A failure of vigilance   (August 27, 2008)

Last evening (August 26, 2008), the manufacturer of Byetta (Amylin Pharmaceuticals, Inc., partnering with Eli Lilly) released information attempting to reassure patients, physicians, and investors that Byetta was safe, despite increasing evidence that may link the drug to a sometimes-fatal complication of diabetes called pancreatitis. In my opinion, they failed.

First, let me remind everyone that the possible association between Byetta and pancreatitis is not proven, and there definitely is an association between T2DM and pancreatitis. What's not yet understood is whether Byetta increases the risk of developing pancreatitis. Amylin and Lilly have announced that they "continue to pursue a comprehensive drug safety program that includes extensive internal and external review of individual cases, and clinical and epidemiologic studies."

A brief history of Byetta-associated pancreatitis follows.

The U.S. prescribing information for Byetta (also known as exenatide) includes  information about the possibility of developing pancreatitis. The present wording is in the PRECAUTIONS section of the label, and reads: "Postmarketing cases of acute pancreatitis have been reported in patients treated with BYETTA. Patients should be informed that persistent severe abdominal pain, which may be accompanied by vomiting, is the hallmark symptom of acute pancreatitis. If pancreatitis is suspected, BYETTA and other potentially suspect drugs should be discontinued, confirmatory tests performed and appropriate treatment initiated. Resuming treatment with BYETTA is not recommended if pancreatitis is confirmed and an alternative etiology for the pancreatitis has not been identified."

In October 2007, the FDA announced that they had reviewed 30 postmarketing reports of acute pancreatitis in patients taking Byetta, and that an association between Byetta and acute pancreatitis was suspected in some of these cases.

A week ago, on August 18, the FDA had stated there were two deaths from pancreatitis in patients taking Byetta:

Since issuing Information for Healthcare Professionals in October 2007, FDA has received reports of 6 cases of hemorrhagic or necrotizing pancreatitis in patients taking Byetta. Byetta is a medicine given by subcutaneous injection to help treat adults with type 2 diabetes. Of the 6 cases of hemorrhagic or necrotizing pancreatitis, all patients required hospitalization, two patients died and four patients were recovering at time of reporting.  Byetta was discontinued in all 6 cases.   

But now we're finding out that there were not 2 pancreatitis-associated deaths, but 6.

Apparently, the four newly-publicized deaths involved patients who had complicating medical conditions in addition to type 2 diabetes, ranging from leukemia to gallstones, which were likely the primary cause of death. According to the AP, Lilly's medical director, James Malone, said that the FDA was aware of these four additional deaths last week, but did not make them public because they involved a milder form of pancreatitis.

Huh? Milder? "Milder" deaths? Really? This statement from the manufacturer's representative, if accurately reported, shows a clear violation of ethical behavior. Reporting to the FDA is mandatory, and had it not been done, legal sanctions would occur. No one is doubting that Amylin and Lilly appropriately reported these cases to the FDA. But failing to report them to an anxious public? Parsing the information so as to hide 4 fatal cases? Sorry, I'm not happy about this.

Nor was the financial community happy. Both Amylin and Lilly took a hit on their stock prices today. As one analyst put it, "We believe more information about these patients with severe pancreatitis could help investors put them in the right perspective. Non-disclosure only leaves more room for speculation." Left unspoken and still to be evaluated is the effect on the confidence of patients and physicians, which also surely took a hit.

What was the failure? It was a failure of pharmacovigilance. Pharmacovigilance is a mandatory activity that every drug company routinely does; it's called "drug safety" by some companies. Entire departments of PV specialists, including physicians, nurses, and other health professionals process incoming reports of "adverse events" and put them into huge computerized safety databases. These PV specialists then forward the reports to the FDA and other Health Authorities in other countries. Then they categorize the reports: how many cases were there of nausea, how many cases of pancreatitis, how many deaths, how many of anything that's reported. So Amylin and Lilly knew (or should have known) precisely how many cases of pancreatitis had been reported.

And one of the first responsibilities of PV physicians is notifying the FDA and their company's management when a "signal" occurs. Fatal cases of an event that is in the label, but not labeled as causing death (such as pancreatitis in the present Byetta label) is one example of a signal. Was the signal missed? Or was the public and perhaps the FDA misled?

Ethics and integrity means that the company should promptly disclose to the FDA that a signal has been spotted. But I believe it's in the best interest of the manufacturer to come clean immediately with full disclosure of what is in their safety databases. Otherwise, as the information slowly leaks out (as it did in this case), one must wonder what else might be in the database. For example, how frequent have abnormalities of the lab tests for pancreatitis (the enzymes amylase and lipase) been reported which were not labeled as pancreatitis by the reporting physicians? Have there been other deaths of patients on Byetta with abnormalities of these enzymes, where the reporting physicians didn't mention "pancreatitis"?

What next? I'd expect the following might occur.

1) Amylin/Lilly might prepare a "Dear Doctor" letter, addressed to health care professionals, outlining what's known to date. These letters take a few days or weeks to prepare, as they need to be reviewed by the FDA and the manufacturer's legal and marketing departments.

2) A change in the Prescribing Information (the "label") seems likely. This can take months, depending on whether the manufacturer wishes to procrastinate and nitpick every suggestion that the FDA proposes, or occur almost immediately if an outbreak of integrity happens.

3) An FDA Advisory Committee meeting might be called. This is an extraordinary event for a drug that's already on the market, but a public review of the safety of Byetta seems possible in light of the importance of these events vs. the upcoming planned expansion of the use of this drug as monotherapy and as once-weekly therapy. At such a meeting, representatives from the FDA and the manufacturer (and concerned citizens) present their data to a panel of experts, who review the data and offer their opinion to the FDA about what next: continuing the present label (unlikely here), changing the label (and how to reword it), or withdrawing the drug from the market (an extreme step granted the current knowledge of Byetta's risk profile).

In the meantime, if you are taking Byetta, my advice from last autumn remains unchanged:

1. Don't stop your Byetta if you're taking it simply because of this issue, unless your blood tests (amylase, lipase) are positive for pancreatitis!
2. Since gut side effects are common with Byetta, and your doctor probably has had dozens or hundreds of people calling about tummy upsets on this drug, you'll have to push your doc -- ask him/her to read the FDA warning.
3. Ask for a pair of blood tests called "amylase and lipase" to confirm or rule out pancreatitis.
4. And finally, if you've had acute pancreatitis while on Byetta, be sure your physician reports your case to the manufacturer and/or the FDA.