Abstract
Objective: Adverse eventreports (AERs) submitted to the US Food and Drug Administration (FDA)
were reviewed to assess the bleeding complications induced by the
administration of antiplatelets and to attempt to determine the
rank-order of the association.
Methods: After a deletion of
duplicated submissions and the revision of arbitrary drug names, AERs
involving warfarin, aspirin, cilostazol, clopidogrel, ethyl
icosapentate, limaprost alfadex, sarpogrelate, and ticlopidine were
analyzed. Authorized pharmacovigilance tools were used for the
quantitative detection of signals, i.e., drug-associated adverse events,
including the proportional reporting ratio, the reporting odds ratio,
the information component given by a Bayesian confidence propagation
neural network, and the empirical Bayes geometric mean.
Results:
Based on 22,017,956 co-occurrences, i.e., drug-adverse event pairs,
found in 1,644,220 AERs from 2004 to 2009, 736 adverse events were
listed as warfarin-associated adverse events, and 147 of the 736 were
bleeding complications, including haemorrhage and haematoma. Both
aspirin and clopidogrel were associated with haemorrhage, but the
association was more noteworthy for clopidogrel. As for bleeding
complications related to the gastrointestinal system, e.g., melaena and
haematochezia, the statistical metrics suggested a stronger association
for aspirin than clopidogrel. The total number of co-occurrences was not
large enough to compare the association with bleeding complications for
the other 5 antiplatelets.
Conclusions: The data strongly
suggest the necessity of well-organized clinical studies with respect to
antiplatelet-associated bleeding complications.
Keywords: adverse events, AERS, warfarin, antiplatelets, pharmacovigilance.
Introduction
Arterial platelet-rich thrombosis differs from venous fibrin-richthrombosis in terms of pathogenesis and prevention strategy, but both
have considerable medical impact [1, 2].
Myocardial infarction or stroke is often caused by arterial thrombosis,
and venous thrombosis accounts for considerable mortality from
cardiovascular events [1, 2].
Antiplatelets are the basis for prevention of arterial thrombosis,
whereas anticoagulants are effective for venous thrombosis; however,
recent molecular investigations suggest the interdependence of platelets
and the coagulation system in both forms of thrombosis [2], and physicians often recommend their combinations especially in patients with atrial fibrillation [3].
Currently,
a myriad of patients are receiving combined warfarin-aspirin therapy to
prevent thrombosis, but there is little clinical evidence of a
therapeutic benefit compared with either alone, except for patients with
acute coronary syndrome and mechanical heart valves [3-5].
The combination of warfarin and aspirin is more effective than aspirin
alone for the prevention of recurrent cardiovascular events in patients
with acute coronary syndrome [6, 7]
and is more effective than warfarin alone for the prevention of
thromboembolic events in patients with mechanical heart valves [8, 9].
On the other hand, there is compelling evidence that the
warfarin-aspirin combination results in an increase in risk for serious
bleeding complications [6-12].
Collectively, we should contemplate the risk-benefit balance of this
combination, especially when data from randomized controlled trials are
lacking [3-5].
Most
reports on bleeding complications caused by anticoagulants and/or
antiplatelets are of warfarin and/or aspirin, with little information
available for other antiplatelets, e.g., cilostazol, clopidogrel, ethyl
icosapentate, limaprost alfadex, sarpogrelate and ticlopidine This study
was conducted to assess the bleeding complications induced by the
administration of antiplatelets and to attempt to determine the
rank-order of the association, using more than a million case reports on
adverse events (AERs) submitted to the US Food and Drug Administration
(FDA). Authorized pharmacovigilance methods were used for quantitative
signal detection [13-19],
where a signal means a drug-associated adverse event or an association
between a drug and an adverse event. Here, 7 antiplatelets were compared
with warfarin in terms of susceptibility to bleeding complications.
Methods
Data sources
Input data for this study were taken from thepublic release of the FDA's Adverse Event Reporting System (AERS)
database, which covers the period from the first quarter of 2004 through
the end of 2009. The total number of reports used was 2,231,029. This
database relies on reports of spontaneous adverse events by health
professionals, consumers, and manufacturers. The data structure of AERS
is in compliance with international safety reporting guidance, ICH E2B,
consisting of 7 data sets; patient demographic and administrative
information (DEMO), drug/biologic information (DRUG), adverse events
(REAC), patient outcomes (OUTC), report sources (RPSR), drug therapy
start and end dates (THER), and indications for use/diagnosis (INDI).
The adverse events in REAC are coded using preferred terms (PTs) in the
Medical Dictionary for Regulatory Activities (MedDRA) terminology. Here,
version 13.0 of MedDRA was used.
Prior to analysis, duplicated
reports were deleted according to the FDA's recommendation of adopting
the most recent CASE number, resulting in a reduction in the number of
reports from 2,231,029 to 1,644,220. All drug names were unified into
generic names by a text-mining approach, because AERS permits the
registering of arbitrary drug names, including trade names and
abbreviations. Spelling errors were detected by a spell checker
software, GNU Aspell, and carefully confirmed by working pharmacists.
The total number of errors was 223,239. Foods, beverages, treatments
(e.g. X-ray radiation), and unspecified names (e.g. beta-blockers) were
omitted for this study, and the total number of omissions was 164,384. A
total of 22,017,956 co-occurrences were found in 1,644,220 reports,
where a co-occurrence was a pair of a drug and an adverse drug event.
Definition of bleeding complications
The Standard MedDRAQueries (SMQs) are groupings of PT terms, which relate to defined
medical conditions or areas of interest. A total of 82 SMQs have been
released by the MedDRA maintenance and support services organization (http://www.meddramsso.com/).
The heamorrhage SMQ consists of 421 PT terms; 91 heamorrhage laboratory
PT terms (e.g., PT10022595: international normalised ratio (INR)
increased, PT10022592: INR abnormal) and 330 heamorrhage PT terms (e.g.,
PT10055798: heamorrhage, PT10018852: haematoma). Here, 421 PT terms
found in the heamorrhage SMQ were defined as bleeding complications.
Aspirin as antiplatelets
Aspirin is indicated for preventionof arterial thrombosis, but is also used to treat mild to moderate pain,
fever and inflammatory diseases. Inclusion criteria for analysis
included 1) “bayaspirin”, “baby aspirin” or “children aspirin” for drug
name, 2) term “low dose” in drug name, 3) PT terms located within a
system organ class (SOC) of cardiac disorders (SOC10007541) or vascular
disorders (SOC10047065), 4) PT terms located within a high level group
term (HLGT) of vascular therapeutic procedures (HLGT10003184), 5) terms
“stroke”, “infarction”, “thrombosis”, “ischemia”, “attack”, “artery”,
“vascular” or “sclerosis” in PT terms, and 6) daily dose of 325mg or
less.
Data mining
In pharmacovigilance analyses, data miningalgorithms have been developed to identify drug-associated adverse
events as signals that are reported more frequently than expected by
estimating expected reporting frequencies on the basis of information on
all drugs and all events in the database [17-19]. For example, the proportional reporting ratio (PRR) [13], the reporting odds ratio (ROR) [14], the information component (IC) [15], and the empirical Bayes geometric mean (EBGM) [16]
are widely used, and indeed, the PRR is currently employed by the
Medicines and Healthcare products Regulatory Agency (MHRA), UK, the ROR
by the Netherlands Pharmacovigilance Centre, the IC by the World Health
Organization (WHO), and the EBGM by the FDA.
All of these
algorithms extract decision rules for signal detection and/or calculate
scores to measure the associations between drugs and adverse events from
a two-by-two frequency table of counts that involve the presence or
absence of a particular drug and a particular event occurring in case
reports. These algorithms, however, differ from one another in that the
PRR and ROR are frequentist (non-Bayesian), whereas the IC and EBGM are
Bayesian. In this section, only the scoring thresholds used in the
present study are given, and the reader is referred to review articles
for more extensive details of each statistical test [17-19].
Here,
we define how a drug and associated adverse event is classified as a
signal when using each statistical test. Using the PRR, a signal is
detected if the count of co-occurrences is 3 or more and the PRR is 2 or
more with an associated χ2 value of 4 or more [13]. For the ROR, a signal is detected, if the lower bound of the 95% two-sided confidence interval exceeds 1 [14].
Signal detection using the IC is done using the IC025 metric, a
criterion indicating the lower bound of the 95% two-sided confidence
interval of the IC, and a signal is detected if the IC025 value exceeds 0
[15].
Finally, the EB05 metric, a lower one-sided 95% confidence limit of
EBGM, is used and a signal is detected when EB05 is greater than or
equal to the threshold value 2.0 [16].
In this study, the adverse events were listed as drug-associated, when
at least 1 of 4 indices met the criteria indicated above.
Results
Table 1lists the total number of co-occurrences, and the number of adverse
events listed as drug-associated adverse events with co-occurrences. The
total number of co-occurrences with warfarin was 156,357, and 59,855
for aspirin and 121,166 for clopidogrel, representing 0.710%, 0.272% and
0.550% of all co-occurrences in the database, respectively. In total,
736, 848 and 838 adverse events were listed as drug-associated adverse
events with 64,289, 24,536, and 55,079 co-occurrences, respectively. The
total number of co-occurrences was not large enough to compare the
association with adverse events for cilostazol, ethyl icosapentate,
limaprost alfadex, sarpogrelate and ticlopidine.
Of 736 warfarin-associated adverse events, 147 were bleeding complications, the worst 20 being listed in Table 2.
Based on the number of co-occurrences, the worst is an increase of INR,
followed by heamorrhage, gastrointestinal heamorrhage, epistaxis, and a
prolongation of prothrombin time in this order. Aspirin was also
associated with these bleeding complications with exception of an
abnormal INR. In the case of clopidogrel, exceptions included an
increase of INR, a prolongation of prothrombin time, an abnormal INR and
a prolongation of activated partial thromboplastin time (statistical
data not shown).
Table 3
lists the statistical data on the association of warfarin, aspirin, and
clopidogrel with heamorrhage and haematoma. The association with
heamorrhage was more noteworthy for clopidogrel than aspirin, but the
signals were weaker than for warfarin. A stronger signal for clopidogrel
than aspirin was also observed for contusion (statistical data not
shown).
Table 4
lists the data on melaena and haematochezia. The statistical metrics
suggested a stronger association for aspirin than clopidogrel. This
order was also found for gastrointestinal haemorrhage, a decrease of
haematocrit value, cerebral haemorrhage, rectal haemorrhage,
haemoptysis, subdural haematoma, haematemesis and gastric haemorrhage
(statistical data not shown).
Discussion
The AERS database is considered a valuable tool; however, somelimitations inherent to spontaneous reporting have been pointed out [17].
First, the data occasionally contain misspelling and miswords, although
the structure of AERS is in compliance with the international safety
reporting guidance. Second, the system was started more than 10 years
ago, and reporting patterns have changed over time. Third, the adverse
events are coded using hierarchical terms of PTs of MedDRA, and changes
in terminology over time also might affect the quality of the database.
Last, there are a number of duplicate entries in the database. To
overcome problems with data quality, we manually corrected mistakes in
the data entities and deleted duplicates according to the FDA's
recommended method. Previously, this system has been used to assess
adverse events accompanying the use of platinum agents [20], statins [21], 5-fluorouracil and capecitabine [22], tigecycline [23] and omeprazole and esomeprazole [24], and anticancer agent-associated hypersensitivity reactions [25, 26].
The reproducibility of clinical observations was suggested, but the
number of co-occurrences should be large enough to detect a signal.
Table 1
Number of warfarin- and antiplatelet-associated adverse events.
| Co-occurrences in database a) | Adverse events b) | Co-occurrences with signal detected b) | |
|---|---|---|---|
| warfarin | 156,357 | 736 | 64,289 |
| aspirin | 59,855 | 848 | 24,536 |
| cilostazol | 8,410 | 459 | 3,337 |
| clopidogrel | 121,166 | 838 | 55,079 |
| ethyl icosapentate | 1,838 | 292 | 909 |
| limaprost alfadex | 1,052 | 227 | 521 |
| sarpogrelate | 1,081 | 223 | 540 |
| ticlopidine | 8,867 | 500 | 4,361 |
a) the total number of co-occurrences in the database.
b) the number of adverse events listed as drug-associated adverse events with co-occurrences.
Table 2 b) the number of adverse events listed as drug-associated adverse events with co-occurrences.
Worst 20 ranking bleeding complications reported for warfarin.
| N | bleeding complications |
|---|---|
| 4753 | International normalised ratio increased |
| 1072 | Haemorrhage |
| 1014 | Gastrointestinal haemorrhage |
| 724 | Epistaxis |
| 689 | Prothrombin time prolonged |
| 630 | Contusion |
| 595 | International normalised ratio abnormal |
| 526 | Haematocrit decreased |
| 522 | Haematuria |
| 494 | Haematoma |
| 489 | Cerebral haemorrhage |
| 468 | Rectal haemorrhage |
| 319 | Melaena |
| 312 | Haemoptysis |
| 283 | Haemorrhage intracranial |
| 268 | Subdural haematoma |
| 268 | Haematemesis |
| 254 | Haematochezia |
| 237 | Activated partial thromboplastin time prolonged |
| 184 | Gastric haemorrhage |
The total number of co-occurrences with warfarin was
156,357 in the AERS database. The adverse events were listed when at
least 1 of 4 indices met the criteria, and 736 adverse events were
listed as warfarin-associated adverse events with 64,289 co-occurrences
in total. Among the 736 events, 147 were bleeding complications. The
worst 20 were ranked according to the number of co-occurrences (N), with
the official PT terms of MedDRA ver. 13.0.
Table 3 156,357 in the AERS database. The adverse events were listed when at
least 1 of 4 indices met the criteria, and 736 adverse events were
listed as warfarin-associated adverse events with 64,289 co-occurrences
in total. Among the 736 events, 147 were bleeding complications. The
worst 20 were ranked according to the number of co-occurrences (N), with
the official PT terms of MedDRA ver. 13.0.
Signal detections for warfarin-, aspirin- and clopidogrel-associated haemorrhage and haematoma.
| N | PRR (χ2) | ROR (95% two-sided CI) | IC (95% two-sided CI) | EBGM (95% one-sided CI) | |
|---|---|---|---|---|---|
| Haemorrhage | |||||
| Warfarin | 1072 | 4.938* (3363.124) | 5.080* (4.779, 5.381) | 2.290* (2.202, 2.378) | 4.902* (4.659) |
| Aspirin | 209 | 2.506* (187.540) | 2.517* (2.196, 2.837) | 1.309* (1.112, 1.505) | 2.455* (2.188) |
| Clopidogrel | 541 | 3.208* (820.735) | 3.247* (2.982, 3.513) | 1.670* (1.547, 1.793) | 3.170* (2.952) |
| Haematoma | |||||
| Warfarin | 494 | 4.226* (1218.110) | 4.326* (3.955, 4.697) | 2.065* (1.936, 2.194) | 4.186* (3.881) |
| Aspirin | 118 | 2.634* (118.048) | 2.646* (2.207, 3.084) | 1.370* (1.109, 1.631) | 2.542* (2.180) |
| Clopidogrel | 254 | 2.801* (293.327) | 2.829* (2.499, 3.159) | 1.471* (1.292, 1.650) | 2.750* (2.477) |
N: the number of co-occurrences.
*: signal detected, and a signal means a drug-associated adverse event (see “Methods” for the criteria of detection).
Haemorrhage and haematoma were coded as PT10055798 and PT10018852, respectively.
PRR: the proportional reporting ratio, ROR: the reporting odds ratio,
IC: the information component, EBGM: the empirical Bayes geometric
mean.
Table 4 *: signal detected, and a signal means a drug-associated adverse event (see “Methods” for the criteria of detection).
Haemorrhage and haematoma were coded as PT10055798 and PT10018852, respectively.
PRR: the proportional reporting ratio, ROR: the reporting odds ratio,
IC: the information component, EBGM: the empirical Bayes geometric
mean.
Signal detections for warfarin-, aspirin- and clopidogrel-associated melaena and haematochezia.
| N | PRR (χ2) | ROR (95% two-sided CI) | IC (95% two-sided CI) | EBGM (95% one-sided CI) | |
|---|---|---|---|---|---|
| Melaena | |||||
| Warfarin | 319 | 3.019* (430.996) | 3.064* (2.742, 3.385) | 1.581* (1.421, 1.741) | 2.972* (2.708) |
| Aspirin | 334 | 8.288* (2126.968) | 8.455* (7.583, 9.326) | 3.011* (2.854, 3.168) | 8.231* (7.516) |
| Clopidogrel | 236 | 2.882* (289.243) | 2.913* (2.561, 3.264) | 1.511* (1.325, 1.696) | 2.825* (2.535) |
| Haematochezia | |||||
| Warfarin | 254 | 2.196* (165.111) | 2.215* (1.956, 2.473) | 1.124* (0.945, 1.303) | 2.165 (1.951) |
| Aspirin | 143 | 3.232* (218.231) | 3.252* (2.757, 3.746) | 1.663* (1.425, 1.900) | 3.123* (2.712) |
| Clopidogrel | 153 | 1.706 (44.202) | 1.713* (1.461, 1.965) | 0.759* (0.529, 0.988) | 1.679 (1.468) |
N: the number of co-occurrences.
*: signal detected, and a signal means a drug-associated adverse event (see “Methods” for the criteria of detection).
Melaena and hematochezia were coded as PT10027141 and PT10018836, respectively.
PRR: the proportional reporting ratio, ROR: the reporting odds ratio,
IC: the information component, EBGM: the empirical Bayes geometric
mean.
In this study, a total of 736 adverse events were *: signal detected, and a signal means a drug-associated adverse event (see “Methods” for the criteria of detection).
Melaena and hematochezia were coded as PT10027141 and PT10018836, respectively.
PRR: the proportional reporting ratio, ROR: the reporting odds ratio,
IC: the information component, EBGM: the empirical Bayes geometric
mean.
listed as warfarin-associated adverse events, and 147 of the 736 were
bleeding complications. In 2007, Wysowski et al. reported the worst 30
ranking adverse events with use of warfarin using the AERS database from
January 1993 to July 2006 [27].
Based on the number of reports, the adverse events ranked at a
relatively high position included an increase of INR, drug interaction,
gastrointestinal haemorrhage, haemorrhage, haematuria, anaemia,
epistaxis, and melaena [27].
These adverse events also ranked at a higher position in this study,
although the ranking was based on the number of occurrences and those
without signals detected were excluded.
Both aspirin and
clopidogrel were also associated with bleeding complications. The
association with haemorrhage was more noteworthy for clopidogrel than
aspirin; however, for gastrointestinal bleeding complications, the
statistical metrics suggested a stronger association for aspirin than
clopidogrel. Recently, Hausen et al. reported the bleeding risk in
patients with atrial fibrillation, treated with warfarin, aspirin,
clopidogrel, and combinations thereof [10].
They performed a cohort study using the data in the Danish National
Patient Registry, a system started in1978, and employed the Cox model to
estimate the risk of bleeding. Using warfarin monotherapy as a
reference, the hazard ratio (95% confidential interval) for fatal
bleeding was 1.37 (1.13-1.65) and 2.22 (1.30-3.77) for aspirin and
clopidogrel monotherapy, respectively, but for gastrointestinal
bleeding, it was 1.28 (1.17-1.41) and 1.18 (0.84-1.67), respectively.
The data obtained in the present study does not conflict with the report
by Hausen et al.
The incidence of bleeding with use of warfarin
was variable among the clinical reports, and the variation might be
explained by many factors, including definition of bleeding, patient
mixes with indications and risks of bleeding, targeted INR, treatment
protocol, treatment setting and length of follow-up [27].
The ARES database has an advantage in the use of well-organized
authorized terms of MedDRA, although the incidence cannot be calculated
in this analysis. Additionally, it should be noted that there is no
credible counterfactual means, e.g., a randomized control group, to
extract drug-associated adverse events as signals, and therefore
disease-oriented adverse events can be listed as signals. The results
can be biased by unmeasured confounding factors. Although the comparison
of aspirin with clopidogrel possibly offsets them, a statistically
well-organized methodology should be established to minimize their
effects. In conclusion, the data strongly suggest the necessity of
well-organized clinical studies with respect to antiplatelet-associated
bleeding complications.
No comments:
Post a Comment