ORIGINAL PAPER

Preprocedural coronary CT angiography significantly improvessuccess rates of PCI for chronic total occlusion

Andreas Rolf • Gerald S. Werner • Annika Schuhback • Johannes Rixe • Helge Mollmann •

Holger M. Nef • Constantin Gundermann • Christoph Liebetrau • Gabriele A. Krombach •

Christian W. Hamm • Stephan Achenbach

Received: 1 March 2013 / Accepted: 14 June 2013

� Springer Science+Business Media Dordrecht 2013

Abstract Chronic total occlusions of coronary arteries

occur in about 20 % of patients with suspected coronary

artery disease and are more frequent with increasing age.

The success rate of interventions is lower (55–80 %)

compared to conventional lesions ([90 %). Coronary CT

angiography (coronary CTA) provides information about

the occluded segment, which cannot be obtained from

invasive angiograms (XA). We therefore hypothesized that

preprocedural coronary CTA may improve success rates of

percutaneous coronary intervention (PCI) for coronary

arteries (CTO). 30 patients with chronic total coronary

artery occlusions (mean age 73 years, 26 men) and pre-

dicted high complexity were imaged by coronary CTA

prior to PCI for CTO. CT data sets were acquired with a 64

detector row dual source scanner and retrograde ECG

gating, 0.6 mm collimation and z-flying focal spot,

yielding isovoxel spatial resolution of about 0.4 mm.

Based on the CT data sets, established complexity criteria

for CTO (Euro CTO club, Di Mario et al. in EuroInter-

vention 3(1):30–43, 2007) were evaluated and compared to

invasive coronary angiography. Three-dimensional vol-

ume-rendered images of the occluded coronary artery were

displayed in the catheterization lab during PCI to guide the

advancement of the wire. PCI success, defined as the

ability to advance the guide wire into the distal lumen with

thrombolysis in myocardial infarction III flow was com-

pared to 43 controls without coronary CTA using propen-

sity score matching based on established criteria of

procedural success. The course of the occluded segments

was visualized by coronary CTA in all cases. Calcification,

lesion length, stump morphology and presence of side

branches were underestimated by invasive angiograms

when compared to coronary CTA. PCI success rate in 30

patients who underwent pre-procedural CTA was signifi-

cantly higher than in patients without prior coronary CTA

[unmatched: CT 90 % (27/30) vs. no CT 63 % (27/43),

p = 0.009; matched: CT 88 % (22/25) vs. no CT 64 % (16/

25) p = 0.03]. Through information not readily seen on

invasive coronary angiography, coronary CTA can signif-

icantly enhance success rates of PCI for CTO.

Keywords Coronary computed tomography angiography �Chronic total occlusion � PCI

Introduction

Chronic total occlusions of coronary arteries (CTO) occur

in about 20 % of patients with known or suspected coro-

nary artery disease [1], are more frequent in the right

coronary artery (RCA), and increase with advancing

A. Rolf (&) � H. Mollmann � H. M. Nef � C. Gundermann �C. Liebetrau � C. W. Hamm

Department of Cardiology, Kerckhoff Heart Center,

Benekestrasse 2-8, 61231 Bad Nauheim, Germany

e-mail: a.rolf@kerckhoff-klinik.de; andreas.rolf@me.com

A. Rolf � J. Rixe � H. Mollmann � H. M. Nef � C. W. Hamm

Department of Cardiology, University of Giessen, Klinikstr. 33,

35392 Giessen, Germany

G. S. Werner

Department of Cardiology, Klinikum Darmstadt, Grafenstr. 9,

64283 Darmstadt, Germany

A. Schuhback � S. Achenbach

Department of Cardiology, University of Erlangen,

Ulmenweg 18, 91054 Erlangen, Germany

G. A. Krombach

Department of Radiology, University of Giessen, Klinikstr. 33,

35392 Giessen, Germany

123

Int J Cardiovasc Imaging

DOI 10.1007/s10554-013-0258-y

patient age [2, 3]. The definition of CTO is an obstruction

of a native coronary artery for at least 3 months showing

no ante grade blood flow as assessed by coronary angiog-

raphy [thrombolysis in myocardial infarction (TIMI) grade

0] [3–5]. Revascularization of CTOs through percutaneous

coronary intervention (PCI) is challenging. Reported suc-

cess rates are between 55 % and 80 % [3, 6–10]. In com-

parison, the success rate for PCI of coronary artery stenoses

is more than 90 % [11]. However, patients benefit from

successful CTO recanalization through reduced long-term

cardiac mortality and decreased need for coronary artery

bypass surgery [12]. The most common cause for PCI

failure is the inability to successfully pass a guidewire

across the occluded segment [8]. Many predictors for PCI

failure have been identified, such as length of the occluded

segment, severe calcification, tortuosity of the occlusion,

stump morphology, distal vessel opacity, presence of

sidebranches and presence of bridging collaterals [7, 8, 13–

15]. The Euro CTO club has defined categorical com-

plexity criteria by which to estimate the expected proce-

dural success (compare Table 1 [5]). The success rate of

procedures can be expected as high as 90 % with pre-

dominantly favourable characteristics but may drop to 60

% in view of several unfavourable characteristics [5].

Even after crossing the proximal cap, further advance-

ment of the guidewire is challenging as the course of the

vessel is not known and the intraluminal position of the

wire cannot be verified.

Coronary computed tomography angiography (coronary

CTA) has been suggested to provide useful information for

procedure planning and prediction of procedure success

[16–21]. One hallmark of coronary CTA is the ability to

visualize the complete course of the occluded segment of

the vessel. This additional information can help to control

the guidewire position during advancement through the

lesion.

We therefore hypothesized that CT may provide useful

additional information beyond invasive coronary angiog-

raphy for CTO procedures. First we determined differences

in CTO classification by CT and invasive angiography and

in a second step, we analysed whether making volume-

rendered coronary CTA images available in the catheteri-

zation laboratory during the CTO procedure might increase

success rates of the procedure.

Patients and methods

Patients and ethics

Between December 2008 and December 2009 30 patients

were referred to our institution to undergo coronary CT

angiography prior to elective interventional recanalization

of a chronic total occlusion at Darmstadt Community

Hospital. Patients were referred because of the expected

high complexity of the intervention, 10 of them had failed

attempts at other institutions. All patients had symptomatic

angina despite optimal medical treatment, the majority of

patients had either undergone exercise testing or had

inducible ischemia on SPECT or cardiac MRI. Viability

diagnostics were available for some patients, those who

had not undergone viability testing had no regional wall

motion abnormalities on echo. All patients had diagnostic

quality of their scans and were therefore included in this

retrospective, explorative analysis. In order to match these

patients with a control cohort, we also included 43 patients

in the study who underwent PCI for CTO at the same

institution who had a similar complexity, but without prior

coronary CTA. A chronic total coronary occlusion was

defined as the presence of TIMI 0 flow within the occluded

segment and occlusion duration C3 month.

The study was approved by the ethics committee of the

University of Giessen.

Coronary CTA

All patients were examined using a dual source CT system

(Somatom Definition�, Siemens Healthcare, Forchheim,

Germany). The CTA was performed 2 ± 0.2 days prior to

the intervention. After placing an 18-G intravenous cath-

eter into an antecubital vein, and placing the ECG

Table 1 Euro CTO club complexity criteria

Simple Complex

Vessel diameter (mm) [3.0 \3.0

Occlusion length (mm) \20 [20

Calcium occluded segment None to moderate Severe

Tortuosity occluded segment Minimal to

moderate

Severe

Occlusion stump Tapered Blunt or

absent

Distal vessel opacification Good to excellent Poor

Distal vessel disease Absent or

moderate

Severe

Tandem/multiple occlusions No Yes

Tortuosity proximal to

occlusion

Minimal to

moderate

Severe

Disease of the proximal

segment

Absent or

Moderate

Severe

Expected guiding catheter

support

Good Poor

Ostial location No Yes

Previous attempts No Yes

Renal insufficiency Yes No

Expected patient tolerance Good Poor

Int J Cardiovasc Imaging

123

electrodes in standard position, all patients received 0.8 mg

of glycerole trinitrate sublingually immediately before the

scan. To achieve sufficient heart rate control patients were

administered between 5 and 15 mg of metroprolol i.v., with

a target heart rate below 65/min. Initially a topogram and

an unenhanced CT acquisition for calcification scoring

(64 9 0.6 collimation, variable pitch of 0.2–0.43,

depending on heart rate, reconstructed slice thickness

3.0 mm, reconstruction kernel B35f, respectively) were

obtained. A test bolus approach was performed determin-

ing the contrast agent transit time (10 ml of iopamidole,

containing 370 mg iodine/ml, followed by 50 ml of iso-

tonic saline, both at 6 ml/s). CT angiography was then

performed after injection of 80–120 ml of contrast agent at

6 ml/s, followed by 50 ml of saline. The exact amount of

contrast agent to be injected was calculated by the expected

duration of CT data acquisition and thus varied between

different patients. Data acquisition was performed from the

level of the tracheal bifurcation to the diaphragm in cranio-

caudal direction using a detector collimation of

32 9 0.6 mm and slice acquisition 64 9 0.6 mm by means

of a z-flying focal spot. Gantry rotation time was 330 ms

and pitch was 0.2–0.43. Tube current was adapted auto-

matically to each patient’s weight using CareDose 4D�

automatic exposure control and a reference tube current of

320 mAs. Tube voltage was 120 kV for both tubes, and the

ECG pulsing window was fixed to 35–70 % of the RR

interval for all patients.

Image data were reconstructed with a slice thickness of

0.75 mm and increment of 0.3 mm using a single RR-

interval reconstruction approach, resulting in a temporal

resolution of 83 ms. For synchronization of data recon-

struction with the ECG signal, retrospective gating was

performed. Image reconstructions were rendered in 5 %

increments from 35 to 70 % of the RR cycle and for each

patient, the data set with least motion was identified and

used for further processing and analysis.

In addition to the evaluation of axial and oblique mul-

tiplanar reconstructions (MPR), oblique maximum inten-

sity projections (MIP) and curved multiplanar

reconstructions (curved MPR) as well as three dimensional

surface weighted volume-rendering technique reconstruc-

tions (VRT) were rendered (see Figs. 1, 2).

Invasive coronary angiography and intervention

All invasive procedures were performed by an experienced

CTO interventionist at Darmstadt Community Hospital

(GSW). First, a diagnostic angiogram was obtained. A

single plane digital angiography system (Philips Medical

System, Netherlands) was used to obtain on average 5

cineangiograms for each patient at 25 frames/s. The gantry

angle of the projections and the diameter of the catheters

were chosen at the discretion of the cardiologist performing

the study. Most patients underwent right femoral 7F cath-

eterization. For contralateral injections 4F sheaths were

introduced in the left femoral artery and exchanged for 7F

sheaths in case the strategy was changed to a retrograde

approach. A ‘‘step-up-stiffness’’ strategy was chosen to

select the appropriate guide wire to be used for crossing the

proximal cap. The following guide wires were used: Mir-

acle family 3/6/12, Pilot family (50/150), Confianza,

Confianza Pro 12 G, PT2, Runthrough, and Whisper fam-

ily. In case of a retrograde approach via septal collaterals,

the collaterals were accessed via the patent contralateral

coronary artery and passed with a floppy, hydrophilic wire

and a Finecross microcatheter. Once the collateral was

successfully passed with the wire, the Finecross catheter

Fig. 1 Curved and straightened

curved MPR for measurement

of true occlusion length (patient

example 2—LAD)

Int J Cardiovasc Imaging

123

was exchanged for an over-the-wire (OTW) balloon. Low-

pressure dilation of the collateral was achieved with a

1.25–1.5 mm OTW balloon to enable the penetration of a

balloon to the segment distal to the CTO. Penetration of the

distal CTO cap, initially with a soft wire and subsequently

with increasingly stiff wires, was then attempted. Wire

penetration was supported by the OTW balloon catheter.

Once the distal CTO cap had been penetrated, the wire was

advanced as far as possible and a distal channel was created

by repetitive balloon inflations. Next, an attempt was made

to cross the proximal CTO cap and the middle portion of

the CTO. The wire was then advanced to the created distal

channel and finally to the distal true lumen. In case the first

wire had penetrated the false lumen a second guide wire

would sometimes be used to follow another course to the

true lumen (parallel wire technique).

Both groups were treated according to the same inter-

ventional strategy.

Procedural success of the intervention was defined as the

ability to cross the lesion with the guide wire and advance

the guide wire into the distal true lumen, with subsequent

dilatation and stent placement with final TIMI3 flow.

To guide the procedure in the catheterization lab, 3D-

rendered VRT reconstructions of the occluded vessels were

projected in the catheterization suite. Orientations were

automatically aligned with the angulation of the c-arm.

Image processing and analysis

Processing of the coronary cineangiograms was performed

on an offline workstation and all metric measures were

calibrated against the diameter of the diagnostic or guide

catheter. Qualitative assessments were performed visually

by two experienced observers blinded to the treatment

group in consensus reading. For each patient, standardized

complexity criteria defined by the Euro CTO club (com-

pare Table 1 [5]) were measured and recorded. The com-

plexity criteria are categorical with two or three tiers in

each variable. Patients were stratified into each category of

each variable in consensus reading of two experienced

physicians familiar with both CT images and cineangio-

grams blinded for the corresponding measurement of the

CTA/cineangiogram respectively. There was no clear cut

off between categories. In case of calcification for example

lesions were rated not calcified, moderately calcified or

severely calcified. Depending on the category of each

variable the lesion was rated as complex or not complex.

The processing of the coronary CTA images was per-

formed on a dedicated workstation (syngo� workplace,

Siemens Medical Solutions, Forchheim, Germany). Vessel

and lumen diameters were measured in orthogonal pro-

jections on the curved MPR images and segment lengths

were measured on curved MPR images to yield the true

length of the occlusion. Qualitative assessments were

performed by two experienced observers. All lesions were

categorized according to the complexity criteria of the Euro

CTO club.

Propensity score matching and statistical analysis

To match patients between those who underwent CT and

those who did not, we chose 8 variables, which had been

Fig. 2 Invasive coronary angiogram of the RCA and corresponding VRT image shows that the true course of the RCA is much longer than the

invasive angiogram suggests

Int J Cardiovasc Imaging

123

shown to predict procedural success and with special rel-

evance to CT imaging: visibility of the occluded segments

(for example through microchannels in case of angiograms

or by CT), degree of calcification, occlusion length (as

categorical variable (B20 or [20 mm), tortuosity of the

occluded segment (as categorical variable (minimal, mod-

erate, severe), stump morphology (as categorical variable

(tapered, blunt, absent), opacity of the distal vessel seg-

ment, presence of side branches and ostial/non ostial

lesion. Also compare Table 1.

We performed nearest neighbour matching without

replacement and a caliper of 0.2. We used standardized

differences to report the baseline balance of the matched

variables. Differences between the two groups regarding

the complexity criteria were tested for statistical signifi-

cance using student’s t test for paired variables, Wilcoxon

signed rank test for non independent variables and

McNemar’s test, where appropriate depending on the scale

of the variable [22]. Additionally, multivariate logistic

regression analysis was performed to investigate the effect

that preprocedural coronary CTA has on procedural suc-

cess. Vessel course visibility, calcification, occlusion

length, tortuosity, stump morphology, distal vessel opacity,

presence of ostial lesions and sidebranches were included

in the model, as they had been identified as predictive of

procedural success in prior publications [7, 8, 13–16, 18].

Results are presented as odds ratios and 95 % confidence

intervals. A p value of B0.05 was considered statistically

significant. All computations were made using STATA11

(StataCorp LP, College Station, Texas, USA), propensity

score matching was performed with the psmatch2 proce-

dure by Leuven and Sianesi [23] balance testing of stan-

dardized differences was performed with the pstest

procedure by the same authors.

Results

The mean heart rate during coronary CT angiography was

61 ± 11 bpm and the mean dose length product of coro-

nary CT angiography was 640.7 ± 256.5, corresponding to

an estimated effective dose of 10.9 ± 4.2 mSv. In all 30

patients who underwent coronary CT angiography, data

sets were considered fully diagnostic and were included in

the analysis. Baseline characteristics of the patients are

displayed in Table 2.

The overall safety of the procedure was very good.

Creatinine levels increased significantly after the procedure

but not clinicially relevant (1.05 ± 0.22 vs.

1.15 ± 0.25 mg/dl, p = 0.001). However there was no

significant difference in postprocedural creatinine levels

between groups (1.18 ± 0.22 mg/dl CTA group vs.

1.12 ± 0.23 mg/dl no prior CTA group, p = 0.21). Three

patients had dissections (CTA group 2, no prior CTA 1

patients) all three could be covered by stents. 3 patients had

minor bleeding complications at the access site (CTA

group 1, without prior CTA 2 patients), there was one

major bleeding complication necessitating one blood

transfuion in the group without prior CTA. One temporary

acute kidney injury was documented in the CTA group,

maximum creatinine was 2.2 mg/dl, which returned to

normal at demission. There was one tamponade docu-

mented in the group without prior CTA, which could be

drained by pericardiocentesis without residual complica-

tions. The overall p for procedural complications was 0.6.

The course of the occluded segment was visible in all of

the 30 coronary CTA data sets but only in 8 of the 30

invasive coronary angiograms (p = 0.0001). The occlusion

length appeared to be significantly shorter on XA angio-

grams than in the projection-free measurements on the

curved MPR CT images (coronary CTA 30.2 ± 3.7 mm

vs. XA 21.9 ± 3.3 mm, p = 0.00001). There was no sig-

nificant difference in the measurements of the proximal

vessel diameter (coronary CTA 2.4 ± 0.25 mm vs. XA

2.3 ± 0.1 mm, p = 0.7289). The degree of calcification

(none to moderate vs. severe) was estimated to be signifi-

cantly smaller on XA angiograms compared to coronary

CTA (coronary CTA 8 of 30 complex calcifications, XA 2

of 30 complex calcifications, p = 0.004). There was a

trend towards more unfavourable stump morphology (blunt

or absent vs. tapered stump) by coronary CTA (coronary

CTA 17 complex morphologies, XA 11 complex mor-

phologies, p = 0.08). The presence of side branches was

correctly detected in a larger number of patients using

coronary CTA (coronary CTA 12 sidebranches, XA 6

sidebranches, p = 0.04). No statistically relevant differ-

ences were found for distal opacification (good or excellent

vs. poor) (coronary CTA 1 complex morphologies, XA 4

complex morphologies, p = 0.76), distal disease (absent or

moderate vs. severe) (coronary CTA 4 complex morphol-

ogies, XA 1 complex morphology, p = 0.07), presence of

multiple or tandem occlusions (coronary CTA 4 complex

morphologies, XA 4 complex morphologies, p = 1.0),

proximal tortuosity (minimal to moderate vs. severe)

(coronary CTA 2 complex morphologies, XA 7 complex

morphologies, p = 0.2), disease of the proximal segment

(absent or moderate vs. severe) (coronary CTA 9 complex

morphologies, XA 3 complex morphologies, p = 0.401) or

the identification of ostial lesions (coronary CTA 2 com-

plex morphologies, XA 2 complex morphologies,

p = 1.00) between coronary CTA and XA angiograms.

The additional burden of ionizing radiation by coronary

CTA was 10.8 ± 0.4 mSv and 73.2 ± 7.2 ml of contrast

agent had been administered for the coronary CTA.

After propensity matching, 25 patients in each group

remained for further analysis. 5 patients in the CT group

Int J Cardiovasc Imaging

123

were not eligible for matching because they were outside

the caliper of 0.2. Matching variables were well balanced

between the two groups (compare Table 3). The recom-

mended bias B10 % was met in all but two variables. A

significant bias remained with respect to vessel tortuosity,

which was in favour of the coronary CTA group (21.1 %).

Both in direct and in matched comparison, the inter-

ventional success rate was significantly higher in the group

of patients which had undergone pre-interventional coro-

nary CTA [unmatched: CT group 90 % (27/30) vs. no CT

group 63 % (27/43)], p = 0.009; matched: CT group 88 %

(22/25) vs. no CT group 64 % (16/25), p = 0.03]. Multi-

variate analysis on all 73 patients identified prior CT

as only independent predictor of procedural success

(coefficient 1.7 (0.087–3.23) p = 0.039. Four of the 27

successful recanalizations in the coronary CTA group were

achieved using a retrograde approach, while only two ret-

rograde approaches had been made in the group without

coronary CTA (p = 0.39).

The total procedure time was comparable between the

CTA- versus invasive coronary angiography groups

(107.7 ± 53.7 vs. 98.6 ± 35.2, p = 0.419). Similarly,

there were no significant differences in procedural time

among patients with successful versus unsuccessful PCI of

CTO (102.2 ± 43.6 vs. 102.8 ± 45, p = 0.957).

There was a trend to higher procedural doses of inon-

izing radiation in the coronary CTA group as compared to

patients without CT (18.9 ± 0.9 vs. 13.9 ± 0.5 mSv,

p = 0.09). There wasalso a trend towards more extensive

use of procedural contrast in the coronary CTA group

(287.7 ± 22.5 vs. 242.3 ± 13.4 ml, p = 0.08).

Discussion

Several publications have speculated about the effect of

preprocedural coronary CTA on the success rate of CTO

interventions and its role for the future. While some CT

parameters have been shown to predict procedural out-

come, there are some concerns regarding the radiation dose

that comes along with the additional CT imaging study [20,

24, 25]. Hence, it is necessary to clarify whether using CT

prior to attempted interventional CTO recanalization

results in clinical benefits. Kaneda et al. compared success

rates of procedures with and without prior coronary CTA

and found the success rate in the CT group to be higher

than that in the group without prior CT. However they did

not match both cohorts and did not use hypothesis testing.

From the early era of CTO interventions till today

several parameters have been identified as independent

predictors of procedural success by multivariate regression

analysis: occlusion length, calcification of the proximal

cap, stump morphology, tortuosity of the vessel and pres-

ence of a side branch as well as bridging collaterals and a

prior failed attempt [13, 14, 26–29]. All factors have been

accounted for in this analysis apart from bridging collat-

erals and failed attempts. The presence of bridging collat-

erals reflects the duration of the CTO and was therefore

found to be predictive of success. Since the duration of the

CTO did not differ in both groups we omitted this factor.

Previous attempts (predictive of failure) had almost

exclusively been made in the coronary CTA group (10

previous attempts vs. 1), therefore matching based on this

variable was not possible. But since 10 failed previous

attempts in the coronary CTA group were in favour of the

control cohort we thought it justified to perform matching

irrespective of this variable.

Table 2 Baseline characteristics

Number of patients Preprocedural

CT

Without

preprocedural

CT

p value

30 42

Clinical

Age (± SD) 63.3 (12) 62.5 (9.9) 0.75

Sex male 26 (87) 33 (76) 0.29

CCS Class I 6 (20) 3 (7) 0.15

CCS Class II 15 (50) 29 (67)

CCS Class III/IV 9 (30) 11 (26)

Inducible Ischemia 24 (80) 26 (60) 0.12

Viability positive (rest

not tested, echo

normal)

17 (56) 24 (55) 0.94

Risk factors

Smoker 20 (66) 20 (46) 0.09

Diabetes 4 (13) 14 (32) 0.06

Hyperlipemia 18 (60) 29 (67) 0.51

Hypertension 25 (83) 33 (77) 0.49

Body mass index (±SD) 27.6 (4.2) 26.8 (4.4) 0.4

Angiographic/procedural

History of myocardial

infarction month

(±SD)

44.1 (55.4) 39.2 (36.8) 0.78

Prior attempt at CTO 10 (33)** 1 (2)** 0.0001

Retrograde approach 5 (17) 6 (43) 0.75

Number of diseased

vessels

0.82

1 3 (10) 5 (11)

2 14 (46) 15 (34)

3 13 (43) 18 (42)

CTO vessel 0.59

Left anterior

descending artery

6 (20) 5 (11)

Circumflex artery 8 (26) 15 (34)

Right coronary artery 16 (53) 22 (51)

** Significant at 0.001 level

Int J Cardiovasc Imaging

123

Thus both groups had very similar baseline likelihoods

of procedural success and the greater success rate can

confidently be attributed to the additional information

yielded by coronary CTA.

We found, that among these factors calcification, lesion

length and stump morphology as well as presence of

sidebranches were underestimated by invasive angiograms.

These findings add to our hypothesis, that coronary CTA

yields additional information that assists in choosing the

right revascularization strategy and thus can help increas-

ing the success rates.

The most difficult issue in CTO interventions is to track

the right course through the lesion, beginning with the

proximal stump morphology and the crossing of the cap

over navigating the right way through the lesion to pene-

trating the distal cap.

It was possible to visualize the true and complete course

of the occluded vessel in all cases, by aligning the 3D

Dataset orientation to the orientation of the C-arm it was

possible to verify the right direction of the guidewire.

Figure 2 illustrates that the invasive angiogram can be

quite misleading when compared to the VRT image from

coronary CTA.

Interestingly the CT guided procedures tended to be

slightly longer and used higher dosages of the contrast

agent. We can only speculate about the reason but this is

probably due to more frequent rechecking of the guidewire

position in accordance with 3D VRT Dataset. This supports

our theory that the higher success rate is largely due to

better guidewire steering through the lesion.

In the present study we found significantly higher suc-

cess rates within the CT group in two well-matched

cohorts, yielding a success rate of 88 versus 64 %. To our

best knowledge this is the first study to make a direct

comparison between two treatment groups based on pro-

pensity score matching which is considered to be the gold

standard for matched pair analyses. The matching variables

were thoroughly chosen and well balanced between the

groups, which is paramount to this analysis. To avoid a

selection bias, we chose 8 variables for the propensity score

matching.

These findings are in good agreement with results of

Otsuka et al. [17] who also found, that 3D VRT images

helped to find optimal projection angles with least fore-

shortening and additional information about sidebranches

and lesion length. Similar results were found by Yokoyama

et al. [19] who found severely bent CTO segments in 13 %

of patients, which were not readily visible on invasive

angiograms.

Our data suggest, that success rates of PCI for CTO can

be improved by use of preprocedural coronary CTA. This

effect can be attributed to better vessel tracking on the

coronary CTA images within the occluded segment and to

the improved estimation of the complexity criteria.

Study limitations

This study is limited by the relatively small number of

patients, however we were able to perform thorough

Table 3 Covariate balance

Variable Sample Mean %Reduct t test

Treated Control %Bias Bias t p [ t

Visibility Unmatched 0.26667 0.09302 45.7 2.00 0.050

Matched 0.08333 0.125 -11 76 -0.46 0.645

Length (cat) Unmatched 0.5 0.53488 -6.9 -0.29 0.773

Matched 0.5 0.45833 8.2 -19.4 0.28 0.778

Calcification Unmatched 0.4 0.2093 32.3 1.37 0.175

Matched 0.29167 0.29167 0.0 100 0.00 1.000

Tortuosity Unmatched 0.53333 0.48837 7.6 0.31 0.757

Matched 0.54167 0.66667 -21.1 -178.0 -0.71 0.484

Stump morphology Unmatched 0.46667 0.69767 -32.0 -1.32 0.192

Matched 0.58333 0.54167 5.8 82.0 0.21 0.835

Distal opacity Unmatched 0.6 0.81395 -27.6 -1.15 0.256

Matched 0.625 0.66667 -5.4 80.5 -0.19 0.851

Ostial lesion Unmatched 0.06667 0.16279 -30.1 -1.22 0.225

Matched 0.08333 0.08333 0.0 100 0.00 1.000

Sidebranch Unmatched 0.2 0.32558 -28.4 -1.18 0.242

Matched 0.20833 0.20833 0.0 100 0.00 1.000

Int J Cardiovasc Imaging

123

matching of the coronary CTA and control cohort and

found a significant impact of coronary CTA on the success

rate of revascularization.

Conclusion

In this retrosepective propensity score matched analysis we

could demonstrate a significantly higher success rate of

CTO recanalization procedures with additional coronary

CTA. These results warrant a prospective, randomized

study comparing procedures with and without prior coro-

nary CTA to confirm our results.

Conflict of interest None.

References

1. Fefer P, Knudtson ML, Cheema AN, Galbraith PD, Osherov AB,

Yalonetsky S et al (2012) Current perspectives on coronary

chronic total occlusions: the canadian multicenter chronic total

occlusions registry. J Am Coll Cardiol 59(11):991–997

2. Cohen HA, Williams DO, Holmes DR Jr, Selzer F, Kip KE,

Johnston JM et al (2003) Impact of age on procedural and 1-year

outcome in percutaneous transluminal coronary angioplasty: a

report from the NHLBI Dynamic Registry. Am Heart J

146(3):513–519

3. Hoe J (2009) CT coronary angiography of chronic total occlu-

sions of the coronary arteries: how to recognize and evaluate and

usefulness for planning percutaneous coronary interventions. Int J

Cardiovasc Imaging 25(Suppl 1):43–54

4. Stone GW, Kandzari DE, Mehran R, Colombo A, Schwartz RS,

Bailey S et al (2005) Percutaneous recanalization of chronically

occluded coronary arteries: a consensus document: part I. Circ

Cardiovasc Imaging. 112(15):2364–2372

5. DiMario C, Werner G, Sianos G, Galassi A, Buttner J, Dudek D

et al (2007) European perspective in the recanalisation of Chronic

Total Occlusions (CTO): consensus document from the EuroCTO

Club. EuroIntervention 3:30–43

6. Suero JA, Marso SP, Jones PG, Laster SB, Huber KC, Giorgi LV

et al (2001) Procedural outcomes and long-term survival among

patients undergoing percutaneous coronary intervention of a

chronic total occlusion in native coronary arteries: a 20-year

experience. J Am Coll Cardiol 38(2):409–414

7. Olivari Z, Rubartelli P, Piscione F, Ettori F, Fontanelli A, Sa-

lemme L et al (2003) Immediate results and one-year clinical

outcome after percutaneous coronary interventions in chronic

total occlusions: data from a multicenter, prospective, observa-

tional study (TOAST-GISE). J Am Coll Cardiol 41(10):1672–

1678

8. Stone GW, Reifart NJ, Moussa I, Hoye A, Cox DA, Colombo A

et al (2005) Percutaneous recanalization of chronically occluded

coronary arteries: a consensus document: part II. Circ Cardiovasc

Imaging 112(16):2530–2537

9. Mehran R, Claessen BE, Godino C, Dangas GD, Obunai K,

Kanwal S et al (2011) Long-term outcome of percutaneous cor-

onary intervention for chronic total occlusions. JACC Cardiovasc

Interv 4(9):952–961

10. Galassi AR, Tomasello SD, Reifart N, Werner GS, Sianos G,

Bonnier H et al (2011) In-hospital outcomes of percutaneous

coronary intervention in patients with chronic total occlusion:

insights from the ERCTO (European Registry of Chronic Total

Occlusion) registry. EuroIntervention 7(4):472–479

11. Soon KH, Selvanayagam JB, Cox N, Kelly AM, Bell KW, Lim

YL (2007) Percutaneous revascularization of chronic total

occlusions: review of the role of invasive and non-invasive

imaging modalities. Int J Cardiol 116(1):1–6

12. Joyal D, Afilalo J, Rinfret S (2010) Effectiveness of recanaliza-

tion of chronic total occlusions: a systematic review and meta-

analysis. Am Heart J 160(1):179–187

13. Noguchi T, Miyazaki MS, Morii I, Daikoku S, Goto Y, Nonogi H

(2000) Percutaneous transluminal coronary angioplasty of

chronic total occlusions. Determinants of primary success and

long-term clinical outcome. Catheter Cardiovasc Interv

49(3):258–264

14. Tan KH, Sulke N, Taub NA, Watts E, Karani S, Sowton E

(1993) Determinants of success of coronary angioplasty in

patients with a chronic total occlusion: a multiple logistic

regression model to improve selection of patients. Br Heart J

70(2):126–131

15. Di Mario C, Werner GS, Sianos G, Galassi AR, Buttner J, Dudek

D et al (2007) European perspective in the recanalisation of

Chronic Total Occlusions (CTO): consensus document from the

EuroCTO Club. EuroIntervention 3(1):30–43

16. Mollet NR, Hoye A, Lemos PA, Cademartiri F, Sianos G,

McFadden EP et al (2005) Value of preprocedure multislice

computed tomographic coronary angiography to predict the out-

come of percutaneous recanalization of chronic total occlusions.

Am J Cardiol 95(2):240–243

17. Otsuka M, Sugahara S, Umeda K, Nakamura M, Nakamura A,

Bonkohara Y et al (2008) Utility of multislice computed

tomography as a strategic tool for complex percutaneous coro-

nary intervention. Int J Cardiovasc Imaging 24(2):201–210

18. Kaneda H, Saito S, Shiono T, Miyashita Y, Takahashi S, Domae

H (2007) Sixty-four-slice computed tomography-facilitated per-

cutaneous coronary intervention for chronic total occlusion. Int J

Cardiol 115(1):130–132

19. Yokoyama N, Yamamoto Y, Suzuki S, Suzuki M, Konno K,

Kozuma K et al (2006) Impact of 16-slice computed tomography

in percutaneous coronary intervention of chronic total occlusions.

Catheter Cardiovasc Interv 68(1):1–7

20. Garcia–Garcia HM, van Mieghem CA, Gonzalo N, Meijboom

WB, Weustink AC, Onuma Y et al (2009) Computed tomography

in total coronary occlusions (CTTO registry): radiation exposure

and predictors of successful percutaneous intervention. EuroIn-

tervention 4(5):607–616

21. Soon KH, Cox N, Wong A, Chaitowitz I, Macgregor L, Santos

PT et al (2007) CT coronary angiography predicts the outcome of

percutaneous coronary intervention of chronic total occlusion.

J Interv Cardiol 20(5):359–366

22. Austin PC (2008) A critical appraisal of propensity-score

matching in the medical literature between 1996 and 2003. Stat

Med 27(12):2037–2049

23. Leuven E, Sianesi B (2010) PSMATCH2: stata module to per-

form full Mahalanobis and propensity score matching, common

support graphing and covariate imbalance testing. 4.0.4. Nov

2010 ed2003

24. Schwartz RS, Cardiac CTA (2008) three-dimensions, and the

chronic total occlusion: a window to the future. Catheter Car-

diovasc Interv 71(6):790–791

25. Hecht HS (2008) Applications of multislice coronary computed

tomographic angiography to percutaneous coronary intervention:

how did we ever do without it? Catheter Cardiovasc Interv

71(4):490–503

26. Puma JA, Sketch MH Jr, Tcheng JE, Harrington RA, Phillips HR,

Stack RS et al (1995) Percutaneous revascularization of chronic

coronary occlusions: an overview. J Am Coll Cardiol 26(1):1–11

Int J Cardiovasc Imaging

123

27. Dong S, Smorgick Y, Nahir M, Lotan C, Mosseri M, Nassar H

et al (2005) Predictors for successful angioplasty of chronic

totally occluded coronary arteries. J Interv Cardiol 18(1):1–7

28. Morino Y, Abe M, Morimoto T, Kimura T, Hayashi Y, Mura-

matsu T et al (2011) Predicting successful guidewire crossing

through chronic total occlusion of native coronary lesions within

30 minutes: the J-CTO (Multicenter CTO Registry in Japan)

score as a difficulty grading and time assessment tool. JACC

Cardiovasc Interv 4(2):213–221

29. Tomasello SD, Costanzo L, Campisano MB, Barrano G, Ca-

podanno D, Tamburino C et al (2011) Does occlusion duration

influence procedural and clinical outcome of patients who

underwent percutaneous coronary intervention for chronic total

occlusion? J Interv Cardiol 24(3):223–231

Int J Cardiovasc Imaging

123