The efficacy of different nimodipine administration route for treating subarachnoid hemorrhage: A network meta-analysis

Abstract

Background:

A systematic review and network meta-analysis (NMA) were conducted to explore the optimal administration route of nimodipine for treatment subarachnoid hemorrhage.

Methods:

Electronic databases (Pubmed, Embase, Web of Science and Cochrane databases) were systematically searched to identify randomized controlled trials evaluating different administration route of nimodipine (intravenous and enteral) versus placebo for treatment subarachnoid hemorrhage. Outcomes included case fatality at 3 months, poor outcome measured at 3 months (defined as death, vegetative state, or severe disability), incidence of delayed cerebral ischemia (DCI), delayed ischemic neurological deficit. A random-effect Bayesian NMA was conducted for outcomes of interest, and results were presented as odds ratios (ORs) and 95% credible intervals. The NMA was performed using R Software with a GeMTC package. A Bayesian NMA was performed and relative ranking of agents was assessed using surface under the cumulative ranking (SUCRA) probabilities.

Results:

Nine randomized controlled trials met criteria for inclusion and finally included in this NMA. There was no statistically significant between intravenous and enteral in terms of case fatality, the occurrence of DCI, delayed ischemic neurologic deficit and poor outcomes (P > .05). Both intravenous and enteral could reduce case fatality, the occurrence of DCI, delayed ischemic neurologic deficit and poor outcomes (P < .05). The SUCRA shows that enteral ranked first, intravenous ranked second and placebo ranked the last for case fatality, the occurrence of DCI and poor outcomes. The SUCRA shows that intravenous ranked first, enteral ranked second and placebo ranked the last for delayed ischemic neurologic deficit.

Conclusions:

It is possible that both enteral and intravenous nimodipine have comparable effectiveness in preventing poor outcomes, DCI, and delayed ischemic neurological deficits. However, further investigation may be necessary to determine the exact role of intravenous nimodipine in current clinical practice.

1. Introduction

Spontaneous subarachnoid hemorrhage (SAH) is a neurosurgical emergency.^[1] SAH is a disease that affects a large percentage of the world population, the overall case fatality rate is 41.7% within 28 days of onset, according to a World Health Organization (WHO) survey.^[2,3] Cerebral vasospasm is a frequently observed complication following SAH, with the constriction of the major conduit arteries (such as the basilar artery and the arteries of the Circle of Willis) near the location of a ruptured aneurysm being particularly intense.^[4] Cerebral vasospasm has always been considered a severe complication following SAH.^[5]

If a patient develops cerebral vasospasm within the first 2 weeks after SAH, the mortality rate increases by 1.5 to 3.0 times.^[6]

Over the years, in order to solve this growing problem, a lot of research has been performed to investigate calcium antagonist for treatment cerebral vasospasm.^[7] Nimodipine is a 1,4-dihydropyridine calcium channel blocker, which is commonly used to treat conditions such as cerebrovascular spasm, stroke, and migraine.^[8,9] Nimodipine has been shown to dilate cerebral arterioles, increase cerebral blood flow without a concurrent increase aerobic metabolism of brain.^[10,11] Thus, nimodipine plays a very important role in the prevention and treatment of cerebral vasospasm and its associated brain damage. As research on nimodipine continues to deepen, many treatment methods have been summarized, including local and systemic administration.^[12,13] The administration of nimodipine through the enteral route is commonly utilized and is recommended in the current guidelines for treating patients with SAH.^[14] In contrast, fewer and smaller studies have examined the effects of intravenous nimodipine, and only 3 studies have directly compared its effectiveness with enteral nimodipine.^[9]

Based on the current available evidence, intravenous nimodipine is only considered as an alternative for patients who are unable to receive the enteral formulation.^[15,16] Nonetheless, some authors argue that enteral bioavailability might be lower than anticipated in specific groups of patients, as reported in patients with high-grade SAH who have shown negligible serum concentrations.^[9] Thus, it is suggested that in certain cases, parenteral administration could be more effective. At present, there is no definitive evidence available regarding the superiority of intravenous nimodipine over the enteral route in treating SAH.

Network meta-analysis (NMA), which is also referred to as mixed treatment comparison or multiple treatments comparison meta-analysis, is a statistical technique that enables the direct and indirect comparison of the effects of 2 or more treatments, and facilitates the ranking of different treatment options.^[17,18]

Therefore, our objective was to conduct a systematic review and NMA comparing the efficacy of different administration route of nimodipine for treatment SAH. We hypothesized that intravenous and enteral had similar clinical effect when treatment with SAH.

2. Methods

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines and the PRISMA Network Meta-Analysis Extension statement will be followed during the conduct of this systematic review and NMA.^[19]

2.1. Search strategy

A search was conducted by 2 reviewers (Gang Lei and Zhongxian Rao) for potentially relevant randomized controlled trials (RCTs) about different administration route of nimodipine (intravenous and enteral) versus placebo for treatment SAH on Pubmed, Embase, Web of Science and Cochrane databases up to November 11, 2022. A structured search was performed using the following search string: “intravenous” OR “enteral” AND (“subarachnoid hemorrhage” and “nimodipine”). Detailed search strategy for Pubmed was listed in Supplement file 1, http://links.lww.com/MD/J650. The titles, abstracts, and full publications were screened against pre-defined criteria by 2 independent reviewers. Disagreements were either resolved through discussion leading to a consensus or by seeking the intervention of a third reviewer. For this systematic review and meta-analysis, manual searches will be conducted for reference lists, related citations, and gray literature from websites. No ethics approval was necessary as there was no patient contact involved.

2.2. Inclusion criteria and exclusion criteria

To be included in the meta-analysis, studies had to fulfill the PICOS (population, intervention, comparator, outcome, study design) criteria as follows: population (P) - patients with subarachnoid hemorrhage; intervention (I) - the intervention group received intravenous or enteral nimodipine, or placebo; outcomes (O) - case fatality at 3 months, poor outcome measured at 3 months (defined as death, vegetative state, or severe disability), incidence of delayed cerebral ischemia (DCI) and delayed ischemic neurological deficit (DIND); study design (S)-RCTs. Retrospective studies, cadaver studies, comments, letters, editorials, protocols, guidelines, surgical registries, and review papers were excluded.

2.3. Literature selection

After collecting all relevant studies, they were imported into Endnote X7 and duplicate literature was removed. Two researchers (Zhongxian Rao and Yuping Hu) then independently reviewed the titles and abstracts to exclude studies that did not meet the PICOS criteria. Any remaining irrelevant studies were subsequently removed. In case of any disagreement regarding which literature to include, a senior reviewer was consulted.

2.4. Data extraction

Two independent reviewers (Gang Lei and Yuping Hu) extracted the available data from the included studies, which consisted of information such as author, study design, publishing year, age, sample size, gender, intervention, and control procedure. The primary outcomes assessed were case fatality at 3 months, poor outcome measured at 3 months (defined as death, vegetative state, or severe disability), incidence of DCI and DIND. DIND was defined by a delayed decrease of consciousness by at least 2 Glasgow Coma Scale levels and/or an increase ≥2 points on the abbreviated National Institutes of Health Stroke Scale and/or a new focal neurological deficit.^[20] DCI was defined as confirmed new hypodensities on the computed tomography scan that were only attributable to cerebral vasospasm and DCI.^[21] If any data was missing, we contacted the corresponding author of the study to obtain the necessary information.

2.5. Quality assessment

Two reviewers (Zhongxian Rao and Yuping Hu) evaluated the risk of bias in RCTs based on the Cochrane Handbook for Systematic Reviews of Interventions version,^[22] which included items such as sequence generation, allocation concealment, blinding of participants and outcome assessors, incomplete outcome data, reporting bias, and other bias. Any discrepancies in the evaluations between the 2 reviewers were resolved by a third reviewer.

2.6. Data analysis and statistical methods

We will use Bayesian methods, specifically JAGS via R with the R package gemtc (https://cran.r-project.org/web/packages/gemtc/gemtc.pdf), for the network meta-analyses of efficacy outcomes. To ensure the reliability of our results, we performed sampling simulations and MCMC calculations using a random effects model. We evaluated the convergent diagnostic results using diagnostic plots such as trajectory plots and density plots. We calculated the means under the random effects model and fixed effects model and examined the homogeneity in the literature using the BlandAltmanLeh package. Good homogeneity is indicated when the distances between all points are within 95% of the LoA. To test consistency between direct and indirect comparisons, we used the node-splitting approach and deemed P values >.05 to be favorable. We investigated further heterogeneity when it was detected, with I² values over 50% indicating heterogeneity, and we considered the total I².pair and I².cons for the overall results. We also performed sensitivity analyses by excluding 1 study at a time and combining the remaining studies for analysis to evaluate the potential impact on the results. We calculated the mean surface under the cumulative ranking (SUCRA) curve for each intervention, with a higher SUCRA indicating a higher rank of the protocol.

3. Results

3.1. Search results

We identified 966 relevant studies from various databases (Pubmed, Embase, Web of Science and Cochrane databases) based on our search strategies. No additional records were identified through other sources such as reference lists. Using Endnote Software (Version X7, Thompson Reuters, CA), we removed 110 duplicate studies. After screening the titles and abstracts, we excluded 843 studies and then removed 4 more studies by reading the full text (Fig. 1). Finally, 9 studies were included in our meta-analysis.^[23–31]

Figure 1.

Flow diagram of the literature selection process.

3.2. Study characteristics

The general characteristic of the included RCTs were listed in Table 1. All of them evaluated the efficacy of intravenous, enteral and placebo for SAH. All 9 included literatures were published between 1983 and 2012. Two studies originated from United Kingdom, 2 studies originated from Finland, 2 studies originated from Swedan, 1 from USA, 1 from France. Eight of 9 RCTs were graded based on Hunt and Hess clinical grade, most of the studies graded at Hunt and Hess I–II. Nimodipine dose ranged from 28 mg/day to 360 mg/day for enteral administration. And nimodipine dose mostly administrated with 48 mg/day. Duration of treatment ranged from 7 days to 21 days.

Table 1.

General characteristic of the included studies.

Author	Country	Clinical grade	Nimodipine dose	Duration of treatment	Outcomes	Comparator	Control	Comparator	Control
Allen 1983[23]	USA	Hunt and Hess I–II	3 mg/kg divided in 6 daily doses	21 days	DIND, death, severity of neurological deficit	Enteral	Placebo	58	63
Philippon 1986[29]	France	Hunt and Hess I–III	360 mg/d	21 days	DIND, death, GOS at 21 days after SAH	Enteral	Placebo	39	42
Messeter 1987[25]	Swedan	Hunt and Hess I–III	48 mg/d	> 9 days	DIND, death, functional outcome without time indication	Intravenous	Placebo	13	7
Neil-Dwyer 1987[26]	United Kingdom	Hunt and Hess I–IV	360 mg/d	21 days	Death, DIND, good or poor functional outcome at 3 months	Enteral	Placebo	38	37
Petruk 1988[28]	Canada	Hunt and Hess III–IV	540 mg/d	21 days	Evaluation of GOS on day 21 days and 3 months after SAH. DCI, DIND, case fatality	Enteral	Placebo	91	97
Ohman 1991	Finland	Hunt and Hess I–III	0.5 μg/kg/min	7–10 days	Outcome at 3 months, death, DCI, DIND	Intravenous	Placebo	104	109
Pickard 1989[30]	United Kingdom	World Federation of Neurosurgeons I–V	360 mg/d	21 days	DCI, death, GOS at 3 months	Enteral	Placebo	278	276
Kronvall 2009[24]	Swedan	Hunt and Hess I–V	48 mg/d	10 days	Incidence of DINDs, DCI, clinical outcome based on the GOS at 3 months, death	Intravenous	Enteral	57	49
Soppi 2012[31]	Finland	Hunt and Hess I–V	48 mg/d	10 days	Incidence of DIND, DCI, clinical outcomes at 3 and 12 months after SAH, death	Intravenous	Enteral	87	84

DCI = delayed cerebral ischemia, DIND = delayed ischemic neurological deficit, SAH = spontaneous subarachnoid haemorrhage.

3.3. Risk of bias

Risk of bias summary and risk of bias graph can be seen in Figures 2 and 3 respectively. For random sequence generation, 5 of the 9 studies were classified as high risk of bias, 2 studies were rated as low risk of bias and the rest 2 studies were identified as unclear risk of bias. Six studies were rated as unclear risk of bias for allocation concealment. All included studies were listed as low risk of bias for blinding of participant and personnel. Only one study was rated as low risk of bias for blinding of outcome assessment. Only 2 studies were listed as low risk of bias for incomplete outcome data. Four studies were listed as low risk of bias for selective reporting and other bias.

Figure 2.

Risk of bias summary of the included studies.

Figure 3.

Risk of bias graph of the included studies.

3.4. Case fatality

A total of 9 studies, including 3 treatments (intravenous, enteral and placebo) contributed to the clinical outcome of the case fatality. As displayed in Figure 4A, the network structure diagrams detailed the direct comparisons between different treatment in the case fatality.

Figure 4.

(A) Network structure diagrams of case fatality; (B) Forest plot of the case fatality as compared with placebo; (C) Heterogeneity of the included studies; (D) Surface under the cumulative ranking curve (SUCRA) probabilities of different drugs for case fatality.

In head-to-head comparison, enteral (OR 0.55, 95% CrI 0.40–0.75) and intravenous (OR 0.49, 95% CrI 0.32–0.74) was more effective than the placebo in terms of case fatality, and the difference was statistically significant (Fig. 4B and Table 2). However, there was no statistically significant between intravenous and enteral in terms of case fatality (OR 1.12, 95% CrI 0.74–1.71). NMA showed no heterogeneity with global I² = 0% (Fig. 4C).

Table 2.

Efficacy of different comparisons of drugs for case fatality by odds ratios (ORs) and corresponding 95% credible intervals (CrIs).

Enteral	0.89 (0.58, 1.35)	1.83 (1.33, 2.5)
1.12 (0.74, 1.71)	Intravenous	2.06 (1.35, 3.11)
0.55 (0.4, 0.75)	0.49 (0.32, 0.74)	Placebo

The SUCRA shows that enteral ranked first (SUCRA, 86.1%), intravenous ranked second (SUCRA, 63.7%), placebo ranked the last (SUCRA, 0.1%, Fig. 4D).

3.5. The occurrence of DCI

A total of 9 studies, including 3 treatments (intravenous, enteral and placebo) contributed to the clinical outcome of the occurrence of DCI. As displayed in Figure 5A, the network structure diagrams detailed the direct comparisons between different treatment in the occurrence of DCI.

Figure 5.

(A) Network structure diagrams of the occurrence of delayed cerebral ischemia (DCI); (B) Forest plot of the occurrence of DCI as compared with placebo; (C) Heterogeneity of the included studies; (D) Surface under the cumulative ranking curve (SUCRA) probabilities of different drugs for the occurrence of DCI.

In head-to-head comparison, intravenous (OR 0.14, 95% CrI 0.076–0.27) and enteral (OR 0.15, 95% CrI 0.068–0.31) was more effective than the placebo in terms of case fatality, and the difference was statistically significant (Fig. 5B and Table 3). However, there was no statistically significant between intravenous and enteral in terms of case fatality (OR 0.99, 95% CrI 0.52–1.90). NMA showed no heterogeneity with global I² = 0% (Fig. 5C).

Table 3.

Efficacy of different comparisons of drugs for the occurrence of delayed cerebral ischemia (DCI) by odds ratios (ORs) and corresponding 95% credible intervals (CrIs).

Enteral	1.01 (0.53, 1.94)	6.91 (3.7, 13.13)
0.99 (0.52, 1.9)	Intravenous	6.84 (3.25, 14.68)
0.14 (0.08, 0.27)	0.15 (0.07, 0.31)	Placebo

The SUCRA shows that enteral ranked first (SUCRA, 75.8%), intravenous ranked second (SUCRA, 74.1%), placebo ranked the last (SUCRA, 0.1%, Fig. 5D).

3.6. Delayed ischemic neurologic deficit

A total of 5 studies, including 3 treatments (intravenous, enteral and placebo) contributed to the clinical outcome of the delayed ischemic neurologic deficit. As displayed in Figure 6A, the network structure diagrams detailed the direct comparisons between different treatment in the delayed ischemic neurologic deficit. NMA showed no heterogeneity with global I² = 0% (Fig. 6B).

Figure 6.

(A) Network structure diagrams of delayed ischemic neurologic deficit; (B) Forest plot of the delayed ischemic neurologic deficit as compared with placebo; (C) Heterogeneity of the included studies; (D) Surface under the cumulative ranking curve (SUCRA) probabilities of different drugs for delayed ischemic neurologic deficit.

In head-to-head comparison, intravenous (OR 0.084, 95% CrI 0.049–0.14) and enteral (OR 0.085, 95% CrI 0.058–0.12) was more effective than the placebo in terms of delayed ischemic neurologic deficit, and the difference was statistically significant (Fig. 6C and Table 4). However, there was no statistically significant between intravenous and enteral in terms of case fatality (OR 1.02, 95% CrI 0.63–1.67).

Table 4.

Efficacy of different comparisons of drugs for delayed ischemic neurologic deficit by odds ratios (ORs) and corresponding 95% credible intervals (CrIs).

Enteral	0.98 (0.6, 1.6)	11.72 (8.1, 17.23)
1.02 (0.63, 1.67)	Intravenous	11.92 (7.03, 20.61)
0.09 (0.06, 0.12)	0.08 (0.05, 0.14)	Placebo

The SUCRA shows that intravenous ranked first (SUCRA, 76.3%), enteral ranked second (SUCRA, 73.6%), placebo ranked the last (SUCRA, 0.0%, Fig. 6D).

3.7. Poor outcome

A total of 5 studies, including 3 treatments (intravenous, enteral and placebo) contributed to the clinical outcome of the poor outcomes. As displayed in Figure 7A, the network structure diagrams detailed the direct comparisons between different treatment in the poor outcomes. NMA showed no heterogeneity with global I² = 0% (Fig. 7B).

Figure 7.

(A) Network structure diagrams of poor outcomes; (B) Forest plot of the poor outcomes as compared with placebo; (C) Heterogeneity of the included studies; (D) Surface under the cumulative ranking curve (SUCRA) probabilities of different drugs for poor outcomes.

In head-to-head comparison, intravenous (OR 0.17, 95% CrI 0.098–0.32) and enteral (OR 0.16, 95% CrI 0.098–0.27) was more effective than the placebo in terms of poor outcomes, and the difference was statistically significant (Fig. 7C and Table 5). However, there was no statistically significant between intravenous and enteral in terms of poor outcomes (OR 0.96, 95% CrI 0.54–1.69).

Table 5.

Efficacy of different comparisons of drugs for poor outcomes by odds ratios (ORs) and corresponding 95% credible intervals (CrIs).

Enteral	1.05 (0.59, 1.85)	6.11 (3.71, 10.21)
0.96 (0.54, 1.69)	Intravenous	5.82 (3.08, 11.24)
0.16 (0.1, 0.27)	0.17 (0.09, 0.32)	Placebo

The SUCRA shows that enteral ranked first (SUCRA, 78.6%), intravenous ranked second (SUCRA, 71.3%), placebo ranked the last (SUCRA, 0.0%, Fig. 7D).

4. Discussion

This NMA revealed that enteral and intravenous nimodipine may be similarly effective for SAH patients. The analysis of both direct and indirect evidence suggests that both forms of nimodipine can reduce the incidence of poor outcome, DCI, and DIND. Enteral nimodipine has been in use for more than 3 decades, but evidence regarding the efficacy of intravenous nimodipine in preventing vasospasm is inconclusive, and current guidelines only recommend its use as “good clinical practice” for specific patients.^[32,33] There is limited evidence on the use of intravenous nimodipine compared to placebo, and only a few relatively small trials have directly compared the 2 formulations, reporting no difference in outcome and no relative benefit for intravenous nimodipine.^[25,34]

The purpose of this NMA was to investigate whether intravenous nimodipine could serve as an effective alternative to enteral nimodipine in the management of SAH patients. Recent studies have raised concerns about the efficacy of enteral nimodipine, which may not be suitable for all patients. For example, patients who are unable to swallow have been found to have reduced or negligible serum concentrations when nimodipine is administered through a nasogastric tube. This problem appears to be more prevalent in patients with “high-grade SAH,” and in such cases, intravenous administration of nimodipine may be a viable alternative to the enteral formulation. Additionally, recent evidence suggests that intravenous nimodipine may be associated with higher serum and spinal fluid concentrations. However, the use of intravenous nimodipine has also been linked to a higher incidence of hypotension, necessitating more dose adjustments to maintain mean arterial pressure.^[35]

Previous meta-analyses have confirmed the effectiveness of enteral nimodipine on poor outcomes and intravenous nimodipine on DCI and DIND, but they could not make recommendations on the benefit of the intravenous route for patient outcomes.^[36,37] Although 3 recent network meta-analyses have been published on the treatment of patients with SAH, none of these studies compared the efficacy of enteral and intravenous nimodipine.^[38] The studies have different conclusions about the best agents to use after SAH. For instance, Yu et al^[39] concluded that nimodipine and cilostazol are effective, while Dayyani et al^[40] suggested that nimodipine, nimodipine with magnesium, and high-dose clazosentan could potentially prevent morbidity and mortality. Lastly, Mishra et al^[41] found that nicardipine prolonged-release implants and cilostazol could improve outcomes after SAH.

In our study, we compared the efficacy of intravenous and enteral nimodipine and included 3 studies that directly compared these 2 formulations. Our NMA, which incorporated these direct comparisons, revealed that intravenous nimodipine may potentially prevent poor outcomes in SAH patients. These findings imply a need to reassess the role of intravenous nimodipine in clinical practice, particularly given that it is available in Europe but not in the USA. Although our results are not sufficient to recommend the universal use of intravenous nimodipine, they strongly suggest that it could be a viable alternative to enteral nimodipine and should be carefully considered for future use. Further high-quality RCTs are required to evaluate the efficacy of each formulation in different subsets of SAH patients and to compare their adverse effects.

There are several limitations to this study that require discussion. Firstly, all of the studies we analyzed were found to have some degree of bias, with 6 showing some concerns and 4 showing a high risk of bias. Secondly, while the duration and dose of intravenous nimodipine were similar across the studies, it was less standardized than the enteral formulation. Thirdly, we only searched 4 databases, so it is possible that we missed other relevant studies. Fourthly, the studies we included were published over a 30-year period, and changes in clinical protocols and outcome definitions may have impacted the reliability of our findings. Fifthly, our results represent an average outcome among patients with different characteristics and clinical presentations. Therefore, further research is necessary to determine the optimal use of intravenous nimodipine within the context of precision medicine. Furthermore, most of the articles we included did not provide supplementary analyses to assess the efficacy of nimodipine in patients with varying clinical or radiological presentations, preventing us from stratifying our results by SAH severity.

5. Conclusions

Our study indicates that both intravenous and enteral nimodipine have the potential to be similarly effective in preventing poor outcomes, DCI, and ischemic neurological deficits in patients with SAH. However, additional high-quality RCTs are required to assess whether there are any differences in effectiveness concerning outcomes and specific subsets of patients.

Author contributions

Investigation: Zhongxian Rao.

Methodology: Zhongxian Rao.

Software: Zhongxian Rao.

Visualization: Gang Lei, Yuping Hu.

Writing – original draft: Gang Lei, Yuping Hu.