Abstract
Background
The efficacy of different cardioplegia solutions on outcomes of complex cardiac operations such as triple valve surgery (TVS) is scarce. Here we compared the outcomes in TVS patients receiving either crystalloid (Bretschneider) or blood (Calafiore) cardioplegia.
Methods
Screening of our institutional database with prospectively entered data identified 471 consecutive patients (mean age 70.3 ± 9.2 years; 50.9% male), who underwent TVS (replacement or repair of aortic, mitral and tricuspid valve) between December 1994 and January 2013. In 277 patients, cardiac arrest was induced with HTK-Bretschneider solution (HTK, n = 277, 58.8%), whereas 194 received cold blood cardioplegia (BCP) according to Calafiore (n = 194, 41.2%). Comparisons of perioperative and follow up outcomes were made between cardioplegia groups.
Results
Preoperative patient characteristics and comorbidities were equally balanced between groups. 30-days mortality was similar between groups (HTK: 16.2%; BCP: 18.2%; p = 0.619). Incidence of the cumulative endpoint (30days mortality, myocardial infarction (MI), arrhythmia, low cardiac output syndrome or need for permanent pacemaker implantation) was also comparable (HTK: 47.6%; BCP: 54.8%, p = 0.149). In patients with reduced left ventricular ejection fraction (LVEF <40%), 30days mortality was higher in the HTK group (HTK 18/71 22.5%; BCP 5/50 10%; p = 0.037). Five-year survival was similar between groups (52 ± 6% for HTK and 55 ± 5% for BCP patients). In-Hospital mortality was best predicted by length of surgery and reperfusion ratio. Decreased age, shorter bypass time, preserved LVEF and concomitant surgical procedures have been found to be protective from long-term mortality.
Conclusions
Myocardial protection with HTK shows equivalent outcomes compared to BCP during TVS. Patients with reduced left ventricular function may benefit from BCP during TVS.
Introduction
Complex open-heart surgery such as triple valve surgery (TVS) can be associated with high operative morbidity and mortality.1,2,3 Recently it has been suggested that improvements in myocardial protection and perioperative care have markedly reduced the operative risk for patients undergoing TVS. 3 However, particularly the manifestation of a low cardiac output syndrome (LCOS) is known to be associated with increased morbidity and mortality 4,5 and inadequate intraoperative myocardial protection can contribute developing LCOS. Surprisingly few publications have investigated the impact of different myocardial preservation techniques on postoperative outcome after complex valve surgery such as TVS.
Despite improved perioperative care TVS remains challenging and continues to be associated with high operative risk. 6 The background of the present study was therefore to examine the cardioplegia solutions most commonly used at our center, namely repetitively applied cold blood cardioplegia modified according to Calafiore and Custodiol (HTK-Bretschneider) with regard to a patient population who are undergoing complex cardiac surgery such as TVS.
Patients and Methods
Overall 487 patients were identified who underwent TVS at our institution between December 1994 and January 2013. Patient data was collected prospectively and retrospectively analyzed regarding the intraoperative myocardial protection. For the majority of patients (56.8%; n = 277) cold crystalloid cardioplegia (Custodiol® or HTK-Bretschneider) was used for myocardial protection. Cold BCP according to Calafiore was used in 39.8% (n = 194) of the patients. In 16 patients (3.2%) miscellaneous strategies for myocardial protection were implemented. These patients were omitted from further analysis.
Main indications for surgery were valve regurgitation, stenosis or a combined pathology of the aortic, mitral and tricuspid valve. Patients suffering endocarditis, myxoma or intracardiac tumor were not excluded, whereas patients with pulmonary heart valve defect indicating triple valve repair were not considered for further statistical examination.
This study was approved by the ethics committee of the medical faculty of the University of Leipzig (No. 266-15–13,072,015) and was performed according to the Declaration of Helsinki. Written informed consent was provided by all patients who were in an adequate conscious state preoperatively.
Surgical technique and myocardial protection
Single aortic cross-clamp technique was used in all patients. Biological and/or mechanical prostheses were implanted according to guidelines, age, comorbidities and surgeon or patient preferences. Valve reconstruction was preferred over replacement.
During cardiopulmonary bypass (CPB) mild hypothermia was established (between 28°C and 34°C). There was a trend towards warmer temperatures during the observation period, data not shown. The cardioplegic solution was administered through a needle-vent that was placed in the ascending aorta proximal to the aortic cross-clamp for antegrade perfusion or through selective coronary perfusion. Access for retrograde application of the cardioplegic solution was through the coronary sinus.
For induction of elective cardiac arrest Custodiol (HTK-Bretschneider; Dr. Franz Köhler Chemie, Alsbach-Hähnlein, Germany) as a crystalloid substitute was used with an administration temperature of 4°C, as originally described by Bretschneider. 7 If cross-clamp times exceeded 90 min, reapplication of the solution was routinely conducted. When prolonged arrest was achieved with additional BCP patients were excluded from further analysis.
BCP was prepared and applied according to the Calafiore modification by Caputo. 8 Cooling was achieved with a heat exchanging device. Induction of cardiac arrest was achieved with high-dose (18-20 mEq/l [K+]) infusion of the potassium/magnesium/blood solution and arrest was maintained with repeated delivery every 20 min with decreasing concentrations of potassium as described before. 8
In our institute, the reperfusion time is set at a period corresponding to approximately one third of the cross-clamp time. In the following the relation between cardiopulmonary bypass time and cross-clamp time, is assessed as reperfusion-ratio.
Follow up
Patients were followed by the referring cardiologist and contacted periodically by our research personnel either through outpatient visit and telephone contact with patients, family members, or both. Supplemental information was gathered from family physicians and referring cardiologists. Our research personnel collected all follow-up data. Median follow up time for survival was 2.1 years.
Statistical analysis
Continuous variables were expressed as mean and standard deviation. Categorical variables were expressed as absolute numbers and percentages. Differences between groups were analyzed using the student t-test and the Fishers exact test as appropriate. Hospital mortality was defined as death in house or within 30 days post discharge. To evaluate which preoperative characteristics could possibly motivate the surgeon to choose a certain cardioplegic solution, the Odds ratio (OR) was calculated for each categorical variable. OR´s were also calculated to assess the risk of adverse outcomes depending on the intraoperative treatment for myocardial protection. Subgroup analyses were conducted according to the comorbidities which could have triggered the use of a certain cardioplegic solution such as prior myocardial infarction (MI), coronary artery disease (CAD), and concomitant CABG or heart failure with reduced ejection fraction (HFrEF, ≤40%). Kaplan-Meier survival curves were calculated and tested by log rank to describe the differences in survival. A multivariable binomial logistic (in-hospital death) and semi-parametric cox-proportional hazards (long-term survival) regression model were used to predict outcomes. Variables introduced into these models were chosen based on clinical relevance. Bootstrapped stepwise backward elimination onto these variables and relevant statistical interactions by cutoff p < 0.05 was applied. Odds-Ratio (in-hospital death) and Hazards Ratio (long-term survival) as well as confidence intervals and p-value per covariate were calculated. Additionally goodness of fit was evaluated by Harrell’s c-statistic and confidence interval as well as likelihood-ratio test. All analyses were performed with Sigma plot (Sigma Plot 12.5, Germany). p-values <0.05 were considered to be statistically significant. Data were reported according to Akins and colleagues. 9
Results
Patient characteristics
Baseline demographics.
Abbreviations: COPD, chronic obstructive pulmonary disease; CAD, coronary artery disease; CVA, cerebrovascular accident; ESRD, end stage renal disease; BCP, blood cardioplegia; BMI, body mass index; HTK, crystalloid cardioplegia; LV-EF, left ventricular ejection fraction.
Operative data.
Abbreviations: ASD, Atrial Septal Defect; CABG, Coronary Artery Bypass Grafting; CPB, cardiopulmonary bypass; X-time, aortic cross-clamp time.
Operative data
Mean CPB- time (HTK: 157.3 ± 55 min; BCP: 170 ± 51 min; p = 0.003) and CX-time (HTK: 103 ± 32 min; BCP: 116.3 ± 36 min; p < 0.001) were significantly shorter in the HTK group (Table 2). Mean body temperature during CPB was significantly lower in the HTK-group compared to the BCP group (HTK: 32.3 ± 2.1°C; BCP: 32.8 ± 1.8°C; p = 0.003). Reperfusion-ratio was significantly increased in HTK-group (1.45 vs 1.41, p = 0.008).
Examination of concomitant procedures (Table 2) revealed that in 55.6% (n = 262) of the patients additional surgery was necessary (n = 262). More patients in the BCP group needed extra surgical treatment (HTK: 50.9%, n = 141; BCP: 68.3%, n = 121; p < 0.001). CABG was more frequently conducted in the BCP group (HTK: 19.9%, n = 55; BCP: 29.9%, n = 59; p = 0.015). While replacement of the ascending aorta was more often observed in the HTK group (HTK: 20.6%, n = 57; BCP: 11.3%, n = 22; p = 0.008).
Early postoperative morbidity and mortality
Postoperative complications.
Abbreviations: LCOS, low-cardiac-output syndrome; ECMO, extracorporeal membrane oxygenation; IABP, intra-aortic balloon counter pulsation; CVA, cerebrovascular accident; RBC, red blood cell; cumulative endpoint (30 days mortality, low cardiac output, myocardial infarction, arrhythmia, permanent pacemaker implantation).
Depiction of reperfusion-ratio by bypass-time, stratified by earyl mortality.
Overall survival and subgroup analyses
Kaplan-Meier analysis for overall survival (Figure 2) showed no significant differences between both groups (p = 0.962). One-year survival was 73 ± 2.6% for HTK and 72 ± 3.3% for BCP. Survival at 5-years follow-up was 56 ± 3.4% for HTK and 55 ± 4.2% for BCP. Multivariate Cox regression resulted in age at surgery (HR: 1.05, CI: 1.03 – 1.07, p < 0.001), LVEF (HR: 0.98, CI: 0.97 – 1.00, p = 0.004), additional procedures without CABG (HR: 0.62, CI: 0.43 – 0.88, p = 0.007) and bypass time (HR: 1.01, CI: 1.00 – 1.01, p < 0.001) being predictive. Other variables were eliminated by the stepwise backward approach. C-Statistic was 0.67 (CI: 0.62 – 0.71). Subsequent evaluation of in-hospital mortality for patients with recent MI (HTK 5/18, 27.7%; BCP 7/25, 29.1%; p = 1.000), CAD (HTK 6/26, 23.1%; BCP 2/18, 11.1%; p = 0.435) or concomitant CABG (HTK 16/55 29.1%; BCP 13/58 23.4%; p = 0.519) revealed no significant differences. The occurrence of postoperative complications, in this subpopulation, was also not statistically different between HTK and BCP. Only manifestation of postoperative renal failure requiring hemodialysis (HTK: 4/18 22.2%; BCP: 10/24 41.6%) was more often observed in patients with prior MI and arrested with BCP intraoperatively (p = 0.041). In patients suffering HFrEF in-hospital mortality was higher in the HTK group (HTK 18/71 22.5%; BCP 5/50 10%; p = 0.037), while long-term survival was not different (Kaplan-Meier p = 0.655). Kaplan-Meier analysis for overall survival after TVS with BCP (mod. Calafiore) versus HTK (HTKBretschneider); Abbreviations: BCP, blood cardioplegia; HTK, crystalloid cardioplegia.
Discussion
In this single-center study with 471 consecutive patients receiving TVS, in 277 patients cardioplegic arrest with crystalloid HTK-Bretschneider (HTK) solution was established, 194 patients underwent myocardial protection with cold blood cardioplegia (BCP). No significant differences were found for hospital mortality and 1-year survival between BCP and HTK. Also no significant differences were found regarding postoperative complications between both groups. In the subgroup analysis for patients presenting with reduced left ventricular ejection fraction a higher operative mortality in the HTK-group was detected.
In the present study, there were no significant differences in the preoperative risk factors between the groups. Except for history of recent myocardial infarction which was more frequently in the blood-cardioplegia group. This balance of preoperative variables between the groups has prompted us to forego a propensity score matching. However, this should not hide the fact that a TVS cohort as a whole is very heterogeneous and such a study can and should only serve to formulate new hypotheses. In a recent single center experience Noack and colleagues were able to show that TVS is still associated with high operative mortality and that the development of LCOS is a relevant complication. 6 The development of postoperative LCOS has been described as multifactorial with insufficient myocardial protection intraoperatively being one of the major contributors. 2 In the present analysis the development of LCOS was not different between patients who received HTK compared to BCP.
Recently Viana and colleagues reported a propensity score matched comparison of patients receiving either HTK-Bretschneider or blood cardioplegia during complex cardiac surgery. 10 They found no significant differences between both cardioplegic strategies concerning operative mortality and morbidity. The working group of Viana presented a cohort with prolonged cross-clamp times suggesting extended cardiac and aortic repairs with only n = 6 triple valve patients. However, the application of the blood cardioplegia was tepid, administered repeatedly and was applied antegradely and retrogradely whereas HTK was given as a single dose. 10
Despite the many years of experience with HTK comparisons with other widely used cardioplegic techniques are rare. To the best of our knowledge this is the first report on the comparison of HTK with cold BCP according to Calafiore in a large and relatively well outlined patient cohort receiving TVS. However, it must also be mentioned that the patients in the cohort underwent complex heart surgery, but are also in some cases to be regarded as high-risk patients per se.
Backward stepwise binomial logistic regression model for full cohort.
Abbreviations: OR, Odds ratio; LCI, Lower confidence interval; UCI, upper confidence interval; CVA, cerebrovascular accident; CPB, cardiopulmonary bypass, GOF, goodness of fit; AUC, Area under the curve.
Remarkable developments of cardioplegic solutions have not been made over the last 20 years, whereas systemic hypothermia during cardiopulmonary bypass (CPB) shows a trend culminating in the progressive practice of moderate-to-mild temperatures and even normothermia in recent years. The impetus for progressive temperature elevation is the limitation of adverse effects of hypothermia and probably especially because of significantly shorter cooling and rewarming periods while on CPB. 12 Hypothermia, however, has been proposed as one of the essentials of myocardial protection. 13 There were no significant differences according to the systemic temperature during extracorporeal circulation between both groups in this series. The influence of normothermia without topical cooling on myocardial protection is probably underreported but has been emphasized in experimental set-ups and thus warrants further investigation. 7
Conclusions
We conclude that myocardial protection with HTK-Bretschneider in high-risk patients undergoing TVS is safe and convenient. We observed a progressive trend towards mild to normothermic systemic temperatures during CPB throughout the observation period. We need prospective clinical trials focusing on the comorbidities of the patients in order to minimize any potential perioperative myocardial damage.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
