Do beating heart techniques applied to combined valve and graft operations reduce myocardial damage?☆ Free

Abstract

We examined the outcomes of combined beating heart CABG and valve surgery (hybrid) and compared these to conventional CABG and valve surgery (conventional). Between April 1997 and March 2006, 388 patients received combined CABG and valve surgery. Patient characteristics and cardiac enzyme release were collected prospectively. To account for differences in case-mix we used logistic regression to develop a propensity score for hybrid group membership and then performed a propensity-matched analysis. One hundred and forty patients underwent hybrid operation with a mean logistic EuroSCORE of 13.5%, compared to 248 who underwent conventional operation with a mean logistic EuroSCORE of 10.9% (P=0.006). Eighty-two patients from each group were successfully matched. The mean logistic EuroSCORE after matching was similar between the groups (11.3% vs. 12.9%; P=0.48). The median number of grafts per patient was also similar, three in each group (P=0.98). Post-op CK-MB levels were found to be significantly lower for hybrid patients (44 U/I vs. 29.5 U/I; P=0.037). In-hospital mortality was not statistically different (9.8% vs. 6.1%; P=0.39). Survival at 5 years was 74% for hybrid and 71% for conventional group (P=0.92). CK-MB levels in patients receiving hybrid CABG and valve surgery are reduced compared to conventional CABG and valve surgery.

1. Introduction

Duration of cardiopulmonary bypass and aortic cross-clamp time (ischaemic time of myocardium) are two major intra-operative factors that influence morbidity and mortality after cardiac surgery. In the last few years, a number of different techniques have emerged which are directed towards minimising the deleterious effects of cardiopulmonary bypass and optimising myocardial protection. This includes mini-bypass and heparin bonded circuits to minimise the inflammatory response as well as a variety of myocardial protection strategies including cardioplegia solutions, temperature and mode of delivery. In recent years, off-pump coronary artery bypass surgery (OPCAB) has become well established as an alternative method to surgical revascularisation without the need for cardiopulmonary bypass.

Combining the use of beating heart techniques to the coronary artery bypass graft (CABG) part of combined CABG and valve procedures will reduce the global ischaemic time. There is very little published literature examining the effects of reducing cross-clamp time in the setting of combined CABG and valve surgery. Gersak and Sutlic [1], described the technique of performing CABG and valve procedure with beating heart supported with a cardiopulmonary bypass machine and cross-clamping the aorta. The metabolic requirement of the heart was met by continuous retrograde warm blood only.

Since 1999 we have adopted a ‘hybrid’ approach by performing CABG on the beating heart supported by cardiopulmonary bypass followed by aortic cross-clamping and cardiac arrest prior to the valve procedure being undertaken. This report seeks to quantify any differences in myocardial damage (enzyme release), other postoperative outcomes and mid-term survival.

2. Methods

2.1. Patient population and data

We conducted a cohort study using prospectively collected data on 388 consecutive patients undergoing combined CABG and valve surgery between 1 April 1997 and 31 March 2006 at the Cardiothoracic Centre-Liverpool. All operations (conventional and hybrid) were performed by two surgeons (DMP and BMF).

Preoperative data were collected prospectively during patient admission as part of routine clinical practice. Methods of data collection and definitions have been previously published [2] and are also available from www.nwheartaudit.nhs.uk. Preoperative and operative data collected are listed in Table 1 . In-hospital outcomes collected included mortality, cardiac enzyme release, intensive care unit and postoperative length of stay. In-hospital mortality was defined as death within the same hospital admission regardless of cause.

2.2. Cardiac enzyme release

Cardiac enzyme release (CK-MB fraction) within the first 24 h post-op was abstracted from a routinely recorded electronic clinical biochemistry archive, blind to clinical data. The cardiac enzymes were measured on the morning after the operation, approximately 12–18 h after the patient's arrival in the intensive care unit. The assay used was CK-MB measured by the immuno-inhibition method performed on the Roche 917 Analyser (reference range 5–24 U/l at 37 °C).

2.3. Patient follow-up

Patient records were linked to the National Strategic Tracing Service (NSTS), which records all deaths in the UK. To establish vital status at one year after operation, patients were matched to the NSTS based on patient name, National Health Service number, date of birth, gender, and postcode.

2.4. Surgical technique

In all operations normothermic (37 °C) cardiopulmonary bypass was used. In the hybrid group, the CABG procedure was done on the beating heart supported by cardiopulmonary bypass. The Octopus IV (Medtronic Inc.) stabilisation device and intracoronary shunts (Medtronic Inc.) were used during construction of distal anastomoses. Thereafter, the heart was arrested with antegrade cold blood cardioplegia and maintained in an arrested state with continuous retrograde cold blood cardioplegia. With the heart arrested, the concomitant aortic or mitral valve procedure was performed. All the graft top ends to the ascending aorta were performed with the cross-clamp applied. Upon completion of the procedure, a terminal hot shot of blood cardioplegia was given via the coronary sinus and the cross-clamp released.

In the conventional group, CABG and valve surgery were done with the cross-clamp on and the heart arrested and maintained in the arrested state with the above technique. Upon completion of the procedure, a terminal hot shot of blood cardioplegia was given via the coronary sinus and the cross-clamp released.

2.5. Statistical methods

Continuous variables not normally distributed are shown as median with 25th and 75th percentiles. Categorical data are shown as percentages. Univariate comparisons were made with Wilcoxon rank sum tests and χ²-tests as appropriate. Deaths occurring over time were described using Kaplan–Meier survival curves [3].

To account for differences in case-mix we developed a propensity score for hybrid group membership [4]. The propensity for hybrid group membership was determined without regard to outcome, using multivariable logistic regression analysis [5]. A full non-parsimonious model was developed that included all variables listed in Table 1 and the year of surgery. The goal is to balance patient characteristics by incorporating everything recorded that may relate to either systematic bias or simply bad luck. This model yielded a C statistic of 0.82, indicating a good ability to differentiate between patients with or without hybrid CABG and valve surgery. We then used a macro (available at: http://www2.sas.com/proceedings/sugi29/165-29.pdf) to perform propensity-matching.

In all cases a P-value <0.05 was considered significant. All statistical analysis was performed retrospectively using SAS for Windows Version 8.2.

3. Results

3.1. Un-matched patients

Patient characteristics based on the operation received are shown in Table 1. One hundred and forty patients underwent hybrid approach with a mean logistic EuroSCORE of 13.5%, compared to 248 who underwent conventional surgery with a mean logistic EuroSCORE of 10.9% (P=0.006). Patients who received hybrid CABG and valve surgery were more likely to have triple-vessel disease, left main stem disease and poor ejection fraction. The hybrid group was also more likely to undergo non-elective surgery and received significantly more grafts during the procedure.

In-hospital mortality and cardiac enzyme release were not statistically significant between the hybrid and conventional groups as shown in Table 2 . However, intensive care and postoperative length of stay were different. No difference existed in follow-up mortality between the groups. Freedom from death for patients receiving hybrid CABG and valve surgery at 30 and 60 months was 78.8% and 74.0%, respectively, compared to 80.0% and 71.3% for patients receiving conventional CABG and valve surgery (Fig. 1 ).

3.2. Propensity-matched patients

Eighty-two patients from each group were successfully matched. The mean logistic EuroSCORE after matching was similar between the groups (11.3% vs. 12.9%; P=0.48). Patient characteristics were well matched as shown in Table 3 . The proportion of patients in each group with triple-vessel disease, left main stem disease, poor ejection fraction and received non-elective surgery was no longer statistically significant. The median number of grafts per patient was also similar, three in each group (P=0.98).

The median duration on cardiopulmonary bypass was 144 min (25th and 75th percentiles: 123–173) for patients receiving hybrid CABG and valve surgery compared to 154 min (25th and 75th percentiles: 134–168), which just failed to reach statistical significance (P=0.058). The median duration of aortic cross-clamp time was significantly reduced when using the hybrid approach compared to conventional CABG and valve surgery [66 min (25th and 75th percentiles: 59–75) vs. 106 min (25th and 75th percentiles: 96–119); P<0.001].

Table 4 shows that after propensity-matching, postoperative CK-MB levels were found to be significantly lower for hybrid patients (29.5 U/I vs. 44 U/I; P=0.037). No differences in outcomes existed between the groups with respect to in-hospital mortality and intensive care and postoperative length of stay. Follow-up mortality was also still not significantly different. Freedom from death for patients receiving hybrid CABG and valve surgery at 30 and 60 months was 78.4% and 70.2%, respectively, compared to 81.3% and 72.7% for patients receiving conventional CABG and valve surgery (Fig. 2 ).

4. Discussion

The primary aim of this study was to determine whether there was any significant difference in myocardial damage between the hybrid and conventional groups of combined CABG and valve surgery. We found that the hybrid group had significantly lower CK-MB release in the first twenty-four hours post surgery compared with the conventional group.

The initial analysis of the un-matched data found that there was no difference in postoperative CK-MB levels, and patients receiving the hybrid technique for combined CABG and valve surgery had longer intensive and postoperative stays. However, the hybrid patients were sicker patients with poorer left ventricular ejection fraction and more extensive coronary disease; receiving on average two more bypass grafts than the group treated with conventional CABG and valve surgery. After performing propensity-matching, which aims to balance out differences between the two groups, the CK-MB levels in the hybrid group were significantly lower compared to the conventional group (29.5 U/I vs. 44 U/I; P=0.037).

This evidence for decreased myocardial necrosis did not, however, translate to a decreased intensive care and total in-hospital stay or mortality. Although the cardiac enzyme release is significantly different between the two groups, the average enzyme rise for both groups is less than three times the upper limit of reference range (ULR). Previous work from this institution has shown that intermediate (3–6 times ULR) and high (>6 times ULR) perioperative cardiac enzyme release are independently associated with increased one-year mortality following isolated CABG [6].

Further work from our centre has also shown that cardiac enzyme release is reduced with OPCAB compared to CABG done on cardiopulmonary bypass but this was not associated with better 30-day and one-year survival [7]. Thus, in this report although we have shown a significantly lower cardiac enzyme release in the hybrid group, the cardiac enzyme release in both groups is actually ‘low’ and probably reflects adequate myocardial protection. It is therefore not surprising that both peri-operative and mid-term results are not significantly different.

The major limitation of the present study is its retrospective nature and the non-randomised allocation of the surgical intervention technique. The treatment groups do have differing baseline characteristics and even though we used propensity-matching to balance out these differences, this cannot categorically exclude subtle selective influences on our measured outcomes. As such we can only demonstrate association and not causality for our observations. Other limitations to the current study include the low numbers of cases for individual aortic and mitral valve pathologies, which preclude us from sub-analysing the CK-MB values and outcomes of these individual groups of patients. It should be noted that the low number of patients in the propensity-matched analysis could also have a risk of type II error. The long time period covered by the study could also have an impact on our findings. However, we attempted to minimise the impact this would have by incorporating the year of surgery into the propensity-matching. It could also be argued that a single measure of cardiac enzyme release on the morning after their operation may not reflect the peak CE levels which might have been reached. However, in this study, the emphasis is not on a specific value of cardiac enzyme release, rather it is a comparison of cardiac enzyme release between two techniques. The cardiac enzyme levels were measured at around the same duration after the operation (12–18 h) for both patient groups; hence, comparison of the measured values between the two groups should be possible. Another limitation is that we did not use more cardio-specific markers of cardiac myocyte injury such as troponin T or I, which although less widely accepted and used as markers of myocardial injury following cardiac surgery, are less likely to be raised by non-cardiac injury and may have additional prognostic value over CK-MB measurements [8]. A final limitation is the fact that we do not have data available regarding cause of death.

In conclusion, we have showed that hybrid CABG and valve surgery does result in lower CK-MB release compared with conventional CABG and valve surgery. Although significant, the difference in CK-MB levels between the two groups is weak (P=0.037) and therefore a randomised controlled trial may be warranted to confirm our findings. In this experience in-hospital outcomes and mid-term survival were similar between the two groups.

References

Conference discussion

Dr. R. Dion (Genk, Belgium): The authors compared the hybrid vs. conventional techniques in combined valve and CABG operations using propensity-matched analysis. As the authors themselves state, it is a very useful strategic tool, but it does not replace a true prospective randomized study. I also have some doubt about the statistical power. For instance, a difference in mortality of 9.8% vs. 6.1% is still non-significant. Also the authors claim a difference in the CK-MB levels, 44 for the conventional and 29 for the hybrid group, but P=0.037, which is not so highly significant.

The described reduction in myocardial damage did not translate to decreased intensive care or total in-hospital stay. The proposed explanation is that the limited difference in myocardial damage between the two groups did not significantly impair ventricular function postoperatively.

I have a few questions and remarks. The CK-MB levels were measured approximately 12–18 h after arrival in intensive care. We know that the CK-MB levels vary with time: why then not display a time course of variation of the CK-MB, and not only after arrival in intensive care: it would have probably been better to start it as from the end of the cross-clamping time. Instead of the CK-MB, for which the time course is important, why did the authors not choose to follow troponin I for which they could have taken only the maximal value?

Also the authors did not study the influence of concomitant valve surgery on the enzymatic production. In our practice we have seen a variation in troponin I production postoperatively according to the type of aortic valve pathology (stenosis or insufficiency), the type of mitral valve exposure, whether transseptal or directly through the left atrium, and even the type of mitral valve procedure (repair vs. replacement). Therefore, I would like to know whether the influence of valve surgery was taken into account in their measurement.

I believe that the authors have proven the feasibility of the hybrid technique. It is at least as good as the conventional method. But they did not convince me that it is better.

Finally, what did this work change in their clinical practice? In other words, who are the patients that you would still treat conventionally in your practice?

Dr. Lu: I'll take the questions one at a time. With regards to timing of the CK-MB, although you are absolutely right in stating that the serial CK-MBs are of most use in terms of knowing the time course of it, at the cardiothoracic center we always have a day 1 CK-MB release done prior to the patient being discharged from the intensive care. However, having said that, if there was a significant rise in the CK-MB, we will continue to monitor that over a few days. So the value that we quoted is actually the peak value over the course of the whole period.

Secondly, with regards to troponin T usage for assessing myocardial damage, there is no doubt, as you rightly say, that it is a far more sensitive and specific measurement compared with the CK-MB. The reason why we do CK-MB studies in our institution is for only one simple reason, that we've been doing it for many years, and not only that, but there were two papers that we did recently in our institution in which we were able to identify the degree of myocardial damage based on CK-MB and relate that to the outcome in the patients, and that is the reason why we persist in using the CK-MB, so that we can relate this to how well the patient does postoperatively in the midterm outcome.

To answer your second question regarding the type of operation, you are absolutely right in saying that the type of operation, the type of pathology significantly influences the CK-MB release level. Obviously aortic stenosis where the myocardium is significantly hypertrophied, the myocardial protection strategies and the endocardial ischemia will result in an increased chance of CK-MB release compared with a heart that has mitral regurgitation. However, the problem with this study, as you can see, is that the numbers are actually quite small once it's propensity-matched, so we could not actually meaningfully subanalyze these groups of patients.

The final thing with regards to suitability and in terms of what the clinical implications are of this, the one thing we can say is that this particular technique in our hands applies very well when you know that you have a patient who has a sick heart, who is going to have a prolonged cardiopulmonary bypass time, and requiring numerous grafts. Although CK-MB release is not that much different between the two groups, it is in the sicker hearts that the marginal gain is significant.

Dr. P. Menasche (Paris, France): I have a question. The duration of cross-clamping is certainly important, but the duration of cardiopulmonary bypass is equally important. Have you tried to push the concept a little bit further and have a totally off-pump bypass surgery, then followed by conventional valve replacement under cardiopulmonary bypass?

Dr. Lu: I can relate to that question as well. Again, the most important thing in assessing what methodology to use is to make sure we understand the pathology of the heart to start off with. If you have a patient who has significant aortic stenosis and multiple-vessel disease, this approach is not very safe. We noticed at the beginning of our modification of techniques that when you do things like this, you end up with two problems: you end up with global myocardial ischemia because of the significant hypertrophied LV that does not tolerate the OPCAB maneuver, and not only that, but with the significant aortic stenosis, you have a deleterious effect on the cardiac output, and that can sometimes translate to decreased end-organ perfusion, and we noticed that particularly in this group. So caution must be taken. On the other hand, however, if you have a patient who has mitral regurgitation and the LV is well preserved, you are absolutely right in saying that they do tolerate it better, although some of them do have the expense of having pulmonary edema postop. So it's all the balances we need to deal with.