Protocol and registration
This protocol was designed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P). A summary of the protocol has been registered on PROSPERO: CRD42022374914. We aimed to answer the following questions: can edaravone promote the neurological function recovery and ameliorate tissue damage in rats following SCI? What is the mechanism underlying neuroprotective effect of edaravone in treating SCI?
Eligibility criteria
Types of studies
The controlled studies evaluating the neurological roles of edaravone on experiment SCI in rat models will be included. The clinical studies, case reports, in vitro studies, reviews and comments will be excluded. No restrictions are specified on the language, date, or publication status of published studies [26].
Types of animal models
We will include laboratory rats with traumatic SCI caused by compression or contusion that are not restricted by age, sex or strain. Any other modelling modality that can cause SCI will be excluded, such as: non-traumatic ischemia, laceration, transection, traumatic root avulsion injury or genetic modifications, as these models do not represent the formation mechanisms of typical human crush SCI [27].
Types of intervention
The assessed intervention will be edaravone administration used in rats with SCI. The dosage, formulation and administration methods of edaravone were unrestricted. The multidrug combination intervention group (e.g., edaravone plus bone marrow mesenchymal stem cells) will be excluded.
Types of comparators
The edaravone-treated populations will be compared with the placebo or control group, which receives saline, saline with diluted dimethyl sulfoxide, vehicle, or no treatment. Studies will be excluded in the absence of placebo or control group.
Types of outcome measures
Primary outcome
The 21-point Basso, Beattie, and Bresnahan (BBB) locomotor rating scale was considered as primary outcomes. Briefly, the researchers independently documented the limb movements and walking characteristics of the study subjects in the open field at the same time, then assigned a BBB scale score for indicating the basic locomotion function. The scales were scored from 0 (complete paralysis)-21 (normal locomotion), representing the grading of the animal's locomotor outcome after SCI [28]. Only the data at same time points will be used in the analyses of BBB scores.
Secondary outcomes
The priori secondary outcomes included spared white matter area and malondialdehyde. White matter damage is one of the main causes of motor loss after SCI, and the residual white matter is critical for the recovery of motor function in the hind limbs of animals [29]. Malondialdehyde is the end product of free radical-mediated lipid peroxidation reaction in tissues. It can not only respond to the rate and intensity of peroxidation in the organism, but also indirectly reflect the degree of tissue peroxidative damage [30]. To reduce the risk of data heterogeneity, only data obtained by the same detection method was used in the analyses.
Information sources and search strategy
Information sources
The following electronic databases will be searched from their inception date: PubMed, Embase, Web of Science, Scopus and Cochrane Library. We will also screen the reference lists of selected studies and reviews for covering the additional eligible studies not retrieved by the search. There were no restrictions on language, publication date or publication status.
Search strategy
Search strategies will be constructed by combining the Medical Subject Headings terms and free-text terms related to the interested disease, intervention and animal. The key terms “spinal cord injuries”, “spinal cord injury”, “spinal cord diseases”, “spinal cord compression”, “spinal cord trauma”, “edaravone”, “edaravone”, “MCI 186”, “rats” and “rat” were used. A preliminary search strategy for PubMed can be viewed in Appendix 1.
Study selection and data extraction
Procedure for study selection
The results retrieved by the initial searches will be imported into NoteExpress 3.2. Following exclusion of duplicates, two reviewers will independently screen titles and abstracts of retrieved studies to identify studies that potentially meet the inclusion criteria. Then, the full text of these potentially eligible studies will be accessed by the two reviewers independently to identify whether it could be included (Fig. 1). A senior author will involve to make the final decision when there is a disagreement between two researchers.
Data extraction
Two reviewers independently performed data extraction and records of study information, collecting data parameters including: the name of the first author, publication year, study population (strain, number, sex, age and weight), model characteristics (method of model, injury level), intervention details (dosage, administration route, timing and times) and measurements. Moreover, the proposed mechanisms and changes of related molecules will be also extracted from included studies for investigating the neuroprotective mechanism behind edaravone [26, 31]. In studies involving multiple intervention groups, only data from edaravone and negative control groups will be extracted for analyses. If values were not explicitly provided in the article, but rather the study data were represented graphically, we will use the GetData Graph Digitizer 2.26 (http://getdata-graph-digitizer.com) to extract the corresponding statistics from the graphs. If data were missing, authors then will be contacted.
Risk of bias assessment
The SYstematic Review Centre for Laboratory animal Experimentation risk of bias (SYRCLE’s RoB) tool will be used to assess the methodological quality of the included studies and to determine the intrinsic veracity of the studies [32]. The checklist includes 10 items regarding selectivity bias, implementation bias, measurement bias, missed visit bias, reporting bias, and other biases. According to the SYRCLE’s risk of bias tool, two investigators independently scored those domains in each study with “yes”, “no” or “unclear”, representing a low, high, and unclear risk of bias, respectively. Any discrepancy will be resolved by consensus and the involvement of a third collaborator [33].
Data synthesis
Summaries and analyses of data from included studies will be performed using RevMan 5.3 software (provided by the Cochrane Collaboration). When at least 3 eligible studies are identified, a meta-analysis comparing edaravone group to control group will be performed. In accordance with the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions, for studies with multiple intervention groups, combining those groups will be performed to enable a single pair-wise comparison. The mean, standard deviation, as well as the sample size of animals in each group will be used for comparisons. Summaries of intervention effects for each study by weighted mean differences or standardized mean differences and 95% confidence intervals will be presented. For the assessment of heterogeneity of included studies, we employ the P value in the χ2 test and Cochrane's I2. Heterogeneity is presumed in the event that the P value is less than 0.10, and is considered to be high when the I² value is more than 50%. A random effects model will be used to account for anticipated heterogeneity due to the exploratory nature of animal studies. A P value less than 0.05 was considered a statistically significant. Linear graphs were constructed by GraphPad Prism software to highlight the dynamic weighted mean differences of BBB scores and dynamic BBB score improvements in both groups.
If sufficient data are available, subgroup analyses by species, age, sex, duration of intervention, dose or route of administration will be carried out to explore the factors modifying on BBB scores. Meanwhile, sensitivity analyses on BBB scores are planned to be performed by excluding either all single studies, non-investigator-blind studies or other type of studies, which is intended to test the robustness of our findings and analyze the heterogeneous source. If we include at least 10 studies in a meta-analysis related to primary outcomes, funnel plots will be employed to test the potential risk of publication bias.
In addition, to explore the appropriate dose of edaravone, we will use a network meta-analysis approach according to the Bayesian method using Stata 12.0 (StataCorp LP, College Station, Texas, USA) to assess absolute differences by simultaneously comparing direct or indirect evidence from the treatment regimens in included studies. For every treatment dose, we plan to calculate the probabilities of its efficacy, and then ranked the treatments according to surface under the cumulative ranking curve.