Residual SYNTAX score following coronary artery bypass grafting†

Abstract

INTRODUCTION

Despite current interventional and surgical procedures designed to achieve complete myocardial revascularization as the standard approach in coronary artery disease (CAD), one cannot always reach this goal [1–2]. Previous investigations reported higher mortality rates in series of patients with incomplete revascularization (IR) compared to those with complete revascularization (CR), for both percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) [3–6]. Nevertheless, in a recent study, Kim et al. [2] demonstrated that neither angiographically monitored CR with drug-eluting stents (DESs) nor surgery improved clinical outcomes in patients with multi-vessel disease. In addition, a Leipzig registry of coronary bypass patients showed that a ‘reasonable’ IR did not adversely affect survival [1]. These controversial conclusions about the impact and prognosis of completeness of revascularization on clinical outcomes are probably due to the different definitions existing in the literature [1]. The Synergy between PCI with TAXUS and Cardiac Surgery (SYNTAX) trial has introduced a more accurate method of quantifying and defining the complexity of CAD [7], which has been associated with clinical outcomes in both PCI and CABG settings [8, 9]. The SYNTAX score (SS) has recently been applied in a series of studies to determine the degree and complexity of the residual disease, quantified as the residual SYNTAX score (rSS), left untreated in patients undergoing an invasive coronary procedure [10–14]. Moreover, a subgroup analysis of the SYNTAX trial has shown that IR is associated with adverse events during follow-up after PCI but not after CABG [8]. It is thought that applying this method of quantifying the residual disease in patients undergoing CABG may be a useful tool for risk stratification.

The aim of the present study was therefore to determine the residual CAD after CABG measured using the SS and its relationship with outcomes 1 year after the operation.

MATERIALS AND METHODS

Patients

The study population comprised 1608 consecutive patients who underwent isolated first-time CABG for multi-vessel coronary disease between May 2004 and April 2015. The primary outcome of interest was major adverse cardiac and cerebrovascular events (MACCE) at 1 year, including all-cause mortality, myocardial infarction, cerebrovascular accident (CVA) and repeat revascularization. Mortality was defined as death from any cause, cardiac and non-cardiac. Myocardial infarction was defined as cardiac enzyme levels three times or more the upper normal limit and associated with the presence of ischaemic symptoms or new electrocardiographic changes. A CVA was defined as a stroke, transient ischaemic attack, or a coma. All clinical, angiographic, operative and outcome data are routinely collected prospectively from independent research personnel and available from a dedicated surgical database. As part of our postoperative practice, patients are clinically followed up at 1 and 12 months via routine outpatient visits. Follow-up was 100% complete. When data were not available from our outpatient clinical records, patients were contacted via telephone and outcomes were obtained.

Operative details

After having a full median sternotomy, all patients underwent on-pump CABG using standard techniques. The left internal mammary artery was used to graft the left anterior descending artery (LADA) and revascularization was completed using saphenous vein grafts (SVG) to the right coronary and left circumflex artery segments. In 192/1608 (12%) patients, a bilateral internal mammary artery revascularization was performed to the circumflex (N = 191/192) or right territory (N = 1/192) with or without adjunct SVG. Aspirin was prescribed indefinitely for all patients after the operation.

SYNTAX score calculation

For the purpose of this study, three experienced interventional cardiologists retrospectively reviewed preoperative diagnostic coronary angiograms of all patients and calculated the baseline SYNTAX score (bSS) by consensus using the SS algorithm [7]. A postoperative SS (rSS) was then assessed, in agreement with the surgeons, by comparing the diagnostic angiograms with the surgical procedure reports. In case of disagreement, the opinion of a fourth analyst was obtained, and the final decision was achieved by consensus. In particular, in the case of significant left main disease and the left internal mammary artery to the LADA, the rSS was calculated taking into account the left main lesion (subtracting the impact of the grafted LADA) and/or any other non-revascularized vessel of the circumflex. Moreover, despite the graft received, individual vessel disease was still quantified in terms of vessel characteristics (tortuosity, calcification) and/or any residual disease distal to the graft anastomosis. Patients were then divided into three groups according to tercile cut points of rSS: low (rSS_low), intermediate (rSS_mid) and high rSS (rSS_high).

Statistical analysis

Statistical analyses were performed using the Statistical Package for the Social Sciences, version 11.0 (SPSS Inc., Chicago, IL). Data are reported as counts and percentages for categorical variables and as mean ± standard deviation (SD) for continuous variables. Continuous variables were compared by analysis of variance. Comparisons between categorical variables were done by Fisher’s exact test. All variables were stratified according to groups of rSSs. The normality assumption was evaluated by the Kolmogorov–Smirnov test. The Spearman correlation coefficient was used to assess correlation between bSS and rSS. Cumulative event rates through follow-up were estimated by the Kaplan–Meier method, and the log-rank test was used to evaluate differences between groups. The effect of rSS on 1-year MACCE was investigated by means of the Cox model by forcing rSS as a categorical variable according to tercile cut points. The final model was adjusted for other variables associated (P < 0.1) with MACCE by univariate analysis. Due to the anticipated correlation between rSS and bSS, we repeated the analysis, forcing bSS instead of rSS. The discrimination power of the two models was compared using C-statistics (areas under the receiver operating characteristics (ROC) curves). The Schoenfeld residuals test was used to check for proportional hazard violation. Variance inflation factors were used to assess multicollinearity; factors with a value >4 were excluded. All statistical tests were two-sided and a P-value of <0.05 was considered statistically significant.

RESULTS

Baseline characteristics

The mean preoperative SS was 26.6 (SD 9.4) (range 10.1–53) and after CABG, the mean residual SS was 15.3 (SD 8.4) (range 0–34). The correlation (Spearman coefficient 0.96, P < 0.001) and distribution of the preoperative bSS and the postoperative rSS are illustrated in Fig. 1. The stratification of patients according to tercile cut points resulted in three groups: rSS_low (0–11, N = 537), rSS_mid (>11–18.5, N = 539) and rSS_high (>18.5, N = 532). Table 1 shows the baseline and operative characteristics stratified for the rSS groups. A higher rSS was associated with progressively increasing clinical comorbidity such as older age (P < 0.001), chronic obstructive pulmonary disease (P = 0.037), previous CVA (P = 0.037), previous MI (P = 0.048), greater EuroSCORE II (P = 0.031), higher number of patients taking β-blockers (P < 0.001) or antiplatelet therapies (P = 0.022) and a higher rate of urgent/emergent operations (P = 0.048). Similarly, a greater rSS was associated with progressively higher bSSs (P < 0.001).

Figure 1:

Correlation between the baseline and the residual SYNTAX score [Local Regression Smoothing (LOESS) trend line in blue].

Open in new tabDownload slide

Table 1:

Open in new tab

Baseline and operative characteristics stratified for residual SYNTAX score groups

Variable .	rSSlow  (N=537) .	rSSmid  (N=539) .	rSShigh  (N=532) .	P-value .
Age, years	65.5  (SD: 9.7)	67.0  (SD: 10 .1)	69.1  (SD: 9.9)	<0 .001
Baseline SYNTAX score	15.8  (SD: 3.0)	25.5  (SD: 3.7)	38.5  (SD: 10 .4)	<0 .001
Residual SYNTAX score	5.9  (SD: 2.8)	14.9  (SD: 1.9)	24.9  (SD: 4.6)	<0 .001
Male gender, n	472  (87.9%)	453  (84%)	463  (87%)	0 .156
Diabetes mellitus, n	264  (49.2%)	244  (45.3%)	264  (49.6%)	0 .292
Hypertension, n	419  (78%)	405  (75.1%)	421  (79.1%)	0 .212
Creatinine clearance, ml/min	58.6  (SD: 11.8)	59.1  (SD: 12.4)	59.8  (SD: 13.1)	0 .344
COPD, n	62  (11.5%)	54  (10%)	80  (15%)	0 .037
Dyslipidaemia, n	397  (73.9%)	394  (73.1%)	387  (72.7%)	0 .904
History of smoking, n	306  (57%)	315  (58.4%)	317  (59.6%)	0 .688
Previous CVA, n	3  (0 .6%)	59  (10 .9%)	69  (13%)	0 .037
Previous MI, n	293  (54.6%)	332  (61.6%)	298  (56%)	0 .048
Unstable angina, n	178  (33.1%)	169  (31.4%)	166  (31.2%)	0 .321
NYHA class III or IV, n	109  (20 .3%)	112  (20 .8%)	112  (21.1%)	0 .841
LVEF, %	46.9  (SD: 13.7)	46.8  (SD: 13.6)	46.6  (SD: 14.3)	0 .471
EuroSCORE II	2.1  (SD: 1.7)	2.7  (SD: 1.9)	3.2  (SD: 1.5)	0 .031
Preoperative medications, n
ACE inhibitors	249  (46.4%)	281  (52.1%)	259  (48.7%)	0 .164
Beta-blockers	310  (57.7%)	349  (64.7%)	368  (69.2%)	<0 .001
Antiplatelet therapies	135  (25.1%)	145  (26.9%)	170  (32%)	0 .022
Three vessel disease, n	370  (68.9%)	370  (68.7)	379  (71.2%)	0 .421
Left main disease, n	230  (42.8%)	227  (42.1%)	249  (46.8%)	0 .391
Urgency/emergency, n	90  (16.8%)	92  (17.1%)	117  (22%)	0 .048
CPB time, min	113.2  (SD: 53.9)	116.2  (SD: 54.9)	117.2  (SD: 60 .9)	0 .525
Aortic cross-clamp time, min	80.8  (SD: 38.3)	80.6  (SD: 34.7)	82.9  (SD: 37.2)	0 .551
Total number of grafts, n	2.7  (SD: 0 .9)	2.5  (SD: 0 .9)	2.3  (SD: 0 .9)	0 .102
Number of arterial grafts, n	1  (SD: 0 .2)	1  (SD: 0 .2)	1  (SD: 0 .2)	0 .856

Variable .	rSSlow  (N=537) .	rSSmid  (N=539) .	rSShigh  (N=532) .	P-value .
Age, years	65.5  (SD: 9.7)	67.0  (SD: 10 .1)	69.1  (SD: 9.9)	<0 .001
Baseline SYNTAX score	15.8  (SD: 3.0)	25.5  (SD: 3.7)	38.5  (SD: 10 .4)	<0 .001
Residual SYNTAX score	5.9  (SD: 2.8)	14.9  (SD: 1.9)	24.9  (SD: 4.6)	<0 .001
Male gender, n	472  (87.9%)	453  (84%)	463  (87%)	0 .156
Diabetes mellitus, n	264  (49.2%)	244  (45.3%)	264  (49.6%)	0 .292
Hypertension, n	419  (78%)	405  (75.1%)	421  (79.1%)	0 .212
Creatinine clearance, ml/min	58.6  (SD: 11.8)	59.1  (SD: 12.4)	59.8  (SD: 13.1)	0 .344
COPD, n	62  (11.5%)	54  (10%)	80  (15%)	0 .037
Dyslipidaemia, n	397  (73.9%)	394  (73.1%)	387  (72.7%)	0 .904
History of smoking, n	306  (57%)	315  (58.4%)	317  (59.6%)	0 .688
Previous CVA, n	3  (0 .6%)	59  (10 .9%)	69  (13%)	0 .037
Previous MI, n	293  (54.6%)	332  (61.6%)	298  (56%)	0 .048
Unstable angina, n	178  (33.1%)	169  (31.4%)	166  (31.2%)	0 .321
NYHA class III or IV, n	109  (20 .3%)	112  (20 .8%)	112  (21.1%)	0 .841
LVEF, %	46.9  (SD: 13.7)	46.8  (SD: 13.6)	46.6  (SD: 14.3)	0 .471
EuroSCORE II	2.1  (SD: 1.7)	2.7  (SD: 1.9)	3.2  (SD: 1.5)	0 .031
Preoperative medications, n
ACE inhibitors	249  (46.4%)	281  (52.1%)	259  (48.7%)	0 .164
Beta-blockers	310  (57.7%)	349  (64.7%)	368  (69.2%)	<0 .001
Antiplatelet therapies	135  (25.1%)	145  (26.9%)	170  (32%)	0 .022
Three vessel disease, n	370  (68.9%)	370  (68.7)	379  (71.2%)	0 .421
Left main disease, n	230  (42.8%)	227  (42.1%)	249  (46.8%)	0 .391
Urgency/emergency, n	90  (16.8%)	92  (17.1%)	117  (22%)	0 .048
CPB time, min	113.2  (SD: 53.9)	116.2  (SD: 54.9)	117.2  (SD: 60 .9)	0 .525
Aortic cross-clamp time, min	80.8  (SD: 38.3)	80.6  (SD: 34.7)	82.9  (SD: 37.2)	0 .551
Total number of grafts, n	2.7  (SD: 0 .9)	2.5  (SD: 0 .9)	2.3  (SD: 0 .9)	0 .102
Number of arterial grafts, n	1  (SD: 0 .2)	1  (SD: 0 .2)	1  (SD: 0 .2)	0 .856

Continuous variables are shown as mean and standard deviation; categorical variables are shown as absolute number and percentage.

Significant P-values in bold.

rSS: residual SYNTAX score; COPD: chronic obstructive pulmonary disease; CVA: cerebrovascular accident; MI: myocardial infarction; NYHA: New York Heart Association; LVEF: left ventricular ejection fraction; ACE: angiotensin converting enzyme; CPB: cardiopulmonary bypass; SD: standard deviation.

Table 1:

Open in new tab

Baseline and operative characteristics stratified for residual SYNTAX score groups

Variable .	rSSlow  (N=537) .	rSSmid  (N=539) .	rSShigh  (N=532) .	P-value .
Age, years	65.5  (SD: 9.7)	67.0  (SD: 10 .1)	69.1  (SD: 9.9)	<0 .001
Baseline SYNTAX score	15.8  (SD: 3.0)	25.5  (SD: 3.7)	38.5  (SD: 10 .4)	<0 .001
Residual SYNTAX score	5.9  (SD: 2.8)	14.9  (SD: 1.9)	24.9  (SD: 4.6)	<0 .001
Male gender, n	472  (87.9%)	453  (84%)	463  (87%)	0 .156
Diabetes mellitus, n	264  (49.2%)	244  (45.3%)	264  (49.6%)	0 .292
Hypertension, n	419  (78%)	405  (75.1%)	421  (79.1%)	0 .212
Creatinine clearance, ml/min	58.6  (SD: 11.8)	59.1  (SD: 12.4)	59.8  (SD: 13.1)	0 .344
COPD, n	62  (11.5%)	54  (10%)	80  (15%)	0 .037
Dyslipidaemia, n	397  (73.9%)	394  (73.1%)	387  (72.7%)	0 .904
History of smoking, n	306  (57%)	315  (58.4%)	317  (59.6%)	0 .688
Previous CVA, n	3  (0 .6%)	59  (10 .9%)	69  (13%)	0 .037
Previous MI, n	293  (54.6%)	332  (61.6%)	298  (56%)	0 .048
Unstable angina, n	178  (33.1%)	169  (31.4%)	166  (31.2%)	0 .321
NYHA class III or IV, n	109  (20 .3%)	112  (20 .8%)	112  (21.1%)	0 .841
LVEF, %	46.9  (SD: 13.7)	46.8  (SD: 13.6)	46.6  (SD: 14.3)	0 .471
EuroSCORE II	2.1  (SD: 1.7)	2.7  (SD: 1.9)	3.2  (SD: 1.5)	0 .031
Preoperative medications, n
ACE inhibitors	249  (46.4%)	281  (52.1%)	259  (48.7%)	0 .164
Beta-blockers	310  (57.7%)	349  (64.7%)	368  (69.2%)	<0 .001
Antiplatelet therapies	135  (25.1%)	145  (26.9%)	170  (32%)	0 .022
Three vessel disease, n	370  (68.9%)	370  (68.7)	379  (71.2%)	0 .421
Left main disease, n	230  (42.8%)	227  (42.1%)	249  (46.8%)	0 .391
Urgency/emergency, n	90  (16.8%)	92  (17.1%)	117  (22%)	0 .048
CPB time, min	113.2  (SD: 53.9)	116.2  (SD: 54.9)	117.2  (SD: 60 .9)	0 .525
Aortic cross-clamp time, min	80.8  (SD: 38.3)	80.6  (SD: 34.7)	82.9  (SD: 37.2)	0 .551
Total number of grafts, n	2.7  (SD: 0 .9)	2.5  (SD: 0 .9)	2.3  (SD: 0 .9)	0 .102
Number of arterial grafts, n	1  (SD: 0 .2)	1  (SD: 0 .2)	1  (SD: 0 .2)	0 .856

Variable .	rSSlow  (N=537) .	rSSmid  (N=539) .	rSShigh  (N=532) .	P-value .
Age, years	65.5  (SD: 9.7)	67.0  (SD: 10 .1)	69.1  (SD: 9.9)	<0 .001
Baseline SYNTAX score	15.8  (SD: 3.0)	25.5  (SD: 3.7)	38.5  (SD: 10 .4)	<0 .001
Residual SYNTAX score	5.9  (SD: 2.8)	14.9  (SD: 1.9)	24.9  (SD: 4.6)	<0 .001
Male gender, n	472  (87.9%)	453  (84%)	463  (87%)	0 .156
Diabetes mellitus, n	264  (49.2%)	244  (45.3%)	264  (49.6%)	0 .292
Hypertension, n	419  (78%)	405  (75.1%)	421  (79.1%)	0 .212
Creatinine clearance, ml/min	58.6  (SD: 11.8)	59.1  (SD: 12.4)	59.8  (SD: 13.1)	0 .344
COPD, n	62  (11.5%)	54  (10%)	80  (15%)	0 .037
Dyslipidaemia, n	397  (73.9%)	394  (73.1%)	387  (72.7%)	0 .904
History of smoking, n	306  (57%)	315  (58.4%)	317  (59.6%)	0 .688
Previous CVA, n	3  (0 .6%)	59  (10 .9%)	69  (13%)	0 .037
Previous MI, n	293  (54.6%)	332  (61.6%)	298  (56%)	0 .048
Unstable angina, n	178  (33.1%)	169  (31.4%)	166  (31.2%)	0 .321
NYHA class III or IV, n	109  (20 .3%)	112  (20 .8%)	112  (21.1%)	0 .841
LVEF, %	46.9  (SD: 13.7)	46.8  (SD: 13.6)	46.6  (SD: 14.3)	0 .471
EuroSCORE II	2.1  (SD: 1.7)	2.7  (SD: 1.9)	3.2  (SD: 1.5)	0 .031
Preoperative medications, n
ACE inhibitors	249  (46.4%)	281  (52.1%)	259  (48.7%)	0 .164
Beta-blockers	310  (57.7%)	349  (64.7%)	368  (69.2%)	<0 .001
Antiplatelet therapies	135  (25.1%)	145  (26.9%)	170  (32%)	0 .022
Three vessel disease, n	370  (68.9%)	370  (68.7)	379  (71.2%)	0 .421
Left main disease, n	230  (42.8%)	227  (42.1%)	249  (46.8%)	0 .391
Urgency/emergency, n	90  (16.8%)	92  (17.1%)	117  (22%)	0 .048
CPB time, min	113.2  (SD: 53.9)	116.2  (SD: 54.9)	117.2  (SD: 60 .9)	0 .525
Aortic cross-clamp time, min	80.8  (SD: 38.3)	80.6  (SD: 34.7)	82.9  (SD: 37.2)	0 .551
Total number of grafts, n	2.7  (SD: 0 .9)	2.5  (SD: 0 .9)	2.3  (SD: 0 .9)	0 .102
Number of arterial grafts, n	1  (SD: 0 .2)	1  (SD: 0 .2)	1  (SD: 0 .2)	0 .856

Continuous variables are shown as mean and standard deviation; categorical variables are shown as absolute number and percentage.

Significant P-values in bold.

rSS: residual SYNTAX score; COPD: chronic obstructive pulmonary disease; CVA: cerebrovascular accident; MI: myocardial infarction; NYHA: New York Heart Association; LVEF: left ventricular ejection fraction; ACE: angiotensin converting enzyme; CPB: cardiopulmonary bypass; SD: standard deviation.

Study population analysis

As shown in Table 2, cumulative 1-year rates of MACCE were significantly increased among patients with higher rSSs (P < 0.001). More specifically, there was a significantly increased rate of complications such as MI, stroke and repeat revascularization, whereas mortality rates were similar both in hospital and at 1 year (Table 2). The Kaplan–Meier analysis in Fig. 2 shows a statistically significant difference for MACCE-free survival at 1 year between the different groups (log-rank test, P < 0.001). The estimated MACCE-free survival rate at 1 year for each group was 98.1% (SE: 1.6) for rSS_low, 95.5% (SE: 1.9) for rSS_mid and 90.5% (SE: 1.3) for rSS_high (Fig. 2).

Figure 2:

Kaplan–Meier analysis of MACCE-free survival (vertical axis starts at 80%).

Open in new tabDownload slide

Table 2:

Open in new tab

Incidence of MACCE stratified for residual SYNTAX score groups

Outcomes .	rSSlow (N=537) .	rSSmid (N=539) .	rSShigh (N=532) .	P-value .
In-hospital mortality rate	1  (0 .2%)	1  (0 .2%)	1  (0 .2%)	0 .867
One year mortality rate	4  (0 .7%)	7  (1.3%)	7  (1.3%)	0 .324
Myocardial infarction	2  (0 .4%)	5  (0 .9%)	13  (2.4%)	0 .046
Stroke	1  (0 .2%)	8  (1.5%)	16  (3%)	0 .002
Repeat revascularization	1  (0 .2%)	4  (0 .7%)	11  (2.1%)	<0 .001
Cumulative MACCE	8  (1.5%)	24  (4.5%)	47  (8.8%)	<0 .001

Outcomes .	rSSlow (N=537) .	rSSmid (N=539) .	rSShigh (N=532) .	P-value .
In-hospital mortality rate	1  (0 .2%)	1  (0 .2%)	1  (0 .2%)	0 .867
One year mortality rate	4  (0 .7%)	7  (1.3%)	7  (1.3%)	0 .324
Myocardial infarction	2  (0 .4%)	5  (0 .9%)	13  (2.4%)	0 .046
Stroke	1  (0 .2%)	8  (1.5%)	16  (3%)	0 .002
Repeat revascularization	1  (0 .2%)	4  (0 .7%)	11  (2.1%)	<0 .001
Cumulative MACCE	8  (1.5%)	24  (4.5%)	47  (8.8%)	<0 .001

Results are shown as absolute number and percentage.

Significant P-values in bold.

rSS: residual SYNTAX score; MACCE: major adverse cardiac and cerebrovascular events.

Table 2:

Open in new tab

Incidence of MACCE stratified for residual SYNTAX score groups

Outcomes .	rSSlow (N=537) .	rSSmid (N=539) .	rSShigh (N=532) .	P-value .
In-hospital mortality rate	1  (0 .2%)	1  (0 .2%)	1  (0 .2%)	0 .867
One year mortality rate	4  (0 .7%)	7  (1.3%)	7  (1.3%)	0 .324
Myocardial infarction	2  (0 .4%)	5  (0 .9%)	13  (2.4%)	0 .046
Stroke	1  (0 .2%)	8  (1.5%)	16  (3%)	0 .002
Repeat revascularization	1  (0 .2%)	4  (0 .7%)	11  (2.1%)	<0 .001
Cumulative MACCE	8  (1.5%)	24  (4.5%)	47  (8.8%)	<0 .001

Outcomes .	rSSlow (N=537) .	rSSmid (N=539) .	rSShigh (N=532) .	P-value .
In-hospital mortality rate	1  (0 .2%)	1  (0 .2%)	1  (0 .2%)	0 .867
One year mortality rate	4  (0 .7%)	7  (1.3%)	7  (1.3%)	0 .324
Myocardial infarction	2  (0 .4%)	5  (0 .9%)	13  (2.4%)	0 .046
Stroke	1  (0 .2%)	8  (1.5%)	16  (3%)	0 .002
Repeat revascularization	1  (0 .2%)	4  (0 .7%)	11  (2.1%)	<0 .001
Cumulative MACCE	8  (1.5%)	24  (4.5%)	47  (8.8%)	<0 .001

Results are shown as absolute number and percentage.

Significant P-values in bold.

rSS: residual SYNTAX score; MACCE: major adverse cardiac and cerebrovascular events.

As shown in Table 3, patients in the rSS_high were associated with an increased incidence of MACCE at 1 year relative to the rSS_low group after adjusting for other confounding factors (full-model C-statistic 0.83). When the analysis was repeated using bSS instead of rSS, bSS was also found to be a predictor of MACCE at 1 year, but with a trend towards a lower discrimination power of the full model (C-statistic 0.76, comparison between areas under the curves P = 0.070). The Schoenfeld residuals test showed that neither the bSS nor the rSS group violated the proportionality hazard assumption (P = 0.71 and 0.85, respectively, for bSS and rSS).

DISCUSSION

The results of this retrospective study showed that the rSS is a useful tool for risk stratification of patients who have undergone CABG and is associated with MACCE at 1 year. In addition, 1-year MACCE was related to other relevant variables such age, diabetes mellitus, gender, left ventricular ejection fraction and preoperative creatinine clearance.

It has long been recognized that IR is associated with poor short- and long-term outcomes following CABG [15–23]. In the past, surgeons strived to graft every diseased vessel, sometimes placing 6 or 7 grafts in a patient [3–5]. On the other end, it has also been clearly demonstrated that a ‘reasonable’ IR may not influence long-term mortality of patients undergoing CABG [1]. In addition, outcomes of patients undergoing either DES PCI or surgery could not be discerned on the basis of the completeness of anatomical revascularization [2]. That said, the term IR has never been clearly defined but has been coined to denote lack of revascularization to any angiographically diseased vessel irrespective of its size, diffusely diseased and by definition unsuitable for grafting or whether it is occluded and either not or poorly visualized when filled retrogradely via collaterals. In surgical series, IR varies between 9% and 39%, which is, among others, according to a non-standardized definition [15–23] and because of different CABG era reports [21]. In an endeavour to clarify such a complex and grey area, we applied the residual SYNTAX scoring system used in cardiology [8, 11–14, 23]. Farooq et al. [23] recently developed the CABG SS in a pilot study of 115 patients who underwent coronary angiography 15 months after surgery. Using a complex calculation, they showed a trend towards higher adverse events in patients with high SS [23]. In the present study, we retrospectively assessed for the first time the rSS post CABG surgery and determined its impact on 1-year outcomes. Calculation of the SS after surgery was measured on the coronary vessels left untreated by correlating the actual revascularization with the planned preoperative strategy. In addition, we considered the residual disease distal to the graft anastomosis. We were able to do this by carefully examining the actual surgical report and the close collaboration between the interventional cardiologists who measured the rSS and the surgeons who performed the operation. Patients were then stratified into rSS groups according to tercile cut points showing that those belonging to the first group of low rSS experienced lower rates of MACCE compared to those with higher rSSs. It was hard work, also because it was done retrospectively, but we strongly believe that this study may represent the first step towards a systematic calculation of the rSS right at the end of each CABG operation in a formal investigation with the setup of a multicentre registry in Italy and Europe, possibly with longer follow-up. We think that this method, which appears easy to apply, may represent a ‘real’ measure of the residual coronary disease after surgery. It is time then to start using the concept of residual disease instead of complete versus IR, suggesting that the term ‘adequate’ revascularization of suitably graftable vessels would be more appropriate. In fact, the SS has changed the old perspective of revascularization completeness by adding an objective measure of coronary disease. In addition, the correlation seen between the bSS and the rSS can be crucial to the choice of treatment strategy for each patient, including any further preoperative assessment (i.e. search for myocardial viability) and/or postoperative care with different drug therapies. The rSS performed better than the bSS in predicting MACCE at 1 year in this cohort, with a trend towards statistical significance. Based on the present results, it is then reasonable to expect patients with increasing rSS to be at higher risk of cardiovascular events at least 1 year after CABG surgery. The Cox regression model showed that the rSS did correlate with the 1-year MACCE together with common clinical risk factors such as age, creatinine level, ejection fraction, male gender and diabetes. Our findings are similar to those from three recent large studies on the impact of complete/IR [1–2, 8], as compared with older studies showing that better clinical outcomes were obtained when each and every angiographically visible stenosis was bypassed by a distal anastomosis [3–6, 16–23]. It is important to emphasize that complete revascularization, whether interventional or surgical, does not necessarily equate with better long-term clinical outcomes. As shown here and in other investigations, other factors play an important role in prognosis such as age, left ventricular function, renal failure and other significant comorbidities such as pulmonary dysfunction, CVA and other vascular diseases. These factors should be taken into consideration when advocating for and striving to achieve CR, which could result in longer bypass and ischaemic times, requiring significant inotropic support to the detriment of the patient. A high rSS, consisting of anatomical characteristics of the coronary vessel, may reflect the clinical risk profile of the patient and the likelihood of native coronary (and possibly non-coronary) atherosclerotic disease progression, which, importantly, may actually target the bypass grafts.

Potential applications

The SS is an important tool for long-term risk stratification of different groups of patients after surgery because it permits accurate measurement of the extent of CAD and its clinical application to quantify the residual atherosclerosis following CABG. This method is being used in post-PCI patients and therefore should be extended to surgical series to aid in the identification of those patients at high risk for future clinical events and who may benefit from modern interventional approaches such as hybrid coronary revascularization or extended use of more durable conduits to achieve the most adequate revascularization. Future investigations may focus on patients with high rSSs who may benefit from more aggressive control of risk factors using both established and emerging drugs to treat atherosclerosis, including its potential application in a research setting where CABG patients into PCI with DESs trials measuring the SS. Such a study could be compared with recent guidelines on myocardial revascularization [24–25], which have updated and expanded the risk score section. The SYNTAX and other scores are mentioned as reasonable tools to guide the decision making of the revascularization modality and may be an integral part of a heart team discussion incorporating all relevant cardiac specialities and heart care providers.

Study limitations

One limitation of the present investigation is that it does not take into account the progression of native CAD or the failure of one or more of the grafts. However, it has been shown that the clinical manifestations of the progression of native coronary artery and bypass disease, especially with SVGs, become more apparent over time and probably are not applicable to the present study. Another limitation is the retrospective, non-randomized design of the study in which selection bias and unmeasured confounders may have affected the results. In fact, it is important to point out that the objective calculation of complete versus IR is complex and includes several factors, some of which may be difficult to measure statistically. Therefore, results should be considered descriptive and not definitive. Larger studies are needed to confirm the observational results presented here.

Conflict of interest: none declared.

REFERENCES

APPENDIX. CONFERENCE DISCUSSION

Dr S. Ensminger (Bad Oeynhausen, Germany):

I think one question that is obvious is did you find actually a correlation between your preoperative angiographic findings of the SYNTAX score and then the postoperative calculated score? Did you find a correlation?

Dr Melina: No. We didn’t look at that.

Dr Ensminger: The other question is, if you say it is a nice predictive factor, my question to you would be what the clinical relevance is? I mean this is interesting for all of us. What do you do with patients who have a high residual SYNTAX score? Do you give them dual antiplatelet, anticoagulation? Do you do redo or whatever? In your opinion, what are the treatment strategies in Rome?

Dr Melina: I think this is the first study to look at the residual SYNTAX score in the coronary artery bypass graft population. In addition, I think that it could be useful to calculate the residual SYNTAX score soon after the operation to start and then, on the basis if this result, make a plan with one of these strategies. This is basically a retrospective analysis on prospectively collected data of patients who underwent coronary artery bypass graft. We believe that measuring the residual SYNTAX score at the end of the operation may allow for a more accurate analysis of the outcomes and undertake the most appropriate strategies to prevent MACCE. Looking at the single variables of major adverse cardiac events, there was a trend towards a higher repeat revascularization rate in the higher risk groups. In addition, there might be a bias due to progression of major coronary artery disease after the operation, which is not calculated here.

Dr Ensminger: In my opinion, it actually looks like a nice base for calculating and for developing an algorithm and say, okay, if the patient has a high residual SYNTAX score, then depending on the score, I would do this and this and this treatment strategy.

Dr Melina: I agree.

Dr D. Pagano (Birmingham, UK): I have a couple of comments and a couple of questions. Just a bit of warning. It does indicate potentially that incomplete revascularization may affect long-term outcome. From our own experience, when we look at the long-term outcome for coronary bypass grafting, there are other factors, other confounders, for which you didn't have the data that actually may influence the outcome, for instance, social deprivation, smoking status, aggressiveness of secondary prevention. I just wondered whether you had that data available and you put more confounders, potential confounders, and you still found the same results. I don't know the result of that because we haven't done that study, but it is just a word of caution.

Dr Melina: Yes. Thank you.

Dr Pagano: The question I have is a methodological question. Your operating room, how do you get the one-year survival and major adverse coronary events from these patients?

Dr Melina: There is an out-patient visit at 6 months and 1 year.

Dr Pagano: Just the routine follow-up?

Dr Melina: Yes, just routine follow-up.

Author notes