Xiaomei Ma, Ph.D. - US grants

DESCRIPTION (provided by applicant): This is an application for a Cancer Prevention, Control, Behavioral and Population Sciences Career Development Award using the K07 mechanism. The purpose of the award is to provide Dr. Xiaomei Ma protected time and additional training and mentoring so she could become an independent investigator. The myelodysplastic syndromes (MDS) are a group of clonal proliferative bone marrow disorders resulting in dysmyelopoiesis and peripheral blood cytopenias. Although MDS has long been considered a pre-leukemic disorder, it is better viewed as a form of cancer because of the clonal nature. Despite MDS's poor prognosis and a likely increasing incidence, understanding about the etiology and factors affecting survival and quality of life is very limited, which may be due, at least in part, to the biological heterogeneity of the disease. In 2003-2004, the US congress identified MDS as understudied and urged that the NCI and other agencies to further MDS research. Building upon a large series of more than 1,600 primary MDS patients with detailed information on biological and clinical characteristics, we propose to 1) classify MDS cases into subgroups that are more homogeneous; 2) illustrate whether cases in different subgroups have distinct patterns of previous exposures (e.g. family history, medical history, cigarette smoking, alcohol drinking, exposure to occupational and environmental chemicals, and exercise); 3) describe the survival patterns of cases in different subgroups; 4) evaluate which factors determine the duration of survival; and 5) assess the health-related quality of life among surviving MDS patients. Compared with previous studies, the proposed research is distinguished by a remarkably large sample size and the comprehensiveness of data. Because of the availability of extensive data from previous research, only few resources are needed for additional data collection in this study. During the course of the study, Dr. Ma will be trained in cancer-related quality of life research, assessment of dietary factors, and application of spatial statistics methods and bioinformatics. Findings from the study will advance the knowledge about the biological heterogeneity of MDS, help guide future etiologic studies of MDS, offer insight into the determinants of survival, and better understand the quality of life of MDS survivors and factors affecting the quality of life. In addition, the award will provide a promising junior investigator a unique opportunity for career development in cancer research.

DESCRIPTION (provided by applicant): The myelodysplastic syndromes (MDS) are a group of clonal proliferative bone marrow disorders that result in dysmyelopoiesis and peripheral blood cytopenias. Population-based cancer registry data indicate that over 10,000 new cases of MDS are diagnosed in the US annually. Eighty percent of MDS cases are 65 years or older at the time of diagnosis, and observed three-year survival is only 35%. The Congress has identified MDS as being understudied and urged the National Cancer Institute and other federal agencies to further MDS research. The availability of the SEER-Medicare linked database provides a unique opportunity to study the patterns of care and outcomes in MDS patients, both of which are poorly understood. By the end of 2008, the SEER-Medicare database will include SEER data on MDS patients who were diagnosed during 2001 - 2005 and Medicare claims information on these patients up through the end of 2007. We propose to use these data to (1) assess the patterns of care, especially the number of bone marrow tests MDS patients received after diagnosis and the use of various treatment strategies, included the recently approved DNA demethylating agents, azacitidine and decitabine;(2) identify factors that influence the survival of MDS patients after accounting for comorbid conditions;and (3) estimate the direct medical costs related to the treatment of MDS patients. A population-based cohort of approximately 6,716 MDS patients diagnosed at 66 years or older are expected to be included in the study, so will non-cancer controls matched to patients on age, gender, race, SEER region, and comordid conditions. For MDS patients, the frequency of bone marrow tests and the percentage of patients receiving various treatments (e.g. blood transfusions, iron chelation therapy, growth factors, chemotherapy, and demethylating agents) will be evaluated. We will also assess whether the patterns of care are influence by individual or community-based demographic and socioeconomic characteristics. In addition, we will evaluate which factors/characteristics affect the survival of MDS patients, after taking into account comorbid conditions that are important in predicting mortality in MDS patients. Furthermore, we will estimate the Medicare costs incurred by MDS patients from diagnosis through the end of 2007, or date of death, whichever is earlier, and assess which patient characteristics influence costs. By comparing the costs of MDS patients with the costs of non-cancer controls who are matched to MDS patients on age, gender, race, SEER region, and comorbid conditions, we will estimate the medical costs that are attributable to MDS. All analyses will be conducted separately for MDS as one group and for major MDS subtypes. To account for potential changes in the coverage for certain treatments (e.g. growth factors) after Medicare Part D took effect in 2006, we will stratify analyses by two different time periods, 2001 - 2005 (Pre Part D) and 2006 - 2007 (Post Part D). The proposed study will be the first ever to use a large population-based sample of MDS patients to evaluate the patterns of care and MDS survival adjusting for comorbid conditions. It will also be the first to provide a national estimate of the direct medical costs related to the treatment of MDS. Findings from the proposed study will be valuable to clinicians making decisions about the care and treatment of MDS patients and very helpful to policy makers in the allocation of health resources. PUBLIC HEALTH RELEVANCE: The myelodysplastic syndromes (MDS), a group of understudied malignancies of the blood system, are most common in elderly people 65 years or older. The investigators propose to use linked cancer registry files and Medicare claims of about 6,716 MDS patients to study the patterns of care and survival and estimate the direct medical costs related to the treatment of MDS. Findings from the study will be valuable to both clinicians caring for MDS patients and policy maker allocating health resources.

DESCRIPTION (provided by applicant): Although acute lymphoblastic leukemia (ALL) is the most common malignancy in children (0 - 14 years), there are few known risk factors. In the United States, Hispanic children have higher incidence of ALL than any other ethnic/racial groups; they have 1.3 times of the risk for ALL than non-Hispanic white children. Two small genome-wide association studies (GWAS) recently identified 2-3 genetic risk factors for childhood ALL, but only in European white populations with non-population based ascertainment. High birth weight and other markers of fetal growth have been consistently linked to an increased risk of childhood ALL in most studies, and Hispanics appear to have accelerated fetal growth early on during pregnancy. In the current application, we seek to significantly advance leukemia research by clarifying the roles of genetic susceptibility and fetal growth in the etiology of childhood ALL in a large population-based sample of Hispanics and non-Hispanic whites in California, using high dimension arrays specifically designed for the two ethnic groups. We will (1) link California birth records with California Cancer Registry data (1988 - 2008) to assemble a population-based case-control study with an unprecedented size (4,000 cases and 4,000 controls, about half of which will be Hispanics and half non-Hispanic whites); (2) obtain the archived neonatal dried blood spots of all cases and controls from the California Department of Public Health; (3) genotype the study population with custom Affymetrix whole genome arrays designed for the California population, with optimized designs for Hispanics and non-Hispanic whites separately; (4) use data from these cases and controls to identify distinct and shared genetic risk alleles for Hispanics and non-Hispanic whites; and (5) validate top GWAS hits in the California Childhood Leukemia Study (an ongoing NIH-funded case-control study with a separate and distinct subject ascertainment, >1000 cases, > 1000 controls, approximately half of which will be Hispanics and half non- Hispanic whites) and in a Brazilian childhood leukemia study with Latin American cases (n = 200) and controls (n = 400). In addition, we will utilize data on the characteristics of all births in California during 1988 - 2008 to develop an algorithm to predict birth weight based on factors such as gestational age, gender, birth order, parental race, parental Hispanic status, and parental ages, and calculate a proportion of expected birth weight (PEBW) for all cases and controls (n = 8000) by dividing their actual observed birth weight by the expected birth weight derived from the algorithm. Subsequently, we will determine whether the PEBW or birth weight is associated with (i) childhood ALL, (ii) the top GWAS hits, and (iii) the SNPs in genes potentially related to fetal growth, in Hispanics and non-Hispanic whites separately. Given the unprecedented sample size, especially the large number of Hispanics included, and ethnic-specific genotyping panels with great genome coverage, this study will clarify the role of genetic susceptibility in childhood ALL, the mechanism involving fetal growth, and in the process identify potential reasons for the puzzling ethnic difference in disease incidence.

DESCRIPTION (provided by applicant): This proposal is to continue a highly successful graduate partnership program (GPP) whose overall goal is to train the best and brightest students in the area of lifestyle and environmental determinants of human cancer risk. The proposal consists of a partnering between academicians and NIH intramural scientists, with the primary partners in this arrangement being Yale University's Department of Epidemiology and Public Health (EPH), and the National Cancer Institute's Division of Cancer Epidemiology and Genetics (DCEG). This unique partnership draws upon the academic strengths of Yale University, the resources and expertise at NCI, and the research strengths of these two institutions, both of which are internationally known for their research in cancer epidemiology and genetics. The cooperative training program trains students in modern methodologies for evaluating lifestyle determinants of human cancer risk, with an emphasis on nutritional, environmental, and occupational determinants including their interactions with genetic factors. Pre-doctoral candidates selected into this unique training program are jointly mentored by Yale faculty and NCI investigators. Upon matriculation, pre-doctoral candidates receive up to 2 years of didactic instruction at EPH, which is an accredited School of Public Health. Coursework emphasizes methods training in epidemiology and biostatistics, with content training in a variety of areas including cancer epidemiology, nutritional epidemiology, environmental / occupational epidemiology, molecular and genetic epidemiology, and cancer biology. At the end of the didactic experience, students take their qualifying exams at Yale, and then subsequently relocate to the NCI. Students identify a thesis topic, write a thesis prospectus, and identify a Dissertation Advisory Committee comprised of two faculty members from Yale (at least one of whom is a senior mentor) and two from NCI. Students meet all requirements for the Ph.D. at Yale, which includes a one-year teaching assistant (TA) experiential requirement. The pre-doctoral training program is administered by a Steering Committee with representation from both EPH and DCEG. Graduates of this program are uniquely qualified to be future leaders in the field of cancer epidemiology and genetics.

DESCRIPTION (provided by applicant): The primary goal of this project is to develop novel statistical methods to handle spatial uncertainty in the event locations when conducting cancer risk estimation. We consider two different types of spatial uncertainty, specifically, 1) coarsenin due to the practice of releasing location information at an area level but not the point level and 2) geocoding error resulting from the use of geographic information systems software to convert residential addresses to geographic coordinates (i.e. longitudes and latitudes). Cancer epidemiologists can extract data from many different sources such as census, statewide health surveys, tumor registries and population-based case-control studies, and each source may yield data with different types of spatial uncertainty. Analytic methods are usually adversely affected by the presence of spatial uncertainty, resulting in biased parameter estimates, inflated standard errors, and reduced statistical power to detect spatial clustering and trends. To address these challenges, we propose a set of highly versatile estimation procedures to account for the spatial uncertainty and to efficiently combine data obtained from multiple sources. These procedures are based upon established theories on estimating equations and as such they can be easily implemented in practice. Compared with existing methods, the proposed methods are novel because 1) they permit the inclusion of individual-level risk factors for subjects with spatially uncertain locations, 2) the proposed intensity model admits a flexible semiparametric form and hence removes potentially restrictive assumptions such as the population density being constant over small geographic areas, and 3) they explicitly account for spatial correlation in the disease locations in both parameter estimation and statistical inference. In the substantive applications, we propose to supplement population-based case-control data with tumor registry data, census data and statewide health survey data. To the best of our knowledge, such an approach would be the first in the field and unparalleled. We will implement our proposed methods in a free, user-friendly R package. Our package will provide much- needed tools for more objective investigations of cancer risk factors by accounting for spatial uncertainty in the event locations. It will allow researchers to take advantage of the full spectrum of available data and use the data more effectively to reduce the burden of disease.

DESCRIPTION (provided by applicant): Childhood leukemia is the most common childhood cancer but the causes of the disease are uncertain in the vast majority of cases. Epidemiology studies have pointed to patterns of infection being influential risk factors for leukemia - with exposure to a wide variety and number of childhood contacts (e.g., daycare and older siblings) being protective for leukemia. These associations are strongest for the pre-B cell immunophenotype of leukemia, which we will study here. New epidemiologic evidence suggests that vigorous response to infections, meaning infections that require care from medical professionals, is a risk factor for leukemia. Additionally, children who contract leukemia have a deficit of the immunosuppressive cytokine IL10 at birth when compared to control children who did not get leukemia, meaning that they enter the world with a congenital immune aberration. We hypothesize, based on these observations, that features of neonatal immune function, namely a lack of immunosuppression, are important in risk of pre-B cell acute lymphoblastic leukemia (pre-B ALL). In the current study we will further define this immunosuppressive phenotype in more detail and examine the developmental fetal genetic interaction and maternal environmental factors that influence this fetal phenotype. We will assemble a series of 500 ALL case children and their mothers, and 500 controls with their mothers to assess whether two genetic factors (HLA-C and KIR) critical in developing a supportive immune environment for a fetus within its mother can influence fetal immune status and leukemia case/control status. We will additionally assess, with 200 case mothers and 500 control mothers, whether neonatal immune function phenotype is influenced by maternal immune status during pregnancy. We will define fetal immune phenotype by additional markers besides only IL10 levels, including TGF-?, IFN-?, and T- regulatory cell marker - a differentially-methylated region of the critical transcription factor FOXP3. The impact of maternal cytokines and IGE levels, and KIR genotypes (along with HLA-C genotypes in the neonate) will first be assessed on neonatal immune function among controls. Both maternal and fetal immune phenotypes will then be assessed for their impact in prediction of case status. The results of this study will further defie critical characteristics in immune development that influence a cancer of the immune system. These results will benefit efforts to understand the etiology of leukemia as well aid predictive and preventative modalities for childhood leukemia, and set the stage for follow-up studies that examine postnatal factors that combine together with congenital immune immunosuppression to produce leukemia.

Acute lymphoblastic leukemia (ALL) is the most common type of cancer in children (age 0-14 years), and the incidence of childhood ALL has continuously increased in the United States since 1975. Genome-wide association studies (GWAS) have identified common variants in several genes that are significantly associated with the risk of developing ALL including ARID5B, IKZF1 and CEBPE. However, there are likely additional variants contributing to the heritability of ALL that current GWAS may not have been powered to detect. There is mounting evidence to suggest that stable and plastic epigenetic modification may also play an essential role in heritability of many cancers including leukemia. MicroRNA (miRNA), one epigenetic factor, post- transcriptionally regulates many biological mechanisms, such as apoptosis, adaptive and innate immunity. MicroRNA regulation has recently emerged as an important, yet under-explored, mechanism that can change the expression of multiple genes implicated in carcinogenesis. We hypothesize that miRNA modification is heritable and associated with ALL risk. Further, we postulate that variants within miRNAs and variants in corresponding miRNA binding sites contribute to the susceptibility of ALL and that interactions between variants in these two regions may contribute to ALL risk in a non-additive manner. To test these hypotheses, we will conduct extensive bioinformatics analyses of genotyping data from two large GWAS of childhood ALL completed in 2016, with more than 4500 cases and nearly 5000 controls across the Discovery and Replication datasets comprised primarily of Hispanics/Latinos and Whites, but also contain subjects of Asian and African- American ancestry. These subjects have been genotyped for more than 700,000 single-nucleotide polymorphisms across the genome. To increase the chance of observing polymorphisms within miRNAs, and to increase comparability between subjects genotyped on different genotyping platforms, we will conduct whole genome imputation prior to analysis. We will first identify variants within miRNA coding regions and miRNA binding sites that are associated with risk of ALL. We will then take a gene-based approach to identify rare variants (minor allele frequency less than 1%) in and around miRNA coding regions and miRNA binding sites that are associated with risk of ALL in aggregate. Finally, we will investigate if interactions between variants in miRNA coding regions and their specific miRNA biding sites and associated target gene contribute to ALL susceptibility. These aims investigating the role of inherited genetic variation in miRNA coding regions and miRNA binding sites done on a genome-wide scale will hopefully provide new insights into the role of miRNAs in the etiology of childhood ALL, a research area of public health significance given the continuously increasing incidence.

PROJECT SUMMARY/ABSTRACT The Yale Cancer Prevention and Control (CPC) Training Program requests five years of support for an innovative, multi-disciplinary and translational training program at the Yale School of Public Health (YSPH) and Yale Cancer Center (YCC). Currently, Yale only has one T32 training grant in cancer research, and it focuses on cancer biology training for basic scientists. Given the long history of successful CPC research and training by Yale faculty, and a critical need to train outstanding CPC scientists to advance our fight against cancer, we are excited to re-establish a CPC Training Program at Yale. Our Yale CPC Training Program will leverage strength from a Pre-Doctoral Program within YSPH, and a Post-Doctoral Program within YCC that draws from many disciplines but is administratively housed within YSPH. The Yale Cancer Prevention and Control (CPC) Training Program aims to educate, train and mentor pre- and post-doctoral fellows in five thematic areas critical to CPC: cancer etiology, cancer outcomes, lifestyle behavioral interventions, implementation science and community-engaged research. These five areas were selected because they are critical domains in the spectrum of CPC research, they leverage strengths of our faculty, and they will ensure that T32 trainees with diverse interests will achieve pivotal growth through the program and become leaders in novel, impactful CPC research. Fellows will also directly participate and interact with the seven YCC Research Programs and 14 YCC Disease-Aligned Research Teams, which are composed of scientists and clinicians conducting cancer research and delivering cancer care. Fellows will work with 22 Yale CPC Training Program Primary Faculty members, who collectively have expertise in our five thematic areas. The Primary Faculty members come from 13 different departments across Yale, and have trained 50 pre-doctoral and 126 post-doctoral fellows over the last ten years, with a strong majority working in cancer-related careers. The 22 Primary Faculty currently have more than a combined $25M of research funds. Our training program will include a comprehensive mentoring program, participation in active CPC research internships/rotations through the five thematic areas, a structured CPC Fellows Seminar Course, and additional integration with resources from Yale. Our program will select the most highly qualified and diverse candidates who wish to pursue a career in cancer prevention and control, and numerous procedures have been put in place to ensure the recruitment, quality mentorship, and retention of under-represented minority fellows. At the conclusion of their training, it is our goal that individuals supported by this T32 will have developed the knowledge and skills necessary to function independently, and that they will develop and carry out innovative, impactful studies filling key gaps in CPC research, potentially changing the standard of care and impacting policy. Ultimately, public health will benefit through the preparation of an exceptional cadre of investigators committed to careers in cancer prevention and control.