Spotlight - Franziska Michor

My most important collaborator at Princeton was the theoretical biologist Yoh Iwasa from Kyushu University in Japan. Yoh has spent his career studying how evolutionary theory can be applied to biology, ecology, and immunology, examining diverse aspects of life using mathematical approaches. Together, Yoh and I were thinking about questions concerning the dynamics of tumor suppressor gene inactivation and the emergence of genetic instabilities. Mathematical models of these processes provide insights into the mechanisms of cancer initiation. During this time, we developed a stochastic process model that allowed us to calculate the probabilities of accumulating different types of mutations, and to evaluate the importance of genetic instabilities for tumor initiation. 

CML Collaboration

My first term at Harvard saw me continuing my studies in evolutionary theory. I became particularly interested in a blood cancer known as chronic myelogenous leukemia (CML). One of the principle researchers working on CML was Charles Sawyers, who at the time was at UCLA but is now at Memorial Sloan-Kettering. I contacted Charles and proposed a collaboration that would allow me to look mathematically at the dynamics of treatment response in CML.

Unlike many cancers, CML is caused by a single mutant protein, BCR-ABL, which results from a genetic abnormality called the Philadelphia chromosome. In the 1990s, understanding how BCR-ABL activates the cascade of intracellular signaling that leads to CML, investigators including Dr. Sawyers began testing the drug imatinib (Gleevec®), which blocks the mutated protein's activity. Gleevec's initial response was remarkable, but over time patients were developing resistance to the drug or, in the absence of resistance mutations, continued to display low levels of cancer cells in their blood streams.

Dr. Sawyers put me in contact with Tim Hughes, a molecular biologist at the Institute of Medical and Veterinary Science in Adelaide, Australia. Tim was involved in a large international clinical trial comparing Gleevec to one of the earlier drugs against CML.

I received a dataset of 169 patients from the resulting patient database. I was interested in studying the treatment response to Gleevec and in looking at the CML cells that remained after successful treatment with Gleevec. Were these remaining cells cancer stem cells, the subset of cancer cells that we believe drives a cancer? Why wasn't Gleevec successful against them? And what are the dynamics of relapse due to resistance mutation?

Treatment Resistant CML Stem Cells

Using a DNA amplifying process known as quantitative polymerase chain reaction (PCR), combined with our detailed understanding of precisely how Gleevec works, I wanted to develop a mathematical approach to answer those questions. What we learned was that Gleevec caused a biphasic exponential decline in leukemic cells during the first year of treatment. The molecular response to the drug seemed to suggest that CML can be described by a mathematical model consisting of four separate subpopulations of cells: progenitor cells, which survive on average 125 days during treatment; differentiated cells, which appear to live for 20 days; terminally differentiated cells, which live an average of one day; and leukemic stem cells, which are not depleted by Gleevec therapy.

In patients who discontinue Gleevec therapy, the leukemic cell count rises within weeks to levels equal to or greater than their pretreatment levels -- suggesting that these leukemic stem cells do not significantly decrease during treatment and are responsible for the residual disease.

CML as Guinea Pig

Because CML is a relatively simple cancer to study, it can serve as sort of a guinea pig for cancer research. If we understand how this simple cancer develops, we can take that knowledge and apply it to more complicated cancers. I hope that the mathematical model I developed for CML will have practical applications, such as helping to find the cause of CML stem cell insensitivity and to design clinical trials.

For my research, after completing my PhD in 2005, I was granted a three-year independent research position in Harvard's Society of Fellows. The freedom of this position allowed me to continue my cancer research while working at the Dana-Farber Cancer Institute in Boston.

When the fellowship ended in 2007, I wanted to find the best place to continue and expand my work. I chose Sloan-Kettering Institute because it is uniquely connected to Memorial Hospital. As patients are being treated, researchers have access to clinical samples and clinical data, and we are able to collaborate with physicians.

Collaborations Galore

Memorial Sloan-Kettering and its people are very friendly and willing to collaborate. I continue to work on CML with other investigators at SKI, and I've also begun collaborative work on a brain cancer known as glioblastoma. I'm interested in designing an evolutionary framework that would allow us to investigate the role of a particular mutation in tumorigenesis. Right now we have some statistical approaches to these questions, but I would like to find an evolutionary theory based on a mathematical model.

Seeking to Optimize Treatment Strategies

One final facet of my research involves the search for a framework that would allow us to optimize treatment strategies for cancer patients. The two examples of this would be: first, delivering a chemotherapy drug at a low enough dose that can be sustained over time to avoid side effects; or, second, delivering high doses of a chemotherapy drug over short periods of time followed by a drug holiday to let the patient recover from toxicity. To investigate which of these options is less likely to produce resistance mutations, a mathematical model based on stochastic processes is necessary. I am now collaborating on a lung cancer mouse model to test the predictions of this model in an in vivo setting. 

I am happy to take advantage of all the tremendous opportunities available here at Memorial Sloan-Kettering. It is an exciting time to be working in the growing field of evolutionary cancer biology, and I hope that some of my findings will lead to clinical advances.