Our Research

Developing and validating tests of socioemotional cognition for patients with cognitive deficits
 

The behavioral variant of frontotemporal dementia (bvFTD) is one of only a handful of neurologic syndromes in which initial clinical diagnosis relies entirely on assessment of patients’ social and emotional symptoms (Ranasinghe, 2016). Because this striking behavioral syndrome is driven by focal brain dysfunction, bvFTD is an ideal disease model with which to investigate the neural underpinnings of human socioemotional behavior. Yet until recently, no tools were available that were designed specifically to measure socioemotional functioning in these patients, who often also have cognitive deficits.

We have addressed this problem by successfully developing measures of socioemotional behavior that are valid for use in patients with neurodegenerative disease, and reflect structural changes to the specific brain circuits that degenerate in bvFTD (Sollberger, 2014). Already, a subset of these tests has been incorporated into an FTD testing module promoted by the National Institute of Aging through the National Alzheimer’s Coordinating Centers. Dementia centers all over the country are now using the tools that we have validated in the course of this project to improve diagnostic accuracy, more precisely characterize disease phenotypes, and clarify brain-behavior relationships. The poor understanding of neuroanatomy, pathology, genetics, and environmental risk factors behind cognitive and socioemotional functioning hinders the ability of clinicians and researchers to make progress improving evaluations, treatment and prevention of such symptoms.

Now that these tests have been developed, we are using them to better understand disease progression in our patients, and to model the neural systems underlying human socioemotional behavior in both healthy and disease states.

Understanding the structural and functional neuroanatomy of complex social processes
 

The recent discovery that each major neurodegenerative disorder initially targets a distinctive intrinsically connected functional network (ICN) in the brain, has led to the identification of three initial sites of dysfunction in different clinical subtypes of bvFTD (Seeley, 2009). These include the ventral salience network (SN), the task control network (TCN), and the “limbic” or semantic-appraisal network (SAN).

Still, little is known about the specific socioemotional behaviors driven by these ICNs, either in normal cognition or in disease. By developing our understanding of the network underpinnings of normal social functioning throughout adulthood, we will have a foundation on which to model how these circuits malfunction to cause the behavior symptoms of bvFTD and other neurologic and psychiatric disorders. Identifying the specific relationships between behavior and network connectivity in bvFTD patients will also maximize our ability to measure symptom progression in upcoming clinical trials for FTD.

Specifically, we aim to elucidate the contribution of these three identified neural networks to normal social behavior and to the socioemotional symptoms of bvFTD. In particular, we hope to clarify how the three bvFTD networks correspond to socioemotional behavior and cognition in healthy normal adults, determine how these three networks relate to severity of social dysfunction in bvFTD, and examine socioemotional and network dysfunction occurring at the earliest stage of bvFTD.

Understanding the structural and functional neuroanatomy of neuropsychiatric symptoms
 

An area related to our investigations of socioemotional functioning is the anatomic basis of neuropsychiatric symptoms in our patients. With the progression of these neurodegenerative diseases, specific circuits in the brain are disrupted, and our patients show a full range of psychiatric symptoms such as anxiety, depression, mania, apathy, emotion dysregulation, and even psychosis (Rankin, 2006). However, still very little is known about the specific mechanisms involved in the appearance of these symptoms. We are measuring specific psychiatric symptoms in our patients through questionnaires and interviews, and are analyzing those symptoms in conjunction with our patients’ structural and functional brain scans to identify causal relationships.

Psychosis, for example, is understudied and widely misunderstood. Clinicians have relied on the premature assumption that psychosis in dementia is similar to psychosis in psychiatric illnesses that affect the young, such as schizophrenia. Although phenomenologically related, the two may have completely difference biologies, as evidenced by the limited benefit of classical antipsychotics in dementia. Psychosis is often the first sign of diseases, which has direct relevance to clinician's ability to identify the earliest brain networks affected in diseases, and thus the ability to predict the underlying neuropathology (Khan, 2012). Similarly, NDG can occur in patients with a lifetime history of depression, but patients can present with new depressive symptoms as their first sign of NDG, or can develop depressive symptoms during the course of their disease.

Using our findings from brain autopsy, clinical, and imaging data concurrently, we hope to both elucidate the specific clinical presentations of psychopathologies in our patients, and to clarify the underlying neural mechanisms for those symptoms.

Improving early and accurate diagnosis of bvFTD in the community
 

Current health care systems are unprepared to meet the growing complexity of Alzheimer’s disease and related dementias. Often, families are ill prepared for the caregiving challenges they encounter (Brown, 2018), and diagnostic evaluations are frequently incomplete or poorly formulated, resulting in missed or wrong diagnoses. These errors are a critical problem, because early detection can prompt evaluation for reversible causes of cognitive loss, offer an opportunity for better management that can optimize prognosis (e.g., by adding symptomatic medications and removing inappropriate medications), and provide time for the patient and family to prepare for future care while still at a stage when the patient can participate in these decisions.

Most patients with dementia are diagnosed and cared for by primary care providers. Problems in accurate diagnosis of dementia arise when physicians not only perceive themselves as lacking competence in dementia diagnosis, but also lack access to valid tools that can be administered during a short primary care office visit. Furthermore, often diagnostic evaluations and treatment recommendations are not consistent with best practices.

There is a major opportunity to improve the accuracy of dementia diagnosis and the quality of treatment by offering decision guidance to primary care physicians, neurologists, and psychiatrists who do not specialize in atypical presentations of neurodegenerative disease. We are working on a number of tools related to this problem, including tablet-based cognitive evaluations (Possin, 2018), automated quantification of neuroimaging, dashboard visualizations for single patient and aggregate group data, and computerized decision trees to aid clinical diagnosis.

Accelerating scientific discovery and cures by creating tools for better data linking, analytics and visualization of clinical data

 

The scientific process will be accelerated to the degree that researchers are able to link data together across disciplines and laboratories, connect with publicly available databases containing existing scientific knowledge about relationships among genes, proteins, and other aspects of cell biology, and enhance communication about scientific activities throughout the larger scientific and patient community. In particular, by connecting phenotypic and genotypic data, investigators will be able to more easily recognize potential biological mechanisms and identify promising new research pathways that could lead to treatments.

One of the major activities in the Rankin Lab is our work supporting the Neuroscience Precision Medicine Knowledge Network Pilot, a project in which we are developing Data Management/Integration tools, specifically two classes of informatics tools designed to enhance the integration of phenotypic (patient clinical) data with genetic data. These tools are designed to allow researchers to perform ad hoc linking of patient data regardless of the source data management platform or data type, in order to create complex, multilevel clinical/imaging/genetic datasets richly describing patients shared across investigator labs. We also have prototyped an automated brain imaging analysis pipeline built to allow investigators to explore brain structure-function relationships. We have been developing multiple dashboards for dynamic visualizations, designed to support interpretation of complex, integrated genetic and clinical data, both in single patients and across groups of patients.

In the next year, we hope to expand use of the Knowledge Network architecture and tools to investigators outside of the Memory and Aging Center, including at other institutions, with a focus on linking data across labs to expand access to multilevel patient-based data.