Microplegia in cardiac surgery: Systematic review and meta‐analysis

Consensus on the optimum choice of cardioplegia remains elusive. One possibility that has been suggested to have beneficial properties is microplegia, a cardioplegia of reduced crystalloid volume. The aim of this meta‐analysis is to comprehensively investigate microplegia against a range of clinical outcomes.


| INTRODUCTION
Cardiopulmonary bypass (CPB) provides physiological support through an extracorporeal circuit, circumventing the heart, which is afforded a bloodless field by aortic cross-clamp. Uncontrolled myocardial oxygen consumption in these conditions and at the time of unclamping could predispose the myocardium to injury. By perfusing the heart with a cardioplegic solution to induce and maintain electromechanical arrest, Sarkar et al 1 acknowledge that cardioprotective strategies arise from managed myocardial oxygen consumption.
Cardioplegia modalities were based on three main principles: chemical arrest, hypothermia, and the addition of extra antiischaemic chemicals. 2 Over the years, there have been many variations of cardioplegia in terms of the ingredients of crystalloid cardioplegia, the temperature and the direction of delivery. 3 One of the most enduring debates is whether to use blood or crystalloid cardioplegia.
The argument for blood cardioplegia centres around its buffering ability, normal oncotic pressure, free radical scavenging potential, and improved oxygen delivery. 3,4 The two meta-analyses that have been carried out, suggest a lower risk of perioperative myocardial infarction (MI), 4 a lower incidence of low cardiac output syndrome (LCOS), and lower creatine kinase myocardial band (CK-MB) release with blood cardioplegia. 5 Standard formulation of blood cardioplegia tends to be either four or eight parts blood to one part crystalloid. 6 Pure blood cardioplegia of reduced volume and delivered intermittently, 7 otherwise known as microplegia is an alternative to this. The main advantages reported here include less haemodilution, reduced myocardial oedema, and rapid recovery of ventricular function. [6][7][8] In addition, the intermittent dosing in microplegia could play a vital role in managing harmful inflammatory responses to global ischemia and regional reperfusion, whereas diluted cardioplegia could have less control in its single-dose modality. 1 Early clinical trials reported the use of less vasopressors 8 and better postoperative left ventricular function 9 with microplegia. More recent trials have consistently shown a higher establishment of spontaneous heartbeat return using microplegia when compared to standard diluted blood cardioplegia. 10,11 One drawback to these studies was low statistical power suggesting the need for a larger analysis/meta-analysis to investigate for true differences. There has been one meta-analysis conducted, which investigated studies up to May 2013. This old meta-analysis contained a limited number of studies and outcomes. 12 The aim of the current comprehensive metaanalysis (CMA) was to investigate the effect of microplegia in a range of clinical and biochemical outcomes to other types of cardioplegia for myocardial protection for patients undergoing cardiac surgery.

| METHODS
This systematic review and meta-analysis were performed in accordance with the PRISMA guidelines (see Supporting Information Files).

| Search strategy
To identify potential studies systematic searches were carried out using: EMBASE, PubMed, Web of Science, and the Cochrane Central Registry of Controlled Trials. The search was supplemented by scanning the reference lists of eligible studies. The search strategy included the key concepts of "microplegia" OR "mini-cardioplegia" OR "miniplegia" AND "cardiac surgery" (see Supporting Information Files). All identified papers were assessed independently by two reviewers. A third reviewer was consulted to resolve disputes. Searches of published papers were conducted up until 1st February 2020.
Data were extracted using predesigned extraction tables.

| Types of trials to be included and excluded
Included trials were those that directly compared the use of microplegia versus any other type of cardioplegia in open-heart surgery.
There were no language restrictions. Animal studies and review papers were excluded. Studies that did not have any of the desired outcome measures were excluded. Incomplete data, or data from an already included study, were excluded. Other treatment modalities and interventions for coronary artery disease such as percutaneous coronary intervention and valvular disease such as transcatheter aortic valve intervention were excluded.

| Participants/population
This meta-analysis analysed randomized controlled trials (RCTs) and case-control trials of both male and female adult (≥18 years) patients with coronary artery disease or valvular disease who were undergoing cardiac surgery using either microplegia or other types of cardioplegia.

| Intervention(s), exposure(s)
This meta-analysis considered all RCTs and case-control trials where microplegia or another type of cardioplegia were used in either coronary artery bypass graft (CABG) or in patients treated for valvular disease.
More specifically, all RCTs and retrospective trials where the intervention of carrying out cardiac surgery using microplegia was performed.

| Search results
Our initial search found 46 articles of which 25 duplicate studies were excluded. A further four studies were excluded as they were investigating paediatric cases; one study was excluded as it had no reported outcomes; one paper was excluded because it was a review; one paper was excluded because it contained duplicate data; and, three studies were excluded because they had no 2738 | comparator group (Figure 1). Eleven studies were included in our analysis. 6,[8][9][10][11][13][14][15][16][17][18]

| Risk of bias (quality) assessment
Risk of bias was assessed using a modified Jadad scale for randomized controlled trials 19 and the Newcastle-Ottawa Scale (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp) for case-control studies. Publication bias was investigated using funnel plots. 20

| Data synthesis
In the following analyses, the non-microplegia group is referred to as diluted cardioplegia. Odds ratios (OR) were calculated for dichotomous data. Continuous data were compared using mean differences (MDs). Where necessary medians and interquartile ranges were converted into means and standard deviations using the method of Hozo et al. 21 Subgroup analyses were carried out to investigate differences in the diluted cardioplegia arm. All analyses were conducted using CMA version 3. Heterogeneity was quantified using the Cochrane Q test 22 and all comparisons were made using random effects. We used a 5% level of significance and 95% confidence intervals (CIs); figures were produced using CMA. Funnel plots were used to analyse pub-   Table S1. The results are summarized with mortality leading, followed by intra-operative and then postoperative variables. A summary of the results is presented in Table 2.

| Mortality
Five studies involving 5358 patients reported the incidence of mortality.
The OR for the comparison was 1.36 (95% CI: 0.95-1.93; I² = 0%; P = .090) (Figure 2). There was no significant difference in the odds of mortality between the two groups. The funnel plot was symmetrical.  Figure S1). The volume of crystalloids used in the microplegia group was significantly lower than the diluted cardioplegia group. The funnel plot was symmetrical.

| CPB time
Nine studies involving 5403 patients reported the length of time on CPB.
The MD between the two groups was −0.87 minutes (95% CI: −4.51 to 2.78; I 2 = 92%; P = .640) ( Figure S2). There was no significant difference in CPB time between patients receiving microplegia and patients receiving diluted cardioplegia. The funnel plot was asymmetrical.

| Aortic cross-clamp time
Nine studies involving 4983 patients reported the time of aortic cross-clamping. The MD between the groups was −1.07 minutes (95% CI: −4.54 to 2.39; I 2 = 90%; P = .544) ( Figure S3). There was no significant difference in aortic cross-clamp time between the two groups. The funnel plot was asymmetrical.

| Intra-aortic balloon pump
Four studies involving 4939 patients reported the requirement for IABP. The OR for the comparison was 0.81 (95% CI: 0.50-1.31; I 2 = 0%; P = .386) ( Figure S4). The two groups showed no significant difference in their requirement for IABP. The funnel plot was symmetrical.

| Spontaneous heartbeat recovery
Six studies involving 1122 patients reported the incidence of spontaneous heartbeat recovery after aortic unclamping. The OR was 4.27 (95% CI: 1.94-9.42; I 2 = 77%; P < .001) (Figure 3). There was a significantly higher incidence of spontaneous heartbeat recovery in the microplegia group than in the diluted cardioplegia group. The funnel plot was asymmetrical.

| Inotropic support
Five studies involving 1233 patients reported the need for postoperative inotropic support. The OR was 0.67 (95% CI: 0.47-0.94; I 2 = 4%; P = .021) ( Figure S5). There was a significantly lower requirement for inotropic support in the microplegia group compared to the diluted cardioplegia group. The funnel plot was symmetrical.

| Low cardiac output syndrome
Four studies involving 4149 patients reported the incidence of LCOS. The OR for the comparison was 0.67 (95% CI: 0.37-1.20; I 2 = 27%; P = .175) ( Figure S6). The incidence of LCOS was not statistically significant between the two groups. The funnel plot was symmetrical.

| Myocardial infarction
Three studies involving 4230 patients reported the incidence of MI as a postoperative complication. The OR for the comparison was 0.86 (95% CI: 0.26-2.92; I 2 = 0%; P = .811) ( Figure S7). There was no significant difference in MI incidence between the two groups. The funnel plot was symmetrical.

| Acute renal failure
Five studies involving 5358 patients reported the incidence of acute renal failure. The OR for the comparison was 0.96 (95% CI: 0.56-1.65; I 2 = 43%; P = .876) ( Figure S8). There was no significant difference in the odds of postoperative acute renal failure between the two groups. The funnel plot was symmetrical.

| Atrial fibrillation
Three studies involving 300 patients reported the incidence of postoperative AF. The OR for the comparison was 0.88 (95% CI: 0.34-2.26; I 2 = 43%; P = .784) ( Figure S9). There was no significant difference in the odds of triggering AF between the two groups. The funnel plot was symmetrical.

| Reoperation for bleeding
Three studies involving 1103 patients reported the incidence of reoperation for bleeding. The OR for the comparison was 1.15 (95% CI: 0.51-2.58; I 2 = 21%; P = .741) ( Figure S10). The two groups showed no significant difference in the need to reoperate for bleeding. The funnel plot was symmetrical.

| Hospital stay
Four studies involving 4310 patients reported the postoperative length of stay in hospital. The MD between the groups was −0.78 days (95% CI: −1.73 to 0.17; I 2 = 57%; P = .109) ( Figure S12). There was a trend toward the microplegia patients having less hospital stay time, however, this did not reach statistical significance. The funnel plot was asymmetrical.

| Subgroup analyses
A subset analysis was carried out examining all parameters in studies under 10 years old. These results are presented in Table S2. It was found that spontaneous return of heartbeat and LCOS were significantly in favour of the microplegia group with all other comparisons nonsignificant.
Subgroup analyses were also performed for RCTs. These results are shown in Table S3. The results for CK-MB, CPB time, aortic cross-clamp time, ICU time, inotropic support, and spontaneous return of heartbeat were all significantly in favour of the microplegia group.
Further subgroup analyses were carried out with the results presented in Table S4. These results largely followed the main data.

| Study quality
Two study quality tests were utilized: The Jadad Scale for RCTs and the Newcastle-Ottawa Scale for case-controlled trials. The former revealed a median score of 4 (Table S5). The Newcastle-Ottawa Scale  Figure 1 except the statistics presented are mean differences reported largely good results with a median score of 8 out of 9 (Table S6). No differences were found in mortality incidence, No significant differences were found in mortality incidence, although the odds leant toward the diluted cardioplegia group. The main contributing trial of mortality as a key outcome was Albacker et al. 17 In this trial, a significant crude mortality figure became insignificant when preoperative and operative differences were accounted for.

| CONCLUSIONS
The consensus that mortality incidence is not dissimilar in whole blood cardioplegia and diluted cardioplegia is concluded in a meta-analysis by Zeng et al 4 stating a clearer clinical correlation to reduced myocardial injury by decreased release of enzymatic markers than to mortality incidence.
ICU time was significantly reduced in microplegia patients. Independent predictors of the length of time in ICU are multifactorial but if we focus on the role of cardioplegia in protecting the myocardium from inflammatory response during ischemia and reperfusion, any reduction in myocardial oedema could, according to Onorati et al 16 improve LV functioning. This would in turn shorten the time taken for the patient to reach haemodynamic stability suitable for discharge from the ICU. A study on factors influencing prolonged ICU stay in heart surgery found a significant association between length of ICU stay and low LVEF less than 40%. 25 A shortened stay in ICU is also beneficial with respect to the mobilization of the patient reducing the risk of development of ICU-acquired weaknesses reported in 50% of patients discharged from the ICU and associated with increased 90-day mortality incidence. 26 In addition to the significantly shorter ICU stay, the hospital stay time was approximately threequarters of a day shorter. These shortened stays could indicate a significant financial saving.

| Limitations
Unfortunately, there is no consensus surrounding the choice of cardioplegia. This, therefore, created difficulty in finding a consistent control arm for this study. To overcome this subgroup, analyses were carried out for the significant results. This showed that some of microplegia's benefits were constrained under certain conditions. It was noticeable though that no matter how the data were segregated, there was always a lower CK-MB release in the microplegia group.
In the majority of the significant analyses, the funnel plot was symmetrical. The exception was spontaneous heartbeat recovery suggesting the possible existence of some publication bias in this finding.
Prospectively randomizing patients to either treatment technique was performed in six of the studies. [8][9][10][11]15,16 As the sample sizes in these studies were generally smaller, this meant that only 4.7% of the 5798 participants were randomized.

| Conclusions
A clear consensus on the optimal cardioplegic solution has been challenged by extensive variation in the composition, volume, temperature, and delivery technique of cardioplegia across surgical practices. Microplegia is one of several different alternatives mooted. This meta-analysis shows that microplegia is similar to other cardioplegia choices and may offer some advantages in terms of early return to spontaneous heartbeat, less requirement for inotropic support, shorter ICU stay, and lower CK-MB release.