·Clinical
Research·
Acute
changes in ganglion cell layer thickness in ischemic optic neuropathy compared
to optic neuritis using optical coherence tomography
Peter
W. MacIntosh1, S. Vijay Kumar2, V.R. Saravanan2,
Virna M. Shah2
1Illinois Eye and Ear Infirmary,
University of Illinois at Chicago, Chicago, IL 60612, USA
2Neuro-Ophthalmology Service, Aravind
Eye Hospital, Coimbatore 641014, India
Correspondence to: Peter W. MacIntosh. 1855 W. Taylor
Street, M/C 648, Chicago, IL 60612, USA. pmacint1@uic.edu
Received:
Abstract
AIM: To elucidate the
changes of different ganglion cell layer (GCL) thinning patterns between the
optic neuritis (ON) and non-arteritic anterior ischemic optic neuropathy
(NAION).
METHODS: A prospective,
observational study was conducted to evaluate the timing of GCL changes between
acute ON and NAION using optical coherence tomography.
RESULTS: Thinning on optical
coherence tomography in the NAION group occurs as early as 11d after symptomatic
onset of vision loss and follows an altitudinal pattern. The mean
superior-inferior GCL thickness difference in the NAION cohort was clinically
significant at 5.7 µm in the NAION cohort compared to controls of 0.8 µm (P=0.032),
but not significant in the ON group compared to controls with both groups
measuring 1.1 µm. Global thinning was significant for the ON group compared to
controls at 7.2 µm (P=0.011) but not the NAION group compared to
controls at 1.35 µm.
CONCLUSION: These findings suggest
that future treatments for NAION should be given early, and possibly before 11d
in order to prevent GCL and irreversible vision loss.
KEYWORDS: optic neuritis;
non-arteritic ischemic optic neuropathy; optical coherence tomography; ganglion
cell layer
DOI:10.18240/ijo.2020.01.17
Citation: MacIntosh
PW, Kumar SV, Saravanan VR, Shah VM. Acute changes in ganglion cell layer
thickness in ischemic optic neuropathy compared to optic neuritis using optical
coherence tomography. Int J Ophthalmol 2020;13(1):120-123
INTRODUCTION
In adults, the two most common optic
neuropathies are optic neuritis (ON) and non-arteritic anterior ischemic optic
neuropathy (NAION). ON responds well to steroids, but often patients are still
left with subtle visual dysfunction.
One study in India has placed the incidence of NAION at 37.8% and ON at
28.9% of patients presenting with acute disk edema[1].
There is no proven treatment for NAION, though clinical trials are currently
underway. Still, the exact treatment and timing of treatment remain unknown. A
better understanding of the changes to the optic nerve and ganglion cell layer
(GCL) may help better direct therapies in the future.
ON and NAION also share some
overlapping clinical features, which poses some difficulty in distinguishing
between the two. However, the diagnosis may generally be made based on age,
pain with eye movements, pattern of visual field loss and visual recovery. More
recently, different thinning patterns between these two entities have been
demonstrated on optical coherence tomography (OCT) by others[2].
Optic disc edema is most commonly associated with NAION, although it may also
be seen with one-third of ON cases[3-4].
In this study, we will further evaluate the timing of GCL changes between acute
ON and NAION using OCT.
SUBJECTS AND METHODS
Ethical Approval The Aravind Eye Care System
Institutional Review Board approved the study. The study was conducted in
accordance with the principles of the Declaration of Helsinki. Informed consent
was obtained from all patients.
A prospective, observational, single
centre study of consecutive patients was performed in the Outpatient Department
of Aravind Eye Hospital, Coimbatore, India between January to May 2017 to
compare the mean GCL thickness between eyes in two groups: ON and NAION.
Healthy, age-matched controls, evaluated with OCT were used for a case control
analysis.
Forty patients with acute ON and
their age-matched controls as well as 40 patients with acute NAION and their
age-matched controls were enrolled. Dilated fundus exam was done to assess for
disk edema followed by HD-OCT (Heidelberg Spectralis, Heidelberg Engineering,
Heidelberg, Germany) to assess the mean GCL thickness. Special attention was
paid to the superior and inferior quadrants and the differences in absolute
values in ON and NAION were compared between.
Patients had to meet all the
following criteria to be included in the NAION group: disc edema; no eye pain
or headache; sudden visual loss; Humphrey visual field with altitudinal field
defect; colour deficit; haemorrhages of the optic nerve head; crowding of the
optic disc in the other eye; normal erythrocyte sedimentation rate
[<(age+10)/2 for women and <age/2 for men] and C-reactive protein (<3
mg/L); age>18y; interval<3wk from onset of symptoms; no prior treatment.
Patients had to meet all the
following criteria to be included in the ON group: no previous ON;
interval<3wk from onset of symptoms; pain with eye movements; Humphrey
visual field with a central scotoma; colour deficit checked using Ishihara
plates; mild or no optic disc swelling; improvement on subsequent examination;
magnetic resonance imaging of the orbits with optic nerve enhancement;
age>18y; interval<3wk from onset of symptoms; no prior treatment.
Based on the diagnostic criteria,
patients were selected after a complete neuro-ophthalmological examination
which included the assessment of the best corrected visual acuity, using a
standard Snellen chart. Colour vision discrimination was assessed using
Ishihara charts. Fundus examination was done using Heidelberg Spectralis high
definition OCT (HD-OCT). Using standard acquisition protocol, each subject’s
eyes were scanned on the day of presentation to our institution. The macular
cube line scans were used to assess GCL.
The scanned area was
GCL analysis was done by calculating
the thickness of the GCL and the inner plexiform layer. The GCL thickness was
measured around the fovea in seven locations: supero-temporal, superior,
superior-nasal, infero-nasal, inferior, infero-temporal and global. The
average, minimum and sectorial thickness of the GCL were measured in an
elliptical annulus around the fovea. Global GCL thickness values, and those in
each of the said locations, expressed in micrometers (µm), were used for
analysis.
MRI was performed on either a
Siemens Skyra 3T or Philips ingenia 1.5T. Sequences obtained for the optic
nerve evaluation incuded T2 fat saturated axial, IR coronal, DWI, post contrast
fat saturated T
RESULTS
Each cohort (NAION and ON) included
40 eyes. In the NAION cohort, the mean age was 61.3±10.6y while in the ON
cohort it was 35.9±11.6y. For the age-matched control cohorts, the mean age was
62.0±11.8y and 35.0±11.6y, respectively. Of the 40 subjects in the NAION
cohort, 13 were women while 21 were women in the ON cohort. For the age-matched
controls in the ON cohort, 14 were women while 23 were in the NAION age-matched
control group. Mean time from the start of symptoms to presentation was
10.7±6.6d in the NAION cohort while it was 11.7±8.6d in ON (P=0.67) with
a range of 3-22d.
There was a significant
superior-inferior difference in GCL thickness in the NAION cohort but not in the
ON cohort. The mean superior-inferior GCL thickness difference in the NAION
cohort was clinically significant at 5.7 µm in the NAION cohort compared to
controls of 0.8 µm (P=0.032), but not significant in the ON group
compared to controls with both groups measuring 1.1 µm (Table 1). However,
global thinning was significant for the ON group compared to controls at 7.2 µm
(P=0.011) but not the NAION group compared to controls at 1.35 µm (P=0.685;
Table 1).
Table 1 Mean GCL thickness in superior
and inferior quadrants and global GCL thickness in ON, NAION and controls µm
|
Parameters |
ON |
ON-control |
NAION |
NAION-control |
|
Mean superior GCL thickness |
63.4 |
71.7 |
66.9 |
71.5 |
|
Mean inferior GCL thickness |
64.5 |
70.6 |
72.6 |
70.7 |
|
Superior-inferior GCL thickness
difference |
1.1 |
1.1 |
5.7 |
0.8 |
|
Mean global GCL thickness |
63.95 |
71.15 |
69.75 |
71.1 |
|
Global GCL thickness difference |
7.2 |
|
1.35 |
|
GCL: Ganglion cell layer; ON: Optic neuritis; NAION:
non-arteritic anterior ischemic optic neuropathy.
DISCUSSION
In this study we have shown that
ganglion cell loss occurs an altitudinal pattern in NAION, resulting in an
asymmetric thinning of the superior and inferior GCL hemispheres of the macula.
This asymmetric thinning is not seen in ON or in healthy controls. This
asymmetric change had a mean of 5.7 µm in NAION subjects. Moreover, this change
is evident as early as a mean of 11d after onset of symptoms. In ON, the GCL
thinning is global, but also present at a mean of 11d. The global thinning in
NAION, on the other hand, does not reach significance, since the hemispheric
thinning is countered by the lack of thinning in the contralateral hemisphere
in the GCL.
The most common visual field defect
in NAION is an altitudinal one, whereas in ON, the visual field loss is more
likely to a central scotoma or diffuse[5-7]
and both of these clinical findings correspond with the GCL thinning patterns
we have demonstrated in each respective group. The segmental ischaemia of the
optic nerve in NAION commonly presents with disc edema, and may be followed by
damage to other areas of the optic nerve[8-9].
Aggarwal et al[2] demonstrated in NAION
that at six months after presentation, if the visual field loss was superior or
inferior, it correlated with GCL thinning. However, there was both superior and
inferior GCL hemisphere thinning in 16 out of 23 eyes in their study.
In this study, we demonstrated that
thinning of the GCL may be seen in NAION and ON as early as a mean of 11d after
symptoms begin when compared to age-matched controls, which is consistent with
a retrospective study by Erlich-Malona et al[10]
and a prospective study by De Dompablo et al[11].
However, to the best of our knowledge, this early change in GCL thickness has
not been reported before in a prospective study of both NAION and ON.
Previous studies have also shown
similar asymmetric retinal nerve fibre layer (RNFL) thinning in NAION, but
demonstrated this thinning at 6mo after symptom onset, when the optic nerve
edema had resolved[7-12].
Others have suggested that GCL thickness may be a better measure of damage in
both ON[13-16] and NAION[2,17] during the acute phase, since the
GCL is not affected by the initial edema the way the retinal nerve fibre layer
frequently is in ON and universally is in NAION. Our findings suggest that GCL
loss can be seen at a mean of less than 2wk.
Kupersmith et al[16] studied patients with multiple sclerosis (MS) and
found a significant difference in GCL thickness between affected and unaffected
eyes at 1mo after onset of ON. However, there was no significant difference in
GCL thickness at presentation[16]. The initial difference
may have been underestimated due to progressive, subclinical GCL thinning that
has been seen in clinically normal eyes of patients with MS[14,18]. Our prospective study, however, did find GCL
thinning as early as 11d after symptom onset in both NAION and ON. Our study
compared subject eyes with control eyes from normal patients, thus removing the
effect of subclinical contralateral GCL thinning in MS patients.
Some authors have shown that disc
edema in NAION patients may resolve faster when systemic corticosteroids are
administered within 2wk of onset, and that they may also enjoy better visual
acuity and field recovery[14]. Experimental
models have suggested that the initial disc swelling in NAION results in a type
of compartment syndrome from the surrounding nerve sheath that leads to
ischaemic damage[19-21]. This
theory supports the early treatment of this initial swelling to reduce the
deleterious effect of the compartment syndrome on the neurons and visual loss.
Our study has demonstrated that
there is an asymmetric thinning of the GCL, as measured with OCT, and that this
asymmetry is significantly greater in NAION than in ON. This finding may be used to help
differentiate the two. More
importantly, our findings help in the recognition of acute GCL loss in both
NAION and ON which may help guide appropriate and early treatment to prevent
permanent vision loss. Currently there
is no known treatment for NAION and the treatment protocol for ON in the ON
treatment trial required intervention within 8d of symptom onset. However, our
results suggest that by as early as a mean of 11d, GCL thinning has occurred,
and possibly future treatments should aim at intervention within days or even
hours of symptom onset to prevent early GCL loss.
Our study has a few limitations
including not differentiating the various causes of ON, small sample size, lack
of contrast in all patient MRI studies and lack of randomization. Future
studies on this topic should include a prospective, randomized study with
serial OCTs to help in detailed evaluation of the GCL thickness in both NAION
and ON, particularly in the early days after symptom onset.
In conclusion, though
clinical assessment remains the mainstay of diagnosing ON and NAION, OCT-GCL
thinning patterns might aid in differentiating the two when the diagnosis is
not certain. Moreover, the thinning that occurs in both NAION and ON appears to
occur as early as 11d after symptom onset. Our findings suggest that the
changes to the GCL occur early, and successful treatments to prevent GCL and
vision loss may need to be applied within days or even hours after symptom
onset.
ACKNOWLEDGEMENTS
Foundations: Supported by Vision Research
National Eye Institute (NEI) P30 EY001792; Unrestricted Research to Prevent
Blindness (RPB) Departmental Grant.
Conflicts of Interest: MacIntosh PW,
None;
Kumar SV, None; Saravanan VR, None; Shah VM, None.
REFERENCES