You are here

  • Home
  • Archive
  • Volume 101, Issue 8
  • Retinal segmented layers with strong aquaporin-4 expression suffered more injuries in neuromyelitis optica spectrum disorders compared with optic neuritis with aquaporin-4 antibody seronegativity detected by optical coherence tomography

Article Text

Article menu

Download PDFPDF

Clinical science
Retinal segmented layers with strong aquaporin-4 expression suffered more injuries in neuromyelitis optica spectrum disorders compared with optic neuritis with aquaporin-4 antibody seronegativity detected by optical coherence tomography
  1. Chun xia Peng1,
  2. Hong Yang Li2,
  3. Wei Wang3,
  4. Jun qing Wang1,
  5. Lei Wang1,
  6. Quan gang Xu4,
  7. Shan shan Cao1,
  8. Huan fen Zhou1,
  9. Shuo Zhao1,
  10. Shi hui Wei1
  1. 1Department of Ophthalmology, Chinese PLA General Hospital, Beijing, China
  2. 2Department of Ophthalmology, Beijing Friendship Hospital, Capital Medical University, Beijing, China
  3. 3Zhongshan Ophthalmic Center, Sun yat-sen University, Guangzhou, China
  4. 4Neurology Department, Chinese PLA General Hospital, Beijing, China
  1. Correspondence to Dr Shi Hui Wei, Ophthalmology Department Chinese PLA General Hospital, Fuxing Road No. 28, Haidian District, Beijing, 100853, China; weishihui706{at}hotmail.com

Abstract

Purpose To evaluate retinal segmented layer alterations in optic neuritis (ON) in an AQP4-Ab seropositive (AQP4-Ab+/ON) cohort and in neuromyelitis optica (NMO) with ON eyes (NMO-ON) compared with an AQP4-Ab seronegative ON (AQP4-Ab−/ON) cohort using optical coherence tomography (OCT).

Methods We recruited 109 patients with ON (161 eyes) and 47 healthy controls. All patients with ON were subdivided into three subcohorts: 37 patients (54 eyes) with AQP4-Ab+/ON, 45 patients (65 eyes) with AQP4-Ab−/ON and 27 patients (42 eyes) with NMO-ON. All subjects were evaluated for their peripapillary retinal nerve fibre layer (pRNFL) and inner macular segmented layer using OCT.

Results AQP4-Ab+/patients with ON had the same structural injury patterns as patients with NMO-ON, and the injury patterns were distinct from those of AQP4-Ab−/patients with ON. NMO-ON and AQP4-Ab+/ON preferentially damaged the pRNFL (all p=0.000), the macular retinal nerve fibre layer (mRNFL; p=0.000 and 0.032, respectively), and the inner plexiform layer (IPL; p=0.000 and 0.006, respectively) without differences in the retinal ganglion cell layer (p=0.106 and 0.374, respectively) compared with AQP4-Ab−/patients with ON. The thickness of the inner nuclear layer (INL) increased in NMO-ON (p=0.043) compared with that of AQP4-Ab−/ON without a significant difference in AQP4-Ab+/ON versus AQP4-Ab−/ON (p=0.353). When the thickness of the inferior nasal quadrant (NI) of the pRNFL was reduced to ≤46.5 μm (area under the curve 0.772, sensitivity 89.2% and specificity 57.5%) 6 months after ON onset, NMO was considered.

Conclusions AQP4-Ab+/ON produced similar structural injury patterns as NMO-ON. The pRNFL, mRNFL and IPL in the two types of ON and the INL in NMO-ON suffered more damage than those in AQP4-Ab−/ON, which could be associated with strong aquaporin-4 expression. The thickness of the NI of the pRNFL could be a potential clue for predicting ON progression to definite NMO.

  • Imaging
  • Optic Nerve
  • Retina
  • Pathology
  • Diagnostic tests/Investigation

https://doi.org/10.1136/bjophthalmol-2016-309412

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Neuromyelitis optica (NMO) and multiple sclerosis (MS) are autoimmune demyelination diseases that mainly affect the central nervous system (CNS)1 and optic neuritis (ON) is the most common initial manifestation of NMO and MS. Isolated ON with aquaporin-4 antibody (AQP4-Ab) seropositivity (AQP4-Ab+/ON) has a high risk of converting to definite NMO in clinical practice; therefore, AQP4-Ab+/ON is presumed to have a similar pathogenesis as definite NMO such that when AQP4-Ab binds to its receptors it causes complement deposition, which results in astrocyte injury and secondary axonal demyelination from oligodendrocyte injury.2 ,3 Therefore, AQP4-Ab+/ON is considered an early or limited type of NMO and has been named an NMO spectrum disorder (NMOSD). Our previous study also demonstrated that AQP4-Ab+/patients with ON had injury patterns distinct from those of AQP4-Ab seronegative ON (AQP4-Ab−/ON) patients in the peripapillary retinal nerve fibre layer (pRNFL) and inner nuclear layer (INL) and measurements of AQP4 -Ab levels were more important in patients with ON who will develop NMO compared with other autoantibodies.4 ,5 However, the pathogenesis of definite NMO involves more than AQP4-Ab+/ON. Other specific NMO-immunoglobulin antibodies or myelin-oligodendrocyte glycoprotein antibodies (MOG-Ab) also play substantial pathogenic roles, especially in AQP4-Ab-seronegative NMO.6 ,7 Furthermore, AQP4-Ab+/ON usually completely fulfils the criteria of NMO after many occurrences of relapse-remission and long-term disease. A critical unanswered question is whether the structural injury to the optic nerve for patients with NMO is similar to the injuries in AQP4-Ab+/patients with ON. Furthermore, whether these injury patterns are different in NMO or AQP4-Ab+/ON from AQP4-Ab−/ON associated with AQP4 expression in the retina has yet to be determined. These structural alterations and their associations with AQP4 could be very useful for uncovering the complicated pathogenesis of NMO. To date, there have been no studies focusing on the structural changes associated with AQP4.

    Most previous studies have focused on the distinct structural injury patterns of NMO with ON (NMO-ON) compared with MS with ON (MS-ON)8–13 and sought structural biomarkers to predict if ON would progress to MS or NMO.4 ,14 However, these studies often pooled NMO and AQP4-Ab+/ON, which is one type of NMOSD, as a cohort, regardless of the differences between NMO and AQP4-Ab+/ON (NMOSD). Additionally, in previous studies, the macular retinal nerve fibre layer (mRNFL), retinal ganglion cell layer (RGCL) and inner plexiform layer (IPL) were often combined due to the poor resolution of optic coherence tomography (OCT) or an older version of the analysis tools from Microsoft, and previous studies may have neglected the feature alterations in mRNFL, RGCL and IPL. Previous researchers also failed to consider whether these distinct injury patterns correlated with AQP4 expression or ethnicity.15–17 In addition, in most previous studies, segmented macular layers were measured with manual analysis methods that lowered the repeatability of OCT measurements and the accuracy of the results.18

    In this study, we recruited patients with ON >6 months after ON onset from a Chinese population to avoid undue influence from the factors of disease duration and ethnicity. To maintain better homogeneity among the subjects, these ON patents were grouped into NMO-ON, AQP4-Ab+/ON (NMOSD) and AQP4-Ab−/ON cohorts. All the subjects underwent Spectralis-OCT with completely automated analysis to detect the pRNFL, every inner macular segmented layer and microcystoid macular oedema (MME) occurrence. Additionally, structural alterations associated with AQP4 expression in the retina and the diagnostic value of structural biomarkers to predict ON progression to NMO were evaluated.

    Subjects and methods

    The present study included 109 patients with ON (161 eyes) diagnosed in the Ophthalmology Department of the Chinese People's Liberation Army General Hospital according to the optic neuritis treatment trial (ONTT), and age-gender matched healthy controls (HCs) due to OCT measurements are influenced by subjects' age and gender.19 Based on serum AQP4-Ab test results in a cell-based assay, 37 patients (54 eyes) were grouped into the AQP4-Ab+/ON cohort and 45 patients (65 eyes) were placed into the AQP4-Ab−/ON group. Overall, 27 patients (42 eyes) were diagnosed with definite NMO by senior neurologists according to their clinical features, serum AQP4-Ab tests and manifestations observed in MRI scans.

    The inclusion criteria were as follows: patients with ON were required to meet the ON criteria for ONTT without having ON attacks for at least six months because the optic nerve should remain relatively stable beyond 6 months after an ON attack.20 patients with ON with AQP4-Ab seropositivity, excluding patients diagnosed with definite NMO according to the Wingerchuk diagnostic criteria for NMO, which were revised in 2007,21 were included in the AQP4-Ab+/ON cohort. After excluding patients who had been diagnosed with definite NMO, AQP4-Ab seronegative patients with ON were recruited into the AQP4-Ab−/ON cohort. patients with ON who met the revised Wingerchuk diagnostic criteria for definite NMO21 were recruited into the NMO-ON cohort. HCs were recruited from the staff of the Chinese PLA General Hospital and among patients' relatives.

    The exclusion criteria were as follows: subjects with a refractive error ≥±6.00 DS or ≥±2.00 DC due to the retinal aquaporin-4 antibody seronegativity thickness being disturbed by axial length and refractive error;22 papillary oedema; intraocular pressure ≥21 mm Hg; ocular disorders or a history of ocular surgery; and CNS disorders. OCT images with segmentation error in the automatic software analysis were excluded.

    Ophthalmology examinations

    Ophthalmology examinations included best-corrected visual acuity (BCVA), intraocular pressure, slit lamp microscopy and ocular fundus examinations conducted by professional ophthalmologists. Neural system examinations were performed on all patients with ON by senior neurologists. For the BCVA evaluations, the BCVA values were obtained using a Snellen Visual Chart and then transformed into logarithm of the minimum angle of resolution (logMAR) values. Among these values, counting the number of fingers held before the eye was transformed as a log MAR value of 1.85, hand movement as a logMAR value of 2, light perception as a logMAR value of 2.7 and no light perception as a logMAR value of 3.0.23

    OCT examinations

    OCT examinations were performed on all subjects with Spectralis OCT (Heidelberg Incorporation, Germany). The pRNFL thickness was evaluated with the 3.4 mm circle scanning mode around the optic head, and the inner macular segmented layers were measured with a multilayer segmentation algorithm for the macular volume scanning mode. All the images were analysed using completely automatic standardised optic nerve analysis software V.6.0.9 for Microsoft (Nisite) (Heidelberg Incorporation). The pRNFL thickness and inner macular segmented layers (including the mRNFL, RGCL, IPL and INL) were evaluated using OCT. Furthermore, the pRNFL was divided into six quadrants: the superior nasal (NS), inferior nasal (NI), superior temporal (TS), inferior temporal (TI), nasal (N) and temporal (T) quadrants. The mRNFL, RGCL, IPL and INL were divided into eight sectors according to the ETDRS (circle diameters: 1, 3 and 6 mm), which excluded the central fovea due to its few retinal ganglion cells. MME was defined according to the criteria that cystic lesions were located in the INL, which at least expanded to two adjacent B-scans on macular volume scanning.24 ,25

    Statistical analysis

    Cohort differences in age were analysed using the Kruskal-Wallis test, and the Pearson χ2 tests was used to control for gender, MME involved in the INL and bilateral eyes included dependency. To analyse the differences in OCT measurements and adjust for the interocular dependency of ON, multivariate linear regression models and least significant difference tests of analysis of variance were performed. Diagnostic specificity and sensitivity of structural biomarkers were analysed with a receiver operating characteristic (ROC) curve and calculated cut-off points. All statistical analyses were performed using Statistical Package for the Social Sciences software V.19.0 (IBM Corporation). Statistical significance was defined as p<0.05.

    Results

    Demographic features of subjects

    The present study recruited 109 patients with ON (161 eyes), including 37 AQP4-Ab+/patients with ON (54 eyes), 43 AQP4-Ab−/patients with ON (65 eyes) and 27 patients with NMO-ON (42 eyes) as well as age-matched and gender-matched HCs (47 subjects, 94 eyes). The gender distributions of the four groups were match tested using the Pearson test (χ2=5.665, p=0.129), and their ages were matched using the Kruskal-Wallis test (χ2=5.336, p=1.49). The median logMAR values were 1.525±1.243 (mean±SD), range 0–3.7 for the NMO-ON cohort for BCVA; 1.218±1.192 (mean±SD), range 0–3.7 for the AQP4-Ab+/ON cohort; and 0.860±0.974 (mean±SD), range −0.079 to 3.7 for AQP4-Ab−/ON. The BCVA in AQP4-Ab−/ON eyes was better than the other cohorts by the Kruskal-Wallis test (χ2=7.033, p=0.030). The disease duration and number of episodes of ON for the NMO-ON cohort were greater than those of the other two cohorts. In the ON groups, the rates of concomitant MME were 3/54 (5.56%) for the AQP4-Ab+/ON cohort, 6/65 (9.23%) for the AQP4-Ab−/ON cohort and 5/42 (11.90%) for the NMO-ON cohort, respectively, and their differences were not significant (p=0.528; table 1).

    View this table:
    Table 1

    Subjects' demographic and clinical characteristics

    Structural injury to the pRNFL and inner macular segmented layers for AQP4-Ab+/ON compared with NMO-ON

    In contrast to HCs (mean±SD: 106.15±10.39 μm), the pRNFL thickness in the AQP4-Ab+/ON (mean±SD: 48.61±17.43 μm) and NMO-ON (mean±SD: 47.88±21.16 μm) cohorts decreased sharply 6 months after ON onset. Comparing AQP4-Ab+/ON to NMO-ON, the pRNFL thickness, as well as its six quadrants, showed no differences after controlling for interocular and bilateral eye dependencies. In contrast to HCs, the volumes of the mRNFL, RGCL and IPL in the AQP4-Ab+/ON and NMO-ON eyes 6 months after ON onset were reduced remarkably. However, the INL volume of the macula in AQP4-Ab+/ON and NMO-ON eyes was thicker compared with HCs (p=0.000). The volumes of the mRNFL (p=0.802), RGCL (p=0.583), IPL (p=0.624) and INL (p=0.252) as well as their average thicknesses of eight sectors by ETDRS in the AQP4-Ab+/ON cohort were similar to those in the NMO-ON cohort after controlling for interocular and bilateral eye dependencies (Table 2 and figure 1).

    View this table:
    Table 2

    pRNFL thickness and inner macular segmented layers measurement in AQP4-Ab+/ON, AQP4-Ab−/ON and patients with NMO-ON

    Figure 1
    Figure 1

    The spatial distribution of the peripapillary retinal nerve fibre layer (pRNFL) and segmented macular layer thickness loss in optic neuritis patients with serum AQP4-Ab positive (AQP4-Ab+/ON), optic neuritis patients with serum AQP4-Ab negative (AQP4-Ab−/ON) and neuromyelitis optica patients with optic nerve affected (NMO-ON). HC, healthy controls; INL, inner nuclear layer; IPL, inner plexiform layer; mRNFL, macular retinal nerve fibre layer; RGCL, retinal ganglion cell layer.

    Distinct structural injury patterns in pRNFL and inner macular segmented layers in AQP4-Ab+/ON and NMO-ON compared with AQP4-Ab−/ON

    The pRNFL in the three ON cohorts decreased substantially relative to the HC cohort (p=0.000). Further evaluation showed that for the AQP4-Ab+/ON and NMO-ON eyes the pRNFL was thinner than in AQP4-Ab−/ON eyes (p=0.001), which was mainly distributed in the NS, N, NI and TI quadrants after controlling for interocular and bilateral eye dependencies (table 2 and figure 1). For the inner macular segmented layers, the volumes of the mRNFL, RGCL and IPL in three ON cohorts decreased substantially relative to HC eyes (all p=0.000). Inversely, the INL volume in three types of ON grew thicker compared with HC eyes (p=0.000). Compared with AQP4-Ab−/ON, the volumes of the mRNFL and IPL in both AQP4-Ab+/ON and NMO-ON were reduced. For the INL volume, NMO-ON lost more than that of AQP4-Ab−/ON (p=0.043), and there was no difference for AQP4-Ab−/ON versus AQP4-Ab+/ON (p=0.353). The average thicknesses for eight sectors divided by ETDRS of the inner macular segmented layers presented with similar alterations to their volumes in the three ON cohorts, as shown in figure 1.

    ROC analysis of the pRNFL and inner macular segmented layers for the diagnosis of NMO-ON

    ROC analysis was performed on OCT measurements with significant differences for AQP4-Ab+/ON versus NMO-ON, mainly including global and four quadrants of the pRNFL, the volumes of the mRNFL, IPL and INL. The outcomes demonstrated that the global pRNFL and the N NI of pRNFL thicknesses (all area under the curve (AUC)>0.7) had small values, which allowed differential diagnosis of NMO from common ON. Among them, when the NI of the pRNFL thickness was reduced to ≤46.5 μm (AUC 0.772, sensitivity 89.2% and specificity 57.5%) 6 months after ON onset, NMO should be considered (table 3 and figure 2).

    View this table:
    Table 3

    The AUC and cut-off values of pRNFL and inner macular segmented layers for differential diagnosis of NMO-ON

    Figure 2
    Figure 2

    Receiver operating characteristic curves of the peripapillary retinal nerve fibre layer (pRNFL) and inner segmented macular layers for predicting values of optical coherence tomography that could diagnose neuromyelitis optica. INL, volume of inner nuclear layer; IPL, volume of inner plexiform layer; mRNFL, volume of retinal nerve fibre layer; N, nasal quadrant of pRNFL; NI, nasal inferior quadrant of pRNFL; NS, nasal superior quadrant of pRNFL; TI, temporal inferior quadrant of pRNFL .

    Discussion

    AQP4-Ab+/ON, as an early stage or limited type of NMO, is classified as an NMOSD. It will pass through long-term remission-relapse before progressing to definite NMO. The autoimmune injuries initiated by AQP4 binding with AQP4-Ab play a key role in the pathogenesis of NMO and AQP4-Ab+/ON or NMOSD. However, definite NMO might be more complicated, involving other pathogenic factors, such as MOG-Ab, than AQP4-Ab+/ON or NMOSD, especially for definite NMO with AQP4-Ab seronegativity. Whether there are any differences in the structural injuries in the optic nerve and retina in these types of ON, as well as their pathogenic mechanism, remains a question.

    The outcomes of the present study revealed that AQP4-Ab+/patients with ON had the same structural injury patterns as patients with NMO-ON, and these patterns were distinct from the injuries in AQP4-Ab−/patients with ON. AQP4-Ab+/ON and patients with NMO-ON suffered more damage in the pRNFL and inner macular segmented layers than did AQP4-Ab−/patients with ON. These outcomes showed that AQP4-Ab+/ON and NMO-ON could have a common pathway for pathogenesis even if the AQP4-Ab+/ON underwent long-term remission-relapse before converting to definite NMO or other pathogenic factors apart from AQP4 need to be involved. Additionally, NMO-ON and AQP4-Ab+/ON (NMOSD) caused worse structural injuries in the optic nerve than AQP4-Ab−/ON. The latter ON is likely to progress to MS or isolated ON after frequent recurrences of ON in clinical practice. Previously, studies about structural injuries by OCT in NMO-ON and AQP4-Ab+/ON (NMOSD) versus AQP4-Ab−/ON were scarce.4 However, for NMO-ON versus MS-ON, there have been many studies8–13 in recent years, as well as the present study, that showed that NMO-ON caused more damage in the pRNFL and inner macular segmented layers. The possible reasons are that, in contrast to MS with demyelination alone, NMO and NMOSD not only cause demyelination injury but also produce neural necrosis and cavities.26 ,27

    The outcome of the present study also showed that compared with AQP4-Ab−/ON, NMO-ON and AQP4-Ab+/ON(NMOSD) preferentially damaged the superior, nasal and inferior quadrants of the pRNFL. Regarding the inner macular segmented layers, they caused more damage to the mRNFL, IPL and INL than did AQP4-Ab−/ON. In studies12 ,13 ,16 about structural alterations of NMO-ON versus MS-ON, Ratchford's13 OCT study demonstrated that NMO-ON caused more loss of the pRNFL thickness (>15 μm) than MS-ON. Naismith's12 study showed that pRNFL thickness loss was mainly distributed in the superior, nasal and inferior quadrants when controlling for contrasting visual acuity. Together, these results were quite similar to the differences observed in AQP4-Ab+/ON (NMOSD) and NMO-ON versus AQP4-Ab−/ON in the present study and in AQP4-Ab+/ON versus AQP4-Ab−/ON in our previous study.4 These differences might be explained by NMO-ON and NMOSD minimally affecting the small diameter axons, which are mainly distributed in the temporal quadrant and are preferentially affected in MS-ON.28 ,29 The other reason could be that the superior, nasal and inferior quadrants of the pRNFL show strong AQP4 expression due to AQP4 being mainly expressed by astrocytes or Müller cells in perivascular regions and facing the vitreous body side.30–32 Consequently, the pRNFL in the superior, nasal and inferior quadrants suffered more severe injuries in NMO or NMOSD.

    Previous studies have shown that inner macular segmented layers suffered more severe damage in NMO or NMOSD than MS-ON in general, but for each macular segmented layer, there are controversial studies.4 ,10 ,16 ,17 ,33 Schneider's10 study showed that the average thickness in the mRNFL, RGCL and IPL was lower in NMOSD than MS-ON but was not significantly different in the INL when controlling for MME influences. Fernandes' study16 revealed that the average INL thickness in NMO was thicker without differences in the mRNFL and RGCL plus IPL than MS-ON, and this implied that the alteration of the INL could be a potential marker to diagnose NMO. Our previous study4 showed that the volume of RGCL plus IPL was lower than that of the INL increase in AQP4-Ab+/ON (NMOSD) compared with AQP4-Ab−/ON. In the present study, the outcomes demonstrated that volumes of the mRNFL and IPL in NMO-ON or AQP4-Ab+/ON (NMOSD) were lower than those in AQP4-Ab−/ON without a difference in the RGCL, and the INL volume increased in NMO-ON compared with AQP4-Ab−/ON when controlling for MME influences. These different outcomes could be due to the differences in the criteria for grouping, controlling variates, sizes of the samples and OCT machines. The mechanism underlying the increased damage to the mRNFL, IPL and INL in NMO-ON compared with AQP4-Ab−/ON remained unclear. According to previous studies, AQP4 is most strongly expressed by Müller cells in perivascular regions and the inner limiting membrane and decreases towards the outer limiting membrane.31 ,32 As is known, in the retina, shallow retinal vessels mainly lie in the RNFL and are abundant in the superior and inferior quadrants around the optic disc. The deeper microvascular net mainly lies in the INL. Therefore, we inferred that these injury patterns could be associated with AQP4 expression.

    ROC curve analysis showed that when the NI quadrant of the pRNFL thickness was reduced to ≤46.5 μm (AUC 0.772, sensitivity 89.2% and specificity 57.5%) 6 months after ON onset, NMO was considered. Previous studies demonstrated that monocular ON caused the pRNFL to lose >15 μm in thickness in comparison with the contralateral eye, which is more likely to occur in NMO (75%);13 with each micrometre of pRNFL thickness lost after ON onset, the odds of progressing to NMO increased by 8%.12 These results could be potential clues to the differential diagnosis of NMOSD or NMO. However, due to the individual differences of OCT imaging and the small size of the sample, all the findings must be interpreted with caution.

    There were some limitations to this study. Even though we adjusted for age, gender, interocular, MME and bilateral eye dependencies, there were variations in BCVA and episodes of ON onset that remained uncontrollable, which would worsen the reliability of the outcomes of the study. Additionally, due to the cross-sectional study, unavoidable individual differences in OCT imaging and small sample size, the results should be validated in future longitudinal studies with larger sample sizes.

    In conclusion, AQP4-Ab+/ON produced similar structural injury patterns to NMO-ON. The pRNFL, mRNFL and IPL in ON and the INL in NMO-ON suffered more damage than those in AQP4-Ab−/ON, which could be associated with strong aquaporin-4 expression. The NI of pRNFL thickness could be a potential clue to predicting ON progression to definite NMO. Due to the size of the sample and individual differences in the OCT parameters, the results need to be verified in a larger sample size and in longitudinal studies.

    References

    1. ↵
      1. Levin MH,
      2. Bennett JL,
      3. Verkman AS
      . Optic neuritis in neuromyelitis optica. Prog Retin Eye Res 2013;36:159–71.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Jarius S,
      2. Wildemann B,
      3. Paul F
      . Neuromyelitis optica: clinical features, immunopathogenesis and treatment. Clin Exp Immunol 2014;176:149–64. doi:10.1111/cei.12271
      OpenUrlCrossRefPubMed
    3. ↵
      1. Papadopoulos MC,
      2. Verkman AS
      . Aquaporin 4 and neuromyelitis optica. Lancet Neurol 2012;11:535–44. doi:10.1016/S1474-4422(12)70133-3
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Peng C,
      2. Wang W,
      3. Xu Q, et al
      . Structural Alterations of Segmented Macular Inner Layers in Aquaporin4-Antibody-Positive Optic Neuritis Patients in a Chinese Population. PLoS ONE 2016;11:e0157645.
      OpenUrl
    5. ↵
      1. Li H,
      2. Zhang Y,
      3. Yi Z, et al
      . Frequency of autoantibodies and connective tissue diseases in Chinese patients with optic neuritis. PLoS ONE 2014;9:e99323. doi:10.1371/journal.pone.0099323
      OpenUrl
    6. ↵
      1. Kitley J,
      2. Waters P,
      3. Woodhall M, et al
      . Neuromyelitis optica spectrum disorders with aquaporin-4 and myelin-oligodendrocyte glycoprotein antibodies: a comparative study. JAMA Neurol 2014;71:276–83. doi:10.1001/jamaneurol.2013.5857
      OpenUrl
    7. ↵
      1. Kitley J,
      2. Woodhall M,
      3. Waters P, et al
      . Myelin-oligodendrocyte glycoprotein antibodies in adults with a neuromyelitis optica phenotype. Neurology 2012;79:1273–7. doi:10.1212/WNL.0b013e31826aac4e
      OpenUrlCrossRefPubMed
    8. ↵
      1. Papais-Alvarenga RM,
      2. Carellos SC,
      3. Alvarenga MP, et al
      . Clinical course of optic neuritis in patients with relapsing neuromyelitis optica. Arch Ophthalmol 2008;126:12–16. doi:10.1001/archophthalmol.2007.26
      OpenUrlCrossRefPubMed
    9. ↵
      1. Syc SB,
      2. Saidha S,
      3. Newsome SD, et al
      . Optical coherence tomography segmentation reveals ganglion cell layer pathology after optic neuritis. Brain 2012;135:521–33. doi:10.1093/brain/awr264
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Schneider E,
      2. Zimmermann H,
      3. Oberwahrenbrock T
      . Optical coherence tomography reveals distinct patterns of retinal damage in Neuromyelitis Optica and MultipleSclerosis. PLoS ONE 2013;8:e66151. doi:10.1371/journal.pone.0066151
      OpenUrlCrossRefPubMed
    11. ↵
      1. Ziemssen T,
      2. Ziemssen F
      . Perspectives of an innovative ophthalmological technology: optical coherence tomography (OCT)—what should be of interest to the neurologist? Clin Neurol Neurosurg 2013;115(Suppl 1):S55–9. doi:10.1016/j.clineuro.2013.09.022
      OpenUrl
    12. ↵
      1. Naismith RT,
      2. Tutlam NT,
      3. Xu J
      . Optical coherence tomography differs in neuromyelitis optica compared with multiple sclerosis. Neurology 2009;72:1077–182. doi:10.1212/01.wnl.0000345042.53843.d5
      OpenUrlCrossRefPubMed
    13. ↵
      1. Ratchford JN,
      2. Quigg ME,
      3. Conger A, et al
      . Optical coherence tomography helps differentiate neuromyelitis optica and MS optic neuropathies. Neurology 2009;73:302–8. doi:10.1212/WNL.0b013e3181af78b8
      OpenUrlCrossRef
    14. ↵
      1. Garcia-Martin E,
      2. Pueyo V,
      3. Almarcegui C, et al
      . Risk factors for progressive axonal degeneration of the retinal nerve fibre layer in multiple sclerosis patients. Br J Ophthalmol 2011;95:1577–82. doi:10.1136/bjo.2010.199232
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Sotirchos ES,
      2. Saidha S,
      3. Byraiah G, et al
      . In vivo identification of morphologic retinal abnormalities in neuromyelitis optica. Neurology 2013;80:1406–14. doi:10.1212/WNL.0b013e31828c2f7a
      OpenUrlCrossRefPubMed
    16. ↵
      1. Fernandes DB,
      2. Raza AS,
      3. Nogueira RG, et al
      . Evaluation of inner retinal layers in patients with multiple sclerosis or neuromyelitis optica using optical coherence tomography. Ophthalmology 2013;120:387–94. doi:10.1016/j.ophtha.2012.07.066
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Monteiro ML,
      2. Fernandes DB,
      3. Apóstolos-Pereira SL, et al
      . Quantification of retinal neural loss in patients with neuromyelitis optica and multiple sclerosis with or without optic neuritis using Fourier-domain optical coherence tomography. Invest Ophthalmol Vis Sci 2012;53:3959–66. doi:10.1167/iovs.11-9324
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Droby A,
      2. Panagoulias M,
      3. Albrecht P, et al
      . A novel automated segmentation method for retinal layers in OCT images proves retinal degeneration after optic neuritis. Br J Ophthalmol 2016;100:484–90. doi:10.1136/bjophthalmol-2014-306015
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Zhao L,
      2. Wang Y,
      3. Chen CX, et al
      . Retinal nerve fibre layer thickness measured by Spectralis spectral-domain optical coherence tomography: The Beijing Eye Study. Acta Ophthalmol 2014;92:e35–41. doi:10.1111/aos.12240
      OpenUrlCrossRef
    20. ↵
      1. Gabilondo I,
      2. Martínez-Lapiscina EH,
      3. Fraga-Pumar E, et al
      . Dynamics of retinalinjuryafteracute optic neuritis. Ann Neurol 2015;77:517–28. doi:10.1002/ana.24351
      OpenUrlCrossRefPubMed
    21. ↵
      1. Wingerchuk DM,
      2. Lennon VA,
      3. Lucchinetti CF, et al
      . The spectrum of neuromyelitis optica. Lancet Neurology 2007;6:805–15.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Jonas JB,
      2. Xu L,
      3. Wei WB, et al
      . Retinal Thickness and Axial Length. Invest Ophthalmol Vis Sci 2016;57:1791–7. doi:10.1167/iovs.15-18529
      OpenUrl
    23. ↵
      1. Schulze-Bonsel K,
      2. Feltgen N,
      3. Burau H, et al
      . Visual acuities “hand motion” and “counting fingers” can be quantified with the freiburg visual acuity test. Invest Ophthalmol Vis Sci 2006;47:1236–40.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Gelfand JM,
      2. Nolan R,
      3. Schwartz DM, et al
      . Microcystic macular oedema in multiple sclerosis is associated with disease severity. Brain 2012;135(Pt 6):1786–93. doi:10.1093/brain/aws098
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Pott JW,
      2. de Vries-Knoppert WA,
      3. Petzold A
      . The prevalence of microcystic macular changes on optical coherence tomography of the macular region in optic nerve atrophy of non-neuritis origin: a prospective study. Br J Ophthalmol 2016;100:216–21.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Lucchinetti CF,
      2. Guo Y,
      3. Popescu BF, et al
      . The pathology of an autoimmune astrocytopathy: lessons learned from neuromyelitis optica. Brain Pathol 2014;24:83–97. doi:10.1111/bpa.12099
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Kerrison JB,
      2. Flynn T,
      3. Green WR
      . Retinal pathologic changes in multiple sclerosis. Retina 1994;14:445–51. doi:10.1097/00006982-199414050-00010
      OpenUrlPubMedWeb of Science
    28. ↵
      1. Petzold A,
      2. de Boer JF,
      3. Schippling S, et al
      . Optical coherence tomography in multiple sclerosis: a systematic review and meta-analysis. Lancet Neurol 2010;9:921–32. doi:10.1016/S1474-4422(10)70168-X
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Bock M,
      2. Brandt AU,
      3. Dörr J, et al
      . Patterns of retinal nerve fiber layer loss in multiple sclerosis patients with or without optic neuritis and glaucoma patients. Clin Neurol Neurosurg 2010;112:647–52. doi:10.1016/j.clineuro.2010.04.014
      OpenUrlCrossRefPubMed
    30. ↵
      1. Nagelhus EA,
      2. Veruki ML,
      3. Torp R, et al
      . Aquaporin-4 water channel protein in the rat retina and optic nerve: polarized expression in Müller cells and fibrous astrocytes. J Neurosci 1998;18:2506–19.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Li J,
      2. Patil RV,
      3. Verkman AS
      . Mildly abnormal retinal function in transgenic mice without Müller cell aquaporin-4 water channels. Invest Ophthalmol Vis Sci 2002;43:573–9.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Amann B,
      2. Kleinwort KJ,
      3. Hirmer S, et al
      . Expression and Distribution Pattern of Aquaporin 4, 5 and 11 in Retinas of 15 Different Species. Int J Mol Sci 2016;17:E1145. doi:10.3390/ijms17071145
      OpenUrl
    33. ↵
      1. Park KA,
      2. Kim J,
      3. Oh SY
      . Analysis of spectral domain optical coherence tomography measurements in optic neuritis: differences inneuromyelitis optica, multiple sclerosis, isolated optic neuritis and normal healthy controls. Acta Ophthalmol 2014;92:e57–65.
      OpenUrlCrossRef

    Footnotes

    • Chunxia Peng and Hongyang Li contributed equally.

    • Contributors CP and HYL are co-first authors and contributed equally. Design and conduct of the study by CP, HYL and SW. Collection, management, analysis and interpretation of the data by CP, LW, JW, SC, HZ and SZ. Preparation of the manuscript by CP, WW and QX. Critical revision of the manuscript was performed by CP and HL. Review and final approval of the manuscript by all the authors.

    • Funding This study was supported by the National High Technology Research and Development Program of China (863 Programme, NO. 2015AA020511).

    • Competing interests None declared.

    • Patient consent Obtained.

    • Ethics approval This study was performed with the approval of the ethics committee of the Chinese People's Liberation Army General Hospital, and the study adhered to the Declaration of Helsinki (the 2013 revision), the guidelines of the International Conference on Harmonisation of Good Clinical Practice and applicable Chinese laws.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement The data involved in this study are owned by the Department of Ophthalmology of Chinese PLA General Hospital. If you want to share the data, please contact to Prof. Shi hui Wei (weishihui706@hotmail.com).

    Linked Articles

    • At a glance
      Highlights from this issue
      Keith Barton James Chodosh Jost B Jonas
      British Journal of Ophthalmology 2017; 101 i-ii Published Online First: 18 Jul 2017. doi: 10.1136/bjophthalmol-2017-310943