Prognostic Features of Preoperative OCT in Retinal Detachments: A Systematic Review and Meta-analysis.

TOPIC: To evaluate the prognostic association between preoperative features seen on OCT imaging and postoperative visual acuity (VA) outcomes in rhegmatogenous retinal detachments (RRDs). CLINICAL RELEVANCE: Currently, there is limited literature on the prognostic value of preoperative RRD OCT features. METHODS: A literature search was conducted on Ovid MEDLINE, Ovid Embase, and Cochrane CENTRAL from inception to September 15, 2022. A meta-analysis was performed using a random-effects model. Quality of studies and evidence were assessed using the Joanna Briggs Institute tools and the Grading of Recommendations, Assessment, Development and Evaluation framework, respectively. RESULTS: A total of 1671 eyes of 1670 patients from 29 observational studies were included. Of these, 89% of eyes had a macula-off RRD at presentation. The mean average duration of detachment was 15 ± 10 days. Most eyes (62%) underwent pars plana vitrectomy. Six preoperative OCT features were analyzed: height of retinal detachment (HRD) at the fovea, central macular thickness (CMT), disruption of the ellipsoid zone (EZ) and/or external limiting membrane (ELM), intraretinal cystic cavities (ICCs), outer retinal corrugations (ORCs), and macular detachment. A greater HRD was weakly associated with postoperative VA (Pearson correlation r = 0.35; 95% confidence interval [CI], 0.20-0.48; P < 0.01), and there was no change in this association throughout the postoperative follow-up period. The CMT was not associated with postoperative VA. Eyes with disruption of the EZ and/or ELM had a postoperative VA worse by 0.35 logarithm of the minimum angle of resolution (logMAR) (95% CI, 0.15-0.54; P < 0.01) or 3 Snellen lines. Eyes with ICCs had a postoperative VA worse by 0.14 logMAR (95% CI, 0.01-0.26; P < 0.01) or 2 Snellen lines. Eyes with ORCs did not have a significantly different postoperative VA than eyes without ORCs. Eyes with macular detachment had a postoperative VA worse by 0.15 logMAR (95% CI, -0.31 to 0.00; P = 0.02) or 2 Snellen lines. Overall, the quality of studies ranged from moderate to good (73%-100%). All associations had a low quality of evidence, with CMT being of very low quality. CONCLUSION: Despite the low quality of evidence, a greater HRD, disruption of the EZ and/or ELM, presence of ICCs, and macular detachment were associated with a poor postoperative VA. We propose a standardized nomenclature for consistency and accuracy in reporting preoperative RRD OCT features for future studies. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.