OUP user menu

Comparison of sirolimus-eluting and bare metal stent for treatment of patients with total coronary occlusions: results of the GISSOC II-GISE multicentre randomized trial

Paolo Rubartelli, Anna Sonia Petronio, Vincenzo Guiducci, Paolo Sganzerla, Leonardo Bolognese, Mario Galli, Imad Sheiban, Fabio Chirillo, Angelo Ramondo, Sandro Bellotti
DOI: http://dx.doi.org/10.1093/eurheartj/ehq199 2014-2020 First published online: 20 June 2010


Aims Percutaneous coronary intervention with bare metal stent (BMS) in chronic total coronary occlusions (CTOs) is associated with a higher rate of angiographic restenosis and reocclusion than that observed in subtotal stenoses. Preliminary reports have suggested a better performance of drug-eluting stents in CTO. In this multicentre, randomized trial, we compared the mid-term angiographic and clinical outcome of sirolimus-eluting stent (SES) or BMS implantation after successful recanalization of CTO.

Methods and results Patients with CTO older than 1 month, after successful recanalization, were randomized to implantation of SES (78 patients) or BMS (74 patients) in 13 Italian centres. Clopidogrel therapy was prescribed for 6 months. The primary endpoint was in-segment minimal luminal diameter (MLD) at 8-month follow-up. Secondary clinical endpoints included death, myocardial infarction (MI), target lesion revascularization (TLR), and target vessel revascularization (TVR) at 24 months. Patients treated with SES showed, at in-segment analysis, a larger MLD (1.98 ± 0.57 vs. 0.98 ± 0.80 mm, P < 0.001), a lower late luminal loss (−0.06 ± 0.49 vs. 1.11 ± 0.79 mm, P < 0.001), and lower restenosis (9.8 vs. 67.7%, P < 0.001) and reocclusion (0 vs. 17%, P = 0.001) rates. At 24-month follow-up, patients in the SES group experienced fewer major adverse cardiac events (50.0 vs. 17.6%, P < 0.001) mainly due to a lower rate of both TLR (44.9 vs. 8.1%, P < 0.001) and TVR (44.9 vs. 14.9%, P < 0.001).

Conclusion In CTO, SES is markedly superior to BMS in terms of restenosis and reocclusion rate, and incidence of repeat revascularization at 24 months.

Clinicaltrials.gov identifier: NCT00220558

  • Coronary artery disease
  • Percutaneous transluminal coronary angioplasty
  • Stent
  • Total coronary occlusion


Percutaneous coronary intervention is an established treatment for selected patients with chronic total coronary occlusions (CTOs), although the long-term results are limited by a higher restenosis rate than that observed in subtotal stenosis. At present, due to the availability of dedicated equipment and increased operator experience, the acute success rate of percutaneous intervention in CTOs has improved.1 However, a more aggressive approach,2 while increasing the success rate, may provoke long dissections and require implantation of long or multiple stents, with an anticipated increase in restenosis risk.

Several studies, mostly observational, have evaluated drug-eluting stents (DES) implantation in CTOs310 with favourable results, reporting restenosis rates varying between 08 and 19.2%7 and total reocclusion rates up to 11.5%.7 However, even using DES, CTOs present a higher restenosis rate than subtotal lesions.11 In addition, to date, no DES has an unlimited worldwide license for use in CTO despite the practice being widespread.

Given the paucity of dedicated randomized data, we thus designed in 2004 a prospective, multicentre, randomized study— Gruppo Italiano di Studio sullo Stent nelle Occlusioni Coronariche —supported by GISE (Società Italiana di Cardiologia Invasiva), or GISSOC II-GISE, in order to compare the angiographic and clinical outcome after implantation of bare metal stent (BMS) or sirolimus-eluting stent (SES) in CTOs.


The GISSOC-2 trial was performed in 13 Italian centres in order to compare a SES (Cypher, Cordis, Miami, FL, USA) with the BMS with similar design (Bx Sonic, Cordis) in CTOs, which were defined as total occlusions with no visible luminal continuity of at least 30 days of duration. The duration of the occlusion was estimated from previous angiographic data, from the date of a myocardial infarction (MI) in the distribution of the occluded vessel, or from beginning or worsening of anginal symptoms.

The study was conducted according to the principles of the Declaration of Helsinki, approved by the institutional review board at each participating centre, and registered at clinicaltrials.gov (NCT00220558) . All patients gave written informed consent.

Patient selection

Patients older than 18 years, with a de novo CTO on a native coronary artery, with an estimated vessel size between 2.75 and 3.75 mm, and with a clinical presentation including angina, silent ischaemia, or myocardial viability, were considered for the study. The occlusion had to be dilated by the balloon and the lesion had to be fully coverable, according to the operator judgment, by ≤2 stent of ≤33 mm in length.

Main exclusion criteria included inability to give informed consent, MI within 30 days in the territory of the target CTO or within 3 days in another territory, left ventricular ejection fraction <30%, renal failure with serum creatinine >3 mg/dL, other comorbid condition with life expectancy <2 years, contraindications to aspirin or clopidogrel therapy, women with child bearing potential, left main disease, and CTO involving a bifurcation with side branch larger than 2 mm.


Eligible patients were randomly assigned in a 1:1 ratio to treatment with SES or BMS by means of sealed envelopes. All patients received aspirin 100–300 mg/day and clopidogrel 75 mg/day. A loading dose of 300 mg of clopidogrel was administered the day before the procedure in patients who were not pre-treated. Clopidogrel discontinuation was recommended 6 months after the index procedure in all patients, while aspirin was prescribed indefinitely. During the procedure, unfractionated heparin was given to maintain the activated clotting time >250 s. Creatinine kinase and MB isoenzyme levels were measured before, and 8 and 24 h after the procedure.

Clinical and angiographic follow-up

A clinical evaluation was scheduled at 1, 8, 12, and 24 months. Coronary angiography was repeated at 8 months or earlier when clinically indicated in all consenting patients. An unscheduled angiogram obtained any time before 6 months substituted for the follow-up angiogram only if an in-segment diameter stenosis ≥50% by quantitative coronary angiography was found, otherwise the patients were required to undergo a repeat angiography at 8 months.

Quantitative coronary angiography

Quantitative coronary angiography was performed off-line at an independent core laboratory (Mediolanum Cardio Research, Milan, Italy) by experienced personnel who was unaware of the type of the stent implanted. The projection that best showed the stenosis severity was selected. Matched projections were obtained at baseline, at the end of the procedure, and at the 8-month follow-up after intracoronary injection of nitroglycerin and were analysed using a validated edge-detection software (CMS Medis Medical Imaging Systems, Leiden, the Netherlands). The tip of the coronary catheter was used for calibration. Quantitative parameters included interpolated reference diameter, minimal luminal diameter (MLD), diameter stenosis, and late luminal loss.

On the pre-intervention angiogram, reference vessel diameter and lesion length were measured. The CTO length was assessed either using the visualization of the distal vessel by collaterals or after balloon pre-dilatation. In the post-procedural and follow-up angiograms, both in-stent and in-segment analyses were performed. In-segment analysis included both the proximal and distal edges of the stent(s), each 5 mm long.

Endpoints and definitions

The primary endpoint of the study was the in-segment MLD at 8 months.8,12 Secondary angiographic endpoints were in-segment late luminal loss and restenosis rate, the latter defined as the rate of patients showing an in-segment diameter stenosis greater ≥50% at follow-up.

Secondary clinical endpoints were major adverse cardiac event rate at 24 months, defined as death, MI (with or without ST elevation), emergent bypass surgery, or repeat target lesion revascularization (TLR) or target vessel revascularization (TVR). Target vessel revascularization was defined clinically driven when core laboratory quantitative coronary angiography identified a >70% stenosis in the target vessel, or alternatively a >50% stenosis associated with angina or evidence of myocardial ischaemia.13 Myocardial infarction related to a revascularization procedure was defined as elevation of the creatine kinase-MB greater than three times the upper normal limit. Myocardial infarction during follow-up was defined as elevation of the creatine kinase-MB greater than two times the upper normal limit with either ischaemic symptoms or ischaemic electrocardiographic changes. In addition, MI was defined as ST elevation when ≥1 mm ST elevation was recorded in two contiguous electrocardiographic leads. Stent thrombosis was defined as an acute coronary syndrome with angiographic documentation of either occlusion of the target lesion or thrombus within the previously stented segment. Stent thrombosis was also adjudicated using the Academic Research Consortium definitions13 that was not available at the time of conception of the study.

Data management

Independent study monitors verified all study data at the participating centres. All major adverse cardiac events and stent thrombosis were adjudicated by an independent committee blinded to treatment allocation after review of original source documentation.

Statistical analysis

Follow-up MLD after BMS implantation in CTOs was preliminarily estimated to be 1.7 ± 0.8 mm.12,1417 Assuming that the in-segment MLD at follow-up after SES implantation in CTOs should be 2.18 mm,4,6,9 a sample size of 120 patients (60 in each arm) will have the power of 0.80 to detect a difference between the two group with a two-tailed alpha error of 0.05. The sample size was prudentially increased to 150 patients to compensate for losses at follow-up.

Categorical variables are expressed as absolute numbers and percent value. Continuous variables are expressed as mean value ± standard deviation or as median and interquartile range for non-Gaussian distributed data. Differences between groups for categorical variables were assessed with the Fisher's exact test. Differences between groups for continuous variables were evaluated with the two-tailed unpaired Student's t-test or the Mann–Whitney U test for non-Gaussian distributed data. Event-free survival rates were estimated using the Kaplan–Meier method, with differences between the two treatment groups assessed with the use of the log-rank test of significance. A two-sided P-value < 0.05 was considered significant. All evaluations were performed according to the intention to treat principle. All analyses were performed using BMDP 2009 Statistical Software (Statistical Solutions Ltd, Cork, Ireland).


From May 2005 to September 2007, 152 patients were enrolled in the study: 78 were randomized to BMS implantation and 74 assigned to SES treatment, with no violation of allocation (Figure 1). As shown in Table 1, no significant difference was observed in the baseline clinical characteristics of the two patient groups. The angiographic and procedural variables (Table 2) were comparable in both groups, except for a slightly larger maximal balloon diameter in the BMS group (P = 0.025). Also, patients treated with BMS showed a slightly larger postprocedural MLD (Table 3, P = 0.023). These differences are possibly related to the unblinded design of the study. One patient treated with SES was lost to follow-up.

View this table:
Table 1

Baseline clinical characteristics

BMS (n = 78)SES (n = 74)
Age (years)63.9 ± 9.863.9 ± 9.6
Male68 (87.1)58 (78.3)
Diabetes mellitus15 (19.2)19 (25.7)
Hypertension51 (65.4)49 (66.2)
Smoking41 (52.6)45 (60.8)
Hypercholesterolaemia59 (75.6)54 (73.0)
Prior myocardial infarction24 (30.8)24 (32.4)
Prior coronary bypass surgery4 (5.1)5 (6.7)
Prior percutaneous revascularization15 (19.2)11 (14.9)
 None24 (30.8)18 (24.3)
 Class I or II30 (38.5)25 (33.8)
 Class III or IV9 (11.5)14 (18.9)
 Unstable angina15 (19.2)17 (23.0)
  • Values expressed as number (%) or mean ± SD. Angina severity according to Canadian Cardiovascular Society classification. All differences between the study groups are not significant.

  • BMS, bare metal stent; SES, sirolimus-eluting stent.

View this table:
Table 2

Angiographic and procedural variables

BMS (n = 78)SES (n = 74)
Ejection fraction (%)55.4 ± 11.355.1 ± 8.8
One vessel disease25 (32.0)26 (35.1)
Two vessels disease28 (35.9)21 (28.4)
Three vessels disease25 (32.0)27 (36.5)
Target vessel
 Left anterior descending20 (25.6)24 (32.4)
 Left circumflex21 (26.9)15 (20.3)
 Right coronary37 (47.4)35 (47.3)
Duration of the occlusion
 1–3 months4 (5.1)6 (8.1)
 >3 months23 (29.5)28 (37.8)
 Not determinable51 (65.4)40 (54.1)
Moderate/severe calcification15 (19.2)8 (10.8)
Baseline TIMI flow grade 061 (78.2)52 (70.2)
Baseline TIMI flow grade 115 (19.2)20 (27.0)
No. of coronary guidewires/occlusion1.68 ± 0.901.67 ± 0.75
Use of specialized stiff or hydrophilic guidewires69 (88.5)67 (90.5)
Contralateral injection9 (11.5)9 (12.2)
No. of stents on the CTO1.55 ± 0.701.64 ± 0.75
Total stent length on the CTO (mm)37.5 ± 16.640.5 ± 18.0
Max balloon diameter (mm)3.02 ± 0.282.92 ± 0.27*
Max inflation pressure (atm)14.3 ± 2.714.8 ± 2.5
  • Values expressed as number (%) or mean ± SD.

  • BMS, bare metal stent; SES, sirolimus-eluting stent; CTO, chronic total coronary occlusion.

  • *P = 0.025, other differences between the study groups are not significant.

View this table:
Table 3

Quantitative coronary angiography parameters

BMS (n = 78)SES (n = 74)P-value
 Proximal vessel diameter (mm)2.65 ± 0.382.62 ± 0.310.582
 Occlusion length (mm)15.5 ± 8.415.3 ± 7.30.880
After stenting
 Reference vessel diameter (mm)2.91 ± 0.342.80 ± 0.450.080
 MLD (mm)
  Proximal edge2.77 ± 0.532.73 ± 0.580.607
  In-stent2.59 ± 0.352.46 ± 0.350.023
  Distal edge2.08 ± 0.451.95 ± 0.560.116
  In-segment2.13 ± 0.461.94 ± 0.500.015
 Diameter stenosis (% of vessel diameter)
  Proximal edge13.7 ± 12.713.5 ± 14.40.915
  In-stent14.2 ± 8.116.2 ± 7.80.139
  Distal edge22.8 ± 12.926.6 ± 14.20.086
  In-segment26.9 ± 12.230.8 ± 12.60.055
 Reference vessel diameter (mm)2.79 ± 0.492.86 ± 0.480.360
 MLD (mm)
  Proximal edge2.64 ± 0.742.65 ± 0.600.893
2.71 (2.24–3.12)2.69 (2.32–3.03)
  In-stent1.01 ± 0.842.26 ± 0.61<0.001
0.77 (0.26–1.78)2.37 (1.93–2.72)
  Distal edge1.75 ± 0.932.21 ± 0.470.012
2.06 (1.22–2.39)2.16 (1.88–2.55)
  In-segment0.98 ± 0.801.98 ± 0.57<0.001
0.77 (0.26–1.62)1.97 (1.69–2.31)
 Diameter stenosis (% of vessel diameter)
  Proximal edge15.3 ± 17.614.7 ± 15.90.839
10.2 (3.0–21.6)9.6 (3.0–23.3)
  In-stent65.4 ± 28.022.4 ± 17.6<0.001
71.6 (39.1–91.1)17.8 (10.6–27.8)
  Distal edge30.9 ± 33.316.3 ± 12.50.040
19.0 (7.2–40.0)13.4 (8.1–25.3)
  In-segment66.4 ± 26.831.0 ± 17.2<0.001
71.6 (42.3–91.1)26.7 (18.0–40.3)
 Late luminal loss (mm)
  Proximal edge0.11 ± 0.590.03 ± 0.470.499
−0.01 (−0.28–0.39)−0.09 (−0.24–0.20)
  In-stent1.57 ± 0.850.20 ± 0.49<0.001
1.66 (0.85–2.27)0.07 (−0.11–0.45)
  Distal edge0.29 ± 0.83−0.29 ± 0.43<0.001
0.19 (−0.34–0.71)−0.31 (−0.52–0.06)
  In-segment1.11 ± 0.79−0.06 ± 0.49<0.001
1.10 (0.50–1.65)−0.10 (−0.35–0.16)
 Binary restenosis
  Proximal edge4 (6.15)1 (1.64)0.186
  In-stent44 (67.7)5 (8.2)<0.001
  Distal edge12 (18.5)1 (1.64)0.001
  In-segment44 (67.7)6 (9.8)<0.001
 Total reocclusion11 (16.9%)0 (0%)0.001
 Restenosis pattern
  Focal14 (21.5)6 (9.8)0.072
  Diffuse19 (29.2)0<0.001
  • Values expressed as number (%), mean ± SD or median (IQR). The follow-up measurements refer to 66 patients in the BMS group and 62 patients in the SES group, because 12 patients in each group did not undergo the angiographic follow-up.

  • IQR, interquartile range; MLD, minimal luminal diameter; BMS, bare metal stent; SES, sirolimus-eluting stent.

Figure 1

Patient-flow diagram of the GISSOC-II GISE Study. QCA, quantitative coronary angiography.

Angiographic follow-up was obtained in 128 patients (84%) 8.4 ± 1.3 months after the procedure. Twelve patients in the BMS group and 11 patients in the SES group, all asymptomatic, refused the repeat angiography. All the baseline, procedural, angiographic, and clinical data were re-analysed excluding patients without angiographic follow-up, but the differences between the study groups remained substantially unchanged. These results are reported in the supplementary material online, Tables S1–S4.

As shown in Table 3 and Figure 2, patients in the BMS group showed, at follow-up, a smaller MLD, and larger late luminal loss and binary restenosis rate at both in-stent and in-segment analyses. In addition, both diffuse and occlusive patterns of restenosis were significantly higher in the BMS group. No stent fracture was identified in either group. No coronary artery aneurysm developed in the target vessel.

Figure 2

Cumulative rate of in-segment minimal luminal diameter.

The clinical events are displayed in Table 4, while Figure 3 shows the freedom from major cardiac adverse event. Two sudden cardiac deaths occurred, one in the BMS group and the other in the SES group, respectively, 23 months and 11 months after the procedure. One patient in the SES group died from metastatic cancer 22 months after the procedure. In the BMS group, four periprocedural non-ST elevation MIs were observed. One ST elevation MI in the SES group was caused by a subacute stent thrombosis occurring 4 days after the procedure. One non-ST elevation MI in the SES group was related to a non-target vessel PCI occurring during follow-up. The major adverse cardiac event rate was significantly higher in the BMS group: this difference was mainly due to a higher rate of TLR and TVR. Moreover, 80% of the TVR was clinically driven. Three patients underwent elective coronary bypass surgery, two in the BMS group and one in the SES group, all determining TLR. Only two TVR (one in each group) occurred after 12 months. Overall, two stent thromboses were observed, one acute (BMS group) and one subacute (SES group). They were both definite according to the ARC definition.13 No event qualifying as probable stent thrombosis was observed, while the two patients with sudden deaths (one in each group) could meet the ARC criteria for late or very late possible stent thrombosis.13 Lastly, angina or inducible silent ischaemia occurring during follow-up was significantly more frequent in the patients assigned to BMS.

View this table:
Table 4

Clinical events and symptomatic status at 24 months

BMS (n = 78)SES (n = 74)aP-value
Death, all causes1 (1.3)2 (2.7)0.613
Death, cardiac1 (1.3)1 (1.4)1.0
Myocardial infarction4 (5.1)2 (2.7)0.682
Coronary bypass surgery2 (2.6)1 (1.4)1.0
TLR35 (44.9)6 (8.1)<0.001
TVR35 (44.9)11 (14.9)<0.001
Clinically driven TVR29 (37.2)8 (10.8)<0.001
Stent thrombosis1 (1.3)1 (1.4)1.0
 Acute1 (1.3)01.0
 Subacute01 (1.4)0.483
 Late or very late00
Stroke1 (1.3)01.0
Any major adverse cardiac events39 (50.0)13 (17.6)<0.001
Angina or silent ischaemia during follow-up25 (32.1)11 (15.1)0.014
  • Values expressed as number (%). In-hospital events are included. Four patients had more than one event. In the BMS group, one patient had an acute thrombosis and a TLR, and another patient had a MI and a TLR. In the SES group, one patient had a subacute thrombosis, a MI and a TLR, and another patient had a MI and a TVR).

  • BMS, bare metal stent; SES, sirolimus-eluting stent; TLR, target lesion revascularization; TVR, target vessel revascularization.

  • aOne patient in the SES group was lost to follow-up: therefore, post-discharge events and symptomatic status in the SES group refer to 73 patients.

Figure 3

Freedom from major adverse cardiac events (death, MI, emergent bypass surgery, TLR, or TVR).


The present study shows that implantation of SES in CTOs confers mid-term angiographic and clinical results much more favourable than those obtained using a BMS of similar design. At 8-month follow-up, the patients treated with SES presented a two-fold increase of in-segment MLD, which was the primary endpoint. In the SES group, the in-segment late luminal loss was slightly negative, when compared with an in-segment late luminal loss of 1.11 mm in the BMS arm. Likewise, SES use resulted in an 85% relative risk reduction of in-segment restenosis (9.8 vs. 67.7%), In addition, restenoses were all focal in the SES group, but predominantly diffuse or occlusive in the BMS group. In particular, the total reocclusion rate was observed in 16.9% of patients in the BMS group and in none of the SES treated patients (P < 0.001). It should be noted that a total occlusion found at angiographic follow-up could result either from asymptomatic stent thrombosis or from restenosis progressing to 100% occlusion. Indeed, the latter appears to be the most probable cause, being much more frequent in the BMS group.

This better angiographic outcome translated into a reduced occurrence of TLR and TVR together with a significantly lower incidence of angina or inducible silent ischaemia in the SES group at 24-month follow-up.

The adverse clinical events other than repeat revascularization and the stent thromboses were low in both group, supporting the safety of SES in CTO. However, the trial was not designed or powered to assess the incidence of such relatively rare events nor the possible difference of them between the study groups.

Drug-eluting stents and bare metal stent in chronic total coronary occlusions

Since market approval, the benefit of DES on restenosis reduction in simple coronary lesions has prompted their use in increasingly complex lesions that were not tested in the pivotal trials. In particular, CTOs were excluded or were marginally represented in the main trials comparing DES to BMS.

Drug-eluting stents implantation in CTOs has been described in several observational, often retrospective, studies,37,10 in a post hoc subgroup analysis of a trial including various types of complex lesions,8 and in a single prospective randomized trial.9 In these studies, the 6-month recurrence rates are generally low, with a restenosis rate varying between 08 and 19.2%,7 and a total reocclusion rate varying from 03,8 to 11.5%.7

However, these reports are not fully comparable because of several methodological differences, regarding inclusion criteria, definitions of angiographic parameters, concomitant therapy, angiographic follow-up, single centre vs. multicentre design, and independent assessment of angiographic parameters and clinical events. On the other hand, the present study is comparable to the only published randomized trial of DES vs. BMS in CTOs, the Primary Stenting of Totally Occluded Native Coronary Arteries II (PRISON II),9 although a few differences regarding the study design and the patients characteristics should be noted.

The duration of the double antiplatelet therapy was 6 months in both studies: however, in the PRISON II study, the follow-up angiography was scheduled at 6 months, while in the GISSOC II study the angiography was performed at 8 months, 2 months after stopping clopidogrel. This timing could have favoured the detection of possible late stent thromboses associated to discontinuation of such therapy. When compared with PRISON II, GISSOC II patients presented markers of a more extensive coronary artery disease that may denote a higher risk of restenosis. In fact, they were about 4 years older, with a higher prevalence of diabetes mellitus (22 vs. 13%), and three-vessel disease (34 vs. 9%).

Moreover, the total stent length was longer in GISSOC II patients (39 vs. 30 mm) with a smaller stent diameter (3.0 vs. 3.25 mm). These differences may explain the higher binary restenosis rate (68 vs. 41%) and late luminal loss (1.11 vs. 0.64 mm) observed in the BMS arm of the GISSOC II when compared with the PRISON II.

On the other hand, focusing on the SES arm, the restenosis rate is very similar in GISSOC II and PRISON II (9.9 and 11%, respectively) with an almost identical late luminal loss (−0.06 and −0.07 mm, respectively).

It is noteworthy that, after BMS implantation in CTOs, the restenosis seems to be highly variable but consistent with the risk associated to baseline and procedural characteristics. On the contrary, after SES implantation in CTOs, the restenosis seems to be not only much lower, but also less variable. In other words, SES seems to be less influenced than BMS by patients and lesion factors known to increase the restenosis risk, such as diabetes, lesion length, and vessel size.

Duration of the chronic total coronary occlusion

According to the current definition,1,18 a coronary occlusion is defined as chronic when older than 3 months. Although in the present study, coronary occlusions older than 1 month could be included, the occlusions with the duration between 1 and 3 months were only 6.6%.

In most CTOs (60%), the duration could not be determined, but there was no clinical or angiographic reason to estimate them as more recent than 3 months. Then, these CTOs have a high likelihood to be older than 3 months.18 In addition, according to the available evidence, the duration of the occlusion does not appear to influence the restenosis risk.9

Study limitations

The sample size may be an important limitation, because the study did not have the power to detect potential differences between the two study groups with regard to infrequent clinical events, such as death, MI, and stent thrombosis. However, the study size was adequate to assess differences of the primary endpoint between the study groups.

The unblinded design of the study is another relevant limitation that may have driven a few differences in procedural variables, such as slight but significantly larger maximal balloon diameter and post-stenting MLD in the BMS group, when compared with the SES group. However, both quantitative coronary angiography and adjudication of all major adverse cardiac events and stent thrombosis were carried out by independent personnel blinded to treatment allocation.

Twelve patients in both groups (16%) did not undergo the follow-up angiography, preventing the assessment of the primary endpoint. However, the equivalence of the rate of control angiography in both study groups, and the marked difference in the study endpoints between the two groups, makes very unlikely a biasing effect from incomplete angiographic follow-up on the main results of the study.

Although the patients of the BMS group showed a significantly higher incidence of angina or inducible silent ischaemia at follow-up, we acknowledge that the difference in repeat revascularization was probably amplified by the protocol mandated angiography at 8 months.


Implantation of a SES after successful PTCA of a CTO results in a larger MLD at follow-up and decreases the restenosis and reocclusion rates. Accordingly, patients treated with a SES experience a significant reduction of repeat revascularization procedures. Thus, SES implantation could be the preferred treatment option in patients with CTO that can be recanalized percutaneously.


This work was supported by an unrestricted grant from Cordis Italia, Sesto San Giovanni, Italy.

Conflict of interest: none declared.


We thank the participating investigators to the GISSOC II–GISE Trial (see Supplementary material online).


View Abstract