The protocol of this systematic review will adhere to the desired anecdote matters for systematic reviews and meta-analyses for protocols (PRISMA-P), recommendations for reporting of systematic reviews and meta-analyses of animal experiments [15-17]. This protocol is registered at the International Prospective Register of Systematic Reviews (PROSPERO—CRD42020168304) at https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=168304.
Types of studies
All animal intervention studies that included a control group will be enrolled in this systematic review, which evaluates human and non-human primate PSC-derived DA progenitor transplantation in preclinical animal models of PD. Only published studies published in English without any date restrictions will be assessed.
Types of animal models
All studies that used animal models of experimental PD (6-OHDA or MPTP) in which the animals developed characteristic motor deficits will be considered in this systematic review.
Types of comparators
The comparison group includes animal models of PD that did not receive cells.
Types of intervention
The intervention group includes animals that received human and non-human primate PSC-derived DA progenitors, and have been investigated for the treatment of PD.
Only studies that used iPSCs and ESCs of a primate origin to derive DA progenitors, neural stem cells (NSCs), or NPCs following the establishment of the PD model will be included in the study.
Non-intervention studies such as case reports, congress abstracts, letters to the editor, human studies, and studies that include in vitro experiments will be excluded. Studies that do not evaluate behavior as an outcome will be excluded, in addition to studies that included PSCs from a non-primate source.
Types of outcome measures
The rescued motor deficits are defined as the primary outcome. They will be measured by the apomorphine-induced rotation test (APO-IR), amphetamine-induced rotation test (AMP-IR), stepping test, spontaneous forelimb use test, cylinder test, neurological scores, spontaneous movement, spontaneous rotation, time in movement, or rotarod test.
Secondary outcomes include histological and neuroimaging data obtained before and after cell injections.
Searching methods for identification of studies
Electronic databases used by MEDLINE via PubMed, Web of Science and SCOPUS from their foundation until the end of December 2019 without any search filters and restrictions of language, date of publication, or publication status are utilized.
The main terms are: "pluripotent stem cell" OR "dopaminergic progenitor" OR "DA progenitor" OR "neural stem cell" OR NSC OR "neural progenitor" OR NPC OR "embryonic stem cell" OR ESC OR "induced pluripotent stem cell" OR iPSC OR "pluripotent stem cell" OR PSC) AND (Parkinson's). Supplement 1 lists the details of search strategies to be used in the electronic databases.
Data collection and analysis
Selection of studies
All studies are imported into Covidence, which is a not-for-profit service established in 2013, and run by a team in Melbourne, Australia . At first, duplicate studies will be removed by Covidence. Then, two reviewers (AAK and ZS) will independently screen titles and abstracts. Any conflicts between reviewers will be solved by agreement or by a third reviewer (HB). After screening the titles and abstracts, the same reviewers will independently evaluate the full text of the studies by using a standardized form that contains the inclusion and exclusion criteria.
Two reviewers (ZS and ME) will independently extract all available sources in the text and graphs of each article. If only graphical data are available, values for mean and standard deviation or standard error will be obtained by using GraphClick (Arizona Software, Phoenix, AZ, USA) under high magnification. The calibration exercises will be conducted to ensure consistency between reviewers before starting data extraction.
The data will be extracted using standardized extraction forms: 1) study characteristics (author, year of publication); 2) features of the included animals and animal models (animal species, PD model, age, gender, farming situations, numbers of animals in intervention, and comparison group); 3) interventions (time and description of preparation); and 4) outcomes of interest.
Risk Assessment of bias (Assessment of bias risk)
Two reviewers (ZS and ME) will independently use the SYstematic Review Centre for Laboratory animal Experimentation (SYRCLE) risk of bias tool to evaluate the quality of the studies and risk of any bias. The SYRCLE risk of bias tool includes ten defined criteria assessment domains related to biases of selection, performance, identification, attrition, and reporting. For each included study, all domains will be scored as low, high, or unclear risk of bias .
The Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies (CAMARADES) checklist will be used to evaluate the selected studies. This checklist evaluates publication in a peer-reviewed journal, recording of temperature control, randomized treatment allotment, blinded evaluation of results, reporting of blinding of the operator, suitable animal models, reporting of a sample size calculation, agreement with animal wellbeing principles, statement of potential conflict of interest, and a comprehensive follow-up [20, 21].
Assessment of the treatment efficacy
The calculations for all included variables are as follows: standardized mean difference (SMD), odds ratio (OR), and related 95% confidence interval (CI) for each result using Cohen's method to normalize the different animal species. Binary data will be calculated as the risk ratio (RR) with 95% CI.
Assessment of heterogeneity and data synthesis
The heterogeneity between studies will be estimated by calculating “I2 inconsistency values and Cochran’s Q statistical test”, where I2 > 50% and/or p < 0.10 recommend high heterogeneity. Heterogeneity will be defined according to the I2 range: 0% -40% (minor heterogeneity), 40% -60% (moderate heterogeneity), 60% -90% (substantial heterogeneity), and >90% (significant heterogeneity) .
The analyses will be performed using Review Manager 5.3 . In cases where the Review Manager statistical software is not sufficient, data analyses will be performed by STATA® statistical software, version 14.2 (Stata Corp., College Station, TX, USA) .
Assessment of publication bias
A graphical funnel plot will be used where at least ten studies contribute to a pooled analysis to investigate the publication bias, which will be presented in the studies .