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
- Appendix: Charts & Graphs
APPENDIX: CHARTS & GRAPHS
Table 1: Naranjo adverse drug reaction probability scale.[a]
|Yes||No||Do Not Know|
|1. Are there previous conclusive reports on this reaction?||+1||0||0|
|2. Did the adverse event appear after the suspected drug was administered?||+2||-1||0|
|3. Did the adverse reaction improve when the drug was discontinued or a specific antagonist was administered?||+1||0||0|
|4. Did the adverse reaction reappear when the drug was readministered?||+2||-1||0|
|5. Are there alternative causes (other than the drug) that could on their own have caused the reaction?||-1||+2||0|
|6. Did the reaction reappear when a placebo was given?||-1||+1||0|
|7. Was the drug detected in the blood (or other fluids) in concentrations known to be toxic?||+1||0||0|
|8. Was the reaction more severe when the dose was increased or less severe when the dose was decreased?||+1||0||0|
|9. Did the patient have a similar reaction to the same or similar drugs in any previous exposure?||+1||0||0|
|10. Was the adverse event confirmed by any objective evidence?||+1||0||0|
[a] Total scores can range from -4 to +13, which are then used, as follows, to assign a probability category for causality with the drug: ≤0 (“doubtful”), 1-4 (“possible”), 5-8 (“probable”), ≥9 (“definite”). Source: Naranjo CA, Busto U, Sellers EM, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther 1981; 30(2): 239-45.
Table 2. Liraglutide and acute pancreatitis: patient and case characteristics.[b]
|Age (n=220)||56 (10.4)|
|Weight (kg; n=150)||104 (27.7)|
|Number of Cases||Percentage of Cases|
|Time to onset of pancreatitis[c] (n=220; median 62 days)|
|Positive dechallenge[d] (n=278)||190||68%|
|Objective evidence[e] (n=278)||132||47%|
[b] Percentages represent a proportion of reported values (n) for each characteristic, indicated in parentheses. In some cases, percentages do not add up to 100% due to rounding.
[c] Times represent total time on drug before discontinuation. In virtually all cases, patients were discontinued from the medication immediately following the reaction; therefore, these times represent very close approximations of the time between first ingestion of the drug and onset of pancreatitis.
[d] Pancreatitis symptoms and/or signs improved upon drug discontinuation.
[e] See criteria for objective evidence in Methods.
Table 3. Liraglutide and acute pancreatitis: Naranjo scores and likelihood of causality (n = 278).[f]
|Naranjo Category||Naranjo Score||Number of cases||Percentage of cases|
[f] There were no cases that scored below 2 or above 9.
Table 4. Cases with “probable” causality and a Naranjo score of 7 (n=12).
|Age/ Sex||Relevant medical history||Dose at time of event (mg/d)||Duration on drug before event[g] (days)||Reported symptoms||Amylase (IU/L)[h]||Lipase (IU/L)[h]||Confirmed by imaging[h]||Disposition||Positive dechallenge|
|69F||No h/o alcohol abuse or gallstones||0.6||4||Severe abdominal pain||508||5,401||Yes||Hospitalized||Yes|
|42M[i]||No h/o alcohol abuse or pancreatitis||0.6||20||Not reported||103||391||No||Hospitalized||Yes|
|39F||No h/o alcohol abuse, gallstones, or pancreatitis||1.2||29||Substernal pain radiating to back||“elevated”||“elevated”||Yes||Other (Serious)||Yes|
|73F||No h/o alcohol abuse, gallstones, or pancreatitis||1.8||32||Abdominal pain radiating to back||217||799||No||Hospitalized||Yes|
|54F[i]||No h/o alcohol abuse, gallstones, peptic ulcer disease, or recent infections; previously was hospitalized after similar reaction to exenatide||1.2||35||Abdominal pain, dyspepsia||36||n/a||No||Other (Serious)||Yes|
|68M||No h/o alcohol abuse, gallstones, or pancreatitis||1.8||42||Abdominal pain, nausea, vomiting||122||281||No||Hospitalized||Yes|
|63F||“Physician stated the patient did not have any known risk factors for the development of pancreatitis”||1.8||46||Not reported||204||1,563||No||Hospitalized||Yes|
|51M||No h/o pancreatitis||1.8||46||“Sick feeling”||447||866||No||Other (Serious)||Yes|
|58F||No h/o alcohol abuse, gallstones, or pancreatitis||1.8||67||Constant abdominal pain radiating to back||n/a||1,853||No||Hospitalized||Yes|
|43M[i]||No h/o pancreatitis; “No biliary tract obstruction”||0.6||74||Abdominal pain||112||n/a||Yes||Hospitalized||Yes|
|55F||No h/o alcohol abuse, gallstones, or pancreatitis||1.2||81||Not reported||959||15,254||Yes||Hospitalized||Yes|
|75M||Triglycerides “controlled”; “No other risk factors for pancreatitis”||1.8||134||No abdominal pain||230||2,300||No||Other (Serious)||Yes|
Abbreviations: h/o, history of; M, male; F, female.
[g] Ordered by duration on drug before discontinuation. In all cases, patients were discontinued from the medication immediately following the reaction; therefore, these times represent very close approximations of the time between first ingestion of the drug and onset of pancreatitis.
[h] See criteria for objective evidence in Methods. Objective evidence was defined as an elevation of either amylase or lipase ≥3x the upper limit of the normal range (ULN) of the reporting laboratory or, if no reference range was provided, of the American College of Physicians’ Medical Knowledge Self-Assessment Program, 16th edition (ULN of 130 IU/L for amylase and 95 IU/L for lipase) (Source: American College of Physicians. Medical Knowledge Self-Assessment Program. 16th ed. 2012), or as a finding consistent with acute pancreatitis on imaging.
[i] Foreign reports (Lebanon [42yo male], UK [54yo female], and Japan [43yo male]).
Table 5. Acute pancreatitis rechallenge with liraglutide (“definite” causality: Naranjo score of 9).[j]
|Age/Sex||Relevant medical history||Pancreatitis episode/ interim recovery||Dose at time of event (mg/d)||Duration on drug before event/ duration until recovery (days)||Presenting symptoms||Amylase (IU/L)||Lipase (IU/L)||Positive dechallenge||Disposition|
|68 M||No h/o alcohol abuse, gallstones, or pancreatitis||1st (first started liraglutide)||1.8||44||Epigastric pain||179||904||n/a||Other (Serious)|
|Interim (off liraglutide)||n/a||8||None||60||184||Yes||Recovered|
|2nd (restarted liraglutide)||0.6||20||Epigastric pain||100||648||n/a||Other (Serious)|
|Final (off liraglutide)||n/a||6||None||89||369||Yes||Recovered|
[j] Reappearance of acute pancreatitis upon reinitiating liraglutide. The presiding practitioner diagnosed acute pancreatitis based on abdominal pain plus the objective evidence of an increase in lipase to ≥3x the upper limit of the normal range (ULN). Resolution of symptoms and a decrease in lipase following liraglutide withdrawal constituted a positive dechallenge; reappearance of abdominal pain and elevated lipase to ≥3x ULN constituted a positive rechallenge (ULN of 95 IU/L for lipase and 130 IU/L for amylase; Source: American College of Physicians. Medical Knowledge Self-Assessment Program. 16th ed. 2012).
Figure 1. Duration on liraglutide before the onset of acute pancreatitis.[k]
[k] Times represent total time on drug before discontinuation. In virtually all cases, patients were discontinued from the medication immediately following the reaction; therefore, these times represent very close approximations of the time between first use of the drug and the onset of pancreatitis (median 2.1 months).
Figure 2. Acute pancreatitis cases reported to the FDA: incretin mimetics versus non-incretin oral antihyperglycemics.[l]
[l] Arrows indicate year of approval. Analysis comparing the frequency of adverse event reports of acute pancreatitis with incretin mimetic drugs and non-incretin diabetes drugs. Using the same methodology as for our liraglutide search, we searched for unique cases of “pancreatitis acute” reported to the FDA between November 1997 (the earliest date for which these AERS data were available) and December 31, 2011, for all oral diabetes medications deemed “primary suspect” drugs. Incretin mimetic drugs approved through 2011 included liraglutide, exenatide, sitagliptin, saxagliptin, and linagliptin. Non-incretin mimetic oral diabetes drugs included metformin, glyburide, glimepiride, glipizide, rosiglitazone, pioglitazone, nateglinide, repaglinide, acarbose, and miglitol. Arrows indicate the year of FDA approval for drugs approved during or after 1997.
Non-incretin mimetic diabetes drugs are collectively much more widely prescribed than incretin mimetics. Metformin constituted 58% of all cases of acute pancreatitis cases reported with non-incretin mimetic oral diabetes drugs. In 2011, there were 59 million prescriptions filled in the U.S. for metformin. (Source: IMS Health. Top 25 Medicines by Dispensed Prescriptions (U.S.). Updated March 22, 2013. https://www.citizen.org/sites/default/files/top_25_medicines_dispensed_prescriptions_u.s.pdf. Accessed September 4, 2013.)
In the same year, there were approximately 1.4 million prescriptions for exenatide, 1.4 million for liraglutide, 7.1 million for sitagliptin, 1.4 million for saxagliptin, and 100,000 for linagliptin. (Source: IMS Health.)