The following is a transcript of the presentation video, edited for clarity. Click the PDF icon to download the presentation handout and selected references.
So, a little overview, I’m going to be talking mostly about progressive apraxia of speech as it occurs in neurodegenerative disease and sometimes as the primary manifestation of neurodegenerative disease in which case we would call it primary progressive apraxia of speech.
I’m going to talk a little bit about the relationship to primary progressive aphasia and I think we had an excellent introduction to PPA this morning for those who were here this morning. I’ll talk about the demographics of these group of people, clinical features and measures of them — both perceptual and some acoustic measures. I’ll talk about non-speech findings that may go along with the problem, neuroimaging correlates. We’ll talk something about disease evolution, the clinical and neurologic correlates of that. And I’ll say something about the underlying pathology.
I should let you know that the research upon which most of what I’m going to talk about this afternoon is based on a couple of studies with a somewhat broader scope that have been helping us to characterize the speech and language difficulties associated with PPA and PPAOS, have allowed us to characterize the syndrome of PPAOS and its relationship to PPA, and describe the clinical neurological and/or psychologic correlates of those problems and neuroimaging correlates — also, and more recently, to look at the evolution of deficits and progress apraxia of speech and primary progressive apraxia of speech.
There have been some secondary dividends of this research overtime. It’s allowed us to kind of look at the validity and current consensus criteria for primary progressive aphasia which Dr. Hillis reviewed in a really fine way this morning. And it’s allowed us to think a little bit about what some modifications might be for those current criteria. And then the research has also allowed us to develop some simple clinical tools for characterizing and quantifying progressive apraxia of speech and associated deficits. So, I’ll be able to say a little bit this afternoon about this apraxia of speech rating scale that we have developed, a very simple articulatory error score that we’ll refer to in some of the data I’ll present, and some simple temporal acoustic measures. So, they’ve been little secondary dividends of this work that we’ve done.
I want to acknowledge our research team, particularly Keith Josephs who is PI of one of the major grants that we have and a neurologist. He was actually supposed to be here today and have half of this presentation time, but because of illness, he was unable to make it. So, I’m sorry we won’t be able to hear him talk particularly about the neuroimaging and underlying pathology a little more, but I’ll muddle my way through. And I have an excuse for saying I don’t know some things you might ask me.
Jennifer Whitwell in neuroimaging is a PI on one of our projects. My speech pathology colleagues, Edythe Strand and Heather Clark, Mary Machulda, neuropsychology, Sarah Boland is our research coordinator and then a large number of speech pathology fellows and neuropsychology fellows and neurology residents and fellows who have contributed to a number of our studies.
Just some basic definitions to get us started. I think we all know what apraxia of speech is, although we might disagree with its clinical diagnosis, I guess. But, we kind of accept that it’s an impaired capacity to program commands necessary for directing movements for phonetically and prosodically normal speech. And we know that it can occur in the absence of weakness, slowness, or in coordination when used for reflex or automatic acts that basically says it’s not dysarthria, it’s a apraxia. And it can occur in the absence of any language problem. Even though, it often does occur with both of those things.
Progressive apraxia of speech, PAOS is apraxia of speech of insidious onset and gradual progression, sometimes without aphasia for a substantial period of time due to a neurodegenerative condition.
And primary progressive apraxia of speech is progressive apraxia of speech in which apraxia of speech is the first, the only, or the most salient feature of a neurodegenerative disease in which criteria are not met for the diagnosis of another neurodegenerative disease.
So, we know that some people with corticobasal syndrome for example, have apraxia of speech. They would not be said to have primary progressive apraxia of speech because the apraxia of speech is just part of a broader constellation of symptoms. Primary progressive apraxia of speech is when that is what you have and criteria are not met for another specific neurologic diagnosis. You may evolve to meet criteria for some of those other conditions, in which case we might agree that you would no longer use the term PPAOS. It’s become something else.
So, I’d like to start with an illustrative case. I’ll sort of set the stage. This is a woman who I met 25 or so years ago. And she taught me and my colleagues a great deal about this condition. She was 71 years old, a retired social worker. She had a two and a half year history of occasional slurring or mispronouncing of words. She was active socially, basically leading a very normal and vigorous life.
When she was seen for a neurology evaluation, her clinical examination was normal with the exception of her speech problem and some slowness of eye movements. EEG, MRI, lab tests were all unremarkable. She was seen by neuropsychology. She had normal memory and learning. And basically was average to above average on everything that they asked her to do.
For speech and language assessment, her language was normal in all modalities except in verbal expression, maybe a rare self-corrected semantic error and very infrequent self-corrected function word errors or omissions in her conversation or sentence repetition. And the neurologist and I both felt that the appropriate diagnosis was apraxia of speech greater than equivocal evidence of aphasia. And she did have a nonverbal oral apraxia as well. So that was the diagnosis at that time.
So, at three years post onset, she had a diary that she kept over years, and I have over a 100 pages of her diary that she kept through periods of her illness. She sent me a card:
It’s just a card to thank you for the suggestion that I contact Virgil’s (which was a service station). If my car problem was more than starting in this frigid weather, Virgil’s was a veritable beehive with all mechanics seemingly in good spirits as they tended to tasks.
>> Repeat these words. Stethoscope.
>> Methodist Episcopal Church.
>> Methodist Episcopal Church.
>> Statistical Analysis.
>> Tastistical Analysis.
>> OK. And Massachusetts.
>> All right, describe what’s happening in that picture form.
[This is the cookie-theft picture.]
>> Because the children are stealing the cookies jar– from that cookie jar and the uh mother, I think, is um washing the dishes while the border flows over the sink. And uh she’s obliverous to children stealing cookies.
>> Very good.
>> OK. So I think you’ll agree that she’s got some features of what we would agree are apraxia of speech. She has some distortions, are somewhat inconsistent, some occasional distorted substitutions. She occasionally will tiptoe through a multisyllabic word. Her language is really very good, so a mild apraxia of speech.
At four to five years post onset, I got a note and she said in it,
No matter how I handle the silverware, it seems to have a mind of its own. Sometimes the food does not get to my mouth.
And she was developing a limb apraxia at that time. But language is still pretty good right there.
At six years post onset, this is what she was like. And I will tell you about her– the rest of her workup after you take a look at this.
>> Now, cough.
>> Cough. Cough.
>> All right. And now, blow.
>> Blo. Blo.
>> Watch me.
>> All right, pop, pop, pop, pop, pop, pop, pop, pop, pop, pop, pop.
>> Pop, p, p, p, p, p, p, p, p, p, p.
>> And now, ta, ta, ta, ta, ta, ta, ta, ta.
>> dah, dah, dah, dah.
>> These words: Mom.
>> Bo bah.
>> Bay beee.
>> Bye bob.
>> Eee. Aye.
>> Good. Say animal.
>> Spesegese [laughs].
OK. So she now has severe motor speech problem that we labeled apraxia of speech and a nonverbal oral apraxia, it’s very, very apparent as well. And maybe some other problems in addition, but we can talk about that during the question and answer if you like.
The astounding thing though that– at that point in time, she was still reading biographies and novels for pleasure. Her writing was her primary means of communication. Her writing was slow and laborious and distorted motorically and there were some grammatical deficits present. So examples are she said,
This is write, I’ll moving to LA in late September… and I’m still going autopsy with Mayo… I’ll still send the copies of my diary to you.
PPAOS: An Evolving Concept
PPAOS is an evolving concept. Back in the 1970s and 80s, if there’s anybody here as old as I am, we were sort of taught that aphasia and apraxia of speech maybe don’t occur in degenerative neurologic disease that if you see trouble with language or see trouble with speech, it’s just part of the bigger picture of impairment that you’re seeing. So in this chapter by Wertz, a very influential chapter back then, he’s got question marks under aphasia and apraxia of speech under the ideology of neurodegenerative disease basically.
So historically, this is a problem that’s been largely ignored, but as I’ll mention in a few minutes, it’s probably been hiding somewhere.
In 1982, when Mesulam presented the first cases of slowly progressive aphasia — modern cases of slowly progressive aphasia. One of his five or six patients had running speech that was described as labored, diminished in quantity, and dysarthric. And this individual had profound limb and buccal facial apraxia. So I think it’s conceivable that at least one of his initial patients had apraxia of speech based on that definition. Of course, we don’t know, but there’s some flavor of that in that early description.
And around that time, I joined Mayo and I started seeing people with primary progressive aphasia. And some of them had apraxia of speech. And I started to see some people who seemed only to have apraxia of speech. So I began just throwing those names into a little folder that I had kept of interesting cases. And after about 20 years, I had 80 cases that I was able to describe who had neurodegenerative apraxia of speech. And these were patients whose apraxia of speech was never less than any– less severe than any aphasia that might have been present.
Speech and language complaints for the initial symptoms, the only initial symptoms in 80% of these people, and 56% speech and language was the only symptom present when they were first evaluated for the problem by us. And that usually occur about two and a half years post onset. Nine percent of them had apraxia of speech only, 31% had apraxia and aphasia, 26% had dysarthria, 34% had apraxia of speech and dysarthria and aphasia. And in those who had apraxia of speech and aphasia, the apraxia of speech was worse than the aphasia in 78%. And the apraxia of speech was worse than the dysarthria in 71% of those who also had some dysarthria. So apraxia, the dominant, the dominant problem in that group of people.
In terms of their neurologic workup, most of these 20 patients — and this is a long time ago, so even diagnostic categorization of neurologic diseases changed since then. But in those patients who got a fairly specific neurologic diagnosis, it most often was with conditions that are associated with prominent motor, rather than cognitive deficits. So, corticobasal degeneration, progressive supranuclear palsy syndrome, occasionally ALS were the kinds of more specific diagnosis that some of the patients would get, although I will say sometimes based only on their apraxia of speech.
Where has it been hiding?
So in general, where’s this problem been hiding? And as Dr. Hillis reviewed this morning when she talked about the criteria for categorizing PPA, it’s likely that this apraxia of speech syndrome is hiding within one or more of these categories. And this is the 2011 consensus criteria paper that laid out the criteria for categorizing the PPA variance. And I’m going to go through these fairly quickly because Argye did that this morning.
So in the semantic variant of PPA, you’ve got to have impaired confrontation naming and impaired single word comprehension and basically some loss of word meaning. And then you may have three of four other features. You should have at least three of these four other features. And one of them is paired speech production. So, that basically said you shouldn’t have any motor speech problem if you have semantic variant PPA.
And we reviewed a fairly recently a large number of cases. And in fact, motor speech is typically paired in cases in the literature with semantic variant PPA perhaps with rare exceptions. And it seems to be holding up well enough. So, you could argue that if apraxia of speech or dysarthria is present in someone with a diagnosis of semantic PPA, you probably are to think– you probably ought to question the diagnosis of semantic variant or maybe the person has more than one condition. So it really should be very rare. In the logopenic variant, you have to have impaired single word retrieval in spontaneous speech and naming, impaired repetition of sentences and phrases. So you’d have to have both of those things, and then you have to have at least three of four additional features. And one of them is spared motor speech. So this is basically saying we don’t typically expect motor speech impairment and logopenic PPA, but it permits its presence.
And in fact when you review cases in the literature, you find that dysarthria is very uncommon, apraxia of speech has been noted in some patients, but when it has been, it’s usually subtle or very mild. So if apraxia of speech is prominent in someone with a diagnosis of logopenic PPA, one should probably question that classification or wonder if there’s another condition also present.
So where has it been hiding? Here is where it’s been hiding. If you look at the agrammatic variant of PPA, you have to have one of two features: agrammatic difficulties in language production, or effortful or halting speech or apraxia of speech. You have to have one of those two. Now that means many patients have both, but you have to have one of those two. And then you have to have two of three other features. One is impaired comprehension of syntactically complex sentences, spared single word comprehension and spared object knowledge. So spared single word comprehension and spared object knowledge basically mean you don’t have the semantic variant. And impaired comprehension of syntactically complex sentences means you don’t have — or impaired comprehension — it means that you are meeting another criteria that’s not present in everybody with, but not in fluent variant.
So what this means if you look at it carefully and interpret it literally is that if one of your core features and you only have to have one as apraxia of speech and you then meet two of the other three criteria. And if those are spared single word comprehension and spared object knowledge. And what it means is that you can have apraxia of speech, and no aphasia, and get a diagnosis of agrammatic PPA. So it’s one of the shortcomings of the way the consensus criteria are laid out right now. So, we would argue that if you have apraxia of speech and you don’t have aphasia then you probably shouldn’t be sad to have primary progressive aphasia because you’re not aphasic.
The fact is that apraxia of speech is probably present in a majority to a substantial majority of those who get the diagnosis of agrammatic PPA. And the median prevalence across a number of studies of apraxia of speech in agrammatic PPA is 78% and the ranges 17 to 100%. And in case series or small case series, its frequency and its severity in individual cases may exceed that of the aphasia when they’re both present. And if you kind to look at the literature as a whole, it may be that apraxia of speech may be present in 20% of all higher level progressive speech in language disturbances, so probably not as infrequent as the literature would lead us to suspect.
Our data kind of backed that up. So we have tested 169 different people with PPA or PPAOS. In that big cohort, 23% had agrammatic PPA and of them 85% had apraxia of speech. So 85% of our patients with agrammatic aphasia have apraxia of speech. Thirty eight percent of our whole cohort had apraxia of speech and 22% of them were also aphasic, and when that was the case, the apraxia of speech was judged as worse than the aphasia in 65% of them. And 18% of our entire cohort, or 31 people, had primary progressive apraxia of speech meaning they did not have aphasia did not meet criteria for any other neurologic diagnosis. Twenty nine percent of them with PPAOS did have dysarthria that was judged to be less severe than the apraxia of speech.
So, progressive apraxia of speech and PPAOS are probably not as uncommon as the literature recognizes or implies.
What are the basic demographics of this group? Age of onset is usually in the late 60s to early 70s, the range can be from 40s to 80s. It’s pretty uncommon before the age of 65, but I’m going to show you a couple of samples of patients who were well under 65 when this disorder begun. As a group though, they tend to be older than people with agrammatic PPA whether they have apraxia of speech or not, and in some studies as much as a decade. So, we’re not quite sure what that means, but it’s an interesting finding and maybe related to underlying pathology or variables that we don’t really understand.
Gender maybe more frequent in women and in men in case theories. It runs from 50 to 70% female to male. Nothing surprising that had and this basically matches that of the general population, 90% right handed. Fairly well-educated group of people about 15 years with a wide range. And they usually present– had presented to us for initial evaluation between two and three years post onset. So this is presentation to a tertiary care outpatient setting. It doesn’t mean this is the first time people have sought a diagnosis. But they are seeking a diagnosis or a second opinion about a diagnosis. Many people we’ve seen have simply not had a diagnosis.
In terms of the speech features associated with PPAOS, if you can read the slide, the usual suspects are there. So, you know, the first four or five characteristics slow over all speech rate lengthened in through segment durations, increase sound distortions or distorted sound substitutions, increasing with increasing length syllable segmentation, sound distortion and then on down the list with features that have been noted in many studies of apraxia of speech due to stroke.
With the exception of the last one there which is the least frequently occurring, but I point it out because it’s not a feature that I’ve seen mentioned in any study of stroke related apraxia of speech, and that is reduced words per breath group relative to maximum vowel duration. So these are individuals who may be able to prolong a vowel for 10 or more seconds, but not produce more than one to three syllables per breath group when speaking. And it seems that what’s going on is that they can only program one or two syllables at a time and that that’s yoked to their programming of breathing for speech. We don’t know that, but that’s a reasonable hypothesis. And something that at least all the other features here you see in stroke-related AOS, but not that one that I’m aware of anyway.
We have wondered over time if there may not be subtypes of progressive apraxia of speech. And anybody who studied apraxia of speech knows that for many years, people have wondered if there aren’t subtypes of apraxia of speech, although I think at least in past years people might have been arguing about the difference between what we now call apraxia of speech and would now call phonological errors. I’m not sure that all of those discussions about different types of apraxia of speech are really about different types of apraxia of speech, it may have been apraxia of speech versus phonological problems that were once considered to be apraxia of speech.
But, we’ve wondered in our cohort if maybe there aren’t two types or three types. Type 1 is — we think is or what were basing this on is that these are patients who seemed to have a predominance of articulatory or segmental difficulties. So they’ve got distortions and distorted substitutions. They repeat sounds. They attempt to self correct. They grope for articulatory postures. So, this is very much related to articulatory accuracy. This subtype seems to be more frequent when aphasia is also present and it’s usually agrammatic aphasia and when the aphasia is more severe than the apraxia of speech. That’s the tendency.
Type 2 is predominated by prosodic abnormalities. So this is a predominance of slow rate and segmentation within or between words or syllables within multisyllabic words. And this seems to be more associated with PPAOS, that is no aphasia or if aphasia is present, the apraxia of speech is more prominent than the aphasia is.
I have to say that that so far our– putting people into these two types is purely clinical judgment. So, do you think that the articulatory problem predominates over the prosodic problem or vice versa?
Or is it Type 3, which is they’re both there and one doesn’t seem to be more prominent than the other. I would suspect that in stroke related AOS, most people are Type 3 if you’re using this categorization. I don’t know that. I don’t think anybody has looked at that, but it be– my gut guess, anyway. But, these are very subjective judgments that are made. And I’ll get to some more objective things data in a little bit.
Let me show you a few more samples of patients just to go over the lay of the land here.
This man is a 64 years old and he is three years post onset of his difficulties and he is aphasic. He would meet criteria for the agrammatic variant of PPA. But he also has apraxia of speech that we judged to be about equivalent to the aphasia in severity. And I don’t know if this sample would allow you to make that judgment but that’s what we concluded.
>> Have you been here before?
>> I went to some Mayo Clinic before.
>> Tell me what’s happening in this picture and–
>> Drink wine. Listening to music. Tom is flying a kite. John and Mary are sailing the boat.
>> Sixty two and a half.
>> Sitty two and a half.
>> The pastry cook was satisfied.
>> The pastry cook was satisfied.
>> The telephone is ringing.
>> The telephone is ringing.
>> Delicious freshly baked bread.
>> Des dislicious freshly baked bread.
>> Artillery, artillery, artilley.
>> Stethoscope, stethoscope, stethoscope.
>> tethes-scope, tethescope, tethescope.
>> OK. Rhinoceros.
>> Rhinoceros, rhinoceros, rhinoceros.
>> A couple of more. Statistics.
>> Testi– De– Destictics. Dest-stictics. Desticsics.
>> The municipal judge sentenced the criminal.
>> The minicipal judge senced the criminals.
>> Tell me about the speech problem. What is this like? Can you describe it for me?
>> It’s frustrating. I speak in short senteces. Not long sentences. I can’t–
>> Why is that?
>> I’m going to first [inaudible]– I can’t create the words.
So, I think his grammatical difficulties are pretty evident, but he’s got some sound level difficulties going on as well. His score on the WAB was close to 90, so his language difficulty fairly mild. OK. But that’s not primary progressive apraxia of speech, that’s apraxia of speech with the agrammatic variant of PPA. So when we say PPAOS, he is not what we’re talking about.
This woman, 60 years old, with a nine year history of her speech difficulty. Her initial diagnosis was conversion disorder. We’re just starting to do a little case history review of all of the diagnosis that patients come to us with. And conversion disorder is on the list, and way more than one person. So, she’s been unable to work for four years. Her language exam is clean as a whistle. So there’s no evidence of aphasia.
>> Say, animal.
>> Perspic– Boys — Poyspiction.
>> Meemis– Meemis-ipal
[ Inaudible Remark ]
Her language skills, to the degree that she is intelligible, are good, and she’s having substantial difficulty with motor planning and programming based on her distortions, distorted substitutions, her attempts at self correction and so on. We– I suppose could argue about whether all of her errors are apraxic and whether some might not be phonologic, but there’s no other evidence of aphasia.
She is now 14 years post onset, and is about mute at this point. Still communicating through writing adequately, although agrammatically and with evidence of aphasia. She may have some dysarthria, but that’s a little hard to know. And her neurologic examination beyond her apraxia of speech and aphasia is normal.
This is a 52 year old man with a five year history of his speech problem. No evidence of aphasia at this time.
>> The boys getting cookies from the cookie jar–
[ Inaudible Remark ]
>> Say, we saw several wild animals.
>> We saw several wil– wild amals.
>> My physician wrote out a prescription.
>> My presi– phys sh– My physison wrote out a prescrition.
>> The municipal judge sentenced the criminal.
>> The mes– menicipal judge ses– sentenced the crimal.
>> OK. And now, some words. Animal.
>> Mesi– Menispal.
>> Try that again.
So, he has many features of apraxia of speech. We would have rated him– we did rate him Type 1, articulatory predominant. But keep in mind that when we’re talking about a predominance of articulatory problems versus rate and prosodic, not presence versus absence. So, it’s a rare patient who has only articulatory and no prosodic problem and vice versa. And that’s the point — that when we say this is articulatory predominant, we’re not saying you don’t have any prosodic problems, we’re saying the articulatory problems predominate over that.
This woman, 79 years old, is four years post onset, no evidence of aphasia. And she is what we would call a Type 2. And she is among the two or three patients who are the best examples because it’s really dominant. But that is a relative absence of the prominent articulatory problems that we saw on the– in the last two cases.
>> What’s happening in that picture? Talk in sentences.
>> The people are enjoying the lake. A couple is sitting on a blanket listening to the radio. He doesn’t have his shoes on. The boy is flying his kite, and enjoying it he’s smiling.
>> Are you speaking slowly because you are unable to speak more rapidly or because you feel you have to go slowly to prevent things from going wrong?
>> I try to prevent things from going wrong.
>> And if you try to speak very rapidly, do you think it’s going wrong?
>> Yes. They do.
>> OK. Let me have you just repeat another sentence, and I want you to say it as fast as you can. We saw several wild animals.
>> We saw, several, wild, animals.
>> OK. Did anything go wrong there?
>> Yes. Three times.
>> What three things went wrong?
>> Animal, wild, several.
>> Mm-hmm. Mm-hmm. And what was–
>> I couldn’t get it out without pausing.
So, that’s sort of the clinical picture. What are other accompanying deficits that might be present initially? The main thing is that there are maybe no other language or cognitive deficits early in the course. And in our research cohort, among 65 people who had apraxia of speech, 48% had PPAOS that is no on ambiguous evidence of aphasia. And 34% had no evidence of aphasia or dysarthria.
About 50% had evidence of a nonverbal oral apraxia. And that proportion increases with disease progression. So, as time goes on, a higher percentage of patients will acquire nonverbal oral apraxia. So far, it looks like nonverbal oral apraxia is more frequent when aphasia is also present. So when aphasia is also present, nonverbal oral apraxia is present about 80 to 90% of the cases. So it’s not that high in PPAOS. So I think it’s another piece of evidence that those problems can be dissociated — not that they may not be related in other ways.
One third of our patients with PPAOS have dysarthria. And that frequency increases with disease progression. The dysarthria type is most often spastic more frequent than hypokinetic or a mix of the two.
And dysphasia is usually not evident unless dysarthria is also present.
In terms of neuropsychological findings, consistent neuropsychological or neurobehavioral deficits are usually absent early on. Cognitive deficits that you see in other — I shouldn’t say other, I should say in PPA variants, are typically not evident. So, these patients do not have memory loss or poor calculations, trouble with facial recognition, visual spatial or visual perceptual deficits that you may see in PPA variants.
There is evidence of executive dysfunction with longer disease duration. And you see that in people with agrammatic PPA as well.
And then later in the disease, other cognitive behavioral and motor signs may become evident, essentially will become evident, as you would expect in almost any neurodegenerative disease.
Development of Useful Tools
Let me just step back a little bit. I said, one of the– one of the little side effects of doing this research is you are almost forced to develop some tools to help you out a little bit. And one of the tools that we originally put together is what we call the apraxia speech rating scale or the ASRS. And we’ve really developed it for the research protocol. We wanted to come up with some quantitative index of the prevalence and severity of apraxia speech beyond a simple mild, moderate, marked or severe rating. So we wanted to try to quantitate this problem, as well as describe it well.
So, we put together the 16 point scale, the ASRS. And the 16 items on the scale are features that are known to be associated with apraxia of speech. So distorted sound substitutions, distorted sound additions increased distortions with increasing length and complexity and so on. Some of these features are pretty unique to apraxia of speech that is you wouldn’t expect them in dysarthria or aphasia and then other features are overlapping, that is you can see them in apraxia speech, but they are not distinguishing features relative to aphasia and dysarthria.
So there’s these features. And we rate them, each of those 16 features on a five point scale. Zero is the feature is not present. One is detectible but infrequent. Two is frequent, but not pervasive. Three is nearly always evident but not severe. Four is nearly always evident and severe. So those 16 features get rated on that four point scale.
I should say, too, that we tried to keep all of these simple. So, our ratings are based on what patients do when they’re describing a pictured scene or engage in some general conversation. And then their performance while repeating 13 words, repeated three times, and repeating three sentences. So that’s what we based the rating on.
So, we didn’t devise items to look at this, it was part of what we would routinely do to evaluate apraxia of speech of aphasia.
And this– we’ve given it to more than this number, but 56 patients with apraxia of speech, some of who had PPAOS and some who were also aphasic; 77 patients with some variant of PPA, but no apraxia of speech. And interjudge reliability is good, and intrajudge reliability good. That’s among us. We still have work to do on this scale. And one of them is we need to improve reliability so that reliability would be good among a large number of clinicians, not just the people doing the research. But, we’re able to establish reliability.
And looking at the ASRS, among these patients we’ve given a test to, and comparing those who were clinically judged not to have apraxia of speech and clinically judged to have apraxia of speech, the ASRS is 96% sensitive and 100% specific using a cutoff score of 8. And the ASRS, the lower the score, the more normal you are. So, zero is best, 64 theoretically is the worst. We’ve not seen anybody who gets a score worse than in the low 40s on the scale. I think because once you get to that point, you’re having so much trouble with speech, you almost can’t make the ratings anymore because they’re not able to do things that let you judge certain characteristics. So, pretty good sensitivity and specificity relative to a clinical judgment of apraxia of speech versus not.
And the correlation between the ASRS score and independent clinical judgments about severity, zero to four scale, not present, mild, moderate, marked and severe, the correlation is 0.88. So there’s some concurrent validity for that as well.
This has been very useful in our research. We’re thinking now as we’ve developed this, that this maybe has potential to become a more widely applied clinical tool. We have not looked at this in the stroke population, so we don’t know if it applies well to that. And that’s an empiric question that needs to be answered. But at least, it’s a measure that gives you some quantitative index of the prevalence and severity of features of the disorder.
And then another very simple measure that we were beginning to use is what we call the articulatory error score. And this is simply the percentage of 56 words in which one or more segmental or articulatory errors occur. And they are defined as distorted or undistorted sound substitutions, additions or repetitions, sound omissions, sound prolongations (beyond the slow rate in general), false starts, successful or unsuccessful attempts to correct sound errors. Note that this could capture phonological errors as well. It’s not– you don’t have to decide whether it’s apraxia or not. It’s just– if you make one of these kind of error on this task, you get dinged.
And the tasks are three consecutive repetitions of 13 words, and repetition of three sentences repeated once. And that gives you the measure. And I’ll come back to this in a few minutes and talk about some data that suggest that may be useful in some ways. It could be useful as, you know, index of severity although it’s based only on repetition. So, one might question that.
Acoustic correlates. I want to recognize Holly Hanley acoustic analysis for us. And I’m going to talk mostly this afternoon just about some very simple temporal measures focusing mostly on rate or syllable per second.
So, the data here– Let me just orient you. So, this is a repetition of cat. And there are three groups. There’s 11 control subjects normal speakers, six patients with agrammatic PPA with no evidence of apraxia of speech. So they are aphasic, but they are not apraxic. And then 21 patients with primary progressive apraxia of speech.
So, the comparison across the three groups here, there’s no difference between them on the single syllable word, which theoretically you probably– you’d accept easily because the programming demands are pretty simple. So, we’re not detecting any difference here and I believe that’s aphasic or apraxic.
The triangles here are the means and vertical– the vertical line is the range not the standard deviation. So this is a pretty strict comparison here. And I will tell you that the comparisons between the controls and the agrammatic aphasic patients — there’s no difference between them on any of the measures I’ll show you now.
But on all of the other measures I’ll show you now, the PPAOS falls out. So you go to the word catnip, and that’s statistically significant — you’ve added some programming complexity. You go to the word catapult, you find differences. No difference between the agrammatics, but the apraxics are different. And you go to the word catastrophe and you’re finding probably a somewhat larger difference because complexity has increased further. So, once you go beyond the single syllable word level, a simple temporal measure of duration, word duration is sensitive to this problem.
The same holds true for sentences. So, on some sentence repetition tests, looking at syllable rate per second the PPAOS group distinguishes itself. And, we’ve also just created a composite word sentence syllable per second rate just as a composite measure over all these words. Basically, we’ve looked at five different words and two different sentences. And the five words are– it’s an average over three repetitions of the five words. For the sentences, it’s just one repetition of the sentence. So, the simple temporal measure seemed to be pretty sensitive.
We’ve also looked at the pairwise variability index. How many people are familiar with the PVI? OK, not too many. The PVI is basically– it’s way to measure stress on syllables. So, the word catastrophe, the first syllable the ca is unstressed, second syllable ta is stressed. And you would– And if you measure the vowel in each of those syllables which is what we did here, in normal speakers there should be a big difference in the duration between those two vowels. They’re two vowels, which is I supposed a problem, but one syllable stress the other is not. So, in normals because of the way you look at the ratio basically between the first and second syllables, so there should be a big difference among normals if you’re truly producing an unstressed and then a stressed syllable. So, the more abnormal score here is toward zero which means no difference between the first and second syllable. And, it looks here like there’s no difference. But the PPAOS group is statistically different than the agrammatic or the control groups. So that’s a significant difference. There’s some overlap.
People have looked at the PVI in the stroke population with apraxia of speech and it looks like a pretty good sensitive measure. We will look at the data carefully enough yet to know if it’s better than any of the simple word duration measures that we’ve used. So those are some pretty simple easy to measure acoustic things that can characterize the problem. And perhaps even serve as a measure of severity or a change overtime.
PPAOS Subtypes: Quantifying Differences
I’m a little bit confused about the people who are Type 1 who seem to have very low articulatory error scores. Is that just that they’re very mildly Type 1?
It was interesting when you asked the woman to purposely speak, the Type 2 case you showed us when she purposely sped up her rate, she started producing more articulatory errors, and I wonder if you did some sort of challenge task of “speak faster” if this would– if they would get more errors and how that would play out.
Neuroimaging Correlates and Underlying Pathology
Neuroimaging correlates and underlying pathology. We looked at in one study 12 people with PPAOS, they had no aphasia. Two of them did have a spastic dysarthria that was less severe than the apraxia of speech. And imaging wise they underwent voxel-based morphometry, diffusion tensor imaging analysis, FDG positron emission tomography. And the primary neuroanatomic correlates looking at gray matter is the brunt of the pathology is in the superior lateral premotor cortex and supplementary motor area.
In white matter analyses, basically overlaps with the gray matter findings but somewhat more extensive. So you add to that inferior premotor cortex body of the corpus callosum, superior longitudinal fasciculus, especially the premotor components of that. The primary composite is superior lateral premotor cortex and supplementary motor area.
And looking at the images in terms of gray matter loss, here you have superior premotor area and supplementary motor area, you look at white matter loss patterns and it overlaps with gray matter loss, but is a bit more widespread. Note that the composite images here show bilateral involvement which is pretty true for all the group analyses that we’ve done. And if you look at hypometabolism the PET results, it looks pretty much like the gray matter loss pattern. So as a group, fairly focal and different than what we learned about where PPA is localized. Also a little different than where we learned stroke induced apraxia of speech is localized.
I think it’s important to keep in mind that when you see group results, it doesn’t look like the individual results that you’re seeing a composite. So these are the data for all 12 subjects. So, you’ve got right lateral, left lateral, right medial, left medial superior view for each of the subjects. And if you look at the individual data, this one I would say that looks pretty much like the composite — kind of superior premotor supplementary motor area some bilateral involvement. But you look at this subject, and that is remarkably normal. And one thing we’ve learned is that — don’t go right to the imaging for the diagnosis. These people have met criteria for primary progressive apraxia of speech and imaging basically normal, basically normal. And that’s true particularly early on in at least a subset of patients.
And then other patients have more widespread impairment than the composite demonstrate. So, don’t think when you see a composite view that it’s based on homogeneous findings. There’s quite a bit of variability.
If you compare the findings for PPAOS with what we– a group of patients we called dominant apraxia of speech. These are patients with the apraxia of speech and aphasia, but the apraxia of speech dominates. And if you look at those findings, they look pretty much like the PPAOS profile with maybe a little more widespread involvement moving sort of south toward Broca’s area. But the basic pattern looks pretty much the same as opposed to a group we called agrammatic PPA, all of whom had apraxia of speech, but the aphrasia worse than the apraxia of speech. And then you see a profile sort of includes the PPAOS, but is much more widespread frontally as well as temporal parietally, so these groups sort of separate out neuroimaging-wise.
So, you know, what we concluded from this preliminary study is that dominant apraxia of speech neuroimaging-wise looks more like PPAOS than it looks like agrammatic PPA.
These are just a few sort of individual data images that fit the predicted pattern. So, here’s someone with PPAOS and that’s right about where the pathology belongs as opposed to the logopenic variant which is temporal parietal, as opposed to the semantic variant which is interior temporal, as opposed to the nonfluent agrammatic variety which is frontal, but a little more than frontal. And you put that all together and then you sort of get a composite view where the blue here seems to capture logopenic, the orange semantic, the green agrammatic, and the yellow PPAOS. So, imaging wise, PPAOS does separate out from PPA, just like it does clinically.
Progression of PPAOS
Disease Course Neuroimaging
Let me just summarize. Progressive apraxia speech exists. It’s probably not as rare as the literature implies. One of the reasons maybe that it’s buried within classifications of primary progressive aphasia.
Progressive apraxia of speech can be the only or the primary or the most salient feature of neurodegenerative disease, in which case we call it primary progressive apraxia of speech. It can be associated and often is associated with dysarthria, spastic or hypokinetic. Many have nonverbal oral apraxia. So when you say PPAOS, those other things are permissible accompaniments, as far as we’re concerned.
And when it’s primary, it should not be subsumed under classifications of PPA because apraxia of speech is not aphasia.
It reflects left or left greater than right hemisphere abnormalities, typically frontal lobes superior and mid premotor cortex and supplementary motor area. It tends to eventually be associated with conditions with prominent motor rather than cognitive deficits meaning, progresses supranuclear palsy syndrome or corticobasal syndrome.
And it tends to predict pathology consistent with taupathy and findings compatible with progressive supranuclear palsy and corticobasal degeneration.
There might be subtypes of PPAOS. We speculate that right now and how some evidence to support it. The presence prominence, severity and progression over time can be measured. And it can be measured perceptually, and it can be measured acoustically. And, those measures can– it looks like– can be simple. Might we learn more if we use more complex measures? Sure we will. But that wasn’t what we set out to do. One of our goals was really to try to keep measures as simple as possible for clinical application. But that doesn’t mean we may not need to go to more complex things to understand things a little bit better.
I think these findings have implications for our understanding of apraxia of speech in general, meaning regardless of etiology, but while apraxia of speech associated with neurodegenerative disease, and apraxia of speech associated with stroke or other acute onset disorders almost by definition have to share a lot of characteristics, the features may not be entirely identical.
And I guess, my thinking right now is that if one were to do a study of apraxia of speech, it would probably not be a good idea to lump stroke induced apraxia of speech with progressive apraxia of speech without parsing them apart and looking at the possible differences. So– But, it’s all apraxia of speech presumably. OK, questions, comments, discussion?
Questions and Discussion
I’m Julie Dalmasso from Western Michigan University. I have a question about what evidence we have so far if there’s any– Let me rephrase this. Is there any evidence to support the intervention could help to maintain with PPAOS like we’ve heard that it can with the PPA variants?
So those of us who have been around for a while, started down our career without cases of primary progressive aphasia — as you all point out Mesulam, his 1982 paper was I guess the first to really formalize this diagnosis. I can remember the late ’70s, I was brought into Lahey Clinic on a case of a 44 year old waitress form the Midwest who had come to the Lahey. And she had a progressive aphasia which of course they thought to have a conversion reaction and all that, but the neurologist there I worked with was also a behavioral neurologist, so he brought me in to look at her aphasia. However, if you look back in our literature, nobody is even talking about anything that resembled this primary progressive stuff. Although, Argye Hillis said this morning, that she thinks that Broca’s Tan tan had primary progressive aphasia. I said, “Why didn’t you say that because boy, that’s an interesting topic?”
But, you are in a unique position at the Mayo. If you look back at Darley’s files and Harrison’s files, would you find cases that now you would label as having, you know, either primary progressive apraxia of speech or primary progressive aphasia?
I would. I guess I would. But, I don’t know because I haven’t done it. And the issue would be what neurologic diagnosis did they get? So, you know, I don’t know. It would be very interesting.
I certainly know that in, you know my first 20 years at Mayo, I didn’t just see 80 people with progressive apraxia of speech. I saw 80 people who had progressive apraxia of speech that was worse than anything else they had. So I mean, I’ve kept the file. I’m up to about 350 now. But, that’s not PPAOS, that’s progressive apraxia of speech.
Because one of the things people ask me: Is this some new thing? And are there more and more cases of it, you know, like autism? What do you think?
Yeah. Well, one little extra comment, I agree, Broca’s Tan tan may have had progressive apraxia of speech. Of course, we can never know because we’re still arguing about whether he was aphasic or apraxic. So we don’t know. But whatever he had was progressive. Now, what’s complicated is he also had a seizure disorder. So, you don’t know if it was the progressive damage, the lesion burn from his seizures, but it was still– it was progressive. And I didn’t know that until two years ago when I reread the paper.
So– But is the incidence of this increasing? You know, when I asked my neurology colleagues, they think probably not. But I think, if it’s not, there’s a whole bunch of explanations. One it’s been buried elsewhere, but it’s now– the grave has been dug up. So people are recognizing it for what it is more readily.
Just the fact that society as a whole recognizes now that there are higher level neurodegenerative diseases beyond Alzheimer’s disease opens the door to kind of subgrouping people a little bit more. I’ve seen a load of people with PPAOS have been given outside diagnose of Alzheimer’s disease. Now, when you look and listen, you know, you’re kind of astounded, but I think for some people, you know, probably neurodegenerative seems high level Alzheimer’s disease, but that’s less true now. I think a lot of these people probably misdiagnosed as dysarthric, many of them probably. And that’s on the list of wrong diagnoses in our files, I know. Just written off as dysarthric. We found, if you look at the AOS population and my study of 80 patients, seven eventually have diagnosis of the ALS.
Wow. Yeah. And one thing I always have to ask nowadays is have you looked at toxin exposure, exposure to toxins in your population because we’re, you know, we have been increasingly toxic environment in this world. And I wonder to what extent some of these things like autism and maybe ABA–
Yeah. While we’ve not– it’s always recorded in the histories, but I don’t think anybody has looked at it systematically relative to PPAOS.
Hello, my name is Martha. I’m from the University of Illinois. I have a question regarding your perceptual scales. The ASRS and AES — given patient awareness of error and frustration, do you have or are you considering adding a patient self report of severity or impact on life?
Nadine Martin, ASHA’s incoming VP for science and research. I have question about the last two case samples that you showed, the videos, the way you have them produce words in isolation and as well as in sentences like “my physician wrote out a prescription.” I could be wrong about this, but the first person I was really surprised that it seemed to me he did better and there would be fewer errors when the words were produced in the context of a sentence. I was thinking about that, and then all of a sudden went to the next one and she was clearly the opposite. She did better when she said them in isolation. I’m just wondering whether you see context affects in some or not others or– Are there two patterns that I’m seeing or is it just one pattern and I’m wrong?
Yeah. I don’t know about the words versus the sentence repetition. I’m not quite sure about that. And, of course, you have issues of control because the words are not the same that you’re looking at. But, there are some differences between conversational and narrative speech and what people do on repetition. And when we do our ratings, we don’t– if you’re only noting difficulties on repetition and it far surpasses difficulty in conversational speech, we don’t allow ourselves to give a rating more than two on that zero to four scale.
Yeah. I haven’t been impressed with the difference between the words and the sentences. But that doesn’t mean it’s not there. It means you haven’t looked at it.
For the woman with the Type 2 variant, the 70 year old, the missionary. Did a motor speech disorder ever crossed your mind with her?
Apraxia of speech.
No. Just no. It actually– No, it didn’t. But you raise an excellent point and relative to apraxia of speech in general. I mean as we’ve refined how we are defining the characteristics of this problem and the work of Mick McNeil really helped us separate out apraxia of speech from aphasic phonological problems. It really helped us do that. But in doing so, it’s moved the criteria or the characteristics closer to overlap with dysarthria. So in my mind, you know, the– I don’t feel– I’m still very challenged by it, but I sort of think I know what I need to hear to separate phonologic from the apraxic errors, but there are some, you know, slow rate. I mean, really, that’s almost pervasive. So that doesn’t– that’s an overlap feature. Excess and equal stress to some degree, that’s an overlap feature. You see it in dysarthria. But this is– When you see slow rate and you see some excess and equal stress, you’re usually thinking about spastic dysarthria. But she doesn’t have strain voice, a strained harsh voice and she doesn’t have someone– she didn’t have any hyper nasality. She’s no resonance difficulties. So does that have to be — Does that have to be part of what you have say spastic dysarthria? Now in my mind, they ought to go together. And if you’ll only see one, you’re confused. But that’s an excellent point, I mean. And we debate all the time in our consensus meetings, at some point with these patients over time where we’ve said they’re just apraxic, no dysarthric and they come in the next time. And we’re saying, “Are they also dysarthric now?” What are we hearing here that– And it can be challenging. And you often fall back on the things like voice quality and resonance which we do not tend to associate with apraxia of speech, so.
Hi. I’m Melissa Johnson. And my question sort of follows up on that. Recently in the SIG 2 discussions, there was some back and forth about this: Can you have apraxia with conduction aphasia? Is that phonological? Motor speech? Back and forth, back and forth. And this apraxia of speech rating scale and what I think you said, you have only looked at it with your progressive apraxia participants. But I’m wondering if you think that would be something useful to help with those distinctions in a broader population like that with helping students see the difference, et cetera.?
I’m Becky Khayum. And regarding the management of progressive apraxia of speech, and thank you again, this is so wonderful. I would be happy to share a little bit with you the Northwestern Communication Bridge Study. This is unpublished but we’ve been working with around 40 participants via internet-based treatment. And I would say of those between 10 and 15 have apraxia of speech, and I would say probably five of those, I would call PPAOS.
And the two strategies that we have found that are more impairment based that have been helpful are a personally relevant multisyllabic word program. We’ve been using syllable segmentation in Rosenbek’s eight step hierarchy to help them pronounce those words. And we have seen dramatic effects. They’ve gone from, you know, they can’t pronounce things, personally relevant multisyllabic words to nearly over 90% accuracy and script training for prayers and ordering food.
With that said, what I found is compared to the other participants without apraxia of speech, at the year check in, they are most likely to be so frustrated by the exercises that they are likely to not be practicing the home programs anymore because they just– it helps, but they get so depressed and frustrated by them. It’s not across the board, but for most of them. So I, with this population immediately began to develop different types of communication aids very early on. So, anyways, I just wanted to share that.
Yeah. No, thank you. I think that’s excellent. And I agree with the script training. We certainly recommended that to some patients with some anecdotal feedback that that makes a difference. And the multisyllabic word drill approach, we often do that too, and the patient gets to select the words. So, those are very logical things to be doing. I also agree that, you know, we’re really talking about staging management. So, at one point, maybe early on when things are relatively mild, more formal, real work on speech to improve or maintain it makes perfect sense. It makes less sense as time goes on. And we just need to figure out what’s the threshold and when do you start doing the other things and so on.
Thank you. Thank you very much.