To assess the effectiveness of HLA matching for corneal transplantation. This will be achieved through the completion of a systematic review of the studies:
- Assessing the effect of major antigen matching on graft outcomes
- Assessing the effect of minor antigen matching on graft outcomes
The protocol for this systematic review was registered with the PROSPERO database (reference CRD42020198882).18 The review and its findings are reported in accordance with PRISMA guidelines.19 A PRISMA-P checklist for this protocol is shown in Additional File 1.
The following sources will be searched between summer and autumn 2020, with no date restriction applied:
- The Cochrane Library (CENTRAL Register of Controlled Trials)
- MEDLINE, MEDLINE in process (Ovid)
- Web of Science
Registers of clinical trials
Abstract and conference proceedings:
- Conference Proceedings Citation Index (Web of Science)
- British library ZETOC
Dissertations, theses and grey literature:
For bibliographic databases, the search strategy will combine index and free terms for the surgical procedure and distinctive lamellar techniques.
A sample strategy from MEDLINE has been formulated to collate all relevant evidence, and this has been included as Appendix 1. For each of the databases above, the search strategy may be adapted as deemed appropriate. An iterative manner will be applied to complete the search from these sources. The bibliographic references of the 1995 systematic review and any appropriate evidence reviews will be hand searched to ensure that no relevant primary study has been missed. Furthermore, a clinical expert will be contacted to ensure no similar systematic reviews are currently ongoing. To collate a comprehensive range of evidence, no restrictions will be placed by either publication date or language. RefWorks and Rayyan will be used to manage the search results. This will also enable exclusion of any duplicate entries, study details and references. Grey literature will also be searched alongside electronic databases to reduce the risk of publication bias being introduced into the systematic review.
The following criteria will be utilised to select studies for this review:
Patients (humans) undergoing corneal transplantation
Donor-recipient HLA matching
Patients receiving unmatched/ selectively matched/ or randomly allocated donor corneas
Occurrence of rejection and failure
Table 1: PICO framework used to generate this review
- Study design:
- RCTs, non-RCT trial-based studies and cohort studies.
- Patients of any age, gender or ethnicity undergoing any form of corneal transplant. No restriction on date of transplantation will be applied.
- Intervention and comparator
- Comparing the use of major antigen matching to antigen mismatching.
- Comparing the use of minor antigen matching to antigen mismatching.
- Primary outcome
- Corneal graft prognosis in the post-operative period: number of graft rejections, immunoreactions, failure and/or survival
With the exception of limbal, endothelial and tectonic transplants, all type of corneal transplant will be analysed.
Major antigen studies will be defined as those that discuss the effect of MHC class I (HLA-A, HLA-B and HLA-C), class II (HLA-DP, HLA-DQ and HLA-DR) and/or class III genes.20 As there are an abundance of minor antigen sub-types 21, any studies including histocompatibility complexes not concerning the aforementioned antigens will be considered as minor.
All types of corneal transplant will be eligible for inclusion regardless of the underlying disease it was used to treat. However, it is important to note that grafts for keratoconus and other non-inflammatory conditions are likely to have better outcomes, compared to outcomes in patients with inflammatory diseases and re-grafts. The analysis of the studies may therefore be grouped based on the underlying disease, should the studies permit such stratification. Studies that include both the assessment of major and minor antigen matching will also be included in this review.
Selection of studies will be in two stages:
- Abstracts and titles of each study will be screened to exclude unnecessary data.
- Potentially relevant studies will have their full texts extracted and assessed against the selection criteria.
The appropriateness of articles will be assessed independently by two reviewers (JPC and SSK). A third reviewer (GB) will resolve any conflicts of opinion between each assessment. This process will be outlined through a PRISMA flow diagram. Exclusion of studies will be recorded and discussed in this review, and any non-English language studies will be translated to allow for a fuller inclusion of relevant studies.
Relevant data from the suitable studies will be extracted independently by two individual authors. Any differences in opinion will be settled by discussion between both authors. If insufficient, this will be followed by a referral to a third author to resolve the matter at hand. A standardised data collection form in Microsoft Excel will be created and used by the authors to summarise the extracted data. The study authors and publishing bodies may be contacted if any relevant information is missing from the reviewed studies. For each study the following information, but not limited to, will be extracted:
- Study characteristics
- Authors, publication year, title and journal
- Study design
- Sample size
- Length of follow-up and variability in post-operative treatment
- Participant characteristics
- Patient selection and recruitment criteria
- Demographic data-number, age, gender, socioeconomic status and ethnicity
- Past ocular history
- Intervention and comparator
- Donor-recipient HLA matching
- Comparator: patients receiving unmatched/randomly allocated donor corneas
- Any differences in underlying care between the treatment groups
- Outcomes and findings
- Number of graft rejections/reactions/failures at pre-defined follow-up intervals
- Graft survival times (time to rejection and time to failure)
- Adverse events (including side effects and complications of treatment)
- Precision and statistical test results for each outcome
- Completeness of follow-up for each outcome
The quality of all the included studies will be assessed independently by two reviewers. Any disagreements will be resolved by discussion between the two individuals. If necessary, a third reviewer will act as an impartial mediator. RCTs will be assessed using the Cochrane Handbook Risk of Bias tool (RoB 2).22 This will also be used to assess non-randomised trials, hence it is understood that the criteria present in this tool for allocation concealment and randomisation will not be relevant. Any prospective controlled observational studies will be assessed using the guidelines present in Chapter 13 of the Cochrane Handbook.22 The RoB 2 tool may also be used as a minimum assessment - again without all criteria of the tool being relevant for this type of study. In prospective observational studies, the most relevant information to evaluate is the group selection criteria, differences in patient characteristics, losses to follow-up, biases and other confounders. The assessment of the biases present in each of the included studies will be collated in a findings table. In particular, the authors acknowledge that there may be variation in the definition of the study endpoints across the literature and this will, therefore, be interrogated as a source of bias.
Included studies will be grouped based on the type of corneal transplant used (intervention) and the outcome parameters measured. It may be necessary to further stratify the studies based on the underlying disease state. The data will be tabulated and a narrative synthesis of the relevant evidence compiled for each outcome of relevance to the review. This will aid with the summary of findings from each study and help to identify patterns in the data. Bias within studies will be quantitatively assessed and tabulated using risk of bias tools.
A decision on whether a fixed-effects or random-effects meta-analysis is to be performed will be made based on study heterogeneity and whether there is sufficient homogeneity to warrant its effective completion.23 Heterogeneity will be determined in the context of percentage variation in study results (using the I2 statistic) and variance of effect size (using the tau-squared statistic).
Data pooling will be carried out for the purpose of generating a Forest Plot, which will serve as a visual representation of the pooled effect of the said data. However, data from studies with variable study designs will not be pooled together.
Where heterogeneity is significant in studies, subgroup analysis may be conducted in order to investigate the source of heterogeneity, if the completeness of a study’s data collection and reporting allows for this. Given the high variability in follow-up periods across the studies to be analysed, time-specific data from the post-operative period might have to be grouped where appropriate. Alternatively, outcome data may be grouped across all post-operative periods if better suited.
The results of this systematic review and associated meta-analysis will be reported using guidance laid out by the PRISMA reporting tool.19 This will be done with the intention of ensuring the reporting of results is both complete and transparent, under a well-recognised checklist.
The robustness of the review methodology will be discussed. An examination of both the internal and external validity of the results will also be conducted, for a complete picture of the integrity of the evidence base on which this review will be founded.
Following this, the implications of this review for future research, practice, health guideline revision and implementation will be considered.