Acute Pancreatitis with Liraglutide – Results

An analysis of MedWatch adverse event reports submitted to the Food and Drug Administration in the first two years following approval

June 5, 2014

• Cover Letter to the FDA (pdf)
• Full Report (pdf)
• View the Press Release
• June 17, 2014 – FDA Response
• Executive Summary
• Part 1: Background
• Part 2: Methods
• Part 3: Results
• Part 4: Discussion
• Part 5: Conclusion
• Acknowledgements
• Appendix: Charts & Graphs

PART III: RESULTS

We initially received from the FDA a total of 390 MedWatch case report forms of acute pancreatitis associated with liraglutide therapy during the study period. Ninety-eight (98) cases were duplicates, and 13 others were excluded because it was not clear that liraglutide had been used or because no acute pancreatitis adverse event was reported. One additional case was excluded because the case report form was misplaced and a replacement could not be obtained from the FDA. As a result, 278 unique cases represented the final population for analysis.

Patient and case characteristics

Table 2 presents patient and case characteristics. The mean age of patients was 56 years (standard deviation [SD] 10.4), with a mean weight of 104 kilograms (SD 27.7). There were approximately equal numbers of males (49%) and females (51%). Seventy-seven percent of the patients in the reports came from the U.S., with 18% from Europe, 3% from Japan, and 2% from other countries.

Of the cases for which the dose of liraglutide at the time of the acute pancreatitis event was reported, equal numbers of patients (41%) were on 1.2 mg and 1.8 mg daily doses, with 14% on the lower daily dose of 0.6 mg and 4.6% on other doses.

Most patients presented with symptoms of acute pancreatitis soon after beginning liraglutide: 30% during the first month, 50% within the first two months, and 72% within the first four months (Table 2; Figure 1); median duration of therapy before onset of acute pancreatitis was 2.1 months. Approximately half (47%) of the cases reported a confirmation of the diagnosis through a serum amylase or lipase ≥3x ULN, imaging studies, or both (Table 2). In most cases (68%), the patient recovered from the acute pancreatitis attack following discontinuation of liraglutide (positive dechallenge). The majority (54%) of cases had no reported alternative cause (not shown in Table 2).

Of the 273 cases that reported outcomes, 164 patients (60%) required hospitalization, 106 (39%) had other (unspecified) serious outcomes, and three (1%) died. Two of the three deaths were due to complications stemming from acute pancreatitis. One of these two patients died of multi-system organ failure and the other from acute respiratory distress syndrome, both deemed by the patients’ physicians to be complications of severe acute pancreatitis. Neither patient had a prior history of pancreatitis, gallstone disease, or alcohol abuse; in one of the two cases, the physician stated that there was no other plausible etiology.

Naranjo analyses

Of the 278 unique cases of acute pancreatitis, we classified one as “definitely” caused (Naranjo score of 9), 51 as “probably” caused (score of 5-8), and 226 as “possibly” caused (score of 1-4) by liraglutide (Table 3).

Of the 51 “probable” cases (Naranjo score of 5-8), 12 had a Naranjo score of 7, in which a causal link with liraglutide appeared particularly convincing (Table 4). In these 12 cases, there was no reported history of pancreatitis or evidence of alternative causes, and symptoms resolved following discontinuation of the drug (positive dechallenge). In 11 of the 12 cases, acute pancreatitis occurred within three months of starting liraglutide, and 10 of the 12 cases were confirmed with objective evidence. One of the patients reportedly had a similar reaction to exenatide previously, though the exact nature of that reaction was not further specified.

Positive rechallenge case

There were three cases in which acute pancreatitis recurred following reexposure to liraglutide. However, only one case had no confounding factors, leading us to consider it a positive rechallenge with a “definite” causal relationship to liraglutide (Naranjo score of 9). This case is described here (and in Table 5) in further detail.

A 68-year-old man with a 10-year history of type 2 diabetes mellitus, chronic kidney disease, hypertension, spinal stenosis, and restless leg syndrome began liraglutide and self-titrated up to a daily dose of 1.8 mg. The patient had no prior history of pancreatitis, alcohol abuse, gallstones, or recent infections, and his serum triglyceride levels were stated to always be below 300 mg/dL. The patient had been on exenatide but had been switched to liraglutide due to a lack of efficacy.

Forty-four days after starting liraglutide, he presented with complaints of epigastric pain. Laboratory tests showed an amylase of 179 IU/L (3X ULN = 390) and lipase of 904 IU/L (3X ULN = 285). He was not hospitalized but told to discontinue liraglutide and begin a clear liquid diet. Eight days later, repeat lab values showed that both amylase and lipase had decreased (60 and 184 IU/L, respectively), and he was considered recovered from acute pancreatitis. That same day, the patient was restarted on liraglutide.

Epigastric pain recurred three weeks later, at which time lab work again showed a slightly elevated amylase and markedly increased lipase (100 and 648 IU/L, respectively). He discontinued the drug and, one week later, symptoms had resolved and serum amylase and lipase were again decreasing (89 and 369 IU/L, respectively).

Next Page » Part IV: Discussion