| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Cardiac Surgery, Heart Center Leipzig, Struempellstr. 39, 04289 Leipzig, Germany
b Department of Cardiology, Heart Center Leipzig, Germany
c Department of Cardiothoracic Surgery, Erasmus Medical Center, Rotterdam, The Netherlands
d Department of Interventional Cardiology, Erasmus Medical Center, Rotterdam, The Netherlands
Received 8 November 2007; received in revised form 6 April 2008; accepted 23 April 2008.
* Corresponding author. Tel.: +49 341 865 1319; fax: +49 341 865 1452. (Email: rastan{at}rz.uni-leipzig.de).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
Key Words: Coronary artery disease • CABG • PCI • Multivessel disease • Left main stem stenosis • SYNTAX study
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
With the advent of drug-eluting stents (DES) and resultant significant reduction in angiographic restenosis compared to BMS [8], there is again an increasing use of PCI with DES for 3-VD/LM patients. By reducing the incidence of restenosis, a major disadvantage of BMS technology, and by treating DES patients with modern antiplatelet agents, it is now supposed that PCI can challenge CABG in the treatment of patients with complex coronary artery disease (CAD). Consequently, the frequency of off-label stenting outside the current guideline indications has increased and was recently estimated to account for 20–30% of routine interventional cardiology practice [9,10]. However, in recent years CABG surgery has also made significant progress as a result of better perioperative management, refinements in surgical skills, the use of more arterial grafts and different operative options including minimally invasive and off-pump surgery allowing an individualized strategy in an aging patient population. The optimal revascularization therapy for patients with 3-VD/LM is therefore once again being put forward for discussion, even though there is no evidence to date that DES offers a mortality benefit in comparison to BMS [11–14].
Comparative studies on PCI with DES versus CABG for the treatment of 3-VD/LM are very limited thus far [15–18]. Furthermore, their validity is restricted because of small study populations, inclusion of highly selected patients and a limited follow-up period. The SYNTAX study (synergy between PCI with Taxus drug-eluting stent and cardiac surgery) is a multinational, prospective, randomized trial comparing CABG and PCI with the Taxus paclitaxel-eluting stent in the treatment of de-novo 3-VD and/or LM disease [19]. In contrast to most previous randomized studies, the SYNTAX trial has an all comers design by replacing the most traditional inclusion and exclusion criteria with a real world decision of interventionalists and cardiac surgeons during a common heart team conference (HTC). Study exclusion criteria were accordingly restricted to prior CABG or PCI procedures, acute myocardial infarction (AMI) with creatine kinase (CK-MB) more than two-fold upper limit of normal, planned need for additional major surgery, age less than 21 years, pregnancy, and uncertainty of study protocol adherence. Beyond these criteria, patients with significant cardiac or extracardiac morbidities, and in particular patients with complex coronary anatomy, could be included in the SYNTAX trial without restrictions. The study investigators therefore hope to produce a real world picture of how CABG and PCI with DES compare in patients with 3-VD or LM disease.
Between 03/2005 and 05/2007 a total of 85 sites in Europe (62) and North America (23) took part in the SYNTAX study enrolment. During the enrolment phase, all consecutive patients with de-novo 3-VD and/or LM disease were screened for study suitability by a local heart team that was composed of an interventionalist, a cardiac surgeon, and the study co-ordinator. If their consensus decision was that comparable revascularization could be achieved by either PCI or CABG therapy, the patient was allocated to the randomized study arm. This arm included a total of 1800 patients; 900 patients were assigned to PCI with Taxus stents and 900 patients were assigned to CABG surgery. The primary endpoint for the randomized cohort was the overall major adverse cardiac and cerebral event (MACCE) rate at 12-month follow-up. The SYNTAX study also includes two nested registries, one for PCI-ineligible patients (CABG registry) and one for CABG-ineligible patients (PCI registry) to define patient characteristics and outcome of these patients compared to the randomized cohort [19].
The purpose of this manuscript is to describe a single center experience concerning frequency and study inclusion rate of patients with de-novo 3-VD/LM disease in a superregional cardiac catheterization unit during the SYNTAX study enrolment phase, as well as to identify comorbidities and cardiac pathologies limiting randomization based on the local HTC decision.
|
2. Patients and methods |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
21 years of age received coronary angiography in our cardiac catheterization unit. These patients were primarily assessed for the extent of CAD based on the three major coronary territories, LM disease, and significant heart valve disease. Patients were reviewed regarding previous interventional or surgical revascularization to identify the original pool of patients with de-novo 3-VD or LM stenosis. Baseline information of these patients was reassessed during the regular HTC meetings and the patients were screened for other exclusion criteria and potential SYNTAX study participation. Patients that met the study inclusion criteria were then assessed by the entire team and a decision was made regarding their eligibility for the randomized cohort, or for the CABG or PCI registry. This was followed by a HTC visit to the patient, describing details on the study purpose, explaining the HTC decision and to ask for the patient's written consent. Ethics approval for the study was obtained from our institutional research ethics board.
Detailed characteristics of all study participants were then recorded according to the study protocol. This included demographic data, surgical risk profile (EuroSCORE) and complexity of the individual coronary vasculature disease [19]. The latter was assessed by the SYNTAX score, an angiographic tool grading the severity of CAD [20]. The SYNTAX score is a comprehensive anatomical assessment derived from various pre-existing classifications and allows quantification of not only the number and location of coronary lesions, but also the complexity of each individual lesion. In brief, each lesion producing 50% luminal obstruction in vessels 1.5 mm is defined based on the modified AHA coronary tree segment classification and separately scored regarding characteristics such as bifurcational or trifurcational or aortic ostial localization, chronic occlusion, vessel tortuosity, length, calcification and thrombus formation [20]. Finally, the score of each lesion was added to obtain the patient's SYNTAX score. Thus, higher SYNTAX scores are indicative of complex coronary disease with more demanding revascularization procedures and potentially worse prognoses [21].
To verify the hypothesis that complex coronary anatomy was a major characteristic that resulted in patients being deemed ineligible for PCI, we compared the randomized cohort with CABG registry patients. On the other hand, it could also be argued that surgical revascularization was refused in patients with very severe comorbidities, which is why we compared PCI registry patients with all other (potentially surgical) candidates. To identify treatment differences, we finally compared PCI and CABG procedures between the randomized and the registry arms.
2.1 Data analysis
Data were prospectively recorded. Continuous variables are expressed as mean ± SD and categorical data as proportions throughout the manuscript. Categorical variables were compared using the chi-square test or Fisher's exact and independent continuous variables were compared by two-tailed Student's test or Mann–Whitney U-test as appropriate. A p value less than 0.05 was considered statistically significant. SPSS 13.0 and Microsoft Excel software package were used for statistical calculation.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
|
|
|
The mean age of all enrolled patients was 69.2 ± 9.7 years with 28.3% being female and 37.3% having medically treated diabetes. The logistic EuroSCORE predicted risk of operative mortality was mean 5.6%. Regarding cardiac risk factors, 43.2% of patients had significant LM stenosis and 40.3% presented with an occlusion of at least one major coronary vessel. The number of coronary artery lesions was 4.9 ± 1.6/patient and the mean SYNTAX score was 29.2 ± 9.5. Ejection fraction was 52.3 ± 16.2%.
As demonstrated in Table 2 , extracardiac risk factors of all randomized patients were comparable to CABG registry patients, but the latter had more complex CAD indicated by a higher number of coronary artery lesions (p = 0.048), total coronary vessel occlusions (p < 0.001) and occlusion of the circumflex (p = 0.007) and right (p = 0.044) coronary artery. CABG registry patients therefore had a higher SYNTAX score (p < 0.001). However, the different cohorts were comparable regarding LM disease. Furthermore there was a trend towards lower EF (p = 0.063) and a higher EuroSCORE in CABG registry patients (p = 0.274). The small number of PCI registry patients (n = 3) had markedly increased extracardiac risk factors and a significantly higher operative risk score, but also demonstrated more complex diseased coronary arteries compared to surgical candidates (Table 2).
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
Coronary 3-VD and LM stenosis represent prognostically relevant CAD with a significant effect on long-term survival, as was shown in the early era of coronary revascularization. In these patients CABG significantly improves both symptoms and patient survival [1]. In contrast, PCI resulted in symptom relief, but failed to demonstrate a significant effect on long-term survival. This might be explained by the fact that PCI was predominantly applied in patients with 1- or 2-VD, lesions that are known to be of minor prognostic relevance in stable angina patients. In risk-adjusted analyses of patients with 3-VD/LM, it could be shown that CABG was associated with a survival benefit compared to PCI with BMS [2–4]. Since a clear survival benefit of DES over BMS has not been demonstrated thus far, it is open to discussion whether or not PCI with DES may result in a survival benefit that is comparable to CABG in 3-VD/LM patients. It is expected that the SYNTAX trial will be able to more definitively answer this question.
During SYNTAX study enrolment we analyzed a total of 3319 adult patients undergoing coronary angiography during a 4-month period in order to characterize the current patient population regarding extent and pattern of CAD in a superregional catheterization laboratory practice. CAD was present in 63.8% of patients with an even distribution of 1-, 2-, and 3-VD. Even though the proportion of patients undergoing coronary angiography who have CAD varies between centers and countries depending on the volume of performed PCI and CABG procedures, medical insurance systems, demography of the local population and research activities, our results were almost identical to 229,462 coronary angiographies recorded between 2003 and 2005 in all of Germany [24]. An interesting finding of our study was that the vast majority of patients with 1- and 2-VD presented with de-novo CAD, while 72.4% of 3-VD/LM patients had prior coronary intervention either by PCI or CABG. This could reflect the center-specific case load of CABG surgery and progressive PCI treatment and might not be representative of cath lab populations in general. However, since all SYNTAX participating centers are high-volume centers with extensive experience in complex PCI procedures, our finding might be representative of all SYNTAX study centers [9].
One major finding of our analysis was that a total of only 191 patients, i.e. 5.8% of all catheterized patients, demonstrated de-novo 3-VD/LM emphasizing that this subset of patients is a minority in the routine cath lab practice. Based on the decision of the HTC, 87 of these patients had medical reasons making them ineligible for SYNTAX study enrolment. The reasons were: no option for invasive revascularization, AMI with 2-fold elevated CK, or indications for non-coronary additional cardiac surgery. Non-coronary indications for cardiac surgery were present in 35 of 191 de-novo 3-VD/LM patients (18.3%), consisting predominantly of moderate aortic valve stenosis and moderate mitral valve incompetence. Since these patients were initially assessed for the presence of CAD only, this underlines the importance of joint decision-making between cardiologists and cardiac surgeons in patients being considered for revascularization.
The remaining 104 patients were potentially suitable for SYNTAX study participation. After a detailed description of the study purpose by the heart team, a considerable number of patients (12.5%) refused to participate on the study. Most of these patients were scheduled for randomization based on the HTC decision, but preferred surgical revascularization. This demonstrates that a significant proportion of patients will opt for a more invasive, but more proven therapy, after an unbiased exchange of information.
Another interesting finding of our study was the unexpected high number of patients that were excluded from study participation because of uncertain protocol adherence. This was mainly due to concerns regarding the ability of patients to tolerate dual antiplatelet drug therapy after DES treatment, particularly in older patients and in those requiring invasive or surgical procedures within the next 12 months. Since premature discontinuation of antiplatelet therapy after PCI with DES is well known to be associated with a significant increase in the risk of stent thrombosis, these patients were not considered for study participation [25].
Finally, a total of 67 patients were included in the SYNTAX study with randomization achieved in 40 patients. Although this represents a small rate of inclusion (2.0%) and randomization (1.2%) of all catheterized patients, the inclusion (35.1%) and randomization rates (20.9%) of all screened patients were respectable. Furthermore, the inclusion and randomization rates of patients that were potentially suitable for SYNTAX study participation (64.4 and 38.5%, respectively) were very good. These rates are significantly better than in previous randomized controlled trials of PCI versus CABG surgery, in which randomization rates ranged between 2% and 9% of screened patients [5]. Based on our single center experience, this confirms the all comers strategy of the SYNTAX trial and should result in the inclusion of patients with representative coronary artery complexity of a real world population.
The majority of our SYNTAX patients were randomized. Of the patients who were not randomized, all but three entered the CABG registry. Our 60% randomization rate (40/67 eligible patients) is comparable to the 58.5% randomization rate (1800/3076 patients) reported for the final SYNTAX study enrolment data at the recent EuroPCR meeting. However, this randomization rate was considerably higher than the 35% rate initially assumed in the SYNTAX protocol [19].
We found identical characteristics for randomized PCI and randomized CABG patients. However, there were significant differences between the randomized cohort and CABG registry patients concerning complexity of CAD, while extracardiac risk factors were similar. Complete vessel occlusion was the predominant reason that the heart team declined randomization and recommended CABG. LM stenosis and low left ventricular ejection fraction played a minor role in this decision process. Patients that were recommended for PCI registry represented a small, but high risk group. These patients had multiple comorbidities that made the risk of surgery prohibitively high.
The SYNTAX score was recently developed as a comprehensive angiographic scoring tool to quantify the complexity of CAD and to assist in patient selection and risk stratification [20]. Although experience with this score is limited to date, it is a promising classification system for detailed description of the anatomical severity of CAD. It was recently demonstrated that PCI patient presenting with SYNTAX scores >26 had a significantly higher post-procedural MACCE rate and that the SYNTAX score independently predicted outcome [21]. However, its prognostic impact for CABG patients is still unknown. Post-hoc analyses of SYNTAX score subgroups may help define coronary characteristics that favor either CABG or PCI in 3-VD/LM patients.
|
5. Conclusions |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
5.1 Study limitations
The data of this paper reflect a single center SYNTAX study enrolment experience and might not be representative for the entire SYNTAX study patient population. All presented data are based on the assessments of the local heart team.
|
Appendix A |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
Dr S. Nashef (Cambridge, UK): One particular interesting finding in your slides is that when patients were allocated to their nested registries because they were felt to be more appropriate for PCI or more appropriate for CABG, if I remember right, I think 24 went to CABG and only 3 went to PCI.
And this really raises the question that one of the major problems with randomized control trials in this field is that what we call suitability for randomization is very often suitability for PCI. I think almost everybody is suitable for CABG unless they’re dead, and sometimes if the death has occurred in the cath lab, then even that contraindication is removed.
So, I mean, what this essentially means is that whenever you do a randomized control trial of CABG versus PCI, there is an inherent bias towards PCI because there is a tendency to choose the patients who are most suitable for PCI for the randomization. And, therefore, we have to be cautious about extrapolating the outcomes from randomized control trials to the ischemic heart disease population as a whole.
Nevertheless, I think it is a very, very good study in comparison with other randomized control trials of CABG versus PCI.
I only have one question for you, and that is the percentage of randomization when compared with the original pool. Because if I remember right, you said that 40 out of 104 were randomized, which is 38.5%, an excellent rate. However, I’m not too sure how you defined the original pool. Is the original pool 104 or is it your 694 or is it the 3000 you started off with, and I would appreciate some guidance on that.
Dr Rastan: The most important limitation to randomize the patients surely was the complexity of coronary arteries not allowing to treat them by PCI. This is obvious. We could operate on all of these patients. However, we had only three patients who we refused to operate on because of too strong comorbidities. So the most important reason not to randomize CABG registry patients was the complexity of the coronary artery disease making these patients not eligible for percutaneous coronary intervention.
However, if you remember the data from our hospital, we could show that more than 50% of our screened patients could be enrolled into the SYNTAX study and of them the majority could be randomized. We defined the number of screened patients as the original pool of patients, because they fulfilled the major criteria of revascularizable de-novo coronary three-vessel and/or left main disease, no severe myocardial infarction and required no additional surgery.
So even if the randomized patients do not represent all of our screened patients I think they more or less represent a significant amount of our daily CABG patient population. In other words it is a real world population compared to other studies which just make their analyses on highly selective patients. However, as we could demonstrate here, the CABG registry patients had significantly more complex coronary artery disease.
I cannot exclude that there might be a strong enrolment difference from center to center, because I can here just present our single center experience from Leipzig. Thus, it remains speculative whether or not we will get similar results until we know the whole SYNTAX data in the next year.
Dr J. Alvarez (Claremont, Australia): Just briefly, we’ve just had the RAPCO study from Australia and so forth. Your length of stay for a very select group of patients, low risk I would say, was 12 days. Could you explain 12 days? That's something peculiar to your German administration. It seems like an extraordinary length of stay for this group of patients.
Dr Rastan: I’m very sorry. I just didn’t catch all of your question. Could you repeat it, please?
Dr Alvarez: Your length of stay for the CABG patients was 12 days. It seems an extraordinary long period of time. For example, the RAPCO study and so forth and most of us, I would imagine, would be about 5 or 6 days, particularly for this group of patients. Could you amplify on that because that will be a big thing, 12 days or the 6 days. But 12 days I will tell you is not the common length of stay around most parts of the world.
Dr Rastan: I think this is a very important point, and we have to look in detail on this issue by making a country-by-country analysis. In Germany we have a very specific medical system which will allow all routine patients to stay more or less 10–12 postoperative days in the hospital, sometimes longer. This is, I suppose significantly longer than in the most of the European countries and in the United States. However, 10–12 days is the average stay we know for CABG patients in Germany.
I think it is also interesting and important to note that the PCI patients also stayed 6 days in the hospital. The question arises why did they? I suggest we can answer this only if we know the incidence of the serious adverse events, but this is not my focus in the present analysis.
During the SYNTAX study analysis it is also planned to make a country-by-country analysis on the length of hospital stay.
Dr M. Sousa Uva (Lisbon, Portugal): I have a comment and a question. First everyone is expecting the SYNTAX trial, and I wonder what is the impact when it comes out? I guess nothing will occur because the cardiologists will say they have new technologies, of course, as usual. But nevertheless, it's a very important study and everyone is expecting it.
So I wanted to ask you two things. Who is responsible for the classification in giving this SYNTAX score, and when will we have the first results of this study?
Dr Rastan: To answer your first question the SYNTAX score was inaugurated by the group of Patrick Serruys in Rotterdam. They published the design of the SYNTAX score which is based on their experience during the ARTS II and ARTS I trial in 2005. They also could demonstrate that it is an extremely reliable tool to identify patients who will require repeated revascularization.
The latter is a very interesting paper that was recently published this year.
To your second question. As it is planned at the moment the first results will be presented at the EuroPCR meeting or during the ESC meeting in the next year.
Dr A. Kappetein (Rotterdam, Netherlands): I’m a member of the executive committee of the SYNTAX trial, and you have perfectly explained what the objective is of the SYNTAX trial and gave some insights into the enrolment of the patients.
A core lab is calculating the SYNTAX score for each patient. Subsequently you have to compare the outcome of the core lab with the outcome of the investigators. It might be that the SYNTAX score is not very useful if the core lab yields a totally different outcome than the investigators. In that case we have failed in making a new coronary lesion score. The corelab is an independent body not related to the investigators.
What is also crucial is that in the SYNTAX trial we have an independent body, the clinical events committee that reviews the adverse clinical events that occur. So none of the investigators, none of the members of the executive committee or steering committee are involved with defining the clinical endpoints. It's an independent body with surgeons and cardiologists involved.
So they conclude if a myocardial infarction occurred or if some of the other endpoint were met.
|
Footnotes |
|---|
Presented at the 21st Annual Meeting of the European Association for Cardio-thoracic Surgery, Geneva, Switzerland, September 16–19, 2007. |
References |
|---|
|
TopAbstract1. Introduction2. Patients and methods3. Results4. Discussion5. ConclusionsAppendix AReferences |
|---|
75 years). Eur Heart J 2007;28:2714-2719.This article has been cited by other articles:
|
|
 |
|
  S. Silber, A. Colombo, A. P. Banning, K. Hauptmann, J. Drzewiecki, E. Grube, D. Dudek, and D. S. Baim Final 5-Year Results of the TAXUS II Trial: A Randomized Study to Assess the Effectiveness of Slow- and Moderate-Release Polymer-Based Paclitaxel-Eluting Stents for De Novo Coronary Artery Lesions Circulation, October 13, 2009; 120(15): 1498 - 1504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K. von Segesser Clinical databases - a double-edged sword! Eur. J. Cardiothorac. Surg., May 1, 2009; 35(5): 749 - 750. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
