Review of amblyopia treatment : Are we overtreating children with amblyopia ?

Aim: Amblyopia is a common reason for outpatient treatment in childhood. This review aims to provide an overview of recent research on amblyopia treatment. Methods: A literature-based review was carried out of evidence available on amblyopia management with additional personal comments on some issues. Results: Recent research evidence has shown that: refractive adaptation is a significant component of the therapeutic response, compliance with patching is problematic, most of the response to patching occurs within 6–8 weeks, and recurrence after cessation of treatment is less frequent if patching is weaned. However, this evidence from clinical trials has yet to be fully incorporated into practice: we are overtreating children. Conclusions: Utilising information gleaned from recent research will enable clinicians to reduce the amount of patching which children being treated for amblyopia receive, with benefits to compliance with treatment and use of health service resources.


Background
In 2007/8 there were almost 500,000 consultations in UK eye clinics and orthoptic departments for children aged 9 years and under, accounting for 14% of all attendances for this age group (http://www.hesonline.nhs.uk).It is thought that 90% of this workload relates to the treatment of amblyopia and strabismus. 1 Amblyopia treatment is therefore one of the commonest reasons for NHS outpatient treatment.
It has long been assumed that almost all amblyopia is fully treatable, given compliance or concordance with the regime recommended by health professionals, and there has been some evidence to support this. 2 What has become apparent, however, from many recent large clinical trials, is that a significant proportion of children do not regain normal visual function, or even driving vision, following amblyopia treatment.
Patching the preferred eye of a child who has amblyopia in the other eye, renders the child visually impaired, and is not a treatment to be undertaken lightly.Treatment-induced visual impairment is a major reason for lack of compliance with treatment, and a major reason for loss of children to follow-up.It is a particular problem for those children whose vision remains poor despite treatment.
There is currently no way to accurately predict the outcome of amblyopia treatment at presentation, except perhaps in those cases where there are significant anatomical abnormalities.One of the challenges facing clinicians treating amblyopia is to identify those children who are not responding to treatment, and to have the courage to abandon treatment if alternative strategies are exhausted.This is not done often enough; and children either suffer ineffectual treatment which impairs their vision, or appear on statistics of patients who have failed to attend their appointments.
The purpose of this article is to indicate to clinicians how to achieve the best possible visual acuity for children being treated for amblyopia while causing a minimum of disruption to the child.

Definitions
The original meaning of amblyopia is 'dull sight', and so originally the term covered all forms of impaired vision.This has led to the distinction in the orthoptic literature between 'organic' amblyopia, in which poor vision is associated with disease of the eye, and 'functional' amblyopia, in which it is not.The term 'organic amblyopia' is not used in medical, as opposed to orthoptic, practice and the term 'functional' amblyopia does not distinguish between the various causes of cerebral visual impairment, of which amblyopia is one.This way of subdividing amblyopia should be abandoned.
The term amblyopia, as currently used, refers to a condition in which visual development is disrupted, in one or both eyes, because of the absence of a focused image on the fovea, with consequent failure of normal stimulation of the developing visual cortex and associated brain areas.This leads to a failure of formation of normal cerebral visual neuronal connectivity.In other words, so-called functional amblyopia is caused by an organic problem, but that problem is in the brain rather than in the eye. 3nlike most forms of cerebral visual impairment, amblyopia is potentially reversible, at least in part, provided the disorder causing the degradation of the foveal image is rectified at a sufficiently early stage.

Diagnosis
Amblyopia is often defined as a two-line difference in corrected visual acuity, in the absence of ocular pathology (although amblyopia can complicate structural ocular abnormalities).While this definition may be adequate (although arbitrary) for prevalence studies in adults, it is not adequate for children.In order to know that a measured reduction in corrected visual acuity is significant, age-related normal values for the test used have to be known, and it is only recently that these have begun to be defined. 4Furthermore, a cause for degradation of the foveal image has to be present.These causes are classified as stimulus deprivation (cataract, ptosis, etc.), refractive, strabismic (which is always unilateral), and mixed.
It is important to understand that visual screening of children does not specifically detect amblyopia, but rather identifies a visual deficit which may be due to refractive error alone 5 or other abnormality.
It is easy to assume, because amblyopia is common and other causes of visual impairment in children are relatively rare, that all visual loss in children is due to amblyopia.It is not uncommon to see children with undiagnosed cataract or retinal dystrophy who are being treated for 'amblyopia'.

Rationale for amblyopia treatment
It seems obvious that amblyopia should be treated; however, this is not as straightforward a matter as it would first appear.Not only is amblyopia treatment not uniformly successful, but also, with the exception of pure refractive correction, amblyopia treatment renders the child visually impaired for the duration of therapy.This leads to high rates of non-concordance with treatment. 6here is little evidence that monocular developmental visual impairment causes significant functional disability or impact on quality of life, as opposed to measurable deficits on psychophysical tests. 7Apart from a general desire to maximise visual function, two reasons are commonly given for treating amblyopia: maximising employment prospects and lowering the risk of bilateral visual impairment resulting from injury to the nonamblyopic eye.
Various visual standards apply to different forms of employment, from the lifeboat service to heavy goods vehicle driving. 8None of these standards are evidencebased, all are potentially subject to challenge under disability discrimination legislation, and all are likely to change in the UK with the introduction of new European legislation.
Individuals with monocular amblyopia are clearly more at risk of visual impairment following ocular trauma and pathology, as they do not have a normal 'spare eye', although most ocular pathology, as opposed to trauma, affects both eyes anyway. 9mprovement in the vision of amblyopic eyes following loss of vision in the normal eye is well documented in some adult, even elderly, individuals.Rahi et al. identified 254 individuals aged 11 years or older with unilateral amblyopia who were visually impaired after loss of vision in their non-amblyopic eye but had no other disorder affecting their amblyopic eye.Twentyfive (10%) of these had improved visual acuity in their amblyopic eye. 10 One unresolved question is whether amblyopia treatment should cease once the maximum visual acuity has been reached, as opposed to being recommenced if acuity subsequently deteriorates.If the goal of amblyopia treatment is to maximise the vision in the amblyopic eye in the event of deterioration of vision in the normal eye, then it may be that the maximum potential vision may be represented by the best acuity seen during treatment, as opposed to the vision at some arbitrary age of discharge from follow-up (typically 5-7 years).

How is amblyopia treated?
The initial phase of amblyopia treatment is to correct the cause of degradation of the retinal image, for example by removal of cataract or correction of refractive error.
][13][14][15][16] Following unilateral cataract removal, there is a continuing amblyogenic stimulus from aphakia or pseudophakia with lack of accommodation.While limited patching may suffice in the first year of life subsequent to unilateral cataract removal (with benefit on the development of binocular vision), 14 beyond the first year, !6 hours a day of patching are required, depending on acuity, up to at least the age of 5 years, in order to obtain good visual results. 15These onerous amounts of patching often affect the vision of the normal eye. 16

Refractive correction
The absence of a significant degree of bilateral refractive error or anisometropia should raise the suspicion that the cause of the visual loss is not refractive amblyopia.The degree of anisometropia should also be consistent with the acuity: for example, anisometropia of 0.75D would be consistent with a corrected acuity of 0.3 logMAR in the worse eye, but an acuity of 1.0 logMAR should be viewed with some suspicion.It is easy to generate apparent refractive errors in eyes with poor vision which do not fix well, leading to off-axis retinoscopy.
Assuming the visual loss is consistent with the refractive error, then correcting the refractive error becomes the first treatment.
8][19] Following correction of refractive error in children with refractive and strabismic amblyopia, patching or atropine treatment should be deferred for up to 24 weeks, assuming continued improvement during that period. 20,21tching Patching the normal eye of a small child with amblyopia is challenging, as anybody who has tried to do it will tell you.Failure to comply with recommended patching treatment is not evidence of moral deficiency on the part of the parents or the child.Before commencing treatment there should be a discussion with the parents concerning the intended goals of treatment, and in the absence of any accurate way to predict the outcome of treatment, an undertaking should be given to the parents that treatment will be abandoned at an early stage if it is proving ineffective.
How can we make patching easier for parents?The most effective way is to reduce the amount of patching they need to do.This includes reducing the amount done per day (and evidence from studies where patching has been monitored suggests most parents cannot manage more than 3-4 hours per day of patching anyway); reducing the duration of patching treatment; and abandoning patching which is demonstrably having no significant effect on vision.Fortunately, there are recent studies which provide evidence that reducing the daily dose of patching does not adversely affect the outcome.Many of these have been carried out by the US Pediatric Eye Disease Investigator Group (PEDIG).PEDIG studies have demonstrated that, for children <7 years with amblyopia of 20/40 to 20/80 (0.3-0.6 logMAR), 2 hours of patching per day is as effective as 6 hours per day. 22A further study showed that for children <7 years with amblyopia of 20/100 to 20/400 (0.7-1.3 logMAR), 6 hours of patching per day is as effective as 'full time' (all but 1 hour per day) patching. 23It is worth noting that approximately 50% of patients in both groups in this latter study failed to improve beyond 20/50 (0.4 logMAR).
Supportive evidence that 6 hours of patching per day is as effective as 12 hours per day comes from the Randomized Occlusion Treatment for Amblyopia Study (ROTAS). 24This study monitored the actual dose of patching received by children in the study and demonstrated that actual occlusion doses were substantially less than those prescribed.
In a separate series of studies from the same group, it was shown that the bulk of treatment effect from patching was achieved with a cumulative dose of between 150 and 250 hours. 25Specific patient characteristics (especially age) modified the steepness of this function: for example, a 0.20 logMAR gain in visual acuity required a cumulative dose of 170 hours for children at age 48 months and 236 hours at age 72 months.They also showed that visual improvement was correlated with >1 hour per day of patching. 26he implication of this is that, for strabismic and refractive amblyopia, there is little point in patching children for 3 hours per day (assuming this is achieved) for longer than around 80 days -say 12 weeks.However, local review of our own practice shows some children undergoing patching for years, and this is likely to be the case in other departments too.Partly this is due to a reluctance to accept that no further significant improvements in vision are occurring, a situation complicated by the test/test variability of quantitative visual acuity measurements in small children, which is greater at lower levels of acuity.

Atropine
Atropine used to be reserved for children who could not tolerate patching at all and so ended up being used for a selected group of patients who were likely to have a poor outcome (as one of the reasons patching was not tolerated was that visual acuity was poor and not improving).It does, however, have advantages as an amblyopia treatment compared with patching in that it cannot be reversed by the patient, and it allows some use of the better seeing eye.It would seem intuitive that atropine would not work for children who have a worse acuity in the amblyopic eye than they do in their atropinised normal eye; however, it has been shown that atropine is equally as effective as patching for acuities from 20/40 to 20/100 (0.3 to 0.7 logMAR), 27 and atropine given only at weekends is an effective treatment for severe amblyopia (0.75-1.3 logMAR). 28here are concerns about systemic side-effects of topical atropine; however, it appears to be extremely safe, particularly when used once or twice per week only.Drops or ointment and 1% or 0.5% concentrations are equally effective and safe.Atropine is therefore a particularly useful treatment for toddlers, who are likely to remove their patches.The usefulness of atropine for sensory deprivation amblyopia has not been formally tested.

Visual stimulation
Treatment of amblyopia by attempting to stimulate development of vision in the amblyopic eye has a long history.The CAM stimulator, a rotating grating pattern, was popular for a time in the 1960s but unfortunately was shown to be ineffective in one of the first randomised controlled trials in this area.Near activities are often recommended during amblyopia treatment but their effectiveness is not clear.More recently, an interactive computer system has been developed which has shown some promise and is the subject of further investigation. 29,30For the time being, however, these methods remain investigational.

Stopping treatment: abruptly or gradually?
There is some evidence that regression from maximum acuity following cessation of treatment is more likely to occur if treatment is stopped abruptly rather than weaned, 31 is more common in children with mixed strabismic and refractive amblyopia, 32 and is more likely to occur in those children in whom treatment had been most successful. 33

Treatment of children over 7 years
Traditionally, it has been assumed that amblyopia treatment is only possible in children under the age of 7 years, as beyond this age there was no potential for cortical plasticity.This is clearly an oversimplification, and a proportion of children will show a treatment response beyond the age of 7. 34 The age at which there is potential for treatment depends, amongst other things, on the age of onset and type of amblyopia.For example, plasticity in a case of unilateral complete congenital cataract lasts for 2 months following birth.It is possible that some adults with amblyopia would show visual improvement if they were patched; however, the induced visual impairment which would result would be unlikely to be tolerated.

Outcomes
In my view, the major issue confronting clinicians treating children with amblyopia is when to stop treatment.Most children currently undergo far more than the 150-250 hours of patching beyond which, studies suggest, little further improvement occurs. 25,35his is partly due to uncertainties about visual acuity measurements in children under 4 years, such as whether normal visual acuity measurements on single optotype tests will translate into normal acuities on crowded tests.For children over 4 years, however, these considerations should not apply.For these children, the difficulty is usually in accepting a subnormal level of acuity as the most which can be achieved with treatment.It is important to point out, though, that a subnormal level of acuity is the norm following treatment for many children.
In 1994, Woodruff et al. demonstrated in a multicentre UK study that only 48% of children treated for amblyopia achieved an acuity of 6/9 or better in the treated eye. 36While this appeared to reflect a poor standard of practice in the UK compared with Scandanavia, 2 Woodruff's findings have been confirmed, although not highlighted, by more recent studies.For example, in the PEDIG study ATS1 (atropine versus patching) 37% of children in the patching group and 47% in the atropine group had visual acuities 20/40 at outcome. 27This pattern is reflected across other trials, and is greater in trials of treatment of children over 7 years of age. 37,38y do some children not achieve normal visual acuity following treatment?
There are a number of reasons why children might not achieve normal visual acuity following amblyopia treatment.Some children may have structural abnormalities of the retina, optic nerve or central visual apparatus which are either not sought, or are not readily apparent with current imaging techniques. 39Others may have developed amblyopia at an early age and reversal of the changes which have occurred may no longer be possible by the time of diagnosis.

Key points
. The terms organic and functional amblyopia should be abandoned. .Amblyopia should only be diagnosed when: . acuity falls below the age-related normal range; . an amblyogenic stimulus consistent with the measured acuity is present. .Patching or atropine treatment should be deferred for up to 24 weeks following correction of refractive error in refractive and strabismic amblyopia, assuming continued improvement. .Patching is challenging: Do not recommend more than is required to obtain the maximum predicted acuitywhich is usually not 0.0 logMAR. .If there has been no significant change in acuity following 6-8 weeks of treatment with atropine or patching, stop treatment and reassess. .Total achieved doses of patching over 250 hours (3 hours per day for 12 weeks or 6 hours per day for 6 weeks) are unlikely to produce significant further improvements in acuity in strabismus and refractive amblyopia. .Deterioration of acuity following cessation of treatment occurs in some patients and may be less frequent if treatment is weaned. .Whether the acuity regained in an amblyopic eye is the maximum achieved during treatment or the acuity at age 7 years is not known.