More trouble for Glaxo. New studies in the Journal of the American Medical Association say that Avandia increased the chances of having a heart attack, while Takeda’s competing drug lowered the risk. The Glaxo pill increased heart attacks by 42 percent and doubled rates of heart failure. Takeda’s Actos reduced heart attacks, strokes and deaths by 18 percent, though it also increased heart failure, according to a separate study funded by the drugmaker.

There’s little doubt this will make it much more difficult for Glaxo to win back market share. Avandia was the world’s best-selling diabetes pill until May, when a controversial study was reported by the Cleveland Clinic’s Steve Nissen in The New England Journal of Medicine that the pill increased heart attack risks by 43 percent.

“This again confirms, as have all the other meta analyses, that rosiglitazone seems to increase the risk of heart attacks,” Harvard Medical School Professor David Nathan tells Bloomberg News. “Whether it kills you or leads to more cardiovascular death is an open question.”

These are the Avandia and Actos abstracts (subscriptions are required for the complete studies).

Each JAMA article is a meta analysis, meaning earlier studies were pooled and analyzed. The Avandia study looked at four trials that enrolled 14,291 people for at least 12 months, Bloomberg notes. They found no increased rate of death from cardiovascular causes. The Actos study analyzed data from 16,390 people treated in 19 trials that ranged from 16 weeks to 3.5 years.

Last month, the Avandia and Actos labels were given Black Box warnings by the FDA, which is still considering a July 30 advisory panel recommendation that the agency add a caution about heart attacks to Avandia’s prescribing info.

Glaxo execs say the JAMA findings conflict with accumulated data on Avandia from 116 clinical trials in more than 52,000 patients, which they said don’t show the drug reduces blood flow to the heart compared to other medicines, including Actos. Reduced blood flow to the heart can cause heart attacks. “I don’t think it adds any new information on the safety of Avandia at all,” gripes Glaxo vp Alastair Benbow to Bloomberg. “We believe Avandia remains an important option for doctors and their patients.”

Avandia brought in $3.3 billion for Glaxo in 2006. After the heart-attack risk was first reported in May, sales of the medicine fell 22 percent in the second quarter. US scrips for Avandia fell by more than half to 146,000 a week since the New England Journal study was published on May 21, according to IMS Health. Avandia now accounts for 5.6 percent of new US scrips for diabetes pills, compared to nearly 14 percent for Actos.

Steven Nissen, a co-author of the JAMA article about Actos and the May 21 New England Journal study of Avandia, said regulators should act on the two JAMA articles. “It’s quite a contrast to look at the two analyses side by side,” Nissen said. “It puts pressure on regulatory agencies around the world to take some sort of action on rosiglitazone,” or Avandia.

Yoon Loke, a clinical pharmacologist at the University of East Anglia in Norwich, England, and a co-author of the Avandia study in JAMA, says docs at Norfolk and Norwich University Hospitals, where he works, are being advised to switch their patients from Avandia to Actos if they require a drug to lower blood sugar.

“We believe it is inappropriate for the author to advise doctors to disregard the FDA’s advice which is to keep patients who are effectively controlling their diabetes on Avandia,” Glaxo said in a statement.

Takeda won’t seek to add the new data to prescribing information for Actos, says Robert Spanheimer, the drugmaker’s senior director of diabetes. “This should give physicians and patients the confidence that Actos does not increase” heart attacks or death rates, he crows.

Takeda provided a $25,000 grant to the Cleveland Clinic team and collected the data used in the study of Actos, also called pioglitazone. Glaxo wasn’t involved in either Avandia study. The drugmaker conducted its own analysis of Avandia that suggested the drug increased the risk of cardiovascular complications by 31 percent and submitted it to regulators in 2006. The data don’t provide a clear picture of the drug’s risks and benefits, Glaxo has argued.

Loke worked with two researchers at Wake Forest University, including FDA advisory committee member Curt Furberg, who’s said the FDA should consider withdrawing Avandia. He wasn’t allowed to vote at the July 30 meeting because the FDA said he has a conflict of interest as a member of another government drug- safety review panel whose research is partly funded by several drugmakers, including Glaxo.