Abstract
Objectives. To evaluate the impact of volatile anesthetic choice on clinically relevant outcomes of patients undergoing cardiac surgery. Methods. Major databases were systematically searched for randomized controlled trials (RCTs) comparing volatile anesthetics (isoflurane versus sevoflurane) in cardiac surgery. Study-level characteristics, intraoperative events, and postoperative outcomes were extracted from the articles. Results. Sixteen RCTs involving 961 patients were included in this meta-analysis. There were no significant differences between both anesthetics in terms of intensive care unit length of stay (SMD −0.07, 95% CI −0.38 to 0.24, ), hospital length of stay (SMD 0.06, 95% CI −0.33 to 0.45, ), time to extubation (SMD 0.29, 95% CI −0.08 to 0.65, ), S100 (at the end of surgery: SMD 0.08, 95% CI −0.33 to 0.49, ; 24 hours after surgery: SMD 0.21, 95% CI −0.23 to 0.65, ), or troponin (at the end of surgery: SMD −1.13, 95% CI −2.39 to 0.13, ; 24 hours after surgery: SMD 0.74, 95% CI −0.15 to 1.62, ). CK-MB was shown to be significantly increased when using isoflurane instead of sevoflurane (SMD 2.16, 95% CI 0.57 to 3.74, ). Conclusions. The volatile anesthetic choice has no significant impact on postoperative outcomes of patients undergoing cardiac surgery.
1. Introduction
Multiple studies have shown the potential benefits of volatile anesthetics over the intravenous anesthetics in cardiac surgery [1]. Several meta-analyses have demonstrated that patients anesthetized with volatile anesthetics tend to have lower postoperative levels of troponin as compared with intravenous anesthesia [2–5], and consequently evidenced cardioprotective effects result in a decreased morbidity (i.e., reductions of myocardial infarctions, intensive care unit and hospital stay, time on mechanical ventilation, and incidence of long-term cardiac events) and mortality [6]. These benefits have also been extrapolated in noncardiac surgery according to a recent meta-analysis [7].
There are several opinions regarding what is the preferred volatile anesthetic. It is known that the newer drug sevoflurane has some chemical advantages over isoflurane (e.g., less solubility resulting in a faster onset and offset of action, less irritating to the airway, and being not as pungent as isoflurane; thus it can be used for inhalational induction of anesthesia [1, 6]. Moreover, there appears to be an opinion among cardiac anesthesiologists that sevoflurane is superior to isoflurane) [8]. However, there is still no a definite response if there is a volatile anesthetic with the best profile, neither a consensus nor recommendation about this topic. We undertook this meta-analysis comparing volatile anesthetics (isoflurane versus sevoflurane) in cardiac surgery, with the aim of evaluating the impact of volatile anesthetic choice on clinically relevant outcomes of patients undergoing cardiac surgery.
2. Methodology
This meta-analysis was performed according to Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) and following the recommendations of the Cochrane Collaboration [24, 25].
2.1. Eligibility Criteria
2.1.1. Types of Studies
Two authors reviewed the literature and screened the abstracts independently. They selected all relevant articles in full text for detailed comprehension and further assessment of the quality and agreement of inclusion criteria. This meta-analysis focused on randomized controlled trials (RCTs). We did not restrict our selection criteria to studies developed at specific regions nor studies with very low sample size.
2.1.2. Types of Participants
The participants included in this meta-analysis were adult patients (age > 18 years) who underwent cardiac surgery with either isoflurane or sevoflurane anesthesia.
2.1.3. Types of Intervention
The use of isoflurane for maintenance of general anesthesia was considered the intervention in this meta-analysis, and the control group consisted of sevoflurane anesthesia.
2.1.4. Types of Outcome Measures
The primary outcomes in this meta-analysis were the intensive care (ICU) length of stay and hospital length of stay. Secondary outcomes included the time to extubation, S100, CK-MB, and troponin at the end of surgery and 24 hours after surgery.
2.2. Inclusion Criteria
(i)RCTs compared sevoflurane with isoflurane used for maintenance anesthesia in cardiac anesthesia.(ii)RCTs should recruit adult patients (age > 18 years) undergoing cardiac surgery (on-pump and off-pump).
2.3. Exclusion Criteria
We excluded studies that compared volatile anesthesia with nonvolatile anesthesia.
2.4. Literature Search
The MEDLINE/PubMed (from 1950 to Feb 2017), Google Scholar (from 1960 to Feb 2017), EMBASE (from 1980 to Feb 2017), and Cochrane library (from 1990 to Feb 2017) were searched for randomized controlled trials (RCTs) comparing the effects of isoflurane with sevoflurane anesthesia in cardiac surgery.
2.5. Search Strategy
The terms included in the search strategy were “cardiac surgery”, “volatile anesthesia”, “isoflurane”, “sevoflurane”, “troponin”, “hospital stay”, “intensive care unit stay”, and “randomized controlled trial”. We did not restrict for language. In addition, we reviewed citations of included articles in order to ensure inclusion of relevant studies not captured in our initial literature search.
2.6. Data Extraction and Management
Two authors verified and extracted the data of the eligible articles. They completed a predefined database in Excel that contained all the possibly relevant variables for this meta-analysis (year of publication, sample size, mean age, anesthetic regimen, type of surgery, and the outcomes).
2.7. Assessment of Methodological Quality
Two authors performed the methodological quality assessment and no disagreement arose. The quality of each study included in this meta-analysis was assessed by the Cochrane review criteria for randomized studies. The score was calculated for each study based on seven items (random sequence generation, allocation concealment, blinding of personnel who performed anesthesia, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias). Each item was scored between 2 and 0 (being 2 “positive,” 1 “unclear,” and 0 “negative”).
2.8. Statistical Analysis
First, an exploratory qualitative analysis was conducted to describe the characteristics of the studies included in this meta-analysis. The meta-analysis was performed using the Review Manager 5.3 (Cochrane Collaboration, Oxford, UK) with random-effect model (DerSimonian & Laird method) [26]. The duration of analgesia, pain scores and sensory and motor block duration, and opioid consumption were extracted as continuous variables and compared using standardized mean difference (SMD) with their respective 95% confidence intervals (CI). We used forest plots to illustrate the estimations and overall effect sizes (pooled SMD represented as a solid diamond at the bottom of the forest plot). PONV was extracted as a dichotomous outcome (present or absent) and compared using risk ratios (RRs). Heterogeneity of each meta-analyzed value was assessed by () with the correspondent chi-squared test (% and % were considered insignificant and significant heterogeneity, resp.). Publication bias was calculated using Stata version 13.0 (Stata, College Station, TX) with Begg’s and Egger’s test [27]. Funnel plots were constructed to represent any tendency for publishing in favour of the positive effect. Significant publication bias was considered when there was asymmetry in the funnel plot (meaning that smaller studies tend to show larger SMDs). values < 0.05 were considered as statistically significant in all statistical analyses.
3. Results
3.1. Literature Search Results
In total sixteen RCTs were included in the meta-analysis. The trial flow diagram illustrates the number of excluded and included articles in detail (Figure 1).
3.2. Study Characteristics
Sixteen potentially eligible articles were reviewed [8–23]. All the studies were published as original articles. Table 1 summarizes the characteristics of the studies included in the analysis.
3.3. Meta-Analysis
There were no significant differences between both volatile anesthetics in terms of intensive care unit length of stay (SMD −0.07, 95% CI −0.38 to 0.24, ; Figure 2), hospital length of stay (SMD 0.06, 95% CI −0.33 to 0.45, ; Figure 3), time to extubation (SMD 0.29, 95% CI −0.08 to 0.65, ; Figure 4), S100 (at the end of surgery: SMD 0.08, 95% CI −0.33 to 0.49, ; 24 hours after surgery: SMD 0.21, 95% CI −0.23 to 0.65, ; Figure 5), or troponin (at the end of surgery: SMD −1.13, 95% CI −2.39 to 0.13, ; 24 hours after surgery: SMD 0.74, 95% CI −0.15 to 1.62, ; Figure 6). CK-MB was shown to be significantly increased when using isoflurane instead of sevoflurane (SMD 2.16, 95% CI 0.57 to 3.74, ; Figure 7); however this result was strongly influenced by only one study. Subgroup analysis by type of surgery (CABG or valvular surgery) showed no differences; there was only one study that included valvular surgery [11]. Therefore, the results did not change significantly.
3.4. Publication Bias
Funnel plots were conducted to assess the publication bias in this meta-analysis of included studies. As shown in Supplemental File in Supplementary Material available online at https://doi.org/10.1155/2017/7073401, there was no evident asymmetry in the funnel plots. Therefore, the result suggested a low probability of publication bias. The quality assessment criteria ranged from 13 to 7 points for evidence synthesis (Figure 8).
4. Discussion
In this study, the use of isoflurane and sevoflurane was analyzed to obtain powerful conclusions regarding their outcomes in cardiac surgery. To our knowledge, this is the first meta-analysis providing a comparison between two inhaled anesthetics in patients undergoing cardiac surgery, showing that the difference was not statistically significant between the use of isoflurane and sevoflurane.
In recent years there have been several studies comparing anesthetics used in cardiac surgery because the latter represents a remarkably cause of morbidity and mortality worldwide, especially in developing countries [19, 28, 29]. Importantly, in cardiac anesthesia volatile anesthetics it is still one of the most important pharmacologic resources for anesthesia maintenance; it is because this type of anesthesia has a better profile (cardioprotective and neuroprotective) than nonvolatile anesthesia. In this scenario, some authors have hypothetically considered the implication of specific inhaled anesthetics in perioperative and postoperative complications. In terms of the properties of volatile anesthetics, relative blood/gas solubility isoflurane is higher than sevoflurane (1.38 and 0.66, resp.); therefore the last has the longer half-life in plasma; and the estimated minimum alveolar concentration (MAC) is the same in both anesthetics (1.15% and 2.05% per hour correspondingly) [22, 30]. Sevoflurane generally costs the most and isoflurane the least. Nevertheless, it is difficult to estimate the precise cost of each inhaled anesthetic due to sharing of devices used to vaporize the medication and the individual dosage according to the weight of each individual [31]. In summary, although the general properties of individual inhaled anesthetics are different, there is not a preferred volatile anesthetic so far in clinical practice.
The overall results of the randomized clinical trials included in this study did not show statistically significant difference between the use of isoflurane and sevoflurane in terms of the primary clinical outcomes ICU length of stay and time of extubation (SMD = −0.07; 95% CI = −0.43, 0.28; and SMD = 0.29 95% IC = 0.08, 0.65; 95% CI; , resp.). Secondary clinical outcomes are discussed as follows. The difference of isoflurane and sevoflurane in hospital length of stay was not statistically significant (SMD = 0.06; 95% CI = −0.33, 0.45; ). Neurological dysfunction is one of the complications associated in patients undergoing cardiopulmonary bypass (CPB) and occurs in 50–80% and may persist several months in 20–30% of these patients [13]. S100 is a nonspecific cerebral tissue protein commonly used in clinical research as a biomarker of neurological impairment, in the setting of cardiac surgery, when it crosses the blood-brain barrier through the bloodstream after glial damage [21]. It has been documented that volatile anesthetics would provide neuroprotection through different mechanisms, especially against cerebral ischemic injury [18, 21]. It has been documented that volatile anesthetics would provide neuroprotection, especially against cerebral ischemic injury, through different mechanisms such as inducible nitric oxide synthase (iNOS), mitochondrial channels, and ubiquitin conjugated protein aggregation [18, 32–34]. Difference in levels of S100 using isoflurane or sevoflurane was not statistically significant (SMD = 0.08; 95% CI = −033 to 0.49; and SMD = 0.21; CI 95% = −0.23, 0.65; , at the end of the surgery and 24 hours after the surgery, resp.). Volatile anesthetics have been shown to be cardioprotective from ischemic injury through different mechanisms, similar to neuroprotectives, such as key roles of channels and adenosine A1 receptors, which improved coronary perfusion mediated by increased nitric oxide production [18, 21, 35, 36]. Difference was statistically significant favouring sevoflurane SMD = 2.16; 95% CI = 0.57, 3.74; . However, the number of included RCTs regarding the outcome cardiac troponin-T (cTnT) () is small; the heterogeneity is substantial (%) and a single study was influencing the pooled results [19]. cTnT is part of the contractile apparatus in myofibrils [37] and is used as an indicator of the severity of damage after cardiac surgery [37]. Levels of TnT were also analyzed; the difference between the two inhaled anesthetics was not statistically significant at the end of the surgery (SMD = −1.13; 95% CI = −2.39, 0.13; ) and 24 hours after surgery (SMD = 0.74; 95% CI = −0.15, 1.62; ).
We consider that this study follows a comprehensive retrospective analysis of the included RCTs and exhaustive assessment of the identified primary clinical outcomes. The systematic search in major databases was wide and exhaustive and the results are consistent. Quality analysis of this study did not evidence substantial publication bias. However, this meta-analysis has limitations, listed as follows. High heterogeneity was detected comparing the difference between both anesthetics for primary clinical outcomes: time of extubation and ICU length of stay (%; for heterogeneity < 0.01 and %; for heterogeneity = 0.01, resp.). Substantial heterogeneity was also found in secondary clinical outcomes such as Hospital LOS (%; for heterogeneity = 0.02), CK-MB levels (%; for heterogeneity < 0.01), and troponin levels (%; for heterogeneity = 0.02). In addition, when aiming to estimate effect sizes, frequently a single study with small sample size was influencing the results from the pool of RCTs included. Nonetheless, larger RCTs were suggesting no significantly difference in the overall results. Furthermore, a small number of high-quality RCTs were included for the estimation of clear statistical parameters in clinical outcomes such as plasmatic levels of CK-MB and S100 in perioperative and postoperative instances.
5. Conclusions
The volatile anesthetic choice has no significant impact on postoperative outcomes of patients undergoing cardiac surgery. Other practical considerations (availability, costs, and preference) may be influential factors into the decision regarding which anesthetic to use.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Supplementary Materials
Funnel plots illustrating the publication bias for each outcome evaluated in this meta-analysis.