Article Text


Charles Bonnet syndrome in patients with glaucoma and good acuity
  1. S A Madill1,
  2. D H Ffytche1
  1. 1Eye Department, King’s College Hospital, Denmark Hill, London SE5 9RS, UK
  1. Correspondence to: Mr Stephen Madill Eye Department, King’s College Hospital, Denmark Hill, London SE5 9RS, UK;
  1. Colin SH Tan

    Statistics from

    We read with interest the paper by Tan et al1 on Charles Bonnet syndrome (CBS) in Asian patients. Their finding of a lower CBS prevalence than European or North American surveys demands further investigation, although this may reflect the stringent criteria of hallucination complexity they used in making the diagnosis (thus excluding the commonest CBS hallucinations of coloured blobs and grid-like “tesselloptic” patterns2,3) and, as pointed out in the accompanying editorial comment, the relatively low prevalence of macular disease in their cohort. However, it is not this aspect of the report we found most intriguing—it was the observation that CBS occurred with good acuity. In fact, three of the four CBS patients described had a degree of impairment which placed them at risk for CBS (best eye acuity 0.3 or worse4). It is the remaining patient (patient three, a 72 year old man) who is of particular importance as his relative preservation of acuity bilaterally (20/30 RE, 20/40 LE) challenges the view that significant acuity loss is a prerequisite for “ophthalmological” visual hallucinations. This case mirrors four patients we have recently studied with CBS secondary to glaucoma and bilaterally good acuity. We describe the cases below and offer a pathophysiological mechanism for the association.

    In one sense, the finding that CBS occurs with preserved acuity is hardly novel. As cited by Tan et al,1 several previous reports have found such an association. However, all is not as it seems, the term CBS being used in different ways by different authors. Some use the term to describe visual hallucinations with insight, irrespective of the presence of eye disease, age or clinical context.5,6 Others use the term to describe the association of visual hallucinations with age and intact cognition, without reference to eye disease or hallucination phenomenology.7,8 Under these definitions it is hardly surprising that a patient with “CBS” has preserved acuity, the patients in these studies having a diverse range of conditions from delirium to Parkinson’s disease and beyond. In contrast, ophthalmologists and neurologists have used CBS to emphasise eye or visual pathway disease, with the phenomenology of the hallucinations and age being of secondary importance.9,10 Although each definition of CBS has its merits, the ophthalmological definition reminds us best of Bonnet’s original description and helps characterise a distinctive subgroup of visually hallucinating patients with predicable prognosis and specific pathophysiology.2,11 However, even CBS as defined ophthalmologically carries with it an inherent ambiguity: is it eye disease itself or the loss of acuity that is the important factor? The consistent finding of acuity loss as a risk factor4,12,13 suggests the latter, or at least that the central retina has a key role in the underlying pathophysiological mechanism.

    As part of a larger study into the visual phenomenology of CBS, we have recruited four patients with advanced glaucoma (three POAG and one chronic narrow angle) but preserved visual acuity. The age range of the patients was 81–91 years, three men and one woman. Their visual acuities ranged from 6/6 to 6/12 monocularly with all patients having 6/9 or better in their better eye. All had extensive field defects bilaterally and cup to disc ratios of 0.8 or greater in both eyes. Two patients had bilateral trabeculectomies now off treatment, one was on bimatoprost and Trusopt to both eyes and one on timolol 0.25% to both eyes. Two patients were bilaterally pseudophakic. The patient with chronic narrow angle glaucoma had previous surgical iridectomies. The duration of their hallucinations ranged from 6 months to 6 years. Three patients hallucinated in colour and one in black and white. The most common hallucination was of tessellopsia2 experienced by all the patients, with two patients seeing, in addition, formed buildings and two patients, letter-like shapes. There were also single reports of hallucinations of groups of people, animals, branching shapes (dendropsia2) and one patient described visual allesthesia.14 In three of the patients the hallucinations encompassed the entire visual field, in the fourth they were restricted to the visual field defect. None had hallucinations in other sensory modalities and all had insight into the nature of the experiences. The phenomenology of the hallucinations and the relative frequency of the different hallucination categories are consistent with previous descriptions of ophthalmologically defined CBS.2,3 Non-ophthalmological causes of visual hallucinations2,3 were excluded. As far as we are aware this is the largest case series of patients with visual hallucinations secondary to eye disease and bilaterally preserved visual acuity yet to be reported.

    Current aetiological theories of CBS emphasise the importance of deafferentation15 (both “physiological” through ganglion cell loss and “functional,” for example, related to blindfolding or cataract), the loss of visual input resulting in a change in cortical excitability.2 Although it has been assumed that deafferentation of sufficient severity to precipitate CBS implies a consequent loss of acuity, our cases and that of Tan et al1 suggest otherwise. Patients with advanced glaucoma can have a significant degree of ganglion cell loss and consequent physiological deafferentation without a loss of acuity, placing them at risk for CBS. This contrasts with age related macular disease where the loss of central retinal ganglion cells leads, indirectly, to an association of CBS with acuity loss. We conclude that reduced acuity is not a necessary prerequisite for ophthalmologically defined CBS and that ophthalmologists should be aware that patients with preserved acuity but significant deafferenting ocular disease are at risk of the syndrome.


    Author’s reply

    I thank Madill and Ffytche for their interesting comments on our paper1 and am grateful for the opportunity to respond to some of the issues raised. The first relates to possible variations in the prevalence rate of Charles Bonnet syndrome (CBS) in different populations, on which there is currently very little substantive data. We agree that additional population based studies are necessary to determine whether the differences found between ethnic groups are consistent. We wish to highlight that our findings1 are supported by another recently published study of 1000 Japanese patients by Shiraishi et al2 who reported a prevalence rate of 0.5% compared to 0.4% in our study population. Since the methodology and diagnostic criteria used in that study are very similar to ours, the comparable prevalence rates reported in both studies may be a true reflection of a lower prevalence of CBS in Asians compared to Western populations.

    We would like to point out that the large variation in prevalence rates of CBS in the various cross sectional studies may be explained by the characteristics of the different populations being screened, especially the degree of visual impairment. When patients were drawn from either a general ophthalmology1–3 or general medical4 clinic, presumably comprising patients with relatively good visual acuity, the prevalence rates ranged from 0.4–2%. In patients with poorer visual acuity3,5–8 the prevalence rates varied from 11–12.3%. This observation suggests that poor visual acuity may be a factor that increases the risk of developing CBS. Indeed, Teunisse et al showed that the risk of CBS was higher in patients with a logMAR visual acuity of worse than 0.33 and Holroyd et al also found an association with bilateral visual acuity of 20/60 or less.9

    Nevertheless, even if reduced visual acuity is indeed a risk factor for the development of CBS, we agree that a reduction in visual acuity is not essential and that CBS can occur in patients with relatively good visual acuity. Gold et al thought that reduced visual acuity is not a prerequisite for the diagnosis of CBS10 and Holroyd et al suggested that it may be the presence of visual impairment and not the severity of visual loss that predisposes to CBS.4 Schultz and Melzack, in an excellent review of the cases reported in the literature, found that the visual acuities of patients diagnosed with CBS ranged from normal to no light perception.11 It is becoming apparent that CBS can develop in patients with visual field defects but otherwise good central acuity. The patients described by Madill and Ffytche and our own patient (patient 3) had glaucomatous visual field defects. A recent paper by Freiman et al12 documented the presence of hallucinations manifesting within visual field defects that occurred following neurosurgery. Two interesting points can be noted from Freiman et al’s series: firstly, the hallucinations were confined to the area of visual loss; secondly, they began soon after the surgery and resolved within days to 6 months after their onset.4 In contrast, in three of four patients reported by Madill and Ffytche as well as our own patient, the hallucinations were not restricted to the visual field defect. The relation of the onset of CBS symptoms to the development of visual field defects as well as the localisation of hallucinations within visual fields are issues that warrant further study as they may provide additional insight into the pathophysiology of CBS.


    View Abstract

    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.