YAJEM-56392; No of Pages 6

American Journal of Emergency Medicine xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

journal homepage: www.elsevier.com/locate/ajem

Anticoagulants in ischemia-guided management of non-ST-elevation acute

coronary syndromes
Martin Mayer
East Carolina University, 4310N Health Sciences Building, Greenville, NC 27858-4353, United States
Vidant Medical Center, 2100 Stantonsburg Rd, Greenville, NC 27834, United States

a r t i c l e i n f o a b s t r a c t

Article history: Background: The most recent joint guidelines from the American Heart Association (AHA) and American College
Received 17 October 2016 of Cardiology (ACC) on the management of non-ST-elevation acute coronary syndromes (NSTE-ACS) are a result
Received in revised form 26 November 2016 of a substantial and considered undertaking, and those involved deserve much recognition for their efforts. How-
Accepted 28 December 2016 ever, the handling of anticoagulants seems somewhat inadequate, and this is a highly-relevant matter when
Available online xxxx
managing NSTE-ACS.
Objective of the review: Among areas of potential uncertainty, emergency medicine professionals might still be left
Keywords:
NSTE-ACS
wondering about the particulars of anticoagulant therapy when pursuing ischemia-guided management of
Non-ST-elevation acute coronary syndromes NSTE-ACS (that is, managing NSTE-ACS without an intent for early invasive measures, such as coronary angiog-
Unstable angina raphy and revascularization). This review seeks to provide insight into this question.
Non-ST-elevation myocardial infarction Discussion: Relevant clinical trials are appraised and translated into clinical context for emergency medicine pro-
Anticoagulant fessionals, including the implications of noteworthy advancements in the management of NSTE-ACS.
Conclusions: Although current guidelines from the AHA and ACC suggest enoxaparin has better evidence than other
anticoagulants in the setting of NSTE-ACS management, careful review of the evidence shows this is not actually
clearly supported by the available evidence in the era of contemporary management. Unless and until better con-
temporary data emerge, emergency medicine professionals must carefully weigh the available evidence, its limita-
tions, and the possible clinical implications of the various anticoagulant options when managing NSTE-ACS.
© 2017 Elsevier Inc. All rights reserved.

Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2. Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.1. Enoxaparin versus UFH: seeking potentially-relevant trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.2. Enoxaparin versus UFH: ESSENCE and TIMI 11B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.3. ESSENCE and TIMI 11B: issues with generalizability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.4. Fondaparinux versus enoxaparin: OASIS-5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
2.5. Heparins versus placebo anticoagulant or no anticoagulant: the Cochrane review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
3. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
Sources of support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
Prior presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

1. Introduction management of non-ST-elevation acute coronary syndromes (NSTE-

ACS) [1,2] are a result of a substantial and considered undertaking,
The most recent joint guidelines from the American Heart Associa- and those involved deserve much recognition for their efforts. However,
tion (AHA) and American College of Cardiology (ACC) on the the handling of anticoagulants seems somewhat inadequate, and this is

E-mail address: mayerm@ecu.edu.

http://dx.doi.org/10.1016/j.ajem.2016.12.070
0735-6757/© 2017 Elsevier Inc. All rights reserved.

Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070
2 M. Mayer / American Journal of Emergency Medicine xxx (2017) xxx–xxx
a highly-relevant matter when managing NSTE-ACS. A recent publica-
tion addressed several questions pertaining to anticoagulation in
NSTE-ACS [3], but even with this article, clinicians might still be left
wondering about the particulars of anticoagulant therapy when manag-
ing NSTE-ACS without an intent for early invasive measures (i.e., coro-
nary angiography and revascularization). Such an approach is
currently being called “ischemia-guided,” [1,2] but other common
terms for this strategy include “conservative,” “noninvasive,” or “medi-
cal” management. Bivalirudin has not been studied in this setting and it
will thus not be considered further. As for the other anticoagulant op-
tions, clinical experience suggests the primary choice for most clinicians
is whether they will use unfractionated heparin (UFH) or the low-mo-
lecular-weight heparin (LMWH) enoxaparin, and thus this will be con-
sidered in detail. Another option, however, is fondaparinux. The AHA
and ACC suggest enoxaparin has a higher level of evidence than both
UFH and fondaparinux [1,2], but careful consideration of the best-avail-
able evidence reveals a far more nuanced picture.
2. Results and discussion
2.1. Enoxaparin versus UFH: seeking potentially-relevant trials
Of the six randomized controlled trials (RCTs) that have investigated
the use of enoxaparin versus UFH in NSTE-ACS, two seem most applica-
ble to the situation of NSTE-ACS being managed without an early inva-
sive strategy: ESSENCE [4] and TIMI 11B [5]. To briefly address the other
RCTs: ACUTE II [6] and INTERACT [7] were relatively small (525 and 746
patients, respectively); ACUTE II was designed with primary intent to
evaluate differences in bleeding, but this assessment was hindered by
lower-than-anticipated bleeding rates; INTERACT sought to assess
both noninferiority in bleeding and superiority in efficacy as assessed
by ischemic changes on electrocardiogram (it was not adequately de-
signed to provide robust data for hard clinical efficacy outcomes); A to
Z used a nonstandard definition for early invasive methods (cardiac
catheterization within 108 h) [1-3,8], with 55.2% of investigators in
the enoxaparin arm and 55.4% of investigators in the UFH arm declaring
intent for an early invasive strategy and 42.5% of patients in the
enoxaparin arm and 43.8% of patients in the UFH arm having coronary
angiography or percutaneous coronary intervention (PCI) within 48 h
(data are not provided for a 24-hour cutoff); ACUTE II, INTERACT, and
A to Z all involved all patients receiving a glycoprotein IIb/IIIa inhibitor
as part of their management; ACUTE II, INTERACT, and A to Z were con-
ducted before widespread use of P2Y12 antagonists (e.g., INTERACT re-
ported 15% utilization of P2Y12 antagonists, specifically clopidogrel,
and A to Z did not even record use of P2Y12 antagonists); and SYNERGY
[1-3,9] was conducted specifically in the setting of high-risk patients for
whom early invasive management (consistent with current standards)
was intended, with 92.1% of enoxaparin recipients and 92.0% of UFH re-
cipients receiving coronary angiography in a median of 21.7 h (inter-
quartile range [IQR], 6.3 to 43.6 h) and 21.5 h (IQR, 6.3 to 42.6 h),
respectively. These elements all obstruct one's ability to use these trials
to elucidate the present clinical question. What, then, of ESSENCE and
TIMI 11B?
2.2. Enoxaparin versus UFH: ESSENCE and TIMI 11B
ESSENCE and TIMI 11B both studied enoxaparin versus UFH in NSTE-
ACS when the initial management strategy did not include early inva-
sive management; in fact, the protocols functionally [4] or explicitly
[5] excluded such patients [10]. ESSENCE randomized 3171 patients
with NSTE-ACS to receive a mandated minimum of 48 h of anticoagu-
lant therapy (either IV infusion of UFH and subcutaneous injection of
placebo or IV infusion of placebo and subcutaneous injection of
enoxaparin) and allowed for up to eight days of anticoagulant therapy;
discontinuation occurred at hospital discharge, a new myocardial in-
farction (MI), a revascularization procedure, or death. Approximately
Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070
2.1% (66/3171) of the patients in ESSENCE had their anticoagulant
discontinued prior to the specified 48-hour minimum duration of ad-
ministration due to needing a procedure. TIMI 11B randomized 3910
patients to either: [1] IV infusion of UFH or matching placebo for at
least 3 days and a maximum of 8 days (with discontinuation after day
3 and prior to day 8 occurring at the discretion of the treating physi-
cian), and [2] subcutaneous injection of enoxaparin or matching placebo
for 8 days or until hospital discharge (whichever came first). TIMI 11B
also gave an initial 30 mg IV bolus of enoxaparin to those allocated to
the enoxaparin arm. The 30 mg IV bolus is mentioned by the guidelines
as being used “in selected patients”,1(e365,e367),2(e162,e164,e165) but evi-
dence does not support its use [3]. The median duration of both thera-
pies in ESSENCE was 2.6 days; in TIMI 11B, the median duration of
UFH was 3.0 days (IQR, 2.99 to 3.98 days) and the median duration of
therapy for enoxaparin was 4.6 days (IQR, 2.97 to 6.59 days).
Both trials found benefit for their primary composite efficacy end-
point, which was comprised of death [4,5], nonfatal MI [4,5], and recur-
rent angina [4] or urgent revascularization [5]. The data from each trial
considered individually [4,5] might understandably lead one to propose
the benefit was most clearly for reducing the incidence of recurrent an-
gina or urgent revascularization. Indeed, appraising the components of a
composite outcome in this manner is appropriate and important
[11-14], and while these trials were not intended to provide definitive
evidence for the components of the composite, they still provide useful
information for such by virtue of the effect estimates and confidence in-
tervals provided [15-17]. Judicious meta-analysis of these two trials of-
fers at least some degree of additional insight into the individual
components. Meta-analyses of the trials [10,18] (including one that ex-
amined follow-up data at one year [18]) standardized the third compo-
nent of the tripartite composite to be urgent revascularization based on
available data in the ESSENCE and TIMI 11B databases. These meta-anal-
yses also found benefit for enoxaparin compared to UFH with respect to
the composite efficacy endpoint (see Table 1). Additionally, these meta-
analyses suggest possible early benefit for the composite of death and
nonfatal MI [10] that might be attenuated by one year [18], with death
seeming to contribute less to the composite than nonfatal MI (though
data for nonfatal MI as a solitary outcome are not provided; see Table 1).
The primary safety outcome in ESSENCE and TIMI 11B was hemor-
rhage. Hemorrhage was divided into major and minor hemorrhage,
which were defined identically in the two trials [4,5]. Major hemorrhage
included bleeding resulting in death; transfusion of at least 2 units of
blood; a fall in hemoglobin of at least 3 g/dL; or a retroperitoneal, intra-
cranial, or intraocular hemorrhage. Minor hemorrhage was any clinical-
ly important bleeding that did not qualify as major (e.g., epistaxis,
ecchymosis or hematoma, or macroscopic hematuria). Minor hemor-
rhage was increased in the enoxaparin arm of both trials (see Table 1)
[4,5,10], with injection site ecchymosis being qualitatively reported as
the most common event (ESSENCE reported injection-site ecchymosis
as the most frequent event, whereas TIMI 11B actually listed two events
as comprising the majority of events: injection-site ecchymosis and he-
matoma at the site of a sheath insertion for those who did eventually go
on to have catheterization) [4,5].
The balance of these results might reasonably be seen as supporting
the use of enoxaparin over UFH when the intent is to manage NSTE-ACS
without early invasive measures. A key question remains, however: Are
these data applicable to current practices?
2.3. ESSENCE and TIMI 11B: issues with generalizability
Unfortunately, one cannot make strong knowledge statements
about the above question. Like some of the aforementioned trials, ES-
SENCE and TIMI 11B were both conducted prior to the widespread use
of P2Y12 antagonists, and indeed, this presents a large problem in apply-
ing these results to contemporary care. In other words, P2Y12 antagonist
therapy is now a well-established component of managing acute coro-
nary syndromes (ACS, a term encompassing NSTE-ACS and ST-elevation
 M. Mayer / American Journal of Emergency Medicine xxx (2017) xxx–xxx 3

Table 1 management [23-26]. Although many may think of statin therapy as

Effectsa of enoxaparin versus UFH on efficacy and safety outcomes in NSTE-ACS based on having benefits over the long run (which is accurate), the MIRACL trial
data from ESSENCE and TIMI 11B. [4,5,10,18]
[24] also showed potential short-term benefit of early institution of
Outcome REE (95% CI) NNT (95% CI)b high-intensity statin therapy in ACS (specifically, atorvastatin 80 mg
Death, nfMI, and UR, day 2c OR 0.77 (0.63 to 0.94) NNTB 73 (45 to 282) compared to placebo). The limited generalizability of ESSENCE and
Death, nfMI, and UR, day 8c OR 0.79 (0.69 to 0.91) NNTB 40 (27 to 94) TIMI 11B can be thought of thusly: The efficacy benefits once seemingly
Death, nfMI, and UR, day 14c OR 0.79 (0.69 to 0.90) NNTB 35 (24 to 75) offered by enoxaparin compared to UFH when managing NSTE-ACS
Death, nfMI, and UR, day 43c OR 0.80 (0.71 to 0.91) NNTB 32 (22 to 72)
without early invasive measures might well be partly or completely
Death, nfMI, and UR, 1 yeard HR 0.88 (0.80 to 0.97) NNTB 37 (22 to 150)
Death and nfMI, day 2c OR 0.80 (0.55 to 1.16) NNTB 282 (NNTH 354 eclipsed by the aforementioned evolutions in management of NSTE-
to ∞ to NNTB 125) ACS. Put differently, to the extent these evolutions in management
Death and nfMI, day 8c OR 0.77 (0.62 to 0.95) NNTB 86 (52 to 398) lower a person's risk for adverse non-hemorrhagic outcomes, it follows
Death and nfMI, day 14c OR 0.79 (0.65 to 0.96) NNTB 78 (46 to 411) that any benefits that might otherwise be offered by enoxaparin be-
Death and nfMI, day 43c OR 0.82 (0.69 to 0.97) NNTB 70 (40 to 423)
Death and nfMI, 1 yeard HR 0.92 (0.81 to 1.04) NNTB 98 (NNTH 197
come smaller and smaller. Likewise, to the extent P2Y12 antagonist ther-
to ∞ to NNTB 41) apy increases bleeding risk, this could tip the risk-benefit balance even
Death, day 2c OR 1.40 (0.68 to 2.90) NNTH 836 (NNTB 1044 further from the side of benefit with respect to using enoxaparin (as op-
to ∞ to NNTH 176) posed to UFH) as the anticoagulant when managing NSTE-ACS without
Death, day 8c OR 0.80 (0.56 to 1.16) NNTB 282 (NNTH 354
early invasive measures.
to ∞ to NNTB 128)
Death, day 14c OR 0.81 (0.59 to 1.12) NNTB 224 (NNTH 356 In light of the above considerations about efficacy and safety in the
to ∞ to NNTB 104) current era of NSTE-ACS management, the SYNERGY trial [9] of patients
c
Death, day 43 OR 0.84 (0.66 to 1.08) NNTB 166 (NNTH 334 with NSTE-ACS randomized to either enoxaparin (n = 4993) or UFH
to ∞ to NNTB 78) (n = 4985) warrants at least some additional consideration. First, it
Death, 1 yeard HR 0.90 NNTB 140 (NNTH 176
must be emphasized that applicability to the current question is
(0.75 to 1.08) to ∞ to NNTB 56)
Minor hemorrhage, day 8c OR 2.38 NNTH 18 (14 to 25) admittedly limited, perhaps most notably due to the difference in
(1.98 to 2.85) management strategy: as aforementioned, SYNERGY was conducted spe-
Minor hemorrhage, day 30e RR 1.65 NNTH 21 (14 to 37) cifically in the setting of high-risk patients for whom an early invasive
(1.32 to 2.07)
c
strategy was intended, with 92.1% of enoxaparin recipients and 92.0% of
Major hemorrhage, day 8 OR 1.23 NNTH 400 (NNTB 459
(0.80 to 1.89) to ∞ to NNTH 104)
UFH recipients receiving coronary angiography in a median of 21.7 h
Major hemorrhage, day 30e RR 0.93 NNTB 202 (NNTH 78 (IQR, 6.3 to 43.6 h) and 21.5 h (IQR, 6.3 to 42.6 h), respectively [1-3,9].
(0.71 to 1.21) to ∞ to NNTB 44) With that caveat, the rate of P2Y12 antagonist use in SYNERGY more close-
CI, confidence interval; HR, hazard ratio; nfMI, nonfatal myocardial infarction; NNT, num- ly matches current practice, albeit perhaps imperfectly: 3.5% of both the
ber needed to treat; NNTB, number needed to treat to benefit one person; NNTH, number enoxaparin and UFH groups received ticlopidine, and approximately
needed to treat to harm one person; OR, odds ratio; REE, relative effect estimate; RR, rel- 62.5% and 63.3% of the enoxaparin and UFH groups received clopidogrel.
ative risk; UR, urgent revascularization.
a It is therefore worth noting enoxaparin does not offer any clear benefit
Only certain endpoint effect estimates are provided, and the reader is encouraged to
consult the text of this article and the corresponding cited publications to appreciate more over UFH in the setting of early invasive intent with use of a P2Y12 antag-
completely all the details of these studies. As the title of the table implies, all effect esti- onist for the primary composite outcome of all-cause mortality or non-
mates listed are for those of enoxaparin compared to UFH. fatal MI at 30 days (14.0% versus 14.5%, respectively, if rounding to one
b
With the exception of major hemorrhage at day 30, the NNT estimates are derived decimal; hazard ratio 0.96; 95% confidence interval [CI], 0.86 to 1.06;
from standard meta-analytic calculations using a patient expected event rate (PEER) and
the corresponding REE. For all calculations in the table, the PEER is calculated as the event
p = 0.40). Assessment of the primary composite outcome at multiple
rate among the participants receiving UFH, and the PEER is not rounded prior being used other time points – including 48 h, 14 days, 6 months, and 12 months –
in conjunction with the corresponding REE to calculate NNT estimates. Consistent with revealed similar results [3,27]. To assess safety, SYNERGY measured
convention, all NNTHs are rounded down to the nearest whole number, whereas all rates of TIMI (Thrombolysis in Myocardial Infarction) major bleeding,
NNTBs are rounded up to the nearest whole number.
c
GUSTO (Global Utilization of Streptokinase and t-PA for Occluded Arter-
Data for outcomes at day 2, 8, 14, and 43 are based on the ESSENCE and TIMI 11B
meta-analysis results (using the Peto method) published in Circulation [10]. Caution is re- ies) severe bleeding, and transfusion. Enoxaparin caused an approximate
quired for interpreting findings past day 8, as TIMI 11B included a trial extension that in- 1.5% absolute risk increase in TIMI major bleeding compared to UFH (9.1%
volved continuing the enoxaparin in the outpatient setting for patients who were versus 7.6% if rounding to one decimal; no relative effect estimate [REE] or
originally randomized to enoxaparin and eligible for the extension after the acute phase. CI data reported; number needed to treat to harm [NNTH], 66 [68 if not
This failed to show any benefit for extending enoxaparin therapy, and indeed, it only
rounding percentages to one decimal]; p = 0.008), whereas rates were
showed potential harm in the form of excess bleeding (which is not terribly surprising).
Heterogeneity analysis revealed no important heterogeneity in the outcomes other than more comparable for GUSTO severe bleeding (2.7% versus 2.2%; no REE
minor hemorrhage (χ2 = 6.99; p = 0.008; I2 not reported). or CI data reported; p = 0.08) and transfusion (17% versus 16%; no REE
d
Data for outcomes at one year are based on the meta-analysis (using a Cox proportional or CI data reported; p = 0.16).
hazards model) of one-year follow-up data from ESSENCE and TIMI 11B published in Europe-
But does any of this necessarily mean UFH is the preferred anticoagu-
an Heart Journal [18]. Caution is required for interpreting findings past day 8, as TIMI 11B in-
cluded a trial extension that involved continuing the enoxaparin in the outpatient setting for
lant in this setting? One can put forward considerations like those seen
patients who were originally randomized to enoxaparin and eligible for the extension after here, but the available evidence still does not provide one with robust, un-
the acute phase. This failed to show any benefit for extending enoxaparin therapy, and in- equivocal evidence favoring UFH, either. The main point here is that the
deed, it only showed potential harm in the form of excess bleeding (which is not terribly sur- concept of generalizability is a basic component of evidence-based prac-
prising). Heterogeneity analysis revealed no important heterogeneity in the outcomes.
e tice, and the ESSENCE and TIMI 11B data have noteworthy limitations in
Data for major hemorrhage at day 30 are based on data from ESSENCE [4] (comparable
data from TIMI 11B are not available). The 30-day data for ESSENCE include absolute event generalizability when compared to contemporary practice. Of course,
rates only, so all other parts of the row are based on standard calculations for the listed there may be a case where these data become more applicable (e.g., a pa-
metrics. tient who does not receive a P2Y12 antagonist for some reason), but such
cases are not the rule.
MI) [19-22], and thus, ESSENCE and TIMI 11B have limited generaliz-
ability to current practice. Although the now-widespread use of P2Y12 2.4. Fondaparinux versus enoxaparin: OASIS-5
antagonists is perhaps the single biggest impediment to the applicabil-
ity of ESSENCE and TIMI 11B to contemporary practice, matters are not Fondaparinux has not been compared to UFH in a head-to-head man-
made any better by the fact that statin therapy – especially high-inten- ner in NSTE-ACS, but OASIS-5 [28] compared fondaparinux to enoxaparin
sity statin therapy – is now another well-established part of in 20,708 patients with NSTE-ACS (approximately 2.9 times the number

Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070
4 M. Mayer / American Journal of Emergency Medicine xxx (2017) xxx–xxx
of patients in ESSENCE and TIMI 11B combined). Revascularization oc-
curred at the discretion of participating clinicians. Although the utilization
of P2Y12 antagonists (clopidogrel or ticlopidine use in 67.2% and 67.6% of
the enoxaparin and fondaparinux groups, respectively) and high-intensi-
ty statins (use of a “lipid-lowering agent”28(p1467) in 78.4% and 79.4% of
the enoxaparin and fondaparinux groups, respectively) might still leave
something to be desired, utilization was clearly better than that seen in
ESSENCE and TIMI 11B. Indeed, the utilization in OASIS-5 seems roughly
comparable to that of SYNERGY (though SYNERGY provides utilization
data for ticlopidine, clopidogrel, statins, and “other”9(p49) lipid-lowering
agents separately, whereas OASIS-5 only provides utilization data as just
described). In OASIS-5, using fondaparinux instead of enoxaparin did
not result in any significant decrement in efficacy outcomes,
fondaparinux offered a bleeding advantage over enoxaparin, and this pat-
tern was maintained when results were analyzed according to whether
revascularization occurred within nine days of randomization. In sum,
the OASIS-5 data suggest fondaparinux would be worth considering in-
stead of enoxaparin in NSTE-ACS if one were, for whatever reason, leaning
toward using enoxaparin instead of UFH [3,28]. (Again, comparisons be-
tween UFH and fondaparinux are made difficult by the lack of adequate
head-to-head data.) If one does use fondaparinux and the patient ends
up being switched to an invasive strategy with PCI, the patient should re-
ceive additional anticoagulant (e.g., UFH) in the standard fashion to pre-
vent catheter thrombosis [1-3].
Table 2
Effects of heparin (UFH or LMWH) versus placebo anticoagulant or no anticoagulant on ef-
ficacy and safety outcomes in NSTE-ACS based on data from the Cochrane SRMA.a [29]
Outcome RR (95% CI) NNT (95% CI)b
c
Death 0.84 (0.36 to NNTB 675 (NNTH 110 to ∞ to NNTB 169)
1.98)
MI d
0.40 (0.25 to NNTB 35 (28 to 57)
0.63)
Major 2.05 (0.91 to NNTH 205 (NNTB 2399 to ∞ to NNTH 59)
hemorrhagee 4.60)
CI, confidence interval; MI, myocardial infarction; NNT, number needed to treat; NNTB,
number needed to treat to benefit one person; NNTH, number needed to treat to harm
one person; RR, risk ratio.
a (Grading of Recommendations Assessment, Development, and Evalua-
Of 3118 patients included in the SRMA, 142 received UFH and warfarin concurrently.
Only the primary outcomes are shown in the table. The reader is encouraged to consult the
text of this article and the Cochrane SRMA to appreciate more completely all the details of
the SRMA and the included studies.
b
The NNT estimates are derived from standard meta-analytic calculations using a pa-
tient expected event rate (PEER) and the corresponding RR. For all calculations in the table,
the PEER is calculated as the event rate among the participants receiving placebo anticoag-
ulant or no anticoagulant, and the PEER is not rounded prior to being used in conjunction
with the corresponding RR to calculate NNT estimates. Consistent with convention, all
NNTHs are rounded down to the nearest whole number, whereas all NNTBs are rounded
up to the nearest whole number.
c trial publication dates span from 1988 to 2000, and even the trial pub-
The RR for death is based on a fixed-effects model meta-analysis involving 9 events
among those receiving a heparin (n = 1238) and 11 events among those receiving place-
bo anticoagulant or no anticoagulant (n = 1188). The total number of participants in this
analysis does not total 3118; this is due to the fact that not all study data are useable for all
analyses. Analyses of heterogeneity and of heparin type subgroups did not suggest impor-
tant heterogeneity (p = 0.82; I2 = 0%) or subgroup differences (p = 0.82; I2 = 0%).
d
The RR for MI is based on a fixed-effects model meta-analysis involving 24 events
among those receiving a heparin (n = 1238) and 57 events among those receiving place-
bo anticoagulant or no anticoagulant (n = 1188). The total number of participants in this
analysis does not total 3118; this is due to the fact that not all study data are useable for all
analyses. Analyses of heterogeneity and of heparin type subgroups did not suggest impor-
tant heterogeneity (p = 0.63; I2 = 0%) or subgroup differences (p = 0.30; I2 = 18%).
e
The RR for major hemorrhage (per the Cochrane SRMA “e.g. fall in hemoglobin level of
N2 g/dL, requires transfusion, is intracranial, retroperitoneal, or intraocular, or causes
death or cessation of the study treatment”29(p7)) is based on a fixed-effects model meta-
analysis involving 17 events among those receiving a heparin (n = 1607) and 7 events
among those receiving placebo anticoagulant or no anticoagulant (n = 1511). All included
study data were usable for this analysis. Analyses of heterogeneity and of heparin type
subgroups did not suggest important heterogeneity (p = 0.93; I2 = 0%) or subgroup dif-
ferences (p = 0.64; I2 = 0%).
Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070
2.5. Heparins versus placebo anticoagulant or no anticoagulant: the
Cochrane review
It is also relevant to consider what benefit – if any – has been demon-
strated by UFH or LMWH over placebo anticoagulant or no anticoagulant
in the setting of NSTE-ACS. The Cochrane systematic review and meta-
analysis (SRMA) on the matter [29] provides a detailed examination of
this question, and the findings for the primary outcomes (death, MI, and
major hemorrhage) are shown in Table 2. Although the best-available
data suggest UFH or LMWH reduces the risk of MI compared to placebo
anticoagulant or no anticoagulant, it must be made clear the best-avail-
able data suffer from considerable weaknesses. While the data for death
and major hemorrhage do not show a statistically significant difference,
one must be careful to avoid oversimplifying and obfuscating research
findings by unduly focusing on the dichotomization of findings as statisti-
cally significant or nonsignificant. For instance, the CI provides consider-
able information beyond whether it includes the possibility of no effect
(which is functionally no more than a surrogate for seeing if the p value
reaches the threshold for statistical significance). As evidenced by the
width of the CI for death, the data for death are notably imprecise: the
data cannot statistically exclude anything from a relative risk reduction
(RRR) of 64% to a relative risk increase (RRI) of 98% (see Table 2 for
NNT estimates and further information). This imprecision introduces con-
siderable additional uncertainty. The analysis for major hemorrhage
shows the data cannot statistically exclude anything from a 9% RRR to a
360% RRI (see Table 2 for NNT estimates and further information), but
there is no reason to believe the heparins would actually reduce major
hemorrhage by any amount.
Although a full analysis of the SRMA or the included trials is beyond
the scope of this review, one notes the evidence base suffers from note-
worthy weaknesses. The SRMA includes a total of 3118 patients from
eight trials, with a trial of dalteparin versus placebo published in 1996
accounting for 1506 patients (48.3% of the patient data). UFH was ad-
ministered to 794 patients (142 of whom received warfarin concurrent-
ly), 814 patients received LMWH (enoxaparin is not represented; 746
received dalteparin and 68 received nadroparin), 1510 patients re-
ceived placebo anticoagulant or no anticoagulant, and a total of 20
deaths, 81 MIs, and 24 major hemorrhages occurred (with the smaller
number of deaths and major hemorrhages obviously contributing to
the aforementioned width of the CIs for these outcomes). By GRADE
tion) standards, one of the clinically-relevant outcomes is considered
to have very low-quality evidence (recurrent angina), with all others
having low-quality evidence (including the primary outcomes of
death, MI, and major hemorrhage). Only one trial could be considered
at low risk of bias: a trial of UFH versus placebo published in 1988 and
contributing only 243 patients to the aforementioned total. The remain-
ing seven trials all have two or more areas where risk of bias was high or
could not be assessed due to inadequate reporting. Furthermore, the
lished in 2000 only contributed 191 patients. As aforementioned, the
management of NSTE-ACS has evolved from the time when these trials
were published, and the generalizability of all these trials is therefore
limited in the same manner as described for ESSENCE and TIMI 11B.
The evidence base for the heparins versus placebo is, in a word, dis-
appointing. Considering the pathophysiology of NSTE-ACS and the
pharmacology of the heparins, the prior probability of heparins being
beneficial in NSTE-ACS is certainly noteworthy, but prior probability
by itself is insufficient, and the available clinical trial data leave much
to be desired. In this light, high-quality, placebo-controlled RCT evi-
dence would be ideal, but given the endorsement [1,2] and widespread
use of LMWH (specifically, enoxaparin) and UFH in the management of
NSTE-ACS, it is rather unclear whether the medical community will ever
see such evidence (despite the weaknesses in present evidence contrib-
uting to clinical equipoise, which serves as a fundamental justification
for further study).
 M. Mayer / American Journal of Emergency Medicine xxx (2017) xxx–xxx
3. Conclusions
What are clinicians to do in the setting of this uncertainty? Evidence-
based medicine does not exist in a vacuum, and one must try to make the
best of the best-available evidence. As described in this review,
anticoagulation is considered part of the standard of care of NSTE-ACS,
and the recent AHA and ACC guidelines [1,2] suggest enoxaparin has bet-
ter evidence than other anticoagulants with respect to managing NSTE-
ACS. However, careful appraisal of the evidence reveals a far more nu-
anced picture: there are noteworthy weaknesses in the evidence for the
heparins versus placebo anticoagulant or no anticoagulant, and given
how other aspects of pharmacotherapeutic management have evolved
since ESSENCE and TIMI 11B, there is considerable uncertainty regarding
what differences might exist between enoxaparin and UFH in the man-
agement of NSTE-ACS when there is no intent for early invasive measures.
Nevertheless, one must still attempt to weigh benefits versus risks. The
benefits seen for enoxaparin in ESSENCE and TIMI 11B may be greatly re-
duced or entirely eclipsed by the aforementioned evolutions in
pharmacotherapeutics. The risks include bleeding and cost. Bleeding
was worse in the enoxaparin groups in ESSENCE and TIMI 11B, and
P2Y12 antagonists also increase the risk of bleeding [19-22]. As for cost,
enoxaparin is more expensive than UFH, and this is only accentuated by
the possibility for enoxaparin to offer no efficacy advantage and merely
increase the risk of bleeding. It is also worthwhile to consider possible
situations where UFH has pharmacokinetic and pharmacodynamic
advantages over enoxaparin. For instance, one can stop a UFH infusion
easily and the UFH will be out of a patient's system more quickly than
after giving an injection of enoxaparin, which is potentially useful for
patients who are at a higher risk of bleeding or for patients who are unsta-
ble or considered to be at high risk for deterioration and may therefore
ultimately require invasive management. Additionally, one can reliably
reverse UFH if needed with protamine sulfate, which is potentially useful
for patients at a higher risk of bleeding (a protamine sulfate infusion can
also be attempted if one desires to reverse the effects of enoxaparin, but
the results are not as guaranteed as they are with UFH). Finally, one can
confidently use UFH in the setting of concern about the adequacy of sub-
cutaneous absorption, and one can use UFH in the setting of advanced
renal disease without any concern about renal-adjusted dosing or unde-
sirable accumulation of the anticoagulant. Additionally, if a clinician is
leaning toward enoxaparin for whatever reason, OASIS-5 suggests the cli-
nician should at least consider fondaparinux as an alternative. Unless and
until better evidence emerges, this is where things stand, and although it
is sometimes a reflexively-offered statement, further research does in-
deed seem indicated to help truly elucidate this situation.
Sources of support
None.
Prior presentation
None.
Acknowledgements
None.
References
[1] Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC guideline for the man-
agement of patients with non-ST-elevation acute coronary syndromes: a report of
the American College of Cardiology/American Heart Association Task Force on Prac-
tice Guidelines. Circulation Dec. 23 2014;130(25):e344–426. http://dx.doi.org/10.
1161/CIR.0000000000000134 [Epub 2014 Sep 23. Erratum in: Circulation. 2014
Dec 23;130(25):e433–434].
[2] Amsterdam EA, Wenger NK, Brindis RG, et al. 2014 AHA/ACC guideline for the man-
agement of patients with non-ST-elevation acute coronary syndromes: a report of
Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070
5
the American College of Cardiology/American Heart Association Task Force on Prac-
tice Guidelines. J Am Coll Cardiol Dec. 23 2014;64(24):e139–228. http://dx.doi.org/
10.1016/j.jacc.2014.09.017 [Epub 2014 Sep 23. Erratum in: J Am Coll Cardiol. 2014
Dec 23;64(24):2713–2714].
[3] Mayer M. Elucidating the use of enoxaparin in non-ST-elevation acute coronary syn-
dromes (NSTE-ACS). Evid Based Med 2016 Feb;21(1):4–6. http://dx.doi.org/10.
1136/ebmed-2015-110275 [Epub 2015 Oct 20].
[4] Cohen M, Demers C, Gurfinkel EP, et al. A comparison of low-molecular-weight hep-
arin with unfractionated heparin for unstable coronary artery disease. Efficacy and
safety of subcutaneous enoxaparin in non-Q-Wave Coronary Events Study Group.
N Engl J Med 1997;337:447–52. http://dx.doi.org/10.1056/NEJM199708143370702.
[5] Antman EM, McCabe CH, Gurfinkel EP, et al. Enoxaparin prevents death and cardiac
ischemic events in unstable angina/non-Q-wave myocardial infarction. Results of
the thrombolysis in myocardial infarction (TIMI) 11B trial. Circulation 1999;100:
1593–601. http://dx.doi.org/10.1161/01.CIR.100.15.1593.
[6] Cohen M, Théroux P, Borzak S, et al. ACUTE II investigators. Randomized double-
blind safety study of enoxaparin versus unfractionated heparin in patients with
non-ST-segment elevation acute coronary syndromes treated with tirofiban and as-
pirin: the ACUTE II study. The antithrombotic combination using tirofiban and
enoxaparin. Am Heart J 2002;144:470–7. http://dx.doi.org/10.1067/mhj.2002.
126115.
[7] Goodman SG, Fitchett D, Armstrong PW, et al. Integrilin and enoxaparin randomized
assessment of acute coronary syndrome treatment (INTERACT) trial investigators.
Randomized evaluation of the safety and efficacy of enoxaparin versus
unfractionated heparin in high-risk patients with non-ST-segment elevation acute
coronary syndromes receiving the glycoprotein IIb/IIIa inhibitor eptifibatide. Circula-
tion 2003;107:238–44. http://dx.doi.org/10.1161/01.CIR.0000050144.67910.13.
[8] Blazing MA, de Lemos JA, White HD, et al. ‘A to Z’ investigators. Safety and efficacy of
enoxaparin vs unfractionated heparin in patients with non-ST-segment elevation
acute coronary syndromes who receive tirofiban and aspirin: a randomized con-
trolled trial. JAMA 2004;292:55–64. http://dx.doi.org/10.1001/jama.292.1.55.
[9] Ferguson JJ, Califf RM, Antman EM, et al. SYNERGY trial investigators. Enoxaparin vs
unfractionated heparin in high-risk patients with non-ST-segment elevation acute
coronary syndromes managed with an intended early invasive strategy: primary re-
sults of the SYNERGY randomized trial. JAMA 2004;292:45–54. http://dx.doi.org/10.
1001/jama.292.1.45.
[10] Antman EM, Cohen M, Radley D, et al. Assessment of the treatment effect of
enoxaparin for unstable angina/non-Q-wave myocardial infarction. TIMI 11B-ES-
SENCE meta-analysis. Circulation 1999;100:1602–8. http://dx.doi.org/10.1161/01.
CIR.100.15.1602.
[11] Freemantle N, Calvert M, Wood J, Eastaugh J, Griffin C. Composite outcomes in ran-
domized trials: greater precision but with greater uncertainty? JAMA May 21 2003;
289(19):2554–9.
[12] Montori VM, Permanyer-Miralda G, Ferreira-González I, et al. Validity of composite
end points in clinical trials. BMJ Mar. 12 2005;330(7491):594–6.
[13] Ferreira-González I, Busse JW, Heels-Ansdell D, et al. Problems with use of composite
end points in cardiovascular trials: systematic review of randomised controlled tri-
als. BMJ Apr 14 2007;334(7597):786 [Epub 2007 Apr 2].
[14] Freemantle N, Calvert M. Composite and surrogate outcomes in randomised con-
trolled trials. BMJ Apr. 14 2007;334(7597):756–7.
[15] Goodman SN, Berlin JA. The use of predicted confidence intervals when planning
experiments and the misuse of power when interpreting results. Ann Intern Med
1994;121:200–6. http://dx.doi.org/10.7326/0003-4819-121-3-199408010-
00008.
[16] Hoenig JM, Heisey DM. The abuse of power: the pervasive fallacy of power calcula-
tions for data analysis. Am Stat 2001;55(1):19–24.
[17] Schulz KF, Grimes DA. Sample size calculations in randomised trials: mandatory and
mystical. Lancet Apr 9–15 2005;365(9467):1348–53.
[18] Antman EM, Cohen M, McCabe C, et al. TIMI 11B and ESSENCE investigators.
Enoxaparin is superior to unfractionated heparin for preventing clinical events at
1-year follow-up of TIMI 11B and ESSENCE. Eur Heart J 2002;23:308–14. http://dx.
doi.org/10.1053/euhj.2001.2779.
[19] Yusuf S, Zhao F, Mehta SR, et al. Clopidogrel in unstable angina to prevent recurrent
events trial investigators. Effects of clopidogrel in addition to aspirin in patients with
acute coronary syndromes without ST-segment elevation. N Engl J Med Aug 16
2001;345(7):494–502.
[20] Mehta SR, Bassand JP, Chrolavicius S, et al. CURRENT-OASIS 7 investigators. Dose
comparisons of clopidogrel and aspirin in acute coronary syndromes. N Engl J Med
Sep. 2 2010;363(10):930–42. http://dx.doi.org/10.1056/NEJMoa0909475.
[21] Wallentin L, Becker RC, Budaj A, et al. PLATO investigators. Ticagrelor versus
clopidogrel in patients with acute coronary syndromes. N Engl J Med Sep. 10
2009;361(11):1045–57. http://dx.doi.org/10.1056/NEJMoa0904327 [Epub 2009
Aug 30].
[22] Wiviott SD, Braunwald E, McCabe CH, et al. TRITON-TIMI 38 investigators. Prasugrel
versus clopidogrel in patients with acute coronary syndromes. N Engl J Med Nov. 15
2007;357(20):2001–15 [Epub 2007 Nov 4].
[23] Sacks FM, Pfeffer MA, Moye LA, et al. The effect of pravastatin on coronary events
after myocardial infarction in patients with average cholesterol levels. Choles-
terol and recurrent events trial investigators. N Engl J Med Oct. 3 1996;
335(14):1001–9.
[24] Schwartz GG, Olsson AG, Ezekowitz MD, et al. Myocardial ischemia reduction with
aggressive cholesterol lowering (MIRACL) study investigators. Effects of atorvastatin
on early recurrent ischemic events in acute coronary syndromes: the MIRACL study:
a randomized controlled trial. JAMA Apr. 4 2001;285(13):1711–8.
[25] Cannon CP, Braunwald E, McCabe CH, et al. Pravastatin or atorvastatin evaluation
and infection therapy-thrombolysis in myocardial infarction 22 investigators.
6 M. Mayer / American Journal of Emergency Medicine xxx (2017) xxx–xxx

Intensive versus moderate lipid lowering with statins after acute coronary syn-
dromes. N Engl J Med Apr 8 2004;350(15):1495–504 [Epub 2004 Mar 8].
[26] Cannon CP, Steinberg BA, Murphy SA, et al. Meta-analysis of cardiovascular out-
comes trials comparing intensive versus moderate statin therapy. J Am Coll Cardiol
Aug. 1 2006;48(3):438–45 [Epub 2006 Jul 12].
[27] Mahaffey KW, Cohen M, Garg J, et al. SYNERGY trial investigators. High-risk patients
with acute coronary syndromes treated with low-molecular-weight or
unfractionated heparin: outcomes at 6 months and 1 year in the SYNERGY trial.
JAMA 2005;294:2594–600.
[28] Fifth Organization to Assess Strategies in Acute Ischemic Syndromes Investigators,
Yusuf S, Mehta SR, et al. Comparison of fondaparinux and enoxaparin in acute coro-
nary syndromes. N Engl J Med Apr. 6 2006;354(14):1464–76 [Epub 2006 Mar 14].
[29] Andrade-Castellanos CA, Colunga-Lozano LE, Delgado-Figueroa N, Magee K. Heparin
versus placebo for non-ST elevation acute coronary syndromes. Cochrane Database
Syst Rev Jun. 27 2014;6:CD003462. http://dx.doi.org/10.1002/14651858.CD003462.
pub3.

Please cite this article as: Mayer M, Anticoagulants in ischemia-guided management of non-ST-elevation acute coronary syndromes, American
Journal of Emergency Medicine (2017), http://dx.doi.org/10.1016/j.ajem.2016.12.070