Pasteris is one of 25 executives appointed last year to shepherd the British drugmaker's experimental medicines. He consults insurance companies and former officials from national health agencies about the Alzheimer's drug on the best way to show its value to patients.

In the past, that meant proving that a drug worked, and did so safely, so that health regulators would approve it. But as governments in the United States and Europe look to slash spending and avert a debt crisis, Glaxo and its rivals want to make sure their medicines are a must-have for patients.

To do it, they are seeking input from the people who hold the purse strings earlier than ever in the clinical research process, in some cases five years or more before regulators would even look at a product, executives told Reuters.

"The ultimate goal was not optimal reimbursement and access," Pasteris said. "Today it is."

These views are shaping more clinical trials, such as which products to test against and study goals to pursue. And that's having major ramifications for the business of Big Pharma.

"If you're going to go out there with a drug that you don't know whether it's better than what's out there, what are you trying to do? Who are we all trying to kid?" said Angus Russell, CEO of British drugmaker Shire Plc. His company has more than doubled its "pharmaco-economic" staff focusing on the value of medicine in the past few years.

Russell said companies "all over the industry" are dropping experimental products they fear will not gain strong reimbursement. For example, Glaxo abandoned a diabetes treatment in mid-stage development in 2009.

Pharmaceutical investors are also a huge source of pressure, with little forgiveness on Wall Street when it comes to medicines that cost hundreds of millions of dollars to develop, but do not get widely used once they reach the market.

Even smaller players are changing their ways. Ron Cohen, CEO of Acorda Therapeutics, regrets not consulting insurers early about its Ampyra, the first drug to help multiple sclerosis patients walk better.

Now, Acorda plans to hold discussions with health insurers once products reach mid-stage development and is getting informal input earlier -- including for a potential multiple sclerosis treatment yet to enter human testing.

"I have no question that the entire industry is moving toward this sort of model," Cohen said.

As drug manufacturers invite marketing input earlier than before, some fear they risk the very innovation that leads to landmark new medicines.

Industry experts point to advances that took time to prove their worth or worry that drugmakers may abandon categories where "good enough" medicines already exist, like depression, partly because it's not worth the economic risk.

"There is a concern that in five, 10 years we won't have anything really new for patients with major mental illnesses, and that would be absolutely a tragedy," said Dr. Alan Schatzberg, former president of the American Psychiatric Association. "It's an unfortunate outcome that we are slowing drug development."

Glaxo, which brought antidepressants Paxil and Wellbutrin to market, is one company to pull away from the field. Atul Pande, who leads Glaxo's neuroscience research, says the science has not advanced enough to identify new ways to significantly improve treatment, but he acknowledged the reimbursement fears.

CLOSER TIES

In this climate of soaring healthcare costs, the drug industry has been sharply criticized for launching expensive new medicines that proved only slightly better than their predecessors. Health insurers and government agencies pushed back, and now drug companies are forging closer ties with those "payors".

This year alone, Pfizer Inc allied with insurer Humana Inc to research elderly health; AstraZeneca Plc and HealthCore, a unit of insurer WellPoint Inc, agreed to study how to economically treat disease; and Sanofi SA signed on pharmacy benefit manager Medco Health Solutions Inc.

Sanofi may soon overtake Pfizer as the world's top drugmaker. CEO Chris Viehbacher says the industry's new crop of drugs must demonstrate "why is this better than what we've already got."

"In defining value -- in whose eyes? So you need a payor perspective," he said.

A closer relationship to payors allows access to vast databases of medical claims to see how drugs are used once they are approved. Drugmakers can learn which medicines they should be comparing their own products to and what goals they should seek in clinical trials.

That can help demonstrate the value of new treatments over a growing pool of cheaper generic drugs. For example, instead of testing a drug just to show whether it lowered blood pressure, the manufacturer could also show that it helped reduce stroke rates or hospitalizations.

Medco, which manages prescription benefits for millions of Americans, is discussing Sanofi's drugs in their initial human trials.

"We're there at late Phase I, stress-testing their entire development program, and it's a very rich conversation," said Medco's chief clinical research and development officer, Robert Epstein.

He believes more companies should not wait until a drug is being tested in hundreds, or thousands, of people by Phase II or III, when they are close to being submitted to regulators.

"Most of the companies still show up in mid-Phase III saying they have a cake in the oven and it's baking and you're going to love it when it comes out," Epstein said.

FAILURE TO LAUNCH

The motivation for these tie-ups strengthened with a series of new drugs that did not catch on widely.

Dendreon Corp's Provenge therapy for advanced prostate cancer is one cautionary tale. Provenge is a novel treatment for patients who have limited options, but sales of the $93,000 therapy have been a huge disappointment.

That's largely because doctors fear they will not be reimbursed after they buy and administer the drug in their offices. The company's market value has plunged about 80 percent.

Glaxo is grappling with similar issues with its lupus drug marketed with Human Genome Sciences Inc.

To try to avoid such hurdles in the future, Glaxo and other companies hold advisory board meetings on their experimental products with representatives of insurers and others with expertise on payment for medicines.

Pfizer's advisory panels also include pharmacy benefit managers and employers to review many of the products in the company's mid to late-stage portfolio, said James Harnett, Pfizer's senior director of U.S. health economics and outcomes research.

But some are concerned that bringing in market considerations too soon will undermine the serendipity of scientific discovery, such as medicines that were found to treat a different condition than initially intended.

Pfizer might have never developed a rheumatoid arthritis medicine, now one of its most important experimental products, according to former research chief John LaMattina. The drug was first developed to prevent rejection of organ transplants, he said, a far less enticing market.

Only when Pfizer began human testing of tofacitinib did it see the potential for treating rheumatoid arthritis with a conventional pill, a more attractive option to the injectable medicines that dominate the market, he said.

"I don't think people dreamed that you could have a small molecule that would successfully challenge the large biologics ... so you never would have gone down that pathway," said LaMattina, Pfizer's research chief from 2004 to 2007.

LaMattina likened this to how Steve Jobs, Apple Inc's founder, said that people often did not know what products they wanted until they saw them.

"When you're doing market research, people are basing their responses and judgments on what's known and not necessarily looking out of the box," LaMattina said. "The earlier you get in the discovery/development continuum, the less valuable I think it is."

Drugs also can prove more useful when used with other therapies, which has been true in the cancer field. That testing may never occur if they look marginally beneficial on their own.

"You find other uses, you find combinations, you find other indications -- so many things can happen," said Nils Behnke, a healthcare partner with Bain & Co.

PAYOR NEEDS

Glaxo has also changed how it rewards researchers. Bonuses are linked to winning reimbursement, not just drug approval.

"This balance between cost control and innovation arguably has swung a little bit more to cost control," said Jack Bailey, Glaxo's senior vice president for institutional customers.

In his discussions about the Alzheimer's drug, SB-742457, Pasteris learned that in some countries, Pfizer's Aricept is the standard therapy, while in others, Forest Laboratories' Namenda is being used with Aricept. Both drugs likely will be sold more cheaply in generic form by the time SB-742457 would reach the market.

"If we have to satisfy the regulators, all we need is a trial versus Aricept," Pasteris said. "To satisfy the payors in some countries, we may need to show evidence versus the combination of Aricept plus Namenda."

Researchers presented mid-stage data on SB-742457 this summer and are designing Phase III studies, the last stage before seeking approval.

"That's really where we shape the asset the most," Pasteris said. "We want to make sure that the endpoints, the trial design, the clinical development plan, the medicine development strategy -- it's all aligned, and it's all trying to meet the payor needs."

(Editing by Michele Gershberg and Edward Tobin)

• A Permanent Economic Emergency