Peripapillary Choroidal Thickness in Nonarteritic Anterior Ischemic Optic Neuropathy

Abstract

Purpose.: To compare peripapillary choroidal thickness, macular thickness, and peripapillary retinal nerve fiber layer (RNFL) thickness between the affected eyes of patients with nonarteritic anterior ischemic optic neuropathy (NAION) and their unaffected eyes and the eyes of healthy control participants.

Methods.: Thirty eyes with NAION, 30 unaffected fellow eyes, and 25 healthy control participants who met the inclusion criteria were included in this prospective comparative study at a single academic institution. The optic disc area, macular thickness, and RNFL thickness were evaluated in the last follow-up session using spectral-domain optical coherence tomography (OCT). Enhanced depth imaging OCT was obtained to evaluate peripapillary choroidal thickness (PCT).

Results.: The mean follow-up time was 4.9 ± 0.9 months and all data was collected from the last follow-up session. The mean optic disc area in the NAION eyes and their unaffected fellow eyes was significantly smaller than healthy control participants (P < 0.001). The mean PCT in the NAION eyes, unaffected fellow eyes, and the control group was 201.6 ± 42.4, 212.1 ± 45.1, and 166.9 ± 49.9 μm, respectively (ANOVA, P = 0.01). All regional PCT values were significantly greater in the NAION eyes and unaffected fellow eyes than the control eyes. In all participants, the optic disc area was correlated with the average PCT (r = −0.33, P = 0.002). In addition, RNFL thickness and macular thickness values were significantly lower in the NAION eyes versus the unaffected fellow eyes and control eyes.

Conclusions.: Peripapillary choroidal thickness is thicker in NAION eyes and unaffected fellow eyes than control eyes. Thick peripapillary choroid might contribute to the development of NAION.

Nonarteritic anterior ischemic optic neuropathy (NAION) is the most common clinical presentation of acute ischemic damage to the optic nerve and is characterized by a sudden painless visual loss, optic disc edema, and peripapillary hemorrhage.¹ 

Branches of the short posterior ciliary arteries provide most of the blood supply to the optic nerve head and NAION is associated with histopathologically verifiable occlusion of these arteries. Medial and lateral paraoptic short posterior ciliary arteries form the circle of Zinn-Haller and its anatomical variations could possibly make this part of the optic nerve head circulation particularly vulnerable to ischemic attacks. Paraoptic short posterior ciliary arteries also supply the peripapillary choroid. On the other hand, peripapillary choroid contributes to the blood supply of the optic nerve head.² Therefore, choroidal abnormalities might also happen in NAION and measurement of the peripapillary choroidal thickness (PCT) may be highly relevant to this disease. 

Spectral-domain optical coherence tomography (SD-OCT) can quantify macular thickness, ganglion cell complex thickness, peripapillary retinal nerve fiber layer (RNFL) thickness, and even peripapillary choroidal thickness (PCT) using the enhanced depth imaging OCT (EDI-OCT) technique. Previous studies including ours have shown that defects in the RNFL and macular thicknesses occur in NAION.^3–8 Choroidal thickness might be important in the pathophysiology of several eye disorders including glaucoma and retinal diseases.^9–16 

The aims of the current study were to quantitatively compare PCT, RNFL, macular thickness, and optic disc area between the affected eyes of patients with NAION and their unaffected eyes and the eyes of healthy control participants. 

Methods

Between January 2012 and January 2014, patients with NAION who met the inclusion criteria at Farabi Eye Hospital (Tehran, Iran) were included in this prospective, comparative study. The study was carried out after the approval of the research ethics committee was received. Written informed consent was obtained from all patients, and the study was conducted in accordance with the Declaration of Helsinki. Inclusion criteria for this study were: (1) visual acuity and/or field loss, (2) optic disc swelling and/or hemorrhage that resolved in the 3-month follow-up period, and (3) patient's cooperation for imaging studies after the 3-month follow-up. 

The exclusion criteria of the study were any clinical and laboratory finding of arteritic AION, any history or clinical evidence of retinal disease, neurologic disease, glaucoma, IOP of more than 21 mm Hg, intraocular surgery or laser therapy, refractive error beyond ±4.00 diopters (D), and less than 3 months of follow-up. Initial evaluation at presentation included the measurement of best-corrected visual acuity, applantation tonometry, slit-lamp biomicroscopy, and fundus examination. These examinations were repeated at least 3 months after acute event by the time which optic disc border clearly delineated and follow-up data was used for analysis. The best-corrected visual acuity was measured using a Snellen chart at 6 m and was converted into the logMAR for statistical analysis. Perimetry was performed with the standard Swedish Interactive Thresholding Algorithm (SITA) using the 24-2 pattern on the Humphrey Field Analyzer (Carl Zeiss Meditec, Dublin, CA, USA). Patients also underwent a refractive error examination using an automated refractometer (RM-8900; Topcon, Tokyo, Japan) and axial length measurements using partial optical coherence interferometry (IOL Master; Carl Zeiss Meditec). Photographs of the optic disc were taken with a fundus camera. 

All OCT measurements were performed using SD-OCT (Heidelberg Spectralis SD OCT, Spectralis software version 5.3.2; Heidelberg Engineering, Heidelberg, Germany) for each eye of the patients and controls by the same operator (MK). Quality scores for scans are expressed as the signal-to-noise ratio in decibels (dB). The quality of the scans above 20 dB was considered acceptable. For macular OCT with automatic real-time of 100 frames, three circular lines representing 1-, 3-, and 6-mm scan diameters were obtained. The data of the innermost circle that defined the fovea was not used. The outer macular thickness (outer nasal, outer temporal, outer superior, and outer inferior) is delineated by the two outer circles (ring-shaped area defined by black arrows), whereas the inner macular thickness is delineated by the area between the inner circle and the fovea (inner nasal, inner temporal, inner superior, and inner inferior). The RNFL thickness was measured using a 360° peripapillary circle scan around the optic nerve head with a diameter of 3.4 mm. The automatic real-time of 100 frames was performed to obtain good-quality images. The peripapillary circle scan (the scan around the optic disc with a diameter of 3.4 mm) was done using EDI for the choroidal thickness measurement.^10,11,15 For this purpose, one author who was blind to the clinical data of all examined eyes (PA) manually delineated the lower and upper segmentation lines in such a way that they corresponded to the sclerochoroidal interface and posterior edge of the retinal pigment epithelium to represent the outer and inner boundaries of the choroid, respectively. Two other authors (SM and MAF) reviewed and confirmed the segmentations. Then, the RNFL thickness sectors algorithm was used to obtain the choroidal thickness in the corresponding sectors. The intraobserver and interobserver reproducibility of the PCT measurements were assessed using a random subset of 20 images by two examiners (RSM and AK) who were blind to the initial results. The interval of separation for intraobserver assessments was 2 months. 

The optic disc contour was marked at the inner border of the scleral ring on infrared optic nerve images of OCT by an experienced clinician (SM) while viewing the optic disc photographs similar to author's previous work.¹⁷ Optical coherence tomography software calculated the optic disc area parameter. Another author (MAF) performed the same measurement. The average of the two values was recorded as the optic disc area. 

In addition, the right eyes of age- and sex-matched healthy control participants without optic nerve head or retinal or neurologic abnormalities were recruited. All ophthalmologic examinations were conducted with the same methods as in the patients with NAION. 

Statistical analysis was performed using SPSS software version 17 (SPSS, Inc., Chicago, IL, USA). 

Sex distribution of the groups was compared by the χ² test. The intraobserver and interobserver reproducibility were evaluated by the intraclass correlation coefficient. The Shapiro-Wilk test was used to determine the normal distribution of the data. If the data was distributed normally, the means of ANOVA and Student's t-test were used to detect the significance of any difference between the groups. Post hoc comparisons were performed using the Tukey's test. In the absence of normality, nonparametric statistical analysis (Mann-Whitney U test) was used. The association between the optic disc area and the choroidal thickness was analyzed using the Pearson's correlation coefficient test. Differences were significant if P was less than 0.05. 

Results

Among 41 patients with NAION, 30 patients were included in the study. Six patients were excluded because of poor OCT quality and/or poor choroidal segmentation. Two patients with bilateral disease and three patients with only the data of the NAION eyes were also excluded. Therefore, the data of 30 NAION eyes and 30 unaffected fellow eyes was analyzed. The mean follow-up duration was 4.9 ± 0.9 months and all the data was collected in the last follow-up session. The data of 25 healthy eyes was also used as the control group. 

There were not any differences in the demographic data (including age, sex, visual acuity, and mean deviation) between the excluded group and the included group. The demographic data of the included NAION eyes, unaffected fellow eyes of the patients with NAION, and the eyes of healthy control participants was not significantly different except for the corrected visual acuity, disc area, and the mean deviation (Table 1). 

Table 1

 

Demographic Data of the Patients With NAION and Their Unaffected Fellow Eyes and Healthy Control Participants

 

There were a correlation of optic disc areas between the affected and unaffected fellow eyes of the patients with NAION (r = 0.86, P < 0.001). The mean ± SD of optic disc area in the NAION eyes was 1.86 ± 0.30 mm², which was significantly smaller than normal control disc area measurements (2.23 ± 0.44 mm²; P < 0.001, ANOVA test). When the optic disc area was compared between the unaffected fellow eyes and the control eyes, the optic disc area was also significantly smaller in the NAION fellow eyes (P < 0.001, Table 1). 

For PCT measurements, the sclerochoroidal junction was visualized in all eyes. Table 2 shows intraobserver and interobserver reproducibility of PCT in different sectors. All the parameters showed excellent reproducibility in the intraobserver and interobserver tests with a range of 0.970 to 0.998 for the former and 0.943 to 0.998 for the later. The average PCT in the NAION eyes, unaffected fellow eyes, and the control group was 201.6 ± 42.4 μm, 212.1 ± 45.1 μm, and 166.9 ± 49.9 μm, respectively (Table 3). The average PCT and all regional values were significantly greater in the NAION eyes when compared with the control eyes (ANOVA, P = 0.01 for average PCT). Similarly, PCT values were significantly greater in the unaffected fellow eyes in comparison with the control group (P = 0.001). No significant difference in the PCT values was found between NAION and unaffected eyes (P = 0.64). The analysis also demonstrated that the inferior peripapillary regional choroid (the average of inferior nasal and inferior temporal sectors in the NAION eye, fellow eye, and control eye equivalent to 176.9 ± 47.9, 185.6 ± 51.5, 145.2 ± 54.4 μm, respectively) was significantly thinner than the superior choroidal region (the average of superior nasal and superior temporal regions in the NAION eye, fellow eye, and control eye equivalent to 209.7 ± 51.1, 220.8 ± 55.3, 178.5 ± 49.4 μm, respectively; t-test; P = 0.01, P = 0.01, P = 0.02, respectively; Fig. 1A). 

Table 2

 

Intraclass Correlation Coefficients for Peripapillary Choroidal Thickness in a Random Subset of 20 Eyes

 

Table 3

 

Mean ± SD of OCT Measurement of RNFL Thickness and PCT in the Three Groups (NAION, Unaffected Fellow Eyes, and Control Group)

 

Figure 1

 

Profile of peripapillary choroidal thickness. (A) Lines show the mean peripapillary choroidal thickness profiles of control subjects, NAION eyes, and unaffected fellow eyes. (B) Scatterplot of the simple linear regression analysis between the disc area and peripapillary choroid all subjects. T, temporal; TS, temporal-superior; S, superior; NS, nasal-superior; N, nasal; NI, nasal-inferior; I, TI, temporal-inferior.

In all subjects, optic disc area was correlated with average PCT (r = −0.31, P = 0.004; Fig. 1B). 

In addition, the RNFL thickness values were significantly less in all sectors in the NAION eyes versus the unaffected fellow eyes and control eyes (P < 0.001; Table 3). Eyes with NAION in comparison with the unaffected fellow eyes also showed significantly thinner macular thickness values in each sector except for the outer inferior sector (Table 4). 

Table 4

 

Mean ± SD of OCT Measurement of Macular Thickness in the Three Groups (NAION, Unaffected Fellow Eyes, and Control Group)

 

Figure 2 shows the examples of RNFL thickness, PCT, and macular thickness measurements in one participant from each of the three groups. 

Figure 2

 

(A) An OCT image shows the manual segmentation of the Bruch's membrane (upper red line) and sclerochoroidal border (lower red line). Peripapillary total choroid profile was measured in a healthy subject. Thickening of peripapillary choroid (B) and measurement of nerve fiber layer (C) and macular thickness (D) in an unaffected fellow eye of ischemic optic neuropathy eye. (E–G) Thickening of choroid and thinning of nerve fiber layer and macula in an ischemic optic neuropathy eye.

Discussion

In this study, according to the OCT images (EDI technique for PCT) taken in the last follow-up, we found that the mean optic disc area in NAION eyes and unaffected fellow eyes was smaller than the control eyes, PCT was thicker in the NAION eyes and unaffected fellow eyes than the control eyes, and RNFL and macular thickness were thinner in the NAION eyes than unaffected fellow eyes and control eyes. 

Nonarteritic anterior ischemic optic neuropathy is associated with the occlusion of paraoptic branches of the short posterior ciliary arteries and the infarction is predominantly located in the retrolaminar region of the optic nerve head. This pattern suggests that the vascular occlusion in NAION does not lie within the choroidal circulation since the choroid majorly supplies anterior prelaminar and laminar layers.¹ Postmortem studies have demonstrated that the main vascular supply of the prelaminar portion is from circle of Haller and Zinn coursing through the choroid with a minimal role of the choroidal circulation itself.¹⁸ Therefore, it seems that the choroid does not have a substantial role in the NAION pathogenesis.¹ On the other hand, some studies have demonstrated separation of the larger recurrent choroidal branches from the circle of Zinn-Haller, which supplies the immediate peripapillary choroid.¹⁹ Therefore, disturbed perfusion of the peripapillary choroid might also happen in the NAION. Based on this information, we expected to observe no changes in the peripapillary choroidal thickness or a mild thinning in NAION because of the choroidal vessel loss. A decrease in the PCT and choroidal perfusion has been proposed in the pathophysiology of glaucomatous optic neuropathy, as well.^9,15 However, several studies have reported no differences in PCT between glaucoma patients or perimetrically unaffected fellow eyes and the eyes of healthy controls.^11–14 Surprisingly, in this study, we found thickening of the peripapillary choroid not only in NAION eyes but also in the unaffected fellow eyes of NAION patients without any significant differences between them. Because the contralateral unaffected eyes of the patients with unilateral NAION could be involved, one may suppose that the thick choroid may have been present before the onset of the disease in our NAION patients. Therefore, it seems that the choroid is not affected pathologically by the NAION and it rather contributes to the NAION development. In other words, alteration in the choroid is not the effect, but the cause of NAION. A similar and important contributing factor for developing NAION is the cupless optic nerve, as well. Previous studies using stereoscopic photographs have found a smaller optic disc area and a smaller cup-to-disc area ratio of the unaffected fellow eyes of patients with unilateral NAION when compared with healthy controls.^20,21 Similar finding have been reported using OCT.^3,22 A small optic disc and scleral canal opening, abnormally stiff lamina cribrosa, and the associated small cup in NAION patients result in crowding of the optic nerve fibers as they pass through a restricted space in the optic disc and lamina cribrosa. Because in this study choroidal thickness was negatively correlated with the optic disc area, we believe that choroidal thickening, as an additional factor, contributes to the structural crowding in NAION. In fact, the thick choroid could restrict the optic disc space even more. Swollen axons from ischemia are crowded more in the restricted space resulting from the small disc, small cup, and thick choroid, particularly at the most crowded region, the cribriform plate.²³ This produces a condition similar to the “compartment syndrome” in which more compression of the capillaries and ischemia happens.²⁴ On the other hand, the thick peripapillary choroid might be due to a local impairment of autoregulatory mechanisms of the optic nerve head. In healthy control participants, the optic nerve head flow is maintained constant despite the variations in the perfusion pressure and metabolic conditions.¹ The abnormal autoregulatory mechanisms and the following thick choroid in the NAION eyes and their unaffected eyes may be a predisposing factor. This hypothesis was raised by Dias-Santos et al.,²⁵ who found increased macular choroidal thickness in NAION eyes compared with healthy control eyes. Interestingly, another study found an abnormally thin macular choroid in the eyes affected by NAION and in the contralateral eyes unaffected by the disease.²⁶ In contrast to the two studies, we investigated the peripapillary choroid instead of macular choroid. It seems that the peripapillary area is involved more than the macular area in NAION as an optic neuropathy. In this study, similar to the previous studies, we also showed an average choroidal thickness of 167 μm in healthy controls and the regional PCT tended to be greater superiorly and smaller inferiorly.^27–29 In fact, our data in control eyes was very similar to the findings of a study by Huang et al.²⁹ because we used the same OCT equipment and method of measurement. The inferior choroidal regions showed a thinning trend with respect to the superior regions in NAION and their fellow eyes similar to the control eyes, which again indicates that the abnormal thick choroid with a similar thickness profile was present before the disease, not after NAION. 

However, regarding the role of the choroid, further research is required to examine its contribution to ischemic optic neuropathy. Recently, Jia et al.³⁰ used OCT angiography to quantify human disc perfusion in glaucoma. The use of choroidal angiography would also be helpful in the evaluation of NAION. In addition, in contrast to NAION, peripapillary choroidal ischemia and poor filling of the choroid in fluorescein angiography have been observed in arteritic ischemic optic neuropathy due to vasculitis of the short posterior ciliary vessels.^1,31 Therefore, it would be valuable to measure PCT in arteritic ischemic optic neuropthy and compare it with NAION and control eyes. 

In addition to PCT assessment, we found thinning of the RNFL and macular thickness in NAION eyes when compared with the control eyes. Similarly, we previously reported defects in the RNFL and posterior pole retinal thickness in eyes affected with NAION. Both of these parameters were correlated with the corresponding visual field defects.⁸ Aggarwal et al.³² measured the macular ganglion cell complex thickness in NAION and showed the correlation of the hemispheric ganglion cell complex loss with altitudinal visual field loss. Papchenko et al.⁷ found the correlations of macular thickness parameters (using Stratus time-domain OCT) with the mean deviation of the visual field in NAION. 

Our study had several limitations. First, our sample size was small but it was in line with prior studies; moreover, given the very significant difference in the thicknesses observed between the groups, the overall results of the study could be generalized. Second, 14% (6 of 41) of the scans in the NAION group could not be analyzed since the choroid–scleral junction was not distinct. Given that the mean PCT was thicker in NAION and unaffected eyes when compared with the control participants, the depth penetration necessary for adequate visualization of the choroidal region in the control eyes may be less than that for the choroidal region in NAION and their unaffected eyes. The prevalence of poor choroidal visualization may also be minimized by limiting the scans to ones with automatic real-time of 100 frames and good-quality images. Finally, disc swelling and edema preclude choroidal measurements during the acute phase of NAION. Therefore, we only recruited patients with more than 3 months of follow-up. 

In conclusion, we showed an increase in the choroidal thickness by EDI-OCT in NAION and the unaffected fellow eyes versus the control eyes, which might be a contributing factor for NAION. 

Acknowledgments

Disclosure: M.A. Fard, None; P. Abdi, None; A. Kasaei, None; R. Soltani Moghadam, None; M. Afzali, None; S. Moghimi, None