Frontotemporal Dementia and Language Processing
An Interview with Murray Grossman, M.D.


by Brenda Patoine

January, 2006

Q: You’ve been at the forefront of efforts to improve diagnostic accuracy and medical management of fron­totemporal dementia (FTD). What is FTD and why is better diagnosis a concern?

Grossman: Frontotemporal dementia is an umbrella term that encompasses several different neurodegenerative conditions, any of which can affect the frontal and temporal parts of the brain and impair language and cognitive functioning. In the past it was called Pick’s disease. A definitive diagnosis of FTD can be made only after examining brain tissue on autopsy, and this has hampered our ability to provide patients with the best possible treatments. Very important work is now underway to try to sort out what kinds of techniques we can develop and what kinds of biomarkers we can use to identify the specific histopathologic conditions causing FTD during life.

Improving the diagnostic accuracy of FTD is not a trivial issue, for several reasons. It’s certainly important from a research perspective to improve our understanding of brain-behavior relations, but it’s also crucially important because we are making rapid advances in developing substances that can treat the conditions that cause FTD. It is essential that doctors have the tools necessary to make an accurate diagnosis during life, so that therapeutic substances can be administered to the appropriate patients.

We’re also trying to make the diagnosis earlier. This too is very important, not only from a purely scientific perspective but from a treatment perspective as well. It’s clear in any progressive neurodegenerative condition that the earlier we intervene, the earlier we start a treatment, the greater the likelihood that a better outcome will be achieved for the patient. We’d like to be able to identify people earlier, when they don’t have full-blown disease but only have some of the initial inklings of what might be coming in the future. If we can intervene at that early point, we’re going to do much better than if we wait for the disease to fully manifest.

We can draw an analogy to Alzheimer’s disease. There is a pre-Alzheimer’s state called Mild Cognitive Impairment, or MCI. This condition is marked by memory difficulty but little impairment in other cognitive domains and reasonable independence in activities of daily living, so people can manage relatively well. Alzheimer’s involves memory difficulty as well as deficits in other cognitive domains sufficient to interfere with day-to-day functioning. If we can slow the onset of Alzheimer’s disease by five years, it is estimated that we can substantially reduce the frequency of the condition in the population.

We’re trying to do the same thing in FTD, that is, to identify people who have some of the very earliest signs of FTD so that we can treat the condition early. This may be even more important in FTD because of the relatively early onset of the condition, when families are still young and haven’t yet devel­oped the social and financial resources to deal with a slowly progressive neurodegenerative condition. Our strategy for identifying the earliest features of FTD is to study families where there is a higher risk for the development of this condition. We are also hoping to study the subgroup of MCI patients without memory difficulty, that is, the MCI patients with subtle problems in other cognitive domains.

Q: You’ve recently published a report suggesting a new technique for determining the levels of the protein tau in the cerebral spinal fluid of patients with FTD. How might this kind of work improve clinical practice?

A: The crucial issue is that we need to find better ways to identify the earliest signs of dementia, and we want to do this with as much specificity as possible. Since tau plays such a crucial role in the pathophysiology of FTD, we think that developing a diagnostic marker for tau has great promise in FTD. If, for example, we can determine during life that the protein tau is accumulating abnormally in the brain, then we can administer a treatment specifically directed at tau, as my colleagues John Trojanowski and Virginia Lee are currently proposing at the University of Pennsylvania (see below). Knowing the underlying cause of a dementia—whether it is tau or something else—is the first step toward curing the condition. That is really our goal. One step is learning to deliver the substance to the appropriate patients, and a bio­marker like the tau level in CSF can help us identify the appropriate patients.

Knowing the underlying cause of a dementia— whether it is tau or something else—is the first step toward curing the condition. That is really our goal.

Other biomarkers also may be useful, including neuropsycho­logical approaches, clinical questionnaires, brain imaging scans, and other biological markers in the cerebrospinal fluid and blood. These also can be used to determine whether or not an etiologically specific treatment is successful in chang­ing the progression of the condition, since biomarkers also can be used to monitor whether there is a response to the treatment. In this way, biomarkers can be used longitudinally to follow individuals over time to track the disease.

Q: Why is tau a target?

A: Tau is one of the abnormal proteins implicated in the cause of FTD. Tau accumulation occurs in about 40 percent of patients with FTD, although the accumulation of tau has several different manifestations. In other FTD patients, there is a remarkable depletion of tau.

Tau is a microtubule-associated protein. Microtubules are microscopic networks of tube-like structures that run along the inside of a nerve cell’s surface. They help distribute meta­bolic substrate around the neuron, and help neurons maintain their shape. In FTD and other neurodegenerative conditions, these proteins can degrade or they can accumulate abnormal­ly and block physiologic function. Eventually this process leads to neuronal atrophy and neuronal death.

My colleagues at the University of Pennsylvania, Virginia Lee and John Trojanowski, are working on the development of substances that will help compensate for the abnormal accumulation of tau in neurons. The substances they are developing are designed to stabilize the microtubules and restore normal function, supporting the role that tau plays in the stabilization and aggregation of microtubules. They published a paper in 2005 in Proceedings of the National Academy of Sciences demonstrating the efficacy of this class of substances in an animal model of FTD that they developed. The substances under investigation are attenuated versions of the anti-cancer treatment Taxol. Their development was triggered by the realization that Taxol, when used at doses needed to treat cancer patients, has a side effect: it causes microtubule aggregation. At much lower doses, this kind of substance may be useful for aggregating microtubules, essentially serving the function that tau normally performs in neurons.

Q: Your basic laboratory research is focused on understanding the neural basis of complex language processing. How does that relate to FTD?

A: I became interested in this area because the two major manifestations of FTD include a disorder of language and a disorder of social comportment and personality. I’m interested in both of those areas—how our brains let us communicate linguistically and socially—so it was a natural progression to study diseases that affect those crucial centers of the brain.

There are very specific kinds of difficulties in language that can be seen in patients who have FTD. This includes patients who have an aphasia—a disorder of language—and patients who do not have an obvious aphasia. In the subgroup of FTD patients with aphasia, we think that there are two major manifestations. One involves impaired semantic memory, or difficulty with the meaning of words and objects. FTD aphasics thus can have significant difficulty understanding the meaning of common words like “chair.” A person with this manifestation of FTD might say, “Chair? I’ve heard that word. What is chair?” It’s as if the page from the patient’s mental dictionary for chair has been ripped out. We hope to study this difficulty so that we can develop effective speech and language therapies for these patients. These therapies are based on understanding the nature of the semantic impair­ment and using functional imaging techniques like fMRI to help understand which brain regions are not contributing to word meaning. Moreover, in our work we’ve found that other healthy brain regions are helping to compensate for some of these difficulties, and we want to support this mechanism with speech and language therapy. 

Another kind of language problem in some aphasics with FTD is difficulty with grammar. We use grammar to construct the sentences we use when we speak. Language is more than just single words that reference an object in the world; language also involves stringing these words together in sentences so we can talk about complex concepts, describe who does what to whom, in a manner that does not depend on the order of word use or the order in which things occur in the world. That kind of grammatical process can fall apart in another group of aphasic patients with FTD.

A third group of FTD patients doesn’t have an obvious lan­guage deficit. They can understand the meanings of words, and their grammar is quite good as well. So these FTD patients have no obvious difficulty speaking or understanding single words and sentences. Instead, they have difficulty with discourse, the ability to tell the story. We recently completed a study with a University of Pennsylvania colleague, Sherry Ash, that looked at the ability to narrate a story based on a word­less picture book, the kind of book you might use with chil­dren who are at the very beginning stages of reading. When these patients try to tell the story by looking at the pictures, all they can do is describe what is on the page. They don’t have any sense that the pages can be integrated into a story. This holds true for their day-to-day conversations as well. Their speech can be tangential and meander from topic to topic without much organization.

To help us understand the neuroanatomic basis for difficulties such as this, we take detailed MRI images of their brains and correlate their difficulty integrating elements of a story into a larger whole with specific brain abnormalities we identify on the scans. For example, this type of language deficit is associated with the frontal and temporal regions of the right hemisphere, the side of the brain that we don’t ordinarily associate with language.

I am confident that we will develop treatments in the near future that will change the natural history of progressive neurodegenerative conditions like FTD.

Q: How might knowledge of language processing in the brain impact clinical management of FTD?

A: I can give you a couple of examples of how we can use language to help understand and treat patients with FTD. If we want to make an early diagnosis of FTD, for example, it becomes critical to discover exactly the right measures of language to see whether there are some specific problems marking a risk for developing FTD. So the idea is to take what we know about these language processing deficits and apply that to some early diagnostic procedures. We can’t put every body in an MRI machine or do a lumbar puncture on every­body to sample cerebrospinal fluid. Instead, we can interview people, talk with them, and try to ascertain a pattern that is sensitive enough to accurately select those who are more like­ly to have FTD. Then we would be in very good shape to iden­tify people for whom it would make sense to follow up with more invasive or more expensive techniques to better deter­mine whether FTD is likely.

I am confident that we will develop treatments in the near future that will change the natural history of progressive neurodegenerative conditions like FTD. If that confidence is well founded, we have to start developing therapies for treat­ing the residual language difficulties that may remain. Doing this effectively requires a detailed understanding of the neural basis of language and figuring out ways to help people. We’re just beginning to understand the neural systems that are compromised when we see a language processing impairment. The goal, of course, is to try to repair the brain, so that these networks can be utilized as fully as possible.

Some patients seem to tap into compensatory mechanisms ... other parts of the brain that are functioning reasonably well can assume some responsibility for the parts of the brain that are not functioning well.

Our imaging studies have shown that some patients seem to tap into compensatory mechanisms that optimize their remaining language skills. In other words, other parts of the brain that are functioning reasonably well can assume some responsibility for the parts of the brain that are not functioning well. It seems clear that there is great opportunity for compen­sation, and some of the behavioral therapies in development are trying to take advantage of this. Essentially, we’re talking about a specific set of mental exercises. We are in the early stages of developing such a program of behavioral therapy. It is incredibly exciting to work in an area that takes advantage of the rapid growth in our knowledge of both the biological basis for a profound neurodegenerative condition and the source of difficulty for the most human of acts, speaking. It is satisfying to know that we are moving forward in our ability to help individuals with these profound conditions.