Off-pump coronary artery bypass grafting (OPCABG) has been increasingly performed in the last few years, thanks to the development of cardiac stabilization devices and increased surgeons' experience [1-3]. It has gained remarkable clinical acceptance in an effort to eliminate the morbidity associated with cardiopulmonary bypass (CPB). Recent studies [4,5] and meta-analyses of randomized controlled studies [1-3] showed no difference in major outcomes between off-pump and the standard CPB technique.
Adverse haemodynamic changes during OPCABG are frequent and, occasionally, the patient must be placed on CPB in order to safely complete the operation. The reasons for switch from off-pump to on-pump procedures include failure to adequately expose the target vessels, deep intramyocardial course of the target vessels or circulatory collapse due to severe ischaemia, arrhythmias or heart failure. The incidence of conversion from off-pump to CPB might be underestimated in randomized controlled studies  and is almost never reported in case series. Patients who require conversion experience higher morbidity and mortality rates. Although some previous studies [7-14] have focused on patients requiring conversion, the debate on off-pump vs. on-pump procedures is ongoing. We therefore evaluated the outcome of patients who were converted owing to haemodynamic instability, with the aim of identifying predictors for the need to convert and of evaluating whether temporal trends and learning curve exist.
The study was approved by the institutional human investigation committee and written informed consent was obtained from all subjects for participation in the prospective data collection that identified patients who required conversion from off-pump to standard CPB technique on an emergency basis in the period June 2001-July 2003. The indications for conversion were: haemodynamic instability unresponsive to Trendelenburg position, atrial pacing, intravenous fluid infusion, or inotropes (dopamine 5-8 μg kg−1 min−1 or norepinephrine 0.025-0.05 μg kg−1 min−1), arrhythmias, severe myocardial ischaemia and cardiac arrest. To allow a rapid conversion to standard CPB, our management did not include intra-aortic balloon pumping.
In all, 988 patients underwent CABG in the study period. Sixty patients were excluded (54 patients owing to emergency operations and six owing to associated carotid surgery). Thus, we studied 928 consecutive patients. Four hundred and thirteen patients (44.5%) completed the operation off-pump and 478 patients (51.5%) were operated on using the CPB as planned. Thirty-seven patients required conversion on an emergency basis from an off-pump to a standard CPB technique.
The decision to perform an off-pump or a standard CPB CABG was based on the surgeon's preference on the basis of the site and diameter of the target vessels as well as on patients' co-morbidities. All five surgeons had performed off-pump CABG for at least 2yrs before the study period (total number of off-pump procedures = 263).
All patients received a standard premedication (morphine 0.1 mg kg−1 and scopolamine 0.25 mg intramuscularly; diazepam per os pro re nata) 1 h before surgery. Standard monitoring included intra-arterial and central venous pressures, but not routinely transoesophageal echocardiography, pulmonary artery catheter or neurologic monitoring. Induction of anaesthesia was performed with fentanyl-propofol. Orotracheal intubation was facilitated by pancuronium. Anaesthesia was maintained with propofol (2-4 mg kg−1 h−1), isoflurane (end-tidal concentration <1 MAC) and fentanyl up to a total of 20 μg kg−1 (off-pump) or 25 μg kg−1 (CPB). The initial dosage of heparin was 300 IU kg−1 for the CPB and 150 IU kg−1 for the off-pump patients. The activated clotting time was maintained at ≥480 s for CPB and at ≥300 s for off-pump procedures. Heparin was reversed with protamine sulphate at the end of the procedure.
CPB was established at mild hypothermia (32°C) with the use of a heparin-coated membrane oxygenator and a roller pump. Arterial flow was adjusted to 2.2-2.4 L min−1 m−2, and blood pressure was maintained between 50 and 70 mmHg. Haematocrit level was kept at >22%. Myocardial preservation was achieved by intermittent (every 20 min) antegrade and retrograde cold (4°C) blood cardioplegia (4 : 1 ratio). The proximal anastomoses to the ascending aorta were constructed during a single cross clamp period.
In the off-pump group the Octopus (Medtronic, Inc., Minneapolis, MN, USA) suction stabilizers were used. Posterolateral wall anastomoses were facilitated by a deep pericardial stitch or a Starfish (Medtronic, Inc., Minneapolis, MN, USA) stabilizer. The right pleura was routinely opened. Intracoronary shunts were used routinely. Distal anastomoses were usually constructed before proximal anastomoses. The left internal thoracic artery to left anterior descending coronary artery anastomosis was always performed first. The proximal anastomoses were constructed during tangential side clamping of the aorta.
Patients who converted had full heparinization and cannulation performed while supportive measures were taken for the few minutes required to shift to CPB (100% oxygen ventilation in all patients, internal cardiac massage and attempts to cardioversion if necessary).
Study outcomes included prolonged hospital stay (>7 days), in-hospital all-cause mortality, stroke (new onset of global or focal brain injury that persisted for more than 72 h), renal failure (postoperative serum creatinine ≥200% of preoperative value), need for continuous renal replacement therapy, reoperation for bleeding, use of blood products and myocardial infarction (specific myocardial band creatine phosphokinase >50 μg L−1 and significant electrocardiogram (ECG) abnormalities).
We tested our primary hypothesis that conversion would worsen patients' outcome in terms of prolonged length of postoperative stay as compared with routine off-pump or CPB procedures. Sample-size calculation was based on a two-sided alpha error of 0.05 and 80% power. On the basis of previous data investigating survival after CABG, we anticipated an incidence of prolonged hospital stay of 10% in the general population and assumed a 40% incidence of prolonged hospital stay in the patients who were converted. We calculated that we would need at least 30 patients in the conversion group.
Preoperative patient characteristics and perioperative data were compared by univariate analysis (χ2-test, Fisher's exact test or t-test for variance when appropriate). A multivariate analysis (stepwise multivariate logistic regression analysis), including all the preoperative values listed in Table 1, was used to identify the independent predictors of conversion and of hospital death (data are reported as odds ratios with 95% confidence intervals (CI)) using Epi Info 2002 software (CDC) and SAS (release 8.00 by SAS Institute, Cary, NC) statistical software. The univariate analysis was then repeated considering the patients who required conversion as belonging to the off-pump group (following an ‘intention to treat' analysis).
A Kolmogorov-Smirnov test and cluster analysis (Wald-Wolfowitz runs test) were performed to study temporal trends in the occurrence of conversion from off-pump to CPB. These ordinal tests assess the likelihood that the two groups come from a single ordering (so that differences observed are caused by chance fluctuations), or that the two groups come from two different orderings.
Patient characteristics and intraoperative data are shown in Table 1. Off-pump surgery was attempted in 450 (48.5%) of the 928 patients. However, 37 patients in the off-pump group (8.2%) had to be converted to CPB on an emergency basis. We could not identify any statistically significant preoperative or intraoperative risk factors for emergent conversion to CPB. However, there was a trend for greater incidence in chronic obstructive pulmonary disease (COPD) (18.9% vs. 9.9%, P = 0.08) patients and in females (24.3% vs. 14.0%, P = 0.1) in the converted patients, as compared with those who had successfully completed off-pump surgery. A multivariate analysis (entry values: P < 0.3) confirmed the trend towards a higher incidence of conversion in females (OR 2.2 95% CI 0.99-5.1, P = 0.051) and COPD patients (OR 2.2 95% CI 0.88-5.4, P = 0.08).
Postoperative data are given in Table 2. Overall, the duration of postoperative mechanical ventilation was 10 h (median, interquartile range=8-15), the duration of ICU stay was 1 day (median, interquartile range=1-2) and the duration of hospital stay was 5 days (median, interquartile range=4-6). The overall hospital mortality was 1.1% (10/928) with significant differences between the three groups: patients who were converted 2/37 (5.4%), off-pump patients 6/413 (1.5%) and CPB 2/478 (0.4%), P=0.02. A multivariate analysis (entry values P < 0.05) that included preoperative and intraoperative factors described in Table 1 and the event ‘conversion' showed that conversion (OR 7.2, 95% CI 1.4-36.5, P=0.02) and chronic renal failure (OR 6.3, 95% CI 1.7-23.1, P=0.006) were independently associated with hospital death. Accordingly, patients who converted from off-pump to CPB had a more complicated hospital course (Table 2). This led to significant differences between the three groups in the number of patients requiring prolonged hospitalization (>7 days): patients who converted 10/37 (27%), off-pump patients 51/413 (12.3%) and CPB 84/478 (17.6%), P=0.02.
The conversion rate was uniformly distributed in the study period and no clusters were noted. A Kolmogorov-Smirnov test was performed to study temporal trends in the occurrence of conversion from off-pump to CPB. It did not show any decrease in the occurrence of conversion during the study period (P=0.1). A Wald-Wolfowitz runs test analysis confirmed that there was no temporal clustering (P=0.3).
Notably, the two patients who died after conversion from off-pump to CPB surgery were operated on within the first 3 months of the data collection period. No converted patients died in the consecutive 22-month period of the study. The mortality rate in the converted patients showed a trend (P=0.1) towards a higher incidence in the first semester of the study unlike time on mechanical ventilation (P=0.9), length of ICU stay (P=0.3) and duration of hospital stay (P=0.7).
We therefore compared mortality of the two groups (off-pump and CPB) based on the ‘intention to treat analysis' (Table 3). The off-pump group had a statistically significant higher mortality than the CPB group (1.8 vs. 0.4%, P=0.04 with Fisher's exact test). However, this difference was not confirmed at a multivariate analysis, looking for risk factors of hospital death that included preoperative factors described in Table 1 (entry values P=0.05) and the surgical technique (intended off-pump vs. CPB surgery).
This study assesses the outcome of patients requiring emergency conversion to CPB during OPCABG. We found higher morbidity, mortality and length of postoperative hospital stay in these patients as compared with patients who complete their operation following the planned technique. Patients who require conversion to CPB on an emergency basis for haemodynamic instability are challenging: this scenario represents a necessity rather than a choice. Interestingly, we noted no temporal trend over a 2-yr period in a group of surgeons already experienced in off-pump surgery. This is in contrast with previous reports that noted a marked reduction of conversion over time [8,10]. The reported incidence of conversion from off-pump technique to CPB in the literature may be underestimated not only in case series but also in randomized controlled studies : sometimes, patients who required conversion from off-pump to CPB are only briefly mentioned and either excluded from the analysis [15-18] or included in the CPB group [19,20], not following the intention to treat analysis.
The incidence of conversions is variable, ranging from 1.6% to 13.3% in reports focusing on this aspect [7-14], with a high mortality rate (8.3-18%). We report a conversion rate in our OPCABG patients of 8.2% with a mortality rate of 5.4%.
Conversion to CPB during attempted OPCABG is a potential confounder in many of the previous retrospective comparisons between OPCABG and on-pump CABG, as patients who required conversion have been included in the CPB group [21,22]. In contrast, we included the converted patients in the off-pump group following an ‘intention to treat' analysis as suggested by Patel and colleagues . With this approach, we found a statistically significant different mortality between the two groups: 1.8% in the OPCABG vs. 0.4% in the CPB group (P=0.04). Even if this result was not confirmed at the multivariate analysis, we recommend that studies comparing OPCABG with conventional CABG should include converted patients in the OPCABG group.
More recently, in some prospective studies [23-25] comparing OPCABG with conventional CPB, the authors included these patients in the OPCABG group, but the outcomes and perioperative data of these patients were not mentioned. Indeed, the most important finding of our study is that converted patients have a poorer postoperative outcome, as shown by prolonged time on mechanical ventilation, prolonged hospital stay and increased hospital death.
The adverse outcomes in this subset of patients dramatically alter the results of the off-pump technique as a whole; any short-term benefits of avoiding CPB will only be relevant if patients requiring conversion are analysed as belonging to the OPCABG group. Currently, emergency conversion is an unpredictable event in our setting. Temporal trend analysis showed that conversion occurs at a stable rate in our hospital, failing to link this event to a learning curve of the operating team. Probably, there is a learning curve in the expertise to manage conversion, especially its timing. The worse outcome in our converted patients might be attributed to the short time of safety between the beginning of haemodynamic deterioration and collapse: an expert team recognizes patients who cannot tolerate luxation or coronary snaring early and safely switches rapidly to CPB operation.
Limitations of the study
The most important limitation of the present study is the small number of patients requiring conversion, which prevented us from identifying independent risk factors that would predict which patient may require conversion. However, we identified some trends: emergency conversion was more frequent in females, in patients with COPD, in patients who had already had cardiac surgery and in patients requiring more grafts. In contrast to our findings, it should be noted that Norris and colleagues  observed a higher male-female ratio among conversion cases.
Another important limitation of the study is that we cannot answer the question of whether it is the act of conversion to CPB itself that leads to the worse outcome or whether conversion merely identifies a high-risk group of patients. Furthermore, we recognize that important data have not been prospectively collected in our database such as the New York Heart Association classification, preoperative medications, incidence of elective conversion from off-pump to CPB (e.g. calcified, deep vessels) and from CPB to off-pump (e.g. calcified aorta), time to perform emergent CPB, percentage of patient that required internal cardiac massage or cardioversion, acid-base status and ECG changes immediately preceding the decision to convert.
In conclusion, converted patients have a higher morbidity and mortality. We suggest that the outcome of patients who are converted from OPCABG to CPB should be separately reported in future trials or that converted patients should be analysed within the OPCABG group. As far as our experience (8.2%) and the review of the literature (from 1.6% to 13.3%) is concerned, conversion rate does not depend on surgical ability (there was no temporal trend over a 2-yr period in a group of surgeons already experienced in off-pump surgery).
We thank Giuseppe Giardina, RN, Mariano Lampugnani, RN, Giordano Gherbi, RN, Adams Castello, RN, Fabiola Sana, RN, Mariano Fichera, RN, from the Department of Cardiovascular Anesthesia, IRCCS San Raffaele Hospital, Milan, Italy for the support in data entry and for the care provided to these patients. The study was supported only by departmental funds.
1. Wijeysundera DN, Beattie WS, Djaiani G et al
. Off-pump coronary artery surgery for reducing mortality
and morbidity: meta-analysis of randomized and observational studies. J Am Coll Cardiol
2005; 46: 872-882.
2. Cheng DC, Bainbridge D, Martin JE, Novick RJ. Does off-pump coronary artery bypass reduce mortality
, morbidity and resource utilization when compared with conventional coronary artery bypass? A meta-analysis of randomised trials. Anesthesiology
2005; 102: 188-203.
3. Parolari A, Alamanni F, Cannata A et al
. Off-pump vs. on-pump coronary artery bypass: metaanalysis of currently available randomized trials. Ann Thorac Surg
2003; 76: 37-40.
4. Schwann NM, Horrow JC, Strong MD, Chamchad D, Guerraty A, Wechsler AS. Does off-pump coronary artery bypass reduce the incidence of clinically evident renal dysfunction after multivessel myocardial revascularization? Anesth Analg
2004; 99: 959-964.
5. Gamoso MG, Phillips-Bute B, Landolfo KP, Newman MF, Stafford-Smith M. Off-pump vs. on-pump coronary artery bypass surgery and postoperative renal dysfunction. Anesth Analg
2000; 91: 1080-1084.
6. Zangrillo A, Annalisa F, Crescenzi C et al
. Underreporting of conversion from off-pump coronary artery bypass surgery. Anesthesiology
2005; 103: 902.
7. Legare JF, Buth KJ, Hirsch GM. Conversion to on pump from OPCAB is associated with increased mortality
: results from a randomized controlled trial. Eur J Cardiothorac Surg
2005; 27: 296-301.
8. Patel NC, Patel NU, Loulmet DF, McCabe JC, Subramanian VA. Emergency conversion to cardiopulmonary bypass
during attempted off-pump revascularization results in increased morbidity and mortality
. J Thorac Cardiovasc Surg
2004; 128: 655-661.
9. Norris JQ, Howells S, Browne J et al
. Failed off-pump CABG predictors and outcomes. Can J Anesth
2003; 50: A74.
10. Edgerton JR, Dewey TM, Magee MJ et al
. Conversion in off-pump coronary artery bypass grafting: an analysis of predictors and outcomes. Ann Thorac Surg
2003; 76: 1138-1143.
11. Iaco AL, Contini M, Teodori G et al
. Off or on bypass: what is the safety threshold? Ann Thorac Surg
1999; 68: 1486-1489.
12. Soltoski P, Salerno T, Levinsky L et al
. Conversion to cardiopulmonary bypass
in off-pump coronary artery bypass grafting: its effect on outcome. J Card Surg
1998; 13: 328-334.
13. Mishra M, Shrivastava S, Dhar A et al
. A prospective evaluation of hemodynamic instability during off-pump coronary artery bypass surgery. J Cardiothorac Vasc Anesth
2003; 17: 452-458.
14. Jin R, Hiratzka LF, Grunkemeier GL, Krause A, Page US III. Aborted off-pump coronary artery bypass patients have much worse outcomes than on-pump or successful off-pump patients. Circulation
2005; 112 (Suppl 9): 1332-1337.
15. Wehlin L, Vedin J, Vaage J, Lundahl J. Activation of complement and leukocyte receptors during on and off pump coronary artery bypass surgery. Eur J Cardiothorac Surg
2004; 25: 35-42.
16. Dorman BH, Kratz JM, Multani MM et al
. A prospective, randomized study of endothelin and postoperative recovery in off-pump vs. conventional coronary artery bypass surgery. J Cardiothorac Vasc Anesth
2004; 18: 25-29.
17. Wandschneider W, Thalmann M, Trampitsch E, Ziervogel G, Kobinia G. Off-pump coronary bypass operations significantly reduce S100 release: an indicator for less cerebral damage? Ann Thorac Surg
2000; 70: 1577-1579.
18. Lönn U, Peterzen B, Carnstam B, Casimir-Ahn H. Beating heart coronary surgery supported by an axial blood flow pump. Ann Thorac Surg
1999; 67: 99-104.
19. Carrier M, Perrault LP, Jeanmart H, Martineau R, Cartier R, Page P. Randomized trial comparing off-pump to on-pump coronary artery bypass grafting in high-risk patients. Heart Surg Forum
2003; 6: E89-E92.
20. Czerny M, Baumer H, Kilo J et al
. Inflammatory response and myocardial injury following coronary artery bypass grafting with or without cardiopulmonary bypass
. Eur J Cardiothorac Surg
2000; 17: 737-742.
21. Al-Ruzzeh S, Ambler G, Asimakopoulos G et al
. Off-pump coronary artery bypass (OPCAB) surgery reduces risk stratified morbidity and mortality
: a United Kingdom multi-center comparative analysis of early clinical outcome. Circulation
2003; 108 (Suppl 1): II1-II8.
22. Meharwal Zs, Mishra YK, Kohli V, Bapna R, Singh S, Trehan N. Off-pump multivessel coronary artery surgery in high-risk patients. Ann Thorac Surg
2002; 74: 1353-1357.
23. Van Dijk D, Nierich AP, Jansen EW et al
. Early outcome after off-pump vs. on-pump coronary bypass surgery: results from a randomized study. Circulation
2001; 104: 1761-1766.
24. Khan NE, De Souza A, Mister R et al
. A randomized comparison of off-pump and on-pump multivessel coronary artery bypass surgery. N Engl J Med
2004; 350: 21-28.
25. Nathoe HM, van Dijk D, Jansen EW et al
. A comparison of on-pump and off-pump coronary bypass surgery in low risk patients. N Engl J Med
2003; 348: 394-402.