UPDATE 1-US panel questions wider use of PET scans in cancer

(Recasts with panel comments, votes)

By Susan Heavey

BALTIMORE, Aug 20 (Reuters) - A U.S. Medicare advisory panel expressed doubts on Wednesday that available data on the use of PET scans to diagnose and evaluate nine cancers could apply more widely to other cancers.

The panel of outside experts also said it did not have a lot of confidence that patient data gathered from a national registry showed positron emission tomography (PET) scans could help doctors make better decisions or improve patient outcomes.

"The early data is suggestive, it's not definitive," said panelist Peter Juhn, head of Medco Health Solutions Inc's (MHS.N) Therapeutic Resource Centers division.

In several votes, a majority of the panelists said they had "little confidence" or were "equivocal" that information gathered so far supported wider use or patient benefit.

Medicare, which provides health insurance for 44 million elderly or disabled patients, is seeking the panel's advice before deciding whether to ease payment restrictions that could allow wider use of the scans.

The medical imaging industry initially asked the agency to lift the current payment requirement that patients be enrolled in the registry if receiving a PET scan for one of nine cancers: brain, cervical, bladder, small-cell lung, ovarian, testicular, prostate, kidney and pancreatic cancer.

Eliminating the registration requirement would allow greater use of PET among Medicare patients, in part because it would be easier for doctors to order the scans.

That could be a boost for PET scan makers such as General Electric Co (GE.N), Siemens AG (SIEGn.DE) and Philips Electronics NV (PHG.AS).

But at the advisory panel meeting on Wednesday representatives of the Academy of Molecular Imaging and other industry and medical groups said the data collected so far supported even wider use of the tests among other cancers.

Data showed the scans helped doctors adjust their treatment plans for roughly one-third of enrolled patients, the groups said.

"This is a broadly applicable technique to cancers in general," Dr. David Mankoff, a University of Washington radiologist, said on behalf of several industry associations.

Concerned about insufficient evidence, Medicare conditionally agreed to cover some PET scans in 2005, provided patient data was collected via the registry.

Medicare recently commissioned a separate analysis that found the quality of available studies on PET's use in cancer patients was poor to moderate.

Some panelists said it was hard to reconcile that finding with the industry's conclusion from the registry data.

"I don't see how you make the leap," said panel member and consumer representative Linda Bergthold "The quality of the data ... was really stunningly poor."

The one panelist who said the data strongly supported PET for other cancers was Dr. Richard Wahl, head of nuclear medicine at Johns Hopkins University. He said proving its effectiveness had become "a statistical exercise" that unfairly impacted those with rarer cancers that were unlikely to accumulate much data.

More than 2 million PET scans are given in the United States each year, according to Philips, which along with its rivals does not break out PET scan sales figures. A spokeswoman for Medicare said figures on how much the agency has spent for the test in cancer patients so far was not readily available.

Various industry estimates show administering the scan can cost between $2,500 and $6,000 each time.

There has been disagreement over the effectiveness of PET scans compared to other imaging techniques such as computed tomography (CT) or magnetic resonance imaging (MRI) scans in detecting cancer cells.

With PET, patients are injected with radioactive sugars that collect in metabolically active parts of the body, which could signal growth and possibly cancer.

The ability to pinpoint biologically active areas is a key difference between PET and CT scans or MRI scans. While PET can find some cancers that CT and MRI miss, its images can be less detailed or trigger false positives.

Medicare is expected to make a draft decision in January before making its final ruling in April. (Editing by Tim Dobbyn)