Funded Grants

Cancer therapy: Perturbing a complex adaptive system

With few exceptions, metastatic cancers remain incurable. The personal and societal burden from cancer cannot be overstated. The World Health Organization estimates that in the next 10 years, 84 million people will die of the disease.

Eradicating the large, diverse and adaptive populations found in most cancers presents a formidable challenge. One cc of cancer contains about 109 transformed cells, so that there are more cancer cells in 10 grams of tumor (a modest size by oncology standards) than there are people on Earth. Unequal cell division and differences in genetic lineages and micro-environmental selection pressures produce complex evolutionary dynamics that result in genetically and phenotypically diverse tumor cell populations. Additionally, tumors are complex ecosystems that include normal cells and blood vessels. That latter are typically dysfunctional resulting in regions of diminished blood flow and oxygen and increased acid concentrations. Cancer, in other words, is a complex, evolutionary system.

While the biomedical research community’s “war on cancer” has, in many ways, been remarkably successful, these advances have had only slight impact on the mortality rates from most common disseminated cancers. This apparent disconnect is both an illustration and result of the complexity and diversity of cancers. Each individual tumor represents a new and novel disease generated by an idiosyncratic pathway of in-vivo selection and cellular adaptation. As a result, genetic lineages differ between cancers and even within populations of the same tumor. It is, thus, not surprising that the corresponding experimental data is both vast and difficult to synthesize.

This situation is not unprecedented and is, in fact, typical of experimental observations in complex systems. In the 16th century Tycho Brahe amassed huge data sets on planetary motions but their orbits seemed to follow no apparent rules within the prevailing paradigm that the earth was the center of the universe. In the early 20th century, the Balmer lines were observed. These were inexplicable within the conventional model of the atom. In the 1950’s, experimental physicists uncovered a bewildering and unruly “zoo” of subatomic particles. Understanding, in each case, required development of comprehensive theoretical models that encompassed the experimental observations but were framed mathematically from first principles (heliocentricity and gravity, quantum mechanics, and the “standard model” respectively). This lockstep of first principles and empiricism (observations and experiments) leads to paradigm shifts, and the understanding of complex system dynamics as the manifestation of generally simple rules operating in real systems with many interacting parts (the kaleidoscope phenomenon).

These experiences in reconciling seemingly impossible masses of data lead with first principles leads to our fundamental hypothesis: Cancer is clearly very complex, but it is not incomprehensibly so. Our proposal is that substantive advances in cancer therapy will require a similar comprehensive, quantitative theoretical model (4,5). Theodosius Dobzhansky famously stated "Nothing in biology makes sense except in the light of evolution". Similarly, we propose that cancer, perhaps uniquely among human illnesses, is a disease that arises through Darwinian interactions of microenvironmental selection and phenotypic adaptation, which are both causes and consequences of its complexity and heterogeneity. Interestingly, Dobzhansky also (and less famously) stated “Scientists often have a naive faith that if only they could discover enough facts about a problem, these facts would somehow arrange themselves in a compelling and true solution.” A key lesson from previous scientific investigation of complex systems is that data alone are not sufficient for understand the underlying dynamics. Complex, evolutionary dynamics can only be understood through application of suitable non-linear mathematical models.

The specific goal of this proposal is to integrate the first principles of evolution by natural selection (heritable variation in tumor cell lines, a struggle for existence within and among tumor cell populations, and this variation influences how cell lines fair in the struggle) into quantitative and experimental methods, and into new strategies for cancer therapy. We propose that cancers are subject to the same governing principles found in the evolution of populations of plants and animals in the natural world. These rules must, of course, be adapted to cancer’s unique in-vivo adaptive landscapes, but nevertheless can provide the necessary framework to guide therapeutic strategies. Such a process will not be easy