Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The effect of silicone oil tamponade on retinal layers and choroidal thickness in patients with rhegmatogenous retinal detachment: a systematic review and meta-analysis

The effect of silicone oil tamponade on retinal layers and choroidal thickness in patients with... Background: To evaluate the effects of intravitreal silicone oil (SO) on the retinal and choroidal thickness in eyes with rhegmatogenous retinal detachment (RRD). Methods: A literature search was performed in Web of Science, Scopus, ProQuest, Embase, Clinical Key, Science Direct, Cochrane Library, and Springer, as well as Persian databases, including IranDoc, MagIran, SID, MOH thesis, and MOH articles until June 2020. Two reviewers independently searched and extracted the data. Results: Sixteen studies (n = 391) met the inclusion criteria. The meta-analysis showed that the SO tamponade could significantly reduce the central macular thickness (CMT ) in patients with RRD as compared to gas tamponade WMD = − 14.91; 95% CI: − 22.23, − 7.60; P < 0.001, I = 71%). No significant change was found in CMT between the eye with SO tamponade (after SO removal) and the fellow healthy eye in patients with RRD ( WMD = − 3.52; 95% CI: − 17.63, 10.59; I = 68.6%). Compared to the preoperative stage, the SO tamponade could significantly reduce the subfoveal choroidal thickness in patients with RRD ( WMD = − 18.67, 95% CI: − 30.07, − 1.28; I = 80.1%). However, there was no significant difference in the subfoveal choroidal thickness before and after SO removal ( WMD = − 1.13, 95% CI: − 5.97, 3.71; I = 87.6%). Conclusion: The SO tamponade had a significant effect on the reduction of retinal layers and the subfoveal choroidal thickness. Keywords: Rhegmatogenous retinal detachment, Silicone oil, Retinal layer thickness, Choroidal thickness Background Retinal detachment (RD) is one of the leading causes of permanent vision loss. It can be classified as rhegma - togenous, tractional, or exudative. Rhegmatogenous retinal detachment (RRD) is the most common type of *Correspondence: [email protected] RD, occurring in one per 10,000 people each year. Reti- Isfahan Eye Research Center, Isfahan University of Medical Sciences, nal surgery is increasingly used for the repair of RRD, Isfahan, Iran and recent developments in techniques and instruments Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 2 of 14 have improved the outcomes [1, 2]. Silicone oil (SO) and a study by Karimi et  al. showed that the subfoveal cho- gas are two of the most commonly used tamponading roidal thickness (SFCT) was reduced in patients receiv- agents in vitreoretinal surgery, each with its own advan- ing SO tamponade [16]. Conversely, a study by Zhou tages and disadvantages. Although there is a consensus et  al. reported a decrease in the retinal layer thickness, on what necessitates gas or SO tamponade, they can be while no significant change was observed in the choroidal still selected at the discretion of the surgeon or based on thickness [17]. regional differences [3]. Studies have been conducted regarding the effect of SO The long-term SO tamponade complications include tamponade on retinal and choroidal thickness in recent secondary glaucoma, cataract, optic neuropathy, band years. However, there is no systematic review of studies keratopathy, and retinal toxicity [4]. Mild vision loss may investigating the effect of SO tamponade on retinal lay - occur after SO removal, usually due to postoperative cys- ers. Therefore, in the present study, we systematically toid macular edema (CME), hypotony, optic neuropathy, reviewed the effect of SO tamponade on the central mac - and development of epiretinal membranes; neverthe- ular thickness (CMT), different retinal layers, and choroi - less, there are some cases of visual loss that do not have dal thickness. a certain cause [4–6] The possible mechanisms include macular dysfunction, mostly of ganglion and bipolar cell Methods synapses, and optic nerve damage due to SO infiltration Search strategy [5]. In this systematic review, we unanimously approved the Optical coherence tomography shows that long-term main concepts using the PICO format (Table 1) Based on SO tamponade may have a negative effect on the thick - the PICO search format, a free search strategy was used ness of retinal layers in some patients [7, 8]. Also, some to perform a more comprehensive and accurate search of researchers have shown that SO tamponade may have a information sources in different databases. Also, in this negative long-term effect on retinal microcirculation [9, study, all published articles were included without any 10]. Besides, in RRD patients without macular involve- time limits. A search of the following international data- ment, the prevalence of unexplained vision loss is higher bases was performed until June 2020: Web of Science, after SO tamponade as compared to gas tamponade [8, PubMed, Scopus, ProQuest, Clinical Key, Science Direct, 11].Recent studies also suggest that visual abnormalities, Embase, Cochrane Library, and Springer, as well as Per- such as decreased foveal sensitivity, central scotoma, and sian databases, such as MagIran, SID, IranDoc, MOH macular dysfunction, may be related to SO tamponade thesis, and MOH article databases. The included studies [12, 13]. were clinical trials, clinical interventions (before-and- Recent studies examining the effect of retinal and cho - after analysis), and cohorts that met the inclusion criteria roidal layers have reported controversial results, includ- after screening. ing a study by Goker et al., which found that the thickness The inclusion criterion in this systematic review was of the inner nuclear layer (INL) and outer plexiform layer all studies investigating the effect of SO tamponade on (OPL) increased in the eyes with SO tamponade in com- retinal and choroidal layers in patients with RRD. There parison with the fellow healthy eyes [14]. However, in was no age, sex, magazine type, or publication limit, and a study by Lee et  al. on patients with RD, it was found all articles meeting the inclusion criteria were evalu- that in the SO tamponade group, there was a marked ated. The exclusion criteria were as follows: a macular reduction in the inner and outer retinal layer thickness hole, a macular pucker, or other maculopathies; RRD compared to the gas tamponade group [15]. Moreover, with high myopia, uveitis, glaucoma, macular edema, or Table 1 Electronic database search terms Synonyms Keywords “Retinal detachment” OR “Retinal detachment surgery” OR “Rhegmatogenous retinal detachment” Retinal detachment Population/disease “Silicone oil” OR “Silicone oil endotamponade” OR “Endotamponade” OR “Intravitreal silicone oil” Silicone oil tamponade Intervention OR “Intravitreal tamponade” OR “Intraocular tamponade” OR “Silicone oil removal” “Gas tamponade” AND “Fellow eye” Gas tamponade and fellow eye Comparison “Retinal layer thickness” OR “Retinal thickness” OR “Macular layers” OR “Macular layer thickness” OR Retinal layers Consequences “Central macular thickness” OR “Central retinal thickness” OR “Retinal nerve fiber layer” OR “Gan- glion cell layer” OR “Inner plexiform layer” OR “Outer plexiform layer” OR “Inner nuclear layer” OR “Outer nuclear layer” OR “Rod and cone inner and outer segments (IS/OS)” OR “Inner retinal layers” OR “Outer retinal layers” OR “Choroidal thickness” OR “Subfoveal choroidal thickness” Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 3 of 14 postoperative endophthalmitis; traumatic RD; and other plot asymmetry was performed by Egger’s and Begg’s optical nerve diseases. Also, case reports and systematic regression tests [23]. Both the funnel plot and the Egger reviews were excluded. and Begg tests do not have good power to assess publi- cation bias in meta-analysis with less than 10 studies. It Data extraction and quality assessment tried to do the bias analysis, but it doesn’t have enough After searching for studies based on the keywords, eligi- studies. When there was publication bias, the trim-and- ble articles were extracted from the scientific databases. fill method was applied to detect the contribution of bias The flowchart of the articles was plotted by one of our to the overall effect. Statistical analyses were performed colleagues (A.R.). In the first stage, by reviewing the title in Stata version 11 (Stata Corp., College Station, TX, and abstract of studies, a number of unrelated and dupli- USA). P-values less than 0.05 were considered statisti- cate studies were eliminated. In the next step, a detailed cally significant. evaluation of the full-text of the remaining articles was performed. All steps were performed by two independent Results researchers (H.Gh. and A.R.J.), and the third researcher Search results and study characteristics (F.K.) supervised and studied their findings [18–20]. The As shown in Table  2, the initial search yielded a total of following data were also extracted from eligible studies 563 publications. Based on the inclusion and exclusion by two researchers (H.Gh. and A.R.J.): the first author’s criteria, 16 articles were found to be eligible: five studies name, publication date, sample size, age, the duration of had arms only evaluating the CMT [24–29], three studies SO tamponade dressing, time spent after SO removal, only evaluating the inner retinal layer thickness [7, 8, 28]. type of RRD (on/off ), CMT, retinal layer segmentation five studies evaluating both CMT and retinal layer seg - thickness, and SFCT. The quality of studies was also mentation thickness [14, 15, 17, 30, 31], and three stud- assessed by completing the CONSORT checklist for ies only evaluating the SFCT [16, 32, 33].In the search of clinical trials and STROBE checklist for observational Persian databases, no relevant studies were found. Fig- studies; based on the score that each study received, we ure 1 presents the flow chart of the selection process used decided to remove or include it. to identify studies. Out of 16 articles reviewed, the findings of 11 articles Data synthesis and statistical analysis were meta-analyzed [13–18, 20, 22–25]. The findings of Two rounds of data analysis were conducted. For all the other five articles are presented only quantitatively included studies, whether they had a control group or due to the lack of necessary information in the statistical not, the effect of SO was evaluated by comparing the out - analysis [5, 6, 19, 21, 26]. Of the selected articles, 13 stud- comes before and after the intervention. Also, for studies ies were retrospective, two were prospective, and one was with a control group, the mean difference in the out - cross-sectional. A total of 391 patients were included in comes before and after the intervention was calculated these studies, and their mean age was 55.36 years. Table 2 for each group and compared between the SO group summarizes the characteristics of the included studies in and the control group. Moreover, the effect size was this systematic review. estimated as the weighted mean difference (WMD) or standardized mean difference (SMD) and 95% confidence Meta‑analysis and data synthesis interval (95% CI). Heterogeneity between studies was Eec ff t of SO tamponade on the retinal layer thickness 2 2 evaluated by Cochran’s Q and I-square (I ) tests; I above There are two types of articles on the effect of SO tam - 50% was regarded as substantial heterogeneity [21]. ponade on the retinal layers. Some studies only examined To calculate the effect sizes and their correspond - the effect of SO tamponade on the CMT or inner retinal ing 95% CIs, a fixed effects model was adopted when layers, while some studies, in addition to CMT, exam- heterogeneity was low. On the other hand, a random ined the effect of SO tamponade separately on the retinal effects model was used in case of moderate or high het - layers. erogeneity, considering the between-study variations in the DerSimonian and Laird method [22]. To explore the Eec ff t of SO tamponade on CMT: (Fig. 2a–o) source of heterogeneity, a meta-regression analysis was We examined the effect of SO tamponade on CMT in performed to evaluate the confounding role of age and several sections: duration of SO tamponade. Besides, sensitivity analyses were performed to evaluate the extent to which infer- Comparison of CMT during SO tamponade application ences might be related to a particular study. Publication and after SO removal bias was also examined by the visual inspection of fun- In this comparison, due to the small number of arti- nel plots [23]. Moreover, a statistical evaluation of funnel cles and incomplete information, no meta-analysis was Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 4 of 14 Table 2 The general characteristics of the included studies in the systematic review First author Publication year Quality of study Age (mean) Sample size Type of Duration of SO Time after CMT Inner Retinal layer SFCT RRD (on/ tamponade SO removal retinal layer segmentation off ) (month) (month) thickness thickness Hostovsky [24] 2020 Good 57.7 SO = 20 Off 1.5 -2.5 6 * Rabina [27] 2019 Excellent 56.1 SO = 41 On 3- 6 1 * Takkar [25] 2017 Good 39.6 SO = 32 Off 3–6 6–9 * Roohipoor [26] 2020 Excellent 54 SO = 45 Off 3 – * Kheir [29] 2018 Good 48.1 SO = 10 On 3 9 * Tode [28] 2016 Fair 61 SO = 15 On 1.5 36 * Caramoy [34] 2014 Fair 67 SO = 9 On 5 – * Christensen [8] 2011 Fair 51 SO = 9 On 5 46 * Gas = 7 Inan [31] 2016 Good 60.7 SO = 28 Off 4 8 * * Gas = 30 Goker [14] 2017 Good 56.35 SO = 20 Off 4 2 * * Gas = 16 Lee [15] 2017 Good 46.83 SO = 33 On 3.5 6 * * Gas = 31 Purtskhvanidze [30] 2017 Good 59.3 SO = 20 Off 5.5 35 * * Gas = 20 Zhou [17] 2020 Good 53.86 SO = 7 On 4 – * * Gas = 14 Mirza [32] 2018 Fair 60.4 SO = 24 Off 3–6 1 * Karimi [16] 2018 Good 52 SO = 60 Off 6–9 3 * Odrobino [33] 2016 Good 62.44 SO = 18 Off 6 - * A scoring scale for this study based on Strobe modified (≥ 85 = Excellent, 70 to < 85 = Good, 50 to < 70 = Fair, < 50 = poor) Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 5 of 14 563 arcles idenfied in total 521 arcles excluded on tle and abstract review 42 arcles retrieved for fulltext review 26 arcles excluded due to not meeng inclusion criteria 16 arcles sasfied all inclusion criteria 5 arcles evaluated 5 arcles evaluated 3 arcles evaluated 5 arcles evaluated Renal layers segmentaon Inner renol layers Choroidalthickness Central macular thickness Fig. 1 The results of the literature search strategy. The flow diagram depicts the screening process of articles and the number of exclusions Comparison of CMT after SO removal with CMT after gas conducted, and the data were only qualitatively exam- tamponade (Fig. 2a) ined. This section includes three articles. The results Four studies with five effect sizes [14, 15, 17, 30] exam- of a study by Hostovsky [24] showed that during SO ined the effect of SO tamponade on the CMT after tamponade application, the CMT was significantly SO removal compared to gas tamponade in patients lower, with a thickness recovery after SO removal (from 248 ± 48.88 to 277 ± 46.46 μ; P = 0.001). There with RRD. The SO tamponade could significantly was no significant difference in the CMT between reduce the CMT (WMD = −  14.91; 95% CI: −  22.23, the eye with SO tamponade and the fellow eye in the −  7.60; P < 0.001, I = 71%) compared to gas tampon- last follow-up after the SO removal (278 ± 49.40 and ade in patients with RRD, with substantial heteroge- 264 ± 34.92 μ, respectively; P = 0.265). neity between studies (Fig.  2a). The meta-regression Another study by Rabina [27] showed that the mean analysis showed that the outcomes were not dependent CMT in the operated eye increased from 249 ± 50 µ on age (P = 0.632) or the time spent after SO removal during SO tamponade application to 279 ± 48 µ after (P = 0.328). However, they were dependent on the dura- SO removal (P = 0.001). Also, the CMT was 281 ± 21 tion of SO tamponade application (P = 0.026). Although µ in the fellow healthy eye. The results showed the Egger’s (P = 0.094) and Begg’s (P = 0.221) regression apparent thinning of retinal thickness, mainly in the tests did not show significant publication bias, substan - inner layers, during SO tamponade application, which tial asymmetry was seen in the funnel plot. Accordingly, resolved after SO removal. However, a study by Kheir we performed a trim-and-fill analysis and found that the WJ [29] reported that the mean CMT decreased from CMT reduced significantly in the SO tamponade group a baseline of 255.7 µ to 247.5 µ during SO tampon compared to the gas tamponade group (WMD = − 14.91; 95% CI: − 22.23, − 7.60; P < 0.001, I = 71%). The sensitiv - ade application, which was followed by an increase ity analysis showed that exclusion of a study did not affect to 262.1 µ after SO removal; however, these changes the results obtained. were not statistically significant (P = 0.44). B esides , Another study by Christensen UC, comparing the quantitative thinning of only the inner retinal layers was observed. Based on these three studies, it seems SO tamponade group with the gas tamponade group, that SO tamponade causes a transient decrease in the was not included in our meta-analysis for two rea- CMT, mainly in the inner layers. After SO removal, the sons. First, in this study, only the inner retinal layers CMT resembled that of the fellow eye. were examined. Second, the unit of measurement was Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 6 of 14 Fig. 2 Eec ff t of SO (silicone oil) tamponade on retinal layers in patients with rhegmatogenous retinal detachment (RRD). SO silicone oil, RRD rhegmatogenous retinal detachment, CMT central macular thickness, NFL nerve fiber layer, GCL ganglion cell layer, IPL inner plexiform layer, INL inner plexiform layer, OPL outer plexiform layer, ONL outer nuclear layer, RPE retinal pigment epithelium reported in pixels, without mentioning the standard regarding CMT (WMD = − 3.52; 95% CI: − 17.63, 10.59; deviation. In this study, thinning of the inner retinal I = 68.6%), with substantial heterogeneity between stud- layers was observed in the SO-operated eyes (5148 ies (Fig.  2b). The meta-regression analysis showed that pixels) compared to gas-operated eyes (6897 pixels) the results were not dependent on age (P = 0.214), the (P = 0.002) [8]. time spent after SO removal (P = 0.346), or duration of SO tamponade application (P = 0.992). Although Egger’s Comparison of CMT after SO removal with the fellow healthy (P = 0.338) and Begg’s (P = 0.060) regression tests did eye (Fig. 2b) not show significant publication bias, substantial asym - Six studies [14, 24, 25, 27, 30, 31] examined the effect metry was observed in the funnel plot. The adjusted of SO tamponade on CMT in the eyes of patients with value, based on the trim-and-fill method, showed no RRD compared to the fellow healthy eyes. No signifi - significant change in the eyes with SO tamponade (after cant change was found in the eyes with SO tamponade SO removal) in comparison with the fellow healthy eyes (after SO removal) as compared to the fellow healthy eyes (WMD = − 7.83; 95% CI = − 22.15, 6.48; P = 0.284). The Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 7 of 14 Fig. 2 continued sensitivity analysis showed that exclusion of a study did one study examining CMT, which was not included not affect the results obtained. in our meta-analysis due to heterogeneity. In the first study by Tode [28], significant thinning of foveal and parafoveal nerve fibers, ganglion cells, and inner plex - Comparison of the inner retinal layer thickness between eyes iform layer (IPL) was observed in the affected eyes with SO tamponade and the healthy fellow eyes (mean: 58.3 ± 13 μm) as compared to the healthy fellow This section includes three studies, with two studies eyes (mean: 84.5 ± 12.3  μm) (P < 0.01) [28]. Moreover, only examining the inner retinal layer thickness and Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 8 of 14 Fig. 2 continued in a study by Caramoy, the inner retinal layers became Eec ff t of SO tamponade on retinal layer segmentation thinner after the use of SO-based endotamponade In this section, the effect of SO tamponade on differ - (retinal layer volume: 1.127 ± 0.160 mm ) compared to ent layers of the retina was compared after SO removal the fellow healthy eye (1.363 ± 0.150 mm ) (P = 0.012) with gas tamponade and the fellow healthy eye. [34]. Besides, the results of these two studies showed a decrease in the inner retinal layers in eyes with SO tam- Nerve fiber layer (NFL) (Fig. 2c, d) ponade as compared to the fellow healthy eyes. Moreo- Four studies with five effect sizes [14, 17, 30, 31] exam- ver, a study by Roohipoor et  al. showed that CMT was ined the effect of SO tamponade on NFL in patients significantly reduced after SO tamponade removal with RRD after SO removal compared to gas tampon- compared to the fellow healthy eye (P = 0.002) [26]. ade. No significant change was found in NFL between Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 9 of 14 SO tamponade (after SO removal) and gas tamponade removal (WMD = − 0.43; 95% CI: − 1.61, 2.46; I = 0.0%) in patients with RRD (WMD = −  0.92; 95% CI: −  2.11, and the fellow healthy eyes (Fig. 2h); there was no hetero- 0.27; I = 12.6%), and there was no heterogeneity between geneity between studies. The results of Egger’s and Begg’s studies (Fig.  2c). Although Egger’s (P = 0.068) and Begg’s regression tests revealed no evidence of publication bias (P = 0.066) regression tests did not show significant pub - in comparison with gas tamponade (Egger’s test P = 0.803 lication bias, substantial asymmetry was observed in the and Begg’s test P = 1.000) and the fellow healthy eyes funnel plot. The adjusted value, based on the trim-and- (Egger’s test P = 0.603 and Begg’s test P = 0.296). The sen - fill method, showed no significant change in the effect sitivity analysis showed that exclusion of a study did not of SO tamponade on NFL in patients with RRD after SO affect the results obtained. removal compared to gas tamponade (SMD = −  0.220; 95% CI = −  0.625, 0.185; P = 0.287). The results of Egg - Inner nuclear layer (INL) er’s (P = 0.068) and Begg’s (P = 0.066) regression tests Four studies with an effect size of five [14, 17, 30, 31] revealed no evidence of publication bias. The sensitivity examined the effect of SO tamponade on NFL in patients analysis showed that exclusion of a study did not affect with RRD after SO removal compared to gas tampon- the results obtained. ades. No significant change was found in INL between Three studies [14, 30, 31] examined the effect of SO the eyes with SO tamponade (after SO removal) and gas tamponade on NFL in patients with RRD compared to tamponade in patients with RRD (WMD = 0.32; 95% the fellow healthy eye. The SO tamponade could signifi - CI: −  1.57, 2.22; I = 31.8.0%); there was no heterogene- cantly increase NFL in patients with RRD compared to ity between studies (Fig.  2i). The results of Egger’s test the fellow healthy eye (WMD = 1.86; 95% CI: 0.65, 3.06; (P = 0.279) and Begg’s test (P = 0.221) revealed no evi- I = 0.0%); there was no heterogeneity between studies dence of publication bias. (Fig.  2d). The results of Egger’s (P = 0.560) and Begg’s Three studies [12, 28, 29] examined the effect of SO (P = 1.000) regression tests revealed no evidence of publi- tamponade on INL in patients with RRD compared to cation bias. The sensitivity analysis showed that exclusion the fellow healthy eyes. The SO tamponade could sig - of a study did not affect the results obtained. nificantly increase the INL in patients with RRD com - pared to the fellow healthy eyes (WMD = 5.51; 95% CI: Ganglion cell layer (GCL) (Fig. 2e, f ) 2.88, 8.15; I = 0.0%); there was no heterogeneity between Three studies [12, 27, 28] examined the effect of SO tam - studies (Fig.  2j). The results of Egger’s (P = 0.629) and ponade on GCL in comparison with gas tamponade and Begg’s (P = 1.000) regression tests revealed no evidence the fellow healthy eyes. The SO tamponade could sig - of publication bias. The sensitivity analysis showed that nificantly reduce GCL in patients with RRD compared to exclusion of a study did not affect the results obtained. gas tamponade (WMD = − 3.41, 95% CI: − 5.58, − 1.25; I = 0.0%); there was no heterogeneity between studies Outer plexiform layer (OPL) (Fig. 2I) (Fig.  2e). Also, no significant change was found in the Four studies with an effect size of five [12, 15, 28, 29], operated eyes after SO removal (WMD = 0.02, 95% CI: as well as three studies [12, 28, 29], examined the OPL −  5.19, 5.23; I = 70.6%) compared to the fellow healthy thickness changes after SO removal in comparison with eyes; there was moderate heterogeneity between studies gas tamponade and the fellow healthy eyes, respec- (Fig. 2f ). The results of Egger’s and Begg’s regression tests tively. No significant change was found in OPL after SO revealed no evidence of publication bias in comparison removal compared to gas tamponade (WMD = −  0.77; with gas tamponade (Egger’s test P = 0.589 and Begg’s 95% CI: −  4.58, 3.05; I = 65.8%), with substantial het- test P = 1.000) and the fellow healthy eyes (Egger’s test erogeneity between studies (Fig.  2k), or with the fel- P = 0.296 and Begg’s test P = 0.629). The sensitivity anal - low healthy eyes (WMD = 0.40; 95% CI: −  5.17, 5.97; ysis showed that exclusion of a study did not affect the I = 79.5%), with substantial heterogeneity between stud- results obtained. ies (Fig.  2l). The results of Egger’s and Begg’s regression tests revealed no evidence of publication bias in com- Inner plexiform layer (IPL) (Fig. 2g, h) parison with gas tamponade (Egger’s test P = 0.224 and Three studies [12, 27, 28] examined the effect of SO tam - Begg’s test P = 0.221) and the fellow healthy eyes (Egger’s ponade on IPL in comparison with gas tamponades and test P = 0.147 and Begg’s test P = 0.296). The sensitivity the fellow healthy eyes. The SO tamponade could sig - analysis showed that exclusion of a study did not affect nificantly reduce the IPL in patients with RRD as com - the results obtained. pared to gas tamponade (WMD = − 2.50, 95% CI: − 4.62, − 0.39; I = 0.0%) (Fig.  2g). No significant change was found between the eyes with SO tamponade after SO Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 10 of 14 Outer nuclear layer (ONL) (Fig. 2m, n) trim-and-fill analysis method and found that the SFCT Three studies [12, 28, 29] examined the ONL thick- reduced significantly in the eyes with SO tamponade ness changes after SO removal in comparison with gas compared to the preoperative period (WMD = − 24.077; tamponade and the fellow healthy eyes. No significant 95% CI = −  41.360, −  6.795; P = 0.006). The sensitivity change was found in the ONL after SO removal, either analysis showed that exclusion of a study did not affect compared to gas tamponade (WMD = −  3.09; 95% CI: the results obtained. −  15.38, 9.20; I = 74.7%) (Fig.  2m) or the fellow healthy eyes (WMD = − 8.58; 95% CI: − 24.27, 7.10; I = 89.5%), with substantial heterogeneity between studies (Fig.  2n). Comparison of SFCT during SO tamponade application Comparison of SO tamponade with gas tamponade after and after SO removal (Fig. 3b) removal by Egger’s (P = 0.502) and Begg’s (P = 1.000) Two studies with an effect size of four [16, 32] examined regression tests revealed no evidence of publication bias. and compared the SFCT changes during and after SO In comparison of SO removal with the fellow healthy application. No significant change was found in the SFCT eyes, although the results of Egger’s (P = 0.758) and after SO removal, compared to the time of SO tampon- Begg’s (P = 1.000) regression tests did not indicate sig- ade application (WMD = −  1.13, 95% CI: −  5.97, 3.71; nificant publication bias, substantial asymmetry was I = 87.6%); there was substantial heterogeneity between seen in the funnel plot. The adjusted value, based on studies (Fig.  3b). Our meta-regression analysis showed the trim-and-fill method, showed no significant change that the results were not dependent on age (P = 0.039) (SMD = − 0.483; 95% CI = − 1.440, 0.475; P = 0.323). The or duration of SO tamponade application (P = 0.662). sensitivity analysis showed that exclusion of a study did The results of Egger’s regression test (P = 0.002) and the not affect the results obtained. funnel plot revealed evidence of publication bias, but the adjusted value, based on the trim-and-fill method, Retinal pigment epithelium (RPE) (Fig. 2o) showed no significant change after SO removal compared Two studies [14, 31] examined the RPE thickness changes to the time when SO was removed (WMD = −  3.711; after SO removal compared to gas tamponade and the 95% CI = − 10.045, 2.622; P = 0.251). The sensitivity anal - fellow healthy eye. No significant change was found in ysis showed that exclusion of a study did not affect the the RPE after SO removal, either compared to gas tam- results obtained. ponade (WMD = 0.43; 95% CI: −  1.17, 2.03; I = 0.0%) (Fig.  2o) or the fellow healthy eyes (WMD = 0.62; 95% CI: −  0.86, 2.11; I = 0.0%); there was no heterogene- Comparison of SFCT after SO tamponade removal ity between studies (Fig.  2p). Publication bias was not with the fellow healthy eyes (Fig. 3c) reviewed due to the small number of studies. Two studies with an effect size of four [32, 33] examined the effect of SO tamponade on the SFCT after SO tam - Eec ff t of SO tamponade on SFCT (Fig. 3a–c) ponade removal and compared it with the fellow healthy There were three studies regarding the effect of SO tam - eyes. After SO removal, the SFCT significantly reduced ponade on SFCT, which are described in the following in patients with RRD compared to the fellow healthy eyes three sections. (WMD = − 34.54; 95% CI: − 50.74, − 18.34; I  = 38.4%); there was no heterogeneity between studies (Fig. 3c). The meta-regression analysis showed that the results were Comparison of SFCT before and during SO tamponade not dependent on age (P = 0.855) or duration of SO tam- application (Fig. 3a) ponade application (P  =  0.716). Although the results of Two studies with an effect size of four [32, 33] exam- Egger’s (P  =  0.694) and Begg’s (P  =  0.308) regression ined the effect of SO tamponade on SFCT and compared tests did not show significant publication bias, substantial the results before and during SO application. Com- asymmetry was observed in the funnel plot. The adjusted pared to the preoperative period, the SO tamponade value, based on the trim-and-fill method, showed that could significantly reduce SFCT in patients with RRD the SFCT reduced significantly after SO removal com - (WMD = −  18.67; 95% CI: −  30.07, − 1.28; I = 80.1%), pared to the fellow healthy eyes (VMD  =  −  34.54; 95% with substantial heterogeneity between studies (Fig.  3a). CI: −  50.74, −  18.34; P  <  0.001). The sensitivity analysis The meta-regression analysis showed that the results showed that exclusion of a study did not affect the results were not dependent on age (P = 0.743) or duration of obtained. SO tamponade application (P = 0.730). Although Egger’s (P = 0.711) and Begg’s (P = 0.734) regression tests did not show significant publication bias, substantial asymmetry was observed in the funnel plot. Accordingly, we used the Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 11 of 14 a. Effect of SO tamponade on subfoveal choroidal thickness (SFCT) compared to the preoperative period in patients with RRD b. Comparison of SFCT during SO application and after SO removal in patients with RRD c. The SFCT after SO removal in patients with RRD compared to the fellow healthy eyes Fig. 3 Eec ff t of SO tamponade on SFCT in patients with RRD. SO silicone oil, RRD rhegmatogenous retinal detachment, SFCT subfoveal choroidal thickness Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 12 of 14 Discussion is almost similar to that of the fellow healthy eye. Based For the first time, the results of this systematic review on this comparison, age, duration of SO application, and and meta-analysis revealed that in patients with RRD, time spent after SO removal had no effects on the out - the SO tamponade reduced CMT, and after SO removal, comes [14, 24, 25, 27, 30, 31]. the retinal layer thickness almost returned to its range In our meta-analysis, CMT was compared between before surgery. Also, the SO tamponade in these patients patients after SO removal and patients with gas tampon- reduced SFCT, which remained reduced following the ade. The results showed that CMT significantly reduced SO removal. Overall, the SO tamponade seems to reduce after SO removal compared to gas tamponade. This CMT. In this regard, studies by Hostovsky (P = 0.001) and reduction in CMT was not related to age or the time Rabina (P = 0.001) showed that the SO tamponade caused spent after SO removal, but was related to the duration a significant reduction in CMT, which returned to the of SO application; in other words, the longer the SO tam- normal range after SO removal [24, 27]. Another study ponade was applied, the more CMT was reduced [14, 15, by Kheir WJ, (P = 0.44) showed that CMT decreased 17, 30]. The decrease in CMT after SO removal, com - when the SO tamponade was applied, but increased after pared to gas tamponade, can be explained by the fact that SO removal; however, these changes were not significant in some studies on RRD patients with gas tamponade, a [29]. Moreover, in two studies by Tode et al. and Caramoy long-term increase in NFL, GCL, and IPL was reported et al., the thickness of the inner retinal layers significantly compared to the fellow healthy eyes. It was assumed reduced in the presence of SO tamponade as compared that the retinal nerve fiber layer (RNFL) edema may be to the fellow healthy eyes [28, 34]. a result of surgery itself or may be related to the relative Previous studies have revealed that CMT decreases retinal ischemia associated with RRD [22, 35]. On the significantly in the presence of SO tamponade, especially other hand, in the SO group, these layers showed relative in the inner retinal layers. The causes of retinal thinning atrophy in the presence of SO tamponade. associated with the use of SO tamponade have not been It seems that SO induces different effects on differ - fully elucidated. However, several factors may be involved ent layers of the retina. Inan et  al. reported thinning of in this phenomenon. One of these factors is the mechani- GCL, OPL, and ONL in the SO group compared to the cal stress of SO tamponade on the retinal layers in the group with gas tamponade [31]. Moreover, Goker et  al. macula [35]. Among other mechanisms, the subretinal reported that in the gas tamponade group, only INL was migration of SO may induce inflammation, resulting in significantly increased. In the SO tamponade group, the the thinning of the retina [36, 37]. Also, emulsified SO INL and OPL thicknesses significantly increased, and the may induce an internal limiting membrane defect and be ONL thickness significantly decreased compared to the toxic to the retina, causing retinal thinning when enter- fellow healthy eyes [14]. Purtskhvanidze et  al. showed ing the intraretinal space [38]. Another explanation for that GCL and IPL were significantly thinner in the SO the thinning of the retina in the presence of SO tam- group compared to the gas tamponade group [30]. More- ponade is that the hydrophobic SO tamponade replaces over, Lee SH et  al. reported that in the SO group, there the natural environment of the hydrophilic vitreous cav- was a significant decrease in the thickness of all retinal ity. This waterproofing effect causes retinal dehydration layers, except for the photoreceptor layer after primary and reduces the thickness of retinal layers, especially RD surgery [15]. the inner layers [27]. Besides, the indirect mechanisms In the study of changes in the thickness of different include changes in the concentration of potassium retinal layers after SO removal in comparison with the because of failure in potassium siphoning and changes in fellow healthy eye and gas tamponade, the results of our cytokine levels, which may induce apoptosis and result in meta-analysis revealed that the thickness of NFL and the thinning of retinal layers [5]. INL increased significantly in patients after SO removal It seems that CMT is not significantly different compared to the fellow healthy eyes, but no significant between the operated eye and the fellow healthy eye after changes were observed in the GCL, IPL, ONL, OPL, SO removal. In studies by Rabina et al. [27] and Purtskh- or RPE. On the other hand, in patients with RRD after vanidze et  al. [30, 38], CMT was not significantly differ - SO removal, the thickness of GCL and IPL significantly ent from the fellow healthy eye after SO removal, while reduced compared to patients for whom gas tamponade in a study by Takker et  al. [25], CMT was significantly was used, while the thickness of NFL, INL, OPL, ONL reduced compared to the fellow healthy eye after SO and RPE did not change significantly. removal. Our meta-analysis showed that CMT was not The increase in NFL thickness after SO removal can significantly different from the fellow healthy eye after be related to surgical NFL edema that may persist for SO removal. In other words, the SO tamponade causes a several months after vitrectomy [31]. The increase in transient decrease in CMT, and after SO removal, CMT INL thickness can be justified by the fact that Müller Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 13 of 14 Authors’ contributions cells are the radial glial cells of the retina, and their HG, FK and AD contributed to the conception and design of the review. AJM, nuclei are located in the INL. The Müller cells play AR, AF and ANB contributed to the drafting and critical revision of the manu- an important role in structural restoration after reti- script. All authors read and approved the final manuscript. nal reattachment; therefore, increasing their activity Funding can increase the thickness of INL [14]. Several studies The author declares that no funding or research Grant was obtained for this have reported the degeneration of ganglion cells in the manuscript. detached retina [39]. In this regard, a rabbit model of Availability of data and materials RD showed the progressive loss of ganglion cell axons Not applicable. in the detached retina [40]; this may explain the thin- ning of GCL observed in our meta-analysis. Declarations In the present study, the effect of SO tamponade on Ethics approval and consent to participate SFCT was investigated in three sections. The SFCT This study does not involve human tissues or animals. This study design is showed a significant decrease in the presence of SO a narrative review with ethical committee approval number IR.MUI.MED. tamponade compared to the period before SO injec- REC.1399.889. tion. However, comparison of SFCT during and after Consent for publication SO tamponade application did not show a signifi - This manuscript does not contain any individual patient data. cant change. It was also found that SFCT significantly Competing interests reduced after SO removal compared to the fellow Authors do not have any financial interest in any products mentioned in this healthy eyes [16, 32, 33]. These findings suggest that article. This study received no external funding. The author declares that he in the presence of SO tamponade, SFCT significantly has no competing interest. decreases, which persists after SO removal. Author details To the best of our knowledge, this is the first meta- Isfahan Eye Research Center, Isfahan University of Medical Sciences, Isfahan, analysis that review the effect of SO tamponade on the Iran. Clinical Informationist Resarch Group, Health Information Technology Research Center, Isfahan University of Medical Sciences, Isfahan, Iran. Depar t- central macular thickness (CMT), different retinal lay - ment of Biostatistics and Epidemiology, School of Health, Isfahan University ers, and choroidal thickness. of Medical Sciences, Isfahan, Iran. This study had several limitations. First, only a limited Received: 31 July 2021 Accepted: 25 November 2021 number of articles, investigating the effect of SO tam - ponade on the retinal layers and choroidal thickness, were included in our meta-analysis. Second, in some of the studies included in this systematic review, only the References thickness of the inner retinal layers was reported, and 1. Dehghani A, Razmjoo H, Fazel F, et al. The comparison of retinal blood CMT was not mentioned. Also, in some studies exam- flow after scleral buckling surgery with or without encircling procedure. J Res Med Sci. 2013;18(3):222. ining the retinal layers separately, the sum of two adja- 2. Akhlaghi M, Zarei M, Ziaei M, et al. Sensitivity, specificity, and accuracy cent layers was reported, which in some cases, made it of color Doppler ultrasonography for diagnosis of retinal detachment. J impossible to compare the layers. On the other hand, Ophthalmic Vis Res. 2020;15(2):166. 3. Azen SP, Scott IU, Flynn HW Jr, et al. Silicone oil in the repair of complex the small sample size can be an important factor, as it retinal detachments: a prospective observational multicenter study. cannot provide us with accurate information. On the Ophthalmology. 1998;105(9):1587–97. other hand, our meta-analysis had its strengths, as it is 4. Herbert EN, Laidlaw DAH, Williamson TH, et al. Loss of vision once silicone oil has been removed. Retina (Philadelphia, Pa). 2005;25(6):808–9. the first study summarizing the effect of SO tamponade 5. Williams PD, Fuller CG, Scott IU, et al. Vision loss associated with the use on the retinal layers and choroidal thickness. and removal of intraocular silicone oil. Clin Ophthalmol (Auckland, NZ). 2008;2(4):955. 6. Michel G, Meyer L, Naoun O. Perte brutale d’acuité visuelle post-ablation de silicone: à propos de trois patients traités pour déchirure géante. J Fr Conclusion Ophtalmol. 2009;32(2):104–11. Based on the results, the SO tamponade causes a tem- 7. Menke MN, Kowal JH, Dufour P, et al. Retinal layer measurements after successful macula-off retinal detachment repair using optical coherence porary decrease in CMT, mainly in the inner retinal tomography. Invest Ophthalmol Vis Sci. 2014;55(10):6575–9. layers. It also causes a significant reduction in SFCT; 8. Christensen UC, la Cour M. Visual loss after use of intraocular silicone however, the mechanisms of these effects on the ret - oil associated with thinning of inner retinal layers. Acta Ophthalmol. 2012;90(8):733–7. ina are unclear. It is recommended to conduct further 9. Eer ff t R, Wolf S, Arend O, et al. Retinal hemodynamics after pars plana vit - research on more patients with a long-term follow-up. rectomy with silicone oil tamponade. Ger J Ophthalmol. 1994;3(2):65–7. 10. Kubicka-Trzaska A, Kobylarz J, Romanowska-Dixon B. Macular microcircu- Acknowledgements lation blood flow after pars plana vitrectomy with silicone oil tamponade. Very special thanks to Ms Pegah Noorshargh for all your help and support. Klin Oczna. 2011;113(4–6):146–8. Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 14 of 14 11. Scheerlinck LM, Schellekens PA, Liem AT, et al. Incidence, risk factors, and 34. Caramoy A, Droege KM, Kirchhof B, et al. Retinal layers measurements in clinical characteristics of unexplained visual loss after intraocular silicone healthy eyes and in eyes receiving silicone oil-based endotamponade. oil for macula-on retinal detachment. Retina. 2016;36(2):342–50. Acta Ophthalmol. 2014;92(4):e292–7. 12. Herbert E, Habib M, Steel D, et al. Central scotoma associated with 35. Dooley I, Treacy M, O’Rourke M, et al. Serial spectral domain ocular coher- intraocular silicone oil tamponade develops before oil removal. Graefes ence tomography measurement of outer nuclear layer thickness in rheg- Arch Clin Exp Ophthalmol. 2006;244(2):248–52. matogenous retinal detachment repair. Curr Eye Res. 2015;40(10):1073–6. 13. Scheerlinck LM, Schellekens PA, Liem AT, et al. Retinal sensitivity following 36. Grzybowski A, Pieczynski J, Ascaso FJ. Neuronal complications of intravit- intraocular silicone oil and gas tamponade for rhegmatogenous retinal real silicone oil: an updated review. Acta Ophthalmol. 2014;92(3):201–4. detachment. Acta Ophthalmol. 2018;96(6):641–7. 37. Majid M, Hussin H, Biswas S, et al. Emulsification of Densiron-68 used in 14. Goker YS, Yuksel K, Turan MF, et al. Segmental analysis of macular layers inferior retinal detachment surgery. Eye. 2008;22(1):152–7. in patients with Rhegmatogenous retinal detachment treated with 38. Miller JB, Papakostas TD, Vavvas DG, et al. Complications of emulsified Perfluoropropane or silicon oil. Ophthalmic Surg Lasers Imaging Retina. silicone oil after retinal detachment repair. Seminars in ophthalmology; 2018;49(1):41–7. 2014: Taylor & Francis. 15. Lee SH, Han JW, Byeon SH, et al. Retinal layer segmentation after silicone 39. Coblentz FE, Radeke MJ, Lewis GP, et al. Evidence that ganglion cells react oil or gas tamponade for macula-on retinal detachment using optical to retinal detachment. Exp Eye Res. 2003;76(3):333–42. coherence tomography. Retina. 2018;38(2):310–9. 40. Faude F, Francke M, Makarov F, et al. Experimental retinal detachment 16. Karimi S, Entezari M, Nikkhah H, et al. Eec ff ts of intravitreal silicone oil on causes widespread and multilayered degeneration in rabbit retina. J subfoveal choroidal thickness. Ophthalmologica. 2018;239(2–3):159–66. Neurocytol. 2002;30(5):379–90. 17. Zhou Y, Zhang S, Zhou H, et al. Comparison of fundus changes following silicone oil and sterilized air tamponade for macular-on retinal detach- Publisher’s Note ment patients. BMC Ophthalmol. 2020;20(1):1–9. Springer Nature remains neutral with regard to jurisdictional claims in pub- 18. Isfahani SN, Farajzadegan Z, Sabzghabaee AM, et al. Does hemoperfusion lished maps and institutional affiliations. in combination with other treatments reduce the mortality of patients with paraquat poisoning more than hemoperfusion alone: a systematic review with meta-analysis. J Res Med Sci. 2019;24. 19. Izadi Laybidy S, Shahmahmoodi T, Rahimi A, et al. Therapeutic electrical stimulation currents in chronic non-specific low back pain: designing a systematic review. Muscles, Ligaments & Tendons J (MLTJ ). 2020;10(3). 20. Shahmahmoodi T, Izadi Laybidy S, Rahimi A, et al. Rehabilitation in sub- jects with chronic non-specific low back pain with sacroiliac joint origin: protocol for a systematic review. Muscles, Ligaments & Tendons J (MLTJ ). 2020;10(4). 21. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–58. 22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clin Trials. 1986;7(3):177–88. 23. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34. 24. Hostovsky A, Mandelcorn MS, Mandelcorn ED. Transient macular thin- ning during the use of heavy silicone oil, Densiron 68. Curr Eye Res. 2021;46(3):350–4. 25. Takkar B, Azad R, Kamble N, et al. Retinal nerve fiber layer changes follow- ing primary retinal detachment repair with silicone oil tamponade and subsequent oil removal. J Ophthalmic Vis Res. 2018;13(2):124. 26. Roohipoor R, Tayebi F, Riazi-Esfahani H, et al. Optical coherence tomogra- phy angiography changes in macula-off rhegmatogenous retinal detach- ments repaired with silicone oil. Int Ophthalmol. 2020;40(12):3295–302. 27. Rabina G, Azem N, Barequet D, et al. Silicone oil tamponade effect on macular layer thickness and visual acuity. Retina. 2020;40(5):998–1004. 28. Tode J, Purtskhvanidze K, Oppermann T, et al. Vision loss under silicone oil tamponade. Graefes Arch Clin Exp Ophthalmol. 2016;254(8):1465–71. 29. Kheir WJ, Mehanna C-J, Koaik M, et al. Macular changes on optical coherence tomography before, during, and after silicone oil tampon- ade for macula-on retinal detachment: a case series. J VitreoRetinal Dis. 2018;2(5):297–301. 30. Purtskhvanidze K, Hillenkamp J, Tode J, et al. Thinning of inner retinal layers after vitrectomy with silicone oil versus gas endotampon- Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : ade in eyes with macula-off retinal detachment. Ophthalmologica. 2017;238(3):124–32. fast, convenient online submission 31. Inan S, Polat O, Ozcan S, Inan UU. Comparison of long-term automated thorough peer review by experienced researchers in your field retinal layer segmentation analysis of the macula between silicone oil and gas tamponade after vitrectomy for rhegmatogenous retinal detach- rapid publication on acceptance ment. Ophthalmic Res. 2020;63(6):524–32. support for research data, including large and complex data types 32. Mirza E, Şatırtav G, Oltulu R, et al. Subfoveal choroidal thickness • gold Open Access which fosters wider collaboration and increased citations change following pars plana vitrectomy with silicone oil endotam- ponade for rhegmatogenous retinal detachment. Int Ophthalmol. maximum visibility for your research: over 100M website views per year 2019;39(8):1717–22. 33. Odrobina D, Golebiewska J, Maroszynska I. Choroidal thickness changes At BMC, research is always in progress. after vitrectomy with silicone oil tamponade for proliferative vitreoretin- Learn more biomedcentral.com/submissions opathy retinal detachment. Retina. 2017;37(11):2124–9. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Retina and Vitreous Springer Journals

The effect of silicone oil tamponade on retinal layers and choroidal thickness in patients with rhegmatogenous retinal detachment: a systematic review and meta-analysis

Download PDF
14 pages

Loading next page...
 
/lp/springer-journals/the-effect-of-silicone-oil-tamponade-on-retinal-layers-and-choroidal-A0ug7wRy5C

References (53)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2056-9920
DOI
10.1186/s40942-021-00348-y
Publisher site
See Article on Publisher Site

Abstract

Background: To evaluate the effects of intravitreal silicone oil (SO) on the retinal and choroidal thickness in eyes with rhegmatogenous retinal detachment (RRD). Methods: A literature search was performed in Web of Science, Scopus, ProQuest, Embase, Clinical Key, Science Direct, Cochrane Library, and Springer, as well as Persian databases, including IranDoc, MagIran, SID, MOH thesis, and MOH articles until June 2020. Two reviewers independently searched and extracted the data. Results: Sixteen studies (n = 391) met the inclusion criteria. The meta-analysis showed that the SO tamponade could significantly reduce the central macular thickness (CMT ) in patients with RRD as compared to gas tamponade WMD = − 14.91; 95% CI: − 22.23, − 7.60; P < 0.001, I = 71%). No significant change was found in CMT between the eye with SO tamponade (after SO removal) and the fellow healthy eye in patients with RRD ( WMD = − 3.52; 95% CI: − 17.63, 10.59; I = 68.6%). Compared to the preoperative stage, the SO tamponade could significantly reduce the subfoveal choroidal thickness in patients with RRD ( WMD = − 18.67, 95% CI: − 30.07, − 1.28; I = 80.1%). However, there was no significant difference in the subfoveal choroidal thickness before and after SO removal ( WMD = − 1.13, 95% CI: − 5.97, 3.71; I = 87.6%). Conclusion: The SO tamponade had a significant effect on the reduction of retinal layers and the subfoveal choroidal thickness. Keywords: Rhegmatogenous retinal detachment, Silicone oil, Retinal layer thickness, Choroidal thickness Background Retinal detachment (RD) is one of the leading causes of permanent vision loss. It can be classified as rhegma - togenous, tractional, or exudative. Rhegmatogenous retinal detachment (RRD) is the most common type of *Correspondence: [email protected] RD, occurring in one per 10,000 people each year. Reti- Isfahan Eye Research Center, Isfahan University of Medical Sciences, nal surgery is increasingly used for the repair of RRD, Isfahan, Iran and recent developments in techniques and instruments Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 2 of 14 have improved the outcomes [1, 2]. Silicone oil (SO) and a study by Karimi et  al. showed that the subfoveal cho- gas are two of the most commonly used tamponading roidal thickness (SFCT) was reduced in patients receiv- agents in vitreoretinal surgery, each with its own advan- ing SO tamponade [16]. Conversely, a study by Zhou tages and disadvantages. Although there is a consensus et  al. reported a decrease in the retinal layer thickness, on what necessitates gas or SO tamponade, they can be while no significant change was observed in the choroidal still selected at the discretion of the surgeon or based on thickness [17]. regional differences [3]. Studies have been conducted regarding the effect of SO The long-term SO tamponade complications include tamponade on retinal and choroidal thickness in recent secondary glaucoma, cataract, optic neuropathy, band years. However, there is no systematic review of studies keratopathy, and retinal toxicity [4]. Mild vision loss may investigating the effect of SO tamponade on retinal lay - occur after SO removal, usually due to postoperative cys- ers. Therefore, in the present study, we systematically toid macular edema (CME), hypotony, optic neuropathy, reviewed the effect of SO tamponade on the central mac - and development of epiretinal membranes; neverthe- ular thickness (CMT), different retinal layers, and choroi - less, there are some cases of visual loss that do not have dal thickness. a certain cause [4–6] The possible mechanisms include macular dysfunction, mostly of ganglion and bipolar cell Methods synapses, and optic nerve damage due to SO infiltration Search strategy [5]. In this systematic review, we unanimously approved the Optical coherence tomography shows that long-term main concepts using the PICO format (Table 1) Based on SO tamponade may have a negative effect on the thick - the PICO search format, a free search strategy was used ness of retinal layers in some patients [7, 8]. Also, some to perform a more comprehensive and accurate search of researchers have shown that SO tamponade may have a information sources in different databases. Also, in this negative long-term effect on retinal microcirculation [9, study, all published articles were included without any 10]. Besides, in RRD patients without macular involve- time limits. A search of the following international data- ment, the prevalence of unexplained vision loss is higher bases was performed until June 2020: Web of Science, after SO tamponade as compared to gas tamponade [8, PubMed, Scopus, ProQuest, Clinical Key, Science Direct, 11].Recent studies also suggest that visual abnormalities, Embase, Cochrane Library, and Springer, as well as Per- such as decreased foveal sensitivity, central scotoma, and sian databases, such as MagIran, SID, IranDoc, MOH macular dysfunction, may be related to SO tamponade thesis, and MOH article databases. The included studies [12, 13]. were clinical trials, clinical interventions (before-and- Recent studies examining the effect of retinal and cho - after analysis), and cohorts that met the inclusion criteria roidal layers have reported controversial results, includ- after screening. ing a study by Goker et al., which found that the thickness The inclusion criterion in this systematic review was of the inner nuclear layer (INL) and outer plexiform layer all studies investigating the effect of SO tamponade on (OPL) increased in the eyes with SO tamponade in com- retinal and choroidal layers in patients with RRD. There parison with the fellow healthy eyes [14]. However, in was no age, sex, magazine type, or publication limit, and a study by Lee et  al. on patients with RD, it was found all articles meeting the inclusion criteria were evalu- that in the SO tamponade group, there was a marked ated. The exclusion criteria were as follows: a macular reduction in the inner and outer retinal layer thickness hole, a macular pucker, or other maculopathies; RRD compared to the gas tamponade group [15]. Moreover, with high myopia, uveitis, glaucoma, macular edema, or Table 1 Electronic database search terms Synonyms Keywords “Retinal detachment” OR “Retinal detachment surgery” OR “Rhegmatogenous retinal detachment” Retinal detachment Population/disease “Silicone oil” OR “Silicone oil endotamponade” OR “Endotamponade” OR “Intravitreal silicone oil” Silicone oil tamponade Intervention OR “Intravitreal tamponade” OR “Intraocular tamponade” OR “Silicone oil removal” “Gas tamponade” AND “Fellow eye” Gas tamponade and fellow eye Comparison “Retinal layer thickness” OR “Retinal thickness” OR “Macular layers” OR “Macular layer thickness” OR Retinal layers Consequences “Central macular thickness” OR “Central retinal thickness” OR “Retinal nerve fiber layer” OR “Gan- glion cell layer” OR “Inner plexiform layer” OR “Outer plexiform layer” OR “Inner nuclear layer” OR “Outer nuclear layer” OR “Rod and cone inner and outer segments (IS/OS)” OR “Inner retinal layers” OR “Outer retinal layers” OR “Choroidal thickness” OR “Subfoveal choroidal thickness” Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 3 of 14 postoperative endophthalmitis; traumatic RD; and other plot asymmetry was performed by Egger’s and Begg’s optical nerve diseases. Also, case reports and systematic regression tests [23]. Both the funnel plot and the Egger reviews were excluded. and Begg tests do not have good power to assess publi- cation bias in meta-analysis with less than 10 studies. It Data extraction and quality assessment tried to do the bias analysis, but it doesn’t have enough After searching for studies based on the keywords, eligi- studies. When there was publication bias, the trim-and- ble articles were extracted from the scientific databases. fill method was applied to detect the contribution of bias The flowchart of the articles was plotted by one of our to the overall effect. Statistical analyses were performed colleagues (A.R.). In the first stage, by reviewing the title in Stata version 11 (Stata Corp., College Station, TX, and abstract of studies, a number of unrelated and dupli- USA). P-values less than 0.05 were considered statisti- cate studies were eliminated. In the next step, a detailed cally significant. evaluation of the full-text of the remaining articles was performed. All steps were performed by two independent Results researchers (H.Gh. and A.R.J.), and the third researcher Search results and study characteristics (F.K.) supervised and studied their findings [18–20]. The As shown in Table  2, the initial search yielded a total of following data were also extracted from eligible studies 563 publications. Based on the inclusion and exclusion by two researchers (H.Gh. and A.R.J.): the first author’s criteria, 16 articles were found to be eligible: five studies name, publication date, sample size, age, the duration of had arms only evaluating the CMT [24–29], three studies SO tamponade dressing, time spent after SO removal, only evaluating the inner retinal layer thickness [7, 8, 28]. type of RRD (on/off ), CMT, retinal layer segmentation five studies evaluating both CMT and retinal layer seg - thickness, and SFCT. The quality of studies was also mentation thickness [14, 15, 17, 30, 31], and three stud- assessed by completing the CONSORT checklist for ies only evaluating the SFCT [16, 32, 33].In the search of clinical trials and STROBE checklist for observational Persian databases, no relevant studies were found. Fig- studies; based on the score that each study received, we ure 1 presents the flow chart of the selection process used decided to remove or include it. to identify studies. Out of 16 articles reviewed, the findings of 11 articles Data synthesis and statistical analysis were meta-analyzed [13–18, 20, 22–25]. The findings of Two rounds of data analysis were conducted. For all the other five articles are presented only quantitatively included studies, whether they had a control group or due to the lack of necessary information in the statistical not, the effect of SO was evaluated by comparing the out - analysis [5, 6, 19, 21, 26]. Of the selected articles, 13 stud- comes before and after the intervention. Also, for studies ies were retrospective, two were prospective, and one was with a control group, the mean difference in the out - cross-sectional. A total of 391 patients were included in comes before and after the intervention was calculated these studies, and their mean age was 55.36 years. Table 2 for each group and compared between the SO group summarizes the characteristics of the included studies in and the control group. Moreover, the effect size was this systematic review. estimated as the weighted mean difference (WMD) or standardized mean difference (SMD) and 95% confidence Meta‑analysis and data synthesis interval (95% CI). Heterogeneity between studies was Eec ff t of SO tamponade on the retinal layer thickness 2 2 evaluated by Cochran’s Q and I-square (I ) tests; I above There are two types of articles on the effect of SO tam - 50% was regarded as substantial heterogeneity [21]. ponade on the retinal layers. Some studies only examined To calculate the effect sizes and their correspond - the effect of SO tamponade on the CMT or inner retinal ing 95% CIs, a fixed effects model was adopted when layers, while some studies, in addition to CMT, exam- heterogeneity was low. On the other hand, a random ined the effect of SO tamponade separately on the retinal effects model was used in case of moderate or high het - layers. erogeneity, considering the between-study variations in the DerSimonian and Laird method [22]. To explore the Eec ff t of SO tamponade on CMT: (Fig. 2a–o) source of heterogeneity, a meta-regression analysis was We examined the effect of SO tamponade on CMT in performed to evaluate the confounding role of age and several sections: duration of SO tamponade. Besides, sensitivity analyses were performed to evaluate the extent to which infer- Comparison of CMT during SO tamponade application ences might be related to a particular study. Publication and after SO removal bias was also examined by the visual inspection of fun- In this comparison, due to the small number of arti- nel plots [23]. Moreover, a statistical evaluation of funnel cles and incomplete information, no meta-analysis was Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 4 of 14 Table 2 The general characteristics of the included studies in the systematic review First author Publication year Quality of study Age (mean) Sample size Type of Duration of SO Time after CMT Inner Retinal layer SFCT RRD (on/ tamponade SO removal retinal layer segmentation off ) (month) (month) thickness thickness Hostovsky [24] 2020 Good 57.7 SO = 20 Off 1.5 -2.5 6 * Rabina [27] 2019 Excellent 56.1 SO = 41 On 3- 6 1 * Takkar [25] 2017 Good 39.6 SO = 32 Off 3–6 6–9 * Roohipoor [26] 2020 Excellent 54 SO = 45 Off 3 – * Kheir [29] 2018 Good 48.1 SO = 10 On 3 9 * Tode [28] 2016 Fair 61 SO = 15 On 1.5 36 * Caramoy [34] 2014 Fair 67 SO = 9 On 5 – * Christensen [8] 2011 Fair 51 SO = 9 On 5 46 * Gas = 7 Inan [31] 2016 Good 60.7 SO = 28 Off 4 8 * * Gas = 30 Goker [14] 2017 Good 56.35 SO = 20 Off 4 2 * * Gas = 16 Lee [15] 2017 Good 46.83 SO = 33 On 3.5 6 * * Gas = 31 Purtskhvanidze [30] 2017 Good 59.3 SO = 20 Off 5.5 35 * * Gas = 20 Zhou [17] 2020 Good 53.86 SO = 7 On 4 – * * Gas = 14 Mirza [32] 2018 Fair 60.4 SO = 24 Off 3–6 1 * Karimi [16] 2018 Good 52 SO = 60 Off 6–9 3 * Odrobino [33] 2016 Good 62.44 SO = 18 Off 6 - * A scoring scale for this study based on Strobe modified (≥ 85 = Excellent, 70 to < 85 = Good, 50 to < 70 = Fair, < 50 = poor) Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 5 of 14 563 arcles idenfied in total 521 arcles excluded on tle and abstract review 42 arcles retrieved for fulltext review 26 arcles excluded due to not meeng inclusion criteria 16 arcles sasfied all inclusion criteria 5 arcles evaluated 5 arcles evaluated 3 arcles evaluated 5 arcles evaluated Renal layers segmentaon Inner renol layers Choroidalthickness Central macular thickness Fig. 1 The results of the literature search strategy. The flow diagram depicts the screening process of articles and the number of exclusions Comparison of CMT after SO removal with CMT after gas conducted, and the data were only qualitatively exam- tamponade (Fig. 2a) ined. This section includes three articles. The results Four studies with five effect sizes [14, 15, 17, 30] exam- of a study by Hostovsky [24] showed that during SO ined the effect of SO tamponade on the CMT after tamponade application, the CMT was significantly SO removal compared to gas tamponade in patients lower, with a thickness recovery after SO removal (from 248 ± 48.88 to 277 ± 46.46 μ; P = 0.001). There with RRD. The SO tamponade could significantly was no significant difference in the CMT between reduce the CMT (WMD = −  14.91; 95% CI: −  22.23, the eye with SO tamponade and the fellow eye in the −  7.60; P < 0.001, I = 71%) compared to gas tampon- last follow-up after the SO removal (278 ± 49.40 and ade in patients with RRD, with substantial heteroge- 264 ± 34.92 μ, respectively; P = 0.265). neity between studies (Fig.  2a). The meta-regression Another study by Rabina [27] showed that the mean analysis showed that the outcomes were not dependent CMT in the operated eye increased from 249 ± 50 µ on age (P = 0.632) or the time spent after SO removal during SO tamponade application to 279 ± 48 µ after (P = 0.328). However, they were dependent on the dura- SO removal (P = 0.001). Also, the CMT was 281 ± 21 tion of SO tamponade application (P = 0.026). Although µ in the fellow healthy eye. The results showed the Egger’s (P = 0.094) and Begg’s (P = 0.221) regression apparent thinning of retinal thickness, mainly in the tests did not show significant publication bias, substan - inner layers, during SO tamponade application, which tial asymmetry was seen in the funnel plot. Accordingly, resolved after SO removal. However, a study by Kheir we performed a trim-and-fill analysis and found that the WJ [29] reported that the mean CMT decreased from CMT reduced significantly in the SO tamponade group a baseline of 255.7 µ to 247.5 µ during SO tampon compared to the gas tamponade group (WMD = − 14.91; 95% CI: − 22.23, − 7.60; P < 0.001, I = 71%). The sensitiv - ade application, which was followed by an increase ity analysis showed that exclusion of a study did not affect to 262.1 µ after SO removal; however, these changes the results obtained. were not statistically significant (P = 0.44). B esides , Another study by Christensen UC, comparing the quantitative thinning of only the inner retinal layers was observed. Based on these three studies, it seems SO tamponade group with the gas tamponade group, that SO tamponade causes a transient decrease in the was not included in our meta-analysis for two rea- CMT, mainly in the inner layers. After SO removal, the sons. First, in this study, only the inner retinal layers CMT resembled that of the fellow eye. were examined. Second, the unit of measurement was Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 6 of 14 Fig. 2 Eec ff t of SO (silicone oil) tamponade on retinal layers in patients with rhegmatogenous retinal detachment (RRD). SO silicone oil, RRD rhegmatogenous retinal detachment, CMT central macular thickness, NFL nerve fiber layer, GCL ganglion cell layer, IPL inner plexiform layer, INL inner plexiform layer, OPL outer plexiform layer, ONL outer nuclear layer, RPE retinal pigment epithelium reported in pixels, without mentioning the standard regarding CMT (WMD = − 3.52; 95% CI: − 17.63, 10.59; deviation. In this study, thinning of the inner retinal I = 68.6%), with substantial heterogeneity between stud- layers was observed in the SO-operated eyes (5148 ies (Fig.  2b). The meta-regression analysis showed that pixels) compared to gas-operated eyes (6897 pixels) the results were not dependent on age (P = 0.214), the (P = 0.002) [8]. time spent after SO removal (P = 0.346), or duration of SO tamponade application (P = 0.992). Although Egger’s Comparison of CMT after SO removal with the fellow healthy (P = 0.338) and Begg’s (P = 0.060) regression tests did eye (Fig. 2b) not show significant publication bias, substantial asym - Six studies [14, 24, 25, 27, 30, 31] examined the effect metry was observed in the funnel plot. The adjusted of SO tamponade on CMT in the eyes of patients with value, based on the trim-and-fill method, showed no RRD compared to the fellow healthy eyes. No signifi - significant change in the eyes with SO tamponade (after cant change was found in the eyes with SO tamponade SO removal) in comparison with the fellow healthy eyes (after SO removal) as compared to the fellow healthy eyes (WMD = − 7.83; 95% CI = − 22.15, 6.48; P = 0.284). The Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 7 of 14 Fig. 2 continued sensitivity analysis showed that exclusion of a study did one study examining CMT, which was not included not affect the results obtained. in our meta-analysis due to heterogeneity. In the first study by Tode [28], significant thinning of foveal and parafoveal nerve fibers, ganglion cells, and inner plex - Comparison of the inner retinal layer thickness between eyes iform layer (IPL) was observed in the affected eyes with SO tamponade and the healthy fellow eyes (mean: 58.3 ± 13 μm) as compared to the healthy fellow This section includes three studies, with two studies eyes (mean: 84.5 ± 12.3  μm) (P < 0.01) [28]. Moreover, only examining the inner retinal layer thickness and Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 8 of 14 Fig. 2 continued in a study by Caramoy, the inner retinal layers became Eec ff t of SO tamponade on retinal layer segmentation thinner after the use of SO-based endotamponade In this section, the effect of SO tamponade on differ - (retinal layer volume: 1.127 ± 0.160 mm ) compared to ent layers of the retina was compared after SO removal the fellow healthy eye (1.363 ± 0.150 mm ) (P = 0.012) with gas tamponade and the fellow healthy eye. [34]. Besides, the results of these two studies showed a decrease in the inner retinal layers in eyes with SO tam- Nerve fiber layer (NFL) (Fig. 2c, d) ponade as compared to the fellow healthy eyes. Moreo- Four studies with five effect sizes [14, 17, 30, 31] exam- ver, a study by Roohipoor et  al. showed that CMT was ined the effect of SO tamponade on NFL in patients significantly reduced after SO tamponade removal with RRD after SO removal compared to gas tampon- compared to the fellow healthy eye (P = 0.002) [26]. ade. No significant change was found in NFL between Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 9 of 14 SO tamponade (after SO removal) and gas tamponade removal (WMD = − 0.43; 95% CI: − 1.61, 2.46; I = 0.0%) in patients with RRD (WMD = −  0.92; 95% CI: −  2.11, and the fellow healthy eyes (Fig. 2h); there was no hetero- 0.27; I = 12.6%), and there was no heterogeneity between geneity between studies. The results of Egger’s and Begg’s studies (Fig.  2c). Although Egger’s (P = 0.068) and Begg’s regression tests revealed no evidence of publication bias (P = 0.066) regression tests did not show significant pub - in comparison with gas tamponade (Egger’s test P = 0.803 lication bias, substantial asymmetry was observed in the and Begg’s test P = 1.000) and the fellow healthy eyes funnel plot. The adjusted value, based on the trim-and- (Egger’s test P = 0.603 and Begg’s test P = 0.296). The sen - fill method, showed no significant change in the effect sitivity analysis showed that exclusion of a study did not of SO tamponade on NFL in patients with RRD after SO affect the results obtained. removal compared to gas tamponade (SMD = −  0.220; 95% CI = −  0.625, 0.185; P = 0.287). The results of Egg - Inner nuclear layer (INL) er’s (P = 0.068) and Begg’s (P = 0.066) regression tests Four studies with an effect size of five [14, 17, 30, 31] revealed no evidence of publication bias. The sensitivity examined the effect of SO tamponade on NFL in patients analysis showed that exclusion of a study did not affect with RRD after SO removal compared to gas tampon- the results obtained. ades. No significant change was found in INL between Three studies [14, 30, 31] examined the effect of SO the eyes with SO tamponade (after SO removal) and gas tamponade on NFL in patients with RRD compared to tamponade in patients with RRD (WMD = 0.32; 95% the fellow healthy eye. The SO tamponade could signifi - CI: −  1.57, 2.22; I = 31.8.0%); there was no heterogene- cantly increase NFL in patients with RRD compared to ity between studies (Fig.  2i). The results of Egger’s test the fellow healthy eye (WMD = 1.86; 95% CI: 0.65, 3.06; (P = 0.279) and Begg’s test (P = 0.221) revealed no evi- I = 0.0%); there was no heterogeneity between studies dence of publication bias. (Fig.  2d). The results of Egger’s (P = 0.560) and Begg’s Three studies [12, 28, 29] examined the effect of SO (P = 1.000) regression tests revealed no evidence of publi- tamponade on INL in patients with RRD compared to cation bias. The sensitivity analysis showed that exclusion the fellow healthy eyes. The SO tamponade could sig - of a study did not affect the results obtained. nificantly increase the INL in patients with RRD com - pared to the fellow healthy eyes (WMD = 5.51; 95% CI: Ganglion cell layer (GCL) (Fig. 2e, f ) 2.88, 8.15; I = 0.0%); there was no heterogeneity between Three studies [12, 27, 28] examined the effect of SO tam - studies (Fig.  2j). The results of Egger’s (P = 0.629) and ponade on GCL in comparison with gas tamponade and Begg’s (P = 1.000) regression tests revealed no evidence the fellow healthy eyes. The SO tamponade could sig - of publication bias. The sensitivity analysis showed that nificantly reduce GCL in patients with RRD compared to exclusion of a study did not affect the results obtained. gas tamponade (WMD = − 3.41, 95% CI: − 5.58, − 1.25; I = 0.0%); there was no heterogeneity between studies Outer plexiform layer (OPL) (Fig. 2I) (Fig.  2e). Also, no significant change was found in the Four studies with an effect size of five [12, 15, 28, 29], operated eyes after SO removal (WMD = 0.02, 95% CI: as well as three studies [12, 28, 29], examined the OPL −  5.19, 5.23; I = 70.6%) compared to the fellow healthy thickness changes after SO removal in comparison with eyes; there was moderate heterogeneity between studies gas tamponade and the fellow healthy eyes, respec- (Fig. 2f ). The results of Egger’s and Begg’s regression tests tively. No significant change was found in OPL after SO revealed no evidence of publication bias in comparison removal compared to gas tamponade (WMD = −  0.77; with gas tamponade (Egger’s test P = 0.589 and Begg’s 95% CI: −  4.58, 3.05; I = 65.8%), with substantial het- test P = 1.000) and the fellow healthy eyes (Egger’s test erogeneity between studies (Fig.  2k), or with the fel- P = 0.296 and Begg’s test P = 0.629). The sensitivity anal - low healthy eyes (WMD = 0.40; 95% CI: −  5.17, 5.97; ysis showed that exclusion of a study did not affect the I = 79.5%), with substantial heterogeneity between stud- results obtained. ies (Fig.  2l). The results of Egger’s and Begg’s regression tests revealed no evidence of publication bias in com- Inner plexiform layer (IPL) (Fig. 2g, h) parison with gas tamponade (Egger’s test P = 0.224 and Three studies [12, 27, 28] examined the effect of SO tam - Begg’s test P = 0.221) and the fellow healthy eyes (Egger’s ponade on IPL in comparison with gas tamponades and test P = 0.147 and Begg’s test P = 0.296). The sensitivity the fellow healthy eyes. The SO tamponade could sig - analysis showed that exclusion of a study did not affect nificantly reduce the IPL in patients with RRD as com - the results obtained. pared to gas tamponade (WMD = − 2.50, 95% CI: − 4.62, − 0.39; I = 0.0%) (Fig.  2g). No significant change was found between the eyes with SO tamponade after SO Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 10 of 14 Outer nuclear layer (ONL) (Fig. 2m, n) trim-and-fill analysis method and found that the SFCT Three studies [12, 28, 29] examined the ONL thick- reduced significantly in the eyes with SO tamponade ness changes after SO removal in comparison with gas compared to the preoperative period (WMD = − 24.077; tamponade and the fellow healthy eyes. No significant 95% CI = −  41.360, −  6.795; P = 0.006). The sensitivity change was found in the ONL after SO removal, either analysis showed that exclusion of a study did not affect compared to gas tamponade (WMD = −  3.09; 95% CI: the results obtained. −  15.38, 9.20; I = 74.7%) (Fig.  2m) or the fellow healthy eyes (WMD = − 8.58; 95% CI: − 24.27, 7.10; I = 89.5%), with substantial heterogeneity between studies (Fig.  2n). Comparison of SFCT during SO tamponade application Comparison of SO tamponade with gas tamponade after and after SO removal (Fig. 3b) removal by Egger’s (P = 0.502) and Begg’s (P = 1.000) Two studies with an effect size of four [16, 32] examined regression tests revealed no evidence of publication bias. and compared the SFCT changes during and after SO In comparison of SO removal with the fellow healthy application. No significant change was found in the SFCT eyes, although the results of Egger’s (P = 0.758) and after SO removal, compared to the time of SO tampon- Begg’s (P = 1.000) regression tests did not indicate sig- ade application (WMD = −  1.13, 95% CI: −  5.97, 3.71; nificant publication bias, substantial asymmetry was I = 87.6%); there was substantial heterogeneity between seen in the funnel plot. The adjusted value, based on studies (Fig.  3b). Our meta-regression analysis showed the trim-and-fill method, showed no significant change that the results were not dependent on age (P = 0.039) (SMD = − 0.483; 95% CI = − 1.440, 0.475; P = 0.323). The or duration of SO tamponade application (P = 0.662). sensitivity analysis showed that exclusion of a study did The results of Egger’s regression test (P = 0.002) and the not affect the results obtained. funnel plot revealed evidence of publication bias, but the adjusted value, based on the trim-and-fill method, Retinal pigment epithelium (RPE) (Fig. 2o) showed no significant change after SO removal compared Two studies [14, 31] examined the RPE thickness changes to the time when SO was removed (WMD = −  3.711; after SO removal compared to gas tamponade and the 95% CI = − 10.045, 2.622; P = 0.251). The sensitivity anal - fellow healthy eye. No significant change was found in ysis showed that exclusion of a study did not affect the the RPE after SO removal, either compared to gas tam- results obtained. ponade (WMD = 0.43; 95% CI: −  1.17, 2.03; I = 0.0%) (Fig.  2o) or the fellow healthy eyes (WMD = 0.62; 95% CI: −  0.86, 2.11; I = 0.0%); there was no heterogene- Comparison of SFCT after SO tamponade removal ity between studies (Fig.  2p). Publication bias was not with the fellow healthy eyes (Fig. 3c) reviewed due to the small number of studies. Two studies with an effect size of four [32, 33] examined the effect of SO tamponade on the SFCT after SO tam - Eec ff t of SO tamponade on SFCT (Fig. 3a–c) ponade removal and compared it with the fellow healthy There were three studies regarding the effect of SO tam - eyes. After SO removal, the SFCT significantly reduced ponade on SFCT, which are described in the following in patients with RRD compared to the fellow healthy eyes three sections. (WMD = − 34.54; 95% CI: − 50.74, − 18.34; I  = 38.4%); there was no heterogeneity between studies (Fig. 3c). The meta-regression analysis showed that the results were Comparison of SFCT before and during SO tamponade not dependent on age (P = 0.855) or duration of SO tam- application (Fig. 3a) ponade application (P  =  0.716). Although the results of Two studies with an effect size of four [32, 33] exam- Egger’s (P  =  0.694) and Begg’s (P  =  0.308) regression ined the effect of SO tamponade on SFCT and compared tests did not show significant publication bias, substantial the results before and during SO application. Com- asymmetry was observed in the funnel plot. The adjusted pared to the preoperative period, the SO tamponade value, based on the trim-and-fill method, showed that could significantly reduce SFCT in patients with RRD the SFCT reduced significantly after SO removal com - (WMD = −  18.67; 95% CI: −  30.07, − 1.28; I = 80.1%), pared to the fellow healthy eyes (VMD  =  −  34.54; 95% with substantial heterogeneity between studies (Fig.  3a). CI: −  50.74, −  18.34; P  <  0.001). The sensitivity analysis The meta-regression analysis showed that the results showed that exclusion of a study did not affect the results were not dependent on age (P = 0.743) or duration of obtained. SO tamponade application (P = 0.730). Although Egger’s (P = 0.711) and Begg’s (P = 0.734) regression tests did not show significant publication bias, substantial asymmetry was observed in the funnel plot. Accordingly, we used the Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 11 of 14 a. Effect of SO tamponade on subfoveal choroidal thickness (SFCT) compared to the preoperative period in patients with RRD b. Comparison of SFCT during SO application and after SO removal in patients with RRD c. The SFCT after SO removal in patients with RRD compared to the fellow healthy eyes Fig. 3 Eec ff t of SO tamponade on SFCT in patients with RRD. SO silicone oil, RRD rhegmatogenous retinal detachment, SFCT subfoveal choroidal thickness Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 12 of 14 Discussion is almost similar to that of the fellow healthy eye. Based For the first time, the results of this systematic review on this comparison, age, duration of SO application, and and meta-analysis revealed that in patients with RRD, time spent after SO removal had no effects on the out - the SO tamponade reduced CMT, and after SO removal, comes [14, 24, 25, 27, 30, 31]. the retinal layer thickness almost returned to its range In our meta-analysis, CMT was compared between before surgery. Also, the SO tamponade in these patients patients after SO removal and patients with gas tampon- reduced SFCT, which remained reduced following the ade. The results showed that CMT significantly reduced SO removal. Overall, the SO tamponade seems to reduce after SO removal compared to gas tamponade. This CMT. In this regard, studies by Hostovsky (P = 0.001) and reduction in CMT was not related to age or the time Rabina (P = 0.001) showed that the SO tamponade caused spent after SO removal, but was related to the duration a significant reduction in CMT, which returned to the of SO application; in other words, the longer the SO tam- normal range after SO removal [24, 27]. Another study ponade was applied, the more CMT was reduced [14, 15, by Kheir WJ, (P = 0.44) showed that CMT decreased 17, 30]. The decrease in CMT after SO removal, com - when the SO tamponade was applied, but increased after pared to gas tamponade, can be explained by the fact that SO removal; however, these changes were not significant in some studies on RRD patients with gas tamponade, a [29]. Moreover, in two studies by Tode et al. and Caramoy long-term increase in NFL, GCL, and IPL was reported et al., the thickness of the inner retinal layers significantly compared to the fellow healthy eyes. It was assumed reduced in the presence of SO tamponade as compared that the retinal nerve fiber layer (RNFL) edema may be to the fellow healthy eyes [28, 34]. a result of surgery itself or may be related to the relative Previous studies have revealed that CMT decreases retinal ischemia associated with RRD [22, 35]. On the significantly in the presence of SO tamponade, especially other hand, in the SO group, these layers showed relative in the inner retinal layers. The causes of retinal thinning atrophy in the presence of SO tamponade. associated with the use of SO tamponade have not been It seems that SO induces different effects on differ - fully elucidated. However, several factors may be involved ent layers of the retina. Inan et  al. reported thinning of in this phenomenon. One of these factors is the mechani- GCL, OPL, and ONL in the SO group compared to the cal stress of SO tamponade on the retinal layers in the group with gas tamponade [31]. Moreover, Goker et  al. macula [35]. Among other mechanisms, the subretinal reported that in the gas tamponade group, only INL was migration of SO may induce inflammation, resulting in significantly increased. In the SO tamponade group, the the thinning of the retina [36, 37]. Also, emulsified SO INL and OPL thicknesses significantly increased, and the may induce an internal limiting membrane defect and be ONL thickness significantly decreased compared to the toxic to the retina, causing retinal thinning when enter- fellow healthy eyes [14]. Purtskhvanidze et  al. showed ing the intraretinal space [38]. Another explanation for that GCL and IPL were significantly thinner in the SO the thinning of the retina in the presence of SO tam- group compared to the gas tamponade group [30]. More- ponade is that the hydrophobic SO tamponade replaces over, Lee SH et  al. reported that in the SO group, there the natural environment of the hydrophilic vitreous cav- was a significant decrease in the thickness of all retinal ity. This waterproofing effect causes retinal dehydration layers, except for the photoreceptor layer after primary and reduces the thickness of retinal layers, especially RD surgery [15]. the inner layers [27]. Besides, the indirect mechanisms In the study of changes in the thickness of different include changes in the concentration of potassium retinal layers after SO removal in comparison with the because of failure in potassium siphoning and changes in fellow healthy eye and gas tamponade, the results of our cytokine levels, which may induce apoptosis and result in meta-analysis revealed that the thickness of NFL and the thinning of retinal layers [5]. INL increased significantly in patients after SO removal It seems that CMT is not significantly different compared to the fellow healthy eyes, but no significant between the operated eye and the fellow healthy eye after changes were observed in the GCL, IPL, ONL, OPL, SO removal. In studies by Rabina et al. [27] and Purtskh- or RPE. On the other hand, in patients with RRD after vanidze et  al. [30, 38], CMT was not significantly differ - SO removal, the thickness of GCL and IPL significantly ent from the fellow healthy eye after SO removal, while reduced compared to patients for whom gas tamponade in a study by Takker et  al. [25], CMT was significantly was used, while the thickness of NFL, INL, OPL, ONL reduced compared to the fellow healthy eye after SO and RPE did not change significantly. removal. Our meta-analysis showed that CMT was not The increase in NFL thickness after SO removal can significantly different from the fellow healthy eye after be related to surgical NFL edema that may persist for SO removal. In other words, the SO tamponade causes a several months after vitrectomy [31]. The increase in transient decrease in CMT, and after SO removal, CMT INL thickness can be justified by the fact that Müller Ghanbar i et al. International Journal of Retina and Vitreous (2021) 7:76 Page 13 of 14 Authors’ contributions cells are the radial glial cells of the retina, and their HG, FK and AD contributed to the conception and design of the review. AJM, nuclei are located in the INL. The Müller cells play AR, AF and ANB contributed to the drafting and critical revision of the manu- an important role in structural restoration after reti- script. All authors read and approved the final manuscript. nal reattachment; therefore, increasing their activity Funding can increase the thickness of INL [14]. Several studies The author declares that no funding or research Grant was obtained for this have reported the degeneration of ganglion cells in the manuscript. detached retina [39]. In this regard, a rabbit model of Availability of data and materials RD showed the progressive loss of ganglion cell axons Not applicable. in the detached retina [40]; this may explain the thin- ning of GCL observed in our meta-analysis. Declarations In the present study, the effect of SO tamponade on Ethics approval and consent to participate SFCT was investigated in three sections. The SFCT This study does not involve human tissues or animals. This study design is showed a significant decrease in the presence of SO a narrative review with ethical committee approval number IR.MUI.MED. tamponade compared to the period before SO injec- REC.1399.889. tion. However, comparison of SFCT during and after Consent for publication SO tamponade application did not show a signifi - This manuscript does not contain any individual patient data. cant change. It was also found that SFCT significantly Competing interests reduced after SO removal compared to the fellow Authors do not have any financial interest in any products mentioned in this healthy eyes [16, 32, 33]. These findings suggest that article. This study received no external funding. The author declares that he in the presence of SO tamponade, SFCT significantly has no competing interest. decreases, which persists after SO removal. Author details To the best of our knowledge, this is the first meta- Isfahan Eye Research Center, Isfahan University of Medical Sciences, Isfahan, analysis that review the effect of SO tamponade on the Iran. Clinical Informationist Resarch Group, Health Information Technology Research Center, Isfahan University of Medical Sciences, Isfahan, Iran. Depar t- central macular thickness (CMT), different retinal lay - ment of Biostatistics and Epidemiology, School of Health, Isfahan University ers, and choroidal thickness. of Medical Sciences, Isfahan, Iran. This study had several limitations. First, only a limited Received: 31 July 2021 Accepted: 25 November 2021 number of articles, investigating the effect of SO tam - ponade on the retinal layers and choroidal thickness, were included in our meta-analysis. Second, in some of the studies included in this systematic review, only the References thickness of the inner retinal layers was reported, and 1. Dehghani A, Razmjoo H, Fazel F, et al. The comparison of retinal blood CMT was not mentioned. Also, in some studies exam- flow after scleral buckling surgery with or without encircling procedure. J Res Med Sci. 2013;18(3):222. ining the retinal layers separately, the sum of two adja- 2. Akhlaghi M, Zarei M, Ziaei M, et al. Sensitivity, specificity, and accuracy cent layers was reported, which in some cases, made it of color Doppler ultrasonography for diagnosis of retinal detachment. J impossible to compare the layers. On the other hand, Ophthalmic Vis Res. 2020;15(2):166. 3. Azen SP, Scott IU, Flynn HW Jr, et al. Silicone oil in the repair of complex the small sample size can be an important factor, as it retinal detachments: a prospective observational multicenter study. cannot provide us with accurate information. On the Ophthalmology. 1998;105(9):1587–97. other hand, our meta-analysis had its strengths, as it is 4. Herbert EN, Laidlaw DAH, Williamson TH, et al. Loss of vision once silicone oil has been removed. Retina (Philadelphia, Pa). 2005;25(6):808–9. the first study summarizing the effect of SO tamponade 5. Williams PD, Fuller CG, Scott IU, et al. Vision loss associated with the use on the retinal layers and choroidal thickness. and removal of intraocular silicone oil. Clin Ophthalmol (Auckland, NZ). 2008;2(4):955. 6. Michel G, Meyer L, Naoun O. Perte brutale d’acuité visuelle post-ablation de silicone: à propos de trois patients traités pour déchirure géante. J Fr Conclusion Ophtalmol. 2009;32(2):104–11. Based on the results, the SO tamponade causes a tem- 7. Menke MN, Kowal JH, Dufour P, et al. Retinal layer measurements after successful macula-off retinal detachment repair using optical coherence porary decrease in CMT, mainly in the inner retinal tomography. Invest Ophthalmol Vis Sci. 2014;55(10):6575–9. layers. It also causes a significant reduction in SFCT; 8. Christensen UC, la Cour M. Visual loss after use of intraocular silicone however, the mechanisms of these effects on the ret - oil associated with thinning of inner retinal layers. Acta Ophthalmol. 2012;90(8):733–7. ina are unclear. It is recommended to conduct further 9. Eer ff t R, Wolf S, Arend O, et al. Retinal hemodynamics after pars plana vit - research on more patients with a long-term follow-up. rectomy with silicone oil tamponade. Ger J Ophthalmol. 1994;3(2):65–7. 10. Kubicka-Trzaska A, Kobylarz J, Romanowska-Dixon B. Macular microcircu- Acknowledgements lation blood flow after pars plana vitrectomy with silicone oil tamponade. Very special thanks to Ms Pegah Noorshargh for all your help and support. Klin Oczna. 2011;113(4–6):146–8. Ghanbari et al. International Journal of Retina and Vitreous (2021) 7:76 Page 14 of 14 11. Scheerlinck LM, Schellekens PA, Liem AT, et al. Incidence, risk factors, and 34. Caramoy A, Droege KM, Kirchhof B, et al. Retinal layers measurements in clinical characteristics of unexplained visual loss after intraocular silicone healthy eyes and in eyes receiving silicone oil-based endotamponade. oil for macula-on retinal detachment. Retina. 2016;36(2):342–50. Acta Ophthalmol. 2014;92(4):e292–7. 12. Herbert E, Habib M, Steel D, et al. Central scotoma associated with 35. Dooley I, Treacy M, O’Rourke M, et al. Serial spectral domain ocular coher- intraocular silicone oil tamponade develops before oil removal. Graefes ence tomography measurement of outer nuclear layer thickness in rheg- Arch Clin Exp Ophthalmol. 2006;244(2):248–52. matogenous retinal detachment repair. Curr Eye Res. 2015;40(10):1073–6. 13. Scheerlinck LM, Schellekens PA, Liem AT, et al. Retinal sensitivity following 36. Grzybowski A, Pieczynski J, Ascaso FJ. Neuronal complications of intravit- intraocular silicone oil and gas tamponade for rhegmatogenous retinal real silicone oil: an updated review. Acta Ophthalmol. 2014;92(3):201–4. detachment. Acta Ophthalmol. 2018;96(6):641–7. 37. Majid M, Hussin H, Biswas S, et al. Emulsification of Densiron-68 used in 14. Goker YS, Yuksel K, Turan MF, et al. Segmental analysis of macular layers inferior retinal detachment surgery. Eye. 2008;22(1):152–7. in patients with Rhegmatogenous retinal detachment treated with 38. Miller JB, Papakostas TD, Vavvas DG, et al. Complications of emulsified Perfluoropropane or silicon oil. Ophthalmic Surg Lasers Imaging Retina. silicone oil after retinal detachment repair. Seminars in ophthalmology; 2018;49(1):41–7. 2014: Taylor & Francis. 15. Lee SH, Han JW, Byeon SH, et al. Retinal layer segmentation after silicone 39. Coblentz FE, Radeke MJ, Lewis GP, et al. Evidence that ganglion cells react oil or gas tamponade for macula-on retinal detachment using optical to retinal detachment. Exp Eye Res. 2003;76(3):333–42. coherence tomography. Retina. 2018;38(2):310–9. 40. Faude F, Francke M, Makarov F, et al. Experimental retinal detachment 16. Karimi S, Entezari M, Nikkhah H, et al. Eec ff ts of intravitreal silicone oil on causes widespread and multilayered degeneration in rabbit retina. J subfoveal choroidal thickness. Ophthalmologica. 2018;239(2–3):159–66. Neurocytol. 2002;30(5):379–90. 17. Zhou Y, Zhang S, Zhou H, et al. Comparison of fundus changes following silicone oil and sterilized air tamponade for macular-on retinal detach- Publisher’s Note ment patients. BMC Ophthalmol. 2020;20(1):1–9. Springer Nature remains neutral with regard to jurisdictional claims in pub- 18. Isfahani SN, Farajzadegan Z, Sabzghabaee AM, et al. Does hemoperfusion lished maps and institutional affiliations. in combination with other treatments reduce the mortality of patients with paraquat poisoning more than hemoperfusion alone: a systematic review with meta-analysis. J Res Med Sci. 2019;24. 19. Izadi Laybidy S, Shahmahmoodi T, Rahimi A, et al. Therapeutic electrical stimulation currents in chronic non-specific low back pain: designing a systematic review. Muscles, Ligaments & Tendons J (MLTJ ). 2020;10(3). 20. Shahmahmoodi T, Izadi Laybidy S, Rahimi A, et al. Rehabilitation in sub- jects with chronic non-specific low back pain with sacroiliac joint origin: protocol for a systematic review. Muscles, Ligaments & Tendons J (MLTJ ). 2020;10(4). 21. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–58. 22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled Clin Trials. 1986;7(3):177–88. 23. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–34. 24. Hostovsky A, Mandelcorn MS, Mandelcorn ED. Transient macular thin- ning during the use of heavy silicone oil, Densiron 68. Curr Eye Res. 2021;46(3):350–4. 25. Takkar B, Azad R, Kamble N, et al. Retinal nerve fiber layer changes follow- ing primary retinal detachment repair with silicone oil tamponade and subsequent oil removal. J Ophthalmic Vis Res. 2018;13(2):124. 26. Roohipoor R, Tayebi F, Riazi-Esfahani H, et al. Optical coherence tomogra- phy angiography changes in macula-off rhegmatogenous retinal detach- ments repaired with silicone oil. Int Ophthalmol. 2020;40(12):3295–302. 27. Rabina G, Azem N, Barequet D, et al. Silicone oil tamponade effect on macular layer thickness and visual acuity. Retina. 2020;40(5):998–1004. 28. Tode J, Purtskhvanidze K, Oppermann T, et al. Vision loss under silicone oil tamponade. Graefes Arch Clin Exp Ophthalmol. 2016;254(8):1465–71. 29. Kheir WJ, Mehanna C-J, Koaik M, et al. Macular changes on optical coherence tomography before, during, and after silicone oil tampon- ade for macula-on retinal detachment: a case series. J VitreoRetinal Dis. 2018;2(5):297–301. 30. Purtskhvanidze K, Hillenkamp J, Tode J, et al. Thinning of inner retinal layers after vitrectomy with silicone oil versus gas endotampon- Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : ade in eyes with macula-off retinal detachment. Ophthalmologica. 2017;238(3):124–32. fast, convenient online submission 31. Inan S, Polat O, Ozcan S, Inan UU. Comparison of long-term automated thorough peer review by experienced researchers in your field retinal layer segmentation analysis of the macula between silicone oil and gas tamponade after vitrectomy for rhegmatogenous retinal detach- rapid publication on acceptance ment. Ophthalmic Res. 2020;63(6):524–32. support for research data, including large and complex data types 32. Mirza E, Şatırtav G, Oltulu R, et al. Subfoveal choroidal thickness • gold Open Access which fosters wider collaboration and increased citations change following pars plana vitrectomy with silicone oil endotam- ponade for rhegmatogenous retinal detachment. Int Ophthalmol. maximum visibility for your research: over 100M website views per year 2019;39(8):1717–22. 33. Odrobina D, Golebiewska J, Maroszynska I. Choroidal thickness changes At BMC, research is always in progress. after vitrectomy with silicone oil tamponade for proliferative vitreoretin- Learn more biomedcentral.com/submissions opathy retinal detachment. Retina. 2017;37(11):2124–9.

Journal

International Journal of Retina and Vitreous – Springer Journals

Published: Dec 20, 2021

Keywords: Rhegmatogenous retinal detachment; Silicone oil; Retinal layer thickness; Choroidal thickness

There are no references for this article.