27-gauge direct endoillumination–assisted scleral buckling using a heads-up 3D visualization system: surgical feasibility and outcomes in primary rhegmatogenous retinal detachment

Abstract

Background

To evaluate the feasibility, anatomical outcomes, visual results, and safety profile of a modified scleral buckling (SB) technique utilizing 27-gauge direct endoillumination under a heads-up three-dimensional (3D) visualization system for primary rhegmatogenous retinal detachment (RRD).

Methods

This retrospective case series included consecutive patients with primary RRD who underwent modified SB surgery at a tertiary referral center. Retinal breaks were identified using a 27-gauge endoillumination probe through a valved trocar under a heads-up 3D visualization system with a non-contact wide-angle viewing lens. Anatomical success after a single surgery was the primary outcome. Secondary outcomes included final best-corrected visual acuity (BCVA) and surgical complications.

Results

Twenty-seven eyes of 27 patients were included, with a mean age of 34.6 ± 12.1 years. Fifteen eyes (56%) were macula-off at presentation. Single-surgery anatomical success was achieved in all cases during a mean follow-up of 9.4 ± 5.9 months. Mean BCVA improved significantly from 0.57 ± 0.55 logMAR preoperatively to 0.21 ± 0.30 logMAR postoperatively (p < 0.001). Twelve eyes (44.4%) achieved a final visual acuity of 20/20. No major intraoperative or postoperative complications, including endophthalmitis, iatrogenic retinal breaks, cataract formation, or vitreous prolapse, were observed.

Conclusions

This modified SB technique using 27-gauge direct endoillumination and heads-up 3D visualization appears to be a feasible and safe approach for selected cases of primary RRD, demonstrating encouraging anatomical and functional outcomes. However, given the retrospective design, small sample size, and absence of a control group, further prospective comparative studies are warranted to validate these findings.

Introduction

Rhegmatogenous retinal detachment (RRD) is one of the most common surgical indications in the vitreoretinal field [1]. Functional success after surgery correlates with macular status, duration of macular involvement, and the location of retinal breaks [1, 2]. Thus, appropriate surgical strategy remains essential.

Scleral buckling (SB) had been the gold standard for RRD for over 60 years [1]. The procedure involves identifying and treating retinal breaks using indirect ophthalmoscopy, combined with cryopexy and external support, which can be technically demanding and time-consuming. With the advent of pars plana vitrectomy (PPV), surgical preference has shifted toward PPV due to its efficiency and predictability. However, in young patients, phakic eyes, and selected simple RRD cases, SB remains an important treatment option [2, 3].

This study introduces a modified SB technique incorporating 27-gauge endoillumination and heads-up 3D visualization to enhance surgical visualization while maintaining the advantages of conventional SB.

Methods

This retrospective case series included consecutive patients who underwent modified SB surgery for primary RRD at China Medical University Hospital, the largest tertiary medical center in central Taiwan, from December 1, 2021, to December 1, 2024. This study was approved by the Institutional Review Board of China Medical University Hospital (IRB No. CMUH114-REC1-068). The requirement for informed consent was waived by the Institutional Review Board due to the retrospective nature of the study and the use of anonymized clinical data. All procedures were conducted in accordance with the Declaration of Helsinki.

Inclusion criteria were as follows: (1) primary rhegmatogenous retinal detachment; (2) retinal breaks considered suitable for scleral buckling, typically located anterior to the equator; and (3) absence of advanced proliferative vitreoretinopathy (PVR grade C or higher).

Exclusion criteria included: (1) complex retinal detachment requiring pars plana vitrectomy; (2) presence of advanced PVR (≥ grade C); and (3) cases deemed unsuitable for scleral buckling based on predefined surgical considerations, including posterior pole breaks, extensive vitreoretinal traction, or anticipated need for pars plana vitrectomy.

Lens status (phakic or pseudophakic) was not used as a selection criterion in this study.

All surgeries were performed by a single surgeon (Hsieh, MC). The modified scleral buckling technique was performed under a heads-up 3D visualization system (ARTEVO 800, Carl Zeiss Meditec) combined with a non-contact wide-angle viewing system (Resight 700, Carl Zeiss Meditec).

A 27-gauge valved trocar was inserted 3.5–4.0 mm posterior to the limbus. The insertion site was selected based on the location of the suspected retinal breaks, typically positioned opposite to the break or at a location that allowed optimal ergonomics for manipulation of the handheld light probe.

A 27-gauge endoillumination probe (Alcon Constellation) was then introduced through the trocar. The probe was gently advanced into the mid-vitreous cavity and directed toward the peripheral retina. Care was taken to avoid a horizontal trajectory and to maintain awareness of the crystalline lens at all times, similar to standard pars plana vitrectomy techniques, in order to minimize the risk of lens touch.

With the aid of the non-contact wide-angle viewing system and heads-up 3D visualization, minimal scleral indentation was sufficient to achieve clear visualization of the peripheral retina, including the pars plana region. Retinal breaks were identified using a combination of controlled indentation and dynamic directional illumination with the handheld light probe. The use of the 3D visualization system allowed high magnification without loss of image clarity and enabled real-time visualization for the entire surgical team.

After localization of the breaks, external cryopexy was applied. Segmental scleral buckling was then performed using 5 − 0 Mersilene sutures, followed by placement of a sponge to achieve adequate external indentation. External drainage was performed at the surgeon’s discretion. Figure 1 demonstrates the step-by-step surgical procedure.

Fig. 1.

Step-by-step demonstration of the modified scleral buckling technique using 27-gauge direct endoillumination under a heads-up 3D visualization system. (A) Conjunctival peritomy is performed according to the location of the suspected retinal break. (B) A 27-gauge valved trocar is inserted 3.5–4.0 mm posterior to the limbus at a location selected based on surgical ergonomics and the position of the retinal break. (C) A handheld light pipe is introduced through the trocar. With controlled scleral indentation, the retinal break is identified using dynamic directional illumination. (D) After localization of the break, 5 − 0 Mersilene sutures are placed to mark the buckle position. (E) A sponge is inserted and secured to create a segmental scleral buckle. (F) The appearance after completion of the procedure, demonstrating a clean surgical field and stable buckle positioning, without vitreous prolapse, lens trauma or other intraoperative complications

The primary outcome measure was the anatomical success rate after single surgery, while secondary outcomes included final visual acuity (VA) and surgical complications. VA was assessed using Snellen’s charts and converted to logarithm of minimal angle of resolution (logMAR) units for statistical analysis. Non-numerical visual acuities were categorized as follows: counting fingers (CF) = 1.7 logMAR, hand movement (HM) = 2.0 logMAR, light perception (LP) = 2.3 logMAR, and no light perception (NLP) = 3.0 logMAR [4]. Phthisis, evisceration, and enucleation were also represented by 3.0 logMAR [4].

Statistical analysis was conducted using SPSS 26.0 software (SPSS, Chicago, IL, USA). Mann-Whitney U tests were employed for continuous variables, while chi-square tests or Fisher’s exact tests were used for categorical variables. Factors influencing final visual outcomes were analyzed using linear regression, with a significance level set at P < 0.05.

Results

During the study period, 27 eyes from 27 patients were included. Among these, 17 were right eyes and 10 were left. Fifteen eyes (56%) were classified as macular off, while 12 (44%) were macular on. The majority had one retinal break (N = 21, 77%), with four having two breaks and two cases exhibiting multiple breaks. Breaks were located in the superior part in 12 cases, in the inferior part in 7 cases, and in the temporal/nasal area in 7 cases, with one case having breaks almost circumferentially. External drainage was performed in only 3 cases (11%). Detailed characteristics are summarized in Table 1. The mean age of the participants was 34.56 ± 12.08 years, with a median age of 33 years (ranging from 18 to 54 years). Notably, 11% of the patients were younger than 20 years old, with the majority (88%) under 50.

Table 1.

Detail characteristics of the enrolled population

id	age	eye	macula	Break	Breaks location	Pre-op VA*	drain	Post-op VA	F/U+
1	26	OS	on	1	nasal upper	0.05	-	0.00	5
2	18	OS	On	1	temporal upper	0.00	-	0.00	3
3	36	OS	Off	1	temporal upper	0.70	-	0.00	18
4	21	OS	Off	1	temporal	0.30	-	0.10	7
5	22	OD	Off	1	upper	0.52	-	0.00	32
6	42	OD	On	1	temporal lower	0.00	-	0.00	12
7	39	OS	On	1	temporal	0.70	-	0.30	14
8	19	OD	Off	1	temporal	0.40	-	0.40	12
9	46	OD	Off	1	temporal	1.00	-	0.15	12
10	48	OS	Off	2	temporal	0.70	-	0.70	12
11	20	OD	Off	2	upper	1.00	-	1.00	2
12	42	OD	On	1	temporal upper	0.15	-	0.00	11
13	19	OD	On	2	temporal	0.15	-	0.00	12
14	21	OD	Off	> 2	temporal	0.52	-	0.30	7
15	31	OD	On	1	Temporal lower	0.00	-	0.00	12
16	49	OD	On	1	Temporal lower	0.00	Yes	0.00	12
17	54	OD	On	1	Temporal upper	0.40	-	0.10	6
18	44	OS	Off	1	Upper	1.70	Yes	0.70	8
19	41	OD	Off	1	Upper	0.52	-	0.05	8
20	29	OD	On	1	Temporal upper	0.00	-	0.00	4
21	51	OS	Off	1	Temporal lower	1.70	-	1.00	5
22	30	OD	Off	> 3	360 DEGREE	2.00	Yes	0.30	4
23	33	OD	On	1	Lower	0.40	-	0.40	2
24	51	OS	Off	1	Lower	0.70	-	0.15	1
25	51	OS	Off	2	Upper	0.70	-	0.15	1
26	26	OD	On	1	Lower	0.00	-	0.00	4
27	24	OD	off	1	nasal upper	1.00	-	0.00	4

*VA: visual acuity in logMAR; ⁺f/u: follow-up preiod

All cases achieved anatomical success during the follow-up period (9.44 ± 5.89 months). The mean pre-operative VAlogMAR was 0.57 ± 0.55, while the mean post-operative VAlogMAR significantly improved to 0.21 ± 0.30 (Wilcoxon test, p < 0.001). Twelve cases (44.44%) reached a VA of 1.0 on the Snellen chart by the end of the follow-up, while only 5 cases did not reach 0.5 on the Snellen chart post-surgery, all of whom were macular off prior to the procedure.

Linear logistic regression identified baseline vision as the risk factor influencing final VA, indicating that poorer baseline vision correlated with worse final VA. Neither macular status nor the location and number of breaks influenced the outcomes (Table 2).

Table 2.

Linear regression analysis of factors associated with final visual acuity

factor	beta	T	Significant	95%CI of B
				Lower limit	Upper limit
Age	-0.006	-0.029	0.977	-0.010	0.010
eye	0.078	0.431	0.672	-0.186	0.281
Macular on prior to op	0.016	0.064	0.949	-0.298	0.317
breaks	0.139	0.642	0.529	-0.156	0.294
drain	-0.191	-0.974	0.343	-0.567	0.208
Breaks location	-0.116	-0.622	0.542	-0.164	0.089
Pre-op VA	0.723	2.582	0.019	0.073	0.714
f/u	-0.118	-0.684	0.503	-0.022	0.011

*VA: visual acuity; ⁺f/u: follow-up preiod

No cases of proliferative vitreoretinopathy, epiretinal membranes, or cataract formation were noted. Surgical complications such as endophthalmitis, iatrogenic retinal breaks, traumatic cataracts, strabismus, refractive changes, or scleral perforation were also absent. Furthermore, trocar wounds healed well without suturing, and no vitreous prolapse occurred.

Discussion

RRD occurs due to retinal breaks, leading to subretinal fluid accumulation and separation of the neurosensory retina from the retinal pigment epithelium (RPE) [5]. Historically, scleral buckling had been the gold standard for repairing RRD, with success largely dependent on accurately locating and indenting retinal breaks via indirect ophthalmoscopy, a process that is both time-consuming and challenging to the operator. With the advent of PPV, SB has become less favored [1, 2, 3, 5], as PPV offers enhanced efficacy and safety with reduced surgical times. To be clear, according to the American Society of Retinal Specialists Preferences and Trends Survey, scleral buckling surgery for primary RRD was preferred by only 28.7% of vitreoretinal surgeons in the United States and 39.5% in European countries [6].

Nevertheless, SB continues to play a crucial role, particularly in minimizing cataract formation [2, 3, 5]. Recent advancements have incorporated surgical microscopes and chandelier lighting systems [6–9], which makes retinal break identification more easily and have shown promising primary anatomical success rates up to 91.7% with shorter surgical durations [8, 9]. However, chandelier systems can lead to complications such as vitreous prolapse and lens contact [8], which are less common in traditional SB.

In our technique, we replace the chandelier with a 27-gauge light pipe with a valve trocar to decrease risk of vitreous prolapse and enhance the control of direct visualization during the surgery. The probe was consistently maintained within the mid-vitreous cavity, and insertion was performed with an oblique rather than horizontal trajectory to minimize the risk of lens contact. This approach is analogous to standard pars plana vitrectomy techniques. The potential intraocular risks associated with the use of a handheld endoillumination probe was therefore be carefully considered, particularly in phakic eyes.

Compared with chandelier-assisted scleral buckling, the use of a handheld light pipe provides greater flexibility in directing illumination. Unlike a fixed chandelier system, the light probe can be dynamically adjusted to target different peripheral regions, which may facilitate faster and more efficient identification of retinal breaks.

Additionally, because the light probe is only introduced transiently during the identification phase, the duration of intraocular manipulation is relatively short, potentially reducing the risk of vitreous traction or related complications. Importantly, no cases of lens injury, cataract formation or vitreoretinal complications were observed during the follow-up period in this series. Therefore, lens status (phakic or pseudophakic) was not considered a limiting factor in patient selection in our cohort.

Although previous studies have described the use of 27-gauge endoilluminator-assisted scleral buckling combined with a wide-field viewing system [10], our technique further incorporates a heads-up 3D visualization system, which provides additional advantages: high-resolution magnified imaging allows clear visualization of the peripheral retina with minimal scleral indentation, even in challenging regions such as areas near the pars plana. In addition, the shared display enables simultaneous visualization by assistants, fellows, and residents, which may enhance intraoperative communication and teaching. Furthermore, the heads-up system improves surgeon ergonomics by reducing the need for prolonged forward-leaning posture. Intraoperative optical coherence tomography also provides valuable insights for determining the need for external drainage, which may reduce the risk of unnecessary drainage or scleral perforation. However, due to the retrospective nature of this study, surgical time and learning curve were not quantitatively assessed. In addition, as the primary surgeon routinely performs surgery using a 3D system, it is difficult to directly compare the learning curve between conventional microscopy and 3D visualization. These factors represent limitations of the current study.

Our findings indicate favorable anatomical outcomes, with no cases of redetachment observed during the follow-up period, with encouraging visual outcomes: 12 of 27 cases achieved 20/20 vision at the end of follow-up, despite a 56% pre-operative macular off rate. In the previously published article, only 53.6% of eyes with macular off RRD achieved a final visual acuity of 20/50 or better by having traditional SB [11]. Recent study also demonstrated only a mean VA of 6/24 in 6 months follow up after SB [12]. Additionally, 81.5% of cases in our study achieved vision above 20/40. The only significant risk factor for suboptimal visual outcomes was baseline visual acuity, as neither macular status nor the location and number of breaks impacted the final VA.

No surgical adverse effects or complications, including traumatic cataracts or iatrogenic breaks, were recorded. The trocar wounds healed well without the need for suturing, and the risks associated with chandelier endoillumination [8] were effectively decreased through the 27-gauge light pipe approach.

But we have to acknowledge that this study has several limitations. First, the sample size was relatively small. Second, the retrospective design may introduce inherent selection bias. Third, the absence of a control group limits direct comparison with conventional scleral buckling or alternative techniques. Additionally, surgical time and learning curve were not systematically evaluated. Furthermore, the impact of transitioning from conventional microscopy could not be assessed. These limitations may affect the generalizability of our findings, and further prospective comparative studies are warranted.

Conclusion

This study demonstrates that scleral buckling assisted by 27-gauge direct endoillumination and a heads-up 3D visualization system is a feasible and safe technique for selected cases of primary rhegmatogenous retinal detachment. The approach provides encouraging anatomical and functional outcomes while offering potential advantages in visualization and surgical workflow. However, given the retrospective nature of this study and the lack of a control group, further prospective and comparative studies are required to confirm these findings.

Author contributions

MCH performed the surgeries. YJC and MCH analyzed and interpreted the patent data. YJC and MCH wrote the manuscript. HYL, CCC, YWC, Jennifer HCH, and YYT revised the manuscript. All authors read and approved the manuscript.

Funding

Nil.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of China Medical University Hospital (IRB No. CMUH114-REC1-068). The requirement for informed consent was waived by the Institutional Review Board due to the retrospective nature of the study and the use of anonymized clinical data. All procedures were conducted in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.