Clinical Research in Cardiology

, Volume 99, Issue 9, pp 541–544 | Cite as

Best way to revascularize patients with main stem and three-vessel lesions. Patients should be operated!

  • H. Reichenspurner
  • L. Conradi
  • J. Cremer
  • F. W. Mohr
Clinical Debate


Despite established guidelines for the treatment of coronary artery disease (CAD) by either coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI), everyday clinical practice has proven to differ substantially with even the most complex coronary lesions being targeted by PCI today. However, an abundancy of clinical trials, both observational and randomized has proven the superiority of coronary surgery over PCI in almost every type of multivessel or left main CAD in symptomatic patients. This holds true also for ‘hard’ endpoints like cardiac death or myocardial infarction as recently demonstrated by the landmark SYNTAX trial. These results have lead to the wording of appropriateness criteria, which integrate current guidelines, evidence from clinical trials and interdisciplinary expert opinion and which express essentially the same message as the SYNTAX trial: “CABG remains the standard of care for patients with three-vessel or left main coronary artery disease” (Serruys et al. in N Engl J Med 360:961–972, 2009).


Coronary artery bypass grafting Percutaneous coronary intervention Multi vessel coronary artery disease Left main coronary artery disease Patient safety 


When percutaneous techniques for the treatment of patients with coronary artery disease (CAD) were first introduced in the late 1970s [1], coronary artery bypass grafting (CABG) was the only effective therapy for coronary revascularization and had proven to be superior to medical treatment in many patients. Since then, advances in coronary surgery such as arterial revascularization, minimally invasive and off-pump procedures, technical refinements enhancing intraoperative myocardial protection or modern postoperative care have further improved outcomes, underlining the role of CABG as the gold standard and treatment of choice for multivessel CAD in symptomatic patients. Ten years later, however, the procedural volume of percutaneous coronary interventions (PCI) had surpassed that of CABG. Technical advances such as the advent of bare-metal (BMS) and drug-eluting stents (DES) have led to a more aggressive attitude among interventional cardiologists, so that today even the most complex coronary lesions are being targeted by PCI [2]. Consequently, a survey investigating current clinical practise found about one-third of all patients with three-vessel CAD in Europe to be treated by PCI instead of CABG [3], although current guidelines suggest CABG as the superior treatment for patients with three-vessel or left main coronary disease. Thus, clinical reality and guidelines often do not correspond.

Appropriateness criteria for coronary revascularization

Therefore, the American College of Cardiology Foundation (ACCF) and the American Heart Association (AHA) along with seven other surgical and non-surgical speciality societies (AATS, STS, SCAI, ASNC, ASE, HFSA, SCCT) recently gathered clinical evidence and expert opinion in order to provide a review of almost 180 common clinical scenarios resulting in the appropriateness criteria for coronary revascularization [4]. In summary, CABG was judged appropriate for patients with two- or three-vessel disease and for left main CAD in all clinical settings where revascularization was deemed necessary. PCI was rated appropriate in acute myocardial injury and in patients with one- or two-vessel disease with involvement of the proximal LAD, but rated uncertain or inappropriate for all patients with three-vessel disease and/or left main stenosis (Fig. 1). Furthermore, the appropriateness criteria provide extensive and detailed information on optimal treatment strategies in many other clinical circumstances such as diabetes, prior bypass surgery (with or without patent grafts), high surgical risk due to non-cardiac comorbidities or acute coronary syndromes.
Fig. 1

Proposed indications for revascularization strategies according to extent of coronary artery disease

Evidence from clinical trials

There is an abundancy of both randomized and observational trials that have compared outcomes of patients with CAD after treatment with either angioplasty alone, PCI using BMS or CABG, which usually demonstrate that CABG is superior to PTCA/PCI in terms of longevity of effectiveness (need for reintervention) and equal or even superior in terms of mortality [5]. In 2003, a metaanalysis integrated the results of 13 randomized trials including almost 8,000 patients with single- or multivessel CAD and found not only a significant advantage for CABG in terms of reintervention, but also a significant survival benefit after 5 years [6]. Thus, when sufficiently large patient numbers were studied over a sufficiently long period of time, a significant survival benefit became apparent for the use of CABG. Some of the included trials compared CABG to angioplasty only (without stenting) but similar results were found in registry studies comparing BMS to CABG in multivessel disease [7].

After the introduction of DES into clinical practise, superiority compared to angioplasty only or stenting with BMS in terms of restenosis rates was soon demonstrated [8]. In an observational study, PCI using DES was compared to CABG in patients with two- or three-vessel CAD with involvement of the proximal LAD. The analysis of the New York Department of Health database demonstrated CABG to be associated with a lower incidence of separate endpoints like death, myocardial infarction and repeat revascularization after 18 months [9]. However, this was a non-randomized, observational analysis and in spite of large patient numbers and careful risk adjustment, results may at least in part have been biased regarding patient risk factors. Randomized trials on the other hand also have important limitations. Many represent single-centre experience and comprise of low patient numbers only. Due to study protocols and inclusion criteria the vast majority of patients receiving a defined treatment is usually eliminated, leading to a highly selective patient population of only about 2–10% of initially screened patients [10] resulting in a poor generalizability for everyday practise.

The SYNTAX trial

In a recently conducted randomized, prospective controlled trial, these limitations were partially addressed. The composite endpoint of major adverse cardiac and cerebrovascular events (MACCE) defined as death, myocardial infarction, stroke or reintervention was assessed after PCI with DES or CABG for patients with three-vessel and/or left main coronary disease [11]. The SYNTAX trial was designed in an all-comers fashion resulting in a high 58.5% (1,800 of 3,075 patients) randomization of all screened patients to either cohort. Patients who were deemed eligible for one treatment option only by both interventional cardiologist and cardiac surgeon (the so-called heart team) were entered into a parallel CABG or PCI registry. The study was statistically underpowered to demonstrate safety other than MACCE, where a significant advantage for CABG was shown. Thus, the primary endpoint of non-inferiority for PCI with regard to MACCE was not met and the authors concluded that CABG remains the standard of care for patients with de-novo three-vessel and/or left main CAD.

After 24 months of follow-up, there was a significantly higher rate of myocardial infarction in the PCI cohort as compared to CABG (5.9 vs. 3.3%, p = 0.01), resulting in a significantly higher rate of cardiac death (4.5 vs. 2.7, p = 0.04). However, regarding all-cause mortality there was only a trend in favour of CABG after 12 months which became more pronounced after 2 years but still remained insignificant. Remarkably, this advantage was present in the perioperative phase and at 30 days already, questioning the paradigm of PCI as the less invasive mode of revascularization. Given the fact that in former studies potentially fatal complications tended to occur early after CABG (e.g. perioperative mortality) but late after PCI (e.g. late stent thrombosis), it seems possible that during longer follow-up the mortality trend might increase further and may eventually translate into a statistically significant survival advantage for CABG. Possibly, the fact that complete revascularization was more likely in the CABG cohort will contribute to this observation.

In the SYNTAX trial, CABG was clearly superior to PCI regarding repeat revascularization, with 17.4% of patients in the PCI group undergoing reintervention as compared to 8.6% (p < 0.001) in the CABG group after 2 years. Even though state-of-the-art PCI using DES was used, and results were better than those reported from earlier studies, the advantage of CABG remained highly significant.

Concerning stroke rate, there was an excess hazard for CABG with an incidence of 2.8% when compared with PCI with 1.4% (p = 0.03) at 24 months. In part, strokes occurred early after or during surgery, most likely attributable to atheroembolism from the ascending aorta. However, approximately 50% of strokes occurred more than 30 days after surgery in the SYNTAX trial. As there was a pronounced difference in postprocedural medication, especially significantly lower numbers of patients treated with aspirin (PCI 91.2% vs. CABG 84.3%) or clopidogrel (PCI 71.1% vs. CABG 15.0%) these secondary factors appear to be at least partially responsible for this finding. The occurrence of transient postoperative atrial fibrillation which is not an issue after PCI may also play a role in this context. For the treatment of high-risk patients, where off-pump coronary artery bypass surgery (OPCAB) has proven to be beneficial in many aspects [12], avoidance of complete or partial clamping of the aorta is gaining clinical acceptance. In Germany, the share of OPCAB procedures in coronary surgery has almost doubled during the last 5 years reaching over 11% of all cases in 2008 [13].

Another important issue in the controversy over the appropriate therapy for multivessel or left main CAD is that of reintervention due to stent restenosis or graft occlusion. Although failure rates of the initial revascularization procedure were found to be not significantly different for PCI and CABG in the SYNTAX trial, the consequences of the two respective events are known to differ enormously. While stent thrombosis frequently results in myocardial infarction (up to 80% of cases) with reported mortality rates between 30 and 45% [14, 15] graft occlusion usually becomes apparent with recurring angina, which then leads to repeat revascularization. This effect is well known from the literature and did indeed translate into increased rates of myocardial infarction and cardiac death in the PCI group of the SYNTAX trial. Interventionalists should be aware of this phenomenon, especially when implementing PCI in cases of left main stenosis or advanced CAD.

The advantage of CABG over PCI became more pronounced with increasing anatomical complexity of CAD as documented by the SYNTAX Score. In the PCI cohort, MACCE rates were significantly higher in the intermediate score tercile and became excessively high in severe CAD (high score tercile). In the corresponding CABG groups, MACCE rates even decreased with higher scores possibly due to lower competitive flow rates that occur when highly stenosed coronary arteries are grafted and that are known to have a beneficial effect on bypass patency. In the SYNTAX trial, more than two thirds of patients were assigned intermediate or high scores in the PCI and CABG cohorts (66.8 and 69.2%, respectively).

In the subset of diabetic patients, a similar effect was observed. The difference in MACCE became more pronounced with increasing severity of diabetes. In diabetic patients treated with insulin, the combined endpoint of death, myocardial infarction and cerebrovascular events after 1 year occurred in 14.8% in the PCI group compared to 8.0% in the CABG group. This difference was however not statistically relevant (p = 0.16).


The Appropriateness Criteria worded by the American College of Cardiology and the results of the landmark SYNTAX trial have equipped us with excellent new tools to guide decision making for revascularization strategies in multivessel and left main CAD. CABG should remain the standard of care for symptomatic patients with three-vessel or left main CAD since the main goal of non-inferiority was not achieved in the recent SYNTAX trial. The same recommendation is expressed in the Appropriateness Criteria, which rate PCI as uncertain to increase health outcomes or survival for three-vessel disease and as inappropriate for left main disease.

The indication for treatment by PCI or CABG depends not only on the formal classification as one-, two- or three-vessel disease but also on the anatomical complexity of coronary lesions. As the SYNTAX trial has impressively demonstrated, PCI should not be attempted in multivessel CAD patients with intermediate or high SYNTAX scores (which account for more than two-thirds of all patients in the SYNTAX PCI cohort) due to inferior outcomes. Superiority of CABG compared to PCI in symptomatic patients with multivessel or left main disease regarding isolated endpoints like death or myocardial infarction has been suggested by observational trials. In the meantime, this has been demonstrated regarding cardiac death and myocardial infarction in the 2-year results of the SYNTAX trial. Further proof may be expected from the long-term follow-up (SYNTAX 5-year data).

In cases where CABG is likely associated with high surgical risk due to patient comorbidities, PCI should be considered. The latter are usually complex patients which should be assessed by both cardiologist and surgeon to define the optimal treatment strategy following the heart team concept described in the SYNTAX trial. In the setting of ongoing myocardial injury, available evidence is ambiguous with good outcomes reported after either interventional or surgical revascularization [16, 17]. The GRACE registry database (Global Registry of Acute Coronary Syndromes; [18] provides extensive data such as characteristics, treatment modalities and outcomes from more than 100,000 patients revascularized in the setting of acute coronary syndromes.

In an ideal setting, every patient with three-vessel or left main CAD should be discussed between cardiologists and surgeons within a heart centre. The final decision making should be a formal process where both specialities define the further treatment mode depending on the specific case, coronary anatomy and comorbidities and independently from any financial or budget-related bias.


  1. 1.
    Grüntzig A (1978) Transluminal dilatation of coronary-artery stenosis. Lancet 1:263CrossRefPubMedGoogle Scholar
  2. 2.
    Zahn R, Hamm CW, Schneider S et al (2007) The sirolimus-eluting coronary stent in daily routine practice in Germany: trends in indications over the years. Results from the prospective multi-center german cypher stent registry. Clin Res Cardiol 96:548–556CrossRefPubMedGoogle Scholar
  3. 3.
    Kappetein AP, Dawkins KD, Mohr FW et al (2006) Current percutaneous coronary intervention and coronary artery bypass grafting practices for three-vessel and left main coronary artery disease: insights from the SYNTAX run-in phase. Eur J Cardiothorac Surg 29:486–491CrossRefPubMedGoogle Scholar
  4. 4.
    Patel MR, Dehmer GJ, Hirshfeld JW et al (2009) ACCF/SCAI/STS/AATS/AHA/ASNC 2009 Appropriateness criteria for coronary revascularization. J Am Coll Cardiol 53:530–553CrossRefPubMedGoogle Scholar
  5. 5.
    Daemen J, Boersma E, Flather E et al (2008) Long-term safety and efficacy of percutaneous coronary intervention with stenting and coronary artery bypass surgery for multivessel coronary artery disease: a meta-analysis with 5-year patient-level data from the ARTS, ERACI-II, MASS-II and SoS trials. Circulation 118:1146–1154CrossRefPubMedGoogle Scholar
  6. 6.
    Hofmann SN, TenBrook JA, Wolf MP et al (2003) A Meta-analysis of randomized controlled trials comparing coronary artery bypass graft with percutaneous transluminal coronary angioplasty: one- to eight-years outcomes. J Am Coll Cardiol 41:1293–1304CrossRefGoogle Scholar
  7. 7.
    Hannan EL, Racz MJ, Walford G et al (2005) Long-term outcomes of coronary-artery bypass grafting versus stent implantation. N Engl J Med 352:2174–2183CrossRefPubMedGoogle Scholar
  8. 8.
    Moses JW, Leon MB, Popma JJ et al (2003) Sirolimus-eluting stents versus standard stents in patients with stenosis in native coronary artery. N Engl J Med 349:1315–1323CrossRefPubMedGoogle Scholar
  9. 9.
    Hannan EL, Wu C, Walford G et al (2008) Drug-eluting stents vs. coronary artery bypass grafting in multivessel coronary disease. N Engl J Med 358:331–341CrossRefPubMedGoogle Scholar
  10. 10.
    Morrison DA, Sethi G, Sacks J et al (2001) Percutaneous coronary intervention versus coronary artery bypass graft surgery for patients with medically refractory myocardial ischemia and risk factors for adverse outcomes with bypass: a multicenter, randomized trial: Investigators of the Department of Veterans Affairs Cooperative Study #385, the Angina With Extremely Serious Operative Mortality Evaluation (AWESOME. J Am Coll Cardiol 38:143–149CrossRefPubMedGoogle Scholar
  11. 11.
    Serruys PW, Morice MC, Kappetein AP et al (2009) Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med 360:961–972CrossRefPubMedGoogle Scholar
  12. 12.
    Puskas J, Cheng D, Knight J et al (2004) Off-pump versus conventional coronary artery bypass grafting: a meta-analysis and consensus statement from the 2004 ISMICS consensus Conference. Innovations 1:3–27Google Scholar
  13. 13.
    German Society for Thoracic and Cardiovascular Surgery Annual Report 2008Google Scholar
  14. 14.
    Mauri L, Hsieh WH, Massaro JM et al (2007) Stent thrombosis in randomized clinical trials of drug-eluting stents. N Engl J Med 356:1020–1029CrossRefPubMedGoogle Scholar
  15. 15.
    Iakovou I, Schmidt T, Bonizonni E et al (2005) Incidence, predictors and outcome of thrombosis after successful implantation of drug-eluting stents. JAMA 293:2126–2130CrossRefPubMedGoogle Scholar
  16. 16.
    Alexiou K, Kappert U, Staroske A et al (2008) Coronary surgery for acute coronary syndrome: which determinants for outcome remain? Clin Res Cardiol 97:601–608CrossRefPubMedGoogle Scholar
  17. 17.
    Dibra A, Tiroch K, Schulz S et al (2010) Drug-eluting stents in acute myocardial infarction: updated meta-analyses of randomized trials. Clin Res Cardiol 99:345–357CrossRefPubMedGoogle Scholar
  18. 18.
    Fox KA, Eagle KA, Gore JM et al (2010) The global registry of acute coronary events, 1999 to 2009-GRACE. Heart [Epub ahead of print]Google Scholar

Copyright information

© Springer-Verlag 2010

Authors and Affiliations

  • H. Reichenspurner
    • 1
  • L. Conradi
    • 1
  • J. Cremer
    • 2
  • F. W. Mohr
    • 3
  1. 1.Department of Cardiovascular Surgery, University Heart Center HamburgUniversity Medical Center Hamburg EppendorfHamburgGermany
  2. 2.Department of Cardiac and Vascular SurgeryUniversity Medical Center Schleswig-Holstein, Campus KielKielGermany
  3. 3.Department of Cardiac Surgery, Heart Center LeipzigUniversity of LeipzigLeipzigGermany

Personalised recommendations