- Email: [email protected]
Multiple Subretinal Fluid Blebs After Successful Retinal Detachment Surgery: Incidence, Risk Factors, and Presumed Pathophysiology
Multiple Subretinal Fluid Blebs After Successful Retinal Detachment Surgery: Incidence, Risk Factors, and Presumed Pathophysiology YONG-KYU KIM, JEEYUN AHN, SE JOON WOO, DUCK JIN HWANG, AND KYU HYUNG PARK PURPOSE:
To investigate the incidence and the clinical factors associated with the occurrence of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment (RD) repair. DESIGN: Retrospective, observational case series. METHODS: We retrospectively investigated the medical records of 185 eyes of 184 patients who had undergone successful RD surgery, either vitrectomy or scleral buckling. Each patient had undergone spectral-domain optical coherence tomography (SDOCT) combined with infrared reflectance (IR) imaging every 3 months postoperatively. We carefully examined postoperative SDOCT and fundus IR images, in an effort to identify any SRF blebs present. RESULTS: Multiple (‡3) SRF blebs were observed in 40 of 185 cases (21.6%). SRF blebs were first detected 1.7 ± 1.8 months postoperatively. In 22 cases that could be fully followed up, SRF blebs were completely absorbed 13.1 ± 6.1 months postoperatively. Multiple logistic regression analysis showed that only young age (<30 years) was significantly associated with the occurrence of multiple SRF blebs (odds ratio, 5.1; 95% confidence interval, 1.5-17.6; P [ .010). Serial measurements of SRF bleb size using SDOCT showed that SRF bleb height was greatest at postoperative 2.9 ± 0.9 months, while SRF bleb width tended to decrease gradually over time. The SRF blebs typically spared large retinal vessels. CONCLUSIONS: Multiple SRF blebs are commonly found after successful RD surgery, especially in young patients. The serial morphologic features evaluated in this study indicate that multiple SRF blebs may result from the active reattachment of retinal pigment epithelium and photoreceptors during the resolution of RD. (Am J Ophthalmol 2014;157:834–841. Ó 2014 by Elsevier Inc. All rights reserved.) Accepted for publication Dec 10, 2013. From the Department of Ophthalmology, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam, South Korea (Y.K.K., J.A., S.J.W., D.J.H., K.H.P.); Department of Ophthalmology, Seoul National University College of Medicine, Seoul Metropolitan Government–Seoul National University Boramae Medical Center, Seoul, South Korea (J.A.); and Department of Ophthalmology, Han Gil Eye Hospital, Incheon, South Korea (D.J.H.). Inquiries to Se Joon Woo, Department of Ophthalmology, Seoul National University Bundang Hospital, #300, Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, South Korea; e-mail: sejoon1@snu. ac.kr
834
Ó
2014 BY
O
PTICAL COHERENCE TOMOGRAPHY (OCT) HAS
been used to elucidate the pathophysiology of many retinal diseases, which previously could only have been investigated through the use of histopathologic techniques. By incorporating OCT into the postoperative assessment of retinal detachment (RD) patients, we were able to detect subclinical levels of subretinal fluid (SRF) that is usually not detected through routine binocular ophthalmoscopic examination.1–7 Aside from the persistence of SRF, bleb-like lesions may arise in the previously detached retina during the postoperative follow-up period after successful retinal reattachment. These lesions appeared similar to those associated with serous detachment of the retinal pigment epithelium (RPE), and using fluorescein angiography, histopathologic review, and OCT, they were determined to be focal sensory retinal detachment.8–10 The described pockets were observed in subfoveal and peripheral areas, whereas the persistent SRF described previously was found primarily in dependent positions, the fovea, or the inferior part of the retina. Moreover, bleblike SRF pockets usually arise in clusters (more than 3), in contrast to confluent SRF or single SRF blebs. Kang and associates reported the incidence and presumed mechanism of SRF blebs. They observed SRF blebs in 11 of 118 patients (9.3%) who underwent successful scleral buckling (SB) and cryotherapy for rhegmatogenous RD. They proposed choroidal vascular damage caused by cryotherapy as the main mechanism of SRF bleb occurrence.11 However, when we evaluated multiple SRF blebs with serial spectral-domain OCT (SDOCT), we noticed additional characteristics that have not been suggested before. In this study, we aimed to investigate the incidence and clinical risk factors associated with multiple SRF blebs that could be found after successful rhegmatogenous RD surgery and also to elucidate the presumed mechanism of SRF bleb occurrence.
METHODS WE RETROSPECTIVELY REVIEWED THE MEDICAL RECORDS OF
all the patients who underwent either pars plana vitrectomy (PPV) or SB surgery for rhegmatogenous RD at Seoul National University Bundang Hospital from January 1,
ELSEVIER INC. ALL
RIGHTS RESERVED.
0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.12.030
2009, to June 30, 2012. This study followed the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University Bundang Hospital. A total of 340 patients (343 eyes) underwent rhegmatogenous RD repair during that period, and ultimately 185 eyes from 184 patients were included in this study. We included only patients with primary spontaneous RD who achieved a successful postoperative course, and we excluded those with complicated RD, primary failure cases, or re-detachment cases. Specifically, we excluded patients who underwent silicone oil injection for retinal reattachment. We also excluded those with preoperative retinal pathology that could have influenced retinal reattachment—for example, trauma history, epiretinal membrane, vitreomacular traction syndrome, combined tractional RD, proliferative vitreoretinopathy (subretinal gliosis), and macular pathology. We excluded reoperation cases as well as patients with incomplete courses of postoperative OCT follow-up. All surgeries were performed by 4 experienced retinal surgeons at Seoul National University Bundang Hospital (P.K.H., W.S.J., A.J., and H.D.J.). The surgeons decided the method of operation, either PPV or SB, based on the clinical characteristics of each patient and RD. Younger patients and those with small definite tears, especially those located inferiorly, were likely to undergo SB, whereas elderly patients with cataracts or pseudophakia and suspicious or hidden tears were more likely to undergo PPV. A 23 gauge transconjunctival sutureless vitrectomy system (Accurus; Alcon Laboratories, Inc, Fort Worth, Texas, USA) was used for PPV. In brief, we used perfluorocarbon liquids during the surgery, and internal SRF drainage was performed through the original tear. After the surgery, the eye was filled with gas (18% SF6 or 14% C3F8). In SB surgery, we identified the retinal tear and then performed cryotherapy. The explant material was silicone sponge with width of 5 or 7.5 mm (No. 506 or 507; MIRA, Waltham, Massachusetts, USA) for the segmental SB and silicone tire (No. 287; MIRA) and silicone band (No. 240; MIRA) for the encircling SB. External SRF drainage was performed during SB surgery in 7 of 82 cases (8.5%). Each patient underwent a thorough ophthalmologic examination, including best-corrected visual acuity (BCVA; Snellen visual acuity chart), slit-lamp examination, binocular indirect ophthalmoscopy, and SDOCT combined with infrared reflectance (IR) confocal scanning laser ophthalmoscopy (Spectralis OCT; Heidelberg Engineering, Heidelberg, Germany) at 1 month after surgery and every 3 months thereafter until any SRF disappeared. For the visual acuity analysis, we converted Snellen visual acuity values to the logarithm of the minimal angle of resolution (logMAR). We evaluated only patients without any preoperative ocular disorders that might influence visual function (eg, amblyopia and untreated severe cataracts). Final visual acuity was defined as visual acuity obtained at least 6 months after the complete resolution VOL. 157, NO. 4
FIGURE 1. Survival curve of subretinal fluid blebs after successful rhegmatogenous retinal detachment surgery. Tick marks indicate censored (incomplete optical coherence tomography follow-up) cases.
of any subfoveal SRF and at least 3 months after cataract surgery in patients who underwent cataract surgery during the follow-up period. SUBRETINAL FLUID BLEB AND OPTICAL COHERENCE TOMOGRAPHY ANALYSIS: We carefully searched for the
presence of SRF bleb using both postoperative SDOCT images and fundus IR images. We retrieved fundus IR images to identify any bleb-like lesion that was dark and round, and then examined SDOCT volume scans spanning an area of 30 3 20 degrees for any SRF bleb lesion. If there were peripherally located SRF blebs beyond the scope of routine imaging, we obtained peripheral images. We defined multiple SRF blebs as 3 or more localized fluid pockets under the extrafoveal retina with or without subfoveal bleb. We excluded those cases with only a single subfoveal bleb. When preoperative SDOCT images were available, we evaluated whether there was any severe undulation (>3 undulations in a single section) in the outer portion of the detached retina. To characterize the serial morphologic changes typical of SRF bleb, we selected 1 SRF bleb from each patient that could be traced with SDOCT throughout the clinical course, then measured its width and height at baseline (when the SRF bleb first appeared) and at its most prominent stage (when the SRF bleb was highest in OCT images and prominent in IR images) using the caliper tool in the OCT software. STATISTICAL ANALYSIS:
We compared preoperative and intraoperative characteristics between those with and without multiple SRF blebs. Some of these factors were significantly associated with the occurrence of SRF bleb in the univariate analysis, which was performed using either the Student t test for continuous variables or the x2 or Fisher exact test for categorical variables. We performed multiple logistic regression analysis with these
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
835
FIGURE 2. Subretinal fluid blebs after successful rhegmatogenous retinal detachment surgery. (Left) One month after the surgery, multiple bleb-like lesions are visible on superior temporal area. (Middle) Three months after the surgery, blebs became more prominent, represented as dark, round lesions on infrared reflectance (IR) scanning laser ophthalmoscopy. (Right) These lesions are revealed to be subretinal fluid (SRF) pockets in spectral-domain optical coherence tomography (SDOCT). The SDOCT images were obtained along the green arrows in IR images. Note that subfoveal fluid bleb (Right, lower) is located apart from multiple SRF blebs in the superior temporal area (Right, upper).
FIGURE 3. Comparison of the incidence of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment surgery in different age groups. *P value < .05, Fisher exact test.
factors to eliminate confounding effects. All statistical analyses were performed using PASW version 18.0 (SPSS, Inc, Chicago, Illinois, USA), and P values < .05 were considered statistically significant.
RESULTS THE 185 EYES (184 PATIENTS) THAT UNDERWENT SUCCESS-
ful primary rhegmatogenous RD repair were ultimately included. Multiple SRF blebs were observed postoperatively in 40 eyes (40 patients) out of 185 eyes (21.6%). In most cases (33 cases, 82.5%), SRF blebs were first 836
detected at 1 month postoperatively; another 5 cases (12.5%) exhibited SRF blebs within 6 months after the surgery. In 2 cases (5%) with diffuse persistent subclinical SRF, SRF blebs first occurred at 8 months postoperatively, as SRF was absorbed. Overall, SRF bleb was first identified on SDOCT 1.7 6 1.8 (mean 6 SD) months postoperatively. In 22 cases, we were able to confirm the complete absorption of SRF blebs with SDOCT during the followup period. The SRF blebs were completely absorbed within 1 year after surgery in 10 cases (45.5%) and within 2 years postoperatively in another 10 cases (45.5%). The longest duration of SRF blebs was 28 months postoperatively (13.1 6 6.1 months postoperatively). For those 18 cases that we could not follow fully until the SRF blebs were completely absorbed, the OCT follow-up period was relatively short (3.6 6 2.4 months). The cumulative survival curve of SRF bleb during the follow-up period is represented in Figure 1. Figure 2 shows an example of a color fundus photograph and fundus IR and SDOCT images of multiple SRF blebs. The SRF blebs are more readily visible on IR images, as round dark lesions. They can be found anywhere in the previously detached retina, either submacularly or peripherally. In SDOCT images, these dark lesions were revealed to be small subretinal fluid pockets. In Figure 2, the subfoveal fluid, separated from multiple SRF blebs in the superior temporal area, is apparent. In this study, we did not count eyes with subfoveal fluid alone as a multiple SRF bleb case. Demographics and clinical characteristics were compared between eyes with and without multiple SRF blebs. Patients with multiple SRF blebs were younger (36.6 6 20.2 years vs 47.2 6 15.9 years, P ¼ .003) than those in the control group. When we stratified patient age by quartile, multiple SRF blebs were most prevalent in the youngest age group (<30 years), while the other 3 age groups did not show any significant differences
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
TABLE. Demographic and Clinical Characteristic Comparison Between Patients With and Without Multiple Subretinal Fluid Blebs After Successful Retinal Detachment Surgery
Age (y), mean 6 SD <30, n (%) > _30, <45, n (%) > _45, <60, n (%) > _60, n (%) Male, n (%) Follow-up period (months), mean 6 SD Symptom onset (days), mean 6 SD < _7 days, n (%) >7 days, n (%) Refractive error (SEQ, diopters), mean 6 SD Phakia, n (%) Macula off, n (%) Number of tears, mean 6 SD Tear location, n (%) Superior Inferior Combined Extent of RD (clock hours), mean 6 SD 1-6 o’clock, n (%) 7-12 o’clock, n (%) Preoperative severe retinal undulation, n (%) Operation methods, n (%) Pars plana vitrectomy Scleral buckling Combined Preoperative BCVA (logMAR), mean 6 SD 1 month BCVA (logMAR), mean 6 SD Final BCVA (logMAR), mean 6 SD
Multiple SRF Blebs (þ) (N ¼ 40)
Multiple SRF Blebs (-) (N ¼ 145)
36.6 6 20.2 20 (50) 7 (17.5) 6 (15) 7 (17.5) 24 (60) 15.4 6 9.1 13.9 6 24.6 25 (62.5) 15 (37.5) 2.9 6 5.3 (N ¼ 25) 39 (97.5) 35 (87.5) 2.1 6 1.4
47.2 6 15.9 22 (15.2) 37 (25.5) 53 (36.6) 33 (22.8) 88 (60.7) 14.1 6 9.4 10.2 6 15.8 102 (70.3) 43 (29.7) 3.4 6 4.1 (N ¼ 75) 120 (82.8) 93 (64.1) 2.1 6 1.5
24 (60) 12 (30) 4 (10) 5.5 6 2.1 33 (82.5) 7 (17.5) 7/27 (25.9) 12 (30) 28 (70) 0 1.13 6 0.83 (N ¼ 32) 0.40 6 0.30 (N ¼ 33) 0.21 6 0.34 (N ¼ 23)
Odds Ratio
95% CI
4.3 0.9 0.5 1.0
1.6-11.8 0.3-2.8 0.2-1.7 0.5-2.0
0.6 -
0.3-1.3 -
8.1 3.9
1.1-61.9 1.4-10.6
102 (70.3) 26 (17.9) 17 (11.7) 5.2 6 1.9 123 (84.8) 22 (15.2) 13/100 (13)
1.0 2.0 -
0.3-3.2 0.5-7.1 -
0.8 2.3
0.3-2.1 0.8-6.6
88 (60.7) 54 (37.2) 3 (2.1) 0.94 6 0.88 (N ¼ 122) 0.38 6 0.32 (N ¼ 121) 0.14 6 0.18 (N ¼ 105)
0.3 -
0.1-0.6 -
-
-
P Valuea
.003 .005 .845 .294 .937 .460 .395 .230 .628 .018 .005 .823 .247 >.999 .305 .312 .720 .135 .001 .001 .231 .457 .982
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; LogMAR ¼ logarithm of the minimal angle of resolution; RD ¼ retinal detachment; SD ¼ standard deviation; SEQ ¼ spherical equivalent; SRF ¼ subretinal fluid. a P values by Student t test or Mann-Whitney test for continuous variables and by x2 test or Fisher exact test for categorical variables. P values that are statistically significant (<.05) are represented in bold.
(Figure 3, Table). Patients with multiple SRF blebs were also associated with phakia (97.5% vs 82.8%, P ¼ .018), preoperative macula-off (87.5% vs 64.1%, P ¼ .005) and SB surgery (70% vs 37.2%, P ¼ .001) compared to the control group. However, there were no significant differences between groups in terms of RD chronicity, the number or location of tears, the extent of the RD, or visual outcomes (Table). We also performed multiple logistic regression analysis using the 5 most significant factors as identified by the univariable analysis: age, lens status, preoperative macula status, preoperative severe undulation in OCT images, and operation method. We converted age into a dichoto_30 years), according to the results mous variable (<30 or > of the univariable analysis. Only younger age (<30 years) was significantly associated with the occurrence of multiple SRF blebs after RD surgery (odds ratio, 5.1; 95% confidence interval, 1.5-17.6; P ¼ .010). VOL. 157, NO. 4
We have presented representative serial fundus IR and SDOCT images showing the dramatic resolution of a prominent cluster of SRF blebs after RD surgery (Figure 4). An 11-year-old boy visited the emergency room complaining of decreased vision with shadowing in the right eye that had started 1 day earlier. Visual acuity was hand motion (OD), 20/25 (OS); and refractive error was 5.75 Dsph ¼ 0.75 Dcyl 3 Axis 180 (OD), 5.0 Dsph ¼ 1.25 Dcyl 3 Axis 180 (OS). The fundus examination revealed a temporal macula-off RD with 3 peripherally located tears at the 8, 9, and 10 o’clock position. He underwent SB and cryotherapy with a 506 silicone sponge. Visual acuity in the right eye gradually improved after surgery, ultimately reaching 20/25 postoperatively. In preoperative OCT images, the outer portion of the detached retina showed severe undulation (Figure 4, Top left). One month after surgery, the detached retina was approximated on RPE, and multiple SRF blebs were observed. However, these blebs were not readily visible
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
837
FIGURE 4. Natural course of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment surgery as presented by serial infrared reflectance (IR) scanning laser ophthalmoscopy and spectral-domain optical coherence tomography (SDOCT) images. An 11-year-old boy underwent scleral buckling and cryotherapy for macula-off rhegmatogenous retinal detachment in his right eye. (Top left) Preoperatively, severe outer retinal undulation was observed in SDOCT. Visual acuity was hand motion. (Second row left) At 1 month postoperatively, the early SRF blebs were observed (magnified in Bottom middle). Visual acuity was 20/63. (Third row left) At 3 months postoperatively, they became most prominent, that is, they are readily observed in fundus IR images and the height of SRF bleb is highest in SDOCT (magnified in Bottom right). Visual acuity was 20/32. (Bottom left) Six months after surgery. Visual acuity was 20/32. (Top right) Nine months after surgery. Visual acuity was 20/32. (Second row right) One year after surgery. Visual acuity was 20/25. (Third row right) The SRF blebs gradually regressed and finally completely disappeared in both SDOCT and fundus IR images at 2 years postoperatively. Visual acuity was 20/25. All SDOCT images were obtained along the green arrows in IR images.
on fundus IR images (Figure 4, Second row left and Bottom middle). Three months after surgery, the SRF blebs were prominently visible on fundus IR images, and SRF bleb height on OCT images had increased as well (Figure 4, Third row left and Bottom right). The SRF bleb size decreased thereafter, regressing gradually until complete absorption 2 years after the surgery (Figure 4, Third row right). We selected 1 representative SRF bleb from each patient and serially measured the width and height of it, then compared the values at baseline and at its most prominent phase. Of 40 patients with multiple SRF blebs, we were 838
able to follow SRF bleb evolution in 23 cases. The SRF blebs were first recognizable at 1.1 6 0.6 months postoperatively, becoming most prominent (ie, highest in OCT and most prominent in fundus IR images) at 2.9 6 0.9 months postoperatively. Compared to the initial stage, SRF blebs became narrower (576.0 6 338.3 mm vs 759.5 6 530.5 mm, P ¼ .007) and higher (55.5 6 19.5 mm vs 46.8 6 22.6 mm, P ¼ .048) at its prominent stage (Figure 5). We also found that the boundaries of the SRF bleb respected major retinal vessels. Although the margins of the SRF bleb sometimes invaded the narrow, terminal branches of retinal vessels, in
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
FIGURE 5. Serial size measurement of the subretinal fluid (SRF) bleb after successful rhegmatogenous retinal detachment surgery. We selected 1 representative SRF bleb and followed its size in spectral-domain optical coherence tomography (SDOCT) in 23 cases. The size of SRF bleb was compared between its initial stage (when the SRF bleb is first detected in infrared reflectance [IR] and SDOCT images, at an average of 1.1 ± 0.6 months postoperatively) and its most prominent stage (when the SRF bleb is most prominently visible in fundus IR images and highest in SDOCT, at an average of 2.9 ± 0.9 months postoperatively). P values by Wilcoxon signed rank test.
most cases they did not cross over the medium to large retinal vessels, which were the first or the second branches of main arcade retinal vessels (Figure 6).
DISCUSSION OUR STUDY SHOWED THAT MULTIPLE SRF BLEBS ARE
commonly found after the successful treatment of primary rhegmatogenous RD. Multiple SRF blebs are numerous (more than 3) bleb-like SRF pockets, which can occur anywhere in previously detached retina, and they are especially prevalent in young patients. These blebs can be visible as early as postoperative 1 month; however, they are most easily recognized at an average of 3 months after the surgery, and SRF blebs usually respect medium to large retinal vessels. These lesions can easily be differentiated from fullthickness retinal folds that might cause visual impairment,12–14 outer retinal folds commonly found after vitrectomy and gas injection,15 or subclinical persistent SRF, which usually occurs in the subfoveal area.1,5 In this study, the incidence of multiple SRF blebs was 21.6% (40 out of 185 cases), which was high compared to the previously reported incidences: 15.3% (15 out of 98 patients) by Benson and associates4 and 9.3% (11 out of 118 cases) reported by Kang and associates.11 This discrepancy might be attributable to the differences in the sensitivity of the imaging tools used for detecting SRF blebs. VOL. 157, NO. 4
We used SDOCT, which can scan a broader area of retina in a shorter time with higher resolution compared to the previously used time-domain OCT. We also used the fundus IR image, which was combined with the OCT image, and it was much easier to detect SRF blebs using fundus IR images, as they appear as dark ovoid lesions that are highly contrasted to the non-SRF bleb area. Kang and associates reported that the SRF bleb was first detected 8.7 6 5.5 weeks after retinal reattachment which was confirmed funduscopically at 4.7 6 3.8 days after surgery. Here we first detected SRF blebs as early as 1.1 6 0.6 months postoperatively; the SRF blebs became most prominent at 2.9 6 0.9 months postoperatively. There are possibilities that they found SRF blebs only after they became prominent, and what they thought of as complete retinal reattachment which was assessed only by funduscopy would have been in reality, just slight approximation of sensory retina to RPE with shallow SRF, if they had examined with SDOCT. The most consistent and significant factor associated with occurrence of multiple SRF blebs was young age. Especially when we divided age groups into quartiles, we found that postoperative SRF blebs were most prevalent in the youngest age group (<30 years); however, there were no significant differences between the other age groups. The other clinical factors, such as lens status, preoperative macula status, and operation method, which were significantly associated with the occurrence of SRF blebs in univariable analysis, did not show significance in multivariable analysis. The SRF blebs showed peculiar serial morphologic changes, as their width decreased and height increased slightly. These temporal morphologic changes of SRF blebs, in combination with the findings that SRF blebs are more prevalent in younger patients, suggest that there might be an active interaction between healthy RPE and photoreceptors. Normally, the RPE apical villi interdigitate with photoreceptor outer segments, and this plays a crucial role in disc phagocytosis and renewal. However, their role in adhesion is uncertain.16 It might provide frictional resistance or an electrostatic force that opposes separation.16,17 In Figure 7, we represent a schematic diagram of the presumed mechanism of SRF bleb formation. In the initial stage, the photoreceptor is just weakly approximated on RPE. The appearance of SRF bleb is rather diffuse, that is, broad in base and short in height. During the natural course, the SRF is pumped through the RPE to the choroid, and hydrostatic pressure inside the SRF bleb decreases. We also postulated an imaginary hydrostatic pressure increase caused by volume decreases in subretinal space as RPE and photoreceptors reattach. If the reattachment process between RPE and photoreceptors were a passive mechanism that occurs only after SRF is removed through RPE, the hydrostatic pressure change caused by SRF absorption and that caused by RPE and photoreceptor reattachment would be in balance. However, if there is an active reattachment process between RPE and photoreceptors at the margin of
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
839
FIGURE 6. Infrared reflectance image and schematic drawings demonstrating the relationship between subretinal fluid bleb margins and retinal vessels. The subretinal fluid bleb margins usually respect medium to large retinal vessels.
FIGURE 7. Schematic diagrammatic representation of the presumed mechanism of subretinal fluid bleb formation. In natural course, subretinal fluid (SRF) is absorbed through retinal pigment epithelium (RPE) to the choroid, and hydrostatic pressure inside SRF bleb is decreasing (DPAbs). DPRPE is an imaginary hydrostatic pressure increase caused by RPE and photoreceptor reattachment, and thereby diminished SRF bleb space. If the reattachment between RPE and photoreceptor is just a passive process, then DPAbs and DPRPE will be balanced (DPAbs [ DPRPE). However, if there is an active reattachment process between RPE and photoreceptor, beyond the speed of SRF absorption, the hydrostatic pressure increase caused by RPE and photoreceptor reattachment (DPRPE) will surpass the decrease in hydrostatic pressure achieved by SRF absorption (DPAbs), leading to increased hydrostatic pressure inside the SRF bleb (Phydrostatic £ P9 hydrostatic). This will result in prominent SRF bleb with diminished width and increased height, and also an increased height-to-width ratio.
residual SRF and the reattachment speed is greater than the speed of natural SRF absorption, the hydrostatic pressure inside the subretinal space will increase and this might create a prominent SRF bleb, which appears to be shorter in width and higher in height on SDOCT. Another remarkable finding regarding multiple SRF blebs is that their margins usually respect the retinal vessels. We found that although invasion of smaller retinal vessels was possible, the bleb did not invade the large retinal vessels, such as the first and second branches of the major arcade retinal vessels. This suggests that compared to the other retinal parts, the less-pliable retina under the large retinal vessels reattaches first during the RD resolution and comprises the SRF bleb boundary. On the other hand, SRF bleb formation was not associated with detached retinal undulation, which is occasionally observed on preoperative OCT. Intraretinal cyst formation, separation, and undulation of outer retina are the characteristics of detached retina investigated by OCT,18–20 and undulation of detached retina is thought to be secondary to cystoid degeneration of the retina and intraretinal or subretinal proliferation 840
of non-neural cell types.21,22 However, despite its morphologic similarities, preoperative retinal undulation does not seem to be related to postoperative SRF bleb formation. Benson and associates also observed wrinkling of the outer retinal surface in detached retinas in 15 patients, and found that this wrinkling persisted even after the retina was opposed to the RPE. However, there was no association between the presence of wrinkling and SRF observed on OCT at 6 weeks postoperatively.4 Thus, we can postulate that the main cause of multiple SRF blebs is an active interaction between healthy RPE and photoreceptor during a normal reattachment process, whereas a preexisting retinal undulation may be only partly associated. Regarding the persistence of SRF after RD surgery, it is well known that patients undergoing PPV show much faster SRF absorption compared to those undergoing SB surgery.2,3,7 We also investigated risk factors associated with persistent submacular fluid after successful SB surgery, and found that preoperative macula-off status (vs maculasplitting) and placement of segmental SB (vs encircling) were significantly associated with the 1-month-postoperative
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
persistent submacular fluid in multiple logistic regression analysis.6 However, these significant risk factors did not correspond with risk factors of multiple SRF blebs in this study. This difference suggests that the pathogenesis between residual SRF and multiple SRF blebs after successful RD surgery are different, in that residual SRF is caused by passive accumulation of fluid from delayed fluid absorption and multiple SRF blebs are caused by active reattachment between RPE and photoreceptors. This study is limited by its retrospective design. The follow-up interval for each patient was different; and in some cases, the OCT follow-up duration was relatively short. Therefore, in these cases, we were unable to trace the natural course of the SRF blebs until their complete absorption. The 3-month interval period between OCT
examinations might have been too long to enable detection of any detailed dynamic changes in the SRF blebs. However, our study included sufficient cases to characterize the clinical association (more prevalent in young patients) as well as a sufficient number of serial OCT images to determine the bleb’s morphologic features. In addition, considering the relatively high prevalence of postoperative SRF blebs in our study as compared to that observed in previous reports, we performed more detailed postoperative examinations. In conclusion, multiple SRF blebs are commonly found after successful RD surgery, especially in young patients. Their serial morphologic features indicate that they might be caused by active reattachment between RPE and photoreceptor, which can be one of the mechanisms of RD resolution.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. This study was supported in part by a grant from the Korea Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (Grant No. A111161). Contributions of authors: involved in design of study (Y.-K.K., J.A., S.J.W., D.J.H., K.H.P.); acquisition of data (Y.-K.K., J.A.); analysis of data (Y.-K.K., J.A.); interpretation of data (Y.-K.K., J.A., S.J.W.); preparation of manuscript (Y.-K.K., J.A.); review of manuscript (Y.-K.K., J.A., S.J.W., D.J.H., K.H.P.).
REFERENCES 1. Hagimura N, Iida T, Suto K, Kishi S. Persistent foveal retinal detachment after successful rhegmatogenous retinal detachment surgery. Am J Ophthalmol 2002;133(4):516–520. 2. Wolfensberger TJ. Foveal reattachment after macula-off retinal detachment occurs faster after vitrectomy than after buckle surgery. Ophthalmology 2004;111(7):1340–1343. 3. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after vitrectomy surgery for retinal detachment. Ophthalmology 2006;113(7):1179–1183. 4. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after scleral buckle surgery for retinal detachment. Ophthalmology 2007;114(1):108–112. 5. Seo JH, Woo SJ, Park KH, et al. Influence of persistent submacular fluid on visual outcome after successful scleral buckle surgery for macula-off retinal detachment. Am J Ophthalmol 2008;145(5):915–922. 6. Kim YK, Woo SJ, Kim KE, Park KH. Subretinal fluid resorption. Ophthalmology 2010;117(8):1655.e1–1655.e4. 7. Kim YK, Woo SJ, Park KH, et al. Comparison of persistent submacular fluid in vitrectomy and scleral buckle surgery for macula-involving retinal detachment. Am J Ophthalmol 2010;149(4):623–629.e1. 8. Machemer R. Experimental retinal detachment in the owl monkey. IV. The reattached retina. Am J Ophthalmol 1968; 66(6):1075–1091. 9. Leaver PK, Chester GH, Saunders SH. Factors influencing absorption of subretinal fluid. Br J Ophthalmol 1976;60(8):557–560. 10. Kaga T, Fonseca RA, Dantas MA, et al. Optical coherence tomography of bleb-like subretinal lesions after retinal reattachment surgery. Am J Ophthalmol 2001;132(1):120–121. 11. Kang SW, Kim JH, Shin WJ, Kim JI. Subretinal fluid bleb after successful scleral buckling and cryotherapy for retinal detachment. Am J Ophthalmol 2008;146(2):205–210.
VOL. 157, NO. 4
12. van Meurs JC, Humalda D, Mertens DA, Peperkamp E. Retinal folds through the macula. Doc Ophthalmol 1991; 78(3-4):335–340. 13. Larrison WI, Frederick AR Jr, Peterson TJ, Topping TM. Posterior retinal folds following vitreoretinal surgery. Arch Ophthalmol 1993;111(5):621–625. 14. Wong R. Longitudinal study of macular folds by spectraldomain optical coherence tomography. Am J Ophthalmol 2012;153(1):88–92.e1. 15. dell’Omo R, Tan HS, Schlingemann RO, et al. Evolution of outer retinal folds occurring after vitrectomy for retinal detachment repair. Invest Ophthalmol Vis Sci 2012;53(13):7928–7935. 16. Marmor MF. Mechanisms of normal retinal adhesion. In: Hinton DF, Ryan SJ, eds. Retina. 5th ed. Los Angeles, CA: Elsevier; 2013:447–464. 17. Gingell D, Fornes JA. Demonstration of intermolecular forces in cell adhesion using a new electrochemical technique. Nature 1975;256(5514):210–211. 18. Hagimura N, Suto K, Iida T, Kishi S. Optical coherence tomography of the neurosensory retina in rhegmatogenous retinal detachment. Am J Ophthalmol 2000;129(2):186–190. 19. Lee SY, Joe SG, Kim JG, et al. Optical coherence tomography evaluation of detached macula from rhegmatogenous retinal detachment and central serous chorioretinopathy. Am J Ophthalmol 2008;145(6):1071–1076. 20. Nakanishi H, Hangai M, Unoki N, et al. Spectral-domain optical coherence tomography imaging of the detached macula in rhegmatogenous retinal detachment. Retina 2009;29(2):232–242. 21. Fisher SK, Erickson PA, Lewis GP, Anderson DH. Intraretinal proliferation induced by retinal detachment. Invest Ophthalmol Vis Sci 1991;32(6):1739–1748. 22. Sethi CS, Lewis GP, Fisher SK, et al. Glial remodeling and neural plasticity in human retinal detachment with proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 2005; 46(1):329–342.
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
841
To investigate the incidence and the clinical factors associated with the occurrence of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment (RD) repair. DESIGN: Retrospective, observational case series. METHODS: We retrospectively investigated the medical records of 185 eyes of 184 patients who had undergone successful RD surgery, either vitrectomy or scleral buckling. Each patient had undergone spectral-domain optical coherence tomography (SDOCT) combined with infrared reflectance (IR) imaging every 3 months postoperatively. We carefully examined postoperative SDOCT and fundus IR images, in an effort to identify any SRF blebs present. RESULTS: Multiple (‡3) SRF blebs were observed in 40 of 185 cases (21.6%). SRF blebs were first detected 1.7 ± 1.8 months postoperatively. In 22 cases that could be fully followed up, SRF blebs were completely absorbed 13.1 ± 6.1 months postoperatively. Multiple logistic regression analysis showed that only young age (<30 years) was significantly associated with the occurrence of multiple SRF blebs (odds ratio, 5.1; 95% confidence interval, 1.5-17.6; P [ .010). Serial measurements of SRF bleb size using SDOCT showed that SRF bleb height was greatest at postoperative 2.9 ± 0.9 months, while SRF bleb width tended to decrease gradually over time. The SRF blebs typically spared large retinal vessels. CONCLUSIONS: Multiple SRF blebs are commonly found after successful RD surgery, especially in young patients. The serial morphologic features evaluated in this study indicate that multiple SRF blebs may result from the active reattachment of retinal pigment epithelium and photoreceptors during the resolution of RD. (Am J Ophthalmol 2014;157:834–841. Ó 2014 by Elsevier Inc. All rights reserved.) Accepted for publication Dec 10, 2013. From the Department of Ophthalmology, Seoul National University College of Medicine, Seoul National University Bundang Hospital, Seongnam, South Korea (Y.K.K., J.A., S.J.W., D.J.H., K.H.P.); Department of Ophthalmology, Seoul National University College of Medicine, Seoul Metropolitan Government–Seoul National University Boramae Medical Center, Seoul, South Korea (J.A.); and Department of Ophthalmology, Han Gil Eye Hospital, Incheon, South Korea (D.J.H.). Inquiries to Se Joon Woo, Department of Ophthalmology, Seoul National University Bundang Hospital, #300, Gumi-dong, Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, South Korea; e-mail: sejoon1@snu. ac.kr
834
Ó
2014 BY
O
PTICAL COHERENCE TOMOGRAPHY (OCT) HAS
been used to elucidate the pathophysiology of many retinal diseases, which previously could only have been investigated through the use of histopathologic techniques. By incorporating OCT into the postoperative assessment of retinal detachment (RD) patients, we were able to detect subclinical levels of subretinal fluid (SRF) that is usually not detected through routine binocular ophthalmoscopic examination.1–7 Aside from the persistence of SRF, bleb-like lesions may arise in the previously detached retina during the postoperative follow-up period after successful retinal reattachment. These lesions appeared similar to those associated with serous detachment of the retinal pigment epithelium (RPE), and using fluorescein angiography, histopathologic review, and OCT, they were determined to be focal sensory retinal detachment.8–10 The described pockets were observed in subfoveal and peripheral areas, whereas the persistent SRF described previously was found primarily in dependent positions, the fovea, or the inferior part of the retina. Moreover, bleblike SRF pockets usually arise in clusters (more than 3), in contrast to confluent SRF or single SRF blebs. Kang and associates reported the incidence and presumed mechanism of SRF blebs. They observed SRF blebs in 11 of 118 patients (9.3%) who underwent successful scleral buckling (SB) and cryotherapy for rhegmatogenous RD. They proposed choroidal vascular damage caused by cryotherapy as the main mechanism of SRF bleb occurrence.11 However, when we evaluated multiple SRF blebs with serial spectral-domain OCT (SDOCT), we noticed additional characteristics that have not been suggested before. In this study, we aimed to investigate the incidence and clinical risk factors associated with multiple SRF blebs that could be found after successful rhegmatogenous RD surgery and also to elucidate the presumed mechanism of SRF bleb occurrence.
METHODS WE RETROSPECTIVELY REVIEWED THE MEDICAL RECORDS OF
all the patients who underwent either pars plana vitrectomy (PPV) or SB surgery for rhegmatogenous RD at Seoul National University Bundang Hospital from January 1,
ELSEVIER INC. ALL
RIGHTS RESERVED.
0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.12.030
2009, to June 30, 2012. This study followed the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Seoul National University Bundang Hospital. A total of 340 patients (343 eyes) underwent rhegmatogenous RD repair during that period, and ultimately 185 eyes from 184 patients were included in this study. We included only patients with primary spontaneous RD who achieved a successful postoperative course, and we excluded those with complicated RD, primary failure cases, or re-detachment cases. Specifically, we excluded patients who underwent silicone oil injection for retinal reattachment. We also excluded those with preoperative retinal pathology that could have influenced retinal reattachment—for example, trauma history, epiretinal membrane, vitreomacular traction syndrome, combined tractional RD, proliferative vitreoretinopathy (subretinal gliosis), and macular pathology. We excluded reoperation cases as well as patients with incomplete courses of postoperative OCT follow-up. All surgeries were performed by 4 experienced retinal surgeons at Seoul National University Bundang Hospital (P.K.H., W.S.J., A.J., and H.D.J.). The surgeons decided the method of operation, either PPV or SB, based on the clinical characteristics of each patient and RD. Younger patients and those with small definite tears, especially those located inferiorly, were likely to undergo SB, whereas elderly patients with cataracts or pseudophakia and suspicious or hidden tears were more likely to undergo PPV. A 23 gauge transconjunctival sutureless vitrectomy system (Accurus; Alcon Laboratories, Inc, Fort Worth, Texas, USA) was used for PPV. In brief, we used perfluorocarbon liquids during the surgery, and internal SRF drainage was performed through the original tear. After the surgery, the eye was filled with gas (18% SF6 or 14% C3F8). In SB surgery, we identified the retinal tear and then performed cryotherapy. The explant material was silicone sponge with width of 5 or 7.5 mm (No. 506 or 507; MIRA, Waltham, Massachusetts, USA) for the segmental SB and silicone tire (No. 287; MIRA) and silicone band (No. 240; MIRA) for the encircling SB. External SRF drainage was performed during SB surgery in 7 of 82 cases (8.5%). Each patient underwent a thorough ophthalmologic examination, including best-corrected visual acuity (BCVA; Snellen visual acuity chart), slit-lamp examination, binocular indirect ophthalmoscopy, and SDOCT combined with infrared reflectance (IR) confocal scanning laser ophthalmoscopy (Spectralis OCT; Heidelberg Engineering, Heidelberg, Germany) at 1 month after surgery and every 3 months thereafter until any SRF disappeared. For the visual acuity analysis, we converted Snellen visual acuity values to the logarithm of the minimal angle of resolution (logMAR). We evaluated only patients without any preoperative ocular disorders that might influence visual function (eg, amblyopia and untreated severe cataracts). Final visual acuity was defined as visual acuity obtained at least 6 months after the complete resolution VOL. 157, NO. 4
FIGURE 1. Survival curve of subretinal fluid blebs after successful rhegmatogenous retinal detachment surgery. Tick marks indicate censored (incomplete optical coherence tomography follow-up) cases.
of any subfoveal SRF and at least 3 months after cataract surgery in patients who underwent cataract surgery during the follow-up period. SUBRETINAL FLUID BLEB AND OPTICAL COHERENCE TOMOGRAPHY ANALYSIS: We carefully searched for the
presence of SRF bleb using both postoperative SDOCT images and fundus IR images. We retrieved fundus IR images to identify any bleb-like lesion that was dark and round, and then examined SDOCT volume scans spanning an area of 30 3 20 degrees for any SRF bleb lesion. If there were peripherally located SRF blebs beyond the scope of routine imaging, we obtained peripheral images. We defined multiple SRF blebs as 3 or more localized fluid pockets under the extrafoveal retina with or without subfoveal bleb. We excluded those cases with only a single subfoveal bleb. When preoperative SDOCT images were available, we evaluated whether there was any severe undulation (>3 undulations in a single section) in the outer portion of the detached retina. To characterize the serial morphologic changes typical of SRF bleb, we selected 1 SRF bleb from each patient that could be traced with SDOCT throughout the clinical course, then measured its width and height at baseline (when the SRF bleb first appeared) and at its most prominent stage (when the SRF bleb was highest in OCT images and prominent in IR images) using the caliper tool in the OCT software. STATISTICAL ANALYSIS:
We compared preoperative and intraoperative characteristics between those with and without multiple SRF blebs. Some of these factors were significantly associated with the occurrence of SRF bleb in the univariate analysis, which was performed using either the Student t test for continuous variables or the x2 or Fisher exact test for categorical variables. We performed multiple logistic regression analysis with these
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
835
FIGURE 2. Subretinal fluid blebs after successful rhegmatogenous retinal detachment surgery. (Left) One month after the surgery, multiple bleb-like lesions are visible on superior temporal area. (Middle) Three months after the surgery, blebs became more prominent, represented as dark, round lesions on infrared reflectance (IR) scanning laser ophthalmoscopy. (Right) These lesions are revealed to be subretinal fluid (SRF) pockets in spectral-domain optical coherence tomography (SDOCT). The SDOCT images were obtained along the green arrows in IR images. Note that subfoveal fluid bleb (Right, lower) is located apart from multiple SRF blebs in the superior temporal area (Right, upper).
FIGURE 3. Comparison of the incidence of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment surgery in different age groups. *P value < .05, Fisher exact test.
factors to eliminate confounding effects. All statistical analyses were performed using PASW version 18.0 (SPSS, Inc, Chicago, Illinois, USA), and P values < .05 were considered statistically significant.
RESULTS THE 185 EYES (184 PATIENTS) THAT UNDERWENT SUCCESS-
ful primary rhegmatogenous RD repair were ultimately included. Multiple SRF blebs were observed postoperatively in 40 eyes (40 patients) out of 185 eyes (21.6%). In most cases (33 cases, 82.5%), SRF blebs were first 836
detected at 1 month postoperatively; another 5 cases (12.5%) exhibited SRF blebs within 6 months after the surgery. In 2 cases (5%) with diffuse persistent subclinical SRF, SRF blebs first occurred at 8 months postoperatively, as SRF was absorbed. Overall, SRF bleb was first identified on SDOCT 1.7 6 1.8 (mean 6 SD) months postoperatively. In 22 cases, we were able to confirm the complete absorption of SRF blebs with SDOCT during the followup period. The SRF blebs were completely absorbed within 1 year after surgery in 10 cases (45.5%) and within 2 years postoperatively in another 10 cases (45.5%). The longest duration of SRF blebs was 28 months postoperatively (13.1 6 6.1 months postoperatively). For those 18 cases that we could not follow fully until the SRF blebs were completely absorbed, the OCT follow-up period was relatively short (3.6 6 2.4 months). The cumulative survival curve of SRF bleb during the follow-up period is represented in Figure 1. Figure 2 shows an example of a color fundus photograph and fundus IR and SDOCT images of multiple SRF blebs. The SRF blebs are more readily visible on IR images, as round dark lesions. They can be found anywhere in the previously detached retina, either submacularly or peripherally. In SDOCT images, these dark lesions were revealed to be small subretinal fluid pockets. In Figure 2, the subfoveal fluid, separated from multiple SRF blebs in the superior temporal area, is apparent. In this study, we did not count eyes with subfoveal fluid alone as a multiple SRF bleb case. Demographics and clinical characteristics were compared between eyes with and without multiple SRF blebs. Patients with multiple SRF blebs were younger (36.6 6 20.2 years vs 47.2 6 15.9 years, P ¼ .003) than those in the control group. When we stratified patient age by quartile, multiple SRF blebs were most prevalent in the youngest age group (<30 years), while the other 3 age groups did not show any significant differences
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
TABLE. Demographic and Clinical Characteristic Comparison Between Patients With and Without Multiple Subretinal Fluid Blebs After Successful Retinal Detachment Surgery
Age (y), mean 6 SD <30, n (%) > _30, <45, n (%) > _45, <60, n (%) > _60, n (%) Male, n (%) Follow-up period (months), mean 6 SD Symptom onset (days), mean 6 SD < _7 days, n (%) >7 days, n (%) Refractive error (SEQ, diopters), mean 6 SD Phakia, n (%) Macula off, n (%) Number of tears, mean 6 SD Tear location, n (%) Superior Inferior Combined Extent of RD (clock hours), mean 6 SD 1-6 o’clock, n (%) 7-12 o’clock, n (%) Preoperative severe retinal undulation, n (%) Operation methods, n (%) Pars plana vitrectomy Scleral buckling Combined Preoperative BCVA (logMAR), mean 6 SD 1 month BCVA (logMAR), mean 6 SD Final BCVA (logMAR), mean 6 SD
Multiple SRF Blebs (þ) (N ¼ 40)
Multiple SRF Blebs (-) (N ¼ 145)
36.6 6 20.2 20 (50) 7 (17.5) 6 (15) 7 (17.5) 24 (60) 15.4 6 9.1 13.9 6 24.6 25 (62.5) 15 (37.5) 2.9 6 5.3 (N ¼ 25) 39 (97.5) 35 (87.5) 2.1 6 1.4
47.2 6 15.9 22 (15.2) 37 (25.5) 53 (36.6) 33 (22.8) 88 (60.7) 14.1 6 9.4 10.2 6 15.8 102 (70.3) 43 (29.7) 3.4 6 4.1 (N ¼ 75) 120 (82.8) 93 (64.1) 2.1 6 1.5
24 (60) 12 (30) 4 (10) 5.5 6 2.1 33 (82.5) 7 (17.5) 7/27 (25.9) 12 (30) 28 (70) 0 1.13 6 0.83 (N ¼ 32) 0.40 6 0.30 (N ¼ 33) 0.21 6 0.34 (N ¼ 23)
Odds Ratio
95% CI
4.3 0.9 0.5 1.0
1.6-11.8 0.3-2.8 0.2-1.7 0.5-2.0
0.6 -
0.3-1.3 -
8.1 3.9
1.1-61.9 1.4-10.6
102 (70.3) 26 (17.9) 17 (11.7) 5.2 6 1.9 123 (84.8) 22 (15.2) 13/100 (13)
1.0 2.0 -
0.3-3.2 0.5-7.1 -
0.8 2.3
0.3-2.1 0.8-6.6
88 (60.7) 54 (37.2) 3 (2.1) 0.94 6 0.88 (N ¼ 122) 0.38 6 0.32 (N ¼ 121) 0.14 6 0.18 (N ¼ 105)
0.3 -
0.1-0.6 -
-
-
P Valuea
.003 .005 .845 .294 .937 .460 .395 .230 .628 .018 .005 .823 .247 >.999 .305 .312 .720 .135 .001 .001 .231 .457 .982
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; LogMAR ¼ logarithm of the minimal angle of resolution; RD ¼ retinal detachment; SD ¼ standard deviation; SEQ ¼ spherical equivalent; SRF ¼ subretinal fluid. a P values by Student t test or Mann-Whitney test for continuous variables and by x2 test or Fisher exact test for categorical variables. P values that are statistically significant (<.05) are represented in bold.
(Figure 3, Table). Patients with multiple SRF blebs were also associated with phakia (97.5% vs 82.8%, P ¼ .018), preoperative macula-off (87.5% vs 64.1%, P ¼ .005) and SB surgery (70% vs 37.2%, P ¼ .001) compared to the control group. However, there were no significant differences between groups in terms of RD chronicity, the number or location of tears, the extent of the RD, or visual outcomes (Table). We also performed multiple logistic regression analysis using the 5 most significant factors as identified by the univariable analysis: age, lens status, preoperative macula status, preoperative severe undulation in OCT images, and operation method. We converted age into a dichoto_30 years), according to the results mous variable (<30 or > of the univariable analysis. Only younger age (<30 years) was significantly associated with the occurrence of multiple SRF blebs after RD surgery (odds ratio, 5.1; 95% confidence interval, 1.5-17.6; P ¼ .010). VOL. 157, NO. 4
We have presented representative serial fundus IR and SDOCT images showing the dramatic resolution of a prominent cluster of SRF blebs after RD surgery (Figure 4). An 11-year-old boy visited the emergency room complaining of decreased vision with shadowing in the right eye that had started 1 day earlier. Visual acuity was hand motion (OD), 20/25 (OS); and refractive error was 5.75 Dsph ¼ 0.75 Dcyl 3 Axis 180 (OD), 5.0 Dsph ¼ 1.25 Dcyl 3 Axis 180 (OS). The fundus examination revealed a temporal macula-off RD with 3 peripherally located tears at the 8, 9, and 10 o’clock position. He underwent SB and cryotherapy with a 506 silicone sponge. Visual acuity in the right eye gradually improved after surgery, ultimately reaching 20/25 postoperatively. In preoperative OCT images, the outer portion of the detached retina showed severe undulation (Figure 4, Top left). One month after surgery, the detached retina was approximated on RPE, and multiple SRF blebs were observed. However, these blebs were not readily visible
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
837
FIGURE 4. Natural course of multiple subretinal fluid (SRF) blebs after successful rhegmatogenous retinal detachment surgery as presented by serial infrared reflectance (IR) scanning laser ophthalmoscopy and spectral-domain optical coherence tomography (SDOCT) images. An 11-year-old boy underwent scleral buckling and cryotherapy for macula-off rhegmatogenous retinal detachment in his right eye. (Top left) Preoperatively, severe outer retinal undulation was observed in SDOCT. Visual acuity was hand motion. (Second row left) At 1 month postoperatively, the early SRF blebs were observed (magnified in Bottom middle). Visual acuity was 20/63. (Third row left) At 3 months postoperatively, they became most prominent, that is, they are readily observed in fundus IR images and the height of SRF bleb is highest in SDOCT (magnified in Bottom right). Visual acuity was 20/32. (Bottom left) Six months after surgery. Visual acuity was 20/32. (Top right) Nine months after surgery. Visual acuity was 20/32. (Second row right) One year after surgery. Visual acuity was 20/25. (Third row right) The SRF blebs gradually regressed and finally completely disappeared in both SDOCT and fundus IR images at 2 years postoperatively. Visual acuity was 20/25. All SDOCT images were obtained along the green arrows in IR images.
on fundus IR images (Figure 4, Second row left and Bottom middle). Three months after surgery, the SRF blebs were prominently visible on fundus IR images, and SRF bleb height on OCT images had increased as well (Figure 4, Third row left and Bottom right). The SRF bleb size decreased thereafter, regressing gradually until complete absorption 2 years after the surgery (Figure 4, Third row right). We selected 1 representative SRF bleb from each patient and serially measured the width and height of it, then compared the values at baseline and at its most prominent phase. Of 40 patients with multiple SRF blebs, we were 838
able to follow SRF bleb evolution in 23 cases. The SRF blebs were first recognizable at 1.1 6 0.6 months postoperatively, becoming most prominent (ie, highest in OCT and most prominent in fundus IR images) at 2.9 6 0.9 months postoperatively. Compared to the initial stage, SRF blebs became narrower (576.0 6 338.3 mm vs 759.5 6 530.5 mm, P ¼ .007) and higher (55.5 6 19.5 mm vs 46.8 6 22.6 mm, P ¼ .048) at its prominent stage (Figure 5). We also found that the boundaries of the SRF bleb respected major retinal vessels. Although the margins of the SRF bleb sometimes invaded the narrow, terminal branches of retinal vessels, in
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
FIGURE 5. Serial size measurement of the subretinal fluid (SRF) bleb after successful rhegmatogenous retinal detachment surgery. We selected 1 representative SRF bleb and followed its size in spectral-domain optical coherence tomography (SDOCT) in 23 cases. The size of SRF bleb was compared between its initial stage (when the SRF bleb is first detected in infrared reflectance [IR] and SDOCT images, at an average of 1.1 ± 0.6 months postoperatively) and its most prominent stage (when the SRF bleb is most prominently visible in fundus IR images and highest in SDOCT, at an average of 2.9 ± 0.9 months postoperatively). P values by Wilcoxon signed rank test.
most cases they did not cross over the medium to large retinal vessels, which were the first or the second branches of main arcade retinal vessels (Figure 6).
DISCUSSION OUR STUDY SHOWED THAT MULTIPLE SRF BLEBS ARE
commonly found after the successful treatment of primary rhegmatogenous RD. Multiple SRF blebs are numerous (more than 3) bleb-like SRF pockets, which can occur anywhere in previously detached retina, and they are especially prevalent in young patients. These blebs can be visible as early as postoperative 1 month; however, they are most easily recognized at an average of 3 months after the surgery, and SRF blebs usually respect medium to large retinal vessels. These lesions can easily be differentiated from fullthickness retinal folds that might cause visual impairment,12–14 outer retinal folds commonly found after vitrectomy and gas injection,15 or subclinical persistent SRF, which usually occurs in the subfoveal area.1,5 In this study, the incidence of multiple SRF blebs was 21.6% (40 out of 185 cases), which was high compared to the previously reported incidences: 15.3% (15 out of 98 patients) by Benson and associates4 and 9.3% (11 out of 118 cases) reported by Kang and associates.11 This discrepancy might be attributable to the differences in the sensitivity of the imaging tools used for detecting SRF blebs. VOL. 157, NO. 4
We used SDOCT, which can scan a broader area of retina in a shorter time with higher resolution compared to the previously used time-domain OCT. We also used the fundus IR image, which was combined with the OCT image, and it was much easier to detect SRF blebs using fundus IR images, as they appear as dark ovoid lesions that are highly contrasted to the non-SRF bleb area. Kang and associates reported that the SRF bleb was first detected 8.7 6 5.5 weeks after retinal reattachment which was confirmed funduscopically at 4.7 6 3.8 days after surgery. Here we first detected SRF blebs as early as 1.1 6 0.6 months postoperatively; the SRF blebs became most prominent at 2.9 6 0.9 months postoperatively. There are possibilities that they found SRF blebs only after they became prominent, and what they thought of as complete retinal reattachment which was assessed only by funduscopy would have been in reality, just slight approximation of sensory retina to RPE with shallow SRF, if they had examined with SDOCT. The most consistent and significant factor associated with occurrence of multiple SRF blebs was young age. Especially when we divided age groups into quartiles, we found that postoperative SRF blebs were most prevalent in the youngest age group (<30 years); however, there were no significant differences between the other age groups. The other clinical factors, such as lens status, preoperative macula status, and operation method, which were significantly associated with the occurrence of SRF blebs in univariable analysis, did not show significance in multivariable analysis. The SRF blebs showed peculiar serial morphologic changes, as their width decreased and height increased slightly. These temporal morphologic changes of SRF blebs, in combination with the findings that SRF blebs are more prevalent in younger patients, suggest that there might be an active interaction between healthy RPE and photoreceptors. Normally, the RPE apical villi interdigitate with photoreceptor outer segments, and this plays a crucial role in disc phagocytosis and renewal. However, their role in adhesion is uncertain.16 It might provide frictional resistance or an electrostatic force that opposes separation.16,17 In Figure 7, we represent a schematic diagram of the presumed mechanism of SRF bleb formation. In the initial stage, the photoreceptor is just weakly approximated on RPE. The appearance of SRF bleb is rather diffuse, that is, broad in base and short in height. During the natural course, the SRF is pumped through the RPE to the choroid, and hydrostatic pressure inside the SRF bleb decreases. We also postulated an imaginary hydrostatic pressure increase caused by volume decreases in subretinal space as RPE and photoreceptors reattach. If the reattachment process between RPE and photoreceptors were a passive mechanism that occurs only after SRF is removed through RPE, the hydrostatic pressure change caused by SRF absorption and that caused by RPE and photoreceptor reattachment would be in balance. However, if there is an active reattachment process between RPE and photoreceptors at the margin of
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
839
FIGURE 6. Infrared reflectance image and schematic drawings demonstrating the relationship between subretinal fluid bleb margins and retinal vessels. The subretinal fluid bleb margins usually respect medium to large retinal vessels.
FIGURE 7. Schematic diagrammatic representation of the presumed mechanism of subretinal fluid bleb formation. In natural course, subretinal fluid (SRF) is absorbed through retinal pigment epithelium (RPE) to the choroid, and hydrostatic pressure inside SRF bleb is decreasing (DPAbs). DPRPE is an imaginary hydrostatic pressure increase caused by RPE and photoreceptor reattachment, and thereby diminished SRF bleb space. If the reattachment between RPE and photoreceptor is just a passive process, then DPAbs and DPRPE will be balanced (DPAbs [ DPRPE). However, if there is an active reattachment process between RPE and photoreceptor, beyond the speed of SRF absorption, the hydrostatic pressure increase caused by RPE and photoreceptor reattachment (DPRPE) will surpass the decrease in hydrostatic pressure achieved by SRF absorption (DPAbs), leading to increased hydrostatic pressure inside the SRF bleb (Phydrostatic £ P9 hydrostatic). This will result in prominent SRF bleb with diminished width and increased height, and also an increased height-to-width ratio.
residual SRF and the reattachment speed is greater than the speed of natural SRF absorption, the hydrostatic pressure inside the subretinal space will increase and this might create a prominent SRF bleb, which appears to be shorter in width and higher in height on SDOCT. Another remarkable finding regarding multiple SRF blebs is that their margins usually respect the retinal vessels. We found that although invasion of smaller retinal vessels was possible, the bleb did not invade the large retinal vessels, such as the first and second branches of the major arcade retinal vessels. This suggests that compared to the other retinal parts, the less-pliable retina under the large retinal vessels reattaches first during the RD resolution and comprises the SRF bleb boundary. On the other hand, SRF bleb formation was not associated with detached retinal undulation, which is occasionally observed on preoperative OCT. Intraretinal cyst formation, separation, and undulation of outer retina are the characteristics of detached retina investigated by OCT,18–20 and undulation of detached retina is thought to be secondary to cystoid degeneration of the retina and intraretinal or subretinal proliferation 840
of non-neural cell types.21,22 However, despite its morphologic similarities, preoperative retinal undulation does not seem to be related to postoperative SRF bleb formation. Benson and associates also observed wrinkling of the outer retinal surface in detached retinas in 15 patients, and found that this wrinkling persisted even after the retina was opposed to the RPE. However, there was no association between the presence of wrinkling and SRF observed on OCT at 6 weeks postoperatively.4 Thus, we can postulate that the main cause of multiple SRF blebs is an active interaction between healthy RPE and photoreceptor during a normal reattachment process, whereas a preexisting retinal undulation may be only partly associated. Regarding the persistence of SRF after RD surgery, it is well known that patients undergoing PPV show much faster SRF absorption compared to those undergoing SB surgery.2,3,7 We also investigated risk factors associated with persistent submacular fluid after successful SB surgery, and found that preoperative macula-off status (vs maculasplitting) and placement of segmental SB (vs encircling) were significantly associated with the 1-month-postoperative
AMERICAN JOURNAL OF OPHTHALMOLOGY
APRIL 2014
persistent submacular fluid in multiple logistic regression analysis.6 However, these significant risk factors did not correspond with risk factors of multiple SRF blebs in this study. This difference suggests that the pathogenesis between residual SRF and multiple SRF blebs after successful RD surgery are different, in that residual SRF is caused by passive accumulation of fluid from delayed fluid absorption and multiple SRF blebs are caused by active reattachment between RPE and photoreceptors. This study is limited by its retrospective design. The follow-up interval for each patient was different; and in some cases, the OCT follow-up duration was relatively short. Therefore, in these cases, we were unable to trace the natural course of the SRF blebs until their complete absorption. The 3-month interval period between OCT
examinations might have been too long to enable detection of any detailed dynamic changes in the SRF blebs. However, our study included sufficient cases to characterize the clinical association (more prevalent in young patients) as well as a sufficient number of serial OCT images to determine the bleb’s morphologic features. In addition, considering the relatively high prevalence of postoperative SRF blebs in our study as compared to that observed in previous reports, we performed more detailed postoperative examinations. In conclusion, multiple SRF blebs are commonly found after successful RD surgery, especially in young patients. Their serial morphologic features indicate that they might be caused by active reattachment between RPE and photoreceptor, which can be one of the mechanisms of RD resolution.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. This study was supported in part by a grant from the Korea Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea (Grant No. A111161). Contributions of authors: involved in design of study (Y.-K.K., J.A., S.J.W., D.J.H., K.H.P.); acquisition of data (Y.-K.K., J.A.); analysis of data (Y.-K.K., J.A.); interpretation of data (Y.-K.K., J.A., S.J.W.); preparation of manuscript (Y.-K.K., J.A.); review of manuscript (Y.-K.K., J.A., S.J.W., D.J.H., K.H.P.).
REFERENCES 1. Hagimura N, Iida T, Suto K, Kishi S. Persistent foveal retinal detachment after successful rhegmatogenous retinal detachment surgery. Am J Ophthalmol 2002;133(4):516–520. 2. Wolfensberger TJ. Foveal reattachment after macula-off retinal detachment occurs faster after vitrectomy than after buckle surgery. Ophthalmology 2004;111(7):1340–1343. 3. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after vitrectomy surgery for retinal detachment. Ophthalmology 2006;113(7):1179–1183. 4. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after scleral buckle surgery for retinal detachment. Ophthalmology 2007;114(1):108–112. 5. Seo JH, Woo SJ, Park KH, et al. Influence of persistent submacular fluid on visual outcome after successful scleral buckle surgery for macula-off retinal detachment. Am J Ophthalmol 2008;145(5):915–922. 6. Kim YK, Woo SJ, Kim KE, Park KH. Subretinal fluid resorption. Ophthalmology 2010;117(8):1655.e1–1655.e4. 7. Kim YK, Woo SJ, Park KH, et al. Comparison of persistent submacular fluid in vitrectomy and scleral buckle surgery for macula-involving retinal detachment. Am J Ophthalmol 2010;149(4):623–629.e1. 8. Machemer R. Experimental retinal detachment in the owl monkey. IV. The reattached retina. Am J Ophthalmol 1968; 66(6):1075–1091. 9. Leaver PK, Chester GH, Saunders SH. Factors influencing absorption of subretinal fluid. Br J Ophthalmol 1976;60(8):557–560. 10. Kaga T, Fonseca RA, Dantas MA, et al. Optical coherence tomography of bleb-like subretinal lesions after retinal reattachment surgery. Am J Ophthalmol 2001;132(1):120–121. 11. Kang SW, Kim JH, Shin WJ, Kim JI. Subretinal fluid bleb after successful scleral buckling and cryotherapy for retinal detachment. Am J Ophthalmol 2008;146(2):205–210.
VOL. 157, NO. 4
12. van Meurs JC, Humalda D, Mertens DA, Peperkamp E. Retinal folds through the macula. Doc Ophthalmol 1991; 78(3-4):335–340. 13. Larrison WI, Frederick AR Jr, Peterson TJ, Topping TM. Posterior retinal folds following vitreoretinal surgery. Arch Ophthalmol 1993;111(5):621–625. 14. Wong R. Longitudinal study of macular folds by spectraldomain optical coherence tomography. Am J Ophthalmol 2012;153(1):88–92.e1. 15. dell’Omo R, Tan HS, Schlingemann RO, et al. Evolution of outer retinal folds occurring after vitrectomy for retinal detachment repair. Invest Ophthalmol Vis Sci 2012;53(13):7928–7935. 16. Marmor MF. Mechanisms of normal retinal adhesion. In: Hinton DF, Ryan SJ, eds. Retina. 5th ed. Los Angeles, CA: Elsevier; 2013:447–464. 17. Gingell D, Fornes JA. Demonstration of intermolecular forces in cell adhesion using a new electrochemical technique. Nature 1975;256(5514):210–211. 18. Hagimura N, Suto K, Iida T, Kishi S. Optical coherence tomography of the neurosensory retina in rhegmatogenous retinal detachment. Am J Ophthalmol 2000;129(2):186–190. 19. Lee SY, Joe SG, Kim JG, et al. Optical coherence tomography evaluation of detached macula from rhegmatogenous retinal detachment and central serous chorioretinopathy. Am J Ophthalmol 2008;145(6):1071–1076. 20. Nakanishi H, Hangai M, Unoki N, et al. Spectral-domain optical coherence tomography imaging of the detached macula in rhegmatogenous retinal detachment. Retina 2009;29(2):232–242. 21. Fisher SK, Erickson PA, Lewis GP, Anderson DH. Intraretinal proliferation induced by retinal detachment. Invest Ophthalmol Vis Sci 1991;32(6):1739–1748. 22. Sethi CS, Lewis GP, Fisher SK, et al. Glial remodeling and neural plasticity in human retinal detachment with proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 2005; 46(1):329–342.
SUBRETINAL FLUID BLEB AFTER RETINAL DETACHMENT SURGERY
841