2018 Volume 82 Issue 1 Pages 123-130
Background: The present study aimed to clarify the current use and outcomes of coronary artery bypass grafting (CABG) for acute coronary syndrome (ACS) based on the Japan Adult Cardiovascular Surgery Database (JACVSD) in stratified risk categories, and also to provide guidance on selection of optimal surgical strategies for ACS.
Methods and Results: From January 2008 through December 2012, 7,867 isolated CABG procedures for ACS were identified from the JACVSD. Patients were stratified into 3 subgroups (<2%, 2–9.9%, ≥10%) according to preoperative risk estimations based on this database. Off- and on-pump CABG surgical outcomes were evaluated in each subgroup. Off-pump CABG (OPCAB) was the predominant surgical strategy in all subgroups. The proportion of on-pump beating CABG increased in the higher-risk groups. Although average observed mortality rates were compatible with preoperative estimated risk in all subgroups, those after OPCAB were significantly lower in the medium (2–9.9%) risk group with lower incidence of major complications. In the low (<2%) and high (≥10%) risk groups, observed mortality rates did not show statistically significant differences between off- and on-pump CABG.
Conclusions: In this study in Japan, OPCAB was mainly performed in patients with ACS, particularly those with estimated risk <10%, with lower mortality rates, whereas on-pump beating CABG was selected for higher-risk patients with ACS, with reasonable mortality rates.
Acute coronary syndrome (ACS) includes unstable angina pectoris (uAP) and acute myocardial infarction (AMI). Affected patients requiring surgical coronary revascularization present with various cardiac and hemodynamic conditions and comorbidities. Although the surgical strategy should be determined on the basis of the patient’s condition, the selection of the optimal coronary artery bypass grafting (CABG) strategy for individual patients with ACS remains controversial.1,2
Fattouch et al conducted a randomized trial in a single institute, and reported better early outcomes of off-pump CABG (OPCAB) within 48 h of the onset of ST-segment elevation MI.3 The annual report of The Japanese Association for Thoracic Surgery indicates that 63.9% of isolated CABG surgeries including ACS cases were performed as OPCAB in Japan in 2013.4 However, OPCAB cannot always be used in patients with severely impaired hemodynamics, such as severe ACS, and the conventional on-pump CABG technique with or without cardiac arrest may be performed in patients with unstable hemodynamics.5–7 Comparisons between the various surgical strategies for ACS are often difficult to interpret, given existing differences and range in patients’ ACS status.
To collect a large number of ACS cases and cover the wide range of ACS patients’ conditions, we utilized a nationwide surgical registry data and stratified the patients’ categories based on preoperative risk estimation. The present study aimed to clarify the current use and outcomes of OPCAB and on-pump beating CABG (obCAB) techniques in risk-stratified patients with ACS, and may also provide surgeons with helpful information on the selection of CABG strategy.
Initiated in 2000, the Japan Adult Cardiovascular Surgery Database (JACVSD) is a nationwide registry of cardiovascular surgery.8 The database includes 255 variables that are nearly identical to those in the database of The Society of Thoracic Surgeons. Informed consent to register clinical data in the JACVSD was obtained from each patient. The JACVSD Organization Review Board approved the present study.
From the JACVSD, we identified urgent or emergency isolated CABG cases for uAP or AMI within 2 weeks of onset (Figure S1), from January 2008 through December 2012. We excluded patients with AMI-related mechanical complications, such as left ventricular rupture, ventricular septum perforation, or acute mitral regurgitation (MR) caused by papillary muscle dysfunction. As a first step, we analyzed the overall ACS cohort, and then stratified the patients into 3 risk groups (<2%, 2–9.9%, ≥10%) according to their preoperative risk estimation (JapanSCORE) based on the JACVSD9–11 followed by analysis of the stratified cohorts.
EndpointsThe primary endpoint of this study was 30-day operative death, which was defined as death at 30 days or before hospital discharge. We also evaluated the incidence of major operative complications that occurred in hospital or within 30 days after surgery.
Statistical AnalysisCategorical variables were compared using Pearson’s chi-square tests or Fisher’s exact tests. Logistic regression analysis was used to identify independent predictors for surgical outcomes. Predictors (listed below) associated with a P-value <0.2 in the univariate analysis were entered into the multivariate analysis using a stepwise selection method. Results are expressed using odds ratios (OR) and 95% confidence intervals (CI). Continuous variables are expressed as mean±standard deviation, and were compared by analysis of variance. All P-values were two-sided, and P<0.05 was considered statistically significant. The variables included in the logistic regression model for 30-day operative death were as follows: OPCAB, obCAB, age category, sex, smoking, diabetes, preoperative renal function, preoperative dialysis, hypertension, preoperative cerebrovascular event, carotid lesion, active infective endocarditis, respiratory function, aortic lesion, peripheral vascular lesion, neurological disorder, previous percutaneous coronary intervention, MI, congestive heart failure (CHF), shock, arrhythmia, aortic stenosis, mitral stenosis, aortic regurgitation, MR, tricuspid regurgitation, body mass index ≥30, New York Heart Association (NYHA) status, Canadian Cardiovascular Society (CCS) status, 3-vessel disease, left main trunk lesion, and left ventricular ejection fraction (LVEF).
From January 2008 through December 2012, 69,481 isolated CABG surgeries were registered in the JACVSD. Of these, 8,833 emergency or urgent surgeries, indicated by uAP or AMI, were identified as CABG for ACS. After excluding 966 cases with incomplete data, we analyzed 7,867 isolated CABG cases for ACS: 4,751 (60.4%) urgent and 3,116 (39.6%) emergency CABG cases. Of these 4,464 (56.7%) were OPCAB cases, 1,647 (20.9%) were obCAB cases, and 1,756 (22.3%) were on-pump arrested CABG (caCAB) cases. There were 4,673 cases of uAP and 3,194 of AMI. The patients’ baseline characteristics are shown in Table 1, with a comparison of CABG techniques. In this ACS cohort, preoperative renal dysfunction was remarkable, with a mean serum creatinine level of 1.6 mg/dL; 15.3% of patients showed high creatinine levels (≥2.0 mg/dL), and 8.9% underwent hemodialysis preoperatively. The obCAB group had the worst preoperative cardiac factors and hemodynamic status and the severest comorbidities.
| Parameters | OPCAB (n=4,464) |
obCAB (n=1,647) |
caCAB (n=1,756) |
P value |
|---|---|---|---|---|
| Age (mean±SD) | 70.4±10.0 | 70.2±10.2 | 68.7±9.9 | <0.001 |
| Male (%) | 75.2 | 76.3 | 74.3 | 0.378 |
| Smoking (%) | 53.9 | 56.3 | 52.6 | 0.087 |
| COPD (≥moderate) (%) | 2.6 | 3.8 | 1.9 | 0.003 |
| Hypertension (%) | 74.5 | 74.3 | 73.2 | 0.581 |
| Diabetes mellitus (%) | 46.0 | 49.0 | 45.0 | 0.045 |
| Dyslipidemia (%) | 52.6 | 48.5 | 52.8 | 0.0095 |
| Renal dysfunction (Cre ≥2.0) (%) | 15.3 | 17.9 | 12.9 | <0.001 |
| Serum Cre (mean±SD) | 1.61±2.3 | 1.82±2.5 | 1.49±2.1 | <0.001 |
| Hemodialysis (%) | 9.2 | 10.4 | 6.9 | 0.001 |
| Carotid lesion (%) | 5.4 | 4.7 | 3 | <0.001 |
| Peripheral vascular disease (%) | 13.2 | 17.5 | 11.0 | <0.001 |
| Previous PCI (%) | 24.5 | 25.7 | 22.6 | 0.088 |
| 3-vessel disease (%) | 68.3 | 75.5 | 76.2 | <0.001 |
| Left main trunk lesion (%) | 56.9 | 55.5 | 56.9 | 0.594 |
| LVEF <30% (%) | 8.9 | 25.5 | 11.0 | <0.001 |
| AMI (%) | 36.5 | 55.9 | 37.4 | <0.001 |
| CHF (%) | 27.1 | 45.0 | 26.1 | <0.001 |
| Shock (%) | 16.3 | 31.0 | 19.9 | <0.001 |
| Arrhythmia (%) | 12.1 | 19.9 | 13.0 | <0.001 |
| NYHA ≥3 (%) | 57.7 | 70.0 | 59.6 | <0.001 |
| CCS4 (%) | 43 | 54.2 | 47.8 | <0.001 |
ACS, acute coronary syndrome; AMI, acute myocardial infarction; CABG, coronary artery bypass grafting; caCAB, on-pump arrested CABG; CCS, Canadian Cardiovascular Society; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; Cre, creatinine; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; obCAB, beating on-pump CABG; OPCAB, off-pump CABG; PCI, percutaneous coronary intervention.
The operative factors of CABG for ACS are summarized in Table 2A. In the obCAB group, the percentages of redo and emergency surgeries and the absence of internal thoracic artery use were significantly higher, and the operation time was significantly longer. The number of anastomoses during OPCAB was 2.9±1.3, 2.9±1.0 during obCAB, and 3.2±1.0 during caCAB, and was significantly higher in the caCAB group (P<0.001).
| OPCAB (n=4,464) |
obCAB (n=1,647) |
caCAB (n=1,756) |
P value | |
|---|---|---|---|---|
| (A) Parameters | ||||
| Redo (%) | 0.9 | 2.6 | 0.7 | <0.001 |
| Urgent (%) | 64.6 | 49.2 | 60.3 | <0.001 |
| Emergency (%) | 35.4 | 50.8 | 39.7 | <0.001 |
| Operation time (min) (mean±SD) | 287.4±89.5 | 353.9±114.8 | 336.5±95.9 | <0.001 |
| No. of anastomosis (mean±SD) | 2.9±1.3 | 2.9±1.0 | 3.2±1.0 | <0.001 |
| Bilateral ITAs (%) | 29.3 | 14.5 | 13.0 | <0.001 |
| No usage of ITA (%) | 7.8 | 16.2 | 10.9 | <0.001 |
| (B) Complications | ||||
| Reoperation for bleeding (%) | 1.5 | 3.9 | 2.3 | <0.001 |
| Reoperation for any reasons (%) | 2.8 | 6.6 | 3.7 | <0.001 |
| Stroke (%) | 2.1 | 4.1 | 2.8 | <0.001 |
| Renal failure* (%) | 6.9 | 13.4 | 8.5 | <0.001 |
| New dialysis required (%) | 3.4 | 9.2 | 5.8 | <0.001 |
| Deep sternum infection (%) | 1.8 | 2.7 | 2.0 | 0.069 |
| Prolonged ventilation ≥24 h (%) | 14.7 | 28.1 | 18.6 | <0.001 |
| ICU stay ≥8 days (%) | 15.4 | 31.9 | 18.5 | <0.001 |
*Renal failure defined as acute or worsening renal failure resulting in ≥1 of the following: (1) increase in serum creatinine to ≥2.0 mg/dL and twice the most recent preoperative creatinine level; (2) new requirement for dialysis postoperatively. ICU, intensive care unit; ITA, internal thoracic artery. Other abbreviations as in Table 1.
The overall average observed 30-day operative mortality rate (including in-hospital deaths) for patients with ACS was 7.1%. The 30-day operative mortality rate after OPCAB was 5.3%, 12.8% after obCAB, and 6.4% after caCAB (P<0.001). The incidence of major complications in each CABG strategy is shown in Table 2B. All major complications, with the exception of deep sternal infection, occurred more frequently in the obCAB group, resulting in that group having the longest intensive care unit (ICU) stay after CABG.
Risk Factors for Death After CABG for ACSLogistic regression analysis (Table 3) showed that the independent risk factors for 30-day operative death were older age, preoperative renal dysfunction, neurological disorder, thoracic aortic lesion, MI, CHF, shock, arrhythmia, redo surgery, MR ≥3, and lower LVEF. OPCAB was an independent favorable factor for operative death after CABG (OR 0.738, 95% CI 0.605–0.899, P=0.003).
| Risk factors | OR | 95% CI | P value |
|---|---|---|---|
| Age (5-year intervals) | 1.34 | 1.259–1.427 | <0.0001 |
| Renal dysfunction (Cre ≥2.0) | 1.54 | 1.132–2.091 | 0.006 |
| Preoperative dialysis | 2.46 | 1.700–3.561 | <0.0001 |
| Neurological disorder | 2.33 | 1.716–3.153 | <0.0001 |
| Thoracic aortic lesion | 2.24 | 1.458–3.443 | <0.001 |
| Myocardial infarction | 1.36 | 1.067–1.724 | 0.013 |
| CHF | 1.55 | 1.248–1.932 | <0.0001 |
| Shock | 3.10 | 2.483–3.870 | <0.0001 |
| Arrhythmia | 1.50 | 1.199–1.884 | <0.001 |
| Medium LVEF (30–60%) | 1.73 | 1.250–2.380 | <0.0001 |
| Poor LVEF (<30%) | 5.13 | 3.608–7.304 | <0.0001 |
| Mitral regurgitation ≥3 | 1.87 | 1.275–2.734 | 0.001 |
| Redo | 2.24 | 1.225–4.102 | 0.009 |
| OPCAB | 0.74 | 0.605–0.899 | 0.003 |
CI, confidence interval; OR, odds ratio. Other abbreviations as in Table 1.
After stratifying ACS patients based on preoperative JapanSCORE, we found that patients with uAP were more dominant in groups with a JapanSCORE <10%, whereas patients with AMI were dominant in the high-risk group (JapanSCORE ≥10%) (Figure A). In terms of the distribution of ACS cases, the <2% risk group comprised 34.7% of all ACS patients, 46.0% of patients had a preoperative risk between 2.0% and 9.9%, and 19.2% were higher-risk patients (≥10%) (Figure B). OPCAB was primarily performed in the low- to medium-risk groups (JapanSCORE <10%). In the highest risk group (JapanSCORE ≥10%), the ratio of OPCAB decreased to 46% and caCAB decreased to 16.5%, whereas the ratio of obCAB increased to 37.5% (Figure B).
Proportions of diseases, surgical strategies and deaths in each stratified group. (A) Percentage of cases. After stratifying patients based on their JapanSCORE, patients with unstable angina (uAP) were more dominant in groups with a JapanSCORE <10%, whereas patients with acute myocardial infarction (AMI) were dominant in the high-risk group (JapanSCORE ≥10%). (B) Numbers of the 3 CABG techniques. OPCAB was most commonly performed in all groups, whereas the proportion of obCAB increased in higher-risk groups. (C) Observed operative mortality rate in each stratified group. Statistical analyses used Pearson’s chi-square tests. caCAB, on-pump arrested coronary artery bypass grafting; obCAB, on-pump beating coronary artery bypass grafting; OPCAB, off-pump coronary artery bypass grafting.
The average observed 30-day operative mortality rates for each CABG strategy in the risk-stratified subgroups are shown in Figure C. In the medium-risk (2–9.9%) group, the observed mortality rate after OPCAB was 3.2%, 6.0% after obCAB, and 6.5% after caCAB, presenting a statistically significant difference (P<0.001). However, there were no statistical differences in mortality rates between the CABG surgical strategies in the low- (<2%) and high-risk (≥10%) groups.
In the medium-risk (2–9.9%) group (n=3,622), patients’ baseline characteristics in the OPCAB group showed significantly higher age, more severe renal dysfunction, and more carotid artery lesions. Preoperative cardiac status such as coronary artery lesions, LVEF, CHF, shock, NYHA status, and CCS status were more severe in the obCAB group (Table 4). The operative factors in this group were similar to those in the overall ACS cohort (Table 5A). OPCAB showed the lowest observed mortality rate and the lowest incidence of major complications (e.g., reoperations for bleeding, renal failure, newly required dialysis, prolonged ventilation, and shortest duration of ICU stay) (Table 5B). The logistic regression analysis for the medium-risk group (2–9.9%) (Table 6) showed that age, renal dysfunction, thoracic aortic lesion, CHF, shock, medium-low LVEF (<60%), and MR ≥2 were independent risk factors, whereas OPCAB was an independent favorable factor (OR 0.51, 95% CI 0.368–0.710, P<0.0001).
| Parameters | OPCAB (n=2,117) |
obCAB (n=702) |
caCAB (n=803) |
P value |
|---|---|---|---|---|
| Age (mean±SD) | 73.3±9.2 | 71.1±9.9 | 71.7±8.9 | <0.001 |
| Male (%) | 73.0 | 74.2 | 70.6 | 0.264 |
| Smoking (%) | 50.8 | 53.7 | 48.9 | 0.179 |
| COPD (≥ moderate) (%) | 2.3 | 3.4 | 2.0 | 0.164 |
| Hypertension (%) | 77.3 | 74.4 | 75.5 | 0.235 |
| Diabetes mellitus (%) | 45.7 | 47.6 | 43.7 | 0.321 |
| Dyslipidemia (%) | 50.1 | 47.3 | 49.8 | 0.423 |
| Renal dysfunction (Cre ≥2.0) (%) | 17.0 | 13.2 | 14.9 | 0.045 |
| Serum Cre (mean±SD) | 1.70±2.4 | 1.67±2.4 | 1.65±2.3 | 0.853 |
| Hemodialysis (%) | 9.5 | 8.3 | 8.0 | 0.322 |
| Carotid lesion (%) | 6.3 | 3.1 | 3.6 | <0.001 |
| Peripheral vascular disease (%) | 12.1 | 11.7 | 10.7 | 0.583 |
| Previous PCI (%) | 25.7 | 25.9 | 21.5 | 0.051 |
| 3-vessel disease (%) | 69.2 | 77.8 | 79.1 | <0.001 |
| Left main trunk lesion (%) | 58.3 | 57.8 | 58.7 | 0.949 |
| LVEF <30% (%) | 5.7 | 13.5 | 9.2 | <0.001 |
| AMI (%) | 38.1 | 53.6 | 40.0 | <0.001 |
| CHF (%) | 29.3 | 39.9 | 29.8 | <0.001 |
| Shock (%) | 15.8 | 19.2 | 20.0 | 0.01 |
| Arrhythmia (%) | 12.1 | 14.5 | 14.7 | 0.085 |
| NYHA ≥3 (%) | 60.9 | 66.1 | 65.5 | 0.012 |
| CCS4 (%) | 46.0 | 52.0 | 50.8 | 0.006 |
Abbreviations as in Table 1.
| OPCAB (n=2,117) |
obCAB (n=702) |
caCAB (n=803) |
P value | |
|---|---|---|---|---|
| (A) Parameters | ||||
| Redo (%) | 0.8 | 2.3 | 1.4 | 0.005 |
| Urgent (%) | 58.1 | 50.3 | 51.6 | <0.001 |
| Emergency (%) | 41.9 | 49.7 | 48.4 | <0.001 |
| Operation time (min) (mean±SD) | 282.7±87.2 | 354.5±115.1 | 333.0±100.3 | <0.001 |
| No. of anastomosis (mean±SD) | 2.9±1.4 | 3.0±1.0 | 3.2±1.0 | <0.001 |
| Bilateral ITAs (%) | 26.2 | 14.5 | 8.1 | <0.001 |
| No usage of ITA (%) | 7.5 | 12.8 | 11.7 | <0.001 |
| (B) Complications | ||||
| Reoperation for bleeding (%) | 1.3 | 2.7 | 2.4 | 0.024 |
| Reoperation for any reasons (%) | 2.8 | 4.1 | 3.7 | 0.17 |
| Stroke (%) | 2.2 | 3.8 | 3.1 | 0.052 |
| Renal failure* (%) | 6.7 | 9 | 9.8 | 0.0078 |
| Newly dialysis required (%) | 3.1 | 6.1 | 6.6 | <0.001 |
| Deep sternum infection (%) | 1.7 | 2.1 | 2.2 | 0.55 |
| Prolonged ventilation ≥ 24 h (%) | 14.2 | 24.5 | 18.6 | <0.001 |
| ICU stay ≥8 days (%) | 15.7 | 28.5 | 19.9 | <0.001 |
*Renal failure defined as acute or worsening renal failure resulting in ≥1 of the following: (1) increase in serum creatinine to ≥2.0 mg/dL and twice the most recent preoperative creatinine level; (2) new requirement for dialysis postoperatively. Abbreviations as in Tables 1,2.
| Risk factors | OR | 95% CI | P value |
|---|---|---|---|
| Age (5-year intervals) | 1.18 | 1.056–1.323 | 0.004 |
| Renal dysfunction (Cre ≥2.0) | 3.69 | 2.440–5.566 | <0.001 |
| Thoracic aortic lesion | 3.93 | 1.977–7.808 | <0.001 |
| Myocardial infarction | 1.38 | 0.961–1.988 | 0.081 |
| CHF | 1.55 | 1.095–2.193 | 0.013 |
| Shock | 2.09 | 1.398–3.118 | <0.0001 |
| Medium LVEF (30–60%) | 1.85 | 1.168–2.921 | 0.009 |
| Poor LVEF (<30%) | 3.68 | 1.961–6.923 | <0.0001 |
| Mitral regurgitation ≥2 | 1.51 | 1.036–2.198 | 0.032 |
| OPCAB | 0.51 | 0.368–0.710 | <0.0001 |
Abbreviations as in Tables 1,3.
A logistic regression analysis for the high-risk (≥10%) group (n=1,506) showed that neither OPCAB nor obCAB was an independent risk factor, but age, preoperative dialysis, neurological disorder, shock, redo surgery, low LVEF (<30%), and MR ≥3 were independent risk factors (Table 7).
| Risk factors | OR | 95% CI | P value |
|---|---|---|---|
| Age (5-year intervals) | 1.21 | 1.103–1.317 | <0.0001 |
| Male sex | 0.77 | 0.58–1.022 | 0.07 |
| Preoperative dialysis | 1.95 | 1.401–2.714 | <0.0001 |
| Neurological disorder | 2.26 | 1.622–3.145 | <0.001 |
| Myocardial infarction | 1.37 | 0.985–1.905 | 0.061 |
| Shock | 2.76 | 2.028–3.763 | <0.0001 |
| Arrhythmia | 1.28 | 0.983–1.657 | 0.067 |
| Redo | 2.37 | 1.251–4.483 | 0.008 |
| LVEF <30% | 2.51 | 1.911–3.291 | <0.001 |
| Mitral regurgitation ≥3 | 1.63 | 1.08–2.462 | 0.02 |
| OPCAB | 1.14 | 0.782–1.657 | 0.499 |
| On-pump beating CABG | 1.16 | 0.896–1.496 | 0.263 |
Abbreviations as in Tables 1,3.
A single surgical strategy cannot be used for all patients with ACS because of the variation in patients’ conditions. Although surgeons should use the right CABG technique for the right patient with ACS, even major randomized control trials12–15 comparing OPCAB and obCAB in particular settings could not provide useful information for surgical technique selection in ACS cases.16 The present study was a retrospective analysis of registry data from a nationwide cardiovascular database in Japan (JACVSD).8 According to the annual report of The Japanese Association for Thoracic Surgery,17 63,800 cardiovascular operations were performed at 583 institutions during 2012. In the same period, 53,612 (84% of total) surgical cases from 336 institutes were registered in the JACVSD. Therefore, this report could both reflect the real-world practice of CABG for ACS in Japan, with the largest number of patients to date, and clarify the current use and outcomes of CABG techniques (OPCAB, obCAB, and caCAB) for ACS. In addition to the overall cohort analysis, we stratified this ACS cohort according to the preoperative risk estimation (JapanSCORE), and analyzed the data in each risk-stratified subgroup to specify the optimal range for each surgical strategy.
Although the overall 30-day operative mortality rate after CABG was relatively high (7.1%), it was comparable with mortality outcomes of 4.1–10.2% in CABG for ACS reported in previous studies.6,7,18 Some speculation about mortality data from the Japanese ACS registry should be discussed. These were nationwide registry data, and included patients with the severest ACS with shock hemodynamics. In fact, 1,588 (20.2%) cases of cardiogenic shock were registered in this ACS cohort. Furthermore, preoperative renal dysfunction, which is a well-known risk factor for CABG, was severe in this ACS cohort. The more severe preoperative characteristics in this ACS cohort affected CABG outcomes in the emergency and urgent settings. The mean incidence of complications such as reoperations was 3.8%, new dialysis was 5.1%, and stroke was 2.7%. However, these morbidity rates may reflect the severity of the condition of patients with ACS.
In the present study, we stratified patients’ risk using the JapanSCORE, a preoperative risk prediction system based on the JACVSD.8–11 The observed mortality rate in each stratified group was comparable with those predicted using the JapanSCORE, which suggests that the JapanSCORE preoperative risk estimate model is reliable, even in patients with ACS requiring isolated CABG. This is also reasonable, as the JapanSCORE model comprises various preoperative factors10 that were mostly comparable with the independent risk factors (shown in Table 3) for CABG in ACS cases in this study, although the ORs were not perfectly identical.
As for the remarkable findings after risk stratification, in the medium-risk group (2–9.9%), OPCAB showed the lowest mortality rate of the 3 surgical strategies, with the lowest incidences of reoperation for bleeding, renal failure, postoperative dialysis, and prolonged ventilation, and shorter ICU stay after CABG, even though the extra-cardiac comorbidities (e.g., renal dysfunction and vascular diseases) in the OPCAB group were more severe preoperatively. Consequently, OPCAB was identified as an independent favorable factor for operative death after CABG for ACS in this risk range. Puskas et al also reported that OPCAB was beneficial for patients with a higher Society of Thoracic Surgeons-predicted risk of death >3%),19 which is consistent with the present study results in terms of the preoperative estimated risk range in which OPCAB showed a survival benefit.
We understand that OPCAB-dominant surgery is a peculiarity of Japanese CABG practice, and accept that there are some concerns, especially about long-term outcomes after OPCAB. Although 2 meta-analyses have reported worse long-term outcomes after OPCAB for standard-risk coronary revascularization,20,21 Puskas et al reported that OPCAB and obCAB were associated with similar early and late graft patency rates, incidence of recurrent or residual myocardial ischemia, need for reintervention, and long-term survival.22 Furthermore, a recent review article by Parissis et al noted the potential benefit of OPCAB in reducing stroke by avoiding the ascending aortic manipulation.23 Those studies suggest that conclusions may vary depending on patient selection, technique performed, and study endpoints.
For patients with ACS, the primary purpose/aim of CABG should be “survival of ACS” without severe complications. Therefore, we should also acknowledge that OPCAB provided a short-term survival benefit, at least for medium-risk (2.0–9.9%) patients with ACS as far as surgeons judge that OPCAB is applicable. Similarly, Harling et al performed a meta-analysis involving 3,001 patients with ACS to determine the beneficial effect of OPCAB on 30-day death in hemodynamically stable patients undergoing emergency CABG.2
In contrast, some studies have reported the safety of obCAB for emergency myocardial revascularization in critically ill patients.5–7,24 The present study showed that the most severely ill patients with ACS in Japan were treated with obCAB. Although the overall operative mortality rate of obCAB for ACS was high (12.8%), the strategy itself was not identified as an independent risk factor in either ACS patients overall or in the highest risk group (Tables 3,7), and the higher mortality rate was probably caused by patient condition bias.
Study LimitationsFirst, as this was a retrospective study using registry surgical data and did not include long-term outcomes, we could only set operative death as an endpoint. Second, factors used to distinguish ACS status (e.g., non-ST-elevated MI and ST-elevated MI, and preoperative and postoperative creatine phosphokinase MB isozyme levels) are not included in this database. Consequently, our study did not provide sufficient findings to discuss the advantages and disadvantages of conducting cardioplegic arrest during CABG in patients with AMI, especially in those with STEMI. Further analysis based on the next version of the database, in which we can identify STEMI patients by the parameter of “MI type (STEMI, NSTEMI, or Unknown)”, would be required. Third, the present study lacked data for completeness of revascularization, which might have affected both short- and long-term outcomes after CABG for ACS. Although the numbers of anastomoses during OPCAB were significantly fewer in both the overall and medium-risk cohorts, the ratio of 3-vessel disease in the OPCAB group was also significantly lower in both cohorts. Therefore, fewer anastomoses may not always result in less complete revascularization.
In the present study, the current surgical outcomes of isolated CABG for patients with ACS in Japan were clarified using a nationwide cardiovascular surgery database. OPCAB was the most common technique performed in all patient groups, particularly in those with estimated risk <10%, and showed the lowest incidence of major complications.
The authors thank the data managers in each of the cardiovascular institutes participating in the JACVSD for their great efforts in registering clinical data, and Ms. Eriko Fukuchi in the Division of Healthcare Quality Assessment, University of Tokyo for her great support in data analysis. The authors have no conflicts of interest regarding this study.
Supplementary File 1
Figure S1. Patient selection and definition flow chart of acute coronary syndrome (ACS) based on pre operative cardiac status (Part F) in JACVSD ver.4.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-17-0561
