Background
Giant retinal tears (GRT) are defined as full-thickness breaks of three clock hours or more in circumferential extent associated with a posterior vitreous detachment (PVD) [
1,
2]. Mostly GRTs arise idiopathic and affect middle-aged men [
1]. Apart from spontaneous cause several risk factors like trauma, high myopia, large areas of lattice degeneration or hereditary vitreoretinopathies have been identified [
2,
3]. The management of a GRT-associated rhegmatogenous retinal detachment (RRD) is challenging for vitreoretinal (VR) surgeons [
4]. In macula-on situations the preoperative visual acuity can still be quite good despite the presence of an extensive tear in the outer periphery of the retina. The standard procedure for RRDs caused by a GRT is pars-plana-vitrectomy (PPV) with endotamponade [
2,
5]. Because of the extent of the break, the risk of slippage of its posterior edge and the higher rate of proliferative vitreoretinopathy (PVR) silicone oil (SO) tamponade is often used [
1]. SO is a well-established long-lasting endotamponade, which has been used for decades in difficult surgical situations. However, SO also has several disadvantages e.g. the need for a second PPV for its removal, the event of unexplained visual loss (UVL), the risk of secondary glaucoma and problems with SO emulsification and SO-associated keratopathy [
6]. Therefore, some VR surgeons prefer a gas tamponade even in complex situations with a GRT [
7]. There is no consensus whether SO or gas should be preferred in GRT detachments [
3,
5,
7]. Therefore, we decided to evaluate our cases of GRT-associated RRDs, which had VR surgery at our clinic within the last fifteen years, reading their functional and anatomical results depending on the used endotamponade. To enhance comparability only cases with macula-on status were included.
Methods
All consecutive cases, with a macula-on GRT detachment, which had VR surgery at our unit between 2007 and 2022 were retrospectively assessed. The medical records and surgical logbooks were studied in detail. The analysis comprised baseline characteristics (age, sex, site, spherical equivalent, lens status, extent and localisation of GRT), surgical parameters (cut-suture-time, laser spots, cryopexy, endotamponade) and imaging data obtained from optical coherence tomography (Spectralis®, Heidelberg Engineering). The functional outcome measured by best-corrected visual acuity (BCVA) using logarithmic decimal charts at the final follow-up and the anatomical results including redetachment rate, eventual success rate (defined as reattached retina in the absence of any endotamponade at the last follow-up visit), central foveal thickness (CFT) and presence of postoperative complications (cataract, epiretinal membrane, macular edema) were compared with regard to the primary endotamponade (SO versus gas tamponade). To define a fair endpoint postoperative BCVA values at 2 to 4 months after the last VR surgery were assessed. The following cases were included in the study: presence of a GRT ≥ three clock hours, macula-on status and at least 2 months of follow-up after the last VR surgery. Cases with a macula-off situation, PVR grade C or higher, hereditary vitreoretinopathies, history of trauma or previous VR surgery were excluded. The review of the medical records was done at least 6 months after the GRT repair.
The statistical analysis was done with SPSS statistics 28 (IBM, USA). For statistical calculation the pre- and postoperative BCVA equivalents were converted to logMAR values. Categorical variables are presented as frequency counts with percentages and continuous values as means with range and standard deviation (SD). Continuous variables were compared using student’s t-Test. For categorical variables, the x2 test or fisher’s exact test were used. For all statistical tests a p-value < 0.05 was considered significant.
Discussion
GRT-associated detachments are challenging because of the great mobility of the retina, the possible complication of slippage of the posterior edge of the tear with fold formation and the higher risk of PVR [
2]. There are different opinions about the best surgical approach for GRT fixation. Schepens was the first to describe the entity of GRTs and recommended surgical management depending on GRT size and mobility, lens status and presence of PVR. For GRTs smaller than two quadrants without slippage a 360° scleral buckle and cryopexy with or without internal gas tamponade was done in earlier days and resulted in a success rate of 50–61% [
2]. However, a GRT of more than 180° had a success rate of only 11–14% [
2,
3]. With the introduction of PPV in the 1970ies the initial success rate raised to about 86% with a gas tamponade, but required intraoperative manoeuvres like rotation of the operating table and kneeling of the surgeon on the floor below the patient during gas injection [
8]. Still, the secondary development of PVR was an issue and resulted in a redetachment rate of up to 65% [
3,
8]. With the use of SO as endotamponade success rates of up to 93% were reported in the 1990ies [
3]. Recently novel surgical techniques, the introduction of small-gauge-vitrectomy and the use of PFCL has considerably improved the outcome [
9,
10].
Nowadays, a GRT-associated detachment is mostly treated with PPV, PFCL and endotamponade optionally combined with a 360° scleral buckle [
11]. In cases with limited GRT extent lens-sparing vitrectomy is possible (one patient with a three clock hour tear in our cohort). In most cases the PPV is also combined with a lens extraction to attain good peripheral visualization and better accessibility of the vitreous base [
4]. The need for an encircling band is controversial [
12]. While Goezinne et al. reported lower redetachment rates due to encircling band placement in GRTs, other authors found no improvement in recurrence rates with the combination of PPV and a buckle [
11,
13,
14]. In our cohort, additional scleral buckling was not done, which alleviates comparison of the SO and the gas group.
In many centres around the world SO is preferred as primary tamponade [
10,
11,
15]. SO has some advantages like a long-lasting tamponade effect, lower risk of secondary PVR, less slippage and immediate postoperative visualization of the retina and better vision for the patient in the early postoperative phase [
6]. But, SO also has relevant disadvantages like the need for a SO removal procedure, corneal changes, intraocular pressure (IOP) elevation, SO emulsification, macular edema and UVL [
6,
16].. In our series, eyes after primary SO tamponade had a statistically significant worse visual outcome than after gas tamponade, which is in concordance with other reports [
5,
7]. Especially considering the high incidence of UVL in fovea-sparing GRTs treated with SO, Banerjee et al. support the idea of a “move away from silicone oil” towards long-lasting gas especially in eyes with a macula-on situation [
5]. Other reports did not find any significant difference in final BCVA or rate of complications after GRT repair with SO compared to gas tamponade, but in these studies also macula-off detachments were included [
9,
17]. In our series, the eyes after SO tamponade had significantly more postoperative complications, this being mainly generated by the high percentage of UVL (50%). Some authors discuss, whether the duration of SO may be a risk factor for UVL [
18,
19]. In our series, the mean duration of SO tamponade was similar to other reports ranging between 3 and 5 months [
5,
18]. According to the literature, the minimum duration of SO tamponade in eyes, which experienced UVL, was about 3 months [
5,
18]. Our cases with UVL had in general a longer SO tamponade duration (124 days) than eyes without UVL (99 days), but the difference was statistically not significant. Maybe, the quicker removal of SO (within 3 months) could be an option in GRT cases with a macula-on status to reduce the risk of UVL. However, the data availability here is poor to identify the ideal SO duration to prevent UVL and to not risk a higher redetachment rate.
Considering redetachment rate, the overall redetachment rate after GRT repair was 20%, which is similar to other reports [
11,
14,
20]. In agreement with other authors, eyes with a primary SO tamponade had a lower redetachment rate compared to eyes with gas [
20]. The main reason for recurrent RRD in our study was the development of PVR, which is in concordance with other reports [
11]. The use of SO is known to lower the risk of PVR, which could explain the lower redetachment rates in eyes with SO tamponade. In a retrospective series by Li e al. from a centre in the US a comparable redetachment rate of 31% for GRT repair with a primary gas tamponade was found [
21]. Interestingly, Li et al. also included eyes with complex GRTs after trauma or with PVR grade C and achieved a single surgery success rate of 75%, while most cases (81%) were treated with PPV, PFCL and gas. However, functional outcome was not reported separately for eyes with gas versus eyes with SO in this study [
21].
Several studies reporting the outcome of GRT management have been published [
1,
4,
7]. But alongside with Banerjee et al. [
5], our study is the only one, which compares solely the outcome of GRT-associated detachments with a macula-on status. We acknowledge the limitations of our study, which are restricted by its monocentric character, the retrospective design, the inclusion of both eyes in patients with bilateral GRT and the case-specific selection of the endotamponade, which is strongly biased by the intraoperative situation and the experience of the surgeon. A prospective randomized clinical trial would be desirable, but remains difficult due to the rarity of GRTs. The strengths of our study are the balanced baseline characteristics like preoperative BCVA, GRT extent and localisation, which were homogenously distributed between the gas and the SO group.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.