Drug-Eluting Stents Versus Bilateral Internal Thoracic Grafting for Multivessel Coronary Disease

Transcription

1 Drug-Eluting Stents Versus Bilateral Internal Thoracic Grafting for Multivessel Coronary Disease Itzhak Herz, MD, Yaron Moshkovitz, MD, Dan Loberman, MD, Gideon Uretzky, MD, Rony Braunstein, PhD, Alberto Hendler, MD, Einat Zivi, BmedSc, Yanai Ben-Gal, MD, and Rephael Mohr, MD Departments of Cardiology and Cardiothoracic Surgery, Tel Aviv Sourasky Medical Center and Sackler Faculty of Medicine, Tel Aviv University, Departments of Cardiology and Cardiac Surgery, Assuta Medical Center, Tel Aviv, and Center for Quality, Safety and Data, Hadassah Hebrew University Hospital, Ein Karem, Jerusalem, Israel Background. Reduction of restenosis and reinterventions was recently reported with percutaneous interventions (PCI), including drug-eluting stents (Cypher; Cordis, Miami Lakes, FL). This study compares results of multivessel Cypher stenting with those of bilateral internal thoracic artery (BITA) grafting. Methods. From January 2002 to June 2004, 768 consecutive patients underwent multivessel myocardial revascularization; 138 by PCI including Cyphers and 630 by BITA. After matching for age, sex, ejection fraction, extent of coronary disease, and congestive heart failure, two groups (113 patients each) were used to compare the two revascularization modalities. Results. Both groups were similar; however, left main and intraaortic balloon were more prevalent in the BITA group. The number of coronary vessels treated per patient was higher in the BITA group (2.87 vs 2.22, p < 0.001). Follow-up ranged between 6 and 34 months. Thirty-day mortality was 0.9% in the BITA and zero in the PCI group (p 0.32). There were no late deaths in the BITA and three (2.7%) in the Cypher group (p 0.08). Angina returned in 28.3% of the Cypher and 12.4% of the BITA group, p A Cox proportional hazard model revealed assignment to the Cypher group to be the only predictor of angina recurrence (odds ratio 2.78, 95% confidence interval ). There were 16 (14.2%) reinterventions in the Cypher group compared with six (5.3%) in the BITA group. One-year reintervention-free survival (Kaplan-Meier) of the BITA was 96% compared with 86.6% in the Cypher group (p 0.005, log-rank test). Conclusions. Despite improved results of PCI with Cyphers, midterm clinical outcome of multivessel patients treated with BITA is still better. (Ann Thorac Surg 2005;80: ) 2005 by The Society of Thoracic Surgeons Surgical revascularization of the left anterior descending (LAD) with the left internal thoracic artery (ITA) is still the only proven method of improving event-free survival of patients with multivessel coronary artery disease having LAD stenosis [1, 2]. Percutaneous intervention (PCI) has failed to match the outcome obtained with bypass grafting primarily because of a greater need for repeat reinterventions as a result of in-stent restenosis [3 5]. Considerable reduction of restenosis and reintervention rates was recently reported with the introduction of drug-eluting stents (DES) [6, 7]. In arecent report of 459 patients with single LAD stenosis, DES significantly decreased restenosis and reintervention rates as compared with bare-metal stents [8]. Bilateral skeletonized ITA with left-sided (LAD circumflex) arterial revascularization [9] is our preferred method of surgical revascularization of patients with multivessel disease. We therefore decided to compare our initial clinical results of bilateral ITA (BITA) grafting Accepted for publication May 17, Address correspondence to Dr Mohr, Dept of Cardiothoracic Surgery, Tel Aviv Sourasky Medical Center, 6 Weizman Street, Tel Aviv 64239, Israel; marion@tasmc.health.gov.il. with those of multivessel revascularization with PCI, including at least one DES. Patients and Methods Between January 2002 and June 2004, 630 consecutive patients with multivessel disease underwent left-sided arterial myocardial revascularization with skeletonized BITA and 138 underwent multivessel PCI, including at least one DES (Cypher; Cordis, Miami Lakes, FL). Most preoperative characteristics of BITA and PCI patients were similar. However, BITA patients were older and had increased prevalence of triple-vessel disease (66% vs 41%). On the other hand, more patients in the PCI group had congestive heart failure (CHF) (8% vs 1.5%). In order to control for these differences between BITA and PCI patients, we decided to compare results between the two treatment modalities only after matching for age, sex, CHF, ejection fraction (EF), and the extent of coronary artery disease involvement (Table 1). Each of the two matched groups thus created contained 113 patients. Preoperative characteristics of patients in both groups are depicted in Table 2. During the study period, selection criteria for surgery vs PCI were mainly technical. In principle, there was a 2005 by The Society of Thoracic Surgeons /05/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg HERZ ET AL 2005;80: CYPHER VS BITA FOR MULTIVESSEL DISEASE 2087 Table 1. Extent of Coronary Disease and Number of Vessels Treated a BITA (n 113) Cypher (n 113) p Left Main 30 (26.5%) 4 (3.5%) Three-vessel disease 61 (54%) 61 (54%) 1.0 Two-vessel disease 52 (46%) 52 (46%) 1.0 Vessels treated: 2 41 (36.3%) 88 (77.9%) or more 72 (63.7%) 25 (22.1%) Revascularization of RCA 50 (44.2%) 63 (55.8%) a Grafts or stents. BITA bilateral internal thoracic artery; RCA right coronary artery. preference to refer patients for surgery for the following reasons. 1. Comorbid diseases such as diabetes, renal failure, etc. 2. In-stent restenosis or thrombosis of a coronary artery. 3. Complex type C lesions (calcified coronary arteries, lesion length over 20 mm, twisted arteries, and suspicion of a thrombus in an artery). 4. Left main disease. 5. Nonavailability of a Cypher, including cases in which the patient was unable to fund a Cypher. 6. Patient s preference. Acute myocardial infarction (within the previous 48 hours), poor EF ( 25%), calcification or thrombus within the LAD, long lesions ( 30 mm), total occlusion and bifurcation lesions, were major criteria for exclusion of patients from the group treated with stents. On the other hand, patients with diabetes or ostial LAD lesions were not excluded. During the study period, about 55% of the patients treated surgically in our institution were operated on without extracorporeal circulation (off-pump coronary artery bypass [OPCAB]). Most decisions to perform open-heart surgery were made intraoperatively. The OPCAB was preferred in patients over 70 years of age; in patients with a calcified aorta, patients with damage to other systems (kidneys, lungs, liver, peripheral vascular disease [PVD], etc.); coronary artery bypass grafting (CABG) with pump was preferred, particularly in the presence of atheromatotic, calcified, and especially small coronary arteries, or in diffuse coronary artery disease that required four or more anastomoses. In the PCI group, stent implantation was performed after balloon angioplasty dilatation. All patients received aspirin (325 mg daily) before and after the procedure, and clopidogrel (Plavix; Sanofi, New York City, NY) (a loading dose of 300 mg the day before the procedure and 75 mg daily for three months thereafter). During the procedure, all patients were treated intravenously with heparin. Intravenous platelet glycoprotein IIb/IIIa inhibitor (Integrilin [eptifibatidel, Schering-Plough, Belgium] or Aggrestat [Tirofiban, Merck, Sharpe & Dohme, Holland]) were used only in 12 of the PCI group. All LAD lesions in the PCI group were treated with the Cypher. In addition to DES to the LAD, the Cypher was also used in 32 of the 63 (51%) right coronary arteries (RCA) and in 53 of the 78 (68%) circumflex (Cx) marginals treated. In most patients, only one Cypher was used for the vessel treated. However, more than one stent was used if required (long lesion, dissection, bifurcation, etc). Bare-metal stents or plain balloon angioplasty were used for non-lad lesions that were not stented with DES. Bare-metal stents (25 in RCA and 16 in Cx) were used in patients with tortuous or calcified coronary vessels, or in patients who could not fund two Cypher stents. A percutaneous transluminal coronary angioplasty with plain balloon (4 RCA and 9 Cx) was used in vessels smaller than 2.25 mm, or in patients with in-stent restenosis, with focal in-stent lesions. In the BITA group, all internal thoracic arteries were dissected as skeletonized vessels. In order to achieve left-sided (LAD circumflex) arterial revascularization, we used the right ITA either as an in situ graft to the LAD or as a free graft attached end-to-side to the left ITA (composite T graft). In the group of patients treated with BITA, right coronary system (posterior descending artery or posterolateral branch of the right coronary artery) revascularization was performed with saphenous vein grafts in 36 Table 2. Patient Characteristics BITA (n 113) Cypher (n 113) p Mean age (years) Age (18.6%) 21 (18.6%) 1.0 Female 5 (4.4%) 5 (4.4%) 1.0 Diabetes 37 (32.7%) 41 (36.3%) Hypertension 67 (59.3%) 69 (61.1%) Hyperlipidemia 80 (70.8%) 82 (72.6%) COPD 3 (2.7%) 5 (4.4%) CRF 4 (3.5%) 4 (3.5%) 1.0 Peripheral vascular 8 (6.7%) 10 (8.3%) disease Old myocardial 40 (35.4%) 44 (38.9%) infarction Acute MI (7 days) 9 (8%) 8 (7.1%) Ejection fraction 1 (0.9%) 1 (0.9%) % Prior IABP 4 (3.5%) 0 (0%) Prior PCI 32 (28.3%) 34 (30.1%) In-stent restenosis 8 (7.1%) 13 (11.5%) Emergency 6 (5.3%) 5 (4.4%) Repeat operation 1 (0.9%) 3 (2.7%) Congestive heart failure 7 (6.2%) 7 (6.2%) 1.0 BITA bilateral internal thoracic artery grafting; CRF chronic renal failure (Cr 1.8); COPD chronic obstructive pulmonary disease; IABP intra-aortic balloon pump; MI myocardial infarction; PCI percutaneous coronary intervention angioplasty or stent.

3 2088 HERZ ET AL Ann Thorac Surg CYPHER VS BITA FOR MULTIVESSEL DISEASE 2005;80: patients, with radial artery in six, and with the distal end of the free right ITA (composite T graft) in eight patients. To decrease the risk of spasm of the arterial grafts, we treated all BITA patients with a high-dose intravenous infusion of isosorbide dinitrate (Isoket) 4 20 mg/hour during the first postoperative 24 to 48 hours [10]. Systolic blood pressure was maintained above 100 mm Hg. From the second postoperative day, radial and gastroepiploic artery (RGEA) patients were treated postoperatively with calcium channel blockers (diltiazem mg/day orally). Patients were advised to have a routine nuclear scan within three months of revascularization (PCI or surgery); however, many of the thallium scans (mainly in the PCI group) were performed only upon return of chest pain. Follow-up was obtained by telephone questionnaire. Patients data were collected and analyzed according to Society of Thoracic Surgeon definitions. Statistical Analysis Data are expressed as mean SD or proportions, as appropriate. The two groups were matched for age, sex, and extent of coronary artery disease. The 2 test and Fisher exact test were used to compare discrete variables. The Cox proportional hazards model was used to evaluate risk factors for early return of angina and reintervention. Odds ratio (OR) and 95% confidence interval (CI) were given. Postoperative angina-free survival and reintervention-free survival are expressed by the Kaplan- Meier method, and comparison between groups is made by the log-rank test. All analyses were performed by SPSS 12 software (SPSS Inc, Chicago, IL). Results After matching for age, sex, CHF, EF, and extent of coronary artery disease (Table 1), the two groups were similar in most preoperative characteristics. However, left main disease (Table 1), and preprocedural use of intraaortic balloon pump (Table 2) were more prevalent in the BITA group. In 76 of the PCI group, two or more vessels were treated with the Cypher. In the remaining 37, the LAD was treated with the Cypher and at least one more vessel was treated with PCI (bare-metal stent or plain balloon angioplasty). The average number of coronary vessels treated in the BITA and PCI groups was vs , respectively, with p less than More patients in the BITA group had revascularization of three vessels (63%) and in most of the patients in the Cypher group (77.9%) only two vessels were revascularized (Table 1). Thirty-day mortality was 0.9% (one patient) in the BITA group. There was no periprocedural mortality among the PCI group (p 0.32). Postoperative morbidity in the BITA group included deep sternal infection (1.8%), stroke (0.9%), perioperative myocardial infarction (MI) (3.5%), and revision for bleeding (4.2%). In the Cypher group, in addition to reangina and reintervention, major morbidity included one periprocedural and two late nonfatal MIs, exacerbation of heart failure (pulmonary edema) in two patients, one life-threatening arrhythmia, and one pacemaker implantation. One patient required dialysis, and one, who underwent CABG three months after Cypher implantation, sustained deep sternal wound infection. Follow-up ranged between six and 34 months (mean 16 months). There were no late deaths in the BITA group and three late deaths (2 MIs and one stroke, 2.7%) in the Cypher group (p 0.081). Angina returned in 32 patients (28.3%) of the Cypher group compared with 14 (12.4%) in the BITA group (p 0.003). Sixty-five of the Cypher group and 61 of the BITA group underwent postoperative thallium SPECT scintigraphy, which was found to be positive in 13 of the Cypher compared with eight of the BITA group patients (p 0.252). During the follow-up period, 22 of the Cypher and 10 of the BITA group were referred for coronary angiography. There were 16 (14.2%) reinterventions in the Cypher group: one surgical intervention and fifteen coronary angioplasties (including eight to a Cypher-treated vessel: target vessel reintervention for Cypher, 9 of %). There were six (5.3%) reinterventions in the surgical group: five PCIs and one surgical intervention (p Cypher vs BITA). Only three of the reinterventions were to ITA-treated vessels (target vessel reintervention for ITA was 3/ %), three were in native coronary arteries, and one was in a saphenous vein graft (SVG)-treated vessel. Two (1.8%) of the PCI patients and one of the BITA patients sustained a nonfatal MI during the follow-up period. Angina-free survival (Kaplan-Meier) of the BITA patients was better than that of the PCI patients (p 0.001, log-rank test, Fig. 1). The BITA patients also had significantly better reintervention-free survival (p 0.005, logrank test, Fig. 2). Cox proportional hazards model and univariate analysis did not define any specific preoperative (Table 2) or operative (Table 1) characteristics, including the use of the off-pump technique, to be associated with increased risk of reangina or reintervention in the BITA group. In the PCI group, independent predictors of reangina were peripheral vascular disease (PVD) (OR % CI ) and triple-vessel disease (OR % CI ). All other pre-pci patient characteristics, as well as the number of vessels treated with DES, did not have any influence on angina recurrence. Fig 1. Angina-free survival of the Cypher versus bilateral internal thoracic artery (BITA) groups (Kaplan-Meier).

4 Ann Thorac Surg HERZ ET AL 2005;80: CYPHER VS BITA FOR MULTIVESSEL DISEASE 2089 Fig 2. Intervention-free survival of patients undergoing multivessel percutaneous intervention, including the Cypher to left anterior descending versus bilateral internal thoracic artery (BITA) (Kaplan-Meier). The only independent predictor (Cox proportional hazards model) of reintervention in the PCI group was PVD (OR % CI ). On the other hand, Cx treatment with the Cypher had a protective effect against reintervention (OR % CI ). To determine whether the difference between groups in reangina and reinterventions is affected by the difference in baseline characteristics or the number of vessels treated, we performed multivariable analysis (Cox) with the patient group (PCI or BITA) as an independent variable. We first included the variables to be controlled (the number of vessels treated, left main disease, and right system revascularization) and then the treatment group. The Cox proportional hazards model showed that, after controlling for the above risk factors, assignment to the PCI group was the only independent predictor of reangina (OR % CI ) and reintervention (OR 3.74, 95% CI ). Comment In several retrospective studies comparing PCI and CABG [11 13],atrendwasfoundforbetterlong-termsurvivalwith CABG in patients with multivessel disease and proximal LAD stenosis. However, in later reports with more liberal use of stents to the proximal LAD, the only advantage of CABG was alower rate of reinterventions [5, 14]. Significant reduction of restenosis and reintervention rateswaswidelyreportedrecentlywiththeuseofdes[6, 8, 15]. After the initial reported success of DES for single-vessel disease, more patients with multivessel disease and LAD stenosis are currently being referred for PCI rather than for CABG. In this report, all PCI patients were treated with the Cypher to the LAD, and in 76 of the 113 patients (67%) two or more vessels were treated with DES. A major finding of this study comparing BITA grafting and PCI with DES for multivessel disease is the significantly lower recurrence of angina and reintervention rate in the surgical group treated with BITA. This is a retrospective cohort study; the two groups were matched for age, sex, CHF, EF, and extent of coronary artery disease, and the study findings were strongly supported by the Cox proportional hazards model that was used in order to control for the major differences between groups (left main, number of vessels treated, and RCA revascularization). After controlling for the above differences between groups, assignment to the PCI group remained the only significant predictor of reangina and reinterventions. Only nine of the reinterventions in the PCI group were performed to a Cyphertreated vessel. Thus, target vessel reintervention to Cypher-treated vessels was only 4.76%. Moreover, circumflex treatment with the Cypher had a protective effect against reintervention. These observations, together with the fact that, despite a similar extent of coronary vessel involvement, the number of vessels treated was significantly lower in the PCI group, may suggest that reangina and reintervention were related to revascularization of non-lad vessels with bare-metal stent or plain balloon angioplasty or to incomplete revascularization. Another explanation for the high recurrence of angina and higher reintervention rate in the Cypher group may be the relatively increased number of patients with diabetes, peripheral vascular disease, in-stent restenosis, and triple-vessel disease, who were not considered in the past to be ideal candidates for PCI [16]. This study reflects the current treatment situation with the two revascularization techniques. The groups were matched for age, sex, CHF, EF, and extent of coronary artery disease. They were also similar with regard to diabetes mellitus, which is a major determinant of restenosis [15]. However, other important covariates (mainly technical), such as vessel diameter, lesion length, occluded or thrombosed vessels, and bifurcation lesions, were not included in the analysis. These technical parameters, which are less important in the surgical group, may affect restenosis, reangina, and reintervention in the PCI group. Larger, prospective multicenter studies are required to determine their importance in patient selection for PCI or surgery in this evolving era of DES. Another limitation of this study is the relatively short follow-up period (mean 16 months). There is growing evidence that DES may develop delayed thrombosis related to delayed endothelialization, hypersensitivity to the stent polymer, or discontinuation of antiplatelet treatment [17, 18]. Longer follow-up is therefore required. In conclusion, despite the significantly higher early morbidity associated with BITA grafting, midterm clinical outcome is better than that of multivessel PCI with DES. The reintervention gap [8] between surgery and percutaneous interventions may further be reduced by better patient selection according to known angiographic criteria to each of the revascularization modalities. References 1. Loop DF, Lytle BW, Cosgrove DM, et al. Influence of internal-mammary-artery-graft on 10 year survival and other cardiac events. N Engl J Med 1986;314:1 6.

5 2090 HERZ ET AL Ann Thorac Surg CYPHER VS BITA FOR MULTIVESSEL DISEASE 2005;80: Barner HB, Swartz MI, Mudd JG, et al. Late patency of the internal mammary artery as a coronary artery bypass conduit. Ann Thorac Surg 1982;34: Goy JJ, Kaufmann U, Goy-Eggenberger D, et al. A prospective randomized trial comparing stenting to internal mammary artery grafting for proximal, isolated de novo left anterior coronary artery stenosis: the SIMA trial. Stenting vs internal mammary artery. Mayo Clin Proc 2000;75: Serruys PW, Unger F, Sousa EJ, et al. Comparison of coronary artery bypass surgery and stenting for treatment of multivessel disease. N Engl J Med 2001;344: Rodriguez A, Rodriguez AM, Baldi J, et al. Coronary stenting versus coronary bypass surgery in patients with multivessel disease and significant proximal LAD stenosis: results from the ERACI II study. Heart 2003; 89: Morice MC, Serruys PW, Sousa E, et al. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med 2003;346: Sousa JE, Costa MA, Abizaid A, et al. Lack of neointimal proliferation after implantation of sirolimus-coated stents in human coronary arteries: a quantitative coronary angiography and three-dimensional intravascular ultrasound study. Circulation 2001; 103: Sawhney N, Moses JW, Leon MB, et al. Treatment of left anterior descending coronary artery disease with Sirolimuseluting stents. Circulation 2004;110: Lev-Ran O, Mohr R, Uretzky G, et al. Graft of choice to right coronary system in left-sided bilateral internal thoracic artery grafting. Ann Thorac Surg 2003;75: Gurevitch J, Miller HI, Shapira I, et al. High-dose isosorbide dinitrate for myocardial revascularization with composite arterial grafts. Ann Thorac Surg 1997;63: Jones RH, Kessler K, Phillips HR, et al. Long-term survival benefits of coronary artery bypass grafting and percutaneous transluminal angioplasty in patients with coronary artery disease. J Thorac Cardiovasc Surg 1996;111: Arvinder S, Kurbaan AF, Rickards CDJ, et al. Relation between coronary artery disease, baseline clinical variables, revascularization mode, and mortality. Am J Cardiol 2000;86: Hannan EL, Racz MJ, McCallister BD, et al. A comparison of three-year survival after coronary artery bypass graft surgery and percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1999;33: O Keefe JH Jr, Kreamer TR, Jones PG, et al. Isolated left anterior descending coronary artery disease. Percutaneous transluminal coronary angioplasty versus stenting versus left internal mammary artery bypass grafting. Circulation 1999;100 (suppl I):II Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in native coronary artery. N Engl J Med 2003;349: Smith SC Jr, Dove JT, Jacobs AK, et al. ACC/AHA guidelines for percutaneous coronary intervention (revision of the 1993 PTCA guidelines) executive summary: a report of the American College of Cardiology/American Heart Association task force of practice guidelines (committee to revise the 1993 guidelines for percutaneous transluminal coronary angioplasty) endorsed by the Society for Cardiac Angiography and Interventions. Circulation 2001;103: Virmani R, Farb A, Guigliumi G, et al. Drug-eluting stents: caution and concerns for long-term outcome. Coron Artery Dis 2004;15: McFadden EP, Stabile E, Regar E, et al. Late thrombosis in drug-eluting coronary stents after discontinuation of antiplatelet therapy. Lancet 2004;364: