Comparison of bypass surgery with drug-eluting stents in diabetic patients with left main coronary stenosis.

PURPOSE: Several studies have compared the effects of coronary stenting and coronary- artery bypass grafting (CABG) on left main coronary artery (LMCA) disease. However, there are limited data on the long-term outcomes of these two interventions in diabetic patients.
MATERIALS AND METHODS: We evaluated 56 patients with LMCA stenosis who underwent drug-eluting stent (DES) implantation and 116 patients who underwent CABG in a single hospital in China between January 2004 and December 2006. We compared long-term major adverse cardiac events (death; a "serious outcome" composite of death, myocardial infarction, or stroke; and target-vessel revascularization).
RESULTS: In-hospital (30-day) mortality was 0% for the DES group and 3.4% for the CABG group (p=0.31). There was no difference between the two groups in terms of risk of death [hazard ratio for stenting group, 0.49; 95% confidence interval (CI), 0.13-1.63; p=0.55] or risk of serious outcome (hazard ratio for DES group, 1.11; 95% CI, 0.39-1.45; p=0.47). The target-vessel revascularization rate was higher in the DES group than in the CABG group (hazard ratio, 3.67; 95% CI, 1.24-11.06; p=0.018).
CONCLUSION: In this cohort of diabetic patients with LMCA stenosis, there was no difference in composite endpoints between patients receiving DESs and those undergoing CABG. However, stenting was associated with higher rates of target-vessel revascularization than CABG. DES implantation in diabetic patients with LMCA disease was found to be at least as safe as CABG.

INTRODUCTION

As known, coronary-artery bypass grafting (CABG) has been considered standard therapy for patients with left main coronary artery disease and is recommended by current practice guidelines.1,2 However, these guidelines have been challenged by the widespread use of coronary stents in combination with improved antiplatelet treatment and by the recent introduction of drug-eluting stents (DESs), which virtually eliminate the recurrence of acute coronary occlusion and also limit the occurrence of restenosis.3,4 Percutaneous coronary intervention (PCI) is therefore being used increasingly in patients with left main coronary artery (LMCA) stenosis.5

In patients with coronary artery diseases, diabetes mellitus (DM) increases the risk of cardiac mortality approximately two- to four-fold.6,7 Although the acute outcomes after PCI are similar in patients with and without diabetes, because of excessive neointimal proliferation, diabetic patients have higher rates of restenosis and repeat revascularization after PCI with or without stenting.8,9 Preliminary results from registries show that the implantation of DESs for unprotected LMCA disease in patients without diabetes is a feasible and safe approach.10 However, until now, no study has analyzed the treatment of LMCA lesions with PCI versus CABG in diabetic patients. The purpose of this study was to compare treatment of LMCA lesions with PCI and DES implantation versus surgical revascularization in patients with DM during their hospital stay (30 days) and long-term follow-up.

MATERIALS AND METHODS

Study population

All diabetic patients with LMCA disease treated with PCI with DES implantation or CABG between January 2004 and December 2006 at Beijing's Anzhen Hospital (Capital University of Medical Sciences, Beijing, China) were considered for inclusion in this study. In all cases, the selected revascularization approach appeared to be suitable for guaranteeing complete revascularization. We excluded patients who had undergone previous CABG or who underwent concomitant valvular or aortic surgery. We also excluded those who had primary cardiomyopathy or myocardial infarction (MI) with ST-segment elevation, or who presented with cardiogenic shock. During the same period, one of the patients who underwent PCI received a bare-metal stent and was therefore excluded from the study. DM was diagnosed on the basis of a history of using glucose-lowering medications or insulin, or a fasting plasma glucose concentration of ≥126 mg/mL.11 All patients were considered to have type 2 DM. The study protocol was approved by the appropriate local institutional ethics committee, and written informed consent was obtained from all patients.

Revascularization procedure

Patients underwent PCI instead of CABG due to their own or their physician's preference or the high risk associated with CABG. Previous studies described about the methods of stent implantation for patients with LMCA disease.12,13 Standard interventional techniques were processed for all procedures. According to the operators' discretion, the use of pre-dilation, an intra-aortic balloon pump, or intravascular ultrasound and the choice of the specific type of DES were applied for each patient treatment. Periprocedurla anticoagulation and antiplatelet therapy followed standard regimens. All patients undergoing stenting were prescribed clopidogrel for at least 12 months. Treatment beyond this duration was administered at the discretion of the physician (clopidogrel 79.2% at the end of follow-up in the PCI group). Aspirin was prescribed indefinitely for all patients who underwent either PCI with DES or CABG treatment. Surgical revascularization was performed with the use of standard bypass techniques.14 Whenever possible, the internal thoracic artery was used preferentially for revascularization of the left anterior descending artery. Complete revascularization was performed when possible with arterial conduits or saphenous vein grafts. A recognized standard of post-interventional care was recommended to the patients.15 Other procedural details are given in Table 1.

Table 1

Procedural Characteristics of Patients Treated with CABG and PCI

PCI, percutaneous cardiac intervention; CABG, coronary-artery bypass grafting; LMCA, left main coronary artery disease.

Percentages may not total 100 because of rounding.

Follow-up, end points, and definitions

Clinical follow-up after PCI with DES and after CABG was stopped in March 2008, and major adverse cardiac events (MACE) that occurred during the follow-up period were studied. During the 30 days of in-hospital follow-up, renal events were also evaluated by biological analysis and calculated for major adverse cardiac cerebrovascular and renal events (MACCRE). Routine angiographic follow-up was recommended by the operators in PCI-group patients (22/56, 39.3%) 6 to 12 months after the procedure. However, patients who were at high risk of procedural complications of angiography and had no symptoms or signs of ischemia, as well as patients who declined to comply with this recommendation, did not undergo routine follow-up angiography. For CABG-group patients, angiographic follow-up was recommended only if there were ischemic symptoms or signs during follow-up (29/116, 25.0%). This is consistent with previous studies in which a low threshold for control coronary angiography was maintained.16

The endpoints of the study were death, the "serious outcomes" composite (death, Q-wave myocardial infarction, or stroke), and target-vessel revascularization (TVR). Clinical events were assessed annually by mail and/or telephone contact with the patients. The medical records of those patients who reported events were collected and the events adjudicated by the patient's physician. Death was defined as death from any cause. Q-wave MI was defined as documentation of a new abnormal Q wave after the index treatment. Stroke, as indicated by neurologic deficits, was confirmed by a neurologist on the basis of imaging studies. TVR was defined as repeat revascularization of the treated vessel, including any segments of the left anterior descending artery and the left circumflex artery. Both PCI with DES and repeat CABG were judged as revascularization.

Ontario score calculation

The Ontario Province Risk system for cardiac surgery operative-risk evaluation (Ontario score, available at http://www.sfar.org/scores2/ontario2.htm) was used to stratify the risk of death during hospitalization in patients undergoing cardiac surgery.17 Relevant factors that contributed to the Ontario score calculation include age, sex, ejection fraction, urgency of surgery, type of surgery, and repeat operation.

Statistical analysis

Data are presented as mean±standard deviation (SD). We compared the baseline covariates between the DES and CABG groups. Comparisons of the continuous baseline characteristics were made using an unpaired t-test. Comparisons of categorical baseline characteristics were made using χ2 tests. Statistical significance and the effects of both therapies on the outcomes of in-hospital (up to 30 days) or long-term follow-up were estimated. The Kaplan-Meier method was used to estimate the incidence of the clinical endpoints, death (overall survival), the "serious outcomes" composite (death, MI or stroke), and TVR, during the follow-up period. The Log-rank test was used to compare the Kaplan-Meier results.

Multivariable analysis was performed using Cox proportional hazards models adjusted for age, sex, body mass index (BMI), left ventricular ejection fraction, and Ontario scores.18 Cox proportional hazard regression analysis was performed to calculate the hazard ratios and 95% confidence intervals (CIs) for the clinical outcomes.

A p-value of <0.05 was considered to indicate statistical significance. Data processing and statistical analysis was performed using the SPSS 13.0 statistical program (SPSS Inc., Chicago, IL, USA).

RESULTS

Baseline characteristics

Between January 2004 and December 2006, 172 patients with LMCA disease met the criteria for inclusion. Fifty-six of these patients received PCI with DES and 116 received CABG. A total of 168 patients (97.7%: 52 DES, 116 CABG) had unprotected LMCA disease.

In the DES group, 53 patients received rapamycin-eluting stents (94.6%) and three received zotarolimus-eluting stents (5.4%). The mean (±SD) total length of stents was 27.8±27.1 mm, the total stent diameter was 3.4±0.6 mm, and the maximum stent implantation pressure was 17.0±1.6 atm. The mean total number of stents implanted in a patient (including left and other vessels) was 2.8±1.6.

In the CABG group, 108 patients (93.1%) underwent off-pump surgery. All of them received at least one arterial conduit, which was used in 115 of the patients (99.1%) towards the revascularization of the left anterior descending artery. One patient underwent concomitant left ventricular aneurysmectomy and one underwent pericardial cystectomy. The mean number of grafts used was 3.0±0.8 (17.2% venous and 82.8% arterial).

The baseline characteristics of the study patients in each group are shown in Table 1. The numbers of patients in the two groups were well balanced with regard to most of the baseline demographic and clinical characteristics. Patients undergoing PCI had a significantly higher prevalence of family history of hypertension and of diabetic, coronary artery, and cerebral disease (p<0.001); they were also significantly more likely to have had previous coronary bypass grafting than those receiving CABG (p=0.005). The PCI group included a significantly larger number of poorly controlled diabetic patients (hemoglobin A1c, 8.1%±2.0% vs. 7.0%±1.7%, p=0.016) (Table 2). Patients undergoing CABG had significantly lower ejection fractions (p=0.010) and Ontario scores (p=0.04), and had significantly more instances of double- or quadruple-vessel disease and involvement of the right coronary artery (p<0.05). There was no significant difference in the proportion of patients in the two groups receiving insulin treatment (39.3% for the PCI group and 38.8% for the CABG group, p=0.950). There were no significant differences in any other preoperative characteristics between the PCI and CABG groups (p>0.05) (Table 2).

Table 2

Baseline Patient Characteristics

PCI, percutaneous cardiac intervention; CABG, coronary-artery bypass grafting.

Percentages may not total 100 because of rounding. Data are shown as mean±SD for continuous variables and absolute numbers (percentages) for dichotomous variables.

P values are based on the unpaired t-test for continuous variables and on the χ2 tests for categorical variables.

In-hospital events

The rates of MACCRE and acute MI (AMI) were significantly higher in the CABG group (31.0% and 29.3%, respectively) than in the PCI group (both 8.9%, p<0.001) (Table 3), during the in-hospital (up to 30-day) follow-up period. In the CABG group, in-hospital cardiac and stroke death occurred in four patients with acute coronary syndrome. One patient, who underwent concomitant pericardial cystectomy, died from low cardiac output syndrome; two patients died from fatal AMI involving postoperative stroke; and one died from cardiac tamponade. The rates of cardiac tamponade, acute heart failure, permanent pacemaker implantation, ventricular tachycardia and fibrillation, postoperative pneumothorax, shock, dialysis requirement, repeat thoracotomy for bleeding or suppuration, and vascular hematoma requiring repair were higher in the CABG group, but the differences between the two groups were not significant (Table 3). The duration of the post-procedural hospital stay was significantly longer in the CABG group than in the PCI group (p<0.001) (Table 3).

Table 3

In-Hospital Events

PCI, percutaneous coronary intervention; CABG, coronary-artery bypass grafting.

Percentages may not total 100 because of rounding. P values are based on the unpaired t-test for continuous variables and on the χ2 tests for categorical variables.

Long-term outcomes

The median follow-up was 28.5 months (interquartile range, 18.5-29.5) in the PCI group and 28.4 months (17.9-38.9) in the CABG group (Table 4). Complete follow-up data for major clinical events were obtained in 97.4% of the subjects overall (98.2% for the PCI group and 96.6% for the CABG group). During follow-up, 10 patients (6.1%) died; seven of these (one from the PCI group and six from the CABG group) died of cardiovascular causes. Twelve (7.0%) patients (14.5% in the PCI group and 3.7% in the CABG group, p=0.022) underwent TVR. The incidence rates of MACE and AMI tended to be greater in the PCI group (23.6% and 10.9%, respectively) than in the CABG group (13.0% and 8.3%, p>.05), and the rate of cerebrovascular events tended to be greater in the CABG group (5.6%) than in the PCI group (0%, p=0.098) (Table 4). There was no significant difference between the PCI and CABG groups in terms of risk of death (hazard ratio for PCI group, 1.33; 95% CI, 0.68-2.59; p=0.27) or risk of serious outcome (hazard ratio for PCI group, 1.37; 95% CI, 0.75-2.67; p=0.59). The rate of TVR was significantly higher in the PCI group than in the CABG group (hazard ratio, 4.89; 95% CI, 1.78-13.02; p<0.001). Following adjustment for age, sex, BMI, EF, and Ontario score, there was still no significant difference between the PCI and CABG groups in terms of risk of death (hazard ratio for PCI group, 0.49; 95% CI, 0.13-1.63; p=0.55) (Fig. 1A, Table 5) or risk of serious outcome (hazard ratio for PCI group, 1.11; 95% CI, 0.39-1.45; p=0.47) (Fig. 1B, Table 5). The rate of TVR was still significantly higher in the PCI group than in the CABG group (hazard ratio, 3.67; 95% CI, 1.24-11.06; p=0.018) (Fig. 1C, Table 5). On other hand, the rate of statins-treatment tended to be higher for the PCI group (51.2%) than for the CABG group (39.8%, p=0.39), and the rate of clopitogrel treatment was significantly higher for the PCI group (79.2%) than for the CABG group (17.8%, p<0.001).

Table 4

Follow-Up Outcomes

PCI, percutaneous coronary intervention; CABG, coronary-artery bypass grafting.

Percentages may not total 100 because of rounding. Data are shown as mean±SD for continuous variables and absolute numbers (percentages) for dichotomous variables. P values are based on unpaired t-tests for continuous variables and on χ2 tests for categorical variables. Data are shown as mean±SD for continuous variables and absolute numbers (percentages) for dichotomous variables.

Fig. 1

Kaplan-Meier curves of the PCI (n=56) and the CABG (n=116) groups for death, composite risk of serious outcomes, and the rate of target-vessel revascularization (TVR). There were no differences in death and composite risk between the two groups (A and B), whereas the rate of TVR was higher in the PCI group than in the CABG group (C). PCI, percutaneous coronary intervention; CABG, coronary-artery bypass grafting; MI, myocardial infarction.

Table 5

Hazard Ratios for Clinical Outcomes after Stenting as Compared to after CABG

CABG, coronary-artery bypass grafting; MI, myocardial infarction; BMI, body mass index; CI, confidence interval.

Values in parentheses are 95% CIs. CIs were estimated from Cox proportional-hazard regression with or without covariate adjustment.

DISCUSSION

During the in-hospital (up to 30-day) follow-up, DES implantation for diabetic patients with LMCA lesions was shown to be safe and was associated with a low rate of in-hospital events (including MACCER and AMI). Our long-term observations showed that the risks of death and serious outcomes (death, Q-wave MI, or stroke) were similar between the PCI and CABG groups. In contrast, the rate of TVR was significantly higher in the PCI group than in the CABG group.

Recent observational studies have reported that patients undergoing CABG have a significantly higher 30-day mortality rate than those receiving PCI.19,20 Our observations here showed that the patients in the CABG group tended to have higher rates of MACCRE and AMI than CABG patients. Ninety-four percent of the patients in our CABG group were found to have coronary stenosis involving more than one vessel in addition to the LMCA segment (p<0.05 v. PCI group), and the CABG group was found to have significantly lower Ontario scores and ejection fractions; these results were similar to those of previous studies.19-21 It is likely that the patients who underwent PCI were those for whom the surgical risk was considered prohibitive, or those who were candidates for CABG but in whom PCI was considered feasible and relatively low-risk. Differences in the clinical baseline demographics as well as the extent of coronary stenosis might have caused the greater proportion of in-hospital events in the CABG group.

The assumption that CABG surgery is the best therapy for unprotected LMCA stenosis is based on the results of many historical studies performed two decades ago22 and on the disappointing results registered in early experience with PCI.23,24 Another large observational study, published before the development of the DES, suggested that patients with LMCA disease did significantly better with CABG than with PCI.25 Although this was a risk-adjusted analysis, patient-selection factors probably influenced the results. In contrast, favorable initial outcomes after LMCA intervention using DESs have been reported in select low-risk patients.26 DES implantation has recently been shown to be safe with regard to acute and midterm complications and the prevention of restenosis in patients with unprotected LMCA stenosis.10 More recent observational studies have shown similar mortality rates and similar risks of major cardiac and cerebrovascular events in both patients receiving DESs and those undergoing CABG.27,28 These studies have all been limited by a lack of consideration of coronary disease risk factors (including diabetes and hypertension), a limited duration of follow-up, and selective use of PCI in patients considered to be poor candidates for CABG.

Despite endeavors to decrease in-stent restenosis after LMCA intervention with BMSs, such as by the use of aggressive debulking atherectomy, the restenosis rate remains high at 20% to 30%.26,29,30 Generally, before the development of the DES, BMS-related in-stent restenosis was widely recognized as the most important reason for which bypass surgery was the first choice for treating LMCA stenosis. On the other hand, DES implantation for LMCA lesions has recently been shown to reduce rates of in-stent restenosis and the rates of target-vessel revascularization in LMCA lesions, or both.10,11 However, recent observational studies have reported that the restenosis rate with DESs was still higher in patients with DM than in those without DM.31 The FREDOM trial was designed to define the optimal revascularization strategy for diabetic patients with multi-vessel coronary disease.32 Here, we observed that the rate of TVR was higher in diabetic patients who underwent PCI than in patients undergoing CABG, although we did not evaluate the impact of diabetic coronary risk factors on in-stent restenosis. Given the limitations in the trials to date and the inherent risk of DES restenosis, results from randomized trials are still needed to definitively establish the role of DES implantation in place of the reference treatment surgery for diabetic patients with LMCA disease.

A number of single- and multi-center cohort studies have shown that DES stenting is similar to CABG for patients with unprotected LMCA disease in terms of short- and long-term mortality and rates of death, Q-wave MI, or stroke.33-36 Notably, Sheiban and other groups reported that EDS stents showed favorable early and long-term results in selected diabetic and nondiabetic individuals undergoing PCI for unprotected LMCA disease.37,38 Furthermore, recent subgroup analyses by Banning and colleagues suggest that the composite safety end point (death/stroke/MI) is comparable between the two treatment options for both diabetic and nondiabetic patients, even though the one-year major adverse cardiac and cerebrovascular event rate is higher among diabetic patients with left main and/or 3-vessel disease treated with paclitaxel-eluting stents compared with CABG, due to the increase in repeat revascularization.39 We observed no increases in the risk of death or of composite serious outcomes. These findings suggest the early and long-term safety and efficacy of stenting versus CABG for unprotected LMCA disease in patients with or without DM. This notion is further supported by a recent study showing that the adjusted risks of death and of the composite (death, Q-wave MI, or stroke) were similar in diabetic patients who received DES and those who underwent CABG, and that diabetes had a minimal prognostic impact on the long-term treatment effects in patients who underwent DES or CABG.40 It should be noted that not only non-diabetic but also diabetic patients treated with DES show higher rates of TVR as compared to those treated with CABG. Effectiveness and durability are concerns related to stenting in coronary atherosclerotic stenosis, as is safety with regards to potential in-stent thrombosis in DES implantation. Although one angiographic study that lasted for up to four years suggested that DES implantation has a low in-stent restenosis rate,41 most other studies have suggested that it has a high TVR rate,3,21 and it is not known whether PCI with DESs further increases the TVR rate. In CABG, the grafts usually bypass the proximal epicardial coronary segment, in which more obstructive lesions occur and progress;42 the alternative route provided by the surgery can perfuse the distal myocardial territory,8 thus protecting, to some degree, against progressive atherosclerosis.

There are inherent limitations to our study. First, this was a single-center retrospective study; therefore, the results are not adequate to draw definite wide-ranging conclusions. Second, there were baseline clinical differences between the study groups, highlighting the inevitable bias related to patient selection. This was similar to previous studies of clinical outcomes following CABG and PCI for LMCA stenosis.43,44 Patients who underwent CABG had more multi-vessel lesions and a lower Ontario score and cardiac output, while patients who underwent PCI had a higher incidence of poorly controlled diabetes and a higher prevalence and family history of cardiovascular disease with previous coronary bypass grafting. Third, the number of study patients was too small to generalize our results to all diabetic patients with LMCA lesions.

In conclusion, these results provide new information regarding the safety and effectiveness of DES implantation in diabetic patients with LMCA lesions. These data highlight the need for a large, long-term, multi-center randomized study comparing DES implantation and bypass surgery in diabetic patients with LMCA disease.