Purpose To report the prevalence of outer retinal layer (ORL) damage after macula-off rhegmatogenous retinal detachment (RRD) surgery and to determine its associated preoperative risk factors.
Methods 253 eyes successfully operated for macula-off RRD were included in the study. The integrity of the external limiting membrane (ELM), ellipsoid zone (EZ) and cone interdigitation zone (CIZ) of the photoreceptors was assessed at 1 month and 6 months using spectral-domain optical coherence tomography. Risk factors were analysed using univariate and multivariate logistic regression. The correlation between ORL integrity and visual outcomes was also evaluated.
Results CIZ, EZ and ELM defects were found in, respectively, 198 (93.4%) eyes, 100 (47.2%) eyes, 64 (30.2%) eyes at 1 month and in 160 (63.2%) eyes, 44 (17.4%) eyes and 18 (7.1%) eyes at 6 months. In multivariate analysis, duration of macular detachment was the only factor associated with ORL damage at 6 months (p=0.007). Best-corrected visual acuity significantly improved from 0.5±0.3 at 1 month to 0.3±0.3 logarithm of minimal angle of resolution at 6 months (p<0.001) and was strongly correlated with the number of affected bands (p<0.001).
Conclusion Prevalence of outer retinal band defects substantially decreased through the study period, confirming the ability of photoreceptors to recover over time. However, shorter interval to surgery and better visual outcomes were significantly associated with fewer defects within the ORL at 6 months. These findings suggest that earlier surgery may limit RRD-associated photoreceptor degeneration and improve the patient’s visual prognosis.
- treatment surgery
Statistics from Altmetric.com
Rhegmatogenous retinal detachment (RRD) is a sight-threatening condition with an annual incidence of 17.9 per 100 000 people.1 Advances in surgical techniques over recent decades have greatly improved the anatomical results, with a primary success rate currently up to 85%.2 However, despite successful retinal reattachment, visual recovery may remain incomplete, especially in cases of macula-off RRD. Reduced postoperative visual acuity may result from clinically detectable lesions including epiretinal membranes (ERMs), cystoid macular oedema, residual macular hole, retinal folds and pigment migration.3 Nevertheless, it can also occur in a normal-looking retina, suggesting the existence of microstructural macular alterations that fundus examination cannot detect.
Experimental studies in both animal models and human eyes have indeed demonstrated that retinal detachment causes a shortening of the outer segments and gradual death of the photoreceptor cells, which may contribute to the poor vision in these patients.4–6
The advent of high-resolution spectral-domain optical coherence tomography (SD-OCT) has allowed detection of photoreceptor microstructural abnormalities in the macula, in histologic detail, and better understanding of the discrepancies between anatomical and visual outcomes after RRD surgery. Many previous studies have demonstrated that disruptions of the external limiting membrane (ELM), ellipsoid zone (EZ) and cone interdigitation zone (CIZ) of the photoreceptors are correlated with decreased visual acuity after retinal reattachment.7–11 Furthermore, other SD-OCT reports have shown evidence of possible photoreceptor regeneration and concomitant visual recovery over time, with gradual restoration of the CIZ, the EZ and the ELM integrity, thickening of the fovea and increase in photoreceptor volume.7 12–15 Thus, prevalence of outer retinal band disruption varies widely in the literature, ranging from 5.6% to 55.0% for the ELM, 11.1% to 77.6% for the EZ and 26.7% to 98.9% for the CIZ, and mainly depends on the time of SD-OCT assessment after surgery.7–12 16 However, there is a lack of research on factors that might affect the integrity of outer retinal layers (ORLs) after RRD surgery.
The aim of this study was to determine, in a large series, the prevalence of ORL damage after macula-off RRD surgery and potential preoperative risk factors. We also evaluated the correlation between ORL integrity and visual outcomes.
Patients and study design
A retrospective review of medical records was performed on 918 consecutive patients who underwent surgery for RRD at Nancy University Hospital by retinal specialists from January 2011 to June 2017. All patients had complete information about the risks and benefits of the surgical procedure and gave their written consent before surgery.
Inclusion criteria were as follows: (1) patients with successfully repaired primary macula-off RRD by a single scleral buckling procedure or pars plana vitrectomy and (2) a minimum follow-up period of 6 months after surgery without retinal redetachment. Exclusion criteria were as follows: (1) patients with macula-on RRD, (2) patients with traumatic RRD, (3) patients with pre-existing macular pathologic features, such as macular hole, age-related macular degeneration, myopic maculopathy, diabetic maculopathy, retinal vein occlusion and (4) patients with unavailable or low-quality and unreliable postoperative OCT images.
All patients underwent a detailed ophthalmological examination before and after surgery, including best-corrected visual acuity (BCVA) measured with projected-light Snellen charts, axial length measurement using IOLMaster (Carl Zeiss Meditec, Dublin, CA, USA), biomicroscopy with anterior segment evaluation, fundus, careful peripheral retina examination and macular imaging using the Spectralis HRA-OCT with determination of outer retinal band integrity (Heidelberg Engineering, Heidelberg, Germany). An Amsler-Dubois scheme was systematically established for each patient, specifying the extent of the retinal detachment (RD), number, type and location of retinal breaks, existence of vitreous haemorrhage and preoperative proliferative vitreoretinopathy (PVR) grading according to Machemer et al. 17
All patients underwent either scleral buckling or three-port pars plana vitrectomy (combined or not with phacoemulsification and posterior chamber intraocular implantation), with gas or silicon oil tamponade depending on the surgeon’s preferences. After surgery, complete retinal attachment was achieved in all cases.
Preoperative, intraoperative and postoperative data
Preoperative data included patient age and sex, axial length, lens status, BCVA, characteristics of the RRD and available SD-OCT findings including height of retinal detachment, existence of outer nuclear layer or inner nuclear layer cystoid cavities, ELM and EZ integrity. Macula-off duration was also noted and was defined as the time from the onset of central vision loss to surgery.
Intraoperative data included the type of surgery and type of tamponade.
Postoperative data were the BCVA and SD-OCT findings including outer retinal band (ELM, EZ and CIZ) integrity and presence of ERM, intra-retinal or subretinal fluid (SRF).
Macular imaging with SD-OCT
Microstructural imaging analysis of the fovea was performed using Spectralis HRA-OCT (Heidelberg Engineering Spectralis, Heidelberg, Germany). The protocol consisted of two high-resolution horizontal and vertical scans, ART5, of 6 mm and a Volume scan using 25 equally spaced horizontal B-scans centred on the fovea and covering an area of 20° horizontally and vertically.
The height of the macular detachment was manually measured between the inner border of the retinal pigment epithelium and the outer border of the backreflection at the fovea using the digital callipers provided by Heidelberg Spectralis OCT software. When it was not possible to evaluate the height of the macular detachment because of a high RD, a value of 1500 µm was considered for analysis.
The three hyperreflective outer retinal bands (ELM, EZ and CIZ) were defined according to the classification proposed by the International Nomenclature for Optical Coherence Tomography panel.18 The integrity of the foveal ELM, EZ, and CIZ was assessed as follows: line not visible or disrupted in at least one scan (band defect) and continuous line in both the horizontal and vertical scans (intact band). The same analysis was performed in the fellow eyes at each follow-up visit to minimise potential bias related to artefacts.
For statistical analysis, ORL status was divided into four subgroups based on the integrity of each of the three hyperreflective bands: intact CIZ, EZ and ELM, CIZ defect with intact EZ and ELM, combined CIZ and EZ defects with intact ELM, and combined CIZ, EZ and ELM defects (figure 1).
When postoperative SD-OCT was available at both 1 and 6 months, the restoration of each outer retinal band was studied. ELM, EZ or CIZ was considered as recovered if the band was either invisible or disrupted at 1 month and intact at 6 months on the same follow-up scans.
All qualitative OCT evaluations and measurements were performed independently by two masked readers (LM and JBC), both unaware of patients’ visual outcomes. When there was a disagreement, a third investigator was consulted for the final decision (JPB).
Main outcome measures
The primary endpoint of the study was to determine the prevalence of outer retinal band defects at 6 months and their associated preoperative risk factors. We also analysed the correlation between ORL integrity and visual acuity.
Statistical analysis was performed using SAS (V.9.4, SAS Institute, Cary, NC, USA). Snellen visual acuity was converted to logarithm of minimal angle of resolution (logMAR) units for analysis. Continuous variables were expressed as mean±SD and categorical variables were expressed as numbers and percentages.
A bivariate logistic regression was used to estimate the relationship between the potential influencing preoperative factors and the number of alterations within the ORL at 6 months after RRD surgery. Variables with p<0,10 in bivariate logistic regression were included in the final multivariate logistic regression. Collinear variables were excluded in the final model.
The Kruskal-Wallis test was carried out to analyse differences in BCVA according to the number of alterations within the ORL at 6 months. The interrater agreement was compared using a weighted κ-coefficient. Statistical significance was set at p<0.05.
During the study period, 918 patients underwent surgery for RRD. Of these, 665 were excluded for the following reasons: macula-on RRD (n=257), postoperative retinal redetachment (n=156), preexisting macular pathologic features (n=54) and absence of or low-quality postoperative SD-OCT images (n=198). As a result, 253 eyes of 253 patients were included in the study.
Baseline characteristics and intraoperative data
Baseline characteristics and intraoperative data are given in table 1.
Of the 253 patients, 156 (61.7%) were men and 97 (38.3%) were women, with a mean age of 61.8±11.0 years at the onset of RRD. The mean preoperative BCVA was 1.7±0.6 logMAR. The mean extent of RD was 2.7±0.8 quadrants and was associated with grade B or C PVR in 45.7% of cases. The mean macula-off duration was 10.1±16.0 days with a median of 5.0 days (1–90). At the end of the follow-up, 205 (81.0%) eyes were pseudophakic.
Prevalence of outer retinal band defects at 1 and 6 months
Prevalence of outer retinal band defects and distribution of the different subgroups are listed in table 2.
At 1 month, SD-OCT was available in 212 of 253 eyes. CIZ, EZ and ELM defects were found in, respectively, 198 (93.4%) eyes, 100 (47.2%) eyes and 64 (30.2%) eyes. At 6 months, CIZ, EZ and ELM were absent or discontinued in, respectively, 160 (63.2%) eyes, 44 (17.4%) eyes and 18 (7.1%) eyes.
The other postoperative SD-OCT findings were the presence of ERM in 35 (13.8%) eyes, macular oedema in 31 (12.3%) eyes with a mean central macular thickness of 425.4±93.8 µm and SRF in 23 (9.1%) eyes.
The interrater agreement for the assessment of outer retinal band integrity was excellent, with κ ranging from 0.88 to 1.00 at both 1 and 6 months.
Preoperative risk factors for ORL damage at 6 months
Results of bivariate and multivariate regression analysis are displayed in table 3a and b.
In the bivariate analysis, preoperative BCVA (p=0.010), macula-off duration (p=0.007), PVR grade (p=0.001), extent of RRD (p=0.003) and height of macular detachment (p=0.020) were significantly associated with the number of defects within the ORL at 6 months.
Preoperative SD-OCT findings were excluded from the multivariate model because of the shortage of available data. Collinearity between variables was tested and collinear variables were excluded in the final model (PVR grade and extent of RRD). After adjusting for influential variables, only macula-off duration and preoperative BCVA showed significant correlation with the number of defects within the ORL (p=0.007 and p=0.030). The odds of having two or three band defects versus one and none increased, respectively, by a factor of 1.03 (95% CI, 1.01 to 1.06) and 1.04 (95% CI, 1.01 to 1.08) per day of macular detachment. As regards the BCVA, patients with preoperative BCVA >1.7 logMAR had a 2.53 higher risk of having one band affected (95% CI, 1.25 to 5.13).
Interestingly, restoration of the CIZ and the EZ integrity was significantly correlated with the number of affected bands at 1 month (p<0.001 and p=0.041). Eyes with combined ELM, EZ and CIZ defects at 1 month had a fivefold increased risk of not restoring the EZ compared with those that had only EZ and CIZ defects (p=0.041, 95% CI, 0.0 to 0.9). Similarly, the probability of restoring the CIZ was 5.9-fold higher in eyes with an isolated CIZ defect at 1 month compared with those that had EZ, ELM and CIZ defects (p<0.001, 95% CI, 2.3 to 14.8). These results suggest that ELM integrity is a significant factor to consider in predicting the chances of subsequent EZ and CIZ recovery.
Functional outcomes and correlation between ORL integrity and final visual acuity
The mean BCVA significantly improved through the 6-month follow-up from 0.5±0.3 logMAR at 1 month to 0.3±0.3 logMAR at 6 months (p<0.001). At 6 months, visual acuity was strongly associated with the number of affected bands within the ORL (p<0001) (table 4).
Despite successful retinal reattachment, visual outcomes may still be disappointing in eyes that have suffered from macula-off RRD. With the introduction of SD-OCT, many authors have shown that damage to the outer photoreceptor layers represents a major cause of poor vision following macula-off RRD.7–11 However, to the best of our knowledge, no study has focused on factors that might affect their integrity.
In this study, we sought to determine the prevalence of ORL damage after successful macula-off RRD surgery and potential preoperative risk factors.
CIZ, EZ and ELM defects were, respectively, found in 93.4%, 47.2% and 30.2% of cases at 1 month and in 63.2%, 17.4% and 7.1% of cases at 6 months. These findings are concordant with those of previous studies which reported disruption of the CIZ, the EZ and the ELM in, respectively, 98.9%, 37.2% to 77.6% and 10.0% to 38.9% of cases at 1 month and in 45.6% to 65.0%, 13.9% to 70.0% and 5.6% to 55.0% of cases at 6 months.7–12 Nonetheless, it is very difficult to make comparisons across studies given the wide variations in terms of design, sample size and preoperative characteristics of RRD.
In the bivariate analysis, preoperative BCVA, extent of RRD, PVR grade, macula-off duration and height of macular detachment were significantly associated with the number of defects within the ORL. This is not surprising as all these factors have already been related to visual recovery after macula-off RRD surgery.11 19–22 However, after adjusting for influential variables and collinearity, two factors—macula-off duration and preoperative BCVA—were independently associated with ORL damage. It should be noted that preoperative SD-OCT findings were not included in the multivariate model, which has certainly influenced the result regarding BCVA. Indeed, preoperative BCVA has already been associated with the height of retinal detachment and retinal microstructures.11 23 Therefore, we assume that duration of macular detachment is the main factor influencing ORL integrity. We found that the odds of having several affected bands substantially increased with the duration of RRD. This result confirms Joe et al’s study, in which shorter delays to surgery were associated with better recovery of EZ and ELM.23 Similarly, Narala et al demonstrated that photoreceptor volume was positively correlated to time to surgical repair, while Saleh et al, in an adaptive optics imaging study, showed a significant correlation between cone density and duration of RRD.15 24
These findings are in line with those of experimental studies. Indeed, animal models have shown that RD causes photoreceptor cell death by apoptosis and that apoptosis occurs within hours, peaks at 2 to 3 days, and drops to a low level 7 days after RD induction.4 5 This time course may also explain the impact of macula-off duration on visual outcomes. Most initial studies defined 7 days as a cut-off after which final visual acuity was not affected by further delay of surgery.25–27 However, two recent studies have reported that eyes treated within 3 days of macular detachment achieve better visual acuity than those with longer time lag.28 29 More recently, Greven et al have indicated that even within the first 3 days, the duration of macular detachment affects the long-term visual outcomes.30 Taken together, these results emphasise the importance of early surgical repair of macula-off RRD.
After successful reattachment, many SD-OCT reports have provided evidence suggesting the photoreceptor’s ability to recover over time.7 12–15 Consistent with previous studies, prevalence of outer retinal band defects substantially decreased from 1 month to 6 months after surgery in this series. Interestingly, we found that restoration of the CIZ and the EZ was strongly associated with the number of affected bands within the photoreceptor layers at 1 month. Eyes with intact ELM had greater chances of recovering their EZ and CIZ compared with those with initial ELM defect. Similarly, eyes with intact ELM and EZ were more likely to completely recover than those with ELM, EZ and CIZ defects. These results are consistent with those of prior reports, which found that photoreceptor recovery depended on the integrity of the ELM.15 16 31–33 Indeed, ELM is located at the boundary between the cell bodies and the inner segments of the photoreceptors and comprises clusters of junctional complexes and microvilli between the Müller cells and the photoreceptors.34 The presence of ELM defect may therefore indicate damage or loss of both cell types, which may compromise the regeneration process.10 11 16 32
Finally and importantly, we assessed the correlation between ORL integrity and visual acuity. BCVA improved significantly through the study period, along with photoreceptor recovery. At 6 months, it was strongly associated with the number of affected bands within the ORL. This is consistent with the studies by Wakayabashi et al and Lai et al, which revealed that combined ELM and EZ and combined ELM, EZ and CIZ disruptions resulted in poorer visual outcomes.9 10 Overall, this finding, in conjunction with the correlation between ORL damage and duration of macular detachment, further confirms the importance of considering early surgery in patients with macula-off RRD.
We acknowledge several limitations to the present study, mainly due to its retrospective design. Preoperative SD-OCT images were not available in all patients and we were not able to enter these data into the multivariate regression model, especially the height of RD. Moreover, alterations of the outer retinal bands were assessed only on a qualitative basis. Decreased band reflectivity due to fragmentation or thinning of the band itself may indeed occur in eyes without any abnormalities.35 Although we excluded patients with decreased band reflectivity in the fellow eye, it is possible that some of the alterations observed in the eyes studied were the result of artefacts. Finally, another factor that may have affected our outcomes is the difficulty in accurately determining the duration of macular detachment. Even though this is a highly subjective variable, the onset of central vision loss that we used as a marker for foveal detachment is often well identified by the patients in recent cases of RRD.
In conclusion, prevalence of outer retinal band defects substantially decreased through the study period, confirming the ability of photoreceptors to recover over time. However, shorter interval to surgery and better visual outcomes were significantly associated with fewer defects within the ORL at 6 months. These findings suggest that earlier surgery may limit RRD-associated photoreceptor degeneration and improve the patient’s visual prognosis.
Contributors Substantial contributions to conception and design: LM, JBC, HR, CB and JPB; acquisition of data: LM and JBC; writing the statistical analysis plan: HR and CB; analysis and interpretation of data: LM, JBC, HR, CB and BL; drafting the article: LM and JBC; revising it critically for important intellectual content: JPB, KAD and FS; and final approval of the version to be published: all authors.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The study adhered to the tenets of the Declaration of Helsinki and the protocol was approved by the Ethics committee of the French Society of Ophthalmology.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.