Strict application of diagnostic criteria and diagnosis reassessment during follow-up have reduced the clinicopathological dissociation in the field of movement disorders. After several clinicopathological studies, dysautonomia, REM behaviour disorder (RBD) and visual hallucinations (VH) have been proposed as clinical markers of underlying synucleinopathy.1^–3 We report a patient who presented with a synucleinopathy but had a pathological diagnosis of progressive supranuclear palsy (PSP).

This woman was first seen in 1986, age 46 years, after a 5 year history of right-hand intention tremor and slowness of movements, evolving from fine manoeuvres (ie, buttoning clothes) to turning in bed and entering or leaving the car. She was hypomimic and hypophonic. Ocular motility was normal. Intention tremor, rigidity and moderate bradykinesia were present in her right arm. Her posture was slightly stooped and her gait short-stepped, with no freezing, and unaided recovery at the pull test. Her cognition was normal except for mild attention deficit. A brain CT showed global atrophy.

Parkinson’s disease (PD) was suspected, and she was started on levodopa with subsequent improvement of 34% in motor score. During the ensuing 5 years she remained stable but complained of frequent unpleasant nightmares and her husband reported that she had violent dream enacting behaviours. In the sixth year of follow-up she had an isolated transient episode of VH after initiation of selegiline that resolved after drug withdrawal. The following year she developed slight wearing off without dyskinesias. Levodopa was increased to 600 mg/day and pergolide was added without further VH, but her motor state remained unchanged.

In 1994, she progressively developed urinary and faecal incontinence, symptomatic orthostatic hypotension, marked antecollis, jerky tremor and motor worsening despite increasing levodopa doses. Multiple system atrophy (MSA) was considered. Laryngeal stridor was not detected on either laryngoscopy or video polysomnography, the latter disclosing REM sleep without muscle atonia (fig 1A). The EMG study demonstrated anal sphincter denervation. A brain MRI (1996) showed no focal lesions or atrophies. Between 1998 and 2001, she developed complex and frightening VH (her dead mother or non-relative persons, mainly at night), severe memory and executive deficits (Mini-Mental State Examination 20/30), worsening of instability with frequent falls and progression of dysarthria and dysphagia. The combination of a poor levodopa responsive parkinsonism, dysautonomia, RBD, VH and cognitive decline led us to consider a Lewy body disorder, such as PD with dementia.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Polysomnographic recording during 1 min of REM sleep showing a lack of the normal atonia in the chin. EOG, electro-oculogram; C3, C4, O1, O2, electroencephalographic channels from central and occipital areas; Chin, EMG recording from the chin; L and R Bic, EMG from left and right biceps; L and R Tib, EMG from left and right anterior tibialis; Nasal and Oral, airflow from nose and mouth; Thor and Abd, thoracic and abdominal respiratory movement. Courtesy of Dr J Santamaria and Dr A Iranzo (Sleep Unit, Hospital Clínic de Barcelona). (B–D) Findings of immunohistochemistry for tau, using AT8 antibodies, in some regions proposed to be involved in REM behaviour disorder, dysautonomia and visual hallucinations (200×). (B) Neurofibrillary tangles and pre-tangles (arrows), along with coiled bodies, in the locus coeruleus. (C) Coiled bodies (arrows), tufted astrocytes and neurofibrillary tangles in the intermediate reticular zone of the medulla oblongata. (D) Tufted astrocytes (arrows) with neurofibrillary tangles and pre-tangles in the amygdala of the hippocampus.

At the latest follow-up she was wheelchair bound and had meaningless vocalisations, severe antecollis and bilateral blepharospasm. Supranuclear gaze limitation was never seen. After recurrent infections, she died on November 2006, 25 years after the onset of symptoms.

On macroscopic examination, severe global atrophy and depigmentation of the substantia nigra and the locus coeruleus were observed. Microscopically there was complete loss of pigmented neurons in both nuclei, with scarce free neuromelanic pigment within the neuropil. Ubiquitin, α-synuclein and β-amyloid immunostaining was negative whereas widespread brainstem phospho-τ aggregates were identified, with abundant neurofibrillary tangles, neuropil threads, coiled bodies and tufted astrocytes. Brainstem nuclei presumably involved in RBD (locus coeruleus, raphe nuclei, pedunculo-pontine nucleus) and those with autonomic function (intermediate reticular zone, nucleus raphe obscurus, dorsal nucleus of the vagus), showed severe neuronal loss and tau-pathology, mainly glial inclusions (fig 1B, C). The pre-central gyrus, hippocampus and amygdala (the latter shown in fig 1D) were severely involved. Tau isoform immunohistochemistry was markedly positive for 4R tau isoform in contrast to 3R. The PSP-tau score4 was nine out of 12. Mutations in tau and LRRK2 (G2019S and 1441) genes were ruled out.

DISCUSSION

This case afforded a combination of proposed clinical markers of synucleinopathy (dysautonomia, RBD and VH), first fulfilling the diagnostic criteria of probable PD and, later, of probable MSA. Severe dysautonomia accompanying parkinsonism most often indicates MSA, dementia with Lewy bodies and advanced PD and, when present in PSP, is either subclinical or slight.1 Our patient suffered from severe dysautonomia (anal sphincter denervation) that might be explained by severe autonomic brainstem nuclei degeneration. RBD and REM without atonia have also been found in patients with clinically probable PSP in a recent clinical-polysomnographic study5 but our patient represents, to the best of our knowledge, the first detailed case report of pathologically confirmed PSP with RBD, showing severe involvement of brainstem nuclei that have been related to RBD generation.2 As for VH, a retrospective brain bank study found a frequency of VH of 50% in PD compared with 7% in PSP.3 Our patient, apart from an early episode of VH attributed to medications, had well formed VH in the advanced stage of the disease. Even though the possible anatomical correlate of VH in parkinsonism is uncertain, a clinicopathological correlation study found that well formed VH in Lewy body disorders correlated with high Lewy body densities in the amygdale, parahippocampus and inferior temporal cortex.6 In our case, amygdala and the hippocampus were among the most involved cortical areas.

Clinical clues to the pathological diagnosis were easy to overlook. Falls appeared and were frequent only in the late stage. Neck rigidity of the antecollis type is expected more in MSA than in PSP. Blepharospasm, perhaps more frequent in PSP, appeared also in advanced disease. Vocalisations have not been studied in depth in parkinsonism but in our experience seem to be more common in PSP. Supranuclear gaze palsy, the clinical hallmark of PSP, was never noted while the patient was still collaborative.

Following the combination of initially asymmetric parkinsonism with mild response to levodopa, dysautonomia, RBD and VH, as well as long disease duration, this case was thought to be PSP-parkinsonism (PSPp).4 The tau isoform profile and high PSP-tau score were, nevertheless, closer to Richardson’s syndrome than to PSPp, as an example of the still uncertain molecular basis underlying the clinical phenotypes of PSP.4

Even though each pathological entity tends to involve susceptible areas, hence providing helpful clinical clues for diagnosis, the unexpected clinicopathological association here shows how difficult the clinical diagnosis of parkinsonism can be and lends support to the notion that the downstream topography of the lesions rather than the type of protein aggregate could be the basis of the clinical picture.