Abstract
Purpose :
Tractional retinal detachment (TRD), the separation of the neurosensory retina from the retinal pigment epithelium (RPE), is a common consequence of proliferative diabetic retinopathy (PDR). TRD cases are complex and have varied surgical outcomes. In this study, we performed a retrospective observational study to observe trends in TRD surgical outcomes with the goal of aiding in preoperative prognostication.
Methods :
Medical records of 193 patients with concurrent diagnoses of PDR and TRD requiring surgery between December 2016 and September 2021 were reviewed. Macula involvement and vitreous hemorrhage at presentation as well as visual acuity (VA) prior to, 1 month after, and 6 months after surgery were recorded. LogMAR VA was used to calculate the mean, standard deviation, and two-tailed t-tests.
Results :
Two hundred and seventeen eyes from 193 subjects were studied (n = 217). Mean VA was 1.66 logMAR (~20/900) preoperatively, 1.54 logMAR (~20/700) 1-month post-operatively, and 1.40 logMAR (~20/500) 6-months postoperatively. Of those with a preoperative VA of 20/200 or less, 31% achieved a postoperative VA of 20/70 or better, 12% fell between 20/70 and 20/200, and 57% did not exceed 20/200 at 6-month follow up. Seventy-four percent of TRDs involved the macula. A statistically significant difference in preoperative VA between the macula involving and macula sparing groups (p > 0.15) was not observed. However, at 6-month follow-up, 63% of eyes with macula sparing TRDs regained VA of at least 20/70 compared to only 37% of eyes with macula involving TRDs. This suggests that preoperative VA and presence of macula involvement are independent factors that can help prognosticate VA outcome. However, eyes without vitreous hemorrhage (VH) and with preoperative VA of 20/200 or worse were less likely to improve to 20/70 or better versus eyes with VH prior to surgery (14% vs 39%). Therefore, presence of VH may be a confounding factor to consider when using preoperative VA as a prognostication tool.
Conclusions :
TRDs are challenging to repair and have a wide array of outcomes, making preoperative prognosis difficult. This study quantifies the prognostic advantage conferred by a preoperative VA of 20/200 or better and by the presence of macula sparing disease. It also demonstrates the effectiveness of TRD repair as a whole. Further analysis will be aimed at elucidating the presence of other confounding variables.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.