New Drug for Diabetes Shows Few Heart Risks

By JENNIFER CORBETT DOOREN and SHIRLEY S. WANG

A new diabetes drug being developed by Bristol-Myers Squibb Co. and AstraZeneca PLC was found to have low rates of cardiovascular events like heart attacks, a safety concern associated with some other diabetes treatments.

The Food and Drug Administration, which released its findings in a review Monday, has been concerned about cardiovascular side effects since a research paper was published in 2007 linking GlaxoSmithKline PLC's widely-used Avandia to increased risk of heart attack.

The FDA issued new guidance in December requiring drug makers to conduct more stringent clinical trials for type-2 diabetes drugs to better assess heart-attack and stroke risks. The FDA suggested studies of at least two years, far longer than the typical three to six months previously used for approval.

On Wednesday, an outside panel of medical experts will vote on whether Bristol-Myers and AstraZeneca have provided enough data to rule out an "unacceptable excess cardiovascular risk relative to competitors." A positive panel vote amounts to a recommendation that the FDA approve the drug. The agency usually follows its panels' advice but isn't required to.

Bristol-Myers and AstraZeneca began conducting human testing of their drug, saxagliptin, well before the new FDA guidelines and didn't design studies to meet the new standards. Instead, the companies reanalyzed data after the studies were completed at the FDA's request to investigate saxagliptin's cardiovascular safety.

Earlier this month, the FDA told Takeda Pharmaceutical Co., which has a similar diabetes drug up for approval, that the amount of data the company submitted wasn't sufficient to meet the new standards.

The FDA review of saxagliptin focused primarily on six core studies submitted by Bristol-Myers and AstraZeneca that followed patients for at least one year after the initial three-to-six month study period. It found an overall low rate of major cardiovascular events, according to Monday's documents. In higher-risk patients, the rates appeared to be similar between patients taking saxagliptin and those getting a placebo or, in one study, the popular diabetes drug metformin, at about 2%.

Write to Jennifer Corbett Dooren at jennifer.corbett-dooren@dowjones.com and Shirley S. Wang at shirley.wang@wsj.com