The study, which examined insurance records for more than 2,500 diabetics between February 2005 and December 2008, found that patients hospitalized with pancreatitis were twice as likely to be taking either of the drugs than a control group of Type 2 diabetics who did not have pancreatitis. However, the study did not examine other meds, such as Novo Nordisk's Victoza (NVO), that were not yet available at that time (here is the abstract).
The question now is whether these results will alter treatment practice by physicians, even though the specter of pancreatitis has hovered over these drugs for years. As far back as 2007, the FDA reported receiving a large number of cases associated with Byetta (back stories here and here) and a similar alert was issued for Januvia two years later (read more here).
Since then, other studies have offered reminders and indicated that physician prescribing habits have been affected (see this). The issue also placed a cloud over Victoza last year when the Public Citizen consumer group filed a petition with the FDA to ask the agency to withdraw the drug (look here).
But the study authors say their analysis moves the ball forward. "This is the first independent confirmation of the suspected link between GLP 1 based therapies and hospitalizations for acute pancreatitis and demonstrates a significant 2 fold increased risk with these drugs," Sonal Singh, the lead study author and an assistant professor of medicine at Johns Hopkins University in Baltimore, tells us. "Clinicians and patients need to balance the benefits of glucose lowering with these risks." Both drugs increase GLP-1, a hormone that stimulates insulin production from the pancreas.
Certainly, the drugs are widely used. Januvia, in particular, generated $4 billion in sales last year, which represented about 9 percent of overall revenue for Merck (MRK) (see page 7). Byetta, which Bristol-Myers Squibb (BMY) acquired with its purchase of Amylin Pharmaceuticals, generated $149 million in revenue last year (see page 3); the drug notched another $159 million for Lilly (LLY) , which had a marketing partnership with Amylin. Bristol-Myers also sells Bydureon, a longer-acting version of Byetta.
One analyst, however, does not believe the study will change much. "We do not believe that the results of this study alone will result in a sudden change in treatment practices," writes ISI Group analyst Mark Schoenebaum in a research note. "The study is limited by its design and methodology. Both DPP-4 drugs and GLP-1 agonists have been on the market long enough that doctors have developed their own experience and comfort with the drugs. In addition, the drug options for diabetics beyond metformin and sulfonylureas, particularly for those who do not want insulin, are currently limited. The use of Actos has already declined significantly, due in part to concerns regarding side effects."
He also noted some of the study limitations. For instance, he writes there were statistically significant and meaningful imbalances favoring the patients in the control group in several characteristics that are known to be important risk factors for acute pancreatitis, particularly alcohol and gallstones. "The authors try to statistically control for these factors but that has its limitations," he writes. And he adds that the study did not consider differences in other medications, and pancreatitis diagnoses were unconfirmed.
In response, Singh writes us that, "like all studies, this study has some limitations that are noted. However, even after adjusting for risk factors a large more than 2 fold increased risk of hospitalization for acute pancreatitis was seen. In light of evidence from animal studies and case reports, this association is causal."
We asked Merck and Bristol-Myers for comment and will update you accordingly.
[UPDATE: A Bristol-Myers spokesman sends us this: "The Singh study is another data point in the larger body of research to date. Five previous retrospective cohort studies used a longitudinal design and insurance claims data to evaluate the risk of acute pancreatitis in patients using exenatide. None found a significantly increased risk of acute pancreatitis in users of GLP-1-based therapies compared with users of other antidiabetic agents. Those studies used a variety of different claims databases and analytic methods. The only study that showed an association between GLP-1-based therapies and acute pancreatitis was based on a cross-sectional analysis of spontaneously reported adverse events in the FDA Adverse Event Reporting System.
"...Based on spontaneous post-marketing reports, Byetta has been associated with acute pancreatitis, including fatal and non-fatal hemorrhagic or necrotizing pancreatitis. Byetta has not been studied in patients with a history of pancreatitis. It is unknown whether patients with a history of pancreatitis are at increased risk for pancreatitis while using Byetta. Other antidiabetic therapies should be considered in patients with a history of pancreatitis.]
[UPDATE: And now, Merck responds: "We are confident in the safety profile of (Januvia), which is an important medicine to help adults with type 2 diabetes lower their blood sugar levels. The safety is actively monitored by Merck and by regulatory agencies around the world; the findings from this analysis should not outweigh the large collection of data that supports the products safety profile. In our extensive evaluation of data, we have seen no compelling evidence of a causal relationship between the use of (Januvia) and pancreatitis.]
shock pic thx to ogimogi on flickr