (82% at five years, and 55% at 15 years from diagnosis), Hodgkin lymphoma, testis, brain and central nervous system, bone and connective tissue. Many patients with these types of cancers (often occurring in young people) can be considered "cured", i.e., with a life expectancy overlapping that of the general population.The estimated proportions of prevalent cases emerging from this study in Italy were quite similar to those reported in Northern Europe, but at least 15%lower than those in the United States. CONCLUSIONS: in 2006, the number of prevalent cases nearly doubled compared to 1992. The increase over time in the proportion of elderly patients, related to population ageing, requires adequate health policies. Knowing the number of people alive many years after cancer diagnosis (either cured or long-term survivors) provides the scientific bases for the definition of health policies focusing on them. Furthermore, it promotes the conduction of studies aimed at improving the present knowledge on the quality of life of these patients during and after the active phase of treatments, in addition to studies on the long-term effects of treatments.

Objectives: The effect of modifiable stroke risk factors in terms of prevented cases remains unclear due to sex-specific disease rate and risk factors prevalence. Our aim was to estimate their impact on stroke by gender through population-attributable fraction (PAF), preventive fraction (PF) and their combination in EPIC-Italian cohort. Methods: 43,976 participants, age 34–75, and free of cardiovascular disease at baseline (1993–1998) were followed up for almost 11 years. Adjusted hazard ratios and PAF were estimated using Cox models. Results: We identified 386 cases. In males, the burden for stroke was 17% (95% CI 4–28%) for smoking and 14% (95% CI 5–22%) for alcohol consumption. In females, hypertension was carrying the biggest burden with 18% (95% CI 9–26%) followed by smoking 15% (95% CI 7–22%). Their combination was 46% (95% CI 32–58%) in males and 48% (95% CI 35–59%) in females. PF for current smokers was gender unequal [males 21% (95% CI 15–27%) females 9% (95% CI 1–17%)]. Conclusions: Half of strokes are attributable to potentially modifiable factors. The proportion of prevented cases is gender unbalanced, encouraging sex-specific intervention.

Background DNA methylation profiles are responsive to environmental stimuli and metabolic shifts. This makes DNA methylation a potential biomarker of environmental-related and lifestyle-driven diseases of adulthood. Therefore, we investigated if white blood cells’ (WBCs) DNA methylation profiles are associated with myocardial infarction (MI) occurrence. Whole-genome DNA methylation was investigated by microarray analysis in 292 MI cases and 292 matched controls from the large prospective Italian European Prospective Investigation into Cancer and Nutrition (EPIC) cohort (EPICOR study). Significant signals (false discovery rate (FDR) adjusted P

The crosstalk between microRNAs (miRNAs) and other epigenetic factors may lead to novel hypotheses about carcinogenesis identifying new targets for research. Because a single miRNA can regulate multiple downstream target genes, its altered expression may potentially be a sensitive biomarker to detect early malignant transformation and improve diagnosis and prognosis. In the current study, we tested the hypothesis that altered methylation of miRNA encoding genes, associated with deregulated mature miRNA expression, may be related to dietary and lifestyle factors and may contribute to cancer development. In a case-control study nested in a prospective cohort (EPIC-Italy), we analysed DNA methylation levels of miRNA encoding genes (2191 CpG probes related to 517 genes) that are present in the Infinium Human Methylation450 BeadChip array in prediagnostic peripheral white blood cells of subjects who developed colorectal cancer (CRC, n = 159) or breast cancer (BC, n = 166) and matched subjects who remained clinically healthy. In the whole cohort, several differentially methylated miRNA genes were observed in association with age, sex, smoking habits and physical activity. Interestingly, in the case-control study, eight differentially methylated miRNAs were identified in subjects who went on to develop BC (miR-328, miR-675, miR-1307, miR-1286, miR-1275, miR-1910, miR-24-1 and miR-548a-1; all Bonferroni-adjusted P < 0.05). No significant associations were found with CRC. Assuming that altered methylation of miRNAs detectable in blood may be present before diagnosis, it may represent a biomarker for early detection or risk of cancer and may help to understand the cascade of events preceding tumour onset.

Background The relationship between coffee consumption and coronary heart disease (CHD) has been investigated in several studies with discrepant results. We examined the association between Italian-style (espresso and mocha) coffee consumption and CHD risk. Methods We investigated 12,800 men and 30,449 women without history of cardiovascular disease recruited to the EPICOR prospective cohort study. Coffee consumption was assessed at baseline. In a random sub-cohort of 1472 subjects, plasma triglycerides, and total, LDL and HDL cholesterol were determined to investigate the effect of coffee consumption on plasma lipids. Results After a mean follow up of 10.9 years, 804 cases of CHD (500 acute events, 56 fatal events and 248 revascularizations, all first events) were identified. Multivariable adjusted hazard ratios for CHD were: 1.18 (95% CI 0.87–1.60) for drinking 1–2 cups/day, 1.37 (95% CI 1.03–1.82) for >2–4 cups/day and 1.52 (95% CI 1.11–2.07) for over 4 cups/day (P trend

Background Colorectal cancer is the third most common cancer worldwide. Diet has been hypothesized as involved in colorectal cancer etiology, but few studies on the influence of total dietary antioxidant intake on colorectal cancer risk have been performed. Methods We investigated the association between colorectal cancer risk and the total antioxidant capacity (TAC) of the diet, and also of intake of selected antioxidants, in 45,194 persons enrolled in 5 centers (Florence, Naples, Ragusa, Turin and Varese) of the European Prospective Investigation into Cancer and Nutrition (EPIC) Italy study. TAC was estimated by the Trolox equivalent antioxidant capacity (TEAC) assay. Hazard ratios (HRs) for developing colorectal cancer, and colon and rectal cancers separately, adjusted for confounders, were estimated for tertiles of TAC by Cox modeling, stratifying by center. Results Four hundred thirty-six colorectal cancers were diagnosed over a mean follow-up of 11.28 years. No significant association between dietary TAC and colorectal cancer incidence was found. However for the highest category of TAC compared to the lowest, risk of developing colon cancer was lower (HR: 0.63; 95% CI: 0.44–0.89, P trend: 0.008). By contrast, increasing TAC intake was associated with significantly increasing risks of rectal cancer (2nd tertile HR: 2.09; 95%CI: 1.19–3.66; 3rd tertile 2.48 95%CI: 1.32–4.66; P trend 0.007). Intakes of vitamin C, vitamin E, and ß-carotene were not significantly associated with colorectal cancer risk. Conclusions Further prospective studies are needed to confirm the contrasting effects of high total antioxidant intake on risk of colon and rectal cancers.

Background and Purpose— Tissue factor (TF) expression is increased in inflammatory atherosclerotic plaques and has been related to their thrombogenicity. Blood-borne TF has been also demonstrated to contribute to thrombogenesis. However, few studies have evaluated the association of circulating levels of TF with stroke. We investigated the association of baseline circulating levels of TF with stroke events occurred in the European Prospective Investigation into Cancer and Nutrition-Italy cohort. Methods— Using a nested case–cohort design, a center-stratified random sample of 839 subjects (66% women; age range, 35–71 years) was selected as subcohort and compared with 292 strokes in a mean follow-up of 9 years. Blood samples were collected at baseline in citrate, plasma was stored in liquid nitrogen and TF was measured by ELISA (IMUBIND, TF ELISA, Instrumentation Laboratory, Milan, Italy). The odd ratios and 95% confidence intervals, adjusted by relevant confounders (covariates of TF) and stratified by center, were estimated by a Cox regression model using Prentice method. Results— Individuals in the highest compared with the lowest quartile of TF plasma levels had significantly increased risk of stroke (odds ratio IVvsI quartile , 2.01; 95% confidence interval, 1.25–3.23). The association was independent from several potential confounders (odds ratio IVvsI quartile , 1.91; 95% confidence interval, 1.15–3.19). No differences were observed between men and women. The increase in risk was restricted to ischemic strokes (odds ratio IVvsI quartile , 2.13; 95% confidence interval, 1.10–4.12; fully adjusted model), whereas high levels of TF were not associated with the risk of hemorrhagic stroke (odds ratio IVvsI quartile , 1.12; 95% confidence interval, 0.49–2.55; fully adjusted model). Conclusions— Our data provide evidence that elevated levels of circulating TF are potential risk factors for ischemic strokes.

Background Metabolic syndrome (defined as at least three among abdominal obesity, high blood triglycerides, low high-density lipoprotein cholesterol, high blood glucose, and high blood pressure) is emerging as a risk factor for breast cancer; however few studies – most confined to postmenopausal women – have investigated associations between breast cancer risk and metabolic syndrome. The purpose of this study was to examine the association between metabolic syndrome and its components, and risk of breast cancer in postmenopausal and premenopausal women. Methods We performed a case-cohort study on 22,494 women recruited in 1993-1998 to four Italian centres (Turin, Varese, Naples, Ragusa) of the European Prospective Investigation into Cancer and Nutrition (EPIC) and followed-up for up to 15 years. A random subcohort of 565 women was obtained and 593 breast cancer cases were diagnosed. Hazard ratios (HR) with 95% confidence intervals (CI), adjusted for potential confounders, were estimated by Prentice-weighted Cox proportional hazards models. Results Presence of metabolic syndrome was associated with significantly increased breast cancer risk in all women (HR 1.52, 95%CI 1.14-2.02). When the analyses were repeated separately for menopausal status, the association was limited to postmenopausal women (HR 1.80, 95%CI 1.22-2.65) and absent in premenopausal women (HR 0.71, 95%CI 0.43-1.16); P for interaction between metabolic syndrome and menopausal status was 0.001. Of metabolic syndrome components, only high blood glucose was significantly associated with increased breast cancer risk in all women (HR 1.47, 95%CI 1.13-1.91) and postmenopausal women (HR 1.89, 95%CI 1.29-2.77), but not premenopausal women (HR 0.80, 95%CI 0.52-1.22; P interaction=0.004). Conclusions These findings support previous data indicating that metabolic syndrome is an important risk factor for breast cancer in postmenopausal women, but not in premenopausal women, and suggest that prevention of metabolic syndrome through lifestyle changes could confer protection against breast cancer.