Author + information
- Received September 14, 2012
- Accepted September 27, 2012
- Published online January 1, 2013.
- Stephan Blazek, MD⁎,
- David Holzhey, MD†,
- Camelia Jungert, BSc⁎,
- Michael A. Borger, MD, PhD†,
- Georg Fuernau, MD⁎,
- Steffen Desch, MD⁎,
- Ingo Eitel, MD⁎,
- Suzanne de Waha, MD⁎,
- Philipp Lurz, MD⁎,
- Gerhard Schuler, MD⁎,
- Friedrich-Wilhelm Mohr, MD† and
- Holger Thiele, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Holger Thiele, Deputy Director, Department of Internal Medicine/Cardiology, University of Leipzig–Heart Center, Strümpellstrasse 39, 04289 Leipzig, Germany
Objectives The aim of this prospective, randomized trial was to assess the 10-year long-term safety and effectiveness of percutaneous coronary intervention (PCI) and minimally invasive direct coronary artery bypass surgery (MIDCAB) for the treatment of proximal left anterior descending (LAD) lesions.
Background Long-term follow-up data comparing PCI and MIDCAB surgery for isolated proximal LAD lesions are sparse.
Methods Patients with significant isolated proximal LAD stenoses were randomized either to PCI with bare-metal stents (n = 110) or MIDCAB (n = 110). At 10 years, data were obtained with respect to the primary endpoint (death, myocardial infarction, target vessel revascularization). Angina was assessed by the Canadian Cardiovascular Society classification.
Results Follow-up was conducted for 212 patients at a median time of 10.3 years. There were no significant differences in the binary primary composite endpoint (47% vs. 36%; p = 0.12) and hard endpoints (death and infarction) between PCI and MIDCAB. However, a higher target vessel revascularization rate in the PCI group (34% vs. 11%; p < 0.01) was observed. Clinical symptoms improved significantly from baseline and were similar between both treatment groups.
Conclusions At 10-year follow-up, PCI and MIDCAB in isolated proximal LAD lesions yielded similar long-term outcomes regarding the primary composite clinical endpoint. Target vessel revascularization was more frequent in the PCI group.
- bypass surgery
- left anterior descending artery
- long-term follow-up
- percutaneous coronary intervention
Previous trials have demonstrated that clinical prognosis in patients with untreated high-grade proximal stenosis of the left anterior descending (LAD) artery is poor due to the large territory of myocardium at risk (1,2).
Established treatment options are conventional bypass grafting, minimally invasive direct coronary artery bypass (MIDCAB) surgery, and percutaneous coronary intervention (PCI) either with bare-metal stents (BMS) or drug-eluting stents (DES) (3–6). In randomized trials (using BMS) and a meta-analysis comparing stenting versus MIDCAB, a significantly higher reintervention rate after PCI and similar results for mortality and myocardial infarction at 6-month follow-up have been demonstrated (4,7,8). At 5-year follow-up, these results were confirmed (9).
To further elucidate advantages or disadvantages of PCI against MIDCAB in the long term, we report the 10-year follow-up of the largest randomized, prospective trial comparing MIDCAB surgery with PCI using BMS for treatment of patients with isolated lesions of the proximal LAD.
The detailed design of this trial has been described previously (4). In summary, patients with isolated high-grade lesions (≥75% diameter stenosis on visual assessment, no total occlusion, no bifurcation lesion of the left main and the first diagonal branch) of the proximal LAD were included. A cardiac surgeon and a cardiologist had to find a consensus on patient eligibility. Balanced randomization was performed after written informed consent and patients were assigned to stenting or surgery.
Exclusion criteria were acute coronary syndromes, additional significant coronary lesions, significant valvular heart disease, and previous interventional or surgical treatment for coronary artery disease. The study was approved by the Institutional Ethics Committee.
Stenting with BMS was performed according to standard practice (4). Patients received aspirin 100 mg/day indefinitely and ticlopidine or clopidogrel for 4 weeks (following a loading dose the day before the procedure).
Minimally invasive direct coronary artery bypass surgery
MIDCAB surgery was performed through a left anterolateral minithoracotomy on the beating heart without cardiopulmonary bypass (4). The left internal thoracic artery was used as bypass graft. Aspirin (100 mg/day) treatment was recommended indefinitely.
Data were collected by a structured patient interview at 5 years—results have been published previously (9)—and again 10 years after randomization. Reported clinical events were confirmed by contact with the general practitioner, referring cardiologist, and/or the treating hospital. Death was defined as death from any cause. Death was regarded as cardiac in origin unless obvious noncardiac causes could be identified. In case of any doubt, death was counted as cardiac. For in-hospital periprocedural myocardial infarction, the previously chosen definition was used with increase of the creatine kinase-myocardial band (CK-MB) value 3× the normal value or if the ratio of CK-MB to total CK exceeded 0.1 (4). At follow-up, the standard universal definitions of myocardial infarction criteria were applied (10).
Any new revascularization by either PCI or bypass grafting was further divided into target lesion revascularization, target vessel revascularization (TVR), and nontarget vessel revascularization categories. For stent thrombosis definition, the Academic Research Consortium criteria were applied (11).
The initial study protocol at 6-month follow-up mandated a complete clinical work-up, including symptom-limited exercise stress test and coronary angiography. For asymptomatic patients, no further scheduled routine noninvasive or invasive studies were performed at long-term follow-up. Repeated interventions were clinically driven and performed in case of recurrence of angina and/or a positive stress test. Quantitative angiography was not performed in these patients, except for the mandatory initial 6-month angiography. The choice of all recurrent revascularization procedures was left to the discretion of the investigator.
All events were adjudicated by an event monitoring committee consisting of an experienced cardiologist and cardiovascular surgeon.
The primary composite endpoint was freedom from major adverse cardiac events (MACE), defined as death from any cause, myocardial infarction, and the need for repeated TVR. Secondary endpoints were each individual component of the primary endpoint. For the combined MACE rate, to avoid double counting of patients with more than 1 event, each patient contributed only once to the composite MACE endpoint and the first occurring event was counted regardless of severity. The clinical status was assessed by the Canadian Cardiovascular Society's classification, and the need for antianginal drugs at 10-year follow-up was documented.
The initial sample size chosen to achieve 95% statistical power with a type I error of 5% by use of 2-sided Fisher exact test was based on the assumption that 30% of the patients treated by stenting with BMS and 9% treated by MIDCAB surgery would reach the combined primary endpoint. Accounting for losses to follow-up, 10 more patients in each group were included, resulting in 220 randomized patients. All analyses were performed according to the intention-to-treat principle. Patients lost to follow-up were included in the analysis until their last known status. Continuous parameters were described as mean ± SD or as median (interquartile range [IQR]) if not normally distributed. Categorical variables were expressed as number and percentage of patients. Differences between the treatment groups were assessed by Fisher exact test or the chi-square test for categorical variables with nominal scales. For continuous data with normal distribution, unpaired Student t test was used.
Additionally, exploratory landmark analyses using the Kaplan-Meier method were performed for the primary composite endpoint, myocardial infarction, and TVR. The landmark was chosen at 7 months; therefore, the first analysis covers the period from 0 to 7 months and the second the period from 7 months to 10 years. The rationale for choosing the 7-month landmark was that it became evident that most TVR events in the PCI group occurred early during the first months of follow-up. Any patient with an event before the 7-month landmark was excluded from further analysis for the second period regarding the composite endpoint. For the TVR landmark analysis, those patients with prior TVR or death were excluded.
The statistical analyses were conducted using SPSS (version 17.0, SPSS Inc., Chicago, Illinois). A 2-tailed p value <0.05 was considered statistically significant.
Between June 1997 and June 2001, 220 consecutive patients with isolated proximal LAD stenosis were randomized to either surgery (n = 110) or stenting (n = 110). Initially, 71% of patients had type B2 or C lesions and the percentage diameter stenosis was 82.8 ± 7.8%. All patients received the assigned treatment. The 10-year long-term follow-up was completed for 212 patients (96%). Despite intensive effort, long-term follow-up information could not be obtained for 8 patients. For survival analysis, data until their last known status were used; median follow-up time was 10.3 years (IQR: 9.4 to 11.4 years) (Fig. 1).
Baseline characteristics were not significantly different between the treatment groups and have been published previously (4).
Primary composite endpoint
In the binary event analysis at 10 years, there was no significant difference in the incidence of the combined primary endpoint in PCI versus MIDCAB (relative risk: 0.81, 95% confidence interval [CI]: 0.62 to 1.05; p = 0.12) (Table 1). However, when a time-to-event analysis was used, the difference in event rates was statistically significant (hazard ratio [HR]: 0.65, 95% CI: 0.43 to 0.98; p = 0.04) (Fig. 2A).
Landmark analysis showed that this difference was mainly driven by a higher event rate in the PCI group in the first months (period 0 to 7 months: HR: 0.31, 95% CI: 0.17 to 0.57; p < 0.01) (Fig. 2B). In contrast, for the period from 7 months to 10 years, the difference between the event curves was no longer statistically significant (HR: 1.29, 95% CI: 0.71 to 2.3; p = 0.40) (Fig. 2C).
The binary incidence of each individual endpoint at 10-year follow-up is shown in Table 1.
In the PCI group, the need for TVR was significantly higher compared with that of the MIDCAB group at 10 years (34% vs. 11%, p < 0.01) (Table 1). This was also true when a time-to-event method of analysis was used (Fig. 3A).
Landmark analysis revealed that this difference was largely caused by a higher TVR rate in the first analysis period from 0 to 7 months (HR: 0.25, 95% CI: 0.13 to 0.47; p < 0.01) (Fig. 3B), whereas in the second period from 7 months to 10 years, the difference in event curves was no longer significant (HR: 0.71, 95% CI: 0.22 to 2.23; p = 0.53) (Fig. 3C).
Some patients had multiple TVRs. The mean number of TVR events per patient in the stenting group who had repeated revascularization was 1.37 ± 0.69 (26 patients had 1, 8 had 2, 1 had 3, and 1 had 4 TVRs). Overall, in the interventional group, 49 TVRs were performed in 36 patients (30 [83%] PCIs, 4 [11%] surgical, 2 [6%] PCIs and surgical).
In the surgical group, 8% required TVR at 6-month follow-up; additional revascularization was necessary in 1% at 5-year follow-up and increased by another 1% at long-term follow-up. The mean number of TVRs per patient in the MIDCAB group was 1.18 ± 0.60. A total of 13 TVRs (9 [69%] PCIs and 4 [31%] surgical) were performed in 11 patients in the surgery group (10 patients had 1 and 1 patient had 3 TVRs).
Proximal LAD restenosis or bypass dysfunction with a necessity for target lesion revascularization occurred in 33 patients in the PCI group and 9 in the MIDCAB group (30% vs. 9%, p < 0.001).
Nontarget vessel revascularization procedures at 10-year follow-up were necessary in 19 patients in the stenting group and 12 in the MIDCAB group (17% vs. 11%, p = 0.24), reflecting the progressive nature of coronary artery disease.
Death and Myocardial Infarction
All-cause mortality did not differ between the 2 treatment groups at 10-year follow-up (PCI 23% vs. MIDCAB 23%, p = 1.00) (Table 1). In addition, cardiac death rates were similar (Table 1). The cardiac deaths in the stenting group included 1 death from unknown cause, 2 infarctions, and 6 deaths due to congestive heart failure. The 10 cardiac deaths after surgery included 2 perioperative deaths and 1 death at follow-up from unknown cause. Three patients died of heart failure and 4 of acute myocardial infarction.
The rate of myocardial infarctions was not statistically different between stenting and surgery at 10-year follow-up (5% vs. 11%, p = 0.08). Landmark analysis showed, despite the numerical higher event rates in the MIDCAB group, no statistically significant differences in the 0 to 7 months period (HR: 0.59, 95% CI: 0.14 to 2.48; p = 0.49) or 7 months to 10 years period (HR: 0.29, 95% CI: 0.06 to 1.39; p = 0.10).
Except for 2 definitive stent thromboses at 6-month follow-up, no subacute or late stent thromboses were observed at longer follow-up in the PCI group.
Anginal Symptoms and Antianginal Medication
After PCI, the median Canadian Cardiovascular Society classification score improved from 2 (IQR: 2 to 3) to 0 (IQR: 0 to 1) (p < 0.01) at 10-year follow-up, and 69% of the patients were completely free from angina. In the surgery group, the median angina class improved from 2 (IQR: 2 to 3) to 0 (IQR: 0 to 1) (p < 0.01 for the comparison with baseline), with 65% of patients free of anginal symptoms (p = 0.77 for the intergroup comparison).
The rate of antianginal medication was not different at 10-year follow-up (20% vs. 19%, p = 0.99). Use of additional cardiac medication was similar between the 2 treatment groups (Table 2).
To the best of our knowledge, this prospective, randomized study is the first to compare long-term results at 10 years in patients with isolated lesions of the proximal LAD treated by either PCI with BMS or MIDCAB surgery. No significant differences in the composite endpoint and individual “hard” endpoints, such as death and myocardial infarction, could be found between both treatment groups. However, consistent with short- and midterm follow-up results, there was a significantly lower TVR rate in favor of the surgery group. In the time frame of the landmark analysis from 7 months to 10 years, no significant differences in the primary composite endpoint and secondary endpoints could be observed.
Several clinical trials have demonstrated an initial higher complication rate in cardiac surgery due to perioperative risk, which at 6-month follow-up is offset by the stenting group due to the higher rate of TVR (7,8). At longer follow-up, the MACE rates align, which is reflected in Figure 2. Notably, the higher TVR rate in the stenting group was primarily driven during the first 6 months, as demonstrated in Figure 3. An important reason for the higher TVR rate in the PCI group might have been the use of BMS and the mandatory follow-up angiography at 6 months. Routine angiographic follow-up leads to additional revascularizations in comparison with sole clinical follow-up (12,13).
Previous studies have shown restenosis rates up to 50% (7,8) in proximal LAD stenoses, and proximal LAD lesion location has been identified as a risk factor for restenosis after BMS implantation (14). Since the introduction of DES, restenosis rates have been dramatically reduced in comparison with BMS for all lesion subsets in short- (15) as well as midterm follow-up (6,16). Importantly, with DES, the proximal LAD lesion location is no longer an isolated risk factor for restenosis (5,17,18). Moreover, second-generation DES seem to reduce the risk of in-stent restenosis at any location even further in comparison with first-generation DES, with target lesion revascularization rates ranging from only 2.2% to 3.1% versus 4.6% to 5.5% (16,19). MACE rates might also be lower by using second-generation compared with first-generation DES (5% to 6% vs. 9%) (16,19) as well as mortality at 2-year follow-up (16). Thus, the use of DES might close the gap in repeat revascularization compared with MIDCAB for significant proximal LAD disease. Of note, in the present study, there was a high prevalence of type B2 and C lesions, which are well known for their higher in-stent restenosis rates in comparison with type A and B1 lesions (20,21).
It should also be noted that the present study was performed with earlier generations of coronary artery stabilization devices for MIDCAB surgery and reflects early experience with this technique. We have previously demonstrated that a learning curve for MIDCAB surgery exists, with fewer complications occurring with increasing operator experience (22). Therefore, TVR rates may be lower for both PCI and MIDCAB in a more contemporary setting.
The effect of PCI with DES versus MIDCAB surgery in proximal LAD lesions has been evaluated in only 1 randomized clinical trial showing noninferiority of DES in comparison with MIDCAB surgery at 12-month follow-up (5). However, the value of each treatment option needs to be further clarified in future randomized trials with longer follow-up.
The lack of a significant difference in mortality or myocardial infarction rates between PCI and surgery observed in the present study is in line with other trials in patients with isolated proximal LAD coronary artery lesions (23,24).
Relief of angina was preserved at 10-year long-term follow-up. In the PCI group, 69% of the patients were free from angina, and in the MIDCAB group, 65% were also free of angina. Similar results have been published for proximal LAD lesions in other trials ranging from 61% to 94% for PCI and also for surgery (24–26).
Several studies have demonstrated the long-term patency of internal mammary arterial grafting and its superiority compared with vein grafts anastomosed to the LAD in terms of survival, freedom from infarction, angina, and repeated revascularization (22,27). There is evidence that the duration of benefit extends for >10 years; therefore, long-term follow-up of revascularization trials is essential to further evaluate the magnitude of benefit over time (27,28).
Some limitations of the present trial need attention. First, some events might have been missed due to 7 patients being lost to long-term 10-year follow-up. Also, results are from a single high-volume tertiary care center and may not be generalizable. Due to the trial design, blinding was not possible. Finally, both PCI and MIDCAB techniques have improved over time, and DES use is the current gold standard in patients undergoing PCI of the proximal LAD. Therefore, the results of this long-term analysis may not be applicable to current best practice.
At the 10-year follow-up, stenting with BMS or MIDCAB for isolated proximal LAD lesions was associated with similar long-term outcomes regarding hard endpoints (death and infarction) and the combined composite endpoint at the cost of a higher TVR rate in the PCI group mainly within the first 6 months.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Blazek and Holzhey contributed equally to this paper. Drs. Mohr and Thiele should be considered as shared senior authors.
- Abbreviations and Acronyms
- bare-metal stent(s)
- confidence interval
- creatine kinase-myocardial band
- drug-eluting stent(s)
- hazard ratio
- interquartile range
- left anterior descending
- major adverse cardiac events
- minimally invasive direct coronary artery bypass
- percutaneous coronary intervention
- target vessel revascularization
- Received September 14, 2012.
- Accepted September 27, 2012.
- American College of Cardiology Foundation
- Thiele H.,
- Neumann-Schniedewind P.,
- Jacobs S.,
- et al.
- Kishi K.,
- Kimura T.,
- Morimoto T.,
- et al.,
- j-Cypher Registry Investigators
- Drenth D.J.,
- Winter J.B.,
- Veeger N.J.,
- et al.
- Jaffery Z.,
- Kowalski M.,
- Weaver W.D.,
- Khanal S.
- Thiele H.,
- Oettel S.,
- Jacobs S.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- White H.D.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.,
- Academic Research Consortium
- Cutlip D.E.,
- Chauhan M.S.,
- Baim D.S.,
- et al.
- Uchida T.,
- Popma J.,
- Stone G.W.,
- et al.
- Kastrati A.,
- Schömig A.,
- Elezi S.,
- et al.
- Sarno G.,
- Lagerqvist B.,
- Fröbert O.,
- et al.
- Sawhney N.,
- Moses J.W.,
- Leon M.B.,
- et al.
- Dangas G.,
- Ellis S.G.,
- Shlofmitz R.,
- et al.,
- TAXUS-IV Investigators
- Kasaoka S.,
- Tobis J.M.,
- Akiyama T.,
- et al.
- Drenth D.J.,
- Veeger N.J.,
- Grandjean J.G.,
- Mariani M.A.,
- van Boven A.J.,
- Boonstra P.W.
- Goy J.J.,
- Kaufmann U.,
- Hurni M.,
- et al.,
- SIMA Investigators