Long-Term Quantitative Analysis of Inner Retinal Dimples and Visual Function Post Internal Limiting Membrane Peeling in Macular Diseases

Abstract

Introduction Inner retinal dimples (IRDs) are frequently detected after internal limiting membrane (ILM) peeling. However, the distribution of IRDs and its effect on postoperative visual function remain unclear. We aim to quantify the distribution of IRDs after ILM peeling in different macular diseases and analyze its influence on postoperative visual function. Methods We retrospectively reviewed patients undergoing vitrectomy with ILM peeling and followed up until 12 months in our center. The distribution of IRDs were quantitatively determined using optical coherence tomography (OCT) and OCT angiography in a different sector of Early Treatment Diabetic Retinopathy Study grid. Visual function was evaluated by retinal sensitivity (RS) using microperimetry. Spearman correlation was performed between RS and IRDs within the same sectors. Multivariate linear regression analysis was performed to analyze the association between baseline characteristics and IRDs. Results A total of 43 idiopathic macular hole (iMH) cases, 56 idiopathic epiretinal membrane (iERM) cases and 42 myopic foveoschisis (mFS) cases were included. IRDs increased gradually at ILM-peeled area, interrupting ganglion cell layer. Most IRDs were observed in temporal sector. A negative correlation was depicted between the increase of IRDs and the progress of RS at both perifovea and parafovea in iERM, but only at perifovea in iMH. No significant correlation between the change of IRDs and RS was found in mFS. Multivariable linear regression model showed that preoperative axial length was significantly associated with postoperative IRDs in all patients. Conclusions IRDs distributed mostly at temporal sector after ILM peeling, interrupting ganglion cell layer. IRD progression may influence postoperative RS only in iMH and iERM. Ophthalmologists may avoid temporal sector especially in eyes with normal axial length or strong ILM–ERM adherence.