H. Lee Moffitt Cancer Center & Research Institute
List of Collaborating Institutions
Mayo Clinic - Scottsdale
The Moffitt PSOC embraces the physical science research paradigm that has combined empirical studies with mathematically-based theoretical models that integrate and organize diverse data. We view cancer as an open, complex, dynamic, adaptive system. “Complex” because it contains a large number of components linked through a variety of mostly non-linear interactions. “Adaptive” because the tumor cells evolve and coevolve phenotypic strategies in response to their micro-environmental circumstance, each other, and eventually to therapy. “Dynamic” because all of the elements of the tumor, the cancer cell population sizes, adaptive strategies, and their interactions with each other and their environment change through space and time. “Open” because it continuously interacts with normal tissue in large part through its vascular network. Similar to Riparian ecosystems, blood vessels provide nutrients and growth factors to and remove metabolites from surrounding cancer population as well as providing a highway for the transport of tumor cells to distant organs.
We investigate the cancer as a complex system through an integrated, multidisciplinary approach that includes physicists, applied mathematicians, cancer biologists, computer scientists, oncologists, and evolutionary biologists. Like all living systems, we assume that cancer populations are ultimately governed by Darwinian dynamics which we view as first principles. Complex systems are notoriously difficult to predict as illustrated by the famous “butterfly effect.” However, we have found that predictive models can be developed through the integration of large data sets (similar to weather forecasting), evolutionary first principles, and sophisticated computational methods that ultimately lead to a deeper understanding of the underlying dynamics. Furthermore, we have demonstrated both theoretically and empirically, that the tendency of complex systems to magnify small perturbations can be exploited to guide cancer towards less aggressive outcomes.
Back To Top
Robert A. Gatenby, MD
Robert A. Gatenby, MD is the Chairman of the departments of Radiology at H . Lee Moffitt Cancer Center. He joined Moffitt in 2008 from the University of Arizona where he was Professor, Department Radiology and Professor, Department of Applied Mathematics since 2000. He received a B.S.E. in Bioengineering and Mechanical Sciences from Princeton University and an M.D. from the University of Pennsylvania in 1977. He completed his residency in radiology at the University of Pennsylvania where he served as chief resident. Bob remains an active clinical radiologist specializing in body imaging. While working at the Fox Chase Cancer Center after residency, Bob perceived that cancer biology and oncology were awash in data but lacked coherent frameworks of understanding to organize this information and integrate new results. Since 1990, most of Bob’s research has focused on exploring mathematical methods to generate theoretical models for cancer biology and oncology. His current modeling interests include: 1. the tumor microenvironment and its role in tumor biology. 2. evolutionary dynamics in carcinogenesis, tumor progression and therapy. 3. information flow in living systems and its role in maintaining thermodynamic stability.
Robert J. Gillies, Ph.D.
Dr. Gillies received his Ph.D. in Zoology from UC Davis in 1979, where he studied “Intracellular pH and Cell Cycle Regulation in Yeast, Tetrahymena and Sea Urchin Eggs” for his dissertation. Following this, he had a 3-year post-doctoral appointment at the Bell Labs and at Yale University under the same mentor (1979-1982), where he learned the basics of In Vivo NMR. He has held faculty positions in Biochemistry at Colorado State University (1982-1988), and in Biochemistry and Radiology at the University of Arizona (1988-2008), prior to his appointment as a member of the Moffitt Cancer Center (2008-pres), where he is chair of the Department of Cancer Imaging and Metabolism, vice-chair of Radiology and director of the Center of Excellence in Imaging and Technology. He has published over 243 peer-reviewed articles in the use of imaging and spectroscopy to interrogate tumor physiology, and is currently PI or co-PI on eight NCI grants or sub-contracts focused on cancer imaging. Professor Gillies has received local, national and international awards for his teaching and research, including the Furrow award for innovative teaching and Kettel award for mentoring (U. Arizona), the Yuhas award (U. Penn), the TEFAF professorship (U. Maastricht), distinguished Basic Scientist awards from the Society of Molecular Imaging and the Academy of Molecular Imaging, Researcher of the Year at Moffitt, and numerous named lectureships. He has served continuously on NIH Study Sections since 1998. Dr. Gillies’ log-standing research interests are focused on the detailed analysis of non-invasive imaging and spectroscopy to inform evolutionary models of carcinogenesis, cancer progression and therapy.
Alexander R. A. Anderson, Ph.D.
Alexander R. A. Anderson, Ph.D. is chair of the Integrated Mathematical Oncology (IMO) Department and Senior member at Moffitt Cancer Center. Dr. Anderson performed his doctoral work on hybrid mathematical models of nematode movement in heterogeneous environments at the Centre for Nonlinear Systems in Biology, a joint venture between the Scottish Crop Research Institute in Dundee and Dundee University, UK. His postdoctoral work was on hybrid models of tumor-induced angiogenesis with Prof. Mark Chaplain at Bath University, UK. He moved back to Dundee in 1996 where he worked for the next 12 years on developing mathematical models of many different aspects of tumor progression and treatment, including anti-angiogenesis, radiotherapy, tumor invasion, intra-tumor heterogeneity evolution of aggressive phenotypes and the role of the microenvironment. He is widely recognized as one of only a handful of mathematical oncologists that develop truly integrative models that have both changed the way biologists do experiments but also the way in which models are developed. Due to his belief in the crucial role of mathematical models in cancer research he moved his group to the Moffitt Cancer Center in 2008 to help establish the Integrated Mathematical Oncology (IMO) department. Since his arrival at Moffitt his focus has shifted to developing organ specific models of tumor initiation and progression that examine the key role of the microenvironment as a selective force in the growth and evolution of cancer. A common theme of these organ specific models is the importance of understanding normal organ form and function particularly in relation to homeostatic regulation and how cancer disrupts and exploits these mechanisms. Building models that can generate testable hypothesis and utilizing experimental data to parameterize
Back To Top
Cancer is an open and complex adaptive system evolves at is grows. This evolution is driven by phenotypic/genomic plasticity in combination with highly selective microenvironments. As cancers grow they can, in turn, alter the microenvironment and hence, can influence the trajectory of their own evolutionary dynamics. These microenvironments occur in spatially distinct regions within tumors that can be characterized as distinct “habitats”, which can be identified by modern imaging at multiple scales. The goal of Project 1 is to characterize the evolutionary dynamics using mathematical models informed by multi-scalar imaging and molecular analyses. The choice of model systems is driven by those with high translational potential, in consultation with Project 2 and the Bench-to-Bedside Core. We will focus our initial efforts on characterizing the evolutionary dynamics of prostate and breast cancers. Specifically, both prostate and breast cancers can be effectively treated with anti-hormonal therapies, but in current clinical practice, these approaches inevitably select for resistance. Forestalling the emergence of resistance using treatments based on evolutionary principles is a goal of the whole PSOC. In support of current and planned clinical trials, we will generate in vitro and in vivo data from co-cultures of hormone-receptor positive and -negative cells grown in 3-D culture and in mouse xenografts, and use these data to inform models being developed within the B2B core. In addition, we have identified distinct sub-types of cancer cells that co-exist within prostate cancers that are adapted to growth and survival under distinct microenvironmental conditions of pH and oxygenation. Importantly, these sub-types have distinct prognostic characteristic, i.e. some are more indolent while others are aggressive and invasive. We have observed in preliminary in vitro and in vivo studies that the evolutionary trajectory of some cancers can be altered towards a more benign phenotype by subtle microenvironmental perturbations. Hence, we will continue to investigate these approaches in a broader array of animal models in vivo and 3-D culture system in vitro with the ultimate goal of developing math model-driven clinical trials using this approach. A substantial effort will be expended to identify clinically achievable biomarkers to predict and monitor response to such approaches. Finally, in recognition of the fact that the immune system plays a critical role in the establishment and therapeutic response of cancers, we will undertake the above studies in immune competent mouse models and investigate the evolutionary trajectory of immune system components during treatment of the animal patients, and these will be followed up with targeted in vitro studies with isolated immune cells.
In the Moffitt PSOC, we view cancer as an open complex adaptive system that can be characterized using necessary and sufficient data, eco-evolutionary first principles and sophisticated computational methods. In Project 2, we examine the clinical application of that conceptual model. Here, we will build computational models of the system dynamics in individual cancer patients undergoing therapy with the ultimate goal of using those models to optimize the outcomes of increasingly complex cancer treatments We will focus our initial efforts on three active clinical trials in different tumors (multiple myeloma, prostate cancer and glioblastoma) treated with three different therapeutic strategies (multidrug chemotherapy, hormonal treatment, and immunotherapy). In each trial a multidisciplinary team will develop a patient-specific computational model of the intratumoral evolutionary and ecological dynamics, based on available clinical data, which govern response and resistance to therapy in each patient. Using an iterative approach, each team will work with the Bench-to-Bedside (B2B) core and Project 1 to optimize the models’ predictive power by exploring alternative mathematical methods and inclusion of novel data elements from each trial (e.g. molecular analysis of primary tumors and circulating tumor cells). In follow-on trials using the same tumors and treatment strategies, we will both validate model predictions and apply variations in the treatment approaches (or combinations of approaches) suggested by pre-clinical experiments in Project 1 and model simulations in the B2B core. Finally, we will develop the infrastructure necessary to develop and oversee evolutionarily-informed, computationally guided clinical trials at Moffitt.
Back To Top
The process of using mechanistic models to assist with understanding experimental results and informing clinical decisions depends on a suitable framework for developing, analyzing, implementing, adjusting and presenting the models and results. Furthermore, models must be calibrated so that they recapitulate experimental and clinical observations both in terms of their range of behavior as well as the intrinsic variability of outcomes before they can be used to specifically impact one or more clinical decisions (e.g. predict novel treatment strategies). There is also a practical aspect of turning theoretical predictions into actionable clinical decisions that requires tools to generate patient cohorts, mimic clinical trials and present results in a visual and easy to use manner such that clinicians can immediately understand and manipulate them. The central focus of the Bench-to-Bedside Core is therefore to integrate experiments, clinical data, and models to assist with hypothesis generation, testing, and validation (Project 1), and translate successes into a clinical setting using clinically-relevant tools (Project 2). This will be accomplished through: Core models that explore unifying principles of evolution, heterogeneity, and response to treatment; Core tools, Radiomics, Pathomics, and a generalized set of analysis tools for integrating experimental and clinical data to facilitate mathematical model development and calibration; Core decision support tools including Phase “i" trials and dynamically optimized therapy for use in clinical trials. The tools and methods we develop here are generalized and therefore suitable to a wider class of experimental and clinical data and will serve as a key legacy from this project to drive bench-to-beside science.
Back To Top