SAN FRANCISCO - Remember leptin? The once-ballyhooed hormone and its receptor technology that Amgen Inc. licensed from Progenitor Inc. way back in 1997 had gained the limelight as a potential obesity cure two years earlier.

Molecular geneticist Jeffrey Friedman, of Howard Hughes Medical Institute, published a paper in Science after finding the gene in mice of which leptin is the product. Its name derives from the Greek leptos, which means "thin," and excitement over the possibilities for big-time tippers of the scale knew almost no bounds. For a while, anyway.

Last week, however, leptin was a topic for a lesson in drug development and comic relief for the crowd at Allicense 2004, an annual event sponsored by Recombinant Capital. Roger Perlmutter, vice president of research and development at Amgen, told listeners that companies often approach research unwisely, and he used leptin as an example.

Although neither Perlmutter nor the company has given up altogether on the hormone, he acknowledged how things might have been done differently.

In the traditional environment, he said, the question typically asked is: "Gee, what is the target for this disease process that is approachable by a potent, orally bioavailable, once-a-day, highly selective molecule that I can package in a traditional format - oval-shaped tablet, wax coated, with white lettering, etc.?"

Perlmutter said that's not the right question.

"The right question is, What's the best target for this disease?'" he said. "I don't care whether that's a target that should be approached by a protein therapeutic, an antibody, a small molecule, an infectious therapy, nucleic acid - I don't care what it is. This business is so hard, why would you want to circumscribe your research activities to just one modality? Fit the tool to the task."

In leptin's case, the tool was preclinical mouse work that seemed to promise obesity aid for human beings, said Perlmutter, who has been involved with Abbott Laboratories, Bristol-Meyers Squibb Co., Pfizer Inc. and Merck & Co. Inc.

"I've had the opportunity to survey the industry in a variety of different ways, and I can tell you that any one of those companies and all the others as well could provide similar kinds of stories," he said.

What Perlmutter called the "brilliant" first research - experimentally linking the blood circulations in mice - proved that a wild-type animal could confer the wild-type phenotype for thinness on an obese mouse but not a diabetic mouse, whereas something the diabetic animal made caused the obese animal to lose weight. From that emerged the assumption that the obese mouse does not produce the satiety factor to regulate its food consumption, whereas the diabetic mouse produces the satiety factor but cannot respond because it has a defective satiety center.

The assumption was right - in mice.

"If you use leptin in animal models, it causes weight loss under all circumstances," Perlmutter pointed out, noting that the hormone acts on receptors in the central nervous system and can be provided peripherally or centrally.

However in humans, researchers found those with higher body fat have more naturally occurring leptin in their systems.

"This makes it seem less likely that giving more leptin is going to cause you to lose fat," Perlmutter said, prompting chuckles from the audience. "It just kind of makes you nervous." But the opportunity seemed ripe in the population of obese people who were deficient in leptin. "That could be a large percentage of obese people," he said. "Who knew?"

The crowd's amusement grew when he added that "there are about 10 of those people identified worldwide - which is a market, but it's a small market."

Perlmutter showed before-and-after slides of three obese people low on leptin whom researchers managed to locate in Turkey. In the second slide, they had "slimmed down very substantially." When Perlmutter added, "I must say I'd feel a lot better if they hadn't looked so happy on the first slide," the audience burst into full laughter.

"There's no doubt that in a leptin-deficient state, providing leptin causes weight loss," he said. The hormone has been administered to more than 1,500 subjects. "I think we can say with great confidence that leptin is safe," Perlmutter said. "It is really safe. Man, oh man, is it safe." It just isn't particularly efficacious in an unselected obese population.

The leptin misadventure showed, among other things, how important trial design is, Perlmutter said.

"If you're going to introduce the molecule into people, you owe it to those people to do an experiment you can interpret," he said, noting that researchers often seek only to get exposure in the blood at the levels gained during preclinical trials.

"That's really not acceptable in this day and age," he said, calling for "appropriate surrogates to know that we have hit the target, so we can reject the targets when they don't work, because most of the time, they won't."

For example, with Amgen's angiogenesis inhibitor AMG 706, the company designed Phase I studies based on preclinical work. Researchers used digital-contrast-enhanced magnetic resonance imaging to measure the decrease in tumors' "vascular blush" when the drug was administered - an approach that provided the basis for later studies showing objective responses in terms of tumor-mass shrinkage.

"That is an extremely useful way to think about studying molecules, just one example," Perlmutter said. "Often, it's extremely challenging to come up with appropriate surrogates, but I believe it's necessary. I think that's the right way to do drug discovery."

As for leptin, the hormone "may rise like a phoenix from the ashes," since a number of leptin-deficiency states other than obesity have been identified. One is lipodystrophy, in which people with little fat mass have such low leptin levels that they suffer "terrible metabolic disturbances," Perlmutter said, such as those endured by a 17-year-old girl who was undergoing regular plasmaphoresis to reduce triglycerides, so abundant in her blood (because of low leptin), that "her blood was nearly as thick as motor oil." Leptin dosage fixed it.

Every compound in the course of development "dies a thousand deaths," as champions within companies struggle to keep interest alive, even when initial results are discouraging, he said. "This death [of leptin] was a little more profound than most, but we'll wait and see."