Clinical event rate in patients with and without left main disease undergoing isolated coronary artery bypass grafting: results from the European DuraGraft Registry

Abstract

OBJECTIVES

Left main coronary artery disease (LMCAD) is considered an independent risk factor for clinical events after coronary artery bypass grafting (CABG). We have conducted a subgroup analysis of the multicentre European DuraGraft Registry to investigate clinical event rates at 1 year in patients with and without LMCAD undergoing isolated CABG in contemporary practice.

METHODS

Patients undergoing isolated CABG were selected. The primary end point was the incidence of a major adverse cardiac event (MACE) defined as the composite of death, myocardial infarction (MI) or repeat revascularization (RR) at 1 year. The secondary end point was major adverse cardiac and cerebrovascular events (MACCE) defined as MACE plus stroke. Propensity score matching was performed to balance for differences in baseline characteristics.

RESULTS

LMCAD was present in 1033 (41.2%) and absent in 1477 (58.8%) patients. At 1 year, the MACE rate was higher for LMCAD patients (8.2% vs 5.1%, P = 0.002) driven by higher rates of death (5.4% vs 3.4%, P = 0.016), MI (3.0% vs 1.3%, P = 0.002) and numerically higher rates of RR (2.8% vs 1.8%, P = 0.13). The incidence of MACCE was 8.8% vs 6.6%, P = 0.043, with a stroke rate of 1.0% and 2.4%, P = 0.011, for the LMCAD and non-LMCAD groups, respectively. After propensity score matching, the MACE rate was 8.0% vs 5.2%, P = 0.015. The incidence of death was 5.1% vs 3.7%, P = 0.10, MI 3.0% vs 1.4%, P = 0.020, and RR was 2.7% vs 1.6%, P = 0.090, for the LMCAD and non-LMCAD groups, respectively. Less strokes occurred in LMCAD patients (1.0% vs 2.4%, P = 0.017). The MACCE rate was not different, 8.5% vs 6.7%, P = 0.12.

CONCLUSIONS

In this large registry, LMCAD was demonstrated to be an independent risk factor for MACE after isolated CABG. Conversely, the risk of stroke was lower in LMCAD patients.

Clinical trial registration number

ClinicalTrials.gov NCT02922088.

Left main disease, Coronary artery bypass grafting, Major adverse cardiac events, Endothelial damage inhibitor

INTRODUCTION

Due to the large myocardial territory at risk, high ischaemic burden and the overall higher atherosclerotic burden in the aorta or other peripheral arteries, coronary artery disease (CAD) with the involvement of the left main (LM) coronary artery is commonly considered to be associated with significant morbidity and mortality [1–3]. Similarly, surgical revascularization of patients with left main coronary artery disease (LMCAD) is associated with a higher risk for adverse cardiac events. However, clinical outcomes have improved over the last decade [4, 5] and published results regarding LMCAD as a risk factor for coronary artery bypass grafting (CABG) vary. Due to the surrounding controversies with surgical revascularization of patients with LMCAD assumed to be associated with a higher risk for adverse cardiac events and in recent years and surgical revascularization as the gold standard for LMCAD being challenged by percutaneous coronary intervention with evolving next-generation drug-eluting stents [6], we have conducted a subgroup analysis of the European DuraGraft Registry to investigate the impact of LMCAD on clinical outcomes after CABG.

PATIENTS AND METHODS

Ethics statement

The study complied with the provisions of the Declaration of Helsinki. The ethical committee at each study centre approved the study protocol. Ethical committee approval was obtained for the first participating site on 28 September 2016 in Vitoria, Spain (No. PI2016118) and consecutively at all new participating sites. All patients provided written informed consent.

Study design and patient population

The design of the European DuraGraft Registry has been described earlier [7]. Briefly, in this multicentre observational study, patients with an age ≥18 years undergoing an isolated CABG procedure or a combined CABG plus valve procedure with ≥1 saphenous vein grafts (SVGs) or ≥1 free arterial graft(s) were enrolled. All SVGs and free arterial grafts were treated intraoperatively with DuraGraft. Surgical techniques such as on- or off-pump, harvesting technique and grafting configuration were at the discretion of the surgeon. Patient follow-up was at 1 month and 1 year. This large study was initiated to collect outcome data and perform subgroup analysis after all-comer CABG surgery. For this subgroup analysis, patients who underwent isolated CABG surgery were grouped according to the presence of LMCAD as assessed during coronary angiography.

Graft treatment

DuraGraft (Marizyme, Jupiter, FL, USA) is a treatment solution applied intraoperatively to protect the structure and the function of the conduit’s endothelium against ischaemic and reperfusion damage during storage and graft placement [7]. All free grafts (i.e. right internal mammary artery, radial artery and SVGs) have been flushed and stored with the Duragraft solution prior to anastomosis (see Supplementary Material for further details).

Event definitions and adjudication

The primary end point is the incidence of major adverse cardiac events (MACE) defined as death, myocardial infarction (MI) or need for repeat revascularization (RR). The secondary outcome measure is the occurrence of major adverse cardiac and cerebrovascular events (MACCE) defined as death, MI, RR or stroke. Definitions of end point events are described in the Supplementary Material, Appendix. All primary and secondary end point-related adverse events were adjudicated by an independent clinical event committee comprised of 2 interventional cardiologists.

Statistical analysis

Continuous variables are summarized with mean ± standard deviation and categorical variables are presented with frequencies and percentages. The chi-square test was used to compare categorical variables and the t-test to compare continuous variables.

Patients were stratified by LMCAD for propensity score matching (PSM). A list of the factors taken into account for the construction of the propensity model is provided in Supplementary Material, Table S1. Baseline demographic variables to match LMCAD and no LMCAD patients were chosen a priori based upon a selection of EuroScore II variables. Using a standard propensity analysis using a 1:1 greedy-match algorithm, patients were matched based on the logit of the propensity score using a calliper of 0.20. In total, 1015 LMCAD patients were matched with 1015 no LMCAD patients for a total sample size of 2030 patients. Standardized differences of <10% were used to determine covariate balance between PSM groups (Supplementary Material, Fig. S4 and Supplementary Material, Table S1). Cox proportional hazards regression model was performed prior to and after propensity match, stratifying on the matched pair. Event rates were calculated using Kaplan–Meier methods prior to propensity match and compared by means of the log-rank test. Multivariable Cox proportional hazards regression analysis was performed to identify independent predictors of MACE (two-sided P-value 0.05 for significance). The multivariable model was built with candidate variables selected for clinical interest and satisfying the criterion of P < 0.2 in the univariate analysis. No correction has been made for multiple testing. All statistical analyses were performed using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA).

RESULTS

A total of 2964 patients were enrolled between December 2016 and August 2019 from 45 hospitals in 8 European countries (Supplementary Material). Of these patients, 2532 underwent isolated CABG (Fig. 1). Information on the presence of LMCAD was missing for 22 (0.7%) patients. LMCAD was present in 1033 (41.2%) patients and 1477 (58.8%) had no LMCAD. The mean age was higher in the LM group: 68.2 ± 9.0 vs 66.8 ± 9.3 years, P < 0.001 (Table 1). Moderate renal impairment was more prevalent in the LMCAD group (normal renal function, moderate and severe renal impairment: 43.2%, 45.2% and 11.6 vs 48.7%, 39.9% and 11.3%, respectively, P = 0.017). There were more patients at a critical operative state in the LMCAD group (3.5% vs 2.1%, P = 0.034). Overall, the median EuroSCORE II was higher in patients with LMCAD (1.5 [interquartile range 1.0–2.5] vs 1.3 [0.9–2.3], P < 0.001). There were less elective and more urgent and emergency procedures in the LMCAD group (68.9%, 28.9% and 2.1% vs 78.6%, 20.6% and 0.7%, P < 0.001). The percentage of patients who received an SVG was similar (91.0% vs 88.6%, P = 0.054) (Table 2). The left (92.4% vs 93.4%, P = 0.38) and right internal mammary arteries (16.5% vs 18.5%, P = 0.19) were used equally often, but the radial artery was used less often in LMCAD patients (9.4% vs 12.7%, P = 0.011). The total number of grafts per patient was less in the LMCAD group (2.70 ± 0.78 vs 2.77 ± 0.80, P = 0.033), due to less arterial grafts used in the LMCAD group (1.19 ± 0.56 vs 1.25 ± 0.60, P = 0.018). There were less distal anastomoses performed (2.95 ± 0.86 vs 3.07 ± 0.90, P = 0.001). The circumflex territory was revascularized more often (87.5% vs 81.4%, P < 0.001) and the right coronary artery territory less often (63.6% vs 75.2%, P < 0.001) in patients with LMCAD.

Figure 1:

Study flowchart.

Table 1:

Baseline patient characteristics for isolated coronary artery bypass grafting patients with and without left main coronary artery disease

Characteristics	LMCAD, n = 1033	No LMCAD, n = 1477	P-Value
Age (years), mean ± SD (n)	68.2 ± 9.0 (1033)	66.8 ± 9.3 (1477)	<0.001
Male % (n)	82.7 (854/1033)	82.2 (1214/1477)	0.76
Caucasian ethnicity % (n)	90.1 (929/1031)	86.2 (1269/1472)	0.003
BMI (kg/m2), mean ± SD (n)	28.3 ± 4.3 (1032)	28.7 ± 4.5 (1477)	0.008
(Ex)smoker % (n)	60.4 (623/1032)	63.1 (931/1476)	0.17
Diabetes mellitus % (n)	43.1 (445/1033)	45.9 (678/1477)	0.16
Insulin dependent diabetes mellitus % (n)	13.4 (138/1033)	15.4 (228/1477)	0.15
Hypertension % (n)	85.4 (877/1027)	84.3 (1237/1467)	0.46
Dyslipidaemia % (n)	80.0 (817/1021)	75.3 (1100/1461)	0.006
Renal impairment % (n)			0.017
Normal function	43.2 (446/1033)	48.7 (720/1477)
Moderate impairment	45.2 (467/1033)	39.9 (590/1477)
Severe impairment	11.6 (120/1033)	11.3 (167/1477)
Cerebrovascular disease % (n)	8.8 (91/1033)	7.8 (115/1477)	0.36
Peripheral vascular disease % (n)	19.2 (196/1023)	14.3 (209/1460)	0.001
Pulmonary hypertension % (n)	7.6 (79/1033)	8.5 (126/1477)	0.43
Previous myocardial infarction % (n)	43.0 (442/1027)	42.2 (620/1468)	0.69
Angina status % (n)			0.001
CCS 0–II	64.1 (662/1033)	69.2 (1022/1477)
CCS III	24.5 (253/1033)	23.6 (349/1477)
CCS IV	11.4 (118/1033)	7.2 (106/1477)
NYHA classification			0.52
NYHA I–II	75.6 (781/1033)	75.6 (1117/1477)
NYHA III	20.3 (210/1033)	21.1 (312/1477)
NYHA IV	4.1 (42/1033)	3.3 (48/1477)
Atrial fibrillation/flutter % (n)	8.0 (83/1033)	7.6 (112/1477)	0.68
Previous PCI % (n)	24.6 (254/1033)	24.5 (362/1477)	0.96
Previous cardiac surgery % (n)	1.1 (11/1033)	1.2 (18/1477)	0.72
Extent of coronary vessel disease % (n)			0.001
1-Vessel disease	3.1 (32/1026)	1.1 (16/1475)
2-Vessel disease	17.1 (175/1026)	15.7 (232/1475)
3-Vessel disease	79.8 (819/1026)	83.2 (1227/1475)
Left ventricular ejection fraction % (n)			0.31
Good (>50%)	67.9 (701/1033)	68.0 (1004/1477)
Fair (31–50%)	27.2 (281/1033)	27.2 (401/1477)
Poor (21–30%)	3.5 (36/1033)	4.1 (61/1477)
Very poor (≤20%)	1.5 (15/1033)	0.7 (11/1477)
Critical preoperative state % (n)	3.5 (36/1033)	2.1 (31/1478)	0.034
EuroSCORE II, median (IQR)	1.5 (1.0–2.5) 1033	1.3 (0.9–2.3) 1477	<0.001

BMI: body mass index; CCS: Canadian Cardiovascular Society; IQR: interquartile range; LMCAD: left main coronary artery disease; NYHA: New York Heart Associatio; PCI: percutaneous coronary intervention; SD: standard deviation.

Table 2:

Surgical characteristics in isolated coronary artery bypass grafting patients with and without left main coronary artery disease

Characteristics	LMCAD, n = 1033	No LMCAD, n = 1477	P-Value
Operative status % (n)			<0.001
Elective	68.9 (712/1033)	78.6 (1161/1477)
Urgent	28.9 (299/1033)	20.6 (305/1477)
Emergency	2.1 (22/1033)	0.7 (11/1477)
Off-pump	15.2 (157/1033)	18.1 (268/1477)	0.053
Vein harvesting % (n)			0.077
Open	74.1 (723/976)	70.4 (961/1365)
Endoscopic	13.5 (132/976)	15.6 (213/1365)
Both open and endoscopic	3.0 (29/976)	3.2 (43/1365)
Bridge technique	8.6 (84/976)	8.8 (120/1365)
No touch	0.8 (8/976)	2.1 (28/1365)
Graft use per patient % (n)
Saphenous vein	91.0 (939/1032)	88.6 (1306/1474)	0.054
LIMA	92.4 (954/1032)	93.4 (1376/1474)	0.38
RIMA	16.5 (170/1032)	18.5 (273/1474)	0.19
Radial artery	9.4 (97/1032)	12.7 (187/1474)	0.011
Number of total grafts per patient, mean ± SD (n)	2.70 ± 0.78 (1032)	2.77 ± 0.80 (1474)	0.033
Number of arterial grafts per patient, mean ± SD (n)	1.19 ± 0.56 (1032)	1.25 ± 0.60 (1474)	0.018
Number of vein grafts per patient, mean ± SD (n)	1.51 ± 0.84 (1032)	1.52 ± 0.92 (1474)	0.74
Number of distal anastomoses, mean ± SD (n)	2.95 ± 0.86 (1025)	3.07 ± 0.90 (1461)	0.001
Coronary territory supplied by graft % (n)
Left anterior descending	94.9 (973/1025)	95.6 (1397/1461)	0.42
Circumflex	87.5 (897/1025)	81.4 (1189/1461)	<0.001
Right coronary artery	63.6 (652/1025)	75.2 (1098/1461)	<0.001
Complete revascularization % (n)	80.9 (834/1031)	81.3 (1200/1476)	0.80
Duration of surgery (min), mean ± SD (n)	256 ± 76 (940)	256 ± 78 (1369)	0.93

LIMA: left internal mammary artery; LMCAD: left main coronary artery disease; RIMA: right internal mammary artery; SD: standard deviation.

At 1 year, the MACE rate was greater for LMCAD patients (8.2% vs 5.1%, P = 0.002), due to significantly higher rates for all-cause death (5.4% vs 3.4%, P = 0.016) and MI (3.0% vs 1.3%, P = 0.002). The incidence of all RR was equally low for both groups (2.8% vs 1.8%, P = 0.13).

The incidence of MACCE was higher in the LMCAD group (8.8% vs 6.6%, P = 0.043) due to less strokes in the LMCAD group (1.0% vs 2.4%, P = 0.011).

After PSM, the difference in the incidence of MACE remained higher in the LMCAD group (8.0% vs 5.2%, P = 0.015), but there was no longer a statistically significant difference in all-cause death (5.1% vs 3.7%, P = 0.10). The incidence of stroke remained lower in the LMCAD group (1.0% vs 2.4%, P = 0.017). The MACCE rate was not different at 1 year (8.5% vs 6.7%, P = 0.12) (see Tables 3 and 4, Fig. 2 and Supplementary Material, Fig. S1).

Figure 2:

Kaplan–Meier curves for isolated coronary artery bypass grafting patients with and without left main coronary artery disease after propensity score matching. (A) Major adverse cardiac event ; (B) all-cause deaths; (C) myocardial infarction; and (D) repeat revascularization.

Table 3:

Clinical outcomes at 1 year by Kaplan–Meier method for isolated coronary artery bypass grafting patients with and without left main disease (unmatched and matched)

	Unmatched			Matched
	Left main disease, n = 1033	No left main disease, n = 1477	P-Value log rank	Left main disease, n = 1011	No left main disease, n = 1011	P-Value log rank
MACE, (%) n	8.2 (84)	5.1 (74)	0.002	8.0 (79)	5.2 (52)	0.015
MACCE, (%) n	8.8 (90)	6.6 (97)	0.043	8.5 (85)	6.7 (67)	0.12
All death, (%) n	5.4 (55)	3.4 (50)	0.016	5.1 (51)	3.7 (37)	0.10
Cardiovascular death	5.0 (51)	2.7 (39)	0.002	4.8 (48)	3.1 (31)	0.053
Non-cardiovascular death	0.4 (4)	0.8 (11)	0.27	0.3 (3)	0.6 (6)	0.33
MI, (%) n	3.0 (30)	1.3 (18)	0.002	3.0 (29)	1.4 (14)	0.020
All RR, (%) n	2.8 (27)	1.8 (26)	0.13	2.7 (26)	1.6 (16)	0.12
Percutaneous RR, (%) n	2.5 (24)	1.5 (21)	0.086	2.4 (23)	1.3 (13)	0.090
Surgical RR, (%) n	0.3 (3)	0.3 (5)	0.84	0.3 (3)	0.3 (3)	0.99
Stroke, (%) n	1.0 (10)	2.4 (35)	0.011	1.0 (10)	2.4 (24)	0.017

MACCE: major adverse cardiac and cerebrovascular events; MACE: major adverse cardiac events; MI: myocardial infarction; RR: repeat revascularization.

Table 4:

Hazard ratios for a clinical event at 1 year by Kaplan–Meier method for isolated coronary artery bypass grafting patients with and without left main coronary artery disease

	Unmatched		Matched
	HR (95% CI)	P-Value	HR (95% CI)	P-Value
MACE	1.65 (1.21–2.26)	0.002	1.53 (1.08–2.18)	0.016
MACCE	1.34 (1.01–1.78)	0.044	1.27 (0.93–1.76)	0.13
All death	1.59 (1.08–2.33)	0.017	1.39 (0.91–2.12)	0.13
Cardiovascular death	1.88 (1.24–2.86)	0.003	1.55 (0.99–2.44	0.055
Non-cardiovascular death	0.53 (0.17–1.67)	0.28	0.51 (0.13–2.03)	0.33
MI	2.41 (1.35–4.33)	0.003	2.08 (1.10–3.95)	0.024
All RR	1.51 (0.87–2.58)	0.14	1.63 (0.87–3.05)	0.12
Percutaneous RR	1.66 (0.92–2.98)	0.090	1.78 (0.90–3.53)	0.095
Surgical RR	0.9 (0.2–3.6)	0.84	1.01 (0.20–5.02)	0.99
Stroke	0.41 (0.20–0.83)	0.014	0.42 (0.20–0.87)	0.021

The unmatched cohort includes 1033 patients from the LMCAD group and 1477 patients from the non-LM group. The propensity score matched cohort includes 1011 patients from each group.

CI: confidence interval; HR: hazard ratio; LM: left main; LMCAD: left main coronary artery disease; MACCE: major adverse cardiac and cerebrovascular events; MACE: major adverse cardiac events; MI: myocardial infarction; RR: repeat revascularization.

In the multivariable analysis, age (per increments of 10 years), extracardiac arteriopathy, critical preoperative state, ejection fraction ≤50% and LMCAD were independent predictors for MACE at 1 year after PSM (Supplementary Material, Table S2 and Fig. 3).

Figure 3:

Hazard ratios for a major adverse cardiac event at 1 year after adjustment for differences in baseline patient characteristics by multivariable Cox proportional hazards regression.

A separate risk analysis for stroke revealed age (per increments of 10 year), peripheral vascular disease and LMCAD as predictors for stroke at 1 year after PSM (Supplementary Material, Table S3).

30-Day landmark Kaplan–Meier curves for isolated CABG patients with and without LMCAD after PSM are provided in Supplementary Material, Figs. S2 and S3.

DISCUSSION

Due to the large amount of myocardium at risk and high ischaemic burden, CAD with the involvement of the LM stem is considered to be associated with significant morbidity and mortality [1, 2]. Similarly, surgical revascularization of patients with LMCAD is associated with a higher risk for adverse cardiac events. However, clinical outcomes have improved over the last decade [4, 5] and published results regarding LMCAD as a risk factor for CABG vary.

The 2018 European Society of Cardiology/ European Association for Cardio-Thoracic Surgery (ESC/EACTS) Guidelines on myocardial revascularization maintained a class I, level A recommendation for CABG in all patients with LMCAD regardless of anatomic complexity, while on the basis of 3-year data of the EXCEL [8] trial and the NOBLE [9] trial, percutaneous coronary intervention was issued a class I, level A indication in patients with LMCAD and low anatomical complexity, i.e. a low SYNTAX score 0–22 (the SYNTAX score reflects a comprehensive angiographic assessment of the coronary vasculature with higher scores indicating a more complex coronary anatomy) [2]. Following the publication of the 5-year outcomes of the EXCEL trial [10], a heated controversy over the interpretation of the results erupted, ultimately leading to a formal withdrawal of EACTS’s support for their current treatment recommendations for LMCAD and further revision of their current guidelines due to a range of scientific, statistical, and professional issues raised in the conduct of the EXCEL trial and the guideline development process [11]. This was echoed by other leading international societies demanding open data and transparent analyses [12, 13].

In light of these events, to that end, we performed a subgroup analysis of the DuraGraft registry to evaluate the impact of LMCAD on clinical outcomes after isolated CABG. At 1 year, the MACE rate was significantly higher in LMCAD patients, due to statistically significant higher all-cause death and MI rates, and numerically more RR. Also, the MACCE rate was numerically higher, but failed to reach statistical significance due to less strokes in the LM group. After PSM to adjust for differences in baseline characteristics, the MACE rate remained significantly higher in the LMCAD group, while the MACCE rate was comparable between the groups. There was no longer a difference in all-cause death, but the stroke rate in the LMCAD group remained lower. Accordingly, LMCAD is an independent risk factor for MACE in the multivariable analysis.

Whether LMCAD itself is an independent risk factor for morbidity and mortality in CABG remains controversial. Cosgrove et al. [14] concluded that LMCAD has been neutralized as an independent risk factor for operative mortality. Accordingly, d'Allonnes et al. [15] reported that postoperative outcomes of isolated LMCAD patients did not differ significantly from other CABG patients in a small study. In the ASCERT study including over 348 000 patients, Shahian et al. presented mortality hazard ratios for 4 postoperative time intervals. LMCAD was found to have a modest mortality impact in the intervals after 30 days with hazard ratios between 1.05 and 1.12 [16]. Jonsson et al. [17] analysed LMCAD’s effect on early and late mortality after CABG during a 30-year period of 1970–1999 from a Swedish national registry. While the proportion of patients with LMCAD increased from 7% during the 1970s to 26% in 1999, a neutralization of LMCAD as an independent risk factor for early and late mortality during the late 90s was observed. Surgical advancements and strict adherence to guideline-directed medical therapy most likely contributed to this observation. Similarly, significant improvement in event-free survival after CABG for LMCAD at 3 years was noted between the SYNTAX and EXCEL trials [4].

Suzuki et al. [5] investigated the impact of LMCAD on outcomes of off-pump CABG in a propensity matched analysis of 237 patients. With freedom from all-cause death at 6 years of 87.3% and 60.7% in the LMCAD group and non-LMCAD group, respectively (P < 0.17), they concluded that LMCAD was not a risk factor for mortality after off-pump CABG.

Conversely, a substudy of the SYNTAX trial comparing 348 patients with LMCAD versus 549 patients without LMCAD in the surgical arm revealed a trend towards an increased MACCE rate of 13.7% in the LMCAD group vs 11.5% in the non-LMCAD group at 1 year of follow-up [18]. In addition, the hazard ratio of LMCAD for mortality at 4 years was 1.55 (95% confidence interval 0.98–2.46, P = 0.059) in the univariate analysis and was not considered an independent variable for mortality [19].

The above-mentioned studies assessed LMCAD’s impact on mortality. LMCAD was only an independent risk factor in the larger ASCERT study, despite mortality events accumulating during multiple postoperative years in some of the other studies. Insufficient power in relationship to the end point likely underlies this. Our analysis used a composite end point (MACE) and may be more sensitive to detect differences between the 2 groups.

Another observation of our study was the lower stroke rate at 1 year in LMCAD patients compared to non-LMCAD patients with note that the majority of the strokes occurred in the first 30 postoperative days. The lower stroke rate in the LMCAD group contrasts earlier reports in which LMCAD was identified as an independent risk factor for a stroke [20]. The cause of the lower stroke rate in our LMCAD group is speculative and might be related to differences in operative techniques such as the use of more grafts in the non-LMCAD group. The use of off-pump was not significantly different after matching and is therefore unlikely a causative factor.

A different perspective on the discussion as to whether LMCAD poses as an independent risk factor has recently been brought up by Gaudino and others [21, 22]. They argue that LMCAD is not a separate entity defined as a biological and statistical subgroup but rather share the same underlying cause as in non-LMCAD and, thus, should be treated accordingly. The majority of patients with LMCAD in the recent LM-trials also have CAD affecting other territories, and isolated LMCAD accounts for only 12.9–14.6% of the patients [10, 21, 23]. In a recent editorial, Gomes attributes the outcomes in patients with LMCAD to the pathophysiologic understanding of CAD in which the prognosis of patients with CAD is mostly affected by acute coronary syndromes that occur as a result of rupture or erosion of non-flow limiting stenosis, rather than by the extent of ischaemia caused at the site of severe stenosis seen on a conventional angiogram [22, 24]. Gomes further regards LMCAD to be a marker for the atherosclerotic burden, such that advanced atherosclerotic CAD carries an increased risk of MI and death [22]. This is in line with a previous study of Doonan et al., which suggested that LMCAD is generally linked to an overall greater calcific burden of the aorta and other arteries, and also recent evidence from the ISCHEMIA trial seems to be in support of this hypothesis [3, 25].

The earlier mentioned controversy over the 5-year results of the EXCEL trial resulting in the withdrawal of EACTS’s support for the current guidelines and the recent publication of the US Coronary Revascularization Guidelines without the endorsement of the 2 major surgical societies have further added fuel to the fire [26, 27]. Instead, it is now of utmost important that the cardiovascular society with all parties involved finds an honest and transparent collaboration to overcome these controversies with the ultimate goal of achieving the best outcomes for patients with or without LMCAD.

Limitations

All inherent limitations of registries also apply here. Randomized clinical trials are traditionally considered to be a prerequisite to establish the risk–benefit ratio for a given therapy; however, they are usually limited by strict inclusion criteria not necessarily reflecting the general target population and common practice. Well-designed all-comer, observational registries generating longitudinal data of large sample size may complement randomized clinical trials. However, protocol requirements are typically less stringent as in the present study where systematic biomarker collection was not performed, and therefore, the rates for MI may also be underestimated. Furthermore, a systematic collection of the SYNTAX score was not part of the registry protocol and thus missing in the outcome analysis. The Duragraft solution was applied in both groups; however, even if unlikely, it cannot be excluded if and to what extent Duragraft may have influenced the outcomes in both groups since we had no control group without the application of Duragraft.

Unmeasured confounding variables could not be included in the PSM. For more insight from this dataset, longer follow-up is highly warranted and planned.

CONCLUSION

In conclusion, in this large multicentre European registry, LMCAD patients have higher rates of myocardial infarctions, which are tied to higher mortality rates than non-LMCAD patients implying a more severe disease manifestation but lower perioperative stroke rates probably due to a lower burden of non-cardiac atherosclerosis. LMCAD was an independent risk factor for MACE in patients undergoing isolated CABG. To the contrary, the risk of stroke was lower in LMCAD patients. Further research as to whether LMCAD poses as an independent risk factor for outcomes in CABG surgery is highly warranted since the evidence in the literature remains scarce and controversial.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

ACKNOWLEDGEMENT

The authors thank Clara Fitzgerald, Boston Clinical Research Institute, Boston, USA, for the statistical analysis.

Funding

This work was supported by Marizyme, Jupiter, Florida, USA.

Conflict of interest: Etem Caliskan, Sigrid Sandner, Martin Misfeld, Jose Aramendi, Sacha P. Salzberg, Yeong-Hoon Choi and Andreas Böning are members of the registry advisory committee (RAC). Louis P. Perrault is a member of the RAC and is a consultant for Marizyme. Maximilian Y. Emmert is the principal investigator of the registry, the chair of the RAC and a consultant for Marizyme. Enrico Ferrari received research grants from Somalution, a Marizyme company. Other authors have nothing to disclose.

Data Availability Statement

Data are held by the study sponsor and are available upon reasonable request.

Author contributions

Etem Caliskan: Conceptualization; Formal analysis; Investigation; Methodology; Supervision; Validation; Visualization; Writing—original draft; Writing—review & editing. Martin Misfeld: Writing—review & editing. Sigrid Sandner: Writing—review & editing. Andreas Böning: Writing—review & editing. Jose Aramendi: Writing—review & editing. Sacha P. Salzberg: Writing—review & editing. Yeong-Hoon Choi: Writing—review & editing. Louis P. Perrault: Writing—review & editing. Ilker Tekin: Writing—review & editing. Gregorio P. Cuerpo: Writing—review & editing. Luca P. Weltert: Writing—review & editing. Johannes Böhm: Writing—review & editing. Jose Lopez-Menendez: Writing—review & editing. Markus Krane: Writing—review & editing. José M. González-Santos: Writing—review & editing. Juan-Carlos Tellez: Writing—review & editing. Enrico Ferrari: Writing—review & editing. Tomas Holubec: Writing—review & editing. Maximilian Y. Emmert: Conceptualization; Formal analysis; Investigation; Methodology; Resources; Supervision; Validation; Visualization; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Nikolaos Bonaros and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

ABBREVIATIONS

ABBREVIATIONS
 
• CABG

  Coronary artery bypass grafting

 
• CAD

  Coronary artery disease

 
• EACTS

  European Association for Cardiothoracic Surgery

 
• LM

  Left main

 
• LMCAD

  Left main coronary artery disease

 
• MACCE

  Major adverse cardiac and cerebrovascular events

 
• MACE

  Major adverse cardiac events

 
• MI

  Myocardial infarction

 
• PSM

  Propensity score matching

 
• RR

  Repeat revascularization

 
• SVGs

  Saphenous vein grafts

Author notes

The European DuraGraft Registry investigators are listed in the Supplementary Material.

Presented at the Annual Meeting of the European Association for Cardiothoracic Surgery (EACTS) October 8, 2020 in Barcelona, SPAIN.