Early and Mid-Term Results of Drug-Eluting Stent Implantation in Unprotected Left Main
Background— The safety and efficacy of percutaneous coronary intervention in unprotected left main (ULM) coronary arteries are still a matter of debate.
Methods and Results— All consecutive patients who had a sirolimus-eluting stent (Cypher, Cordis, Johnson and Johnson Co) or a paclitaxel-eluting stent (Taxus, Boston Scientific) electively implanted in de novo lesions on unprotected left main were analyzed. Patients treated with a drug-eluting stent (DES) were compared with the historical group of consecutive patients treated with bare metal stent (BMS). Eighty-five patients were treated with DES; 64 had BMS implantation. Patients treated with DES had lower ejection fractions (51.1±11% versus 57.4±13%, P=0.002) and were more often diabetics (21.2% versus 10.9%, P=0.12) with more frequent distal left main involvement (81.2% versus 57.8%, P=0.003). Furthermore, in the DES group, smaller vessels (3.33±0.6 versus 3.7±0.7 mm, respectively; P=0.0001) with more lesions (2.94±1.6 versus 2.25±1.3, P=0.004) and vessels (2.03±0.69 versus 1.8±0.72, P=0.05) were treated with longer stents (24.3±12 versus 15.8±8.6 mm, P=0.0001). Despite the higher-risk patients and lesion profiles in the DES group, the incidence of major cardiac events at a 6-month clinical follow-up was lower in the DES than in the BMS group (20.0% versus 35.9%, respectively; P=0.039). Moreover, cardiac deaths occurred in 3 DES patients (3.5%), as compared with 6 (9.3%) in the BMS group (P=0.17).
Conclusions— In this early experience with DES in unprotected left main, this procedure appears safe with favorable and improved clinical results as compared with historical control subjects with a BMS. A randomized study comparing surgery appears justified at present.
Received July 9, 2004; revision received October 31, 2004; accepted November 5, 2004.
Lesions in the unprotected left main (ULM) are one of the most challenging lesion subsets in percutaneous coronary intervention (PCI). Lesions in the ULM are considered a standard indication for surgical revascularization.1,2 Some studies have demonstrated that stenting of the ULM is feasible and appears to be a promising strategy in selected patients.3–8
Recently, use of a drug-eluting stent (DES), either a sirolimus-eluting stent (SES; Cypher, Cordis, Johnson and Johnson Co) or a paclitaxel-eluting stent (PES; Taxus, Boston Scientific), has been associated with a low restenosis rate.9–13 A single-center observational study recently reported encouraging results with the treatment of left main stenosis, either protected or unprotected, with SES.14,15 No data are available on the use of PES in ULM disease. To provide insight into this controversial field, we report our experience with elective DES implantation in lesions involving the ULM. The aim of this study was to evaluate the safety and the efficacy of DES implantation in the ULM as compared with a bare metal stent (BMS).
All consecutive patients electively treated from April 2002 to April 2004 in the EMO Centro Cuore Columbus and San Raffaele Hospital with implantation of an SES or a PES in the ULM on de novo lesions were analyzed. Patients treated with a DES were compared with the historical group of consecutive patients treated with a BMS from April 1993 to June 2001.6
The decision to perform PCI instead of surgery was considered when 1 of these 2 conditions was present: (1) suitable anatomy for stenting and preference by patient and by referent physician for a percutaneous approach or (2) suitable anatomy for stenting and contraindication for surgery because of the presence of comorbidity.
Coronary angioplasty and DES implantation were performed according to the practice of fully covering the diseased segment.11,12,16 At the start of the procedure, a bolus of unfractionated heparin was administered at 100 IU/kg to achieve an activated clotting time >250 seconds. Glycoprotein IIb/IIIa inhibitors were administered at the discretion of the operator. Patients were asked to start combined antiplatelet therapy with aspirin (≥100 mg QD) and ticlopidine 250 mg BID or clopidogrel 75 mg QD ≥3 days before the procedure and to continue for ≥6 months. A loading dose of 300 mg clopidogrel was administered to those not previously taking thienopyridines.
Coronary angiograms were analyzed with a semiautomated edge-contour-detection computer analysis system (MEDIS QCA CMS, version 4)17 at baseline, after the procedure, and at follow-up.
Clinical follow-up was scheduled for all patients at 1, 6, and 12 months. Angiographic follow-up was scheduled between 4 and 8 months or earlier if noninvasive evaluation or clinical presentation suggested the presence of ischemia.
The events analyzed for follow-up in this study were death, CABG, myocardial infarction (MI; Q wave and non–Q wave), restenosis, target lesion revascularization (TLR), and target vessel revascularization (TVR).
Procedural success was defined as revascularization in the target lesion with <30% residual stenosis according to angiography and with the patient leaving hospital free of any of these events: death, Q-wave MI, or CABG. The exclusion of non–Q-wave MI in the definition of procedural success was done to eliminate difficulties in diagnosing MI after surgery, which is a needed comparison.
Deaths were classified as either cardiac or noncardiac. Deaths that could not be classified were considered cardiac. Non–Q-wave MI was defined as elevation of total creatine kinase 2 times above the upper limit of normal with a positive MB fraction in the absence of pathological Q waves.
Restenosis was defined as >50% luminal narrowing at the segment site (stent and 5 mm proximal and distal) demonstrated at the follow-up angiography, regardless of clinical symptoms of the patient.
TLR was defined as any revascularization performed on the treated segment; TVR was defined as any reintervention performed on the treated vessel. Major cardiac events (MACE) were evaluated as the occurrence of cardiac death, MI, TLR, and TVR. Cumulative MACE was considered the cumulative occurrence of MACE in hospital and during follow-up.
The Parsonnet score18 and European system for cardiac operative risk evaluation (Euroscore)19 were used to stratify the risk of death at 30 days. The patients were stratified as high risk in the presence of a Euroscore >6 and/or Parsonnet >15 and as very high risk if Euroscore was >13 and/or Parsonnet was >20.18,19
The study hypothesis was that implantation of a DES in the ULM is safer and more effective than BMS implantation. The primary end point of the study was a MACE or the occurrence of any of the following within 6 month from the index procedure: cardiac death, MI, TLR, or TVR. All other comparisons and outcomes were considered secondary study objectives. Because of the exploratory nature of the study and the consequent lack of a priori knowledge about effect size, no formal sample size computations were performed.
Data are presented as percentages and mean±SD. In general, differences in proportions were tested with χ2 test or Fisher’s exact test, and differences in location parameters of continuous variables were tested with a Student t test. The relationship of clinical outcomes to covariates was assessed with exact logistic regression models.20 Because of the nonrandomized nature of the study, to minimize any selection bias, the propensity score analysis was performed. After the propensity scores were computed by means of multivariable logistic regression, estimates adjusted for the propensity to be treated with a DES were computed by means of a conditional logistic regression, with stent type as a dummy fixed effect and propensity score as a confounding factor. Because of the relatively small sample size and low event rates, Saddlepoint inference21 was used to obtain a surrogate of exact inference.20
The Statistical Analysis System Program version 8.2 (SAS Institute) and S-Plus HOA’s library21 were used for the analysis.
Baseline clinical, lesion, and angiographic characteristics are summarized in Tables 1 and 2⇓. Eighty-five patients were treated with a DES (41 with an SES, and 44 with a PES); 64 received a BMS. In the DES group, no patients were treated with >1 type of stent.
Compared with BMS patients, those treated with a DES had a lower ejection fraction (51.1±11% versus 57.4±13%, respectively; P=0.002) and were more often diabetic (21.2% versus 10.9%, P=0.12) with more frequent distal left main involvement (81.2% versus 57.8%, P=0.003). High mortality risk scores (Euroscore >6 and/or Parsonnet >15) were present in 38 DES patients (45%), as compared with 22 BMS patients (34%; P=0.23). Moreover, in the DES group, smaller vessels (3.33±0.6 versus 3.7±0.7 mm; P=0.0001) with more lesions (2.94±1.6 versus 2.25±1.3, P=0.004) and vessels (2.03±0.69 versus 1.8±0.72, P=0.05) were treated during the index procedure with longer stents (24.3±12 versus 15.8±8.6 mm, P=0.0001).
In the DES group, an intra-aortic balloon pump was used less frequently (21.2% versus 57.8%, P=0.001), whereas a trend toward a greater use of inhibitors of glycoprotein IIb/IIIa was observed (28.5% versus 15.6%, P=0.07). The use of atherectomy devices before stenting was significantly higher in the BMS than in the DES group (Table 1). In the DES group, the nominal stent diameter was 3.26±0.28 mm, and maximum balloon diameter was 3.39±0.37 mm.
During hospitalization, 2 patients in BMS group had CABG, as compared with 0 in the DES group (P=0.18); in the BMS group, 3 had non–Q-wave MI and 2 had Q-wave MI, as compared with 5 with non–Q-wave MI in the DES group. One patient had repeat PCI in the BMS group, as compared with 0 DES patients. No patient in either group died. Procedural success was significantly higher in the DES than in the BMS group (100% versus 93%, P=0.03).
Follow-Up Cardiac Events
At the 6-month clinical follow-up, the incidence of MACE was significantly lower in the DES group (17 of 85) than in the BMS group (23 of 64) (20.0% versus 35.9%, respectively; P=0.039). In addition, the incidence of cumulative MACE was lower in the DES than in the BMS group (24.7% versus 42.1%, respectively; P=0.03).
Nine patients (14.1%) in the BMS group died, as compared with 3 (3.5%) in the DES group (P=0.03). Cardiac deaths occurred in 3 patients (3.5%) in the DES group, as compared with 6 (9.3%) in the BMS group (P=0.17), resulting in a 63% relative reduction in the occurrence of cardiac death in the patients treated with DES.
Three patients were classified as having cardiac death in the DES group: 1 died after discontinuing antiplatelet therapy 53 days after the procedure because of acute pancreatitis, 1 died as a result of pulmonary edema (this patient had severe aortic and mitral regurgitation, and stent patency was demonstrated 20 days before death), and 1 died of complications related to elective CABG.
At 3 months, while on double antiplatelet therapy, 1 patient in the DES group had thrombosis in the proximal left anterior descending artery with consequent MI treated with thrombolysis and PTCA. TLR was performed in 12 patients (14.1%) in the DES group, as compared with 15 (24.2%) in the BMS group (P=0.13). TVR occurred in 16 DES patients (18.8%), as compared with 19 BMS patients (30.6%; P=0.11). Of the 16 TVRs performed, 12 were due to TLRs and 4 were for disease progression. Restenosis occurred in 12 DES patients (19%) and in 15 BMS patients (30.6%; P=0.18). All restenotic lesions in the DES group were focal (<10 mm in length), whereas the mean lesion length in the restenotic BMS lesions was 6.5±7.6 mm. All restenosis in the DES group occurred in patients with treatment of distal left main and were focal, involving only the circumflex or left anterior descending artery in 58% of the patients (7 of 12). In 2 patients, edge restenosis occurred in the ostial left main, and restenosis occurred in the body of the left main in only 3 patients. Table 3 gives the exact location of these angiographic restenoses and the most relevant patient characteristics. Late loss was significantly lower in the DES than in the BMS group (0.58 versus 1.08 mm, respectively; P=0.01).
At the multivariate analysis, only final maximum pressure (atm) (OR, 0.85; 95% CI, 0.75 to 0.96; P=0.007) correlated to the occurrence of TLR.
After adjustment for the propensity score, which takes into account differences in baseline and lesion characteristics, the reduction in the occurrence of MACE with DES implantation was statistically significant (OR, 0.27; 95% CI, 0.09 to 0.73; P=0.007). Moreover, adjustment for propensity score showed that treatment with a DES led to a significant reduction in TLR (OR, 0.28; 95% CI, 0.09 to 0.81; P=0.01) and TVR (OR, 0.34; 95% CI, 0.12 to 0.91; P=0.03).
High mortality risk scores (Euroscore >6 and/or Parsonnet >15) were present in 38 patients (45%) in the DES group, as compared with 22 (34%) in the BMS group (P=0.23). In this subgroup of patients, there was still a significant reduction in the occurrence of MACE in the DES as compared with the BMS group (15.8% versus 45.4%, P=0.01). Moreover, a trend was observed in the DES group toward a lower occurrence of cardiac death as compared with the BMS group (5.2% versus 22.7%, P=0.08).
Because of the retrospective nature of the subgroup analysis, we cannot exclude the possibility that the statistical significance is due to the occurrence of a type I error inflation.
The main findings of this study are that (1) use of a DES to treat ULM lesions appears feasible with a very high procedural success rate (100%); (2) a significant reduction was seen in the occurrence of MACE at 6 months and in cumulative MACE in DES as compared with BMS patients; (3) compared with prior experience with BMS, with the use of a DES, the incidences of death and cardiac death were lower despite the lack of elimination of restenosis and TLR; and (4) all restenotic lesions in DES patients were focal.
The high procedural success rate further confirms the technical feasibility of treating ULM lesions with the current generation of DES. The lack of serious short-term events supports this statement.
Compared with prior experience with BMS, the patients and lesions treated with a DES had a higher risk profile and complexity testified by a high Parsonnet and Euroscore, a lower ejection fraction, and the presence of smaller vessels and longer lesions with more bifurcations. The fact that only 2 DES patients, as compared with 25 BMS patients, required any adjunctive technology such as directional or rotational atherectomy may have contributed to lower procedural complications. Other groups reported very high procedural success with BMS implantation in ULM lesions.5,8 Our findings are particularly relevant because of the higher lesion complexity as compared with the BMS experience (Table 1), supported by the 81.2% prevalence of patients with distal bifurcational disease.
The higher risk profile of patients treated with DES is likely to have contributed to the marginal improvements in terms of angiographic restenosis and need for revascularization. After adjustment for propensity score, which takes into account the differences in clinical and lesion characteristics of the 2 groups, DES implantation resulted in a lower rate of TLR and TVR than in BMS implantation. An important consideration about the 3 patients treated with a DES who died is that only 1 of them died of an event related to DES implantation; the discontinuation of antiplatelet agents 53 days after stenting was an important factor. Despite the small number of patients studied, a 44% relative reduction in MACE and a 62% relative reduction in cardiac death with the use of a DES as compared with a BMS are an encouraging finding. The antiproliferative properties of DES are demonstrated by a significant lower late loss than that associated with BMS and by the fact that all restenotic lesions, which occurred with DES, were focal (<10 mm). The relatively “benign” features of the restenotic process occurring with DES may have somehow contributed to the lower death rate seen at follow-up as compared with BMS. As prior studies demonstrated,6 stent restenosis in ULM may present with the occurrence of cardiac death.
One element that is still unresolved is the high incidence of restenosis when bifurcations in left main lesions are treated. The issue becomes more difficult to understand because, of the 12 restenotic lesions presenting in patients treated on bifurcations, only 5 occurred at the bifurcation (circumflex ostium, left anterior descending ostium, or both); all the others occurred proximal or distal to the bifurcation. Use of a specific technique for bifurcations lesions was driven by the anatomy of the bifurcations (V technique in very short left mains and other techniques in longer left main trunks). Thus, it may be difficult and questionable to elaborate on outcome according to the technique used. We cannot exclude the possibility that elements specific to the left main artery may play a role in predisposing these lesions to restenosis. We should also take into account that the left main artery is usually a very large vessel despite the small intraluminal measurements and that specific stents with a more favorable metal-to-artery ratio may be needed. A study with data available from intravascular ultrasound evaluations would help to clarify some of these issues.
In our study, a higher final balloon pressure to deploy or postdilate the stent was associated with a lower TLR. We could explain this finding by the correlation found among final balloon pressure and reference diameter, final stent mean luminal diameter, and percent stenosis in the stent (multiple regression R2=0.15, P=0.0005). A higher final balloon pressure may also guarantee a better apposition of the stent struts to the vessel wall and hence more complete drug delivery.
Compared with most surgical series,1,2 the important achievements in our experience with DES are the lack of in-hospital mortality and the low mortality during follow-up. The high angiographic follow-up rate may have contributed to the increased use of revascularization when many of these focal restenotic lesions were not associated with angina and few patients had noninvasive testing for ischemia. There is general concern about leaving a focal angiographic narrowing at the ostium of a major epicardial vessel that may trigger an “oculostenotic reflex.” In most surgical series, angiographic follow-up is rarely performed routinely, and this type of problem is not present.
All the above issues should be taken into consideration when these results are compared with any surgical series and, most importantly, when a prospective study comparing CABG is planned.
Implantation of a DES on ULM lesions appears to be a feasible and safe approach. Compared with prior experience with BMS, there is a reduction in MACE, including death rate, during the 6-month follow-up. The occurrence of angiographic restenosis is usually focal and treatable with repeat PCI. In addition, the finding of a relative low mortality despite a high risk profile in patients treated with DES may allow a randomized study comparing DES with surgery for ULM disease to be performed.
Caracciolo EA, Davis KB, Sopko G, Kaiser GC, Corley SD, Schaff H, Taylor HA, Chaitman BR. Comparison of surgical and medical group survival in patients with left main coronary artery disease: long-term CASS experience. Circulation. 1995; 91: 2325–2334.
d’Allonnes FR, Corbineau H, Le Breton H, Leclercq C, Leguerrier A, Daubert C. Isolated left main coronary artery stenosis: long term follow up in 106 patients after surgery. Heart. 2002; 87: 544–548.
Ellis SG, Tamai H, Nobuyoshi M, Kosuga K, Colombo A, Holmes DR, Macaya C, Grines CL, Whitlow PL, White HJ, Moses J, Teirstein PS, Serruys PW, Bittl JA, Mooney MR, Shimshak TM, Block PC, Erbel R. Contemporary percutaneous treatment of unprotected left main coronary stenoses: initial results from a multicenter registry analysis 1994–1996. Circulation. 1997; 96: 3867–3872.
Takagi T, Stankovic G, Finci L, Toutouzas K, Chieffo A, Spanos V, Liistro F, Briguori C, Corvaja N, Albero R, Sivieri G, Paloschi R, Di Mario C, Colombo A. Results and long-term predictors of adverse clinical events after elective percutaneous interventions on unprotected left main coronary artery. Circulation. 2002; 106: 698–702.
Kelley MP, Klugherz BD, Hashemi SM, Meneveau NF, Johnston JM, Matthai WH Jr, Banka VS, Herrmann HC, Hirshfeld JW Jr, Kimmel SE, Kolansky DM, Horwitz PA, Schiele F, Bassand JP, Wilensky RL. One-year clinical outcomes of protected and unprotected left main coronary artery stenting. Eur Heart J. 2003; 24: 1554–1559.
Park SJ, Park SW, Hong MK, Lee CW, Lee JH, Kim JJ, Jang YS, Shin EK, Yoshida Y, Tamura T, Kimura T, Nobuyoshi M. Long-term (three-year) outcomes after stenting of unprotected left main coronary artery stenosis in patients with normal left ventricular function. Am J Cardiol. 2003; 91: 12–16.
Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M, Colombo A, Schuler G, Barragan P, Guagliumi G, Molnar F, Falotico R. A randomized comparison of a sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med. 2002; 346: 1773–1780.
Colombo A, Drzewiecki J, Banning A, Grube E, Hauptmann KE, Silber S, Dudek D, Fort S, Schiele F, Zmudka K, Guagliumi G, Russell ME, for the group ftTIs. Randomized Study to Assess the Effectiveness of Slow- and Moderate-Release Polymer-Based Paclitaxel-Eluting Stents for Coronary Artery Lesion. Circulation. 2003; 108: 788–794.
Moses JW, Leon MB, Popma JJ, Fitzgerald PJ, Holmes DR, O’Shaughnessy C, Caputo RP, Kereiakes DJ, Williams DO, Teirstein PS, Jaeger JL, Kuntz RE. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med. 2003; 349: 1315–1323.
Schofer J, Schluter M, Gershlick AH, Wijns W, Garcia E, Schampaert E, Breithardt G. Sirolimus-eluting stents for treatment of patients with long atherosclerotic lesions in small coronary arteries: double-blind, randomised controlled trial (E-SIRIUS). Lancet. 2003; 362: 1093–1099.
Arampatzis CA, Lemos PA, Tanabe K, Hoye A, Degertekin M, Saia F, Lee CH, Ruiter A, McFadden E, Sianos G, Smits PC, van der Giessen WJ, de Feijter P, van Domburg R, Serruys PW. Effectiveness of sirolimus-eluting stent for treatment of left main coronary artery disease. Am J Cardiol. 2003; 92: 327–329.
Arampatzis CA, Lemos PA, Hoye A, Saia F, Tanabe K, van der Giessen WJ, Smits PC, McFadden E, de Feyter P, Serruys PW. Elective sirolimus-eluting stent implantation for left main coronary artery disease: six-month angiographic follow-up and 1-year clinical outcome. Catheter Cardiovasc Interv. 2004; 62: 292–296;discussion 297.
Colombo A, Orlic D, Stankovic G, Corvaja N, Spanos V, Montorfano M, Liistro F, Carlino M, Airoldi F, Chieffo A, Di Mario C. Preliminary observations regarding angiographic pattern of restenosis after rapamycin-eluting stent implantation. Circulation. 2003; 107: 2178–2180.
Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999; 16: 9–13.
Davison AC. Approximate conditional inference in generalized linear models. J Royal Stat Soc. 1988; 50: 445–461.