Outcomes of Eyes Undergoing Multiple Surgical Interventions after Failure of Primary Rhegmatogenous Retinal Detachment Repair

Purpose: To assess outcomes among eyes undergoing surgery for recurrent rhegmatogenous retinal detachment (RRD) and investigate variables that correlate with visual and anatomic outcomes. Design: Retrospective, consecutive case series. Participants: Eyes in which initial RRD repair failed that underwent ≥2 surgeries for RRD repair between January 1, 2016, and December 31, 2020. Methods: A retrospective analysis of eyes that underwent ≥2 RRD repairs. Eyes with etiologies of retinal detachment other than RRD, previous vitreoretinal surgery, and proliferative diabetic retinopathy were excluded. One-way analysis of variance with Tukey honestly significant difference testing and multivariate regression analyses were used to assess statistical significance. Main Outcome Measures: Anatomic success rate and habitual best recorded visual acuity (VA) at postoperative time intervals and at final follow-up. Results: The single-operation success rate (SOSR) among eyes that underwent primary RRD repair over a 5-year period was 92.4% (2021 of 2187 eyes). The remaining 166 eyes (164 patients) in which primary RRD repair failed were included in this study. Although the anatomic success rates after the second and third RRD repairs were lower than the SOSR (71.7% and 68.1%, respectively), the final anatomic success rates did not significantly differ between eyes that underwent ≥2 RRD repairs (range, 90.6%–100%) (P > 0.05), and final anatomic success was achieved in 95.8% of the eyes. The average VA at both postoperative time intervals and final follow-up decreased with an increased total number of RRD repairs. Eyes that received silicone oil endotamponade during the second RRD repair were 20.3% (P = 0.03) and 38.4% (P = 0.04) less likely to require a third RRD repair compared with eyes that received octafluoropropane and sulfur hexafluoride gases, respectively. Eyes that developed a third RRD due to proliferative vitreoretinopathy (PVR) (n = 43) were 110% more likely to require a fourth RRD repair than eyes that developed a third RRD due to a new tear or failure to reattach (n = 4) (P = 0.04). Conclusions: In this series, the rates of anatomic success appeared consistent after each surgery when multiple reoperations were required for RRD. The VA decreased with an increased total number of RRD repairs, and the endotamponade choice might have affected the risk of additional reoperations. Proliferative vitreoretinopathy was a significant prognostic factor for RRD recurrence.