Comparison of Long-Term Clinical and Angiographic Outcomes Following Implantation of Bare Metal Stents and Drug-Eluting Stents in Aorto-Ostial Lesions

Comparison of Long-Term Clinical and Angiographic Outcomes Following Implantation of Bare Metal Stents and Drug-Eluting Stents in Aorto-Ostial Lesions

Comparison of Long-Term Clinical and Angiographic Outcomes Following Implantation of Bare Metal Stents and Drug-Eluting Stents in Aorto-Ostial Lesions...

200KB Sizes 2 Downloads 39 Views

Comparison of Long-Term Clinical and Angiographic Outcomes Following Implantation of Bare Metal Stents and Drug-Eluting Stents in Aorto-Ostial Lesions Rasha Al-Lamee, MDa,b,c, Alfonso Ielasi, MDa, Azeem Latib, MDa,b, Cosmo Godino, MDa,b, Marco Mussardo, MDa, Francesco Arioli, MDa, Filippo Figina, Daniela Pirainoa, Mauro Carlino, MDa, Matteo Montorfano, MDa, Alaide Chieffo, MDa, and Antonio Colombo, MDa,b,* Percutaneous coronary intervention (PCI) to aorto-ostial (AO) lesions is technically demanding and associated with high revascularization rates. The aim of this study was to assess outcomes after bare metal stent (BMS) compared to drug-eluting stent (DES) implantation after PCI to AO lesions. A retrospective cohort analysis was conducted of all consecutive patients who underwent PCI to AO lesions at 2 centers. Angiographic and clinical outcomes in 230 patients with DES from September 2000 to December 2009 were compared to a historical control group of 116 patients with BMS. Comparison of the baseline demographics showed more diabetics (32% vs 16%, p ⴝ 0.001), lower ejection fractions (52.3 ⴞ 9.7% vs 55.0 ⴞ 11.5%, p ⴝ 0.022), longer stents (17.55 ⴞ 7.76 vs 14.37 ⴞ 5.60 mm, p <0.001), and smaller final stent minimum luminal diameters (3.43 ⴞ 0.53 vs 3.66 ⴞ 0.63 mm, p ⴝ 0.001) in the DES versus BMS group. Angiographic follow-up (DES 68%, BMS 66%) showed lower restenosis rates with DES (20% vs 47%, p <0.001). At clinical follow-up, target lesion revascularization rates were lowest with DES (12% vs 27%, p ⴝ 0.001). Cox regression analysis with propensity score adjustment for baseline differences suggested that DES were associated with a reduction in target lesion revascularization (hazard ratios 0.28, 95% confidence interval 0.15 to 0.52, p <0.001) and major adverse cardiac events (hazard ratio 0.50, 95% confidence interval 0.32 to 0.79, p ⴝ 0.003). There was a nonsignificantly higher incidence of Academic Research Consortium definite and probable stent thrombosis with DES (n ⴝ 9 [4%] vs n ⴝ 1 [1%], p ⴝ 0.131). In conclusion, despite differences in baseline characteristics favoring the BMS group, PCI with DES in AO lesions was associated with improved outcomes, with lower restenosis, revascularization, and major adverse cardiac event rates. © 2011 Elsevier Inc. All rights reserved. (Am J Cardiol 2011;108:1055–1060) Various techniques have been tried in an attempt to improve angiographic and clinical outcomes in aorto-ostial (AO) lesions, such as the use of laser-based systems,1 directional atherectomy,2 cutting balloon predilatation,3 and implantation of polytetrafluoroethylene stents.4 The introduction of drug-eluting stents (DES) led to improved revascularization rates compared to bare-metal stent (BMS) implantation in AO lesions, but initial studies were relatively small and were performed before the availability of large DES with diameters ⬎3.5 mm, leading to a possible discrepancy between the stent and reference vessel sizes.5–7 The aims of this study were to assess long-term angiographic and clinical outcomes after DES implantation in AO lesions in the contemporary percutaneous coronary inter-

a

Interventional Cardiology Unit, San Raffaele Scientific Institute; bInterventional Cardiology Unit, EMO-GVM Centro Cuore Columbus, Milan, Italy; and cImperial College Healthcare NHS Trust, London, United Kingdom. Manuscript received March 28, 2011; revised manuscript received and accepted June 6, 2011. *Corresponding author: Tel: 39-0248712422; fax: 39-0248193433. E-mail address: [email protected] (A. Colombo). 0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2011.06.004

vention (PCI) era and to compare these results to a historical control group of BMS-treated lesions. Methods We analyzed data that had been prospectively collected after PCI at 2 institutions, San Raffaele Scientific Institute and EMO Centro Cuore Columbus Hospital (Milan, Italy). In total, 230 consecutive unselected patients with de novo AO disease treated exclusively with DES implantation from September 2002 to December 2009 were enrolled (Figure 1). Long-term outcomes after PCI to AO lesions with DES were compared to a historical control cohort of 116 consecutive unselected patients treated with BMS before the availability of DES. Exclusion criteria were PCI for ST-segment elevation myocardial infarction, PCI for restenosis, and PCI to an AO lesion as part of “full metal jacket” stent implantation. An AO lesion was defined as a lesion involving the junction between the aorta and the orifice of the left main stem, right coronary artery, or a saphenous vein graft within 3 mm of the ostium of the vessel. Angiographic success was defined as Thrombolysis In Myocardial Infarction (TIMI) www.ajconline.org

1056

The American Journal of Cardiology (www.ajconline.org)

Aorto-ostial lesions (n=628) Excluded Aorto-ostial lesions (n=464) Excluded

Full-metal jacket with aorto-ostial involvement, aorto-ostial segment was not index lesion (n=164) Restenotic lesions (n=84)

Aorto-ostial lesions (n=380) Excluded

POBA (n=10) PTFE stent (n=13)

De novo aorto-ostial lesions (n=357)

Figure 1. Study flowchart. POBA ⫽ percutaneous angioplasty; PTFE ⫽ polytetrafluoroethylene.

grade 3 flow with ⬍30% residual stenosis at the end of the procedure. Periprocedural myocardial infarction was defined as a 3-fold increase in creatine kinase-MB. Angiographic restenosis was defined as diameter stenosis ⱖ50% by quantitative coronary angiography within a previously stented segment or within 5 mm of the distal or proximal edge of the stent on follow-up angiography. The Mehran classification system was used to define the pattern of restenosis.8 Target lesion revascularization (TLR) was defined as the need for any repeat revascularization, either surgical or percutaneous, for a stenosis within the stent or within the 5-mm borders adjacent to the stent. Target vessel revascularization (TVR) was defined as the need for any repeat revascularization on a treated vessel. Major adverse cardiac events (MACE) were defined as a combination of all-cause mortality, nonfatal myocardial infarction, TLR, TVR, and need for coronary artery bypass grafting. Stent thrombosis was defined on the basis of the Academic Research Consortium definitions and cumulative stent thrombosis as a combination of all episodes of stent thrombosis during follow-up. All patients were treated with aspirin and a loading dose of a thienopyridine before the procedure. During the procedure, all patients were then treated with intravenous heparin with an initial 100 U/kg bolus, followed by further heparin as required to achieve a target activated clotting time ⱖ250 seconds. Glycoprotein IIb/IIIa inhibitors were given at the operator’s discretion. Coronary intervention was then performed as usual with predilatation and stent implantation using standard techniques through the femoral artery, as previously described.9 Although the final interventional strategy was left to the discretion of the primary operator, most stents were implanted to protrude 1 mm into the aorta and were postdilated to high pressure with compliant or noncompliant balloon inflation. Postoperatively, all patients received aspirin unless there was a specific contraindication, and those patients receiving intracoronary stents received dual-antiplatelet therapy with aspirin and a thienopyridine as determined by contemporary guidelines. After the cessation of thienopyridine therapy, all patients continued to take aspirin indefinitely. Cine angiograms were assessed using a validated edge detection system (CMS version 5.2; Medis Medical Imag-

ing, Leiden, The Netherlands). All coronary angiograms were assessed by 2 experienced angiographers who were not involved in the stenting procedure and were blinded to the type of stent implanted. Angiographic analysis was performed using standard quantitative and qualitative methods, as previously described.10 Paired cine frames of 2 orthogonal frames in the end-systolic frames showing the stenosis in its most severe projection were selected. The minimum luminal diameter, reference vessel diameter, and percentage stenosis before and after coronary intervention and, when available, at follow-up were calculated. In addition, lesion length was analyzed before PCI. Angiographic restenosis was defined as ⬎50% diameter stenosis by quantitative coronary angiography within a previously stented segment. Late luminal loss was calculated as the difference between minimum luminal diameter at the end of the procedure and the minimum luminal diameter on the initial angiogram at follow-up. As is our regular clinical practice, follow-up was obtained at regular intervals with clinic visits or telephone interviews at 3, 6, and 12 months in the first year and at intervals of 6 months thereafter. Additional data were obtained from primary care physicians, referring cardiologists, or relatives as necessary. All repeat interventions and complications were prospectively entered into a dedicated database. All adverse events were clarified by reviewing the medical notes or contacting the patient’s physician. Clinical follow-up was obtained in 94% (n ⫽ 326). Our regular clinical practice is to follow up all high-risk PCI procedures with angiographic surveillance when possible at 6 to 9 months after the index procedure. Angiographic follow-up was obtained in 67% of lesions (n ⫽ 227). Functional test results were negative in those patients in whom angiographic follow-up was not performed. Adverse procedural and in-hospital events were defined as the need for cardiopulmonary resuscitation, myocardial infarction, acute stent thrombosis, necessity for urgent coronary artery bypass grafting, and death. The long-term primary end points were defined as death from any cause, myocardial infarction, TVR, TLR, need for coronary artery bypass grafting, and MACE at any time during the inhospital stay or at follow-up. We defined the secondary end points as the incidence of restenosis, and stent thrombosis, defined as probable and definite. Statistical analysis was performed using SPSS version 16.0 (SPSS, Inc., Chicago, Illinois). Continuous variables are expressed as mean ⫾ SD or as medians with interquartile ranges as appropriate. Categorical variables are expressed as counts and percentage. Continuous variables were compared using Student’s t tests. Categorical variables were compared using chi-square statistics. A p value ⬍0.05 was considered statistically significant, and all reported p values are 2 sided. To account for potential differences between the 2 groups, a propensity analysis was performed on a patientbased setting for MACEs and a lesion-based setting for TLR.11,12 The propensity to be treated with DES was estimated using a nonparsimonious multivariate logistic regression model. The individual propensity score was then incorporated into Cox proportional-hazards regression models as a covariate as well as a treatment group (i.e., DES vs

Coronary Artery Disease/DES Versus BMS in Aorto-Ostial Lesions Table 3 Baseline procedural characteristics

Table 1 Baseline clinical characteristics Variable Age (years) Men Ejection fraction (%) Previous myocardial infarction Previous PCI Previous coronary artery bypass grafting Hypertension* Hypercholesterolemia* Current smoker Diabetes mellitus Unstable angina pectoris (CCS class IV) Stable angina pectoris (CCS class I–III) Silent myocardial ischemia (CCS class 0) Multivessel coronary disease Number coronary arteries diseased 1 2 3

BMS (n ⫽ 116)

DES (n ⫽ 230)

p Value

Variable

BMS (n ⫽ 118)

68.6 ⫾ 10.6 79 (68%) 55.0 ⫾ 11.5 41 (35%) 28 (24%) 41 (35%)

65.7 ⫾ 10.88 174 (76%) 52.3 ⫾ 9.7 101 (44%) 128 (56%) 86 (66%)

0.019† 0.135 0.022† 0.126 ⬍0.001† 0.709

12 (10%) 98 (83%) 1.02 ⫾ 0.13

70 (60%) 68 (59%) 13 (11%) 18 (16%) 44 (38%)

162 (70%) 167 (73%) 13 (6%) 73 (32%) 51 (22%)

0.059 0.009† 0.064 0.001† 0.002†

57 (49%)

99 (43%)

0.282

14 (12%)

59 (26%)

0.003†

97 (84%)

197 (86%)

0.401

19 (16%) 35 (30%) 62 (53%)

33 (14%) 67 (29%) 130 (57%)

0.618 0.841 0.587

Glycoprotein IIb/IIIa inhibitors Multivessel stenting Number of stents per lesion DES implanted Sirolimus-eluting stent Paclitaxel-eluting stent Everolimus-eluting stent Zotarolimus-eluting stent Biolimus-eluting stent Total stent length (mm) Final stent minimal luminal diameter (mm) Direct stenting Balloon predilatation Cutting balloon predilatation Directional coronary atherectomy Intravascular ultrasound guidance Rotablation Noncompliant balloon postdilatation Postdilatation balloon diameter (mm) Postdilatation balloon length (mm) Maximum balloon inflation (atm)

Data are expressed as mean ⫾ SD or as number (percentage). CCS ⫽ Canadian Cardiovascular Society. * Defined as being treated before coronary angiography. † Statistically significant.

AO lesion Left main coronary artery Right coronary artery Saphenous vein graft Lesion characteristics Type B1 Type B2 Type C Total occlusion Before intervention Reference vessel diameter (mm) Minimum luminal diameter (mm) Stenosis (%) Lesion length (mm)

DES (n ⫽ 239)

16 (7%) 200 (84%) 1.05 ⫾ 0.23 246 — 114 (46%) — 89 (36%) — 24 (10%) — 16 (7%) — 3 (1%) 14.37 ⫾ 5.60 17.55 ⫾ 7.76 3.66 ⫾ 0.63 3.43 ⫾ 0.53

p Value 0.252 0.880 0.127

⬍0.001* 0.001*

23 (20%) 88 (75%) 5 (4%) 10 (9%)

74 (31%) 116 (49%) 22 (9%) 0

0.022* ⬍0.001* 0.095 ⬍0.001*

37 (31%)

41 (17%)

0.002*

16 (14%) 65 (55%)

1 (1%) 99 (41%)

⬍0.001* 0.015*

3.43 ⫾ 1.37

2.74 ⫾ 1.72

⬍0.001*

14.55 ⫾ 7.48 11.29 ⫾ 8.77

⬍0.001*

15.31 ⫾ 7.25 15.72 ⫾ 10.67

0.666

Data are expressed as mean ⫾ SD or as number (percentage). * Statistically significant.

Table 2 Baseline lesion characteristics Variable

1057

BMS (n ⫽ 118)

DES (n ⫽ 239)

p Value

40 (34%) 67 (57%) 11 (9%)

91 (38%) 108 (45%) 40 (17%)

0.441 0.039* 0.060

32 (27%) 64 (55%) 22 (19%) 1 (1%)

85 (36%) 110 (46%) 44 (18%) 4 (2%)

0.110 0.144 0.957 0.532

3.49 ⫾ 0.83

3.32 ⫾ 0.84

0.097

loon diameter, direct stenting, and intravascular ultrasound guidance. In the lesion-based analysis, because observations recorded in the same patient cannot be considered independent,13 the sandwich estimator of the variance-covariance matrix was used to take into account clustered data (more lesions within the same subject). The discrimination and calibration ability of the propensity score model was assessed by means of the C-statistic and the Hosmer-Lemeshow statistic.14 The results of the Cox proportional-hazards analyses are reported as unadjusted and adjusted HRs with associated 95% confidence intervals (CIs) and p values.

1.36 ⫾ 0.68

1.31 ⫾ 0.61

0.582

Results

62.59 ⫾ 16.35 6.63 ⫾ 4.68

60.96 ⫾ 16.99 7.87 ⫾ 5.13

0.434 0.057

From a total of 628 AO lesions recorded in our database, we excluded 271 lesions (Figure 1). This resulted in a final group of 357 de novo AO lesions in 346 patients, with 239 lesions (230 patients) in the DES group and 118 lesions (116 patients) in the BMS group. The baseline clinical demographics of these patients are presented in Table 1. There were some baseline differences in clinical demographics between the 2 groups, with a higher incidence of previous PCI, hypercholesterolemia, and diabetes mellitus in the DES group. The baseline lesion characteristics are listed in Table 2. In the BMS and DES groups, most lesions were located within the right coronary artery. There were no significant differences

Data are expressed as mean ⫾ SD or as number (percentage). * Statistically significant.

BMS) to calculate the adjusted hazard ratios (HRs). Variables included in the propensity model were age, gender, the ejection fraction, hypertension, hypercholesterolemia, diabetes, unstable angina, current smoking, previous PCI, family history of coronary artery disease, left main stem treatment, right coronary artery treatment, total stent length, number of stents per lesion, rotablation, postdilatation bal-

1058

The American Journal of Cardiology (www.ajconline.org)

Table 4 Angiographic outcomes during the follow-up period Angiographic Outcome

BMS (n ⫽ 107)

DES (n ⫽ 231)

p Value

Angiographic follow-up obtained Postintervention Reference vessel diameter (mm) Minimum luminal diameter (mm) Stenosis (%) Follow-up Reference vessel diameter (mm) Minimum luminal diameter (mm) Stenosis (%) Lesion length (mm) Late luminal loss (mm) Restenosis Restenosis type Focal Diffuse Occlusive

70 (66%)

157 (68%)

0.726

3.85 ⫾ 0.68

3.79 ⫾ 0.57

0.449

3.66 ⫾ 0.63

3.43 ⫾ 0.53

0.001*

3.99 ⫾ 11.01

9.23 ⫾ 7.91

⬍0.001*

3.50 ⫾ 0.65

3.64 ⫾ 0.57

0.196

2.21 ⫾ 1.15

2.97 ⫾ 0.95

⬍0.001*

38.83 ⫾ 28.62 5.42 ⫾ 4.52 1.54 ⫾ 1.21 32 (47%)

18.03 ⫾ 21.24 4.94 ⫾ 3.59 0.42 ⫾ 0.84 32 (20%)

⬍0.001* 0.519 ⬍0.001* ⬍0.001*

22 (69%) 7 (22%) 3 (9%)

19 (59%) 9 (28%) 4 (13%)

0.434 0.564 0.689

Data are expressed as mean ⫾ SD or as number (percentage). * Statistically significant. Table 5 Clinical outcomes in-hospital and during the follow-up period Variable In-hospital outcomes Myocardial infarction Acute stent thrombosis Emergency coronary artery bypass grafting Death Follow-up Length of follow-up (years) Dual-antiplatelet therapy (months) Total death Cardiac death Myocardial infarction Need for coronary artery bypass grafting TLR (per patient) TLR (per lesion) TVR (per patient) TVR (per lesion) Definite stent thrombosis Probable stent thrombosis MACEs

BMS (n ⫽ 104)

DES (n ⫽ 222)

p Value

6 (5%) 1 (1%) 0

14 (6%) 1 (1%) 0

0.765 0.574 —

1 (1%)

1 (1%)

0.609

2.5 (1.2–4.8) 1 (0.9–1)

3.3 (2.0–4.7)

0.134

12.4 (9.5–32.7) ⬍0.001*

19 (18%) 14 (14%) 2 (2%) 4 (4%)

25 (11%) 15 (7%) 8 (4%) 5 (2%)

0.084 0.047* 0.412 0.413

28 (27%) 28 (26%) 37 (36%) 37 (35%) 0 1 (1%) 49 (47%)

26 (12%) 28/231 (12%) 52 (23%) 54 (23%) 2/222 (1%) 7 (3%) 76 (34%)

0.001* 0.001* 0.022* 0.027* 0.332 0.233 0.026*

Data are number (percentage) or as median (interquartile range). * Statistically significant.

in preintervention lesion characteristics between the groups as assessed by quantitative coronary angiography. The baseline procedural characteristics are listed in Table 3. Multivessel PCI was performed in most patients in the 2 groups. The total stent length was longer in the DES group

than in the BMS group (17.55 ⫾ 7.76 vs 14.37 ⫾ 5.60 mm, p ⬍0.001). The final stent minimum luminal diameter was significantly smaller in the DES group (3.43 ⫾ 0.53 vs 3.66 ⫾ 0.63 mm, p ⫽ 0.001). Balloon predilatation and intravascular ultrasound were performed less frequently in the DES group. Table 4 lists the angiographic outcomes at follow-up. At follow-up, the minimum luminal diameter was significantly larger in the DES group than in the BMS group (2.97 ⫾ 0.95 vs 2.21 ⫾ 1.15 mm, p ⬍0.001). In addition, the incidence of binary restenosis was 20% (32 of 157) in the DES group compared to 47% (32 of 70) in the BMS group (p ⬍0.001). Table 5 lists the clinical outcomes in the hospital and during the follow-up period. There was 1 episode of acute stent thrombosis in the BMS group in a patient after PCI to the left main stem complicated by intraprocedural thrombosis within the left main stem and left anterior descending artery. The patient was treated with glycoprotein IIb/IIIa inhibitors and dual-antiplatelet therapy with aspirin and clopidogrel with TIMI grade 3 flow and resolution of thrombus at the end of the procedure. However, she went on to have cardiac arrest and died ⬍24 hours after the procedure. There was also 1 episode of acute stent thrombosis in the DES group in a patient who had PCI with sirolimus-eluting stent implantation in the ostium of the right coronary artery with intravascular ultrasound guidance. Despite dual-antiplatelet therapy with aspirin and clopidogrel, the patient had sudden cardiac arrest 24 hours after the index procedure; repeat coronary angiography showed acute occlusion of the right coronary artery, and the patient subsequently died. Clinical follow-up was achieved, with a median follow-up of 2.5 years (interquartile range 1.2 to 4.8) in the BMS group and a median follow-up of 3.3 years (interquartile range 2.0 to 4.7) in the DES group. The incidence of TLR per lesion was significantly lower in the DES group at 12% (n ⫽ 28) than in the BMS group (26% [n ⫽ 28]) (p ⫽ 0.001). There was a significantly lower rate of TVR per patient in the DES group compared to the BMS group (23% [n ⫽ 52] vs 36% [n ⫽ 37] p ⫽ 0.022). The MACE rate at follow-up was lower in the DES group (34% [n ⫽ 76]) compared to the BMS group (47% [n ⫽ 49]) (p ⫽ 0.208), although this value was not statistically significant. Propensity-adjusted Cox proportional-hazards analysis was used to adjust for baseline differences in lesion characteristics between the 2 groups. Unadjusted analysis showed a significantly lower TLR per lesion rate in the DES group (HR 0.38, 95% CI 0.23 to 0.65, p ⬍0.001) and a significantly lower MACE rate in the DES group (HR 0.65, 95% CI 0.45 to 0.94, p ⫽ 0.02). The propensity-adjusted analyses confirmed that the use of DES was associated with significantly lower TLR per lesion rates (HR 0.28, 95% CI 0.15 to 0.52, p ⬍0.001) and MACE rates (HR 0.50, 95% CI 0.32 to 0.79, p ⫽ 0.003) during the follow-up period. The C-statistic for the propensity score model was 0.83, indicating excellent discrimination. The Hosmer-Lemeshow goodness-of-fit test p value was 0.25, confirming good calibration and fit of the multivariate model that estimated the propensity score. Definite stent thrombosis occurred in 1% of patients (n ⫽ 2) in the DES group and no patients in the BMS group (p ⫽

Coronary Artery Disease/DES Versus BMS in Aorto-Ostial Lesions

0.332); 1 was an acute stent thrombosis and the other was a late stent thrombosis. Probable stent thrombosis occurred in 3% of the DES group (n ⫽ 7) and 1% of the BMS group (n ⫽ 1) (p ⫽ 0.233). In the DES group, probable stent thrombosis was composed of 2 subacute stent thromboses, 1 late stent thrombosis, and 4 very late stent thromboses, while in the BMS group, the probable stent thrombosis event was acute. Discussion The main findings of this study were (1) DES implantation during PCI to AO lesions resulted in lower restenosis rates compared to BMS implantation, (2) TLR and TVR rates were significantly reduced by the use of DES, and (3) MACE rates were significantly lower with DES use. From previous reports, PCI of lesions in an AO location is associated with higher rates of restenosis than nonostial lesions.15 The main factor responsible for restenosis is excess intimal hyperplasia and occasionally vessel recoil in the acute and chronic phase.16 The presence of a greater degree of sclerosis and calcification may contribute to a final suboptimal result.17 Notably, up to 4.5% of restenotic lesions at the AO site are due to mechanical problems with stent deployment, such as a crushed stent, missing of the lesion during stent deployment, or the stent being stripped off the balloon during implantation, with the remainder of lesions secondary to stent underexpansion during the index procedure.18 Furthermore, to address the possibility of missing an AO lesion, PCI is usually performed with slight stent protrusion into the aorta; protruding stent struts may result in increased platelet activation, thrombus formation, and distal embolization and make future procedures more challenging because of difficulty in catheter reengagement of the coronary ostium.19 –21 Concerns have therefore been raised as to the optimal treatment strategy for these lesions, particularly given their high risk status because of the large myocardial area that would potentially be affected by an adverse event at this site.22 Limited data are available on restenosis and clinical adverse event rates after intervention to AO lesions, largely because of small study sample sizes in previous publications. To our knowledge, this report represents the largest study with the longest reported follow-up of angiographic and clinical outcomes after PCI to AO lesions. Park et al6 reported 1-year outcomes after DES use in 184 patients versus BMS use in 174 patients, showing a restenosis rate of 11% and a TLR rate of 4% in the DES group. In addition, restenosis rates of 4% after sirolimus-eluting stent implantation and 8.7% after paclitaxel-eluting stent implantation, with a combined TLR rate of 7% in the DES group compared to 32% in the BMS group, were shown by Barlis et al.23 A study from our group showed TLR in 6% and MACEs in 19% after sirolimus-eluting stent use, compared to 28% and 44%, respectively, in the BMS group at 10month follow-up.5 The longest previously published clinical follow-up period of 2 years reported a 20% restenosis rate at 7-month angiographic follow-up and a 5% TLR rate after paclitaxel-eluting stent use.7 Our results compare relatively favorably to these reports, with a 20% restenosis rate and a

1059

12% TLR rate in the DES group, in the context of a median follow-up period of ⬎3 years. Notably, there was no statistically significant difference in the rates of stent thrombosis between the DES and BMS groups, despite a significant difference in the duration of dual-antiplatelet therapy between the 2 groups. Although there was a higher incidence of definite and probable stent thrombosis after DES (4%) compared to BMS (1%), this was not statistically significant. Our study was underpowered to detect a significant difference in this relatively rare event. Previous studies have suggested that revascularization rates in AO lesions may be reduced by the use of cutting balloon predilatation in an attempt to overcome resistant stenoses and create more organized plaque distribution.24 However, this observation could not be assessed in our study, because a cutting balloon was used relatively infrequently. Furthermore, the use of adjunctive technology such as directional atherectomy, rotablation, and intravascular ultrasound was significantly more common in the BMS group, because of historical changes in PCI techniques, and therefore did not appear to have an impact of rates of restenosis, unlike the use of DES. Interestingly, restenosis rates in DES were significantly lower despite a significantly larger mean final stent minimum luminal diameter and postdilatation balloon diameter in the BMS group, combined with a longer mean stent length in the DES group. These results therefore highlight the advantageous long-term effects of DES implantation in AO lesions and demonstrate that DES should be used when possible in this high-risk lesion subset. This study had some limitations. First, it had the known limitations of a retrospective analysis. Second, there were differences in the baseline clinical demographics between the 2 groups, but these were generally in favor of the BMS group; a propensity-adjusted analysis to account for baseline differences showed that the long-term outcomes in this group were worse than in the DES group. Third, clinical follow-up was achieved in a greater proportion of the DES group; despite this, more favorable clinical outcomes were achieved in this group. Fourth, angiographic follow-up was not performed in all patients. Last, the mean clinical follow-up periods were different between the groups but was longest in the DES group, therefore adding to the strength of the findings of this study in that long-term outcomes after DES implantation are more favorable than BMS despite longer follow-up in this group. Acknowledgment: We thank all the members of the cardiac catheterization teams at Columbus and San Raffaele hospitals who assisted with the data input and analysis. 1. Eigler NL, Weinstock B, Douglas JS Jr, Goldenberg T, Hartzler G, Holmes D, Leon M, Margolis J, Nobuyoshi M, O’Neill W. Excimer laser coronary angioplasty of aorto-ostial stenoses. Results of the excimer laser coronary angioplasty (ELCA) registry in the first 200 patients. Circulation 1993;88:2049 –2057. 2. Koller PT, Freed M, Grines CL, O’Neill WW. Success, complications, and restenosis following rotational and transluminal extraction atherectomy of ostial stenoses. Cathet Cardiovasc Diagn 1994;31: 255–260.

1060

The American Journal of Cardiology (www.ajconline.org)

3. Kurbaan AS, Kelly PA, Sigwart U. Cutting balloon angioplasty and stenting for aorto-ostial lesions. Heart 1997;77:350 –352. 4. Gagnor A, Varbella F, Rubartelli P, Luceri S, Conte MR. Recurrent restenosis after implantation of sirolimus-eluting stents in aorto-ostial lesions: successful treatment with polytetrafluoroethylene-covered stents. J Cardiovasc Med (Hagerstown) 2008;9:201–204. 5. Iakovou I, Ge L, Michev I, Sangiorgi GM, Montorfano M, Airoldi F, Chieffo A, Stankovic G, Vitrella G, Carlino M, Corvaja N, Briguori C, Colombo A. Clinical and angiographic outcome after sirolimus-eluting stent implantation in aorto-ostial lesions. J Am Coll Cardiol 2004;44: 967–971. 6. Park DW, Hong MK, Suh IW, Hwang ES, Lee SW, Jeong YH, Kim YH, Lee CW, Kim JJ, Park SW, Park SJ. Results and predictors of angiographic restenosis and long-term adverse cardiac events after drug-eluting stent implantation for aorto-ostial coronary artery disease. Am J Cardiol 2007;99:760 –765. 7. Tsuchida K, Daemen J, Tanimoto S, Garcia-Garcia HM, Kukreja N, Vaina S, Ong AT, Sianos G, de Jaegere PP, van Domburg RT, Serruys PW. Two-year outcome of the use of paclitaxel-eluting stents in aorto-ostial lesions. Int J Cardiol 2008;129:348 –353. 8. Mehran R, Dangas G, Abizaid AS, Mintz GS, Lansky AJ, Satler LF, Pichard AD, Kent KM, Stone GW, Leon MB. Angiographic patterns of in-stent restenosis: classification and implications for long-term outcome. Circulation 1999;100:1872–1878. 9. Colombo A, Tobis J. Techniques in Coronary Artery Stenting. London, United Kingdom: Martin Dunitz, 2000. 10. Lansky AJ, Dangas G, Mehran R, Desai KJ, Mintz GS, Wu H, Fahy M, Stone GW, Waksman R, Leon MB. Quantitative angiographic methods for appropriate end-point analysis, edge-effect evaluation, and prediction of recurrent restenosis after coronary brachytherapy with gamma irradiation. J Am Coll Cardiol 2002;39:274 –280. 11. D’Agostino RB Jr. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998;17:2265–2281. 12. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127:757–763. 13. Kastrati A, Schomig A, Elezi S, Schuhlen H, Wilhelm M, Dirschinger J. Interlesion dependence of the risk for restenosis in patients with

14. 15.

16.

17. 18.

19. 20. 21. 22. 23. 24.

coronary stent placement in multiple lesions. Circulation 1998;97: 2396 –2401. Hosmer DW, Hosmer T, Le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med 1997; 16:965–980. Mavromatis K, Ghazzal Z, Veledar E, Diamandopoulos L, Weintraub WS, Douglas JS, Kalynych AM. Comparison of outcomes of percutaneous coronary intervention of ostial versus nonostial narrowing of the major epicardial coronary arteries. Am J Cardiol 2004;94:583–587. Tsunoda T, Nakamura M, Wada M, Ito N, Kitagawa Y, Shiba M, Yajima S, Iijima R, Nakajima R, Yamamoto M, Takagi T, Yoshitama T, Anzai H, Nishida T, Yamaguchi T. Chronic stent recoil plays an important role in restenosis of the right coronary ostium. Coron Artery Dis 2004;15:39 – 44. Stewart JT, Ward DE, Davies MJ, Pepper JR. Isolated coronary ostial stenosis: observations on the pathology. Eur Heart J 1987;8:917–920. Castagna MT, Mintz GS, Leiboff BO, Ahmed JM, Mehran R, Satler LF, Kent KM, Pichard AD, Weissman NJ. The contribution of “mechanical” problems to in-stent restenosis: an intravascular ultrasonographic analysis of 1090 consecutive in-stent restenosis lesions. Am Heart J 2001;142:970 –974. Harb TS, Ling FS. Inadvertent stent extraction six months after implantation by an entrapped cutting balloon. Catheter Cardiovasc Interv Jul 2001;53:415– 419. Abhyankar A, Gai L, Bailey BP. Angioplasty through a stent side door. Int J Cardiol 1996;55:107–110. Chetcuti SJ, Moscucci M. Double-wire technique for access into a protruding aorto-ostial stent for treatment of in-stent restenosis. Catheter Cardiovasc Interv 2004;62:214 –217. Rissanen V. Occurrence of coronary ostial stenosis in a necropsy series of myocardial infarction, sudden death, and violent death. Br Heart J 1975;37:182–191. Barlis P, Kaplan S, Dimopoulos K, Ferrante G, Di Mario C. Comparison of bare-metal and sirolimus- or paclitaxel-eluting stents for aortoostial coronary disease. Cardiology 2008;111:270 –276. Nassar H, Gotsman I, Gerganski P, Moseri M, Lotan C, Gotsman M. Cutting balloon angioplasty and stent implantation for aorto-ostial lesions: clinical outcome and 1-year follow-up. Clin Cardiol 2009;32: 183–186.