25-Gauge pars plana vitrectomy combined with air tamponade for primary rhegmatogenous retinal detachment

Abstract

Objective

This study was performed to evaluate the outcomes of 25-gauge (25-G) pars plana vitrectomy (PPV) with air tamponade for primary rhegmatogenous retinal detachment (RRD).

Methods

This retrospective consecutive case series included 126 eyes of 125 patients with primary RRD who underwent 25-G PPV with air tamponade. The patients were followed up for at least 6 months following surgery. The main outcome measures were the primary and final anatomical success rates and postoperative complications.

Results

The mean age of the 125 patients (80 men and 45 women) was 53.7 ± 10.0 years. The mean follow-up period was 8.3 ± 2.2 months (range, 6–18 months). Twenty-four eyes (19.0%) presented with high myopia, and 13 eyes (10.3%) were pseudophakic. Of the 126 eyes, 37 (29.4%) had inferior breaks, 2 (1.6%) had choroidal detachment, and 86 (68.3%) had macular detachment. The single- and final-operation success rates were 96.0% and 100%, respectively. Postoperative complications included macular hole formation in two eyes. During follow-up, secondary cataract surgery was performed in 27 (23.9%) of the 113 phakic eyes.

Conclusion

25-G PPV with air tamponade is effective and safe in treating selected patients with primary RRD with a high anatomical success rate.

Introduction

Rhegmatogenous retinal detachment (RRD) is a common sight-threatening disease. The three key factors for the development of RRD are liquefied vitreous, abnormal vitreoretinal traction that produces a retinal break, and the ingress of fluid into the subretinal space through the retinal break.¹ Surgical treatment of RRD is performed to close and seal all retinal breaks, relieve any traction on the edges of the tears, and reattach the neurosensory retina.

The standard modalities for the treatment of RRD include scleral buckling, pneumatic retinopexy, and pars plana vitrectomy (PPV).^2–4 In recent years, improvements in vitrectomy instrumentation, development of wide-angle viewing systems, and advancements in microincisional techniques have expanded the role of PPV, which has consequently become the most important surgical technique for complex detachments.^2,5,6 After removing the vitreous, long-acting gas or silicone oil is injected into the vitreous cavity to support closure of the retinal breaks and assure successful reattachment of the retina.

Because of legislation changes, long-acting gases have not been commercially available in China since April 2016. The use of silicone oil increases the likelihood of postoperative intraocular pressure (IOP) elevation.⁷ Therefore, air tamponade is a reasonable alternative for the management of RRD.^6,8,9 The much shorter half-life of air allows faster visual recovery and may reduce risk of cataract progression compared with other agents. Recent studies have shown that air tamponade is comparable to long-acting gas in terms of the surgical outcome.^8,10 However, air tamponade is generally indicated for RRDs with small or superior breaks^6,11,12; large, multiple, or inferior breaks always pose a challenge when performing air tamponade. Therefore, we herein report the surgical outcomes of consecutive patients with primary RRD treated with 25-gauge (25-G) PPV using air tamponade only.

Materials and methods

The medical records of consecutive patients who were treated for primary RRD from May 2018 to November 2021 were reviewed. Primary RRD was considered newly developed RRD. The exclusion criteria were grade ≥C2 proliferative vitreoretinopathy (PVR), traumatic RRD, myopic macular hole-associated retinal detachment, pediatric RRD (birth to 18 years of age), a history of vitreoretinal surgery, other serious eye diseases, and less than 6 months of follow-up. Written informed consent was obtained from all patients who were included in the study and all appropriate participant consent forms were obtained. This study adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Committee of the Second People’s Hospital of Foshan (approval number: KJ2019009, date of approval: 5 May 2019). Attached please find the approval document. The patients’ key information will not be published to conceal their identity. The reporting of this study conforms to the STROBE guidelines.¹³

Comprehensive ophthalmic examinations were performed, including Snellen best corrected visual acuity (BCVA), noncontact tonometry, slit-lamp examination, and optical coherence tomography (OCT). The characteristics (number, size, type, and location) of retinal breaks and the procedural details (e.g., combined lens extraction and drainage retinotomy) were assessed from the surgical records. PVR was diagnosed according to the classification established by the American Retina Society Terminology Committee (1983).¹⁴

All procedures were performed under retrobulbar anesthesia by a single experienced surgeon (P.S.). Standard three-port 25-G PPV was performed using the Alcon Constellation Vision System (Alcon Laboratories, Fort Worth, TX, USA) and a noncontact wide-angle viewing system (RESIGHT 700; Carl Zeiss Meditec AG, Jena, Germany). After central and peripheral vitreous removal, all eyes underwent 360° scleral indentation to shave the peripheral vitreous up to the ora serrata. After complete release of vitreous traction from the retinal tears, the retina was flattened by fluid–air exchange, and the subretinal fluid (SRF) was drained off through the original breaks or drainage retinotomy. All retinal breaks were then treated by endolaser photocoagulation. Cryotherapy, 360° prophylactic laser photocoagulation, or perfluorocarbon liquid (PFCL) was not applied in all cases. The sclerotomies were carefully closed at the end of surgery. Phacoemulsification and intraocular lens implantation were performed simultaneously if lens opacification was visually significant.

Postoperatively, the patients were instructed to maintain an alternative prone or lateral position. Those patients with inferior breaks maintained an alternative supine or lateral position at least 12 hours per day for no less than 5 days. The axial length of highly myopic eyes was measured by a Lenstar LS 900 with software version 1.1 (Haag-Streit AG, Köniz, Switzerland).

Statistical analysis

The data were analyzed using SPSS for Windows 21.0 (IBM Corp., Armonk, NY, USA) and are reported as mean ± standard deviation or n (%). BCVA in the Snellen value was converted to the logarithm of the minimum angle of resolution (logMAR). Visual acuity of light perception, hand movements, and counting fingers was assigned 2.9, 2.6, and 2.3, respectively.¹⁵ BCVA was compared using the Mann–Whitney U test. A P value of <0.05 was considered statistically significant.

Results

Baseline characteristics

In total, 126 eyes of 125 consecutive patients (80 men and 45 women) were analyzed. The patients’ baseline characteristics are summarized in Table 1. Their mean age was 53.7 ± 10.0 years (range, 30–78 years), and their mean follow-up was 8.3 ± 2.2 months (range, 6–18 months). Their mean preoperative IOP was 13.0 ± 2.7 mmHg (range, 7.7–18.9 mmHg). There were 113 (89.7%) phakic eyes and 13 (10.3%) pseudophakic eyes. The posterior lens capsule was intact in five eyes (38.5%) and broken with a sulcus lens in one eye (7.7%). Yttrium–aluminum–garnet laser capsulotomy was performed in the other seven eyes (53.8%).

Table 1.

Clinical characteristics of 126 eyes of 125 patients treated by 25-gauge PPV with air tamponade for primary RRD.

Characteristic
Sex, male/female	80 (64.0%)/45 (36.0%)
Age, years	53.7 ± 10.0 (range, 30–78)
Disease course, days	15.1 ± 15.2 (median, 10; range, 2–90)
Preoperative logMAR BCVA, Snellen equivalent	1.49 ± 0.94 (20/618)
Lens status, phakic/pseudophakic	113 (89.7%)/13 (10.3%)
Vitreous status, transparent/haze/hemorrhage	43 (34.1%)/42 (33.3%)/41 (32.5%)
High myopia, number of eyes	24 (19.0%)
Axial length of highly myopic eyes, mm	27.64 ± 1.51 (range, 25.73–31.12)
PVR grading, A or none/B/C1	61 (48.4%)/57 (45.2%)/8(6.3%)
RRD extent, clock hours	5.5 ± 2.4 (range, 1–12)
Inferior quadrants involved	74 (58.7%)
Macula involved, on/off	40 (31.7%)/86 (68.3%)
Drainage retinotomy, no/yes	75 (59.5%)/51 (40.5%)

Data are presented as n (%) or mean ± standard deviation unless otherwise noted.

PPV, pars plana vitrectomy; RRD, rhegmatogenous retinal detachment; logMAR, logarithm of minimum angle of resolution; BCVA, best corrected visual acuity; PVR, proliferative vitreoretinopathy.

Overall, there were 240 breaks, of which 185 (77.1%) were located in the superior retina (between the 8- and 4-o’clock meridians, including the 8- and 4-o’clock meridians) and 55 (22.9%) were located in the inferior retina (between the 4- and 8-o’clock meridians). Of the 240 breaks, 175 (72.9%) were horseshoe tears and 65 (27.1%) were atrophic holes. The clinical characteristics of the retinal breaks are summarized in Table 2. Of the 126 eyes, 37 (29.4%) had inferior breaks, 2 (1.6%) had choroidal detachment, and 86 (68.3%) had macular detachment. Ten eyes underwent PPV combined with cataract surgery.

Table 2.

Clinical characteristics of retinal breaks.

Characteristic
Breaks, n	1.9 ± 1.5 (range, 1–12)
Single/two/three or more (eyes)	63 (50.0%)/36 (28.6%)/27 (21.4%)
Type of breaks
Horseshoe tears	85 (67.5%)
Atrophic holes	19 (15.1%)
Horseshoe tears + atrophic holes	22 (17.5%)
Size of breaks
Small breaks, ≤2 PD	81 (64.3%)
Medium breaks, >2 to 5 PD	36 (28.6%)
Large breaks, >5 PD	9 (7.1%)
Location of breaks
Superior	88 (69.8%)
Inferior	11 (8.7%)
Superior + inferior	26 (20.6%)
Superior + macular hole	1 (0.8%)

Data are presented as n (%) or mean ± standard deviation unless otherwise noted.

Anatomical outcome

Single-operation success was achieved in 121 eyes (96.0%). Table 3 shows the clinical characteristics of the five eyes that developed re-detachment. Of these, the re-detachments in four eyes were attributed to new breaks and that in the remaining eye was associated with postoperative progression to PVR 3 weeks following surgery. Silicone oil tamponade was performed as a salvage treatment in all five eyes, and none developed re-detachment from the time of silicone oil removal to the last follow-up.

Table 3.

Clinical characteristics of patients with primary reattachment failure.

Characteristics	Case 1	Case 2	Case 3	Case 4	Case 5
Sex	Male	Male	Female	Male	Male
Age (years)	49	44	74	64	44
Disease course (days)	10	5	21	16	4
High myopia	No	Yes (AL, 26.74 mm)	No	No	Yes (AL, 29.55 mm)
Preoperatively
Number of quadrants involved	5	9	6	6	6
Inferior quadrants involved	Yes	Yes	Yes	Yes	Yes
Number of retinal breaks	4	3	1	1	3
Location of retinal breaks	Inferior	Superior	10 o’clock of right eye	9 o’clock of right eye	2 in 9 o’clock and 1 in 7 o’clock of right eye
Type of retinal breaks	Horseshoe tears	Horseshoe tears	Atrophic hole	Atrophic hole	1 horseshoe tear + 2 atrophic holes
Size of retinal breaks (PD)	1	3	0.5	0.3	1
Macula involved	Off	Off	Off	Off	Off
PVR	None or A	None or A	B	B	None or A
Final success	Yes	Yes	Yes	Yes	Yes

PVR, proliferative vitreoretinopathy; PPV, pars plana vitrectomy; AL, axial length.

Visual acuity outcome

The mean postoperative logMAR BCVA (0.38 ± 0.27; Snellen equivalent, 20/48) was significantly better than the mean preoperative logMAR BCVA (1.49 ± 0.94; Snellen equivalent, 20/618) (P < 0.001). Of the 40 eyes with macula-on RRD, the mean preoperative logMAR BCVA (Snellen equivalent, 20/92) was 0.66 ± 0.86 (range, 0.0–2.6; median, 0.2) and the mean postoperative logMAR BCVA (Snellen equivalent, 20/26) was 0.11 ± 0.16 (range, 0.0–0.6; median, 0.05). Of the 86 eyes with macula-off RRD, the mean preoperative logMAR BCVA (Snellen equivalent, 20/1517) was 1.88 ± 0.70 (range, 0.6–2.6; median, 2.3) and the mean postoperative logMAR BCVA (Snellen equivalent, 20/65) was 0.51 ± 0.21 (range, 0.1–1.1; median, 0.5). Both postoperative BCVAs showed significant improvements (P < 0.001).

Complications

The air was completely absorbed 9 to 12 days following surgery. Postoperative transient IOP elevation (>21 mmHg) developed in 10 highly myopic eyes 1 week after surgery and was well controlled using topical medications without any permanent damage. One highly myopic eye (Case 20) complicated with grade C1 PVR developed a macular hole 1 week after surgery. A second surgery with internal limiting membrane peeling and air tamponade was then performed. Two weeks postoperatively, OCT showed resolution of the macular hole, and the final BCVA was 1 logMAR (Snellen equivalent, 20/200). During follow-up, 27 (23.9%) of the 113 phakic eyes underwent secondary cataract surgery.

Discussion

In this retrospective study, 25-G PPV with air tamponade achieved high success rates in the treatment of primary RRD. The single-operation success rate of retinal reattachment was high at 96.0% and was accompanied by with early visual recovery and few complications.

Intraocular tamponade serves as a barrier, preventing fluid in the vitreous cavity from entering the subretinal space. Once a chorioretinal adhesion has formed, the barrier is no longer needed. Therefore, a short-acting gas such as air might be a reasonable alternative to tamponade. The intraocular air bubble can seal retinal breaks, preventing accumulation of fluid in the subretinal space. Air usually remains in the eye for about 1 week,¹⁶ thus allowing quicker visual recovery and fewer complications such as PVR and elevated IOP.^17,18 Theoretically, the chorioretinal adhesion establishes within 24 hours in eyes without SRF.¹⁹ After this time, fluid will not enter the subretinal space through breaks. Therefore, long-acting tamponade may be excessive. Additionally, the short duration of air in the eye may mitigate the development of an epiretinal membrane or postoperative PVR.²⁰

Martinez-Castillo et al.²¹ reported a high reattachment rate (93.3%) in their prospective study of primary pseudophakic RRD. Li et al.⁸ retrospectively analyzed 59 eyes and found that the single-operation success rate was 94.9%. A randomized comparative study also showed that air tamponade had effects equivalent to those of C₃F₈ tamponade in the treatment of RRD with inferior breaks.¹⁷ Consistent with previously reported results, our findings (single-operation success rate of 96.0%, final-operation success rate of 100%) provide further proof that air tamponade is adequate for the formation of a stable chorioretinal adhesion and is effective in treating selected cases of primary RRD.

The location of retinal breaks is considered closely related to the surgical outcomes.⁸ Tan et al.¹¹ reported that RRD involving the inferior quadrants had significantly lower primary success rates when using air tamponade than when using SF₆ tamponade. Gas tamponade is superior to air tamponade for RRD involving the lower quadrants. In contrast, Martinez-Castillo et al.^21,22 demonstrated in their prospective study that PPV with air tamponade was effective in the management of pseudophakic RRD with inferior breaks without a facedown position. Gozawa et al.²³ measured the intraocular gas contact rates of the retina using magnetic resonance imaging. They reported that the gas can adequately support inferior parts of the retina in the supine position with a gas contact rate exceeding 90%. Hence, we instructed the patients with inferior breaks to maintain an alternative supine or lateral position, allowing the air bubble in the vitreous cavity to adequately seal inferior breaks. This study included 37 cases (29.4%) with at least one inferior break (between the 4- and 8-o’clock meridians). Two cases (5.4%) re-detached because of development of new breaks. In line with previous studies,^10,17,21,22 our data suggest that PPV with air tamponade can attain satisfactory outcomes for patients with primary RRD, even those with inferior breaks.

According to the literature, the single-operation success rate of air tamponade ranges from 70.0% to 94.9%.^{8,11,12,17,21,22} This discrepancy could be due to diverse surgical techniques and different patient selection criteria. The pathogenesis of RRD involves vitreoretinal tractional forces that result in a full-thickness retinal break.²⁴ Therefore, in this study, the vitreous was removed as completely as possible. A 360° scleral indentation with a wide-angle viewing system is essential for complete shaving of the peripheral vitreous. Triamcinolone was also routinely used to detect the residual vitreous in case of insufficient vitreous liquefaction. Inferior breaks always present a challenge when performing air tamponade because the residual fluid may seep into the original breaks by gravity. However, if the SRF does not reach the original breaks, short-acting tamponade can adequately limit the occurrence of re-detachment (the chorioretinal adhesion was comparable to that of the normal retina at 24 hours after photocoagulation).^10,21,22 Therefore, complete SRF drainage is essential in these cases. Appropriately adjusting the position of the head during fluid–air exchange could help the fluid to flow out more easily. For patients with a small break in the periphery, performing a posterior drainage retinotomy can help fully drain the fluid. Postoperatively, the supine position may provide better coverage than the prone position for inferior breaks located anterior to the equator by the formation of intraocular air bubbles.

In the present study, we avoided excessive interventions to prevent unnecessary complications. PFCL is often employed in vitreoretinal surgery to facilitate the peripheral vitreous shaving and SRF drainage. However, inadvertent subretinal PFCL retention is a common complication. Li et al.⁸ suggested that 360° laser coagulation could reduce the risk of re-detachment. It is important to identify and treat all retinal breaks. Nevertheless, excessive laser therapy may cause retinal necrosis and small, difficult-to-find retinal holes, leading to re-detachment. Such holes are difficult to identify within the patches of chorioretinal atrophy. Moreover, we excluded eyes affected by complex RRD with giant retinal tears, myopic macular holes, and grade ≥C2 PVR, which also contributed to our favorable outcomes. We speculate that short-acting tamponade is insufficient to establish a firm chorioretinal adhesion in these demanding cases; however, further studies are required to evaluate the effectiveness of air tamponade in these cases.

As previously demonstrated,^11,17,25 the leading cause of primary failure is new breaks, which occurred in four eyes of our study. Reopening of the original breaks with development of a subretinal band occurred in another eye. Two highly myopic eyes developed a macular hole postoperatively. One eye underwent epiretinal membrane peeling, which carried a risk of damage to the thinned central foveal tissue. The other eye subsequently re-detached at 2 weeks postoperatively. Because a small amount of submacular fluid would be left at the end of surgery, a potential explanation may be that the high surface tension of air presses against the fovea and the submacular fluid. A previous study indicated that exposure of the lens to abnormally high oxygen levels can lead to nuclear sclerosis in vitrectomized eyes.²⁶ In our study, 27 eyes (23.9%) underwent cataract surgery during follow-up. The quicker resolution of air in the vitreous cavity could result in less disturbance to the lens and may decrease the risk of cataract progression.

This study had two main limitations. First, its retrospective design has an inherent risk of selection bias. Second, a control group treated with long-acting gas tamponade was not included because these gases were no longer commercially available in China at the time of this writing.

Conclusions

This study shows that 25-G PPV with air tamponade has a satisfactory success rate in repairing primary RRD, even in demanding cases with grade C1 PVR, choroidal detachment, and large, multiple, or inferior retinal tears. In addition, this technique provides faster visual recovery, has a shorter positioning period, and is associated with fewer complications. The results of our study indicate that air tamponade is effective and safe in the management of selected patients with primary RRD. Prospective comparative studies to assess the effectiveness of this technique for complex RRD are needed.

Representative cases

A 57-year-old man with high myopia presented with a 2-week history of darkness and decreased vision in the left eye. The preoperative evaluation revealed poor vision in the left eye, with BCVA of 2.6 logMAR (hand movements). The axial length of the left eye was 27.86 mm (measured by a Lenstar system 2 weeks after surgery). Fundoscopy revealed a superior RRD (range: 5 clock hours) with a large superotemporal U-shaped tear (7.5 PD), and OCT showed macular involvement (Figure 1(a)). Seven days after surgery, the retina reattached (Figure 1(b)). Two months after surgery, the BCVA of the left eye was 0.5 logMAR (Snellen equivalent, 20/63) (Figure 1(c)).

Figure 1.

Preoperative and postoperative fundus photographs of a patient with RRD (Patient No. 15) undergoing 25-G PPV with air tamponade. (a) Superior RRD (range: 5 clock hours) with a large U-shaped tear (7.5 PD) and vitreous hemorrhage. (b) At the 1-week postoperative follow-up, the retina had reattached. An air bubble was present in the vitreous cavity and (c) At the 2-month postoperative follow-up, the retina had reattached and firm chorioretinal adhesion had been induced by laser coagulation.

A 42-year-old man with high myopia was referred to our clinic. He complained of decreased vision of the left eye, which had persisted for about 2 weeks before his initial visit to our hospital. The patient’s BCVA was 1.1 logMAR (Snellen equivalent, 20/250). The axial length of the left eye was 26.58 mm (measured by a Lenstar system 1 month after surgery). Fundoscopy revealed an inferior RRD (range: 5 clock hours) with an inferonasal horseshoe tear (2 PD) and three atrophic holes in the inferior degenerative area, and OCT showed macular involvement (Figure 2(a)). Six days after surgery, the retina reattached (Figure 2(b)). Three months after surgery, the BCVA of the left eye was 0.3 lozgMAR (Snellen equivalent, 20/40) (Figure 2(c)).

Figure 2.

Preoperative and postoperative fundus photographs of a patient with RRD (Patient No. 7) undergoing 25-G PPV with air tamponade. (a) Inferior RRD (range: 5 clock hours) with an inferonasal horseshoe tear (2 PD) and three atrophic holes in the inferior degenerative area. (b) At the 6-day postoperative follow-up, the retina had reattached. An air bubble was present in the vitreous cavity and (c) At the 3-month postoperative follow-up, the retina had reattached and firm chorioretinal adhesion had been induced by laser coagulation.

Availability of data and materials

The data can be obtained by request to the corresponding author. There are no confidential data and no restriction on accessing the original data.

ORCID iD

Peiyang Shen https://orcid.org/0000-0001-6679-1691