When a patient comes into the ER with a fever and a pain in her side, she’s often identified as having a potential case of appendicitis. But a definitive diagnosis requires more work, more money, and more time. Ultrasounds and high-dose radiation CT scans can help narrow it down, but even with these tools, at least one in four women diagnosed with appendicitis has a healthy appendix removed.
In this and a multitude of other conditions, faster and more accurate diagnoses can save time, money, even lives. That’s the premise on which Applied Proteomics was founded. And although it’s still in its very early days, the company’s first trials appear to have gotten off to a promising start.
Applied Proteomics was founded by two men with no shortage of accomplishments after their names: oncologist David Agus, who leads the prostate cancer center and the molecular medicine center at the University of Southern California; and inventor and computer engineer Danny Hillis, a one-time VP of research and development at Disney Imagineering and creator of one of the world’s first parallel supercomputers. Together, they’re doing what no one else has successfully accomplished: They have created a system that is capable of identifying suites of proteins in blood, saliva, and other bodily fluids that are unique to specific diseases.
There are millions of different proteins in our bodies, and levels of these proteins can fluctuate on a daily or even hourly basis. They change from morning to evening, from pre- to post-exercise, they can even vary depending on what you ate for breakfast. Proteins are also incredibly sensitive markers of disease—they’re one of the first things to change when the body’s systems aren’t working quite right. “Proteins are a great place to look for diagnostics because they’re the sum of both genetics and environment,” says Peter Klemm, CEO of Applied Proteomics.
The problem is that accurately measuring the levels of hundreds of thousands of proteins at once has been a tricky thing to do. Many have tried but the ability to filter out signal from noise, and the processing required to do so, proved remarkably difficult to get right. “Agus and Hillis saw it as something that required amazing attention to detail—they figured out all the noise, all the steps, and built a tool that allows us to dial it so that the noise decreases dramatically,” Klemm says.
From a sample of blood, the researchers remove all of the body’s most abundant proteins. Then they break the remaining proteins down into their smaller building blocks and run them through a machine, called a mass spectrometer, that separates the proteins by both their weight and how slippery they are. Then, they dive into the resulting data to find patterns.
So far they believe they’ve successfully identified markers for appendicitis, polycystic ovarian syndrome, and colon polyps—the last of which is a risk factor for the future development of colon cancer—and they have their sights set on many more, including various cancers.
Beyond that, Hillis and Agus believe that it should be possible to use their protein data to better understand why cells are turning cancerous, and which treatments might be best suited to treat specific cancers. “With this discovery engine, we’re in a position to really capture the full potential of the proteome,” Klemm says.