Retinal re-detachments after removal of silicone oil: Frequency and timings in a retrospective clinical study

Objective: To analyse anatomical outcome after removal of silicone oil endotamponade in complicated cases of retinal detachment with reference to frequency and timings, and to study factors affecting the outcome. Methods: The retrospective study was conducted at one public-sector and one private-sector hospital in Karachi, and comprised data related to patients who underwent pars plana vitrectomy with silicone oil endotamponade for complicated retinal detachment and subsequent removal of silicone oil between 1996 and 2015. The surgery had been carried out by both active and passive methods. Outcome variable was retinal re-detachment. Timings and frequency of re-detachment were noted and regression analysis for pre-dictive factors was carried out. Results: The study comprised 355 eyes relayed to as many patients. Of the total, 239(67.30%) were males and 116(32.70 %) were females. Overall mean age was 43.29± 17.15years. Mean duration of silicone oil endotamponade was 5±1.34 months. Indications of injection were retinal detachment complicated by proliferative vitreoretinopathy in 290(81.69%) eyes and proliferative diabetic retinopathy in 65(18.31%). Retinal re-detachment was seen in 51(14.36%) cases. Among them, 15(29.41%) were seen on the table, and 40(78%) within the first 3 months. Duration of silicone oil tamponade, technique of removal, type of silicone oil used and number of previous surgeries had no predictive value on anatomical outcome (p>0.05 each).


Introduction
Silicone oil (SO) has a well-established role in vitreoretinal surgery and is successfully used in the management of complicated retinal detachments (RD). 1,2In order to minimise complications associated with long-term use, SO is removed once retinal status appears stable. 3emoval of silicone oil (ROSO) is associated with serious complications that include re-detachment of retina, hypotony, expulsive haemorrhage, vitreous haemorrhage and unexplained visual loss. 4,5Residual vitreo-retinal traction, especially at vitreous base and re-proliferations, are most likely reasons for retinal re-detachment.][8][9][10][11] ROSO through pars plana by active method has been one of the commonest methods performed.Trans conjunctival sutureless vitrectomy (TCSV) is a safe and effective technique, which makes sutureless active and passive three-port ROSO possible, and obviates the need for conjunctival peritomy and stitching of scleral ports. 12,13e current study was planned to review the frequency and timing of the re-detachment of retina post-ROSO and to discuss the risk factors for this still unresolved complication in vitreoretinal surgery.

Patients and Methods
The retrospective study was conducted at one publicsector and one private-sector hospital in Karachi, and comprised data related to patients who underwent pars plana vitrectomy with silicone oil endotamponade for complicated retinal detachment and subsequent ROSO between 1996 and 2015.These surgeries were performed by a single surgeon at the Department of Ophthalmology, Dow University of Health Sciences, Karachi, and at the private-sector Hashmani Hospital, Karachi.Approval of institutional review boards was obtained after which data was collected.All patients had signed an informed consent form for the procedure and for permission to use data for scientific documentation.Using non-probability convenience sampling, all consecutive cases performed during the study period were included.The sample size calculation was carried out in line with literature. 14,15dical records were reviewed for age, gender, ophthalmic history, preoperative and postoperative visual acuities (VAs), intraocular pressure (IOP), notes of anterior segment and fundus examinations, and intraoperative or postoperative complications.Also noted were aetiology of complicated retinal detachment, number of prior vitreoretinal surgery, duration of tamponade, type of silicone oil used, type of primary pathology, active or passive method of ROSO, timing of re-detachment of retina and final anatomical status of retina at a minimum follow-up of six months.
Indications for performing ROSO were evidence of emulsification of SO (EOSO), persistent increased IOP on full medical treatment, presence of SO in the eye for more than 3-6 months and decreased vision postoperatively due to increasing lens opacities.Data related to patients with complicated retinal detachments secondary to proliferative vitreoretinopathy (PVR) and proliferative diabetic retinopathy (PDR), which are the two conditions commonly dealt with in practice.Patients with other pathologies causing complicated retinal detachments that needed SO injections were excluded.Only-eyed and high-risk patients and those with <5mmHg IOP were not encouraged for ROSO.
Surgeries were performed under general anaesthesia.In phakic eyes with lens opacities, cataract surgery with or without intraocular lens implantation was performed as combined procedure with ROSO.Scan ultrasound was performed to evaluate the status of retina in patient with media opacities prior to surgery.
After localised incision of overlying conjunctiva, 3 sclerotomy ports were made with 20G MVR knife through pars plana, 3-3.5mm from the limbus depending on the eye was phakic, pseudophakic or aphakic.Active aspiration, both for 1000cs and 5000cs SO, was carried out using Associate (D.O.R.C.International, Holland), Pulsar 2 (Optikon, Italy), Accurus/ Consetellation (Alcon Laboratories, Fort Worth, Texas, United States) vitrectomy machines.Binocular ophthalmoscope microscope (BIOM) wide-angle viewing system (Oculus, Inc., Wetzlar, Germany) was used from 2009 onwards.Complete ROSO was ensured with 3-4 spells of fluid-air exchange.Any epiretinal membrane (ERM) seen at the posterior pole as a result of peri-silicone oil fibrosis was removed in the course of the procedure.Vitreous base was inspected with external scleral depressor.Residual vitreous base, if present, was removed with vitreous cutter.Retina was inspected for its anatomical stability, and 2-3 rows of 360degree intraoperative endo-photocoagulation were performed behind visible scars of the previous surgery in all cases as part of the procedure.At the end of the procedure, vitreous cavity was left half-filled with air.Sclerotomy ports and conjunctiva were then closed with 6/0 vicryl.
Passive ROSO was carried out by TCSV technique.For ROSO of 1000cs, SO used was 23G system and, for 5000cs SO, 20G TCSV non-valved cannulae (D.O.R.C.International, Holland) were used.SO egressed through two superior cannulae.Last bubble of SO was made to egress after temporary closure of one of the superior ports temporarily by trocar.The rest of the procedure was not different from what was done during active aspiration method except that at the end of the procedure no stitches were applied to either conjunctival or scleral ports.
In case of re-detachment of retina on table, repeat surgery was carried out that involved removal of visible proliferative tissues, and other necessary procedures including retinotomies and retinectomies (RR) whenever needed.SO was again used during the process of surgery in all the cases.
After ROSO, patients were followed on 1st postoperative day, after a week, after 1 month and then at 3-6 months' interval.On each visit best-corrected visual acuity (BCVA) and IOPs were documented.Anatomical status of retina, whether attached or detached, was noted in all patients.Minimum follow-up period was 6 months.Data was analysed using SPSS 21.Numerical variables were expressed through means and standard deviations (SDs), and categorical variables were expressed by frequencies and percentages.Snellen's VA was converted to logarithm of minimum angle of resolution (Log MAR) for statistical analysis.Student's t-test and Pearson's Chisquare test were used for statistical analysis depending upon the nature of the independent variables.Regression analysis was carried out to study the predictive values of independent variables on the final anatomical status of retina.P<0.05 was considered statistically significant.

Results
The study comprised 355 eyes relayed to as many patients.Of the total, 239(67.30%)were males and 116(32.70%)were females.Overall mean age was 43.29± 17.15 years.Mean follow-up period was 8±2 months.On regression analysis, independent variables did not show statistically significant predictive values (Table -3).

Discussion
7][18] In 1985 Gonvers 16 introduced the concept of SO as a temporary tamponade agent.Most physicians now aim to remove SO from eyes with attached retina with secure chorio-retinal scars without any evidence of traction on the retina on clinical examination. 83,[16][17][18] In the current study, patients were advised ROSO at 3 months.The relationship of re-detachment of retina with the duration of SO tamponade is not clear.Scholda et al. 18 showed relationship of re-detachment to shorter duration of SO tamponade.In other series there was no association between length of oil retention and incidence of recurrent retinal detachment after ROSO. 4,5,19Mean duration of SO tamponade in our series was 5 months and it did not have any predictive value for retinal redetachment (odds ratio [OR] 1.125; Confidence Interval [CI]:0.940-1.346; p0.200).Teke et al. 20 in 2014, Tavares et al. 21in 2015 and Rhatigan et al. 22 in 2018 have reported similar observations.
Reported series in literature do not have a uniform technique for ROSO and vary from manual aspiration through pupil to two-port pars plana infusion and active manual or vitrectomy machine-assisted three-port aspiration. 17,18Two-port and three-port pars plana techniques have been reported to have similar incidence of retinal re-detachments. 23In the present series, ROSO was carried out by both active and passive methods through three-port pars plana surgery as both these methods are commonly used for ROSO. 24,25ROSO by passive method is very cost-effective in resourcerestricted country like Pakistan where reusable instruments are commonly used to reduce cost, especially in repeat surgeries.
The major complication of ROSO is retinal re-detachment, which has reported incidence between 2% and 40%. 10,22he current study encountered re-detachments in 51(14.36%)cases.This is despite the fact that we applied 360° laser application intraoperatively in all cases as a part of the procedure.Laser photocoagulation has been reported to decrease the incidence of re-detachment after ROSO. 5,13,26 Re-detachments seem to be a complication of early postoperative period and are seen in the first 50 days in up to 75% cases. 6Reason for rapidity of re-detachment might be persistent traction that had been counteracted by oil before ROSO. 8,26Risk of re-detachment decreases steeply with increasing time and is unlikely after 3 to 5.5 months. 6,8In the current study, 15/51(29.41%)redetachments were seen on-table during ROSO, and 13(25.49%) in the first month postoperatively.At 3 months this figure increased to 40(78.43%).Seven (13.73%) re-detachments in our series were seen between 2 and 5 years.Time of retinal re-detachment is statistically independent of the method of silicone oil removal, refractive error, VA before silicone oil removal, type of anaesthesia, gender and age of patients. 6 our series age, (p=0.949) and gender (p=0.868) did not have any predictive value on anatomical outcome, and similar to findings by Nagpal et al., 19 we did not find significant predictive value of numbers of previous VR surgeries on the outcome variable (p=0.447).Aetiology of retinal detachment has been reported as risk factor. 20In our series, eyes with PVR (290) were far more than eyes with PDR (65).Although this risk factor has no significant predictive value (p=0.449) in the present study, its interpretation does not reflect our clinical impression of PVR being a higher risk factor.Although the present study pertains to common conditions of PVR and PDR, 5 we believe that these results could be applied to other clinical conditions in which SO is used.This includes complicated retinal detachments caused by trauma, viral retinitis, macular hole in highly myopic eyes, chronic and persistent macular holes, colobomatous retinal detachment, chronic uveitis with hypotony, in patients who have to be airborne and in patients who cannot maintain correct postoperative positioning, such as children, old, and mentally compromised patients. 24,27dditionally, viscosity of SO used did not have any predictive value over anatomical outcome (p=0.763).This finding correlates well with literature. 28 terms of limitations, the retrospective nature of the study.It only reports ROSO in eyes that had SO injected in retinal detachments complicated by PVR and PDR.It is not a detailed comparative study comparing active and passive methods of ROSO.However, large number of cases, uniformity of the surgical skills, no predictive value of different techniques on outcome variable mentioned in the text are strong points for the generalisability of the study.

Conclusions
Most of re-detachments occurred during the first 3 months postoperatively.None of independent variables studied had any influence on the anatomical outcome.
Residual VR traction, especially at the vitreous base, and re-proliferations could be reasons for retinal redetachment.

Table - 2
: Complications in cases of removal of silicone oil (ROSO).

Table - 3
: Regression Analysis Outcome Variable -Retinal Re-detachment in 355 cases of removal of silicone oil (ROSO).