One-year outcomes after successful chronic total occlusion percutaneous coronary intervention

Abstract

We aimed to determine clinical outcomes 1 year after successful chronic total occlusion (CTO) PCI and, in particular, whether use of dissection and re-entry strategies affects clinical outcomes. Hybrid approaches have increased the procedural success of CTO percutaneous coronary intervention (PCI) but longer-term outcomes are unknown, particularly in relation to dissection and re-entry techniques. Data were collected for consecutive CTO PCIs performed by hybrid-trained operators from 7 United Kingdom (UK) centres between 2012 and 2014. The primary endpoint (death, myocardial infarction, unplanned target vessel revascularization) was measured at 12 months along with angina status. One-year follow up data were available for 96% of successful cases (n = 805). In total, 85% of patients had a CCS angina class of 2-4 prior to CTO PCI. Final successful procedural strategy was antegrade wire escalation 48%; antegrade dissection and re-entry (ADR) 21%; retrograde wire escalation 5%; retrograde dissection and re-entry (RDR) 26%. Overall, 47% of CTOs were recanalized using dissection and re-entry strategies. During a mean follow up of 11.5 ± 3.8 months, the primary endpoint occurred in 8.6% (n = 69) of patients (10.3% (n = 39/375) in DART group and 7.0% (n = 30/430) in wire-based cases). The majority of patients (88%) had no or minimal angina (CCS class 0 or 1). ADR and RDR were used more frequently in more complex cases with greater disease burden, however, the only independent predictor of the primary endpoint was lesion length. CTO PCI in complex lesions using the hybrid approach is safe, effective and has a low one-year adverse event rate. The method used to recanalize arteries was not associated with adverse outcomes. © 2017 Wiley Periodicals, Inc.

Full text

Original Studies
One-Year Outcomes After Successful Chronic Total
Occlusion Percutaneous Coronary Intervention: The
Impact of Dissection Re-Entry Techniques
W. M. Wilson,
1
MBBS,FRACP , S.J. Walsh,
2
MD,FRCP, A. Bagnall,
3,4
MBCHB,PhD,FRCP,
A.T. Yan,
5
MD,FRCPC, C.G. Hanratty,
2
MD,FRCPI, M. Egred,
3,4
BSC (HONS),MBCHB,MD,FRCP,
E. Smith,
6
BSC,MBBS,MD, K.G. Oldroyd,
7
MBCHB,MD (HONS),FRCP, M. McEntegart,
7
MD,PhD,
J. Irving,
8
MBCHB,MD,FRCPEDIN, H. Douglas,
2
MB,BCH, J. Strange,
9
MBCHB,MRCP,MD,and
J.C. Spratt,
10
*BSC,MBCHB,MD,FRCP
We aimed to determine clinical outcomes 1 year after successful chronic total occlusion
(CTO) PCI and, in particular, whether use of dissection and re-entry strategies affects clin-
ical outcomes. Hybrid approaches have increased the procedural success of CTO percu-
taneous coronary intervention (PCI) but longer-term outcomes are unknown, particularly
in relation to dissection and re-entry techniques. Data were collected for consecutive
CTO PCIs performed by hybrid-trained operators from 7 United Kingdom (UK) centres
between 2012 and 2014. The primary endpoint (death, myocardial infarction, unplanned
target vessel revascularization) was measured at 12 months along with angina status.
One-year follow up data were available for 96% of successful cases (n5805). In total,
85% of patients had a CCS angina class of 2–4 prior to CTO PCI. Final successful proce-
dural strategy was antegrade wire escalation 48%; antegrade dissection and re-entry
(ADR) 21%; retrograde wire escalation 5%; retrograde dissection and re-entry (RDR) 26%.
Overall, 47% of CTOs were recanalized using dissection and re-entry strategies. During a
mean follow up of 11.5 63.8 months, the primary endpoint occurred in 8.6% (n569) of
patients (10.3% (n539/375) in DART group and 7.0% (n530/430) in wire-based cases).
The majority of patients (88%) had no or minimal angina (CCS class 0 or 1). ADR and RDR
were used more frequently in more complex cases with greater disease burden, however,
the only independent predictor of the primary endpoint was lesion length. CTO PCI in
complex lesions using the hybrid approach is safe, effective and has a low one-year
adverse event rate. The method used to recanalize arteries was not associated with
adverse outcomes. V
C2017 Wiley Periodicals,Inc.
Key words: revascularization; stent; CrossBoss; stingray; angina; hybrid approach
1
Royal Melbourne Hospital, Department of Cardiology, Park-
ville VIC 3050, Melbourne, Australia
2
Department of Cardiology, Belfast Health and Social Care
Trust, Belfast, Ireland
3
Freeman Hospital, Cardiothoracic Services, Newcastle upon
Tyne, England
4
Institute of Cellular Medicine, Newcastle University, Newcas-
tle upon Tyne, England
5
Division Cardiology, St Michael’s Hospital, University of
Toronto, Canada
6
The London Chest Hospital, Barts Health NHS Trust, London,
England, United Kingdom
7
Golden Jubilee National Hospital, West of Scotland Regional
Heart and Lung Centre, Glasgow, Scotland
8
Ninewells Hospital, Dundee, Scotland
9
Bristol Heart Institute, Bristol, United Kingdom
10
Forth Valley Royal Hospital, Larbert, United Kingdom
Conflict of interest: Nothing to report.
*Correspondence to: James C. Spratt, Forth Valley Acute Hospitals,
Stirling Rd, Larbert FK5 4WR, United Kingdom. E-mail: James.
spratt@nhs.net
Received 10 July 2016; Revision accepted 17 January 2017
DOI: 10.1002/ccd.26980
Published online 00 Month 2017 in Wiley Online Library
(wileyonlinelibrary.com)
V
C2017 Wiley Periodicals, Inc.
Catheterization and Cardiovascular Interventions 00:00–00 (2017)

INTRODUCTION
Symptomatic patients with chronic total occlusions
(CTO) undergo percutaneous coronary intervention
(PCI) less commonly than patients with nonocclusive
disease [1]. Multiple factors contribute to this referral
bias, including concerns regarding potential CTO PCI
success, safety and a lack of data demonstrating dura-
bility of outcomes. This is especially pertinent in com-
plex CTO lesions, such as those with nontapered (e.g.,
ambiguous or blunt) proximal caps, tortuosity, calcifi-
cation, or long occlusion segments where antegrade
wire escalation (AWE) or retrograde wire escalation
(RWE) techniques are less effective [2]. Antegrade dis-
section and re-entry (ADR) and retrograde dissection
and re-entry (RDR) strategies for CTO PCI have been
developed to address many of these technical and clini-
cal limitations.
ADR was first described as the subintimal tracking
and re-entry (STAR) technique [3,4]. This involves
pushing a folded or “knuckled” (usually jacketed) wire
in an antegrade direction through the subintimal space
until it re-enters the distal true lumen, which usually
occurs at a bifurcation. Long-term outcomes were poor
with STAR, with reocclusion rates of up to 50% noted
[5,6], most likely related to poor outflow as a result of
uncontrolled re-entry into distal small branches [7].
Whilst STAR has been largely abandoned, ADR has
evolved as a technique. In contemporary ADR, the
dedicated CrossBoss and Stingray system (Boston Sci-
entific, Natick, MA) facilitates safe, controlled ante-
grade dissection in the subintimal space and targeted
re-entry as soon as feasible distal to the occlusion, thus
limiting the length of dissection.
RDR techniques have also been introduced into clin-
ical practice. In the controlled antegrade and retrograde
tracking (CART) and reverse CART (rCART) techni-
ques [8], the proximal and distal true lumen are con-
nected by a common sub-intimal space within the CTO
segment, created by ballooning from a retrograde
direction (CART) or antegrade direction (rCART). As
with contemporary ADR, the length of the subintimal
space created is shorter than typically seen with STAR.
ADR and RDR approaches form an integral part of the
hybrid approach to CTO PCI [9] (Fig. 1). This approach
advocates an initial employment of wire escalation or
dissection and re-entry techniques (DART) based on
anatomical considerations, with rapid switching to alter-
native strategies if there is failure to make progress (Fig.
1). We [2] and others [10–12] have reported success
rates of >90% in complex CTO lesions when using the
hybrid approach. The long-term outcome of DART is
unknown, therefore, we sought to define long-term clini-
cal outcomes of CTO PCI procedures performed
according to the “hybrid” algorithm using contemporary
ADR and RDR techniques where required. In particular,
the influence of stent insertion in the subintimal space
was examined to determine if this was associated with
adverse outcomes.
METHODS
Data were collected for consecutive patients undergo-
ing CTO PCI from 7 centres in the United Kingdom per-
formed between January 1, 2012 and March 30, 2014.
An anonymised online database for clinical audit was
used to collect demographic data, co-morbid conditions,
procedural details, and long-term clinical outcomes.
Clinical outcomes were collected at hospital discharge,
30-days, and 12 months. Data for clinical events were
also collected at the time of a planned review or any
repeat readmission. Outcomes for patients not followed
up at the performing hospital were collected by chart
review or by telephone contact with the patient and/or
general practitioner. The operators performed procedural
data entry within their respective institutions. The CTO
PCI operator did not perform clinical follow-up
assessments.
All CTO operators practiced according to the
“hybrid” model (Fig. 1) and had performed >300 CTO
PCIs in their career. DART were encouraged as a pri-
mary approach in complex, long lesions or as bailout
techniques if AWE or RWE strategies were unsuccess-
ful. All operators were trained in the use of Crossboss
and Stingray devices. Drug-eluting stents were recom-
mended but not mandated. Adherence to guideline-
based treatment duration for dual antiplatelet therapy
was recommended.
Definitions
Angina status was defined according to the Canadian
Cardiovascular Society (CCS) classification [13]. Coro-
nary CTOs were defined as lesions with Thrombolysis
in Myocardial Infarction (TIMI) grade flow of 0 for
Fig. 1. Hybrid algorithm. [Color figure can be viewed at
wileyonlinelibrary.com]
2 Wilson et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

at least 3 months (angiographically confirmed or esti-
mated clinically from time of symptom onset if prior
angiography not available). Lesion complexity was
assessed using the J-CTO score (Multi-Centre CTO
Registry of Japan: 0 easy; 1 intermediate; 2 difficult;
3 very difficult) [14]. Location of the CTO (proximal
cap) within the vessel was defined according to the
AHA classification [15]. Angiographically visible cal-
cification within the occluded segment was coded as
none, mild (spots only), moderate (<50% of vessel cir-
cumference), severe (>50% of vessel circumference).
Tortuosity within the CTO segment was coded as
straight (no bend or <458single bend), slight
(>458single bend), moderate (2 bends >458or 1 bend
>908) or severe (2 bends >908or 1 bend >1208). Dis-
ease proximal and distal to the CTO was classified as
absent, mild, moderate, or severe. Occlusion length
was estimated by dual injections if proximal and distal
caps were defined or from the length of occlusion
apparent after guidewire crossing if not clear on initial
dual injection.
Technical success was defined as restoration of
TIMI 3 antegrade flow with <30% residual diameter
stenosis within the treated segment after CTO PCI.
Events during follow-up were evaluated for those with
successful CTO PCI and included: death, stroke, stent
thrombosis (definite or probable according to Academic
Research Consortium guidelines [16], myocardial infarc-
tion (MI: chest pain 6ECG changes with typical rise
and fall of cardiac biomarkers), target vessel revasculari-
zation (TVR) or asymptomatic re-occlusion. TVR was
defined as attempted PCI of the previously stented seg-
ment/vessel or referral for coronary artery bypass sur-
gery that included the target vessel. The primary
outcome of major adverse cardiovascular events
(MACE) was defined as the occurrence of death, MI or
unplanned TVR.
The primary and final strategies used were recorded,
as was any switch in strategy. DART for purposes of this
analysis included: primary ADR including limited ante-
grade subintimal tracking, STAR, and/or Stingray bal-
loon assisted re-entry; primary RDR (synonymous with
“rCART”); bailout ADR/RDR after unsuccessful AWE
or RWE. DART was also coded if subintimal passage
had occurred during an unsuccessful prior attempt (i.e.,
so-called “investment” procedure) or if ‘true to true’
Crossboss passage was recorded. The presence of a sub-
intimal tract (SIT) was inferred by selection of a strategy
likely to cause a SIT. Use of intravascular ultrasound
was only 15% in this cohort and was not mandated to
define the presence and length of any SIT. As such,
AWE or RWE were considered “no DART,” accepting
that there may have been inadvertent subintimal wire
tracking.
Statistical Analysis
Continuous variables are presented as mean 6
standard deviation (SD) and were compared using the
student-t or Wilcoxan rank-sum tests. Categorical vari-
ables are expressed as percentages and were compared
using the Chi-Square or the Fisher’s exact tests. Data
regarding peri-procedural complications is provided for
the overall cohort (successful and failed cases) so as to
illustrate safety profile for the intention to treat group,
whilst outcome data pertains only to those with suc-
cessful CTO PCI as our focus is the impact of dissec-
tion re-entry techniques with stent insertion in the
subintimal space upon long-term clinical outcomes.
The incidence of clinical events during follow-up was
estimated using the Kaplan-Meier method and compari-
sons between groups made using the log rank test. A
multivariable analysis using a Cox regression model
was used to determine independent predictors of the
primary outcome. Four models were tested using dif-
ferent combinations of predictors, which were selected
on the basis of clinical considerations and findings on
univariate analysis. In an ancillary analysis, we devel-
oped a propensity score model consisting of 22 varia-
bles (C statistic 50.86 and Hosmer-Lemeshow
P50.38); the baseline characteristics were well
matched after stratification by the propensity score. We
verified the proportional hazard assumption by examin-
ing the log-log plot, and by testing for a time-
dependent covariate (time 3treatment group interac-
tion) in the model. Two sided Pvalues of <0.05 were
considered statistically significant. All statistical analy-
ses were performed with SPSS Statistics version 22
(IBM Corporation, Armonk, NY).
RESULTS
Data from 969 patients were collected (Fig. 2).
Patients with >1 CTO treated during a single proce-
dure (n540) were excluded from further analysis. For
patients who required more than one CTO PCI proce-
dure to achieve success (n596), follow-up duration
was measured from the first successful procedure. For
those patients who had a second CTO in a different
vessel treated at a subsequent PCI procedure, only data
relating to the first procedure were included. Overall
technical success rate (TIMI 3 flow with <30% residu-
al stenosis), including those who required more than
one attempt, was 90.3% (839/929). Follow-up data
were available for 96% of patients with successful
CTO PCI. 805 patients were, therefore, included in this
analysis.
Baseline clinical characteristics, lesion characteristics,
and procedural details are presented in Tables (I–III) and
Long-Term Outcomes Post Hybrid CTO PCI 3
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

stratified by those with and without events during follow
up. The mean age was 65.3 610.5 years and 79% were
male. Indications for the procedure included: Stable
angina 78%, acute coronary syndrome 8%, heart failure
3%, staged post STEMI 3%, staged post NSTEMI 7%.
At baseline, 85% of patients had a CCS angina class of
2-4 and 46% had dyspnoea on exertion (NYHA class 2–
4). The majority (72%) of patients had pre-
catheterisation stress testing and intervention was based
upon the presence of symptoms with viability in the
affected territory in those without stress testing. The
mean (6SD) J-CTO score was 2.4 61.4. Second genera-
tion drug-eluting stents were implanted in 98% of
patients. The final successful strategy was classified as
dissection re-entry in 47% of cases.
Procedural Characteristics and Safety for
Overall Group
Procedural characteristics for the overall group
(n5929 patients, 1039 procedures, including repeat
attempts in 110 patients) were: procedure time
107 651 min, contrast 299 6127 ml, X-Ray skin dose
2.3 61.4 Gy, Dose Area Product (DAP) 13226 69547
cGycm
2
, fluoroscopy time 40 624 min. CTO lesions
requiring DART were significantly more complex than
those requiring wire escalation strategies (Table IV);
they had a higher mean J-CTO score (3.2 61.2 vs.
1.8 61.3, P50.001) with longer occlusion lengths
(35.9 620.8 mm vs. 18.5 613.1 mm, P<0.001) and
longer stent lengths were required (86.8 629.1 mm vs.
62.2 630.3 mm, P<0.001). Previous CABG was
more common in the DART group (30% vs. 12%,
P<0.001).
All procedural parameters were higher in the DART
group compared to the no DART group (procedural
time 132 647 vs. 86 643 min, P<0.001; Contrast
335 6124 vs. 267 6122 ml, P<0.001; Skin dose
2.8 61.3 vs. 1.8 61.2 Gy, P<0.001; DAP
16357 610525 vs. 10,632 67645 cGycm
2
,P<0.001;
Fluoroscopy time 52 623 vs. 29 619 min, P<0.001).
The mortality rate for the overall group (successful
and failed cases) was 0.3% at 30 days. MACE
occurred in 1.8% of patients. The respective MACE
rates were not significantly different per strategy
Fig. 2. Patient flow. [Color figure can be viewed at wileyonli-
nelibrary.com]
TABLE I. Baseline Clinical Characteristics of Successful PCI Group.
Variable Overall (n5805) Event (n562) No event (n5743) Pvalue
Age 65.3 610.5 65.6 69.6 65.3 610.5 0.45
Male 79% 84% 79% 0.5
Body mass index 29.1 64.8 29.2 64.8 29.1 64.9 0.71
Hypertension 70% 67% 70% 0.56
Dyslipidemia 68% 70% 68% 0.78
Diabetes Mellitus 27% 35% 26% 0.17
Current smoker 17% 24% 17% 0.20
Heart failure 9% 14% 9% 0.21
LVEF <55% 33% 40% 33% 0.21
Diseased vessels
1 52% 44% 53% 0.36
2 28% 32% 28%
3 20% 24% 19%
Prior MI 55% 66% 56% 0.11
Prior CABG 21% 23% 21% 0.74
Prior PCI 63% 66% 62% 0.36
Prior failed attempt CTO PCI 27% 31% 27% 0.46
Prior stroke 5% 8% 5% 0.23
Chronic kidney disease 14% 21% 13% 0.08
Chronic obstructive airways disease 8% 9% 7.7% 0.81
LVEF: Left ventricular ejection fraction; CABG: Coronary artery bypass graft; PCI: Percutaneous coronary intervention.
4 Wilson et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

(AWE 1.8%, ADR 0.6%, RWE 2.3%, and RDR 2.9%,
P50.34). Ellis grade 3 perforation requiring pericar-
diocentesis occurred in 1.6% of patients overall and
was not statistically different between strategies (AWE
1.0%; ADR 1.2%, RWE 2.3%, and RDR 2.9%,
P50.15).
One-Year Outcomes for Those with Successful
CTO PCI
Mean duration of follow-up was 11.5 63.8 months.
The primary endpoint occurred in 8.6% (n569) of
patients. Components of the primary endpoint are
described in Table V and sub divided according to
TABLE II. Baseline Lesion Characteristics of Successful PCI Group
Variable Overall (n5805) Event (n562) No event (n5743) Pvalue
J-CTO score 2.4 61.4 3.0 61.2 2.4 61.4 <0.001
J-CTO3 49% 71% 47% <0.001
Target vessel
Left anterior descending 27% 24% 28% 0.9
Left circumflex 17% 17% 17%
Right coronary artery 56% 59% 55%
Calcification
Any 61% 71% 60% 0.02
Mod-severe 39% 58% 37% 0.002
Tortuosity (mod-severe) 28% 26% 29% 0.07
Tortuosity proximal (moderate-severe) 18% 13% 18% 0.55
Ostial 14% 13% 14% 0.99
Bifurcation involvement 37% 43% 37% 0.34
Lesion length 26.8 619.2 32.0 618.1 26.6 619.3 0.001
Length >25 mm 41% 62% 40% 0.001
Lesion diameter 3.2 60.4 3.1 60.4 3.2 60.4 0.15
Lesion diameter <3.0mm 22% 20% 22% 0.63
Disease proximal to CTO
Any 71% 79% 70% 0.1
Moderate-Severe 43% 58% 42% 0.014
Disease distal to CTO
Any 84% 91% 84% 0.019
Moderate-severe 65% 78% 64% 0.024
In-stent restenosis 8% 14% 7% 0.12
J-CTO: Japanese Chronic total occlusion.
TABLE III. Procedural Characteristics for Successful PCI Group
Variable Overall (n5805) Event (n562) No event (n5743) Pvalue
Final strategy
AWE 48% 40% 49% 0.11
ADR 21% 18% 21%
RWE 5% 3% 5%
RDR 26% 39% 25%
Final DART 47% 57% 46% 0.14
Any DART 48% 55% 47% 0.29
Any ADR 26% 22% 26% 0.65
Any RDR 28% 40% 27% 0.03
Subintimal tracking 48% 55% 48% 0.29
CrossBoss 19% 16% 19% 0.73
Stingray balloon 14% 3% 13% 0.02
Knuckle 31% 41% 31% 0.11
Number of stents 2.5 61.1 2.6 61.1 2.5 61.1 0.8
Stent length 73.7 632.1 81.2 635.0 73.7 632.2 0.13
Stent length >50 mm 76% 81% 76% 0.43
Stent length >75 mm 48% 56% 47% 0.19
Subintimal track length 30 619 32 619.9 29 618.9 0.1
Sub-intimal stent length 19.1 621.2 25.3 621.8 18.4 620.7 0.037
Intravascular ultrasound use 15.9% 21% 16% 0.83
AWE: Antegrade wire escalation; ADR: Antegrade dissection re-entry; RWE: retrograde wire escalation; RDR: retrograde dissection re-entry;
DART: dissection and re-entry.
Long-Term Outcomes Post Hybrid CTO PCI 5
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

whether the final strategy involved DART. A total of
88% of patients had no or minimal angina (CCS class
0 or 1) and 80% described no dyspnoea. Follow-up
angiography was performed in 24% of patients, more
commonly in the DART group (32 vs. 17%,
P<0.001). This was usually scheduled and most fre-
quently to reassess distal vessel run-off and remodel-
ling. Normal angiographic appearances were seen in
70% of patients. The vessel was occluded in 13% of
those who underwent angiography (one third asymp-
tomatic, all managed conservatively), with no differ-
ence in frequency of vessel occlusion between DART
and wire-based groups.
Predictors of Adverse Events during Follow-Up
Differences between those who had an adverse event
and those who did not are described in Tables (I–III).
Those with events were more likely to have higher J-
CTO scores, longer lesions, greater degrees of calcifica-
tion, greater disease burden (both proximal and distal to
the lesion) and to have had RDR as the final strategy.
Freedom from the primary endpoint according to final
PCI strategy and lesion length is presented in Fig. 3.
Differences between patients according to final suc-
cessful strategy are described in Table IV. As expected
from the hybrid algorithm, those patients with RDR as
the final strategy were more likely to have a J-CTO
score >2 (79 vs. 37%, P<0.001), longer occlusion
length (>25 mm in 77 vs. 28%, P<0.001), longer
stent length (>75 mm in 71 vs. 39%, P<0.001), bifur-
cation involvement (46 vs. 34%, P50.003), disease
proximal to the CTO (55 vs. 39%, P<0.001), and dis-
ease distal to the CTO (72 vs. 62%, P50.02).
There remained no difference in the primary out-
come between DART and no DART groups when the
overall cohort was analysed (i.e., inclusion of failed
cases). Similarly, there was no difference in the prima-
ry outcome if CrossBoss true-to-true passage (n527)
was classified as AWE rather than ADR (DART
10.1% vs. no DART 7.5%, P50.2).
Multivariable Cox and Propensity Analysis
Multivariable analyses using different combinations
of factors predictive of outcome in univariate analysis
are presented in Table VI. The only independent pre-
dictor of the primary endpoint was occlusion length. In
TABLE IV. Differences in Lesion Characteristics According to Final Strategy
AWE ADR RWE RDR P
Mean JCTO 1.7 61.2 3.2 61.3 2.6 61.2 3.35 61.03 <0.001
JCTO>2 26% 64% 59% 77% <0.001
Mean lesion length (mm) 17.9 612.6 31.5 618.9 24.8 615.8 39.4621.8 <0.001
Length >25 mm 18% 56% 29% 74% <0.001
Lesion diameter <3.0 mm 24% 16% 27% 22% 0.16
Disease proximal to CTO (moderate-severe) 45% 38% 38% 55% 0.002
Disease distal to CTO (moderate-severe) 62% 63% 64% 74% 0.04
Calcification (moderate-severe) 31% 43% 39% 50% <0.001
Bifurcation 34% 31% 43% 45% 0.01
In-stent restenosis 4% 3% 0.4% 0.5% <0.001
Mean stent length (mm) 60.4 629.9 82.3 628.5 78.0 630.6 89.9629.3 <0.001
Stent length >75 mm 30% 61% 54% 70% <0.001
Subintimal stent length (mm) 0 30.3 621.5 0 28.2 618.3 0.31
AWE: Antegrade wire escalation; ADR: Antegrade dissection re-entry; RWE: Retrograde wire escalation; RDR: Retrograde dissection re-entry; J-
CTO: Japanese Chronic total occlusion.
TABLE V. One Year Outcomes According to DART Versus Wire-Based Strategy
Variable Overall (n5805) DART (n5375) Wire based (n5430) Pvalue
Composite endpoint 8.6% (69) 10.3% (39) 7.0% (30) 0.1
Death
All-cause 2.5% (21) 2.6% 2.3% 0.82
Cardiac 1.0% (8) 1.1% 0.9% 1.0
MI
All 3.0% (25) 3.7% 1.9% 0.13
Stent thrombosis 1.4% (11) 1.9% 0.9% 0.36
Unscheduled TVR 5.0% (41) 6.1% 3.3% 0.06
Target lesion revascularization 4.5% (37) 5.8% 3.7% 0.19
DART: dissection and re-entry.
6 Wilson et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

the propensity analysis, a final strategy of DART, com-
pared to no DART, was not independently associated
with the primary outcome (adjusted HR 50.77, 95%CI
0.42–1.41, P50.39).
DISCUSSION
We report very high (>90%) freedom from death, MI
and unplanned TVR with durable symptom relief at 12
months in a large cohort of patients with complex CTOs.
DART was used frequently (47%) as part of the hybrid
approach, particularly in complex lesions, however, the
only independent predictor of adverse outcomes was
lesion length. Our findings suggest that the majority of
patients with complex CTOs, including those with
previous bypass grafting, can now expect to achieve suc-
cessful recanalization of their occluded arteries and have
complete or near complete symptom relief out to 1 year.
Adverse Event Rates in Previous CTO PCI
Outcome Studies
We report the largest study to date describing
longer-term clinical outcomes according to revascular-
isation strategy after contemporary CTO PCI. Despite
the complex cohort recruited, the incidence of adverse
events during long-term follow up was favourable with
a TLR of 4.5%. Previous reports from CTO cohorts
report TLR of 6.3–10.7% [17–19] with second genera-
tion DES and 17% with first generation DES [20].
Fig. 3. Survival curves for freedom from the primary endpoint according to strategy (DART
vs. no DART) and Lesion length. [Color figure can be viewed at wileyonlinelibrary.com]
TABLE VI. Multivariate Analysis (Cox Regression Models)
Model 1 Model 2 Model 3 Model 4
Variable Odds ratio Pvalue Odds ratio Pvalue Odds ratio Pvalue Odds ratio Pvalue
Final Approach
AWE 1
ADR 0.66 (0.3–1.5) 0.32
RWE 0.68 (0.16–2.9) 0.61
RDR 1.1 (0.53–2.2) 0.83
DART (any) 0.74 (0.4–1.4) 0.35
ADR (any) 0.67 (0.35–1.3) 0.23
RDR (any) 1.3 (0.67–2.4) 0.47
Lesion length >25 mm 2.5 (1.3–4.8) 0.007 2.9 (1.5–5.6) 0.002 2.6 (1.5–4.8) 0.001 2.3 (1.2–4.3) 0.01
Stent length >50 mm 0.85 (0.4–1.8) 0.85 0.85 (0.4–1.7) 0.65 0.85 (0.41–1.7) 0.66 0.79 (0.39–1.6) 0.52
Disease distal
(moderate or severe)
1.6 (0.82–2.9) 0.17 1.67 (0.85–3) 0.15 1.6 (0.83–3.0) 0.17 1.6 (0.86–3.1) 0.16
Bifurcation 1.2 (0.8–2.2) 0.36 1.3 (0.8–2.2) 0.3 1.3 (0.7–2.2) 0.36 1.3 (0.8–2.2) 0.34
Diabetes 1.4 (0.8–2.4) 0.3 1.4 (0.8–2.4) 0.28 1.4 (0.8–2.4) 0.18 1.3 (0.8–2.4) 0.31
In-stent restenosis 1.5 (0.6–3.6) 0.4 1.3 (0.6–3.1) 0.64 1.4 (0.6–3.3) 0.23 1.4 (0.6–3.5) 0.47
Presented as Hazard ratio (95% confidence interval) and Pvalue.
AWE: Antegrade wire escalation; ADR: Antegrade dissection re-entry; RWE: retrograde wire escalation; RDR: retrograde dissection re-entry;
DART: dissection and re-entry.
Long-Term Outcomes Post Hybrid CTO PCI 7
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

Technical Approaches and Disease Burden/
Complexity
We found no statistically significant difference in
MACE at 12 month follow-up between patients
treated with DART versus no-DART approaches on
multivariable analysis. Directly comparing strategies as
a predictor of outcomes is challenging, particularly as
certain adverse anatomical characteristics may favour
the use of a particular strategy and in turn prejudice
outcomes. Within our database, DART use, particularly
RDR, was strongly associated with greater lesion com-
plexity. Higher TLR rates in retrograde cases have
been previously reported [21] and were postulated to
relate to longer stent length and use in more complex
lesions, rather than the retrograde approach per se.
Whilst RDR approaches had the highest event rate in
our cohort, all baseline characteristics were more
unfavourable in this group.
Previous experience with sub-intimal stenting fol-
lowing use of the STAR technique suggested disap-
pointing longer-term outcomes [5,6]. Data regarding
outcomes with modern techniques are much more
promising, which likely relates to the more favourable
distal run-off achievable with targeted and controlled
re-entry. Three single-centre North American studies
[18,22,23] have demonstrated no difference in adverse
clinical events during intermediate-term follow-up
between those treated with DART (antegrade or retro-
grade) versus conventional wire escalation strategies. A
sub-analysis of the J PROCTOR registry, which used
IVUS to detect subintimal guidewire tracking found
greater late lumen loss in the sub-intimal tracking
group, however, this did not translate into an increase
in clinical events [24]. The majority of events in these
studies were related to TVR rather than death or MI.
Nonetheless, the duration of follow-up within these
reports and in our study is insufficient to exclude an
excess of very late stent thrombosis in patients with
subintimal stent insertion, thus on-going surveillance is
required.
Does the Direction of Dissection Matter?
Both ADR and RDR (“reverse CART”) involve dis-
section and subsequent stenting within the subintimal
space. Concern has been expressed that ADR poses a
higher risk of subintimal hematoma formation and
uncontrolled dissection, which may be associated with
unnecessarily longer stent lengths or compression of
the distal lumen with consequent undersizing of stents.
DART cases in our study were roughly half antegrade
and half retrograde, and the length of estimated subin-
timal stenting was similar in the ADR and RDR
groups. There was no signal that those cases completed
with ADR (primarily using CrossBoss and Stingray)
had higher event rates. Indeed, the ADR event rate
closely approximated that of AWE cases, despite
higher complexity in the ADR cohort. Whilst there
was an excess of events in the patients requiring RDR
as the final strategy, this reflected increased lesion
complexity and target vessel disease burden. When
adjusted for complexity, no independent relationship
existed between RDR and adverse outcomes.
Predictors of Adverse Events during Long-Term
Follow-Up
The only independent predictor of adverse clinical
events in our study was lesion length, The presence of
a longer lesion and consequently longer stent length is
likely to increase the risk of restenosis. It is intuitive
that increased disease burden, as defined by disease
proximal and distal to the CTO, would also contribute
to an increased risk for TVR, and distal disease may
also increase the risk for re-occlusion by contributing
to poor runoff. There was a numerical trend in our
study for excess events was seen in those who had
moderate or severe disease distal to the CTO lesion.
This may have implications for case selection, where
those vessels with very severe and diffuse distal dis-
ease are best managed medically. It is worth acknowl-
edging, however, that it can be difficult to gauge how
the distal vessel will behave following recanalization
in the presence of flow-mediated positive remodelling
[25].
Finally, in the study by Rinfret and colleagues [18],
treatment of occlusive in-stent restenosis was indepen-
dently associated with increased need for TVR. How-
ever, occlusive disease within a previously implanted
stent was not a predictor of the primary endpoint in
our larger study.
Limitations
Whilst this is a large, multicenter study, it is based
upon registry data entered by the respective operators.
There was no core laboratory for independent review
of procedural angiograms, nor independent adjudication
of clinical events during follow-up, however, consensus
regarding the classification of clinical events was
required between at least two authors. Systematic
assessment of angina class was not undertaken and def-
inition of procedural complications is potentially sub-
ject to observer bias. Follow-up was not 100%, thus
the adverse event rate may theoretically be higher. The
proportion of cases with DART may have been higher
than reported; inadvertent subintimal passage of the
wire during AWE or RWE can occur and is generally
only detectable by IVUS, which was used in the
8 Wilson et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

minority of cases. Finally, the results of this study may
not be reproducible amongst low-volume operators or
those who have limited experience of hybrid
approaches to CTO lesions.
CONCLUSIONS
CTO PCI in complex lesions using the hybrid
approach is safe, has a low 1-year adverse event rate,
and produces durable relief from angina. Use of con-
temporary dissection re-entry techniques did not
adversely affect outcomes, with only lesion length pre-
dicting a higher incidence of MACE at 1 year.
ACKNOWLEDGMENTS
The authors would like to acknowledge the follow-
ing for assistance with data entry: Jane Blair, Jane
Geddes, Roshan Paranamana, Aadil Shaukat, and
MurugapathyVeerasamy.
The authors would also like to thank Optima Educa-
tion and Adrian Brown for illustrative input.
DISCLOSURES STATEMENTS
Conflict of interest: W.M. Wilson, A.T. Yan, K.G.
Oldroyd, and M. McEntegart have no conflicts of
interest.
J. Strange and A. Bagnall have received income
from proctoring from Abbott Vascular.
Dr S. Walsh is a Consultant to Abbott Vascular,
Boston Scientific, Medtronic, and Vascular Solutions.
He has also received research funding from Abbott
Vascular, Boston Scientific, and Nitiloop.
Dr Hanratty is a Consultant to Abbott Vascular, Bos-
ton Scientific, Medtronic, and Vascular Solutions.
Dr Egred has received consultancy and honorarium
from Abbott Vascular, Boston Scientific, Volcano,
Vascular Perspectives, and Spectranetics.
Dr E. Smith has received income from proctoring
from Boston Scientific and Vascular Perspectives
Dr J. Irving has received income from proctoring
from Boston Scientific and Vascular Perspectives.
Dr J. C. Spratt is a Consultant to Abbott Vascular,
Boston Scientific, Asahi Intecc, Vascular Perspectives,
and Vascular Solutions. He has received research fund-
ing from Abbott Vascular, Boston Scientific, and
Nitiloop
REFERENCES
1. Grantham JA, Marso SP, Spertus J, House J, Holmes DR, Jr,
Rutherford BD. Chronic total occlusion angioplasty in the Unit-
ed States. JACC Cardiovasc Interv 2009;2:479–486.
2. Wilson WM, Walsh SJ, Yan AT, et al. Hybrid approach
improves success of chronic total occlusion angioplasty. Heart
2016;102:1486–1493.
3. Colombo A, Mikhail GW, Michev I, et al. Treating chronic total
occlusions using subintimal tracking and reentry: The STAR tech-
nique. Catheter Cardiovasc Interv 2005;64:407–411. discussion 412.
4. Carlino M, Godinoa C, Latib A, Moses JW, Colombo A. Subin-
timal tracking and re-entry technique with contrast guidance: A
safer approach. Catheter Cardiovasc Interv 2008;72:790–796.
5. Godino C, Latib A, Economou FI, et al. Coronary chronic total
occlusions: mid-term comparison of clinical outcome following
the use of the guided-STAR technique and conventional antero-
grade approaches. Catheter Cardiovasc Interv 2012;79:20–27.
6. Valenti R, Vergara R, Migliorini A, et al. Predictors of reocclu-
sion after successful drug-eluting stent-supported percutaneous
coronary intervention of chronic total occlusion. J Am Coll Car-
diol 2013;61:545–550.
7. Visconti G, Focaccio A, Donahue M, Briguori C. Elective ver-
sus deferred stenting following subintimal recanalization of cor-
onary chronic total occlusions. Catheter Cardiovasc Interv 2015;
85:382–390.
8. Sumitsuji S, Inoue K, Ochiai M, Tsuchikane E, Ikeno F. Funda-
mental wire technique and current standard strategy of percutane-
ous intervention for chronic total occlusion with histopathological
insights. JACC Cardiovasc Interv 2011;4:941–951.
9. Brilakis ES, Grantham JA, Rinfret S, et al. A percutaneous
treatment algorithm for crossing coronary chronic total occlu-
sions. JACC Cardiovasc Interv 2012;5:367–379.
10. Christopoulos G, Menon RV, Karmpaliotis D, et al. Application
of the “hybrid approach” to chronic total occlusions in patients
with previous coronary artery bypass graft surgery (from a Con-
temporary Multicenter US registry). Am J Cardiol 2014;113:
1990–1994.
11. Michael TT, Mogabgab O, Fuh E, et al. Application of the
“hybrid approach” to chronic total occlusion interventions: a
detailed procedural analysis. J Interv Cardiol 2014;27:36–43.
12. Syrseloudis D, Secco GG, Barrero EA, et al. Increase in J-CTO
lesion complexity score explains the disparity between recanali-
sation success and evolution of chronic total occlusion strate-
gies: insights from a single-centre 10-year experience. Heart
2013;99:474–479.
13. Campeau L. Letter: Grading of angina pectoris. Circulation
1976;54:522–523.
14. Morino Y, Abe M, Morimoto T, et al. Predicting successful
guidewire crossing through chronic total occlusion of native cor-
onary lesions within 30 minutes: The J-CTO (Multicenter CTO
Registry in Japan) score as a difficulty grading and time assess-
ment tool. JACC Cardiovasc Interv 2011;4:213–221.
15. Austen WG, Edwards JE, Frye RL, et al. A reporting system on
patients evaluated for coronary artery disease. Report of the Ad
Hoc Committee for Grading of Coronary Artery Disease, Coun-
cil on Cardiovascular Surgery, American Heart Association. Cir-
culation 1975;51:5–40.
16. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in
coronary stent trials: A case for standardized definitions. Circu-
lation 2007;115:2344–2351.
17. Kandzari DE, Kini AS, Karmpaliotis D, et al. Safety and Effec-
tiveness of Everolimus-Eluting Stents in Chronic Total Coronary
Occlusion Revascularization: Results From the EXPERT CTO
Multicenter Trial (Evaluation of the XIENCE Coronary Stent,
Performance, and Technique in Chronic Total Occlusions).
JACC Cardiovasc Interven 2015;8:761–769.
18. Rinfret S, Ribeiro HB, Nguyen CM, Nombela-Franco L, Urena
M, Rodes-Cabau J. Dissection and re-entry techniques and
Long-Term Outcomes Post Hybrid CTO PCI 9
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).

longer-term outcomes following successful percutaneous coro-
nary intervention of chronic total occlusion. Am J Cardiol 2014;
114:1354–1360.
19. Galassi AR, Sianos G, Werner GS, et al. Retrograde Recanaliza-
tion of Chronic Total Occlusions in Europe: Procedural, In-
Hospital, and Long-Term Outcomes From the Multicenter
ERCTO Registry. J Am Coll Cardiol 2015;65:2388–2400.
20. Mehran R, Claessen BE, Godino C, et al. Long-term outcome
of percutaneous coronary intervention for chronic total occlu-
sions. JACC Cardiovasc Interv 2011;4:952–961.
21. Michael TT, Mogabgab O, Alomar M, et al. Long-term out-
comes of successful chronic total occlusion percutaneous coro-
nary interventions using the antegrade and retrograde approach.
J Interv Cardiol 2014;27:465–471.
22. Mogabgab O, Patel VG, Michael TT, et al. Long-term outcomes
with use of the CrossBoss and stingray coronary CTO crossing
and re-entry devices. J Invasive Cardiol 2013;25:579–585.
23. Amsavelu S, Christakopoulos GE, Karatasakis A, et al. Impact
of Crossing Strategy on Intermediate-term Outcomes After
Chronic Total Occlusion Percutaneous Coronary Intervention.
Can J Cardiol 2016;32:1239 e1–1239 e7.
24. Muramatsu T, Tsuchikane E, Oikawa Y, et al. Incidence and
impact on midterm outcome of controlled subintimal tracking in
patients with successful recanalisation of chronic total occlu-
sions: J-PROCTOR registry. Eurointervention 2014;10:681–688.
25. Galassi AR, Tomasello SD, Crea F, et al. Transient impairment
of vasomotion function after successful chronic total occlusion
recanalization. J Am Coll Cardiol 2012;59:711–718.
10 Wilson et al.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).