What does a company do when there's anecdotal evidence that two of its drugs are equally effective in treating a leading cause of blindness in the elderly, one costing patients $60 per treatment and the other $2,000?

In the case of Genentech Inc., nothing.

The company declined to seek federal approval for the cheaper drug, Avastin, to treat the wet form of age-related macular degeneration. Nor would it help finance -- or cooperate with -- a National Eye Institute study comparing the effectiveness and safety of Avastin, a cancer drug, and the more expensive eye drug, Lucentis.

The financial stakes stemming from the study are huge. Medicare officials estimate there could be 50,000 or more additional cases of macular degeneration a year. Treating just one year's worth of new patients with Lucentis would cost $1.2 billion a year, compared with $60 million if they're treated with Avastin, Medicare officials said.

Genentech is making no promises that it will act upon the trial's final results, which are expected in two to three years.

The company has raised concerns that safety issues were not properly addressed. In particular, the trial doesn't have enough patients to show some of the rare but serious side effects that could occur with use of the cheaper drug, the company contends.

"No matter the outcome, we continue to believe Lucentis is the most appropriate treatment for wet AMD," said Krysta Pellegrino, a company spokeswoman.

Wet AMD occurs when abnormal blood vessels leak blood and fluid affecting the part of the eye that allows you to see fine detail.

Many eye doctors believe Avastin works just as well in treating macular degeneration even though it hasn't been approved for that purpose. It's not unusual for drugs to be used off-label -- treating diseases other than ones the drug was approved for.

Both drugs target a protein that causes blood vessels in the back of the eye to grow, but Lucentis is a much smaller molecule. It was specifically designed -- at great expense -- to penetrate the retina.

Companies routinely help finance clinical trials, but such trials almost never pit two products from the same company against each other.

"It's a very unusual situation where a company would be trying to compare its own drugs," said Dr. Frederick Ferris, director of clinical research at the National Eye Institute. "I'm not sure usual situations are all that relevant in this particular case."

Still, health officials pleaded with Genentech to participate in the clinical trial comparing the two drugs. At one point the company considered doing so by providing the medicines in masked, identical vials, according to e-mail exchanges obtained by the Senate Aging Committee.

"Good news is that the Board supports the proposed studies," said one e-mail sent in June 2007 from Charlie Johnson, a company vice president, to Dr. Daniel Martin, the chairman of the study who works at the Emory University School of Medicine.

In the end, the board did not support the study. Martin made a final plea.

"The fact that we are comparing your drugs and you are not involved is very awkward and can easily give way to anti-Genentech sentiments," Martin said. "The leaders of this study are only interested in answering the many scientific and patient management questions that we face with our patients every day, but some investigators and the press want this study to be more than that. Your involvement would be very helpful to both of our causes."

Genentech routinely provides financial support for clinical trials, Pellegrino said in an interview. But in this case, she said, "Our resources would be better spent looking at other diseases where there are no treatments.

Dr. Philip Rosenfeld, who has treated hundreds of his eye patients in South Florida with Avastin, said Genentech had little economic incentive to help finance the trial -- unless it was confident Lucentis was truly superior.

"By fact that they didn't support the clinical study leads me to conclude that in reality there is no difference between the two drugs," Rosenfeld said. "The result is clearly not in Genentech's best interest."

Avastin was approved to treat colon cancer in February 2004. It's a genetically engineered product that inhibits the growth of blood vessels, thus denying tumors blood, oxygen and other nutrients needed for growth. It's expensive, costing $2,200 for a typical treatment for colon cancer. However, for treating eye disease, pharmacy compounding firms split the drug into many tiny doses suitable for injection into the eye. That's what brings the price down to about $60 per injection.

Rosenfeld said the study could put doctors at ease about potential litigation if they prescribe Avastin instead of the FDA-approved drug.

"I see this as a public health study, not only for us, but for the whole world. It gives everyone the license to use both drugs interchangeably," Rosenfeld said. "Clearly, for Medicare it would make economic sense to put preference on the use of Avastin.

As lawmakers await the results of the trial, they are already considering what steps, if any, could be taken to steer the Medicare program to the less costly drug -- if it's indeed comparable.

An internal memorandum from congressional aides to the Senate Aging Committee's chairman, Herb Kohl, D-Wis., recommends that lawmakers consider urging Medicare officials to pay no more for one drug than the other when it comes to treating the eye disease.

Medicare's contractors already have authority to pay the same amount for items that achieve much the same result -- such as hormones used to treat prostate cancer.

If the drugs are shown to work comparably, "it would surprise me if the contractors did not quickly use that concept," said Dr. Steve Phurrough, director of coverage and analysis at the Centers for Medicare and Medicaid Services.

Pellegrino said it's too early in the comparison trial to comment about the staff's recommendation.

Pellegrino said Genentech's pricing for Lucentis reflects the cost of developing the drug, which the FDA approved in June 2006. The development program included a clinical trial involving more than 6,000 patients at a cost of more than $45,000 a patient.

"It took decades and hundreds of millions of dollars to develop the drug," she said.