Keywords
transcatheter aortic valve replacement - atrial fibrillation - closure time of adenosine
diphosphate - primary hemostasis - bleeding events
Introduction
Transcatheter aortic valve replacement (TAVR) has emerged as an established treatment
in patients with symptomatic severe aortic stenosis (AS).[1]
[2]
[3]
[4] Atrial fibrillation (AF) is common in patients undergoing TAVR and constitutes an
indication for long-term oral anticoagulants (OACs) with a vitamin K antagonist (VKA)
or direct-acting OAC (DOAC). The reported prevalence of pre-existing AF in patients
undergoing TAVR ranges from 16 to 51%.[5]
[6] While AF is a strong predictor of thromboembolic events, the benefits of anticoagulation
therapy require careful balancing against the increased risk of bleeding events. In
fact, numerous studies have stressed that AF was associated with an increased rate
of all-cause mortality and bleeding events in patients undergoing TAVR compared with
those in sinus rhythm.[6] Notably, a recent observational study conducted by Lother et al demonstrated that
the incidence of bleedings defined the in-hospital outcome of patients with AF after
transfemoral TAVR.[5] Although studies have suggested that the postprocedural phase demands particular
care in bleeding prevention, an indicator of future bleeding events in AF patients
is lacking.
Closure time of adenosine diphosphate (CT-ADP) is a point-of-care test used as a surrogate
marker of primary hemostatic disorder. We previously established that postprocedural
CT-ADP >180 seconds was an independent predictor of significant paravalvular leak
(PVL), early- and late-bleeding events, and increased 1-year mortality among patients
undergoing TAVR.[7]
[8]
[9] Interestingly, our observational study highlighted that CT-ADP >180 seconds was
also associated with early ischemic neurological events as well as major bleeding
events.[10] Consistent with prior studies, the paradoxical relationship between bleeding and
ischemic stroke events can be explained by the low hemoglobin (Hb) level which may
induce reduced cerebral oxygenation or blood transfusions contributed to the increased
risk of stroke through a transfusion-enhanced systemic inflammatory reaction, causing
a prothrombotic environment.[11]
[12] In line with this paradigm, strategies aimed at reducing the risk of bleeding constitute
an important area of patient's care improvement.
Thus, the current study aimed to evaluate the impact of ongoing primary hemostatic
disorders, defined by a CT-ADP >180 seconds after TAVR, on bleeding events in TAVR
patients with AF.
Methods
Study Design and Patients
Consecutive patients with symptomatic severe AS[13] who underwent TAVR were enrolled in a prospective registry at our institution (Nouvel
Hôpital Civil, Université de Strasbourg, Strasbourg, France) between February 2012
and May 2019. For the purpose of the present analysis, absence of CT-ADP assays and
patients with missing AF data were excluded ([Fig. 1]). Indications for TAVR and procedural approaches were assessed by the local heart
team. The balloon-expandable Edwards SAPIEN XT or S3 prosthesis (Edwards Lifesciences,
Irvine, California, United States), the self-expandable CoreValve, Evolut-R, or Evolut-PRO
(Medtronic, Irvine, California, United States), and ACURATE neo (Boston Scientific,
Natick, Massachusetts, United States) were used. All patients signed informed consent
before the procedure and agreed to the anonymous processing of their data (France
2 and France TAVI Registries). The study protocol was developed in accordance with
the Declaration of Helsinki and was approved by the France 2 study: 911262.
Fig. 1. Study design. AF, atrial fibrillation; CT-ADP, closure time adenosine diphosphate;
TAVR, transcatheter aortic valve replacement.
Definition of AF
AF was defined as a supraventricular tachyarrhythmia with uncoordinated atrial electrical
activation and consequently ineffective atrial contraction.[14] Atrial flutter was defined as a supraventricular tachyarrhythmia with a continuous
regular electrical activity, most commonly a saw-tooth pattern on electrocardiogram.[15] Patients with a history of AF or atrial flutter (permanent, persistent, or paroxysmal)
or with an AF newly diagnosed during the admission were included in the AF group.
Atrial flutter was included in the AF group considering its frequent coexistence with
AF.[15]
Antithrombotic Regimen
Unfractionated heparin was used during the procedure to achieve an activated clotting
time of 250 to 350 seconds. Heparin was antagonized with 100 IU/kg of protamine at
the end of the procedure. Patients without OAC indication received aspirin (75–160 mg)
and clopidogrel (300 mg for loading dose and 75 mg/day for maintenance dose) before
TAVR, with ongoing dual antiplatelet therapy (DAPT) after the procedure for 3 months.
Clopidogrel loading dose was not provided when the patient was under chronic clopidogrel
therapy. In patients receiving chronic OAC therapy, clopidogrel was not initiated
and OAC and aspirin were continued for 3 months after TAVR. Routinely, OAC therapy
was discontinued 5 days prior to the procedure. The choice of OAC was determined at
the discretion of the attending physician after the heart team considered the risk
for bleeding and thrombotic complications.
Blood Collections
Whole blood samples were collected by venipuncture the day before and 24 hours following
TAVR. PFA-100 (Siemens Healthcare, Marburg, Germany) was used for the CT-ADP assay.
Ongoing primary hemostatic disorder was defined by a postprocedural CT-ADP >180 seconds.[9] The extent of P2Y12 inhibition by clopidogrel was evaluated 24 hours after TAVR by the analysis of vasodilator-stimulated
phosphoprotein (VASP) phosphorylation by flow cytometry as previously described.[16]
Clinical Events
Bleeding complications were assessed according to the Valve Academic Research Consortium-2
(VARC-2) definitions and were classified as follows: life-threatening bleeding, major
bleeding, and minor bleeding. Major adverse cardiac and cerebrovascular events (MACCEs)
were defined as a composite of all-cause death, myocardial infarction, stroke, and
heart failure hospitalization. In-hospital events were collected by careful reviewing
of the patient electronic medical records. Patients' follow-up data were obtained
through telephone interviews from the patients or their family, the cardiologist,
the family physician, or by hospital records.
The primary endpoint of the present analysis was major/life-threatening bleeding complications
(MLBCs) at 1 year after TAVR. Secondary endpoint was MACCE at 1 year.
Statistical Analysis
Categorical variables are represented as frequencies and percentages, and continuous
variables are expressed as mean ± standard deviation or median and interquartile values.
Differences between two groups were assessed with χ2 tests or Fischer's exact tests for categorical variables. Continuous variables with
normal distributions were compared between groups using unpaired Student's t-test. The Wilcoxon test was used to analyze continuous variables with skewed distributions.
The primary and secondary endpoints were compared with the use of a log-rank test
and the hazard ratio (HR), with a 95% confidence interval (CI), was calculated from
a Cox proportional hazard model. Factors associated with a p-value <0.05 in the univariate analysis were included in the multivariate analysis.
In patients with AF, another multivariate analysis was performed, forcing known bleeding
risk factors in the model.[17]
p-Values of <0.05 were considered to indicate statistical significance. All analyses
were performed using JMP 13 software (SAS Institute, Cary, North Carolina, United
States).
Results
Study Population
Among 1,125 consecutive patients undergoing TAVR, 878 patients were enrolled in the
present study ([Fig. 1]). Of these, groups of non-AF and AF included 431 (49%) and 447 (51%) patients, respectively.
In the AF group, 46 (10%) had a new-onset AF following TAVR. The cohort was then split
into four subgroups according to AF and postprocedural CT-ADP >180 seconds ([Fig. 1]). Clinical follow-up at 1 year was completed in 873 patients (99.4%).
Patient Characteristics and Clinical Outcomes in AF and Non-AF Patients
More comorbidities and higher risk scores were found in patients with AF compared
to those without AF ([Supplementary Table S1]). As expected, AF patients were likely to receive OAC therapy on admission. Preprocedural
CT-ADP levels were lower in the AF group, which can partly be explained by the lower
mean aortic gradient before TAVR and the lower incidence of clopidogrel therapy on
admission. Selections of valves and access site were similar between the two groups
([Supplementary Table S2]). Lower postprocedural CT-ADP levels, higher platelet reactivity index (PRI)-VASP
levels, and lower incidence of clopidogrel use at discharge were observed in AF patients.
In contrast, AF patients were likely to receive OAC at discharge. As shown in [Fig. 2], the cumulative event rates of all-cause mortality, heart failure hospitalization,
MACCE, and MLBCs at 1 year were significantly higher in AF patients than in non-AF
patients.
Fig. 2. Cumulative incidence of (A) all-cause mortality, (B) heart failure hospitalization, (C) MACCE, and (D) major/life-threatening bleeding complications were significantly higher in AF patients
than in non-AF patients. AF, atrial fibrillation; MACCE, major adverse cardiac and
cerebrovascular events.
Patient Characteristics in Four Groups According to AF and Postprocedural CT-ADP 180
Seconds
In the non-AF group, lower levels of Hb and higher levels of CT-ADP at baseline were
evidenced in patients with postprocedural CT-ADP >180 seconds compared to those with
postprocedural CT-ADP ≤180 seconds ([Table 1]). Lower levels of PRI-VASP and Hb after TAVR, and higher incidence of significant
PVL were found in patients with postprocedural CT-ADP >180 seconds as well ([Table 2]). Consistent with the non-AF group, higher CT-ADP levels at baseline and lower PRI-VASP
value after TAVR ([Table 2]) were found in AF patients with postprocedural CT-ADP >180 seconds. In line with
this finding, more patients received clopidogrel at discharge and less patients received
OAC at discharge in those patients. Lower Hb levels and higher incidence of significant
PVL at 1 month after TAVR were also found in patients with postprocedural CT-ADP >180 seconds.
Table 1
Baseline clinical characteristics in four groups according to AF and postprocedural
CT-ADP 180 seconds
|
Non-AF (n = 431)
|
AF (n = 447)
|
|
CT-ADP ≤ 180 s
(n = 300)
|
CT-ADP > 180 s
(n = 131)
|
p-Value
|
CT-ADP ≤ 180 s
(n = 361)
|
CT-ADP > 180 s
(n = 86)
|
p-Value
|
|
Age, y
|
82 ± 7
|
82 ± 8
|
0.90
|
83 ± 6
|
84 ± 7
|
0.82
|
|
Men
|
134 (45)
|
61 (47)
|
0.72
|
172 (48)
|
43 (50)
|
0.69
|
|
Body mass index, kg/m2
|
26.9 ± 5.5
|
26.6 ± 6.0
|
0.69
|
27.7 ± 6.3
|
26.5 ± 4.9
|
0.17
|
|
Logistic EuroSCORE
|
17.8 ± 11.4
|
18.0 ± 11.7
|
0.87
|
20.1 ± 13.5
|
23.1 ± 16.7
|
0.13
|
|
EuroSCORE 2
|
4.9 ± 4.2
|
4.8 ± 4.7
|
0.97
|
5.9 ± 6.5
|
6.7 ± 8.1
|
0.33
|
|
STS score
|
5.8 ± 4.6
|
6.3 ± 5.0
|
0.41
|
6.3 ± 5.0
|
7.7 ± 7.5
|
0.06
|
|
NYHA class III or IV
|
175/298 (59)
|
67/130 (52)
|
0.17
|
228/359 (64)
|
61/85 (72)
|
0.15
|
|
Hypertension
|
250 (83)
|
109 (83)
|
0.97
|
301 (83)
|
79 (92)
|
0.048
|
|
Dyslipidemia
|
187 (62)
|
80 (61)
|
0.80
|
213 (59)
|
56 (65)
|
0.30
|
|
Diabetes mellitus
|
102 (34)
|
48 (37)
|
0.60
|
107 (30)
|
34 (40)
|
0.08
|
|
CKD (eGFR <60 mL/min/1.73 m2)
|
153/299 (51)
|
75/131 (57)
|
0.24
|
238 (66)
|
59 (69)
|
0.64
|
|
Hemodialysis
|
5 (2)
|
4 (3)
|
0.46
|
4 (1)
|
4 (5)
|
0.048
|
|
Prior myocardial infarction
|
37 (12)
|
13 (10)
|
0.47
|
44 (12)
|
14 (16)
|
0.31
|
|
Prior PCI
|
102 (34)
|
42 (32)
|
0.69
|
112 (31)
|
35 (41)
|
0.09
|
|
Prior CABG
|
35 (12)
|
13 (10)
|
0.60
|
49 (11)
|
10 (12)
|
0.83
|
|
Peripheral artery disease
|
75 (25)
|
28 (21)
|
0.41
|
117 (32)
|
22 (26)
|
0.22
|
|
Prior stroke
|
48 (16)
|
13 (10)
|
0.10
|
55 (15)
|
16 (19)
|
0.44
|
|
COPD
|
43 (14)
|
16 (12)
|
0.56
|
63 (17)
|
10 (12)
|
0.19
|
|
CHA2DS2VASC score
|
–
|
–
|
–
|
4.7 ± 1.3
|
4.8 ± 1.5
|
0.36
|
|
Medication at baseline
|
|
|
|
|
|
|
|
DAPT
|
96 (32)
|
48 (37)
|
0.35
|
43 (12)
|
19 (22)
|
0.01
|
|
Clopidogrel
|
92 (31)
|
43 (33)
|
0.66
|
50 (14)
|
19 (22)
|
0.06
|
|
OAC
|
18 (6)
|
2 (2)
|
0.04
|
287 (80)
|
50 (58)
|
<0.001
|
|
VKA
|
12 (4)
|
1 (1)
|
0.12
|
195 (54)
|
33 (38)
|
0.01
|
|
DOAC
|
6 (2)
|
1 (1)
|
0.68
|
92 (25)
|
17 (20)
|
0.27
|
|
Echocardiography data at baseline
|
|
Mean aortic gradient, mmHg
|
49 ± 15
|
48 ± 14
|
0.65
|
45 ± 12
|
44 ± 12
|
0.63
|
|
LVEF, %
|
53 ± 16
|
54 ± 14
|
0.68
|
52 ± 15
|
53 ± 16
|
0.63
|
|
LVEDD, mm
|
50 ± 8
|
50 ± 8
|
0.93
|
50 ± 8
|
50 ± 8
|
0.92
|
|
LVESD, mm
|
34 ± 9
|
35 ± 9
|
0.64
|
35 ± 10
|
34 ± 9
|
0.36
|
|
Types of aortic stenosis
|
|
High-gradient
|
255/298 (86)
|
109 (83)
|
0.53
|
287 (80)
|
69 (80)
|
0.88
|
|
Low-flow, low-gradient with reduced ejection fraction
|
31/298 (10)
|
17 (13)
|
0.44
|
48 (13)
|
7 (8)
|
0.19
|
|
Low-flow, low-gradient with preserved ejection fraction
|
4/298 (1)
|
3 (2)
|
0.48
|
20 (6)
|
6 (7)
|
0.61
|
|
Laboratory tests at baseline
|
|
WBC, 103/mm3
|
7.5 ± 2.2
|
7.2 ± 2.1
|
0.09
|
7.6 ± 2.1
|
6.9 ± 2.0
|
0.004
|
|
Hb, g/dL
|
12.4 ± 1.7
|
11.8 ± 1.5
|
<0.001
|
12.2 ± 1.7
|
11.5 ± 1.9
|
<0.001
|
|
Platelet count, 103/mm3
|
235 ± 76
|
223 ± 61
|
0.11
|
233 ± 72
|
201 ± 55
|
<0.001
|
|
CRP, mg/dL
|
0.4 (0.4–0.8)
|
0.4 (0.4–0.6)
|
0.29
|
0.4 (0.4–1.1)
|
0.5 (0.4–1.6)
|
0.19
|
|
Creatinine, µmol/L
|
88 (72–119)
|
95 (77–121)
|
0.16
|
103 (82–134)
|
123 (88–149)
|
0.07
|
|
eGFR, mL/min/1.73 m2
|
59 ± 21
|
56 ± 22
|
0.13
|
52 ± 19
|
48 ± 21
|
0.09
|
|
CT-ADP, s
|
193 ± 76
|
243 ± 72
|
<0.001
|
161 ± 68
|
212 ± 77
|
<0.001
|
Abbreviations: AF, atrial fibrillation; CABG, coronary artery bypass grafting; CKD,
chronic kidney disease; COPD, chronic obstructive pulmonary disease; CRP, C-reactive
protein; CT-ADP, closure time of adenosine diphosphate; DAPT, dual antiplatelet therapy;
DOAC, direct oral anticoagulant; eGFR, estimated glomerular filtration rate; Hb, hemoglobin;
LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction;
LVESD, left ventricular end-systolic diameter; NYHA, New York Heart Association; OAC,
oral anticoagulant; PCI, percutaneous coronary intervention; VKA, vitamin K antagonist;
WBC, white blood cell.
Note: Values are n (%) or n/N (%), mean ± SD, or median (interquartile range).
Table 2
Procedural and postprocedural characteristics in four groups according to AF and postprocedural
CT-ADP 180 seconds
|
Non-AF (n = 431)
|
AF (n = 447)
|
|
CT-ADP ≤ 180 s (n = 300)
|
CT-ADP > 180 s (n = 131)
|
p-Value
|
CT-ADP ≤ 180 s (n = 361)
|
CT-ADP > 180 s (n = 86)
|
p-Value
|
|
Procedural characteristics
|
|
Femoral approach
|
265/299 (89)
|
122/131 (93)
|
0.15
|
322 (92)
|
81 (94)
|
0.49
|
|
Prothesis type
|
|
Balloon-expandable
|
178 (59)
|
78 (60)
|
0.97
|
219 (61)
|
55 (64)
|
0.57
|
|
Self-expandable
|
122 (41)
|
53 (40)
|
0.97
|
142 (39)
|
31 (36)
|
0.57
|
|
Prosthesis diameter, mm
|
27 ± 3
|
26 ± 3
|
0.67
|
27 ± 3
|
26 ± 3
|
0.07
|
|
Sheath diameter, mm
|
15 ± 2
|
15 ± 2
|
0.53
|
15 ± 2
|
15 ± 2
|
0.90
|
|
Postprocedural characteristics
|
|
24 hours after TAVR
|
|
CT-ADP, sec
|
116 ± 28
|
278 ± 39
|
<0.001
|
112 ± 28
|
268 ± 43
|
<0.001
|
|
PRI-VASP, %
|
67 ± 15
|
57 ± 20
|
<0.001
|
73 ± 14
|
67 ± 19
|
0.007
|
|
At discharge
|
|
WBC, 103 /mm3
|
7.8 ± 2.7
|
7.9 ± 4.0
|
0.85
|
7.9 ± 2.7
|
7.3 ± 2.4
|
0.06
|
|
Hb, g/dL
|
10.4 ± 1.4
|
9.9 ± 1.4
|
<0.001
|
10.3 ± 1.4
|
9.9 ± 1.2
|
0.01
|
|
CRP, mg/dL
|
1.0 (1.0–1.1)
|
1.1 (1.0–1.1)
|
0.52
|
1.1 (1.0–1.1)
|
1.1 (1.0–1.1)
|
0.44
|
|
Creatinine, µmol/L
|
54 (28–90)
|
48 (26–85)
|
0.09
|
57 (33–94)
|
48 (25–86)
|
0.12
|
|
Medication at discharge
|
|
Aspirin
|
292/299 (98)
|
129/131 (98)
|
0.73
|
337 (93)
|
78 (91)
|
0.39
|
|
Clopidogrel
|
259/299 (87)
|
122/131 (93)
|
0.05
|
68 (19)
|
35 (41)
|
<0.001
|
|
OAC
|
21/299 (7)
|
5/131 (4)
|
0.20
|
327 (91)
|
58 (67)
|
<0.001
|
|
VKA
|
12/299 (4)
|
1/131 (1)
|
0.12
|
194 (54)
|
39 (45)
|
0.19
|
|
DOAC
|
9/299 (3)
|
4/131 (3)
|
1.00
|
133 (37)
|
19 (22)
|
0.01
|
|
Number of antithrombotic therapies
|
|
|
0.30
|
|
|
0.03
|
|
0
|
3/299 (1)
|
1/131 (1)
|
|
0 (0)
|
1 (1)
|
|
|
1
|
20/299 (7)
|
5/131 (4)
|
|
16 (4)
|
8 (9)
|
|
|
2
|
276/299 (92)
|
124/131 (95)
|
|
319 (88)
|
68 (79)
|
|
|
3
|
0/299 (0)
|
1/131 (1)
|
|
26 (7)
|
9 (10)
|
|
|
Echocardiography at 1 month (n = 820)
|
|
LVEF, %
|
59 ± 12
|
61 ± 11
|
0.15
|
56 ± 12
|
58 ± 10
|
0.19
|
|
LVEDD, mm
|
50 ± 8
|
50 ± 7
|
0.67
|
51 ± 8
|
52 ± 7
|
0.60
|
|
LVESD, mm
|
33 ± 9
|
32 ± 8
|
0.26
|
35 ± 10
|
35 ± 8
|
0.65
|
|
Mean prosthetic valve gradient, mmHg
|
11 ± 5
|
11 ± 7
|
0.23
|
9 ± 4
|
10 ± 5
|
0.10
|
|
Significant PVL (> mild)
|
20/285 (7)
|
18/120 (15)
|
0.01
|
21/335 (6)
|
19/81 (23)
|
<0.001
|
Abbreviations: AF, atrial fibrillation; CRP, C-reactive protein; CT-ADP, closure time
of adenosine diphosphate; Hb, hemoglobin; DOAC, direct oral anticoagulant; LVEDD,
left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction;
LVESD, left ventricular end-systolic diameter; OAC, oral anticoagulant; PRI-VASP,
platelet reactivity index by vasodilator-stimulated phosphoprotein; PVL, paravalvular
leak; TAVR, transcatheter aortic valve replacement; VKA, vitamin K antagonist; WBC,
white blood cell.
Note: Values are n (%) or n/N (%), mean ± SD, or median (interquartile range).
Clinical Outcomes in Four Groups According to AF and Postprocedural CT-ADP 180 seconds
Among non-AF patients, postprocedural CT-ADP >180 seconds was associated with increased
MLBCs during hospitalization, at 1 year, and at 2 year after TAVR ([Supplementary Table S3]). Likewise, the impact of postprocedural CT-ADP >180 seconds on in-hospital, 1-year,
and 2-year MLBCs was evidenced in AF patients ([Supplementary Table S3]). Kaplan–Meier curves in [Fig. 3] clarify that the patients with AF and postprocedural CT-ADP >180 seconds had an
increased rate of cumulative adverse events, remarkably in MLBCs, compared to the
other three groups. Consistently, Kaplan–Meier curves in [Supplementary Fig. S1] demonstrate an increased cumulative bleeding event in patients with prior OAC and
postprocedural CT-ADP >180 seconds, whereas no additional impact of prior DAPT over
postprocedural CT-ADP >180 seconds was found.
Fig. 3. Cumulative event rates of all-cause mortality (A), heart failure hospitalization (B), MACCE (C), and major/life-threatening bleeding events (D) in four groups according to AF and postprocedural CT-ADP >180 seconds. A drastic
increase of bleeding events was evidenced in patients with AF + CT-ADP >180 seconds.
AF, atrial fibrillation; CT-ADP, closure time adenosine diphosphate; MACCE, major
adverse cardiac and cerebrovascular events.
Predictors of MLBCs in AF and Non-AF Patients
In non-AF patients, prior myocardial infarction, mean aortic gradient, Hb levels,
and platelet count were independent predictors of MLBCs at 1 year, whereas postprocedural
CT-ADP >180 seconds turned out to be insignificant after the adjustment ([Supplementary Table S4]). In contrast to the non-AF cohort, the multivariate analysis demonstrates that
postprocedural CT-ADP >180 seconds was significantly associated with MLBCs at 1 year
in AF patients ([Supplementary Table S5], Model 1). Known bleeding risk factors in AF patients including the components of
HASBLED score[17] and OAC administration were forced in [Supplementary Table S5], Model 2, reinforcing the impact of postprocedural CT-ADP >180 seconds on MLBCs
at 1 year.
Predictors of MLBCs and MACCE in the Whole Cohort
After adjustment of clinical variables, Hb levels at baseline and AF + postprocedural
CT-ADP >180 seconds remained independent predictors of 1-year MLBCs ([Fig. 4] and [Supplementary Table S6]). For 1-year MACCE, elevated CRP levels and a nonfemoral approach were predictive
factors, whereas AF + CT-ADP >180 seconds was no longer associated with MACCE after
the adjustment ([Fig. 4] and [Supplementary Table S7]).
Fig. 4. Multivariate Cox regression analyses illustrate that patients with AF and ongoing
primary hemostatic disorder (postprocedural CT-ADP >180 seconds) were associated with
an increased risk of major/life-threatening bleeding complications at follow-up, whereas
no impact was evidenced in MACCE. CT-ADP, closure time adenosine diphosphate; MACCE,
major adverse cardiac and cerebrovascular events.
Late Bleeding Events Following TAVR
[Fig 5] illustrates the cumulative event rates of late bleeding events (>30 days after TAVR).
AF patients were prone to late MLBCs compared to non-AF patients ([Fig. 5A]), which was mainly driven by the patients with postprocedural CT-ADP >180 seconds
([Fig. 5B]). [Fig. 5(C, D)] emphasizes that the late bleeding events in patients with postprocedural CT-ADP
>180 seconds were mainly due to continuous OAC therapy after TAVR. On the other hand,
an additional impact of continuous DAPT over ongoing primary hemostasis disorder on
late bleeding events was not clarified ([Supplementary Fig. S2]). Gastrointestinal bleeding was the most frequent type of late MLBCs (62%), followed
by intracranial and nose or mouth bleedings ([Supplementary Table S8]).
Fig. 5. Cumulative event rates of late major/life-threatening bleeding events (>30 days after
TAVR) in AF and non-AF patients (A), in four groups according to AF and postprocedural CT-ADP >180 seconds (B), in patients with and without OAC (C), and in four groups according to the use of OAC at discharge and postprocedural
CT-ADP >180 seconds (D). AF, atrial fibrillation; CT-ADP, closure time adenosine diphosphate; OAC, oral
anticoagulant; TAVR, transcatheter aortic valve replacement.
Patients with and without Clopidogrel at Discharge
Given the strong relationship between the use of clopidogrel and prolonged CT-ADP
levels ([Table 2]), a sensitivity analysis was performed with patients with and without clopidogrel
at discharge. As shown in [Supplementary Fig. S3], the deleterious effects of prolonged CT-ADP levels (>180 seconds) in AF patients
were consistently observed in patients with and without clopidogrel at discharge.
Although the magnitude of AF + postprocedural CT-ADP >180 seconds in predicting 1-year
MLBCs was higher in patients without clopidogrel at discharge (HR: 4.40; 95% CI: 2.66–7.15)
compared to patients with clopidogrel at discharge (HR: 3.41; 95% CI: 1.50–8.21),
the HRs were not statistically different between the two cohorts (p
interaction = 0.59).
Discussion
The principal findings of this study are summarized as follows: first, patients with
AF after TAVR were associated with a higher rate of long-term adverse clinical events
including bleedings as compared with non-AF patients. Second, postprocedural CT-ADP
>180 seconds was a major determinant of MLBCs in AF patients but not in non-AF patients.
Third, AF + postprocedural CT-ADP >180 seconds was strongly associated with late bleeding
events after TAVR. These findings suggest that ongoing primary hemostatic disorders
as evaluated by CT-ADP could be a reliable predictor of long-term bleeding events
particularly in TAVR patients with AF.
Atrial Fibrillation in TAVR Patients
While the reported incidence of pre-existing AF in TAVR patients ranges from 16 to
51%,[5]
[6]
[18] data on new-onset AF are more scattered compared with pre-existing AF, being reported
between 1 and 32%.[6]
[19] In line with these findings, pre-existing or new-onset AF was found in 51% of our
cohort.
Although previous investigations have shown conflicting results concerning the relationship
between AF and bleeding events,[6] a large registry including 72,660 patients who underwent non-apical TAVR clarified
that both pre-existing and new-onset AF were associated with increased risk of bleeding
events.[19] The bleeding diathesis in AF patients can be partly explained by the comorbidities
and multiple antithrombotic therapy.[6]
[19]
[20] Moreover, a recent report by Lother et al delineated that the incidence of bleeding
defines the outcome of patients with AF after transfemoral TAVR,[5] indicating that particular care in bleeding prevention is indispensable for TAVR
patients with AF. Our current study highlights the high-risk profile of AF patients
treated with multiple antithrombotic therapies following TAVR, which may have led
to an increased risk of MLBCs at follow-up.
Primary Hemostatic Disorder in TAVR Patients
Among various factors that could favor bleeding events following TAVR, special attention
was given to the role played by primary hemostatic disorders. CT-ADP is a primary
hemostasis point-of-care test which can be evaluated by the PFA-100.[21] The system aspirates citrated whole blood at high shear rates through disposable
cartridges containing an aperture within a membrane coated with collagen and ADP which
serves as a platelet-stimulating agent. Studies have indicated that the clinical performance
of CT-ADP is excellent in both inherited and acquired von Willebrand disease (vWD).[9]
[22] Importantly, the value will not be influenced by aspirin or OACs.
Severe AS induces a very turbulent blood flow at the vicinity of the aortic valve,
causing high shear stress that promotes deployment and cleavage of von Willebrand
factor (VWF)'s high-molecular-weight (HMW)-multimers results in an increase of CT-ADP.[23] Moreover, a significant PVL (>mild) can also induce HMW-multimer defects because
of its turbulent flow. Although postprocedural CT-ADP >180 seconds has been predictive
of the presence of significant PVL with early generation TAVR valves,[9] our data demonstrate a stepwise decrease of significant PVL with newer valves across
the last 5 years, whereas the prevalence of patients with postprocedural CT-ADP >180 seconds
holds more than 20% ([Supplementary Fig. S4]). This result suggests that the mild relationship between significant PVL and increased
CT-ADP levels in the current TAVR era may have mitigated the impact of CT-ADP >180 seconds
on adverse cardiac events ([Fig. 4]).[24] Several mechanisms could be responsible for the discrepancy between significant
PVL and postprocedural CT-ADP levels; first, mild PVL may also have an impact on CT-ADP
levels. Our data demonstrate a higher incidence of postprocedural CT-ADP >180 seconds
in patients with mild PVL compared to those without PVL at discharge (29 vs. 19%)
and at 1-month follow-up (28 vs. 19%), suggesting that the defects of VWF HMW-multimers
may also be induced by mild PVL. Second, high shear stress-induced VWF HMW-multimer
defects can be created by other types of valvular heart disease[23]
[25] which were not fully evaluated in our cohort. Third, the CT-ADP may also predict
the bleeding tendency resulting from functional platelet alteration in patients with
defects other than vWD.[22] Taken together, postprocedural CT-ADP levels may exhibit persistent platelet dysfunction
which can mainly affect AF patients who are likely to have multiple comorbidities
and receive OACs after TAVR.
Antithrombotic Therapy in TAVR Patients
While prior guidelines on antithrombotic therapy in TAVR patients have been mostly
based on expert opinion, latest insights were provided by recent randomized trials.
The GALILEO trial[26] randomized patients without an indication of OAC to low-dose rivaroxaban plus aspirin
for 3 months, followed by rivaroxaban alone versus aspirin plus clopidogrel for 3
months, followed by aspirin alone. The trial was terminated because of 69% relative
increase in mortality and 50% increase in MLBCs in the rivaroxaban arm. In POPULAR
TAVI trial cohort B,[20] patients undergoing TAVR with an indication of OAC (mainly AF 94.9%) were randomized
to receive either OAC alone or OAC plus clopidogrel for 3 months. The incidence of
serious bleeding was lower with OAC alone than with OAC plus clopidogrel. Altogether,
these two trials reinforced the detrimental effects of OAC when it is combined with
antiplatelet therapies. Given these results, the 2021 ESC/EACTS (European Society
of Cardiology/European Association for Cardio-Thoracic Surgery) guidelines recommend
single OAC following TAVR in patients with an indication of OAC and single antiplatelet
therapies (SAPTs) in patients without indication of OAC.[13] In addition, a recent consensus document from the ESC Working Group proposes OAC
plus SAPT for patients with an indication of OAC when PCI has been performed within
3 months.[27] Although a tailored antithrombotic therapy is recommended in those patients according
to the bleeding diathesis, an indicator of future bleeding events is lacking and an
optimal dose of OAC remains unclear.
Although the noxious impact of OAC in addition to antiplatelet therapy on bleeding
events was evidenced in recent trials, the extent of platelet functional defect has
not been explored. We previously described the extent of platelet inhibition by clopidogrel
in TAVR patients, showing that an appropriate response to clopidogrel (PRI-VASP ≤50%)
was contributed to major bleeding events.[28] In contrast, CT-ADP reflects a broader spectrum of platelet function, including
defects of VWF HMW-multimers, various platelet disorders, and platelet inhibition
by P2Y12 inhibitors.[7]
[22] On a pragmatical approach, our present data clearly suggest that the inhibition
of secondary hemostasis by OAC over primary hemostatic disorder may hamper the hemostasis
process extensively, resulting in increasing the prevalence of bleeding complications.
Therefore, a reduced amount of OAC may be favorable for patients with prolonged CT-ADP
levels and a shorter duration of dual antithrombotic therapy can be considered in
patients with recent coronary stenting.
It is noteworthy that the latest evidence from ENVISAGE-TAVI AF trial[29] showed a higher incidence of major bleeding events with edoxaban than with VKA in
AF patients who underwent TAVR. Interestingly, patients who received low edoxaban
dose (30 mg once daily) had similar incidence of major bleeding to those who received
VKA, suggesting that that a lower dose of DOAC may mitigate the late bleeding risk
after TAVR. In the latest ESC/EACTS guidelines for the management of AF,[14] a reduced dose of DOAC is recommended based on the bleeding risk profile of AF patients.
While our study emphasizes that ongoing primary hemostatic disorder, evaluated by
postprocedural CT-ADP, may also be a predictor of future bleeding events in TAVR patients
with AF, further trials are required to elucidate whether similar dose reduction can
be safely applied to those patients.
Study Limitations
We acknowledge the following limitations: first, the analyses were performed on the
basis of a single center data set with uncertain generalizability. Second, CT-ADP
prolongation could be influenced by the presence of low Hb levels, low platelet count,
and P2Y12 inhibitors. Nevertheless, the noxious impact of CT-ADP >180 seconds on bleeding events
was confirmed by adjusting the Hb levels and platelet count in the multivariate analysis
and the sensitivity analysis with and without clopidogrel at discharge. Third, assessments
of CT-ADP were only performed 24 hours after TAVR and were not repeated during the
follow-up. Forth, since the majority of patients received dual antithrombotic therapy
at discharge (90%) ([Supplementary Table S9]), we were not able to investigate the relationship between the number of antithrombotic
therapies and bleeding events. Fifth, the antithrombotic regimen at the time of the
events was not captured. Sixth, the HASBLED score, which is an established bleeding
risk score, was not collected in our registry. Thus, the classical bleeding risk in
AF patients with and without CT-ADP >180 seconds remains uncertain. However, several
components of the score were available in our study, showing no difference between
the two groups. Seventh, certain blood parameters such as serum iron concentration[30] or red blood cell distribution[31] are associated with bleeding complications following TAVR but were not collected
in our registry. Eighth, multivariate analysis should be interpreted with caution
because of the limited sample size and the nature of observational study. There are
inherent limitations in this type of study, mainly related to unknown confounding
factors. Thus, further investigations are required to validate the CT-ADP-guided care
after TAVR.[32]
Conclusion
Ongoing primary hemostatic disorder, evaluated by postprocedural CT-ADP, was independently
associated with MLBCs among TAVR patients with AF. A large majority of late MLBCs
occurred in patients with OAC and postprocedural CT-ADP >180 seconds. These findings
indicate that a better individualized and risk-adjusted antithrombotic therapy may
be considered for reducing bleeding events in TAVR patients with prolonged CT-ADP
levels.
What Is Known about This Topic?
-
A tailored antithrombotic therapy in transcatheter aortic valve replacement (TAVR)
patients is needed for atrial fibrillation (AF) patients according to the bleeding
diathesis, but an indicator of future bleeding events is lacking.
-
In TAVR patients, prolonged closure time of adenosine diphosphate (CT-ADP) in AF patients
was associated with major/life-threatening bleeding complications at follow-up.
-
Likewise, an increased bleeding risk was found in patients with oral anticoagulant
(OAC) and prolonged CT-ADP.
What Does This Paper Add?
-
Our present data suggest that the inhibition of secondary hemostasis by OAC over primary
hemostatic disorder may hamper the hemostasis process extensively, resulting in increasing
the prevalence of bleeding complications.
-
A tailored and risk-adjusted antithrombotic therapy after TAVR may be considered particularly
in AF patients with primary hemostatic disorder.
