Association Between Photoreceptor Regeneration and Visual Acuity Following Surgery for Rhegmatogenous Retinal Detachment

Abstract

Purpose: The purpose of this study was to evaluate foveal regeneration and the association between retinal restoration and visual acuity following reattachment surgery for rhegmatogenous retinal detachment (RRD).

Methods: Twenty-nine eyes of 29 patients with successfully reattached macula-off RRD were retrospectively analyzed. We used spectral-domain optical coherence tomography to image macular regions and measure retinal thickness and Snellen chart visual acuity (VA) to evaluate best-corrected VA (BCVA) at 1, 2, 3, 6, 9, and 12 months after vitrectomy. Best-corrected visual acuity data were converted to the logarithm of the minimum angle of resolution scale. Opposite eyes were used as controls.

Results: The thicknesses of the external limiting membrane (ELM)-ellipsoid zone (EZ) and EZ-retinal pigment epithelium (RPE) were significantly thinner in involved eyes than in corresponding unaffected eyes at 1 month after surgery (P < 0.001 for both), and the thickness increased over time (P < 0.001 for both). Best-corrected visual acuity significantly improved over time (P < 0.001), and the improvement correlated with EZ-RPE thickness (r = −0.45, P = 0.021). Multiple regression analysis demonstrated the presence of a foveal bulge as the independent predictor of final BCVA (P < 0.001). Eyes with a foveal bulge had significantly better BCVA and greater EZ-RPE thickness than those without throughout the follow-up period. Significant restoration of the integrity of EZ and cone interdigitation zone (CIZ) was observed over time (P < 0.001 for both) in eyes with a foveal bulge.

Conclusions: The thickness of EZ-RPE and cone density increased during foveal regeneration, as demonstrated by the continuous improvements in CIZ integrity over time, leading to the formation of foveal bulge and good vision following successful reattachment of macula-off RRD.

Rhegmatogenous retinal detachment (RRD) is a sight-threatening pathology. Currently, the only treatment modality for RRD is retinal reattachment.¹ Although the anatomical success rate of retinal reattachment is high,^2–4 patients are often disappointed with improvements in best-corrected visual acuity (BCVA) following surgery, particularly in eyes affected by macular detachment. Photoreceptor volume in cases of macula-off RRD (i.e., separation of the macula from the retinal pigment epithelium [RPE]) has been shown to be reduced compared to that in cases of macula-on RRD.⁵ Persistent functional damage to the macula is observed in a proportion of eyes affected by macula-off RRD.^6–11 Factors reportedly associated with functional recovery following macula-off RRD include preoperative VA,¹² duration of macular detachment,^6,8 height of macular detachment,^13,14 and age.¹² Even among eyes predicted to have good postoperative vision due to good preoperative conditions and the absence of significant complications during surgery, some eyes continue to have poor VA. 

Technologic advantages in spectral-domain optical coherence tomography (SD-OCT) have allowed detailed retinal evaluation and understanding of the foveal microstructure recovery process after RRD. The Spectralis model (Heidelberg Engineering, Heidelberg, Germany), which incorporates software with an eye movement-tracking function, can perform serial scans at the same location, thus allowing precise evaluation of changes occurring at given retinal areas.¹⁵ Many studies using SD-OCT have demonstrated that integrity of the ellipsoid zone (EZ; i.e., the junction between the inner and outer segment of photoreceptors) and external limiting membrane (ELM) is significantly correlated with BCVA following retinal reattachment.^5,16–21 However, these reports had considerable limitations, including inconsistencies between duration and location of images taken, a lack of objective assessment of membrane integrity, and decreased band reflectivity due to fragmentation or thinning, which can be observed in 5% of healthy eyes as a result of artifacts.²² 

Careful examination of SD-OCT images of normal eyes demonstrated bulging of the EZ at the central fovea, termed the foveal bulge. Recent OCT studies have shown that the presence or absence of a foveal bulge at the central fovea is significantly correlated with VA in eyes with albinism,²³ occult macular dystrophy,²⁴ amblyopia,²⁵ and branch retinal vein occlusion.²⁶ Hasegawa et al.²⁷ reported a significant correlation between the presence of foveal bulge and BCVA after successful RRD repair with vitrectomy. The authors supported the utility of evaluating foveal bulge and foveal photoreceptor outer segment (OS) length in determining the visual properties of eyes successfully treated with retinal reattachment.²⁷ However, that study did not evaluate retinal layer thickness over time, including OS length, or the time required for foveal bulge regeneration following successful retinal reattachment. 

Data are lacking in regard to changes in retinal layer thickness and the correlation between thickness changes and BCVA outcomes following successful RRD repair. Dell'Omo et al.²⁸ and Terauchi et al.²⁹ performed serial evaluations of changes in retinal layer thickness at the same location with SD-OCT and observed progressive increases in the thickness of several central retinal layers. However, assessments of the relationship between retinal layer thickness and vision were limited in those studies. The direct correlation between increased retinal layer thickness and improvements in BCVA was unclear as parameters were not directly compared over time. In addition, no significant relationship between foveal bulge regeneration and VA was observed. 

Thus, the goal of the present study was to quantify changes in retinal layer thickness and evaluate its relationship to improvements in BCVA and to investigate the time required for regeneration of the foveal bulge following successful retinal reattachment and any potential association with postoperative BCVA. 

Methods

Ethics Statement

This retrospective, observational, comparative, single-center study followed the principles of the Declaration of Helsinki and was approved by the Institutional Review Board and Ethics Committee of the Nagoya University Graduate School of Medicine. 

Measurement Using Optical Coherence Tomographic Images

A Spectralis model SD-OCT unit was used to obtain all SD-OCT images. We evaluated horizontal cross-sectional images recorded at each visit after successful retinal reattachment. Retinal layer thickness was measured on the same selected central foveal scan throughout follow-up, using the computer-based caliper measurement tool of the SD-OCT system. Central foveal thickness (CFT) was measured as the thickness between the surface of the internal limiting membrane (ILM) and the outer border of RPE at the central fovea (Fig. 1). The thickness of the outer nuclear layer (ONL) was defined as the distance between the outer borders of ILM and ELM. The ELM-EZ thickness (inner segment thickness) was defined as the distance between the outer borders of ELM and EZ.³⁰ The EZ-RPE (OS) thickness was defined as the distance between the outer border of EZ and the inner border of RPE. 

Figure 1

 

Representative SD-OCT image of a normal eye. A horizontal scan through the central fovea was obtained. Retinal zones were visualized by SD-OCT. Central foveal thickness was defined as the distance between the surface of the ILM and the outer border of the RPE at the central fovea. Outer nuclear layer thickness was measured as the distance between the outer border of the ILM and the outer border of the ELM band. External limiting membrane-ellipsoid zone thickness was measured as the distance between the outer border of the ELM band and the outer border of the EZ. Ellipsoid zone-retinal pigment epithelium thickness was measured as the distance between the outer border of EZ and inner border of RPE. The SD-OCT image shows bulging of ELM, EZ, and CIZ at the central fovea, known as the foveal bulge.

Retinal layer thickness was measured manually at the foveal bulge (if visible) by operators masked to VA values and other information including preoperative status. A foveal bulge was defined as an EZ-RPE thickness at the central fovea >10 μm greater than the average EZ-RPE thickness at 250 μm temporal and nasal to the central fovea. In cases where the foveal bulge was not visible, measurements were made along a vertical line passing through the steepest part of the foveal excavation. Identical measurements were made in opposite eyes as controls. 

Integrity of the foveal ELM, the EZ, and the cone interdigitation zone (CIZ) was evaluated in a 1-mm-diameter area for each image on a 4-point scale as follows: 1, line not visible; 2, line disruption >200 μm; 3, line disruption <200 μm; and 4, continuous line. Identical measurements were performed in opposite eyes as controls. 

Subjects

We retrospectively reviewed all patients who had undergone successful RRD repair with vitrectomy at the Nagoya University Hospital from June 2012 to May 2014, in whom the EZ line at the central fovea could be observed in follow-up SD-OCT images. All patients provided an informed consent form prior to surgery. 

Patients were initially divided into two groups according to preoperative macula status and were evaluated using preoperative SD-OCT macular scans: macula-off RRD, that is, retinal detachment involving the macula; and macula-on RRD, that is, retinal detachment not involving the macula. Patients were further divided into two subgroups according to the presence of a foveal bulge, which was evaluated at each follow-up visit. 

All patients underwent comprehensive ophthalmic examinations, including measurements of BCVA, IOP, and axial length; slit-lamp examination; fundus examination; and SD-OCT before surgery and at 1, 3, 6, 9, and 12 months after surgery. Snellen VA values were converted to the logarithm of the minimum angle of resolution (LogMAR) units in order to create a linear scale of VA. 

Surgical Technique

Standard 3-port pars plana vitrectomy was performed with 25-gauge instruments after retrobulbar anesthesia with 2.5 mL each of 2% lidocaine and 0.5% bupivacaine. No patients underwent concurrent scleral buckling surgery. In eyes with cataract, cataract surgery was performed as described below. A 2.4-mm-wide self-sealing superior sclerocorneal tunnel was created at the 12-o'clock position, and a continuous curvilinear capsulorhexis was performed. The lens nucleus was removed, and the residual cortex was aspirated with an irrigation/aspiration tip. Next, a foldable acrylic intraocular lens was implanted in the bag. A trocar was then inserted at approximately 30° parallel to the limbus with the bevel-side up. Once the trocar was past the trocar sleeve, the angle was changed to perpendicular to the surface. After 3 ports were created, vitrectomy was performed using the Constellation system (Alcon Laboratories, Inc., Fort Worth, TX, USA). After fluid–air exchange and subretinal fluid drainage from the causative retinal tear(s) or iatrogenic hole were performed, intraoperative photocoagulation was applied to the causative retinal tear(s) or iatrogenic hole (if present). At completion of vitrectomy, 20% sulfur hexafluoride (SF6) was injected into the vitreous. After IOP was adjusted to a normal tension, cannulae were withdrawn, and the sclera was pressed and massaged with an indenter to close the wound. 

Exclusion Criteria

Exclusion criteria included dense ocular medium (e.g., vitreous hemorrhage, vitreous opacity), pre-existing macular conditions (e.g., macular degeneration, vascular occlusive diseases, or diabetic retinopathy), proliferative vitreoretinopathy grade ≥C,³¹ and clinically evident postoperative change likely to interfere with accurate evaluation of retinal layers (e.g., recurrent RRD, epiretinal membrane, cystoid macular edema, or persistent subretinal fluid). 

Statistical Analysis

Values are means ± standard deviations (SD). Independent t-test was used to compare normally distributed data, and the χ² test was used for categorical data. One-way analysis of variance (ANOVA) was used to evaluate changes in BCVA, retinal layers thickness, and integrity of outer retinal bands over time. After a linear approximate equation was used for calculating the slopes of BCVA and EZ-RPE/ELM-EZ thickness for each eye (Supplementary Fig. S1), Pearson's correlation coefficient test was used to evaluate the association between them. Multiple linear regression analysis was used to evaluate the association between final BCVA and independent variables including the presence of foveal bulge, EZ-RPE thickness, preoperative BCVA, age, axial length, and duration of retinal detachment. A P value <0.05 was considered statistically significant. 

Results

Patient Demographics and Surgical Parameters

Between June 2012 and May 2014, 53 eyes of 53 patients with macula-off RRD and 26 eyes of 26 patients with macula-on RRD underwent vitrectomy at our department for the repair of RRD. Of those, 37 eyes were excluded for presence of proliferative vitreoretinopathy grade ≥C or worse (n = 3), vitreous hemorrhage (n = 1), macular hole (n = 1), diabetic retinopathy (n = 1), postoperative development of dense cataract (n = 2), macular edema (n = 2), subretinal fluid (n = 6), or significant epiretinal membrane (n = 1) at any time of follow-up or an incapacity to attend regular follow-up visits (n = 20). As a result, 29 eyes with macula-off RRD and 13 eyes with macula-on RRD were included in final analysis. Patient demographics and surgical parameters are shown in Table 1. No significant intergroup differences were observed in age, sex, axial length, or surgical procedures, except for preoperative BCVA (logMAR; P < 0.001). 

Table 1

 

Patient Clinical Characteristics

 

Changes in Retinal Thickness and BCVA Over Time Following Surgery for Macula-Off and Macula-On RRD

In eyes affected by macula-on RRD, no significant differences in BCVA or thickness of CFT, ONL, ELM-EZ, or EZ-RPE were observed during the follow-up period, and no significant differences were observed in any parameter compared to control eyes (Fig. 2; Table 2). 

Figure 2

 

A representative SD-OCT image of an eye with macula-on RRD before surgery (A). Thickness and reflectivity of the lines in the corresponding unaffected eye were similar to those of the affected eye prior to surgery (B). The ELM, EZ, and CIZ were visible and without any disruption during the follow-up period, and the foveal bulge was present during the follow-up period in the eye affected by macula-on RRD (C).

Table 2

 

Changes Between Thickness of Different Layers in Eyes With Macula-Off and Those With Macula-On RRD

 

Significantly increased thickness was observed at the level of ELM-EZ (inner segment thickness: 25.2 ± 4.8 to 31.4 ± 2.6 μm; P < 0.001) (Fig. 3A; Table 2) and EZ-RPE (OS thickness: 25.4 ± 10.4 to 41.1 ± 5.6 μm; P < 0.001) between postoperative months 1 and 12 (Fig. 3B) in the macula-off RRD group, and no significant differences were observed in CFT or ONL thickness over time (Figs. 3C, 3D). Mean EZ-RPE thickness was significantly thinner in eyes affected by macula-off RRD than in control eyes throughout the follow-up period (Fig. 3B, **P < 0.001). Mean ELM-EZ thickness was significantly thinner in eyes affected by macula-off RRD than in control eyes until 6 months postoperatively (Fig. 3A, **P < 0.001, *P < 0.01). 

Figure 3

 

Changes in mean retinal layer thickness are shown in eyes with macula-off RRD after surgery and corresponding control eyes. Mean ELM-EZ thickness gradually increased in eyes with macula-off RRD over the postoperative period (A). External limiting membrane-ellipsoid zone thickness was significantly thinner in eyes with macula-off RRD than in control eyes until 6 months postoperatively. Mean EZ-RPE thickness gradually increased in eyes with macula-off RRD over time (B). Ellipsoid zone-retinal pigment epithelium thickness was significantly thinner in eyes with macula-off RRD than in control eyes throughout the follow-up period. Mean CFT in eyes with macula-off RRD was significantly thinner in control eyes at 1 month after surgery (C). No significant differences in ONL thickness were observed any time throughout the follow-up period (D). *P < 0.01; **P < 0.001.

In eyes affected by macula-off RRD, mean postoperative BCVA was significantly improved, from 0.39 ± 0.29 to 0.15 ± 0.14, between postoperative months 1 and 12 (P < 0.001) (Fig. 4A; Table 2) but remained worse than control eyes (0.03 ± 0.05) at 12 months. 

Figure 4

 

Change in mean BCVA after vitreous surgery for macula-off RRD (A). The slope of the regression line from logMAR over time was correlated with the slope of the EZ-RPE thickness over time ([B] r = −0.45, P = 0.021) but not the slope of the regression line for ELM-EZ thickness over time ([C] r = −0.16, P = 0.422). *P < 0.01; **P < 0.001.

The slope of the regression line for change in BCVA over time (Supplementary Fig. S1) was significantly correlated with EZ-RPE thickness over time (r = −0.45; P = 0.021) (Fig. 4B) but not with ELM-EZ thickness (r = −0.16; P = 0.422) (Fig. 4C). 

Multiple stepwise regression analysis for final BCVA (Table 3) revealed only the presence of a foveal bulge as an independent predictor of final VA (P < 0.001). 

Table 3

 

Results of Multiple Stepwise Regression Analysis for Independence of Factors Contributing to Final BCVA

 

Differences Between Eyes With and Without a Foveal Bulge in the Macula-Off RRD Group

Clinical characteristics of the macula-off RRD group with or without the presence of a foveal bulge are shown in Table 4. A foveal bulge was observed in 17 eyes (Fig. 5) and not in 12 (Fig. 6) during the follow-up period. No significant differences in any clinical characteristic, including age, sex, duration of macular detachment, preoperative BCVA, and axial length, were observed between patients with and without a foveal bulge. 

Table 4

 

Patient Clinical Characteristics in Eyes With and Without Foveal Bulge

 

Figure 5

 

Representative SD-OCT images of an eye with thickening in the central fovea and reconstitution of the foveal bulge after repair of macula-off RRD. Spectral-domain optical coherence tomography image demonstrating the preoperative status of the macula prior to surgery, with a visual acuity of 20/60 (A). The thickness and reflectivity of the outer bands were normal in the corresponding unaffected eye (B). The ELM, EZ, and CIZ appeared fragmented and thin, and the foveal bulge was not visible at 1 month after surgery (C). Progressive increases in the reflectivity of the outer bands, along with reconstitution of the foveal bulge were observed between postoperative months 3 and 12. The reflectivity of the outer bands became similar to that of the corresponding unaffected eye at 12 months postoperatively, but the thickness of the outer bands remained thin compared to those of the corresponding unaffected eye (C).

Figure 6

 

Representative SD-OCT image of an eye with thickening of the central fovea but no reconstitution of the foveal bulge after repair of macula-off RRD. Preoperative SD-OCT image of the macula with a visual acuity of 20/200 (A). The thickness and reflectivity of the outer bands were normal in the corresponding unaffected eye (B). All retinal layers were thin, and the ELM, EZ, and CIZ bands were not apparent at 1 month after surgery (C). Each retinal layer became thicker, and ELM and EZ bands were observed to be partially reconstituted at 3 months after surgery. Each retinal layer became thicker at 12 months after surgery, and no differences in retinal layer thickness were observed compared to the corresponding unaffected eye. However, EZ and CIZ were only partially reconstituted at the central fovea, with a visual acuity of 20/40.

Mean postoperative BCVA in eyes with a foveal bulge was significantly better than in eyes without a foveal bulge throughout the follow-up period (P < 0.001) (Table 5). Mean postoperative BCVA significantly improved from 0.23 ± 0.14 to 0.05 ± 0.07 between postoperative months 1 and 12 in eyes with a foveal bulge (P = 0.035) but not in eyes without a foveal bulge (Table 5). In addition, EZ-RPE thickness was significantly greater in eyes with a foveal bulge than in eyes without a foveal bulge throughout the follow-up period (P < 0.001) (Table 5). Furthermore, EZ-RPE thickness significantly increased from 28.0 ± 8.6 to 46.3 ± 5.9 μm between postoperative months 1 and 12 in eyes with a foveal bulge (P < 0.001), but no significant improvement was observed in eyes without a foveal bulge. 

Table 5

 

Changes in Thickness of Different Layers in Eyes With and Without Foveal Bulge in Macula-Off RRD

 

The time at which a foveal bulge was first observed varied between 1 and 12 months postoperatively and was not associated with final BCVA (r = 0.27; P = 0.281) (Fig. 7). 

Figure 7

 

Correlation between the time of first appearance of the foveal bulge and final BCVA. The time of first appearance of the foveal bulge was not found to correlate with LogMAR at 12 months after surgery.

During the follow-up period, significant restoration of the integrity of EZ and CIZ was observed in eyes with a foveal bulge (both: P < 0.001) and of EZ in eyes without a foveal bulge (P < 0.001) (Table 5). The integrity of CIZ significantly differed between these two groups throughout the follow-up period (P < 0.001 to P < 0.05 at different time points). 

Discussion

Our results showed that ELM-EZ thickness (inner segment thickness) and EZ-RPE thickness (OS thickness) were thinned in eyes affected by macula-off RRD compared to those in opposite, unaffected eyes at 1 month after successful attachment, and then the thicknesses significantly increased over time, with partial restoration of the integrity of the outer retinal bands. In eyes affected by macula-off RRD, BCVA significantly improved between postoperative months 1 and 12. In addition, the slope of the regression line for change in EZ-RPE thickness over time was significantly correlated with that of BCVA over time. Foveal bulge was an independent predictor of final BCVA. Eyes with a foveal bulge had significantly better BCVA and greater EZ-RPE thickness than those without a foveal bulge throughout the follow-up period. 

Spectral-domain optical coherence tomography provided direct visualization of in vivo retinal morphology, allowing high-resolution observation of individual layers of the macula, thereby providing greater information for structural postoperative macular changes. Recent OCT studies have reported disruptions of photoreceptor microstructures and integrity of the outer retinal bands at the fovea in cases of macula-off RRD^{5,17–19,32,33} and other retinal diseases (e.g., macular hole^20,34 and central serous chorioretinopathy³⁵). In the present study, SD-OCT facilitated precise measurement of retinal layer thickness and evaluation of outer retinal band integrity. Regarding thinning of the retinal layer following RRD, experimental studies have demonstrated dropout of OS photoreceptors due to RD-induced separation of the OS from the RPE, thereby disrupting normal OS renewal and leading to OS shortening and eventual degeneration.^36–39 Detachment of the neural retina from the RPE induced a variety of changes in several cell types (e.g., photoreceptors, RPE, Müller cell, and so on) throughout the retina.¹⁰ 

There have been more limited time-sequenced data regarding the time course of retinal layer thickness changes and association with BCVA outcomes following successful macula-off RRD repair. Two recent studies reported the thickness of retinal layers was significantly thinned after successful attachment and gradually increased, with only EZ-RPE thickness remaining thinner than the opposite, unaffected eye at 12 months postoperatively.^28,29 Our results corroborate those findings. 

In addition to changes in retinal layer thickness following successful retinal reattachment, BCVA in eyes with macula-off RRD significantly improved between postoperative months 1 and 12 in the present study. Previous studies^28,29 evaluated the relationship between improvements in BCVA and thickness of EZ-RPE or ELM-EZ at particular time points postoperatively (e.g., 1 month or 12 months); however, increased retinal layer thickness and improvements in BCVA were not directly compared over time. Accordingly, the association between improvements in BCVA and increased retinal layer thickness remains unclear. We calculated a linear approximate equation for the association between BCVA and ELM-EZ or EZ-RPE thickness over time and observed a significant correlation between the slopes of the regression lines for change in EZ-RPE thickness and BCVA over time. We found that BCVA improved in parallel with increased EZ-RPE thickness (OS thickness) following successful retinal reattachment. 

Factors previously reported to be associated with functional recovery after macula-off RRD included preoperative VA,¹² duration of macular detachment,^6,8 height of macular detachment,^13,14 and age.¹² In the present study, the presence of a foveal bulge was the only significant independent predictor of final BCVA. Perhaps the number of patients we included was insufficient to evaluate the aforementioned factors. Recent OCT studies have demonstrated the presence or absence of a foveal bulge at the central fovea is significantly associated with VA in eyes with albinism,²³ occult macular dystrophy,²⁴ amblyopia,²⁵ branch retinal vein occlusion,²⁶ and RRD.²⁷ Centripetal migration of cone cells and thinning of individual foveal cone OS reportedly results in an increase in foveal cone OS density.⁴⁰ Hasegawa et al.^26,27 suggested that increased foveal photoreceptor OS length was related to the presence of a foveal bulge on OCT imaging, indicating high foveal cone OS density in eyes with a foveal bulge. We found that BCVA was significantly better in eyes with a foveal bulge than in those without throughout the follow-up period. Our results indicate the presence of a foveal bulge is essential for achieving good final vision, and corroborate the previous findings of an association between formation of a foveal bulge and vision. On the other hand, no association was observed between final BCVA and time until the first appearance of a foveal bulge after surgery. This result indicates that eyes with a foveal bulge are more likely to have better final vision regardless of the time required for the reappearance of a foveal bulge after macula-off RRD. 

In eyes that eventually developed a foveal bulge, BCVA was significantly greater even at 1 month postoperatively, when no eyes were seen to have a foveal bulge. In addition, EZ-RPE thickness and integrity of EZ and CIZ significantly differed between eyes with and without a foveal bulge from 1 month postoperatively. These results indicate regeneration of the foveal structure after macula-off RRD occurs earlier in eyes with a foveal bulge, resulting in better BCVA even before the formation of a foveal bulge. Perhaps eyes with a foveal bulge have less macular damage prior to surgery or increased photoreceptor regeneration in eyes with macula-off RRD. These findings indicate EZ-RPE thickness or integrity of outer bands during the early postoperative period may be good predictors of the formation of a foveal bulge during the follow-up period and the achievement of a good final BCVA. 

Cone photoreceptor density is an important marker for achieving good BCVA. Ooto et al.⁴¹ used an adaptive optics scanning laser ophthalmoscope to determine cone photoreceptor density and compared their findings with microstructures determined by a commercially available SD-OCT. They found the mean cone density in eyes with a disrupted CIZ line was significantly lower than that in eyes with an intact CIZ line and that cone density in the foveal area was correlated with BCVA. Of the outer retinal bands, the integrity of CIZ was better in eyes with a foveal bulge throughout the follow-up period in the present study, corroborating the findings of Ooto et al.⁴¹ as the presence of a foveal bulge on OCT imaging has previously been shown to indicate high foveal cone OS density.^26,27 Taken together, these findings indicate EZ-RPE thickness (the OS thickness) and cone density increase during foveal regeneration, observed as a continuous CIZ line, would be associated with the formation of a foveal bulge and better final vision in eyes following successful reattachment of macula-off RRD. 

There are limitations to the present study. This was a retrospective study with a relatively small sample size, which may have resulted in an insufficient number of participants for adequate comparison of preoperative BCVA and other factors, such as age and duration of macular detachment. Another limitation is the follow-up period of 12 months. As EZ-RPE thickness may increase after 12 months postoperatively, further longitudinal studies are required to confirm complete recovery of EZ-RPE thickness to the same level as that of the corresponding healthy eye. Furthermore, we did not evaluate the preoperative microstructure of retinal layers using SD-OCT imaging. Although no significant differences in preoperative BCVA were observed between eyes with and without a foveal bulge, retinal microstructure might have differed between these groups prior to surgery, particularly as restoration of the integrity of the outer retinal bands differed between these groups even during the early postoperative period. In addition, retinal layer distances were manually measured as automated calculation of retinal layer thicknesses may be technically challenging in eyes with fragmented or poorly visualized retinal layers. Further prospective studies with larger sample sizes and automated calculation of retinal thicknesses and assessments of outer retinal bands are required. 

In conclusion, the findings of the present study demonstrate ELM-EZ (inner segment thickness) and EZ-RPE thickness (OS thickness) in eyes with a reattached RRD are significantly thinner immediately after surgery and progressive recovery of thickness and restoration of the outer retinal layers/bands at the fovea occurs following macula-off RRD repair. During foveal regeneration, increased EZ-RPE thickness (OS thickness) and cone density, observed as a continuous CIZ, would be associated with the formation of a foveal bulge and better final vision in eyes following successful reattachment of macula-off RRD. 

Acknowledgments

Supported by a 2015 Association for Research in Vision and Ophthalmology travel grant (MK), Grant-in-Aid for Scientific Research C; 26462635; TI; Tokyo, Japan, and Grant-in-Aid for Scientific Research B; 15H04994; HT; Tokyo, Japan. 

Disclosure: M. Kobayashi, None; T. Iwase, None; K. Yamamoto, None; E. Ra, None; K. Murotani, None; S. Matsui, None; H. Terasaki, None