Dr. David Agus, the author of The End of Illness and a professor of medicine and engineering at USC who treated Steve Jobs, is a frustrated man. He points out that despite all the billions poured in R&D we’ve barely made a dent in reducing cancer rates. Agus believes we need an entirely different approach to battling cancer. What if you could tell when your body first starts to grow cancer and then cut it off early? One way to do that, he believes, would be to read our proteins — those building blocks of life that can give us a real-time picture of what’s actually going on in our bodies. Says Agus, “we talk about cancer as a noun, as if it’s a one time event: “I’ve got cancer.” But the body is changing all the time. We should think about it instead as a verb: “I’m cancering.” The trick would be to pick up on early signs of when a person started to “cancer” and then nip it in the bud.
In theory that sounds promising. In practice, there’s a huge problem: How do you measure the body’s hundreds of thousands of proteins, which are changing constantly? Dr. Agus figured he needed massive computing power to do that so he called Danny Hillis, the inventor and entrepreneur who founded Thinking Machines, a company that helped to develop parallel supercomputing, and who later worked for Disney’s Imagineering animation division. Hillis didn’t return his calls. The computer whiz, who now runs the consultancy Applied Minds, then got calls on the same day from Al Gore and John Doerr, urging him to return Agus’s call. So Hillis did. “David says to me, ‘I see people dying everyday, and we have to have a window into proteins.'” Hillis was sold.
In 2007, Dr. Agus and Hillis formed Applied Proteomics, to see if they could crack the protein code. Today CEO Peter Klemm says the company has results from ongoing clinical trials that are so promising that they should have a diagnostic test for colon cancer on the market next year.
Here’s how Applied Proteomic’s cancer test works: Your system starts behaving differently when you grow polyps, and that’s reflected in your proteins. The trouble is, proteins are changing constantly and have been beyond the capability of humans to measure them. So Hillis built an automatic assembly line of specialized robots that uses mass spectrometry to capture and catalogue the 360,000 different protein pieces found in a patient’s blood sample. Then he applies the massive power of parallel supercomputers to analyze that data. Says he: “We look for patterns in the population. We look at individuals and compare them to a healthy control group. It’s giving us a picture not just of the disease but also of the health of the whole body.” One reason this is possible is that today’s computers are 10,000 times faster then a few years ago. Adds Hillis: “We couldn’t have done these calculations 10 years ago; and even five years ago it would have been too expensive.”
The stakes are high. Some 150,000 Americans are diagnosed each year with colon cancer, and 50,000 die of it, costing the health care industry as much as $40 billion a year. One reason the death rate is so high is that people won’t get routine colonoscopies because the procedure is so unpleasant. Applied Proteomic’s blood test, which will cost only a few hundred dollars, can identify who really needs the procedure. If the test comes back positive, this should encourage the patient to get a colonoscopy to remove the polyps, which in the vast majority of cases stops the cancer in its tracks. CEO Klemm estimates that the colon cancer testing market could grow to as much as $3 billion annually.
The company, which as raised a total of $57 million, in May closed a $27.8 million series C financing. Applied Proteomics is backed by a bevy of heavyweights including Microsoft co-founder Paul Allen and the life science investment firm Domain Associates, whose founding partner, Jim Blair was an early investor in Amgen. Most recently Genting Berhad — a Malaysian conglomerate involved in gaming, energy and biotech — became a backer. Chairman and CEO of Genting Berhad Tan Sri Lim Kok Thay was appointed to the API Board of Directors.
Going forward, Agus and Hillis believe the technology could potentially be used for early detection of other kinds of cancers, as well as heart disease and Alzheimer’s. Hillis also believes that someday the technology could even be adapted to treat the diseases themselves. That’s a tall challenge. Cancer researchers have spent decades and billions battling cancer through traditional methods such as chemo and radiation. Getting them to accept that proteins hold the key to cancer will be a daunting task. That doesn’t faze Hillis: “This is the future of medicine.” Some big-time investors are betting he’s right.
A shorter version of this story appeared in the September 2, 2013 issue of Fortune.
