Walking around in public without clothes, or baking in the summer sun with much too much on? Laughing at a funeral, or frowning at a funny joke? These are just a few of the odd behaviors that the partners of people with frontotemporal dementia (FTD) first notice about their loved one. Sometimes the abnormalities seem to contradict each other—some people with FTD become hypersensitive to certain stimuli, while others grow a thick skin. How are these diverse symptoms related, and what are the underlying changes in neural circuitry that cause them? How can doctors reliably detect early changes in something as varied as behavior? Could these changes be clues to diagnostic or therapeutic biomarkers? Researchers grappled with these questions, and presented some early answers, at the 9th International Conference on Frontotemporal Dementias, held October 23 to 25 in Vancouver, Canada.
Whatever their nature, FTD-related behavioral changes all represent a deficit in the brain’s ability to process, interpret, or respond to incoming stimuli such as pain, temperature, sound, emotion, or social cues, said Virginia Sturm of the University of California, San Francisco. Sturm and other researchers presented clever ways to measure patients’ responses to these stimuli via physiologic and autonomic responses, such as dilated pupils, a quickening of the pulse, or a spike in blood pressure. They then correlated those findings with neuroimaging and genetics data to uncover the networks that break down in FTD. One common vein in their findings was the disintegration of hubs in the cortex's salience network—such as the insula and the cingulate—in people with FTD who lose their ability to respond properly to emotional or physiologic stimuli. People with C9ORF72 mutations, a major genetic cause of FTD, had trouble even deeper in the brain, in the thalamus. From this coupling of physiological phenotypes with neurodegenerative patterns, key “FTD circuits” started to materialize at the conference.
Jason Warren of University College London broke the ice on this topic during the first session of ICFTD. Warren presented data on pupil changes in response to sound. He compared responses of patients across the spectrum of FTD disorders, as well as people with Alzheimer’s disease and healthy controls. In general, people with FTD responded less robustly to sounds than did the two control groups. However, key differences even between different types of FTD emerged when Warren compared the way people responded to a meaningless sound, such as static noise, to the way their pupils ballooned when hearing a sound with real-world meaning, such as the buzz of a mosquito.
Unexpectedly, people with semantic dementia had the biggest difference in their responses to meaningful versus meaningless sounds. They were aroused by meaningful sounds, even though their semantic deficits prevented them from understanding what those sounds meant. Warren thinks the results suggest that the brain may be hyperstimulated by salient sounds it cannot define. “This could be an autonomic index of things not being quite right with the world, similar to when you hear something that goes bump in the night,” he told Alzforum. Interestingly, this test held up as a way to measure semantic deficits across all the disorders. For example, people with AD who displayed semantic deficits on other tests also tended to have a larger pupil response to salient sounds. This test therefore could be useful as a physiological measure of semantic decline, or perhaps in the setting of therapeutic trials of recovery of semantic memory.
Warren’s group presented a slew of posters on the first day of the conference, displaying data on the way people with FTD responded to music, humor, art, and even sarcasm, all of which were off in some manner. To connect these response deficits to underlying neural circuitry, the London group analyzed the way people with FTD responded to pain and temperature changes. Phillip Fletcher, a graduate student in Warren’s lab, presented these findings in a talk.
Many people with FTD show early abnormalities in their ability to feel or process pain or changes in temperature. To get an idea of how these symptoms present across the FTD spectrum, Fletcher gave caregivers a questionnaire that asked for details about their loved ones’ responses to temperature and pain. Fletcher found that 32 out of 58 patients with FTD had problems with pain and/or temperature awareness, but the problems manifested in different ways between clinical types. People with semantic dementia tended to have exaggerated responses, often complaining of cold or pain, for example. On the other hand, people with behavioral variant FTD tended to have dulled responses. One caregiver recalled having to insist that a patient remove his jacket when it was hot outside; another reported that a patient did not complain of pain even after suffering a bad fall and a black eye. All six patients with C9ORF72 mutations (five with behavioral variant FTD, one with progressive aphasia) had abnormal pain and/or temperature awareness.
To correlate these abnormalities with neuroanatomical changes, Fletcher and colleagues scanned the patients for gray-matter atrophy using voxel-based morphometry. All patients with disturbed temperature or pain awareness had atrophy in their right posterior insula. This makes sense, Fletcher said, because the region has been shown to switch on in response to pain. Patients lacking C9ORF72 mutations also had shrinkage in the right anterior temporal lobe, a region that is important for nonverbal semantic processing. In contrast, the C9ORF72 expansion carriers had normal anterior temporal lobes but a shrunken thalamus—one of the first regions to encounter and process incoming pain and temperature signals. From there, the signals move up toward the posterior insula, which creates a virtual map of the signal’s origin in the body. The posterior insula then communicates with the anterior insula, which integrates this sensory information with signals from the cortex, helping put the pain in context. Together, these regions coordinate the pain or temperature response.
That C9ORF72 mutation carriers have atrophy in the thalamus jibes with other studies. Laura Downey, also from the University College London group, recently found that C9ORF72 mutation carriers have skewed body schema, meaning they have difficulty creating a mental representation of their body or determining where their own body ends and another’s begins (see Downey et al., 2014). Body schema mapping relies heavily upon both the insula and the thalamus. Another recent study led by William Seeley at the University of California, San Francisco, found that C9ORF72 carriers had more thalamic atrophy than patients with other types of FTD, and that atrophy correlated with decreased functional connectivity between the thalamus and key hubs of the salience network, including the anterior insula and cingulate cortex (see Lee et al., 2014).
FTLD has long been considered primarily a cortical disease, and the role of the thalamus has not been appreciated until recently, Warren said. “The thalamus has been forgotten and then ‘rediscovered’ throughout the history of neurology,” he said. The recent combination of physiological and neuroimaging data has built a strong case for the region’s importance in particular dementia diseases. “The physiological markers and the anatomy data are all gelling,” he said.
Downey and Colin Mahoney, another graduate student from Warren’s group at UCL, delved into the changes in structural connectivity that underlie some FTD patients’ difficulties in picking up social cues, such as sarcasm. Warren asked patients to interpret the intent of actors in a video to determine their grasp of sarcasm or other emotional states. Then, the researchers performed diffusion tensor imaging to measure the integrity of white-matter tracts. A defunct “sarcasm radar” correlated with decreased integrity of the uncinate fasciculus, a white-matter tract that connects regions of the limbic system to those in the cortex. A patient’s inability to read emotions in others correlated with an erosion of white-matter tracts emanating from the thalamus and the fornix.
In her talk, Sturm dug deeper into complex emotions and social cues. People with FTD often lack inhibition or a sense of decorum, and Sturm attempted to take stock of this deficit using an amusing test. She presented data from a study where researchers measured embarrassment in FTD patients with the “karaoke task.” Participants were fitted with headphones and asked to sing along to the song “My Girl” while being video-taped. They then had to watch themselves singing the song without the accompanying music—an experience that would make most people cringe. During the playback, the participants were monitored for signs of embarrassment, such as changes in facial expression, as well as autonomic responses, such as increased heart rate, sweatiness, or blood pressure. As expected, people with behavioral variant FTD had fewer changes in facial expression or autonomic responses while watching themselves than other people did (see Sturm et al., 2013). Neuroimaging data revealed that this blunted self-consciousness correlated with atrophy in the anterior cingulate cortex, another key hub in the salience network.
It makes sense that the anterior cingulate cortex (ACC) would be involved in generating feelings of embarrassment, Sturm said, because it has connections with the frontal lobes, which facilitate the understanding and encoding of social context and roles, as well as with regions deeper in the brain that mediate autonomic changes in heart rate and breathing. “The ACC acts as a way station between understanding a social context and mounting an emotional response to it,” Sturm said. “So you could imagine that complex emotions, such as self-consciousness, might be particularly hard hit if this hub is lost.”
Sturm is applying similar methods—measuring facial expressions and autonomic responses—to probe empathy in FTD. She monitored patients’ faces while they watched a heart-warming scene from a movie, looking for signs that the patients felt the same way. People with AD had heightened empathic responses compared to controls, while the scene left people with FTD cold. Sturm is currently searching for neuroimaging correlates of this lack of empathy. Similar regions may crop up, as previous studies have reported that the posterior insula lights up when someone experiences pain, and the anterior insula and ACC are activated when someone watches a loved one experience pain (see Singer et al., 2004).
Linking these physiological measures with changes in brain structure finally anchors emotion in biology, Sturm said. “Many times people don’t think emotion is biological; they think it’s something else,” she said. “This work is important for pinpointing the biological reason in the brain why people change their emotions and behavior, and for differentiating it from just a psychiatric explanation.” She added that such studies will also inform researchers’ understanding about the networks that may be affected in psychiatric disorders.
One major theme at ICFTD was that FTD is a behavioral disease. “In the past, we’ve stressed executive function, but now there is more of a push to look at the social/behavioral context of FTD,” said Nadine Tatton of the Association for Frontotemporal Degeneration (AFTD). While physiological measures have a long way to go before they will be seen as objective, standardized tests employed in many clinics, they could ultimately make behavioral symptoms more tractable, Tatton said.
These symptoms initially may be subtle. Work led by Camilla Clark in the UCL group suggests that changes in a person’s sense of humor may be an early warning sign of FTD. Warren views physiological measurements as a way to offer better explanations to patients and caregivers. For example, a person may be perplexed about why their father suddenly switches from enjoying Woody Allen movies to getting a kick out of the Three Stooges. “When it comes to these more complex behavioral and emotional changes, we’re really not giving people answers, and I’d like to change that,” he said. “Even before we have a treatment, I’d like to give people a useful consultation.”
Warren believes distilling emotional responses into relatively simple tests and measurements will entice neurologists to follow suit. “Clinical neurologists are starting now to genuinely engage with the need to look at things like social cognition and emotion, something psychiatrists are more used to dealing with,” he said.
Warren hopes that the measures may one day aid diagnosis or help monitor progress in clinical trials. “The main point of measuring physiological markers is to deconstruct things that are relatively complex into things that are reproducible, measureable, and may appear prior to structural changes in the brain. Tests that seem off-the-wall, like the karaoke test, might be what we need to know whether an intervention is working or not,” he said. “Otherwise, how else would you measure?”
Dana Hilt of FORUM Pharmaceuticals in Watertown, Massachusetts said that some form of physiological test could serve as a bridge between biochemical markers, such as blood biomarkers or neuroimaging, and clinical outcomes, such as overt changes in behavior and executive function. Hilt used such a measure—evoked-response EEG measurements—in a schizophrenia trial to learn if the drug not only accessed the CNS, but triggered a pharmacological effect (see Preskorn et al., 2014).
“What I think is necessary in this area are new ‘intermediate biomarkers’ that are stronger than biochemical effects and that can then be used to select doses and schedules for full clinical studies,” Hilt wrote in an email to Alzforum. “Assessment of cognition and clinical function are impacted by so many sources of variability that it makes utilizing these for smaller Phase 2 studies quite risky. An ‘intermediate biomarker’ would be useful in this regard.”—Jessica Shugart
- Sturm VE, Sollberger M, Seeley WW, Rankin KP, Ascher EA, Rosen HJ, Miller BL, Levenson RW. Role of right pregenual anterior cingulate cortex in self-conscious emotional reactivity. Soc Cogn Affect Neurosci. 2012 Mar 20; PubMed.
- Singer T, Seymour B, O'Doherty J, Kaube H, Dolan RJ, Frith CD. Empathy for pain involves the affective but not sensory components of pain. Science. 2004 Feb 20;303(5661):1157-62. PubMed.
No Available Further Reading