You’ve all heard of the ten known PSP subtypes. They’re classified into three groups by the general areas of the brain involved (i.e., cortical vs. subcortical). Here’s the list:
Cortical and subcortical (~50% of all PSP)
PSP-Richardson syndrome (PSP-RS)
Cortical (~20%)
PSP-frontal (PSP-F)PSP-speech/language (PSP-SL)
PSP-corticobasal syndrome (PSP-CBS)
Subcortical (~30%)
PSP-Parkinsonism (PSP-P)
PSP-progressive gait freezing (PSP-PGF)
PSP-postural instability (PSP-PI)
PSP-ocular motor (PSP-OM)
PSP-cerebellar (PSP-C)
PSP-primary lateral sclerosis (PSP-PLS)
(As an aside: Neurodegenerative diseases are defined mostly by their pathological (i.e., microscopical) appearance, but each disease so defined may have several possible sets of outward signs and symptoms in the living person, depending on the general locations of the pathology within the brain. We deal with this by hyphenating the names of neurodegenerative diseases, with the pathology first and the clinical picture second.)
Now, to the news: Researchers at the Institute of Science in Tokyo, the Mayo Clinic, and UC San Diego have refined the above subtype classification. First author is Dr. Daisuke Ono, senior author Dr. Dennis Dickson and eight colleagues included 588 autopsy-proven cases of PSP from the Mayo brain bank without evidence of other neurodegenerative diseases. First, they used ChatGPT’s large-language software to extract clinical data from 53,527 pages of medical records, tabulating the order of appearance and progression rate of 12 pre-specified, PSP-related symptoms in each patient. Next, they performed a statistical technique standard for this sort of thing called “cluster analysis,” coupling it with a “decision tree model.” The first revealed groups of symptoms and progression rates that occurred together more often than would be expected by chance. The second worked out a practical, step-by-step way for neurologists to assign an individual patient to a subtype.
The most important result was a new subtype combining some patients with what has been defined as PSP-F with some from PSP-PI. The analysis still found statistical justification for continuing to recognize those two familiar categories as bona fide subtypes on their own. The new subtype, called PSP-PF (the continuous red curve), has the dubious distinction of having the most rapid progression and shortest total survival of all. In the graph below, you can trace a vertical line from where the “median” line crosses the curve for each subtype to find the median survival on the horizontal axis.
The median survival of PSP-PF was 6 years, with a 25-75 interquartile span (i.e., the middle two quarters of the group) of 5-7 years. This compares to PSP-RS, with a median of 7 and a 25-75 span of 6-8. For the subjects remaining in the PSP-PI and PSP-F groups, the median survival figures were 8 and 9 years, respectively.
This re-shuffling isn’t just a statistical detail. In the total group of 588 patients, 188 (32%) had PSP-PF, while only 68 (12%) had PSP-RS. Even considering the bias of any autopsy series toward over- representation of atypical cases, it’s still remarkable that PSP-PF is far more common than the other non-RS subtypes.
All that should be accompanied by the standard scientific conservatism about adopting new findings as textbook-worthy, especially without independent confirmation. Weaknesses in this study, all of which are recognized by the authors, include the following:
When a clinical feature wasn’t mentioned in the records, the analysis treats it as if it were known to be absent.
Quantitative data such as drug dosages, blood tests results, cognitive test results, and imaging details were not considered.
If a symptom onset date was not mentioned in the records, the date of the first relevant physician visit was used as the equivalent.
Having recognized all that, we can still say that an AI-based procedure may have found a pattern in ordinary medical record data that human neurologists and researchers missed.
My title for this post tentatively calls the discovery “unwelcome” only because no one would be glad to learn that their subtype of PSP is more rapidly progressive than they thought. (I’m referring to those people with PSP-PI and PSP-F who would fall within the definition of the new PSP-PF.) But one upside is that the news that a large group of people with PSP has a rate of progression similar to that of PSP-Richardson could allow neurologists to better counsel patients and their families. Another important upside is that perhaps clinical neuroprotection trials could now enroll participants with both PSP-PF and PSP-RS instead of confining themselves to the latter. This could greatly increase the pool of eligible trial participants and shorten the time required for the recruitment and double-blind periods.
The main potential obstacle to enrolling participants with PSP-PF into trials is that the primary outcome measure, the PSP Rating Scale, has not yet been validated for that subtype. But that should be possible to accomplish by identifying people with PSP-PF in existing, longitudinal or retrospective observational cohorts using the decision rubric of Ono et al. Then, one would simply assess the ability of their existing, longitudinally administered PSP Rating Scale scores to track their symptoms over time.
So the “unwelcome” part of this won’t actually change anyone’s PSP for the worse and the upside of speeding up clinical trials would be most welcome.
Yesterday’s post was about the clinical heterogeneity of PSP and how it prompts a theory about the cause(s) of the disease. A couple of hours after I hit “send” I saw a new paper that indirectly supports my idea.
As you probably know, PSP comes in ten known subtypes. The original type, first described in detail in 1963, is called PSP-Richardson syndrome and accounts for about half of all PSP. The other nine have been described since 2005. The new paper reports five subtypes among PSP-Richardson syndrome itself.
The study is from Dr. Mahesh Kumar, a post-doc at the Mayo Clinic, with Dr. Keith Josephs as senior author. They performed statistical tests called “network analysis” and “cluster analysis” on their 118 patients with PSP-Richardson. The five PSP-Richardson “sub-sub-types” emphasize, respectively, tremor; light sensitivity; reduced eye movement (i.e., supranuclear gaze palsy); cognitive loss and slowness/stiffness.
These are not just points on a continuous spectrum. Rather, in each of the five PSP-Richardson sub-sub-types, a group of features and their severities occurs together in individuals in a combination that would not be expected by random combination based their respective frequencies in the total PSP-RS population. For example, people with worse slowness/stiffness tended to have milder eye movement problems and worse cognition than chance would dictate.
Here’s a graphical representation of the results. The features represented by the circles in each group interact with one another in a mutually reinforcing (the black bars) or interfering (the red bars) way. The thickness of the bars represents the strength of the interaction. An explanation in the researchers’ own words follows:
From Kumar et al. Mov Dis Clin Prac 2025 Network Analysis showing 16 signs/symptoms and their associations. Each node in figure represents symptom/sign, Black edges represent positive connection, and red edges represent negative connection; thicker edges represent stronger association. V1, Sensitivity to bright light; V2, MoCA (Cognition Score); V3, Neck Rigidity; V4, Urinary Incontinence; V5, Emotional Incontinence; V6, Upward ocular movement dysfunction; V7, Downward ocular movement dysfunction; V8, Horizontal ocular movement dysfunction; V9, Eye lid dysfunction; V10, Limb apraxia; V11, FAB (Executive Score); V12, Gait dysfunction; V13, Bradykinesia; V14, Postural tremor; V15, Kinetic tremor; V16, Rest tremor.
All this begs the question as to the basis of the specific groups of signs and symptoms. The answer will probably apply as well to the ten PSP subtypes as to the five PSP-Richardson sub-sub-types. It probably has to do with the specific combination of PSP’s menu of causative factors at work in the individual. As I pointed out in my last post, there are 14 known gene variants contributing to PSP risk and that number is growing. Exposure to toxic metals may also be a factor and those exposures could come at different times of life and in various durations, intensities and combinations. The number of genetic/toxic combinations of these factors sufficient to cause PSP would be astronomical, and the likeliest combinations might account for the likeliest PSP subtypes and sub-sub-types.
Then throw in the stochastic factors, meaning random throws of the dice. I’ll get to that in a future post.
“Pheno-“ is the Greek root for “outward appearance” and so far, PSP has ten of them. The differences arise from varying emphasis of the degenerative process among different parts of brain. While standard laboratory methods show that at the cellular level the pathology among the ten phenotypes is identical, a few details are starting to emerge using more recent and sophisticated techniques at the molecular level.
The most common PSP phenotype, called PSP-Richardson’s syndrome (PSP-RS), is the one Steele, Richardson and Olszewski originally described in 1964. The others were published piecemeal starting in the early 2000s. J.C. Richardson was the leader of the trio at the University of Toronto, a senior clinical neurologist who noticed an unusual form of Parkinsonism among his patients in the 1950s and 60s. John C. Steele was his trainee and Jerzy Olszewski was the neuropathologist who described the corresponding microscopical abnormalities. So, it’s altogether fitting and proper that Dr. Richardson should be honored in this way.
Prevalence of the phenotypes. The percentage of the PSP population with each phenotype has not been studied in a true community-based population. The published percentages vary widely across centers and are all from referral-based populations at research institutions, where unusual forms of diseases are over-represented to varying degrees. Even without that issue, several things make it hard to be sure of the prevalence of the various phenotypes:
It’s difficult to estimate the population prevalence of atypical cases of PSP from autopsy series because atypical cases are more likely to come to autopsy, and without autopsy, it’s hard to know that someone with atypical PSP really had PSP.
In their later years, all of the phenotypes tend to merge into a PSP-RS appearance, so the relative frequencies of the phenotypes may depend on the patients’ disease stage when the researchers evaluated them.
Clear diagnostic criteria do not yet exist for many of the phenotypes and many patients satisfy criteria for more than one, even in early stages. A method has been published for how to deal with this, but most of the publications antedate or ignore it.
Although the original differentiation of PSP-RS from PSP-P in 2005 did use a rigorous statistical technique called “factor analysis” to confirm that the two are distinct, this is not usually the case for the other phenotypes relative to PSP-RS or to one another.
Nevertheless, here are my very rough estimates of their contributions to PSP in general, based on a Gestalt impression of the literature:
Richardson’s syndrome
45%
PSP-RS
Parkinsonism
25%
PSP-P
Frontal
10%
PSP-F
Progressive gait freezing
5%
PSP-PGF
Speech/language
5%
PSP-SL
Corticobasal syndrome
3%
PSP-CBS
Postural instability
3%
PSP-PI
Ocular motor
3%
PSP-OM
Cerebellar
<1%
PSP-C
Primary lateral sclerosis
<1%
PSP-PLS
Many recent research articles group these into three categories based on their anatomical predilections in the brain: cortical vs subcortical. PSP-RS falls into neither of these because it has approximately equal degrees of both cortical and subcortical features. One important, practical reason for the grouping of phenotypes is to have groups large enough for meaningful statistical analysis.
PSP-Parkinsonism. The most common “atypical” (i.e., non-PSP-RS) phenotype of PSP is PSP-Parkinsonism (PSP-P). Relative to PSP-RS, it features more asymmetry, generalized bradykinesia, tremor, and levodopa responsiveness, and only later displays falls and cognitive loss. It is usually initially misdiagnosed as Parkinson’s disease. It has perhaps the slowest course among the PSP phenotypes, averaging about 9 years’ survival from symptom onset. This compares with about 6 years for PSP-RS and intermediate figures for the other phenotypes. In fact, the PSP-subcortical group as a whole has a similarly longer average survival duration than the PSP-cortical group as a whole or the PSP-RS + PSP-cortical groups.
PSP-progressive gait freezing. After PSP-P, the most common atypical phenotype is PSP-progressive gait freezing (PSP-PGF). In fact, most patients exhibiting only progressive gait freezing will eventually develop diagnostic features of PSP. The central feature of PSP-PGF is loss of ability to continue ongoing gait, especially after a pause, during a turn, or at a doorway threshold. In advanced cases, the patient cannot initiate gait at all. The picture also includes rapid, small handwriting and rapid, soft speech as frequent or severe features. The anatomic location of the pathology in such cases differs from that of PSP-RS in showing less involvement of the base of the pons (part of the brainstem) and of the dentate nuclei (part of the cerebellum).
PSP-speech/language. This is a composite category. In PSP-nonfluent/agrammatic variant of primary progressive aphasia (PSP-nfaPPA) speech is halting, with poor grammar, syntax, and pronunciation, but with normal comprehension and naming. A mirror-image variant called semantic-variant primary progressive aphasia (svPPA) features difficulty in naming with reduced vocabulary but with normal grammar and syntax. Together, PSP-svPPA and PSP-nfaPPA are referred to as PSP-speech/language disorder (PSP-SL).
PSP-corticobasal syndrome. CBS as a clinical syndrome (meaning a group of signs and symptoms that occur together, although the underlying disease may differ across patients) comprises highly asymmetric rigidity, slowed movement, and apraxia (loss of skilled movement), often with equally asymmetric dystonia (fixed postures), pyramidal findings (weakness and abnormal reflexes), myoclonus (small, rapid, irregular movements), and cortical sensory signs such as astereognosis (inability to identify objects by feeling them) and agraphesthesia (inability to identify figures traced on the skin). Aphasia (difficulty processing language) and other abnormalities localized to specific brain areas may also occur. Dysarthria (difficulty with pronunciation) can be prominent but gaze palsy, postural instability, and cognitive loss tend to be later and milder than in PSP-RS.
PSP-frontal. More formally called PSP-behavioral variant frontotemporal dementia (PSP-bvFTD or simply PSP-frontal), this phenotype features disinhibition, irritability, apathy, and loss of empathy for others, along with impairment in frontal “executive” functions such as ability to maintain attention, to follow instructions, to shift tasks on command, and to inhibit an ongoing action when appropriate. This is the core of the cognitive and behavioral deficits in PSP-RS, but when it appears first and remains worst, the term PSP-F is appropriate.
PSP-ocular motor and PSP-postural instability. Perhaps unsurprisingly given the cardinal features of PSP-RS, PSP can also take the form of a relatively pure ocular motor picture or a relatively pure picture of severe postural instability with falls and little else to suggest PSP. However, reported cases are very sparse to date. These have been designated PSP-ocular motor (PSP-OM) and PSP-postural instability (PSP-PI).
PSP-primary lateral sclerosis. The pathology of PSP can also produce the clinical picture of primary lateral sclerosis. PLS is one of the phenotypes of amyotrophic lateral sclerosis (ALS; Lou Gehrig disease) and can be difficult to distinguish from it, especially as ALS can produce frontal cognitive difficulties in many cases. The clinical picture of PSP-PLS is highly asymmetric and resembles that of CBS but with little or no cortical sensory loss (spatial sensation ability), dystonia (fixed postures), or myoclonus (very quick, small, irregular involuntary movements).
PSP-cerebellar. The classic lurching gait of PSP-RS has a cerebellar appearance, the speech of PSP has an ataxic (or drunken-sounding) component in many cases, and the ocular square-wave jerks of PSP occur commonly in cerebellar disease. For decades, these were considered minor and inconsistent features, but in 2009, neurologists in Japan described a PSP phenotype of PSP with involvement of the cerebellum at autopsy and early, prominent ataxia of the trunk and limbs. Although that original report found PSP-C in 14% of all people PSP at a Japanese center, the figure is much lower in Western populations for unclear reasons.
A word about drug trial eligibility. Until the cause of PSP in general is better understood, neuroprotection trials — those aimed at the fundamental brain cell loss rather than merely at ameliorating the symptoms — will continue to recruit only patients with PSP-RS. Why?
We don’t yet know if the non-PSP-RS phenotypes share – either with PSP-PS or with one another — the same molecular abnormality being targeted by the drug.
While the classic PSP pathology underlies close to 100% of PSP-RS, that figure is much lower for some of the other nine phenotypes. That means that non-PSP-RS may well be a non-PSP pathology, and that admitting participants with non-PSP-RS to a drug study runs a risk that some may not have PSP at all. This could obscure any benefit the drug may have unless the trial is prohibitively large.
The main outcome measure for nearly all PSP trials is the PSP Rating Scale, which was designed for, has been validated for, PSP-Richardson’s syndrome alone.
Trials to slow the progression of PSP use the rate of progression as the “outcome variable” of the trial. As noted above, the phenotypes do vary in their expected survival durations, and by inference, their progression rates. Therefore, including phenotypes with different inherent rates of progression would require a larger, longer and more expensive trial. One solution would be to make sure the active drug group and the placebo group receive similar proportions of the various phenotypes (a time-proven technique in trial design called “pseudo-randomization”). But this doesn’t solve the four preceding problems.
As PSP-RS progresses faster than the other nine phenotypes, a trial enrolling only that phenotype can reach a result in a shorter time or with fewer participants. This isn’t only financially advantageous for the trial’s sponsor. If the drug is ineffective or harmful, fewer patients will have been exposed to it, and if it’s effective and safe, it will reach market approval more quickly.
These graphs are from a review of records of 32 people with autopsy-proven PSP in Spain and Germany. The lead author was Dr. Mar Guasp of Hospital Clínic de Barcelona and the senior author was Dr. Yaroslau Compta of the same department. They show the difference in survival from initial symptom to death between the PSP-RS and non-PSP-RS (top) and between [PSP-RS + PSP] and PSP-cortical (bottom). Graphs A and C: For each box, the horizontal line is the median, the upper and lower borders are the 25th and 75th percentiles and the ends of the “whiskers” are the highest and lowest values. Graphs B and D show the same thing in the form of “survival curves” or “Kaplan-Meier curves.” The vertical axis shows the fraction of the original patients still living at the time (post-onset) shown on the horizontal axis. The p values are the likelihood that the difference could have happened by random chance. The asterisk indicates that this likelihood is low enough for the difference between the groups to be considered “statistically significant.” (Statistical veterans: Sorry to belabor this for the benefit of the statistical novices.)
To-do list:
Let’s figure out why the disease spreads through the brain in PSP-RS and the other PSP-cortical phenotypes more quickly than in PSP-subcortical. Efforts to do that have in fact begun and could provide the key to the whole puzzle of PSP.
Let’s agree on a way to enroll people with non-PSP-RS phenotypes into clinical trials. Current efforts to diagnose PSP using tau-based positron emission tomography (PET) and measures of tau in skin, blood or spinal fluid could potentially identify people with PSP other than PSP-RS who could potentially join a trial.
Let’s educate neurologists to identify, or at least to suspect, the non-PSP-RS phenotypes. This would allow them to avoid or delay fruitless diagnostic testing and to provide their patients with useful prognostic information.
PSP Blog 