- Blood Sugar
- Clinical Studies
- Complementary Medicine
- Diabetes and Men
- Erectile Dysfunction
- Food and Nutrition
- Insulin Pumps
- Meters and Test Strips
- Product Recalls
- Type 1
- Type 2
- Weight Loss Surgery
Gastroenterology Pancreatic Risks No Greater With Victoza
December 3, 2013 (MedPage Today) -- The treatment of type 2 diabetes with liraglutide (Victoza) didn't lead to a greater rate of acute pancreatitis or pancreatic cancer compared with other anti-diabetic drugs, an interim analysis suggested.
The incidence rate for acute pancreatitis among patients receiving the glucagon-like peptide-1 (GLP-1) analog was 187.5 per 100,000 person-years compared with 154.4 per 100,000 for a group of pooled comparator drugs, according to Donnie Funch, PhD, of Optum Epidemiology in Waltham, Mass., and colleagues.
And the incidence rate for pancreatic cancer was 19.9 per 100,000 compared with 33 per 100,000 for the pooled comparator group, the researchers reported online in Diabetes, Obesity and Metabolism.
Recent reports have linked the use of GLP-1 receptor agonists with pancreatitis and hospitalization for acute pancreatitis, and the FDA has been reviewing data about pancreatitis and precancerous changes.
"Before FDA approval, we initiated a prospective surveillance program to evaluate potential adverse effects of liraglutide in the USA," Funch and colleagues wrote.
The Optum Research Database surveillance program includes national claims through commercial health insurance plans for liraglutide and comparator drugs, and is expected to continue through 2014.
The pooled comparator group included metformin, three types of sulfonylureas, and pioglitazone. Exenatide and the dipeptidyl peptidase 4 inhibitors were excluded from this analysis because those agents had previously been reported to have adverse pancreatic outcomes.
Between February 2010 and December 2012, a total of 47 patients on liraglutide developed pancreatitis during 25,072 person-years of treatment, as did 472 on the various comparator drugs during 305,621 person-years.
In the study period, five patients on liraglutide were diagnosed with pancreatic cancer during 25,114 person-years of therapy, as did 101 of patients taking the comparator drugs during 306,064 person-years.
The liraglutide-treated patients were more often women (54.2% versus 49.5%), had higher rates of overweight and obesity (21.2% versus 13.1%), and had more complications including neuropathy, nephropathy, and retinopathy (15.7% versus 8.3%).
Patients receiving liraglutide also were more commonly on insulin (28.1% versus 10.3%).
Median follow-up was 15 months.
For liraglutide versus the comparator drugs, the adjusted relative risks for acute pancreatitis and pancreatic cancer were 1.10 (95% CI 0.81-1.49) and 0.65 (95% CI 0.26-1.60), respectively.
The researchers also found these relative risks for acute pancreatitis for liraglutide versus the individual comparator drugs:
- Metformin, RR 1.14 (95% CI 0.83-1.56)
- Sulfonylureas, RR 1.04 (95% CI 0.73-1.48)
- Pioglitazone, RR 0.95 (95% CI 0.65-1.39)
For pancreatic cancer, the relative risks for the three comparators were:
- Metformin, RR 0.81 (95% CI 0.32-2.05)
- Sulfonylureas, RR 0.40 (95% CI 0.15-1.06)
- Pioglitazone, RR 0.49 (95% CI 0.17-1.41)
For all the drugs, the median time from treatment initiation to pancreatic cancer diagnosis was 270 days.
Because early diagnosis of cancer can reflect pre-existing disease, the researchers also analyzed rates after excluding patients whose pancreatic cancer diagnosis was within the first 3 months of treatment and found an adjusted relative risk of 0.82 (95% CI 0.33-2.05) for liraglutide versus the comparators.
"The use of large insurance claims databases permits the rapid study of large numbers of patients under routine care," Funch and colleagues noted.
However, they acknowledged that their analysis did have limitations. The median follow-up of 15 months may have been adequate to assess the risk for acute pancreatitis, but possibly not for pancreatic cancer.
In addition, their calculations are likely to represent overestimates, because in this interim analysis diagnoses weren't confirmed. There also may have been residual confounding because of population differences.
"Future analyses within this data resource will be based on larger cohorts, more follow-up time, and adjudicated outcomes of interest (through review of medical records)," the researchers concluded.