LONGITUDINAL OPTICAL DENSITY ANALYSIS OF SUBRETINAL FLUID AFTER SURGICAL REPAIR OF RHEGMATOGENOUS RETINAL DETACHMENT AMIR H. KASHANI, MD, PHD,* ALBERT Y. CHEUNG, MD,† JOSHUA ROBINSON, MD,†‡ GEORGE A. WILLIAMS, MD†‡ Purpose: To investigate optical coherence tomography–derived reflectivity and optical density (OD) characteristics of persistent subretinal fluid (SRF) in eyes after surgical repair of macula-off rhegmatogenous retinal detachment. Methods: Retrospective case series of nine eyes with macula-off rhegmatogenous retinal detachment that underwent surgical repair with either scleral buckling or vitrectomy with or without scleral buckling. Major inclusion criteria included 1) availability of high-quality optical coherence tomography scans at 2 or more time points, and 2) sufficient SRF for optical coherence tomography sampling without including tissue edges. Demographic, clinical, and optical coherence tomography imaging data were collected on all eyes. Optical density and SRF height measurements were obtained using a manual image segmentation method with ImageJ. Optical density measurements were standardized by conversion to optical density ratios to facilitate comparison between different visits and eyes. Correlations were assessed for significance through both univariate and multivariate regression analyses. Results: Optical density ratio measurements increased with time after surgery, and this was statistically significant (P = 0.001, R2 = 0.331). Subretinal fluid height measurements decreased in all eyes. There was a significant correlation between optical density ratios and log of SRF height (P # 0.001, R2 = 0.485). In multivariate analysis, neither optical density ratios nor SRF height was a statistically significant predictor of visual acuity. Conclusion: Changes in optical density ratios of the residual SRF after retinal detachment repair may be representative of changes in the SRF composition over time. This is in agreement with previous biochemical studies and may serve as a noninvasive method of assessing SRF content in vivo. RETINA 35:149–156, 2015
R
a full-thickness retinal defect. The source of this subretinal fluid (SRF) is thought to be multifactorial and may include the vitreous,1,2 serum,3,4 and retinal debris.5,6 Biochemical studies of RRD have demonstrated that SRF contains at least in part proteins,7 photoreceptor outer segments,8 lipids,9 cells (primarily inflammatory cells),10 glucose/carbohydrate compounds,11 and various inflammatory mediators.12,13 Although results from early studies associating RRD duration and SRF protein concentration were conflicting,7,11,14–16 it seems that total protein concentration correlates with RRD duration.11,17 Unfortunately, collection of SRF specimens for biochemical analysis is technically
hegmatogenous retinal detachment (RRD) is defined by the accumulation of fluid in the potential space between the neurosensory retina and the retinal pigment epithelium (RPE) in the presence of
From the *Department of Ophthalmology, University of Southern California, Los Angeles, California; †Department of Ophthalmology, Oakland University William Beaumont School of Medicine, Royal Oak, Michigan; and ‡Department of Vitreoretinal Surgery, Oakland University William Beaumont School of Medicine, Associated Retinal Consultants PC, Royal Oak, Michigan. None of the authors have any financial/conflicting interests to disclose. Reprint requests: George A. Williams, MD, Department of Ophthalmology, Oakland University School of Medicine, William Beaumont Hospital, 3535 W. 13 Mile Road, Ste 344, Royal Oak, MI 48073; e-mail: [email protected]
149
150
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
challenging, and analyses based on these techniques may be influenced by a number of confounding factors such as blood contamination, cryotherapy increasing protein concentrations, and sample amounts too small to compare with controls. Noninvasive methods of measuring fluid composition in vivo may provide greater reliability, safety, and longitudinal follow-up of residual SRF. Optical coherence tomography (OCT) provides a noninvasive, readily available, and high resolution means of detecting and potentially evaluating SRF. Optical coherence tomography findings have been demonstrated to be more sensitive than clinical findings by ophthalmoscopy.18 Persistent SRF has been described by OCT after RRD repair, although it may not be evident on clinical ophthalmoscopy examination or by fluorescein angiography.19–22 Studies have noted an incidence of persistent SRF 4 weeks to 6 weeks after scleral buckling (SB) surgery for macula-off RRD ranging from 27% to 94%.19–22 Ricker et al21 demonstrated a progressive resolution of persistent subfoveal fluid in his cohort by OCT with 94%, 72%, 55%, 24%, and 17% of cases showing SRF at 1, 3, 6, 9, and 12 months after SB surgery, respectively. Persistent SRF is thought to slow visual recovery and may take up to 1 year to 2 years to resolve.20 It may be associated with decreased visual acuity and alterations in color vision perception, contrast sensitivity, depth perception, and metamorphopsia. Moreover, visual improvement has been described on its resolution.20–24 Recently, it has been suggested that the use of quantifiable OCT image characteristics may serve as a proxy for the biochemical composition of the fluid and aid in the pathophysiological differentiation of disease processes leading to SRF accumulation.25,26 To date, no one has reported on the change in optical density profiles of rhegmatogenous subretinal fluid features over time. In the current series, spectral domain OCT is used to measure the optical density of residual SRF in eyes after surgical repair of macula-off RRD. We hypothesize that the optical density characteristics of the persistent SRF may help us predict the nature and the clinical behavior of the fluid.
Methods Patient Selection and Data Collection This study was conducted according to a protocol approved by the Western Institutional Review Board. The protocol and methods used also complied with the standards set forth by the Declaration of Helsinki. The
study cohort consists of a retrospective series of eyes treated for primary RRD at a single retina subspecialty practice between October 2011 and November 2013. Inclusion criteria were 1) presence of macula-off RRD repaired with SB and/or vitrectomy procedure, 2) high-quality OCT scans at 2 or more time points through the fovea, and 3) sufficient SRF for OCT sampling without including tissue edges. Patient baseline characteristics and clinical features are listed in Table 1. Some eyes had baseline OCTs before RRD repair, but this was not an absolute inclusion criteria because it was not always possible to reliably obtain high-quality OCTs preoperatively. All OCTs were performed on a Heidelberg Spectralis device (Heidelberg Engineering, Carlsbad, CA). High-resolution singleline scans (averaged 100 frames) were used for reflectivity, optical density, and SRF height measurements. Exclusion criteria included any media opacities that substantially affected OCT quality and poor OCT quality precluding image analysis for any other reason. Demographic, clinical, and imaging data were collected on all patients including age, gender, preoperative Snellen best-corrected visual acuity, bestcorrected visual acuity at each visit for which OCT data were available, duration of preoperative symptoms, duration of residual SRF postoperatively, clinical characteristics of the retinal detachment, type of surgical repair, and basic elements of a complete ophthalmic examination. Optical Coherence Tomography Imaging and Optical Density Ratio Measurements Heidelberg Spectralis (Heidelberg Engineering) OCT images were used for analysis in all cases. Data were exported from the OCT acquisition software as gray-scale TIFF images with brightness level set at “9” to standardize background vitreous enhancement for all eyes. The images were analyzed using ImageJ (National Institutes of Health; http://rsbweb.nih.gov/ij/). The OCT images were cropped and rotated uniformly. Using the region of interest (ROI) manager tool, multiple areas from different locations of the OCT image were selected, and the average gray value within each selection was measured (range between 0 corresponding to pure black and 255 corresponding to pure white). An average value was calculated as the sum of all the pixels in a selection divided by the number of pixels. This average gray value was defined as the absolute reflectivity value of the ROI. Subretinal fluid measurements in this study specifically examined subfoveal fluid pockets. Similar to previous studies,26,27 the SRF compartment reflectivity was expressed as an optical density ratio (ODR),
151
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL Table 1. Summary of the Patient Baseline Characteristics and Clinical Features Eye
Age, Years/ Gender Eye
Detachment Duration, days*
VA Baseline
VA Last Follow-up
PVD† −/−
1
44/M
OD
34
20/150
20/25
2
41/F
OS
374
20/400
20/80
3
77/M
OD
11
CF
20/60
4
33/F
OS
7
20/30
20/20
5
14/M
OD
50
20/80
20/60
6
63/F
OS
22
20/150
20/100
7
59/M
OS
8
20/40
20/40
8
56/F
OS
3
20/30
20/30
+/+
9
61/F
OD
17
20/60
20/200
+/+
Lens Status‡
Phakic/ phakic −/− Phakic/ phakic +/+ Phakic/ phakic −/− Phakic/ phakic −/+ Phakic/ phakic +/+ Phakic/ phakic +/NA IOL/IOL Phakic/ phakic Phakic/ phakic
Fovea On/Off§
Surgery¶
Off
SB/cryo/drain/none
Off
SB/cryo/drain/none
Off Off
SB/cryo/drain/0.5 mL air SB/cryo/drain/none
Off
SB/cryo/drain/none
Off
SB/cryo/no drain/ none PPV/cryo/drain/ retinotomy/SF6 SB/cryo/drain/none
Off On Off
SB, PPV/no cryo/ drain/SF6
*Length of time from onset of symptoms until date of surgery. †PVD status at surgery and at last follow-up. ‡Lens status at surgery and last follow-up. §Based on clinical examination. ¶Details of surgery: whether SB, PPV, or both/whether cryotherapy used/whether subretinal fluid drained/whether gas or air exchanged. CF, counting fingers; cryo, cryotherapy; F, female; HM, hand motion; IOL, intraocular lens; IOP, intraocular pressure; M, male; NA, not available; PPV, pars plana vitrectomy; PVD, posterior vitreous detachment; SF6, sulfur hexafluoride; VA, visual acuity.
which is defined as the ratio of the absolute reflectivity value of the SRF versus the absolute reflectivity value of an equivalent congruous area of overlying vitreous. Vitreous was used as a reference to help control for nonspecific confounding effects of OCT signal strength, media opacities, and other factors. The vitreous ROI was selected such that it was in the same anterior–posterior axis as the SRF to minimize errors associated with inhomogeneities of the various ocular structures (corneal opacifications, cataract, vitreous floaters, or other causes of inhomogeneous signal intensity at the retinal level). Additionally, the vitreous ROI was selected so that it did not include the detached posterior hyaloidal surface or any other visible structure within the vitreous cavity. In cases where the exact shape could not be selected in the vitreous, a region of equivalent vitreous area was selected on the same anterior–posterior axis. In cases where the SRF ROI was small, multiple corresponding vitreous equivalents along the same anterior– posterior axis were measured and averaged for comparison. For each pocket of SRF, the full-fluid area was selected to measure an accurate average ODR of all the fluid present. Multiple methods of SRF demarcation were attempted to evaluate reproducibility and reliability of results. For example, measurements of the entire SRF area excluding
associated-cellular debris and of the entire SRF area plus cellular debris were made separately. Cellular debris included hyperreflective opacities contiguous with the outer retina. Additionally, a segment of fullthickness retina, a band of RPE, and their vitreous volume equivalents were measured as controls. The segment of full-thickness retina was measured and compared to determine relative consistency of reflectivity measurements between OCT images from different dates. Given that the RPE is often one of the brightest signals on OCT, the band of RPE of each image was measured to obtain the maximum gray value that should be present on each image. The following ODRs were calculated: 1. ODRSRF = Reflectivity (SRF)/Reflectivity (Vitreous EquivalentSRF) 2. ODRRetina = Reflectivity (Full Retina)/Reflectivity (Vitreous EquivalentRetina) 3. ODRRPE = Reflectivity (RPE)/Reflectivity (Vitreous EquivalentRPE). Figure 1 shows representative images from an eye with sample measurements and annotations. The subfoveal fluid height (later referenced as SRF height in this article) was also measured for each OCT (Figure 2). The location of the foveal center was defined as the midway point of a straight line
152
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
Fig. 1. Optical coherence tomography of a representative eye (Eye 1) with residual SRF pocket after SB repair of retinal detachment. A. Raw, highresolution, spectral domain OCT image through the fovea demonstrates residual SRF at postoperative Month 2. B. Annotated image of the same OCT section showing measurements of each ROI made from SRF (white asterisks), RPE (black asterisk), and retina (black arrow). Equivalent vitreous ROIs were chosen on the same anterior–posterior axis. C. Optical coherence tomography through the fovea demonstrating 2 subsequent areas (central and temporal) of SRF that persisted from the larger original pocket of SRF after 6.5 months after the baseline image shown in panel A. D. Temporal change in the absolute reflectivity values from the ROI areas illustrated in panel B. E. Temporal change in ODR of the SRF.
connecting the perifoveal terminations of the inner nuclear layers (Figure 2, Line 1). The location of the central foveal photoreceptors was found by drawing a straight line from the midpoint and orthogonal to Line 1 toward the outer retinal surface (Figure 2, Line 2). The height of the subfoveal fluid was defined as the distance between the outer termination of Line 2 and the surface of the RPE along a straight line perpendicular to the surface of the RPE (Figure 2, Line 3). Measurements in microns
Fig. 2. Measuring the SRF height. The location of the foveal center was defined as the midway point of a straight line connecting the perifoveal terminations of the inner nuclear layers (Line 1). The location of the central foveal photoreceptors was found by drawing a straight line from the midpoint and orthogonal to the Line 1 toward the outer retinal surface (Line 2). The height of the subfoveal fluid was defined as the distance between the outer termination of Line 2 and the surface of the RPE along a straight line perpendicular to the surface of the RPE (Line 3).
were obtained using the caliper software built into the Heidelberg system. Statistical Analysis Statistical analysis was performed using Excel (Microsoft, Redmond, WA). Correlations between logMAR visual acuity, SRF height, and ODR were assessed for significance through both univariate and multivariate regression analyses through calculation of Pearson correlation coefficients. Statistical significance
Fig. 3. Graphs of ODRs over time for all eyes in the study cohort. There was a general increasing trend (dotted line) in ODR values over time for all eyes when measuring the ODR of the SRF.
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL
was defined as a P , 0.05. Acuities were converted to logMAR using the following formula: logMAR ¼ −log ðSnellen decimal acuityÞ:
153
The inclusion or exclusion of these outer segments in the measured area did not make a qualitative difference in the ODR trends that we described. As the measurement of these hyperreflective cell debris opacities could be quite subjective and not reliably reproducible, they were not included.
Results Discussion A total of 9 eyes had OCT images that were of adequate image quality and sufficient residual SRF volume for measurements. Baseline patient data are presented in Table 1. For each eye, calculated ODRs, SRF heights, and visual acuities are shown in Table 2. Eight (88.9%) eyes continued to have SRF at the last follow-up, although all eyes demonstrated decreasing subfoveal SRF heights over time. One eye had complete resolution of SRF over 1 month and was not included in subsequent analyses. Compared with preoperative visual acuity, 8 (89%) eyes had the same or better visual acuity at last follow-up. In 5 eyes, an initially contiguous large posterior SRF pocket separated into distinct pockets of SRF (e.g., see Figure 1C). In cases where multiple SRF pockets were present in high-resolution OCT, the SRF pockets were treated as one aggregate value for subsequent analyses. The assumption was made that the SRF in these multiple SRF pockets would have the same ODR/composition as they originated from the same subfoveal SRF pocket. In general, ODR measurements were found to increase with time after surgery (see Figure 3), and this relationship was found to be statistically significant (P = 0.001, R2 = 0.331). This showed that overall there was a 0.716 increase in the ODR per month. There was also a significant inverse correlation between ODR and log of SRF height (P # 0.001, R2 = 0.485). In multivariate analysis, neither ODR nor SRF height was found to be a statistically significant predictor of logMAR visual acuity. No difference was observed in subgroup analyses comparing eyes that underwent vitrectomy versus SB. In most cases, the SRF demonstrated grossly visible discrete areas of increased hyperreflectivity (Figure 1). In addition, in most cases, the outer nuclear layer seemed to have hyperreflective opacities contiguous with the neurosensory retina projecting into the SRF pocket. This appearance has been described previously and is consistent with increasing length of photoreceptor outer segments.28 To accurately assess the ODR of the SRF pockets, we performed two sets of measurements on these images. In one sample, we included the elongated segments in the measured area, and in another sample, we excluded the elongated segments.
Our study demonstrates that reflectivity measurements and ODR calculations of residual SRF can be evaluated over time to assess the optical properties of SRF after repair of RRDs. The reflectivity measurements demonstrate an increasing trend over time, suggesting a change in the composition of the SRF. This trend is present when reflectivity measurements are expressed as ODR to account for potential confounding factors in the reflectivity measurements. We interpret these findings to be consistent with earlier observations that SRF protein content and lipid content increase over time.11,17 Previous studies have demonstrated the reliability of OCT to measure the mean reflectivity of the SRF in various pathologies including RRD.25–27 The reflectivity parameters of both the retina and SRF have been correlated with the type of pathology causing the fluid accumulation.27 Although reflectivity measurement may not reveal the exact composition of the SRF, higher reflectivity signals likely signify higher density of macroscopic particles in the measured fluid and lower reflectivity signals indicate a relatively dilute fluid composition.29,30 Therefore, OCT-measured reflectivity and/or ODRs may serve as a noninvasive and quantitative measure of SRF composition. Such measurements may have diagnostic and prognostic utility in the clinical setting. The overall increasing trend in SRF reflectivity and ODR likely represents a change in the biochemical composition of the SRF over time. A large body of evidence from previous investigations suggests that this change in composition is due to accumulation of proteins,7 photoreceptor outer segments,8 lipids,9 cells (mainly inflammatory cells),10 glucose/carbohydrate compounds,11 vitreous,1,2 serum,3,4 retinal debris,5,6 and various factors involved in the inflammation process.13 Veckeneer et al8 found that cell-rich SRF samples correlated with SRF seen on OCT. Using immunohistochemistry and electron microscopy, they found that photoreceptor outer segment fragments, in particular, contributed to this increased cellularity. Veckeneer et al8 hypothesized that a certain SRF composition along with RPE dysfunction may predispose to fluid persistence in long-standing RDs; they also
Eye 1 Eye 2
Post-op Month 3
—
0.678
— 20/150 0.594
164 mm 20/30
Eye 3 Eye 4 Eye 5 Eye 6
177 mm 20/400 — — 20/1,000 0.568 368 mm 20/30 0.344 7 mm 20/80 —
— 20/150 −2 Eye 0.813 7 170 mm 20/40 Eye — 8 — 20/30 NC Eye — 9 — 20/60
Post-op Month 1
Post-op Month 4
Post-op Month 5
Post-op Month 6
Post-op Month 7
Post-op Month 8
Post-op Month 9
0.318
0.652
0.821
4.901
138 mm 20/25 −1
96 mm 20/30 +2
80 mm 20/25 +2
29 mm 20/25 −1
Post-op Month 10
3.854
133 mm 20/80
65 mm 20/80 7.124
0.986
0.907
144 mm 20/60 −2
147 mm 20/50 −2
109 mm 20/60 9.547
9.996
10 mm 20/25 +1
12 mm 20/20 −2
23 mm 20/60
NA† 0 mm 20/60 +1 0.950
9.885
121 mm 80 mm 20/150 −1 20/100 −2
12 mm 20/100
4.717 15 mm 20/40 1.779
1.585
1.811
1.203
77 mm 20/50 0.816
19 mm 20/40
22 mm 20/30-1
0.566
54 mm 20/40 0.657
281 mm 20/60
300 mm 20/200
320 mm 20/80
Post-op Month 12
0.640
1.206
0.985
Post-op Month 11
0.763
0.801
261 mm 20/80
171 mm 20/200
Values are listed in that order vertically. Of note, some patients only had acuities measured at preoperative visit. *SRF height was subfoveal through measurement described in Figure 2. †All SRF resolved. NA, not available; NC, near card; post-op, postoperatively; pre-op, preoperatively.
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
Post-op Month 2
Pre-op
154
Table 2. Summary of the Raw Values for the ODR, SRF Height (in Micrometers),* and Visual Acuity Over Time
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL
suggested that drainage would decrease the cellularity and viscosity, leading to reduction in the incidence of persistent SRF. Our results show that increasing ODR may reflect the increased concentration of photoreceptor outer segment fragments in the SRF, as the serous component is absorbed, where the serous component is absorbed faster than the cellular component, thereby increasing ODR. At this point, it is not possible to differentiate between different SRF etiologies from OCT measurements, but our results suggest that this may be possible given a large sample of patients and varying pathological processes. Specifically, our study is the first to suggest that longitudinal measurements of OCT reflectivity and ODR may provide useful information about the nature and course of SRF after retinal detachment repair. Neither ODR nor SRF height was found to be significant predictors of measured visual acuity in this cohort. Importantly, nearly all of the eyes in this cohort were phakic at last follow-up visit, and it is likely that visual acuity results were confounded by cataract progression. This trend would be better assessed in a larger series of eyes, preferably pseudophakic at presentation. There are a number of limitations to this study. One such limitation is the protean nature of the vitreous cavity. It is likely that the presence or absence of a posterior vitreous detachment will affect the reflectivity measurements from the vitreous cavity. This variability may conceal small or modest changes in the ODR. For example, in Figure 1D, the reflectivity of the vitreous fluctuates nearly 10-fold but seems to be unchanged after 10 months compared with the constantly increasing reflectivity of the SRF pockets. This variability results in a relatively flat ODR in Figure 1E over the first 5 months to 6 months. This fluctuation is in contrast to the stable reflectivity measurements from the RPE and full-thickness retina, which are anatomically distinct and easily identifiable for measurements. Close inspection of subsequent OCTs (Figure 1C) demonstrates partial posterior vitreous detachment with vitreomacular adhesion at the fovea. We suspect that the posterior boundary of the vitreous is not always clearly visible on OCT in this case, although the OCT was of adequate quality for SRF measurements. Therefore, the large variability in the reflectivity measurements from the vitreous may reflect sampling from the vitreous inside or outside the posterior hyaloidal boundary in the case of partial posterior vitreous detachment. Despite this variability, it is still apparent that the reflectivity and ODR of SRF compartments increase with time. Of note, the preoperative ODR was less than one in all eyes. This does not necessarily imply that the SRF
155
is optically less dense than the overlying vitreous as the intervening retina is of relatively high reflectivity, which would serve to artificially decrease the absolute reflectivity of the SRF, thereby decreasing the ODR. As we chose to investigate temporal changes in the ODR, this confounding factor is minimized. This is supported by our observation that the absolute reflectivity of the retina remained stable over time in all eyes. It is important to note that the phenomenon of persistent SRF after retinal detachment repair is not unique to SB and is rather more simply a reflection of the completeness of SRF drainage at the time of surgery. Two eyes from this study were treated with vitrectomy and were noted to have persistent SRF. These eyes were found to display similar SRF patterns (increasing/stable ODR and decreasing SRF height) to the rest of the cohort, which underwent primary SB. Optical density ratio measurements may be confounded by numerous optical artifacts including, but not limited to, media opacity. In addition, differences in the tear film and other factors may affect the overall signal strength of the image on any given sitting. Therefore, we performed a number of analyses to control for possible variations in measurements both within an imaging session and between two different imaging sessions on the same eye. These analyses are described in the Methods section in detail. In general, the additional analyses normalized the signal intensity of the measured area of SRF to one of a number of “standards.” These standards included the vitreous cavity overlying the area of SRF (low-intensity control), the RPE under the area of SRF (high-intensity control), and the full thickness of the neurosensory retina overlying the area of SRF. All of these analyses confirmed the general trend of increasing SRF ODR with time. In conclusion, residual subfoveal fluid after surgical repair of macula-off RRD is not uncommon and may have implications in terms of clinical course and visual prognosis. Standard OCT imaging may be used to characterize this SRF quantitatively through measurement of absolute reflectivity and calculation of ODR. Changes in the ODR of residual SRF after RD repair can be easily monitored and may be representative of changes in the composition of the SRF over time. With additional study, this parameter may prove a reliable noninvasive method of characterizing SRF content in vivo. Key words: optical coherence tomography, optical density, retina, rhegmatogenous retinal detachment, subretinal fluid.
156
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
References 1. Cooper WC, Halbert SP, Manski WJ. Immunochemical analysis of vitreous and subretinal fluid. Invest Ophthalmol 1963;2: 369–377. 2. Lister W. Holes in the retina and their clinical significance. Br J Ophthalmol 1924;8:i4–20. 3. Dorello U. Electrophoretic studies of the protein content of the subretinal fluid in idiopathic retinal detachment. Am J Ophthalmol 1956;41:564. 4. Smith JL, Douty E. Electrophoresis of subretinal fluid. Arch Ophthalmol 1960;64:114–119. 5. Magitot A. The subretinal fluid in idiopathic detachment of the retina. Arch Ophthalmol 1934;11:159–173. 6. Hara S, Ishiguro S, Hayasaka S, Mizuno K. Immunoreactive opsin content in subretinal fluid from patients with rhegmatogenous retinal detachments. Arch Ophthalmol 1987;105:260–263. 7. Heath H, Beck TC, Foulds WS. Chemical composition of subretinal fluid. Br J Ophthalmol 1962;46:385–396. 8. Veckeneer M, Derycke L, Lindstedt EW, et al. Persistent subretinal fluid after surgery for rhegmatogenous retinal detachment: hypothesis and review. Graefes Arch Clin Exp Ophthalmol 2012;250:795–802. 9. Lam KW, van Heuven WA, Ray GS, Feman S. Subretinal fluids: lipid analyses. Invest Ophthalmol 1975;14:406–410. 10. Feeney L, Burns RP, Mixon RM. Human subretinal fluid. Its cellular and subcellular components. Arch Ophthalmol 1975; 93:62–69. 11. Quintyn JC, Brasseur G. Subretinal fluid in primary rhegmatogenous retinal detachment: physiopathology and composition. Surv Ophthalmol 2004;49:96–108. 12. Williams GA, Reeser F, O’Brien WJ, Fleischman JA. Prostacyclin and thromboxane A2 derivatives in rhegmatogenous subretinal fluid. Arch Ophthalmol 1983;101:463–464. 13. Chen YH, Chen JT, Chien MW, et al. Subretinal fluid from rhegmatogenous retinal detachment and blood induces the expression of ICAM-1 in the human retinal pigment epithelium (ARPE-19) in vitro. Eye (Lond) 2010;24:354–360. 14. Akhmeteli LM, Kasvina BS, Petropavlovskaja GA. Biochemical investigation of the subretinal fluid. Br J Ophthalmol 1975; 59:70–77. 15. Schenk H, Formanek K, Förster O. Untersuchungen der Eiweißkörper in der subretinalen Flüssigkeit bei Netzhautablösung. Graefes Arch Ophthalmologie 1961;164:29–41. 16. Sachsenweger R, Gassler H, Hentsch R. Elektrophoretische Untersuchungen der subretinalen Flüssigkeit bei Netzhautablösungen. Graefes Arch Ophthalmologie 1964;166:432–439.
17. Berrod JP, Kayl P, Rozot P, Raspiller A. Proteins in the subretinal fluid. Eur J Ophthalmol 1993;3:132–137. 18. Brown JC, Solomon SD, Bressler SB, et al. Detection of diabetic foveal edema: contact lens biomicroscopy compared with optical coherence tomography. Arch Ophthalmol 2004;122:330–335. 19. Wang XY, Shen LP, Hu RR, Xu W. Persistent subretinal fluid after successful scleral buckle surgery for macula-off retinal detachment. Chin Med J (Engl) 2011;124:4007–4011. 20. 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;145:915–922. 21. Ricker LJ, Noordzij LJ, Goezinne F, et al. Persistent subfoveal fluid and increased preoperative foveal thickness impair visual outcome after macula-off retinal detachment repair. Retina 2011;31:1505–1512. 22. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after scleral buckle surgery for retinal detachment. Ophthalmology 2007;114:108–112. 23. Wolfensberger TJ, Gonvers M. Optical coherence tomography in the evaluation of incomplete visual acuity recovery after macula-off retinal detachments. Graefes Arch Clin Exp Ophthalmol 2002;240:85–89. 24. Delolme MP, et al. Anatomical and functional macular changes after rhegmatogenous retinal detachment with macula off. Am J Ophthalmol 2012;153:128–136. 25. Barthelmes D, Sutter FK, Gillies MC. Differential optical densities of intraretinal spaces. Invest Ophthalmol Vis Sci 2008; 49:3529–3534. 26. Ahlers C, Golbaz I, Einwallner E, et al. Identification of optical density ratios in subretinal fluid as a clinically relevant biomarker in exudative macular disease. Invest Ophthalmol Vis Sci 2009;50:3417–3424. 27. Neudorfer M, Weinberg A, Loewenstein A, Barak A. Differential optical density of subretinal spaces. Invest Ophthalmol Vis Sci 2012;53:3104–3110. 28. Guérin CJ, Lewis GP, Fisher SK, Anderson DH. Recovery of photoreceptor outer segment length and analysis of membrane assembly rates in regenerating primate photoreceptor outer segments. Invest Ophthalmol Vis Sci 1993;34:175–183. 29. Dubois A, Vabre L, Boccara AC, Beaurepaire E. Highresolution full-field optical coherence tomography with a Linnik microscope. Appl Opt 2002;41:805–812. 30. Nakano. Studies on the subretinal fluid. Report I. Refractometric and paper-electrophoretic study on the subretinal fluid of spontaneous retinal detachment. Jpn J Ophthalmol 1961;3: 23–28.
R
a full-thickness retinal defect. The source of this subretinal fluid (SRF) is thought to be multifactorial and may include the vitreous,1,2 serum,3,4 and retinal debris.5,6 Biochemical studies of RRD have demonstrated that SRF contains at least in part proteins,7 photoreceptor outer segments,8 lipids,9 cells (primarily inflammatory cells),10 glucose/carbohydrate compounds,11 and various inflammatory mediators.12,13 Although results from early studies associating RRD duration and SRF protein concentration were conflicting,7,11,14–16 it seems that total protein concentration correlates with RRD duration.11,17 Unfortunately, collection of SRF specimens for biochemical analysis is technically
hegmatogenous retinal detachment (RRD) is defined by the accumulation of fluid in the potential space between the neurosensory retina and the retinal pigment epithelium (RPE) in the presence of
From the *Department of Ophthalmology, University of Southern California, Los Angeles, California; †Department of Ophthalmology, Oakland University William Beaumont School of Medicine, Royal Oak, Michigan; and ‡Department of Vitreoretinal Surgery, Oakland University William Beaumont School of Medicine, Associated Retinal Consultants PC, Royal Oak, Michigan. None of the authors have any financial/conflicting interests to disclose. Reprint requests: George A. Williams, MD, Department of Ophthalmology, Oakland University School of Medicine, William Beaumont Hospital, 3535 W. 13 Mile Road, Ste 344, Royal Oak, MI 48073; e-mail: [email protected]
149
150
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
challenging, and analyses based on these techniques may be influenced by a number of confounding factors such as blood contamination, cryotherapy increasing protein concentrations, and sample amounts too small to compare with controls. Noninvasive methods of measuring fluid composition in vivo may provide greater reliability, safety, and longitudinal follow-up of residual SRF. Optical coherence tomography (OCT) provides a noninvasive, readily available, and high resolution means of detecting and potentially evaluating SRF. Optical coherence tomography findings have been demonstrated to be more sensitive than clinical findings by ophthalmoscopy.18 Persistent SRF has been described by OCT after RRD repair, although it may not be evident on clinical ophthalmoscopy examination or by fluorescein angiography.19–22 Studies have noted an incidence of persistent SRF 4 weeks to 6 weeks after scleral buckling (SB) surgery for macula-off RRD ranging from 27% to 94%.19–22 Ricker et al21 demonstrated a progressive resolution of persistent subfoveal fluid in his cohort by OCT with 94%, 72%, 55%, 24%, and 17% of cases showing SRF at 1, 3, 6, 9, and 12 months after SB surgery, respectively. Persistent SRF is thought to slow visual recovery and may take up to 1 year to 2 years to resolve.20 It may be associated with decreased visual acuity and alterations in color vision perception, contrast sensitivity, depth perception, and metamorphopsia. Moreover, visual improvement has been described on its resolution.20–24 Recently, it has been suggested that the use of quantifiable OCT image characteristics may serve as a proxy for the biochemical composition of the fluid and aid in the pathophysiological differentiation of disease processes leading to SRF accumulation.25,26 To date, no one has reported on the change in optical density profiles of rhegmatogenous subretinal fluid features over time. In the current series, spectral domain OCT is used to measure the optical density of residual SRF in eyes after surgical repair of macula-off RRD. We hypothesize that the optical density characteristics of the persistent SRF may help us predict the nature and the clinical behavior of the fluid.
Methods Patient Selection and Data Collection This study was conducted according to a protocol approved by the Western Institutional Review Board. The protocol and methods used also complied with the standards set forth by the Declaration of Helsinki. The
study cohort consists of a retrospective series of eyes treated for primary RRD at a single retina subspecialty practice between October 2011 and November 2013. Inclusion criteria were 1) presence of macula-off RRD repaired with SB and/or vitrectomy procedure, 2) high-quality OCT scans at 2 or more time points through the fovea, and 3) sufficient SRF for OCT sampling without including tissue edges. Patient baseline characteristics and clinical features are listed in Table 1. Some eyes had baseline OCTs before RRD repair, but this was not an absolute inclusion criteria because it was not always possible to reliably obtain high-quality OCTs preoperatively. All OCTs were performed on a Heidelberg Spectralis device (Heidelberg Engineering, Carlsbad, CA). High-resolution singleline scans (averaged 100 frames) were used for reflectivity, optical density, and SRF height measurements. Exclusion criteria included any media opacities that substantially affected OCT quality and poor OCT quality precluding image analysis for any other reason. Demographic, clinical, and imaging data were collected on all patients including age, gender, preoperative Snellen best-corrected visual acuity, bestcorrected visual acuity at each visit for which OCT data were available, duration of preoperative symptoms, duration of residual SRF postoperatively, clinical characteristics of the retinal detachment, type of surgical repair, and basic elements of a complete ophthalmic examination. Optical Coherence Tomography Imaging and Optical Density Ratio Measurements Heidelberg Spectralis (Heidelberg Engineering) OCT images were used for analysis in all cases. Data were exported from the OCT acquisition software as gray-scale TIFF images with brightness level set at “9” to standardize background vitreous enhancement for all eyes. The images were analyzed using ImageJ (National Institutes of Health; http://rsbweb.nih.gov/ij/). The OCT images were cropped and rotated uniformly. Using the region of interest (ROI) manager tool, multiple areas from different locations of the OCT image were selected, and the average gray value within each selection was measured (range between 0 corresponding to pure black and 255 corresponding to pure white). An average value was calculated as the sum of all the pixels in a selection divided by the number of pixels. This average gray value was defined as the absolute reflectivity value of the ROI. Subretinal fluid measurements in this study specifically examined subfoveal fluid pockets. Similar to previous studies,26,27 the SRF compartment reflectivity was expressed as an optical density ratio (ODR),
151
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL Table 1. Summary of the Patient Baseline Characteristics and Clinical Features Eye
Age, Years/ Gender Eye
Detachment Duration, days*
VA Baseline
VA Last Follow-up
PVD† −/−
1
44/M
OD
34
20/150
20/25
2
41/F
OS
374
20/400
20/80
3
77/M
OD
11
CF
20/60
4
33/F
OS
7
20/30
20/20
5
14/M
OD
50
20/80
20/60
6
63/F
OS
22
20/150
20/100
7
59/M
OS
8
20/40
20/40
8
56/F
OS
3
20/30
20/30
+/+
9
61/F
OD
17
20/60
20/200
+/+
Lens Status‡
Phakic/ phakic −/− Phakic/ phakic +/+ Phakic/ phakic −/− Phakic/ phakic −/+ Phakic/ phakic +/+ Phakic/ phakic +/NA IOL/IOL Phakic/ phakic Phakic/ phakic
Fovea On/Off§
Surgery¶
Off
SB/cryo/drain/none
Off
SB/cryo/drain/none
Off Off
SB/cryo/drain/0.5 mL air SB/cryo/drain/none
Off
SB/cryo/drain/none
Off
SB/cryo/no drain/ none PPV/cryo/drain/ retinotomy/SF6 SB/cryo/drain/none
Off On Off
SB, PPV/no cryo/ drain/SF6
*Length of time from onset of symptoms until date of surgery. †PVD status at surgery and at last follow-up. ‡Lens status at surgery and last follow-up. §Based on clinical examination. ¶Details of surgery: whether SB, PPV, or both/whether cryotherapy used/whether subretinal fluid drained/whether gas or air exchanged. CF, counting fingers; cryo, cryotherapy; F, female; HM, hand motion; IOL, intraocular lens; IOP, intraocular pressure; M, male; NA, not available; PPV, pars plana vitrectomy; PVD, posterior vitreous detachment; SF6, sulfur hexafluoride; VA, visual acuity.
which is defined as the ratio of the absolute reflectivity value of the SRF versus the absolute reflectivity value of an equivalent congruous area of overlying vitreous. Vitreous was used as a reference to help control for nonspecific confounding effects of OCT signal strength, media opacities, and other factors. The vitreous ROI was selected such that it was in the same anterior–posterior axis as the SRF to minimize errors associated with inhomogeneities of the various ocular structures (corneal opacifications, cataract, vitreous floaters, or other causes of inhomogeneous signal intensity at the retinal level). Additionally, the vitreous ROI was selected so that it did not include the detached posterior hyaloidal surface or any other visible structure within the vitreous cavity. In cases where the exact shape could not be selected in the vitreous, a region of equivalent vitreous area was selected on the same anterior–posterior axis. In cases where the SRF ROI was small, multiple corresponding vitreous equivalents along the same anterior– posterior axis were measured and averaged for comparison. For each pocket of SRF, the full-fluid area was selected to measure an accurate average ODR of all the fluid present. Multiple methods of SRF demarcation were attempted to evaluate reproducibility and reliability of results. For example, measurements of the entire SRF area excluding
associated-cellular debris and of the entire SRF area plus cellular debris were made separately. Cellular debris included hyperreflective opacities contiguous with the outer retina. Additionally, a segment of fullthickness retina, a band of RPE, and their vitreous volume equivalents were measured as controls. The segment of full-thickness retina was measured and compared to determine relative consistency of reflectivity measurements between OCT images from different dates. Given that the RPE is often one of the brightest signals on OCT, the band of RPE of each image was measured to obtain the maximum gray value that should be present on each image. The following ODRs were calculated: 1. ODRSRF = Reflectivity (SRF)/Reflectivity (Vitreous EquivalentSRF) 2. ODRRetina = Reflectivity (Full Retina)/Reflectivity (Vitreous EquivalentRetina) 3. ODRRPE = Reflectivity (RPE)/Reflectivity (Vitreous EquivalentRPE). Figure 1 shows representative images from an eye with sample measurements and annotations. The subfoveal fluid height (later referenced as SRF height in this article) was also measured for each OCT (Figure 2). The location of the foveal center was defined as the midway point of a straight line
152
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
Fig. 1. Optical coherence tomography of a representative eye (Eye 1) with residual SRF pocket after SB repair of retinal detachment. A. Raw, highresolution, spectral domain OCT image through the fovea demonstrates residual SRF at postoperative Month 2. B. Annotated image of the same OCT section showing measurements of each ROI made from SRF (white asterisks), RPE (black asterisk), and retina (black arrow). Equivalent vitreous ROIs were chosen on the same anterior–posterior axis. C. Optical coherence tomography through the fovea demonstrating 2 subsequent areas (central and temporal) of SRF that persisted from the larger original pocket of SRF after 6.5 months after the baseline image shown in panel A. D. Temporal change in the absolute reflectivity values from the ROI areas illustrated in panel B. E. Temporal change in ODR of the SRF.
connecting the perifoveal terminations of the inner nuclear layers (Figure 2, Line 1). The location of the central foveal photoreceptors was found by drawing a straight line from the midpoint and orthogonal to Line 1 toward the outer retinal surface (Figure 2, Line 2). The height of the subfoveal fluid was defined as the distance between the outer termination of Line 2 and the surface of the RPE along a straight line perpendicular to the surface of the RPE (Figure 2, Line 3). Measurements in microns
Fig. 2. Measuring the SRF height. The location of the foveal center was defined as the midway point of a straight line connecting the perifoveal terminations of the inner nuclear layers (Line 1). The location of the central foveal photoreceptors was found by drawing a straight line from the midpoint and orthogonal to the Line 1 toward the outer retinal surface (Line 2). The height of the subfoveal fluid was defined as the distance between the outer termination of Line 2 and the surface of the RPE along a straight line perpendicular to the surface of the RPE (Line 3).
were obtained using the caliper software built into the Heidelberg system. Statistical Analysis Statistical analysis was performed using Excel (Microsoft, Redmond, WA). Correlations between logMAR visual acuity, SRF height, and ODR were assessed for significance through both univariate and multivariate regression analyses through calculation of Pearson correlation coefficients. Statistical significance
Fig. 3. Graphs of ODRs over time for all eyes in the study cohort. There was a general increasing trend (dotted line) in ODR values over time for all eyes when measuring the ODR of the SRF.
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL
was defined as a P , 0.05. Acuities were converted to logMAR using the following formula: logMAR ¼ −log ðSnellen decimal acuityÞ:
153
The inclusion or exclusion of these outer segments in the measured area did not make a qualitative difference in the ODR trends that we described. As the measurement of these hyperreflective cell debris opacities could be quite subjective and not reliably reproducible, they were not included.
Results Discussion A total of 9 eyes had OCT images that were of adequate image quality and sufficient residual SRF volume for measurements. Baseline patient data are presented in Table 1. For each eye, calculated ODRs, SRF heights, and visual acuities are shown in Table 2. Eight (88.9%) eyes continued to have SRF at the last follow-up, although all eyes demonstrated decreasing subfoveal SRF heights over time. One eye had complete resolution of SRF over 1 month and was not included in subsequent analyses. Compared with preoperative visual acuity, 8 (89%) eyes had the same or better visual acuity at last follow-up. In 5 eyes, an initially contiguous large posterior SRF pocket separated into distinct pockets of SRF (e.g., see Figure 1C). In cases where multiple SRF pockets were present in high-resolution OCT, the SRF pockets were treated as one aggregate value for subsequent analyses. The assumption was made that the SRF in these multiple SRF pockets would have the same ODR/composition as they originated from the same subfoveal SRF pocket. In general, ODR measurements were found to increase with time after surgery (see Figure 3), and this relationship was found to be statistically significant (P = 0.001, R2 = 0.331). This showed that overall there was a 0.716 increase in the ODR per month. There was also a significant inverse correlation between ODR and log of SRF height (P # 0.001, R2 = 0.485). In multivariate analysis, neither ODR nor SRF height was found to be a statistically significant predictor of logMAR visual acuity. No difference was observed in subgroup analyses comparing eyes that underwent vitrectomy versus SB. In most cases, the SRF demonstrated grossly visible discrete areas of increased hyperreflectivity (Figure 1). In addition, in most cases, the outer nuclear layer seemed to have hyperreflective opacities contiguous with the neurosensory retina projecting into the SRF pocket. This appearance has been described previously and is consistent with increasing length of photoreceptor outer segments.28 To accurately assess the ODR of the SRF pockets, we performed two sets of measurements on these images. In one sample, we included the elongated segments in the measured area, and in another sample, we excluded the elongated segments.
Our study demonstrates that reflectivity measurements and ODR calculations of residual SRF can be evaluated over time to assess the optical properties of SRF after repair of RRDs. The reflectivity measurements demonstrate an increasing trend over time, suggesting a change in the composition of the SRF. This trend is present when reflectivity measurements are expressed as ODR to account for potential confounding factors in the reflectivity measurements. We interpret these findings to be consistent with earlier observations that SRF protein content and lipid content increase over time.11,17 Previous studies have demonstrated the reliability of OCT to measure the mean reflectivity of the SRF in various pathologies including RRD.25–27 The reflectivity parameters of both the retina and SRF have been correlated with the type of pathology causing the fluid accumulation.27 Although reflectivity measurement may not reveal the exact composition of the SRF, higher reflectivity signals likely signify higher density of macroscopic particles in the measured fluid and lower reflectivity signals indicate a relatively dilute fluid composition.29,30 Therefore, OCT-measured reflectivity and/or ODRs may serve as a noninvasive and quantitative measure of SRF composition. Such measurements may have diagnostic and prognostic utility in the clinical setting. The overall increasing trend in SRF reflectivity and ODR likely represents a change in the biochemical composition of the SRF over time. A large body of evidence from previous investigations suggests that this change in composition is due to accumulation of proteins,7 photoreceptor outer segments,8 lipids,9 cells (mainly inflammatory cells),10 glucose/carbohydrate compounds,11 vitreous,1,2 serum,3,4 retinal debris,5,6 and various factors involved in the inflammation process.13 Veckeneer et al8 found that cell-rich SRF samples correlated with SRF seen on OCT. Using immunohistochemistry and electron microscopy, they found that photoreceptor outer segment fragments, in particular, contributed to this increased cellularity. Veckeneer et al8 hypothesized that a certain SRF composition along with RPE dysfunction may predispose to fluid persistence in long-standing RDs; they also
Eye 1 Eye 2
Post-op Month 3
—
0.678
— 20/150 0.594
164 mm 20/30
Eye 3 Eye 4 Eye 5 Eye 6
177 mm 20/400 — — 20/1,000 0.568 368 mm 20/30 0.344 7 mm 20/80 —
— 20/150 −2 Eye 0.813 7 170 mm 20/40 Eye — 8 — 20/30 NC Eye — 9 — 20/60
Post-op Month 1
Post-op Month 4
Post-op Month 5
Post-op Month 6
Post-op Month 7
Post-op Month 8
Post-op Month 9
0.318
0.652
0.821
4.901
138 mm 20/25 −1
96 mm 20/30 +2
80 mm 20/25 +2
29 mm 20/25 −1
Post-op Month 10
3.854
133 mm 20/80
65 mm 20/80 7.124
0.986
0.907
144 mm 20/60 −2
147 mm 20/50 −2
109 mm 20/60 9.547
9.996
10 mm 20/25 +1
12 mm 20/20 −2
23 mm 20/60
NA† 0 mm 20/60 +1 0.950
9.885
121 mm 80 mm 20/150 −1 20/100 −2
12 mm 20/100
4.717 15 mm 20/40 1.779
1.585
1.811
1.203
77 mm 20/50 0.816
19 mm 20/40
22 mm 20/30-1
0.566
54 mm 20/40 0.657
281 mm 20/60
300 mm 20/200
320 mm 20/80
Post-op Month 12
0.640
1.206
0.985
Post-op Month 11
0.763
0.801
261 mm 20/80
171 mm 20/200
Values are listed in that order vertically. Of note, some patients only had acuities measured at preoperative visit. *SRF height was subfoveal through measurement described in Figure 2. †All SRF resolved. NA, not available; NC, near card; post-op, postoperatively; pre-op, preoperatively.
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
Post-op Month 2
Pre-op
154
Table 2. Summary of the Raw Values for the ODR, SRF Height (in Micrometers),* and Visual Acuity Over Time
OPTICAL DENSITY OF SUBRETINAL FLUID KASHANI ET AL
suggested that drainage would decrease the cellularity and viscosity, leading to reduction in the incidence of persistent SRF. Our results show that increasing ODR may reflect the increased concentration of photoreceptor outer segment fragments in the SRF, as the serous component is absorbed, where the serous component is absorbed faster than the cellular component, thereby increasing ODR. At this point, it is not possible to differentiate between different SRF etiologies from OCT measurements, but our results suggest that this may be possible given a large sample of patients and varying pathological processes. Specifically, our study is the first to suggest that longitudinal measurements of OCT reflectivity and ODR may provide useful information about the nature and course of SRF after retinal detachment repair. Neither ODR nor SRF height was found to be significant predictors of measured visual acuity in this cohort. Importantly, nearly all of the eyes in this cohort were phakic at last follow-up visit, and it is likely that visual acuity results were confounded by cataract progression. This trend would be better assessed in a larger series of eyes, preferably pseudophakic at presentation. There are a number of limitations to this study. One such limitation is the protean nature of the vitreous cavity. It is likely that the presence or absence of a posterior vitreous detachment will affect the reflectivity measurements from the vitreous cavity. This variability may conceal small or modest changes in the ODR. For example, in Figure 1D, the reflectivity of the vitreous fluctuates nearly 10-fold but seems to be unchanged after 10 months compared with the constantly increasing reflectivity of the SRF pockets. This variability results in a relatively flat ODR in Figure 1E over the first 5 months to 6 months. This fluctuation is in contrast to the stable reflectivity measurements from the RPE and full-thickness retina, which are anatomically distinct and easily identifiable for measurements. Close inspection of subsequent OCTs (Figure 1C) demonstrates partial posterior vitreous detachment with vitreomacular adhesion at the fovea. We suspect that the posterior boundary of the vitreous is not always clearly visible on OCT in this case, although the OCT was of adequate quality for SRF measurements. Therefore, the large variability in the reflectivity measurements from the vitreous may reflect sampling from the vitreous inside or outside the posterior hyaloidal boundary in the case of partial posterior vitreous detachment. Despite this variability, it is still apparent that the reflectivity and ODR of SRF compartments increase with time. Of note, the preoperative ODR was less than one in all eyes. This does not necessarily imply that the SRF
155
is optically less dense than the overlying vitreous as the intervening retina is of relatively high reflectivity, which would serve to artificially decrease the absolute reflectivity of the SRF, thereby decreasing the ODR. As we chose to investigate temporal changes in the ODR, this confounding factor is minimized. This is supported by our observation that the absolute reflectivity of the retina remained stable over time in all eyes. It is important to note that the phenomenon of persistent SRF after retinal detachment repair is not unique to SB and is rather more simply a reflection of the completeness of SRF drainage at the time of surgery. Two eyes from this study were treated with vitrectomy and were noted to have persistent SRF. These eyes were found to display similar SRF patterns (increasing/stable ODR and decreasing SRF height) to the rest of the cohort, which underwent primary SB. Optical density ratio measurements may be confounded by numerous optical artifacts including, but not limited to, media opacity. In addition, differences in the tear film and other factors may affect the overall signal strength of the image on any given sitting. Therefore, we performed a number of analyses to control for possible variations in measurements both within an imaging session and between two different imaging sessions on the same eye. These analyses are described in the Methods section in detail. In general, the additional analyses normalized the signal intensity of the measured area of SRF to one of a number of “standards.” These standards included the vitreous cavity overlying the area of SRF (low-intensity control), the RPE under the area of SRF (high-intensity control), and the full thickness of the neurosensory retina overlying the area of SRF. All of these analyses confirmed the general trend of increasing SRF ODR with time. In conclusion, residual subfoveal fluid after surgical repair of macula-off RRD is not uncommon and may have implications in terms of clinical course and visual prognosis. Standard OCT imaging may be used to characterize this SRF quantitatively through measurement of absolute reflectivity and calculation of ODR. Changes in the ODR of residual SRF after RD repair can be easily monitored and may be representative of changes in the composition of the SRF over time. With additional study, this parameter may prove a reliable noninvasive method of characterizing SRF content in vivo. Key words: optical coherence tomography, optical density, retina, rhegmatogenous retinal detachment, subretinal fluid.
156
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 1
References 1. Cooper WC, Halbert SP, Manski WJ. Immunochemical analysis of vitreous and subretinal fluid. Invest Ophthalmol 1963;2: 369–377. 2. Lister W. Holes in the retina and their clinical significance. Br J Ophthalmol 1924;8:i4–20. 3. Dorello U. Electrophoretic studies of the protein content of the subretinal fluid in idiopathic retinal detachment. Am J Ophthalmol 1956;41:564. 4. Smith JL, Douty E. Electrophoresis of subretinal fluid. Arch Ophthalmol 1960;64:114–119. 5. Magitot A. The subretinal fluid in idiopathic detachment of the retina. Arch Ophthalmol 1934;11:159–173. 6. Hara S, Ishiguro S, Hayasaka S, Mizuno K. Immunoreactive opsin content in subretinal fluid from patients with rhegmatogenous retinal detachments. Arch Ophthalmol 1987;105:260–263. 7. Heath H, Beck TC, Foulds WS. Chemical composition of subretinal fluid. Br J Ophthalmol 1962;46:385–396. 8. Veckeneer M, Derycke L, Lindstedt EW, et al. Persistent subretinal fluid after surgery for rhegmatogenous retinal detachment: hypothesis and review. Graefes Arch Clin Exp Ophthalmol 2012;250:795–802. 9. Lam KW, van Heuven WA, Ray GS, Feman S. Subretinal fluids: lipid analyses. Invest Ophthalmol 1975;14:406–410. 10. Feeney L, Burns RP, Mixon RM. Human subretinal fluid. Its cellular and subcellular components. Arch Ophthalmol 1975; 93:62–69. 11. Quintyn JC, Brasseur G. Subretinal fluid in primary rhegmatogenous retinal detachment: physiopathology and composition. Surv Ophthalmol 2004;49:96–108. 12. Williams GA, Reeser F, O’Brien WJ, Fleischman JA. Prostacyclin and thromboxane A2 derivatives in rhegmatogenous subretinal fluid. Arch Ophthalmol 1983;101:463–464. 13. Chen YH, Chen JT, Chien MW, et al. Subretinal fluid from rhegmatogenous retinal detachment and blood induces the expression of ICAM-1 in the human retinal pigment epithelium (ARPE-19) in vitro. Eye (Lond) 2010;24:354–360. 14. Akhmeteli LM, Kasvina BS, Petropavlovskaja GA. Biochemical investigation of the subretinal fluid. Br J Ophthalmol 1975; 59:70–77. 15. Schenk H, Formanek K, Förster O. Untersuchungen der Eiweißkörper in der subretinalen Flüssigkeit bei Netzhautablösung. Graefes Arch Ophthalmologie 1961;164:29–41. 16. Sachsenweger R, Gassler H, Hentsch R. Elektrophoretische Untersuchungen der subretinalen Flüssigkeit bei Netzhautablösungen. Graefes Arch Ophthalmologie 1964;166:432–439.
17. Berrod JP, Kayl P, Rozot P, Raspiller A. Proteins in the subretinal fluid. Eur J Ophthalmol 1993;3:132–137. 18. Brown JC, Solomon SD, Bressler SB, et al. Detection of diabetic foveal edema: contact lens biomicroscopy compared with optical coherence tomography. Arch Ophthalmol 2004;122:330–335. 19. Wang XY, Shen LP, Hu RR, Xu W. Persistent subretinal fluid after successful scleral buckle surgery for macula-off retinal detachment. Chin Med J (Engl) 2011;124:4007–4011. 20. 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;145:915–922. 21. Ricker LJ, Noordzij LJ, Goezinne F, et al. Persistent subfoveal fluid and increased preoperative foveal thickness impair visual outcome after macula-off retinal detachment repair. Retina 2011;31:1505–1512. 22. Benson SE, Schlottmann PG, Bunce C, et al. Optical coherence tomography analysis of the macula after scleral buckle surgery for retinal detachment. Ophthalmology 2007;114:108–112. 23. Wolfensberger TJ, Gonvers M. Optical coherence tomography in the evaluation of incomplete visual acuity recovery after macula-off retinal detachments. Graefes Arch Clin Exp Ophthalmol 2002;240:85–89. 24. Delolme MP, et al. Anatomical and functional macular changes after rhegmatogenous retinal detachment with macula off. Am J Ophthalmol 2012;153:128–136. 25. Barthelmes D, Sutter FK, Gillies MC. Differential optical densities of intraretinal spaces. Invest Ophthalmol Vis Sci 2008; 49:3529–3534. 26. Ahlers C, Golbaz I, Einwallner E, et al. Identification of optical density ratios in subretinal fluid as a clinically relevant biomarker in exudative macular disease. Invest Ophthalmol Vis Sci 2009;50:3417–3424. 27. Neudorfer M, Weinberg A, Loewenstein A, Barak A. Differential optical density of subretinal spaces. Invest Ophthalmol Vis Sci 2012;53:3104–3110. 28. Guérin CJ, Lewis GP, Fisher SK, Anderson DH. Recovery of photoreceptor outer segment length and analysis of membrane assembly rates in regenerating primate photoreceptor outer segments. Invest Ophthalmol Vis Sci 1993;34:175–183. 29. Dubois A, Vabre L, Boccara AC, Beaurepaire E. Highresolution full-field optical coherence tomography with a Linnik microscope. Appl Opt 2002;41:805–812. 30. Nakano. Studies on the subretinal fluid. Report I. Refractometric and paper-electrophoretic study on the subretinal fluid of spontaneous retinal detachment. Jpn J Ophthalmol 1961;3: 23–28.
