- Email: [email protected]
Mortality in Patients with Atrial Fibrillation Randomized to Edoxaban or Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial
Accepted Manuscript Mortality in Patients With Atrial Fibrillation Randomized to Edoxaban or Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial Robert P. Giugliano, MD, SM, Christian T. Ruff, MD, MPH, Stephen D. Wiviott, MD, Francesco Nordio, PhD, Sabina A. Murphy, MPH, Johannes AN. Kappelhof, BSc, Minggao Shi, PhD, Michele Mercuri, MD, Elliott M. Antman, MD, Eugene Braunwald, MD PII:
S0002-9343(16)30246-7
DOI:
10.1016/j.amjmed.2016.02.028
Reference:
AJM 13419
To appear in:
The American Journal of Medicine
Received Date: 16 February 2016 Revised Date:
17 February 2016
Accepted Date: 18 February 2016
Please cite this article as: Giugliano RP, Ruff CT, Wiviott SD, Nordio F, Murphy SA, Kappelhof JA, Shi M, Mercuri M, Antman EM, Braunwald E, Mortality in Patients With Atrial Fibrillation Randomized to Edoxaban or Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial, The American Journal of Medicine (2016), doi: 10.1016/j.amjmed.2016.02.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Mortality in Patients With Atrial Fibrillation Randomized to Edoxaban or
RI PT
Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial
Robert P. Giugliano MD, SM1, Christian T. Ruff, MD, MPH1, Stephen D. Wiviott, MD1, Francesco Nordio, PhD1, Sabina A. Murphy, MPH1, Johannes AN Kappelhof, BSc, Minggao
SC
Shi, PhD2, Michele Mercuri, MD2, Elliott M. Antman, MD1, and Eugene Braunwald, MD1
M AN U
From the 1TIMI Study Group, Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA and 2Daiichi-Sankyo Pharma Development, Edison, NJ
Conflicts of Interest
TE D
Running Title: Mortality with Edoxaban vs Warfarin in AF
EP
Dr. Giugliano reports receiving consulting fees from Bristol Myers Squibb, Daiichi Sankyo, Janssen Pharmaceuticals, Merck, Pfizer, and Portola; lecture fees from Daiichi Sankyo and Merck; and grant support through his institution from Daiichi Sankyo, Merck, Johnson & Johnson, Sanofi, and AstraZeneca.
AC C
Dr. Ruff reports receiving consulting fees from Daiichi Sankyo, Bristol-Myers, Squibb, and Boehringer Ingelheim and grant support through his institution from Daiichi Sankyo. Dr. Wiviott reports receiving consulting fees from AstraZeneca, Bristol-Myers Squibb, Eisai, Arena Pharmaceuticals, Eli Lilly, Daiichi Sankyo, Aegerion, AngelMed, Janssen Pharmaceuticals, Xoma, ICON Clinical Research, and Boston Clinical Research Institute and grant support through his institution from AstraZeneca, Bristol-Myers Squibb, Eisai, Arena Pharmaceuticals, Merck, Eli Lilly, Daiichi Sankyo, and Sanofi. Drs. Nordio and Antman, and Ms. Murphy report receiving grant support through their institution from Daiichi Sankyo. Mr. Kappelhof and Drs. Shi, and Mercuri report being employees of Daiichi Sankyo. Dr. Mercuri also reports holding a pending patent related to the clinical properties of edoxaban.
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF
RI PT
Dr. Braunwald reports receiving consulting fees from Sanofi, Genzyme, Amorcyte, the Medicines Company, and Cardiorentis; lecture fees from Eli Lilly, Menarini, Medscape, and Bayer HealthCare; and grant support through his institution from Daiichi Sankyo, AstraZeneca, Johnson & Johnson, GlaxoSmithKline, Bristol-Myers Squibb, Beckman Coulter, Roche Diagnostics, Pfizer, Merck, and Sanofi. He also reports serving as an unpaid consultant for Merck and providing uncompensated lectures for Merck and CVRx.
Text word count: 2910
AC C
EP
TE D
Total word count: 4681
M AN U
Address for Correspondence Robert P. Giugliano, MD, SM TIMI Study Group 350 Longwood Avenue, 1st Floor Offices Boston, MA 02115 TEL: 617-278-0145 FAX: 617-734-7329 EMAIL: [email protected]
SC
All authors had access to the data and a role in writing the manuscript.
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF Abstract Background: When compared with warfarin, edoxaban significantly reduced cardiovascular mortality in the ENGAGE AF-TIMI 48 trial. We studied the possible reasons
RI PT
leading to this reduction.
Methods: ENGAGE AF-TIMI 48 was a double-blind, double-dummy comparison of warfarin with 2 regimens of once-daily edoxaban in 21,105 patients with atrial fibrillation
SC
followed for 2.8 years (median). Causes of deaths in the intention-to-treat population were
classified as cardiovascular (including fatal bleeding and ischemic stroke), malignancy, or non-
M AN U
cardiovascular/non-malignancy by an independent, blinded, clinical endpoint committee . Deaths also were adjudicated as directly due to bleeding (i.e., fatal), or bleeding contributing to death, or neither.
Results: There were 839 total deaths (4.35%/yr) in the warfarin arm, compared with 773
TE D
(3.99%/yr, p=0.08) with the higher-dose edoxaban regimen, and 737 (3.80%/yr, p=0.006) with the lower-dose edoxaban regimen. No significant differences between treatments were observed in: 1) any of the 3 most common causes of cardiovascular death (sudden cardiac, heart failure,
EP
ischemic stroke), 2) fatal malignancies, 3) other non-cardiovascular death. There were 124 fatal bleeds, 65 with warfarin, significantly fewer with the higher-dose (n=35, p=0.003) and lower-
AC C
dose (n=24, p<0.001) edoxaban regimens. There were 101 bleeding events with warfarin that were either fatal or that contributed to death. There were significantly fewer with the higher-dose (n=59, P=0.001) and lower-dose (n=54, P<0.001) edoxaban regimens. Conclusions: Fewer total and cardiovascular deaths were observed with edoxaban as
compared to warfarin in the ENGAGE AF-TIMI 48 trial, and this predominantly resulted from significantly lower rate of major bleeding with edoxaban. Edoxaban reduces mortality both
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF directly (less fatal bleeding) and indirectly (fewer bleeding-related complications and
RI PT
interruptions in therapy after non-fatal bleeding).
Abstract word count: 266
AC C
EP
TE D
M AN U
SC
Keywords: death, hemorrhage, bleeding, anticoagulant, factor Xa inhibitor, edoxaban, warfarin
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF A systematic review of population-based studies in patients with atrial fibrillation from 21 Global Burden of Disease regions demonstrated that the overall incidence, prevalence, and mortality attributed to atrial fibrillation has increased from 1990 to 2010. This is largely due to
RI PT
aging of the world population, and adds to a growing and significant worldwide public health burden.1 While a meta-analysis of 6 trials involving 2900 patients with non-valvular atrial
fibrillation randomized to warfarin or placebo demonstrated a 26% reduction in total mortality
SC
with warfarin2, there was no significant reduction in total mortality with warfarin compared to antiplatelet therapy in 8 trials involving 3647 patients2, nor in a subsequent trial of 6706 patients
M AN U
randomized to warfarin or dual antiplatelet therapy (hazard ratio (HR) = 0.99).3 Residual mortality risk in patients with atrial fibrillation anticoagulated with warfarin has been demonstrated4, and may be due to limitations in efficacy or increased mortality related to sideeffects of warfarin, or both. For example, warfarin increases the risk of major hemorrhage
TE D
relative to no antithrombotic or aspirin, in particular intracranial hemorrhage, which is frequently fatal or severely disabling, even despite rapid reversal5 of the anticoagulant effect with prothrombin complex concentrates.6
EP
In contrast, in a meta-analysis of four large randomized warfarin-controlled trials involving 71,683 patients with atrial fibrillation, we reported that when compared to warfarin,
AC C
the non-vitamin K oral anticoagulants (NOACs) reduced total mortality by 10% and did not increase major bleeding as compared to warfarin.7 In addition, NOACs demonstrated a consistent and large (approximately 50%) reduction in intracranial hemorrhage compared to warfarin across these four trials, whereas the observed reduction in thromboembolic events was modest and variable across these trials.7
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF In the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation– Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial, the oral factor Xa inhibitor edoxaban, as compared to well-managed warfarin, was non-inferior for prevention of
RI PT
stroke and systemic embolic events, while reducing cardiovascular mortality, major bleeding, and intracranial hemorrhage.8 However, data regarding the causes of death and the relationship of bleeding to death with NOACs as compared to warfarin are sparse. Hence, we analyzed
SC
death in the ENGAGE AF-TIMI 48 trial to explore the factors that contributed to the improved
M AN U
survival observed with edoxaban.
Methods
The ENGAGE AF-TIMI 48 trial was a double-blind, double-dummy, randomized controlled trial in 21,105 patients with atrial fibrillation. Details of the trial design have been
TE D
described.9 Briefly, patients with atrial fibrillation in the prior 12 months with a CHADS2 score10 >2 were randomized to warfarin titrated to an international normalized ratio of 2-3, a higher-dose regimen of edoxaban, or a lower-dose edoxaban regimen. The doses of edoxaban
EP
were 60 mg and 30 mg once daily in the higher-dose and lower-dose edoxaban regimen arms respectively. In both edoxaban arms the dose was reduced by 50% if patients had any one of the
AC C
following at any time during the trial: creatinine clearance <50 mg/dL, body weight < 60 kg, concomitant use of verapamil, quinidine, or dronedarone. Major exclusion criteria were high risk for bleeding, creatinine clearance < 30 mL/min, need for dual antiplatelet therapy, or acute coronary syndrome or stroke within 30 days. The primary efficacy endpoint was the combination of stroke or systemic embolic events. The principal safety endpoint was major bleeding as defined by the International Society of Thrombosis and Haemostasis.11 Patients were followed
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF for a median of 2.8 years (interquartile range 2.4 to 3.2 years). The protocol and amendments were approved by the ethics committee at each participating center. All the patients provided written informed consent.
RI PT
An independent clinical end-point committee, blinded to treatment assignment, adjudicated all deaths and suspected cerebrovascular events, systemic embolic events,
myocardial infarctions, and bleeding. In the adjudication of deaths, the clinical endpoint
SC
committee determined the primary cause of death, categorized as shown in Table 1 (see
Appendix 1 for detailed definitions). Fatal bleeding events (whether intra- or extra-cranial) were
M AN U
classified as cardiovascular deaths. In addition the clinical endpoint committee determined whether: 1) deaths were directly caused by bleeding (i.e., a fatal bleeding event), 2) bleeding contributed to, but was not the direct cause of death within the next 30 days (i.e., a death in which a bleeding event was part of a causal chain of medical events that ultimately led to death
TE D
within 30 days of the bleed, but bleeding was not directly and/or immediately related to subject’s death), and 3) death was not associated with either of the above (See Table S1 for details). The present analysis was conducted using the intention-to-treat principle; it included all
EP
patients randomized and counted all deaths and bleeding events, whether on or off study drug, that occurred between randomization and the end of the study drug treatment period. The one
AC C
exception was the analysis of consequences of a major bleed while on study drug, which was conducted in the on-treatment population. Annualized event rates were calculated as KaplanMeier estimates. P-values <0.05 without adjustment for multiple comparisons, were considered nominally significant.
Role of the Funding Source
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF The ENGAGE AF-TIMI 48 trial was supported by a research grant from Daiichi Sankyo Pharma Development, who participated in the design and conduct of this trial, and in the preparation of
RI PT
this manuscript.
Results
A total of 2349 deaths occurred, of which 1668 (71.0%) were cardiovascular, 271
SC
(11.5%) were due to malignancy, and 410 (17.5%) were due other causes (Table 2). Of these 2349 deaths, 839 deaths (4.35%/year) occurred in patients randomized to warfarin. Fewer deaths
M AN U
occurred with the higher-dose edoxaban regimen (773 deaths, 3.99%/year) and with the lowerdose edoxaban regimen (737 deaths, 3.80%/year) as compared to warfarin. There were a total of 1668 cardiovascular deaths. Of these, 611 (3.17%/year) occurred in the warfarin group. Both edoxaban dose regimens significantly reduced cardiovascular
TE D
mortality: higher-dose edoxaban regimen 530 cardiovascular deaths, 2.74%/year; lower-dose edoxaban regimen 527 cardiovascular deaths, 2.71%/year. There were no significant differences among the treatment groups between deaths due to either malignancy (warfarin 84 (0.44%/year);
EP
higher-dose edoxaban regimen 94 (0.49%/year), lower-dose edoxaban regimen 93 (0.48%/year)) or other non-cardiovascular/non-malignant causes of death, although numerically fewer occurred
AC C
with the lower-dose edoxaban regimen (warfarin 144 (0.75%/year); higher-dose edoxaban regimen 149 (0.77%/year), lower-dose edoxaban regimen 117 (0.60%/year)).
Causes of Cardiovascular Death The four most common causes of cardiovascular deaths were sudden cardiac death (749 (1.29%/year); 44.9% of the 1668 cardiovascular deaths), heart failure (390
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF (0.67%/year); 23.4% of cardiovascular deaths), ischemic stroke (146 (0.25%/year); 8.8% of cardiovascular deaths), and bleeding (124 (0.21%/year); 7.4% of cardiovascular deaths) (Table 2, Figure 1). There were no differences between the treatment groups in the first 3 subcategories of
RI PT
cardiovascular death (sudden cardiac death, heart failure, ischemic stroke), but there was a
significant excess of fatal bleeding events with warfarin (n=65 (0.33%/year)) as compared to both the higher-dose edoxaban regimen (35 (0.18%/year); HR=0.54, p=0.003 vs. warfarin) and
SC
the lower-dose edoxaban regimen (24 (0.12%/year)); HR=0.37, p<0.001 vs. warfarin). The rates of fatal bleeding were low and not dissimilar with the higher-dose edoxaban regimen
M AN U
(0.18%/year) and the lower-dose edoxaban regimen (0.12%/year, p=0.24). The case fatality rates following a major bleeding event were 11.3% with warfarin, 7.7% with the higher-dose edoxaban regimen (p=0.064) and 8.3% with the lower-dose edoxaban regimen (p=0.20).
TE D
Deaths Related to Bleeding
Of the 124 (0.21%/year) fatal bleeding events, 98 (0.17%/year, 79% of the fatal bleeds) were due to intracranial hemorrhage, even though intracranial hemorrhage accounted for fewer
EP
than 20% (294/1592) of the major bleeding events. This can be explained by the nearly 20-fold greater fatality rate after intracranial hemorrhage (33%) as compared to extracranial bleeding
AC C
(2%, Figure 2). There were 155 (0.97%/year) intracranial hemorrhages in the warfarin group. Fewer intracranial hemorrhages occurred in the higher-dose edoxaban regimen group (n=79 (0.50%/year); p<0.001), and fewer yet in the lower-dose edoxaban regimen group (n=60 (0.37%/year); p<0.001 vs. both the higher-dose edoxaban regimen and vs. warfarin). There was a trend toward more major extracranial bleeding events in the warfarin group as compared to the higher-dose edoxaban regimen group (506 vs. 455, (3.16 vs. 2.87%/year) p=0.09), while there
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF were significantly fewer with the lower-dose edoxaban regimen (337 (2.07%/year), p<0.001 compared to either warfarin or the higher-dose edoxaban regimen). In addition to the significant reduction in fatal bleeding events with edoxaban, there also
RI PT
was a numerical, but statistically non-significant, reduction in bleeding events that were
adjudicated as contributing to death in the edoxaban groups (warfarin 36 (0.23%/year); higherdose edoxaban regimen 24 (0.15%/year), p=0.12 higher-dose edoxaban regimen vs. warfarin;
SC
lower-dose edoxaban regimen 30 (0.18%/year), p=0.44 lower-dose edoxaban regimen vs.
warfarin). In contrast with fatal bleeding events, only 17% (16 of 96) of major bleeding events
M AN U
that contributed to death were due to intracranial hemorrhage. The total number of bleeding events that were either fatal or contributed to death was 40-45% lower (relatively) with edoxaban as compared with warfarin (Figure 3). In the warfarin group there were 81 (0.42%/year) excess cardiovascular deaths as compared to the higher-dose edoxaban regimen group, of which 52%
TE D
(42/81) were related to bleeding that was either fatal or contributed to death. Similarly with the lower-dose edoxaban regimen, 54% (47/84) of the excess cardiovascular deaths with warfarin as
EP
compared to the lower-dose edoxaban regimen were related to bleeding.
Consequences of Non-Fatal Major Bleeding While on Study Drug
AC C
Patients on study drug who experienced a major non-fatal bleeding event that did not
contribute to death had nearly twice the relative risk of dying before the end of the trial compared to those who did not have major bleeding during the trial (7.0 vs. 4.1% per year, relative risk unadjusted 1.9, p<0.001). Of the 1080 patients (2.34%/year) on study drug who survived a major bleed (463 (3.03%/year) warfarin, 385 (2.53%/year) higher-dose edoxaban regimen, 232 (1.48%/year) lower-dose edoxaban regimen), study drug was interrupted in 91%
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF for at least 4 days, with similar rates (91% warfarin, 90% higher-dose edoxaban regimen, 91% lower-dose edoxaban regimen) across the 3 treatment groups. Of these 979 who interrupted study drug, 53% were re-challenged with study drug, 37% received open-label anticoagulant, while
RI PT
10% did not receive any further anticoagulant (neither study drug nor open-label anticoagulant) for the remainder of the trial.
Patients who survived a major bleed, but interrupted anticoagulation for > 4 days
SC
experienced high rates of subsequent major adverse cardiac events (7.9% annualized rate of major adverse cardiac events, defined as cardiovascular death, myocardial infarction, stroke or
M AN U
systemic embolic events) and of all-cause death (7.1%/year), regardless of randomized treatment. In the warfarin group, 423 patients (2.72%/year) interrupted anticoagulation after a non-fatal major bleed that did not contribute to death. Since edoxaban reduced major bleeding as compared to warfarin, the corresponding number of patients who interrupted study drug
TE D
following major bleeding were fewer with the higher-dose edoxaban regimen (345 (2.23%/year); p=0.004 vs. warfarin) and, fewer yet with the lower-dose edoxaban regimen (211 (1.32%/year); p<0.001 vs. warfarin and vs. higher-dose edoxaban regimen). Following interruption of study
EP
drug due to a non-fatal major bleeding event, there were numerically more deaths with warfarin (n=81) than with edoxaban (higher-dose edoxaban regimen 64; p=0.18 vs. warfarin, lower-dose
AC C
edoxaban regimen 42; p=0.001 for lower-dose edoxaban regimen vs. warfarin). Similarly, there were more major adverse cardiac events after study drug interruption due to a non-fatal major bleed with warfarin (n=111) as compared to the higher-dose edoxaban regimen (66; p=0.001 vs. warfarin) and lthe ower-dose edoxaban regimen (58, p<0.001 vs. warfarin).
Discussion
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF In this analysis of causes of total mortality and the relationship between bleeding and causes of death in the ENGAGE AF-TIMI 48 trial, we found that more than half of the reductions in all-cause deaths with edoxaban as compared to warfarin were directly attributable
RI PT
to fewer fatal bleeding events and non-fatal major bleeding that contributed to a death within 30 days. There were no differences between edoxaban and warfarin in deaths due to non-
cardiovascular causes, nor for any of the 3 most common cardiovascular causes of death (sudden
SC
cardiac death, heart failure, ischemic stroke). However, warfarin was associated with
approximately twice the risk of fatal bleeding as compared to edoxaban. Since edoxaban (like
M AN U
other NOACs) reduces intracranial hemorrhage by approximately 50% relative to warfarin7, and intracranial hemorrhage carries a high fatality rate12 (1-in-3 in this study), the single greatest factor accounting for the lower total mortality with edoxaban was a reduction in fatal intracranial hemorrhage (by 43% and 72% with the higher-dose edoxaban regimen and lthe ower-dose
TE D
edoxaban regimen, respectively, compared to warfarin).
In addition to a reduction in fatal bleeding events with edoxaban, there also were fewer non-fatal major bleeding events in patients treated with edoxaban. This latter finding also may
EP
have contributed to the reduction in total mortality with edoxaban in two ways. First, patients who initially survive a major bleeding event may subsequently develop another medical
AC C
complication that ultimately proves fatal, for example, an event that is: a) a direct consequence of the bleed (e.g., hypovolemic shock leading to myocardial or cerebral infarction), b) related to management of the bleeding event (e.g., transfusion- or procedure-related complication), or c) due to an in-hospital complication (e.g., nosocomial infection). Secondly, patients who survive a major bleed often interrupt anticoagulation for several days (91% in this study), and even brief interruptions in anticoagulation place patients with atrial fibrillation at an increase risk of
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF thromboembolic events and death.1 In fact, even brief interruptions of anticoagulation for ≤ 5 days have been associated with increased risk.13, 14 Our finding that the majority of deaths among patients with atrial fibrillation treated with
RI PT
a NOAC or well-managed warfarin are cardiovascular in nature is consistent with those reported by the RE-LY investigators.15 The numerically greater proportion of cardiovascular deaths in ENGAGE AF-TIMI 48 (71%) as compared to RE-LY (61%) may be attributable to the greater
SC
burden of risk factors in the ENGAGE AF-TIMI 48 trial (patients with CHADS2 score 4-6
represented 53% of the ENGAGE AF-TIMI 488 and 33% of the RE-LY16 trial populations). In
M AN U
both trials, there were trends favoring the NOACs over warfarin with relative reductions in total mortality of 8-13%. While there were no differences between the NOACs and warfarin in the most common forms of cardiovascular death (sudden cardiac death, heart failure) in either trial, a significant reduction in vascular deaths (including stroke, peripheral or pulmonary embolism and
TE D
bleeding) was observed with dabigatran as compared to warfarin in RE-LY.15 In the current analysis, we extended these findings to show more specifically that the major explanation for lower mortality with edoxaban was related to the large reductions in bleeding events that were
EP
either fatal or contributed to death within 30 days. Both apixaban and rivaroxaban reduce intracranial hemorrhage by ~50% compared to warfarin, however, similar detailed analyses of
AC C
the causes of death in the large phase 3 trials comparing these two Factor Xa inhibitors with warfarin have not been published. 17, 18,19 We believe that our findings provide a strong rationale to consider a NOAC rather than
warfarin, to reduce bleeding and death, while preserving protection from thromboembolism in patients with atrial fibrillation at moderate to high risk for stroke. Our results provide further support to the recommendations from the European Society of Cardiology20 and the Japanese
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF Circulation Society 21 favoring NOACs over vitamin K antagonists for most patients with atrial fibrillation and two or more risk factors for stroke. With the anticipated availability in the near future of antidotes to Factor Xa and thrombin inhibitors,22 the safety advantage of the NOACs
RI PT
over vitamin K antagonists should be enhanced even further, potentially extending the net clinical benefit with NOACs to patients with even lower risk of stroke.23, 24
We acknowledge some limitations. This analysis was limited to patients who met entry
SC
criteria for and agreed to participate in a randomized clinical trial, and the findings may differ in a general population. Secondly, attributing a single cause of death is challenging despite use of
M AN U
prospective definitions since multiple factors may contribute to death, source data are not always available (e.g., for deaths that occur at non-study institutions), and autopsies are infrequently performed. However, since adjudications of death were unaware of treatment assignment, the likelihood of introducing substantial bias is minimized. Classification of bleeding deaths as
TE D
cardiovascular was arbitrary, but was prespecified in the trial protocol. Nonetheless, the validity of our results still stand if bleeding deaths were considered in a separate category. This analysis does not address the question of how much bleeding should be tolerated to prevent a
patients.
EP
thromboembolic event nor which dose regimen of edoxaban should be selected for individual
AC C
Strengths of our analysis include the independent adjudication of causes of death by an
expert committee blinded to treatment assignment, using a detailed prospectively designed charter. The dataset also represents the largest (21,105 patients) and longest (2.8 year median) follow up with the greatest number of total deaths (n=2349) from a randomized trial in patients with atrial fibrillation conducted to date. Follow-up for vital status was complete for all except one patient, and complete information for key clinical endpoints was available in 99.5% of the
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF total 56,346 patient-years of potential follow-up.8 The median time in therapeutic range was 68.4% [56.5%, 77.4%] among the patients treated with warfarin, which is higher than that reported in prior NOAC vs. warfarin trials16-18 (range 58-66%) and in a recent US assessment
RI PT
involving 138,319 Americans with atrial fibrillation (median 57.5%).25
Conclusions
SC
In patients with atrial fibrillation at moderate to high risk of stroke randomized to edoxaban vs. warfarin in the ENGAGE AF-TIMI 48 trial, the significant reduction in
M AN U
cardiovascular mortality observed with edoxaban was mediated primarily by lower rates of
AC C
EP
TE D
bleeding that were either fatal or contributed to death in the subsequent 30 days.
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF References
1.
Chugh SS, Havmoeller R, Narayanan K, Singh D, Rienstra M, Benjamin EJ, Gillum RF,
RI PT
Kim YH, McAnulty JH, Jr., Zheng ZJ, Forouzanfar MH, Naghavi M, Mensah GA, Ezzati M, Murray CJ. Worldwide epidemiology of atrial fibrillation: A global burden of disease 2010 study. Circulation. 2014;129:837-847
Hart RG, Pearce LA, Aguilar MI. Meta-analysis: Antithrombotic therapy to prevent
SC
2.
stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med.
3.
M AN U
2007;146:857-867
Connolly S, Pogue J, Hart R, Pfeffer M, Hohnloser S, Chrolavicius S, Yusuf S. Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the atrial fibrillation clopidogrel trial with irbesartan for prevention of vascular events (active w):
4.
TE D
A randomised controlled trial. Lancet. 2006;367:1903-1912
Senoo K, Lip GY, Lane DA, Buller HR, Kotecha D. Residual risk of stroke and death in anticoagulated patients according to the type of atrial fibrillation: Amadeus trial. Stroke;
5.
EP
a journal of cerebral circulation. 2015;46:2523-2528 Goldstein JN, Refaai MA, Milling TJ, Jr., Lewis B, Goldberg-Alberts R, Hug BA, Sarode
AC C
R. Four-factor prothrombin complex concentrate versus plasma for rapid vitamin k antagonist reversal in patients needing urgent surgical or invasive interventions: A phase 3b, open-label, non-inferiority, randomised trial. Lancet. 2015;385:2077-2087
6.
Dowlatshahi D, Butcher KS, Asdaghi N, Nahirniak S, Bernbaum ML, Giulivi A, Wasserman JK, Poon MC, Coutts SB. Poor prognosis in warfarin-associated intracranial
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF hemorrhage despite anticoagulation reversal. Stroke; a journal of cerebral circulation. 2012;43:1812-1817 7.
Ruff CT, Giugliano RP, Braunwald E, Hoffman EB, Deenadayalu N, Ezekowitz MD,
RI PT
Camm AJ, Weitz JI, Lewis BS, Parkhomenko A, Yamashita T, Antman EM. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: A meta-analysis of randomised trials. Lancet. 2014;383:955-962
Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL,
SC
8.
Ezekowitz MD, Weitz JI, Spinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M,
M AN U
Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M, Antman EM. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2013;369:2093-2104 9.
Ruff CT, Giugliano RP, Antman EM, Crugnale SE, Bocanegra T, Mercuri M, Hanyok J, Patel I, Shi M, Salazar D, McCabe CH, Braunwald E. Evaluation of the novel factor xa
TE D
inhibitor edoxaban compared with warfarin in patients with atrial fibrillation: Design and rationale for the effective anticoagulation with factor xa next generation in atrial fibrillation-thrombolysis in myocardial infarction study 48 (engage af-timi 48). American
10.
EP
heart journal. 2010;160:635-641
Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of
AC C
clinical classification schemes for predicting stroke: Results from the national registry of atrial fibrillation. JAMA. 2001;285:2864-2870
11.
Schulman S, Kearon C. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005;3:692-694
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF 12.
Giugliano RP, Giraldez RR, Morrow DA, Antman EM, Gibson CM, Mohanavelu S, Murphy SA, McCabe CH, Braunwald E. Relations between bleeding and outcomes in
heart journal. 2010;31:2103-2110 13.
RI PT
patients with st-elevation myocardial infarction in the extract-timi 25 trial. European
Healey JS, Eikelboom J, Douketis J, Wallentin L, Oldgren J, Yang S, Themeles E,
Heidbuchel H, Avezum A, Reilly P, Connolly SJ, Yusuf S, Ezekowitz M. Periprocedural
SC
bleeding and thromboembolic events with dabigatran compared with warfarin: Results from the randomized evaluation of long-term anticoagulation therapy (re-ly) randomized
14.
M AN U
trial. Circulation. 2012;126:343-348
Sherwood MW, Douketis JD, Patel MR, Piccini JP, Hellkamp AS, Lokhnygina Y, Spyropoulos AC, Hankey GJ, Singer DE, Nessel CC, Mahaffey KW, Fox KA, Califf RM, Becker RC. Outcomes of temporary interruption of rivaroxaban compared with
TE D
warfarin in patients with nonvalvular atrial fibrillation: Results from the rivaroxaban once daily, oral, direct factor xa inhibition compared with vitamin k antagonism for prevention of stroke and embolism trial in atrial fibrillation (rocket af). Circulation. 2014;129:1850-
15.
EP
1859
Marijon E, Le Heuzey JY, Connolly S, Yang S, Pogue J, Brueckmann M, Eikelboom J,
AC C
Themeles E, Ezekowitz M, Wallentin L, Yusuf S. Causes of death and influencing factors in patients with atrial fibrillation: A competing-risk analysis from the randomized evaluation of long-term anticoagulant therapy study. Circulation. 2013;128:2192-2201
16.
Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, Pogue J, Reilly PA, Themeles E, Varrone J, Wang S, Alings M, Xavier D, Zhu J, Diaz R, Lewis
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF BS, Darius H, Diener HC, Joyner CD, Wallentin L. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-1151 17.
Granger CB, Alexander JH, McMurray JJ, Lopes RD, Hylek EM, Hanna M, Al-Khalidi
RI PT
HR, Ansell J, Atar D, Avezum A, Bahit MC, Diaz R, Easton JD, Ezekowitz JA, Flaker G, Garcia D, Geraldes M, Gersh BJ, Golitsyn S, Goto S, Hermosillo AG, Hohnloser SH, Horowitz J, Mohan P, Jansky P, Lewis BS, Lopez-Sendon JL, Pais P, Parkhomenko A,
fibrillation. N Engl J Med. 2011;365:981-992
Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, Breithardt G, Halperin
M AN U
18.
SC
Verheugt FW, Zhu J, Wallentin L. Apixaban versus warfarin in patients with atrial
JL, Hankey GJ, Piccini JP, Becker RC, Nessel CC, Paolini JF, Berkowitz SD, Fox KA, Califf RM. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891
Hylek EM, Held C, Alexander JH, Lopes RD, De Caterina R, Wojdyla DM, Huber K,
TE D
19.
Jansky P, Steg PG, Hanna M, Thomas L, Wallentin L, Granger CB. Major bleeding in patients with atrial fibrillation receiving apixaban or warfarin: The aristotle trial
EP
(apixaban for reduction in stroke and other thromboembolic events in atrial fibrillation): Predictors, characteristics, and clinical outcomes. Journal of the American College of
20.
AC C
Cardiology. 2014;63:2141-2147
Camm AJ, Lip GY, De Caterina R, Savelieva I, Atar D, Hohnloser SH, Hindricks G, Kirchhof P. 2012 focused update of the esc guidelines for the management of atrial fibrillation: An update of the 2010 esc guidelines for the management of atrial fibrillation. Developed with the special contribution of the european heart rhythm association. European heart journal. 2012;33:2719-2747
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF 21.
Guidelines for pharmacotherapy of atrial fibrillation (jcs 2013). Circulation journal : official journal of the Japanese Circulation Society. 2014;78:1997-2021
22.
Crowther M, Crowther MA. Antidotes for novel oral anticoagulants: Current status and
23.
RI PT
future potential. Arteriosclerosis, thrombosis, and vascular biology. 2015;35:1736-1745 Lip GY, Skjoth F, Nielsen PB, Larsen TB. Non-valvular atrial fibrillation patients with none or one additional risk factor of the cha2ds2-vasc score. A comprehensive net
SC
clinical benefit analysis for warfarin, aspirin, or no therapy. Thrombosis and haemostasis. 2015;114:826-834
Lip GY, Skjoth F, Rasmussen LH, Larsen TB. Oral anticoagulation, aspirin, or no
M AN U
24.
therapy in patients with nonvalvular af with 0 or 1 stroke risk factor based on the cha2ds2-vasc score. Journal of the American College of Cardiology. 2015;65:1385-1394 25.
Dlott JS, George RA, Huang X, Odeh M, Kaufman HW, Ansell J, Hylek EM. National
TE D
assessment of warfarin anticoagulation therapy for stroke prevention in atrial fibrillation.
AC C
EP
Circulation. 2014;129:1407-1414
Resubmit 021616 text only
ACCEPTED MANUSCRIPT
Table 1 – Classification of Death by the Clinical Endpoint Committee
Categories of Causes of Death
2. Malignancy
TE D
Infection Suicide Accidental/traumatic Hepatobiliary Renal Other
EP
3. Other a. b. c. d. e. f.
M AN U
SC
RI PT
1. Cardiovascular a. Fatal Bleeding i. Intracranial hemorrhage ii. Extracranial bleeding b. Non-bleeding cardiovascular deaths i. Ischemic stroke ii. Systemic arterial embolic event iii. Congestive heart failure or cardiogenic shock iv. Directly related to coronary revascularization v. Dysrhythmia vi. Pulmonary Embolism vii. Sudden or unwitnessed death viii. Atherosclerotic vascular disease (excluding coronary artery disease) ix. Other
Relationship of Death to Bleeding
AC C
1. Fatal bleeding. Bleeding event led directly to death within 7 days. Cause of death must be either intracranial or extracranial bleeding. 2. Bleeding contributed to death. Bleeding was part of a causal chain resulting in death within 30 days.
Death unrelated to a bleeding event.
AJM submission 16 Feb 2016
1
ACCEPTED MANUSCRIPT
Table 2. Causes of Death by Treatment Group
65 (0.33) 53 (0.27) 12 (0.06)
35 (0.18) 30 (0.15) 5 (0.03)
1543 (2.66) 749 (1.29) 390 (0.67) 146 (0.25) 51 (0.09) 25 (0.04) 17 (0.03) 12 (0.02) 8 (0.01) 145 (0.25) 271 (0.47) 72 (0.12) 32 (0.06) 22 (0.04) 145 (0.25) 410 (0.71) 256 (0.44) 25 (0.04) 21 (0.04) 108 (0.19)
546 (2.83) 270 (1.40) 143 (0.74) 47 (0.24) 15 (0.08) 8 (0.04) 5 (0.03) 4 (0.02) 3 (0.02) 51 (0.26) 84 (0.44) 18 (0.09) 5 (0.03) 5 (0.03) 56 (0.29) 144 (0.75) 92 (0.48) 10 (0.05) 8 (0.04) 34 (0.18)
AJM submission 16 Feb 2016
2
P-Value LD edoxaban Warfarin vs. (N=7034) HD edoxaban n (%/year) 0.082 737 (3.80) 0.013 527 (2.71)
M AN U
SC
RI PT
124 (0.21) 98 (0.17) 26 (0.04)
AC C
Non-bleeding cardiovascular deaths Sudden cardiac death Heart failure or shock Ischemic stroke Dysrhythmia Atherosclerotic vascular disease Pulmonary embolism Directly related to revascularization Systemic arterial embolism Other cardiovascular Malignancy Lung Pancreatic Small or large bowel All other malignancies Non-cardiovascular/non-malignancy Infection Accidental / traumatic Renal All other
HD edoxaban (N=7035) n (%/year) 773 (3.99) 530 (2.74)
TE D
Fatal Bleeding Intracranial hemorrhage Extracranial bleeding
Warfarin (N=7036) n (%/year) 839 (4.35) 611 (3.17)
EP
All deaths Cardiovascular
All Patients (N=21105) n (%/year) 2349 (4.05) 1668 (2.87)
495 (2.56) 247 (1.28) 129 (0.67) 44 (0.23) 16 (0.08) 5 (0.03) 5 (0.03) 3 (0.02) 0 46 (0.24) 94 (0.49) 29 (0.15) 14 (0.07) 9 (0.05) 42 (0.22) 149 (0.77) 95 (0.49) 10 (0.05) 4 (0.02) 40 (0.21)
P-value Warfarin vs. LD edoxaban 0.006 0.008
0.003 0.012 0.099
24(0.12) 15 (0.07) 9 (0.05)
<0.001 <0.001 0.50
0.10 0.29 0.37 0.74 0.87 0.41 0.48 0.75 -0.59 0.47 0.12 0.049 0.30 0.15 0.81 0.86 0.99 0.25 0.49
502 (2.59) 232 (1.20) 118 (0.61) 55 (0.28) 20 (0.10) 12 (0.06) 9 (0.05) 3 (0.02) 5 (0.03) 48 (0.25) 93 (0.48) 25 (0.13) 13 (0.07) 8 (0.04) 47 (0.24) 117 (0.60) 69 (0.36) 5 (0.03) 9 (0.05) 34 (0.18)
0.13 0.073 0.10 0.46 0.41 0.39 0.30 0.70 0.50 0.71 0.54 0.31 0.072 0.42 0.35 0.080 0.061 0.20 0.83 0.97
ACCEPTED MANUSCRIPT
36 (0.19) 101 (0.52)
24 (0.12) 59 (0.30)
0.12 0.001
AC C
EP
TE D
M AN U
SC
RI PT
Relationship of Deaths to Bleeding Bleeding contributed to death 90 (0.16) Fatal bleeding or bleeding that 214 (0.37) contributed to death % rates shown are annualized Kaplan-Meier estimates
AJM submission 16 Feb 2016
3
30 (0.15) 54 (0.28)
0.44 <0.001
ACCEPTED MANUSCRIPT
Figure 1 # pts 150
A. Sudden Cardiac Death 270 247 232
143
100
129
118
SC
200
50
P = NS for each pairwise comparisons
0
M AN U
100
P = NS for each pairwise comparisons
0
HD Edox
LD Edox
TE D
Warfarin
EP
# pts C. Fatal Ischemic Stroke 75
# pts 75
55
AC C
50
B. Fatal Heart Failure
RI PT
# pts 300
47
44
25
50
Warfarin
HD Edox
D. Fatal Bleeding 65
P<0.001 P=0.003
35
25
24
P = NS for each pairwise comparisons
0
LD Edox
0 Warfarin
HD Edox
LD Edox
Warfarin
HD Edox
LD Edox 1
ACCEPTED MANUSCRIPT
Figure 2 # patients
RI PT
600
Non-fatal
506
Fatal
494
400
450
337* 328
TE D
300
EP
200 102
100
AC C
155 79* 49
30 Warfarin
60* 45
53
0
455†
M AN U
SC
500
HD Edox
15 LD Edox
Intracranial Hemorrhage
9
5 12 Warfarin
HD Edox
LD Edox
Extracranial Major Bleeding 2
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3
ACCEPTED MANUSCRIPT
Highlights
•
An analysis of the causes of death was performed in a clinical trial of NOAC vs warfarin
RI PT
in patients with AF •
Edoxaban reduced mortality primarily by reducing fatal intracranial bleeding
•
Edoxaban also reduced non-fatal bleeding, which indirectly may result in fewer
AC C
EP
TE D
M AN U
SC
subsequent deaths
ACCEPTED MANUSCRIPT
Supplemental Appendix
AC C
EP
TE D
M AN U
SC
RI PT
Appendix S1. Clinical Endpoint Committee Classification of Death................................... 27-31
AJM submission 16 Feb 2016
1
ACCEPTED MANUSCRIPT
Table S1 - Clinical Endpoint Committee Classification of Death
Classification of Death
RI PT
Death will be classified as Cardiovascular, Malignancy, or Non-cardiovascular/Non-
malignancy. The cause of death is determined by the principal condition that caused the death, not the
or a non-cardiovascular cause can be clearly shown.
SC
immediate mode of death. All deaths will be assumed to be cardiovascular in nature unless a malignant
Cardiovascular death is defined as death due to documented cardiovascular cause, including
M AN U
deaths due to bleeding. Causes of cardiovascular deaths include, but are not limited to, deaths resulting from atherosclerotic coronary heart disease (acute myocardial infarction, sudden cardiac death, nonsudden death with gradually worsening cardiac symptoms, unwitnessed death without clear alternate cause, procedural death related to cardiac surgery or coronary angiography), atherosclerotic vascular
TE D
disease (cerebrovascular disease including stroke and hemorrhage, aortic, mesenteric, renovascular, peripheral arterial disease, or complication of a non-coronary vascular procedure), other cardiovascular (pulmonary embolism, endocarditis, congestive heart failure, valvular heart disease, arrhythmia), and
EP
deaths due to bleeding.
Deaths due to malignancy will include deaths that are directly a consequence of a malignancy,
AC C
such as a brain tumor that causes herniation, coma, and respiratory arrest. Deaths due to malignancy will be further subclassified by organ system and timing of diagnosis (before vs. after randomization). Examples of non-cardiovascular/non-malignancy deaths include those caused primarily by infection, pulmonary, gastrointestinal, accidental, renal, trauma, or non-cardiovascular organ system failure.
AJM submission 16 Feb 2016
2
ACCEPTED MANUSCRIPT
For all deaths, the relationship of the death to bleeding and malignancy will be adjudicated as follows: Relationship to Bleeding (categories are mutually exclusive)
RI PT
1. Fatal bleeding – death in which a bleeding event directly led to death within 7 days. Examples of fatal bleeding events are an intracranial hemorrhage that led to herniation of the brain and death within 24 hours, and a massive gastrointestinal hemorrhage that results in
SC
shock, hemodynamic collapse, and death. If a bleeding event is considered fatal, then the cause of death must be either intracranial or non-intracranial bleeding.
M AN U
2. Bleeding contributed to death – a death in which a bleeding event was part of a causal chain of medical events that ultimately led to death within 30 days of the bleed, but bleeding was not directly and/or immediately related to subject’s death. An example of bleeding contributing to death is a large retroperitoneal bleed that leads to surgical evacuation,
TE D
development of a subsequent abscess in the area of bleeding that leads to sepsis, multiorgan failure and death 10 days after the onset of bleeding. If bleeding has contributed to death (but the bleeding was not categorized as “fatal”), then the cause of death must be recorded as
EP
something other than intracranial / non-intracranial bleeding. 3. Deaths unrelated to a bleeding event – The case of death was unrelated to bleeding, either
AC C
because there was no clinical significant bleeding in the month prior to death or the bleeding event did not contribute to the subject’s death. An example of a death unrelated to bleeding is an episode of guaiac positive stools in a patient who dies of post-obstructive pneumonia due to lung cancer. In these cases, the cause of death cannot be intracranial / non-intracranial bleeding.
AJM submission 16 Feb 2016
3
ACCEPTED MANUSCRIPT
Relationship to Malignancy 1. Death directly related to malignancy – Death in which the mode of death can be attributed to the direct effects of a malignancy. In such cases, the cause of death adjudicated by the CEC
RI PT
must be malignancy.
2. Death due to a consequence related to malignancy. This would include deaths due to other processes (e.g., infection in a patient who becomes septic and neutropenic due to acute
SC
leukemia) that are a known complication of the malignancy. The underlying malignancy should be on the causal pathway leading to death, but not the immediate cause of death. In
M AN U
such cases, the cause of death adjudicated by the CEC cannot be malignancy, but instead should be the other process (e.g., infection).
3. Death not related to a malignancy. Either no malignancy has been diagnoses or the malignancy that is present was not related to the cause of death. The cause of death must be
TE D
something other than malignancy.
EP
Sudden Cardiovascular Death
Sudden CV death is defined as a sudden, unexpected death that was either:
AC C
a) witnessed, occurring within 60 min from the onset of new symptoms, in the absence of a clear cause other than cardiovascular; or
b) unwitnessed, within 24 hours of being observed alive, in the absence of pre-existing progressive circulatory failure or other non-cardiovascular causes of death;
AJM submission 16 Feb 2016
4
ACCEPTED MANUSCRIPT
Non-sudden Cardiovascular Death This category refers to a patient who had symptoms of a cardiovascular nature and had gradual deterioration prior to death. It includes all patients with CV death who do not meet criteria for sudden
RI PT
death or unwitnessed CV death.
Un-witnessed CV Death
M AN U
known other major causes of death are identified.
Illustrative examples Cause of Death Cardiovascular death:
A subject found dead at home without known major non-cardiovascular
TE D
•
SC
Death that occurred unexpectedly, without patient being seen within 24 hours, and for which no
illness for which death was expected •
A subject with renal failure on hemodialysis who dies following a recurrent myocardial
EP
infarction
A subject who dies following aortic dissection
•
A subject who dies of complications of endocarditis
•
AC C
•
A subject who dies of complications related to coronary artery bypass surgery or percutaneous coronary intervention
•
A subject who dies within 24 hours since the onset of neurologic symptoms compatible with a stroke
AJM submission 16 Feb 2016
5
ACCEPTED MANUSCRIPT
•
A subject who dies suddenly and autopsy shows either a cerebral infarction or a cerebral hematoma (Note: the latter event would also be classified as a fatal bleeding event)
•
A subject who dies of progressive hypotension in the setting of a severe gastrointestinal
RI PT
hemorrhage. (Note: this event would also be classified as a fatal bleed).
Malignancy death:
A subject with an unresectable tracheal carcinoma that causes airway compromise and
SC
•
respiratory arrest
A subject with metastatic breast cancer that causes a recurrent malignant pericardial effusion and pericardial tamponade.
M AN U
•
Non-cardiovascular/Non-malignancy death:
A subject with progressive interstitial pulmonary fibrosis with home hospice care who is
TE D
•
found dead at home
A subject with severe pneumonia and hypoxia who has a bradycardic arrest
•
A subject who dies of multiple trauma in a motor vehicle accident.
•
A subject with stage IV prostate cancer with widespread metastases who dies of hepatic
AC C
EP
•
encephalopathy. (Note: this event would also be classified as a death due to a consequence of
a malignancy, hepatic failure due to metastatic disease)
AJM submission 16 Feb 2016
6
S0002-9343(16)30246-7
DOI:
10.1016/j.amjmed.2016.02.028
Reference:
AJM 13419
To appear in:
The American Journal of Medicine
Received Date: 16 February 2016 Revised Date:
17 February 2016
Accepted Date: 18 February 2016
Please cite this article as: Giugliano RP, Ruff CT, Wiviott SD, Nordio F, Murphy SA, Kappelhof JA, Shi M, Mercuri M, Antman EM, Braunwald E, Mortality in Patients With Atrial Fibrillation Randomized to Edoxaban or Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial, The American Journal of Medicine (2016), doi: 10.1016/j.amjmed.2016.02.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Mortality in Patients With Atrial Fibrillation Randomized to Edoxaban or
RI PT
Warfarin: Insights from the ENGAGE AF-TIMI 48 Trial
Robert P. Giugliano MD, SM1, Christian T. Ruff, MD, MPH1, Stephen D. Wiviott, MD1, Francesco Nordio, PhD1, Sabina A. Murphy, MPH1, Johannes AN Kappelhof, BSc, Minggao
SC
Shi, PhD2, Michele Mercuri, MD2, Elliott M. Antman, MD1, and Eugene Braunwald, MD1
M AN U
From the 1TIMI Study Group, Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA and 2Daiichi-Sankyo Pharma Development, Edison, NJ
Conflicts of Interest
TE D
Running Title: Mortality with Edoxaban vs Warfarin in AF
EP
Dr. Giugliano reports receiving consulting fees from Bristol Myers Squibb, Daiichi Sankyo, Janssen Pharmaceuticals, Merck, Pfizer, and Portola; lecture fees from Daiichi Sankyo and Merck; and grant support through his institution from Daiichi Sankyo, Merck, Johnson & Johnson, Sanofi, and AstraZeneca.
AC C
Dr. Ruff reports receiving consulting fees from Daiichi Sankyo, Bristol-Myers, Squibb, and Boehringer Ingelheim and grant support through his institution from Daiichi Sankyo. Dr. Wiviott reports receiving consulting fees from AstraZeneca, Bristol-Myers Squibb, Eisai, Arena Pharmaceuticals, Eli Lilly, Daiichi Sankyo, Aegerion, AngelMed, Janssen Pharmaceuticals, Xoma, ICON Clinical Research, and Boston Clinical Research Institute and grant support through his institution from AstraZeneca, Bristol-Myers Squibb, Eisai, Arena Pharmaceuticals, Merck, Eli Lilly, Daiichi Sankyo, and Sanofi. Drs. Nordio and Antman, and Ms. Murphy report receiving grant support through their institution from Daiichi Sankyo. Mr. Kappelhof and Drs. Shi, and Mercuri report being employees of Daiichi Sankyo. Dr. Mercuri also reports holding a pending patent related to the clinical properties of edoxaban.
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF
RI PT
Dr. Braunwald reports receiving consulting fees from Sanofi, Genzyme, Amorcyte, the Medicines Company, and Cardiorentis; lecture fees from Eli Lilly, Menarini, Medscape, and Bayer HealthCare; and grant support through his institution from Daiichi Sankyo, AstraZeneca, Johnson & Johnson, GlaxoSmithKline, Bristol-Myers Squibb, Beckman Coulter, Roche Diagnostics, Pfizer, Merck, and Sanofi. He also reports serving as an unpaid consultant for Merck and providing uncompensated lectures for Merck and CVRx.
Text word count: 2910
AC C
EP
TE D
Total word count: 4681
M AN U
Address for Correspondence Robert P. Giugliano, MD, SM TIMI Study Group 350 Longwood Avenue, 1st Floor Offices Boston, MA 02115 TEL: 617-278-0145 FAX: 617-734-7329 EMAIL: [email protected]
SC
All authors had access to the data and a role in writing the manuscript.
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF Abstract Background: When compared with warfarin, edoxaban significantly reduced cardiovascular mortality in the ENGAGE AF-TIMI 48 trial. We studied the possible reasons
RI PT
leading to this reduction.
Methods: ENGAGE AF-TIMI 48 was a double-blind, double-dummy comparison of warfarin with 2 regimens of once-daily edoxaban in 21,105 patients with atrial fibrillation
SC
followed for 2.8 years (median). Causes of deaths in the intention-to-treat population were
classified as cardiovascular (including fatal bleeding and ischemic stroke), malignancy, or non-
M AN U
cardiovascular/non-malignancy by an independent, blinded, clinical endpoint committee . Deaths also were adjudicated as directly due to bleeding (i.e., fatal), or bleeding contributing to death, or neither.
Results: There were 839 total deaths (4.35%/yr) in the warfarin arm, compared with 773
TE D
(3.99%/yr, p=0.08) with the higher-dose edoxaban regimen, and 737 (3.80%/yr, p=0.006) with the lower-dose edoxaban regimen. No significant differences between treatments were observed in: 1) any of the 3 most common causes of cardiovascular death (sudden cardiac, heart failure,
EP
ischemic stroke), 2) fatal malignancies, 3) other non-cardiovascular death. There were 124 fatal bleeds, 65 with warfarin, significantly fewer with the higher-dose (n=35, p=0.003) and lower-
AC C
dose (n=24, p<0.001) edoxaban regimens. There were 101 bleeding events with warfarin that were either fatal or that contributed to death. There were significantly fewer with the higher-dose (n=59, P=0.001) and lower-dose (n=54, P<0.001) edoxaban regimens. Conclusions: Fewer total and cardiovascular deaths were observed with edoxaban as
compared to warfarin in the ENGAGE AF-TIMI 48 trial, and this predominantly resulted from significantly lower rate of major bleeding with edoxaban. Edoxaban reduces mortality both
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF directly (less fatal bleeding) and indirectly (fewer bleeding-related complications and
RI PT
interruptions in therapy after non-fatal bleeding).
Abstract word count: 266
AC C
EP
TE D
M AN U
SC
Keywords: death, hemorrhage, bleeding, anticoagulant, factor Xa inhibitor, edoxaban, warfarin
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF A systematic review of population-based studies in patients with atrial fibrillation from 21 Global Burden of Disease regions demonstrated that the overall incidence, prevalence, and mortality attributed to atrial fibrillation has increased from 1990 to 2010. This is largely due to
RI PT
aging of the world population, and adds to a growing and significant worldwide public health burden.1 While a meta-analysis of 6 trials involving 2900 patients with non-valvular atrial
fibrillation randomized to warfarin or placebo demonstrated a 26% reduction in total mortality
SC
with warfarin2, there was no significant reduction in total mortality with warfarin compared to antiplatelet therapy in 8 trials involving 3647 patients2, nor in a subsequent trial of 6706 patients
M AN U
randomized to warfarin or dual antiplatelet therapy (hazard ratio (HR) = 0.99).3 Residual mortality risk in patients with atrial fibrillation anticoagulated with warfarin has been demonstrated4, and may be due to limitations in efficacy or increased mortality related to sideeffects of warfarin, or both. For example, warfarin increases the risk of major hemorrhage
TE D
relative to no antithrombotic or aspirin, in particular intracranial hemorrhage, which is frequently fatal or severely disabling, even despite rapid reversal5 of the anticoagulant effect with prothrombin complex concentrates.6
EP
In contrast, in a meta-analysis of four large randomized warfarin-controlled trials involving 71,683 patients with atrial fibrillation, we reported that when compared to warfarin,
AC C
the non-vitamin K oral anticoagulants (NOACs) reduced total mortality by 10% and did not increase major bleeding as compared to warfarin.7 In addition, NOACs demonstrated a consistent and large (approximately 50%) reduction in intracranial hemorrhage compared to warfarin across these four trials, whereas the observed reduction in thromboembolic events was modest and variable across these trials.7
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF In the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation– Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial, the oral factor Xa inhibitor edoxaban, as compared to well-managed warfarin, was non-inferior for prevention of
RI PT
stroke and systemic embolic events, while reducing cardiovascular mortality, major bleeding, and intracranial hemorrhage.8 However, data regarding the causes of death and the relationship of bleeding to death with NOACs as compared to warfarin are sparse. Hence, we analyzed
SC
death in the ENGAGE AF-TIMI 48 trial to explore the factors that contributed to the improved
M AN U
survival observed with edoxaban.
Methods
The ENGAGE AF-TIMI 48 trial was a double-blind, double-dummy, randomized controlled trial in 21,105 patients with atrial fibrillation. Details of the trial design have been
TE D
described.9 Briefly, patients with atrial fibrillation in the prior 12 months with a CHADS2 score10 >2 were randomized to warfarin titrated to an international normalized ratio of 2-3, a higher-dose regimen of edoxaban, or a lower-dose edoxaban regimen. The doses of edoxaban
EP
were 60 mg and 30 mg once daily in the higher-dose and lower-dose edoxaban regimen arms respectively. In both edoxaban arms the dose was reduced by 50% if patients had any one of the
AC C
following at any time during the trial: creatinine clearance <50 mg/dL, body weight < 60 kg, concomitant use of verapamil, quinidine, or dronedarone. Major exclusion criteria were high risk for bleeding, creatinine clearance < 30 mL/min, need for dual antiplatelet therapy, or acute coronary syndrome or stroke within 30 days. The primary efficacy endpoint was the combination of stroke or systemic embolic events. The principal safety endpoint was major bleeding as defined by the International Society of Thrombosis and Haemostasis.11 Patients were followed
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF for a median of 2.8 years (interquartile range 2.4 to 3.2 years). The protocol and amendments were approved by the ethics committee at each participating center. All the patients provided written informed consent.
RI PT
An independent clinical end-point committee, blinded to treatment assignment, adjudicated all deaths and suspected cerebrovascular events, systemic embolic events,
myocardial infarctions, and bleeding. In the adjudication of deaths, the clinical endpoint
SC
committee determined the primary cause of death, categorized as shown in Table 1 (see
Appendix 1 for detailed definitions). Fatal bleeding events (whether intra- or extra-cranial) were
M AN U
classified as cardiovascular deaths. In addition the clinical endpoint committee determined whether: 1) deaths were directly caused by bleeding (i.e., a fatal bleeding event), 2) bleeding contributed to, but was not the direct cause of death within the next 30 days (i.e., a death in which a bleeding event was part of a causal chain of medical events that ultimately led to death
TE D
within 30 days of the bleed, but bleeding was not directly and/or immediately related to subject’s death), and 3) death was not associated with either of the above (See Table S1 for details). The present analysis was conducted using the intention-to-treat principle; it included all
EP
patients randomized and counted all deaths and bleeding events, whether on or off study drug, that occurred between randomization and the end of the study drug treatment period. The one
AC C
exception was the analysis of consequences of a major bleed while on study drug, which was conducted in the on-treatment population. Annualized event rates were calculated as KaplanMeier estimates. P-values <0.05 without adjustment for multiple comparisons, were considered nominally significant.
Role of the Funding Source
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF The ENGAGE AF-TIMI 48 trial was supported by a research grant from Daiichi Sankyo Pharma Development, who participated in the design and conduct of this trial, and in the preparation of
RI PT
this manuscript.
Results
A total of 2349 deaths occurred, of which 1668 (71.0%) were cardiovascular, 271
SC
(11.5%) were due to malignancy, and 410 (17.5%) were due other causes (Table 2). Of these 2349 deaths, 839 deaths (4.35%/year) occurred in patients randomized to warfarin. Fewer deaths
M AN U
occurred with the higher-dose edoxaban regimen (773 deaths, 3.99%/year) and with the lowerdose edoxaban regimen (737 deaths, 3.80%/year) as compared to warfarin. There were a total of 1668 cardiovascular deaths. Of these, 611 (3.17%/year) occurred in the warfarin group. Both edoxaban dose regimens significantly reduced cardiovascular
TE D
mortality: higher-dose edoxaban regimen 530 cardiovascular deaths, 2.74%/year; lower-dose edoxaban regimen 527 cardiovascular deaths, 2.71%/year. There were no significant differences among the treatment groups between deaths due to either malignancy (warfarin 84 (0.44%/year);
EP
higher-dose edoxaban regimen 94 (0.49%/year), lower-dose edoxaban regimen 93 (0.48%/year)) or other non-cardiovascular/non-malignant causes of death, although numerically fewer occurred
AC C
with the lower-dose edoxaban regimen (warfarin 144 (0.75%/year); higher-dose edoxaban regimen 149 (0.77%/year), lower-dose edoxaban regimen 117 (0.60%/year)).
Causes of Cardiovascular Death The four most common causes of cardiovascular deaths were sudden cardiac death (749 (1.29%/year); 44.9% of the 1668 cardiovascular deaths), heart failure (390
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF (0.67%/year); 23.4% of cardiovascular deaths), ischemic stroke (146 (0.25%/year); 8.8% of cardiovascular deaths), and bleeding (124 (0.21%/year); 7.4% of cardiovascular deaths) (Table 2, Figure 1). There were no differences between the treatment groups in the first 3 subcategories of
RI PT
cardiovascular death (sudden cardiac death, heart failure, ischemic stroke), but there was a
significant excess of fatal bleeding events with warfarin (n=65 (0.33%/year)) as compared to both the higher-dose edoxaban regimen (35 (0.18%/year); HR=0.54, p=0.003 vs. warfarin) and
SC
the lower-dose edoxaban regimen (24 (0.12%/year)); HR=0.37, p<0.001 vs. warfarin). The rates of fatal bleeding were low and not dissimilar with the higher-dose edoxaban regimen
M AN U
(0.18%/year) and the lower-dose edoxaban regimen (0.12%/year, p=0.24). The case fatality rates following a major bleeding event were 11.3% with warfarin, 7.7% with the higher-dose edoxaban regimen (p=0.064) and 8.3% with the lower-dose edoxaban regimen (p=0.20).
TE D
Deaths Related to Bleeding
Of the 124 (0.21%/year) fatal bleeding events, 98 (0.17%/year, 79% of the fatal bleeds) were due to intracranial hemorrhage, even though intracranial hemorrhage accounted for fewer
EP
than 20% (294/1592) of the major bleeding events. This can be explained by the nearly 20-fold greater fatality rate after intracranial hemorrhage (33%) as compared to extracranial bleeding
AC C
(2%, Figure 2). There were 155 (0.97%/year) intracranial hemorrhages in the warfarin group. Fewer intracranial hemorrhages occurred in the higher-dose edoxaban regimen group (n=79 (0.50%/year); p<0.001), and fewer yet in the lower-dose edoxaban regimen group (n=60 (0.37%/year); p<0.001 vs. both the higher-dose edoxaban regimen and vs. warfarin). There was a trend toward more major extracranial bleeding events in the warfarin group as compared to the higher-dose edoxaban regimen group (506 vs. 455, (3.16 vs. 2.87%/year) p=0.09), while there
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF were significantly fewer with the lower-dose edoxaban regimen (337 (2.07%/year), p<0.001 compared to either warfarin or the higher-dose edoxaban regimen). In addition to the significant reduction in fatal bleeding events with edoxaban, there also
RI PT
was a numerical, but statistically non-significant, reduction in bleeding events that were
adjudicated as contributing to death in the edoxaban groups (warfarin 36 (0.23%/year); higherdose edoxaban regimen 24 (0.15%/year), p=0.12 higher-dose edoxaban regimen vs. warfarin;
SC
lower-dose edoxaban regimen 30 (0.18%/year), p=0.44 lower-dose edoxaban regimen vs.
warfarin). In contrast with fatal bleeding events, only 17% (16 of 96) of major bleeding events
M AN U
that contributed to death were due to intracranial hemorrhage. The total number of bleeding events that were either fatal or contributed to death was 40-45% lower (relatively) with edoxaban as compared with warfarin (Figure 3). In the warfarin group there were 81 (0.42%/year) excess cardiovascular deaths as compared to the higher-dose edoxaban regimen group, of which 52%
TE D
(42/81) were related to bleeding that was either fatal or contributed to death. Similarly with the lower-dose edoxaban regimen, 54% (47/84) of the excess cardiovascular deaths with warfarin as
EP
compared to the lower-dose edoxaban regimen were related to bleeding.
Consequences of Non-Fatal Major Bleeding While on Study Drug
AC C
Patients on study drug who experienced a major non-fatal bleeding event that did not
contribute to death had nearly twice the relative risk of dying before the end of the trial compared to those who did not have major bleeding during the trial (7.0 vs. 4.1% per year, relative risk unadjusted 1.9, p<0.001). Of the 1080 patients (2.34%/year) on study drug who survived a major bleed (463 (3.03%/year) warfarin, 385 (2.53%/year) higher-dose edoxaban regimen, 232 (1.48%/year) lower-dose edoxaban regimen), study drug was interrupted in 91%
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF for at least 4 days, with similar rates (91% warfarin, 90% higher-dose edoxaban regimen, 91% lower-dose edoxaban regimen) across the 3 treatment groups. Of these 979 who interrupted study drug, 53% were re-challenged with study drug, 37% received open-label anticoagulant, while
RI PT
10% did not receive any further anticoagulant (neither study drug nor open-label anticoagulant) for the remainder of the trial.
Patients who survived a major bleed, but interrupted anticoagulation for > 4 days
SC
experienced high rates of subsequent major adverse cardiac events (7.9% annualized rate of major adverse cardiac events, defined as cardiovascular death, myocardial infarction, stroke or
M AN U
systemic embolic events) and of all-cause death (7.1%/year), regardless of randomized treatment. In the warfarin group, 423 patients (2.72%/year) interrupted anticoagulation after a non-fatal major bleed that did not contribute to death. Since edoxaban reduced major bleeding as compared to warfarin, the corresponding number of patients who interrupted study drug
TE D
following major bleeding were fewer with the higher-dose edoxaban regimen (345 (2.23%/year); p=0.004 vs. warfarin) and, fewer yet with the lower-dose edoxaban regimen (211 (1.32%/year); p<0.001 vs. warfarin and vs. higher-dose edoxaban regimen). Following interruption of study
EP
drug due to a non-fatal major bleeding event, there were numerically more deaths with warfarin (n=81) than with edoxaban (higher-dose edoxaban regimen 64; p=0.18 vs. warfarin, lower-dose
AC C
edoxaban regimen 42; p=0.001 for lower-dose edoxaban regimen vs. warfarin). Similarly, there were more major adverse cardiac events after study drug interruption due to a non-fatal major bleed with warfarin (n=111) as compared to the higher-dose edoxaban regimen (66; p=0.001 vs. warfarin) and lthe ower-dose edoxaban regimen (58, p<0.001 vs. warfarin).
Discussion
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF In this analysis of causes of total mortality and the relationship between bleeding and causes of death in the ENGAGE AF-TIMI 48 trial, we found that more than half of the reductions in all-cause deaths with edoxaban as compared to warfarin were directly attributable
RI PT
to fewer fatal bleeding events and non-fatal major bleeding that contributed to a death within 30 days. There were no differences between edoxaban and warfarin in deaths due to non-
cardiovascular causes, nor for any of the 3 most common cardiovascular causes of death (sudden
SC
cardiac death, heart failure, ischemic stroke). However, warfarin was associated with
approximately twice the risk of fatal bleeding as compared to edoxaban. Since edoxaban (like
M AN U
other NOACs) reduces intracranial hemorrhage by approximately 50% relative to warfarin7, and intracranial hemorrhage carries a high fatality rate12 (1-in-3 in this study), the single greatest factor accounting for the lower total mortality with edoxaban was a reduction in fatal intracranial hemorrhage (by 43% and 72% with the higher-dose edoxaban regimen and lthe ower-dose
TE D
edoxaban regimen, respectively, compared to warfarin).
In addition to a reduction in fatal bleeding events with edoxaban, there also were fewer non-fatal major bleeding events in patients treated with edoxaban. This latter finding also may
EP
have contributed to the reduction in total mortality with edoxaban in two ways. First, patients who initially survive a major bleeding event may subsequently develop another medical
AC C
complication that ultimately proves fatal, for example, an event that is: a) a direct consequence of the bleed (e.g., hypovolemic shock leading to myocardial or cerebral infarction), b) related to management of the bleeding event (e.g., transfusion- or procedure-related complication), or c) due to an in-hospital complication (e.g., nosocomial infection). Secondly, patients who survive a major bleed often interrupt anticoagulation for several days (91% in this study), and even brief interruptions in anticoagulation place patients with atrial fibrillation at an increase risk of
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF thromboembolic events and death.1 In fact, even brief interruptions of anticoagulation for ≤ 5 days have been associated with increased risk.13, 14 Our finding that the majority of deaths among patients with atrial fibrillation treated with
RI PT
a NOAC or well-managed warfarin are cardiovascular in nature is consistent with those reported by the RE-LY investigators.15 The numerically greater proportion of cardiovascular deaths in ENGAGE AF-TIMI 48 (71%) as compared to RE-LY (61%) may be attributable to the greater
SC
burden of risk factors in the ENGAGE AF-TIMI 48 trial (patients with CHADS2 score 4-6
represented 53% of the ENGAGE AF-TIMI 488 and 33% of the RE-LY16 trial populations). In
M AN U
both trials, there were trends favoring the NOACs over warfarin with relative reductions in total mortality of 8-13%. While there were no differences between the NOACs and warfarin in the most common forms of cardiovascular death (sudden cardiac death, heart failure) in either trial, a significant reduction in vascular deaths (including stroke, peripheral or pulmonary embolism and
TE D
bleeding) was observed with dabigatran as compared to warfarin in RE-LY.15 In the current analysis, we extended these findings to show more specifically that the major explanation for lower mortality with edoxaban was related to the large reductions in bleeding events that were
EP
either fatal or contributed to death within 30 days. Both apixaban and rivaroxaban reduce intracranial hemorrhage by ~50% compared to warfarin, however, similar detailed analyses of
AC C
the causes of death in the large phase 3 trials comparing these two Factor Xa inhibitors with warfarin have not been published. 17, 18,19 We believe that our findings provide a strong rationale to consider a NOAC rather than
warfarin, to reduce bleeding and death, while preserving protection from thromboembolism in patients with atrial fibrillation at moderate to high risk for stroke. Our results provide further support to the recommendations from the European Society of Cardiology20 and the Japanese
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF Circulation Society 21 favoring NOACs over vitamin K antagonists for most patients with atrial fibrillation and two or more risk factors for stroke. With the anticipated availability in the near future of antidotes to Factor Xa and thrombin inhibitors,22 the safety advantage of the NOACs
RI PT
over vitamin K antagonists should be enhanced even further, potentially extending the net clinical benefit with NOACs to patients with even lower risk of stroke.23, 24
We acknowledge some limitations. This analysis was limited to patients who met entry
SC
criteria for and agreed to participate in a randomized clinical trial, and the findings may differ in a general population. Secondly, attributing a single cause of death is challenging despite use of
M AN U
prospective definitions since multiple factors may contribute to death, source data are not always available (e.g., for deaths that occur at non-study institutions), and autopsies are infrequently performed. However, since adjudications of death were unaware of treatment assignment, the likelihood of introducing substantial bias is minimized. Classification of bleeding deaths as
TE D
cardiovascular was arbitrary, but was prespecified in the trial protocol. Nonetheless, the validity of our results still stand if bleeding deaths were considered in a separate category. This analysis does not address the question of how much bleeding should be tolerated to prevent a
patients.
EP
thromboembolic event nor which dose regimen of edoxaban should be selected for individual
AC C
Strengths of our analysis include the independent adjudication of causes of death by an
expert committee blinded to treatment assignment, using a detailed prospectively designed charter. The dataset also represents the largest (21,105 patients) and longest (2.8 year median) follow up with the greatest number of total deaths (n=2349) from a randomized trial in patients with atrial fibrillation conducted to date. Follow-up for vital status was complete for all except one patient, and complete information for key clinical endpoints was available in 99.5% of the
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF total 56,346 patient-years of potential follow-up.8 The median time in therapeutic range was 68.4% [56.5%, 77.4%] among the patients treated with warfarin, which is higher than that reported in prior NOAC vs. warfarin trials16-18 (range 58-66%) and in a recent US assessment
RI PT
involving 138,319 Americans with atrial fibrillation (median 57.5%).25
Conclusions
SC
In patients with atrial fibrillation at moderate to high risk of stroke randomized to edoxaban vs. warfarin in the ENGAGE AF-TIMI 48 trial, the significant reduction in
M AN U
cardiovascular mortality observed with edoxaban was mediated primarily by lower rates of
AC C
EP
TE D
bleeding that were either fatal or contributed to death in the subsequent 30 days.
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF References
1.
Chugh SS, Havmoeller R, Narayanan K, Singh D, Rienstra M, Benjamin EJ, Gillum RF,
RI PT
Kim YH, McAnulty JH, Jr., Zheng ZJ, Forouzanfar MH, Naghavi M, Mensah GA, Ezzati M, Murray CJ. Worldwide epidemiology of atrial fibrillation: A global burden of disease 2010 study. Circulation. 2014;129:837-847
Hart RG, Pearce LA, Aguilar MI. Meta-analysis: Antithrombotic therapy to prevent
SC
2.
stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med.
3.
M AN U
2007;146:857-867
Connolly S, Pogue J, Hart R, Pfeffer M, Hohnloser S, Chrolavicius S, Yusuf S. Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the atrial fibrillation clopidogrel trial with irbesartan for prevention of vascular events (active w):
4.
TE D
A randomised controlled trial. Lancet. 2006;367:1903-1912
Senoo K, Lip GY, Lane DA, Buller HR, Kotecha D. Residual risk of stroke and death in anticoagulated patients according to the type of atrial fibrillation: Amadeus trial. Stroke;
5.
EP
a journal of cerebral circulation. 2015;46:2523-2528 Goldstein JN, Refaai MA, Milling TJ, Jr., Lewis B, Goldberg-Alberts R, Hug BA, Sarode
AC C
R. Four-factor prothrombin complex concentrate versus plasma for rapid vitamin k antagonist reversal in patients needing urgent surgical or invasive interventions: A phase 3b, open-label, non-inferiority, randomised trial. Lancet. 2015;385:2077-2087
6.
Dowlatshahi D, Butcher KS, Asdaghi N, Nahirniak S, Bernbaum ML, Giulivi A, Wasserman JK, Poon MC, Coutts SB. Poor prognosis in warfarin-associated intracranial
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF hemorrhage despite anticoagulation reversal. Stroke; a journal of cerebral circulation. 2012;43:1812-1817 7.
Ruff CT, Giugliano RP, Braunwald E, Hoffman EB, Deenadayalu N, Ezekowitz MD,
RI PT
Camm AJ, Weitz JI, Lewis BS, Parkhomenko A, Yamashita T, Antman EM. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: A meta-analysis of randomised trials. Lancet. 2014;383:955-962
Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL,
SC
8.
Ezekowitz MD, Weitz JI, Spinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M,
M AN U
Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M, Antman EM. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2013;369:2093-2104 9.
Ruff CT, Giugliano RP, Antman EM, Crugnale SE, Bocanegra T, Mercuri M, Hanyok J, Patel I, Shi M, Salazar D, McCabe CH, Braunwald E. Evaluation of the novel factor xa
TE D
inhibitor edoxaban compared with warfarin in patients with atrial fibrillation: Design and rationale for the effective anticoagulation with factor xa next generation in atrial fibrillation-thrombolysis in myocardial infarction study 48 (engage af-timi 48). American
10.
EP
heart journal. 2010;160:635-641
Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of
AC C
clinical classification schemes for predicting stroke: Results from the national registry of atrial fibrillation. JAMA. 2001;285:2864-2870
11.
Schulman S, Kearon C. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005;3:692-694
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF 12.
Giugliano RP, Giraldez RR, Morrow DA, Antman EM, Gibson CM, Mohanavelu S, Murphy SA, McCabe CH, Braunwald E. Relations between bleeding and outcomes in
heart journal. 2010;31:2103-2110 13.
RI PT
patients with st-elevation myocardial infarction in the extract-timi 25 trial. European
Healey JS, Eikelboom J, Douketis J, Wallentin L, Oldgren J, Yang S, Themeles E,
Heidbuchel H, Avezum A, Reilly P, Connolly SJ, Yusuf S, Ezekowitz M. Periprocedural
SC
bleeding and thromboembolic events with dabigatran compared with warfarin: Results from the randomized evaluation of long-term anticoagulation therapy (re-ly) randomized
14.
M AN U
trial. Circulation. 2012;126:343-348
Sherwood MW, Douketis JD, Patel MR, Piccini JP, Hellkamp AS, Lokhnygina Y, Spyropoulos AC, Hankey GJ, Singer DE, Nessel CC, Mahaffey KW, Fox KA, Califf RM, Becker RC. Outcomes of temporary interruption of rivaroxaban compared with
TE D
warfarin in patients with nonvalvular atrial fibrillation: Results from the rivaroxaban once daily, oral, direct factor xa inhibition compared with vitamin k antagonism for prevention of stroke and embolism trial in atrial fibrillation (rocket af). Circulation. 2014;129:1850-
15.
EP
1859
Marijon E, Le Heuzey JY, Connolly S, Yang S, Pogue J, Brueckmann M, Eikelboom J,
AC C
Themeles E, Ezekowitz M, Wallentin L, Yusuf S. Causes of death and influencing factors in patients with atrial fibrillation: A competing-risk analysis from the randomized evaluation of long-term anticoagulant therapy study. Circulation. 2013;128:2192-2201
16.
Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, Pogue J, Reilly PA, Themeles E, Varrone J, Wang S, Alings M, Xavier D, Zhu J, Diaz R, Lewis
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF BS, Darius H, Diener HC, Joyner CD, Wallentin L. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-1151 17.
Granger CB, Alexander JH, McMurray JJ, Lopes RD, Hylek EM, Hanna M, Al-Khalidi
RI PT
HR, Ansell J, Atar D, Avezum A, Bahit MC, Diaz R, Easton JD, Ezekowitz JA, Flaker G, Garcia D, Geraldes M, Gersh BJ, Golitsyn S, Goto S, Hermosillo AG, Hohnloser SH, Horowitz J, Mohan P, Jansky P, Lewis BS, Lopez-Sendon JL, Pais P, Parkhomenko A,
fibrillation. N Engl J Med. 2011;365:981-992
Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, Breithardt G, Halperin
M AN U
18.
SC
Verheugt FW, Zhu J, Wallentin L. Apixaban versus warfarin in patients with atrial
JL, Hankey GJ, Piccini JP, Becker RC, Nessel CC, Paolini JF, Berkowitz SD, Fox KA, Califf RM. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891
Hylek EM, Held C, Alexander JH, Lopes RD, De Caterina R, Wojdyla DM, Huber K,
TE D
19.
Jansky P, Steg PG, Hanna M, Thomas L, Wallentin L, Granger CB. Major bleeding in patients with atrial fibrillation receiving apixaban or warfarin: The aristotle trial
EP
(apixaban for reduction in stroke and other thromboembolic events in atrial fibrillation): Predictors, characteristics, and clinical outcomes. Journal of the American College of
20.
AC C
Cardiology. 2014;63:2141-2147
Camm AJ, Lip GY, De Caterina R, Savelieva I, Atar D, Hohnloser SH, Hindricks G, Kirchhof P. 2012 focused update of the esc guidelines for the management of atrial fibrillation: An update of the 2010 esc guidelines for the management of atrial fibrillation. Developed with the special contribution of the european heart rhythm association. European heart journal. 2012;33:2719-2747
Resubmit 021616 text only
ACCEPTED MANUSCRIPT Mortality with Edoxaban vs Warfarin in AF 21.
Guidelines for pharmacotherapy of atrial fibrillation (jcs 2013). Circulation journal : official journal of the Japanese Circulation Society. 2014;78:1997-2021
22.
Crowther M, Crowther MA. Antidotes for novel oral anticoagulants: Current status and
23.
RI PT
future potential. Arteriosclerosis, thrombosis, and vascular biology. 2015;35:1736-1745 Lip GY, Skjoth F, Nielsen PB, Larsen TB. Non-valvular atrial fibrillation patients with none or one additional risk factor of the cha2ds2-vasc score. A comprehensive net
SC
clinical benefit analysis for warfarin, aspirin, or no therapy. Thrombosis and haemostasis. 2015;114:826-834
Lip GY, Skjoth F, Rasmussen LH, Larsen TB. Oral anticoagulation, aspirin, or no
M AN U
24.
therapy in patients with nonvalvular af with 0 or 1 stroke risk factor based on the cha2ds2-vasc score. Journal of the American College of Cardiology. 2015;65:1385-1394 25.
Dlott JS, George RA, Huang X, Odeh M, Kaufman HW, Ansell J, Hylek EM. National
TE D
assessment of warfarin anticoagulation therapy for stroke prevention in atrial fibrillation.
AC C
EP
Circulation. 2014;129:1407-1414
Resubmit 021616 text only
ACCEPTED MANUSCRIPT
Table 1 – Classification of Death by the Clinical Endpoint Committee
Categories of Causes of Death
2. Malignancy
TE D
Infection Suicide Accidental/traumatic Hepatobiliary Renal Other
EP
3. Other a. b. c. d. e. f.
M AN U
SC
RI PT
1. Cardiovascular a. Fatal Bleeding i. Intracranial hemorrhage ii. Extracranial bleeding b. Non-bleeding cardiovascular deaths i. Ischemic stroke ii. Systemic arterial embolic event iii. Congestive heart failure or cardiogenic shock iv. Directly related to coronary revascularization v. Dysrhythmia vi. Pulmonary Embolism vii. Sudden or unwitnessed death viii. Atherosclerotic vascular disease (excluding coronary artery disease) ix. Other
Relationship of Death to Bleeding
AC C
1. Fatal bleeding. Bleeding event led directly to death within 7 days. Cause of death must be either intracranial or extracranial bleeding. 2. Bleeding contributed to death. Bleeding was part of a causal chain resulting in death within 30 days.
Death unrelated to a bleeding event.
AJM submission 16 Feb 2016
1
ACCEPTED MANUSCRIPT
Table 2. Causes of Death by Treatment Group
65 (0.33) 53 (0.27) 12 (0.06)
35 (0.18) 30 (0.15) 5 (0.03)
1543 (2.66) 749 (1.29) 390 (0.67) 146 (0.25) 51 (0.09) 25 (0.04) 17 (0.03) 12 (0.02) 8 (0.01) 145 (0.25) 271 (0.47) 72 (0.12) 32 (0.06) 22 (0.04) 145 (0.25) 410 (0.71) 256 (0.44) 25 (0.04) 21 (0.04) 108 (0.19)
546 (2.83) 270 (1.40) 143 (0.74) 47 (0.24) 15 (0.08) 8 (0.04) 5 (0.03) 4 (0.02) 3 (0.02) 51 (0.26) 84 (0.44) 18 (0.09) 5 (0.03) 5 (0.03) 56 (0.29) 144 (0.75) 92 (0.48) 10 (0.05) 8 (0.04) 34 (0.18)
AJM submission 16 Feb 2016
2
P-Value LD edoxaban Warfarin vs. (N=7034) HD edoxaban n (%/year) 0.082 737 (3.80) 0.013 527 (2.71)
M AN U
SC
RI PT
124 (0.21) 98 (0.17) 26 (0.04)
AC C
Non-bleeding cardiovascular deaths Sudden cardiac death Heart failure or shock Ischemic stroke Dysrhythmia Atherosclerotic vascular disease Pulmonary embolism Directly related to revascularization Systemic arterial embolism Other cardiovascular Malignancy Lung Pancreatic Small or large bowel All other malignancies Non-cardiovascular/non-malignancy Infection Accidental / traumatic Renal All other
HD edoxaban (N=7035) n (%/year) 773 (3.99) 530 (2.74)
TE D
Fatal Bleeding Intracranial hemorrhage Extracranial bleeding
Warfarin (N=7036) n (%/year) 839 (4.35) 611 (3.17)
EP
All deaths Cardiovascular
All Patients (N=21105) n (%/year) 2349 (4.05) 1668 (2.87)
495 (2.56) 247 (1.28) 129 (0.67) 44 (0.23) 16 (0.08) 5 (0.03) 5 (0.03) 3 (0.02) 0 46 (0.24) 94 (0.49) 29 (0.15) 14 (0.07) 9 (0.05) 42 (0.22) 149 (0.77) 95 (0.49) 10 (0.05) 4 (0.02) 40 (0.21)
P-value Warfarin vs. LD edoxaban 0.006 0.008
0.003 0.012 0.099
24(0.12) 15 (0.07) 9 (0.05)
<0.001 <0.001 0.50
0.10 0.29 0.37 0.74 0.87 0.41 0.48 0.75 -0.59 0.47 0.12 0.049 0.30 0.15 0.81 0.86 0.99 0.25 0.49
502 (2.59) 232 (1.20) 118 (0.61) 55 (0.28) 20 (0.10) 12 (0.06) 9 (0.05) 3 (0.02) 5 (0.03) 48 (0.25) 93 (0.48) 25 (0.13) 13 (0.07) 8 (0.04) 47 (0.24) 117 (0.60) 69 (0.36) 5 (0.03) 9 (0.05) 34 (0.18)
0.13 0.073 0.10 0.46 0.41 0.39 0.30 0.70 0.50 0.71 0.54 0.31 0.072 0.42 0.35 0.080 0.061 0.20 0.83 0.97
ACCEPTED MANUSCRIPT
36 (0.19) 101 (0.52)
24 (0.12) 59 (0.30)
0.12 0.001
AC C
EP
TE D
M AN U
SC
RI PT
Relationship of Deaths to Bleeding Bleeding contributed to death 90 (0.16) Fatal bleeding or bleeding that 214 (0.37) contributed to death % rates shown are annualized Kaplan-Meier estimates
AJM submission 16 Feb 2016
3
30 (0.15) 54 (0.28)
0.44 <0.001
ACCEPTED MANUSCRIPT
Figure 1 # pts 150
A. Sudden Cardiac Death 270 247 232
143
100
129
118
SC
200
50
P = NS for each pairwise comparisons
0
M AN U
100
P = NS for each pairwise comparisons
0
HD Edox
LD Edox
TE D
Warfarin
EP
# pts C. Fatal Ischemic Stroke 75
# pts 75
55
AC C
50
B. Fatal Heart Failure
RI PT
# pts 300
47
44
25
50
Warfarin
HD Edox
D. Fatal Bleeding 65
P<0.001 P=0.003
35
25
24
P = NS for each pairwise comparisons
0
LD Edox
0 Warfarin
HD Edox
LD Edox
Warfarin
HD Edox
LD Edox 1
ACCEPTED MANUSCRIPT
Figure 2 # patients
RI PT
600
Non-fatal
506
Fatal
494
400
450
337* 328
TE D
300
EP
200 102
100
AC C
155 79* 49
30 Warfarin
60* 45
53
0
455†
M AN U
SC
500
HD Edox
15 LD Edox
Intracranial Hemorrhage
9
5 12 Warfarin
HD Edox
LD Edox
Extracranial Major Bleeding 2
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3
ACCEPTED MANUSCRIPT
Highlights
•
An analysis of the causes of death was performed in a clinical trial of NOAC vs warfarin
RI PT
in patients with AF •
Edoxaban reduced mortality primarily by reducing fatal intracranial bleeding
•
Edoxaban also reduced non-fatal bleeding, which indirectly may result in fewer
AC C
EP
TE D
M AN U
SC
subsequent deaths
ACCEPTED MANUSCRIPT
Supplemental Appendix
AC C
EP
TE D
M AN U
SC
RI PT
Appendix S1. Clinical Endpoint Committee Classification of Death................................... 27-31
AJM submission 16 Feb 2016
1
ACCEPTED MANUSCRIPT
Table S1 - Clinical Endpoint Committee Classification of Death
Classification of Death
RI PT
Death will be classified as Cardiovascular, Malignancy, or Non-cardiovascular/Non-
malignancy. The cause of death is determined by the principal condition that caused the death, not the
or a non-cardiovascular cause can be clearly shown.
SC
immediate mode of death. All deaths will be assumed to be cardiovascular in nature unless a malignant
Cardiovascular death is defined as death due to documented cardiovascular cause, including
M AN U
deaths due to bleeding. Causes of cardiovascular deaths include, but are not limited to, deaths resulting from atherosclerotic coronary heart disease (acute myocardial infarction, sudden cardiac death, nonsudden death with gradually worsening cardiac symptoms, unwitnessed death without clear alternate cause, procedural death related to cardiac surgery or coronary angiography), atherosclerotic vascular
TE D
disease (cerebrovascular disease including stroke and hemorrhage, aortic, mesenteric, renovascular, peripheral arterial disease, or complication of a non-coronary vascular procedure), other cardiovascular (pulmonary embolism, endocarditis, congestive heart failure, valvular heart disease, arrhythmia), and
EP
deaths due to bleeding.
Deaths due to malignancy will include deaths that are directly a consequence of a malignancy,
AC C
such as a brain tumor that causes herniation, coma, and respiratory arrest. Deaths due to malignancy will be further subclassified by organ system and timing of diagnosis (before vs. after randomization). Examples of non-cardiovascular/non-malignancy deaths include those caused primarily by infection, pulmonary, gastrointestinal, accidental, renal, trauma, or non-cardiovascular organ system failure.
AJM submission 16 Feb 2016
2
ACCEPTED MANUSCRIPT
For all deaths, the relationship of the death to bleeding and malignancy will be adjudicated as follows: Relationship to Bleeding (categories are mutually exclusive)
RI PT
1. Fatal bleeding – death in which a bleeding event directly led to death within 7 days. Examples of fatal bleeding events are an intracranial hemorrhage that led to herniation of the brain and death within 24 hours, and a massive gastrointestinal hemorrhage that results in
SC
shock, hemodynamic collapse, and death. If a bleeding event is considered fatal, then the cause of death must be either intracranial or non-intracranial bleeding.
M AN U
2. Bleeding contributed to death – a death in which a bleeding event was part of a causal chain of medical events that ultimately led to death within 30 days of the bleed, but bleeding was not directly and/or immediately related to subject’s death. An example of bleeding contributing to death is a large retroperitoneal bleed that leads to surgical evacuation,
TE D
development of a subsequent abscess in the area of bleeding that leads to sepsis, multiorgan failure and death 10 days after the onset of bleeding. If bleeding has contributed to death (but the bleeding was not categorized as “fatal”), then the cause of death must be recorded as
EP
something other than intracranial / non-intracranial bleeding. 3. Deaths unrelated to a bleeding event – The case of death was unrelated to bleeding, either
AC C
because there was no clinical significant bleeding in the month prior to death or the bleeding event did not contribute to the subject’s death. An example of a death unrelated to bleeding is an episode of guaiac positive stools in a patient who dies of post-obstructive pneumonia due to lung cancer. In these cases, the cause of death cannot be intracranial / non-intracranial bleeding.
AJM submission 16 Feb 2016
3
ACCEPTED MANUSCRIPT
Relationship to Malignancy 1. Death directly related to malignancy – Death in which the mode of death can be attributed to the direct effects of a malignancy. In such cases, the cause of death adjudicated by the CEC
RI PT
must be malignancy.
2. Death due to a consequence related to malignancy. This would include deaths due to other processes (e.g., infection in a patient who becomes septic and neutropenic due to acute
SC
leukemia) that are a known complication of the malignancy. The underlying malignancy should be on the causal pathway leading to death, but not the immediate cause of death. In
M AN U
such cases, the cause of death adjudicated by the CEC cannot be malignancy, but instead should be the other process (e.g., infection).
3. Death not related to a malignancy. Either no malignancy has been diagnoses or the malignancy that is present was not related to the cause of death. The cause of death must be
TE D
something other than malignancy.
EP
Sudden Cardiovascular Death
Sudden CV death is defined as a sudden, unexpected death that was either:
AC C
a) witnessed, occurring within 60 min from the onset of new symptoms, in the absence of a clear cause other than cardiovascular; or
b) unwitnessed, within 24 hours of being observed alive, in the absence of pre-existing progressive circulatory failure or other non-cardiovascular causes of death;
AJM submission 16 Feb 2016
4
ACCEPTED MANUSCRIPT
Non-sudden Cardiovascular Death This category refers to a patient who had symptoms of a cardiovascular nature and had gradual deterioration prior to death. It includes all patients with CV death who do not meet criteria for sudden
RI PT
death or unwitnessed CV death.
Un-witnessed CV Death
M AN U
known other major causes of death are identified.
Illustrative examples Cause of Death Cardiovascular death:
A subject found dead at home without known major non-cardiovascular
TE D
•
SC
Death that occurred unexpectedly, without patient being seen within 24 hours, and for which no
illness for which death was expected •
A subject with renal failure on hemodialysis who dies following a recurrent myocardial
EP
infarction
A subject who dies following aortic dissection
•
A subject who dies of complications of endocarditis
•
AC C
•
A subject who dies of complications related to coronary artery bypass surgery or percutaneous coronary intervention
•
A subject who dies within 24 hours since the onset of neurologic symptoms compatible with a stroke
AJM submission 16 Feb 2016
5
ACCEPTED MANUSCRIPT
•
A subject who dies suddenly and autopsy shows either a cerebral infarction or a cerebral hematoma (Note: the latter event would also be classified as a fatal bleeding event)
•
A subject who dies of progressive hypotension in the setting of a severe gastrointestinal
RI PT
hemorrhage. (Note: this event would also be classified as a fatal bleed).
Malignancy death:
A subject with an unresectable tracheal carcinoma that causes airway compromise and
SC
•
respiratory arrest
A subject with metastatic breast cancer that causes a recurrent malignant pericardial effusion and pericardial tamponade.
M AN U
•
Non-cardiovascular/Non-malignancy death:
A subject with progressive interstitial pulmonary fibrosis with home hospice care who is
TE D
•
found dead at home
A subject with severe pneumonia and hypoxia who has a bradycardic arrest
•
A subject who dies of multiple trauma in a motor vehicle accident.
•
A subject with stage IV prostate cancer with widespread metastases who dies of hepatic
AC C
EP
•
encephalopathy. (Note: this event would also be classified as a death due to a consequence of
a malignancy, hepatic failure due to metastatic disease)
AJM submission 16 Feb 2016
6