Subretinal Hemorrhage Associated with Scleral Buckle Surgery for Retinal Detachment: A Consecutive Case Series

Abstract

Introduction:

This study aimed to analyze rates of subretinal hemorrhage (SRH), any correlations between between SRH and drainage technique (scleral cut-down and external needle drainage) during scleral buckle (SB) surgery, as well as associated structural outcomes.

Materials and Methods:

A retrospective analysis was conducted on 351 adult patients undergoing SB surgery, of whom 196 (56%) received SRF drainage: 124 via scleral cut-down and 72 via needle drainage. Data extracted included SRF drainage technique, SRH characteristics, and need for additional surgery.

Results:

Intraoperative SRH occurred in 19/196 (9.7%) cases, with no significant difference between techniques (p=0.99). Macular involvement of SRH was observed in 7/19 (37%) cases, all were macula-off pre-operatively. SRH size (p=0.230), resolution time (p=0.097), and surgeon variability (p>0.05) were not significantly different based on drainage technique in our relatively small study. Additional surgery occurred in 3 cases (1.5%). BCVA improved from 20/100 to 20/63 at final follow-up.

Conclusions:

Intraoperative SRH rate was 9.7% and not dependent on surgical technique. Most cases had favorable outcomes and only 1.5% of eyes required further surgical intervention. The overall rate of macula-involving subretinal hemorrhage was 3.6% and found only in preoperative macula-off detachments. While significant differences between surgical techniques were not seen, larger studies are required to draw more definitive conclusions.

Introduction:

The incidence of rhegmatogenous retinal detachments (RD) is estimated at at 6.3–7.9 per 100,000 individuals.¹ A common surgical treatment for phakic patients is scleral buckle (SB) surgery.² This method may be associated with a variety of complications including residual subretinal fluid (SRF), subretinal hemorrhage (SRH), diplopia, retinal incarceration, and scleral rupture.^3–7

SRF drainage is frequently employed as a complementary technique during SB surgery to facilitate retinal reattachment.^8,9 However, SRH is a known complication of this technique, with studies reporting rates from 1.2–15.0%.^10–12 This complication can have a variety of clinical implications as it has been correlated with anatomical failure, recurrent RD with proliferative vitreoretinopathy (PVR), and/or macular hole formation.^13–15 Two common methods for SRF drainage include the external needle drainage technique and the scleral cut-down technique.^12,16 Needle drainage is considered less cumbersome, while the cut-down and diathermy of the choroid has been hypothesized to decrease bleeding.^10,12 Some physicians may have difficulty adopting this method due to poor visualization of the needle in the subretinal space, leading to damage of the retina, choroid, and other ocular strucutres. The guarded needle technique can help prevent these complications by ensuring that the needle cannot penetrate the eye farther than the length of the exposed needle.¹⁷

While studies have reported outcomes for each technique separately, very few have compared these drainage techniques in terms of SRH incidence and structural outcomes.^{4,10–12,16,18–20} Additional data is needed to help guide practice patterns. Our study aims to address this gap by detailing cases with SRH during SB surgery and assessing the long-term structural outcomes by method of intraoperative SRF drainage.

Methods:

A single-center retrospective case series was conducted at Massachusetts Eye and Ear. Electronic health records were reviewed for patients having undergone SB surgery for initial rhegmatogenous RD (CPT 67107) performed between January 1^st, 2017 and January 1^st, 2023. Records were extracted using the Research Patient Data Registry (RPDR) query tool. Patients with intraoperative SRH were isolated and included in the analysis; exclusion criteria were patients without intraoperative SRH or patients with SRH prior to SB surgery (4 patients).

Demographic information and pre-operative and intraoperative characteristics including best-corrected visual acuity (BCVA), location and features of detachments and breaks, presence of lattice degeneration, method of intraoperative SRF drainage (scleral cut-down or needle drainage), and presence of SRH were recorded. Post-operative outcomes included whether the SRH was macula-involving, size of SRH, time to SRH and SRF resolution, as well as the need for additional surgery. The first ultra-widefield fundus photograph(s) taken after SB surgery were used to measure SRH size in millimeters squared (mm²). Size was also recorded in disc diameters if reported in the operative note.

Statistical Analysis

The statistical analysis was conducted using Statistical Package for the Social Sciences (SPSS) version 28.0.0.0 (IBM Corporation, Armonk, NY, USA). A Mann-Whitney U test was used to compare medians for non-normally distributed outcomes. A binary logistic regression was performed to determine if there were significant differences in categorical outcomes between scleral cut-down vs needle drainage of SRF. A linear regression was used to assess for differences in continuous outcomes. Statistical significance was set as p-value < 0.05. This study was approved by the Massachusetts General Brigham Institutional Review Board and adhered to the tenets of the Declaration of Helsinski. The need for informed consent was waived due to the retrospective nature of the study.

Results

Baseline Characteristics

A total of 351 SB surgeries on 333 patients were performed during this period. The median age of the cohort was 34.4 years (IQR: 26.4–45.3 years, range 14.81—90.17 years), and males comprised 54.1% (190/351) of eyes (Table 1). From this cohort, drainage of SRF was done in 55.8% (196/351) of cases per surgeon discretion. SRF drainage was performed via scleral cutdown and needle drainage in 63% (124/196) and 37% (72/196) of procedures, respectively. 5.4% (19/351) of all surgeries and 9.7% (19/196) of drainage cases had intraoperative SRH. The quadrant of SRF drainage was recorded in 11 eyes. Of these, 45.5% (5/11) were drained superotemporally, 36.3% (4/11) inferotemporally, 9.1% (1/11) superonasally, and 9.1% (1/11) inferonasally. The rates of SRH were 8.9% (11/124) among surgeries with scleral cut-down and 11.1% (8/72) with needle drainage (p=0.99). 11 eyes underwent drainage via 30-gauge(g) needle, 5 eyes via 27g. Three eyes did not have a recorded needle gauge for drainage. Four eyes were excluded from analysis for having SRH prior to SB surgery. Overall, 9.7% (19/196) of patients with drainage had SRH and were included in the analysis. Median age in this sub-group was 38.7 ± 12.8 years (Range: 22.5—64.2 years) with 36.8% (7/19) and 63.2% (12/19) of patients being male and female, respectively. The average pre-operative BCVA was 20/100 (logMAR 0.73 +/− 0.85), and 100% of patients were phakic prior to SB operation with 21.1% (4/19) having preoperative PVR Grade C. In comparison, the overall rate of PVR Grade C among the entire patient cohort was 10.0% (35/351) (p=.109). Among the patients with SRF, 21% (4/19) received C3F8 tamponade and were instructed to position themselves face-down. In comparison, 16.2% (57/351) of all eyes that received scleral buckle surgery received intraoperative tamponade: 73% (42/57) C3F8, 23% (13/57) air, and 4% (2/57) SF6. 68.4% (13/19) of patients had RDs involving the macula, and lattice was noted in 47.4% (9/19) eyes. Superior breaks and retinal detachments were noted in 52.6% (10/19) of eyes.

Table 1.

Baseline, operative, and postoperative characteristics.

Characteristic (N = 19)	Value	Characteristic (N = 19)	Value
Age, average (SD), years	38.6 (12.8)	Surgical technique, No. (%)
Sex, Female, No. (%)	12 (63)	Scleral cut-down	11 (58)
Laterality, No. (%)		Needle drainage	8 (42)
Left	9 (47)	Subretinal hemorrhage, size
Right	10 (53)	Size, average (SD), mm2	25.0 (51.8)
Lens status, No. (%)		Expansion post-op, No. (%)	3 (16)
Phakic	19 (100)	No expansion post-op, No. (%)	16 (84)
Lattice present, No. (%)	10 (53)	Subretinal hemorrhage, macula, No. (%)
PVR, No. (%)		Macula involved	7 (37)
None	13 (68)	Subretinal hemorrhage, resolution
Grade A	2 (11)	Resolution post-op	16 (84)
Grade C	4 (21)	No resolution post-op	1 (5)
Retinal detachment, No. (%)		Lost to follow-up	2 (11)
Macula-on	6 (32)	Time-to-resolution, median (IQR), days	35 (24 – 106)
Macula-off	13 (68)	Subretinal fluid
Break Location, No. (%)		Time-to-resolution, median (IQR), days	209 (56 – 394)
Superior break	10 (53)	Additional intervention, No. (%)
Inferior break	13 (68)	Additional intervention after SRH	3 (16)
Both hemispheres	8 (42)	No additional intervention	16 (84)
Retinal detachment, No. (%)		New retinal detachment	1 (5)
Superior	10 (53)	Failure of first surgery	2 (11)
Inferior	14 (74)	BCVA, mean (SD), Log MAR
Both hemispheres	5 (26)	At final follow-up	0.498 (0.546)
Multi-quadrant	15 (79)	Preoperatively	0.726 (0.853)
Retinal detachment
Total clock hours, average (SD)	4.1 (1.5)

PVR = Proliferative vitreoretinopathy; SRH = Subretinal hemorrhage; BCVA = Best-corrected visual acuity; Log MAR = Logarithm of the Minimum Angle of Resolution

Among all 19 surgeons in the study, 8 (42.1%) encountered SRH. When conducting a binary logistic regression to determine if there was an association between surgeon and incidence of SRH, surgeon factor was not found to be significantly correlated with SRH when analyzing all cases that underwent intraoperative drainage (p>0.05). SRH was observed externally in 8 eyes and was observed with indirect ophthalmoscopy in the remaining 11 eyes.

SRH Outcomes and Characteristics

Submacular hemorrhage was observed in 36.8% (7/19) of patients with SRH; there was no significant difference between both drainage techniques in the incidence of submacular hemorrhage (p=0.71). Submacular hemorrhages were noted in 3 eyes that received needle drainage and 4 eyes that received scleral cut-down. Among the 7 eyes that had macula-involving SRH, all had macula-off RDs prior to SB surgery. 2 of the eyes in which SRH was noted via external examination were noted to have macular involvement, in comparison to 5 in which SRH was viewed via indirect ophthalmoscopy. SRH expansion, defined as an increase in the area of hemorrhage at any time point post-SB surgery, was observed in 3 eyes: 2 from needle-drainage and 1 from scleral cut-down. SRH resolution was seen in 84.2% (16/19) by the end of follow-up, with 9 eyes having scleral cut-down and 7 having needle-drainage. The average duration of follow-up in this cohort was 26.12 ± 22.73 months. There was no significant difference in the rate of SRH resolution (p=0.351) and rate of SRH expansion (p=0.237) between eyes with either drainage method.

Additional surgical intervention was performed in 15.8% (3/19) of eyes: 2 with scleral-cutdown and 1 with needle drainage. Two eyes that required surgical intervention had SRH expansion post-operatively, 1 eye with needle drainage and 1 with scleral cut-down. As a result, among the 196 eyes that received RD repair with SRF drainage, only 1.5% required further surgery after SRH was found intraoperatively (Table 2). The first eye had PPV to address recurrent detachment with bullous submacular hemorrhage under the detached retina 28 days after initial SB surgery with needle drainage. Perfluoro-n-octane (PFO) was utilized to flatten the retina and facilitate the evacuation of the SRH through the retinectomy site by exerting direct pressure to displace the hemorrhage. The second eye had persistent retinal detachment with continued SRH that progressed to vitreous hemorrhage two weeks after SB surgery with scleral cut-down. The patient then underwent PPV for new identifiable breaks; PFO was used to stabilize the retina, and subretinal blood was removed using a soft-tip cannula through the retinectomy site. The third eye requiring surgical intervention had PPV with air fluid exchange for persistent SRF 20 days after initial SB surgery with scleral cut-down.

Table 2.

Characteristics of patients requiring additional surgery.

Age	Macula status	PVR-C present	SRF drainage technique	0.25 cc 100% C3F8	SRH Expansion	Indication for PPV	Procedure
23	Off	Yes	Needle	Yes	Yes	Recurrent RD, submacular hemorrhage	PPV/PFO/Retinectomy/Remove subretinal blood/EL/SO
36	Off	Yes	Cut-down	No	Yes	Persistent RD, SRH progressed to vitreous hemorrhage	PPV/PFO/Retinectomy/Remove subretinal blood/FAX/SO
50	Off	No	Cut-down	No	No	Persistent SRF	PPV/FAX/EL/C3F8

PVR = proliferative vitreoretinopathy; SB = scleral buckle; SRF = subretinal fluid; SRH = subretinal hemorrhage; PPV = pars plana vitrectomy; RD = retinal detachment; PFO = perfluoro-n-octane; EL = endolaser; SO = silicone oil; FAX = fluid-air exchange

There was no significant difference in the rates of additional surgical intervention between both drainage methods (p=0.604). The average SRH size among the cohort was 25.0 +/− 51.8 mm², with no significant difference between drainage technique (p=.230). Median time to SRH resolution among the cohort was 35 (IQR: 24–106) days, with no statistically significant difference between macula-off (35 days, IQR: 26–52.5) and macula-on eyes (70.5 days, IQR: 9.5–121.6) (p=0.920). No significant difference was observed with the time to SRH resolution between both drainage methods (p=0.097).

Vision at Final Follow-up.

BCVA among the 19-patient sub-group at final follow-up was 20/63 (logMAR: 0.498 ± 0.546), improving from pre-operative BCVA of 20/100 (logMAR: 0.726 ± 0.853). Final BCVA of all cases was noted to be 20/40 (logMAR: : 0.32 ± 0.45), and the average BCVA of all submacular SRH cases was 20/200 (logMAR: 1.02 ± 0.59). The average time to SRF resolution was 255.27 +/− 215.74 days. No significant difference was seen in final BCVA among patients with macula-on RDs [20/30 (LogMAR:0.180 +/− .200)] and macula-off RDs [20/80 (LogMAR: 0.645 +/− 0.597)] (p=0.134).

Discussion:

This study examined rates and outcomes of subretinal hemorrhage during SB surgery. The findings were 1) low incidence of macula-involving SRH and reoperations, 2) no significant differences between scleral cut-down and needle drainage in SRH incidence, macular involvement, resolution time, post-op expansion, or need for additional surgery, 3) favorable visual outcomes, and 4) no cases of macula-involving hemorrhage in pre-operative macula-on detachments.

Although rates of SRH have been reported in large series of SB surgery as a secondary outcome, the literature is lacking in details of these specific cases. Given the potential risks of SRH and high degree of variation among cases, there is a need to assess outcomes with greater granularity. This study focused on the rate of macular involvement, need for additional interventions, as well as risk factors to guide management and patient counseling. Additionally, we sought to answer a common question regarding the optimal technique, scleral cut-down vs external needle drainage.

There are only two studies, directly comparing this outcome between needle drainage and the conventional cut-down approach.^12,21 The rates of SRH drainage varied from 12.7–17.5% for scleral cut-down and 15.0–22.2% for needle drainage among cohorts of 80 and 100 patients, respectively.^12,21 Our results demonstrated a comparably lower incidence of SRH with both drainage techniques in a larger sample size. The cause for hemorrhage has been attributed to the perforation of the highly vascularized choroid. Intuitively, one might expect lower rates of SRH with cut-down given the opportunity to cauterize prior to draininage. Some potential drawbacks that may outweigh this perceived benefit include partial treatment of the choroid, large incision, and/or inadvertent drainage before cautery. Location of drainage can also influence the development of SRH. Avoidance of the quadrant center, staying near muscle insertions, and minimizing disruption of areas with overlying choroidal vasculature are all important factors when selecting the drainage location.

Overall, the outcomes of patients with subretinal hemorrhage were still favorable. With a high rate of spontanous resolution, observation is an appropriate management choice in most cases. The majority of patients in our study cohort did not require further surgical intervention. The two eyes requiring vitrectomy had macular involvement as well as post-operative expansion of the SRH. The timing of reoperation is controversial. SRH may threaten the vision potential through build-up of blood breakdown products between the retinal pigment epithelium (RPE) and the neurosensory retina or between the RPE and the choroid.²² Therefore, if there is persistent submacular hemorrhage, intervention may be advisable. Options include the use of gas for pneumatic displacement with or without adjunctive tissue plasmogenic activator (tPA) to facilitate the removal of subretinal blood.^23–26

A key finding is that this study found no cases of macular involvement of SRH in eyes with macula-on retinal detachments. The decision for drainage of SRF in SB surgery is multifactorial. If there is little to no risk for macula-involving subretinal hemorrhage in macula-on detachments, the risk-benefit ratio could favor drainage in most cases. There may still be a chance that subretinal hemorrhage can traverse a large break and cause a vitreous hemorrhage, which is suspected to have occurred in one of our cases. Furthermore, it remains unknown if SRH or vitreous hemorrhage could potentially increase the rates of post operative PVR.

The main limitation of our paper is the small sample size of SRHs. This small sample size prevents our study from robustly comparing the two surgical drainage methods. As a result, larger studies are necessary to provide more definitive conclusions. A post-hoc analysis was performed to determine the sample size requirements for detecting differences between surgical techniques. The minimum number of eyes required to detect difference in our binary outcomes is approximately a 190 eyes to achieve 80% power with a significance level of 0.05. In contrast, for our continuous outcomes, a total of approximately 26 eyes would be needed to achieve sufficient power. Additionally, one of the primary limitations in our study is its retrospective nature with data from RDs treated at a single center. Although, to the best of our knowledge, this is the largest study to date, the low case incidence limited the sample size. Future studies could aim to increase power by taking a multi-center approach. Registry or database studies are unlikely to be helpful given the detailed information required from operative reports. Another limitation was the variation in the timing of the post-operative fundus photographs used to assess SRH size. These were taken between a range of 1–41 days, and four patients did not have an adequate post-operative fundus photograph, for whom the documented retinal exam was relied on.

Conclusion:

No significant differences were found in the incidence and outcomes of subretinal hemorrhage between scleral cut-down and needle drainage techniques in our small study. Larger studies would be helpful in further elucidating advantages based on drainage approach. A majority of subretinal hemorrhages had favorable visual and anatomical outcomes with further intervention only required in cases with macula-involvement and post-operative hemorrhage expansion (1.5%). There were no cases of macula-involving subretinal hemorrhages in macula-on detachments.