Aniseikonia and Foveal Microstructure After Retinal Detachment Surgery

Abstract

Purpose.: To quantify aniseikonia after successful surgical repair of rhegmatogenous retinal detachment (RD), and to investigate the relationship between the severity of postoperative aniseikonia and retinal microstructures as well as clinical parameters.

Methods.: The study included 106 eyes of 106 patients, without any history of ocular disease/surgery and with less than 2 diopters of anisometropia, who had undergone successful retinal reattachment surgery. Aniseikonia was measured with the New Aniseikonia Test, and foveal microstructure was assessed with spectral-domain optical coherence tomography (OCT) at 6 months postoperatively.

Results.: Twenty-eight of 106 patients (26%) had micropsia; 17 patients (16%) had macropsia; and 61 patients (58%) had no aniseikonia. The mean absolute value of aniseikonia was 2.3 ± 2.9% (range, −12.5% to +12.0%). Of 57 eyes with macula-on RD, 3 had micropsia and 12 had macropsia. Of 49 eyes with macula-off RD, 25 had micropsia and 5 had macropsia. Eyes with micropsia mostly exhibited persistent or transient cystoid macular edema, subretinal fluid, hyperreflective or disruption of IS/OS line, while most of the eyes with macropsia presented epiretinal membrane. Stepwise multiple regression analysis revealed that postoperative best-corrected visual acuity and the area of RD were significantly relevant to the mean absolute value of aniseikonia.

Conclusions.: These results suggested that approximately half of patients with successful repair of RD had aniseikonia. Eyes with macula-off RD tended to show micropsia, while those with macula-on RD mostly presented macropsia. Micropsia and macropsia were primarily caused by respective specific abnormal structures at the foveal region.

Introduction

The effectiveness of rhegmatogenous retinal detachment (RD) surgery has generally been assessed by the rates of retinal reattachment and the postoperative visual acuity. ^1–6 With the improved anatomical success rate of surgery for RD, increasing attention has been directed toward the quality of postoperative vision. Even after successful retinal reattachment and improvement of visual acuity, the postoperative quality of vision may be unsatisfactory in some cases. Aniseikonia is one of the common postoperative symptoms in patients after RD surgery, with 35% of patients complaining of aniseikonia by questionnaire. ⁷  

Aniseikonia, the difference in perceived image size between the two eyes, is believed to play a significant role in binocular function. Symptoms owing to aniseikonia include headache, asthenopia, photophobia, reading difficulty, and nausea. ^8,9 Symptoms develop in sensitive individuals at typical clinical value for percentage aniseikonia of 1% to 3%; symptoms and binocular impairment develop at 3% to 5%; and binocular vision typically becomes absent at more than 5%. ^10,11  

Aniseikonia is generally associated with anisometropia, such as aphakia, pseudophakia, and corneal refractive surgery. ^12–17 In rare cases, however, aniseikonia can develop due to retinal diseases. Retina-induced aniseikonia may be caused by the stretching or compression of the retina, which changes the perceived image size because of the alteration in spacing between the photoreceptors. ¹⁸ Aniseikonia due to retinal diseases such as epiretinal membrane (ERM), ^10,18–20 reattached retina, ^20–23 and macular edema ^24,25 has been reported. 

Although some studies have reported on aniseikonia in RD patients, they included a relatively small number of patients (4–12 cases). ^20–23 In addition, no studies so far have investigated the relationship between the severity of aniseikonia and the foveal microstructure after RD surgery. 

The purpose of the present study was to quantify the amount of aniseikonia and to assess clinical characteristics of aniseikonia after successful repair of RD. The relationship between aniseikonia and clinical factors, including those obtained with spectral-domain optical coherence tomography (OCT), was also investigated. 

Methods

We analyzed 106 patients (68 men, 38 women) following successful surgery for unilateral RD undertaken at the University of Tsukuba Hospital from October 2010 to December 2012. Their ages averaged 56.1 ± 10.9 years (means ± SD). This prospective study was conducted in accordance with the tenets of the Declaration of Helsinki, and the study protocol was approved by the institutional review committees. Prior to inclusion in the study, the nature of the study was explained to all patients, and their written informed consent was obtained. Exclusion criteria included patients with a previous history of vitreoretinal surgery and ophthalmic disorders except myopia of less than −10.0 diopters and severe cataract of more than grade 3 nuclear sclerosis or cortical opacity, and more than 2.0 diopters of anisometropia postoperatively. Eyes with complicated vitreoretinal diseases, such as proliferative vitreoretinopathy and RD resulting from giant retinal tears, macular hole, or ocular trauma, were also excluded. 

Best-corrected visual acuity (BCVA) was measured, and the degree of aniseikonia was quantified using the New Aniseikonia Test (NAT; Handaya, Tokyo, Japan) 6 months after surgery. Best-corrected visual acuity measured with the Landolt chart was expressed as logarithm of minimum angle of resolution (logMAR) and used for the subsequent analyses. 

The NAT, comprising a book and spectacles, measures aniseikonia by dissociating binocular vision with red and green filters. Each eye perceives a half-moon printed on the book page. Two half-moons of different sizes are arranged in a series in a pair, with the difference varying in increments of 1%. The subject, wearing red–green spectacles, views the plates to allow the right eye to see one of the half-moons in each pair, the left eye the other half-moon. Subjects indicate the pair when two half-moons appear to be of equal size. The actual size difference in the pair of half-moons represents the percent of the subject's aniseikonia. The NAT target size was 4 cm (visual field angle of 5.7°), with measurement from 1% to 24% possible. ¹² We measured subjects at approximately 40 cm in both vertical and horizontal meridians and used their mean values for data analyses. Mean aniseikonia of +2% or more was defined as macropsia and mean aniseikonia of −2% or less as micropsia. Patients with logMAR BCVA > 1.0 were excluded since it was difficult for them to perceive the half-moon due to suppression. ¹²  

Retinal images were obtained with spectral-domain OCT (Cirrus high-definition OCT; Carl Zeiss, Dublin, CA, USA) 3 and 6 months after surgery. We performed the five-line raster scans in a horizontal and vertical manner on each eye using Cirrus analysis software version 3.0. Scans with signal strength of more than 7/10 were considered appropriate; from these a representative image was selected. Diagnosis of a disrupted photoreceptor inner and outer segment junction (IS/OS) was made based on loss and irregularity of the hyperreflexive line corresponding to the IS/OS junction. Two graders (YS and TH) assessed the status of the IS/OS lines. Both graders were masked to the clinical findings of the patients, including their visual acuity and aniseikonia results. 

Clinical data were collected, including age, sex, circumferential dimension of retinal tears, area of RD, macular status (on or off), surgical procedures (scleral buckling or vitrectomy), and difference of postoperative spherical equivalent between the two eyes, to determine factors related to aniseikonia. 

Surgery was performed at our clinic by experienced vitreoretinal surgeons (FO and YO). The buckling surgery consisted of cryopexy and circumferential silicone sponge buckling (no. 506; MIRA, Waltham, MA, USA). The encircling was performed with a silicone band (no. 240; MIRA) or a silicone sponge (no. 506G; MIRA). Subretinal fluid drainage and air injection were performed when required. We carried out 23- or 25-gauge pars plana vitrectomy that comprised release of vitreous traction around the breaks, internal drainage of the subretinal fluid, a total fluid/gas exchange with air or 20% sulfur hexafluoride (SF₆) for internal tamponade, and endolaser photocoagulation. None of the vitrectomy group had any additional external buckling elements. In both groups, patients treated with gas injection were instructed to maintain a facedown position for 2 to 7 days. 

The mean and standard deviations were calculated for degree of aniseikonia and other parameters. A paired t-test was performed to compare horizontal and vertical aniseikonia, and an unpaired t-test was done to compare differences in the amount of aniseikonia between sexes and surgical procedures (scleral buckling and vitrectomy). The associations between aniseikonia and clinical parameters and between horizontal and vertical aniseikonia were examined by the Spearman rank correlation test. Multivariate analysis with stepwise regression was performed to investigate the relationship between aniseikonia and clinical parameters. All tests of associations were considered statistically significant if P < 0.05. The analyses were carried out using StatView (version 5.0; SAS, Inc., Cary, NC, USA). 

Results

The Table shows clinical characteristics, surgical procedures, and visual functions in patients undergoing surgery for RD. Preoperative logMAR BCVA was 0.51 ± 0.76, with 49 of 106 eyes being macula-off RD. Postoperative logMAR BCVA was 0.04 ± 0.18; postoperative spherical equivalent was −2.95 ± 2.97 diopters; and the difference in postoperative spherical equivalent between the two eyes was 0.69 ± 0.52 diopters. 

Table

 

Clinical Characteristics, Surgical Procedures, and Visual Functions in Patients Undergoing Surgery for Retinal Detachment

	Values
Age, y	56.1 ± 10.9
Sex, male/female	68/38
Preoperative best-corrected visual acuity, logMAR	0.51 ± 0.76
Circumferential dimension of retinal tears, deg	17.9 ± 14.2
Area of retinal detachment, deg	128 ± 77
Macular status, on/off	57/49
Surgical procedures, scleral buckling/vitrectomy	27/79
Postoperative best-corrected visual acuity, logMAR	0.04 ± 0.18
Postoperative spherical equivalent, diopters	−2.95 ± 2.97
Difference of postoperative spherical equivalent between the two eyes, diopters	0.69 ± 0.52
Absolute value of postoperative aniseikonia
Horizontal aniseikonia, %	2.3 ± 2.9
Vertical aniseikonia, %	2.5 ± 3.4
Mean aniseikonia, %	2.3 ± 2.9

 Values are presented as the mean ± standard deviation.

Clinical Features of Aniseikonia After RD Surgery

The amount of mean aniseikonia ranged from −12.5% to +12.0% (Fig. 1). Of 106 patients, 28 (26%) had micropsia, 17 (16%) had macropsia, and 61 (58%) had no aniseikonia. The absolute value of mean aniseikonia in all patients was 2.3 ± 2.9%, while horizontal and vertical aniseikonia was 2.3 ± 2.9% and 2.5 ± 3.4%, respectively. No significant difference was observed between horizontal and vertical aniseikonia (P = 0.314), while a significant correlation was found between the two values (r = 0.805, P < 0.0001). 

Figure 1

 

Histogram of the mean aniseikonia score after retinal detachment surgery. Aniseikonia ranged from −12.5% to +12.0%. Of 106 patients, 28 had micropsia and 17 had macropsia.

Relationship Between Preoperative Macular Status and Postoperative Aniseikonia

Of 57 macula-on RD patients, 3 had micropsia, 12 had macropsia, and 42 had no aniseikonia after surgery. Of 49 macula-off RD patients, 25 had micropsia, 5 had macropsia, and 19 had no aniseikonia postoperatively (Fig. 2). 

Figure 2

 

Relationship between preoperative macular status and postoperative aniseikonia. In patients with macula-on RD, 74% patients had no aniseikonia and 21% exhibited macropsia. Approximately half of patients with macula-off RD showed micropsia and 10% presented macropsia.

Relationship Between Surgical Procedures and Postoperative Aniseikonia

Among the 106 RD patients, 79 underwent vitrectomy and 27 received scleral buckling surgery. In 57 patients with macula-on RD, 37 received vitrectomy and 20 scleral buckling. Of 37 macula-on RD patients who underwent vitrectomy, 1 had micropsia, 5 had macropsia, and 31 had no aniseikonia, whereas of 20 macula-on RD patients who underwent scleral buckling, 1 had micropsia, 6 had macropsia, and 13 had no aniseikonia. 

In 49 patients with macula-off RD, 42 underwent vitrectomy and 7 scleral buckling. Of 42 macula-off RD patients who received vitrectomy, 24 had micropsia, 3 had macropsia, and 15 had no aniseikonia. Of 7 macula-off RD patients who underwent scleral buckling, 2 had micropsia, 2 had macropsia, and 3 had no aniseikonia. In macula-off RD patients, the incidence of postoperative micropsia tended to be higher with vitrectomy (57.1%) compared to scleral buckling (28.6%) (P = 0.16). 

OCT Findings in Eyes With Aniseikonia

Of 28 eyes with micropsia after surgery, 6 eyes had cystoid macular edema (CME), 5 eyes had hyperreflective IS/OS line at the foveal region, 4 eyes had disruption of IS/OS, 3 eyes had subretinal fluid (SRF), 2 eyes had ERM, 1 eye had macular hole (MH), and the other 7 eyes exhibited no abnormal morphologic change (Figs. 3A–E). Six of 28 cases had CME at 3 months postoperatively; of these, CME resolved at 6 months postoperatively in 4 cases and no change was observed in the other 2 cases. Five of 28 cases developed hyperreflective IS/OS line at the foveal region at 3 months postoperatively; of these, the hyperreflective IS/OS line disappeared at 6 months postoperatively in 3 cases and no change was found in the other 2 cases. 

Figure 3

 

Spectral-domain optical coherence tomography (OCT) images of eyes with aniseikonia after retinal detachment surgery. (A, B) Images of the right eye of an 80-year-old man with macula-off RD. Cystoid macular edema was seen at 3 months after surgery (A), which resolved by 6 months postoperatively (B). His visual acuity was 0.15 (logMAR), and mean aniseikonia was −10.5% of micropsia 6 months after surgery. (C) An image of the right eye of a 47-year-old man with macula-off RD. Optical coherence tomography showed hyperreflective photoreceptor inner and outer segment junction (IS/OS) line at the foveal lesion. His visual acuity was 0.00 (logMAR), and mean aniseikonia was −9.0% of micropsia. (D) An image of the right eye of a 61-year-old man with macula-off RD. Disrupted IS/OS line is displayed. His visual acuity was −0.08 (logMAR), and mean aniseikonia was −4.5% of micropsia. (E) An image of the right eye of a 47-year-old man with macula-off RD. Optical coherence tomography showed subretinal fluid. His visual acuity was 0.30 (logMAR), and mean aniseikonia was −4.5% of micropsia. (F) An image of the right eye of a 53-year-old woman with macula-on RD. Epiretinal membrane was detected. Her visual acuity was 0.22 (logMAR), and mean aniseikonia was +4.0% of macropsia.

Of 17 eyes with macropsia, 10 eyes had ERM and the other 7 exhibited no abnormal morphologic change (Fig. 3F). 

Relationship Between Postoperative Aniseikonia and Clinical Parameters

The absolute value of mean aniseikonia showed a significant correlation with age (r = 0.265, P < 0.01), postoperative logMAR BCVA (r = 0.396, P < 0.0001, Fig. 4A), and the area of RD (r = 0.385, P < 0.0001, Fig. 4B). In contrast, a significant correlation was not observed between the absolute value of mean aniseikonia and circumferential dimension of retinal tears (r = −0.144, P = 0.150), postoperative spherical equivalent (r = 0.078, P = 0.424), or difference of postoperative spherical equivalent between the two eyes (r = 0.104, P = 0.290). A higher value of mean aniseikonia was significantly associated with disrupted IS/OS line (P < 0.05) and macula-off RD (P < 0.0001). No significant relationship was found between aniseikonia and sex (P = 0.233) or between aniseikonia and surgical procedures (P = 0.959). Multivariate analysis with stepwise regression revealed that the absolute value of mean aniseikonia was significantly correlated with postoperative logMAR BCVA (r = 0.337, F = 10.259) and the area of RD (r = 0.286, F = 7.396). 

Figure 4

 

Correlation between absolute values of mean aniseikonia and postoperative logMAR best-corrected visual acuity (A) and the area of retinal detachment (B).

Discussion

As shown in the above results, 45 of 106 patients (42%) exhibited aniseikonia after successful surgery for unilateral RD at 6 months into follow-up. Wright et al. ⁷ investigated motility and binocularity in 40 patients after RD surgery, and found that 35% of patients complained of aniseikonia by questionnaire. However, their study did not quantify aniseikonia and also did not consider the influence of anisometropia due to silicone oil tamponade or compression of the globe by scleral buckling. In our present study, the influence of refractive error caused by intraocular or external ocular factors was excluded because patients with postoperative anisometropia had been excluded. The range of aniseikonia was broad, from −12.5% to +12.0%, consistent with the findings of previous reports that indicated a range from −12.3% to +10.4%. ^20–23 It has been reported that symptoms appear in binocular impairment at a typical clinical value of percentage aniseikonia at 3% to 5%, with binocular vision generally becoming absent at more than 5%. ^10,11 In the present study, more than 3% and 5% aniseikonia was observed in 32 patients (30%) and 19 patients (18%), respectively, suggesting that postoperative quality of vision remains deteriorated in some patients due to aniseikonia even though retinal reattachment and visual recovery were attained. 

In patients with macula-on RD, 74% had no aniseikonia, but 12 of 15 patients with aniseikonia exhibited micropsia. In addition, approximately half of patients with macula-off RD developed micropsia, while only five presented macropsia. To sum up the previous four reports, most aniseikonia after RD surgery was micropsia (27 of 30 cases). ^20–23 Our results in patients with macula-off RD are consistent with the findings of previous studies that supposedly included macula-off patients. 

As shown in the OCT images, approximately half of eyes with micropsia showed persistent and/or transient CME, SRF, and hyperreflective or disruption of IS/OS. If the photoreceptors are stretched apart, the object image stimulates fewer receptors and appears smaller (micropsia). In contrast, if the photoreceptors are compressed more closely together, the object image stimulates a larger number of receptors and appears larger (macropsia). ¹⁸ It is known that CME and central serous chorioretinopathy accompanied by foveal SRF, in which the retina has been stretched, cause micropsia. ^24,25 Cystoid macular edema is a well-known postoperative complication responsible for secondary visual impairment, with a prevalence of 4% to 11%. ^26–28 In our present study, there were six cases with CME (5.7%), in agreement with the results of previous reports. Interestingly, micropsia still remained in a patient who had exhibited CME for 3 months following surgery and whose CME was resolved at 6 months postoperatively (Figs. 3A, 3B). In the literature, OCT studies have observed persistent SRF in 47% and 52.3% of eyes 1 month after RD surgery. ^29,30 At 6 months postoperatively, SRF still existed in 8% of eyes after vitrectomy and 43% of eyes after scleral buckling. ^31,32 Subretinal fluid gradually disappeared within a year after surgery in most cases, whereas disrupted IS/OS was slow to recover. ^27,30 In our study, four eyes with micropsia showed disruption of IS/OS at 6 months postoperatively. These cases suggested that SRF existed at an early postoperative period and disappeared at 6 months, but micropsia still remained. Five of 28 eyes with micropsia had a hyperreflective IS/OS line at the foveal region at 3 months postoperatively; of these, the hyperreflective IS/OS line resolved at 6 months postoperatively in five cases and no change was observed in the other two cases in our present study. Dell'Omo et al. ³³ demonstrated that hyperreflective lesions formed by the folded hyperreflective bands consisting of the IS/OS and external limiting membrane (ELM) lines, the so-called outer retinal folds, were common findings after RD repair with vitrectomy and related to metamorphopsia. The outer retinal folds tend to resolve within 6 months or less from the operation, leaving behind sharply demarcated skip changes in reflectivity of the IS/OS line, so-called IS/OS skip reflectivity abnormalities. ³⁴ Eyes with micropsia having hyperreflective IS/OS line in our study may have possessed these IS/OS skip reflectivity abnormalities. 

More than half of eyes with macropsia showed ERM on OCT. Macular contraction due to ERM has been considered to distort the distribution of photoreceptors and cause macropsia. Ugarte and Williamson ¹⁹ reported that 11 of 14 cases with ERM had macropsia, while Benegas et al. ¹⁸ found that five of seven cases exhibited macropsia. ¹⁸ The reported ERM incidence on OCT following repair of primary RD ranged between 9.0% and 23%. ^{26–28,34,35} In our study, 14 of 106 eyes (13.2%) developed ERM, of which 12 had aniseikonia. In addition, two cases had ERM associated with CME; 1 had macropsia, and the other had micropsia. The case with macropsia was macula-on RD; moderated ERM developed at 3 months postoperatively, and slight CME appeared at 6 months postoperatively. The case with micropsia was macula-off RD and had persistent postoperative CME and slight ERM at 6 months postoperatively. Epiretinal membrane causes macropsia, while CME causes micropsia. These results suggest that when ERM is more severe than CME, macropsia occurs, and when CME is more severe than ERM, micropsia occurs. 

Fourteen eyes with aniseikonia exhibited normal appearance in OCT without any disruption of IS/OS or abnormal macular contour. In these eyes, aniseikonia appeared to be arising from a transient morphological change in the fovea, displacement of the retina, or foveal dysfunction. Because our observation period was only 6 months, postoperative CME, SRF, and IS/OS disruption, if any, might have been restored during the short postoperative period. Aniseikonia might have persisted owing to this kind of transient change. In addition, displacement of macula is common following vitrectomy for macula-off RD with gas tamponade, ³⁶ and is associated with complaints of distortion. ³⁷ Such displacement of the retina distorts distribution of photoreceptors, resulting in aniseikonia. In our study, the incidence of postoperative micropsia was higher with vitrectomy (57.1%) than with scleral buckling (28.6%) in patients with macula-off RD, and these findings suggested that aniseikonia occurred through displacement of the retina. Furthermore, foveal microcirculation is impaired even in patients with macula-on RD. ³⁸ A previous study in RD patients revealed that retinal electroretinographic responses declined in both the reattached and detached retina. ³⁹ In view of these findings, a slight foveal dysfunction could cause aniseikonia. 

Multivariate analysis showed that the amount of aniseikonia was significantly associated with postoperative visual acuity and the area of RD. Shiragami et al. ³⁶ clarified that the extent of RD and macular status were significantly associated with postoperative displacement of the retina using fundus autofluorescence. They suggested that a large-extent macula-off RD could cause retinal translocation even after successful surgery, resulting in dislocation of the macula from its original location. In addition, there was a significant correlation between the presence of macula displacement and symptoms of distortion in the early period following RD surgery. ³⁷ In consideration of the previous reports, our results suggest that the wider the area of RD, the larger the retinal displacement and the amount of aniseikonia. 

The limitations of this study include a short-term follow-up period and the resolution of OCT. We evaluated patients for 6 months postoperatively. Previous studies reported that visual acuity in patients with RD improved more even at 1 to 5 years postoperatively. ^3,40 Aniseikonia in patients after RD surgery also improves in the period between 7 and 45 months postoperatively. ²³ In addition, postoperative OCT findings, including CME, SRF, and status of the IS/OS line, may possibly change during follow-up of more than 6 months. ^27,30–32 Our OCT measurements were based only on five horizontal B-scan cross sections, and imaging inaccuracy might exist in the photoreceptor layer after RD. Future studies with a large sample size, longer follow-up period, and improved OCT image-capturing technologies will be needed. 

In conclusion, 6 months after successful RD repair, 42% of patients showed aniseikonia of varying degrees. Postoperative micropsia was mainly observed in patients with macula-on RD, and postoperative macropsia in patients with macula-off RD. A major cause of micropsia was persistent and/or transient CME, SRF, and IS/OS change, and the principal cause of macropsia was ERM. The amount of aniseikonia was associated with visual acuity and the area of RD. 

Acknowledgments

Disclosure: F. Okamoto, None; Y. Sugiura, None; Y. Okamoto, None; T. Hiraoka, None; T. Oshika, None