A Sharper Eye for Seeing Within

A man of vision, Michael Phelps first recognized PET imaging technology's potential -- and then fought to get it accepted

Michael E. Phelps has done his share of mingling with the rich and the famous: The late entrepreneur, philanthropist, and art collector Norton Simon was a friend. But the lessons that shaped Phelps's life were the skills he mastered in the rough Irish neighborhood in Port Orchard, Wash., where he boxed and brawled -- and missed out on whatever his school had to teach about chemistry and math because he was kicked out of class.

Now one of the world's most famous scientists, Phelps works at the crossroads of chemistry, biology, and math. At 62, he likes to say that his fighter's instincts got him where he is. "The world is about training and fighting," Phelps declares. "Training is a way of becoming, and fighting is about standing and delivering when the time arrives."


  Phelps delivered his knockout blow in 1974, when he and postdoctoral student Ed Hoffman invented the first positron emission tomography (PET) scanner. Unlike X-rays, which display only the body's structure, PET reveals chemical and biological processes within the body. It illustrates, for example, how the brain remembers and thinks, how the heart beats, and how the pancreas synthesizes insulin.

The invention created the field of brain mapping. It has also saved thousands of lives: PET scans change doctor-prescribed treatments in more than 15% of cancer cases, says Phelps. Where CT (computed tomography imaging) or MRI (magnetic resonance imaging) scans may proclaim a patient cancer-free, PET can take a closer look and detect a malignant growth. The device can also spot Alzheimer's disease 15 years before the first symptoms occur, according to Phelps's most recent research.

Today, only major hospitals own PET scanners, which cost between $800,000 and $3 million. But the worldwide market for the devices should explode from $1 billion in 2002 to $4 billion in 2006, according to a forecast by GE Medical Systems, as prices come down and the range of uses expands. Already, pharmaceutical companies are using miniature PET scanners to test new drugs on mice -- known as "the little patients" in the drug industry.


  PET imaging could eventually become a tool used in yearly physicals, thus helping doctors detect -- and treat -- diseases at their inception, says Richard Hichwa, director of the PET Imaging Center at the University of Iowa Hospitals & Clinics. In a PET scan, a patient is injected with a tiny amount of radioactive glucose, then wheeled through the doughnut-shaped PET camera. As the glucose spreads through the patient's veins, heart, and tissue, the camera takes pictures. Because cancer cells gobble up sugar, an image showing a spot of glucose absorption could indicate a malignant tumor.

Phelps became the father of PET quite literally by accident. After high school, he went to Western Washington State University, which he considered an Irish heaven -- full of women and booze. He partied through his freshman year. But then, Phelps was in a car accident. The front of his skull got cracked, leaving him dizzy and with constant headaches for about a year. "It took away normal things I wanted to do," he says. So, in his sophomore year, "I went back to fight."

He attacked math and chemistry with a vengeance after deciding those fields would give him the most advantage in life. He zoomed through grad school and, in 1970, received his PhD in chemistry. Then, Phelps started teaching at the Washington University School of Medicine in St. Louis, where he co-developed the first PET scanner. "Irish, we think in stories," says Phelps, who tends to end conversations and e-mails with his grandfather's proverbs and toasts. "I needed to put the pieces of the puzzle together to form the picture of our bodies' biological functions."


  Winning acceptance of the technology was a challenge. Insurance companies refused for years to cover PET scans, which can cost about $2,500 and may take as long as 90 minutes. By comparison, a CT scan requires 15 minutes and runs no more than $500, estimates the University of Iowa's Hichwa. But clinical studies showed that PET had an accuracy rate that was 9% to 43% higher than CT and similar methods, depending on the disease and the patient.

Finally, in 1997, the government approved PET for detecting lung cancer in Medicare/Medicaid patients. Since then, PET has been approved for other diagnoses, thanks in part to years of lobbying by Phelps and supporters such as Senator Ted Stevens, (R-Alaska). In 1981, Stevens became so engrossed in a conversation with Phelps about PET that he missed a speech he was supposed to give before the American Legion in Los Angeles. "I was mesmerized," says Stevens. "It seemed to me this was a breakthrough that could lead to other breakthroughs."

That could be the case. Scientists already use the scanner to learn how genes give orders to cells to grow, for instance. Even more advanced applications for PET could be coming, says Phelps, who is now chairman of the department of molecular and medical pharmacology, chief of nuclear medicine, and director of the Center for Molecular Medicine at the University of California at Los Angeles (UCLA) School of Medicine.


  In June, PET-scanner maker CTI Molecular Imaging (CTMI ), where Phelps is a director, went public despite the instability in the financial markets. Next up, Phelps hopes that PET scanners will help doctors discern molecular errors that cause diseases -- and help create the ways of correcting them. He is working with human genome guru Leroy Hood and nanotech great James Heath to create disease-specific drugs in as little as 24 hours, vs. the 10 years to 20 years it can take today.

The trio is developing a tiny "lab," which would allow scientists to examine the molecular nature of both normal and pathological cells. Their technology could be part of a handheld device or simply connected to a computer, but it would allow researchers to learn the nitty-gritty of molecular interactions.

That could come in handy, say, for fast detection of hereditary pathogens. The minilab could also spot a germ and send out an alert, describing the biological agent, which parts of the human body it attacks, and how it can be treated. It's conceivable the minilab might also determine the precise formula for treating diseases.


  Today, researchers have to try out up to 100,000 different molecules to find the handful that fight a particular disease. With the nanolab's precise data on how different molecules interact, the initial sample could be narrowed down to five, Phelps predicts.

In Phelps's perfect world, the PET technology will be used to test the drugs produced by such a streamlined discovery process. First, the molecules of the most promising medications will be tried out on genetically engineered mice to determine how they might react in humans. Later, PET could make human testing itself both safer and more accurate.

In the early stages of testing a drug now, human subjects are asked to take huge amounts, an approach that risks side-effects. With PET, the patient would receive only a small injection -- a 30-second jab -- of the drug laced with a trace of a short-lived radioactive substance that the PET scanner can observe as it either works or fails.


  Speedy creation of drugs is still a dream -- some would even call it a pipe dream. But Phelps likes to quote his art-collector friend Norton Simon. "Life has a natural curve," Simon told him. "You go up, plateau, and you go down. The only way to change this is to be continually starting new curves and, in this way, always remain in a state of becoming."

In 1999, Phelps received one of science's most prestigious honors, the Enrico Fermi Presidential Award. "He is the top living person in PET technology," says Prem Srivastava, a program manager at the Energy Dept., which finances some of Phelps's research. That's why he continues to work seven days a week: Phelps is training for his next fight.

By Olga Kharif in Portland, Ore.

Before it's here, it's on the Bloomberg Terminal.