Impact of inhospital stent thrombosis and

cerebrovascular accidents on long-term prognosis

after percutaneous coronary intervention
Elena Guerra, MD, a Gjin Ndrepepa, MD, a Stefanie Schulz, MD, a Robert Byrne, MD, a Petra Hoppmann, MD, b
Sebastian Kufner, MD, a Tareq Ibrahim, MD, b Tomohisa Tada, MD, a Heribert Schunkert, MD, a,c
Karl-Ludwig Laugwitz, MD, b,c and Adnan Kastrati, MD a,c Munich, Germany

Background Inhospital stent thrombosis (ST) and cerebrovascular accidents (CVA) are rare but serious adverse events after
percutaneous coronary intervention (PCI). The association of ST or CVA with long-term outcome after PCI remains poorly investigated.
Methods The study included 18,334 consecutive patients who underwent PCI. Patients were divided into 3 groups: the group
with ST, the group with CVA, and the group without these events. The primary outcome was all-cause mortality at 3-year follow-up.
Results Inhospital ST or CVA occurred in 59 patients (0.32%) and in 90 patients (0. 49%), respectively. There were 2,149
deaths (11.7%) during the follow-up: 26 deaths among patients with ST, 32 deaths among patients with CVA, and 2,091 deaths
among patients without ST or CVA (Kaplan-Meier estimates of 3-year mortality 45.3%, 38.0%, and 12.9%, odds ratio 6.1, 95% CI
3.6-10.2, P b .001 for ST group vs the group without ST or CVA and odds ratio 4.2 [2.7-6.6], P b .001 for CVA group vs the group
without ST or CVA). There was no significant difference in the 3-year mortality between CVA and ST groups (P = .29). The Cox
proportional hazards model showed that ST (adjusted hazard ratio 4.97, 95% CI 2.58-9.56, P b .001) and CVA (adjusted hazard
ratio 2.25 [1.25-4.04], P = .006) were independently associated with the increased risk of 3-year mortality.
Conclusion Inhospital ST and CVA after PCI are associated with the increased risk of 3-year mortality. Both events seem
to have a similar impact on long-term survival. (Am Heart J 2014;168:862-868.e1.)

Cerebrovascular accidents (CVAs) and stent thrombosis
(ST) are rare but serious complications of percutaneous
coronary intervention (PCI). Despite significant improve-
ments in the device technology and peri-PCI adjunctive
pharmacologic therapy, the incidence of CVA related to PCI
appears to have remained relatively constant over the last
20 years, perhaps due to the increased complexity of
patients undergoing PCI, which may counterbalance the
advantages of improved technology. The incidence of
inhospital CVA was reported to be between 0.22% and
1.4%. 1-4 Although hemorrhagic stroke is closely related to
the use of anticoagulant and thrombolytic therapy, the
mechanisms of ischemic CVA after PCI are poorly under-
stood. Clinical trials and registries have shown that ST is
From the aDeutsches Herzzentrum München, Technische Universität, Munich, Germany,
b
1.Medizinische Klinik rechts der Isar, Technische Universität, Munich, Germany, and
c
DZHK (German Centre for Cardiovascular Research), partner site Munich Heart Alliance,
Munich, Germany.
Submitted April 29, 2014; accepted August 16, 2014.
Reprint requests: Adnan Kastrati, MD, Deutsches Herzzentrum München, Lazarettstrasse 36,
D-80636 Munich, Germany.
E-mail: kastrati@dhm.mhn.de
0002-8703
© 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ahj.2014.08.015
associated with higher rates of morbidity and mortality. 5,6
Although few studies have investigated the impact of
inhospital ST on outcome, there is evidence that early
(inhospital) ST may have a different impact on prognosis
compared with late or very late ST. 6,7 Prior studies have
identified a series of demographic and procedural factors
that predispose for CVA 2,8 and ST 9 after PCI. Although these
studies showed that there is, at least, a partial overlapping of
predisposing factors for CVA and ST and most CVA and ST
are thrombotic in nature, differences in baseline risk and
characteristics of patients across various studies may hamper
a comparative analysis of their predisposing factors. So far, a
comparative analysis of factors associated with an increased
risk of CVA or ST as well as a comparison of their impact on
prognosis has not been investigated.
The aim of this study was to assess factors associated with
increased risk of inhospital CVA or ST and to investigate and
compare the impact of these adverse events on long-term
prognosis of patients treated with PCI.
Methods
Study patients
Consecutive patients with symptomatic coronary artery
disease admitted in 2 German hospitals for treatment with
American Heart Journal
Volume 168, Number 6
Guerra et al 863

Table I. Baseline characteristics

Variable No ST or CVA (n = 18185) ST (n = 59) CVA (n = 90) P

Age (y) 67.62 (59.20-75.34) 66.47 (56.09-72.62) 71.10⁎ (66.12-77.43) .003

Women 4683 (25.8) 16 (27.1) 36 (40.0)⁎ .008
BMI (kg/m2) 26.67 (24.34-29.40) 26.79 (24.35-29.34) 24.89⁎ (23.34-27.67) b.001
Diabetes 4702 (25.9) 21 (35.6) 37 (41.1)⁎ .001
On insulin therapy 1479 (8.1) 8 (13.6) 14 (15.6)⁎ .012
Arterial hypertension 11655 (64.1) 28 (47.6)⁎ 58 (64.4) .029
Current smoking 3825 (21.0) 15 (25.4) 21 (23.3) .62
Hypercholesterolemia 11060 (60.8) 33 (55.9) 46 (51.1) .13
LVEF (%) 56.0 (42.0-62.0) 48.0⁎ (39.5-55.5) 45.0⁎ (37.25-56.0) b.001
Atrial fibrillation 2287 (12.6) 2 (3.4)⁎ 18 (20.0)⁎ .011
Prior cerebrovascular events 576 (3.2) 1 (1.7) 4 (4.4) .64
Prior myocardial infarction 4387 (24.1) 17 (28.8) 22 (24.4) .70
Prior CABG surgery 2113 (11.6) 5 (8.5) 12 (13.3) .66
Clinical presentation b.001
Stable angina 9546 (52.5) 13 (22.0)⁎ 30 (33.3)⁎
NSTE-ACS 4791 (26.3) 15 (25.4) 23 (25.6)
STEMI 3848 (21.2) 31 (52.6)⁎ 37 (41.1)⁎

Data are expressed as median and interquartile range (25th-75th percentiles) or counts (percentages). Abbreviations: BMI, body mass index; CABG, coronary artery bypass graft;
LVEF, left ventricular ejection fraction; NSTE-ACS, non–ST-segment elevation acute coronary syndrome; STEMI, ST-segment elevation myocardial infarction.
⁎ P b .05 compared with patients without ST or CVA.

PCI between January 2000 and January 2011 were included.
Patients undergoing coronary artery bypass surgery during
the index hospitalization were excluded. Clinical and
angiographic data were prospectively collected and stored
in a dedicated database.
Angiographic examination and PCI
Coronary angiography was performed according to the
standard criteria. 10 Off-line analysis of digital angiograms
was performed in the core laboratory using an automated
edge detection system (CMS; Medis Medical Imaging
Systems, Neuen, the Netherlands) by personnel blinded
to the clinical data. Angiographic analysis was performed
according to the modified American College of Cardiology/
American Heart Association Stenosis Morphology Classifi-
cation. 11 The B2 and C lesions are considered as complex.
Coronary stenting was performed as per standard
practice. Before procedure, all patients received 325 to
500 mg of aspirin and a loading dose of 600 mg of
clopidogrel. Unfractionated heparin or bivalirudin was
used periprocedurally. Post-PCI antithrombotic therapy
consisted of aspirin (80-325 mg/d continuously) and
clopidogrel (150 mg/d until discharge but for no longer
than 3 days followed by 75 mg/d for at least 1 month after
bare-metal stent or ≥6 months after drug-eluting stent
implantation). Other medications were left at the
discretion of the patient's attending physician.
Study definitions
Stent thrombosis. Inhospital ST was defined as a
new onset of ischemic symptoms or new electrocardio-
graphic changes that suggested acute ischemia or a
typical rise and fall in cardiac biomarkers at any time after
the PCI during the index hospitalization, associated with
angiographic confirmation of ST according to the
Academic Research Consortium criteria for definite ST. 12
Cardiovascular accidents. A CVA was defined as the
onset of a new neurologic deficit that occurred any time
after the PCI, during the index hospitalization. All the
neurologic deficits were assessed by an expert neurolo-
gist at the time of events. Stroke was diagnosed according
to the World Health Organization Criteria, which defines
a stroke as focal deficit that lasts for N24 hours. 13 The
neurologic deficit lasting b24 hours was classified as a
transient ischemic attack (TIA). In addition, brain imaging
(computed tomography or magnetic resonance imaging)
was obtained in all stroke patients and in most TIA
patients, to confirm the diagnosis and establish the origin
(ischemic or hemorrhagic) of the neurologic deficit.
Mortality. Deaths occurring during the hospital course
(inhospital) and at 30 days and 3 years after PCI were
analyzed. Information on deaths was acquired from the
hospital records, death certificates, or telephone contact
with the referring physician(s), relatives of the patient,
insurance companies, or registration of address office.
Patients were visited by their physician or interviewed
by telephone at 30 days, 6 months, 1 year, and yearly
thereafter after the PCI procedure.
Statistical analysis
Categorical data are presented as count and percentage
(%). Continuous data are presented as median with 25th to
75th percentiles or as mean ± SD. The normality of
distribution of continuous data was tested with the 1-
sample Kolmogorov-Smirnov test. Continuous data are
compared with analysis of variance test or with the Kruskal-
Wallis rank sum test. For variables that showed a significant
difference, intergroup post hoc comparisons were
 American Heart Journal
864 Guerra et al December 2014

Table II. Angiographic data

Variable No ST or CVA (n = 28476) ST (n = 102) CVA (n = 161) P

No. of narrowed coronary arteries⁎ .002

1 4469 (24.6) 6 (10.2) † 15 (16.7)
2 5305 (29.2) 12 (20.3) 28 (31.1)
3 8411 (46.2) 41 (69.5) † 47 (52.2)
Multivessel disease⁎ 13716 (75.4) 53 (89.8) † 75 (83.3) .008
Vessel affected .51
Bypass graft 707 (2.48) 3 (2.9) 2 (1.2)
Left descending coronary artery 12227 (42.9) 40 (39.2) 73 (45.3)
Left main coronary artery 1180 (4.1) 6 (5.9) 9 (5.6)
Left circumflex coronary 6404 (22.5) 25 (24.5) 45 (27.9)
Left internal mammary artery 7 (0.02) 0 (0) 0 (0)
Right coronary artery 7951 (43.7) 28 (47.5) 32 (33.9)
Restenotic lesion 949 (3.3) 6 (5.9) 7 (4.4) .54
Chronic total occlusion 1355 (4.7) 8 (7.8) 7 (4.4) .91
Complex lesions 22331 (78.4) 93 (91.2) † 137(85.1) † .04
Lesion length (mm) 13.15 (8.90-19.34) 15.26 † (9.45-23.85) 12.44 (7.96-18.32) .016
Reference diameter (mm) 2.85 (2.48-3.25) 2.82 (2.38-3.27) 2.78 (2.44-3.15) .28
Baseline percentage stenosis (%) 66.28 (55.19-79.43) 69.82 (58.17-92.04) 66.04 (56.34-82.40) .057
Stent type .03
Bare-metal stent 10618 (37.3) 41 (40.2) 49 (30.4) †
First-generation DES 5800 (20.4) 26 (25.5) 46 (28.6) †
Second-generation DES 12058 (42.3) 35 (34.3) 66 (41.0)
Total stented length (mm) 24.8 ± 12.1 29.5 ± 17.7 23.6 ± 10.2 b.001
Postintervention MLD (mm) 2.67 (2.33-3.04) 2.54 † (2.23-2.91) 2.61 (2.27-2.95) .019
Residual percentage stenosis (%) 9.51 (5.56-14.03) 11.86 † (6.39-17.98) 9.52 (5.81-14.15) .003

Data are expressed as median (25th-75th percentiles), mean ± SD, or counts (percentages). Abbreviations: DES, drug-eluting stent; MLD, minimal lumen diameter.
⁎ Number of narrowed coronary arteries and multivessel disease refer to the number of patients; the rest of analysis is lesion based.
† P b .05 compared with patients without ST or CVA.

performed using the Tukey test. Categorical variables were
compared with the χ 2 test or Fisher exact test when
expected cell values were b5. Multiple logistic regression
model was used to identify the independent predictors of
CVA and ST. Baseline and angiographic data that differed
with a P b .10 in the univariable comparisons were entered
into the model. Adjusted odds ratios (ORs) with 95% CIs
were used as summary statistics. The Kaplan-Meier method
was used to estimate probability of survival after PCI.
Differences in survival were compared with the log-rank
test. The Cox proportional hazards model was used to
assess the association between ST or CVA with 3-year
mortality while adjusting for potential confounders.
Variables of Tables I and II that differed with a P b .10
were entered into the model. All analyses were performed
with the R 2.15.1 (The R foundation for Statistical
Computing, Vienna, Austria). A 2-sided P b .05 was
considered to indicate statistical significance.
No extramural funding was used to support this work.
The authors are solely responsible for the design and
conduct of this study, all study analyses, the drafting and
editing of the manuscript, and its final contents.
Results
The study included 18,334 consecutive patients who
underwent PCI. During the hospital course, ST or CVA
occurred in 59 patients (0.32%) and 90 patients (0.49%),
respectively. Among 90 CVA events, 69 were classified as
stroke, and 21, as TIA. The time interval (median [25th-
75th percentiles]) between PCI procedure and the
occurrence of ST, stroke, and TIA was 1.4 (0.4-3.3), 1.3
(0.5-3.4), and 0.6 (0.4-3.4) days, respectively. Imaging
examination was done in 84 patients with CVA and
showed that CVAs were of ischemic origin in 64 patients
(76.2%; 58 patients with stroke and 6 patients with TIA)
and of hemorrhagic origin in 11 patients (13.1%). Nine
patients had no evidence of brain damage. All 6 patients
without a brain imaging had TIA. Of them, 1 patient has
left ventricular thrombus, 1 patient has paroxysmal atrial
fibrillation, 3 patients were not submitted to brain
imaging because symptoms lasted only few minutes,
and 1 patient declined imaging test. There were no
differences between men and women regarding the type
of stroke: 39 ischemic CVA versus 7 hemorrhagic CVA in
men and 25 ischemic CVA versus 4 hemorrhagic CVA in
women (P = .77). Two patients developed both ST and
CVA, and both patients were included in the ST group.
Baseline data are shown in Table I. Patients with CVA
compared with patients with ST were more likely to be
older and had a higher frequency of female sex, arterial
hypertension, and atrial fibrillation. For patients with
CVA, ST, and those without these events, the hospital
stay duration (median [25th-75th percentiles]) was 8.0
American Heart Journal
Volume 168, Number 6
Guerra et al 865

Table III. Independent predictors of inhospital ST and CVA Figure 1

Variable Odds ratio (95% CI) P

Inhospital ST
Stent type .025
Second-generation DES vs BMS 0.82 (0.47-1.43)
Second-generation DES vs 0.47 (0.27-0.82)
first-generation DES
Multivessel disease 3.68 (1.47-9.26) .006
Clinical presentation b.001
NSTE-ACS (vs stable angina) 1.90 (1.02-3.51)
STEMI (vs stable angina) 4.51 (2.51-8.13)
Complex lesions 2.33 (1.01-5.41) .049
Lesion length (for 1-mm increase) 1.29 (1.04-1.60) .019
Inhospital CVA
Women 1.56 (1.07-2.28) .019
Clinical presentation .003
NSTE-ACS (vs stable angina) 1.67 (1.09-2.55)
STEMI (vs stable angina) 2.13 (1.36-3.33) Inhospital and 30-day mortality among patients with ST, patients with
BMI (for 5 kg/m 2 increase) 0.68 (0.54-0.86) .001 CVA, and patients without ST or CVA.
Left ventricular ejection fraction 1.64 (1.31-2.04) b.001
(for 10% decrease)

Abbreviation: BMS, bare metal stent.

[0.97-2.40], P = .06) and diabetes (OR 1.39 [0.96-2.02], P

(4.1-13.8), 11.1 (6.4-18.8), and 3.2 (2.0-5.8) days, respec-
tively. For patients with stable coronary artery disease,
non–ST-segment elevation acute coronary syndromes
(ACS) and ST-segment elevation myocardial infarction,
the hospital stay duration was 2.9 (2.0-4.1), 3.2 (2.1-5.1),
and 6.8 (4.9-9.6) days, respectively.
Angiographic data are shown in Table II. There were
differences between the groups regarding the extent of
coronary artery disease, proportions of patients with
complex lesions, total stented length, postintervention
minimal lumen diameter, and residual stenosis. Patients
with ST compared with patients with CVA had longer
stented length and higher residual stenosis. Baseline and
angiographic data for patients with stroke and TIA are
shown in online Appendix Supplementary Tables I and II.
Predictors of ST and CVA
Multivariable logistic regression analysis (see Methods
section for variables entered into the analyses) showed
that type of stenting, multivessel disease, clinical presen-
tation (non–ST-segment elevation ACS or ST-segment
elevation myocardial infarction vs stable angina), com-
plex lesions, and length of stented segment were
independently associated with the increased risk for
inhospital ST (Table III). Diabetes (OR 1.51, 95% CI 0.94-
2.43, P = .09) showed a trend toward an association with
the risk of ST. When applied to identify the independent
associates of CVA, the model showed that female sex,
clinical presentation with an ACS, reduced left ventricular
function, and lower body mass index (BMI) were
independently associated with the increased risk for
inhospital CVA (Table III). Atrial fibrillation (OR 1.53
= .08) showed a strong trend toward an association with
the increased risk of inhospital CVA.
Clinical outcome
Inhospital deaths occurred in 14 patients (23.7%) with
ST, 7 patients (7.8%) with CVA, and 368 patients (2.0%)
without ST or CVA (OR 15.0 [8.2-27.7], P b .001 for the
ST group vs the group without ST or CVA and OR 4.1 [1.9-
8.9], P b .001 for the CVA group vs the group without ST
or CVA). At 30 days, there were 17 deaths (28.8%) among
patients with ST, 14 deaths (15.5%) among patients with
CVA, and 655 deaths (3.6%) among patients without ST
or CVA (OR 10.8 [6.1-19.1], P b .001 for the ST group vs
the group without ST or CVA and OR 4.9 [2.8-8.9], P b
.001 for the CVA group vs the group without ST or CVA)
(Figure 1). There were 2,149 deaths (11.7%) at 3 years of
follow-up: 26 deaths among patients with ST, 32 deaths
among patients with CVA, and 2,091 deaths among patients
without ST or CVA. The Kaplan-Meier estimates of 3-year
mortality were 45.3%, 38.0%, and 12.9%, for patients with
ST, with CVA, and without ST or CVA, respectively (OR 6.1
[3.6-10.2], P b .001 for the ST group vs the group without
ST or CVA and OR 4.2 [2.7-6.6], P b .001 for the CVA group
vs the group without ST or CVA). Kaplan-Meier curves of 3-
year mortality are shown in Figure 2. There was no
difference in 3-year mortality between patients with ST or
CVA (OR 1.43 [0.70-2.79], P = .29). Mortality in patients
with stroke and TIA is shown in online Appendix
Supplementary Table III.
The Cox proportional hazards model was used to assess
the association of ST or CVA with 3-year mortality while
adjusting for potential confounders (see Methods section
for variables entered into the model). The yield of the
Cox model is shown in Table IV. The inhospital ST was

866 Guerra et al
Figure 2
Kaplan-Meier curves of 3-year mortality among patients with ST, patients with CVA, and patients without ST or CVA.
associated with an almost 5-fold increase in the adjusted
risk for 3-year mortality. Inhospital occurrence of CVA
was associated with N2-fold increase in the adjusted risk
for 3-year mortality (Table IV).
Discussion
The main findings of this study may be summarized as
follows: (1) Inhospital ST and CVA occurred in 0.32% and
0.49% of patients undergoing PCI. Both conditions are
associated with increased risk for early (inhospital and 30-
day) and late (3-year) mortality. (2) Inhospital ST and CVA
after PCI seem to have mostly distinct predisposing
factors. The use of first-generation drug-eluting stents,
presentation with an ACS, complexity of lesions, and the
longer stented length were independently associated
with the increased risk for inhospital ST, whereas female
sex, presentation with an ACS, lower BMI, and reduced
left ventricular function were associated with increased
risk of inhospital CVA. The novelty of this analysis may
consist in the use of the same sample of patients to
investigate the predisposing factors for ST or CVA. This
approach eliminates any impact that heterogeneity in
baseline data may have on the occurrence of ST or CVA
after PCI.
Stent thrombosis is one of the most feared complica-
tions after stent implantation, and it is associated with
high rates of morbidity and mortality. Stent thrombosis
occurred in approximately 4% of patients over 2 years
after primary PCI, and more than one-third of all ST events
American Heart Journal
December 2014
occurred in the hospital phase. 6,14,15 A study by Dangas
et al 7 showed that 1-year mortality after ST was
significantly higher among patients with inhospital ST
compared with patients with out-of-hospital ST (27.8% vs
10.8%). Delayed neointimal coverage and local wall
inflammation are important aspects that distinguish
different types of stents and their propensity to ST. 16
Our study showed that, also in the acute phase, the type
of stent was an independent predictor of ST. Particularly,
the risk for ST was higher for first-generation drug-eluting
stents compared with second-generation drug-eluting
stents or bare-metal stents. This finding as well as the
increased risk of ST associated with stent length,
complexity of lesions, and multivessel disease has also
been previously reported. 9,17
The incidence of CVA after PCI has remained steady
during the last 2 decades, and the rate of 0.49% as
reported in the present study is in line, albeit slightly
higher than the rate reported in previous studies. 2,3 Well-
established factors predisposing for CVA include athero-
sclerotic risk factors such as older age, diabetes, arterial
hypertension, and impaired renal function, which are not
related to PCI itself. 2,8 It may be suggested that peri-PCI
and spontaneous CVA may have a common physiopa-
thology, that is, embolization of thrombotic-atheroscle-
rotic material from the aortic arch or aortic arterial
trunks. During the PCI, the manipulation of intravascular
catheter and mechanical adjunctive devices may
facilitate disruption of atherosclerotic debris and their
cerebral embolization.
American Heart Journal
Volume 168, Number 6
Guerra et al 867

Table IV. Results of Cox proportional hazards model applied to was commonly reported in recent studies. 2,8,21 A
assess predictors of 3-year mortality significant reduction of the hemorrhagic events from
Variable Adjusted HR (95% CI) P 46.5% in a 2002 study 3 to 12% in a recent study 4 has been
reported. A low rate of hemorrhagic CVA may be
Inhospital ST 4.97 (2.58-9.56) b.001 explained by exclusion of TIAs in prior studies and by a
Inhospital CVA 2.25 (1.25-4.04) .006 change in clinical practice involving infrequent use of
Age (for 10-y increase) 1.84 (1.72-1.97) b.001
thrombolytic drugs in current practice. 4
Stent type
Second-generation DES vs BMS 0.73 (0.64-0.83) b.001 Consistent with our aim to evidence all neurologic
Second-generation DES 0.92 (0.79-1.08) .32 symptoms, all strokes and TIAs occurring over the
vs first-generation DES inhospital course were assessed. It may be due to this
BMI (for 5 kg/m 2 increase) 0.88 (0.82-0.96) .002 reason that the inhospital mortality in our study was lower
Women 1.06 (0.93-1.22) .33
(7.8%) compared with inhospital mortality reported in
Diabetes 1.72 (1.52-1.94) b.001
Arterial hypertension 0.64 (0.57-0.73) b.001 patients with stroke, ranging from 32.6% 22 to 37.2%. 3
Multivessel disease 1.35 (1.15-1.59) b.001 However, over the long-term follow-up, the 3-year
(vs single-vessel disease) mortality after CVA was significantly higher and compara-
Clinical presentation ble with previous studies that have included patients with
NSTE-ACS (vs stable angina) 1.03 (0.89-1.19) .66
STEMI (vs stable angina) 1.50 (1.30-1.72) b.001
stroke. 3,23 This finding seems to concur with a study by
Left ventricular ejection fraction 1.46 (1.40-1.52) b.001 Hoffman et al 8 that has reported no difference in the long-
(for 10% decrease) term survival between patients with TIA and stroke after
Atrial fibrillation 1.53 (1.32-1.77) b.001 PCI. The present study showed that 3-year mortality
Complex lesions 1.19 (1.06-1.34) .004 differed little among patients with ST and CVA. We believe
Lesion length (for 10-mm increase) 1.00 (0.95-1.05) .88
that the present study provides the first comparison of the
Abbreviation: HR, hazard ratio. impact of ST and CVA on mortality. This finding may have
important implications because it offers information that
not only death due to cardiac causes but also death due to
Our study found that presentation with an ACS was CVA makes a significant contribution to early and late
associated with increased risk for ST and CVA. The mortality after PCI.
association between presentation with an ACS and The present study has some limitations. Although data
increased risk for ST may be explained by the presence are prospectively collected, by design, the present study
of acute thrombus at the culprit lesions, which may represents a retrospective analysis. There is a possibility
predispose for stent malapposition and cause modula- that some oligosymptomatic or nonspecific neurologic
tion of drug release kinetics by superimposed throm- events not attracting medical attention might have
botic material and uptake of lipophilic drugs by escaped from the documentation. Because of the long
necrotic core. 16 The Dutch Stent Thrombosis Registry study period, some changes in the practice of PCI
reported a significant higher cumulative incidence of including pharmacological adjunct therapy and stent
early ST when the indication for index stent implantation technology may have occurred. However, it remains
was unstable angina/non–ST-segment elevation myocar- difficult to assess, and thus, they remain unaccounted for
dial infarction/ST-segment elevation myocardial infarc- in the setting of present study.
tion. 9 Other studies have also shown that ACS appears to In conclusion, inhospital ST and CVA after PCI seem to
be related to CVA as well. 1 Patients with ACS have have mostly distinct predisposing factors. Inhospital ST
circulating activated platelets and platelet-monocyte and CVA after PCI are associated with increased risk for
aggregates that may predispose for ST or thrombosis in early (inhospital and 30-day) and late (3-year) mortality,
different vascular sites including cerebrovascular and both ST and CVA seem to be associated with similar
bed. 18,19 Furthermore, patients with ACS may have reduction of long-term survival.
generalized vascular inflammation, which may be associ-
ated with plaque instability, predisposing for vascular
thrombosis both at the stent site, and in other preexisting Disclosure
nonculprit atherosclerotic plaques in coronary and The authors have no conflict of interest to disclose.
cerebral arterial tree. 20
In accordance with previous studies, 8 we showed that
References
sex female and lower BMI were independently associated
1. Aggarwal A, Dai D, Rumsfeld JS, et al. Incidence and predictors of
with the increased risk for inhospital CVA. Both stroke associated with percutaneous coronary intervention. Am J
conditions are known risk factors for hemorrhagic CVA. Cardiol 2009;104(3):349-53.
However, in the present study, the incidence of 2. Dukkipati S, O'Neill WW, Harjai KJ, et al. Characteristics of
hemorrhagic CVA was rather low and did not differ cerebrovascular accidents after percutaneous coronary interventions.
among men and women. A low rate of hemorrhagic CVA J Am Coll Cardiol 2004;43(7):1161-7.

868 Guerra et al
3. Fuchs S, Stabile E, Kinnaird TD, et al. Stroke complicating
percutaneous coronary interventions: incidence, predictors, and
prognostic implications. Circulation 2002;106(1):86-91.
4. Guptill JT, Mehta RH, Armstrong PW, et al. Stroke after primary
percutaneous coronary intervention in patients with ST-segment
elevation myocardial infarction: timing, characteristics, and clinical
outcomes. Circ Cardiovasc Interv 2013;6(2):176-83.
5. Armstrong EJ, Feldman DN, Wang TY, et al. Clinical presentation,
management, and outcomes of angiographically documented early,
late, and very late stent thrombosis. JACC Cardiovasc Interv 2012;5
(2):131-40.
6. Kimura T, Morimoto T, Kozuma K, et al. Comparisons of baseline
demographics, clinical presentation, and long-term outcome among
patients with early, late, and very late stent thrombosis of sirolimus-
eluting stents: observations from the Registry of Stent Thrombosis for
Review and Reevaluation (RESTART). Circulation 2010;122(1):52-61.
7. Dangas GD, Claessen BE, Mehran R, et al. Clinical outcomes
following stent thrombosis occurring in-hospital versus out-of-hospital:
results from the HORIZONS-AMI (Harmonizing Outcomes with
Revascularization and Stents in Acute Myocardial Infarction) trial.
J Am Coll Cardiol 2012;59(20):1752-9.
8. Hoffman SJ, Holmes Jr DR, Rabinstein AA, et al. Trends, predictors,
and outcomes of cerebrovascular events related to percutaneous
coronary intervention: a 16-year single-center experience. JACC
Cardiovasc Interv 2011;4(4):415-22.
9. van Werkum JW, Heestermans AA, Zomer AC, et al. Predictors of
coronary stent thrombosis: the Dutch Stent Thrombosis Registry. J Am
Coll Cardiol 2009;53(16):1399-409.
10. Task Force on Myocardial Revascularization of the European Society
of Cardiology (ESC) and the European Association for Cardio-
Thoracic Surgery (EACTS), European Association for Percutaneous
Cardiovascular Interventions (EAPCI), Wijns W, et al. Guidelines on
myocardial revascularization. Eur Heart J 2010;31(20):2501-55.
11. Ellis SG, Vandormael MG, Cowley MJ, et al. Coronary morphologic
and clinical determinants of procedural outcome with angioplasty for
multivessel coronary disease. Implications for patient selection.
Multivessel Angioplasty Prognosis Study Group. Circulation 1990;82
(4):1193-202.
American Heart Journal
December 2014
12. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in
coronary stent trials: a case for standardized definitions. Circulation
2007;115(17):2344-51.
13. Bonita R. Epidemiology of stroke. Lancet 1992;339(8789):342-4.
14. Iakovou I, Schmidt T, Bonizzoni E, et al. Incidence, predictors, and
outcome of thrombosis after successful implantation of drug-eluting
stents. JAMA 2005;293(17):2126-30.
15. Mauri L, Hsieh WH, Massaro JM, et al. Stent thrombosis in randomized
clinical trials of drug-eluting stents. N Engl J Med 2007;356(10):1020-9.
16. Joner M, Finn AV, Farb A, et al. Pathology of drug-eluting stents in
humans: delayed healing and late thrombotic risk. J Am Coll Cardiol
2006;48(1):193-202.
17. Tada T, Byrne RA, Simunovic I, et al. Risk of stent thrombosis among
bare-metal stents, first-generation drug-eluting stents, and second-
generation drug-eluting stents: results from a registry of 18,334
patients. JACC Cardiovasc Interv 2013;6(12):1267-74.
18. Brambilla M, Camera M, Colnago D, et al. Tissue factor in patients
with acute coronary syndromes: expression in platelets, leukocytes,
and platelet-leukocyte aggregates. Arterioscler Thromb Vasc Biol
2008;28(5):947-53.
19. Kolandaivelu K, Swaminathan R, Gibson WJ, et al. Stent thrombo-
genicity early in high-risk interventional settings is driven by stent
design and deployment and protected by polymer-drug coatings.
Circulation 2011;123(13):1400-9.
20. Lombardo A, Biasucci LM, Lanza GA, et al. Inflammation as a
possible link between coronary and carotid plaque instability.
Circulation 2004;109(25):3158-63.
21. Korn-Lubetzki I, Farkash R, Pachino RM, et al. Incidence and risk
factors of cerebrovascular events following cardiac catheterization.
J Am Heart Assoc 2013;2:e000413. http://dx.doi.org/10.1161/
JAHA.113.000413.
22. Budaj A, Flasinska K, Gore JM, et al. Magnitude of and risk factors for
in-hospital and postdischarge stroke in patients with acute coronary
syndromes: findings from a Global Registry of Acute Coronary Events.
Circulation 2005;111(24):3242-7.
23. Cronin L, Mehta SR, Zhao F, et al. Stroke in relation to cardiac
procedures in patients with non–ST-elevation acute coronary syndrome:
a study involving N18 000 patients. Circulation 2001;104(3):269-74.
American Heart Journal
Volume 168, Number 6
Guerra et al 868.e1

Appendix Supplementary Table II. Angiographic data in patients with

stroke and TIA
Stroke (69 TIA (21
Supplementary Table I. Baseline data in patients with stroke patients/125 patients/36
and TIA Variable lesions) lesions) P

Stroke TIA
No. of narrowed .03
(n = 69 (n = 21
coronary arteries⁎
Variable patients) patients) P
1 11 (15.9) 4 (19.0)
2 17 (24.6) 11 (52.4)
Age 70.6 (65.2-77.0) 71.5 (68.4-78.0) .69 3 41 (59.5) 6 (28.6)
Women 28 (40.1) 8 (38.1) .84 Multivessel disease⁎ 58 (84.1) 17 (80.9) .74
BMI (kg/m 2) 25.0 (23.2-27.4) 26.6 (24.5-29.2) .08 Vessel affected .42
Diabetes 29 (42.0) 8 (38.1) .75 Bypass graft 2 (1.6) 0 (0.0)
On insulin therapy 12 (17.4) 2 (0.49) .38 Left descending coronary 53 (42.4) 20 (55.6)
Arterial hypertension 48 (66.7) 12 (57.1) .42 artery
Current smoking 16 (23.2) 5 (23.8) .95 Left main coronary artery 7 (5.6) 2 (5.6)
Hypercholesterolemia 33 (47.8) 13 (61.9) .28 Left circumflex coronary 39 (31.2) 6 (16.7)
Left ventricular 44.8 (36.0-55.0) 48.2 (39.5-59.5) .38 Left internal mammary artery 0 (0) 0 (0)
ejection fraction (%) Right coronary artery 24 (19.2) 8 (22.1)
Atrial fibrillation 10 (14.5) 8 (38.1) .02 Restenotic lesion 6 (4.8) 1 (2.8) .60
Prior cerebrovascular events 4 (5.8) 0 (0.0) .26 Chronic total occlusion 6 (4.8) 1 (2.8) .95
Prior myocardial infarction 17 (24.6) 5 (23.8) .94 Complex lesions 108 (86.4) 29 (80.6) .39
Prior CABG surgery 11 (15.9) 1 (4.8) .19 Lesion length (mm) 13.6 (7.7-18.0) 14.1 (9.1-18.4) .45
Clinical presentation .06 Reference diameter (mm) 2.8 (2.5-3.1) 2.7 (2.4-3.1) .19
Stable angina 20 (29.0) 10 (47.6) Baseline percentage stenosis (%) 70.1 (56.3-83.6) 69.2 (56.5-78.9) .80
NSTE-ACS 16 (23.2) 7 (33.3) Stent type .1
STEMI 33 (47.8) 4 (19.1) Bare-metal stent 33 (26.4) 16 (44.4)
Data are expressed as median and interquartile range (25th-75th percentiles) or counts First-generation DES 39 (31.2) 7 (19.4)
(percentages). Abbreviations: BMI, body mass index; CABG, coronary artery bypass Second-generation DES 53 (42.4) 13 (36.2)
graft; NSTE-ACS, non–ST-segment elevation ACS; STEMI, ST-segment elevation Total stented length (mm) 24.3 (18.0-28.0) 21.0 (16.0-24.0) .09
myocardial infarction. Postintervention MLD (mm) 2.6 (2.3-2.9) 2.6 (2.2-2.8) .41
Residual percentage stenosis (%) 10.9 (6.2-14.5) 9.9 (5.6-13.9) .57

Data are expressed as median and interquartile range (25th-75th percentiles), or counts
(percentages). Abbreviations: DES, drug-eluting stent; MLD, minimal lumen diameter.
⁎ Number of narrowed coronary arteries and multivessel disease refer to the number
of patients; the rest of analysis is lesion based.

Supplementary Table III. Inhospital, 30-day, and 3-year

mortality in patients with stroke and TIA
Variable Stroke (n = 69) TIA (n = 21) P

Inhospital mortality 7 (10.1) 0 (0.0) .29

30-d mortality 13 (18.8) 1 (4.8) .21
3-y mortality 28 (40.6) 4 (19.1) .12

Data are expressed as number of events (percentages).

Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.