The FDA released a staff review document that found "no significant difference in mortality" up to one year, between aortic stenosis patients who received the Sapien valve and a control group that had aortic valve replacement through open-heart surgery.

But the document said there was only limited data available beyond two years and cautioned that some of the findings were hard to assess partly because some patients in the study did not receive the intended treatments.

The Sapien heart valve, which is threaded to the heart through an incision in the groin or ribs, has already been approved by the FDA for aortic stenosis patients deemed to be too sick for open-heart surgery. A panel of outside advisers to the FDA will vote Wednesday on whether to recommend that the agency approve the product for use among patients whose cases are high-risk but not inoperable.

Wall Street analysts said the FDA staff's cautious comments suggest that Edwards Lifesciences will face a challenge advisory committee hearing on Wednesday, but ultimately win the panel's recommendation and the regulatory agency's approval.

Sapien is considered to be one of the most important future growth drivers for Edwards, whose shares were up about 13 cents at $88.30 in afternoon trading.

"We estimate there will be approximately 5,500 of these (high-risk) procedures performed in the United States in 2013, moving to nearly 40,000 by 2016," Goldman Sachs analyst David Roman said in a note to investors.

Roman added that his 2016 estimate would be double the number of procedures anticipated for patients ineligible for surgery during the same year. The procedure is estimated to cost about $30,000.

In pivotal clinical trials, the company said the Sapien led to a slightly lower death rate and dramatically shorter recovery times and hospital stays than chest-cracking open-heart surgery.

The FDA staff cited findings that showed patients treated with Sapien had a mortality rate of 24.27 percent when the device was inserted through the leg, and 22.24 percent when inserted between the ribs. The control group had mortality rates above 26 percent.

FDA staff also highlighted research showing a significantly higher incidence of stroke. Earlier research suggested a higher stroke risk could potentially make surgery preferable for some, especially younger, stronger patients.

In addition to strokes, the FDA staff pointed to data showing that 53 percent of patients treated with the Sapien system experienced leakage known as aortic regurgitation and said there was mounting evidence of a link between the condition and death among Sapien patients.

"These data also show that the amount of (aortic regurgitation) is appreciable and does not decrease over time," the staff said.

Up to 1.5 million people in the United States suffer from aortic stenosis, a progressive disease that affects the aortic valve of their hearts, according to Edwards.

The company said that approximately 250,000 people have severe symptoms. But only about two-thirds of them undergo surgery.

The federal government's Medicare healthcare program for the elderly and disabled and the joint federal-state Medicaid program for the poor announced last month that they would cover the procedure for patients who cannot have surgery.

(Reporting By David Morgan; editing by Sofina Mirza-Reid and Maureen Bavdek)

• A Permanent Economic Emergency