Cancer drugs make almost everyone sick as a dog from toxic side effects, yet they have a hit-or-miss record when it comes to killing cancer cells. Jim Olson is a pediatric oncologist, so he’s well aware of the suffering and waste created by this approach. He prescribes these nasty toxins for young kids with only a slim chance of benefit.
What if he could invent a way that doctors could tell whether a tumor was responding to a certain kind of treatment, or whether the cancer cells were resisting it? Maybe doctors could move more quickly, quit wasting everyone’s time, and switch to a drug that might actually work.
Olson, a researcher at the Fred Hutchinson Cancer Research Center and physician at Seattle Children’s Hospital, has led a team that has developed a prototype device that works in mice. The technology has been spun out into a new company called Presage Therapeutics. The company hopes to introduce a new way of personalizing cancer treatment that any physician can use, without having to know anything special about a patient’s genetic profile.
Cancer treatment is a big business, with a global market worth $66 billion a year, and predicted to grow to $84 billion by 2012, according to Cowen & Company. There’s a lot of talk about bringing down costs, and waste, by personalizing cancer treatment based on genetic profiles of people who are more likely to respond. One example doctors are debating is whether to prescribe Eli Lilly’s cetuximab (Erbitux) or Amgen’s panitumumab (Vectibix) strictly for patients with a normal form of a tumor growth gene called KRAS. That’s because analyses on clinical trials show that about 60 percent of patients have mutated forms of KRAS, and they get no benefit from the drug, and lots of side effects. Yet a lot of people get these drugs—at tens of thousands of dollars a pop—without any realistic chance of benefit.
“There’s a huge hunger out there among patients about how they’re doing on a given drug,” Olson says. Referring to the genetic approaches toward personalizing cancer treatment, he says, “The promise of that has not been realized.”
But the Presage method doesn’t depend on that kind of sophisticated genotyping like with the KRAS example, or on biomarkers of which genes are turned on or off. Here’s how the technology is supposed to work:
A prototype, which Olson showed me in his lab, has five porous needles with holes along the sides. These five needles can deliver five different kinds of chemotherapy drugs—or combinations of experimental biotech treatment—to different localized regions of the tumor. The drugs are made to seep out within a small radius of the needle, so doctors can see how different regions of the same tumor respond to different drugs. If one part of the tumor is clearly being killed by a certain drug combination, and the other four aren’t having much effect, the doctor will know which drug to prescribe with confidence.
“We see this as dramatic innovation in cancer care,” says Thane Kreiner, Presage’s founding CEO.
Doctors have tried for decades to simulate
