LOS ANGELES » Dr. Peter C. Butler initially declined a request by the drugmaker Merck to test whether its new diabetes drug, Januvia, could help stave off the disease in rats.

"I said, I’m not interested in your money, go away," Butler recalled.

Merck no doubt now wishes it had. When Butler finally agreed to do the study, he found worrisome changes in the pancreases of the rats that could lead to pancreatic cancer. The discovery, in early 2008, turned Butler into a crusader whose follow-up studies now threaten the future of not only Januvia but all the drugs in its class, which have sales of more than $9 billion annually and are used by hundreds of thousands of people with Type 2 diabetes.

"I knew some stuff that I thought was a worry and I was obliged to pursue it," said Butler, chairman of endocrinology at the University of California, Los Angeles.

Based on his latest study, both the Food and Drug Administration and the European Medicines Agency have begun investigations that could lead to new warnings on the drugs, or even to their removal from the market.

Or they could result in no action at all.

Butler faces powerful opponents in the makers of the drugs and many diabetes specialists, who say his studies are contradicted by other evidence.

"The data are inconclusive," said Dr. Robert Ratner, chief scientific and medical officer of the American Diabetes Association. He said even if there were some excess risk, it would be "exceptionally low."

Nancy Thornberry, who heads diabetes drug development at Merck, said clinical trials, the gold standard of medical evidence, had found no increased risk of pancreatic disease from Januvia, even when results of trials were pooled to achieve greater numbers. "In fact, my mother takes sitagliptin," she added, referring to Januvia by its generic name.

Questions about whether the drugs raise the risk of pancreatitis, a painful and possibly lethal inflammation of the pancreas, arose soon after the first one, Byetta, now sold by Bristol-Myers Squibb and AstraZeneca, was approved in 2005. The drugs’ labels contain warnings about that. What is new and potentially more serious is a possible risk of pancreatic cancer, which is virtually untreatable and kills most victims within a year.

Many people in the field compare Butler to Dr. Steven Nissen, the well-known Cleveland Clinic cardiologist whose warnings about Avandia, a different type of diabetes drug, led to its being banned in Europe and highly restricted in the United States.

Both men have faced criticism from those who call them zealots. But opinions differ on whether Nissen, and now Butler, are heroes or zealots. The FDA is about to examine data suggesting that Avandia might not be so dangerous after all. Some critics say Butler overstates his conclusions and that his findings have not been replicated by others.

"Basically, no one in the entire world over the last 10 years, with thousands of animals," has found what Butler found, said Dr. Daniel J. Drucker, a professor of medicine at the University of Toronto and a consultant to many drug companies.

Still, Butler is not easy to write off. He is a former editor of Diabetes, the flagship journal of the American Diabetes Association. And he has some defenders.

"He should be an American hero, actually, a rugged individualist who is not going to be browbeaten," said Dr. Edwin Gale, professor emeritus at the University of Bristol in Britain, who recently wrote a commentary with Butler on the drugs.

Butler was born in Kenya to British parents, though he has worked in the United States since 1987 and is a U.S. citizen. His wife, Dr. Alexandra E. Butler, a pathologist who occupies the office next to his, has also worked on some of the studies.

In the last month, lawyers defending drug companies against a lawsuit claiming that Byetta had caused a patient’s pancreatitis, subpoenaed virtually all of Butler’s records.

"I think the message here is they want him out of business," said Brian Depew, a lawyer representing the plaintiff, Ross Hubert of New Hampshire, who claims that Byetta caused him to get pancreatitis. Butler said UCLA told him not to comment on the subpoena.

More than 100 lawsuits representing 575 plaintiffs around the country are claiming injury from Byetta, mostly pancreatitis, according to the latest quarterly regulatory filing from Bristol-Myers. Forty-three suits claim that Januvia caused pancreatic cancer, according to Merck.

Other drugs in the class, called incretin mimetics, are Bydureon and Onglyza, which are also sold by Bristol-Myers Squibb and AstraZeneca; Victoza from Novo Nordisk; Tradjenta from Eli Lilly and Boehringer Ingelheim; and Nesina from Takeda. By far the biggest, though, is Merck’s Januvia and the related Janumet, which had global sales of $5.7 billion last year.

Butler said that after his group presented its rat findings to Merck, "I never heard from them again," except from company lawyers asking when the study would be published.

He said that studies done by the drug companies that led to the drugs’ approval by the FDA tended to use young healthy animals that would not be expected to get pancreatic cancer.

The concern, he said, was that the drugs work essentially by increasing levels of a hormone called glucagonlike peptide-1. That hormone might accelerate precancerous conditions already present in middle-aged people, much as the hormone estrogen might promote growth of nascent breast tumors.

Three other pieces of evidence raise possible concerns.

One is the side effects reported to the FDA, typically by doctors or companies, after a drug is on the market. Butler and colleagues found far more cases of pancreatitis and pancreatic cancer reported for the incretin drugs than for Avandia.

Public Citizen and the Institute for Safe Medication Practices, two watchdog groups, have since separately found the same thing. Public Citizen has already asked the FDA to remove Victoza from the market.

But these reports are voluntary and can be unreliable. Also, when there is publicity about a safety risk, reports of that side effect can spike.

Several groups have looked at medical records of thousands of patients held by insurance companies. At least three of these studies have found no increased incidence of pancreatitis or pancreatic cancer. But a recent study found roughly a doubling of the risk of acute pancreatitis among users of the drugs.

But what has prompted the reviews by regulatory agencies has been Butler’s study of human pancreases obtained from 34 organ donors who had died for reasons unrelated to pancreatic disease. Seven of the donors happened to have taken Januvia and one had taken Byetta.

The pancreases of those eight people tended to have more precancerous lesions than the organs of the diabetics who had not taken those drugs, or those of the nondiabetics. There was also one case of a neuroendocrine tumor, a type of pancreatic cancer.

Also, the pancreases of the incretin drug users were heavier, with faster growth of certain cells. "There were strange growths" that "you’d never see in a normal human pancreas," Alexandra Butler said.

Critics point out that the incretin users were much older than the other diabetics and had been sick longer than other diabetics. That, not the drugs, could have accounted for the findings, they say.

"There are enormous problems with this paper," said Ratner of the diabetes association.

Dr. Fred Gorelick, professor of medicine and cell biology at Yale, said the precancerous lesions found were early-stage ones. Many middle-age people have these and they often do not lead to cancer. Still, he said, the study "raised several red flags."

More information could come out in June when the National Institutes of Health will hold a two-day meeting on possible links between diabetes, diabetes drugs and pancreatic cancer. Butler will be one of the speakers.

And starting this summer, results will be coming from large randomized clinical trials meant to assess whether the drugs raise the risk of heart attacks. Those trials should also be able to pick up an increased risk of pancreatic cancer that might have been missed by the smaller trials used to win approval of the drugs.

So far the safety concerns have not substantially reduced use of the drugs, though there are signs of "possible softness" recently, said Mark Schoenebaum, a pharmaceutical analyst at ISI Group. He said the evidence of a risk was weak and that the FDA would probably take no action.

Butler said he was not calling for the drugs to be removed from the market, though he does not prescribe them to his own patients. Rather, he said, studies should be done using MRI scans to see if use of the drugs is enlarging the pancreases of patients.

"We have all these people out there taking these drugs," Butler said, "and the problem is: What is happening to their pancreases?"

© 2013 The New York Times Company