JACC: CARDIOVASCULAR INTERVENTIONS
VOL. 12, NO. 16, 2019
ª 2019 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER
Antithrombotic Therapy and Cardiovascular Outcomes After Transcatheter Aortic Valve Replacement in Patients With Atrial Fibrillation Ioanna Kosmidou, MD, PHD,a,b Yangbo Liu, MS,a Maria C. Alu, MS,b Mengdan Liu, MS,a Mahesh Madhavan, MD,a,b Tarun Chakravarty, MD,c Raj Makkar, MD,c Vinod H. Thourani, MD,d Angelo Biviano, MD, MPH,b Susheel Kodali, MD,a,b Martin B. Leon, MDa,b
ABSTRACT OBJECTIVES The study sought to determine the patterns of antithrombotic therapy and association with clinical outcomes in patients with atrial fibrillation (AF) and CHA2DS2-VASc (congestive heart failure, hypertension, age $75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65–74 years, sex category) score $2 following transcatheter aortic valve replacement (TAVR). BACKGROUND The impact of antithrombotic regimens on clinical outcomes in patients with AF and severe aortic stenosis treated with TAVR is unknown. METHODS In the randomized PARTNER II (Placement of Aortic Transcatheter Valve II) trial and associated registries, 1,621 patients with prior AF and CHA2DS2-VASc score $2 comprised the study cohort. Outcomes were analyzed according to antithrombotic therapy. RESULTS During the 5-year enrollment period, 933 (57.6%) patients were discharged on oral anticoagulant therapy (OAC). Uninterrupted antiplatelet therapy (APT) for at least 6 months or until an endpoint event was used in 544 of 933 (58.3%) of patients on OAC and 77.5% of patients not on OAC. At 2 years, patients on OAC had a similar rate of stroke (6.6% vs. 5.6%; p ¼ 0.53) and the composite outcome of death or stroke (29.7% vs. 31.8%; p ¼ 0.33), compared with no OAC. OAC with APT was associated with a reduced rate of stroke (5.4% vs. 11.1%; p ¼ 0.03) and death or stroke (29.7% vs. 40.1%; p ¼ 0.01), compared with no OAC or APT. Following adjustment, OAC with APT and APT alone were both associated with reduced rates of stroke compared with no OAC or APT (hazard ratio for OACþAPT: 0.43, 95% confidence interval: 0.22 to 0.85; p ¼ 0.015; hazard ratio for APT alone: 0.32, 95% confidence interval: 0.16 to 0.65; p ¼ 0.002), while OAC alone was not. CONCLUSIONS Among patients with prior AF undergoing TAVR, antiplatelet with or without anticoagulant therapy was associated with a reduced risk of stroke at 2 years, implicating multifactorial stroke mechanisms in this population. (J Am Coll Cardiol Intv 2019;12:1580–9) © 2019 by the American College of Cardiology Foundation.
From the aClinical Trials Center, Cardiovascular Research Foundation, New York, New York; bDepartment of Medicine, NewYorkPresbyterian Hospital/Columbia University Medical Center, New York, New York; cDepartment of Medicine, Cedars-Sinai Medical Center, Los Angeles, California; and the dDepartment of Cardiac Surgery, Medstar Heart and Vascular Institute/Georgetown University, Washington, DC. The PARTNER II (Placement of Aortic Transcatheter Valve II) trial was funded by Edwards Lifesciences (Santa Clara, California). Ms. Alu has received consulting fees from Cardiac Dimensions. Dr. Makkar has received grants from Edwards Lifesciences and St. Jude Medical; has received consulting fees from Abbott Vascular, Cordis, and Medtronic; and holds equity in Entourage Medical. Dr. Thourani has served on advisory board for Edwards Lifesciences, Abbott Vascular, Gore Vascular, Bard Medical, JenaValve, and Boston Scientific. Dr. Kodali has received consulting fees from Edwards Lifesciences, Medtronic, Claret Medical, Merrill Lifesciences, BioTrace Medical, and Microinterventional Devices; has served on advisory boards for Thubrikar Aortic Valve, Dura Biotech, Abbott Vascular, Paieon Medical, and St. Jude Medical; and owns equity in Thubrikar Aortic Valve, Dura Biotech, BioTrace Medical, and Microinterventional Devices. Dr. Leon has received Institutional Research Support from Boston Scientific, Edwards, and Medtronic; and has served on the PARTNER Trial Executive Committee (no direct compensation). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received March 1, 2019; revised manuscript received May 29, 2019, accepted June 4, 2019.
ISSN 1936-8798/$36.00
https://doi.org/10.1016/j.jcin.2019.06.001
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
R
ecent randomized trials have demonstrated
transient ischemic attack or thromboembo-
ABBREVIATIONS
noninferiority or superiority of transcatheter
lism, vascular disease, age 65–74 years, sex
AND ACRONYMS
aortic valve replacement (TAVR) compared
category] score $2) enrolled in the PARTNER
with surgical aortic valve replacement in patients
II (Placement of Transcatheter Aortic Valve II)
with severe aortic stenosis (AS) at all levels of surgical
trial (3).
AS = aortic stenosis
SEE PAGE 1590
fibrillation (AF) is high and its presence is associated with increased risk for long-term adverse cardiovas-
AF = atrial fibrillation
APT = antiplatelet therapy
risk (1–5). In this population, the prevalence of atrial
METHODS
cular outcomes (6–8). Advances in pharmacotherapy
CI = confidence interval HR = hazard ratio NOAC = non–vitamin K oral
STUDY DESIGN AND ENDPOINTS. The design
anticoagulant
(NOACs) as an alternative to warfarin (9–12); however,
and initial results of the PARTNER II trial have
OAC = oral anticoagulant
current trends in utilization of oral anticoagulant
been published previously (3,5). Briefly, the
therapy (OAC) following TAVR have not been exam-
PARTNER II trial enrolled patients with severe
ined. Further, the impact of different antithrombotic
symptomatic AS into 2 cohorts: those at in-
regimens with OAC as well with or without antiplate-
termediate surgical risk (cohort A) and those consid-
let therapy (APT) on long-term clinical outcomes in
ered inoperable due to severe coexisting conditions
patients with AF following TAVR for severe AS is un-
(cohort B). Patients in cohort A were randomized to
known (13,14). In the present study, we sought to
TAVR with the SAPIEN XT (transfemoral, transapical,
determine current practice patterns of antithrombotic
or transaortic) (Edwards Lifesciences, Irvine, Califor-
therapy and clinical outcomes in patients with AF and
nia) or to conventional surgical aortic valve replace-
an absolute indication for anticoagulation (CHA 2DS2 -
ment. Patients in cohort B were randomized to
have
introduced
non–vitamin
K
anticoagulants
therapy
TAVR = transcatheter aortic valve replacement
hypertension,
transfemoral TAVR with a first-generation heart valve
age $75 years, diabetes mellitus, prior stroke or
system (SAPIEN [Edwards Lifesciences]) or with a
VASc
1581
Antithrombotic Therapy After TAVR in Patients With AF
[congestive
heart
failure,
F I G U R E 1 Study Flowchart
AF ¼ atrial fibrillation; AS ¼ aortic stenosis; CHA2DS2-VASc ¼ congestive heart failure, hypertension, age $75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65–74 years, sex category; NOAC ¼ non–vitamin K oral anticoagulant; pts ¼ patients; TAVR ¼ transcatheter aortic valve replacement.
1582
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019
Antithrombotic Therapy After TAVR in Patients With AF
AUGUST 26, 2019:1580–9
F I G U R E 2 Trends of Antithrombotic and Anticoagulant Therapy Use at Discharge Following TAVR
(A) Patterns of oral anticoagulant therapy use at discharge and (B) oral anticoagulant and antiplatelet therapy use at discharge following TAVR for severe AS in patients with pre-existing AF and an absolute indication to anticoagulation (CHA2DS2-VASc score $2) enrolled in the PARTNER (Placement of Aortic Transcatheter Valve) trial from 2011 to 2015. Abbreviations as in Figure 1.
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
1583
Antithrombotic Therapy After TAVR in Patients With AF
second-generation valve (SAPIEN XT). Several nested registries were embedded into cohort B: inoperable
T A B L E 1 Clinical Characteristics
Anticoagulant (n ¼ 933)
No Anticoagulant (n ¼ 688)
p Value
82.8 6.7
82.9 7.4
0.68
Female
34.4 (321/933)
32.8 (226/688)
0.51
Hyperlipidemia
78.7 (734/933)
83.0 (571/688)
0.03
Hypertension
91.7 (856/933)
92.4 (636/688)
0.61
Prior stroke or TIA
22.0 (205/933)
19.8 (136/688)
0.28
Congestive heart failure
89.6 (836/933)
87.6 (603/688)
0.22
Diabetes mellitus
35.3 (329/933)
36.5 (251/688)
0.61
sciences) in patients at high or prohibitive risk or at
Pulmonary disease
45.9 (428/933)
43.6 (300/688)
0.36
intermediate risk (5,15) (Online Figure 1). For the pre-
History of anemia
21.2 (198/933)
27.0 (186/688)
0.007
sent analyses, patient-level data were pooled in a
History of carotid artery disease
20.2 (188/933)
20.3 (140/688)
0.92
common database including cohort A and cohort B as
Peripheral vascular disease
32.0 (299/933)
33.9 (233/688)
0.44
Renal insufficiency (SCr $2 mg/dl)
8.9 (83/933)
11.3 (78/688)
0.10
and the SAPIEN 3 registries. For the purpose of the
History of alcohol abuse
2.5 (23/933)
2.3 (16/688)
0.86
present study, patients undergoing TAVR with a his-
Liver disease
tory of AF and an absolute indication for OAC
Previous or current cancer
patients undergoing TAVR via the transapical or transaortic access route, patients receiving TAVR within a degenerated bioprosthetic valve in the aortic position, and inoperable patients who required 29-mm valve. All patients in the cohort B nested registries received
the
SAPIEN
XT
valve.
Following
the
completion of enrollment in the 2 randomized cohorts, 2 additional registries were initiated to study the newest generation SAPIEN 3 valve (Edwards Life-
well as the PARTNER II trial cohort B nested registries
(CHA 2DS 2-VASc
score
$2)
were
identified.
The
Age, yrs
2.6 (24/933)
3.8 (26/688)
0.16
30.2 (282/933)
35.3 (243/688)
0.03
Thrombocytopenia
5.3 (49/933)
6.8 (47/688)
0.18
2.4 (22/933)
2.6 (18/688)
0.74
CHA2 DS2-VASc score was calculated for each patient,
Coagulopathy
with a possible total score of 0 to 9 points. Primary
Coronary artery disease
76.3 (712/933)
81.0 (557/688)
0.02
analyses were performed according to OAC prescribed
Prior PCI or CABG
49.1 (458/933)
58.1 (400/688)
0.0003
at discharge. Secondary analyses were performed ac-
Previous or current bleeding
12.8 (71/554)
17.5 (65/371)
0.05
cording to uninterrupted APT, defined for the purpose
Body mass index, kg/m2
28.4 6.1
27.5 5.6
0.003
of the study as continuous APT (single or dual) for at
Left ventricular ejection fraction, %
least 6 months following the index procedure, as
Frailty
suggested in recent guidelines (16) or until a primary endpoint event, whichever occurred first. The primary endpoints of stroke and the nonhierarchical composite outcome of stroke or death at 2 years were defined a priori and were assessed according to the Valve Academic Research Consortium-
52.8 13.8
53.1 13.9
0.66
23.0 (214/932)
29.8 (205/687)
0.002
STS-PROM score
8.2 4.2
8.8 4.5
0.004
CHA2DS2-VASc score
5.6 1.3
5.6 1.2
0.32
Values are mean SD or % (n/N). CABG ¼ coronary artery bypass grafting; CHA2DS2-VASc ¼ congestive heart failure, hypertension, age $75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65–74 years, sex category; PCI ¼ percutaneous coronary intervention; SCr ¼ serum creatinine; STS-PROM¼ Society of Thoracic Surgeons Predicted Risk of Mortality; TIA ¼ transient ischemic attack.
2 consensus definitions (17). Other study definitions were previously described (3,5). Patients were determined to be frail if they met protocol-defined cutpoints for 3 of 4 metrics: slowness (15-foot walk test), weakness (grip strength), wasting and malnutrition
T A B L E 2 Procedural and Hospitalization Characteristics
(serum albumin), and loss of independence (Katz In-
Anticoagulant (n ¼ 933)
No Anticoagulant (n ¼ 688)
p Value
Procedure duration, min
95.2 43.1
94.0 45.2
0.58
Transcatheter valve size 20 mm 23 mm 26 mm 29 mm
3.1 (28/918) 29.3 (269/918) 37.3 (342/918) 30.4 (279/918)
1.8 (12/681) 32.3 (220/681) 38.2 (260/681) 27.8 (189/681)
0.19 0.10 0.20 0.71 0.25
Approach Transfemoral Transthoracic
78.0 (727/932) 22.0 (205/932)
79.0 (543/687) 21.0 (144/687)
58.7 (548/933) 41.3 (385/933) 5.0 (4.0–8.0)
62.9 (433/688) 37.1 (255/688) 6.0 (4.0–9.0)
dex of Independence in Activities of Daily Living). The investigation was approved by the Institutional Review Board or ethics committee at each participating
center,
and
all
patients
signed
written
informed consent. Median follow-up was 2.8 (interquartile range: 1.6 to 3.3) years. STATISTICAL METHODS. Continuous variables are
reported as mean SD or median (interquartile range) and were compared using the Student’s t-test or Wilcoxon rank sum test, as appropriate. Categorical variables are expressed as counts and proportions and were compared with the chi-square or Fisher exact test, as appropriate. Time-updated event rates were estimated using the Kaplan-Meier method, with comparisons made using the log-rank
Transcatheter valve SAPIEN XT SAPIEN 3 Index hospitalization stay, days
0.62
0.09
Values are mean SD, % (n/N), or median (interquartile range).
0.25
1584
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
Antithrombotic Therapy After TAVR in Patients With AF
discharge. To exclude procedure-related factors,
T A B L E 3 Clinical Outcomes at 2 Years According to Oral Anticoagulant Therapy
landmark analyses were performed at 30 days for 2year events. A 2-sided p value <0.05 was considered
Anticoagulant (n ¼ 933)
No Anticoagulant (n ¼ 688)
p Value
Death or stroke
29.7 (273)
31.8 (214)
0.33
All-cause death Cardiovascular
26.1 (240) 18.0 (160)
28.9 (194) 19.7 (126)
0.20 0.42
9.4 (SAS Institute, Cary, North Carolina).
9.3 (75) 6.6 (53)
8.2 (48) 5.6 (33)
0.55 0.53
RESULTS
22.5 (193) 5.9 (51)
19.1 (117) 5.0 (31)
0.10 0.44
PATTERNS OF ANTICOAGULANT USE. The study
Aortic valve reintervention
1.2 (10)
1.3 (8)
0.81
cohort included 1,621 patients with AF and CHA 2DS 2-
Prosthetic valve dysfunction
5.6 (47)
5.0 (30)
0.59
VASc score $2 treated with TAVR from 2011 to 2015; of
Stroke or TIA Stroke Bleeding Major
to be statistically significant for all tests. All statistical analyses were performed with SAS software version
these, 933 (57.6%) patients were discharged on OAC Values are % (n).
(778 [83.4%] on warfarin and 155 [16.6%] on NOACs)
TIA ¼ transient ischemic attack.
and 688 (42.4%) were not discharged on OAC (Figure 1). Overall implementation of OAC following test. Multivariable Cox proportional hazards analyses were used to determine whether anticoagulant therapy influenced the rate of stroke and the composite outcome of death or stroke from discharge to 2 years, compared with patients on no OAC at discharge. The following covariates were defined a priori based on clinical relevance and were included in the main model: OAC, age, sex, medically treated diabetes, hypertension, congestive heart failure, baseline renal insufficiency, coronary artery disease, prior stroke or transient ischemic attack, anemia, CHA 2DS 2-VASc score, and type of implanted valve. To account for
TAVR increased consistently from 46.4% in 2011 to 69.6% in 2015 and warfarin remained the most frequently
used
oral
anticoagulant
(Figure
2A).
Among patients not on OAC at discharge, 67 of 688 (9.7%) were subsequently placed on OAC during the 2-year follow-up period. APT according to the study definitions was used in 544 of 933 (58.3%) of patients discharged on OAC (455 of 778 [58.5%] on warfarin, 89 of 155 [57.4%] on NOAC) and 533 of 688 (77.5%) of patients not discharged on OAC. Antiplatelet utilization did not vary substantially over the study followup period (Figure 2B).
changes in anticoagulant use over time, the time-
BASELINE AND PROCEDURAL CHARACTERISTICS.
adjusted Cox model was also used (18), with antico-
Baseline characteristics according to OAC at discharge
agulants dynamically updated in time. Alternative
are shown in Table 1. Patients not receiving OAC at
analyses were performed according to anticoagulant
discharge were less frequently overweight; had more
therapy
previously
frequent history of hyperlipidemia, anemia, bleeding,
defined, using the same clinically pertinent cova-
coronary artery disease and coronary revasculariza-
riates as in the main model, compared with patients
tion; and higher frailty and Society of Thoracic Sur-
on no antithrombotic therapy (no OAC or APT) at
geons Predicted Operative Mortality scores compared
and
uninterrupted
APT,
as
with patients discharged on OAC. The CHA 2DS2-VASc T A B L E 4 Clinical Outcomes at 2 Years According to Uninterrupted Antiplatelet Therapy
score was similar between groups; overall, 1,568 of
With or Without Concomitant Oral Anticoagulant Therapy After TAVR
1,621 (96.7%) patients in the study cohort had a
Anticoagulant
CHA 2DS2-VASc score $4. Procedural characteristics
No Anticoagulant
were similar among groups (Table 2). Medications at
With APT (n ¼ 544)
Without APT (n ¼ 387)
With APT (n ¼ 533)
Without APT (n ¼ 149)
Overall p Value
Death or stroke
29.7 (160)
29.7 (113)
29.5 (156)
40.1 (57)
0.06
OAC were more frequently discharged on diuretic
All-cause death Cardiovascular
26.6 (143) 18.3 (95)
25.5 (97) 17.7 (65)
27.7 (146) 18.8 (94)
33.1 (47) 23.4 (32)
0.31 0.44
agents and less frequently on dual APT and antiar-
7.7 (37) 5.4 (26)
11.6 (38) 8.3 (27)
6.8 (32) 4.2 (20)
13.4 (16) 11.1 (13)
0.04 0.02
on OAC.
23.7 (119) 6.9 (35)
20.8 (74) 4.5 (16)
17.8 (86) 5.1 (25)
23.6 (30) 4.4 (6)
0.13 0.41
according to OAC at discharge are presented in Table 3.
Aortic valve reintervention
1.9 (9)
0.3 (1)
1.2 (6)
1.6 (2)
0.26
At 2 years, the rates of stroke or the composite of death
Prosthetic valve dysfunction
5.5 (27)
5.8 (20)
4.4 (21)
7.3 (9)
0.59
or stroke did not differ according to OAC. When
Stroke or TIA Stroke Bleeding Major bleeding
discharge are shown in Online Table 1. Patients on
rhythmic
therapy
compared
with
patients
not
CLINICAL OUTCOMES. Unadjusted clinical outcomes
warfarin and NOAC were examined separately, there Values are % (n). APT ¼ antiplatelet therapy; TAVR ¼ transcatheter aortic valve replacement; TIA ¼ transient ischemic attack.
were no differences in the rates of stroke, or the composite outcome of death or stroke compared with no
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
1585
Antithrombotic Therapy After TAVR in Patients With AF
C ENTR AL I LL U STRA T I O N Time-to-Event Curves and Adjusted Hazard Ratios for the 2-Year Endpoint of Stroke According to Antithrombotic Therapy Following TAVR 20
Stroke HR (95% CI)
8.3% 5.4% 4.2%
5
Number at risk: OAC with APT OAC Alone APT Alone No OAC or APT
6
12
18
24
402 288 397 92
371 261 366 81
0.32 (0.16, 0.65) 0.002
OAC Alone
0.61 (0.31, 1.21)
Time (Months) 544 387 533 149
OAC with APT
485 340 470 121
443 318 430 108 OAC Alone
APT Alone
0.16
0.44 (0.22, 0.87) 0.018
0. 1 0. 5 1.0
0
APT Alone
OAC with APT
0
p Value
2. 0 2. 5 3. 0
11.1%
10
1.5
Stroke (%)
15
Reference: no OAC or APT
No OAC or APT
Kosmidou, I. et al. J Am Coll Cardiol Intv. 2019;12(16):1580–9.
Time-to-event curves and adjusted hazard ratios (HRs) for the 2-year endpoint of stroke according to antithrombotic therapy following transcatheter aortic valve replacement for severe aortic stenosis in patients with preexisting atrial fibrillation and an absolute indication to anticoagulation (CHA2DS2-VASc [congestive heart failure, hypertension, age $75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65–74 years, sex category] score $2). Two-year stroke rates were lower in patients treated with antiplatelet therapy (APT) alone or anticoagulant with concomitant APT, and similar in patients with anticoagulant therapy alone compared with no antithrombotic therapy. CI ¼ confidence interval; OAC ¼ oral anticoagulant.
OAC (Online Table 2). In contrast, unadjusted clinical
NOAC were entered as separate covariates, neither
outcomes differed significantly according to APT use
anticoagulant affected the adjusted risk of stroke or
(Table 4). Specifically, compared with no antith-
the
rombotic therapy, patients on OAC and APT had a
(Online Table 3). Consistently, when OAC was
significantly lower unadjusted risk of stroke (hazard
assessed as a dynamically updated variable in time,
ratio [HR]: 0.49; 95% confidence interval [CI]: 0.25 to
OAC resulted in a similar risk of stroke (unadjusted
0.95; p ¼ 0.03) (Central Illustration) and the composite
HR: 1.03; 95% CI: 0.67 to 1.59; p ¼ 0.88; adjusted HR:
outcome of death or stroke (HR: 0.68; 95% CI: 0.50 to
1.04; 95% CI: 0.67 to 1.60; p ¼ 0.86) or the composite
0.92; p ¼ 0.01) (Figure 3) whereas OAC without APT
of death or stroke (unadjusted HR: 0.85; 95% CI: 0.71
conferred a similar unadjusted risk of stroke (HR: 0.71;
to 1.02; p ¼ 0.08; adjusted HR: 0.87; 95% CI: 0.73 to
95% CI: 0.37 to 1.38; p ¼ 0.31) and a lower risk of death
1.05; p ¼ 0.14) and increased adjusted risk of bleeding
or stroke (HR: 0.69; 95% CI: 0.50 to 0.95; p ¼ 0.02).
(adjusted HR: 1.27; 95% CI: 1.01 to 1.61; p ¼ 0.040)
Among patients who had a stroke, all had a CHA 2DS2 -
compared with no OAC.
composite
of
death or
stroke
at
2 years
VASc score $4 and 27 of 86 (31.4%) were on OAC alone
In alternative models, APT with or without OAC
whereas 26 of 86 (30.2%) were on both OAC and APT
was associated with a reduced risk of stroke and the
before the event.
composite outcome of death or stroke compared with
After multivariable adjustment, OAC at discharge
patients on no antithrombotic therapy (no OAC or
was not an independent predictor of the risk of stroke
APT). In contrast, OAC without APT was not associ-
or the composite outcome of death or stroke (Table 5,
ated with a reduction in the risk of stroke at 2 years
model I). There was a trend toward increased overall
but was associated with a reduction in the risk of the
bleeding risk at 2 years in patients on OAC (HR: 1.07;
composite outcome of death or stroke, compared with
95% CI: 0.98 to 1.56; p ¼ 0.07). When warfarin or
no antithrombotic therapy (Table 5, model II).
1586
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
Antithrombotic Therapy After TAVR in Patients With AF
F I G U R E 3 Time-to-Event Curves and Adjusted Hazard Ratios for the 2-Year Composite Endpoint of Death or Stroke According to Antithrombotic Therapy
Following TAVR
Time-to-event curves and adjusted hazard ratios (HRs) for the 2-year composite endpoint of death or stroke according to antithrombotic therapy following TAVR for severe AS in patients with pre-existing AF and an absolute indication to anticoagulation (CHA2DS2-VASc score $2). APT ¼ antiplatelet therapy; CI ¼ confidence interval; OAC ¼ oral anticoagulant; other abbreviations as in Figure 1.
Analyses landmarked at 30 days confirmed the inde-
experience from studies in the general population or
pendent association between OAC and APT, as well as
predominantly the percutaneous coronary stent trials
APT alone, with a reduced risk of stroke at 2 years,
(19–21). Nevertheless, although preexistent AF iden-
compared with no antithrombotic therapy (Table 6).
tifies a high risk patient population that may benefit from anticoagulant therapy following TAVR (6,7),
DISCUSSION
management of such patients is challenging, given the difficulty in balancing the ischemic and bleeding
The major findings from the present analysis from the
risk in a typically elderly population, as well as lack of
PARTNER II trial and nested registries, in which the
conclusive data on the efficacy and safety of different
patterns and clinical impact of antithrombotic thera-
antithrombotic regimens (22). Further, whether ce-
pies among patients with AF undergoing TAVR were
rebrovascular outcomes are linked to AF-related
examined, are as follows: 1) OAC was not prescribed
thromboembolic risk (which could be mitigated by
in 43.4% of patients with AF and an absolute indica-
anticoagulation) or whether alternate stroke mecha-
tion for anticoagulation (CHA 2DS2 -VASc score $2)
nisms are in effect in patients with AS managed with
following TAVR; however, overall use of anticoagu-
TAVR remains unknown. In the present study, we
lant therapy increased steadily over the course of the
identified patients with AF undergoing TAVR for se-
study, reflecting the lower risk population enrolled
vere AS who met criteria for anticoagulant therapy
over time; 2) APT varied according to concomitant use
(23), and we examined the patterns of antithrombotic
of anticoagulant therapy; 3) OAC alone was not
therapies implemented over the 5-year course of the
associated with a reduced risk of stroke; and 4) un-
PARTNER II trial and their impact on long-term
interrupted APT, irrespective of concomitant anti-
outcomes.
coagulation, was associated with a reduced risk of
Regarding management of antithrombotic therapy
stroke and the composite outcome of death or stroke
in patients with AF following TAVR, our results
at 2 years.
suggest that over 40% patients were not prescribed
Antithrombotic strategies following TAVR remain
OAC upon discharge, though this rate decreased over
heterogeneous and largely empirical, drawing from
the course of the trial. Several observations may
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
Antithrombotic Therapy After TAVR in Patients With AF
explain these findings; first, patients not on OAC at discharge had frequent history of bleeding and ane-
T A B L E 5 Independent Predictors of Stroke and the Composite Outcome of
Death or Stroke at 2 Years
mia, as well as prior coronary revascularization with an indication for long-term APT. Second, patients not on OAC had a higher frailty index and operative
Anticoagulant therapy (vs. no anticoagulant therapy)
ferences may simply reflect the inclusion of very
Anemia
high-risk patients in the early stages of the trial;
to
lack
of
was not prescribed in 9.2% of patients in the entire
0.016 0.0004
1.31 (1.10–1.55)
0.002
Anticoagulant therapy (vs. no anticoagulant therapy)
1.17 (0.76–1.82)
0.48
Age (per 10 yrs)
1.53 (1.02–2.28)
0.038
0.64 (0.47–0.87) 0.67 (0.49–0.93) 0.69 (0.50–0.93)
0.005 0.016 0.016
1.28 (1.05–1.57) 0.54 (0.38–0.76) 1.32 (1.12–1.57)
0.015 0.0004 0.001
0.32 (0.16–0.65) 0.61 (0.31–1.21) 0.44 (0.22–0.87)
0.002 0.16 0.018
1.54 (1.03–2.31)
0.037
Stroke
comprehensive data on NOAC utilization in the AS and TAVR population (9–11). Antiplatelet therapy
1.28 (1.05–1.56)
Age (per 10 yrs)
reported in the general population (25). NOACs were due
0.60
0.54 (0.39–0.76)
SAPIEN 3
nevertheless, the OAC trends were similar to those likely
0.95 (0.79–1.14)
Death or stroke
adverse outcomes following TAVR (24). These dif-
infrequently,
p Value
Model I
mortality risk, suggesting a heightened risk for
prescribed
Hazard Ratio (95% Confidence Interval)
Model II Death or stroke
study population and 21.8% of patients not dis-
Antiplatelets (vs. no therapy) Anticoagulant (vs. no therapy) Anticoagulant with antiplatelet therapy (vs. no therapy) Anemia SAPIEN 3 Age (per 10 yrs) Stroke Antiplatelets (vs. no therapy) Anticoagulant (vs. no therapy) Anticoagulant with antiplatelet therapy (vs. no therapy) Age (per 10 yrs)
charged on OAC, indicating both reasonable adherence to guideline-directed APT following TAVR while further suggesting a significant role of bleeding risk assessment on a patient level. Upon examining the impact of antithrombotic regimens on clinical outcomes, we showed that OAC did not affect the long-term risk for stroke or the composite outcome of death or stroke when used without concomitant APT for at least 6 months following TAVR. Interestingly, among patients not on OAC, most were managed with APT, which typically consisted of 2 antiplatelet agents; as such, it is
Additional covariates tested (noted in the Methods section) were not statistically significant in the final models and are not included in the tables.
plausible that any potential benefit from OAC was masked by the benefit provided by APT (26). In support of this hypothesis, when APT was implemented for at least 6 months, the risk of stroke was significantly
reduced
when
compared
with
no
antithrombotic therapy (no OAC or APT), irrespective of OAC. Collectively, our results suggest that, in contrast to the general population (27,28), OAC alone is not sufficient in preventing cerebrovascular events in patients with AF following TAVR. The following hypotheses can be made; first, the mechanisms of stroke following TAVR are multifactorial; acute and subacute events are primarily related to procedural factors, whereas late events (>30 days from the index TAVR) are driven mainly by clinical characteristics,
such
as
AF,
age,
and
the
extent
of
atherosclerosis (21). Importantly, OAC without APT did not reduce the risk of stroke whether examining all events or only late events, as evidenced by adjusted analyses landmarked at discharge and 30 days, respectively. Second, the deployed stent valve can trigger platelet activation and increased
for 50% of stroke in patients treated with TAVR and of those, few are directly linked to left atrial thrombus formation and thereby preventable by OAC alone. Accordingly, in the present study, 61.6% of patients who had a stroke were on OAC before the event and results were confirmed when OAC was examined as a dynamic variable. Consistent with our observations, a recent report showed that among patients who had a stroke within a year following TAVR, 67.8% were on OAC following discharge from the index procedure (30). Last, the beneficial effects of APT may be reflective of the diffuse inflammatory and atherothrombotic process in patients with AS (31), which in turn provides a pathophysiologic stroke substrate unrelated to left atrial thromboembolism. Collectively, the present study suggests that OAC therapy, while indicated for prevention of cardioembolic events in the setting of coexistent AF, should nevertheless be used in conjunction with APT for at least 6 months following TAVR.
thrombogenicity due to endothelial denudation (29),
STUDY LIMITATIONS. As the largest trial to date of
a process that can be halted by APT in the early
patients with AS treated with TAVR, the PARTNER II
stages following implantation of a transcatheter
trial provides clinically relevant insights regarding
valve. Third, thromboembolic events likely account
the
utilization
of
antithrombotic
therapies
and
1587
1588
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
Antithrombotic Therapy After TAVR in Patients With AF
T A B L E 6 Independent Predictors of Stroke and the Composite Outcome of
Death or Stroke at 2 Years (Landmark Analysis, 30 Days to 2 Years)
affected by dementia. Nevertheless, our results were similar when OAC was assessed as a dynamically updated variable. Last, the type of AF (paroxysmal or
Hazard Ratio (95% Confidence Interval)
p Value
persistent) was not specified and patients with new onset AF following TAVR were excluded from the
Model I
present analysis.
Death or stroke Anticoagulant therapy (vs. no anticoagulant therapy)
0.96 (0.79–1.15)
0.64
SAPIEN 3
0.54 (0.38–0.77)
0.0006
Age (per 10 yrs)
1.30 (1.09–1.54)
0.003
Anticoagulant therapy (vs. no anticoagulant therapy)
1.25 (0.78–2.01)
0.35
prior stroke or TIA
1.77 (1.09–2.89)
0.021
0.72 (0.52–1.00) 0.71 (0.50–1.01) 0.76 (0.55–1.06)
0.049 0.057 0.11
0.54 (0.38–0.77) 1.32 (1.10–1.57)
0.0007 0.002
0.39 (0.18–0.85) 0.77 (0.36–1.63) 0.51 (0.24–1.09)
0.017 0.50 0.08
CONCLUSIONS In the PARTNER II trial, among patients with AF and
Stroke
Model II Death or stroke Antiplatelets (vs. no therapy) Anticoagulant (vs. no therapy) Anticoagulant with antiplatelet therapy (vs. no therapy) SAPIEN 3 Age (per 10 yrs) Stroke Antiplatelets (vs. no therapy) Anticoagulant (vs. no therapy) Anticoagulant with antiplatelet therapy (vs. no therapy)
an absolute indication for anticoagulation, anticoagulant therapy was associated with a reduced risk of stroke and the composite of death or stroke when used concomitantly with uninterrupted APT following TAVR. Further studies are needed to better define the pathophysiologic mechanisms of stroke and optimal preventive therapies following TAVR. ADDRESS
FOR
CORRESPONDENCE:
Dr.
Ioanna
Kosmidou, Columbia University Medical Center, Department of Cardiology, Herbert Irving Pavillion-6, New York, New York 10032. E-mail: ik2394@cumc.
Additional covariates tested were not statistically significant in the final models and are not included in the tables.
columbia.edu. Twitter: @IKosmidou, @crfheart. PERSPECTIVES
impact on clinical outcomes. However, several limi-
WHAT IS KNOWN? Current guidelines suggest that
tations should be considered. First, the present
patients with a history of AF undergoing TAVR may be
analysis was post hoc, and should thus be consid-
discharged on OAC with or without single or dual APT.
ered hypothesis-generating. APT was defined a priori
However, practice trends in antithrombotic therapy
as uninterrupted therapy with single or dual anti-
and clinical impact on long-term outcomes in patients
platelets for at least 6 months after TAVR (according
with AF and severe AS managed with TAVR remain
to current guidelines) (16) or until the first occur-
unknown.
rence of a primary endpoint event; as such, our findings may not be applicable to different durations of APT. The absolute number of events was modest, and not all confounders in their described relationships may have been identified. Second, medication dosages were not captured. Similarly, international normalized
ratio
levels
were
not
monitored
throughout the study and subtherapeutic INR levels may have affected the results (32). Third, antithrombotic therapy was based on site investigator preference and selection bias cannot be excluded. NOAC use was limited and most patients received dabigatran. NOACs and P2Y12 inhibitors were not often prescribed at the time of enrollment. Fourth, at 2 years, approximately 7% of patients transitioned from no OAC to OAC or OAC to no OAC in a symmetrical process. The specific reasons for transition were not reported, however patients transitioning from OAC to no OAC were older and more frequently
WHAT IS NEW? Among patients with AF undergoing TAVR, 57.6% of patients with an indication for anticoagulant therapy are discharged on OAC, most commonly warfarin. Antiplatelet therapy for at least 6 months with or without concomitant OAC is associated with a reduced risk of stroke at 2 years, whereas OAC alone does not reduce the risk of stroke. WHAT IS NEXT? The present study suggests that adverse cerebrovascular outcomes in patients with preexisting AF and severe AS treated with TAVR are only partially related to AF-related cardioembolism and APT for at least 6 months following TAVR significantly improves long-term prognosis. Further studies are needed to define the optimal periprocedural and long-term approach to antithrombotic therapy in patients with AF following TAVR.
Kosmidou et al.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 12, NO. 16, 2019 AUGUST 26, 2019:1580–9
Antithrombotic Therapy After TAVR in Patients With AF
REFERENCES 1. Mack MJ, Leon MB, Thourani VH, et al. Transcatheter aortic-valve replacement with a balloonexpandable valve in low-risk patients. N Engl J Med 2019;380:1695–705. 2. Popma JJ, Deeb GM, Yakubov SJ, et al. Transcatheter aortic-valve replacement with a selfexpanding valve in low-risk patients. N Engl J Med 2019;380:1706–15. 3. Leon MB, Smith CR, Mack MJ, et al. Transcatheter or surgical aortic-valve replacement in intermediate-risk patients. N Engl J Med 2016; 374:1609–20. 4. Smith CR, Leon MB, Mack MJ, et al. Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med 2011;364: 2187–98. 5. Thourani VH, Kodali S, Makkar RR, et al. Transcatheter aortic valve replacement versus surgical valve replacement in intermediate-risk patients: a propensity score analysis. Lancet 2016;387:2218–25. 6. Mojoli M, Gersh BJ, Barioli A, et al. Impact of atrial fibrillation on outcomes of patients treated by transcatheter aortic valve implantation: A systematic review and meta-analysis. Am Heart J 2017;192:64–75. 7. Mok M, Urena M, Nombela-Franco L, et al. Clinical and prognostic implications of existing and new-onset atrial fibrillation in patients undergoing transcatheter aortic valve implantation. J Thromb Thrombolysis 2013;35:450–5.
14. O’Malley RG, Mahaffey KW, Fearon WF. Adjuvant antithrombotic therapy in TAVR. Curr Cardiol Rep 2017;19:41. 15. Herrmann HC, Thourani VH, Kodali SK, et al. One-year clinical outcomes with SAPIEN 3 transcatheter aortic valve replacement in high-risk and inoperable patients with severe aortic stenosis. Circulation 2016;134:130–40. 16. Otto CM, Kumbhani DJ, Alexander KP, et al. 2017 ACC Expert Consensus Decision Pathway for Transcatheter Aortic Valve Replacement in the Management of Adults With Aortic Stenosis: a report of the American College of Cardiology Task Force on Clinical Expert Consensus Documents. J Am Coll Cardiol 2017;69:1313–46. 17. Kappetein AP, Head SJ, Genereux P, et al. Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document. J Am Coll Cardiol 2012;60:1438–54. 18. Simon R, Makuch RW. A non-parametric graphical representation of the relationship between survival and the occurrence of an event: application to responder versus non-responder bias. Stat Med 1984;3:35–44. 19. Holmes DR Jr, Mack MJ, Kaul S, et al. 2012 ACCF/AATS/SCAI/STS expert consensus document on transcatheter aortic valve replacement. J Am Coll Cardiol 2012;59:1200–54.
25. Katz DF, Maddox TM, Turakhia M, et al. Contemporary trends in oral anticoagulant prescription in atrial fibrillation patients at low to moderate risk of stroke after guidelinerecommended change in use of the CHADS2 to the CHA2DS2-VASc score for thromboembolic risk assessment: analysis from the National Cardiovascular Data Registry’s Outpatient Practice Innovation and Clinical Excellence Atrial Fibrillation Registry. Circ Cardiovasc Qual Outcomes 2017;10. 26. Connolly SJ, Pogue J, Hart RG, et al. Effect of clopidogrel added to aspirin in patients with atrial fibrillation. N Engl J Med 2009;360: 2066–78. 27. Alexander JH, Lopes RD, Thomas L, et al. Apixaban vs. warfarin with concomitant aspirin in patients with atrial fibrillation: insights from the ARISTOTLE trial. Eur Heart J 2014;35: 224–32. 28. Xu H, Ruff CT, Giugliano RP, et al. Concomitant use of single antiplatelet therapy with edoxaban or warfarin in patients with atrial fibrillation: analysis from the ENGAGE AF-TIMI48 trial. J Am Heart Assoc 2016;5:e002587. 29. Marechaux S, Corseaux D, Vincentelli A, et al. Identification of tissue factor in experimental aortic valve sclerosis. Cardiovasc Pathol 2009;18: 67–76. 30. Kleiman NS, Maini BJ, Reardon MJ, et al. Neurological events following transcatheter aortic
8. Nombela-Franco L, Webb JG, de Jaegere PP,
20. Iung B, Rodes-Cabau J. The optimal management of anti-thrombotic therapy after valve replacement: certainties and uncertainties. Eur Heart J 2014;35:2942–9.
et al. Timing, predictive factors, and prognostic value of cerebrovascular events in a large cohort of patients undergoing transcatheter aortic valve implantation. Circulation 2012;126:3041–53.
21. Mastoris I, Schoos MM, Dangas GD, Mehran R. Stroke after transcatheter aortic valve replacement: incidence, risk factors, prognosis, and preventive strategies. Clin Cardiol 2014;37:756–64.
Nonrheumatic calcific aortic stenosis: an overview from basic science to pharmacological prevention. Eur J Cardiothorac Surg 2009;35: 493–504.
9. Connolly SJ, Ezekowitz MD, Yusuf S, et al.
22. Sherwood MW, Vora AN. Challenges in aortic stenosis: review of antiplatelet/anticoagulant
32. Connolly SJ, Pogue J, Eikelboom J, et al. Benefit of oral anticoagulant over antiplatelet therapy in atrial fibrillation depends on the quality of international normalized ratio control
Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009;361:1139–51. 10. Giugliano RP, Ruff CT, Braunwald E, et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2013;369:2093–104. 11. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011;365:981–92. 12. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011;365:883–91. 13. Abdul-Jawad Altisent O, Durand E, MunozGarcia AJ, et al. Warfarin and antiplatelet therapy versus warfarin alone for treating patients with atrial fibrillation undergoing transcatheter aortic valve replacement. J Am Coll Cardiol Intv 2016;9:1706–17.
valve replacement and their predictors: a report from the CoreValve trials. Circulation Cardiovasc Interv 2016;9:e003551. 31. Parolari A, Loardi C, Mussoni L, et al.
therapy management with transcatheter aortic valve replacement (TAVR): TAVR with recent PCI, TAVR in the patient with atrial fibrillation, and TAVR thrombosis management. Curr Cardiol Rep 2018;20:130. 23. January CT, Wann LS, Alpert JS, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol 2014; 64:e1–76. 24. Green P, Arnold SV, Cohen DJ, et al. Relation of frailty to outcomes after transcatheter aortic valve replacement (from the PARTNER trial). Am J Cardiol 2015;116:264–9.
achieved by centers and countries as measured by time in therapeutic range. Circulation 2008;118: 2029–37.
KEY WORDS antithrombotic therapy, aortic stenosis, atrial fibrillation, clinical outcomes, transcatheter aortic valve replacement
A PP END IX For a supplemental figure and tables, please see the online version of this paper.
1589