The evolution in recommendations concerning the treatment of left main coronary artery disease (LMCAD) undoubtedly reflects the current state of cardiac surgery over recent years. Despite excellent results, the surgical community has been compelled to “defend” itself against publications of questionable evidence and hasty interpretations of certain studies.

Historically, for stable angina, LMCAD has predominantly been treated surgically. Given the high mortality associated with this pathology, classical studies from the early ’90s recommended surgery to address the ischemic myocardium at risk. For decades, myocardial revascularization was primarily surgical. However, advancements in percutaneous treatment have prompted various studies to reconsider the suitability of surgical (CABG) versus percutaneous (PCI) treatment.

Without evaluating the number of CABG procedures, which are undoubtedly below those observed in other countries, it is essential to understand the outcomes of myocardial revascularization surgery in Spain. According to consistent data from the national registry, coronary surgery mortality rates are around 2% (2.24% for CABG with extracorporeal circulation (ECC) in 2022, 1.59% for CABG without ECC in 2022). Although the registry data do not provide specific outcomes for LMCAD, preliminary data from the Spanish Cardiac Surgery Registry (RECC) confirm these figures, placing the mortality rate for myocardial revascularization surgery in Spain between 2-3%, despite an increase in patient risk and the percentage of PCI prior to surgery in recent years.

Changes in treatment guidelines began with the publication of coronary revascularization guidelines. In the United States, the guidelines were released in 2021, stating that surgical revascularization is a Class I recommendation for LMCAD in patients with stable angina to improve survival. For certain patients, PCI is deemed reasonable. Earlier, in 2018, European guidelines took a significant step by equating the recommendation level between CABG and PCI for patients with LMCAD and a Syntax score between 0-22 points (low Syntax score). For patients with intermediate or high Syntax scores, the PCI recommendation drops to IIa and III, respectively. This change in European guidelines was prompted by the development of several clinical trials, most notably the EXCEL study, which will be discussed later.

The first of these notable studies was the SYNTAX study, which analyzed 705 patients with LMCAD. After 5 years, the benefit of surgery was observed only in high-risk Syntax scores (46.5% of major cardiovascular events in PCI vs. 29.7% in surgery). However, the 10-year follow-up showed no surgical advantage over percutaneous treatment. This study reinforced the superiority of surgical treatment for multivessel disease, but not for LMCAD. Various studies conducted in this field sought differences through “new” statistical tools, focusing on combined outcome objectives and non-inferiority studies.

Among these, the NOBLE study, which analyzed 1200 patients, explored differences between both treatment modalities, with a primary outcome that included mortality, need for revascularization, stroke, and non-procedural myocardial infarction. The result showed 28.4% of events in PCI compared to 19% in surgery. The interpretation of this study suggested that CABG “appears superior to PCI for LMCAD treatment.” From this study, the predictable conclusion was that adding up events, non-procedural myocardial infarction and the need for revascularization penalized percutaneous treatment.

The EXCEL study learned from previous “errors” and identified the need for certain improvements to continue analyzing the “problem.” Consequently, it considered that the need for revascularization did not require analysis, and by modifying the protocol, it changed the definition of infarction to include procedural infarctions based solely on enzymatic criteria. Following the third universal definition of infarction, in the Excel study, periprocedural infarctions (PPI) were observed at 2.2%, increasing to 6.1% in the surgical branch following the protocol change. Despite a 40% discrepancy from the initial protocol, mortality differences favoring surgery (13.0% in PCI vs. 9.9% in CABG), many unpublished data, and a non-inferiority obtained based on the expected enzymatic increase after surgical revascularization, the EXCEL study results were published in the New England Journal of Medicine. This likely influenced the clinical guidelines change and contributed to the perception within the cardiological community that surgery and angioplasty were now equivalent for LMCAD treatment.

All these events had a dual impact within the surgical community. The first was the public reaction, with rejection statements covered by non-scientific media. For the first time, “opinion differences” between surgery and cardiology were broadcast to the public. The second impact was the development of new studies and reinterpretations of previous studies that could shed more light on the ideal treatment for LMCAD.

Regarding the reinterpretation of previous data, we highlight, on one hand, the Bayesian approach to the EXCEL study conducted by Gaudino in 2020, and on the other, the creation of a working group between the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) in 2022. The Bayesian approach uses existing study data and modifies the order of data analysis. By focusing on the event itself, the goal is to determine whether the cause was PCI or CABG, an original and distinct approach compared to traditional methods. This allows for analyzing the magnitude of each effect. Without delving into details, as a detailed analysis of this study has already been published, the major shift involved moving away from the frequentist (often simplistic) approach centered on finding a “p.” A more comprehensive interpretation of the results was achieved, highlighting the magnitude of differences between events.

Fortunately, this controversy sparked a reflection that materialized into a return to common sense. Consequently, the ESC and EACTS worked to analyze the therapeutic recommendations for LMCAD in 2022. The results were reviewed by 12 members from both societies, taking into account four major clinical trials (NOBLE, SYNTAX, EXCEL, and PRECOMBAT) and a meta-analysis published by Sabatine (The Lancet, 2021). Overall, 4,394 patients with five-year follow-up data were analyzed. The primary endpoint was mortality, while secondary endpoints included myocardial infarction, stroke, the need for repeat revascularization (NRR), and a composite of death-stroke-myocardial infarction.

In summary, according to a Bayesian analysis, PCI was associated with an 85.7% higher likelihood of increased mortality. However, the impact on early mortality was less pronounced. Regarding secondary outcomes, there were more spontaneous myocardial infarctions in the PCI group (NNT = 29 CABG procedures would prevent one spontaneous myocardial infarction during follow-up), with no significant differences observed in periprocedural infarctions under the third universal definition of infarction. There were no significant differences in stroke (2.7% for PCI, 3.1% for CABG). PCI was associated with a higher need for repeat revascularization (NNT = 14 CABG procedures would prevent one repeat revascularization during follow-up). The combined endpoint of death-stroke-myocardial infarction showed 19.7% of events in the PCI group compared to 15.5% in the CABG group at five years. Approximately 20 events occurred in the PCI group over five years versus 16 in the surgery group.

In summary, the working group concluded that out of 100 patients undergoing PCI for LMCAD, 89 would be alive at five years, and 80 of these would remain event-free. In the surgical branch, out of 100 patients, 90 would be alive and 84 event-free. Based on these results, it was recommended to modify the guidelines. For the low-to-moderate Syntax score group, PCI’s recommendation was downgraded to IIa, while maintaining Class I for surgery.

This appeared to close (for now) an era in LMCAD research characterized by statistical ambiguity. Studies with leading titles, questionable influence from “scientific” sponsors, repeated losses to follow-up, protocol deviations, composite endpoints, non-inferiority with absolute rather than relative margins—all simplified the problem of treating LMCAD and led to flawed conclusions. The interpretation and extrapolation of these studies is undoubtedly the most critical point. For instance, it would be unreasonable to interpret Holger Thiele’s study on ECMO use in cardiogenic shock to mean it should be entirely dismissed from treatment. Among other reasons, studies sometimes seem disconnected from daily clinical practice.

Reflecting on daily clinical practice, we highlight two recent studies to conclude this commentary. The first is a propensity-matched analysis of 1,128 patients with LMCAD undergoing surgery or PCI in Canada. After seven years, the study demonstrated increased mortality with PCI (54%) compared to surgery (35%), as well as more myocardial infarctions in the PCI group (19% versus 11%) and a higher need for repeat revascularization (18% versus 6%). On the other hand, the surgical branch showed more strokes (5.3% in PCI versus 7.6% in CABG). The second study, previously analyzed in this blog, is the SWEDEHEART registry, which examined 11,137 patients over a 10-year period. Mortality and major cardiovascular events were favorable to the surgical branch. However, one of the study’s most interesting findings was the analysis of median survival. Moving beyond the simplification of Kaplan-Meier survival curves and incorporating recommendations to analyze the area under the curve for true survival impact, the Swedish study revealed a median survival difference of 2.58 years. In this study, patients with LMCAD undergoing surgery lived nearly three years longer than those undergoing PCI.

LMCAD can be treated surgically or percutaneously. PCI results make this treatment option feasible in cases that are technically uncomplicated or when surgery is considered high-risk. Generally, for operable patients with a life expectancy of more than five years, current scientific evidence supports the suitability of surgical treatment. In retrospect, the studies on LMCAD treatment have had clear benefits, even for the surgical community. Not only through the scientific knowledge gained, which has required the development of additional statistical skills, but also by motivating the surgical community to continue achieving optimal surgical outcomes and improving techniques and approaches. Just as the “TAVI phenomenon” has prompted surgeons to perform less invasive surgeries with excellent results in terms of mortality and hemodynamics, excellence is essential in coronary disease treatment. Whether prevention or medical treatment calls for a more aggressive option, the boundaries between surgical and percutaneous approaches must disappear, with one ultimate beneficiary always at the center of our actions: the patient.

REFERENCES:

1. Lourdes Montero-Cruces, Daniel Pérez-Camargo, Javier Cobiella-Carnicer, Rosa Beltrao Sial, Cristina Villamor-Jiménez et al. Resultados de la cirugía coronaria en España. Análisis del conjunto mínimo básico de datos 2001-2020. Revista Española de Cirugía Cardiovascular, Vol 30, num. 2, 82-89. DOI: 10.1016/j.circv.2022.08.003
2. Byrne RA, Fremes S, Capodanno D, Czerny M, Doenst T, Emberson J R et al. 2022 Joint ESC/EACTS Review of the 2018 Guideline Recommendations on the Revascularization of Left Main Coronary Artery Disease in Patients at Low  Surgical Risk and Anatomy Suitable for PCI or CABG. Eur J Cardiothorac Surg. 2023 Aug 1;64(2):ezad286. doi: 10.1093/ejcts/ezad286
3. Mohr F, Morice M-C, Kappetein AP, Feldman TE, Stahle E, Colombo A et al. Coronary Artery Bypass Graft Surgery versus Percutaneous Coronary Intervention in Patients with Three-Vessel Disease and Left Main Coronary Disease: 5-Year  Follow-up of the Randomised, Clinical SYNTAX Trial. Lancet. 2013 Feb 23;381(9867):629-38. doi: 10.1016/S0140-6736(13)60141-5.
4. Stone G, Kappetein AP, Sabik J, Pocock SJ, Morice M-C, Puskas J et al. Five-Year Outcomes after PCI or CABG for Left Main Coronary Disease. N Engl J Med. 2019 Nov 7;381(19):1820-1830. doi: 10.1056/NEJMoa1909406.
5. Persson J, Yan J, Angeras O, Venetsanos D, Jeppsson A, Sjögren I et al. PCI or CABG for Left Main Coronary Artery Disease: The SWEDEHEART Registry. Eur Heart J. 2023 Aug 7;44(30):2833-2842. doi: 10.1093/eurheartj/ehad369.
6. Holm NR, Mäkikallio T, Lindsay MM, Spence MS, Erglis A, Menown IBA et al. Percutaneous Coronary Angioplasty versus Coronary Artery Bypass Grafting in the Treatment of Unprotected Left Main Stenosis: Updated 5-Year Outcomes from the Randomised, Non-Inferiority NOBLE Trial. Lancet. 2020 Jan 18;395(10219):191-199. doi: 10.1016/S0140-6736(19)32972-1.
7. Sabatine MS, Bergmark BA, Murphy SA, O´Gara PT, Smith PK, Serruys PW et al. Percutaneous Coronary Intervention with Drug-Eluting Stents versus Coronary Artery Bypass Grafting in Left Main Coronary Artery Disease: An Individual Patient Data Meta-Analysis. Lancet. 2021 Dec 18;398(10318):2247-2257. doi: 10.1016/S0140-6736(21)02334-5.
8. Tam DY, Fang J, Rocha RV, Rao SV, Dzavik V, Lawton J et al. Real-World Examination of Revascularization Strategies for Left Main Coronary Disease in Ontario, Canada. JACC Cardiovasc Interv. 2023 Feb 13;16(3):277-288. doi: 10.1016/j.jcin.2022.10.016.