2016, Wiley Periodicals, Inc.

DOI: 10.1111/joic.12327

PERCUTANEOUS CORONARY INTERVENTION

Comparison of Clinical Results Following the Use of Drug-Eluting
Balloons for a Bare-Metal Stent and Drug-Eluting Stent Instent Restenosis
WEI-CHIEH LEE, M.D., YEN-NAN FANG, M.D., CHIH-YUAN FANG, M.D.,
CHIEN-JEN CHEN, M.D., CHENG-HSU YANG, M.D., HON-KAN YIP, M.D.,
CHI-LING HANG, M.D., F.A.C.C., CHIUNG-JEN WU, M.D., and HSIU-YU FANG, M.D.
From the Division of Cardiology, Department of Internal Medicine, Kaohsiung Chang Gung Memorial Hospital, Chang Gung University
College of Medicine, Kaohsiung, Taiwan, Republic of China

Background: Drug-eluting balloons (DEBs) have emerged as a potential alternative to current treatments of
instent restenosis (ISR). The study aims to investigate the clinical outcomes of a DEB angioplasty to treat baremetal stent (BMS) ISR and drug-eluting stent (DES) ISR at 1-year clinical follow-up period.
Methods: Between November 2011 and December 2014, 312 patients were diagnosed with coronary artery ISR at
our hospital. A total of 426 coronary ISR lesions were treated with DEBs. The clinical outcomes, including target
lesion revascularization (TLR), myocardial infarction, stroke, cardiovascular mortality, and all-cause mortality
were compared between the BMS-ISR group and DES-ISR group. Propensity score matched analysis was used to
minimize bias.
Results: The average age of the patients was 64.99
10.35 years, and 76.9% of the patients were male. After
multivariate Cox regression analyses about 1-year recurrent restenosis in DES-ISR group, only end stage renal
disease (ESRD) (P 0.047) and previous DEB failure (P < 0.001) were identied with signicant difference. After
propensity score matched analysis, the bias of baseline characteristics showed no signicant difference. The DESISR group experienced more myocardial infarctions (2.8% vs 8.3%, P 0.075), more TLR (8.1% vs 15.4%,
P 0.051), especially at nonostial lesion (5.7% vs 14.9%, P 0.030) than the BMS-ISR group. Higher incidence of
major cardiac cerebral adverse events happened in the DES-ISR group. (11.7% vs 22.1 %, P 0.038)
Conclusion: During the 1-year follow-up period, DEBs angioplasty for BMS-ISR had better clinical outcomes and
less TLR than DES-ISR. ESRD and previous DEB failure were associated to TLR in DES-ISR group. (J Interven
Cardiol 2016;29:469479)

Introduction
Instent restenosis (ISR) due to neointimal proliferation has been an important issue following percutaneous coronary interventions (PCIs). Drug-eluting
balloons (DEBs) have emerged as a potential alternative to current treatments of ISR because stents cannot
Yen-Nan Fang contributed equally as first author.
Wei-Chieh Lee and Hsiu-Yu Fang share equal contribution as
correspondence author.
Address for reprints: Dr. Hsiu-Yu Fang, Division of Cardiology,
Department of Internal Medicine Kaohsiung Chang Gung Memorial Hospital 123, Ta Pei Road, Niao Sung District Kaohsiung City,
83301, Taiwan, Republic of China. Fax: 886 7 7322402; e-mail:
ast42aiu@hotmail.com

Vol. 29, No. 5, 2016

be implanted at all sites.1 Compared to drug-eluting

stents (DES), DEBs offer advantages that include
immediate and homogeneous drug release in the
arterial wall and the absence of polymers that can
induce chronic inammatory reactions. To treat ISR
with bare-metal stents (BMSs)2 and treat ISR with
DESs,3,4 several clinical randomized studies showed
that DEBs were superior to uncoated balloon angioplasty. To the best of our knowledge, few studies have
discussed the clinical results between the BMS-ISR
and DES-ISR group after DEB use.
The aim of this study was to compare the differences
in clinical outcomes between the BMS-ISR and DESISR group following DEB use. Because the study
patients were not randomly assigned and had some bias

Journal of Interventional Cardiology

469

LEE, ET AL.

of baseline characteristics, propensity score matched

analysis was used to minimize bias.

Methods
Patient Collection and Groups. Between
November 2011 and December 2014, 312 patients
were diagnosed with coronary artery ISR and 426
coronary ISR lesions were treated with DEB angioplasty at our hospital. Demographics characteristics,
risk factors, lesion sites, lesion type, characteristics of
coronary artery disease, previous stents, and intravascular ultrasound (IVUS) use were compared between
the BMS-ISR group and the DES-ISR group. According to chart review, the occurrence of stroke,
myocardial infarction, target lesion revascularization
(TLR), cardiovascular death, and all-cause death were
collected. Over a 1-year follow-up period, angiographic restenosis happened after DEB use for BMSISR was dened BMS-ISR group. On the other hand,
angiographic restenosis happened after DEB use for
DES-ISR was dened DES-ISR group.
Ethics Statement. The Institutional Review
Committee on Human Research of Chang Gung
Memorial Hospital approved the study protocol
(104-7726B). All patient records/information were
anonymized and de-identied prior to analyses.
Procedure and Protocol. Interventions were
performed from radial or femoral route as our usual
practice. All patients were pretreated with aspirin and
clopidogrel. Unfractionated heparin (initial intravenous bolus 5000IU) was used during the procedure
(with additional intravenous boluses as required for
procedures), targeting an activated clotting time
>250 seconds. The predilation of the ISR lesion using
high-pressure balloons was to achieve adequate
expansion of prior stent and was documented by
subsequent angiogram or intravascular ultrasound. The
SeQuent Please (B Braun Melsungen AG, Melsungen,
Germany) was the only DEB used in our hospital. The
balloon has a shelf life of >1 year. More than 80% of
the drug is retained during balloon delivery to the
target lesion, and 1015% of the initial dose is
delivered to the vessel wall upon 60-seconds ination.
DEB was inated at the ISR site for 3060 seconds
according the patients ability to tolerate this treatment.
During the 1-year follow-up period following DEB
treatment, we used a noninvasive examination half
year later, such as a treadmill test or a thallium scan, for

470

asymptomatic patients to detect possible restenosis and

used coronary angiography thereafter if the patient had
typical angina symptoms. In addition, follow-up
coronary angiography need to be performed if the
patient experience myocardial infarction.
Denitions. ISR can be dened by the presence of
>50% diameter stenosis in the stented segment
angiographically. Cardiovascular mortality was death
related to myocardial infarction, stroke, heart failure,
and cardiac arrhythmia. All-cause mortality was
related to death for any reasons. The denition of
major adverse cerebral cardiac events (MACCEs)
included myocardial infarction, TLR, stroke, and
cardiovascular mortality.
Primary and Secondary End-Points. The primary end-point of this study was TLR. The secondary
end-points of this study were myocardial infarction
(ST elevation myocardial infarction and nonST
elevation myocardial infarction), stroke, cardiovascular mortality, and all-cause mortality.
Angiographic Analysis. Coronary angiograms
were analyzed in a core laboratory by trained personnel
using standard methodology. Lesion characteristics
were assessed after the administration of intracoronary
nitroglycerin if possible. Three orthogonal angiographic views were selected to display the target
segment, avoiding foreshortening and overlap. All
lesion severities were measured by quantitative coronary angiography (QCA) before and after the interventional procedure.
Statistical Analysis. All of the statistical analyses
were performed using SPSS 22.0 software (SPSS, Inc.,
Chicago, IL). The data are expressed as percentages
and means
standard deviations. The categorical
variables were compared using chi-squared tests.
The continuous variables were compared using
t-test. Differences in the continuous variables between
the 2 groups were analyzed using 2-way analyses of
variance. A KaplanMeier survival curve was performed for the outcomes of TLR at 1-year follow-up
duration between the BMS-ISR group and the different
generation DES-ISR groups and the different segment
ISR groups. Univariate and multivariate Cox regression analyses were performed to identify the associations of recurrent restenosis after DEB use. Each
correlation between the variables is expressed as a
hazard ratio with 95% condence interval (CI).
Because the patients were not randomly assigned
and had some bias in baseline characteristics. A
propensity score matched analysis was performed by

Journal of Interventional Cardiology

Vol. 29, No. 5, 2016

CLINICAL RESULTS FOLLOWING DEB USE FOR BMS AND DES ISR

using a logistic regression model for the BMS-ISR

group versus the DES-ISR group to adjust for
differences in baseline characteristics. This analysis
included a number of clinical, angiographic, and
procedural variables and are listed in Table 1. We
performed a 1-to-1 matched analysis without replacement on the basis of the estimated propensity score of
each patient. Using the estimated logits, BMS-ISR
group and DES-ISR group had the closest estimated
logit value. After propensity score matched analysis,
we compared the baseline covariates between the 2
groups. P values below 0.05 were considered to be
statistically signicant.

Results
Baseline Characteristics of the Study Patients
Before Propensity Score Matched Analysis
(Table 1). The average age of the BMS-ISR group
was 65.99
10.52 years, and 78.0% of the patients
were male. The average age of the DES-ISR group was
63.89
10.35 years, and 75.7% of the patients were
male. The demographic characteristics, such as the
average age and gender, were similar between the 2
groups. There were no signicant differences in the
clinical conditions. When considering comorbidities,
the DES-ISR group had a higher prevalence of diabetes
mellitus (48.2% vs 62.2%, P 0.017) and ESRD
(17.1% vs 27.7%, P 0.029), and the BMS-ISR group
had a higher prevalence of current smokers (48.2% vs
33.8%, P 0.011). Serum creatinine except ESRD was
similar between the 2 groups. The BMS-ISR group had
more lesions in the right coronary artery (RCA) (43.7%
vs 36.8%), and the DES-ISR group had more lesions in
the left circumex artery (LCX) (16.7% vs 26.9%). The
DES-ISR group had a high prevalence of multiple
vessel disease (90.2% vs 94.6%, P 0.109) and
more ostial lesions (16.5% vs 33.7%, P < 0.001).
More type C lesions were in BMS-ISR group, and more
type B2 lesions in DES-ISR group (P 0.010). More
total occlusion and subtotal occlusion lesions were in
BMS-ISR group (18.3% vs 15.8%; P 0.501), and
more focal lesions were in DES-ISR group (12.5% vs
28.2%; P < 0.001). BMS-ISR group had more lesion
numbers of target vessel (1.51
0.74 vs 1.47
0.78;
P 0.554) and more prevalence of multiple lesions of
target vessel (37.5% vs 33.2%) than DES-ISR group.
PrePCI angiographic vascular stenosis, minimal
luminal diameter (MLD) and reference luminal

Vol. 29, No. 5, 2016

diameter (RLD) were similar between the 2 groups.

PostPCI angiographic vascular stenosis was higher in
the DES-ISR group (14.94
8.47% vs 18.39
9.54%;
P < 0.001), and postPCI MLD and -PCI RLD were
similar. The diameter of the used DEBs did not
statistically differ, and the length of used DEB was
longer in the BMS-ISR group (27.9
3.58 mm vs
26.6
4.39 mm; P 0.001). There were no signicant
differences in the percentages of the same lesions that
were previously treated by DEB between 2 groups.
The use of periprocedure IVUS was similar between 2
groups. Low rates of complicated procedures were
found in both groups, and there was no signicant
difference. Only 1 patient in the DES-ISR group
experienced a balloon-jailed side branch vessel.
The duration of follow-up was longer in the DESISR group, but there was no signicant difference. In
our study, a total 68% patients received coronary
angiography or noninvasive examination to detect
current restenosis. Follow-up examination was performed for 64.0% patients in the BMS-ISR group, and
72.4% patients in the DES-ISR group (P 0.202).
Treadmill test was arranged for 11.0% patients in the
BMS-ISR group, and 14.9% patients in the DES-ISR
group (P 0.305). Thallium perfusion scan was
arranged for 21.3% patients in the BMS-ISR group,
and 17.6% patients in the DES-ISR group (P 0.401).
Coronary angiography was performed for 31.7%
patients in the BMS-ISR group, and 39.9% patients
in the DES-ISR group (P 0.226).
The KaplanMeier Curves of TLR Between
BMS-ISR and DES-ISR Groups and Between
Ostial Segment and Nonostial Segment: (Figs. 1
and 2). The KaplanMeier curve of TLR differed
between BMS-ISR and DES-ISR groups (Fig. 1) and
P-value of log rank test was 0.042. The KaplanMeier
curve of TLR of ostial segment showed no difference
(P 0.939) (Fig. 2). The KaplanMeier curve of TLR
of nonostial segment had signicant difference
(P 0.023) (Fig. 2).
The Recurrent TLR Rate in BMS-ISR and
Different Generation DES-ISR (Table 2) and the
KaplanMeier Curves of TLR Between BMS-ISR
and Different Generation DES-ISR Groups:
(Fig. 3). The TLR rate was 10.6% BMS-ISR group,
and 16.7% in rst generation DES-ISR group, and
19.8% in second generation DES-ISR group. P value
was 0.078 between 3 groups. The KaplanMeier curve
of TLR also showed signicant difference between
BMS-ISR group and second generation DES-ISR

Journal of Interventional Cardiology

471

LEE, ET AL.
Table 1. Baseline Characteristics of Study Patients Before Propensity Score Match

Patient number
Lesion number
General demographics
Age (year)
Male gender (%)
Clinical condition
STEMI (%)
NSTEMI (%)
Unstable angina (%)
Stable angina (%)
Risk factors
Hypertension (%)
Diabetes (%)
Current smoker (%)
Old myocardial infarction (%)
Old stroke (%)
PAOD (%)
Hyperlipidemia (%)
Heart failure (%)
Prior CABG (%)
ESRD on maintenance hemodialysis (%)
Laboratory examination
Creatinine (mg/dL) (exclude ESRD)
Lesion-related artery (%)
Left anterior descending artery
Left circumflex artery
Right coronary artery
Characteristics of coronary artery disease
Single- or multiple-vessel disease (%)
Single vessel disease
Double vessel disease
Triple vessel disease
Left main disease (%)
Ostial lesion (%)
Lesion type
A
B1
B2
C
Total and subtotal occlusion
Focal lesion (<10 mm)
Lesion numbers of target vessel
Average
Multiple (%)
PrePCI angiography
PrePCI stenosis (%)

472

BMS-ISR

DES-ISR

P-value

164
224

148
202

65.99
10.52
78.0

63.89
10.35
75.7

3.0
17.7
58.5
20.7

0.7
18.2
57.4
23.6

78.0
48.2
48.2
40.9
6.1
6.7
59.8
24.4
7.9
17.1

79.1
62.2
33.8
41.2
6.8
8.8
63.5
26.4
6.8
27.7

0.891
0.017
0.011
0.948
0.822
0.529
0.560
0.698
0.829
0.029

1.32
1.10

1.39
1.23

0.650
0.036

39.6
16.7
43.7

36.3
26.9
36.8

9.8
27.4
62.8
23.2
16.5

5.4
15.5
79.1
33.1
33.7

6.7
10.3
15.2
67.9
18.3
12.5

6.9
7.4
28.2
57.4
15.8
28.2

1.51
0.74
37.5

1.47
0.78
33.2

0.554
0.363

78.34
13.06

79.62
12.77

0.307

0.073
0.687
0.456

0.007

Journal of Interventional Cardiology

0.058
<0.001
0.010

0.501
<0.001

Vol. 29, No. 5, 2016

CLINICAL RESULTS FOLLOWING DEB USE FOR BMS AND DES ISR
TABLE 1. Continued

PrePCI MLD (mm)

PrePCI RLD (mm)
PostPCI angiography
PostPCI stenosis (%)
PostPCI MLD (mm)
PostPCI RLD (mm)
DEB
Diameter (mm)
Length (mm)
Previous DEB use
IVUS use (%)
Complication of PCI (%)
F/U time (days)
F/U examination (%)
Treadmill test (%)
Thallium perfusion scan (%)
Coronary angiography (%)

BMS-ISR

DES-ISR

P-value

1.13
0.83
2.79
0.59

0.59
0.39
2.89
0.57

0.353
0.074

14.94
8.47
2.39
0.52
2.88
0.58

18.39
9.54
2.40
0.51
2.98
0.59

<0.001
0.728
0.069

3.10
0.40
27.9
3.58
2.2
33.5
0
509.25
285.44
64.0
11.0
21.3
31.7

3.13
0.42
26.6
4.39
4.0
37.6
0.5
569.15
296.73
72.4
14.9
17.6
39.9

0.474
0.001
0.226
0.214
0.474
0.071
0.202
0.305
0.401
0.226

DES, drug-eluting stent; ISR, instent restenosis; BMS, bare-metal stent; STEMI, ST segment elevation myocardial infarction; NSTEMI, nonST
segment elevation myocardial infarction; PAOD, peripheral arterial occlusive disease; CABG, coronary artery bypass grafting; ESRD, end stage
renal disease; MLD, minimal luminal diameter; RLD, reference luminal diameter; IVUS, intravascular ultrasound; PCI, percutaneous
intervention; F/U, follow-up. Data are expressed as mean
SD or percentage.

group (P 0.030) (Fig. 3). There was no signicant

difference between BMS-ISR group and rst generation DES-ISR group (P 0.408) and between rst
generation DES-ISR group and second generation
DES-ISR group (P 0.730) (Fig. 3). Three kinds of

different second generation drug-eluting stents were

used in our institution. However, everolimus-eluting
stents were used for most study patients. Therefore,
sample size was different between everolimus-eluting
stent ISR and zotarolimus-eluting stents ISR. Among

Figure 1. The KaplanMeier curve of TLR differed between the

BMS-ISR group and the DES-ISR group (P 0.042).

Figure 2. The KaplanMeier curve of TLR of ostial segment was no

difference (P 0.939). The KaplanMeier curve of TLR of
nonostial segment had signicant difference (P 0.023).

Vol. 29, No. 5, 2016

Journal of Interventional Cardiology

473

LEE, ET AL.
Table 2. The Current TLR Rate of the BMS-ISR and the Different
Generation DES-ISR After DEB Angioplasty

BMS

Patient number
164
Patient lesion (N)
224
Restenosis rate (%) 10.6

First
generation
DES

Second
generation
DES

P-value

24
30
16.7

124
172
19.8

0.078

TLR, target lesion revascularization; BMS, bare-metal stent; DES,

drug-eluting stent; ISR, instent restenosis; DEB, drug-eluting
balloon. Data are expressed as number or percentage.

the patients with DES-ISR, 111 (50%) patients

received everolimus-eluting stents, and 22 (10.9%)
patients received zotarolimus-eluting stents, and other
patients received other second generation DESs or rst
generation DESs. The rate of ISR was 18.0% in the
patients with everolimus-eluting stents, and 9.1 % in
the patients with zotarolimus-eluting stents, and
P-value was 0.303.
The Associations of TLR in the DES-ISR Group:
(Table 3). After univariate and multivariate Cox
regression analyses about 1-year TLR in DES-ISR
group, only ESRD on maintenance hemodialysis
(hazard ratio [HR]: 2.339; 95%CI: 1.0135.404,
P 0.047) and previous DEB failure (HR: 17.090;
95%CI: 4.43065.297, P < 0.001) were identied
signicant difference. Diabetes, current smoker, ostial
lesion, left main bifurcation lesion, types B2 and C

Figure 3. The KaplanMeier curve of TLR also showed no

signicant difference between BMS-ISR group and rst generation
DES-ISR group (P 0.408), and signicant difference between
BMS-ISR group and second generation DES-ISR group (P 0.030),
and no signicant difference between rst generation DES-ISR
group and second generation DES-ISR group (P 0.730).

474

lesion, multiple lesions of the target vessel, rst

generation drug eluting stent, postPCI stenotic
severity, and IVUS use did not have signicant
difference.
Baseline Characteristics of the Patients After
Propensity Score Matched Analysis: (Table 4).
After propensity score matched analysis was performed for the entire population, there were 230
matched pairs of patients. Total 315 lesions were
compared between 2 groups. General demographics,
clinical conditions, risk factors, lab examination,
lesion-related artery, characteristics of coronary artery
disease, lesion type, lesion numbers of target vessel,
pre- and postangiographic parameter became similar
between 2 groups. The prevalence of diabetes mellitus
(50.4% vs 57.4%; P 0.290), heart failure (20.9% vs
27.0%; P 0.354) and ESRD (15.7% vs 21.7%;
P 0.310) were still higher in DMS-ISR. There was
no signicant difference. The prevalence of smoker
was higher in BMS-ISR group (43.5% vs 36.5%;
P 0.346), but became no signicant difference. Total
occlusion and subtotal occlusion lesions were similar
between 2 groups. More focal lesions were still in
DES-ISR group (15.0% vs 25.8%; P 0.017). Higher
prevalence of left main disease and ostial lesions was
in DES-ISR group, but became no signicant difference. PostPCI stenotic severity (15.07
0.69% vs
17.95
9.80%; P 0.006) and the length of used DEB
(28.01
3.61 mm vs 26.98
4.30 mm; P 0.022) still
had signicant difference.
One-Year Clinical Outcomes of the Propensity
Score Matched Patients: (Table 5). The incidence
of myocardial infarction was 2.8% in the BMS-ISR
group and 8.3% in the DES-ISR group (P 0.075).
The current TLR rate was 8.1 % in the BMS-ISR group
and 15.4% in the DES-ISR group (P 0.051). Ostial
segment ISR was similar in the 2 groups and had near
one fth experienced current restenosis over the 1-year
follow-up period. Nonostial segment of BMS-ISR had
5.7% current restenosis rate and nonostial segment of
DES-ISR had 14.9% current restenosis rate
(P 0.030). The incidence of stroke and cardiovascular mortality did not statistically differ between the 2
groups (2.6% vs 0.9%, P 0.622; 2.7% vs 4.5%,
P 0.721). The incidence of major adverse cardiac
cerebral events (MACCEs) was 11.7% in the BMSISR group and 22.1% in the DES-ISR group
(P 0.038). The incidence of all-cause mortality did
not statistically differ between the 2 groups (7.8 % vs
7.7%, P 1.000).

Journal of Interventional Cardiology

Vol. 29, No. 5, 2016

CLINICAL RESULTS FOLLOWING DEB USE FOR BMS AND DES ISR
Table 3. Multivariate Cox Regression Analyses About 1-Year TLR in DES-ISR Group
Variables

Hazard ratio

95%CI

1.356
1.399
2.339

0.5653.250
0.6453.032
1.0135.404

1.757
0.497
1.196
0.650
0.842
0.970
17.090
1.162

0.5915.229
0.1441.708
0.3494.106
0.2481.705
0.2602.730
0.9271.014
4.43065.927
0.5022.688

Risk factors
Diabetes
Current smoker
ESRD on maintenance hemodialysis
Lesion site and type
Ostial lesion
Left main bifurcation lesion
Lesion types B2 and C
Multiple lesions of the target vessel
First generation drug eluting stent
The severity of postPCI stenosis
Previous DEB use
IVUS use

P-value

0.495
0.395
0.047
0.311
0.267
0.776
0.381
0.774
0.180
<0.001
0.726

TLR, target lesion revascularization; DES, drug-eluting stent; ISR, instent restenosis; CI, condence interval; ESRD, end stage renal disease;
PCI, percutaneous intervention; DEB, drug-eluting balloon; IVUS, intravascular ultrasound.

Discussion
Although stent implantation has increased the
accuracy of the results and improved long-term
outcomes, ISR still occurs and is an important problem
that affects prognosis. Additional stenting is not
always desirable, and PCI with DEBs has emerged
as an adjunctive strategy for ISR.5 Paclitaxel has been
identied as the primary drug for use in DEBs because
of its rapid uptake and prolonged retention. Paclitaxel
is embedded in hydrophilic iopromide, which increases the solubility and the transfer of paclitaxel to
the vessel wall.6 Compared to DESs, DEBs offer
advantages that include immediate and homogeneous
drug release in the arterial wall and the absence of
polymers that can induce chronic inammatory
reactions. Other advantages include about a shorter
period of combined antiplatelet therapy.
In several randomized clinical trials, DEBs have
been found to be superior to uncoated balloon
angioplasty to treat ISR with BMSs2 and DESs.3,4
Although drug-eluting stents are still considered to be
the best possible care for treating ISR,7 they may
further reduce the exibility of the vessel and limit the
repeatability of the procedure. In addition, stents
cannot be implanted at all sites, especially at the ISR
site. Repeated stent-in-stent therapy could cause
lumen loss and repeat stenosis. Therefore, repeat DEB
angioplasty was reasonable to use for current

Vol. 29, No. 5, 2016

restenosis if additional stenting was not desirable.

One study had proved that the DEB was superior to the
DES in the primary angiographic end point and was
associated with fewer adverse clinical events.8
Recently, a new meta-analysis9,10 had showed
signicant advantage with DEB use in ISR than prior
balloon angioplasty. Siontis et al. also had published
new meta-analysis11 which showed better outcome in
everlimus-eluting stents (EES) than DEB in angiographic and clinical outcome. But another randomized
trial12 had showed that treatment of BMS ISR using
paclitaxel eluting balloon led to signicant less
12-month late lumen loss than the implantation of
second-generation EES. Furthermore, another metaanalysis13 which published by Lee et al. had showed
better outcome in DEB or DES for ISR lesions than
POBA in preventing TLR, but treatment with DEB
showed a trend of less MI than did treatment with DES.
The use of bioresorbable vascular scaffold (BVS)
implantation for the treatment of complex BMS-ISR
and DES-ISR lesions also showed acceptable longterm clinical outcomes.14
Few studies have compared the clinical results of
DEB use for BMS-ISR and DES-ISR. In one
prospective study, DES-ISR was associated with
poorer outcomes compared with BMS-ISR after
treatment with DEB. In DEB-treated lesions, current
restenosis occurred in 1.1% of patients with BMS-ISR
and in 9.1% of patients with DES-ISR.15 In one

Journal of Interventional Cardiology

475

LEE, ET AL.
Table 4. Baseline Characteristics of Study Patients After Propensity Score Match

Patient number
Lesion number
General demographics
Age (year)
Male gender (%)
Clinical condition
STEMI (%)
NSTEMI (%)
Unstable angina (%)
Stable angina (%)
Risk factors
Hypertension (%)
Diabetes (%)
Current smoker (%)
Old myocardial infarction (%)
Old stroke (%)
PAOD (%)
Hyperlipidemia (%)
Heart failure (%)
Prior CABG (%)
ESRD on maintenance hemodialysis (%)
Laboratory examination
Creatinine (mg/dL) (exclude ESRD)
Lesion-related artery (%)
Left anterior descending artery
Left circumflex artery
Right coronary artery
Characteristics of coronary artery disease
Single- or multiple-vessel disease (%)
Single vessel disease
Double vessel disease
Triple vessel disease
Left main disease (%)
Ostial lesion (%)
Lesion type
A
B1
B2
C
Total and subtotal occlusion
Focal lesion (<10 mm)
Lesion numbers of target vessel
Average
Multiple (%)
PrePCI angiography
PrePCI stenosis (%)

476

BMS-ISR

DES-ISR

P-value

115
160

115
155

65.10
10.41
77.4

63.49
10.34
76.5

4.3
15.7
62.6
17.4

0.9
17.4
59.1
22.6

75.7
50.4
43.5
41.7
7.0
6.1
67.0
20.9
7.8
15.7

75.7
57.4
36.5
39.1
7.0
9.6
66.1
27.0
5.2
21.7

1.000
0.290
0.346
0.788
1.000
0.462
0.889
0.354
0.595
0.310

1.20
0.50

1.31
0.94

0.354
0.092

38.6
16.5
44.9

37.7
26.0
36.4

8.7
20.0
71.3
12.5
18.1

7.0
17.4
75.7
14.2
20.0

5.6
10.0
12.5
71.9
16.9
15.0

8.4
9.0
19.4
63.2
18.1
25.8

0.781
0.017

1.50
0.69
38.8

1.46
0.73
34.8

0.599
0.486

77.91
12.93

80.48
12.04

0.069

0.241
0.876
0.300

0.748

Journal of Interventional Cardiology

0.741
0.774
0.247

Vol. 29, No. 5, 2016

CLINICAL RESULTS FOLLOWING DEB USE FOR BMS AND DES ISR
TABLE 4. Continued

PrePCI MLD (mm)

PrePCI RLD (mm)
PostPCI angiography
PostPCI stenosis (%)
PostPCI MLD (mm)
PostPCI RLD (mm)
DEB
Diameter (mm)
Length (mm)
Previous DEB use
IVUS use (%)

BMS-ISR

DES-ISR

P-value

0.36
0.28
2.76
0.59

0.55
0.35
2.81
0.54

0.303
0.490

15.07
8.70
2.39
0.53
2.88
0.57

17.95
9.80
2.33
0.49
2.87
0.57

0.006
0.275
0.938

3.13
0.41
28.01
3.61
2.5
33.1

3.09
0.42
26.98
4.30
3.9
38.1

0.396
0.022
0.537
0.410

DES, drug-eluting stent; ISR, instent restenosis; BMS, bare-metal stent; STEMI, ST segment elevation myocardial infarction; NSTEMI, nonST
segment elevation myocardial infarction; PAOD, peripheral arterial occlusive disease; CABG, coronary artery bypass grafting; ESRD, end stage
renal disease; MLD, minimal luminal diameter; RLD, reference luminal diameter; IVUS, intravascular ultrasound; PCI, percutaneous
intervention; F/U, follow-up. Data are expressed as mean
SD or percentage.

retrospective study, recurrent restenosis occurred in

13.0% of the BMS-ISR group and in 21.1% of the
DES-ISR group following DEB use a half year later.16
In 2 studies, late restenosis and delayed late lumen loss
were signicantly higher in the DES-ISR group after
DEB angioplasty.15,16 Neointima progression was
observed in both the BMS-ISR and DES-ISR groups,
but the extent of neointima progression was larger in
the DES-ISR group. The mechanisms of DES-ISR
include several drug-specic factors such as localized
hypersensitivity, nonuniform drug deposition,

polymers, and drug resistance.15 Therefore, DEB

angioplasty for DES-ISR could not provide excellent
outcomes. The observation of previous studies that
DES-ISR is more frequently focal, due to the potent
anti-proliferative drug effects and other mechanical
factors, such as focal stent under-expansion, stent
fracture, loss of longitudinal integrity, and/or incomplete stent apposition.17 In our study, we also found
higher incidence of focal lesions in DES-ISR, and
higher incidence of total or subtotal lesion in BMSISR. Our results are consistent with the results of those

Table 5. One-Year Clinical Outcomes of Study Patients After Propensity Score Match
BMS-ISR

Patient number
Lesion number
1-year clinical outcomes
Myocardial infarction (%)
Target-lesion restenosis (%)
Ostial segment
Nonostial segment
Stroke (%)
Cardiovascular mortality (%)
MACCE (%)
All-cause mortality (%)

115
160
2.8
8.1
18.5
5.7
2.6
2.7
11.7
7.8

DES-ISR

P-value

115
155
8.3
15.4
17.5
14.9
0.9
4.5
22.1
7.7

0.075
0.051
0.901
0.030
0.622
0.721
0.038
1.000

DES, drug-eluting stent; ISR, instent restenosis; BMS, bare-metal stent; MACCE, major adverse cardiovascular cerebral event. Data are
expressed as number or percentage.

Vol. 29, No. 5, 2016

Journal of Interventional Cardiology

477

LEE, ET AL.

studies and the KaplanMeier curve showed

signicant difference between BMS-ISR group and
DES-ISR group (P 0.042). Whether rst generation
DES-ISR group or second generation DES-ISR
group, the results of TLR were worse than BMS-ISR
group.
However, no study discussed the restenosis rate in
different segments of current restenosis following
DEB use. In our study, regardless of the BMS-ISR
group or DES-ISR group, the incidence of current
restenosis was higher at the ostial lesions and was
around one fth. There was no signicant difference
between TLR of ostial segment in BMS-ISR group and
DES-ISR group (P 0.939). In addition, signicant
difference was noted between TLR of nonostial
segment in BMS-ISR group and DES-ISR group
(P 0.023). After propensity score matched analysis,
the same results still were noted. The TLR was similar
in DES-ISR no matter which segment was involved.
Nonostial BMS-ISR still had better results of TLR than
ostial BMS-ISR.
In our study, the DES-ISR group had more comorbidities of diabetes mellitus and ESRD, and the
BMS-ISR group had a higher prevalence of current
smokers. More multiple vessel coronary artery disease
and more ostial lesions were presented in the DES-ISR
group, which could affect our clinical results. This
phenomenon was reasonable because the interventionist should treat the patients with complex coronary
lesions and more comorbidity with DES. More focal
lesions were presented in DES-ISR group, even after
propensity score matched analysis. These problem
affected the length of DEB use. On the other hand,
operators selected longer DEB because the BMS-ISR
does not exist any anti-proliferative drugs previously,
and hoped longer DEB could cover whole BMS-ISR to
prevent further ISR like a DES implantation. In the
DES-ISR group, the lesion may already exist antiproliferative drugs, and operators choose shorter
length of DEB to cover the focal DES-ISR. Even if
the incidence of restenosis rate was low in DES,
current restenosis still became an important problem of
these patients. Therefore, propensity score matched
analysis was used to minimize bias of baseline
characteristics between the 2 groups. After propensity
score matched analysis, the current TLR rate was 8.1%
in the BMS-ISR group and 15.4% in the DES-ISR
group over a 1-year follow-up period. The restenosis
rate was higher in the DES-ISR group in our study
1 year later. Different restenosis rate in different

478

medicated stent and brands were presented following

DEB use, despite the small numbers of lesions.
According to previous studies,1,18 treating BMS-ISR
with a DEB persistently reduced the incidence of major
adverse cardiac events compared with the incidence
after treatment with an uncoated balloon or DES
implantation during long-term follow-up. In our study,
the incidence of myocardial infarction and MACCEs
was signicantly higher in the DES-ISR group. Our
hypothesis was that worse results following DEB use
in the DES-ISR group was related to poorer response
of vascular wall to anti-inammation and antiproliferation medication after medication delivered
to vascular wall repeatedly.
In 1 meta-analysis study, Siontis et al. concluded
that 2 strategies should be considered for treatment of
any type of coronary ISR: PCI with everolimus-eluting
stents because of the best angiographic and clinical
outcomes, and DEB because of its ability to provide
favorable results without adding a new stent layer.11
No previous study explored the associations about
current restenosis in DES-ISR group after DEB use.
Therefore, we did not know when to select DES
instead of DEB for ISR. The associations of TLR in
DES-ISR group were ESRD and previous DEB failure.
In addition, previous DEB failure is a strong predictor
about current restenosis following DEB use. Therefore, we may consider DESs use to prevent current
restenosis if the patient was ESRD and the lesions had
been treated by DEB.
In our study, we analyzed clinical results and
provided our extensive experience regarding DEB
angioplasty for BMS-ISR and DES-ISR. Clinicians
could understand clinical outcomes of different
generation DES ISR, and different segment of
coronary artery after DEB use, and the associations
of failure of DEB angioplasty for DES-ISR.

Study Limitations
First, our study was a retrospective single-center
study with an observational analysis. The clinical
results of this study are valuable because we included
all consecutive patients undergoing DEB angioplasty
for ISR lesions. Second, not all patients received a
follow-up angiography if the patients did not have
clinical symptoms or negative stress test. In our study,
a total 68% patients received coronary angiography or
noninvasive examination to detect current restenosis.

Journal of Interventional Cardiology

Vol. 29, No. 5, 2016

CLINICAL RESULTS FOLLOWING DEB USE FOR BMS AND DES ISR

Third, not all patients received an image analysis, such

as IVUS or optical coherence tomography. Four, even
the propensity score matched analysis was performed
to minimize bias, this study still had some selection
bias.

Conclusions
During the 1-year follow-up period, DEBs angioplasty for BMS-ISR had better clinical outcomes and
less TLR than DES-ISR. ESRD and previous DEB
failure were associated to TLR in DES-ISR group.

References
1. Scheller B, Clever YP, Kelsch B, et al. Long-term follow-up
after treatment of coronary in-stent restenosis with a paclitaxelcoated balloon catheter. JACC Cardiovasc Interv 2012;5:
323330.
2. Scheller B, Hehrlein C, Bocksch W, et al. Treatment of coronary
in-stent restenosis with a paclitaxel-coated balloon catheter.
N Engl J Med 2006;355:21132124.
3. Habara S, Mitsudo K, Kadota K, et al. Effectiveness of
paclitaxel-eluting balloon catheter in patients with Sirolimuseluting stent restenosis. J Am Coll Cardiol Intv 2011;4:149154.
4. Rittger H, Brachmann J, Sinha AM, et al. PEPCAD-DES: A
randomized, multicenter, single blinded trial comparing
paclitaxel coated balloon angioplasty with plain balloon
angioplasty in drug-eluting-stent restenosis. J Am Coll Cardiol
2012;59:13771382.
5. Habara S, Iwabuchi M, Inoue N, et al. A multicenter randomized
comparison of paclitaxel-coated balloon catheter with conventional balloon angioplasty in patients with bare-metal stent
restenosis and drug-eluting stent restenosis. Am Heart J 2013;
166:527533.e2.
6. Waksman R, Pakala R Drug-eluting balloon. the comeback kid?
Circ Cardiovasc Interv 2009;2:352358.
7. Iona E, Haager PK, Radke PW, et al. Sirolimus- and paclitaxeleluting stents in comparison with balloon angioplasty for
treatment of in-stent restenosis. Catheter Cardiovasc Interv
2005;64:2834.

Vol. 29, No. 5, 2016

8. Unverdorben M, Vallbracht C, Cremers B, et al. Paclitaxelcoated balloon catheter versus paclitaxel-coated stent for the
treatment of coronary in-stent restenosis. Circulation 2009;119:
29862994.
9. Giacoppo D, Gargiulo G, Aruta P, et al. Treatment strategies for
coronary in-stent restenosis: Systematic review and hierarchical
Bayesian network meta-analysis of 24 randomised trials and
4880 patients. BMJ 2015;351:h5392.
10. Sethi A, Malhotra G, Singh S, Singh PP, Khosla S. Efcacy of
various percutaneous interventions for in-stent restenosis:
Comprehensive network meta-analysis of randomized controlled trials. Circ Cardiovasc Interv 2015;8:e002778.
11. Siontis GC, Stefanini GG, Mavridis D, et al. Percutaneous
coronary interventional strategies for treatment of in-stent
restenosis: A network meta-analysis. Lancet 2015;15(386):
655664.
12. Pleva L, Kukla P, Kusnierova P, et al. Comparison of the
efcacy of paclitaxel-Eluting balloon catheters and everolimusEluting stents in the treatment of Coronary in-Stent restenosis:
The treatment of In-Stent restenosis study. Circ Cardiovasc
Interv 2016;9:e003316.
13. Lee JM, Park J, Kang J, et al. Comparison among drug-eluting
balloon, drug-eluting stent, and plain balloon angioplasty for the
treatment of in-stent restenosis: A network meta-analysis of 11
randomized, controlled trials. JACC Cardiovasc Interv 2015;
8:382394.
14. Moscarella E, Ielasi A, Granata F, et al. Long-Term clinical
outcomes after bioresorbable vascular scaffold implantation for
the treatment of Coronary In-Stent restenosis: A multicenter
italian experience. Circ Cardiovasc Interv 2016;9:e003148.
15. Habara S, Iwabuchi M, Inoue N, et al. A multicenter randomized
comparison of paclitaxel-coated balloon catheter with conventional balloon angioplasty in patients with bare-metal stent
restenosis and drug-eluting stent restenosis. Am Heart J 2013;
166:527533.
16. Habara S, Kadota K, Shimada T, et al. Late restenosis after
paclitaxel-Coated balloon angioplasty occurs in patients with
drug-Eluting stent restenosis. J Am Coll Cardiol 2015;66:
1422.
17. Kang SJ, Mintz GS, Park DW, et al. Mechanisms of in-stent
restenosis after drug-eluting stent implantation: Intravascular
ultrasound analysis. Circ Cardiovasc Interv 2011;4(1):914.
18. Unverdorben M, Vallbracht C, Cremers B, et al. Paclitaxelcoated balloon catheter versus paclitaxel-coated stent for the
treatment of coronary in-stent restenosis: The three-year results
of the PEPCAD II ISR study. EuroIntervention 2015;11(8):
926934.

Journal of Interventional Cardiology

479