Challenges in Causality
Assessment in Spontaneous
Reporting Systems
Syed Rizwanuddin Ahmad, MD, MPH, FISPE, FCP
3rd ICIUM, Antalya, Turkey
November 2011
Disclaimer
• These are my personal views not
shared by the U.S. FDA or the U.S.
government
• I am not representing the FDA
• No conflict of interest to declare
2
Outline
• Spontaneous Reporting Systems
– What
– Why
– Strengths/Limitations
• Evidence to Suggest Causal Relationship
• Approaches to Causality Assessment
– WHO-UMC
– Naranjo
– Bradford Hill
• Case example
• Safety-related Drug Withdrawals
• Conclusions
3
What is Spontaneous Reporting?
• The process of reporting of all unsolicited reports
of adverse events from health care professionals
or consumers to the FDA (or any appropriate
authority) is called spontaneous reporting
-Ahmad SR, Goetsch RA, Marks NS. Spontaneous reporting in the United States. Chapter 9. In Strom’s
Pharmacoepidemiology, 2005 p. 135-159.
4
Why Spontaneous Reporting?
Limitations of Pre-marketing Clinical trials
• Too small --- 2,000-5,000
• Too short --- <1 yr of exposure
• Too narrow --- restricted population
5
Spontaneous Reporting –
Strengths/Limitations
Strengths
• Inexpensive/All patients/drugs
• Generation of hypothesis and signals
• Good for identifying rare, serious drug-induced events with low
background rate
Limitations
•
•
•
•
•
•
Passive surveillance
Adverse event recognition
Underreporting
Duplicate reporting
Report quality
Reporting biases
6
Factors to Consider in Causal
Assessment
• Temporally associated with use of drug
• Biological plausible
• No other likely causes
– Underlying diseases or disease progression
– Concurrent meds
• Event abates after drug is stopped (+ dechallenge)
• Event recurs when drug is restarted (+ rechallenge)
7
Approaches to Causality
Assessment
• Expert Judgement
– individual assessments
– no standardized tool
• Algorithms
– sets of specific questions with or without scores
• Probabilistic or Bayesian methods
– Is based on assigning a prior probability to an event of
interest
Agbabiaka TB, et al. Methods for causality assessment of ADRs. Drug Safety
2008;31: 21-37
Jones JK. Determining causation from case reports. In Pharmacoepidemiology,
Strom BL. 2000, p. 525-538.
8
WHO UMC Causality Categories
•
•
•
•
•
•
Certain
Probable/Likely
Possible
Unlikely
Conditional/Unclassified
Unassessable/Unclassifiable
9
The Naranjo ADR Probability Scale
Questions
Yes
No
Don’t
Know
1) Are there previous conclusive reports on this
reaction?
+1
0
0
2) Did the ADR appear after the suspected drug was
administered?
+2
-1
0
3) Did the ADR improve when the drug was discontinued?
+1
0
0
4) Did the ADR appear with re-challenge?
+2
-1
0
5) Are there alternative causes for the ADR?
-1
+2
0
6) Did the reaction appear when placebo was given?
-1
+1
0
7) Was the drug detected in blood at toxic levels?
+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 drug in any previous exposure?
+1
0
0
10) Was the ADR confirmed by any objective evidence?
+1
0
0
10
Bradford-Hill Criteria
•
•
•
•
•
•
•
•
•
Strength
Consistency
Specificity
Temporality
Biological gradient
Plausibility
Coherence
Experimental evidence
Analogy
11
Case Example Exenatide and Pancreatitis
A 64-year-old, nonalcoholic woman with NIDDM presented with a
1-month history of epigastric pain beginning 2 days after
starting exenatide. Serum lipase concentration was 2700 U/L
(reference range, 114–320 U/L), and serum amylase
concentration was 131 U/L (reference range, 30–110 U/L).
Liver function test results, lipid profile, and serum creatinine
concentration were normal. Abdominal computed tomography
(CT) showed changes consistent with pancreatitis, and the
gallbladder was absent. Exenatide was discontinued.
Conservative therapy resulted in rapid resolution of symptoms,
normal lipase concentration (151 U/L), and normal findings
from CT of the pancreas 90 days later.
Ayoub W., et al. Exenatide-induced Acute Pancreatitis. Endocrine Practice. 2010;16(1):80-83.
12
Case Example Exenatide and Pancreatitis - 2
•
•
•
•
•
A search in FDA’s AERS database to identify additional cases of AP in
association with exenatide and other antidiabetics
Reporting rate was compared with the RR of comparator antidiabetics
Reporting rates are typically based on case counts divided by some measure
of drug’s utilization
If reporting rate of an event is > than background rate we can say that there
is a potential association between the drug & AE
Reporting rate for exenatide was higher compared to other
antidiabetics  labeling change
-Ahmad SR, Swann J. N Engl J Med 2008;358:1971-2.
-Graham DJ, Ahmad SR, Piazza Hepp TD. Spontaneous Reporting – USA. Pp. 237-247. In: Mann R, Andrews E. (eds.). Pharmacovigilance. 2nd ed.
2007
-Ahmad SR, Graham DJ. Exenatide and acute pancreatitis: Time to event analysis. Pharmacoepidemiol Drug Saf 2008;17:S132-3.
-Ahmad SR, Swann J. Reporting rates of hemorrhagic/necrotizing pancreatitis (HNP) in association with selected newer antidiabetics.
Pharmacoepidemiol Drug Saf 2009;18:S79-80.
13
Drug Withdrawals because of Safety Reasons
Drug Name
Year
Reason
Terfenadine (Seldane, Triludan)
1998
Drug interactions/Cardiac arrhythmias
Cisapride (Propulsid)
1999
Drug interactions/arrhythmias
Astemizole (Hismanal)
1999
Drug interactions/Cardiac arrhythmias
Troglitazone (Rezulin)
2000
Hepatotoxicity
Alosetron (Lotronex)
2000
Ischemic colitis; reintroduced 2002 on a restricted basis
Phenylpropanolamine
(Propagest, Dexatrim)
2000
Hemorrhagic stroke
Cerivastatin (Baycol, Lipobay)
2001
Rhabdomyolysis
Lumiracoxib (Prexige)
2007
Hepatotoxicity
Rimonabant (Acomplia)
2008
Depression/suicide
Sibutramine (Reductil/Meridia)
2010
Cardiovascular outcome
Rosiglitazone (Avandia)
2010
Myocardial infarction/death. Still available in the U.S.
14
Conclusions
• Spontaneous reporting systems are the most
common methodology used to generate and
detect new and rare signals
• In spite of challenges in causality
assessment, AE reports submitted to
spontaneous reporting systems have been
instrumental in most safety-related drug
withdrawals and labeling changes
15