Evaluation of bone marrow-derived cell-based therapies in the hindlimb ischaemia model: a protocol for a systematic review and meta-analysis

Objective Bone marrow(BM)-derived cell-based therapies for critical limb ischamia showed less clinical benefit than expected. While this might be due to patient-specific factors, it remains possible that important details were lost in the bench-to-clinic translation. The hindlimb ischaemia model is the golden standard to evaluate cell-based therapies aimed at promoting neovascularisation. To inform future trial design and identify potential knowledge gaps, we propose a systematic review and meta-analysis of preclinical evidence to assess the efficacy of BM-derived cell administration in restoring relative perfusion in the hind limb model and identify determinants of therapeutic efficacy. Search strategy PubMed and EMBASE were searched for prospective studies in which the hindlimb ischaemia model was used to assess BM-derived therapies. Screening and annotation Studies with an outcome measure related to relative perfusion of the hindlimb will be included. Study characteristics which include model-related factors as well as details on BM therapy will be extracted. Data management and reporting For the primary analysis, a random effects model will be constructed using the mean difference calculated from the maximum relative perfusion for each study arm in each study. A separate model will be constructed using the relative perfusion at the latest time point in each study. We will also assess the risk of bias using the SYRCLE tool for internal validity. Subgroup analysis will be performed on animal characteristics, administration route, dose and cell characteristics such as the cell donor. PROSPERO registration number This protocol has been registered at PROSPERO (CRD2021226592).


Specify the outcome measures
Relative perfusion of the limb (after intervention) 16.
State your research question (based on items [11][12][13][14][15] What is the effect of bone marrow derived cell products on limb perfusion in animal models of hindlimb ischemia? We will screen the reference lists of both relevant reviews (identified through our search strategy) and included studies, for additional references to possibly relevant studies. These references will selected on title and are then added to the screening process, and screened as usual (described under 21-21) Study selection

21.
Define screening phases (e.g. prescreening based on title/abstract, full text screening, both) 1). Screening on eligibility on Title/Abstract 2). Screening on final inclusion on full text 22.
Specify (a) the number of reviewers per screening phase and (b) how discrepancies will be resolved Title/abstract: 2 reviewers. Full text screening: 1 reviewer. Discrepancies will be resolved through discussion, a third reviewer will act as tie-breaker. Define all inclusion and exclusion criteria based on:

Type of study (design)
Inclusion criteria: Prospective, controlled, intervention study with separate treatment arms Exclusion criteria: Case reports, Observational studies, Cross-over studies, non primary studies such as reviews 24.
Type of animals/population (e.g. age, gender, disease model) Inclusion criteria: Animals undergoing femoral artery ligation or banding in one limb (co-morbidities and knockouts incl. immunosuppressed animals will annotated as covariable) . Exclusion criteria: Studies in humans, in vitro or in silico studies. Non-permanent disruption of blood flow (such as ischemia-reperfusion models that feature a temporary Timing of the outcome measurement in hours post induction of hindlimb ischemia Relative perfusion as measured by laser Doppler perfusion imaging or laser speckle contrast analysis (this is reported as a ratio of the ischemic:non ischemic leg).

Other (e.g. drop-outs)
Assessment risk of bias (internal validity) or study quality 37.
Specify (a) the number of reviewers assessing the risk of bias/study quality in each study and (b) how discrepancies will be resolved 2 reviewers. Discrepancies will be resolved through discussion.

38.
Define criteria to assess (a) the internal validity of included studies (e.g. selection, performance, detection and attrition bias) and/or ( Methods for data extraction/retrieval (e.g. first extraction from graphs using a digital screen ruler, then contacting authors) 1. Copying reported values if mentioned in paper (visual check whether it is similar to graphs) 2. Extraction from graphs using Plot digitizer 41.
Specify (a) the number of reviewers extracting data and (b) how discrepancies will be resolved 1 reviewer will extract the data. A random selection of 10% will be assessed by a second assessor to determine accuracy of the data extraction. A third assessor will mediate resolution. Data analysis/synthesis 42.
Specify (per outcome measure) how you are planning to combine/compare the data (e.g. descriptive summary, meta-analysis) All outcome data will be reported in a descriptive summary which will include animal characteristics, treatment characteristics, and details on the cells used. A meta analysis will be conducted if at least 10 studies can be included in the analyisis. Subgroup analyses will be performed if at least 5 studies can be included in each stratum. Correction for multiple testing will be applied using the Bonferroni-Holmes method.

43.
Specify (per outcome measure) how it will be decided whether a metaanalysis will be performed relative perfusion: We require at least 10 individual studies in the overall analysis. For sub-group analyses, at least 5 studies per stratum. If a meta-analysis seems feasible/sensible, specify (for each outcome measure):

44.
The effect measure to be used (e.g. mean difference, standardized mean difference, risk ratio, odds ratio) Any sensitivity analyses you propose to perform As a sensitivity analysis, we will re-run the analysis using the latest time point in each study (as opposed to the maximum perfusion)

49.
Other details meta-analysis (e.g. correction for multiple testing, correction for multiple use of control group) The Bonferroni-Holmes correction will be applied to correct for multiple testing.

50.
The method for assessment of publication bias Publication bias will be assessed over the overall effect by funnel plot. A minimum of 10 studies is needed (see above). The trim-and-fill test will be used to assess asymmetry.
Final approval by (names, affiliations): Date: BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)