You see, it's hard to be smart without math. It's not that you don't know anything; it's that what you know doesn't help you solve the problems that are staring you in the face. Take physics, for example. For centuries, physics was a descriptive science. It dedicated itself to understanding the movement of the spheres -- the mysteries of planetary motion -- but it could do little more than say, They move. It wasn't until Isaac Newton came along and invented calculus that physics could not just describe where the planets were yesterday but predict where they were going to be tomorrow (...)

And that's where medicine is right now. "If you like science but don't like math, you go into medicine," Franziska says, and the problem with that is this: Evolution likes math. Cancer likes math. Both of them derive their power from mutations that increase exponentially over time. And so medicine is, well, outnumbered. It has battled cancer heroically for decades, but, as Franziska says, "even with all that time and all that money, it has not made very much progress, right?"--Esquire 2007

As cancers go, CML is pretty simple. It results from a defect in a single gene instead of a complex series of them. Its simplicity is what has allowed medicine to develop a drug for CML that is both very specific and very effective, a drug that targets the protein produced by the single defective gene. The drug is called Gleevec, and it's probably the best cancer drug that medicine's got. The only problem is that once people with CML stop taking it, the number of cancerous cells in their blood goes right back to the levels that existed before they started treatment, or even to much higher levels. It was a frustrating problem in what was considered a medical success story, and to figure out what was going wrong, Franziska did not do any laboratory experiments. Nor did she rely on any invasive procedures with CML patients. Instead, she got the numbers, then she did the math. She got blood counts from a collaborator in Australia and figured out how the cancer was responding to Gleevec over time. Once she had the data, she wrote down a series of equations that established, mathematically, the following:

a) that Gleevec was not affecting the population of stem cells where the cancer originated;

b) that Gleevec, while effective, could never cure CML; and

c) that even when medicine is relatively smart and cancer relatively dumb, cancer is still smarter.

One of the data sets she's been receiving is the data on patients with a type of brain cancer called glioblastoma. It is the most common form of brain cancer, and also the most dangerous, because it is the most diffuse. The tumor opens itself up in the brain like a fist opening to fingers, and by the time it is diagnosed, it is powered, Franziska says, "by hundreds of mutations." Often, glioblastoma is untreatable; generally it is inoperable; and almost inevitably it is fatal, with less than 50 percent of those who have it living six months past diagnosis. And one of the reasons for its intractability is its mathematical complexity; physicians faced with the tumor's brutal brand of exponential reasoning don't even know where to start. But Franziska does. She's been working to devise a mathematical model of glioblastoma, if only, she says, to create "a natural history of the tumor," a systematized vision of the evolution of chaos. It is another thing that she is doing that has never been done, and she is doing it so that maybe "we'd know which mutations to treat.