Keywords
Nimodipine - Subarachnoid Hemorrhage - Meta-Analysis - Treatment Outcome
Palavras-chave
Nimodipina - Hemorragia Subaracnóidea - Metanálise - Resultado do Tratamento
INTRODUCTION
Subarachnoid hemorrhage (SAH) refers to a clinical syndrome caused by sudden rupture
and hemorrhage of blood vessels at the base or surface of the brain, due to various
causes, and the subsequent direct blood flow to the subarachnoid membrane.[1] It is a cerebrovascular disease with rapid onset and frequent recurrence, accompanied
by cerebral vasospasm (CVS) and other high-risk complications.[2] Clinical studies have shown that intracranial aneurysm rupture, cerebrovascular
malformations, and vascular abnormalities at skull base can induce this disease, especially
intracranial aneurysm rupture.[3]
Clinical studies have shown that nimodipine can effectively prevent and treat CVS
after SAH, and therefore reduce mortality without increasing the risk of rebleeding.[4] Nimodipine has high lipid solubility and can smoothly pass through the blood-brain
barrier into the nervous system.[5]
[6]
[7] This drug, as a calcium channel blocker, reduces Ca2 + influx in brain cells and
free radical formation, and encourages the vascular smooth muscle to become relaxed,
thereby reducing vasospasm.[8]
[9] Additionally, nimodipine also has the effects on anti-free radical injury and antagonistic
effect on endothelin neurotoxicity, thus improving the tolerance of nerve cells to
ischemia and hypoxia, and improving neurological functions, reducing cerebral ischemia-caused
death and global cerebral infarction after SAH, and, consequently, effectively improving
prognosis.[10]
[11] Several meta-analyses have evaluated the efficacy of nimodipine in the treatment
of SAH. Vergouwen et al.[12] found no effect of nimodipine on the prognosis of patients with traumatic SAH by
including five studies in their analysis. Liu et al.,[13] in 2011, found by meta-analysis results that nimodipine significantly reduced CVS
and delayed neurological deficits, as well as cerebral infarction, compared with placebo.
However, in recent years there have been fewer studies on the comprehensive evaluation
of the clinical efficacy of nimodipine in the treatment of SAH. Therefore, a critical
systematic review would be very beneficial for clinicians. In this study, a meta-analysis
was used to comprehensively evaluate the efficacy and safety of nimodipine in the
treatment of SAH using a large sample size.
METHODS
Literature search strategy
The electronic databases including China National Knowledge Infrastructure (CNKI),
VIP, SinoMed, China Master's Theses Full-text Database (CMFD), China Doctoral Dissertations
Full-text Database (CDFD), Cochrane Library, PubMed and Embase were used to search
for the efficacy of nimodipine in the treatment of SAH from 2010 to 2021. The keywords
using the following terms: (nimodipine) AND (subarachnoid hemorrhage OR SAH) AND (clinical effect). The Chinese database was also searched using the above search terms in Chinese.
Additionally, a manual search of references to relevant journals and retrieved articles
was conducted by reviewing titles and abstracts. There were no language restrictions
in the literature search process.
Inclusion criteria
1) Types of studies: randomized controlled trials (RCTs). 2) Study subjects: patients
diagnosed with SAH by clinical examination. 3) Types of intervention: the patients
were divided into control group and treatment group according to different treatment
methods. The control group mainly received routine treatment including oxygen inhalation,
hemostasis, lowering intracranial pressure, maintaining stable blood pressure, symptomatic
treatments (such as analgesia, sedation, preventing infection, correcting metabolic
disorders), and bed rest. The treatment group was treated with continuous administration
of nimodipine adjuvant therapy by intravenous pump on the basis of routine treatment.
The treatment lasted for 2 weeks in both groups. 4) Outcome measures: effective rate,
incidence of adverse reactions, comparison of the middle cerebral artery (MCA), blood
flow velocity (BFV), and the Glasgow coma scale (GCS) score before and after treatment
in both groups. Adverse reactions mainly include hemorrhage, CVS, cerebral infarction,
hydrocephalus, death, decreased blood pressure, and intracranial infection.
Exclusion criteria
Literature that met any of the following criteria was excluded: studies in which the
data required by this meta-analysis were not provided and could not be obtained; the
original text could not be obtained; studies with missing data; duplicates; case reports,
systematic reviews, opinion articles, or studies with animal experiments.
Literature screening and quality evaluation
Two reviewers independently initially screened studies by reading titles and abstracts.
Then, the full text of randomized controlled trials was required to be read to determine
whether it met the inclusion criteria, followed by cross-checking. The final included
studies were jointly decided by two reviewers. During this process, the disagreement
was resolved by discussion between the two or by a third party's decision. For the
repeated or extended reports, the recently published ones with complete data were
selected.
The literature quality was evaluated using the Newcastle-Ottawa scale (NOS).[14]
Data extraction
Reference Aid for Medicine v3.0 and Endnote X5 (Clarivate Analytics, London, UK) were
used to manage and extract the data including: ① basic information: the first author,
publication time, number of included population; ② baseline information of patients:
age, diagnosis, etc.; ③ intervention measures: the total number of patients in the
control and treatment groups, the number of males and females, and the treatment method;
④ study results: the number of CVS, rebleeding, and deaths; ⑤ study design type: clinical
randomized controlled trial.
Statistical analysis
The Stata 16.0 (StataCorp LLC., College Station, TX, US) software was used for statistical
analysis of the collected data. The included enumeration data (binary variables),
such as effective rate, incidence of CVS, and incidence of adverse reactions were
presented and analyzed using an odds ratio (OR) and 95% confidence interval (CI),
while measurement data (continuous variables) were analyzed using mean difference
(MD) and 95% CI. The heterogeneity analysis of recruited studies was performed by
I2 statistic. The random-effect model (REM) was adopted for meta-analysis if significant
heterogeneity was assessed (p < 0.10 and I2 > 50%); Otherwise, the fixed-effect model (FEM) was employed. A statistically significant
difference was indicated if p < 0.05.
RESULTS
Results of literature screening and quality evaluation
According to the search strategy, a total of 611 studies were retrieved. Of these,
531 irrelevant studies were excluded after evaluating titles and abstracts. Then,
34 duplicate articles were excluded, and 13 were eliminated after reading the full
text. Finally, 10 studies[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24] were included for meta-analysis. The literature screening process is shown in [Figure 1].
Figure 1 Literature screening process.
The basic characteristics of the included studies are shown in [Table 1]. All studies were RCTs, and the subjects were patients with traumatic SAH in 9 studies,
and non-traumatic SAH in 1 study.[22] A total of 882 patients were included in the study, including 439 in the control
group and 443 in the treatment group. There were no significant differences between
the treatment group and the control group in terms of patient age, and male to female
ratio. All literature had NOS scores greater than 6, which were of high quality and
could be included in the meta-analysis. Finally, the risk of bias graph and risk of
bias summary are observed in [Figure 2A-B].
Table 1
The basic characteristics of inclusion in the literature
|
Study
|
Year
|
Sample time
|
Cases Treat/Con
|
Age (years)
|
Sex ratio (M/FM)
|
Study design
|
Treatment time (day)
|
NOS score
|
Outcome measures
|
|
Treat
|
Con
|
Treat
|
Con
|
|
Liu Y. et al.[15]
|
2014
|
2009.10∼2012.09
|
31/31
|
38 ± 12
|
37 ± 14
|
18/13
|
15/16
|
RCT
|
28
|
7
|
①②③④
|
|
Zheng SB et al.[16]
|
2017
|
2015.06∼2017.01
|
60/60
|
48.3 ± 6.5
|
49.3 ± 6.7
|
40/20
|
38/22
|
RCT
|
14
|
6
|
①②③④⑤⑥
|
|
Ma LJ. et al.[17]
|
2019
|
2017.01∼2018.12
|
40/40
|
54.3 ± 8.3
|
53.1 ± 8.6
|
23/17
|
25/15
|
RCT
|
14
|
6
|
①②③④⑤⑥
|
|
Tian Y. et al.[18]
|
2014
|
2009.01∼2012.01
|
31/31
|
43.8 ± 3.2
|
33.2 ± 2.8
|
20/11
|
18/13
|
RCT
|
14
|
6
|
①②④
|
|
Cao YB. et al.[19]
|
2014
|
2013.01∼2013.12
|
32/28
|
34.1 ± 8.4
|
33.8 ± 7.9
|
19/13
|
16/12
|
RCT
|
21
|
6
|
①②④
|
|
Zuo MX.[24]
|
2017
|
2015.06∼2017.01
|
45/45
|
36.5 ± 2.3
|
36.2 ± 2.1
|
20/25
|
22/23
|
RCT
|
21
|
7
|
①②③④⑤⑥
|
|
Zhu ZY.[20]
|
2019
|
2016.01∼2018.12
|
38/38
|
45.6 ± 7.6
|
44.2 ± 1.8
|
15/23
|
16/22
|
RCT
|
28
|
7
|
①③④⑤⑥
|
|
Wang Y., Zhan JN.[21]
|
2011
|
2008.03∼2010.02
|
46/46
|
47.1 ± 8.9
|
46.5 ± 9.2
|
26/20
|
25/21
|
RCT
|
21
|
7
|
③④⑤⑥
|
|
Dong XF. et al.[22]
|
2019
|
2014.04∼2017.04
|
80/80
|
46.4 ± 4.6
|
47.2 ± 4.3
|
45/35
|
42/38
|
RCT
|
56
|
7
|
①②③④
|
|
Cao YS. et al.[23]
|
2011
|
2008.08∼2009.12
|
40/40
|
49.7 ± 2.4
|
50.2 ± 2.5
|
31/19
|
32/18
|
RCT
|
14
|
6
|
①②④
|
Abbreviations: Treat, treatment group; Con, control group; M, male; FM, female; RCT, randomized
controlled trial; NOS, Newcastle-Ottawa scale. Notes: ①: effective rate of treatment; ②: adverse effects rate; ③: blood flow velocity of
middle cerebral artery before treatment; ④: blood flow velocity of middle cerebral
artery after treatment; ⑤: Glasgow coma scale (GCS) score before treatment; ⑥: GCS
score after treatment.
Figure 2 The risk of bias graph (A) and bias summary (B).
Results of meta-analysis
Effective Rate and Incidence of Adverse Reactions
A total of 9 articles[15]
[16]
[17]
[18]
[19]
[20]
[22]
[23]
[24] reported the effective rate of treatment. No heterogeneity was identified among
these studies (I2 = 0.0%, p = 0.949), so FEM was utilized to pool the effect size. The results showed that the
effective rate in the treatment group (n = 397) was significantly higher than that in the control group (n = 393) (OR = 3.21, 95% CI: 2.25, 4.58; p < 0.001) ([Figure 3A]).
Figure 3 Forest plots of effective rate (A) and incidence of adverse reactions (B).
Additionally, 8 articles[15]
[16]
[17]
[18]
[19]
[22]
[23]
[24] reported the incidence of adverse reactions. Marked heterogeneity was identified
in these studies (I2 = 65.7%, p = 0.005), so REM was utilized to pool the effect size. The results showed that the
incidence of adverse reactions after treatment in the treatment group (n = 359) was significantly lower than that in the control group (n = 355) (OR = 0.35, 95% CI: 0.19, 0.67; p = 0.001) ([Figure 3B]).
MCA blood flow velocity before and after treatment
There were 7 articles[15]
[16]
[17]
[20]
[21]
[22]
[24] which reported MCA blood flow velocity before treatment. No heterogeneity was identified
among these studies (I2 = 0.0%, p = 0.961), so FEM was adopted for pooling the effect size. The results revealed no
significant difference in MCA blood flow velocity before treatment between the two
groups (SMD = −0.01, 95% CI: −0.17, 0.14; p = 0.85) ([Figure 4A]), indicating the comparability of the experiments.
Figure 4 Forest plots of middle cerebral artery (MCA) blood flow velocity of two groups before
(A) and after (B) treatment; forest plots of the Glasgow coma scale (GCS) score of the two groups
before (C) and after (D) treatment.
Furthermore, the heterogeneity (I2 = 96.8%, p < 0.001) was identified in the 10 articles[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24] reporting MCA after treatment, so REM was used to pool the effect sizes. The result
showed that MCA blood flow velocity after treatment in the treatment group (n = 443) was significantly lower than that of the control group (n = 439) (SMD = −1.36, 95% CI: −2.28, −0.49; p = 0.002) ([Figure 4B]).
GCS score before and after treatment
Only half, 5, of the selected articles[16]
[17]
[20]
[21]
[24] reported GCS score before treatment. No heterogeneity was identified among these
studies (I2 = 0.0%, p = 0.462), so FEM was adopted for pooling the effect size. The results revealed no
significant difference in GCS score before treatment between the two groups (SMD = −0.06,
95% CI: −0.25, 0.12; p = 0.504) ([Figure 4C]).
Further, marked heterogeneity (I2 = 90.3%, p < 0.001) was identified in the 10 articles[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24] reporting GCS score after treatment, so REM was used to pool the effect sizes. The
result showed that GCS score after treatment in the treatment group (n = 229) was significantly higher than that of the control group (n = 229) (SMD = 1.24, 95% CI: 0.58, 1.89; p < 0.001) ([Fig. 4D]).
Sensitivity analysis
Due to the heterogeneity of included studies regarding effective rate, incidence of
adverse reactions, as well as MCA blood flow velocity and GCS after treatment, sensitivity
analysis was required. After eliminating each literature one by one, it was found
that the overall heterogeneity did not change significantly ([Figure 5A-F]), indicating stable and reliable results of this study.
Figure 5 Sensitivity analysis of effective rate (A) and incidence of adverse reactions (B); sensitivity analysis of middle cerebral artery (MCA) blood flow velocity of two
groups before (C) and after (D) treatment; sensitivity analysis of Glasgow coma scale (GCS) score of the two groups
before (E) and after (F) treatment.
DISCUSSION
As a clinical syndrome, SAH occurs in a critical condition, and is prone to rebleeding
and CVS, threatening the life of patients. Studies have shown that there is a correlation
between adverse reactions caused by SAH and structural and functional changes in the
vascular wall.[25] Nimodipine has a definite neuroprotective effect, such as antioxidant effect, which
improves cerebral metabolic rate of oxygen and reduces brain injuries due to calcium
overload during cerebral blood flow reperfusion.[26] Its neuroprotective effect can also relieve brain edema and glial cell swelling
after SAH, while effectively reducing the risk of death secondary to CVS.[27]
[28] In this study, we systematically provide data on the clinical efficacy of nimodipine
in the treatment of SAH. After a comprehensive literature search and evaluation, a
total of 10 studies (882 patients with SAH) were included for meta-analysis. The results
showed that patients in the nimodipine group had significantly higher treatment efficiency
and a significantly lower incidence of adverse effects (including hemorrhage, CVS,
cerebral infarction, hydrocephalus, death, decreased blood pressure, and intracranial
infection) compared with the control group. This result is consistent with the findings
of others. Hockel et al.[29] showed that continuous intra-arterial treatment with nimodipine prevented secondary
cerebral ischemia in patients with prolonged severe macrovascular spasm.
Additionally, our study compared the MCA blood flow velocity and found that BFV in
MCA was significantly lower in the nimodipine group after treatment. Sun et al.[30] showed that the use of electroacupuncture for CVS in patients with SAH improved
CVS by significantly reducing BFV in MCA. These studies suggest that nimodipine may
improve CVS by reducing BFV in MCA. The GCS is the most widely used scoring system
for level of consciousness.[31] Likewise, the score is also widely used by neurosurgeons for the initial assessment
of patients with SAH, with higher scores indicating a lower level of consciousness
impairment.[32] Studies by Zheng et al.[16] and Ma et al.[17] then found that the GCS score was significantly higher in the nimodipine combination
therapy group of SAH patients after treatment. In the present study, we found that
the GCS score of patients in the nimodipine-treated group were significantly higher
than those in the control group.
In summary, nimodipine is effective in the treatment of SAH, lowering the incidence
of adverse reactions, reducing the incidence of CVS and improving the prognosis of
patients.
This study is a secondary research, and therefore its quality mainly depends on the
quality of the original researches and may have the following limitations. First,
relevant studies are collected by searching the electronic database and manual screening
literature and references. Therefore, the omission of related articles may be caused
by the possible shortcomings in electronic database collection and search strategy.
Second, only domestic studies based on clinical randomized controlled trials are collected,
and the comprehensive evaluation of literature quality is not high. Therefore, the
statistical results may be biased. Third, the low methodological quality of the included
literature and small sample size of many studies will lead to low power of the test,
and affect the strength of evidence of this study. Collectively, it is still necessary
to carry out a well-designed, scientific, large-sample, multicenter, prospective clinical
study to verify the above conclusions, and to provide more reliable clinical evidence
of medication.
