Advances in Amblyopia: What Have We Learned From PEDIG Trials?

abstract Amblyopia is the most common cause of preventable visual loss in children. This article
reviews treatment options, durations, and efficacy in randomized multicentered trials conducted
by the Pediatric Eye Disease and Investigator Group in the last decade. Parents and patients
should be counseled that many forms of treatment are efficacious, allowing the option of choice of
best-tolerated treatment method. Compliance is key to successful treatment. The course of
treatment is likely at least 6–12 months, with yearly follow-up suggested once amblyopia has been
treated to monitor for regression. Pediatrics 2013;131:540–547

Amblyopia is the most common cause of monocular visual loss in children and young adults with
an incidence of 1% to 3.5% in developed countries. This visual disorder is potentially reversible if
detected and treated at a young age. Amblyopia is the reduction in visual acuity attributable to
neurologic deficits in the visual output of an eye. It is caused by inhibition of neurologic signals in
the visual pathway of the amblyopic eye by the fellow eye during visual development. The
inhibition results in anatomic changes visible in the lateral geniculate nucleus and in the occipital
cortex. The areas in these structures receiving signal from the amblyopic eye are reduced in
volume, as are binocularly driven cells in the occipital cortex. Amblyopia results from any etiology
that creates a disparity in the vision from each eye. The most common causes of a visual disparity
resulting in amblyopia are refractive error (a need for glasses), strabismic (ocular misalignment),
other pathologies that occlude the visual axis, or a combination of these etiologies. Strabismic and
refractive etiologies account for 90% of all amblyopia. Plasticity in the neurologic structures allows
reassignment of the pathways from the amblyopic eye to the fellow eye, resulting in the visual
deficit. This phenomenon only occurs during a sensitive time period from birth to ∼6 to 8 years of
age, called the critical window of visual development. Treatment within the critical window can
also restore the vision to the amblyopic eye by altering the visual pathways. The neurologic
pathways are more susceptible to treatment earlier in the critical window.

In 1997, the Pediatric Eye Disease and Investigator Group (PEDIG) was formed to investigate the
treatment modalities for amblyopia by using randomized trials in a multicenter format. The group
is a collaborative network with .60 participating sites and .120 practitioners participating with
funding by the National Eye Institute. Although PEDIG has broadened its investigative subject
matter, this review focuses on the results of the studies related to amblyopia.

COMPARING TREATMENT OPTIONS FOR AMBLYOPIA

The treatment of amblyopia centers on eliminating the inhibitory signal from the fellow eye to
allow normal neurologic development of the amblyopic eye once refractive, strabismic, or
occlusive etiologies have been resolved.

BEST REFRACTIVE CORRECTION ALONE

Best optical correction improves visual acuity, but amblyopia, or reduced vision despite best
optical correction, was generally held to require more treatment. Some studies had suggested
spectacle correction alone as treatment of anisometropic amblyopia,1 but a prospective study
conclusively demonstrated the effectiveness of patching compared with spectacles alone.2 A
group of 84 children between ages 3 and 7 with refractive etiology for amblyopia were treated
with refractive correction alone. The vision was checked at 5-week intervals with plans for addition
of patching or atropine if the vision failed to improve from the previous examination. Surprisingly,
amblyopia improved with best refractive correction alone by $2 lines in 77% of patients. The
average improvement in visual acuity was 2.9 lines. Most children reached maximum
improvement in visual acuity after 15 weeks of spectacle usage.2 Twenty seven percent of patients
had complete resolution of their amblyopia. Amblyopia was most likely to resolve with glasses
alone in the subgroups with better baseline visual acuity and in children with lesser magnitude of
anisometropia. This study has shifted practice patterns of ophthalmologist to treat patients with
amblyopia with glasses until the visual acuity stops improving with glasses alone with follow-up of
patients at 5- to 6-week intervals before initiating patching or atropine treatment.

A similar improvement in visual acuity with spectacle correction alone in children aged 3 to 7 years
old was seen in patients with strabismus as the etiology for the amblyopia.3 In this small study of
12 patients with untreated strabismic amblyopia, the visual acuity improved by at least 2 lines in 9
patients treated with spectacle correction alone. In older children aged 7 to 17, there is maximum
improvement in visual acuity with penalization of the fellow eye compared with spectacle
correction alone.4

Patients with untreated severe bilateral refractive errors may develop bilateral amblyopia. A study
of patients with both moderate and severe bilateral amblyopia revealed excellent improvement in
visual acuity with glasses alone.5 The probability of binocular acuity of 20/25 after 1 year of
treatment was 74%.

AMBLYOPIA TREATMENT

In the setting of residual visual disparity following the maximum improvement in visual acuity with
spectacles, the treatment options for the remaining amblyopia include patching or atropine
penalization of the fellow eye. Severe amblyopia with visual acuity worse than 20/100, is generally
more difficult to treat than moderate amblyopia with beginning visual acuity better than 20/ 80.
This is in part because of the difficulty of asking the child to continue his or her activities of daily
living with the poor vision of the amblyopic eye while the fellow eye is occluded or atropinized.

These treatment options have advantages and disadvantages. Patching requires complete
occlusion of 1 eye. Disposable adhesive patches are the most common patching method, although
reusable spectacle-mounted eye patches or occluders are also popular. The number of hours of
assigned patching may vary from 1 hour to essentially all day. Patching can be difficult because of
the physical discomfort of the patch or its adhesive material, skin rashes or breakdown,
compliance with the number of hours of patching, and social stigma related to the patch. Atropine
inhibits the focusing ability of the eye, reducing visual acuity at near in all patients and distance
vision in some, for potentially several days. Atropine can be given daily in the fellow eye, only on
weekends, or any interval in between. Atropine causes mydriasis of the pupil, resulting in light
sensitivity in some patients. Reduced near visual acuity may affect schoolwork. Potential
complications of atropine also include a localized allergic reaction, inadvertent administration to
the incorrect eye, and toxicities from overdose.

PATCHING

The control arm of the first PEDIG amblyopia trial was designed to determine the efficacy of
aggressive patching (6 hours up to full-time daily).6 Aggressive patching was considered the gold
standard at that time. Children aged 3 to 7 years with moderate amblyopia (20/40 to 20/100) were
treated with patching either 6 hours daily (43% of patients) or full-time (17% of patients) at the
investigator’s discretion.7 The average visual acuity in the amblyopic eye was 20/63 at the
beginning of treatment. The etiology of the amblyopia was strabismic in 38% of patients,
anisometropic (difference in the refractive error between eyes) in 37%, and a combination of the 2
in 24%. After 6 months of treatment, 79% of patients had visual acuity of at least 20/30. The
overall improvement of visual acuity was 3.16 lines. This provided the evidence-based guideline
for expectations of improvement with amblyopia treatment.

Additional studies investigated the ideal number of hours of patching. Children 3 to 7 years of age
with moderate amblyopia were assigned to 2 hours of patching compared with 6 hours of patching
daily.8 After 4 months of treatment, there was no statistically significant difference in visual acuity
between the 2 groups. Sixty-two percent of patients achieved either 20/30 visual acuity or at least
3 lines of improvement from baseline with either of the patching regimens. Initially there was a
faster rate of improvement in the group patched 6 hours daily compared with 2 hours, but the
final visual acuity was the same for both groups after 6 months of treatment.9,10 This study
proved that a greater number of hours of patching does not have either a clinically or statistically
significant effect after 6 months of treatment of moderate amblyopia.8

Several authors have questioned the results of this study because compliance was monitored by
self-reporting.11,12 Only 6% of patients assigned patching for 6 hours patched for the prescribed
time 75% or less of the time.8 There would be no difference seen in effectiveness of treatment
between 2 hours of patching and 6 hours, if children assigned to 6 hours of patching only patched
2 hours. Monitoring compliance with occlusion monitors reveals that far fewer hours of patching
are preformed than reported to investigators.12,13 Additional PEDIG studies monitored
compliance with logs, telephone calls, and self-reporting to attempt to address these concerns
because electronic devices to measure compliance are not commercially available.12 The question
regarding the effect of the actual number of hours patched still remains unanswered. However,
these studies still provide useful information about the effect on vision of the prescribed number
of hours of patching.8,9,14,15

Other studies evaluated severe amblyopia (visual acuity 20/100 to 20/400).14 Children were
randomized to 6 hours of patching compared with full-time patching. The etiologies of the
amblyopia were similar to previous studies. The average visual acuity was 20/160 at the initiation
of treatment. The average improvement in visual acuity was 4.8 lines in the group patching 6
hours daily and 4.7 lines in the group with fulltime patching. Compliance with treatment was also
subjectively assessed. Subgroup analysis of etiology revealed no statistically significant difference
in the regimens.

Fewer hours of patching have not been compared with 6 hours of patching for severe amblyopia
specifically, but in 1 study of patients with anisometropic amblyopia, 2 hours of patching for both
moderate and severe amblyopia was compared with a control group with only spectacle
correction.15 Patching 2 hours resulted in a 0.8-line improvement in visual acuity after 5 weeks of
treatment compared with no improvement in the spectacle-only group, suggesting that 2 hours
daily patching may treat severe amblyopia. In another study, visual acuity improved 3.7 lines in
children with severe amblyopia treated with 2 hours of daily patching after ∼4 months.16
Although the net improvement in vision was less than in the study comparing 6 hours and full-time
patching in severe amblyopia cited earlier, the methods of treatment varied in the length of time,
and best refractive correction was worn before the start of patching, making comparison difficult.

The effect of age on patching success was evaluated in a randomized trial and in follow-up studies
of previous randomized trials.4,17,18 Age ,5 at the initiation of treatment had greater long-term
success than older ages in degree and stability of visual recovery.17,18 In children aged 7 to 12
years, 53% had $10 letters of improvement in visual acuity testing with glasses combined with
patching 2 to 6 hours daily or atropine penalization.4 In older children, aged 13 to 17 years, only
25% of children had similar improvement if previous amblyopia treatment had been tried, but if
their amblyopia had not been previously treated, 47% of this group had $10 letters of
improvement in visual acuity.4

Addition of near activities while patching revealed no statistically significant difference compared
with children performing distance activities while patching.16 Near activities included tasks such
as crafts, reading, writing, or computer/handheld video games. Compliance with assigned activities
was high, as indicated with monitored weekly logs, telephone interviews, and follow-up visit
discussions. The improvement in visual acuity averaged 2.6 lines in the distance activities group
compared with 2.5 lines in the near activities group.

Patching does not affect the refractive error in the patched eye.19 There was only negligible
change in the refraction in the patched eye over a 2-year period while patching. Patching may
affect the ocular alignment after treatment. In 1 study, 16% of patients with no previous
misalignment developed a microstrabismus after 6 months of patching, 13% of patients with
strabismus had worsening of the alignment, and 16% of patients had resolution of their strabismus
after treatment.20

For both severe and moderate amblyopia, patching is an effective treatment. Compliance is
generally improved with fewer required hours of patching, and fewer hours are equally effective in
moderate and severe amblyopia after 6 months of treatment (Table 1).

ATROPINE

The first PEDIG amblyopia trial compared atropine drops daily in the fellow eye with patching for
moderate visual loss (20/40 to 20/100) in children 3 to 7 years old.6,21 Patching was prescribed
for at least 6 hours daily. Compliance was monitored by subjective reporting at follow-up visits.
The average age of patients was 5.3 years. Visual acuity improved by 3.16 lines in the patching
group and 2.84 lines in the atropine group at 6 months. Although the improvement in visual acuity
was faster for the patching group, both treatments resulted in a statistically equivalent visual
acuity improvement at 6 months. Seventy-four percent of the atropine group achieved 20/30
visual acuity, as did 79% of the patching group after 6 months of treatment. There remained a 1.8-
line difference between the amblyopic eye and the normal eye in both groups. There was no
difference in success of treatment dependent on the etiology of the amblyopia. Neither patching
nor atropine had unexpected side effects of treatment. Atropine had a higher degree of patient
acceptability. In a follow-up study at 2 years, there continued to be no statistically significant
difference in the improvement in vision between atropine treated groups and patched groups.22
Furthermore, in follow-up at 10 years of age, the mean visual acuity in the amblyopic eye was
20/32 for both atropine and patch treated eyes.17 Forty-two percent of patients patched had
20/25 or better visual acuity compared with 49% of patients treated with atropine. Combining
both groups, 64% of patients continued to have .1 line of interocular difference between their
normal and amblyopic eye. The mean difference was 2.0 lines. Younger age at initiation of
treatment was associated with better final visual acuity. There continued to be no difference in
final visual acuity based on treatment. This led to the recommendation that patching 6 hours daily
had equal efficacy to atropine penalization daily for the treatment of moderate amblyopia. Either
modality could be used as the initial treatment of amblyopia because an individual patient may
respond differently to treatment than the PEDIG group.

The Amblyopia Treatment Index was developed to assess the psychosocial impact on the child and
family of amblyopia treatment.23 This 20-question test assesses quality-of-life issues with regards
to differing treatments for amblyopia after 5 weeks of treatment. It has been internally validated
with high reliability.23,24 The Amblyopia Treatment Index revealed that atropine treatment
compared with patching was consistently better tolerated by the child and family with regard to 3
factors: adverse effects of treatment, difficulty with compliance, and social stigma.25 If fewer
hours of patching were prescribed so that patching could be completed at home, away from peers
and friends, the social stigma scores were not as negative.26 Assessing the psychosocial burdens
of treatment may increase compliance with treatment, ultimately resulting in better visual acuity.

Atropine is thought to function by its inhibition of accommodation, preventing the normal eye
from being used at near. Eyes that are hyperopic or farsighted use accommodation, or focusing to
present clear images to the retinas. In the presence of atropine, full cycloplegic correction of the
hyperopia would be needed to maintain clear vision in the distance. Patients with severe
amblyopia may be expected to continue using the normal eye at near if the vision in the amblyopic
eye is worse than the blur achieved by the atropine. Yet a study monitoring the fixation preference
at near revealed improvement in visual acuity in the amblyopic eye even without evidence of
fixation switch to the amblyopic eye.21 The exact mechanism of atropine efficacy, therefore,
remains to be determined.

In addition, although some studies initially suggested that the effect of atropine could be
enhanced by prescribing less than the full cycloplegic refraction o the normal eye (no power in
lens of fellow eye),21,27 this result was refuted by a randomized study designed to investigate this
effect specifically.28 The improvement in moderate amblyopia was 2.8 lines in the atropine with
optical undercorrection (no power in lens of fellow eye) group compared with 2.4 lines in the
atropine with full optical correction group at 18 weeks. This difference was not statistically or
clinically significant. This study did not specifically address whether undercorrected glasses could
be effective if atropine treatment alone failed to produce improvement in the visual acuity.

Although atropine was initially prescribed daily for amblyopia treatment, studies have compared
daily atropine with weekend atropine for moderate amblyopia.29 Less frequent dosing should
improve compliance with treatment of both the child and family. After 4 months of treatment, the
improvement in visual acuity of the amblyopic eye was 2.3 lines for either atropine regimen in
children with moderate amblyopia. The vision improved to 20/25 in 47% of children using daily
atropine and in 53% of children using weekend only atropine. A study of the application of
atropine on weekends only in severe amblyopia revealed similar improvement in visual acuity.27
Patients with full glasses correction and weekend atropine had 4.5 lines of improvement by 18
weeks of treatment. Visual acuity improved to 20/40 or better in 21% of the atropine and optical
correction group. Atropine did not change the refractive correction of the nonamblyopic eye in
patientsovera6-monthto2-yearperiod.19 Atropine had a similar effect on the ocular alignment as
patching.20 Eighteen percent of patients with no previous misalignment developed a
microstrabismus (not clinically significant), 15% had deterioration of their alignment, and 16% had
resolution of their strabismus after treatment with atropine.

Initial therapy with atropine is equally efficacious in the improvement of visual acuity as patching
in moderate amblyopia. Patient expectations with severe amblyopia need to be managed.
Weekend atropine is likely equally efficacious and has greater compliance (Table 2).

CESSATION OF AMBLYOPIA TREATMENT

After adequate recovery of visual acuity has occurred, the ophthalmologist must consider
discontinuation of treatment with the expectation of preservation of visual acuity. The regression
risk after cessation of treatment varies from 6% to 67% in previous studies.30,31 In the follow-up
of children treated with either patching or atropine penalization, the visual acuity was monitored
for 1 year after cessation of either treatment to assess for vision stability.32 Recurrence was
considered a decrease in vision in the previously amblyopic eye of at least one-third of the
treatment effect. The recurrence rate was 25% in children who discontinued patching and 21% in
children discontinuing atropine. Recurrence was 4 times as likely in children who did not have a
gradual taper of their treatment (ie, reducing patching from 6 hours daily to 2 hours daily before
discontinuation resulted in better outcomes than immediate cessation). Other factors linked to
greater likelihood of recurrence included better visual acuity at the end of treatment, a greater
number of lines of improvement in visual acuity during treatment, and previous history of
recurrence.33 Excellent alignment, good stereoacuity, and younger age during treatment were not
associated with reduced risk of recurrence.

The lack of protective effect with older age at cessation of treatment is particularly contradictory
to previous beliefs regarding the plasticity of the visual system. Whereas treatment was previously
often discontinued at age 6 to 8, the PEDIG study supports the continuous monitoring of visual
acuity for at least 1 year after the cessation of amblyopia treatment.33 In an older group of
children aged 7 to 12 treated for amblyopia, the recurrence rate for loss of $2 lines was only 7%.34
These authors speculate that some of the visual gain in this group may have been from better
refractive correction, which would not be susceptible to reversal with cessation of amblyopia
treatment. Importantly, however, most of the visual gain in this older age group is sustained,
making treatment of amblyopia in this older age group worthwhile.

In summary, a portion of the gains in visual acuity from amblyopia is vulnerable to regression with
cessation of amblyopia treatment. Tapering of treatment improves the retention of visual acuity,
but diligent monitoring of visual acuity after cessation of treatment is required.

FUNCTIONAL IMPACT OF AMBLYOPIA TREATMENT

Visual outcomes of amblyopia treatment are often limited to high contrast visual acuity (black
letters on a white background) in the amblyopic eye. Visual function has many aspects such as
varying contrast sensitivity, stereoacuity, reading fluency and reading comprehension that have
recently been assessed.35–37

Contrast Sensitivity

Contrast sensitivity is reduced in amblyopic eyes. After treatment of the amblyopia, there
continues to be a slight loss of contrast sensitivity with fine visual acuity (smaller letters) using low-
contrast letters.35 Yet the overall distribution of contrast sensitivity is similar to children without
amblyopia.

Stereoacuity

Computer-generated 3-dimensional entertainments have increased in popularity. Stereoacuity at

near and distance is reduced in amblyopia. Stereoacuity can be measured with a series of tests
such as the Preschool Randot Test, Distance Randot test, and Frisby-Davis distance and near
test.38 The Frisby-Davis tests use real depth analysis compared with overlapping images
discernable by each eye separately in the Randot testing. The testing methods greatly influence
the outcomes. Studies of the effect on stereoacuity of successful amblyopia treatment using
Randot testing reveal that stereoacuity remains lower in children with treated amblyopia than in
nonamblyopic children of the same age.37 Amblyopia is less severe in eyes with better baseline
stereoacuity and less anisometropia. After treatment of amblyopia, better stereoacuity was
associated with better baseline stereoacuity and better visual acuity of the amblyopic eye at end
of treatment. Therefore, the best chance of improving stereoacuity is successfully treating the
amblyopia. Most children will have improvement in their ability to see computer generated 3-
dimensional images with treatment of amblyopia, but these children will not have the same
degree of computer-generated 3- dimensional ability as children without amblyopia. Functionally,
however, tasks that require depth perception such as sports and driving are not affected by
amblyopia, as shown by testing with the Frisby-Davis methods.

Reading

Reading is a crucial developmental requirement for school-age children. Reading fluency is a

measure of both reading speed and accuracy. Reading wasassessedmonocularlyin10-year-old
patients previously treated for amblyopia by using the Gray Oral Reading Test (4th edition).36 In
this study, amblyopic eyes were slightly slower and less accurate than nonamblyopic eyes, but the
reading comprehension was similar. Poorer visual acuity in the amblyopic eye or slower central
processing by the brain may have contributed to this result. The type size used in the reading
material was far larger than the visual acuity of the amblyopic eye, and the comprehension rates
in the amblyopic eye were comparable to the fellow eye. Studies of binocular reading ability have
shown contradictory results.39,40 Some studies have shown impaired binocular reading in
children with amblyopia, and others have shown no correlation between amblyopia and binocular
reading. There are many confounding variables that make comparisons of reading function with
regard to amblyopia history difficult. This important function needs to be studied further to
develop guidelines for children with regard to potential reading dysfunction.

CONCLUSIONS
In summary, amblyopia treatment is highly successful with ∼75% of children ,7 years of age
achieving resolution of the amblyopia with either patching or atropine. Best optical refraction is
essential to successful treatment. Patching and atropine penalization are equally effective for
treatment of moderate amblyopia. If one treatment method fails, the alternative method may be
initiated. Six hours of patching was equally efficacious to greater numbers of hours in moderate
amblyopia in randomized trials. In severe amblyopia, all patching regimens have not been directly
compared in randomized trials, but as little as 2 hours of patching was effective. Younger age is
associated with better final visual acuity, but treatment should be attempted in older children
without previous treatment because up to 47% of these patients had improvement in visual acuity
that was sustained. Upon cessation of treatment, visual acuity needs to be monitored to identify
and treat regression. Finally, amblyopia therapy has significant social and physical impact on the
child and family. Decisions regarding treatment that factor in the wishes of the family and child are
more likely to lead to successful outcome.