Fovea Sparing ILM Peeling in Diabetic Patients

● Study: Foveal sparing ILM peeling versus complete ILM peeling in diabetic patients indicated for Pars Plana Vitrectomy (comparative study).

● Background and Rationale: .Diabetic retinopathy significantly affects vision nationally and globally. .Proliferative diabetic retinopathy (PDR) is characterized by the growth of new vessels on the retina or the optic disc, leading to vitreous hemorrhage, tractional retinal detachment, rubeosis iridis, and eventually blindness.

.Panretinal photocoagulation has been the treatment for PDR since the 1970s. However, chronic vitreous hemorrhage, tractional, or combined tractional-rhegmatogenous retinal detachments are indications for early vitrectomy.

Some surgeons peel the internal limiting membrane (ILM) after staining to prevent epimacular reproliferation (although not mandatory) aiming for better postoperative visual results. ILM peeling removes the scaffold for ERM proliferation, thereby decreasing recurrence due to traction on the macula.

.However, several potential detrimental effects on the inner retina have been demonstrated after ILM peeling and irregularities in the inner retinal surface.

.It is known that the umbo of the fovea is important for optimum visual function. It is also thought that the Muller cells at the fovea with their extended funnel shape, higher refractive index than the surrounding tissue, and orientation along the direction of light propagation may act as optical fibers to maximize light transmission through the mesh-like inner retina to the photoreceptors.

● Objectives: Investigators aim in this study to compare between foveal sparing ILM peeling and total ILM peeling in diabetic patients indicated for PPV on functional and anatomical outcomes regarding postoperative best corrected visual acuity, measuring central subfield thickness and evaluation of morphology and integrity of inner retinal layers (including vitreoretinal interface) and the outer retinal layers (especially the ellipsoid zone) using macular OCT.

● Study Design: Randomized Clinical Trial

• Study Methods:
• Population of study: Diabetic patients indicated (candidates) for pars plana vitrectomy.
• Study location:

Diabetic patients will be recruited from the retina outpatient clinic at Kasr Al Ainy Hospital, Cairo University.

● Methodology in details: This prospective clinical trial will include 48 eyes of diabetic patients who are candidates for either PPV or combined phaco PPV (in all phakic patients). Preoperatively, all patients will undergo complete ophthalmic examination including uncorrected and corrected visual acuity measurement, slit lamp and dilated fundoscopy.

OCT macula (Optovue, USA) will be done preoperatively if media allows visualization of the macula.

All surgeries will be performed by two experienced vitreoretinal surgeons under local (peribulbar) anaesthesia or general anesthesia.

This study will be performed after obtaining an informed consent from the participants and approval of the study by the Ethical Committee, after explaining the patients' current condition, the surgery needed, benefits, and possible complications of the surgical intervention. Pars-plana vitrectomy (23 G) (Constellation, Alcon, ,USA) will be performed on 48 diabetic patients with advanced diabetic eye disease and the aforementioned inclusion criteria. The steps of vitrectomy will include cataract surgery (in phakic patients), core vitrectomy, removal of the posterior hyaloid membrane, segmentation and delamination of epiretinal membranes, either unimanually using high-speed cutting rates (5000 cuts/min) by the vitreous cutter, or using bimanual techniques for dissection of epiretinal membranes using an end-gripping forceps and curved scissors. After that, removal of residual posterior hyaloid tissue and shaving of the vitreous base will be done, then endo-laser photocoagulation will be performed. Complete ILM peeling from the macular area will be randomly chosen to be done in 24 cases (group I) while foveal sparing ILM peeling (leaving an area of 1 to 2 disc diameter of ILM centered on the fovea) will be done in the other 24 cases (group II). ILM peeling will be performed after removal of the posterior hyaloid face and removal of epimacular traction aided by the use of Brilliant Blue G stain (ILMBLUE; DORC, Zuidland, The Netherlands). The dye will be refluxed on to the macula and aspirated off following 30 seconds of contact time. A macular non contact lens +60D (Resight, Germany) will be used to view the peeling procedure. The ILM will be peeled using a pinch technique. Foveal sparing ILM peeling will be done by starting peeling the ILM in a centripetal manner from periphery towards the center 360 degrees in between the arcades, using 23g ILM forceps (Geuder, Germany) till reaching the circle of foveal sparing ILM peeling (about 1-2 DD around the fovea) then trimming the peeled ILM using the vitreous cutter. Eventually sulfur hexafluoride (SF6) gas 20% will be used as tamponade. All patients will receive postoperative treatment in the form of topical antibiotics (moxifloxacin drops) and topical steroids (prednisolone acetate 1% eye drops). Patient follow-up will be carried out at 1-day, 1- week, 1-month, 3-month, and 6-month intervals. Visual acuity (VA) results, OCT macula (Optovue, USA) at 1,3 and 6 months postoperatively will be done to measure central subfield thickness and to evaluate morphology and integrity of inner retinal layers (including vitreoretinal interface) and the outer retinal layers (especially the ellipsoid zone) then results will be tabulated and compared between the two groups.

● Potential risks: Complications of intraocular surgery

• Intraoperative: iotrogenic tears, retinal detachment and vitreous hemorrhage
• Postoperative: endophthalmitis, glaucoma and intraocular inflammation.
• Confidentiality of data:

Patients' personal data will be kept confidential with no disclosure of their names. Numbers will be used instead of names to refer to the patients.

● Sample size: Calculated total sample size is 48 patients, 24 patients in each group. Using G-Power sample size calculator; with 0.05 alpha error and power of the study 0.80, based on a previous study the post-operative best corrected visual acuity was normally distributed with a pooled standard deviation of 0.31, in CMIP group the mean was 0.6 while in FSIP group was 0.35 accordingly we will need 40 patients to reject the null hypothesis, 20 patients in each group.34 The average number of diabetic patients attending retina clinic with indication for pars plana vitrectomy ranges between 2-5 patients per month and each patient needs to be followed up for 6 months, because of the estimated 20% drop-out rate the total sample will be 48 patients 24 in each group.

Sampling technique: a convenient sample of all diabetic patients who will come to the retina outpatient clinic at Kasr AlAiny Hospital, Cairo University during the study period with the inclusion and exclusion criteria will be assigned into study.

● Statistical analysis: All statistical analyses will be done using IBM SPSS v24.0 statistical software (IBM Corporation, NY, USA). Categorical data will be presented as numbers and percentages. Numerical data will be presented as means ± standard deviations (SD). The chi-square test or the Fisher's exact test will be used to compare between the groups with respect to categorical data. Pairwise analysis will be done to test the normality of variables between the 2 groups and accordingly the student-t-test will be used for normally distributed data and a nonparametric test equivalent to the t-test, will be used in non-normally distributed variables. All p-values will be two-sided. P-values of 0.05 will be considered significant.