Human amblyopia

1
Human amblyopia
With apparently normal looking retinas, and well
corrected optics, a small of the population
have reduced VA in one eye. We call this
Amblyopia. It is typically associated with
anomalous binocular vision strabismus,
anisometropia, or both, reduced or absent
stereopsis, and chronic suppression of the
amblyopic eye. Under normal circumstances,
amblyopia is asymptomatic in that the other eye
appears to have normal (or near normal) vision.
Optometry and ophthalmology have a long history
or treating this condition.
Four big questions (1) What is the
phenomenology and underlying neurology of
amblyopia? (2) Where is the anomaly located in
the visual pathway? (3) What causes it? (4) How
can we treat it?
2
1a. Phenomenology of amblyopia
The prevalence of amblyopia depends upon the
criterion used
Prevelance ()

• VA no better than criterion in amblyopic eye
• Difference in VA between the amblyopic and
  non-amblyopic eye

1-3
20/30
20/40
20/60
20/200
Criterion (unilateral VA no better than)
3
What is abnormal about amblyopic vision?
Amblyopes, by definition, have reduced VA in one
eye which is generally much worse for letters
than for gratings AND they are highly susceptible
to the crowding phenomenon.
Contrast sensitivity is typically reduced at high
and medium SF, but not at low SF
Position acuities are reduced by about the same
amount as letter acuity
Non-amblyopic eye
Vernier acuity
Amblyopic eye
Bisection acuity
4
What is normal about amblyopic vision?
Spectral sensitivity and dark adaptation are
normal as is low spatial frequency vision.
Vl
Log threshold luminance
sensitivity
Time in the dark
Conclusion photoreceptor function looks normal
5
1b The underlying neural anomaly?

• Competing hypotheses
• Low-pass filtering, similar to the effect of
  defocus.
• Re-mapping similar to that observed with ARC.
• Neural under-sampling similar to that observed in
  the peripheral retina.
• Topographical disarray or neural scrambling of
  the amblyopic projection.
• All of these hypotheses are inadequate to
  explain all of the phenomenology, and we are
  still trying to identify the neurological cause
  of amblyopic vision.

6
Psychophysics has been basis for mechanistic
models of human amblyopia.
1. Low Pass Filter model Sensitivity to
spatial contrast is selectively reduced at
higher SF in most amblyopes.
Low spatial frequency vision is spared!
normals
amblyopes
CSF
Explanations a) high SF signal is decreased
during development due to habitual blur. b) high
SF neurons are inherently more susceptible.
Contrast Sensitivity
(low)
(high)
Spatial Frequency
7
Problem for simple Low Pass Filter Model
Although CS and VA of amblyopes are qualitatively
similar to that of low-pass filtering (.e.g.
defocus), many aspects of amblyopic vision are
incompatible with simple low-pass filtering.
8
2. Distorted Retinotopic Map Local distortions
qualitatively similar to ARC are manifest by
strabismic amblyopes.
Task Draw a circle
Task Align small bar with triangles

Sireteanu, et al 1993
Bedell Flom, 1981
These look circular to strabismic amblyopes
9
Problem for Distortion of retinotopic map model
a) Anisometropic amblyopes seem not to exhibit
this re-mapping phenomenon, but they exhibit many
other amblyogenic characteristics. b) Knowledge
of the distortion failed to predict misperception
of gratings seen by the amblyopes, and it is not
clear how it would predict reduced VA and CS.
10
3. Neural undersampling Levi and Klein 1986 Due
to reduced representation of the amblyopic eye
image within the cortex, the neural image
arriving in the cortex is undersampled.
Monocularly deprived cortex
Normal cortex
Clearly there are fewer neurons in V1 responding
to the deprived eye in MD, this may also be the
case in human amblyopia.
11
Problem for Neural Undersampling
Grating studies fail to observe predicted visual
effects of neural (retinotopic) under-sampling
1. High SF above sampling limit should appear
as low SF. They do not (Hess et al, Thibos
Bradley). 2. SF above the sampling limit should
appear to drift in the opposite direction. They
do not. Hess Anderson
12
4. Neural disarray Hess et al, 1978
Hess et al, 1978 observed perceptual distortions
in grating stimuli. Reject a low-pass filtering
model of amblyopia, and gave the name tarachopia
(scrambled vision) as a substitute for the
classic term amblyopia (blunt vision) to
distinguish it from simple low-pass filtering.
Pseudonyms neural scrambling, uncalibrated
distortions, uncalibrated disarray, topographical
disarray, and intrinsic spatial disorder.
13
Problem for Tarachopia
Some of Hess et als own data, and data collected
by, Bradley Freeman, 1985, and Thibos and
Bradley 1993 suggests that many of the anomalous
perceptions reported by amblyopes are very
orderly
Hess et al
Thibos and Bradley
14
2. Where is this anomaly located?
Retina seems normal (fundus, ERG), VEP from V1/V2
is abnormal, fMRI in V1 is abnormal. V1 is top
candidate.
fMIR showing R and L hemispheres from 4 amblyopes
with V1/V2 showing high levels of neural activity
following visual stimulation through each eye
Occipital lobes
Fixing eye
Amblyopic eye
Fixing eye
Amblyopic eye
15
3. What is the cause of amblyopia?
Classical Model of Amblyopia Etiology
Problem few prospective studies!
Sample prospective study of infantile esotropes
normals
100
75
Unilat- eso
w/ stereopsis
w/ amblyopia
Altern-eso
0
0
4
7
10
13
4
7
10
13
Age (months)
Age (months)
16
Causality in human amblyopia is now a rather
confusing issue
17
4. Can we treat amblyopia?
Standard approach Reduce visual input from the
non-amblyopic eye while maximizing visual input
from the amblyopic eye. This is done by
optically correcting the amblyopic eye, and
sometimes forcing the patient to use this eye
to do visual exercises while either patching or
blurring the non-amblyopic eye. Occlusion
therapy, short-term occlusion, patching therapy
atropinization. Results it works in most young
patients and in some older patients, but never
perfectly, and there is often regression.
Beware of creating occlusion amblyopia (due to
monocular deprivation).
18
Animal experiments confirm that reverse occlusion
(monocular deprivation of the originally
non-deprived eye) during the sensitive period can
fully reverse the original effects.
Initial monocular deprivation
Reverse monocular deprivation
Period of reverse occlusion
1
2
4
5
6
7
1
2
4
5
6
7
3
3
19
Evidence of occlusion amblyopia comes from two
sources naturally occuring monocular
deprivation (e.g. unilateral cataracts), and
clinically induced deprivation by patching
therapy. If the patching is too much early in
development, then the patched (deprived) eye can
become blind.
Example Archives of Ophthalmology, 1959 2 yr.
Old boy diagnosed with amblyopia 35P Eso, onset 6
mo of age. Mother instructed to patch good eye
with elastoplast occluder for one month. Boy
seen at 1 month, cannot fixate with occluded eye.
Mother told to shift patch to other (initially
amblyopic) eye and go home. Frantic call, one
hour later. Child appeared to be completely
blind in the now uncovered and previously normal
eye. Back to office immediately. Child could
not see an adult standing next to him. Doctor
Hardesty agrees that the eye that had been
patched was now blind. The father who had been
a guard on a professional football team demanded
to know what I planned to do.His anger increased
For the first time in my practice I feared
violence!! With patch removed from originally
good and now blind eye, the child sat and did
nothing on day 1, gross objects could be made
out on day 2. The child has now been alternately
patched and has good acuity in both eyes.
20
Visual acuity following monocular deprivation and
reverse occlusion
Experiments on cats showed very interesting
results AFTER the reverse occlusion had been
terminated
Initially seeing eye
Initially deprived eye
Is reverse occlusion an effective treatment for
the original deprivation amblyopia? Do these
data show evidence of regression?
VA (c/deg)
Initial period of MD
Period of reverse MD
Post-treatment period
Donald Mitchell
21
By adjusting the ratio of reverse occlusion and
binocular vision, the success of the occlusion
therapy was improved. Conclusion binocular
input is important for successful therapy, at
least in cats who have amblyopia generated by
monocular deprivation
50 patching
70 patching
100 patching
Donald Mitchell
22
Evidence of reverse occlusion and regression in
humans
Patient had vitrectomy OD at 15 months of age and
had 10D of OD esotropia
Patient had cataract removed from RE at 3 days of
age followed by CL correction for aphakia and had
30D of RE esotropia at 3 months
Patch LE 6 days per week for 1 month
Patch RE 5 days per week for 6 months
15
Patch LE 6 days per week
13
24
11
20
9
VA (c/deg)
VA (c/deg)
16
No patching
7
12
5
8
3
4
1
0
15
17
19
21
4
6
8
10
Age (months)
Age (months)
VA measured with cortical VEP
OS
OD
Vernon Odom
23
Amblyopia therapy There are two primary
approaches (1) Patching the non-amblyopic eye,
which removes all retinal image contrast, but is
not 24/7, and (2) atropinization of the
non-amblyopic eye, which removes some of the
retinal image contrast (because of intermittent
defocus) and is 24/7. This method is often
called Atropine Penalization. Therefore, both
of the current standards for amblyopia therapy
share a common feature in that both produce
partial contrast deprivation in the non-amblyopic
eye.
Atropine
Patching
Results Both methods can generate improved
visual performance in the amblyopic eye.
Atropine has the huge compliance advantage.
Neither methods works for every amblyope.
24
CAM Treatment (1980s) Example of variation on
the short-term patching method that also
includes a component of active stimulation of
the amblyopic eye. Previously invented by Arneson
in 1930s in Iowa. The idea is that in addition
to decreasing the visual input from the
non-amblyopic eye, treatment works by actively
stimulating the visual input from the amblyopic
eye (controlled studies suggest that active input
is no more effective than the visual input
normally obtained).
25
Same idea, but 50 years earlier was invented by
Arneson in Iowa.
Does it work?
26
Conclusions

• Amblyopia primarily affects spatial vision at
  high and medium frequencies
• The underlying neurology is still unknown
• The neural anomaly is located in V1 (and possibly
  additional deficits in V2)
• Although monocular deprivation can clearly cause
  amblyopia, as can strabismus, there is some doubt
  that anisometropia is a causative agent.
• Occlusion therapy has proven itself to be
  successful, but lack of carefully controlled
  studies makes definitive conclusions impossible.
• Regression seems a common result in both humans
  and animals
• Binocular experience may be essential for the
  best outcomes.
• Improved vision in amblyopic eye can be achieved
  by treating with partial deprivation of
  non-amblyopic eye.
• Treatment benefits can often regress.