Determinants of Cancer Stage at Diagnosis

Karen Kopciuk

Differences in individual, health system and clinical factors, such as awareness of cancer, health habits, diagnostic delay, screening, stage, comorbidity and access to care, are all potential explanations of differences observed within Canada in cancer survival and mortality. Predictive models designed to support decision making following cancer diagnosis often account for diverse patient risk factors. However, few tools for predicting cancer outcomes include information that may influence probability of early detection such as co-morbidity, family history, lifestyle risk factors, access to screening, or occupational history.

This project will explore patient and system factors associated with stage of cancer at diagnosis, using data from Alberta’s Tomorrow Project (ATP). Nearly 2500 ATP participants have been diagnosed with cancer since joining ATP. Linkage of ATP data with administrative databases will provide screening information at multiple time points, as well as diagnostic information.

The specific aims are to:

1. Develop multivariable regression models using ordinal logistic regression to identify independent and synergistic factors associated with cancer stage at diagnosis;
2. Explore reductions in stage at diagnosis that may be possible by modifying these identified factors (individual and health care system) in simulation studies; and
3. Validate the models using data from the BC Generations Project.

The comprehensive modeling of factors associated with incident cancers in ATP will provide information simply not attainable in other provincial databases. Validation of Alberta results with B.C. data will ensure the methodology performs as expected, while providing additional support for translating the findings into cancer screening practices in Alberta and beyond. Using a simplistic framework of delayed stage at diagnosis, we will identify factors that can be used by screening programs to identify individuals who may benefit from individualized screening practices or from targeted prevention messages, thereby increasing the proportion of cases diagnosed at earlier stages.

Effects of Diet and Exercise in the Years Prior to Cancer Diagnosis on Treatment Outcomes after Cancer Diagnosis

Winson Cheung

A growing body of literature exists exploring the relationship between lifestyle factors, such as diet and exercise, and cancer risk. Fruit and vegetable consumption may be protective against the development of oral cavity, pharynx, larynx, esophagus, stomach and lung (fruit only) cancer. A recent study suggested that approximately 2% of new cancer cases in Alberta in 2010, representing around 290 cases, could be attributed to insufficient fruit and vegetable consumption1. Physical activity has also been linked to decreased risk of cancer. A meta-analysis found that cancer risk was reduced in people who engaged in the highest levels of leisure time physical activity2. Despite growing research into how diet and physical activity influence the development of cancer, limited research exists as to whether lifestyle factors prior to a cancer diagnosis may influence treatment outcomes after a cancer diagnosis.

Some research exists as to how diet and exercise affect cancer recurrence. However, the majority of these studies examined diet and exercise near or after the diagnosis of cancer. One retrospective study showed that higher levels of physical activity one year prior to diagnosis with breast cancer resulted in slightly better hazard ratios for overall survival. In addition, physical activity following diagnosis was found to slow progression of cancer and reduce the risk of secondary life threatening disease. It is plausible that in addition to reducing primary cancer risk, a history of exercise long before a cancer diagnosis could potentially lead to improved disease outcomes in patients newly diagnosed with cancer.

Through linking data from Alberta’s Tomorrow Project (ATP) with data from the Cancer Measurement Outcomes Research and Evaluation (C-MORE) group in Alberta, the relationship between lifestyle factors in the years prior to diagnosis and outcomes following treatment can be assessed.

Development and Validation of a Lung Cancer Risk Prediction Model in Alberta’s Tomorrow Project

Alain Tremblay

Lung cancer is the leading cause of cancer death in Canada and around the world. Early detection and treatment of lung cancer through screening seem to be the most promising strategy to reduce lung cancer mortality. The Canadian Task Force on Preventative Health Care (CTF) has recently published recommendations in favour of low-dose computed tomography (LDCT) screening for high risk individuals as defined in the National Lung Screening Trial (NLST) (i.e., ever smokers aged 55 to 74 years, ≥30 pack-years of smoking and <15 years since quitting). Nevertheless, a risk prediction model based on Prostate, Lung, Colorectal and Ovarian Cancer Screening (PLCO) data has been developed and suggested that use of an accurate risk prediction model for selecting individual for lung cancer screening is more efficient than applying the NLST criteria. Therefore, we aim to validate this lung cancer risk prediction model and compare its performance vs. NLST criteria in a Canadian population to inform optimal inclusion criteria for a clinical screening program using Alberta’s Tomorrow Project (ATP) dataset.

Investigating the role of Dietary Intakes of Red and Processed Meat, Vegetables and Fruit, and Fiber on Cancer Incidence in a Large Cohort of Adults in Alberta’s Tomorrow Project

Katerina Maximova

Background: Epidemiological findings suggest that red and processed meat consumption is positively related to cancer risk and there is an important socioeconomic gradient in following a healthy diet among Canadians. Researchers have only recently begun to examine the presence of interactions between different food items and their relation to cancer risk. These observations underscore the importance of research aimed at understanding the role of simultaneous exposure to the key food items recommended for cancer prevention in order to optimize future prevention efforts to reduce cancer burden. Elucidating how socioeconomic status (SES) confers risk for cancer will inform inequalities in cancer incidence.

Study Aim: The aim of the proposed study is to examine the co-occurrence of adverse intakes of red and processed meat, vegetables and fruit, and fiber, and its impact on cancer incidence and SES inequalities in cancer risk in a large prospective cohort study.

Methods: This study will utilize data from Alberta’s Tomorrow Project a large ongoing prospective cohort of 55,000 Albertan adults aged 35-69. Dietary intake of red and processed meat, vegetables and fruit, and fiber was assessed through a past-year food frequency questionnaire. We will construct a measure of co-occurrence by summing different consumption levels (e.g., high/medium/low) of the estimated mean daily intakes of nutrients and food group servings, including the number of servings or daily intake for red and processed meat, vegetables and fruit, and fiber. Measures of socioeconomic status will be based on education level, annual household income and working status. Cancer incidence will be determined using cancer registry linkage data. Data analyses will be adjusted for gender, physical activity, weight status, current and past smoking, alcohol use, cancer screening tests, first degree family history of cancer and chronic disease, and personal history of chronic disease.

Significance: Results will provide new knowledge about the potential synergies of dietary intakes of red and processed meat, vegetables and fruit, and fiber for the development of cancer risk, and gain insights into SES disparities in this relationship.

Occupational Physical Activity and Lung Cancer Risk

Vikki Ho

Being physically active have been shown to reduce the risk of some cancers. There are different types of physical activity, including activity at work and recreational activities. Participation in recreational activities has been shown to reduce lung cancer risk. However, the role of physical activity at work in affecting the development of lung cancer is not well-established. In fact, some studies have found that people who have physically demanding jobs also have a higher risk of lung cancer. As people spend many hours at work and some jobs are very physically demanding, new studies are needed to fully understand the role of physical activity at work on lung cancer development. In this research, we will examine how lung cancer risk is associated with physical activity levels at work. Our research will be based on the Canadian Partnership for Tomorrow Project (CPTP). Our research will include all men and women who were newly diagnosed with lung cancer since CPTP began. For comparison, we will also select a subgroup of people, who did not have a cancer at the beginning of the study. The interview asked detailed questions on the longest-held job for all participants. Using this information, we will compare activity levels in the longest-held job between those who were diagnosed with lung cancer and those who remained cancer-free. This study offers a valuable opportunity to examine, in a short time frame and at low cost, whether physical activity at work plays an important role in lung cancer development. The role of physical activity (PA) at work in affecting the development of lung cancer is not well established. As people spend many hours at work and some jobs are very physically demanding, new studies are needed to fully understand the role of PA at work on lung cancer development. AIMS: To investigate the relationship between occupational PA levels and lung cancer risk. We will also explore potential effect modification by occupational lung carcinogens.

METHODOLOGY: This research will be conducted among individuals enrolled in the Canadian Partnership for Tomorrow Project (CPTP) using a case-cohort design. Cases will include all participants with an incident diagnosis of lung cancer during a follow-up period from baseline to 2016 (estimated N=1280). A sub-cohort of 5120 individuals will be randomly sampled from the entire CPTP cohort at baseline. All participants with a history of cancer (other than non-melanoma skin cancer) at baseline will be excluded. At baseline, participants of CPTP provided detailed information on their longest-held job, including job title, industry, and age at which the job started/ended. To assign metabolic equivalent of tasks (MET) to the longest-held job, we will use data generated by our team on the energy expenditures associated with almost 3600 job titles. Exposure to occupational lung carcinogen in the longest-held job will be determined by linkage with the Canadian job exposure matrix. A weighted Cox proportional hazards regression model will be used to estimate adjusted hazard ratios and 95% confidence intervals for the associations between occupational PA in the longest-held job and lung cancer risk, separately for men and women. We will also explore analyses considering the main histological types of lung cancer. Potential effect modification by exposure to occupational carcinogens will be analyzed through the inclusion of specified multiplicative interactions.

Dietary Patterns of South Asians in Alberta and their risk for Chronic Diseases

Catherine Chan

Individuals of South Asian origin (from the following countries: India, Pakistan, Bangladesh, Sri Lanka and Nepal) represent the majority of the immigrant population in Canada. Acculturation of dietary practices and adaptation of locally available foods in the diets may be challenging for many new immigrants. Further, South Asians are at risk of early onset cardio vascular diseases (CVD), pre-diabetes and type 2 diabetes (T2D). Lifestyle choices and dietary intake are modifiable risk factors that can alter the risk of developing chronic diseases, however there is limited data examining dietary practices of South Asian in North American or Canadian contest. Thus, the objective of the present study is to describe the dietary patterns among people of South Asian origin and their risk for developing metabolic syndrome and chronic diseases such as CVD and T2D. We will conduct a cross-sectional investigation of data collected from South Asians participating in the “Alberta Tomorrow Project” (ATP). Assuming that ATP recruitment was representative of the Alberta population, and based on South Asians comprising 4% of the Canadian population, we estimate ~1000 South Asians would have participated in this study. The study aims for this project are as follows: 1. Assess dietary intake and alcohol consumption of South Asian population. 2. Compare dietary intake of South Asian population to the Healthy Eating Index score as a means of assessing adherence to dietary recommendations made in “Eating Well with Canada’s Food Guide”. 3. Compare dietary intake of South Asian population to the Alberta Diet Score as a means of assessing acculturation to a dietary pattern typical of the Prairie Provinces. 4. Assess metabolic disease risk based on the eating patterns of South Asian population. Results from this study will help inform and prepare targeted intervention programs to improve dietary and lifestyle choices within these communities.

Role of Lifestyle Factors (Physical Activity, Sedentary Behaviour, and Excess Body Weight) on Overall and Site-Specific Cancer Risk in Alberta’s Tomorrow Project

Darren Brenner

Physical activity has well-established links in the prevention and management of a number of chronic conditions, including cancer and in particular, cancers of the breast, colon and endometrium. Reductions in risk by as much as 20-40%for certain chronic diseases have been shown in adults with the greatest level of physical activity compared to sedentary individuals. Increased sedentary behavior has also been strongly associated with the risk of many chronic health conditions, especially diabetes, obesity and cancer. This relation has been found to be independent of physical activity and is strongest for colon, breast and bladder cancers. Lastly, excess body weight/obesity has been associated with cancer risk, particularly for cancers of the endometrium, gallbladder and kidney in women and cancers of the esophagus, colon and thyroid in men. The objectives of the proposed study are to use data collected by the Alberta’s Tomorrow Project (ATP) from over 50,000 participants to: (1) derive cancer risk estimates associated with physical activity (intensity, frequency and duration), sedentary behaviors, excessive body weight/obesity and stress; (2) determine how additional covariates, such as sociodemographic characteristics, influence the prevalence of physical activity, sedentary behaviors, obesity, stress, tobacco smoking and adverse health outcomes, especially cancer; and (3) assess the joint effects of exposures, as well as the mediation/relationships between exposures. These estimates will be derived using Cox regression, path analysis and structural equation modelling. The proposed study will combine questionnaire and cancer registry data provided by ATP participants to provide accurate and reportable estimates of physical activity, sedentary behavior and excess weight/obesity as they relate to cancer risk outcomes in the adult Albertan population. Results from this study will support the development of approaches to better assess the relationship between cancer risk and these lifestyle factors in human populations.

Cohort Profile: Design, Methods, and Demographics from Phase II of Alberta’s Tomorrow Project, as a regional cohort within the Canadian Partnership for Tomorrow Project

Heather Whelan

Between 2000 and 2015, Alberta’s Tomorrow Project (ATP) enrolled ~50,000 adults aged 35-69y into a prospective cohort established to support future studies in cancer and chronic disease etiology.

In Phase I (2000-2008), participants completed general health and lifestyle, diet and physical activity questionnaires, and gave consent for linkage with administrative databases. Pilot studies investigating the feasibility of blood collection from a geographically dispersed cohort were undertaken. A manuscript describing this period is currently in press (Robson et al. 2016).

In Fall 2007, the Canadian Partnership Against Cancer called for a proposal to develop a cohort of up to 300,000 adults across Canada. At that time, several cohorts (such as ATP) were already underway, and thus it was decided that a pan-Canadian cohort (called the Canadian Partnership for Tomorrow Project, CPTP) could be created by leveraging existing infrastructure and additional funding sources. The joining of ATP to the CPTP initiative is described as Phase II.

For Phase II of ATP (2009-2015), recruitment to the CPTP protocol in Alberta was accomplished in two ways. First, existing ‘active’ Alberta’s Tomorrow Project participants enrolled between 2000 and 2009 (n=31,213) were invited to re-consent to the pan-Canadian protocol; 15,872 (51%) accepted, and, of these, 10,212 people (63% women) attended a study centre to provide biospecimens for banking. The second approach was to recruit new participants using various strategies; 24,321 additional participants were enrolled, and 84% of those attended a study centre for physical measurements and biospecimen donation.

The purpose of the present study is to describe methods for recruitment and to characterize the participants who were enrolled in Phase II. We will summarize and describe Phase II participants based on sociodemographic characteristics and health-related characteristics. In addition, we will compare and contrast ATP participants who consented and those who chose not to consent to the CPTP protocol.

Tobacco use and exposure over time in Alberta’s Tomorrow Project: Evaluating the impact of the 2008 tobacco legislation

Heather Whelan

Tobacco use and exposure have notable implications for disease, disability and death (WHO, 2008). Strong and consistent epidemiologic evidence exists that tobacco smoking and exposure causes cancers and other chronic diseases (Jha et al., 2009). Additionally, changes to tobacco legislation have been shown to impact use and exposure prevalence at a population level and may have implications for disease trajectories (Who, 2008).

The proportion of incident cancer cases and chronic diseases attributable to active and passive tobacco exposure in Alberta has been previously examined using data from the 2000-2007 Canadian Community Health Survey (CCHS) (Poirier et al., 2015). However, CCHS is limited in its ability to examine changes in the same sample over time. Alberta’s Tomorrow Project (ATP) is in a unique capacity to overcome this limitation.

The Alberta provincial government introduced the Tobacco Reduction Act on January 1, 2008 which comprehensively banned smoking in all workplaces and public places (including within a five meter range of any air intake, door, or window) throughout the province (Alberta Provincial Government, 2007). Moreover, it prohibited the sale of tobacco products in all post-secondary institutions, healthcare facilities, and stores containing pharmacies. The purpose of these legislative changes was to influence the sale and use of tobacco products, and consequently, exposure. There is a dearth of studies to date that have deliniated the impact of legislative changes on smoking use and exposure prevalence within the same individuals over time.

Alberta’s Tomorrow Project (ATP) captured smoking status and second hand smoke exposure prevalence in participants at enrollment (2000-2008) and follow-up (2009-2015). This study is in a unique capacity to evaluate the possible impact and implications of 2008 tobacco legislative changes for both active and passive tobacco exposure as well as chronic disease occurrence over a period of time within the same sample of individuals in the province of Alberta, Canada.

This project will estimate the prevalence of active (current or former smoker) and passive (second-hand smoke) tobacco exposure in Albertan adults (aged 35-69 years old) enrolled in phase I of the ATP cohort. Changes on tobacco use, exposure, chronic disease occurrence and quitting behaviors have been captured at enrollment and follow-up from self-report questionnaires. The data will be used to determine the potential impact of the 2008 legislation on tobacco use and exposure prevalence at a population health level, much like a natural intervention, in order to determine possible implications on chronic disease occurrence and trajectory. Quitting behaviours at the 2008 follow-up will also be used to assess cessation intentions and resulting changes in current users at follow-up. Finally, we will identify the socio-demographic factors associated with these exposures and evaluate how these rates change over time within individuals.

The objectives of this study are to determine the:

1. prevalence of and changes in smoking use and second-hand smoke exposure prevalence of ATP cohort participants before (2000-2007) and after (2009-2015) the 2008 tobacco legislative changes took effect in Alberta
2. prevalence of intentions to quit tobacco smoking at Survey 2008, rationale for quitting and smoking status outcome at follow-up
3. change in chronic disease incidence from enrollment to follow-up and relating these to tobacco use and exposure

To our knowledge, these will be one of the first estimates within the same sample of individuals to determine the possible impact and implications of 2008 tobacco legislative changes for both active and passive tobacco exposure as well as chronic disease in the province of Alberta, Canada.

Influence of misreporting on dietary data in Alberta’s Tomorrow Project, and application to the relationship between dietary patterns, glycemic load and risk of cancer

Nathan Solbak

It is well established that dietary measurements are prone to error. Accounting for misreporting error (either underreporting or overreporting both energy intake and food items) is an important step towards understanding and clarifying the potential role diet plays in disease outcomes. The gold standard for assessing misreporting is the measurement of energy balance using doubly labeled water. However, the high expense of this technique makes it unrealistic for use in large cohorts; therefore, alternative statistical methods are available to identify individuals who misreport. The Goldberg method identifies misreporters based on cut-offs between the ratio of energy intake and basal metabolic rate compared to physical activity level (PAL). However, the PAL needs to be assumed in order to use the Goldberg method. In contrast, the updated McCrory method only uses PAL to select the appropriate prediction equation to calculate the estimated energy requirement (EER). The EER is then compared to energy intake as a cut-off in the McCrory method. Depending on which method is used, there could be a different group of individuals identified.

Additionally, the choice of misreporting method could have implications for dietary pattern analysis. The associations between dietary patterns identified and disease outcomes such as cancer could be vastly different whether or not inaccurate data is included in the analysis. Dietary components such as glycemic load are likely to be influenced by the inclusion of misreported data in the analysis. Therefore, associations between glycemic load and cancer risk may have different results if misreported data is present. This project will address the following objectives:

1. Identify energy intake misreporters in ATP using a comparison of the Goldberg and updated McCrory methods.
2. Determine the influence of misreported energy intake on dietary patterns identified in ATP.
3. Determine the relationship between dietary patterns, glycemic load, and both the risk of overall cancer as well as particular sites including breast and colorectal.

Religion/Spirituality and Perceived Cancer Susceptibility as Individual-level Factors Encouraging Cancer Screening Behaviour: A Longitudinal Analysis of the Alberta Tomorrow Project

Mark Oremus

Background: Screening for cancer is a secondary prevention strategy that relies on early detection to identify and halt the pathological development of disease in asymptomatic individuals who are at risk of disease. Religion and spirituality (R/S) may encourage or discourage people from undergoing screening for disease, while perceived susceptibility (PS) – whether an individual feels they are personally vulnerable to a health-related condition or disease – has been shown to be associated with cancer screening uptake.

Overall Objective: We propose to use data from the Health and Lifestyle Questionnaire (HLQ), Survey 2004, Survey 2008, Update Health and Lifestyle Questionnaire (UHLQ), and Core Questionnaire (Core) from the Alberta Tomorrow Project (ATP) to address the following research questions:

1. What is the association between R/S at baseline and the incidence of sigmoidoscopy, colonoscopy, prostate specific antigen (PSA), Pap, and mammography screening tests in the ATP?
2. Is PS to developing cervical, breast, prostate, and colorectal cancer associated with the incidence of PSA, Pap, mammography, sigmoidoscopy, and colonoscopy screening tests in ATP? Does an individual’s PS of developing cancer differ across diseases? Does an individual’s incidence of screening differ between screening tests?

Methods: We will include ATP participants of any age who report being free of chronic conditions at their baseline interviews and who have at least one follow-up interview. We will explore the data descriptively using histograms and bar charts. We will summarize normally distributed continuous variables as means and standard deviations, non-normally-distributed continuous variables as medians and interquartile ranges, and categorical variables as frequencies (n/%). To examine the association between the exposure variables (R/S or PS) and incident screening at follow-up, we will build a series of multivariable logistic regression models for each of the screening tests of interest.

Translational Application of microRNA Profiling for Early Detection of Lung Cancer

Eric Bédard

Lung cancer is the primary cause of cancer mortality. Its presentation at late stages is one of the main factors responsible for the overall poor survival of patients with this disease. Approximately 75% of non-small cell lung cancer (NSCLC) cases are presented at an incurable stage. Micro-ribonucleic acids (miRNAs), non-protein-coding RNA strands, are promising candidates for cancer biomarkers due to them being expressed in a tissue-specific manner. Dysregulation of miRNAs can lead to carcinogenesis as they may function as tumor suppressors or promoters. Using miRNAs would allow us to screen for cancer using body fluid-based tests that are minimally-invasive, relatively low-cost and easily repeatable. There is a high false positive rate with CT screening for lung cancer, and this leads to patient anxiety and unnecessary tests. Circulating biomarkers are attractive for cancer screening since they are body fluid-based tests that are minimally-invasive, relatively low-cost and easily repeatable. In our previous work a miRNA panel was developed and has demonstrated good sensitivity and specificity in early stage lung cancer studies when measured in sputum and bronchial washing with RT-PCR and cluster analysis of miRNA profiling. RT-PCR is used as a standard method to detect miRNA. We hypothesize that miRNA profiling collected from blood plasma can be used for future detection of early stage NSCLC. To do this, we propose a case control study of clinical stage I/II NSCLC, matched with healthy controls. Inclusion of 70 cases and 110 controls matched for age, gender, and smoking history will allow statistical significance, which is recommended for such studies. We also want to characterize the hypothesis that after surgical resection of the cancer tissue, miRNA levels will change. To do this, we will compare miRNA levels from plasma before and after surgery. This can lead to development of future screening for recurrence of cancer.

Paula Robson, Darren Brenner, Alianu Akawung, Tiffany Haig, Will Rosner, Nathan Solbak, Jennifer Vena, Heather Whelan, Jason Xu, Christine Friedenreich

Over the past two decades, several organizations have published series of recommendations and guidelines that provide strategies for cancer risk reduction based on modifiable behaviors and lifestyle factors. Although fine details of the recommendations differ, many focus on the broad themes of achieving and maintaining a body size within the normal range for body mass index (BMI), being physically active, consuming low amounts of alcohol (if consumed at all), avoiding exposure to tobacco, consuming diets that are predominantly plant-based and/or relatively high in fruits and vegetables, and participating in screening programs. Similar types of recommendations exist for reducing risks of cardiovascular disease and type II diabetes.

Currently there is a dearth of comprehensive population-based prevalence data for modifiable cancer and chronic disease risk factors in Alberta, and across Canada as a whole. We propose to describe the prevalence of a series of risk factors (physical inactivity, overweight and obesity, diet related risk factors, alcohol use, and tobacco use) reported by adults enrolled in Alberta’s Tomorrow Project (n=31,157). We will also describe the extent to which participants comply with cancer screening guidelines.

Alberta’s Tomorrow Project used the method of Random Digit Dialing to recruit approximately 30,000 participants across Alberta between 2000-2009. Only people within the ages of 35 and 69 years with no history of cancer were recruited.

The study will provide the descriptive analyses of exposure distribution of cancer and chronic disease risk factors in the ATP population upon enrollment in the cohort. Data concerning time and energy expended in domains of recreation, transportation, household and occupation, the distribution of BMI, waist circumference and waist/hip ratios reported by participants, reported intakes of vegetables, fruits, whole grains, red meat, energy and macronutrients (including sources) and self-reported alcohol and tobacco exposure will be described by sex, age category and other demographic variables.

Analyses of these data will be the first step in helping identify priority areas for the development and implementation of comprehensive cancer and chronic disease risk reduction strategies in Alberta, and potentially across Canada. They may also help inform other projects that aim to calculate up to date population attributable risks/fractions for cancer and chronic disease associated with the prevalence of modifiable risk factors.

Targeted Re-sequencing and Fine Mapping of Breast Cancer Susceptibility Loci at Chromosome 4q31.22 – Identified by Genome Wide Association Study

Sambasivarao Damaraju, John Mackey, Anil A Joy, Jennifer Dufour, Russ Greiner

The aim of the current study is to identify breast cancer susceptibility alleles in populations. Breast cancer prediction models have used single nucleotide polymorphisms or SNPs and currently also use copy number variations (CNVs). The true test of the models is to assess for positive predictive value by comparing any potential “outliers” from the controls who may have been diagnosed with breast cancer since their initial enrolment in Alberta’s Tomorrow Project (ATP). Currently, the models simply show in a confusion matrix, cases and controls that are not assigned to proper bins. Access to the refined data from ATP will help to improvise the models. True applicability of the findings rests on improvising the models so that they will have selectivity and specificity for the potential population screening to identify those at risk for breast cancer, long before the actual diagnoses; thus enabling prophylactic interventional modalities to reduce the burden of disease for individuals and the societal burden in terms of health care dollars.

Christine Friedenreich, Darren Brenner, Anne Grundy, Abbey Poirier, Xin Grevers, Farah Khandwala

Limited research has focused on estimating population attributable fractions/risks (PARs) for modifiable risk factors and cancer in Alberta. Information concerning the fraction of cancer attributable to individual risk factors is essential for both developing and implementing population-based cancer prevention strategies.

The objectives of this project are to: 1) estimate the prevalence of exposures to modifiable risk factors consistently associated with cancer for the Alberta adult population; and 2) estimate the number of incident cases of cancer per year attributable to these factors among Albertans.

Three types of data are required for the estimation of PARs including the magnitude of the risk association between individual exposures and cancer sites, estimates of the population prevalence of individual exposures, and the site specific cancer incidence data. Estimates of risk for each exposure and cancer site of interest will be extracted from a review of reports from international collaborative groups/panels and the current published peer-reviewed literature in PubMed. Data on the prevalence of individual exposures of interest will be obtained through: 1) results from Statistics Canada surveys; 2) publicly available government databases; 3)published peer-reviewed literature; and 4)consultation with relevant experts. Data on cancer incidence in 2012 will be acquired from the Alberta Cancer Registry.

The proportion of cancers at a given cancer site that can be attributed to an individual exposure, will primarily be estimated for both Alberta as a whole and for individual Alberta Health Services zones using the formula given by Levin: PAR = (Pe x (RR-1))/(1 – (Pe x (RR-1))). Here Pe represents the population prevalence of a given exposure and RR represents the relative risk of cancer at a specified site for the given exposure.

These estimates will provide a greater understanding about the main causes of cancer in Alberta and will be useful in informing future cancer prevention efforts.

Ilona Csizmadi, Tram Pham

Anthropometric measurements are used as estimates of excess adiposity and are routinely used in epidemiological studies. Excess adipose tissue is associated with an increased risk of cardiovascular diseases (CVD), hypertension, diabetes and some cancers. Individuals who are overweight or obese have a higher chance of disease risk and even premature death. There exists gender differences in the performance of anthropometric measurements in assessing adiposity. These differences in measurement performance can directly impact the assessment of obesity related disease risk and prevalence in men and women.

The purpose of the project is to examine the performance of various anthropometric measures to assess adiposity and to determine if these results are consistent in both men and women. Specific aims are: i) To describe the consistency with which various measures (BMI, BAI, waist circumference, hip circumference, waist to hip ratio, and waist to height ratio) estimate the prevalence of excess adiposity, ii) To estimate the relation between these measurements (BMI, BAI, waist circumference, hip circumference, waist to hip ratio, and waist to height ratio) and self-reported hypertension, and iii) To determine if there are gender differences in this relation.

A descriptive study approach of self-reported data from approximately 30,000 participants who have completed the Health and Lifestyle Questionnaire, along with the self- reported physical measures will be analyzed using SAS (version 9.1). The various anthropometric measures will be compared across different BMI categories and stratified by gender. The correlation between will also be studied and presented in a correlation matrix. The relation between anthropometric measures and high blood pressure will be studied using logistic regression.

Development of New Diagnostic and Predictive Biomarker Tests for Prostate Cancer

John Lewis, Desmond Pink, Catalina Vasquez, Konstantin Stoletov, Deborah Sosnowski

Currently, the PSA test helps to identify men that are likely to have prostate cancer. However, infection, trauma, and benign prostatic hyperplasia (BPH) are all causes of elevated serum PSA, significantly confounding PSA as a cancer biomarker. An improved test could significantly decrease the number of unnecessary biopsies. The present study aims to develop, validate and translate novel tools to improve prostate cancer testing, to determine the aggressiveness of prostate cancer and to improve our prediction of the onset of metastasis in prostate cancer patients. This study will also help us to identify prognostic biomarkers to differentiate more dangerous vs. less dangerous prostate cancer. This important information could help patients and their physicians choose treatment options for an improved outcome and better quality of life.

This will be a study of biofluid samples from prostate cancer patients that have donated their specimens to different biobanks in North America and Australia to evaluate the utility of novel biomarkers to diagnose and predict biochemical recurrence and-or metastasis in patients with prostate cancer. Associated clinical data related to diagnosis, treatment, outcome will be used to associate with results from the research tests. Samples from subjects without prostate cancer are required to serve as controls.

To achieve this, our team will focus on the validation of Cancer Microparticle (CMP)-based tests to diagnose prostate cancer, to predict biochemical recurrence and/or metastasis, and to identify novel markers for prostate cancer. For these studies, plasma, urine and/or semen samples from patients with confirmed prostate cancer and normal controls will be utilized first in retrospective analyses, using biospecimens from different centres across North America and Australia. Later, when APCaRI biobank is established, we will also use prospectively collected biospecimens from prostate cancer patients that donate their samples.

Predicting Outcomes in the Inflammatory Bowel Diseases: Gene-Environment-Microbe-Serology Interaction Studies

Gilaad Kaplan

The Alberta Inflammatory Bowel Disease (IBD) Consortium is a team of clinicians and scientists performing large-scale clinical outcome and pathogenetic studies in IBD – particularly focusing on gene, microbe, and environment interactions. The goal of the Consortium is to generate new knowledge through research that will assist in our understanding of the determinants and causes of IBD. With an already established framework of patient recruitment and sample analysis, our attention turns to retrospective patient cohorts to help find genetic, environmental, microbial, and serological factors that contribute to Crohn’s disease (CD) and ulcerative colitis (UC). In this proposal, we will identify Albertans living with IBD and focus on identifying gene-environment-microbe-serology interactions as predictors of disease phenotype, prognosis, and drug response in patients with IBD.

The results of such studies have the potential for primary, secondary, and tertiary prevention of IBD. This will not only better the quality of life of those who are susceptible to IBD and the 15,000 Albertans living with chronic IBD, but it will generate relief to Alberta Health Services as early detection and targeted intervention can alleviate much of the pressure on the health care system. By identifying gene-environment interactions in IBD, steps can be taken to prevent initial development of IBD and reduce morbidity by educating the public of risk factors, implementing screening programs for high-risk populations, and providing prompt early diagnosis and treatment.

Adherence to Cancer Prevention Guidelines: Survival Analysis to Estimate the Association Between Albertans’ Adherence to Cancer Prevention Guidelines and Subsequent Incidence of Cancer in Alberta’s Tomorrow Project Cohort

Jason Xu, Paula Robson, Heather Whelan, Will Rosner

Hypothesis: WCRF and the AICR published guidelines on food and nutrition, physical activity, body composition and other modifiable risk factors that may affect the risk of developing cancer and other chronic diseases. We hypothesized that increased overall adherence to these cancer-specific recommendations were associated with a reduced subsequent incidence of cancer. Objective: To estimate the association between Tomorrow Project participants’ adherence to cancer prevention guidelines, as recommended by WCRF and AICR, and the subsequent incidence of cancer observed in the Tomorrow Project cohort. Methodology: The project will involve analysis of data collected from participants in Alberta’s Tomorrow Project. The baseline data (from three baseline questionnaires: Health and Lifestyle Questionnaire (HLQ), Diet History Questionnaire (DHQ) and Past Year Total Physical Activity Questionnaire (PYTPAQ)) were collected over the span of 9 years from 2001 to 2009. Cancer diagnoses in the follow up period will be obtained via a combination of participants’ self report and linkage with the data from the Alberta Cancer Registry.

The final study sample will exclude the following participants: participants who were recruited as ‘second in household’ in the first recruitment wave, participants outside the 35-69 year age range at the time of completing the HLQ, pregnant women, participants with a body mass index (BMI) <18.5kg/m2, participants who reported having history of cancer prior to enrolment, participants not living in Alberta at the time of enrolment, participants who reported implausible energy intakes, and participants with extreme dietary energy intake or total energy expenditure (using the interquartile range method). The project will involve manipulation of the HLQ, DHQ and PYTPAQ datasets to create variables according to the cancer prevention guidelines. Preparation of the datasets and statistical analyses (e.g. descriptive statistics, chi-square tests, logistic regression models, and Cox proportional hazards model) will be conducted at Holy Cross (Calgary) by the Statistical Associate using the SAS software.

Jason Xu, Nathan Solbak, Sanaz Vaseghi, Heather Whelan, S Elizabeth McGregor

Cancer screening increases the chance of survival and reduces morbidity associated with certain cancers. This study aims to determine the extent of adherence of Alberta Tomorrow Project participants to breast, colorectal, cervix, and prostate cancer screening test guidelines and to identify any patterns of utilization to help developing strategies to improve participation rates. This project will involve analysis of the baseline data from the Health and Lifestyle Questionnaire (HLQ) collected between 2000 and 2009 and Survey 2008, a follow up survey collected between 2008 and 2009. All data is based on self-reported screening status included in HLQ and Survey 2008.

Participants diagnosed with breast, cervix, colorectal or prostate cancer between completing the HLQ and Survey 2008 will be excluded from the analysis of screening patterns at Survey 2008. Utilizing screening tests for each site will be classified as follows: Pap test: within 1 year; 2 to 3 years; more than 3 years; or never screened. Mammography: within 2 years; more than 2 years; or never screened. Fecal Occult Blood Test: within 2 years; more than 2 years; or never screened. Sigmoidoscopy and colonoscopy: within 5 years; never or not within 5 years (on HLQ) and on Survey 2008, within 5 years; or not within 5 years for sigmoidoscopy; and within 10 years; never or not within 10 years for colonoscopy. Overall screening status for colorectal cancer will be determined by combining screening using FOBT and/or endoscopy with up-to-date defined as any of: FOBT within the past 2 years; sigmoidoscopy within the past 5 years; colonoscopy within the past 5 (HLQ) or 10 (Survey 2008) years. PSA: within 2 years; more than 2 years; or never screened. Reason for screening categories classified as screening; family history; or indication.

Patterns of screening for each site will be described based on screening status at HLQ and Survey 2008. This data analysis is descriptive and exploratory and the final categories will depend on the patterns observed in the data. Statistical analysis (e.g. descriptive statistics, chi-square tests, logistic regression models) will be conducted using the SAS software.

Weight Gain During Adulthood in Tomorrow Project Participants

Lorraine Shack, S Elizabeth McGregor, Paula Robson, Heather Whelan, Sanaz Vaseghi, Geraldine Lo Siou, Jennifer Vena, Tiffany Haig

Obesity is of major public health concern in Canada. The percentage of Canadians who are overweight or obese has risen dramatically in recent years, mirroring a worldwide phenomenon. Worldwide it is estimated that 25% of all cancer cases are caused by obesity. Specifically, there is convincing evidence that overweight and obesity increase the risk for other chronic diseases, including type 2 diabetes, cardiovascular disease, hypertension and stroke and cancers of the endometrium, breast, colon, esophagus and kidney. The cause of this rapid increase in obesity has been attributed to lifestyle, environmental and cultural factors that have resulted in a shift toward a sedentary lifestyle, with an over consumption of calories and reduced physical activity.

As individuals age there is a tendency to gradually gain weight. This is partially due to increasing age corresponding to a decrease in muscle mass and subsequently metabolism (sarcopenia), lifestyle changes and increasing prevalence of other health conditions that may limit an individual’s ability to exercise. An average annual weight increase of one pound has been commonly reported during middle adulthood. An increase in caloric intake relative to energy expenditure of just 50 to 100 Kcal per day is enough to cause this gradual weight gain. The middle age weight increase may be of importance to cancer prevention as the proportion of the population in the older age categories increases. However, there are relatively few studies which have investigated the factors associated with weight change over time in middle adulthood. To our knowledge there are no studies investigating this topic in Canada. The aim of this study is to investigate weight changes over time associated with increasing age and to explore potential explanatory factors in the Tomorrow Project cohort.