Progressive Supranuclear Palsy

Progressive supranuclear palsy (PSP) is the most common atypical parkinsonian disorder, with an estimated prevalence of 5–7 per 100,000. It is a relentlessly progressive 4-repeat tauopathy characterized by supranuclear vertical gaze palsy, early postural instability with backward falls, levodopa-resistant parkinsonism, and frontal-subcortical cognitive decline. The classic presentation — PSP-Richardson syndrome (PSP-RS) — is readily recognizable, but the 2017 MDS diagnostic criteria now define eight phenotypic variants, many of which initially mimic PD, FTD, or primary progressive aphasia, delaying diagnosis by years. Mean symptom onset is 63 years, with no documented cases under age 40 — early onset should prompt consideration of alternative diagnoses. The MAPT H1 haplotype (particularly H1c) increases PSP risk at least fivefold. Despite intensive effort, all disease-modification trials have failed — including riluzole, tideglusib, davunetide, gosuranemab, and tilavonemab — but the lessons from these failures are shaping a new generation of approaches targeting tau production, aggregation, and neuroinflammation.

Epidemiology, Genetics, and Pathology

Prevalence estimates range from 5–7 per 100,000, though MSA is sometimes cited as the "most common" atypical parkinsonism depending on the cohort — both are far rarer than PD. The sex distribution is approximately equal. No consistent environmental risk factors have been identified, though the ENGENE-PSP study linked lower academic achievement, exposure to well water and industrial waste, and firearm use to higher risk. Clusters have been reported in northern France (near industrial heavy metal exposure) and Guadeloupe (associated with annonacin in tropical pawpaw fruit).

The strongest genetic association is the MAPT H1 haplotype, particularly the H1c subhaplotype, which increases PSP risk at least fivefold. Genome-wide association studies have additionally identified STX6 (syntaxin 6), EIF2AK3/PERK (endoplasmic reticulum unfolded protein response), MOBP (myelin-associated oligodendrocytic basic protein), SLCO1A2 (ion transport), and DUSP10 (tau trafficking). Pathogenic MAPT gene mutations are rare and more commonly associated with frontotemporal lobar degeneration with parkinsonism (FTLD-tau).

Neuropathologically, PSP is defined by abnormal aggregation of 4-repeat (4R) tau in characteristic lesions: tufted astrocytes (pathognomonic), globose neurofibrillary tangles, and coiled bodies in oligodendrocytes, along with neuropil threads and pretangles. Tau pathology preferentially affects subcortical structures — globus pallidus, subthalamic nucleus, substantia nigra, pons, ocular motor nuclei, and dentate nuclei — with cortical involvement in cortical variants (PSP-F, PSP-SL, PSP-CBS). Midbrain and superior cerebellar peduncle atrophy are key pathologic markers that correlate with disease progression.

Clinical Features

Falls and Gait Impairment

Early postural instability with unprovoked falls, typically backward, is the hallmark of PSP-RS. The gait is characteristically stiff, broad-based, and upright — contrasting with the stooped, shuffling gait of PD. Patients are often described as walking "like a toy soldier" or standing stiffly with a retrocollis posture. The rocket sign — rising rapidly and recklessly from a chair despite obvious instability — is characteristic. Falls usually begin within the first 1–3 years of symptom onset (compared with >10 years in PD). Gait freezing is the predominant feature in PSP-PGF, which has a notably longer survival (11–15 years).

Ocular Motor Dysfunction

Ocular motor findings are central to PSP diagnosis, though they are present in only 41% at disease onset and may not develop for years in non-RS variants.

Vertical supranuclear gaze palsy is the most recognized finding, with downgaze affected before upgaze. This is a critical clinical distinction: restricted upgaze alone is nonspecific and common in normal aging, PD, MSA, and other conditions. The supranuclear nature is confirmed by intact vestibulo-ocular reflexes (doll's eyes maneuver), which bypass the supranuclear pathways.

Slowed vertical saccades are the most sensitive and specific early sign — they precede frank gaze palsy and can be detected at the bedside by asking the patient to look quickly between two vertical targets. Horizontal saccades are eventually involved, leading to the characteristic "round-the-house" sign — oblique saccades during attempted horizontal gaze due to combined vertical and horizontal slowing.

Other ocular findings include: square-wave jerks (involuntary saccadic intrusions during fixation), eyelid dysfunction (eyelid opening apraxia, reduced blink rate, blepharospasm), convergence insufficiency (causing reading difficulty and diplopia), and photophobia. The combination of frontalis overactivity (compensating for eyelid apraxia) and reduced blink rate creates the characteristic "astonished" or "worried" facial expression — sometimes called the "sunglass sign" when patients wear dark glasses indoors for photophobia.

Speech and Swallowing

Dysarthria in PSP is characteristically spastic-hypokinetic — a mixed pattern with both strained/strangled quality (spastic component) and reduced volume/monotone (hypokinetic component). This differs from the purely hypokinetic dysarthria of PD. Dysarthria typically develops within 3 years of onset and may be an early presenting feature in PSP-SL (speech-language variant). Dysphagia is the leading cause of mortality in PSP, through aspiration pneumonia. Swallowing evaluations should be performed regularly and early, with diet modification and aspiration precautions. PEG tubes can help with nutrition and medication delivery but do not prevent aspiration.

Cognition and Behavior

The cognitive profile is frontal-subcortical: apathy (the most common behavioral feature), bradyphrenia, impaired verbal fluency (especially phonemic), dysexecutive syndrome, and behavioral changes (disinhibition, impulsivity, perseveration). Memory is relatively preserved early. The frontal assessment battery is a useful bedside tool. When frontal-behavioral features dominate early, patients may be misdiagnosed with frontotemporal dementia (PSP-F variant). The cognitive profile differs fundamentally from the amnestic pattern of Alzheimer disease and the visuospatial/attentional pattern of DLB.

Sleep

RBD is rare in PSP — in stark contrast to PD, DLB, and MSA. Similarly, restless legs syndrome and respiratory disturbances are uncommon. The absence of RBD in a patient with parkinsonism should raise suspicion for a tauopathy (PSP, CBS) rather than a synucleinopathy. Insomnia is common, however, and often attributed to rigidity, pain, inability to turn in bed, and nocturia.

Phenotypic Variants

The 2017 MDS criteria recognize eight phenotypic variants, each defined by a combination of OPAC domain features (Ocular, Postural, Akinesia, Cognitive) at different levels of certainty (1 = highest, 2 = midlevel, 3 = lowest). As disease progresses, most variants converge toward a Richardson-like phenotype.

Emerging variants not yet included in the diagnostic criteria include PSP with cerebellar ataxia and PSP with primary lateral sclerosis. A "protracted course PSP" has also been proposed, characterized by slow progression, 10–15 year duration, and more localized tau pathology.

Diagnostic Criteria

2017 MDS Criteria — Levels of Certainty

The 2017 MDS criteria include four diagnostic levels: Definitive (neuropathologic confirmation), Probable (high specificity), Possible (higher sensitivity, lower specificity), and Suggestive of (earliest suspicion). The criteria require progressive onset at age ≥40 years with primarily sporadic presentation.

Supportive features (lacking specificity but increasing clinical suspicion): levodopa resistance, spastic-hypokinetic dysarthria, early dysphagia, photophobia, midbrain atrophy or hypometabolism, and abnormal DaT-SPECT.

Exclusion criteria: prominent episodic memory deficits (AD-pattern), prominent early dysautonomia (suggests MSA), appendicular ataxia, hallucinations, cognitive fluctuations, motor neuron findings, history of encephalitis, sudden onset or stepwise progression, severe leukoencephalopathy, hydrocephalus, or major stroke on imaging.

OPAC Framework

Diagnosis is built on four core domains — Ocular (O), Postural (P), Akinesia (A), Cognitive (C) — with features graded by severity level (1 = highest certainty, 2 = mid, 3 = lowest).

Multiple Allocations eXtinction (MAX) rules were developed to resolve diagnostic overlap when a patient meets criteria for multiple variants simultaneously. The rules weight toward the initial phenotype and assign diagnoses by certainty level. Most patients progress from "suggestive of" or "possible" to "probable" PSP within a mean of 3.6 years.

Imaging and Biomarkers

Structural MRI

MRI is the most helpful diagnostic tool in PSP, serving both to identify characteristic atrophy patterns and to exclude mimics (chronic small vessel disease, hydrocephalus, mass lesions).

Midbrain atrophy is the hallmark finding: the hummingbird sign (midsagittal) and morning glory / Mickey Mouse sign (axial) reflect selective midbrain tegmental volume loss. Quantitative measures include the midsagittal T1 midbrain-to-pons area ratio (>5 supports PSP).

Superior cerebellar peduncle atrophy is another key marker that correlates with disease progression. The MRPI 2.0 (Magnetic Resonance Parkinsonism Index) integrates pons-to-midbrain area ratios with middle cerebellar peduncle, frontal horn, and third ventricle measurements — differentiating PSP-P from PD with 94.3% specificity. An automated version has been validated for clinical use.

Advanced diffusion MRI techniques, including free water imaging, can accurately distinguish PSP from PD and MSA-P. The automated imaging differentiation in parkinsonism (AID-P) protocol has been validated across multiple sites.

Functional Imaging

FDG-PET: Frontal and midbrain hypometabolism supports PSP diagnosis. DaT-SPECT: Usually abnormal in PSP but nonspecific (also abnormal in PD, MSA, DLB, CBS). A normal DaT-SPECT makes degenerative parkinsonism unlikely. The "suggestive of" PSP diagnosis has a nearly 14% false-positive rate, with differential including MSA, CBS, and PD — highlighting the need for better early biomarkers.

Tau PET

Tau PET remains primarily research-based. The widely used first-generation tracer ¹⁸F-flortaucipir (AV-1451) binds 3R+4R tau well in Alzheimer disease but has weaker binding in 4R-predominant tauopathies like PSP. Second-generation tracers — ¹⁸F-PI-2620, ¹⁸F-MK6240, ¹⁸F-PM-PBB3 — show increased binding in PSP and CBS and may become clinically useful in combination with MRI. Off-target binding remains a limitation across all tracers.

Fluid Biomarkers

NfL (plasma or CSF): Markedly elevated in PSP compared with PD, correlating with disease severity and progression rate. However, NfL is also elevated in MSA and CBS, so it differentiates tauopathies/MSA from PD but cannot distinguish among atypical parkinsonisms.

CSF tau: Total and phosphorylated tau (p-tau181, p-tau217) levels are altered in PSP but are not increased compared with Alzheimer disease, where they are markedly elevated. CSF tau does not reliably differentiate PSP from other atypical parkinsonisms. GFAP is similarly nonspecific.

α-Synuclein SAA: Expected to be negative in PSP (a tauopathy). A positive SAA in a patient suspected of PSP should prompt reconsideration of the diagnosis — favoring PD, DLB, or MSA.

Differential Diagnosis

The differential for PSP is broad and includes conditions that may respond to specific treatments:

Management

Treatment is entirely symptomatic. Multidisciplinary team care — with patient and caregiver at the core — is essential and should include neurology, PT, OT, speech-language pathology, neuropsychology, psychiatry, social work, dietetics, and neuropalliative care.

Disease-Modification Trials

Five major trials and several smaller studies have been completed — all negative. Understanding why they failed is essential for designing the next generation of therapies.

Emerging Therapeutic Approaches

Active immunization: AADvac1, ACI-35, and JACI-35 are tau vaccines in various stages of development, aiming to generate an immune response against pathologic tau species.

Antisense oligonucleotides (ASOs): Target MAPT mRNA to reduce tau production upstream, delivered intrathecally. This approach bypasses the limitation of anti-tau antibodies (which only target extracellular tau) by preventing tau protein synthesis altogether.

O-GlcNAcase inhibitors: ASN90 (FNP-223) blocks tau acetylation, a post-translational modification that promotes aggregation. This oral therapy targets an intracellular mechanism.

Retrotransposon inhibitors: TPN-101, a nucleoside analogue reverse transcriptase inhibitor originally developed for HIV, blocks LINE-1 and other retrotransposons thought to drive neuroinflammation. Preliminary Phase 2 data show reduction in NfL and IL-6 — biomarkers of neurodegeneration and neuroinflammation.

Prognosis

Prognosis varies by phenotype. PSP-RS has a median survival of 7–8 years from symptom onset. PSP-P and PSP-PGF have significantly longer courses, averaging ≥9 years and 11–15 years, respectively. Death most commonly results from aspiration pneumonia, respiratory failure, or complications of falls and immobility. The PSPRS is the standard tool for tracking progression (~5.3 points/year decline in PSP-RS). Early palliative care integration improves quality of life for patients and caregivers.

References

1. Höglinger GU, Respondek G, Stamelou M, et al. Clinical diagnosis of progressive supranuclear palsy: the Movement Disorder Society criteria. Mov Disord. 2017;32(6):853-864.
2. Respondek G, Levin J, Höglinger GU. Progressive supranuclear palsy and corticobasal syndrome: diagnosis and treatment. Curr Treat Options Neurol. 2019;21(10):49.
3. Bensimon G, Ludolph A, Agid Y, et al. Riluzole treatment, survival and diagnostic criteria in Parkinson plus disorders: the NNIPPS study. Brain. 2009;132(Pt 1):156-171.
4. Tolosa E, Litvan I, Höglinger GU, et al. A phase 2 trial of the GSK-3 inhibitor tideglusib in progressive supranuclear palsy. Mov Disord. 2014;29(4):470-478.
5. Boxer AL, Lang AE, Grossman M, et al. Davunetide in patients with progressive supranuclear palsy: a randomised, double-blind, placebo-controlled phase 2/3 trial. Lancet Neurol. 2014;13(7):676-685.
6. Höglinger GU, Litvan I, Mendonca N, et al. Safety and efficacy of tilavonemab in progressive supranuclear palsy: a phase 2, randomised, placebo-controlled trial. Lancet Neurol. 2021;20(3):182-192.
7. Dam T, Boxer AL, Bhatt DL, et al. Safety and efficacy of gosuranemab in progressive supranuclear palsy: a phase 2, randomized, placebo-controlled trial. Nat Med. 2021;27(8):1451-1457.
8. McFarland NR, Veilleux Carpentier A. Progressive supranuclear palsy and corticobasal syndrome. Continuum (Minneap Minn). 2025;31(4):1020-1050.