The Influence of Surgical Timing on Clinical Outcomes in Primary Extramacular Retinal Detachment in a Tertiary Referral Center

Abstract

Purpose:

This article reports the influence of timing on the clinical outcomes in primary extramacular rhegmatogenous retinal detachment (RRD) at a tertiary referral center.

Methods:

A retrospective case series was conducted of all patients presenting between January 1, 2014 and December 31, 2016, with primary extramacular RRD. Retinal detachments with grade C proliferative vitreoretinopathy, combined tractional and RRD, eyes with inflammatory disease, and prior retinal surgery were excluded. The main outcome measures were single-operation anatomic success (SOAS), final anatomic success, and best-corrected visual acuity (BCVA).

Results:

There were 202 eyes of 198 patients with an average of 22 months’ follow-up (range, 6-47 months). Eyes were operated on an average of 1.1 days after initial presentation. At last clinical examination, SOAS had been achieved in 174 (86%) eyes, final anatomic success in 200 (99%) eyes, and average postoperative logarithm of the minimum angle of resolution (logMAR) BCVA was 0.18 (Snellen equivalent, 20/30; SD, 0.36). In those treated the day of presentation, average postoperative logMAR BCVA was 0.18 (Snellen, 20/31) in comparison to 0.18 (Snellen, 20/30) in those treated the day after presentation and 0.14 (Snellen, 20/28) in those treated after 2 days or more (P = .92).

Conclusions:

Regarding timing of surgery, SOAS and BCVA outcomes in primary extramacular RRDs were favorable with an urgent and semiurgent approach to repair. There was no difference in visual and anatomic outcomes between patients who were operated on the day of presentation and those treated a short time later when clinical decisions were made by the treating surgeon on a case-by-case basis.

Introduction

Rhegmatogenous retinal detachment (RRD) has a risk of severe visual loss, especially if the macula is involved. Hence, surgical treatment of extramacular detachments has been considered to be urgent for the best visual outcomes following surgery. ^1,2 There are only a few studies that have examined visual outcomes with immediate, semiurgent, or delayed repair of extramacular RRD. The purpose of the present study is to describe the outcomes of surgery for primary extramacular RRD at an academic tertiary care center with respect to anatomical success and visual acuity outcomes as well as complication rates after different times between presentation and surgical repair as dictated by the treating physician.

Methods

This is a retrospective, interventional, consecutive case series of all patients treated by faculty vitreoretinal surgeons for a primary extramacular RRD between January 1, 2014 and December 31, 2016, at Bascom Palmer Eye Institute in Miami, Florida, and its 4 satellite locations. Approval was obtained by the institutional review board of the University of Miami Miller School of Medicine.

Extramacular status was determined by preoperative clinical examination or B-scan ultrasonography when the view to the posterior pole was limited by vitreous hemorrhage or other media opacity. If the B-scan could not unequivocally assess the macular status, the eye was not included as an extramacular RRD case. Optical coherence tomography was not routinely used.

Those eyes with presumed foveal attachment but macular subretinal fluid elsewhere within the vascular arcades were considered to have macula-involving RRD and were excluded from the study. Additional exclusion criteria were severe trauma; age younger than 18 years; concurrent macular hole (MH); syndromes predisposing to detachment such as Stickler, Marfan, and Ehlers-Danlos; previous retinal surgery; complex cataract surgery with retained lens fragments; chorioretinitis or other inflammatory ocular disease; proliferative vitreoretinopathy grade C at initial presentation; combined tractional and RRD; giant retinal tear; prior or existing severe posterior pole disease such as exudative age-related macular degeneration; amblyopia or optic neuropathy; and follow-up less than 6 months. Of note, there were 43 eyes that had inadequate follow-up and were removed from the subsequent analysis.

Age at diagnosis, sex, initial date of presentation and of surgery, type of surgical procedure, lens status and interval cataract extraction, and tamponade (if used) were recorded. The main outcome measures were single-operation anatomic success (SOAS), final anatomic success (FAS) at last follow-up, and best-corrected visual acuity (BCVA) at last clinical examination. Outcome measures were also recorded at 6 months. The components of initial surgery performed were at the discretion of the treating vitreoretinal surgeon.

SOAS was defined as a permanent retinal reattachment achieved with 1 surgery. Eyes requiring additional surgery for recurrent retinal detachment, MH, or silicone oil removal were considered to be primary reattachment failures and not tabulated as SOAS. Eyes with silicone oil at the postoperative intervals when anatomic success was graded were not considered to have achieved retinal reattachment and were deemed to have failed SOAS.

Eyes re-treated in the clinic with an additional gas or air injection or supplemental laser were not considered primary failures if operating room surgery was not required. Silicone oil removal for when silicone oil was initially placed for travel reasons was not considered a primary failure. When silicone oil was placed for other reasons (eg, as a primary tamponade agent for monocular status, because of a patient’s inability to posture or surgeon’s choice) that eye was considered to have failed SOAS. Subsequent surgery for epiretinal membrane or cataract was also not considered a primary failure. All statistical comparisons between categorical variables were conducted using a Fisher exact or Pearson test. Continuous variables such as logarithm of the minimum angle of resolution (logMAR) BCVA were compared using a 2-sample t test. For the purposes of calculating average logMAR BCVA, numerical values for count fingers, light perception, or no light perception vision were assigned as described in the literature. ^3,4 All analyses were performed in SPSS version 22 (SPSS Inc).

Results

A search of electronic medical records for all extramacular RRDs repaired by vitreoretinal faculty between January 1, 2014 and December 31, 2016, resulted in 454 eyes, of which 202 eyes of 198 patients met the inclusion criteria. The average follow-up was 22 months (range, 6-47 months; SD, 11). Four (2%) patients had bilateral extramacular RRDs. A total of 126 (64%) patients were male. The average age at the time of surgery was 57 years (range, 19-86 years; SD, 12). Eight (4%) eyes had a history of mild dry age-related macular degeneration, 5 (2%) of epiretinal membrane, and 2 (1%) of mild amblyopia. Eleven (5%) eyes had a history of mild blunt trauma preceding the retinal detachment. The average time to repair was 1.1 days after initial presentation (range, 0-11 days; SD, 1.5). Fifty-nine (29%) eyes received same-day surgery, 116 (57%) next-day surgery, and 27 (13%) received surgery more than 1 day after initial presentation (range, 2-11 days). Average preoperative logMAR BCVA was 0.19 (Snellen equivalent, 20/31; SD, 0.37). A total of 140 (69%) eyes were phakic and 62 (31%) were pseudophakic. Baseline characteristics are presented in Table 1.

Table 1.

Demographic and Baseline Characteristics of Patients With Macula-Sparing Rhegmatogenous Retinal Detachments (n = 202).

Age, average y	57
Sex, male, %	64
Follow-up, average mo	22
Lens status
Phakic, %	31
Pseudophakic, %	69
Aphakic, %	0
Preoperative visual acuity, average logMAR (SD) (Snellen equivalent)	0.19 (0.37) (20/31)
Primary scleral buckle, No. (%)	51 (25)
Primary pars plana vitrectomy, No. (%)	60 (30)
Primary scleral buckle/pars plana vitrectomy, No. (%)	89 (44)
Primary pneumatic retinopexy, No. (%)	2 (1)

Abbreviation: LogMAR, logarithm of the minimum angle of resolution.

Overall, 51 (25%) eyes were treated initially with scleral buckle (SB) alone. Sixty (30%) were treated with pars plana vitrectomy (PPV) alone, and 89 (44%) initially received combined scleral buckling and PPV (SB-PPV). Two (1%) received primary pneumatic retinopexy (PR). Of the cases managed with PPV alone, 24 (40%) were phakic eyes. An SB was more likely to be placed in phakic eyes. An SB with or without PPV was used in 116 of 140 (83%) phakic eyes in comparison to 24 of 62 (39%) pseudophakic eyes (P ≤ .01).

When vitrectomy was performed and the gauge of the surgical instruments was recorded, 3 (2%) eyes received 20 gauge, 98 (69%) 23 gauge, 40 (28%) 25 gauge, and 2 (1%) 27 gauge. The initial tamponade was sulfur hexafluoride (SF₆) in 66 (33%) eyes, perfluoropropane (C₃F₈) in 98 (48%), air in 6 (3%), and 1000-mm²/s oil in 6 (3%) (3 of which were for travel reasons and 3 because the patient was monocular). Grade C proliferative vitreoretinopathy (PVR) developed in 10 (5%) eyes postoperatively, of which only 2 of 10 (20%) had SOAS.

SOAS was achieved in 174 (86%) eyes by last clinical examination and FAS in 200 (99%) eyes. SOAS was obtained in 43 (84%) eyes treated with SB, in 51 (85%) eyes treated with PPV, and 80 (90%) eyes treated with SB-PPV. There was no statistically significant difference in SOAS between these 3 primary surgical techniques (P = .55). Of the 2 eyes treated with PR, both required subsequent surgery. SOAS was achieved in 56 (85%) eyes when SF₆ was used and in 88 (90%) eyes when C₃F₈ was used as the initial tamponade (P = .34).

In the phakic eyes (n = 140), 49 (35%) received SB, 24 (17%) PPV, and 67 (48%) SB-PPV. SOAS was achieved in 123 (88%) phakic eyes overall and in 42 (86%) of eyes that received SB, 19 (79%) treated with PPV, and 62 (93%) that received SB-PPV (P = .19).

In the pseudophakic eyes (n = 62), 2 (3%) received SB, 36 (58%) PPV, and 22 (35%) SB-PPV. SOAS was achieved in 51 (82%) pseudophakic eyes overall and in 1 (50%) that received SB, 32 (89%) treated with PPV, and 18 (82%) that received SB-PPV. Among patients with pseudophakic eyes, there was no statistically significant difference in SOAS success rate between PPV and SB-PPV (P = .57). There was an insufficient number of cases to perform statistical analysis of SB alone in these patients (n = 2).

Among all eyes in the study, average logMAR BCVA at last clinical examination was 0.18 (Snellen, 20/30; range, 20/15 to hand motion; SD, 0.36). At last clinical examination, 173 (86%) eyes had BCVA of Snellen 20/40 or better. A total of 147 (73%) eyes maintained their preoperative vision.

Of the entire cohort, eyes treated with SB had an average postoperative logMAR BCVA of 0.12 (Snellen, 20/27; SD, 0.18), those treated with PPV attained 0.07 (Snellen, 20/23; SD, 0.18), and those treated with SB-PPV 0.28 (Snellen, 20/38; SD, 0.49). When each technique was compared, there was no statistically significant difference between those treated with SB and PPV (P = .12), but there was a difference between those treated with SB and SB-PPV (P = .01) and those treated with PPV and SB-PPV (P < .01).

In those eyes that experienced SOAS (n = 174), average postoperative logMAR BCVA was 0.13 (Snellen, 20/27; SD, 0.30), 89% reached 20/40 or better, and 76% maintained their preoperative vision.

In the entire cohort, SOAS was achieved in 47 of 59 (80%) eyes treated the same day as presentation, 103 of 116 (89%) eyes treated the next day, and 24 of 27 (89%) eyes treated more than 1 day after presentation (range, 2-11 days). There was no statistical difference in SOAS between eyes treated the same day, the day after presentation, and more than 1 day after presentation (P = .23). Average final logMAR BCVA was 0.18 (Snellen, 20/31) in eyes treated the same day, 0.18 (Snellen, 20/30) in eyes treated the next day, and 0.14 (Snellen, 20/28) in eyes treated more than 1 day after presentation (P = .92) (Table 2).

Table 2.

Clinical Outcomes and Time to Surgery in Macula-Sparing Rhegmatogenous Retinal Detachment Repair.

No. of Eyes	Time to Surgery	Final Mean Visual Acuity, Average LogMAR (SD) (Snellen Equivalent)	Final Anatomical Success, %
59	Same day	0.18 (0.38) (20/31)	98
116	Next day	0.18 (0.37) (20/30)	99
27	2 days or later	0.14 (0.27) (20/28)	100

Abbreviation: LogMAR, logarithm of the minimum angle of resolution.

There were 28 eyes of 27 patients requiring additional surgery after their initial detachment repair because of failed initial surgery—SOAS failures. Sixteen (64%) patients were male. Seventeen (61%) eyes were phakic and 11 (39%) were pseudophakic. The average time to surgery was 0.8 days.

Eight (29%) of these eyes were treated initially with encircling SB, 9 (32%) with primary PPV, and 9 (32%) with SB-PPV. Two (7%) received primary PR. Of these eyes that had additional operations besides the initial repair, 1 (4%) was only for silicone oil removal, 2 (7%) were for development of MH, and 25 (89%) were performed because of recurrent retinal detachment. Grade C PVR developed in 8 (28%) eyes postoperatively and failed SOAS; these eyes also had recurrent detachments. The remaining 17 eyes that had recurrent detachments had either untreated or new retinal breaks.

By last follow-up, FAS was achieved in 26 (93%) eyes that failed to be reattached in 1 surgery. Two eyes (7%) were still under silicone oil. Average logMAR BCVA at last clinical examination was 0.44 (Snellen, 20/55, range, 20/15 to count fingers; SD, 0.56,) in comparison to 0.13 (Snellen, 20/27; SD, 0.30) in those eyes that had SOAS. In the eyes that failed SOAS, by last clinical examination, 18 (64%) had BCVA of Snellen 20/40 or better. Fifteen (54%) eyes maintained their preoperative vision at last clinical examination.

Lastly, we examined those eyes with poor final BCVA (defined as Snellen 20/200 or worse). There were 11 such eyes in 11 patients. Seven (64%) patients were male. Eight (73%) eyes were phakic and 3 (27%) were pseudophakic. The average time to surgery was 0.7 days, with 4 (36%) patients receiving same-day surgery, 6 (55%) receiving next-day surgery, and 1 (9%) receiving surgery 2 days after presentation. One (9%) of these eyes was treated initially with encircling SB and 10 (91%) with SB-PPV. In this subset of eyes, SOAS was achieved in 5 (45%) eyes and FAS in 9 (82%) eyes. Average logMAR BCVA at last clinical examination was 1.43 (Snellen, 20/541; SD, 0.46).

Conclusions

The present study reports similar anatomic and BCVA outcomes in eyes with extramacular RRD whether they were treated with SB, PPV, or combined SB-PPV, or whether they were treated the same day, the next day, or after the next day. Overall, the average postoperative BCVA was Snellen 20/30 in the entire cohort and a majority maintained vision of Snellen 20/40 or better.

The appropriate timing of repair for extramacular RRD has been a subject of debate, ^{5 -10} although clinical data investigating this question are limited. Wykoff et al ⁷ reported that progression of extramacular detachments to macular involvement was rare when a selective, but not strictly emergent, surgical approach was followed. That series predated the present series at our same center (1989-2004) and found similar rates of SOAS (88%), FAS (99.5%), and final BCVA of Snellen 20/40 or better (73%). Of note, the previous study by Wykoff et al included macula-involving but fovea-sparing RRDs.

In the present series, cases with presumed macular fluid by clinical examination were excluded. Optical coherence tomography was not used preoperatively in most cases and thus fovea on status could not be objectively verified. Fovea-sparing cases may have different requirements for surgical timing from those that are fully extramacular.

A recent study reported the success of a preoperative positioning regimen in slowing the progression of extramacular retinal detachments. ¹¹ In the present study, patients were instructed to avoid strenuous physical activity, but bilateral patching while awaiting surgery was not pursued.

This study did not detect a statistically significant difference in SOAS between eyes treated primarily with SB, PPV, or SB-PPV. Several randomized studies ^12,13 have examined the success rate of different surgical approaches of primary RRD repair. Although results have been conflicting regarding the utility of adjunctive SB when performing a PPV, the use of SB has been shown to be effective, especially in patients with phakic eyes. ¹³ A recent meta-analysis pooling nonrandomized studies suggested that supplemental SB increases primary reattachment rate in PPV. ¹⁴ The cumulative SOAS rate in the present study is comparable to other reports from academic medical centers. ¹⁵ A potentially important difference is that the present study excluded patients with less than 6 months of follow-up.

A notable limitation of this study is the absence of data such as the duration of symptoms, the nature of the detachment being shallow or bullous, the location and number of retinal breaks, and the clock hours and anterior or posterior extent of subretinal fluid. Given the retrospective nature of this series, not all surgeons recorded these metrics preoperatively, clinical drawings were often not recorded, and many physicians did not conduct preoperative imaging. These factors, not captured by our data set, may have had a role in the final decision made by the treating physician on timing of surgical repair and introduced an element of selection bias common in retrospective surgical series. These factors would be of interest in a future standardized, prospective study.

Nonetheless, the purpose of this study is to explore whether all macula on retinal detachments need to be treated emergently rather than urgently, and our data suggest that a semiurgent, rather than a strictly urgent, protocol may be used without sacrificing patient outcomes. A variety of techniques was used by our group of vitreoretinal surgeons and demonstrated anatomical and visual success, including the use of a buckle at the initial surgery in the majority of phakic eyes. In conclusion, the present study, which to our knowledge represents the largest series examining surgical timing and outcomes in extramacular RRD, demonstrates that favorable outcomes can be achieved by allowing the treating surgeon to combine clinical judgment and individual patient considerations to determine how and when to repair this subset of primary RRDs on a case-by-case basis.