Confabulation or hallucination? A case of bilateral occipital infarct blurring the boundaries between Charles Bonnet and Anton’s syndromes 

Abstract

We present the case of a 53 year old man with severe visual impairment after an acquired brain injury who experienced erroneous visual perceptions. These experiences had features of both Charles Bonnet and Anton’s syndromes, two conditions characterised by abnormal visual experiences but traditionally differentiated by the presence or absence of insight. We believe this case provides a window into the processes underpinning perception generation and suggests the boundaries between these two presentations may be looser than is generally thought.  

Key points

1. Charles Bonnet and Anton’s syndromes both arise in the context of visual impairment. In the former visual cortex is preserved (generation of perceptions being understood as release phenomena), whereas the latter arises in the context of bilateral occipital damage (visual experiences being generally understood as confabulations)
2. Traditionally insight has been regarded as preserved in CBS and lost in Anton’s syndrome, patients with the latter condition not acknowledging that they are blind and claiming their perceptions represent reality. These boundaries do not seem absolute however, it being generally acknowledged that transient impairment of insight can be encountered in CBS.
3. This case report suggests that it may be possible for elements of these two syndromes to coexist. Indeed, CBS-related phenomena may further perturb a process of perception generation driven by deranged predictive processing poorly constrained by degraded visual input.

Introduction

Visual hallucinations in visual impairment are generally described as Charles Bonnet syndrome (CBS) and can be simple or complex. The former are commonly basic shapes and grid like patterns, the latter complicated images potentially including disembodied faces and costumed figures [1]. It is hypothesised CBS hallucinations arise from de-afferentation, loss of visual input leading to changes in cortical inhibition that result in spontaneous activity [2]. CBS is typically associated with intact cognition and preserved insight, with insight in this context referring to the fact the visual experiences are recognised as not representing genuine stimuli. Another eponym, Anton’s syndrome, describes a rarer group reporting vivid visual experiences in cortical blindness. These patients reportedly lack insight into their visual impairment, denying or minimising blindness, and (presumably relating to anosognosia for visual loss) seemingly regard their experiences as real. Whether their visual experiences are confabulations or true hallucinations is unresolved. A core distinguishing feature from CBS is loss of insight, (to blindness and the fact perceptions are false), though intriguingly descriptions emphasise the absence of significant cognitive impairment [3]. The two conditions are compared in Table 1 . We present a case straddling these descriptions, postulating it reveals processes central to normal visual perception as well as those underpinning the erroneous visual perceptions occasionally encountered in visual impairment. 

Case presentation  

A 53 year old man was found with a deep neck laceration, GCS 4 and hypothermic.  Diary notes confirmed suicidal intent.  CT showed left vertebral artery dissection, presumed caused by neck hyperextension, and posterior circulation ischaemia.  Subsequent CT showed ischaemic infarcts involving the entire right and partial left posterior cerebral artery territory (i.e. bilateral occipital and right posterior temporal lobes), entire left superior cerebellar artery territory, and right thalamus.  It also showed established gliotic change in right anterior temporal and both inferior frontal lobes. MRI two years post-injury further delineated pathology (Figure 1). 

Following ITU care cortical blindness, left facial droop and left hemiparesis were noted.  Seven weeks post-injury he was transferred to medical rehabilitation, but verbal hostility secondary to visual hallucinations (e.g. of people recording him on mobile phones) necessitated transfer to a neurobehavioral unit.  On admission, aside from CN2, cranial nerves were grossly intact.  Modestly reduced power and hyper-reflexia on the left; also loss of sensation to light touch, vibration and proprioception.  Unable to touch his nose or ear with his left hand unless guided by his right. He could stand unaided and, with physiotherapy guidance, walk using a tripod walking stick.  When walking unawareness of foot contacting ground necessitated ‘stomping’ for compensatory auditory input.

Background revealed a RTA-related TBI in 1999 with associated personality change and post-traumatic epilepsy; this was associated with the established gliotic change detailed above.  In 2008 he lost his job and subsequently did not sustain employment.  His marriage had ended five years prior to suicide attempt and he had only intermittent contact with his children. There had been sporadic contact with psychiatry, generally when there were interpersonal or employment difficulties, and at least three previous suicide attempts. 

Profound visual impairment limited a cognitive assessment, but five months post-injury deficits in verbal memory and some ‘executive’ spheres were evident (see Table 2).  There was limited preservation of vision on the extreme right of his visual field.  Ophthalmology assessment suggested perception at level of hand movements, but even early post-admission he was generally aware of large objects (i.e. though not necessarily able to name it, was aware of a large object such as a piece of furniture was in front him).  These stimuli were seemingly incorporated into perceptions. For example, in the art room he was aware of a worktop-like object, and reported he was in a kitchen and a kettle and toaster were on the worktop.  In reality there was a kitchen-unit style sideboard on which art items (such as pots of pens) were stored. When asked what I was wearing he could clearly locate me in space, but reported a suit, tie and glasses.  I was wearing bright hospital scrubs and no glasses; presumably he ‘expected’ a doctor to wear a suit and glasses.  In addition to these perceptions, seemingly influenced by a combination of genuinely present stimuli and expectations, he continued to report perceptions without obvious environmental contributions. On occasion, for example, he reported seeing piles of building materials (predominantly bricks) on the ward.  When his perceptual errors were corrected he tended not to rigidly adhere to his perception being correct, rather he would rationalise how it could have occurred (generally with the implication an erroneous perception could occur in the context of preserved visual ability).  For example, at times he reported it was too dark to see properly (though in reality the room was brightly lit).  Interestingly he was somewhat firmer on his perception of building materials being correct.

Within a few months it transpired there was (or had developed) more substantial preservation of visual acuity in the periphery of his right visual field.  He was able to identify large letters (around 20cm in height) held at 1-2 metres in preserved visual field.  Initially he required prompting to utilise preserved vision, though over time increasingly did it spontaneously.  Experience of false perceptions diminished, though he continued to have very limited appreciation of the functional implications of impairments.    

Our patient had many preserved cognitive abilities, these possibly best exemplified by a poem he composed (and dicatated) about the unit (see Figure 2). Rehabilitation outcomes were however limited, predominantly by poor appreciation of the severity and implications of his visual impairment.  Experience of formed visual perceptions potentially contributed to his belief that his vision was much better than clinicians reported, but his understanding of the implications of visual impairment remained limited even when these experiences diminished.  In terms of physical constraints, his need to use a tripod for balance precluded using a cane to compensate for visual impairment.  Consequently he needed staff support to walk, generally using a wheelchair to mobilise.  In this he readily orientated himself to the ward and was soon independently going between the bedroom and dining area.  The dissociation between gross preservation of executive abilities (see Table 2) and limited appreciation of the implications of visual impairment was stark.  He regularly stated his intention to independently undertake tasks indisputably requiring support.  Examples included travelling to London independently by train and driving.  This lack of appreciation of support needs ultimately necessitated a Welfare Guardianship Order for ongoing care. 

Discussion 

The traditional belief that perception is determined simply by sensory input is now accepted as illusory. Rather, our visual model of the world emerges from an active process of integration of external (bottom-up) information with pre-existing knowledge (top-down influences) [4]. Top-down biases modulate and guide visual processing, our brain proactively testing multiple hypotheses to identify an object. These biases include the potential emotional salience of the object, associations derived from context, and memory of previously encountered events.  These factors, albeit triggered and constrained by incoming bottom-up signals, will ultimately determine the percept experienced.  The mechanism underpinning this process is likely a repeated comparison of ascending and descending information, the latter being the brain’s predictions of what is being perceived, these predictions repeatedly refined on the basis of sensory input.  The convergence of these processes ultimately, generally at least, achieves successful object recognition.

A core concept in the above model is context-based predictions having substantial influence over perceptual experience.  In a normal functioning system these predictions are extremely useful, saving processing resources and reducing mental effort when dealing with predictable aspects of our surroundings. They allow us to allocate attention towards relevant environmental aspects more efficiently, this usefully guiding our actions and behaviour.  In our patient we see this process going awry.  Impoverished residual visual sensory input means that predictions are generated, but the process of modification by additional sensory input (unless sought from non-visual sources – for example by touching objects, which our patient tended not to do) is impoverished.  An object is identified as likely being a worktop, contextual association suggests a kitchen, and objects on it (pots of pens) are then perceived as a toaster and a kettle. In this dysfunctional model impairment of higher level appraisal of input means contextually unlikely percepts (e.g. building materials in a hospital ward), are accepted as present.  There is then an associated impairment, despite repeated discussions with staff, in appreciating that perceptual errors are being made and the associated functional consequences of visual impairment.  As discussed, visual impairment was not his only disability; our patient also had reduced power and sensory ataxia on the left side.  Though early in his rehabilitation he had a number of falls as he attempted to mobilise independently, he ultimately seemed more accepting of the implications of motor rather than visual impairments. 

What about diagnosis?

In CBS, precepts are dissociated from sensory input, understood as arising from spontaneous cortical activity. These precepts are experienced in external space, seeming identical to those associated with normal seeing.  They are however generally recognised as false, either because of greater detail than real stimuli (they occur in visual impairment) or because of contextually inappropriate content [5]. Two proposed diagnostic criteria for CBS do however recognise that insight into the unreality of percepts may be partial [5,6]. Anton’s syndrome is rarer and less studied than CBS.  It is distinguished from CBS, primarily, by lack of insight into blindness.  Kletenik et al undertook a comprehensive synthesis of studies of Anton’s syndrome, identifying 24 cases where brain injury was associated with: (1) complete visual loss, (2) lack of awareness of vision loss, and (3) a brain image for lesion mapping [7]. These were compared to cases with anosognosia for hemiplegia and patients with visual impairment or hemiplegia with preserved awareness.  Their findings suggested a cross-modal anosognosia network converging on the hippocampus and precuneus.  Lesions more specifically associated with visual anosognosia however were reported as connected to V2, posterior cingulate, retrosplenial cortex, inferior parietal lobe, medial frontal pole, corpus callosum, and lateral occipital lobe/angular gyrus.  Regions associated with both these cross-modal and visually specific networks were damaged in our patient. 

A prerequisite for inclusion in Kletenik et al’s review was ‘lack of awareness of visual loss’; it wasn’t specified if this could be partial or must be absolute.  To explore this further we reviewed the included papers. 18/20 papers could be obtained, these discussing 22 patients. Though 19 patients ‘denied visual loss’, in 3 at initial assessment there is limited acceptance vision is imperfect (e.g. it is ‘blurred’). Most reports do not detail follow up assessments, but 4 cases report insight into visual impairment improved over time. In 2 this was associated with partial recovery of vision, though neither report discusses whether confabulation of visual experiences continued as insight improved.  None of these reports suggested that acceptance of blindness and confabulation are co-occuring, albeit it does seem some awareness of imperfect vision is not incompatible with visual confabulations.  As our patient experienced some visual recovery, acknowledgement of visual impairment improved (though he minimised implications for independent living), and confabulation of visual perceptions diminished.  In stark contrast to Anton’s syndrome, retained insight is generally regarded as core to CBS.  Hamedani and Pelak note however that ‘some patients with what otherwise appears to be CBS lack complete insight into the nature of the visual hallucinations’ [8]. In their systematic review of CBS studies, 18 specified that presence of insight was necessary for a CBS diagnosis. In 12 of these however ‘partial insight’ was regarded adequate. Indeed, a recent consensus guideline on visual hallucinations is explicit in stating that insight is variable and may be absent at the time when hallucinations occur (though restored in retrospect) [9]. It therefore seems in practice that clinicians are not so binary as to regard full insight as essential for CBS and complete absence of insight as a prerequisite for Anton’s syndrome. 

In Anton’s syndrome visual experiences are often described as confabulations rather than, as in CBS, hallucinations. Our patient however, seemed to have elements of both syndromes. The dominant process underpinning his experiences seems to be a derangement of the balance between top down and bottom up processing, with excessive weight placed on predictions of perceptions.  One could, presumably, refer to these as ‘confabulated perceptions’.  In contrast to how this is generally conceptualised in Anton’s syndrome however, residual visual function meant perception was not entirely unrelated to available visual stimuli. This influenced perception generation, meaning perceptions often had some relationship to genuinely present stimuli.  It is interesting to speculate if spontaneous cortical activity (historically conceptualised as ‘release phenomena’) fed into this process.  This possibility is perhaps best exemplified by his perception of building materials, this echoing the geometric patterns commonly reported in CBS.  One fMRI study associated geometric hallucinations with increased activity in the collateral sulcus [10], a region which (on the left side at least) could conceivably be preserved in our patient but deprived of input. Potentially these release phenomena could then be incorporated into the process of perception generation, leading to perception of a building materials. 

Overall we felt that this case was best described as Anton’s syndrome, presentation modified by some residual visual ability, a limited degree of insight and, potentially, elements of CBS.  As suggested by the review above, it would be expected that return of some vision (and/or appreciation of visual impairment) would preclude visual confabulation (top down predictions become constrained by visual input, with the loss of spontaneous perceptions helping foster insight).  In our patient however the pre-existing fronto-temporal brain damage may impede this process, possibly hampering efficient reality testing.   It seems the unusual nature of our patient’s brain injury/injuries reveals some of the processes underpinning visual perception which, of course, in the uninjured brain we are generally oblivious to. 

We utilised a conceptualisation of perception generation being driven by impaired correction of erroneous top-down predictions to guide treatment. By altering perceptual expectations we hypothesised we could increase the likelihood of accurate perception generation.  Consequently we routinely described the immediate environment to our patient, hoping to increase accuracy of top-down predictions. It seemed this, together with some visual recovery and greater utilisation of the preserved peripheral visual field, did indeed reduce the likelihood of erroneous perceptual experiences. Unfortunately however poor appreciation of the implications of visual impairment limited potential for independent living, necessitating ongoing residential care. 

Figure 2: Poem our patient composed about the rehabilitation unit. 

Poem for the 30^th Anniversary, Rehabilitation For YoU 

Robby F laid all the tiers to keep us safe for 30 years, 

His life cut short at 24, his head smashed on cold stone floor. 

But he shared his plans and sowed the seeds, 

Who could take over and deliver our needs. 

Let us not shed a tear, Dr Andrew Duncan brought us here. 

A one stop shop for neuro needs, 

Residential care, it took the lead, 

Appointments, diaries were not a need, all on site including feed. 

Your own bedroom, bathroom, it’s your private den, 

A rehab base to start again. 

Doctors and nurses to coach us till well, 

and four different therapies would make us excel. 

Let us make a start with creating things through art, 

Physios help us walk, 

Speech and Language perfect our talk, 

OTs remind us how to live, 

So as long as you can take and

We could be the greatest, we should be the best, 

We could be the neuro destination of the West. 

It`s 2022, and we`ve Scotland name for brains, 

A refurbishment is needed, not building more with cranes. 

What do we need to be the best, 

To be the neuro destination of the west. 

We need comms, we need links, we need faces to tell, 

A total refurb, includes our doorbell. 

Our fitments are old and rotten, 

And we are feeling a bit forgotten, 

We are ready awaiting a cash injection, 

Please make cash spent a true reflection.

References

1. Santhouse AM, Howard RJ, ffytche DH. Visual hallucinatory syndromes and the anatomy of the visual brain. Brain 2000; 123: 2055-64. https://doi.org/10.1093/brain/123.10.2055
2. Painter, D.R., Dwyer, M.F., Kamke, M.R., and Mattingley, J.B. Stimulus-driven cortical hyperexcitability in individuals with Charles Bonnet hallucinations. Curr. Biol. 2018; 28: 3475-3480.https://doi.org/10.1016/j.cub.2018.08.058
3. Chen, J. J., Chang, H. F., Hsu, Y. C., & Chen, D. L. Anton-B abinski syndrome in an old patient: a case report and literature review. Psychogeriatrics 2015; 15: 58-61. https://doi.org/10.1111/psyg.12064
4. Gilbert, C. D. & Li, W. Top-down influences on visual processing. Nat. Rev. Neurosci. 2013; 350-363. https://doi.org/10.1038/nrn3476
5. Teunisse RJ, Cruysberg JR, Hoefnagels WH, Verbeek AL, Zitman FG. Visual hallucinations in psychologically normal people: Charles Bonnet’s syndrome. Lancet 1996; 347: 794-797. https://doi.org/10.1016/S0140-6736(96)90869-7
6. Gold K, Rabins PV. Isolated visual hallucinations and the Charles Bonnet Syndrome: A review of the literature and presentation of six cases. Compr Psychiatry 1989;30:90-98 https://doi.org/10.1016/0010-440X(89)90122-3
7. Kletenik I, Gaudet K, Prasad S, Cohen AL, Fox MD . Network localization of awareness in visual and motor anosognosia. Ann Neurol. 2023; 94: 434-441 https://doi.org/10.1002/ana.26709
8. Hamedani AG, Pelak VS. The Charles Bonnet syndrome: a systematic review of diagnostic criteria. Current Treatment Options in Neurology. 2019; 21: 1-13. https://doi.org/10.1007/s11940-019-0582-1
9. O’Brien J, Taylor JP, Ballard C, Barker RA, Bradley C, Burns A, Collerton D, Dave S, Dudley R, Francis P, Gibbons A, Harris K, Lawrence V, Leroi I, McKeith I, Michaelides M, Naik C, O’Callaghan C, Olsen K, Onofrj M, Pinto R, Russell G, Swann P, Thomas A, Urwyler P, Weil RS, Ffytche D. Visual hallucinations in neurological and ophthalmological disease: pathophysiology and management. J Neurol Neurosurg Psychiatry. 2020;91(5):512‐519 https://doi.org/10.1136/jnnp-2019-322702
10. Ffytche DH, Howard RJ, Brammer MJ, David A, Woodruff P, Williams S. The anatomy of conscious vision: an fMRI study of visual hallucinations. Nature Neurosci. 1998; 1: 738-742. https://doi.org/10.1038/3738