Interactive CardioVascular and Thoracic Surgery 19 (2014) 1008–1012 doi:10.1093/icvts/ivu297 Advance Access publication 12 September 2014
ORIGINAL ARTICLE – ADULT CARDIAC
Cardiac troponin release following hybrid coronary revascularization versus off-pump coronary artery bypass surgery Ralf E. Harskampa,b,*, Murad Abdelsalamc, Renato D. Lopesa, Gouthami Bogac, Sameer Hirjia, Mrinalini Krishnanc, Lukasz Kiljanekc, Mubashir Mumtazc, Jan G. Tijssenb, Christine McCartyc, Robbert J. de Winterb and William B. Bachinskyc a b c
Duke Clinical Research Institute, Durham, NC, USA Department of Cardiology, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands Pinnacle Health Cardiovascular Institute, Harrisburg, PA, USA
* Corresponding author. Duke Clinical Research Institute, 2400 Pratt Street, Office 7047B, Durham, NC, USA. Tel: +1-919-6688287; fax: +1-919-6687858; e-mail: [email protected] (R.E. Harskamp). Received 15 April 2014; received in revised form 28 July 2014; accepted 7 August 2014
Abstract OBJECTIVES: Cardiac ischaemic marker release is associated with adverse clinical outcomes after cardiac surgery. We sought to compare the release of cardiac troponin I (cTnI) after hybrid coronary revascularization (HCR) with off-pump coronary artery bypass surgery (OPCAB). METHODS: Using data from a prospective single-centre registry, we compared cTnI measured at postoperative day 1 following one-stage HCR and OPCAB among patients with normal baseline cTnI. Multivariable linear regression analysis was used to adjust for variables that may have influenced cardiac marker release other than the used revascularization strategy. RESULTS: Sixty-five consecutive patients underwent elective HCR (n = 33) or OPCAB (n = 32). Overall, no differences were seen in comorbidities, CABG risk scores and the lesion-specific SYNTAX score. Procedural complications were lower (15.2 vs 34.4%, P = 0.072), but 30-day and 1-year clinical outcomes (death, myocardial infarction, and repeat revascularization) were similar between the two groups (3.0 vs 3.1% and 9.1 vs 6.2%, respectively). Post-procedural cTnI release measured at 24 h after surgery was significantly lower following HCR compared with OPCAB [ratio of upper reference level URL: median: 3.5, interquartile range (IQR): 0.8–9.1 vs 12.8, IQR: 6.9–21.8, P = 0.001]. After adjusting for potential confounders, HCR was associated, on average, with cTnI less than half (46%) compared with CABG (P 50% stenosis, and were symptomatic or asymptomatic with an abnormal stress test. Candidates for HCR also met the following angiographic criteria: (i) left anterior descending artery (LAD)
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
R.E. Harskamp et al. / Interactive CardioVascular and Thoracic Surgery
Procedures Surgical and interventional techniques used for HCR and OPCAB were performed per protocol and have been previously published [9]. In short, HCR was performed on aspirin without a loading dose of clopidogrel. The femoral arterial sheath was placed prior to surgery; take-down of the left internal mammary artery (LIMA) was performed using a robotic system (Da Vinci Surgical System; Intuitive Surgical, Mountain View, CA, USA) and thoracic access was obtained by mini-thoracotomy. For patients undergoing OPCAB, the procedure was performed through conventional median sternotomy with the use of the LIMA and other arterial and vein graft conduits. Coronary arteriotomy and bypass graft anastomosis in both groups were performed without the use of cardiopulmonary bypass, and without the use of intracoronary shunts. The use of drug-eluting or bare-metal coronary stents as well as the duration of dual antiplatelet therapy were left at the discretion of the operators.
End-points The primary end-point was to compare peak serum release of cardiac troponin I (cTnI) assessed after 24 h as an indicator for myocardial injury. This time window was chosen since it takes 24 h for cTnI to reach its peak serum level [10]. Secondly, we assessed postoperative length of stay, procedural complications (renal failure, postoperative atrial fibrillation, prolonged ventilation, infection at the access site, and reoperation) and a composite outcome of death, MI and repeat revascularization at 30 days and 1 year. Death was defined as all reported/documented deaths of any cause; myocardial infarction (MI) included procedure-related and spontaneous MI. Procedure-related MI was defined as ≥5 times the upper limit of normal of cardiac markers and new Q-waves on electrocardiography. Repeat revascularization involved any re-PCI or redo-CABG during follow-up.
Statistical analysis Patients were compared based on the use of HCR or OPCAB. Comparisons between continuous variables were reported as median and interquartile range (IQR) and compared using the Wilcoxon rank-sum test. Discrete variables were reported as percent and compared using the Pearson χ2 or exact test. Multivariable linear regression analysis was used to adjust for differences
in baseline characteristics that may have influenced cardiac marker release. Adjustment variables included age, preoperative creatinine level, ejection fraction, total operating time, STS score and SYNTAX score. Linear regression assumptions, including normality and linearity, were assessed. Since postoperative cardiac marker release was not normally distributed, we applied a natural logarithm function transformation for the most appropriate fit. These natural logs of the cTnI values were used instead of the original raw values for our multivariable analysis. All reported P-values were two-sided, and values of 50%) LCx (>50%) RCA (>50%)
65 (58.5, 71.5) 84.8 (28) 30 (27, 33) 30.3 (10) 60.6 (20) 45.5 (15) 84.8 (28) 1.0 (0.9, 1.1) 75.7 (25) 9.1 (3) 15.2 (5) 15.2 (5) 27.3 (9) 24.2 (8) 55 (47.5, 62.5) 0 (0) 93.9 (31) 6.1 (2) 3.0 (1) 6.1 (2) 84.8 (28) 69.7 (23) 32 (25.5, 38.5) 0.4 (0.2, 0.6) 12.1 (4) 100 (33) 81.8 (27) 72.7 (24)
67 (60, 74) 56.3 (18) 29.9 (27.4, 32.4) 25.0 (8) 43.8 (14) 53.1 (17) 87.5 (28) 0.9 (0.7, 1.1) 96.9 (31) 12.5 (4) 12.5 (4) 12.5 (4) 37.5 (12) 21.9 (7) 55 (50, 60) 9.4 (3) 87.5 (28) 6.3 (2) 6.3 (2) 12.5 (4) 90.6 (29) 78.1 (25) 32.5 (27, 38) 0.7 (0.4, 1.0) 25.0 (8) 90.6 (29) 81.3 (26) 78.1 (25)
0.28 0.011 0.64 0.63 0.17 0.54 >0.99 0.35 0.027 0.71 >0.99 >0.99 0.38 0.82 0.89 0.11 0.43 >0.99 0.61 0.43 0.71 0.44 0.85 0.17 0.22 0.11 0.95 0.61
Continuous variables are presented as median and IQR between parentheses; categorical variables are presented as percentage (%) and actual number. BMI: body mass index; CAD: coronary artery disease; PVD: peripheral vascular disease; CVD: cerebrovascular disease; MI: myocardial infarction; PCI: percutaneous coronary intervention; LVEF: left ventricular ejection fraction; CHF: congestive heart failure; ACE: angiotensin converting enzyme; ADP: adenosine diphosphate; STS: society of thoracic surgeons; LM: left main; LAD: left anterior descending artery; LCx: left circumflex artery; RCA: right coronary artery.
ORIGINAL ARTICLE
and/or diagonal disease that was not amenable to percutaneous coronary intervention (PCI), but suitable for bypass surgery, and non-LAD disease that was amenable to PCI or bypass surgery; or (ii) left main disease with LAD and/or diagonal suitable for bypass surgery and left circumflex artery (LCx) and/or RCA disease amenable to PCI and bypass surgery. Patients with elevated baseline cardiac markers, haemodynamically unstable with a need for emergent CABG, or SYNTAX scores [7] above 60 were excluded. The same surgeons performed HCR and OPCAB procedures. Clinical characteristics, preoperative risk factors, medications, angiographic and procedural data, laboratory results and in-hospital events were monitored by research personnel and recorded in an electronic database. Baseline characteristics were defined according to the STS Adult Cardiac Database data specification criteria (http:// www.sts.org/national-database). Patient mortality risk scores were calculated using the EuroSCORE and predicted risk of mortality score [8].
1009
1010
R.E. Harskamp et al. / Interactive CardioVascular and Thoracic Surgery
HCR patients were less frequently female and had less hypertension, and no differences were seen in other comorbidities and medications. Moreover, the STS and the EuroSCORE risk scores as well as the lesion-specific SYNTAX score were all similar. Compared with OPCAB, total operating time (measured from time from incision to closure) was longer in the HCR group (365 vs 264 min, P
ORIGINAL ARTICLE – ADULT CARDIAC
Cardiac troponin release following hybrid coronary revascularization versus off-pump coronary artery bypass surgery Ralf E. Harskampa,b,*, Murad Abdelsalamc, Renato D. Lopesa, Gouthami Bogac, Sameer Hirjia, Mrinalini Krishnanc, Lukasz Kiljanekc, Mubashir Mumtazc, Jan G. Tijssenb, Christine McCartyc, Robbert J. de Winterb and William B. Bachinskyc a b c
Duke Clinical Research Institute, Durham, NC, USA Department of Cardiology, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands Pinnacle Health Cardiovascular Institute, Harrisburg, PA, USA
* Corresponding author. Duke Clinical Research Institute, 2400 Pratt Street, Office 7047B, Durham, NC, USA. Tel: +1-919-6688287; fax: +1-919-6687858; e-mail: [email protected] (R.E. Harskamp). Received 15 April 2014; received in revised form 28 July 2014; accepted 7 August 2014
Abstract OBJECTIVES: Cardiac ischaemic marker release is associated with adverse clinical outcomes after cardiac surgery. We sought to compare the release of cardiac troponin I (cTnI) after hybrid coronary revascularization (HCR) with off-pump coronary artery bypass surgery (OPCAB). METHODS: Using data from a prospective single-centre registry, we compared cTnI measured at postoperative day 1 following one-stage HCR and OPCAB among patients with normal baseline cTnI. Multivariable linear regression analysis was used to adjust for variables that may have influenced cardiac marker release other than the used revascularization strategy. RESULTS: Sixty-five consecutive patients underwent elective HCR (n = 33) or OPCAB (n = 32). Overall, no differences were seen in comorbidities, CABG risk scores and the lesion-specific SYNTAX score. Procedural complications were lower (15.2 vs 34.4%, P = 0.072), but 30-day and 1-year clinical outcomes (death, myocardial infarction, and repeat revascularization) were similar between the two groups (3.0 vs 3.1% and 9.1 vs 6.2%, respectively). Post-procedural cTnI release measured at 24 h after surgery was significantly lower following HCR compared with OPCAB [ratio of upper reference level URL: median: 3.5, interquartile range (IQR): 0.8–9.1 vs 12.8, IQR: 6.9–21.8, P = 0.001]. After adjusting for potential confounders, HCR was associated, on average, with cTnI less than half (46%) compared with CABG (P 50% stenosis, and were symptomatic or asymptomatic with an abnormal stress test. Candidates for HCR also met the following angiographic criteria: (i) left anterior descending artery (LAD)
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
R.E. Harskamp et al. / Interactive CardioVascular and Thoracic Surgery
Procedures Surgical and interventional techniques used for HCR and OPCAB were performed per protocol and have been previously published [9]. In short, HCR was performed on aspirin without a loading dose of clopidogrel. The femoral arterial sheath was placed prior to surgery; take-down of the left internal mammary artery (LIMA) was performed using a robotic system (Da Vinci Surgical System; Intuitive Surgical, Mountain View, CA, USA) and thoracic access was obtained by mini-thoracotomy. For patients undergoing OPCAB, the procedure was performed through conventional median sternotomy with the use of the LIMA and other arterial and vein graft conduits. Coronary arteriotomy and bypass graft anastomosis in both groups were performed without the use of cardiopulmonary bypass, and without the use of intracoronary shunts. The use of drug-eluting or bare-metal coronary stents as well as the duration of dual antiplatelet therapy were left at the discretion of the operators.
End-points The primary end-point was to compare peak serum release of cardiac troponin I (cTnI) assessed after 24 h as an indicator for myocardial injury. This time window was chosen since it takes 24 h for cTnI to reach its peak serum level [10]. Secondly, we assessed postoperative length of stay, procedural complications (renal failure, postoperative atrial fibrillation, prolonged ventilation, infection at the access site, and reoperation) and a composite outcome of death, MI and repeat revascularization at 30 days and 1 year. Death was defined as all reported/documented deaths of any cause; myocardial infarction (MI) included procedure-related and spontaneous MI. Procedure-related MI was defined as ≥5 times the upper limit of normal of cardiac markers and new Q-waves on electrocardiography. Repeat revascularization involved any re-PCI or redo-CABG during follow-up.
Statistical analysis Patients were compared based on the use of HCR or OPCAB. Comparisons between continuous variables were reported as median and interquartile range (IQR) and compared using the Wilcoxon rank-sum test. Discrete variables were reported as percent and compared using the Pearson χ2 or exact test. Multivariable linear regression analysis was used to adjust for differences
in baseline characteristics that may have influenced cardiac marker release. Adjustment variables included age, preoperative creatinine level, ejection fraction, total operating time, STS score and SYNTAX score. Linear regression assumptions, including normality and linearity, were assessed. Since postoperative cardiac marker release was not normally distributed, we applied a natural logarithm function transformation for the most appropriate fit. These natural logs of the cTnI values were used instead of the original raw values for our multivariable analysis. All reported P-values were two-sided, and values of 50%) LCx (>50%) RCA (>50%)
65 (58.5, 71.5) 84.8 (28) 30 (27, 33) 30.3 (10) 60.6 (20) 45.5 (15) 84.8 (28) 1.0 (0.9, 1.1) 75.7 (25) 9.1 (3) 15.2 (5) 15.2 (5) 27.3 (9) 24.2 (8) 55 (47.5, 62.5) 0 (0) 93.9 (31) 6.1 (2) 3.0 (1) 6.1 (2) 84.8 (28) 69.7 (23) 32 (25.5, 38.5) 0.4 (0.2, 0.6) 12.1 (4) 100 (33) 81.8 (27) 72.7 (24)
67 (60, 74) 56.3 (18) 29.9 (27.4, 32.4) 25.0 (8) 43.8 (14) 53.1 (17) 87.5 (28) 0.9 (0.7, 1.1) 96.9 (31) 12.5 (4) 12.5 (4) 12.5 (4) 37.5 (12) 21.9 (7) 55 (50, 60) 9.4 (3) 87.5 (28) 6.3 (2) 6.3 (2) 12.5 (4) 90.6 (29) 78.1 (25) 32.5 (27, 38) 0.7 (0.4, 1.0) 25.0 (8) 90.6 (29) 81.3 (26) 78.1 (25)
0.28 0.011 0.64 0.63 0.17 0.54 >0.99 0.35 0.027 0.71 >0.99 >0.99 0.38 0.82 0.89 0.11 0.43 >0.99 0.61 0.43 0.71 0.44 0.85 0.17 0.22 0.11 0.95 0.61
Continuous variables are presented as median and IQR between parentheses; categorical variables are presented as percentage (%) and actual number. BMI: body mass index; CAD: coronary artery disease; PVD: peripheral vascular disease; CVD: cerebrovascular disease; MI: myocardial infarction; PCI: percutaneous coronary intervention; LVEF: left ventricular ejection fraction; CHF: congestive heart failure; ACE: angiotensin converting enzyme; ADP: adenosine diphosphate; STS: society of thoracic surgeons; LM: left main; LAD: left anterior descending artery; LCx: left circumflex artery; RCA: right coronary artery.
ORIGINAL ARTICLE
and/or diagonal disease that was not amenable to percutaneous coronary intervention (PCI), but suitable for bypass surgery, and non-LAD disease that was amenable to PCI or bypass surgery; or (ii) left main disease with LAD and/or diagonal suitable for bypass surgery and left circumflex artery (LCx) and/or RCA disease amenable to PCI and bypass surgery. Patients with elevated baseline cardiac markers, haemodynamically unstable with a need for emergent CABG, or SYNTAX scores [7] above 60 were excluded. The same surgeons performed HCR and OPCAB procedures. Clinical characteristics, preoperative risk factors, medications, angiographic and procedural data, laboratory results and in-hospital events were monitored by research personnel and recorded in an electronic database. Baseline characteristics were defined according to the STS Adult Cardiac Database data specification criteria (http:// www.sts.org/national-database). Patient mortality risk scores were calculated using the EuroSCORE and predicted risk of mortality score [8].
1009
1010
R.E. Harskamp et al. / Interactive CardioVascular and Thoracic Surgery
HCR patients were less frequently female and had less hypertension, and no differences were seen in other comorbidities and medications. Moreover, the STS and the EuroSCORE risk scores as well as the lesion-specific SYNTAX score were all similar. Compared with OPCAB, total operating time (measured from time from incision to closure) was longer in the HCR group (365 vs 264 min, P
