A 59-year-old woman, recently retired following multiple car accidents and cognitive difficulties, underwent emergent intervention for myocardial infarction, with no hypoxia, arrhythmia or cerebral hypoperfusion. Progressively disoriented, confused and agitated the next day, she experienced visual hallucinations (VH) and “unpleasant” auditory hallucinations, accused staff of inappropriate physical contact, attacked a relative and telephonically lodged a police complaint alleging sexual assault. Agitation and confusion recurred overnight over the next few days, improving after intramuscular haloperidol. Cross-sectionally irritable, disoriented to time and place with rigidity of thought, she had poor attention and right–left confusion, appeared distrustful, with inconsistent visual fields and inability to describe a Snellen chart or colours, but no parkinsonism.
Delirium or a reactive psychosis possibly following acute medical illness was considered; common precipitants of delirium were excluded by appropriate testing, and some improvement was noted with a brief course of risperidone 0.5 mg twice daily, although her husband noted prominent constipation, persistent VH and fluctuating confusion. She was re-admitted a month later with suicidal ideation, anhedonia, low mood, increasing forgetfulness and persistent VH with difficulty identifying faces, constipation and fluctuating confusion. She neglected one side and had jerky pursuits and slowed saccades with obvious difficulty with visual fixation while conversing. Left visual neglect, Bálint syndrome (oculomotor apraxia, optic ataxia and simultanagnosia), bilateral limb apraxia and alexia were prominent in the face of preserved visual acuity. She scored 45/100 on the Addenbrooke’s cognitive examination, performing poorly on fluency and visuospatial function. Neuropsychological testing demonstrated significant visual-spatial impairments, including space and object perception, integration and interpretation, confirming simultanagnosia, construction apraxia and optic ataxia. The cognitive profile was consistent with major neurocognitive disorder, likely posterior cortical atrophy (PCA).
Relative posterior hemispheric cortical atrophy without focal parenchymal lesions was seen on brain MRI performed for evaluation of the initial delirium, and a subsequent fluorodeoxyglucose (FDG) positron emission tomography (PET) scan demonstrated severe bilateral parieto-occipital hypometabolism sparing the posterior cingulate gyrus (Figure 1A-E).
(A) T1-weighted and PET coronal images demonstrating relative lack of atrophy or hypometabolism in the mesial temporal regions. (B) T1-weighted sagittal and FLAIR axial images showing relative posterior hemispheric cortical atrophy (arrows). (C) Axial PET image. (D) Left parasagittal PET image. (E) Right parasagittal PET image showing severe bilateral parietal and occipital hypometabolism (arrows), with sparing of the posterior cingulate gyrus (cingulate island sign, asterisk). The colour sidebar shows increasing degrees of deviation from normal as the signal trends lower (red is normal, while black is maximally abnormal).

Longitudinal review revealed subsequent anxiety; systematised paranoid delusions; sleep disturbances with sundowning, nocturnal agitation, confusional arousals and dream enactment consistent with rapid eye movement (REM) sleep behaviour disorder (RBD); dressing apraxia; and worsening prosopagnosia, fulfilling criteria for a diagnosis of probable dementia with Lewy bodies (DLB). Reference McKeith, Boeve and Dickson1
PCA is a neurodegenerative syndrome characterised by progressive impairment of higher-order visual functions and atrophy involving parietal, occipital and occipitotemporal cortices. The term PCA is used inconsistently, both as a diagnostic label (synonymous with the “visual variant” of Alzheimer’s disease [AD] – referred to as PCA-AD) and – more correctly – as a descriptive syndromic term. Multiple pathologies other than AD underlie the clinico-radiological PCA syndrome, including DLB (called PCA-DLB), corticobasal degeneration and prionoses. Reference Crutch, Schott and Rabinovici2,Reference Crutch, Lehmann, Schott, Rabinovici, Rossor and Fox3 The proposed entity of “PCA-plus” includes individuals (such as our patient) with PCA who also fulfil core criteria for another neurodegenerative syndrome. Reference Crutch, Schott and Rabinovici2
DLB is characterised by four core features: dementia with fluctuating cognition, parkinsonism, recurrent complex VH and RBD. Reference Crutch, Lehmann, Schott, Rabinovici, Rossor and Fox3 Prodromal DLB may present as mild cognitive impairment, particularly affecting visuospatial and executive function, or as delirium-onset or psychiatric-onset DLB. Reference Wyman-Chick, Chaudhury and Bayram4,Reference McKeith, Ferman and Thomas5 Early cognitive-psychiatric symptoms are typical, with episodes of confusion or drowsiness separated by lucid intervals. Reference McKeith, Boeve and Dickson1 Whether delirium-onset DLB represents a greater vulnerability to delirium in this condition, or severe fluctuations with clouding of consciousness are misinterpreted as delirium, is unclear. Reference Wyman-Chick, Chaudhury and Bayram4,Reference McKeith, Ferman and Thomas5 Late-onset primary psychosis, frequently associated with dementia or delirium, can be the initial symptom of prodromal DLB. Reference McKeith, Ferman and Thomas5
Apathy, anxiety, delusions, depression, VH and other hallucinations distinguish the cognitive–behavioural profile of DLB from AD. Reference Wyman-Chick, Chaudhury and Bayram4 While PCA usually resembles AD in its neuropsychiatric profile, VH occur in 10%–25%, usually >4 years after onset, and are often associated with the Bálint syndrome and occipital lobe atrophy. Reference Crutch, Lehmann, Schott, Rabinovici, Rossor and Fox3 Underlying DLB should be considered in any PCA patient with RBD and/or VH – especially if well-formed, non-threatening, of silent people or animals or in association with auditory hallucinations. Reference Crutch, Lehmann, Schott, Rabinovici, Rossor and Fox3,Reference McMonagle, Deering, Berliner and Kertesz6 Anxiety, in combination with one core clinical criterion, provides a specific discriminant to distinguish DLB from AD, and several reports describe an initial psychiatric presentation of DLB, including depression, psychosis, anxiety and catatonia with a lag to the development of cognitive decline. Reference Wyman-Chick, Chaudhury and Bayram4,Reference McKeith, Ferman and Thomas5 Executive, attentional and visuospatial deficits, relative sparing of episodic memory, psychiatric symptoms and neuroleptic sensitivity characterise early DLB. Reference Park, Kim, Cheon, Cha, Kim and Kim7 This is explained by the differential involvement of brain regions in the two diseases: neuropathological changes in the medial temporal cortex and neocortical association areas in AD and frontal and parietal cortices and basal ganglia in DLB. Reference Park, Kim, Cheon, Cha, Kim and Kim7
Neuroimaging shows promise in differentiating DLB from PCA-AD. Occipital hypometabolism on 18F-FDG-PET is a feature of PCA, but it is relatively more symmetric and extensive in DLB. This includes the mesial occipital lobes, orbitofrontal and anterior temporal cortices and caudate nucleus, indicating dysfunction of attentional and executive networks. Reference McKeith, Ferman and Thomas5,Reference Whitwell, Graff-Radford and Singh8 Relative sparing of posterior cingulate metabolism compared with the precuneus and cuneus (i.e. cingulate island sign, Figure 1) has 90% specificity in distinguishing DLB from AD, even in the presence of concomitant AD pathology, and is a supporting criterion for diagnosis. Reference Wyman-Chick, Chaudhury and Bayram4,Reference Whitwell, Graff-Radford and Singh8 Mesial temporal atrophy is a feature of AD, and relative mesial temporal sparing with posterior and/or frontal cortical atrophy and insular thinning and grey matter volume loss suggest DLB (Figure 1). Combining MRI and FDG-PET findings improves diagnostic accuracy. Reference McKeith, Ferman and Thomas5
Diagnostic criteria for “probable” DLB highlight an aggressive cognitive decline with at least two of the four core features, or with one core feature and an indicative marker (polysomnographic confirmation of REM sleep without atonia; slowing, variability and loss of reactivity of the posterior dominant frequency on electroencephalography; abnormal metaiodobenzylguanidine myocardial scintigraphy; or reduced uptake in the basal ganglia on dopamine transporter single photon emission computed tomogrpahy). Reference McKeith, Boeve and Dickson1,Reference Wyman-Chick, Chaudhury and Bayram4,Reference McKeith, Ferman and Thomas5 As there are no biomarkers for DLB currently available, an in vivo diagnosis of “probable” PCA due to DLB can be made if the patient with PCA fulfils clinical criteria for DLB, while a “definite” diagnosis needs pathological confirmation. Reference Crutch, Schott and Rabinovici2
We present a middle-aged woman who presented with agitated delirium and psychotic features during an acute medical illness. Premorbid cognitive and visual-spatial difficulties, detailed neurocognitive examination and imaging confirmed a PCA syndrome, and she subsequently fulfilled diagnostic criteria for DLB. The PCA syndrome may rarely present with delirium as the initial manifestation of underlying DLB. Delirium-onset DLB may be associated with fluctuating cognition and/or VH, warranting a careful search for other DLB features with a low threshold for undertaking the supportive investigations described above. Reference Crutch, Lehmann, Schott, Rabinovici, Rossor and Fox3,Reference Wyman-Chick, Chaudhury and Bayram4 The risk of DLB in these patients – and the concomitant neuroleptic sensitivity – should be integrated into decisions regarding pharmacologic treatment of delirium. Reference Wyman-Chick, Chaudhury and Bayram4,Reference McKeith, Ferman and Thomas5
Data availability statement
The data in this paper have not been previously presented orally or by poster at scientific meetings.
Author contributions
RJ: Data collection, literature review, writing the original manuscript, revision, approval of the final draft. JP: Data collection, critical review of the manuscript, approval of the final draft. AG: Critical review of the original manuscript, revision, approval of the final draft. AS: Data collection, literature review, critical review of the manuscript, approval of the final draft.
Funding statement
No funding was received from any source.
Competing interests
None declared.
Ethical standards
The patient provided verbal informed consent for anonymized patient information to be published in this article. Ethics approval was not required by the Tasmanian Health Service for reporting individual cases or case series.