Report shows Pfizer's hand in Neurontin studies

* Internal documents show many studies not published

* Studies changed after the fact to show benefit-report

* Pfizer denies allegations; claims study is biased

By Julie Steenhuysen

CHICAGO, Nov 11 (Reuters) - A study of internal company documents suggests Pfizer Inc (PFE.N) altered or omitted unfavorable study findings to expand its epilepsy drug Neurontin's market, U.S. researchers said on Wednesday, offering a look at how drugmakers influence scientific research.

Clinical trials are supposed to answer a specific, predetermined scientific question, but a comparison of Pfizer documents and published studies on Neurontin for conditions other than epilepsy found that eight out of 20 study reports never made it into medical journals.

And in eight of the 12 published studies, the primary outcome -- the answer to the main scientific question -- was changed by Pfizer, the world's biggest drugmaker, from the original study design.

"There were a lot of primary outcomes that were shifted around between the planning of the protocol and the reporting of the study," said Kay Dickersin of Johns Hopkins University in Baltimore, whose study appears in the New England Journal of Medicine.

"Some primary outcomes were lost altogether. Some were brand new. Some were secondary outcomes that were upgraded to primary," she said in a telephone interview.

The studies, all funded by Pfizer, showed how the drug worked in people with problems like migraines or pain, which are off-label uses of the drug.

Once a drug is approved, doctors are free to prescribe it as they see fit, and while companies are not permitted to market a drug for anything but the approved use, they can hand out reprints of studies published in medical journals showing how the drugs work in patients with different problems.

Dickersin got the documents while serving as an expert witness against Pfizer, which in 2004 paid $430 million to settle a lawsuit over illegal promotion of Neurontin.

Pfizer was sued again last year by lawyer Thomas Greene, who brought the original case against the company for off-label marketing practices, for holding back negative study results and changing the design of its trials to produce more favorable results.

That case was the latest in a string of allegations against the pharmaceutical industry suggesting it has controlled the flow of clinical trial research to boost its marketing position.

Pfizer spokesman Chris Loder said in a statement the suggestion that the company attempted to mislead the medical community is untrue and was "derived from a report created for litigation and coauthored by plaintiffs' expert witness."

Dickersin said the studies she reviewed are still not publicly available.

She said while there can be legitimate reasons to change a study's primary goal or endpoint, that change needs to be included in a formal amendment and published in a journal.

Not taking that step leaves a false impression, and in the case of companies, reveals the competing interests of scientists and marketing departments, she said.

To improve transparency, Dickersin said study protocols and primary endpoints should be required when companies register their clinical trials on the public database clinicaltrials.gov.

"It's important for us not to give up on trying to understand this because if we don't, the truth of science will not be upheld," she said.

(Editing by Philip Barbara)

(Julie.Steenhuysen@thomsonreuters.com ; +1 312 408 8131)