Department of Biostatistics
Training Program in Cancer Research
The role and future of biostatisticians in cancer research: Rationale for the training program
It is widely recognized that biostatistical collaboration and methodology are essential components of research into the underlying mechanisms, causes, risk factors, and therapeutic interventions for cancer. The long term aim of this program is to increase the participation in cancer research of biostatisticians who are educated not only in the powerful methods of modern statistics, but also in the biology, genetics and epidemiology of cancer, the current body of knowledge about the etiology of the disease, its natural history, prevention and treatment.
Cancer is a leading cause of mortality and morbidity among people throughout the world. It has a devastating effect on individuals and the society in which they live. Over 40 percent of us will develop cancer; over 20 percent of us will die from cancer. There is a tremendous amount of active and ongoing research into understanding the basic mechanisms of cancer and developing methods to prevent and treat cancer. The science of cancer is multifaceted, as illustrated by the breadth of scientists involved in the research efforts to fight cancer which includes pathologists, immunologists, virologists, population scientists, geneticists, radiation biologists, and clinical trialists, to name a few.
There has been a long tradition of involvement by statisticians in cancer research. A major emphasis and achievement of statisticians in cancer research has been in developing the discipline of clinical trials and in developing methods for epidemiological studies. These areas can and will continue to play important roles. The phenomenal expansion in the knowledge of the underlying mechanisms of cancer indicates a need for statistical scientists with a greater understanding of the science of cancer. The astounding advances in molecular biology and genetics will have a substantial impact on the role that biostatisticians play in the future. They are likely to be closer to the cutting edge in advances in basic science. One area where they will play an important and crucial role is in the emerging field of bioinformatics.
In the future, biostatisticians will be working more and more in multidisciplinary cancer research teams. A premise of this training program is that such teams will be greatly enhanced if the biostatistician is biologically knowledgeable. This will be especially the case if these biostatisticians have a substantial knowledge of the molecular biology, the biological mechanisms and the genetics of cancer.
With the advances in medical science and the associated technology, the types of data which statisticians are seeing and the types of issues they face are becoming more complex. As medical science becomes more complex the demands for statistical expertise will increase as well. In modern biostatistical research, the statistician scientist must be trained not only in the theory of statistics, but also be an interested and knowledgeable scientist, as well as having the interpersonal skills to work with investigators who are not trained as statisticians.
The outcomes used in clinical trials and epidemiologic studies are becoming more diverse. There is a need for more advanced statistical methods and individuals that understand their use and interpretation in the context of cancer.
Another example of the advances in medical science, which are likely to have a profound effect on biostatisticians, is in the area of drug development. Advances in combinational chemistry and high through-put assays will inevitably lead to more potential therapies to evaluate on the limited number of patient resources available. These high through-put technologies will lead to very high dimensional datasets, an unfamiliar problem for the classical clinical biostatistician. Future therapies are also likely to be increasingly based on gene and protein targets. Thus, genetic or protein expression data will be the likely type of information used to evaluate therapy as well as the traditional response rate and survival time.
We believe that the training program is quite novel and unique amongst departments of Biostatistics. The program is designed to respond to the changing environment for biostatistics, and the special need for biostatisticians to understand biology more than ever.
The overall purpose of this training program is to provide biostatisticians with the requisite scientific knowledge to understand current issues in cancer research, and to provide training in statistical and epidemiological techniques and research methodology related to cancer. The methods of education will include formal coursework in biostatistics, epidemiology, and biology relating to cancer; interdisciplinary seminars on current research and biostatistical topics in cancer research; a lecture course on “Statistical Methodology in Cancer Research”; a “Statistics in Cancer” workshop/journal club; mentored research in collaboration with cancer investigators; and presentation of research products in national statistical and cancer conferences.
Training environment at the University of Michigan
The primary biostatistical training facility will be the Department of Biostatistics in the School of Public Health at the University of Michigan. The department is a top ranked department in the United States, with many internationally regarded faculty, and a vibrant research and teaching program. It is one of five departments in an outstanding School of Public Health. The department has close links with the Statistics department and the Bioinformatics program at the University of Michigan. It is located in close proximity to the University of Michigan Medical Center and the Cancer Center.
The Department of Biostatistics is a leader in the development and application of statistical methods in the biomedical sciences. It is an expanding department with 25 faculty and over 100 graduate students. The faculty include leading researchers in the areas of survival analysis, missing data, longitudinal and correlated data analysis and statistical genetics. It has extensive collaboration with the University of Michigan Medical Center in many areas, including cancer, diabetes, kidney disease, imaging, genomics and alzheimer’s.
The Comprehensive Cancer Center at the University of Michigan is a very successful Cancer Center with over 70 million dollars in grant funding. In 2008, the University of Michigan was ranked third in cancer research funding from the NCI. It has an extensive research program, including clinical research in many disease sites (prostate, breast, head and neck, skin cancer leukemia, etc.). It has successful programs in basic science, behavioral science, epidemiology and genomics. It has very active research programs in imaging, genomics and proteomics.
Many faculty and students from the department of Biostatistics are involved in cancer related research projects.
The list below gives the cancer related research of key faculty in the department of Biostatistics:
Jeremy Taylor: methods for evaluating biomarkers, cure models, pahe I design, longitudinal modelling and genomics.
Alex Tsodikov: demography of cancer, cellular stochastic models and survival analysis.
Sinae Kim: Bayesian methods for clustering and analysis of gene expression data.
Tom Braun: novel designs for phase I studies.
Tim Johnson: Bayesian modelling of spatial and imaging data.
Bhramar Mukherjee: methods for estimating gene-environment interaction.
Rod Little: missing data, survey methods of casual modelling in cancer.
Kerby Shedden: computational methods for high dimensional genomics and imaging data.
Mousumi Banerjee: health services research, tree based survival analysis methods.
Steve Qin: bioinformatics, methods for sequence data.
Yun Li : casual inference.