Spectral versus time-domain OCT in detecting preoperative epiretinal membranes that accompany macular holes.

Abstract

To compare the sensitivities of spectral-domain optical coherence tomography (SD-OCT) versus time-domain OCT (TD-OCT) in identifying epiretinal membranes (ERM) preoperatively in patients who underwent surgery for full-thickness macular holes (FTMH). This is an interventional retrospective case series of 59 eyes diagnosed with FTMHs who underwent 25-G pars plana vitrectomy with internal limiting membrane peeling between 2009 and 2015. Preoperative OCTs were obtained by SD-OCT (Spectralis, Heidelberg, Germany) or TD-OCT (Stratus, Carl Zeiss Meditec, Dublin, CA, USA). Volume scans were reviewed for ERM accompanying the FTMH. This was compared to indocyanine green-negative staining and intraoperative findings of ERM as the gold standard. Baseline characteristics between the SD-OCT and TD-OCT groups were comparable. Mean duration of postoperative follow-up was 41.4 weeks (±49.0). Of 59 eyes, 33 (55.9%) exhibited an ERM intraoperatively. Four ERMs (SD-OCT group) compared to 12 (TD-OCT group) were not visualized on preoperative OCT (p = 0.003). Sensitivity and specificity of SD-OCT in ERM detection was 79% and 100% compared to 14% and 91% for TD-OCT. Visual acuity improved in both arms (0.5 and 0.3 logMAR units in SD-OCT and TD-OCT, respectively (p = 0.002, 0.0002). We found that SD-OCT was superior to TD-OCT in identifying the presence of ERM preoperatively in patients who underwent macular hole surgery. Since ERMs may decrease the chance of successful pharmacologic vitreolysis, we recommend using SD-OCT over TD-OCT in the evaluation of patients with FTMH to more accurately identify ERMs and allow more comprehensive treatment decisions (pharmacologic versus surgical).