NEUROPHTHALMOLOGY

Early axonal damage detection by ganglion cell complex analysiswith optical coherence tomography in nonarteritic anteriorischaemic optic neuropathy

Begoña Arana Larrea & Marta Galdos Iztueta &

Lorea Martinez Indart & Nerea Martinez Alday

Received: 21 February 2014 /Revised: 12 June 2014 /Accepted: 16 June 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractPurpose To investigate the ability of ganglion cell complex(GCC) analysis by optical coherence tomography (OCT) todetect early axonal damage in nonarteritic anterior ischaemicoptic neuropathy (NAION), and to assess the relationship ofGCC measurements with visual field defects and functionparameters.Methods Twenty-two patients with NAION participated inthis retrospective case-series study. Patients underwentspectral-domain OCTmeasurement of retinal nerve fibre layer(RNFL) and GCC average and minimum thicknesses, best-corrected visual acuity, Ishihara test and Humphrey visualfield (SITA Standard 24–2). These measurements were re-corded in the acute (2–6 weeks after the ischaemic episode)and chronic (≥6 months later) phases. Spearman’s coefficientswere used to assess the relationship between GCC thicknessand visual field defects.Results In the acute phase, none of the patients showed atro-phy of the optic disc, while early damage was observed in theGCC average and minimum thickness in 54.54 % and77.27 % of patients. At 6 months, the rate of patients withRNFL below normal limits increased to 90 % in the RNFL,and 92.85 % and 100 % in the GCC average and minimumGCC respectively. Spearman’s coefficients indicated signifi-cant relationships of GCC in the acute phase with visual fieldindex and mean deviation in both acute and chronic phases. A

significant correlation was also found with location of thedefects.Conclusions GCC thickness measurement by OCT is capableof detecting early axonal damage in NAION eyes in the acutephase that cannot be detected by RNFL. GCC defects aresignificantly correlated with visual field globally and thedefect location.

Keywords Optical coherence tomography (OCT) . Ganglioncell complex (GCC) . Nonarteritic anterior ischaemic opticneuropathy (NAION) . Diagnosis . Axonal damage

Introduction

Nonarteritic anterior ischaemic optic neuropathy (NAION)constitutes 95 % of all anterior ischaemic optic neuropathies,and is the most common cause of acute optic neuropathy inpeople over the age of 50 [1]. It is characterized by a suddenonset of altitudinal visual loss resulting from hypoperfusion ornon-perfusion of the posterior ciliary arteries that feed theoptic nerve head. These arteries are divided into two mainsectors, a superior one and an inferior one. Accordingly, thevisual field (VF) defect primarily found is altitudinal, and itsmagnitude depends on the extension of the infarct of theposterior ciliary arteries.

When the patient is examined at onset, a relative afferentpupillary defect on the affected eye and optic nerve headswelling can be seen on fundoscopy. Over the first 3 months,the optic disc swelling resolves until degeneration of axonsand apoptosis of ganglion cell bodies in the retina result inoptic nerve head atrophy [1].

Spectral-domain optical coherence tomography (SD-OCT)is a non-invasive imaging technique producing high-resolution cross-sectional images of the retina that allows in-vivo measurements of retinal thickness in the macula and

B. A. Larrea (*) :M. G. Iztueta :N. M. AldayDepartment of Ophthalmology, Cruces University Hospital,Plaza de Cruces s/n, Barakaldo, Vizcaya 48903, Spaine-mail: begoimsa@hotmail.com

M. G. IztuetaInstituto Clínico Quirúrgico Oftalmológico (ICQO), Bilbao, Spain

L. M. IndartClinical Epidemiology Unit, Cruces Hospital, Vizcaya, Spain

Graefes Arch Clin Exp OphthalmolDOI 10.1007/s00417-014-2697-0

around the optic nerve. SD-OCT imaging has become a valu-able tool for identifying and monitoring peripapillary retinalnerve fibre layer thickness in optic neuropathies [2–4] .

The appearance of the optic nerve head following NAIONhas been evaluated using OCT measurements, demonstratinga moderate-to-strong correlation between peripapillary retinalnerve fibre layer (RNFL) loss and corresponding VF sensitiv-ity loss measured with automated perimetry [5–9]. However,the characteristics of optic disc oedema as shown by OCT atonset have limited prognostic value: specifically, initial RNFLthickness is not correlated with final RNFL thickness, visualacuity, or VF mean deviation (MD) [5]. On the other hand,some authors have measured the whole macular thickness(MT) in NAION at least 6 months after the acute event, andfound correlations with VF [10, 11].

Recent advances in segmentation algorithms have made itpossible to visualise and measure individual retinal layers withSD-OCT. Ganglion cell–inner plexiform layer (GCIPL) anal-ysis in the macula has become an important examination in theassessment of various optic neuropathies [12, 13]. Recently,GCIPL losses have also been measured in a chronic phase ofNAION (6 months after the acute event) by Fourier-domainOCT, authors reporting a correlation with both magnitude andlocation of VF losses [14, 15].

The primary purpose of our study was to assess the role ofthe GCIPL or ganglion cell complex (GCC) analysis by SD-OCT in detecting early axonal damage, avoiding the artefactcaused by oedema in peripapillar RNFL measurements. Thesecondary aim was to analyse the relationship between GCCmeasurements and the magnitude and location of VF losses.

Methods

Design

This was an observational retrospective study. Patients withdiagnosis of NAION admitted to the Department ofOphthalmology at Cruces University Hospital, Bilbao, Spainbetween October 2010 and June 2013 were revised for thestudy; a total of 28 participants were recruited. Four patientswere excluded because of coexistence of ophthalmic or neuro-logic disease, and another two because they showed initialatrophy of optic disc at first OCTscan. The studywas approvedby the Clinical Research Ethics Committee, and conducted inaccordance with the principles of the Declaration of Helsinki.

All participants underwent a comprehensive neuro-ophthalmic examination, which included measurement of thebest-corrected visual acuity (BCVA) using a Snellen chart, anIshihara test and a stereoscopic fundus examination. AutomatedVF testing was also performed (see below). RNFL, GCIPL, andmacular thickness were measured with the Cirrus OCT system(Carl Zeiss Meditec, Dublin, CA, USA).

The following parameters were recorded in both acute(between 2 and 6 weeks from the NAION acute episode)and chronic (at least 6 months after the acute episode) phases:BVCA, number of correct plates on the Ishihara test (out of17), VF (description of defects) mean deviation (MD), andvisual field index (VFI), as well as SD-OCT measurements ofperipapillary RNFL and macular GCC average and minimumthicknesses.

Diagnosis of NAION

NAION diagnosis was made on the basis of a sudden andpainless visual loss, optic disc swelling and VF defects.Fluorescein angiography was only performed in 40% becausemost patients met clear diagnostic criteria (age, ARPD, con-tralateral axonal crowding, hiperemic edema at onset, nodemyelinating or infectious disease...) and showed weak leak-age from the optic disc. Clinical and laboratory findings ruledout arteritic anterior ischaemic optic neuropathy. Further, allpatients recruited to the study had resolution of the discoedema within 8 weeks and followed a clinical course consis-tent with NAION. Follow-up VF tests and OCT were per-formed between 6 and 8 months after the acute event.

Exclusion criteria

Patients were excluded if they had media opacities that wouldpreclude SD-OCT scanning, glaucoma, coexistence of oph-thalmic or neurological disease, or any other retinal pathology,or if they had undergone previous eye surgery (other thanuneventful cataract surgery).

Visual field testing

VF testing (Humphrey Visual Field Analyzer, 2010 (CarlZeiss Meditec) was performed with the SITA-Fast 24–2 pro-gram with central fixation strategy. VFI and global MD wererecorded. MD for the superior and inferior hemifields werecalculated from the total deviation values. VF defects werecategorised as superior or inferior if the scotoma did not crossthe horizontal middle line and remained in one of the twohemifields or bihemispheric if both hemifields were affected.

Optical coherence tomography scanning

Subjects underwent ocular imaging with dilated pupils; scansonly were included if they were high quality, had focusedimages with the circular ring of the scan centred around theoptic disc and retinal layers were correctly identified. Weperformed Optic Disc Cube 200 × 200 scans covering an areaof 6 x 6 mm around the optic disc in the OCT image (CarlZeiss Meditec, Dublin, CA, USA) and used RNFL OU anal-ysis to record the mean RNFL thickness, as well as thickness

Graefes Arch Clin Exp Ophthalmol

in superior, inferior, temporal, and nasal quadrants. The GCCanalysis obtained from the Macular Cube 512 × 128 analysiswas used to obtain the average GCC (GCCav) and minimumGCC (GCCmin) thicknesses, as well as the thickness in sixsectors around the macular map provided. For each patient,the colour provided in the OCT report characterising thethickness relative to normal values was also recorded (redand yellow for atrophy, green for values within the normalrange, and white for oedema).

Statistical analysis

The qualitative variables were described in percentages, andquantitative variables were described with means and standarddeviations or median and maximum and minimum dependingon their distributional characteristics. Pearson’s coefficientwas used to assess correlations between OCT values andBCVA, RNFL, and correct Ishihara plates. Spearman´s coef-ficient was used to find correlations between OCT and visualfield parameters. Wilcoxon’s test was used to detect signifi-cant changes between the acute and chronic phases within thesame variable. Statistical analysis was performed with SPSS(vers. 21).

Results

Demographics and visual function

Overall, eight patients were female and 14 male, while 16 ofthe eyes included were right and six left. Mean age of partic-ipants was 63.4±11.7 years ,and mean BCVA at onset was0.55±0.35 with a wide range between 0.05 and 1; the samevariability was found in the Ishihara test, the mean number ofcorrect plates being 10±6, with a range of 1 to 17.

In the acute phase, the VFMDwas −17.51±9.58 and meanVFI was 50.57 %±31. VF defects were categorised as supe-rior in four patients (19 %), inferior in six (28.5 %) andbihemispheric in 11 (52.4 %). BCVA and the number ofcorrect Ishihara plates did not change significantly between

the acute and chronic phases (+0.04, 95 % CI (−0.14 to 0.11)p=0.82 and −1.1, 95 % CI (−1.7 to 3.9) p=0.41 respectively),while MD and VFI decreased slightly (−0.77 95%CI (−3.6 to2.1) p=0.55 and −7.84 95 % CI (−0.6 to 16.3), p=0.122respectively) (Table 1).

OCT measurements

Peripapillary RNFL

In the acute phase, OCT showed a mean peripapillary RNFLthickness of 150.35±11.4 μm. OCT measurements indicatedthat 19 patients (86.4 %) showed values above normal limits(white), while three (13.6 %) had values within normal limits(green) and none showed values below normal limits (red oryellow). Correlations between GCC av at onset and finalRNFL, BCVA, and correct Ishihara plates were not significant(Pearson’s coefficients: −0.020, 0.21, 0.41, p=0.93, p=0.43,p=0.11 respectively). Table 2 reports the percentages of pa-tients that showed values above, within, or below normal inthe RNFL and GCC, both the mean and values for the fourquadrants at acute and chronic phases. Figure 1 represents thenumber of patients that showed values below normal limitsconsidering the mean RNFL, GCC av, and GCC min thick-nesses in the two phases of the NAION. It is easy to appreciatethat despite an apparent lack of damage at the onset in theRNFL measurements, the GCC av value shows that there wasganglion cell damage in more than half of the patients, and theGCC min value identifies damage in as many as 17 (77.3 %)of them. However, when NAION is followed by resolution ofthe optic disc oedema and optic nerve atrophy becomes visi-ble, the damage is also reflected in the RNFL, with chronicphase measurements of this layer showing values below nor-mal limits in 20 of the patients (90 %).

Ganglion cell complex (GCC)

With regard to the GCC analysis, the number of patients thatshowed red or yellow color at the same time point was 12(54.5 %) considering GCC av thickness, and as high as 17(77.3 %) considering the GCC min thickness.

Table 1 Mean values and varia-tion of different parameters be-tween acute and chronic phases

RNFL (retinal nerve fibre layer),GCC (ganglion cell complex),BCVA (best-corrected visual acu-ity), MD (mean deviation), VFI(visual field index), 95 % CI(confidence interval)

Acute 95 % CI Chronic 95 % CI Variation 95 % CI

RNFL (μm) 150.35 126−174 64.80 56−72 85.55 62−109GCC av (μm) 68.33 61−75 54.87 50−59 13.46 6−20GCC min (μm) 59.00 49−69 43.14 36−50 15.86 5−25BCVA 0.55 0.4−0.7 0.59 0.4−0.8 0.04 −0.14 to 0.11Ishihara 10.16 6.9−13.4 9.06 5.3−12.7 1.1 −1.7 to 3.9

MD −17.51 −13.1 to −21.8 −18.28 13.6−22.9 0.77 −3.6 to 2.1

VFI 50.58 35.4−65.7 42.74 25.7−59.7 7.84 −0.6 to 16.3

Graefes Arch Clin Exp Ophthalmol

The GCC measurements also showed statistically signifi-cant decrease in the average and minimum thicknesses be-tween acute and chronic phases (p=0.005 for GCC av and p=0.008 for GCCmin) (Table 1). Overall, the percentage reduc-tions in layer thickness from the acute to the chronic phasewere RNFL, 56.9%; GCCmin, 28.88%; and GCC av, 19.7%,that is, there was a greater percentage reduction in RNFLthickness due to the resolution of optic disc edema.

GCC and VF correlations

We found significant correlations between GCC av at onsetand VF parameters (MD and VFI respectively) in both acuteand chronic phases. Spearman’s coefficients (Fig. 2) were−0.50 (p=0.021) and 0.54 (p=0.016) for the acute phase and−0.59 (p=0.005) and 0.57 (p=0.008) for the chronic phase.Correlations between GCC thickness in the chronic phase andVF defect (MD) were −0.6 (p=0.020) for GCC av and −0.54(p=0.05) for GCCmin).

With regard to the GCC analysis and the VF defect loca-tion, Fig. 3 summarizes the results of correlations betweenGCC thickness at onset with VF of the superior and inferiorhemifields. Means of the three superior and inferior valuesprovided by the GCC analysis map were correlated withmeans of total deviation values of the corresponding superiorand inferior hemifields of the VF. Correlations were

significant for both initial and final VFs for the inferior GCCthickness and the corresponding superior visual field [−0.57(p=0.008) and −0.59 (p=0.006)] respectively and betweensuperior GCC thickness and the corresponding inferior visualhemifield in the chronic phase [−0.52 (p=0.02)], but notbetween the superior GCC thickness and the correspondinginferior visual field in the acute phase [−0.41 (p=0.07)]. Incontrast, Spearman’s correlations between location of GCCdefects in the chronic phase and location of final VF defectwere not significant (−0.083, p=0.76 for inferior VF and−0.32, p=0.23 for the superior one).

GCC defect at onset was not significantly correlated withlocation of the final optic nerve atrophy, but we did observe atrend: specifically, all patients that showed bihemispheric atro-phy in the GCC measurements at onset also showedbihemispheric atrophy of the optic nerve at the final assessment.

Figure 4 shows OCT reports for the right eye of one of thepatients in the acute and chronic phases. At first, RNFLmeasurements indicate values above normal limits at the samepoint that the GCC map exhibits damage of the superiorhemisphere corresponding to an inferior VF defect. Sixmonths later, red color appears in the superior pole of theoptic nerve and ganglion cell complex measurements show aslight increase in the defect. Note that VF defects are morestrongly correlated with GCC defects in the acute phase thanthose observed in the chronic phase.

Discussion

The present study used SD-OCT to demonstrate the importantrole of ganglion cell–inner plexiform layer or GCC analysis inthe macular region in detecting early axonal damage in theacute onset of NAION (6 weeks or less after the acuteepisode).

Some authors have measured the whole macular thicknessin NAION in the chronic phase, showing correlations with VFand supporting the role of macular assessment in this neurop-athy [10, 11]. However, recent advances in segmentation

Table 2 Percentages of patientswith values above, within or be-low normal limits of OCTmeasurementsin the acute andchronic phases

RNFL (retinal nerve fibre layer),GCC (ganglion cell complex)

Percent abovenormal values

Percent withinnormal values

Percent belownormal values

Acute Chronic Acute Chronic Acute Chronic

Mean RNFL (μm) 86.4 0 13.6 10 0 90

RNFL sup (μm) 63.6 5 27.3 5 9 90

RNFL inf (μm) 77.3 0 22.7 25 0 75

RNFL nasal (μm) 63.6 10 36.4 65 0 25

RNFL temp (μm) 45 0 50 55 5 45

GCC av (μm) 0 0 45.4 7.15 54.5 92.85

GCC min (μm) 0 0 22.7 0 77.3 100

Fig. 1 Bars show number of patients that showed values below normallimits on the RNFL and GCC thicknesses in the acute and chronic phases

Graefes Arch Clin Exp Ophthalmol

Fig. 2 Upper row: scatter plots showing correlations between mean valuesof ganglion cell complex thickness at acute phase (GCCac) and visual fieldindex at acute (VFI acute), and chronic (VFI chron) phases andmean deviation

at acute phase (MD acute). Lower row: Correlations between GCCac andmean meviation at chronic phase (MD chron) and between ganglion cellcomplex at chronic phase (GCCchron) and MDchron and VFIchron

Fig. 3 Upper row: scatter plots showing correlations between mean valuesof upper ganglion cell complex thickness at acute phase (GCCac sup) andinferior visual field at acute (VF acute inf) and chronic (VF chron inf) phases;and between inferior thickness of GCC at acute phase (GCCac inf) andsuperior visual field at acute stage (VF acute sup). Lower row: correlations

between inferior GCC at acute phase (GCC ac inf) and superior visual field atthe chronic phase (VF chron sup) and between superior and inferior GCC atchronic stage (GCC chron sup, GCC chron inf) and the correspondig visualfield at the chronic stage (VF chron inf, VF chron sup)

Graefes Arch Clin Exp Ophthalmol

algorithms have made it possible to visualise and measureindividual retinal layers with OCT in the macular region [16,17, 19]. Approximately 50 % of the retinal ganglion cells areconcentrated within 4.5 mm of the fovea [18]. SD-OCTcan beused to measure the thickness of the ganglion cell layer orGCC, defined as the combination of retinal nerve fibre, gan-glion cell and inner plexiform layers (GCL+IPL or GCIPL).

Some previous studies support the view that GCC mea-surements have predictive value: Marzoli SB et al. [20] havefound damage in the GCC while the optic disc seemednormal and patients retained good visual acuity in idiopathicintracranial hypertension; and Tan et al. [21] found damage inthe GCC in patients with glaucoma, but a preserved VFwithout scotomas. Choi et al. [22] suggests that, with time,GCC damage results in damage to the photoreceptors thatproduce an irreversible scotoma. In contrast, in inflammatoryneuritis, in which the VF defect is reversible, the GCC isintact.

The RNFL measurements in the acute phase of NAIONindicated values above normal limits in most patients, whilethey were within normal limits in a few cases but no patientsdemonstrated values below normal limits. We also expectperipapillary axonal damage at onset, but this is being maskedby the peripapillary oedema, and so cannot be detected. In

contrast, in the GCC analysis in the macular region, higherpercentages of patients showed values below normal limits atthe same time point: over 50 % with regard to the averageGCC thickness and as many as 77 % with regard to theminimum GCC thickness.

The present results relating to early GCC damage detectedby early SD-OCT measurements are in accordance with SD-OCT imaging in experimental NAION models and histolog-ical measurements of retinal ganglion cell loss and axonalthinning occurring early after the onset of the ischaemia,inducing a gradual thinning over the first few weeks[23–25]. Further, histological analysis showed a decrease inganglion cell layer thickness over the first 4 weeks [26].

In addition, findings with RNFL measurements in thepresent sample were consistent with previous studies, that is,the characteristics of optic disc oedema as shown by OCT atonset have limited prognostic value; specifically, initial RNFLthickness is not correlated with final RNFL thickness, visualacuity, or VF MD [5, 27].

We found a significant correlation between GCC av atonset and VF parameters in both acute and chronic stages,suggesting that the GCC thickness in the acute stage may be adetermining factor to predict final VF defects, while a corre-lation was also found between GCC thickness in the chronic

Acute stage Chronic stage

Fig. 4 Visual field, RNFL, and GCC analysis of the right eye of a patient in the acute and chronic stages of NAION

Graefes Arch Clin Exp Ophthalmol

stage and VF defects. Moreover, there were good correlationsbetween the location of the damage in the GCC in the macularregion and the location of the scotoma in the VF, with superiorGCC defects being significantly correlated with inferior altitu-dinal defects in the chronic phase, and inferior GCC defectswith superior altitudinal defects in the acute and chronic phases.

These findings agree with previous research evaluating theGCC in patients with NAION in a chronic phase that observeda correlation between GCC av and VF in magnitude andlocation [15], and add to previous evidence in that we founda correlation in the acute phase that, to our knowledge, has notpreviously been reported. On the other hand, in our sample,although the GCC in the chronic stage was significantlycorrelated with global VF parameters [MD, −0.59 (p=0.005)and VFI, 0.57 (p=0.008)], correlations between location ofthe GCC defect in the chronic stage and that of the final VFdefect were not significant (−0.083, p=0.76 for the inferiorVF and −0.32, p=0.23 for the superior VF), suggesting thatGCC in the acute stage could be a better prognostic factor thanGCC chronic measurements. This could be explained becauseas seen in Fig. 4, GCC damage becomes more severe anddiffuse with time, and more internal plexus layer cells such asglyal cells may be damaged. Ganglion cell complex decreasedetected by the OCT can be explained by damage of otherstructures of the internal layers but may partially preserveganglion cell function, and so it has less repercussion on thevisual field with the consequent minor correlation.

When NAION is followed by resolution of disc oedemaand optic nerve atrophy becomes visible, it is also reflected inthe RNFL, with values below normal limits being observed in90 % of patients. There was a slight decrease in GCC thick-ness from the acute to the chronic phase 6 months later, but thedifference is significantly larger in the RNFL changes,supporting the view that GCC analysis is a more useful toolfor the early diagnosis of damage because the big changesobserved in the RNFL thickness are mainly due to the resolu-tion of optic disc edema and not to axonal damage.

Despite being a retrospective study with a limited number ofpatients, the current study demonstrates the role of the ganglioncell–inner plexiform layer or ganglion cell complex (GCC)analysis by SD-OCT in detecting early ganglion cell and innerplexiform layers damage in NAION, avoiding the artefactcaused by oedema in peripapillar RNFL measurements. EarlyGCC measurements (within 6 weeks after the acute episode)were found to have a good correlation with quantitative mea-sures of VF (VFI and MD), as well as being a useful tool indetecting hemisphere location of the damage in VF. Althoughmore studies would be necessary to describe more precisely thetime point at which the ganglion cell complex begins to showdamage, this study indicates that early GCC damage does occurin patients with acute NAION, and this damage can be accu-rately measured with SD-OCT and correlated with final VFdefect in the chronic phase.

Acknowledgments We wish to thank Maialen Lopez Aricha, JoseMaría Losada Domingo, Bárbara Berasategui, and Ana Orive for collect-ed data and care of study patients.

None of the authors have financial, commercial, or proprietary interestin any device mentioned. The authors have no conflicts of interest todeclare.

References

1. Hattenhauer MG, Leavitt JA, Hodge DO et al (1997) Incidence ofnonarteritic anterior ischemic optic neuropathy. Am J Ophthalmol123(3):103–107

2. Kernstock C, Friebe K, Tonagel F (2013) Applications of opticalcoherence tomography (OCT) in neuro-ophthalmology. KlinMonatsbl Augenheilkd 230(11):1097–1105

3. Maekubo T, Chuman H, Kodama Y, Nao-I N (2013) Evaluation ofinner retinal thickness around the optic disc using optical coherencetomography of a rodent model of nonarteritic ischemic optic neurop-athy. Jpn J Ophthalmol 57(3):327–332

4. Savini G, Carbonelli M, Barboni P (2011) Spectral-domain opticalcoherence tomography for the diagnosis and follow-up of glaucoma.Curr Opin Ophthalmol 22(2):115–123

5. Contreras I, Noval S, Rebolleda G, Muñoz-Negrete FJ (2007)Follow-up of nonarteritic anterior ischemic optic neuropathy withoptical coherence tomography. Ophthalmology 114:2338–2344

6. Contreras I, Rebolleda G, Noval S, Muñoz-Negrete FJ (2007) Opticdisc evaluation by optical coherence tomography in nonarteriticanterior Ischemic optic neuropathy. Invest Ophthalmol Vis Sci 48:4087–4092

7. Deleón-Ortega J, Carroll KE, Arthur SN, Girkin CA (2007)Correlations between retinal nerve fiber layer and visual field in eyeswith nonarteritic anterior ischemic optic neuropathy. Am JOphthalmol 143:288–294

8. Hood DC, Anderson S, Rouleau J et al (2008) Retinal nerve fibrestructure versus visual field function in patients with ischemic opticneuropathy a test of a linear model. Ophthalmology 115:904–910

9. Bellusci C, Savini G, Carboneli M, Carelli V, Sadun AA, Barboni P(2008) Retinal nerve fiber layer thickness in nonarteritic anteriorischemic optic neuropathy: OCT characterization of the acute andresolving phases. Graefes Arch Clin Exp Ophthalmol 246:641–647

10. Fernández-Buenaga R, Rebolleda G, Muñoz-Negrete FJ,Contreras I, Casas-Llera P (2009). Macular thickness.Ophthalmology 116(8):1587

11. Papchenko T, Grainger BT, Savino PJ, Gamble GD, Danesh-MeyerHV (2012) Macular thickness predictive of visual field sensitivity inischaemic optic neuropathy. Acta Ophthalmol 90:e463–e469

12. Syc SB, Saidha S, Newsome SD, Ratchford JN, Levy M, Ford E,Crainiceanu CM, Durbin MK, Oakley JD, Meyer SA, Frohman EM,Calabresi PA (2012) Optical coherence tomography segmentationreveals ganglion cell layer pathology after optic neuritis. Brain135(Pt 2):521–533

13. Renard JP, Fénolland JR, El Chehab H, Francoz M, Marill AM,Messaoudi R, Delbarre M, Maréchal M, Michel S, Giraud JM(2013) Analysis of macular ganglion cell complex (GCC) withspectral-domain optical coherence tomography (SD-OCT) in glauco-ma. J Fr Ophtalmol 36(4):299–309

14. Aggarwal D, Tan O,HuangD, SadunAA (2012) Patterns of ganglioncell complex and nerve fiber layer loss in nonarteritic ischemic opticneuropathy by Fourier-domain optical coherence tomography. InvestOphthalmol Vis Sci 53(8):4539–4545

15. Gonul S, kokterik BE, Bakbak B, Gedik S (2013) Comparison of theganglion cell complex and retinal nerve fibre layer measurements

Graefes Arch Clin Exp Ophthalmol

using Fourier domain optical coherence tomography to detect gan-glion cell loss in non-arteritic anterior ischaemic optic neuropathy. BrJ Ophthalmol 97:1045–1050

16. Wang M, Hood DC, Cho JS, Ghadiali Q, De Moraes CG, Zhang X,Ritch R, Liebmann JM (2009) Measurement of local retinal ganglioncell layer thickness in patients with glaucoma using frequency-domainoptical coherence tomography. Arch Ophthalmol 127(7):875–881

17. Fabritius T, Makita S, Miura M, Myllylä R, Yasuno Y (2009)Automated segmentation of the macula by optical coherence tomog-raphy. Opt Express 17(18):15659–15669

18. Curcio CA, Allen KA (1990) Topography of ganglion cells in humanretina. J Comp Neurol 300(1):5–25

19. Mwanza JC, Oakley JD, Budenz DL, Chang RT, Knight OJ, FeuerWJ (2011): Macular ganglion cell-inner plexiform layer: automateddetection and thickness reproducibility with spectral domain-opticalcoherence tomography in glaucoma. Invest Ophthalmol Vis Sci52(11):8323–8329

20. Marzoli SB, Ciasca P, Curone M, Cammarata G, Melzi L, CriscuoliA, Bussone G, D´Amico D (2013) Quantitative analysis of opticnerve damage in idiopathic intracranial hypertension (IIH) at diagno-sis. Neurol Sci 34: S143–S145

21. Tan O, Li G, Lu AT, Varma R, Huang D (2008) Mapping of macularsubstructures with optical coherence tomography for glaucoma diag-nosis. Ophthalmology 115:949–956

22. Choi SS, Zawadzki RJ, Keltner JL, Wern JS (2008) Changes incellular structures revealed by ultra-high resolution retinal im-aging in optic neuropathies. Invest Ophthalmol Vis Sci 49:2013–2119

23. Bernstein SL, Johnson MA, Miller NR (2011) Nonarteritic anteriorischemic optic neuropathy (NAION) and its experimental models.Prog Retin Eye Res 30(3):167–187

24. Goldenberg-Cohen N, Guo Y, Margolis F, Cohen Y, Miller NR,Bernstein SL (2005) Oligodendrocyte dysfunction after inductionof experimental anterior optic nerve ischemia. Invest OphthalmolVis Sci 46(8):2716–2275

25. Zhang C, Guo Y, Slater BJ, Miller NR, Bernstein SL (2010) Axonaldegeneration, regeneration and ganglion cell death in a rodent modelof anterior ischemic optic neuropathy (rAION). Exp Eye Res 91:286–292

26. Ho JK, Stanford MP, Shariati MA, Dalal R, Liao YJ (2013) Opticalcoherence tomography study of experimental anterior ischemic opticneuropathy and histologic confirmation. Invest Ophthalmol Vis Sci54:5981–5988

27. Kupersmith MJ, Anderson S, Durbin M, Kardon R (2013)Scanning laser polarimetry, but not optical coherence tomographypredicts permanent visual field loss in acute nonarteritic anteriorischemic optic neuropathy. Invest Ophthalmol Vis Sci 54(8):5514–5519

Graefes Arch Clin Exp Ophthalmol