Epiretinal Membrane Surgery Outcomes in Eyes with Subretinal Drusenoid Deposits

Epiretinal Membrane Surgery Outcomes in Eyes with Subretinal Drusenoid Deposits A Case Control Study Craig Wilde, FRCOphth, PhD,1 Mary Awad, MB ChB,1 Harminder Dua, FRCOphth, PhD,1 Ravi Gandhewar, MB BS, FRCS,2 Hean-Choon Chen, MB BS, FRCOphth,2 Winfried M. Amoaku, FRCOphth, PhD1 Purpose: To evaluate outcomes of epiretinal membrane (ERM) surgery in eyes with subretinal drusenoid deposits (SRDDs) and to compare them with those with isolated ERM. Design: Retrospective case-control study of consecutive patients who underwent pars plana vitrectomy (PPV) with ERM peeling. Participants: Twenty-five patients with SRDDs on spectral-domain (SD)-OCT who underwent surgery for ERM were included in the study. From the same cohort, for each case, we selected 2 age-matched control participants (50 eyes with isolated ERM). Preoperative best-corrected visual acuity (BCVA) also was matched as closely as possible. Methods: All participants underwent PPV and ERM peel for primary ERM. Main Outcome Measures: Postoperative BCVA, improvement in BCVA, preoperative and postoperative central macular thickness, surgical complications, and development of age-related macular degeneration (AMD) were recorded. Results: At final examination, mean postoperative BCVA was significantly less for eyes with SRDDs (0.51 logarithm of the minimal angle of resolution [logMAR] vs. 0.21 logMAR; P ¼ 0.0001). Eyes with SRDDs demonstrated less improvement in BCVA after ERM surgery (0.13 logMAR vs. 0.30 logMAR; P ¼ 0.0032). Eyes with SRDDs were significantly less likely to gain 2 or more Snellen lines of BCVA after ERM surgery (28% vs. 56%; P ¼ 0.028). Three of 25 patients (12%) undergoing ERM surgery showed worsening of Snellen BCVA by 2 lines or more. Three of 25 eyes (12%) with SRDDs demonstrated advanced AMD after surgery, compared with 0 participants in the control group (P ¼ 0.034). Conclusions: Epiretinal membrane surgery in eyes with SRDDs is associated with less favorable visual outcomes. Fewer patients demonstrate gain in BCVA, whereas a significant number show a deleterious decline. After surgery, AMD incidence seems high and patients may have an increased risk of raised intraocular pressure. These findings require further study to establish whether this represents a causal relationship. Surgeons should be vigilant for these complications. Appropriate patient counseling during the consenting process must be made. Ophthalmology Retina 2018;-:1e9 Crown Copyright ª 2018 Published by Elsevier Inc. on behalf of the American Academy of Ophthalmology

Surgery for idiopathic epiretinal membranes (ERMs) has become a common procedure since the first 6 successful cases were reported by Machemer1 in 1978. Prevalence of ERMs, based on color fundus photographs, is reported to be between 6% and 19% across population-based studies.2e5 As well as being relatively common, they are a significant pathologic condition, resulting in impairment in quality of life and, in a minority of cases, in a severe reduction in vision.6,7 Surgical outcomes are well documented, with 70% to 80% of patients achieving an improvement in best-corrected visual acuity (BCVA) of 2 or more Snellen lines.8e12 However, most studies included idiopathic ERM in otherwise healthy eyes, with surgical outcomes in eyes with comorbidities, such as high Crown Copyright  2018 Published by Elsevier Inc. on behalf of the American Academy of Ophthalmology

degrees of myopia13 and retinitis pigmentosa,14 being less well documented. Because ERMs are predominantly the consequence of physiologic posterior vitreous detachment (PVD), they occur mainly in the elderly,5 and consequently, it is not infrequent to identify patients with comorbid agerelated maculopathy. To date, only a limited number of studies specifically have assessed the outcome of ERM surgery in eyes with age-related maculopathy, and the reported results are conflicting.15,16 Furthermore, no literature exists on the outcome of surgery in eyes with reticular pseudodrusen (RPD), now referred to as subretinal drusenoid deposits (SRDDs). The epidemiologic features of SRDDs differ from those of age-related maculopathy and conventional drusen. They https://doi.org/10.1016/j.oret.2018.06.009 ISSN 2468-6530/18

1

Ophthalmology Retina Volume -, Number -, Month 2018 occur more frequently in women,17 impose greater risk for patients to progress to age-related macular degeneration (AMD) when compared with drusen,17,18 have been shown to be associated with reduced retinal sensitivity with electrodiagnostic testing,19 and recently have been associated with visual dissatisfaction when adjusted for age, lens opacities, and grade of AMD.20 They are a frequent finding in eyes with age-related choroidal atrophy (ACA)21 and are associated with choroidal ischemia and thinning.22 They may represent a distinct fundus feature of pathologically thin choroid and outer retinal ischemia, rather than being a manifestation of AMD or a variant of it. Therefore, it is important to assess whether their presence influences the outcome of ERM surgery. The purpose of this study was to evaluate in a large series of patients anatomic and functional outcomes of ERM surgery in eyes with comorbid SRDDs and to compare them with age-matched control participants.

Methods Patients and Study Design This was a retrospective case-control design in a cohort of consecutive patients who underwent pars plana vitrectomy (PPV) and ERM peel for primary ERM between December 2014 and December 2017. Surgery was performed under the care of 2 vitreoretinal surgeons (H.-C.C. and R.G.) at the Royal Derby Hospital (Derby, United Kingdom). Patients were identified from review of operating theater records. Analysis of spectral-domain (SD)-OCT images was undertaken by 2 ophthalmologists (C.W. and M.A.) to identify patients with SRDDs and ERM, without reference to patient notes, including postoperative outcomes. To minimize confounding bias, for each patient, 2 age-matched control participants (age on day of surgery within 12 months), who were patients with ERM without SRDDs were selected. Surgery for patients and control participants was performed within 1 year of each other. If age-matched control participants were not available, which occurred only for the oldest patient (94 years), then the next oldest control participant was chosen. We aimed to match patients and control participants for lens status (pseudophakic, phakic, or combined surgery performed). Pseudophakic individuals were matched for preoperative BCVA within 1 Snellen line. If there were more than 2 potential age-matched control participants, preoperative visual acuity was matched as closely as possible, acting as a determining factor for control selection. Equal numbers of surgeries under the care of each consultant were chosen. Exclusion criteria included eyes with secondary ERM after vitrectomy for retinal detachment, previous retinal vein occlusions,

posterior uveitis, diabetic maculopathy, preoperative geographic atrophy or neovascular AMD (nAMD), and amblyopia. Any patient with permanent structural damage to the fovea was excluded. Ethics committee approval was not required for this study, and it adhered to ethical principles outlined in the Declaration of Helsinki.

Image Grading Images were graded in a masked side-by-side fashion with open discussion. If disagreement existed, one of the authors (W.M.A.) adjudicated and acted as final arbiter. Observers were unaware of patient demographics, including age, preoperative or postoperative BCVA, comorbidities, or surgical complications. Subretinal drusenoid deposits were graded if there were 5 or more discrete hyperreflective collections in the subretinal space, above the retinal pigment epithelium, being sufficient to alter the contour of the ellipsoid zone in a saw-tooth pattern on SD-OCT (Fig 1), as described previously by others.23 Images were acquired using the Spectralis SD-OCT (Heidelberg Engineering, Heidelberg, Germany). The image grading protocol consisted of a volume scan with 25 images, each comprising 9 frames at a scan angle of 20 and a 1024-pixel resolution. Some participants underwent additional imaging using the 3-dimensional OCT1000 instrument (Topcon, Tokyo, Japan). The field of view used was a 6  6-mm area centered on the fovea. A raster scan consisting of 128 frames was performed for each eye. All frames were reviewed for the presence of SRDDs. Given that SRDDs using color fundus photographs are bilateral 80% of the time20 and that ERM and macula edema cause disruption, artifacts, and shadows within the outer retina, to ensure accurate grading, SRDDs had to be present in both eyes. Isolated outer retinal changes in eyes with ERM were not deemed sufficient for inclusion in the study without typical contralateral SRDDs.

Surgical Procedure The decision to perform surgery was based on the severity of symptoms and individual patient requirements. Fully informed written consent was obtained. All surgeries were undertaken with accepted techniques, which were similar between the 2 surgeons, as described below. Part or all of some surgeries may have been performed by other vitreoretinal surgeons in training, under strict supervision, using the same techniques as used by the 2 primary surgeons. Either 23- or 25-gauge PPV was performed, with the department transitioning to 25 gauge halfway through the study period. If a PVD was absent, one would be induced with active aspiration of the posterior vitreous hyaloid with the vitrectomy probe adjacent to or above the optic disc. If PVD induction was difficult, visualization of the posterior hyaloid may have been

Figure 1. Spectral-domain OCT images illustrating subretinal drusenoid deposits creating a saw-tooth pattern of elevation of the ellipsoid zone.

2

Wilde et al



ERM Surgery Outcomes in Eyes with SRDD

enhanced with MembraneBlue-Dual (DORC, Zuidland, The Netherlands), if it was deemed necessary. Staining of ERM with MembraneBlue-Dual (for up to 1 minute) was performed in all patients, with further application as required, for instance, with poor initial staining or if the edge of a membrane were lost. A retractable diamond-dusted membrane scraper (Bausch and Lomb, Rochester, NY) was used primarily to initiate ERM peel by creating a tear and flap edge in the ERM, which then was grasped with end-gripping forceps (Bausch and Lomb). Peeling was performed ideally in a continuous sheet from arcade to arcade and optic disc nasally. If the ERM had a raised edge, peeling would be initiated with end-gripping forceps. Internal limiting membrane was not intentionally peeled.

Outcome Data Data were acquired from case notes review (after masked grading of SD-OCT images), including age, gender, lens status, comorbid ocular diseases, and surgical complications. Patients underwent mydriatic ophthalmic examination before and after surgery. Snellen visual acuity was obtained for each eye uncorrected, with glasses, and pinhole. For study purposes, BCVA was used for analysis and Snellen visual acuity was converted to logarithm of the minimum angle of resolution (logMAR). Postoperative visits varied according to clinical need, with appointments generally arranged on day 1, 2 weeks, 1 month, and 3 months, with further visits on a clinically needed basis. Results of clinical assessments and outcomes were reviewed from each visit in a longitudinal manner. Primary outcomes measured included postoperative BCVA and change in BCVA in comparison with baseline. Secondary outcomes were mean change in central macular thickness, postoperative complication rates, and the number of eyes achieving BCVA of 6/9 or more within each group. Spectraldomain OCT images were acquired before and after surgery at each clinic. For study purposes, the last available SD-OCT image was used for calculations and measurements.

Statistical Analysis Results were presented as proportions for categorical variables and mean  standard deviation for continuous variables. Student’s t test was used to compare continuous data. The unpaired t test was used to evaluate and compare baseline differences between patients and control participants for continuous variables, including baseline logMAR BCVA, SD-OCT macular thickness, and period of follow-up. Mean postoperative BCVA, mean change in BCVA after surgery, and change in SD-OCT macular thickness after surgery were compared with the unpaired Student t test. For binary outcomes, including incidence of AMD or a gain of 2 or more Snellen lines of BCVA, the Fisher exact test was used for intergroup comparisons of proportions. P values of 0.05 or less were considered statistically significant. Statistical analysis was performed using Statistical Package for the Social Sciences for Mac version 24 (SPSS, Inc., Chicago, IL).

Results Baseline Characteristics of Patients A total of 25 patients with ERM and comorbid SRDDs was identified. The mean age of patients was 80.44 years (standard deviation, 6.62 years; range, 63e94 years). Fourteen patients (56%) were women. The mean preoperative BCVA was 0.64 logMAR (standard deviation, 0.2 logMAR). Mean preoperative average central macular thickness was 483.2 mm (standard

deviation, 95.5 mm). Twelve of 25 patients (48%) underwent combined vitrectomy, ERM peel, and cataract surgery. Eight patients (32%) were pseudophakic at the time of surgery, whereas 5 patients (20%) were phakic at the time of vitrectomy and ERM peeling and remained phakic at the termination of the study.

Comparison of Baseline Characteristics of Patients and Control Participants Fifty age-matched patients were selected from the cohort as described previously. Table 1 draws comparison between baseline characteristics of patients and control participants. The mean preoperative BCVA for patients with ERM and SRDDs was 0.64  0.2 logMAR, being slightly worse than that in the control group (0.51  0.24 logMAR; P ¼ 0.034) with isolated ERM. The mean age of patients (80.44  6.62 years) was slightly older than the mean age for control participants (78.46  7.36 years), but the difference was not statistically significant (P ¼ 0.260). Preoperative maximum and mean central retinal thickness did not differ between patient and control groups (Table 1). The mean duration of follow up was 6.04  6.1 months for patients and slightly less for control participants (4.53  3.44 months). The difference was not statistically significant (P ¼ 0.174). Fewer control participants at baseline were pseudophakic (24%) when compared with patients (32%), but the difference did not reach statistical significance (P ¼ 0.580).

Visual Outcomes Five of 25 eyes (20%) with SRDDs and ERM showed worsening of BCVA after surgery. Three of these eyes were pseudophakic, with 1 of these eyes showing a reduction in BCVA of 0.3 logMAR associated with outer retinal atrophy (ORA) that developed after surgery. The other 2 pseudophakic eyes with poor outcomes showed loss of vision secondary to development of a pigment epithelial detachment and the other showed features of ORA (Fig 2). In the 2 phakic patients in this subgroup, the reduced vision was deemed secondary to cataract development. One patient subsequently was listed for cataract surgery, whereas the second declined and was discharged. Four of 50 control eyes (8%) with isolated ERM showed worsening of preoperative BCVA. Two of these eyes (4%) with worsening vision were phakic and 1 eye was re-referred with postvitrectomy cataract development. No pseudophakic control eye showed worsening of BCVA after isolated vitrectomy and ERM peeling. At the final clinic examination, the 2 groups showed statistically significant differences in final postoperative BCVA and improvement in BCVA (Table 2). The mean BCVA for patients was 0.510.24 logMAR, which was worse than that for control participants (0.210.18 logMAR; P ¼ 0.001). At final examination, mean change in BCVA was þ0.130.24 logMAR for the patients (corresponding to a mean gain of 6.5 letters) and þ0.300.22 logMAR for the control group (P ¼ 0.0032), corresponding to a mean gain of 15 letters. When stratified for duration of follow-up available, mean BCVA outcomes (across all available follow-up durations) were worse for patients when compared with control participants (0e<3 months, 0.490.29 logMAR vs. 0.270.17 logMAR; 3e<6 months, 0.570.2 logMAR vs. 0.130.18 logMAR; 6 months, 0.510.22 logMAR vs. 0.200.16 logMAR).

3

Ophthalmology Retina Volume -, Number -, Month 2018 Table 1. Baseline Characteristics of Eyes with Epiretinal Membrane and Subretinal Drusenoid Deposits (Patient Group) versus Eyes with Isolated Epiretinal Membrane (Control Group) Characteristics Gender, no. (%) Male Female Mean age (SD), yrs Preoperative maximum central retinal thickness (mm) Mean (SD) 95% CI Range Preoperative average central retinal thickness (mm) Mean (SD) 95% CI Range Preoperative BCVA (logMAR) Mean (SD) 95% CI Range Preoperative lens status, no. (%) Pseudophakic Phakic Phakic but had a combined procedure Laterality, no. (%) Right Left Follow-up (mos) Mean (SD) 95% CI Range

Patient Group (n [ 25)

Control Group (n [ 50)

11 (44) 14 (56) 80.44 (6.62)

17 (34) 33 (66) 78.46 (7.36)

537.2 (103.2) 493.7e580.8 364e845

532.1 (140.8) 491.7e572.6 330e975

483.2 (95.5) 442.9e523.5 311e775

470.0 (94.8) 442.7e497.1 268e745

0.64 (0.20) 0.55e0.72 0.3e1

0.51 (0.24) 0.45e0.58 0e1

8 (32) 5 (20) 12 (48)

12 (24) 7 (14) 32 (64)

14 (56) 11 (44)

26 (52) 24 (48)

6.04 (6.1) 3.52e8.56 0.25e22

4.53 (3.44) 3.55e5.51 0.5e12

P Value 0.45* 0.260y 0.876y

0.576y

0.034y

0.580* (pseudophakic vs. phakic)

0.810*

0.174y

BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; ERM ¼ epiretinal membrane; logMAR ¼ logarithm of the minimum angle of resolution; SD ¼ standard deviation. *Fisher exact test. y Student t test.

No control eyes showed worsening of BCVA by 2 or more Snellen lines. Among the patients, 3 of 25 eyes (12%) showed a reduction in BCVA by 2 lines or more, including 1 phakic eye in which cataract developed that was clinically responsible for BCVA reduction. The remainder of eyes with poor visual outcome demonstrated ORA, with 1 of these eyes showing features of choroidal atrophy. No significant difference in anatomic outcomes in terms of mean or maximum central retinal thickness was found between the patient and control groups (Table 2). Four of 25 patients (16%) achieved a BCVA of 6/9 or better after surgery, and none

achieved a final BCVA of 6/6 or better. In the control group, 29 of 50 participants (58%) achieved a postoperative BCVA of 6/9 or better, and 11 of the 50 control participants (22%) achieved a final BCVA of 6/6 or better.

Postoperative Complications Three of 25 patients (12%) demonstrated advanced AMD during the follow-up period, including 1 case of foveal-involving geographic atrophy and 2 cases of nAMD. One of these patients was 94 years of age, and during the 14-month follow-up period,

Figure 2. Spectral-domain OCT image showing a small area of outer retinal atrophy.

4

Wilde et al



ERM Surgery Outcomes in Eyes with SRDD

Table 2. Anatomic and Visual Outcomes after Epiretinal Membrane Surgery in Eyes with and without Subretinal Drusenoid Deposits at the Final Visit Outcome Measure Maximum central macula thickness (mm) Mean (SD) Range 95% CI Average central macula thickness (mm) Mean (SD) Range 95% CI BCVA (logMAR) Mean (SD) Range 95% CI Mean change in BCVA (logMAR) Mean (SD) Range 95% CI Standard error mean BCVA, no. (%) Improved 2 lines Unchanged within 2 lines Worsened 2 lines AMD developed after surgery, no. (%) OHT, glaucoma, or glaucoma suspect, no. (%)

Patient Group (n [ 25)

Control Group (n [ 50)

430.1 (64.2) 341e576 403.0e457.2

439.0 (57.1) 296e577 422.6e455.4

370.1 (60.8) 275e543 344.5e395.8

392.8 (53.7) 190e488 377.4e408.3

0.51 (0.24) 0.18e1.0 0.41e0.61

0.21 (0.18) e0.2 to e0.6 0.16e0.26

0.13 (0.24) e0.4 to þ0.54 0.03e0.23 0.049

0.30 (0.22) e0.12 to þ0.82 0.23e0.36 0.033

P Value 0.55*

0.11*

0.0001*

0.0032*

7 15 3 3 5

(28) (60) (12) (12) (20)

28 22 0 0 3

(56) (44) (0) (0) (6)

0.028y (improved BCVA 2 lines vs. unchanged or worse)

0.034y 0.108y

AMD ¼ age-related macular degeneration; BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; logMAR ¼ logarithm of the minimum angle of resolution; OHT ¼ ocular hypertension; SD ¼ standard deviation. *Student t test. y Fisher exact test.

she also demonstrated nAMD in the contralateral eye that had not undergone surgery. In comparison, no patients in the control group demonstrated either geographic atrophy or nAMD. The difference in incident AMD was deemed statistically significant (P ¼ 0.034). The third patient with postoperative incident AMD demonstrated nAMD. Initially, BCVA for this patient improved after combined PPV, ERM peel, and phacoemulsification plus intraocular lens from 6/24 to 6/12 at 1 week after surgery. One year later, after being lost to follow-up, the patient demonstrated a disciform scar with BCVA of 1.2 logMAR. For the purpose of this study, the final BCVA was recorded as 6/12 for analysis. One patient (4%) demonstrated a postoperative pigment epithelial detachment, but this remained stable in appearance and underwent observation and no treatment. Two patients (8%) demonstrated ORA. Five of 25 patients (20%) demonstrated either postoperative ocular hypertension (OHT) or glaucoma or were deemed a glaucoma suspect. Three of 50 control participants (6%) demonstrated postoperative OHT. The higher rate of incident OHT or glaucoma between the 2 cohorts was not deemed significant (P ¼ 0.108). One patient (4%) demonstrated cystoid macular edema after surgery and 1 control participant also demonstrated cystoid macular edema (2%).

Discussion Our surgical results and baseline demographics for control patients are similar to previous reports regarding outcomes of ERM surgery.24 In a study by Kim et al25 assessing

long-term changes in visual outcome after PPV for idiopathic ERM, mean preoperative BCVA was 0.49 logMAR, which is almost identical to that of our control group (0.51 logMAR). Mean gain in BCVA was similar, with improvement of up to 0.35 logMAR at 24 months after vitrectomy. This study for the first time demonstrated that patients with comorbid ERM and SRDDs are less likely to achieve good visual outcomes after PPV and ERM peeling. They are less likely to achieve good BCVA or to demonstrate improvement in BCVA. They are less likely to gain 2 or more Snellen lines after surgery when compared with control participants with isolated ERM. Only a minority of patients (28%; patients with SRDDs and ERM) will gain 2 or more Snellen lines after surgery, which is half as many when compared with patients with isolated ERM (56%). Moreover, 25% demonstrate worsening of BCVA after surgery, and 12% lose 2 or more Snellen lines. Our findings are not surprising, having previously reported that RPD are associated with visual dissatisfaction.20 Reticular pseudodrusen have been associated with ACA,21 which is recognized to cause mild visual impairment. Furthermore, RPD have been associated with reduced retinal sensitivity.19 This raises the question of whether patients with SRDDs and ERM are indeed symptomatic, and to what degree, from both pathologic processes. Visual dysfunction, dissatisfaction, and reduced BCVA before surgery could result from ACA, RPD, and associated

5

Ophthalmology Retina Volume -, Number -, Month 2018 outer retinal changes being attributed falsely in some cases to comorbid ERM. Further study with patient-reported outcomes is warranted. The integrity of the inner segment ellipsoid (ISel) consistently has been reported to correlate with preoperative and postoperative BCVA.25e28 It is not surprising that patients with SRDDs show worse visual outcomes after ERM peeling, because SRDDs cause alterations in the ISel.23,29 Regression of SRDDs has been demonstrated to cause loss of the ISel and to be associated with visual loss.29,30 Subretinal drusenoid deposits have been associated with outer retinal ischemia. Their presence likely represents loss of retinal pigment epithelium, photoreceptor, and choriocapillaris structure and function,31e33 limiting potential for improvement in BCVA after ERM peeling. The influence of vitrectomy and ERM peeling on the evolution of such changes remains to be elucidated. We note that 2 of our patients demonstrated significant reduction in vision associated with incident ORA. Complement fragments and other immune system constituents have been found within SRDD.34 Some authors postulate that inflammation plays a role in the cause.17,35 We cannot conclude whether postoperative inflammation enhanced the rate of progression of the outer retinal changes. Outer retinal changes associated with ERM can be difficult to determine in the presence of gross edema or retinal thickening (Fig 3). We emphasize the importance of examining contralateral eye SD-OCT images for SRDDs because they are predominantly bilateral. Their presence in the contralateral eye should raise caution for the vitreoretinal surgeon. It may be that in some cases, the outer retinal changes become more apparent after surgery, with resolution of edema. Previous ERM surgery studies have been imprecise in their classification of disruption to the ISel, with defects often being categorized crudely as

simply intact, disrupted, missing, or other similar variations. Further studies are warranted to subcategorize the exact nature of disruption and appropriate risk stratification for various defects. Interestingly, we demonstrated a greater incidence of AMD during the follow-up period for eyes with SRDDs that undergo ERM peeling when compared with eyes of age-matched control participants with isolated ERM. Choroidal neovascularization (CNV) development after vitrectomy and ERM removal is a known complication.36e39 The exact incidence is unknown, but it is certainly rare, with Sandali et al40 reporting no case in a retrospective series of 909 surgical patients. Our finding of CNV occurring in 10% of eyes with SRDDs after ERM peeling warrants attention and demands specific mention during any consent process. The vitreoretinal surgeon should be aware of this complication and should have a low threshold for using fundus fluorescein angiography in patients with increasing postoperative intraretinal or subretinal fluid, allowing prompt recognition and treatment. It is known that eyes with SRDDs and RPD are at high risk of AMD developing.20 Whether the risk is increased by ERM peeling cannot be elucidated from this study, but the temporal relationship with incident CNV after surgery suggests a direct relationship. We hypothesize that eyes with SRDDs potentially are more susceptible to mechanical trauma during ERM peeling, having an unhealthy retinal pigment epitheliumeBruch’s membraneechoriocapillaris complex. Tractional forces during peeling allow breaks to form in Bruch’s membrane, permitting development of CNV. This has been postulated previously in eyes without SRDDs,41e43 but our series identified a clear group of patients who are at risk of this rare occurrence. It is possible that inflammation after surgery may stimulate vascular endothelial growth factor release, precipitating CNV development in predisposed eyes, because it is

Figure 3. A, B, Spectral-domain OCT images of eyes with an epiretinal membrane and significant edema. These illustrate that is it not possible to grade the presence or absence of outer retinal changes or subretinal drusenoid deposits (SRDDs) accurately in such eyes. Both patients had typical contralateral SRDDs.

6

Wilde et al



ERM Surgery Outcomes in Eyes with SRDD

known that vitrectomy increases intraocular cytokine levels.44 It may be suggested that eyes with SRDDs are susceptible to phototoxicity from endoillumination, a feature of which is outer retinal changes, although there is no evidence to support this.45 Alternatively, it may be that subclinical CNV was present before ERM peeling. However, no clinical or SD-OCT features suggested this. It is unknown whether histologic or mechanical properties of ERM or posterior hyaloid and prevalence of PVD are different in eyes with SRDDs. We highlight that several secondary membranes occur in association with pathologic features that cause retinal ischemia, including retinal vein occlusions, retinitis pigmentosa, or diabetes mellitus. Choroidal ischemia and vascular insufficiency increasingly are recognized features of eyes with SRDDs. Whether ERM develops secondary to such ischemia and is more prevalent in eyes with SRDDs cannot be elucidated. A more perplexing and unexpected finding is the increased incidence of OHT and glaucoma in patients with SRDDs after ERM surgery. This finding likely reflects chance, given the small numbers involved and lack of statistical significance. However, eyes with ACA (which is associated with SRDDs), are known to have a high glaucoma prevalence.21 Eyes with SRDDs have atrophic avascular choroids, and it may be postulated that these changes result in glaucoma or OHT via reduced capacity for uveoscleral outflow that may raise intraocular pressure after surgery.46 Our findings have important clinical ramifications. The indication and decision to perform ERM peeling is not standardized, and outcomes vary among patients. Predicting outcome is essential for informed consent, forming a crucial part of preoperative discussions, allowing patients to make informed decisions regarding risks and benefits of surgery. This study adds new evidence that vitreoretinal surgeons should adjust their approach to managing patients with SRDDs and ERM. Appropriate adjustment of the explanation of risks and outcomes of surgery within this group, rather than previously published standardized figures of surgical outcomes, is required. Strengths of this study include the large number of patients identified, the case-control design, and the fact that surgical procedures were performed within 1 department with standard accepted techniques that were identical between the 2 groups. There are several limitations, with the most significant being its retrospective nature. All attempts to minimize bias were made, including blind analysis of SD-OCT images for selection of a consecutive series of patients and control detection through use of the operating theater notes only. No clinical notes or outcomes were viewed until after case-control selection. Another is the short follow-up period, which is inevitable within a modern United Kingdom National Health Service working environment. Fortunately, the follow-up duration was similar between the 2 groups and should not introduce significant bias unless eyes with SRDDs take a longer period to achieve full visual recovery after ERM peeling. However, this seems unlikely, given the poor mean BCVA outcome for the patients who seemed not to improve as duration of follow-up increased. Ideally, only surgeries performed entirely by the 2 consultant surgeons would have been included, with

exclusion of cases either partly or fully performed by residents. However, less than 40% of surgeries (equally split between patients and control participants) were performed in part by residents. Outcomes were equal between surgeon groups and the same surgical techniques were used throughout. The addition of patient-reported outcomes of surgery and a third group of eyes with conventional drusen but no SRDDs would have been beneficial. Lens Opacification Classification System grading for all patients would have been required to assess the influence of comorbid cataract on the outcome of surgery. Unfortunately, this was not available and the authors acknowledge there was a higher proportion of pseudophakic eyes in the patient cohort, which could have limited the potential visual gain. However, the authors believe the influence of any cataract effect will be minimal because 3 of the patients with SRDDs with visual deterioration were pseudophakic before undergoing ERM peel. References 1. Machemer R. [The surgical removal of epiretinal macular membranes (macular puckers) (author’s transl)]. Klin Monbl Augenheilkd. 1978;173:36e42. 2. Mitchell P, Smith W, Chey T, et al. Prevalence and associations of epiretinal membranes. The Blue Mountains Eye Study, Australia. Ophthalmology. 1997;104:1033e1040. 3. Fraser-Bell S, Guzowski M, Rochtchina E, et al. Five-year cumulative incidence and progression of epiretinal membranes: the Blue Mountains Eye Study. Ophthalmology. 2003;110:34e40. 4. Fraser-Bell S, Ying-Lai M, Klein R, et al. Prevalence and associations of epiretinal membranes in Latinos: the Los Angeles Latino Eye Study. Invest Ophthalmol Vis Sci. 2004;45:1732e1736. 5. McCarty DJ, Mukesh BN, Chikani V, et al. Prevalence and associations of epiretinal membranes in the visual impairment project. Am J Ophthalmol. 2005;140:288e294. 6. Kinoshita T, Imaizumi H, Okushiba U, et al. Time course of changes in metamorphopsia, visual acuity, and OCT parameters after successful epiretinal membrane surgery. Invest Ophthalmol Vis Sci. 2012;53:3592e3597. 7. Okamoto F, Okamoto Y, Hiraoka T, Oshika T. Effect of vitrectomy for epiretinal membrane on visual function and vision-related quality of life. Am J Ophthalmol. 2009;147: 869e874, 874.e861. 8. Wong JG, Sachdev N, Beaumont PE, Chang AA. Visual outcomes following vitrectomy and peeling of epiretinal membrane. Clin Exp Ophthalmol. 2005;33:373e378. 9. Michels RG. Vitreous surgery for macular pucker. Am J Ophthalmol. 1981;92:628e639. 10. McDonald HR, Verre WP, Aaberg TM. Surgical management of idiopathic epiretinal membranes. Ophthalmology. 1986;93: 978e983. 11. Margherio RR, Cox Jr MS, Trese MT, et al. Removal of epimacular membranes. Ophthalmology. 1985;92:1075e1083. 12. Rice TA, De Bustros S, Michels RG, et al. Prognostic factors in vitrectomy for epiretinal membranes of the macula. Ophthalmology. 1986;93:602e610. 13. El Sanharawi M, Sandali O, Bonnel S, et al. Epiretinal membrane surgery outcomes in highly myopic eyes without traction maculopathy: long-term results of a case-control study. Am J Ophthalmol. 2013;156:319e325.e311.

7

Ophthalmology Retina Volume -, Number -, Month 2018 14. Ikeda Y, Yoshida N, Murakami Y, et al. Long-term surgical outcomes of epiretinal membrane in patients with retinitis pigmentosa. Scientific Reports. 2015;5:13078. 15. Mason 3rd JO, Patel SA. Efficacy of vitrectomy and epiretinal membrane peeling in eyes with dry age-related macular degeneration. Clin Ophthalmol. 2015;9:1999e2003. 16. Roller AB, Mahajan VB, Boldt HC, et al. Effects of vitrectomy on age-related macular degeneration. Ophthalmology. 2010;117:1381e1386. 17. Klein R, Meuer SM, Knudtson MD, et al. The epidemiology of retinal reticular drusen. Am J Ophthalmol. 2008;145:317e326. 18. Finger RP, Wu Z, Luu CD, et al. Reticular pseudodrusen: a risk factor for geographic atrophy in fellow eyes of individuals with unilateral choroidal neovascularization. Ophthalmology. 2014;121:1252e1256. 19. Ooto S, Ellabban AA, Ueda-Arakawa N, et al. Reduction of retinal sensitivity in eyes with reticular pseudodrusen. Am J Ophthalmol. 2013;156:1184e1191.e1182. 20. Wilde C, Poostchi A, Mehta RL, et al. Prevalence of reticular pseudodrusen in an elderly UK Caucasian populationdThe Bridlington Eye Assessment Project (BEAP): a cross-sectional study (2002e2006). Eye. 2018;32:1130e1137. 21. Spaide RF. Age-related choroidal atrophy. Am J Ophthalmol. 2009;147:801e810. 22. Thorell MR, Goldhardt R, Nunes RP, et al. Association between subfoveal choroidal thickness, reticular pseudodrusen, and geographic atrophy in age-related macular degeneration. Ophthalmic Surg Lasers Imaging Retina. 2015;46:513e521. 23. Zweifel SA, Spaide RF, Curcio CA, et al. Reticular pseudodrusen are subretinal drusenoid deposits. Ophthalmology. 2010;117:303e312.e301. 24. Kim J, Rhee KM, Woo SJ, et al. Long-term temporal changes of macular thickness and visual outcome after vitrectomy for idiopathic epiretinal membrane. Am J Ophthalmol. 2010;150: 701–709.e701. 25. Kim JH, Kim YM, Chung EJ, et al. Structural and functional predictors of visual outcome of epiretinal membrane surgery. Am J Ophthalmol. 2012;153:103–110.e101. 26. Falkner-Radler CI, Glittenberg C, Hagen S, et al. Spectraldomain optical coherence tomography for monitoring epiretinal membrane surgery. Ophthalmology. 2010;117:798e805. 27. Inoue M, Morita S, Watanabe Y, et al. Inner segment/outer segment junction assessed by spectral-domain optical coherence tomography in patients with idiopathic epiretinal membrane. Am J Ophthalmol. 2010;150:834e839. 28. Mitamura Y, Hirano K, Baba T, Yamamoto S. Correlation of visual recovery with presence of photoreceptor inner/outer segment junction in optical coherence images after epiretinal membrane surgery. Br J Ophthalmol. 2009;93:171e175. 29. Mrejen S, Sato T, Curcio CA, Spaide RF. Assessing the cone photoreceptor mosaic in eyes with pseudodrusen and soft drusen in vivo using adaptive optics imaging. Ophthalmology. 2014;121:545e551. 30. Curcio CA, Messinger JD, Sloan KR, et al. Subretinal drusenoid deposits in non-neovascular age-related macular

31.

32.

33.

34. 35. 36. 37. 38. 39. 40.

41. 42. 43. 44. 45.

46.

degeneration: morphology, prevalence, topography, and biogenesis model. Retina. 2013;33:265e276. Spaide RF. Outer retinal atrophy after regression of subretinal drusenoid deposits as a newly recognized form of late age-related macular degeneration. Retina. 2013;33: 1800e1808. Alten F, Clemens CR, Heiduschka P, Eter N. Localized reticular pseudodrusen and their topographic relation to choroidal watershed zones and changes in choroidal volumes. Invest Ophthalmol Vis Sci. 2013;54:3250e3257. Xu X, Liu X, Wang X, et al. Retinal pigment epithelium degeneration associated with subretinal drusenoid deposits in age-related macular degeneration. Am J Ophthalmol. 2017;175:87e98. Rudolf M, Malek G, Messinger JD, et al. Sub-retinal drusenoid deposits in human retina: organization and composition. Exp Eye Res. 2008;87:402e408. Pumariega NM, Smith RT, Sohrab MA, et al. A prospective study of reticular macular disease. Ophthalmology. 2011;118: 1619e1625. Katagiri S, Hayashi T, Takashina H, et al. Choroidal neovascularization in angioid streaks following microincision vitrectomy surgery: a case report. BMC Ophthalmol. 2013;13:29. Bae KW, Woo SJ. Choroidal neovascularization following epiretinal membrane peeling. Korean J Ophthalmol. 2015;29: 439e440. Warden SM, Pachydaki SI, Christoforidis JB, et al. Choroidal neovascularization after epiretinal membrane removal. Arch Ophthalmol. 2006;124:1652e1654. Asencio-Duran M, Manzano-Munoz B, Vallejo-Garcia JL, Garcia-Martinez J. Complications of macular peeling. J Ophthalmol. 2015;2015:467814. Sandali O, El Sanharawi M, Basli E, et al. Paracentral retinal holes occurring after macular surgery: incidence, clinical features, and evolution. Graefes Arch Clin Exp Ophthalmol. 2012;250:1137e1142. Tabandeh H, Smiddy WE. Choroidal neovascularization following macular hole surgery. Retina. 1999;19:414e417. Rishi P, Dhupper M, Rishi E. Can retinal microtrauma by internal limiting membrane peeling cause retinal angiomatosis proliferans? Oman J Ophthalmol. 2011;4:144e146. Natarajan S, Mehta HB, Mahapatra SK, Sharma S. A rare case of choroidal neovascularization following macular hole surgery. Graefes Arch Clin Exp Ophthalmol. 2006;244:271e273. Gu R, Zhou M, Jiang C, et al. Elevated concentration of cytokines in aqueous in post-vitrectomy eyes. Clin Exp Ophthalmol. 2016;44:128e134. Oh SH, Kim KS, Lee WK. Outer retinal changes in endoilluminator-induced phototoxic maculopathy evident on spectral-domain optical coherence tomography. Clin Exp Optometry. 2015;98:381e384. Haas P, Esmaeelpour M, Ansari-Shahrezaei S, et al. Choroidal thickness in patients with reticular pseudodrusen using 3D 1060-nm OCT maps. Invest Ophthalmol Vis Sci. 2014;55: 2674e2681.

Footnotes and Financial Disclosures Originally received: March 31, 2018. Final revision: June 9, 2018. Accepted: June 19, 2018. Available online: ---. 1

2

Manuscript no. ORET_2018_152.

Ophthalmology and Vision Sciences, Division of Clinical Neurosciences, Queen’s Medical Centre, University of Nottingham, Nottingham, United Kingdom.

8

Royal Derby Hospital, Uttoxeter Road, Derby, United Kingdom.

Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Wilde et al



ERM Surgery Outcomes in Eyes with SRDD

HUMAN SUBJECTS: Human subjects were included in this. The human ethics committees at Derby Teaching Hospitals NHS Foundation Trust determined that approval was not required for this study. All research adhered to the tenets of the Declaration of Helsinki. All participants provided informed consent. No animal subjects were included in this study. Author Contributions: Conception and design: Wilde Analysis and interpretation: Wilde, Awad, Dua, Gandhewar, Chen, Amoaku Data collection: Wilde, Awad, Gandhewar, Chen Obtained funding: Not applicable. Overall responsibility: Wilde, Dua, Gandhewar, Chen, Amoaku

Abbreviations and Acronyms: ACA ¼ age-related choroidal atrophy; AMD ¼ age-related macular degeneration; BCVA ¼ best-corrected visual acuity; CNV ¼ choroidal neovascularization; ERM ¼ epiretinal membrane; ISel ¼ inner segment ellipsoid; logMAR ¼ logarithm of the minimum angle of resolution; nAMD ¼ neovascular age-related macular degeneration; OHT ¼ ocular hypertension; ORA ¼ outer retinal atrophy; PPV ¼ pars plana vitrectomy; PVD ¼ posterior vitreous detachment; RPD ¼ reticular pseudodrusen; SRDD ¼ subretinal drusenoid deposit; SD ¼ spectral-domain. Correspondence: Craig Wilde, FRCOphth, PhD, Ophthalmology and Vision Sciences, Division of Clinical Neurosciences, B Floor, EENT Centre, Queen’s Medical Centre, University of Nottingham, Nottingham NG7 2UH, United Kingdom. E-mail: [email protected].

9