PURPOSE
To compare visual acuity and photoreceptor integrity following pars plana vitrectomy with drainage from the peripheral retinal break(s) (PRB), posterior retinotomy (PR), or perfluorocarbon liquid (PFCL) for macula-off rhegmatogenous retinal detachment.
DESIGN
Retrospective consecutive interventional comparative clinical study.
METHODS
300 consecutive patients (300 eyes) with primary macula-off rhegmatogenous retinal detachment underwent 23-gauge pars plana vitrectomy with subretinal fluid drainage through PRB (n = 100), PR (n = 100), or with PFCL (n = 100). Visual acuity and spectral-domain optical coherence tomography were performed preoperatively and at 3, 6, and 12 months postoperatively. Primary outcomes were visual acuity and discontinuity of the external limiting membrane, ellipsoid zone, interdigitation zone, and retinal pigment epithelium at 1 year.
RESULTS
Baseline characteristics were similar. Single-operation reattachment rates were as follows: PRB 86%, PR 85%, and PFCL 83% ( P = .9). Mean (±SD) logMAR visual acuity at 1 year was greater with PRB and PR compared with PFCL (PRB 0.6 ± 0.5, PR 0.7 ± 0.6, PFCL 0.9 ± 0.6, P = .002). There was an association between drainage technique and discontinuity of the external limiting membrane (PRB 26%, PR 24%, PFCL 44%, P = .001), ellipsoid zone (PRB 29%, PR 31%, PFCL 49%, P < .001), and interdigitation zone (PRB 43%, PR 39%, PFCL 56%, P = .004). There was an association between drainage technique and risk of cystoid macular edema (PRB 28%, PR 39%, PFCL 46%, P = .003) and epiretinal membrane (PRB 64%, PR 90%, PFCL 61%, P < .001).
CONCLUSIONS
PFCL-assisted drainage is associated with worse visual acuity and greater risk of outer retinal band discontinuity and cystoid macular edema compared with PRB or PR. PR had a greater risk of epiretinal membrane compared with PRB and PFCL. PRB had the best outcomes overall. Drainage technique may impact long-term anatomic and functional outcomes.
R hegmatogenous retinal detachment (RRD) is a common indication for vitreoretinal surgery. The estimated annual incidence of RRD is 10.5 (interquartile range 8.1-13.2) per 100 000 persons, with the true risk being influenced by age, lens status, and refraction. , Single-operation anatomic reattachment rates of operating room procedures for RRD range from 84% to 92% depending on the complexity of the detachment and surgical techniques used. ,
Scleral buckle, pars plana vitrectomy (PPV), pneumatic retinopexy, and a combination of scleral buckle and PPV are the most common surgical approaches for RRD today. The PRO study, a recent large retrospective cohort study comparing different methods of RRD repair, demonstrated similar single-operation reattachment rates for PPV (84% success) and scleral buckle (91% success). Currently, PPV remains the most common method for RRD repair in North America. However, few studies have compared techniques for subretinal fluid (SRF) drainage at the time of PPV.
In a recent survey of retinal specialists by the American Society of Retina Specialists, 49% of American respondents indicated that they drain fluid primarily through a peripheral retinal break (PRB), 38% drain via a posterior retinotomy (PR), and 11% use perfluorocarbon liquid (PFCL) routinely. In contrast, PCFL was used routinely by 43% of international respondents, drainage via PRB was used by 45%, and PR was the most infrequent technique internationally at 10%. Although data comparing drainage techniques are sparse, a recent large retrospective cohort study found similar primary anatomic reattachment rates and visual acuity outcomes in eyes undergoing vitrectomy with PR vs PFCL at approximately 1 year. Similarly, when PFCL was compared to drainage through a break in the retina (either PRB or via PR), there were no significant differences in anatomic outcomes or visual acuity at approximately 1 year. Importantly, only 1 study as of this writing has specifically compared drainage through PRB alone to either PR alone or PFCL alone.
Furthermore, a scarcity of data exist on postoperative photoreceptor microstructural integrity with these different methods of internal SRF drainage. It is becoming more apparent that even with similar anatomical reattachment rates, significant differences in outer retinal microstructural integrity occur with different RRD repair techniques. , There is increasing emphasis on maximizing the integrity of the reattachment, which increases the chance of a more favorable functional outcome. The purpose of this study was to compare the microstructural integrity of the outer retina using spectral-domain optical coherence tomography (SD-OCT) and visual acuity outcomes for drainage of SRF through a PRB, PR, or with PFCL during PPV for primary uncomplicated RRD.
METHODS
STUDY DESIGN
A retrospective consecutive cohort study was performed of primary RRD undergoing PPV using different SRF drainage techniques at St Michael’s Hospital/Unity Health Toronto, Canada, between October 2011 and December 2019. This study was approved by the Research Ethics Board at St Michael’s Hospital/Unity Health Toronto in Toronto, Ontario, Canada, and adhered to the tenets of the Declaration of Helsinki.
ELIGIBILITY
Consecutive patients aged ≥18 years with primary macula-off RRD undergoing treatment with PPV and gas tamponade were eligible for the study. Exclusion criteria included prior history of pneumatic retinopexy or scleral buckle in the affected eye, media opacity, use of silicone oil as primary tamponade, history of trauma, and preexisting macular or retinal pathology or proliferative vitreoretinopathy grade B or worse.
SURGERY
All surgeries were performed by 3 experienced vitreoretinal surgeons who used similar techniques other than SRF drainage: R.H.M. (75% PRB, 15% PR, 15% PFCL), D.T.W. (16% PRB, 15% PR 65% PFCL), and A.B. (PRB 9%, 70% PR, 20% PFCL). All surgeries were completed in the same operating rooms with the same 23-gauge vitrectomy systems (Alcon Constellation; Alcon Canada) and widefield operating microscope system (Carl Zeiss Meditec). All surgeries using PFCL used Alcon Perfluoron liquid (Alcon Canada). Gas tamponade was with isoexpansile concentrations of SF 6 or C 3 F 8 with complete gas fill and postoperative face-down positioning ( Table 1 ).
| Characteristic | PRB (n = 100) | PR (n = 100) | PFCL (n = 100) | P Value |
|---|---|---|---|---|
| Age, y, mean (SD) | 58 (12) | 61 (11) | 58 (15) | .9 |
| Sex, n (%) | ||||
| Male | 69 (69) | 77 (77) | 72 (72) | .9 |
| Female | 31 (31) | 23 (23) | 28 (28) | |
| Visual acuity | ||||
| Snellen equivalent | 20/700 | 20/700 | 20/800 | |
| LogMAR, mean (SD) | 1.42 (0.8) | 1.48 (0.75) | 1.53 (0.79) | .6 |
| Patients with moderate myopia (>–3.00 D), % | 14 | 12 | 14 | .9 |
| Patients with high myopia (>–7.00 D), % | 4 | 3 | 4 | .9 |
| Lens status, n (%) | ||||
| Phakic | 43 (43) | 46 (46) | 54 (54) | .7 |
| Pseudophakic | 55 (55) | 53 (53) | 45 (45) | |
| Aphakic | 2 (2) | 1 (1) | 1 (1) | |
| Macular status, n (%) | ||||
| On | 0 (0) | 0 (0) | 0 (0) | n/a |
| Off | 100 (100) | 100 (100) | 100 (100) | |
| Detachment size, n (%) | ||||
| <2 quadrants | 57 (57) | 63 (63) | 65 (65) | .9 |
| >2 quadrants | 43 (43) | 37 (37) | 35 (35) | |
| Subfoveal fluid height, μm | ||||
| Patients with >1500 μm | 25 | 26 | 29 | .9 |
| Mean (SD) in patients with <1500 μm | 752.1 (797.9)* | 362.7 (496.2) | 420.5 (591.2) | .042 |
| Total visits (12 mo), mean (SD) | 12.8 (8.9) | 10.2 (6.5) | 10.1 (6.5) | .9 |
| Anatomic reattachment (single procedure): attached at 12 mo, n (%) | 86 (86) | 85 (85) | 83 (83) | .9 |
| Tamponade, n (%) | ||||
| C3F8 | 80 (80) | 82 (82) | 81 (81) | .9 |
| SF6 | 15 (15) | 14 (14) | 12 (12) | .9 |
| Air | 5 (5) | 4 (4) | 7 (7) | .9 |
| Patients requiring post-PPV laser retinopexy, n (%) | 9 (9) | 9 (9) | 9 (9) | n/a |
DATA COLLECTION AND IMAGE ANALYSIS
We assessed records of 300 consecutive patients (300 eyes, n = 100 PRB, n = 100 PR, n = 100 PFCL) who met the inclusion and exclusion criteria. OCT data were analyzed at 3 (n = 95 PRB, n = 92 PR, n = 92 PFCL), 6 (n = 93 PRB, n = 96 PR, n = 95 PFCL), and 12 (n = 90 PRB, n = 87 PR, n = 91 PFCL) months postoperatively. Out of these 300 eyes, 268 had gradable (signal strength ≥ 5) SD-OCT 5-line raster scans (Cirrus HD-OCT; Carl Zeiss Meditec) acquired at 12±2 months postoperatively. Subgroup analysis of patients with a single-operation reattachment at 12 months was performed for all outcomes (n = 86 PRB, n = 85 PR, n = 83 PFCL).
Outer retinal photoreceptor integrity was assessed by the 2 masked graders (B.R.M. and A.B.L.), with any disagreement adjudicated by a third senior masked grader (R.H.M.). Images were graded using the central 3-mm foveal and 6-mm foveal and nonfoveal scans of the 5-line raster. Grading of the epiretinal membrane (ERM) was based on the OCT grading system proposed by Govetto and associates to standardize the assessment of ERM and reduce bias that may exist in the clinical diagnosis of ERM.
OUTCOMES
Primary outcomes were visual acuity and discontinuity of the external limiting membrane (ELM), ellipsoid zone (EZ), and interdigitation zone (IDZ) on SD-OCT at 1 year postoperatively. Secondary outcomes included postoperative cystoid macular edema (CME) and ERM formation at 1 year postoperatively.
STATISTICAL ANALYSIS
All statistical analysis was performed using IBM SPSS, version 26 (IBM Corp, Canada). Sample size calculation was based on visual acuity and IDZ analysis based on previous studies. , , A total of 71 patients in each group was found to be sufficient to determine a significant difference in IDZ and visual acuity. We chose to include a total of 100 patients in each group to account for any patients with poor-quality OCT scans and for any other missing data.
Statistical analysis using t test for continuous data and χ 2 test for categorical data was conducted when data were normally distributed. For categorical data analysis, Fisher exact test was used when cell counts were less than 5. The Mann-Whitney U test was reserved for continuous data that were not normally distributed. All inferential statistical analyses were conducted with 2-sided P values. P values were adjusted using the Bonferroni correction for multiple comparisons. The alpha was set at P <.05 and adjusted to P <.0125 for multiple comparisons. Interobserver agreement for OCT grading was assessed using Cohen kappa.
RESULTS
There were no significant differences in age, sex, baseline visual acuity, lens status, or extent of myopia between groups ( Table 1 ). The proportion of patients with subfoveal fluid height greater than 1500 μm was similar between groups (n = 25 PRB, n = 26 PR, n = 29 PFCL). Among those patients with a subfoveal fluid height ≤1500 μm, eyes in the PRB group had a greater subfoveal fluid height compared to both PR and PFCL (mean ±SD: PRB, 752.1 ± 797.9 μm [n = 75]; PR, 362.7 ± 496.2 μm [n = 74]; PFCL, 420.5 ± 591.2 μm [n = 71]; P = .042). There was no statistically significant difference in the extent of the RRD (measured in quadrants) between groups ( Table 1 ).
There were no differences in the number of retinal breaks at the time of surgery between groups (mean ± SD: PRB, 1.8 ± 1.4; PR, 1.5 ± 1.5; PFCL, 1.6 ± 1.3; P = .8). In addition, the type of gas tamponade and extent of intraoperative endolaser use was similar between groups (mean ± SD burns per patient: PRB, 1069±128; PR, 923±136; PFCL, 1055±140; P = .8). The number of patients requiring postoperative laser retinopexy and total number of office visits were also similar between groups ( Table 1 ).
Time to presentation was similar between groups (median [interquartile range] days: PRB, 3 [9.5]; PR, 2 [10]; PFCL, 3 [9]; P = .7). The proportion of patients with gradable SD-OCT at 1 year was similar between groups (PRB, 90% [90/100]; PR, 87% [87/100]; PFCL, 91% [91/100]). Cohen kappa coefficient (95% CI) for the 12-month OCT analysis was 0.95 (0.89-0.98) for ELM, 0.91 (0.78-0.97) for EZ, 0.85 (0.74-0.99) for IDZ, and 1.0 (1.0-1.0) for the retinal pigment epithelium (RPE).
In our cohort, there was no significant difference in mean (±SD) logMAR visual acuity at 3 months postoperatively ( Table 2 ). However, by 6 months, visual acuity was significantly better in both PRB and PR compared with PFCL (mean ± SD logMAR at 6 months: PRB, 0.9±0.5; PR, 0.9±0.7; PFCL, 1.0±0.5; P = .02). This difference persisted at 12 months, with visual acuity being significantly better in the PRB and PR groups compared with PFCL, with no significant difference between the PRB and PR groups (mean ± SD logMAR at 12 months: PRB, 0.6±0.5; PR, 0.7±0.6; PFCL, 0.9±0.6; P = .002).
| Drainage Technique | ||||
|---|---|---|---|---|
| PRB | PR | PFCL | P Value | |
| Visual acuity: 3 mo | ||||
| n | 95 | 96 | 92 | |
| Snellen equivalent | 20/160 | 20/200 | 20/200 | |
| LogMAR, mean (SD) | 0.9 (0.5) | 1.0 (0.6) | 1.0 (0.6) | .09 |
| Visual acuity: 6 mo | ||||
| n | 93 | 96 | 95 | |
| Snellen equivalent | 20/125 | 20/150 | 20/200 | |
| LogMAR, mean (SD) | 0.9 (0.5) | 0.9 (0.7) | 1.0 (0.5)* | .02 |
| Visual acuity: 12 mo | ||||
| n | 100 | 100 | 100 | |
| Snellen equivalent | 20/80 | 20/100 | 20/160 | |
| LogMAR, mean (SD) | 0.6 (0.5) | 0.7 (0.6) | 0.9 (0.8)* | .002* |
| Single-operation anatomic reattachment patients | ||||
| n | 86 | 85 | 83 | |
| Snellen equivalent | 20/80 | 20/100 | 20/160 | |
| LogMAR, mean (SD) | 0.6 (0.4) | 0.7 (0.6) | 0.9 (0.5)* | .011* |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
