There's just a hint of melancholy when Dr. Guy Zimmerman, professor of internal medicine and director of the Program in Human Molecular Biology and Genetics at the U., tells the story of Pafase.

In the early '90s, he was part of a group of researchers who worked with a biotech company to clone an enzyme with anti-inflammatory properties. It had — still has — potential to interrupt key events that may influence the pathological process in various parts of the body.

It was an engineered version of a naturally occurring enzyme. In Phase I clinical trials, it was spectacular.

Here's the time line: For about 10 years, they worked on the biology, cloning it in the mid-'90s. By '97, the biotech company was several million dollars into it and ready for the small Phase 1 trial. Government estimates put the total cost of bringing a new drug to market at more than $300 million.

Which inflammatory disease would be the right target? They debated and finally zeroed in on sepsis (blood poisoning), a common, complicated condition that often kills.

Zimmerman had advised a number of companies against targeting sepsis because it is so complex. But there was a strong rationale, based on preclinical data. "If it worked, it would be a huge breakthrough."

In the Phase II clinical trial, participants were patients who had sepsis and a trauma with complications, such as acute respiratory failure. In such cases, half the patients normally die.

The evidence suggested that, introduced at the right time, Pafase could shut down the cascade of events.

The study took place in multiple centers, with 240 patients. Again, "spectacular" results.

"We were ecstatic about that," physician-researcher Zimmerman said.

The Phase III trial was multicountry, multicenter, mega-expensive. Then, a little over a year in, the regularly scheduled independent review of the 1,000 participants already enrolled found no significant improvement with Pafase. The study slammed to a stop.

The U. officials had no financial interest in the outcome, but they cared. And they wondered: Had the study simply chosen the wrong illness? The wrong timing? The wrong dose? Those questions are still on the table.

The biotech company was fortuitously involved in something else that paid off, so it survived. And moved on.

The U. has an agreement to do some animal model research into sepsis. "It could ride again. But once a drug has a first big mark against it . . .," Zimmerman said.

The hope that drives research is not easily killed.

"I am a physician, and I come at this as a physician. I am not doing investigational work because I don't like patients. It's that there's more we don't know than we do. It's the idea we can make a discovery that will help them."