Outcomes Based on the Pre-operative Integrity of IS/OS Junction in Idiopathic Epimacular Membrane Surgery

In this observational study, patients with idiopathic EMM, undergoing pars plana vitrectomy with Inner Limiting Membrane (ILM) peeling, at the Ophtalmology Clinic of IRCCS Policlinico San Matteo, are enrolled.

The clinical diagnosis of EMM is based on indirect binocular ophthalmoscopy with a + 90 D no-contact lens (Volk Optical Inc., Ohio, USA) and on the SD-OCT examination (see below) which revealed an increase in macular volume associated with optically hyperreflective epiretinal echoes due to the presence of a membrane.

Since several pathologies may induce changes of the macula bioelectrical responses, and their presence may represent potential confounding factors, patients enrolled in the present study were selected from a large EMM cohort of patients after exclusion of presence of dense lens opacities, ocular hypertension or glaucoma, diabetic retinopathy, drusen and age-related macular disease, other general diseases such as systemic hypertension, diabetes, connective tissue or neurological diseases.

Based on that reported by Falkner-Radler et al. , our patient are divided into 2 groups in relationship to the integrity of IS/OS junction: a group in which an intact IS/OS junction was identified by SD-OCT examination (EMM-I Group) and a group in which a disrupted IS/OS junction was observed by SD-OCT examination, EMM-D Group.

Evaluation of photoreceptor status is performed on the central part of the 1.5 mm of two perpendicular scans centered on the fovea. The intact IS/OS junction is identified as a continuous hyperreflective line ; the disrupted IS/OS junction is identified as interruptions of the normal hyperreflective line. Classification was assessed by the agreement of two different grader. In case of disagreement a third assessment was obtained by another expert grader.

VA, mfERG and SD-OCT were assessed in each enrolled patient at baseline (pre-operative) and after 1, 3 and 6 month of follow-up after surgical treatment for EMM in each enrolled patient.

Surgical treatment

All the patients undergo a 25-gauge pars plana vitrectomy with a solution of 0.15% trypan blue, 0.025% brilliant blue G, 4% polyethylene glycol (MembraneBlue Dual) assisted Epiretinal Membrane/Inner Limiting Membrane (ERM/ILM) removal performed by a single surgeon (GV); no air-fluid exchange is performed during surgery. In phakic eyes, phacoemulsification with intraocular lens implantation was performed at the time of vitrectomy.

Visual acuity evaluation

Best-corrected VA is assessed by the modified ETDRS Charts (Lighthouse, Low vision products, Long Island City, N.Y. USA) and expressed in LogMAR values obtained at the distance of 4, 2, 1 and 0.5 meters.

MfERG recordings

The Espion D310 Profile Multifocal System (Diagnosis UK LTD, Histon, Cambridge, UK) is used to record the mfERG.

In all eyes, mfERGs are binocularly recorded under photopic conditions in normal room illumination. Pupils are maximally pharmacologically dilated with 1% tropicamide to a diameter of 7-8 mm. Binocular mfERG recording is preferred for helping subjects to have a stable target fixation. At every mfERG examination, each patient positively reported that he/she could clearly perceive the cross fixation target. The eye's position is continuously monitored by a fixation camera to track patient gaze.

The cornea is anesthetized with 0.4% oxybuprocaine. MfERGs are recorded bipolarly between an active electrode (Dawson-Trick-Litzkow -DTL- thread electrodes) and a reference Ag/AgCl electrode placed on the temple. A small Ag/AgCl skin ground electrode s placed at the center of the forehead. Interelectrode resistance is less than 5 KOhms.

The multifocal achromatic stimulus, consisting of 61 scaled hexagons, is displayed on a high-resolution liquid crystal displays (LCD), black-and-white 32" (diagonal) monitor (size 70 cm width and 39.5 cm height) with a frame rate of 75 Hz (base period for stimulus presentation is 13.3 ms). The hexagon-shaped stimuli are arranged along several concentric rings and are randomly alternated between black (1 cd/m2) and white (400 cd/m2), with the on-off string of each hexagon determined according to a 14-bit M-sequence. This resulted in a contrast of 99%. The luminance of the monitor screen is 200 cd/m2 and a central fixation red cross is used as target in the center of the stimulation field. The eccentricity of the outermost ring of the 61-hexagon array is about 30° (radius from the fixation point to edge of display).

The m-sequence had 213-1 elements and total recording time is approximately 5 min. Total recording time is divided into 9 segments, with a duration of 30 seconds each. Between segments, the subject is allowed to rest for a few seconds.

The signal is amplified (gain 100.000) and filtered (band pass 10-100 Hz) by LPC2000 Flash Utility software V2.2.1. After automatic rejection of artefacts (and blinks) and application of line current (50 or 60 Hz) interference filter (or Runtime Mains Rejection), and automatic rejection of the bioelectrical responses with noise greater than 1 μV root mean square (RMS), the system software (V6.58.2) provided the 1st order kernel response, that was normalized considering the ratio between the bioelectrical responses and the relative retinal area from which the signals were collected. In the 1st order kernel response, different peaks (called with the letter indicating the negative, N and positive, P polarity and the sequential number of appearance) are identified in a time after the 0 ms corresponding to the presentation of the visual stimuli.

In the analysis of normalized mfERG responses, the investigator considered the peak-to-peak amplitude between N1 and P1 peaks, called response amplitude density (RAD) measured in nanoVolt/degree2 (ɳV/deg2).

MfERG ring analysis is selected to differentiate changes of the bioelectrical responses of the foveal region with respect to the more eccentric macular areas. The investigators analyze the averaged response obtained from two concentric annular retinal regions (rings, R) centered on the fovea. Therefore, the investigators analyze the N1-P1 RADs derived from 0 to 5° (R1) and from 5 to 10° (R2).

SD-OCT assessment

All patients undergo structural SD-OCT scan using Heidelberg Spectralis (version 1.10.4.0, Heidelberg Engineering, Heidelberg Germany) after pupil dilation. The SD-OCT imaging protocol consist of at least 20°x15° volume scans of the macula area with 19 B-scans.

A qualitative analysis is performed in order to evaluate the integrity of the inner segment and outer segment photoreceptoral junction . Quantitative analysis of central retinal thickness (CRT), including the overall, the inner (I-CRT) and outer (O-CRT) CRT, and macular volume (MV) are automatically obtained from the macular map centered to the fovea, through an inbuilt software of Heidelberg Spectralis (version 1.10.2.0).

The research follows the tenets of the Declaration of Helsinki. Each patient signed an informed consent. The study was performed according to our routine clinical practice, thus IRB approval was not required.

Statistical Analysis

The Anderson-Darling and Kolmogorov-Smirnovtests are applied to verify that data were normally distributed. Only one eye is chosen for each patient.

Pearson's correlation is used to assess the relationship between the differences (logarithmic values at 1, 3 and 6 months minus logarithmic values at baseline) in VA and mfERG and SD-OCT data. All statistical analyses is performed using SPSS V.26 (Statistical Package for Social Science IBM), and a p-value less than 0.05 is considered as statistically significant.