Article Text
Abstract
Aims: To investigate the relationship between the presence of the photoreceptor inner and outer segment (IS/OS) junction and visual acuity after epiretinal membrane (ERM) surgery.
Methods: Seventy eyes of 70 consecutive patients who had undergone vitrectomy for idiopathic ERM were examined by optical coherence tomography before and 3 and 6 months after surgery. The IS/OS junction was graded into three grades. The time course of recovery of the IS/OS junction, central foveal thickness (CFT) and best corrected visual acuity (BCVA) during the postoperative period was studied.
Results: A normal IS/OS junction was detected in 47.1% of the eyes before surgery and in 65.7% at 3 months and 75.7% at 6 months after. There was a significant correlation between the IS/OS grade and BCVA before and at 3 and 6 months after the operation (p = 0.0001, p<0.0001, p<0.0001, respectively). The preoperative IS/OS junction grade correlated significantly with BCVA at 6 months (p = 0.0239). CFT did not correlate significantly with BCVA at 3 and 6 months.
Conclusions: The presence of a normal IS/OS junction was associated with good visual acuity after ERM surgery. A normal IS/OS junction probably indicates morphological and functional recovery of the photoreceptors.
Statistics from Altmetric.com
Vitrectomy for an idiopathic macular epiretinal membrane (ERM) is a common vitreoretinal procedure. The principal indication for ERM surgery is a decrease in visual acuity with or without metamorphopsia.1 Despite the successful removal of the ERM, the functional results are not always satisfactory. Many prognostic factors for the outcome of ERM surgery have been investigated, the most significant ones being preoperative visual acuity and duration of symptoms.2 3 The correlation between preoperative foveal thickness and postoperative visual acuity is not significant,1 but the prognostic influence of preoperative macular oedema is still disputable.2
Optical coherence tomography (OCT) is a well-established method of examining the retinal architecture in vivo. Recently, a distinct, highly reflective line was detected just vitread of the retinal pigment epithelium (RPE) layer in images obtained by ultra-high-resolution OCT (UHR-OCT).4 This line has been identified as the photoreceptor inner/outer segment (IS/OS) junction. A distinct and continuous line indicates normal alignment of the membranous discs in the photoreceptor outer segments.5 As alignment of the discs is necessary for normal functioning of the photoreceptors, the presence of a normal IS/OS junction strongly suggests that the photoreceptors are functioning normally. At present, it is generally accepted that images obtained by the OCT3 (Zeiss Humphrey, Sun Leandro, California, USA) correspond well to UHR-OCT images,5 and the IS/OS junction is so distinct that it can be observed in OCT3 images as well as in UHR-OCT images.5–8
It has been reported that the presence of the IS/OS line correlates with recovery of good vision after macular hole surgery.6–8 In non-operated ERM cases, the presence of the IS/OS junction has been found to be related to good visual acuity.2 In operated ERM cases, however, the relationship between the IS/OS junction and visual acuity during the postoperative period has not been determined.
The purpose of this study was to investigate the relationship between the presence of an IS/OS junction and visual acuity after ERM surgery. To accomplish this, we obtained OCT3 images at different times after vitrectomy to determine whether an IS/OS junction was present.
SUBJECTS AND METHODS
The medical records of 70 eyes of 70 consecutive patients with an idiopathic ERM (39 women and 31 men), who underwent pars plana vitrectomy at Chiba University Hospital from January 2005 to September 2007, were examined. Cases of moderate or severe cataract that affected visual acuity, high myopia, or secondary ERM due to retinal vascular disease, ocular inflammation, retinal tears and retinal detachment were excluded. The procedures conformed with the tenets of the Declaration of Helsinki, and approval was obtained from the institutional review board of Chiba University.
The age of the patients ranged from 52 to 83 years (mean (SD) 66.7 (6.9)). The interval between the onset of visual disturbance and surgery was estimated from the patients’ history. All subjects underwent a standard ophthalmological examination, including best corrected visual acuity (BCVA) measurements, applanation tonometry, slit-lamp biomicroscopy, indirect ophthalmoscopy and colour fundus photography, before and 3 and 6 months after surgery. BCVA was measured with a standard Japanese Landolt visual acuity chart and converted into logarithm of the minimal angle resolution (logMAR) units for statistical analyses.
In 66 of the 70 eyes, treatment consisted of a pars plana vitrectomy, phacoemulsification and intraocular lens implantation, and ERM peeling. In four cases, the patient was already pseudophakic. Eyes in which the lens was preserved at the time of vitrectomy were excluded from the study. During the operation, the ERM was peeled successfully in all the cases without intentional removal of the internal limiting membrane. An iatrogenic peripheral retinal tear was found in two eyes, and they were successfully treated with endolaser photocoagulation and sulfur hexafluoride gas tamponade.
OCT images of the retina were obtained before and 3 and 6 months after surgery with an OCT3 instrument. Horizontal scans of 5 mm were made through the fovea of the affected and healthy fellow eyes. All OCT images were acquired through a dilated pupil. Greyscale images were used for more precise identification of the IS/OS junction.6 9 10 The appearance of the IS/OS junction in the OCT images at the fovea was graded from 0 to 2: 0, IS/OS line not visible; 1, abnormal (discontinuous) IS/OS line; 2, normal (well-preserved) IS/OS line (fig 1). The grade of the IS/OS line was determined by two of the authors (KH and BT) who were blinded to the visual acuity. Eleven eyes with low-resolution OCT images with scan strength of 4 or less were excluded.
The central foveal thickness (CFT) was measured manually on the horizontal OCT3 images of all eyes. It was measured by placing calibrated callipers at the vitreous–retina and sensory retina–RPE interfaces.
The significance of the differences in BCVA, CFT and estimated preoperative interval in the three IS/OS junction grades was analysed using one-factor analysis of variance (ANOVA) with post hoc comparisons tested by Fisher’s protected least significant difference (PLSD). The difference in the percentage of eyes with a normal IS/OS was tested using χ2 tests. The correlation between CFT and BCVA was determined by Pearson correlation tests. p<0.05 was considered significant.
RESULTS
The estimated preoperative interval between the onset of visual disturbance and surgery was 1–40 months (mean (SD) 14.0 (12.3)). The preoperative BCVA ranged from 0.00 to 1.00 (mean (SD) 0.35 (0.18)) logMAR units. In the preoperative OCT images, the IS/OS was not detected (grade 0) in 10 eyes (14.3%), was abnormal (grade 1) in 27 eyes (38.6%), and was normal (grade 2) in 33 eyes (47.1%) (fig 2, table 1). The IS/OS was normal (grade 2) in all 58 fellow eyes without macular abnormalities.
The postoperative OCT examination revealed a grade 2 IS/OS junction in 46 eyes (65.7%) at 3 months and 53 eyes (75.7%) at 6 months (table 1). A grade 1 IS/OS junction was detected in 20 eyes at 3 months and 13 eyes at 6 months, and a grade 0 IS/OS junction was detected in four eyes at 3 and 6 months. The percentage of eyes with a normal IS/OS gradually increased after surgery (fig 2), and the changes from baseline were significant at 3 and 6 months (p = 0.0408, p = 0.0010, respectively; χ2 test). Table 2 shows changes in grades of IS/OS junction in each eye.
There was a significant difference in preoperative BCVA among the three preoperative IS/OS junction grades (p = 0.0001, one-factor ANOVA; fig 3, table 1). After surgery, there was a significant difference in the BCVA at 3 months among the three IS/OS grades (p<0.0001, one-factor ANOVA). There was also a significant difference in the 6-month BCVA among the three IS/OS grades (p<0.0001, one-factor ANOVA). At the preoperative examination and at 3 and 6 months after surgery, the BCVA was significantly better in eyes with an IS/OS grade 2 than in those with grade 0 or 1, when the mean BCVA of the three groups was compared independently with each of the other groups (Fisher’s PLSD). BCVA was also significantly better in grade 1 than in grade 0 (Fisher’s PLSD).
The preoperative IS/OS junction grade correlated significantly with preoperative CFT (p = 0.0065; one-factor ANOVA; table 1). However, there was no correlation between the IS/OS grade and CFT at 3 months (p = 0.4200) or 6 months (p = 0.3280) (one-factor ANOVA).
The preoperative IS/OS junction grade correlated significantly with the 6-month BCVA (p = 0.0239) but not with the preoperative interval of the symptoms (p = 0.4086) or BCVA improvements at 6 months (6-month logMAR BCVA − preoperative logMAR BCVA, p = 0.7817; one-factor ANOVA; table 1).
There was a direct significant correlation between preoperative CFT and preoperative logMAR BCVA (r = 0.497, p<0.0001; Pearson correlation test; table 3). Preoperative CFT correlated inversely with BCVA improvement at 6 months (r = −0.408, p = 0.0004), but not with the 6-month BCVA (r = 0.001, p = 0.9927; Pearson correlation test). The correlations between CFT at 3 months and BCVA at 3 months and between CFT at 6 months and BCVA at 6 months were not significant (r = 0.092, p = 0.4525 and r = −0.095, p = 0.4373, respectively; Pearson correlation test).
DISCUSSION
Because alignment of the discs is necessary for normal functioning of the photoreceptors, the presence of a normal IS/OS junction on OCT images indicates normally functioning photoreceptors. In non-operated ERM cases, it has been reported that defects in the IS/OS junction are related to poor visual acuity.2 We have reported that, in patients with retinitis pigmentosa, the grade of the IS/OS junction correlates closely with visual acuity.10 In eyes with macular oedema associated with branch retinal vein occlusion, Ota et al11 12 found that the grade of the IS/OS line in the fovea correlated with visual acuity in resolved or persistent macular oedema, and that the IS/OS junction before treatment may be a prognostic indicator of visual outcome. On the basis of these findings, we conclude that a well-preserved IS/OS junction in the fovea is associated with good visual acuity, whereas discontinuous or absent IS/OS lines indicate disorganisation of the photoreceptor outer segments.
Our results indicate that eyes in which a normal IS/OS junction was detected after surgery had significantly better visual acuity than those without a normal IS/OS junction. This correlation between the presence of a normal IS/OS junction and better postoperative visual acuity was probably due to better morphological recovery of the macular photoreceptor cells. We have reported that a normal IS/OS junction is associated with good visual recovery after macular hole surgery.6
The gradual postoperative increase in the incidence of a normal IS/OS junction is probably related to the slow recovery of the photoreceptors over at least 6 months. This slow recovery of the IS/OS line supports the slow recovery of visual acuity after ERM surgery. Pesin et al3 reported that the mean time to the best visual acuity after ERM surgery was slightly less than 1 year.
It would be of great value to be able to predict before surgery the final visual acuity. The grade of the preoperative IS/OS junction correlated significantly with BCVA 6 months after surgery. Thus, the preoperative IS/OS grade might predict the visual outcome of ERM surgery. On the other hand, preoperative CFT correlated significantly with BCVA improvement at 6 months but not with the 6-month BCVA. Massin et al1 also reported that preoperative macular thickness did not correlate with postoperative BCVA. These results indicate that patients with thickened fovea having ERM surgery can expect greater visual improvement but cannot always achieve good postoperative BCVA. One reason is that preoperative BCVA is poor in patients with thickened fovea.
The IS/OS grade correlated significantly with CFT before surgery. In eyes with thickened fovea, the photoreceptor layer may be disarranged or degenerated, leading to the absence or discontinuity of the IS/OS junction. However, there is a possibility that the thickened retina weakened the signal intensity of the outer retinal layers, making it difficult to detect the reflections from the IS/OS junction.11 To reduce this possibility, we excluded eyes with low-resolution OCT images with scan strength of 4 or less from this study. Moreover, if the preoperative absence or discontinuity of the IS/OS junction is only due to weakening of the signal intensity of the outer retinal layers by the thickened retina, the IS/OS grade should be improved by postoperative thinning of the foveal thickness. However, our results show that the IS/OS grade did not significantly correlate with CFT 3 and 6 months after surgery. The reason for this postoperative dissociation, ie, non-significant correlation between IS/OS grade and CFT, may be that more severe oedema in the foveal photoreceptor layer results in significant disarrangement or degeneration of photoreceptor cells even after postoperative thinning of the CFT.
Niwa et al13 investigated focal macular electroretinograms before and after ERM surgery. They found that the preoperative a-wave amplitude correlated significantly with postoperative BCVA. This is consistent with our finding that the preoperative IS/OS grade correlated significantly with postoperative BCVA, because the a-wave originates mainly from the combined activity of the photoreceptor and off-bipolar cells.14
The IS/OS line was clearly delineated on the OCT3 greyscale images in this study. An OCT image is usually examined in false-colour images, but the IS/OS line is sometimes difficult to differentiate from the RPE layer, especially in high-signal OCT images. In such cases, greyscale images are better because they can enhance each layer, and a distinct image of the IS/OS line separated from the RPE layer can be seen.6 9 10 Ishikawa et al9 examined eyes with cystoid macular oedema using both false-colour and greyscale OCT images and concluded that greyscale images showed a finer gradation of signal reflectance.
In summary, our results show that the incidence of the postoperative normal IS/OS line in OCT images increased gradually after ERM surgery. Moreover, the presence of the normal IS/OS line was associated with good visual recovery. The presence of a normal IS/OS line may indicate morphological and functional recovery of the photoreceptor after ERM surgery. These findings suggest that OCT images can provide information on the foveal photoreceptor layer and visual prognosis in eyes undergoing ERM surgery by use of the IS/OS line. Further studies, especially using high-resolution images of UHR-OCT or spectral-domain OCT, are necessary to determine a more precise relationship between the foveal photoreceptor layer and visual prognosis after ERM surgery.
REFERENCES
Footnotes
Funding: None.
Competing interests: None.
Ethics approval: Obtained.
Patient consent: Obtained.