In the words of Russian scientist Theodosius Dobzhanski, "Nothing in biology makes sense except in the light of evolution."
In the words of Dana-Farber scientist Franziska Michor, PhD, "Why should cancer be the one exception to this rule?"
The notion that tumors are subject to the same evolutionary forces as the necks of giraffes and the wings of hawks has intrigued cancer scientists for nearly a century, but only in recent years, with the introduction of technology for speed-reading the DNA of cells, have researchers been able to capitalize on it. An evolutionary perspective allows scientists to study how tumors change and adapt over time – how they respond to stress and, potentially, how their impetuous behavior can be tamed with therapies.
At Dana-Farber's Physical Science-Oncology Center (PS-OC), where Michor is principal investigator, researchers use this approach to tackle some key questions about cancer. One project is examining the mechanisms by which gene alterations accumulate in cancer cells, information that can help doctors predict how the disease will progress and which therapies are likely to be most effective against it.
Another project aims to identify the cells at the root of human cancers. Researchers are exploring whether cancer originates in stem cells – the living clay from which other cells arise – or in the more "differentiated" cells that comprise the body's many organs and tissues.
It is in the realm of drug resistance, however, that the PS-OC is set to have its most immediate impact on the treatment of cancer patients. Resistance, the process by which tumors bounce back from drugs that once were detrimental to them, has long been a weak point in cancer therapy, limiting survival rates and often dampening the promise of new treatments. Using evolutionary principles, with a generous assist from applied mathematics, Michor and her colleagues have devised a strategy for slowing or even halting drug resistance in certain patients.
"We're interested in determining which genetic alterations cause cancer, which increase cancer cells' ability to reproduce and spread, and how we can use that knowledge to improve treatment," says Michor, who helped establish the PS-OC at Dana-Farber in 2009. "By applying evolutionary theory to the study of cancer cells, we want to identify the mutations that are driving the disease – and make the best targets for new therapies – and those that are simply 'passengers.'"
Michor's unconventional approach to biological science reflects her unique background. The daughter of a mathematician father and a nurse mother, she earned a PhD from Harvard at age 22. She "wanted to combine the area of expertise of my father with the humanitarian mission of my mother, to help come up with a quantitative approach to cancer."
To understand how evolution can shed light on drug resistance, it's useful to view cancer from a different vantage point than the customary focus on individual cells. It requires a closer look at group behavior – at how populations of cells are shaped by their environment – and a reckoning with forces that extend far beyond the life of a single cell.
In one sense, cancer represents a turning back of the clock about 3.8 billion years, to a time when life is thought to have first appeared on earth. The original life forms were single-celled creatures whose sole concern was their own survival and reproduction. Footloose and unencumbered, they had no need to coordinate their activities with any other cell. That changed with the rise of multicellular life in the Precambrian period about a billion years ago, when cells had to form "pacts" to cooperate with one another for the good of the entire organism.
"Cancer represents, in essence, the breakdown of this pact," Michor remarks. "With cancer, we have cells that behave as if they're not part of a multicellular organism anymore. They're selfish: they ignore all the commands that are meant to keep the agreement intact."
Tumors result from mutations in genes that are the guardians of this pact. When enough mutations have occurred in genes that restrain a cell's growth or prevent it from invading neighboring tissue, the cell becomes an outlaw, a renegade cancer cell.
If evolution once led cells to organize into multicellular creatures, it also, by a perverse logic, underlies the development of cancer cells, which can sometimes be hardier and scrappier, better equipped for survival than normal cells are.
When thinking about how evolution operates on cancer – and how it can be turned to therapeutic advantage – it helps to recognize that tumors are not collections of identical cancer cells, but aggregations of different kinds of cancer cells, some differing only very slightly from the others. "Cells within a tumor are always evolving, always acquiring new mutations," Michor explains carefully. "Some have Mutation A, some may have Mutation B – there can be an enormous variety. Each population of cells responds to drug agents differently."
In classic evolutionary fashion, the cells not killed by a particular drug gradually come to dominate a tumor as their more-susceptible cousins die. These resilient cells pass their survival skills on to their offspring. Over time, a drug that originally decimated cancer cells becomes powerless against a more transformed tumor.
"Evolutionary pressures drive these tumors to be more and more aggressive and invasive," Michor says. "The thrust of these pressures is toward the benefit of the tumor but the detriment of the patient."
The good news is that unlike earlier generations of cancer scientists, today's researchers have developed models that describe in a very precise way how this evolutionary process unfolds. The core of these models, the language in which they are written and the physical laws they express, is a set of mathematical formulas.
In retrospect, Michor relates, it's bewildering that math has come so late to the study of the basic biology of cancer. After all, "Math permeates everything; it's the foundation of physics, we use it to map the movement of the stars," she comments. "It gives us a set of rules for describing physical forces and making predictions. Why hasn't it been used more widely in medicine?"
In devising equations to study drug resistance in cancer, Michor and her colleagues accounted for several factors: the rates at which mutations arise in populations of cancer cells, and the growth and death rates of cells exposed to different doses of a drug. The formulas – all algebraic symbols and Greek letters, like something scrawled on the blackboard of a physics lab – allow researchers to calculate the chance that drug resistance will occur, and how quickly.
In a recent study, Michor's team used their mathematical modeling strategy to predict how frequently a drug targeting a common form of lung cancer should be taken to slow or even prevent the emergence of resistance. "What if, instead of one pill a day, every day – the dosing schedule approved by the Food and Drug Administration for this agent – patients would be better off taking two pills every second day, or three every third day, or some other combination?" Michor asks. The model allowed researchers to test millions of different possibilities simply by running the variables through a computer.
The results: the one-a-day schedule was not the best of all the alternatives for delaying the emergence of resistance. The optimal schedule, which pushed resistance back the furthest, is now being put to the test in a clinical trial. Patients will take the drug according to either the standard or the new, optimized schedule, and researchers will track which group develops resistance first.
"Mathematical modeling of this type won't replace clinical research or animal studies of cancer," Michor observes. "But it provides a tool for speeding up the process of determining which treatment method may be most effective." If that tool proves to be a hybrid approach as distinctive as Michor's own background and ambition, few will be surprised.
Paths of Progress Fall/Winter 2012 Table of Contents
Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215 | Call us toll-free: