Clinical Neurology and Neurosurgery 114 (2012) 827–832
Contents lists available at SciVerse ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Review
A systematic review of clinical outcomes, perioperative data and selective adverse events related to mild hypothermia in intracranial aneurysm surgery Zhong-xin Zhao, Cong Wu, Min He ∗ Department of Neurosurgery, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, PR China
a r t i c l e
i n f o
Article history: Received 18 April 2011 Received in revised form 10 May 2012 Accepted 12 May 2012 Available online 30 May 2012 Keywords: Aneurysm Mild hypothermia Subarachnoid haemorrhage Systematic review
a b s t r a c t Background: In the last two decades, mild intraoperative hypothermia has become widely accepted as a protective therapy in neurosurgery. However, its effect in intracranial aneurysm surgery remains unclear. Objective: The purpose of this study was to assess the perioperative effects and selected adverse events associated with intraoperative mild hypothermia in aneurysm surgery and to compare those with events in normothermic surgery. Methods: Three literature databases, namely the Cochrane Library, PubMed and EMBASE, were searched for randomised controlled trials (RCTs) of aneurysm surgery that compared intraoperative mild hypothermia and normothermia from January 1965 to August 2010. Three RCTs were identified. We extracted the following information: author names and publication year; clinical outcome (number of deaths and Glasgow outcome scales); perioperative data (number of moderate or severe intraoperative brain swelling occurrences, hypertensive episodes, ruptured or leaking aneurysms, volume of blood loss during surgery, duration of temporary clipping, and number of patients who received protective drugs, who required rewarming and who were intubated); number of adverse events (cerebral infarctions, brain swelling, myocardial ischaemia or infarction, congestive heart failure, meningitis or ventriculitis and pneumonia). Except for author names and publication year, the data were pooled to perform a mean effect size estimate. The effects of intraoperative mild hypothermia were then analysed. Results: The number of patients requiring rewarming in the mild hypothermia group was significantly greater than in the normothermia group (odds ratio, 33.89; 95% confidence intervals, 3.61–318.36). There were no other statistically significant differences. Conclusion: Based on available RCTs, especially involving surgery of low-grade aneurysms, intraoperative mild hypothermia showed no advantages compared with normothermia. © 2012 Elsevier B.V. All rights reserved.
Contents 1. 2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Eligibility criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Data collection and extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Analysis of the quality of the studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Literature search scheme and characteristics of the eligible references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Analysis of the quality of eligible references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Clinical outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Perioperative data and complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Selected adverse events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
∗ Corresponding author at: No. 37, Guoxue Alley, Chengdu, Sichuan Province, China. Tel.: +86 28 85422488; fax: +86 28 85422490. E-mail addresses:
[email protected],
[email protected] (M. He). 0303-8467/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.clineuro.2012.05.008
828 828 828 828 828 828 828 828 828 829 829 829 830
828
4.
Z.-x. Zhao et al. / Clinical Neurology and Neurosurgery 114 (2012) 827–832
3.6. Publication bias assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Mild hypothermia is generally defined as a body temperature in the range of about 33.0–35.0 ◦ C [1], while traditional hypothermia is considered a body temperature between about 24 ◦ C and 33 ◦ C [2,3]. In surgery, mild hypothermia has been reported to prevent ischaemia-induced OH generation [4], stabilise cell membranes [5], reduce platelet function [6], retard coagulation cascade reactions and control infection [7,8]. Thus, mild hypothermia can have a beneficial effect on cerebral ischaemia and traumatic brain injury (TBI). However, the benefits of using mild intraoperative hypothermia are not clear for intracranial aneurysm surgery. To address this, we systematically searched literature databases for relevant studies published between January 1965 and August 2010. Our aim was to determine whether mild intraoperative hypothermia in intracranial aneurysm surgery had advantages over normothermia in perioperative, intraoperative or postoperative situations. 2. Methods 2.1. Eligibility criteria For a study to be included in our systematic review, it had to meet the following criteria: (1) it had to be a randomised controlled trial (RCT) of aneurysm surgery that compared intraoperative mild hypothermia and normothermia; (2) it had to monitor the oesophageal temperature of all patients; (3) it had to define an intraoperative body temperature of about 33.0–35.0 ◦ C for the mild hypothermia group and one of about 36 ◦ C for the normothermia group. We excluded quasi-randomised studies, animal studies, and studies that used duplicate data. 2.2. Search strategy We searched three literature databases for studies published in English between January 1965 and August 2010: the Cochrane Library, PubMed and EMBASE databases. The search terms were ‘intracranial aneurysm’, ‘subarachnoid hemorrhage’, ‘hypothermia’ or ‘mild hypothermia’, and the terms were searched individually and in combination. 2.3. Data collection and extraction Two reviewers independently screened the titles of all papers identified in the search to eliminate duplicate references, and then they read all of the abstracts to select RCTs according to the inclusion criteria. When there was uncertainty about whether a study should be considered further, a third reviewer looked at the study and made the decision. Finally, the full text or responses from the study authors were assessed by the two reviewers to decide whether the study should be included. The two reviewers extracted the following information from the identified studies: (1) general information: author names and publication year; (2) clinical outcomes: the number of deaths at the end of the trial; Glasgow outcome scales; (3) perioperative data: the number of moderate or severe intraoperative brain swelling events, hypertensive episodes, ruptured or leaking aneurysms, volume of blood loss, duration of temporary clipping, patients who received protective drugs during surgery, patients requiring rewarming and
830 830 830 831 831
intubated patients; (4) selected adverse events: the number of cerebral infarctions, brain swelling events, myocardial ischaemia or infarction, congestive heart failure, meningitis or ventriculitis and pneumonia. 2.4. Analysis of the quality of the studies The quality of the eligible studies was assessed using the Cochrane Collaboration’s tool for assessing risk of bias [9]. Using this tool, we evaluated sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting and other potential sources of bias. 2.5. Statistical analysis For dichotomous data, the odds radio (OR) and 95% confidence intervals (CI) were calculated as the pooled mean effect size estimate using the Mantel–Haenszel method; for continuous data, the standardised mean difference (SMD) and 95% CI were calculated. Based on the heterogeneity of the data in these studies, if the P value of the Cochran’s Q test was above 0.10 and I2 (which shows inconsistency) was below 50%, the fixed effects model was used; otherwise, the random effects model was employed. If the pooled estimate included fewer than ten RCTs, publication bias was not required [10]. Differences in study quality and eligibility criteria (e.g., when a study was eligible for inclusion but had poor quality data, or when a study only included low-grade aneurysms) could generate between-study heterogeneity that would impact the stability of the pooled effect size estimates. Thus, for the sensitivity analysis, the studies were divided into subgroups according to the included grades of aneurysms, such as low-grade aneurysm (World Federation of Neurological Surgeons grading scale ≤ 3 or Hunt and Hess grade ≤ 3), poor-grade aneurysm (World Federation of Neurological Surgeons grading scale ≥ 4 or Hunt and Hess grade ≥ 4) or all grades (both low-grade and poor-grade aneurysms). The subgroup meta-analysis method was used to identify such studies. We used Review Manager 5 (5.0.25; Thomson ResearchSoft, Carlsbad, CA, USA) to calculate the OR value with 95% CI, the SMD with 95% CI and the Cochran’s Q test with I2 . The publication bias was calculated with Stata/MP (v. 11.0, StataCorp; College Station, Texas, USA). 3. Results 3.1. Literature search scheme and characteristics of the eligible references As shown in Fig. 1, we searched the Cochrane Library, PubMed database and EMBASE database and obtained an initial pool of 209 references. Preliminary screening eliminated 46 duplicate references, and perusal of the titles and abstracts eliminated 144 more references that did not meet the eligibility criteria. Examination of the full text or contacting the author eliminated 16 more references, one of which [11] was a preliminary report [11,12]. These screening steps resulted in the identification of 3 studies involving a total of 1161 patients that were eligible for meta-analysis and systematic review [12–14]. The characteristics of the 3 studies are shown in Table 1.
Z.-x. Zhao et al. / Clinical Neurology and Neurosurgery 114 (2012) 827–832
829
Table 1 Characteristics of the three studies.a Authors
Year of publication
No. of patients
Age (years)
Sex (% female)
WFNS score or HH classification
Controlled body temperature (◦ C)
Time of GOS scoring (months)
Hindman et al. [12]
1999 2005
Chouhan et al. [14]
2006
M: 48(43–59)/54(46–63) N: 56(47–67)/52(43–64) M: 52 ± 12 N: 51 ± 13 M: 46.32 ± 3.86 N: 49.09 ± 4.55
M: 54/64 N: 61/86 M: 65 N: 66 M: 46 N: 48
M: 16, 7, 1 N: 20, 6, 2 M: 67, 28, 5 N: 66, 30, 5 M: 1, 8, 8, 5, 2 N: 1, 6, 10, 4, 2
M: 33.5 N: 36.5 M: 33 N: 36.5 M: ∼33.0–34.5 N: ∼36.0–37.5
6
Todd et al. [13]
M: 24/33 N: 28/29 M: 499 N: 501 M: 24 N: 23
3 –b
Hindman et al. and Todd et al. used WFNS scoring [scores of 1, 2, 3 (in that order) and no motor deficit in Todd et al.] and Chouhan et al. used the HH classification [scores of 0, 1, 2, 3, 4 (in that order)]. Hindman et al. just provided WFNS score of the subarachnoid haemorrhage (SAH) group. The study by Hindman et al., reported patient number, age and sex separately for SAH and no SAH groups and reported age as the median plus 25th to 75th percentiles. The two other studies described as the mean and standard deviation. a M, mild hypothermia; N, normothermia; GOS, Glasgow outcome scale; WFNS score, the World Federation of Neurological Surgeons grading scale; HH classification, Hunt and Hess classification. b GOS was assessed at the time of discharge from the hospital.
3.2. Analysis of the quality of eligible references We evaluated the eligibility criteria of the 3 identified studies using the Cochrane Collaboration tool. The quality analysis of all 3 studies is shown in Fig. 2. Although the studies by Hindman et al. and Todd et al. looked mainly at low-grade aneurysms, the studies were high quality. The RCT by Chouhan et al. did not describe sequence generation and allocation concealment clearly. Additionally, the interval between the operation and the Glasgow outcome scale (GOS) scoring was short, which may have influenced the quality of the study. 3.3. Clinical outcomes When we considered patients for whom GOS scoring was performed after surgery, data from 3 RCTs involving a total of 1158 patients could be pooled [12–14]. The pooled OR was 0.98 (95% CI 0.60–1.58), and deaths in the normothermia and mild hypothermia groups were not significantly different. For the low-grade and
Fig. 1. Literature search scheme. Three databases, namely the Cochrane Library and the PubMed and EMBASE databases, were searched for randomised controlled trials (RCTs) of aneurysm surgery that compared intraoperative mild hypothermia and normothermia from January 1965 to August 2010.
all-grades subgroups, the ORs were 0.95 (95% CI 0.58–1.55) and 2.00 (0.17–23.70). Thus, the postoperative deaths were the same for the mild hypothermia and normothermia groups. GOS scores can reflect neurological outcome. For the 1158 patients in the RCTs [12–14], we looked specifically at patients with “good GOS” scores i.e. no or minor disability or moderate disability. The pooled OR was 1.26 (95% CI 0.91–1.74). For the subgroups, low grade and all grades did not have significantly different OR values: 1.27 (95% CI 0.91–1.77) and 1.05 (95% CI 0.23–4.82), respectively. For patients undergoing surgery, there was no difference the GOS scores of the mild hypothermia and normothermia groups. Thus, there were no differences in the results for the subgroups versus the entire patient population as reflected by the similar OR values. 3.4. Perioperative data and complications To compare perioperative and intraoperative complications, we pooled the relevant patients from the three studies. To study intraoperative brain swelling, 2 RCTs with 1114 patients were analysed [12,13]. The pooled OR for this group of 1114 was 0.98 (95%
Fig. 2. Risk of bias summary of the eligible references. “+” indicates low risk; “?” indicates unclear risk and “−” indicates high risk. For the study by Chouhan et al., the allocation sequence was generated without description, the method of concealment was not described, and the time of Glasgow outcome scoring was short, possibly leading to a risk of potential bias.
830
Z.-x. Zhao et al. / Clinical Neurology and Neurosurgery 114 (2012) 827–832
CI 0.64–1.51), indicating that mild hypothermia did not reduce intraoperative brain swelling compared to normothermia. Data regarding hypertensive episodes or anti-hypertensive drug use during surgery were available for all 3 RCTs [12–14], and the overall OR was 0.82 (95% CI 0.44–1.51). For the low-grade group, the OR was 0.80 (95% CI 0.41–1.55), and for the all grades group the OR was 0.95 (95% CI 0.17–5.28). This shows that intraoperative mild hypothermia did not help control hypertension during surgery. Data regarding the number of ruptured or leaking aneurysms during surgery were available for 2 RCTs involving a total of 1047 patients [13,14], and the pooled OR was 0.82 (95% CI 0.50–1.36). The OR values of the subgroups were 0.85 (95% CI 0.51–1.41) and 0.46 (95% CI 0.04–5.41), indicating that mild hypothermia did not reduce the occurrence of ruptured or leaking aneurysms during surgery. To study intraoperative blood loss, 1047 patients from 2 RCTs were analysed [13,14]. The overall SMD was −0.16 (95% CI −0.64 to 0.33), and the SMD values of the subgroups were 0.03 (95% CI −0.10 to 0.15) and −0.49 (95% CI −1.07 to 0.09). There was no significant difference in intraoperative blood loss between the mild hypothermia and normothermia groups. The total duration of temporary clipping was analysed for 1047 patients in 2 RCTs [13,14]. The overall SMD was −0.12 (95% CI −0.67 to 0.43), and the SMD values for the subgroups were 0.09 (95% CI −0.03 to 0.22) and −0.48 (95% CI −1.06 to 0.10). This showed that the total duration of temporary clipping was not shorter for the mild hypothermia group compared to the normothermia group. We also looked at the use of neuroprotective drugs during surgery by analysing 1114 patients from 2 RCTs [12,13]. The OR for the pooled patients was 0.80 (95% CI 0.46–1.37), showing that mild hypothermia during surgery was not beneficial for cerebral ischaemia. After surgery, data regarding the number of patients requiring rewarming were available for 2 RCTs (161 patients) [12,14]. The pooled OR was 33.89 (95% CI 3.61–318.36), and the OR values of the two subgroups were 15.37 (95% CI 5.33–44.34) and 166.64 (95% CI 8.66–3205.02). Thus, more patients in the mild hypothermia group needed rewarming. Data regarding postoperative intubation were available for 161 patients in 2 RCTs [12,14]. The pooled OR value was 0.52 (95% CI 0.26–1.01). The OR values of the low-grade and all-grade groups were 0.54 (95% CI 0.25–1.17) and 0.46 (95% CI 0.13–1.68). Therefore, mild hypothermia did not reduce the number of patients needing intubation during recovery. Thus, aside from the number of patients needing rewarming postoperatively, we found no differences in perioperative, intraoperative and postoperative events in the mild hypothermia group compared to the normothermia group. 3.5. Selected adverse events The selected adverse events we analysed were mainly postoperative complications but also included some events that occurred between randomisation and the final follow-up evaluation. There was no significant difference in the number of neurological events (cerebral infarctions) in the mild hypothermia versus the normothermia groups in the 2 RCTs that reported these data, with an overall OR value of 0.84 (95% CI 0.49–1.42) [13,14] and OR values of 0.86 (95% CI 0.50–1.48) and 0.46 (95% CI 0.04–5.41) for the two groups. For brain swelling, assessed for in patients in 2 RCTs, the overall OR was 0.81 (95% CI 0.48–1.37) [12,13]. There were no significant differences in the subgroups, showing that intraoperative mild hypothermia does not benefit patients in terms of reducing cerebral infarction and swelling. To look at the influence of intraoperative mild hypothermia on the heart, 1112 patients in 2 RCTs were analysed [12,13]. The OR for myocardial ischaemia or infarction in this pooled population was 1.55 (95% CI 0.43–5.67), and the OR for congestive heart failure was 0.92 (95% CI 0.41–2.08). These values show that mild hypothermia
in aneurysm surgery does not improve heart function compared to normothermia. For intracranial infection, 1048 patients in 2 RCTs were available for the pooled analysis [13,14]. The pooled OR was 1.51 (95% CI 0.53–4.34), and the OR values of the low grade group and the all grade group were 1.52 (95% CI 0.42–5.40) and 1.50 (95% CI 0.23–9.92). Thus, intraoperative mild hypothermia in aneurysm surgery does not control intracranial infection better than normothermia. Data about pneumonia were available for 1112 patients in 2 RCTs [12,13]. The pooled OR was 0.80 (95% CI 0.31–2.05). This indicated that mild hypothermia does not reduce infections such as pneumonia. 3.6. Publication bias assessment The studies included in our research had fewer than ten RCTs, so we did not assess publication bias. 3.7. Sensitivity analysis To assess the sensitivity of each meta-analysis, we compared the overall pooled mean effect size estimate with the subgroup results. No differences were found. Thus, the results of this analysis were stabile and dependable. 4. Discussion Previous observational studies have confirmed the protective effects of mild hypothermia in a animal brain injury model [15], and mild hypothermia has been found to improve patient prognosis by reducing traumatic intracranial hypertension [16]. Notably, intraoperative mild hypothermia is generally considered efficacious in aneurysm surgery for neuroprotection and for relieving cerebral vasospasm [12,17–21], but the role of intracranial mild hypothermia in aneurysm surgery is controversial. In the RCTs conducted by Hindman et al. and Chouhan et al., mild hypothermia was considered beneficial during surgery [12,14], and lower body temperature led to better outcomes in acute stroke patients [22]. However, in the large RCT conducted by Todd et al., there was no advantage to hypothermia. Indeed, the outcome of severe subarachnoid haemorrhage was not improved by hypothermia treatment [13,23]. This discrepancy in results prompted us to analyse the perioperative and prognostic effects of mild hypothermia in aneurysm surgery and to compare the effects with those in intraoperative normothermia conditions. Mild hypothermia can reduce ischaemic neuronal injury and can therefore lessen neurological deficits due to aneurysmal subarachnoid haemorrhage in the acute phase [24–26]. In terms of cerebral blood flow, Karibe reported that intraoperative mild hypothermia may reduce intraoperative vessel occlusion and cerebral retraction and should thus improve the prognosis of Hunt and Hess Grade II or III patients who undergo aneurysm surgery [18]. However, in our analysis mild hypothermia did not reduce postoperative mortality and did not improve GOS scores, which reflect neurological outcome. There is a similar contradiction for TBI. In the RCT conducted by Marion et al., mild hypothermia was reported to have a beneficial effect on the prognosis of patients with Glasgow Coma Scale (GCS) scores of ∼5–7 [27], but was considered ineffective in patients with GCS scores of ∼3–4. Further researches by Shiozaki et al. confirmed that mild hypothermia was effective only for TBI involving increases in intracranial pressure (ICP) [28–30]. This could be explained by the relationship between lower ICP and better GOS scores [31], and by the fact that mild hypothermia can reduce intracranial hypertension in a rat model [32]. Therefore, further studies are needed that use subgroups that are defined by Hunt and Hess grades or by preoperative ICP values.
Z.-x. Zhao et al. / Clinical Neurology and Neurosurgery 114 (2012) 827–832
Generally, intraoperative brain swelling indicates higher ICP due to events such as brain oedema or a leaking or ruptured aneurysm with bleeding. Because hypothermia has been confirmed to reduce oedema and ICP, it should also decrease the occurrence of intraoperative brain swelling [28–30,32,33]. However, in our analysis we did not find that intraoperative brain swelling was reduced in the mild hypothermia groups or that intraoperative mild hypothermia reduced leaking or ruptured aneurysms. Chi et al. reported that intraoperative mild hypothermia does not affect blood pressure [34], which may explain our findings. In addition, our meta-analysis found no differences between the mild hypothermia and normothermia groups in terms of intraoperative blood loss or the duration of temporary clipping. For surgical patients, we found that mild hypothermia did not reduce the intraoperative use of protective drugs or the number of patients requiring intubation. Only the number of patients requiring rewarming was greater in the intraoperative mild hypothermia group than in the normothermia group, which is to be expected and is acceptable. In spite of the need for rewarming, the operation time was not significantly different for the two groups, which was also noted in the RCT conducted by Todd et al. [13] Taken together, our findings indicate that intraoperative mild hypothermia has no obvious benefit for aneurysm surgery. In terms of surgical complications, problems associated with traditional intraoperative hypothermia (24–33◦ ), such as intracranial or cardiac complications and infection, are well established [2,3]. Accordingly, we pooled the mean effect size estimates for aneurysm surgery to look at the complications mentioned above in the mild hypothermia patient group. In rat models, mild hypothermia regulates the blood–brain barrier and changes the cellular structure of the brain [35,36]. It also reduces brain oedema 48 h after intracerebral haemorrhage [37]. However, we found that intraoperative hypothermia had no effect on cerebral infarction and brain swelling. For cardiac complications, a recent study found that mild hypothermia reduced coronary t-PA release after myocardial ischaemia [38], improved ventricular compliance after myocardial infarction and improved myocardial perfusion in a cardiogenic shock pig model [39,40]. However, intraoperative mild hypothermia did not show any benefits for myocardial ischaemia or infarction or for congestive heart failure compared to normothermia surgery. Regarding infections, research by L’Her et al. indicated that mild hypothermia was a protective factor against infection in animal experiments [41], and Jiang et al. concluded that longer-term mild hypothermia had beneficial effects [42]. Moreover, muscle relaxants and anaesthetic agents used to prevent shivering and calorigenesis were important factors in infection [43]. Because the anaesthesia and the duration of surgery were the same in the two groups [13], we can conclude that there were no differences in the infection rate in the mild hypothermia and normothermia groups that we analysed. The mechanism responsible for any benefits of intraoperative mild hypothermia remains unclear. The most important use of mild hypothermia has been to reduce ICP and improve patient prognosis, as shown in some studies [16,27–32]. Although we found no significant benefits to mild hypothermia versus normothermia when comparing groups of patients, intraoperative mild hypothermia may be useful for some patients with high ICP. The number of published studies that were analysed in our study was limited, and they focused mainly on surgeries involving low-grade aneurysms. However, we found that there was no clear advantage to using intraoperative hypothermia in aneurysm surgery. Funding None.
831
References [1] Bilotta F, Pietropaoli P, La Rosa I, Spinelli F, Rosa G. Effects of shivering prevention on haemodynamic and metabolic demands in hypothermic postoperative neurosurgical patients. Anaesthesia 2001;56:514–9. [2] Ginsberg M, Sternau L, Globus M, Dietrich W, Busto R. Therapeutic modulation of brain temperature: relevance to ischemic brain injury. Cerebrovascular and Brain Metabolism Reviews 1992;4:189–225. [3] Maher J, Hachinski V. Hypothermia as a potential treatment for cerebral ischemia. Cerebrovascular and Brain Metabolism Reviews 1993;5:277–300. [4] Van Hemelrijck A, Hachimi-Idrissi S, Sarre S, Ebinger G, Michotte Y. Postischaemic mild hypothermia inhibits apoptosis in the penumbral region by reducing neuronal nitric oxide synthase activity and thereby preventing endothelin-1-induced hydroxyl radical formation. European Journal of Neuroscience 2005;22:1327–37. [5] Pestel GJ, Kurz A. Hypothermia – it’s more than a toy. Current Opinion in Anaesthesiology 2005;18:151–6. [6] Michelson A, MacGregor H, Barnard M, Kestin A, Rohrer M, Valeri C. Reversible inhibition of human platelet activation by hypothermia in vivo and in vitro. Thrombosis and Haemostasis 1994;71:633–40. [7] Reed 2nd RL, Johnson JT, Hudson JD, Fischer RP. The disparity between hypothermic coagulopathy and clotting studies. Journal of Trauma 1992;33:465–70. [8] Flores-Maldonado A, Medina-Escobedo CE, Rios-Rodriguez HM, FernandezDominguez R. Mild perioperative hypothermia and the risk of wound infection. Archives of Medical Research 2001;32:227–31. [9] Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions, version 5.0.2 [updated September 2009]. The Cochrane Collaboration; 2009. http://www.cochrane-handbook.org. [10] Song F, Eastwood AJ, Gilbody S, Duley L, Sutton AJ. Publication and related biases. Health Technology Assessment 2000;4:1–115. [11] Todd MM, Carras D, Hindman B, Loftus C, Gelb A, Craen R, et al. A randomized prospective pilot trial of mild hypothermia durintracranial aneurysm surgery. Anesthesiology 1996;85(3a), ing http://www.mrw.interscience.wiley.com/cochrane/clcentral/articles/978/CN00318978/frame.html. [12] Hindman BJ, Todd MM, Gelb AW, Loftus C, Craen R, Schubert A, et al. Mild hypothermia as a protective therapy during intracranial aneurysm surgery: a randomized prospective pilot trial. Neurosurgery 1999;44:23–32. [13] Todd MM, Hindman BJ, Clarke WR, Torner JC. Mild intraoperative hypothermia during surgery for intracranial aneurysm. New England Journal of Medicine 2005;352:135–45. [14] Chouhan RS, Dash HH, Bithal PK, Chaturvedi A, Pandia MP, Radhakrishnan M, et al. Intraoperative mild hypothermia for brain protection during intracranial aneurysm surgery. Journal of Anaesthesiology Clinical Pharmacology 2006;22:21–8. [15] Dietrich WD, Atkins CM, Bramlett HM. Protection in animal models of brain and spinal cord injury with mild to moderate hypothermia. Journal of Neurotrauma 2009;26:301–12. [16] Shiozaki T, Sugimoto H, Taneda M, Yoshida H, Iwai A, Yoshioka T, et al. Effect of mild hypothermia on uncontrollable intracranial hypertension after severe head injury. Journal of Neurosurgery 1993;79:363–8. [17] Baker KZ, Young WL, Stone JG, Kader A, Baker CJ, Solomon RA. Deliberate mild intraoperative hypothermia for craniotomy. Anesthesiology 1994;81: 361–7. [18] Karibe H, Sato K, Shimizu H, Tominaga T, Koshu K, Yoshimoto T. Intraoperative mild hypothermia ameliorates postoperative cerebral blood flow impairment in patients with aneurysmal subarachnoid hemorrhage. Neurosurgery 2000;47:594–9, discussion 599–601. [19] Murthy HS, Chidanandaswamy MN, Rao GS, Kolluri S. Spontaneous intraoperative hypothermia and cerebral protection in aneurysmal subarachnoid hemorrhage. Middle East Journal of Anesthesiology 2005;18:313–18332. [20] Nagao S, Irie K, Kawai N, Kunishio K, Ogawa T, Nakamura T, et al. Protective effect of mild hypothermia on symptomatic vasospasm: a preliminary report. Acta Neurochirurgica Supplementum 2000;76:547–50. [21] Seule MA, Muroi C, Mink S, Yonekawa Y, Keller E. Therapeutic hypothermia in patients with aneurysmal subarachnoid hemorrhage, refractory intracranial hypertension, or cerebral vasospasm. Neurosurgery 2009;64:86–92, discussion 92–83. [22] Reith J, Jorgensen HS, Pedersen PM, Nakayama H, Raaschou HO, Jeppesen LL, et al. Body temperature in acute stroke: relation to stroke severity, infarct size, mortality, and outcome. Lancet 1996;347:422–5. [23] Anei R, Sakai H, Iihara K, Nagata I. Effectiveness of brain hypothermia treatment in patients with severe subarachnoid hemorrhage: comparisons at a single facility. Neurologia Medico-Chirurgica 2010;50:879–83. [24] Zhang H, Xu G, Zhang J, Murong S, Mei Y, Tong E. Mild hypothermia reduces ischemic neuron death via altering the expression of p53 and bcl-2. Neurological Research 2010;32:384–9. [25] Li H, Wang D. Mild hypothermia improves ischemic brain function via attenuating neuronal apoptosis. Brain Research 2011;1368:59–64. [26] Thome C, Schubert GA, Schilling L. Hypothermia as a neuroprotective strategy in subarachnoid hemorrhage: a pathophysiological review focusing on the acute phase. Neurological Research 2005;27:229–37. [27] Marion DW, Penrod LE, Kelsey SF, Obrist WD, Kochanek PM, Palmer AM, et al. Treatment of traumatic brain injury with moderate hypothermia. New England Journal of Medicine 1997;336:540–6.
832
Z.-x. Zhao et al. / Clinical Neurology and Neurosurgery 114 (2012) 827–832
[28] Shiozaki T, Sugimoto H, Taneda M, Oda J, Tanaka H, Hiraide A, et al. Selection of severely head injured patients for mild hypothermia therapy. Journal of Neurosurgery 1998;89:206–11. [29] Shiozaki T, Kato A, Taneda M, Hayakata T, Hashiguchi N, Tanaka H, et al. Little benefit from mild hypothermia therapy for severely head injured patients with low intracranial pressure. Journal of Neurosurgery 1999;91:185–91. [30] Shiozaki T, Hayakata T, Taneda M, Nakajima Y, Hashiguchi N, Fujimi S, et al. A multicenter prospective randomized controlled trial of the efficacy of mild hypothermia for severely head injured patients with low intracranial pressure. Mild Hypothermia Study Group in Japan. Journal of Neurosurgery 2001;94:50–4. [31] Mrlian A, Smrcka M, Klabusay M. The use of controlled mild hypothermia and immune system status in patients with severe brain injury. Bratislavske Lekarske Listy 2006;107:113–7. [32] Torok E, Klopotowski M, Trabold R, Thal SC, Plesnila N, Scholler K. Mild hypothermia (33 ◦ C) reduces intracranial hypertension and improves functional outcome after subarachnoid hemorrhage in rats. Neurosurgery 2009;65:352–9, discussion 359. [33] Kollmar R, Staykov D, Dorfler A, Schellinger PD, Schwab S, Bardutzky J. Hypothermia reduces perihemorrhagic edema after intracerebral hemorrhage. Stroke 2010;41(8):1684–9. [34] Chi OZ, Choi YK, Lee DI, Kim YS, Lee I. Intraoperative mild hypothermia does not increase the plasma concentration of stress hormones during neurosurgery. Canadian Journal of Anaesthesia 2001;48:815–8. [35] Kiyatkin EA, Sharma HS. Permeability of the blood–brain barrier depends on brain temperature. Neuroscience 2009;161:926–39.
[36] Chi OZ, Liu X, Weiss HR. Effects of mild hypothermia on blood–brain barrier disruption during isoflurane or pentobarbital anesthesia. Anesthesiology 2001;95:933–8. [37] Kawanishi M, Kawai N, Nakamura T, Luo C, Tamiya T, Nagao S. Effect of delayed mild brain hypothermia on edema formation after intracerebral hemorrhage in rats. Journal of Stroke and Cerebrovascular Diseases 2008;17: 187–95. [38] van der Pals J, Gotberg M, Olivecrona GK, Brogren H, Jern S, Erlinge D. Mild hypothermia markedly reduces ischemia related coronary t-PA release. Journal of Thrombosis and Thrombolysis 2010;29:289–95. [39] Hamamoto H, Sakamoto H, Leshnower BG, Parish LM, Kanemoto S, Hinmon R, et al. Very mild hypothermia during ischemia and reperfusion improves postinfarction ventricular remodeling. Annals of Thoracic Surgery 2009;87: 172–7. [40] Gotberg M, van der Pals J, Olivecrona GK, Koul S, Erlinge D. Mild hypothermia reduces acute mortality and improves hemodynamic outcome in a cardiogenic shock pig model. Resuscitation 2010;81:1190–6. [41] L’Her E, Amerand A, Vettier A, Sebert P. Effects of mild induced hypothermia during experimental sepsis. Critical Care Medicine 2006;34:2621–3. [42] Jiang JY, Xu W, Li WP, Gao GY, Bao YH, Liang YM, et al. Effect of long-term mild hypothermia or short-term mild hypothermia on outcome of patients with severe traumatic brain injury. Journal of Cerebral Blood Flow and Metabolism 2006;26:771–6. [43] Schwab S, Schwarz S, Spranger M, Keller E, Bertram M, Hacke W. Moderate hypothermia in the treatment of patients with severe middle cerebral artery infarction. Stroke 1998;29:2461–6.