Progressive Supranuclear Palsy: Advances in Diagnosis and Management

Abstract

Progressive supranuclear palsy (PSP) is a complex clinicopathologic disease with no current cure or disease modulating therapies that can only be definitively confirmed at autopsy. Growing understanding of the phenotypic diversity of PSP has led to expanded clinical criteria and new insights into etiopathogenesis that coupled with improved in vivo biomarkers makes increased access to current clinical trials possible. Current standard-of-care treatment of PSP is multidisciplinary, supportive and symptomatic, and several trials of potentially disease modulating agents have already been completed with disappointing results. Current ongoing clinical trials target the abnormal aggregation of tau through a variety of mechanisms including immunotherapy and gene therapy offer a more direct method of treatment. Here we review PSP clinicopathologic correlations, in vivo biomarkers including MRI, PET, and CSF biomarkers. We additionally review current pharmacologic and non-pharmacologic methods of treatment, prior and ongoing clinical trials in PSP. Newly expanded clinical criteria and improved specific biomarkers will aid in identifying patients with PSP earlier and more accurately and expand access to these potentially beneficial clinical trials.

Introduction

The clinicopathological syndrome of Progressive supranuclear palsy (PSP) was originally described in 1964 by Drs. Steele, Richardson, and Olszewski where they detailed a series of patients with postural instability, ocular motor abnormalities, facial and cervical dystonia, dementia, and other features [1]. This seminal work also detailed the initial neuropathologic findings including argyrophyllic globose and flame-shaped inclusions in both the gray and white matter throughout brainstem, subcortical and neocortical regions and accompanying neuronal loss and white matter degeneration. Further study using immunohistochemistry ultimately revealed that these inclusions were accumulations of the tau protein affecting neurons, astrocytes, and glia in a variety of morphologies [2]. Tau is a microtubule associated protein which contributes to the stability of the axonal cytoskeleton. Due to alternative splicing of the of microtubule binding site domains E2, E3, and E10, six tau isoforms exist, either with 4 repeated domains (4R, which includes the E10 region), or 3 repeated domains (3R, which excludes E10) [3]. In Alzheimer’s disease, tau pathology exists in a fairly even 3R/4R mix in paired helical filament conformation [4–6] which occurs in combination with amyloid-β plaque deposits. In contrast, PSP is a ‘pure tauopathy’, where tau accumulations composed primarily of 4R tau species are the pathogenic lesions. For this reason, PSP is often targeted for trials for anti-tau therapeutics. PSP and corticobasal degeneration (CBD), are both 4R tauopathies where tau species are arranged in straight filament conformations [7, 8]. The differences in conformations are due to differing orientations of the C-shaped tau subunits that compose them, with 3R/4R tau subunits being directed towards each other resulting in the helical conformation, whereas in 4R tauopathies, the C-shaped subunits are positioned back to back [9]. While PSP and CBD are both 4R tauopathies, the pathological features of PSP are typically tufted astrocytes and globose neurofibrillary tangles, whereas CBD has astrocytic plaques and ballooned pale neurons with thready neuronal tau inclusions that affect cortical regions more severely than subcortical regions and in a distinct conformation from what is seen in PSP [10, 11].

The prevalence of the classic Richardson syndrome presentation of PSP is approximately 6/100,000, with average age of onset in the mid-60s and disease duration of approximately 6 years [12–15]. However, it has become increasingly recognized that multiple clinical phenotypes aside from the originally described Richardson syndrome phenotype may result from PSP pathology. PSP pathology may be found in patients with parkinsonism mimicking Parkinson’s disease (PSP-P), frontotemporal dementia (PSP-F), and corticobasal syndrome (PSP-CBS), and others [16–18]. The growing recognition of the clinical spectrum of PSP pathology has resulted in an expanded research criteria for the diagnosis of PSP which incorporates these clinical phenotypes [19]. Consequently, more recent age-adjusted prevalence estimates in Europe have increased to 8.8–10.8/100,000 patients [13, 20], and in Yonogo Japan, PSP age-adjusted prevalence increased from 5.8/100,000 patients in 1999 to 17/100,000 patients in 2010 [21, 22]. This recognition has also increased the need for specific biomarkers to diagnose patients early during life; advancements have progressed in MRI, PET, and biofluid biomarkers in PSP. Despite these advances, current therapies for PSP remain symptomatic and disease-modulating medications remain elusive despite extensive efforts. Current strategies are focused on targeting the tau protein by different mechanisms including immunotherapy and gene therapy. Here, we review the clinicopathological complexity of PSP, etiopathogenesis, and emerging biomarkers and well as a review of past and current clinical trials.

Clinicopathologic Complexity of PSP

While several early clinical criteria exist [23–26] the first criteria based on a large autopsy-confirmed series and consensus of experts was performed by Litvan et al. in 1996 (the NINDS-SPSP criteria) which outlined the core clinical features of what is currently known as the PSP-Richardson syndrome (PSP-RS) of a gradual progressive disorder affecting patients over 40 years old with a vertical supranuclear gaze palsy or postural instability and falls within the first year of onset [27]. Having both a vertical supranuclear palsy and postural instability is indicative of clinically probable PSP whereas having either an isolated vertical gaze palsy or having slowed saccades with postural instability is indicative of clinically possible PSP [27]. Pathologic confirmation remains necessary for a ‘definite’ diagnosis of PSP. The clinical entity of PSP-RS has a high probability of demonstrating neuropathologic PSP at autopsy, but many other clinical phenotypes may harbor PSP pathology at autopsy (Figure 1) [17, 27, 28]. Up to one third of PSP patients have a presentation initially marked by parkinsonism (PSP-P), which can have asymmetric akinetic rigid features, resting tremor, and even moderate levodopa responsiveness early in the disease with more classic Richardson syndrome developing only years later [15, 17, 18, 29–31]. Patients with PSP-P have a significantly longer survival compared patients with a classic PSP-RS presentation [17, 18, 31, 32] and progression. Other patients may present with frontotemporal dementia features of prominent apathy and/or disinhibition (PSP-F) [33–35], or a progressive non-fluent aphasia with apraxia of speech (PSP-SL) [36–38]. Up to one third of patients with a corticobasal syndrome presentation with limb apraxia, cortical sensory deficits, alien limb phenomena, dystonia and myoclonus, may harbor PSP pathology at autopsy rather than corticobasal degeneration (PSP-CBS) [10, 39–41]. Other, more rare presentations of PSP have also been described (PSP-PGF: PSP with progressive gait freezing [42, 43], PSP-OM: PSP with exclusive ocular motor dysfunction [17, 27], PSP-PLS: PSP with primary lateral sclerosis [44, 45], PSP-C: PSP with a cerebellar ataxia phenotype [46, 47]). As patients progress, they are more likely to begin to exhibit core features of PSP-RS including supranuclear gaze palsy and postural instability, but these features may be delayed by several years making early and accurate diagnosis a continued challenge [17]. The International Parkinson and Movement Disorder’s Society has published recent diagnostic criteria (MDS-PSP) which acknowledges the wide array of clinical symptoms and signs that may be associated with PSP pathology [19]. These criteria are more sensitive but less specific than the NINDS-SPSP criteria and can lead to diagnosing multiple overlapping phenotypes in the same subject [48, 49]. While the MDS-PSP criteria includes a new definition of the PSP-P phenotype, this criteria needs refinement in view of inclusion of PSP-RS in addition to PSP-P[50]. Thus, subsequent rules (Multiple Allocation eXtinction: MAX rules) were recently developed to improve the assignment of individual phenotypes in clinical practice and research settings [49]. Using MAX rules has improved diagnostic overlap of PSP-RS and PSP-P, but unfortunately may still fail to disentangle 40% of these patients [50].

Figure 1: Clinicopathologic Complexity of PSP.

PSP neuropathology may be associated with a number of different clinical phenotypes (shown in light blue boxes with rounded edges). Other neuropathologies that may also present with these clinical phenotypes are shown in the lower row. Green: Tau, Orange: alpha-synuclein, Yellow: Alzheimer’s disease, Purple: TDP-43. Abbreviations: PAGF: pure akinesia freezing of gait, CBS: corticobasal syndrome, bvFTD: behavioral variant FTD, PPA: primary progressive aphasia, PLS: primary lateral sclerosis, PD/DLB: Parkinson’s Disease/dementia with Lewy bodies, MSA: multiple systems atrophy, CBD: corticobasal degeneration, PSP: progressive supranuclear palsy, AD: Alzheimer’s disease, TDP-43: TAR DNA binding protein 43.

The cardinal neuropathologic features of PSP at autopsy include abnormal accumulations of tau in the forms of tufted astrocytes [51] and globose neurofibrillary tangles in grey matter and coiled bodies in oligodendrocytes in white matter [52]. As stated previously, in PSP the accumulated tau is mostly of a 4R variety and is typically phosphorylated [53, 54], often acetylated [55], and thioflavin-S positive [56]. The globus pallidus, subthalamic nucleus, substantia nigra, and dentate nucleus of the cerebellum are core regions affected [2]. Frontal, temporal, and parietal cortices may also show disease [57] and patients who exhibit more cortically localizing clinical signs have been described to have higher burdens of corresponding cortical tau pathology [58, 59] (figure 2). Pathological subtypes based on the relative distribution of tau pathology and its relation to clinical phenotypes have been proposed [60]. As opposed to corticobasal degeneration (CBD) which is also a 4R tauopathy marked by astrocytic plaque pathology where tau accumulates in distal astrocytic processes which, the tufted astrocytes of PSP have a radial distribution of tau which affects astrocyte cell bodies and more proximal processes (Figure 2) [60].

Figure 2: Pathologic Distribution and Clinical Correlations in PSP.

A. Common clinical features of PSP associated with pathology in these affected regions.

B. Regions in red commonly affected by PSP tau pathology and gliosis including, frontal, temporal, parietal lobes, globus pallidus, putamen, caudate, subthalamic nucleus, hippocampus, midbrain tectum and tegmentum, substantia nigra, pontine base and locus ceruleus, inferior olivary nucleus. Darker areas or red are more commonly affected. Certain phenotypes are more likely to exhibit pathology in specific areas (i.e. PSP-SL and temporal lobe pathology and PSP-CBS with parietal lobe pathology)

C. Characteristic microscopic lesions seen in PSP after immunohistochemical staining for phospho-tau with AT8 antibody. Upper row shows two tufted astrocytes, bottom left showing coiled bodies (asterisk), bottom right showing globose neurofibrillary tangles. Scale bar is 50 μm.

Etiopathogenesis

Environmental Factors

While no single cause of PSP has been identified, a number of environmental and genetic associations have been investigated. The ENGENE-PSP study found that lower educational attainment, exposure to well water and industrial wastes, and firearm use was related to higher risk of developing PSP [61, 62]. A cluster of PSP patients was observed in northern France in an area of high industrial waste contamination that also contained heavy metals and an independent study also documented that occupational exposure to heavy metals was associated with risk of developing PSP [61, 63]. Consumption of high levels of annonacin, a mitochondrial complex 1 inhibitor found in the tropical fruit pawpaw was associated with developing PSP or other atypical parkinsonism in studies in Guadeloupe in the Caribbean [64, 65]. There may be a slight male predominance within PSP patients [18, 24], and one study documented that increased estrogen exposure in women may be associated with lower likelihood of developing PSP [66].

Genetics

Genetic mutations in the MAPT gene have been described leading to PSP [67–69] as well as frontotemporal dementia, FTLD with parkinsonism, primary progressive aphasia and CBD [70]. The H1 haplotype specifically elevates 5.6 times the risk for developing PSP, which is comparable to the ApoE ε3/ε4 risk for developing Alzheimer disease [19, 71]. Interestingly, the H1/H1 haplotype is more common in PSP-RS than PSP-P [72]. A genome wide association study in a large pathologically-validated study confirmed the MAPT variants and the H1 haplotype as associated with PSP, and also identified other gene loci including MOBP, STX6, and EIF2AK3 [73]. MOBP, which encodes for myelin basic protein, is also implicated in CBD and highlights to potential importance of white matter in these conditions [74]. STX6 encodes for a SNARE protein implicated in fusing vesicles in the Golgi network and endosomal structures [75]. EIF2AK3 encodes for a protein responsible for inhibiting protein synthesis in the face of excess endoplasmic reticulum stress [76, 77]. These have been validated in a second GWAS study which also identified SLCO1A2, which is involved in ion transport, and DUSP10, which is involved in tau trafficking, as other genes of interest requiring further study [78].

Oxidative Stress, Mitochondrial Dysfunction, and Inflammation

In PSP, as with other neurodegenerative diseases, mitochondrial dysfunction and oxidative stress has been demonstrated in vitro models and in human tissue [79–82]. Mitochondrial enzymatic activity is decreased and lipid peroxidation is increased in PSP which leads to excessive oxidative stress [81–84]. Interestingly, ATP production is decreased in muscle tissue in PSP patients as well as PSP cybrids [79, 85–87]. Elevated oxidative stress, mitochondrial dysfunction and neurodegeneration, leads to inflammation and in PSP, higher levels of the inflammatory cytokine, IL1β is seen in PSP affected brain regions and correlates with levels of microglial activation [88]. Other studies have reported that microglial activation correlates with tau deposition [89] and higher levels of pro-inflammatory cytokines can be demonstrated in the CSF of PSP patients when compared to Parkinson’s disease patients [90]. Superoxide dismutase and glutathione, essential antioxidants, are often seen to be elevated in PSP brain tissue, likely sources of defense [80, 91]. Activated microglia can be visualized using positron emission tomography in PSP which will be discussed below [92]. Such observations have led to disease modifying trials using anti-inflammatory or anti-oxidant agents. CoQ10 showed improvements in cerebral metabolism as well as improvements in the total PSP-Rating Scale scores and frontal assessment battery in a small six week trial[93]. The short duration of this trial made differentiating symptomatic relief from disease modification difficult. A follow up year-long study of CoQ10 showed a nearly significant difference in the progression of the PSP Rating scale compared to placebo (p=0.07), but also had a high drop-out rate of patients with more severe disease making interpretation of results difficult [94]. Regrettably, trials of riluzole and rasagiline have failed to show meaningful benefit in large trials [95, 96].

Prion-like Spread of Tau

Of note, more recent data suggests that abnormally fibrillated tau is capable of acting as a template to further induce misfolding of normal monomeric tau leading to increasing disease spread in a ‘prion-like’ manner. In vivo animal studies using preformed fibrils [97, 98], human diseased brain homogenates [99], and other techniques [100, 101] have shown distal spread of tau pathology via trans-synaptic spread [102, 103]. There may be specific ‘strains’ of tau capable of seeding unique pathologies [51, 100, 104]. In fact, the recent use of cryo-electron microscopy technology to study several tauopathies shows that the molecular structure of tau varies throughout tauopathies which may in part confer strain-specific properties in these conditions [4, 105, 106]. The molecular structure of CBD related tau has been identified [11] but similar studies for PSP related tau have not been published.

Biomarkers

Magnetic Resonance Imaging

Structural neuroimaging biomarkers in PSP include the well described ‘hummingbird sign’ [107], ‘morning glory sign’ [108], or Mickey-Mouse sign [109] all of which result from midbrain atrophy. One study where features of ante mortem MRIs from 48 pathologically-confirmed PSP and MSA cases were compared, 16/22 (73%) of PSP cases could be correctly identified by a radiologist subjectively identifying the ‘hummingbird sign’, with 100% specificity but only 69% sensitivity [109]. Studies that have relied on the qualitative identification of atrophy patterns document high specificity with lower sensitivity, likely because these changes are visually more apparent in late disease[107–109]. Studies making use of quantitative measurements of the midbrain, superior cerebellar peduncles and other brainstem structures, have shown much higher sensitivity and specificity [110–114]. Perhaps the best studied quantitative MRI signature is the magnetic resonance parkinsonism index (MRPI), which has documented >80% sensitivity and specificity differentiating PSP from non-PSP patients and predicting the occurrence of a supranuclear gaze palsy in PSP-P patients [111, 112, 115]. A subsequent version has been reported to predict clinical evolution of PSP-P in patients initially diagnosed with PD, raising the possibility of its utility to differentiate PD from PSP in early disease when clinical diagnosis is often most ambiguous [116, 117].

Positron Emission Tomography

Several tracers are under development that bind to the tau protein including ¹⁸F-5105, ¹⁸F-FDDNP, ¹⁸F-THK523, ¹¹C-PBB3, and others [118]. ¹⁸F-Flortaucipir (formerly AV-1451 and T807) is the most researched tau tracer to date and binds to paired helical filaments in 3R/4R tauopathies such as AD [119] and exhibits expected retention patterns in both amnestic AD [120, 121] and non-amnestic variants [122, 123]. However, retention appears to be less robust in 4R tauopathies [119, 124, 125]. While group-wise differences can still be demonstrated differentiating PSP from healthy controls largely due to increased retention in the basal ganglia and nigral regions [120, 126–128], individual patient-level distinctions remain difficult because of significant overlap and the ligand is not expected to be of use diagnosis PSP subjects in early stages of the disease. PET tracers targeting activated microglia (¹¹C-(R) PK11195) may offer methods to assess inflammation associated with neurodegeneration in PSP and other related diseases [92, 129]

Biofluids

CSF biomarkers for PSP are still under development. Tau fragments, including measures of total tau (t-tau) and phosphorylated tau (p-tau) have been more extensively studied in Alzheimer’s disease [130], but these tau species tend not to be reliably elevated in PSP [7, 131, 132]. One study reported that a ratio of certain tau fragments may aid in distinguishing PSP from healthy controls and different neurodegenerative diseases [133] but the findings were not easily replicated [134]. CSF neurofilament light chain (NfL), an intermediate filament and non-specific measure of neuronal injury [135], shows elevation in PSP and other atypical parkinsonian syndromes [132, 136–139]. New single molecular arrays (SIMoA) are capable of detecting NfL on the ng/L levels, making blood based assays possible. Serum NfL correlates tightly with CSF NfL concentrations [140] and higher baseline levels of serum NfL has been associated with worse clinical and radiologic outcomes in PSP [141]. Thus, serum NfL could be used as secondary outcome measures in future therapeutic trials. Real time quaking induced conversion (RT-QuIC), which was originally pioneered in Creutzfeldt-Jakob disease, makes use of the ability of abnormally aggregated proteins to act as a template to seed further aggregation of monomeric proteins [142]. RT-QuIC are very promising in the diagnosis of synucleinopathies (Parkinson’s disease, dementia with Lewy bodies, multiple systems atrophy) [143–145] and is also expanding into tauopathies including PSP [146, 147].

Neurophysiologic Markers

The slowness of vertical eye movements that is a clinical hallmark of PSP can be demonstrated through electro-oculogram, a version of electromyogram [148] and more recently small studies of non-invasive computerized eye tracking software can reliably differentiate PSP from non-PSP cases using a variety of eye movement features [149, 150]. Spontaneous blink rate is decreased in PSP as expected, and in comparative neurophysiologic studies, blink rate is much lower in PSP than PD or other disorders [151]. Early recovery and enhanced excitability of the blink reflex has been shown as well [152]. Auditory startle reflex is severely decreased in PSP, likely related to damage to the reticulospinal system [153, 154]. Abnormal autogenic inhibition of spinal interneuron circuits has been described as well [155].

Current Therapies

Symptomatic Pharmacologic Therapies

Current pharmacologic therapies for PSP are symptomatic, and tend to show mild to moderate efficacy. Levodopa preparations may be used to treat functionally-impairing bradykinesia and rigidity. In one retrospective study of pathologically confirmed PSP patients, 32% of cases showed a >30% improvement in the Unified Parkinson’s Disease Rating Scale and 4% of cases showed levodopa induced dyskinesias [18]. Other studies have documented similar response rates[156–159]; however, levodopa responses are often milder than what is seen in PD and require higher doses of levodopa to achieve and can often wane over time [26, 160, 161]. It is recommend titrating up to 1.0 gm/day and continuing this dose for at least a month to attempt to achieve benefits before weaning. Marked and prolonged improvement with levodopa therapy is considered an exclusionary criteria for a diagnosis of PSP and makes a diagnosis of Parkinson’s disease more likely. Dopamine agonists have also been trialed in PSP but are generally less effective than levodopa and carry a greater likelihood of causing side effects [157, 162, 163]. A few studies have documented improvement in features of bradykinesia, rigidity, and dystonia with the use of amantadine in PSP but side effects have been also been reported in nearly half of patients treated with this medication, ranging in severity from leg edema and livedo reticularis to hallucinations and worsened cognitive impairments [157, 164–166]. If dose-limiting side effects do occur, we recommend a slow wean of medications, removing 100 mg per week, as a withdrawal syndrome with delirium from abrupt cessation of amantadine has been described [167, 168]. Zolpidem was reported to offer mild improvement of ocular motor deficits and saccadic speed, but this has not been confirmed in other studies [169–171]. Ophthalmic lubricants are useful to treat dry eyes. Sunglasses are also useful for the photosensitivity. Prism glasses may be employed to improve the double vision due to decreased convergence, but when not useful, alternating eye patches may be employed. Problems with sleep initiation and sleep maintenance are common in PSP and there have been no major studies of pharmacologic interventions, but treatment can be attempted with melatonin, clonazepam or trazodone [172–174]. Constipation may be managed with dietary changes, agents that accelerate bowel movements, gentle laxatives, or by increasing fluid secretion. For depression, tricyclic antidepressants, selective serotonin reuptake inhibitors and serotonin-norepinephrine reuptake inhibitors may be employed. However, they do not improve the apathy that in PSP is often prevalent. Amitriptyline, a TCA, in particular has been reported to improve depression in PSP patients and has improved motor parkinsonism in case reports [157, 175, 176]. For cognition, while donepezil may show mild selective benefits on cognition in PSP, currently this medication is not recommended due to potential deleterious effects on gait and dysphagia [177]. Botulinum toxin for blepharospasm, eyelid opening apraxia, retrocollis, or sialorrhea may be considered but must be weighed against the potential to cause side effects including worsening dysphagia [178–181]. Sublingual administration of atropine 1% drops can be considered but carries a significant risk of causing cognitive and urinary side effects if not dosed carefully [182]. Urinary urgency and frequency is a common symptom in PSP and can be treated with medications which are less likely to cross the blood brain barrier including mirabegron and solifenacin. Pseudobulbar affect, abrupt crying or laughing episodes that are not always mood-congruous or proportional to stimuli, can be treated effectively with dextromethorphan-quinidine or antidepressants [183]. Agitation may occur in PSP. Seroquel or clozapine could be used as pharmacologic treatments to augment non-pharmacological interventions. However, there are no therapeutic trials showing beneficial effects in PSP, but clozapine as well as pimavanserin have been shown to improve psychosis in PD, although they are rarely used in PSP. Clozapine therapy must be accompanied by frequent blood monitoring to avoid the rare but life-threatening side effect of agranulocytosis. Other neuroleptics and dopamine blocking anti-emetics should be avoided in PSP given the possibility of worsening parkinsonism. There are no formal studies of medical marijuana or cannabidiols in PSP, but studies in other conditions have shown improvements in sleep, pain, anxiety, and spasticity which are common problems in PSP, so we await formal studies in PSP to guide utility [184–187].

Non-pharmacologic therapies

Patients with PSP benefit from multidisciplinary non-pharmacological care. Physical therapy improves function and decreases falls [188, 189]. Use of weighted walkers are helpful to improve patient’s safe gait and independence. Speech therapy may help with coping strategies for vocal changes, techniques for communication, and safe eating and drinking. When difficulties with swallowing liquids are identified, a modified barium swallow evaluations can diagnose the extent of the problem and help identify proper compensation techniques or diets. Occupational therapists can do safety inspection of the homes. Social workers and palliative care consultants can aid in the management of the PSP patients and families including stress, nursing home placement, end of life care and decision making which positively affects patients and families quality of life [190]. Therefore, PSP patients and families benefit from having a multidisciplinary team knowledgeable in the management of these patients.

Clinical Trials

Drug development for PSP disease modulating trials has focused on inhibiting post translational modifications of tau, enhancing immune mediated clearing, stabilizing microtubules, or reducing levels of expression through gene therapies.

Tau post-translational modifications: Phosphorylation and Acetylation, and Others

In PSP aggregated tau is hyper-phosphorylated [191]; therefore, kinases which phosphorylate tau, including GSK 3β, have been examined as potential therapeutic targets [192]. Both valproic acid and lithium are GSK 3β inhibitors which showed some promise in animal models [193–195]. Valproic acid did not result in significant improvements in human trials and the lithium trial was stopped early because of poor tolerability (NCT00385710, NCT00703677) [196]. Tideglusib is a small molecule inhibitor of GSK 3β which failed to show significant clinical differences in the PSP-Rating Scale between treated patients and patients receiving placebo in a multicenter, randomized, double-blind, placebo-controlled trial [197]. However, in a subset of patients with MRIs, less brain atrophy was observed in treated patients than controls in parietal and occipital cortical regions [198]. These areas are not typically affected by tau pathology in PSP as are the frontal lobes, basal ganglia, or brainstem structures [197, 198]. Possible explanations for this observation include better quality of volumetric MRI data from cortical areas versus subcortical and brainstem regions, variable expression of GSK 3β throughout the brain, or selection bias of the included MRI subgroup [198].

CDK5 is another kinase which phosphorylates tau and inhibitor molecules are in development (NCT04253132). Salsalate inhibits tau acetylation which may alleviate tau pathology as well [199] and is undergoing a phase I pilot study in PSP (NCT02422485). O-GlcNAc modification, which is involved in intracellular tau trafficking, and caspase-mediated cleavage are other potential therapeutic targets [200–202]. Most recently, the company Retrotope has obtained orphan drug designation for a stabilized fatty acid compound to prevent lipid peroxidation for investigation in PSP although no trials in humans have been conducted yet [203].

Microtubule-stabilizing agents

A neuropeptide with neuroprotective and microtubule stabilizing properties, davunetide, which looked promising in animal models [204] was studied in a multicenter phase IIb/III trial of over 300 patients with PSP but regrettably failed to show clinical efficacy on all endpoints [205]. TPI-287 is a taxane derivative that stabilizes microtubules, is able to cross the blood brain barrier, and may decrease cell proliferation in cancer [206]. A preliminary study resulted in some anaphylactoid reactions at the higher dosing arm without significant clinical improvements and drug development on this agent is likely stopped [207]. A separate compound, epothilone D, which also stabilizes microtubules may be of interest, but is still early in development [208].

Tau Immunotherapy

Recent in vitro and in vivo experiments have shown that abnormally folded proteins may be capable of inducing misfolding in normal native proteins and creating spread of pathology through ‘prion-like’ templating [99, 103, 209, 210]. Brain homogenates from human PSP and CBD induce tau inclusions in mice that spread well beyond the injection sites [99, 211]. These aggregations may be transported within the cellular structure and be transferred from cell to cell in a variety of mechanisms [212, 213]. Such experiments have also suggested different therapeutic strategies in PSP including tau-directed immunotherapy to promote clearance of aggregates before further toxicity can occur.

Again, a number of studies in animal models using different immunization strategies, showed reduction of tau pathology with favorable safety profiles [214–219]. Several tau directed passive immunotherapy trials in Alzheimer’s disease have been performed with limited success (for a recent review see [220]). In PSP, Biogen’s antibody product BIIB092 (Gosuranemab), directed against N-terminal fragments of extracellular tau [221], showed a favorable safety profile in a phase I trial and lowered CSF tau levels as a measure of target engagement [222] [223]. However, a phase II study (PASSPORT NCT03068468) was halted as the trial failed to show differences in the primary and secondary endpoints [224]. Abbvie also tested a monoclonal antibody product, ABBV-8E12 with favorable phase I safety results [225] and adequate, dose ascending CSF penetration [226]; however, a phase II study in PSP was recently stopped after failing a futility analysis. Other antibody products are in development (NCT04185415).

A number of challenges and questions remain in the field of immunotherapy and neurodegeneration. Selecting the right epitope target may be challenging as antibodies can be raised to target different regions of tau [227, 228], tau species with specific post-translational modifications [229, 230], or oligomers or specific conformations [120, 228, 231–234]. Ensuring adequate penetration of the blood brain barrier of peripherally administered products is a challenge and several different strategies including viral vector delivery, ultrasound with microbubbles, or antibodies tagged to small molecules are being explored [120]. Development of these and other strategies are critical given that brain penetration of intravenously administered IgG antibodies is typically cited as 0.1–1% [223, 235–239]. Most of the products to date have targeted predominantly extracellular tau, which likely constitutes ghost pathology from deceased cells but may be beneficial in intercepting the transmission of misfolded toxic species from cell to cell. Intracellular tau is more difficult to target with polarized antibody products and it is not certain whether the attendant immune response would be prohibitively damaging [240]. Targeting other tau epitopes may provide different responses compared to the agents that have been tested to date.

Gene Therapy

Tau reduction through genetic strategies may prove beneficial in PSP. In mouse models, there are conflicting reports of whether tau knockouts have preserved function [241, 242] or if they exhibit cognitive or motor symptoms [243–246]. Anti-sense oligonucleotides (ASOs) reduce protein expression by binding to mRNA where it can be degraded by mRNA-ase H to prevent translation. Specific ASOs that are capable of lowering total tau expression or shifting expression from 4R to more 3R tau [247] have been tested in animal models [248–250]. Small interfering RNA (siRNA) also prevent protein expression by binding to mRNA prior to translation and have been examined in mouse models [251]. Similar to antibody therapies, drug delivery to ensure adequate brain penetration is a consideration. Again strategies include tagging molecules to more lipophilic compounds [252] or other small molecules [253, 254], intrathecal injections [255], intraparenchymal or intraventricular injections [256], and the use of viral vectors [257, 258].

Conclusion

Progressive supranuclear palsy is a complex clinicopathologic entity with diverse clinical manifestations which can led to delays in diagnosis. While the Richardson syndrome is the most indicative of underlying PSP pathology at autopsy, growing understanding of the diverse clinical phenotypes resulting from PSP pathology has led to a significant expansion of the PSP diagnostic criteria aimed at increasing sensitivity of diagnosis. Clinical treatment of PSP is supportive and there is mild to moderate efficacy of symptomatic therapies to treat a myriad of associated symptoms that can occur with the disease; however, disease modulating treatments have remained elusive. New generation immunotherapies rationally designed small molecules, and genetic therapies directed against the tau protein are under development. Multiple challenges remain in the drive to establish disease modulating therapies in PSP. The use of better imaging and biofluid biomarkers to promote early and accurate diagnosis continues to be essential to aid in ensuring the enrollment of appropriate patients who are likely to more clearly show beneficial effects into clinical trials. Furthermore, the establishment of sensitive and specific markers to track disease progression in PSP to augment clinical scales is crucial for evaluation of therapies. Lastly, many treatments have failed to make the leap from animal models to human therapies. Underlying differences in physiology, the appropriateness of the original model systems, and the challenges of drug delivery in humans are major considerations as development continues in the search for disease modifying therapies in PSP, but several antibody products, small molecules, and gene therapy strategies are still under development.

Table 1:

Mechanistic Based Trials for PSP

Pathogenic Mechanism	Therapeutic Mechanism	Drug Name	Trials/References	Phase	Main Outcomes	Results	Ongoing
Inflammation	Anti-inflammatory/antioxidants	CoQ10	[93] [94]	II, II	PSP-RS, ADL, MMSE, PDQ-39	No benefit	Completed
		Riluzole	[95]	III	Survival, H/Y, SEADL, MMSE	No benefit	Completed
		Rasagiline	[96]	II	PSP-RS	No benefit	Completed
Lipid Peroxidation	Lipid Stabilization	RT-001	[203]	Pre-Clinical
Tau	GSK-3β Inhibitors	Valproic Acid	NCT00385710	II	PSP-RS	No benefit	Completed
Phosphorylation			[196]
		Lithium	NCT00703677	I (PSP)	Tolerability, PSP-RS, PSP-QOL	Not tolerated	Completed
						No benefit
		Tideglusib	NCT01350362	II	PSP-RS, DRS, SEADL,	No benefit	Completed
			[197] [198]		EuroQol
	CDK5 inhibitors	Tolfenamic Acid	NCT04253132	IIa	Tolerability		Not yet recruiting
Tau Acetylation	Acetylation Inhibitors	Salsalate	NCT02422485	I	Tolerability, PSP-RS	Not posted	Active, not recruiting
O-GlcNAc	O-GlcNAc Inhibitors		[200] [201]	Pre-clinical
Tau Assembly	Microtubule Stabilizing Agents	Davunetide	NCT01056965	II/III	PSP-RS, SEADL, RBANS	No benefit	Completed
			[205]
		TPI-287	NCT02133846	I	Tolerability	Not tolerated	Active not recruiting
			[207]
		Epothilone D	[208]	Pre-clinical
Tau	Passive	ABBV-8E12	NCT02985879	II	PSP-RS, Tolerability	No benefit	Completed
Accumulation/Spread	Immunotherapy		NCT03391765	II Extension	PSP-RS	No benefit	Completed
		BIIB092	NCT03068468	II	PSP-RS, Tolerability	No benefit	Completed
			[224]
		UCB0107	NCT04185415	Ib	Tolerability		Recruiting
	Active Immunotherapy	No agents in PSP
Tau Production	ASO		[251]	Pre-clinical
	siRNA		[251]	Pre-clinical
Neurotrophic	Neurotrophic Inducer	AZP2006	NCT04008355	II			Not yet recruiting
Inducer

Abbreviations: PSP-RS: Progressive Supranuclear Palsy Rating Scale, ADL: activities of daily living, MMSE: minimental status exam, PDQ-39: Parkinson's disease questionnaire-39, H/Y: Hoehn and Yahr scale, SEADL: Schwab and England activities of daily living, PSPQOL: PSP quality of life scale, EuroQol: European quality of life scale, RBANS: repeatable battery for the assessment of neuropsychological status, ASO: antisense oligonucleotides, siRNA: small interfering RNA.

Highlights.

• New PSP diagnostic criteria recognizes diverse phenotypes, but refinement is needed

• Multiple genetic and environmental factors increase the risk for developing PSP

• Improving biofluid and imaging biomarkers will aid in early and accurate diagnoses

• Symptomatic treatments with a multidisciplinary approach are standard of care

• PSP is an ideal disease for novel therapeutic approaches targeting the tau protein