Webonly Supplementary Material
Table w1. RCTs and meta-analyses with data on the risk of MI in patients with non valvular AF treated with NOACs.
(a)Primary RCTs
Author/year
Connolly et al20, 70
Study design
RE-LY trial, comparing
two fixed doses of
dabigatran (110 mg
bid and 150 mg bid),
in patients with non
valvular AF
Size
18113 patients of
RE-LY trial
Summary of findings
First data analysis: MI occurred in 0.72%/year of
patients treated with dabigatran 110 mg bid,
0.74%/year of patients treated with dabigatran 150
mg bid, 0.53%/year of patients treated with warfarin
(HR 1.35, 95% CI 0.98–1.87, p=0.07 for dabigatran
110 vs. warfarin and HR 1.38, 95% CI 1.00–1.91, p=
0.048 for dabigatran 150 vs. warfarin.
Revision of data after blinded adjudication of
additional events following data locking: MI occurred
in 0.82%/year of patients. treated with dabigatran
110 mg bid, 0.81%/year of patients treated with
dabigatran 150 mg bid, 0.64%/year of patients
treated with warfarin (HR 1.29, 95% CI 0.96–1.75,
p=0.09 for dabigatran 110 vs. warfarin and HR 1.27,
95% CI 0.94–1.71, p= 0.12 for dabigatran 150 vs.
warfarin.
Patel et al.21.
ROCKET AF trial,
comparing
rivaroxaban (at adaily
dose of 20 mg) with
warfarin in patients
with non valvular AF .
14264 patients of
ROCKET AF trial
MI occurred in 0.9% /year of patients treated with
rivaroxaban compared with 1.1%/year of patients
treated with warfarin (HR 0.81, 95% CI, 0.63-1.06; p =
0.12).
Connolly et al.65
AVERROES trial,
5599 patients of
comparingapixaban (5 AVERROES trial
mg bid) or aspirin (81
to 324 mg per day) in
MI occurred in 0.8%/year of patients treated with
apixaban compared with 0.86%/year of patients
treated with aspirin (HR 0.86, 95% CI 0.50–1.48, p=
0.59).
Comment
Overall, 16.6% of all
patients had a prior
myocardial infarction.
patients with non
valvular AF for whom
VKA therapy was
unsuitable
Granger et al.22
ARISTOTLE trial,
comparing apixaban
(at a dose of 5 mg
bid) with warfarin
18201 patients of
ARISTOTLE trial
MI occurred in 0.53%/year of patients treated with
apixaban compared with 0.61%/year of patients
treated with warfarin (HR 0.88, 95% CI 0.66–1.17, p=
0.37).
Giugliano et al66
ENGAGE-AF trial,
comparing two doses
of edoxaban (60 and
30 mg, respectively)
21,105 patients of
MI occurred in 0.75%/year of patients treated with
warfarin compared with 0.70%/year of patients
treated with high dose edoxaban (HR 0.94, 95% CI
0.74–1.19) and compared with 0.60%/year of
patients treated with low dose edoxaban (HR 1.19,
95% CI 0.95–1.49)
(b)Posthoc
analyses of RCTs
and metaanalyses
Hohnloser et al71
Detailed analysis of
the RE-LY trial to
report rates of MI,
unstable angina,
cardiac arrest, and
cardiac death
ENGAGE AF trial
MI occurred at annual rates of 0.82% and 0.81% with The analysis was not predabigatran 110 or 150 mg bid compared with 0.64%
specified and are post hoc
with warfarin (HR 1.29, 95% CI 0.96-1.75, P=0.09 for
dabigatran 110 mg; HR 1.27, 95% CI 0.94-1.71, P=0.12
for dabigatran 150 mg).
Annual rates of a composite of MI, unstable angina,
cardiac arrest, and cardiac death were 3.16% per year
with dabigatran 110 mg, 3.33% per year with
dabigatran 150 mg, and 3.41% per year with warfarin
(HR versus warfarin 0.93, 95% CI 0.80-1.06, P=0.28 for
dabigatran 110 mg and HR 0.98, 95% CI 0.85-1.12,
P=0.77 for dabigatran 150 mg).
Uchino et al.67
Dentali et al.68
Mak et al.69
Metanalysis of 7 RCTs
on dabigatran: 2
studies in stroke
prevention in AF,
1 in ACS, 1 in acute
venous
thromboembolism,
and 3 on short-term
prophylaxis of DVT.
Control arms included
warfarin, enoxaparin,
or placebo.
Meta-analysis of 12
phase II and phase III
RCTs comparing
NOACs with vitamin K
antagonists in
patients with AF.
Metanalysis of 28
RCTs on ximelagatran,
dabigatran,
rivaroxaban and
apixaban with the aim
to assess the effect of
these agents on MI or
acute coronary
syndrome (MI/ACS),
major bleeding
Overall 30 514
patients
Dabigatran was significantly associated with a higher
risk of MI or ACS than that seen with agents used in
the control group (1.19% in dabigatran vs. 0.79% in
control, OR 1.33, 95% CI 1.03-1.71; P=.03).
So far Dabigatran has not
been tested in a phase III
study of patients with
recent ACS.
The risk of MI or ACS was similar when using revised
RE-LY trial results (OR 1.27; 95% CI 1.00-1.61; P=.05)
or after exclusion of short-term trials (OR 1.33, 95%
CI, 1.03-1.72; P=.03).
Overall 30 584
patients treated
with NOACs and 23
531 patients
treated with
VKAs
12 studies on non valvular AF (3 on dabigatran, 4 on
rivaroxaban, 2 on apixaban, and 3 on edoxaban)
enrolling a total of 54 875 patients. NOACs
significantly reduced total mortality (5.61% versus
6.02%; RR, 0.89; 95% CI, 0.83– 0.96), cardiovascular
mortality (3.45% versus 3.65%; RR, 0.89; 95% CI,
0.82– 0.98), and stroke/systemic embolism (2.40%
versus 3.13%; RR, 0.77; 95% CI, 0.70–0.86). There was
a trend toward reduced major bleeding (RR, 0.86;
95% CI, 0.72–1.02) with a significant reduction of
intracranial hemorrhage (RR, 0.46; 95% CI, 0.39–
0.56). No difference in myocardial infarction was
observed.
Overall 138 948
The risk for MI/ACS was increased for dabigatran (OR
patients from 28
1.30; 95% CI 1.04-1.63; p=0.021), reduced for
RCTs) on dabigatran rivaroxaban (OR 0.78; 95% CI 0.69-0.89; p<0.001).
(4 on VTE
Apixaban demonstrated a non-significant lower
prophylaxis, 3 on
likelihood (OR 0.94; 95% CI 0.82-1.07; p=0.952).
VTE treatment, 1 on Ximelagatran showed a higher risk for MI/ACS, which
stroke prevention in was not statistically significant.
AF, 1 on ACS
treatment),
Among the RCTs for MI/ACS among the four agents,
rivaroxaban (4 on
only those pertaining to ximelagatran showed
A subgroup analysis
performed separately to
assess each novel drug,
but no statistically
significant difference
between the NOACs and
VKAs
complication and
all-cause mortality.
Control arms included
warfarin, enoxaparin,
aspirin or placebo.
VTE prophylaxis, 1
on VTE treatment, 1
on stroke
prevention in AF, 1
on ACS treatment),
apixaban (3 on VTE
prophylaxis, 2 on
stroke prevention in
AF, 2 on ACS
treatment) or
ximelagatran (2 on
VTE prophylaxis, 1
on VTE treatment, 2
on stroke
prevention in AF)
heterogeneity.
All the agents were associated with a lower all-cause
mortality, without heterogeneity among the studies.
Major bleeding complication rates varied
considerably among different agents.
Table w2 MI event rates in relation to average TTR in warfarin treated patients in recent Phase 3 trials of NOACs in atrial fibrillation
Average TTR
in warfarin
patients
MI events with comparator
n/N [%]
(rate, per 100 patient years)
MI events on warfarin
n /N [%]
(rate, per 100 patient years)
HR (95%CI)
RELY20
Main trial
64%
D110: 98/6015 (0.82)
D150: (97/6076) (0.81)
75/6022 (0.64)
1.28
(0.98-1.67)
ROCKET-AF21
Main trial (OT )
55%
101/7061
(0.91)
126/7082
(1.12)
0.81
(0.63-1.06)
ROCKET-AF
North America* (ITT)
64%
41/1339
(1.51)
36/1342
(1.31)
1.15
(0.74, 1.80)
ARISTOTLE22
62.2%
90/9120
(0.53)
102/9081
(0.61)
0.88
(0.66–1.17)
ACTIVE-W74
63.8%
36/3335
(0·86)
23/3371
(0·55)
1·58
(0·94–2·67)
ENGAGE-AF66
68.0%
High dose: 133/7035 (0.70)
Low dose: 169/7034 (0.89)
141/7036
(0.75)
High dose: 0.94 (0.74–1.19
Low dose: 1.19 (0.95–1.49)
HOKUSAI VTE trial76
63.5%
20 /4118
(0.5%)
13 /4122
(0.3%)
[NB. this is a trial of venous
thromboembolism treatment]
*http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/CardiovascularandRenalDrugsAdvisoryCommittee/UCM272005.
pdf
Table w3 – Recent studies of anticoagulation management and/or bridging therapy in patients on vitamin K antagonists undergoing cardiac device
implantation
Author
Ahmed
Year
2010
Ref.
Study design
Retrospective,
observational
Size
459 patients
undergoing
pacemaker or
defibrillatior
implantation
Ghanbari
2010
106
Retrospective,
observational
123 patients
undergoing CRT
device implantation
Tompkins
2010
108
Retrospective,
observational
2010
109
Retrospective,
observational
1,388 patients
undergoing
pacemaker or
defibrillator
implantation
518 patients
undergoing
pacemaker
implantation
Chow
107
Summary of findings
Pocket hematoma: 0.45% in the
continued warfarin group, 5.7% in the
heparin bridging group, and 1.75% in
the anticoagulation withheld group
(P=0.004 for patients on bridging
therapy vs those who continued taking
warfarin).
TIA: 0% in the continued warfarin
group, 0.8% in the heparin bridging
group, and 3.5% in the anticoagulation
withheld group (P=0.35 for patients on
bridging therapy vs those who
continued taking warfarin).
Pocket hematoma: 5.0% in the
continued warfarin group, 20.7% in the
heparin bridging group, and 4.1% in the
anticoagulation withheld group
(P=0.03).
Bleeding: 14.3% in the heparin bridging
group vs 1.6% in the anticoagulation
withheld group (P<0.0001).
Comment
Continuing warfarin therapy with a
therapeutic INR during a device
procedure is safe and, in terms of
perioperative complications, seems
superior to temporarily holding
warfarin with or without bridging
therapy.
Uninterrupted OAT does not
increase the risk of bleeding in
patients undergoing CRT device
implantation and is a safe
alternative to heparin bridging.
Periprocedural heparin significantly
increases the risk of bleeding
complications in patients undergoing
pacemaker or defibrillator
implantation.
After adjustment by multivariate
Peri-operative therapeutic
analysis, peri-operative anticoagulation anticoagulation is associated with
vs no anticoagulation was significantly
increased risk of pocket haematoma
associated with higher risk of pocket
formation in patients undergoing
hematoma formation (OR 46.51, 95% CI pacemaker implantation, particularly
15.1–143.1; P<0.001). Pocket
with the use of LMWH/UFH bridging
haematomas more commonly
therapy.
Cheng
2011
110
Randomized
100 patients
undergoing
pacemaker or
defibrillator
implantation
Li
2011
111
Retrospective,
observational
766 patients
undergoing
pacemaker or
defibrillator
implantation
Cano
2011
112
Prospective,
observational,
with historical
control
419 patients
undergoing
pacemaker or
defibrillator
implantation
Lee
2012
113
Retrospective,
observational
Birnie
2013
114
Randomized
260 patients
undergoing
pacemaker
implantation
681 patients
developed in patients who received
heparin bridging vs those who did not
(65% vs 0%; P<0.001).
Two pocket hematomas, one
pericardial effusion, one transient
ischaemic attack, and one case of
heparin-induced thrombocytopenia
were observed among patients
randomized to warfarin interruption vs
those continuing warfarin (P=0.056).
Systemic bleeding or pocket
hematoma: 3.7% in the continued
warfarin group, 7.0% in the heparin
bridging group, and 2.1% in the
anticoagulation withheld group
(P=0.029). INR of 2.0-2.5 at time of
procedure did not increase bleeding
risk compared with INR less than 1.5
(3.7% vs 3.4%; P = 0.72), but INR
greater than 2.5 increased the bleeding
risk (10.0% vs 3.4%; P = 0.029).
Pocket hematoma: 2.3% in the
continued warfarin group vs 17.7% in
patients at moderate-to-high
thromboembolic risk undergoing
heparin bridging (P=0.001); 0% in the
anticoagulation withheld group vs 13%
in patients at low thromboembolic risk
undergoing heparin bridging
(P<0.0001).
Significant bleeding complications
occurred in 3.1% of patients, who were
all receiving heparin bridging during the
procedure (p<0.0001).
Pocket hematoma: 3.5% in the
There was a trend toward reduced
complications in patients undergoing
pacemaker or defibrillator
implantation who continued OAT
with warfarin.
Uninterrupted OAT with INR <2.5
does not increase the risk of
bleeding in patients undergoing
device implantation.
A standardized protocol, classifying
patients according to their
thromboembolic risk, resulted in a
significant reduction in bleeding
complications.
Patients undergoing pacemaker
implantation who receive heparin
bridging are at high risk for
significant bleeding complications.
Uninterrupted OAT markedly
undergoing
pacemaker or
defibrillator
implantation
continued warfarin group, 16.0% in the reduces the incidence of pocket
heparin bridging group (P<0.001).
hematoma compared with heparin
Major surgical and thromboembolic
bridging in patients undergoing
complications were rare and did not
pacemaker or defibrillator
differ significantly between the study
implantation.
groups.
115
Chen
2013
Retrospective, 1,093 patients
After multivariate adjustment, heparin Heparin bridging significantly
observational
undergoing
bridging vs no bridging was significantly increases the risk of pocket
pacemaker or
associated with a higher risk of pocket
hematoma complications in patients
defibrillator
hematoma formation (OR 6.04, 95% CI undergoing pacemaker or
implantation
2.43-15.01; P<0.001).
defibrillator implantation.
CI = confidence interval; CRT = cardiac resynchronization therapy; INR = international normalized ratio; LMWH = low molecular weight heparin; MACCE =
major adverse cardiac and cerebrovascular events; OAT = oral anticoagulation therapy; OR = odds ratio; UFH = unfractionated heparin.
Table w4 – Recent studies of anticoagulation management and/or bridging therapy in patients on vitamin K antagonists undergoing surgical or invasive
procedures other than cardiac device implantation
Author
McBane
Year
2010
Ref.
116
Study design
Retrospective,
observational
Size
775 patients with VTE
undergoing invasive
procedures
Steib
2010
117
Randomized
94 patients
undergoing surgery
Billett
2010
118
Retrospective,
observational
Gerson
2010
119
Prospective,
observational
Summary of findings
The 3-month cumulative incidence of
thromboembolism (1.8%), major
hemorrhage (1.8%), and mortality
(1.7%) were low and did not differ
between patients who received heparin
bridging and those who did not.
INR was significantly higher in patients
on VKA treatment until day -2 before
surgery, followed by 1 mg of oral
vitamin K on the day before surgery
compared with patients randomized to
heparin bridging on day -5 before
surgery. An INR ≤1.5 was not achieved
in 66% of patients who received
vitamin K.
4,797 patients with AF After multivariate adjustment, heparin
undergoing invasive
bridging vs no bridging was significantly
procedures
associated with a reduced 30-day risk
of death (OR 0.07, 95% CI 0.01-0.47;
P=0.007) and VTE (OR 0.24, 95% 0.090.64; P=0.005). No strokes occurred in
each group. No difference in the risk of
bleeding was noted with heparin
bridging (OR 0.88, 95% CI 0.31-2.47;
P=0.81).
483 patients
Major hemorrhagic events: 1% in
undergoing
patients who received bridging therapy
endoscopic
vs 1% in those who did not (P = 0.10).
procedures
Comment
Thromboembolism, bleeding, and
death among VTE patients in whom
anticoagulation is temporarily
interrupted for an invasive
procedure is low and does not seem
to be significantly impacted by
heparin bridging.
Oral vitamin K (1 mg) should not be
a substitute for heparin bridging
before surgery.
LMWH bridging in patients with AF
may be associated with lower risks
of VTE and death within 30 days of
discharge.
Hemorrhagic events are not
increased in anticoagulated patients
managed according to the bleeding
risk of different endoscopic
procedures.
Jaffer
2010
120
Prospective,
observational
492 patients
undergoing surgery
Di Biase
2010
121
Retrospective,
observational
6454 patients
undergoing AF
ablation
Hakalahti
2011
122
Retrospective,
observational
193 patients
undergoing AF
ablation
Smoyer-Tomic
2012
123
Retrospective,
observational
6,340 patients
hospitalized with nonvalvular AF
Lahtela
2012
36
Prospective,
observational
451 patients with AF
undergoing coronary
stenting
After adjustment by multivariate
analysis, full dose UFH/LMWH was
identified as a significant predictor of
30-day major bleeding (OR 4.4, 95% CI
1.5-14.7; P<0.05)
Major bleeding: 0.4% in the continued
warfarin group, and 0.4% and 0.8% in
the anticoagulation withheld groups.
Pericardial effusion: 0.5% in the
continued warfarin group, and 0.4%
and 0.8% in the anticoagulation
withheld groups. Periprocedural stroke
or TIA rate: 0% in the continued
warfarin group, and 1.1% and 0.9% in
the anticoagulation withheld groups.
Major bleeding: 0.8% in the continued
warfarin group, 1.0% in the
anticoagulation withheld group
(P=1.00).
Minor bleeding: 11.2% in the continued
warfarin group, 10.7% in the
anticoagulation withheld group
(P=0.90).
Mean length of stay: 6.3±5.9 days for
hospitalized non-valvular AF patients
who received bridging therapy vs
4.2±3.1 days for those who did not; on
multivariate analysis, patients who
received bridging therapy had a 18.2%
higher length of stay than those who
remained on warfarin (P=0.0017).
After adjustment by propensity score,
bridging therapy vs. uninterrupted OAT
was not significantly associated with
bleeding complications (OR, 1.38; 95%
The risk of major bleeding in OAT
patients undergoing surgery is
strongly associated with the use of
postoperative therapeutic doses of
UFH/LMWH
Periprocedural therapeutic
anticoagulation with warfarin
reduced the risk of periprocedural
stroke but not increased the risk of
major bleeding or pericardial
effusion.
Uninterrupted OAT provides a
feasible and safe alternative to
bridging therapy for patients
undergoing AF ablation.
Two-thirds of non-valvular AF
patients receiving warfarin while
hospitalized also received in-patient
bridging therapy. LMWH/PS was the
most common bridging agent,
followed by UFH. Length of stay was
shown to be longer for bridged than
non-bridged patients.
Uninterrupted OAT does not
increase the risk of bleeding or
perioperative thrombotic
complications during coronary
CI: 0.77-2.48, P=0.28) or MACCE (OR,
stenting compared with
1.16; 95%CI: 0.44-3.05, P=0.76).
conventional heparin bridging.
124
Pini
2012
Retrospective, 502 patients
Patients in the continued OAT group
Uninterrupted OAT may be a safe
observational
undergoing carotid
(N=20) did not experience any adverse and effective option in patients
artery stenting
event in term of strokes, death,
undergoing carotid artery stenting.
myocardial infarction or
hemorrhage/hematoma.
125
Tafur
2012
Prospective,
2,182 patients
Three-month major bleeding occurred
Premature heparin re-initiation after
observational
undergoing invasive
more frequently in patients receiving
an invasive procedure is
procedures
bridging therapy (3% vs. 1%; P = 0.017). independently associated with major
Independent predictors of major
bleeding in patients on OAT
bleeding included re-initiation of
undergoing invasive procedures.
heparin therapy within 24 h after the
procedure (HR 1.9; 95% CI 1.1-3.4;
P<0.05).
AF = atrial fibrillation; CI = confidence interval; CRT = cardiac resynchronization therapy; HR = hazard ratio; INR = international normalized ratio; LMWH = low
molecular weight heparin; MACCE = major adverse cardiac and cerebrovascular events; OAT = oral anticoagulation therapy; PS = pentasaccharide; OR = odds
ratio; UFH = unfractionated heparin; VTE = venous thromboembolism.