Language-Network Connectivity Disrupted in Primary Progressive Aphasia

By Will Boggs MD

September 11, 2019

NEW YORK (Reuters Health) - Primary progressive aphasia is associated with perturbations of connectivity within components of the brain's language network, researchers report.

"We found areas of the brain in primary progressive aphasia (PPA) where function of the living networks was decreased beyond the degree of brain shrinkage or atrophy," said Dr. Borna Bonakdarpour of Northwestern University Feinberg School of Medicine, in Chicago.

"That means there are still neurons available within certain brain networks that are alive but dysfunctional," he told Reuters Health by email. "Dead cells cannot be revitalized, but these neural networks are potential targets for therapeutic interventions since they are alive."

Dr. Bonakdarpour and colleagues previously showed that core regions of the language network - the inferior frontal gyrus (IFG), middle temporal gyrus (MTG) and anterior temporal lobe (ATL) - can show functional-connectivity changes in early stages of non-fluent/agrammatic PPA (PPA-G).

In the current study, they extend those findings by examining larger samples of patients with PPA-G and the other variants of PPA: logopenic (PPA-L), distinguished by impairment of word finding and repetition, and semantic (PPA-S), characterized by impairment in naming and single word comprehension.

The functional and structural MRI studies included 36 individuals with PPA-G, 20 with PPA-L, 17 with PPA-S and 33 healthy controls.

Cortical volume was significantly decreased in all PPA groups compared with controls, the team reports in Cortex, online September 1.

All three PPA groups showed decreased IFG-MTG connectivity compared with the control group. The PPA-S group also had significantly reduced IFG-ATL and MTG-ATL connectivity, and MTG-ATL connectivity was lower in the PPA-S group than in the other PPA groups.

After correction for atrophy, IFG-MTG connectivity was still reduced in PPA-G, trended towards significant reduction in PPA-L, and was no longer significantly reduced in PPA-S.

Additional analyses revealed decreased connectivity between IFG and angular gyrus (AG) in the PPA-L and PPA-S groups and a unique decrease in AG-MTG connectivity in PPA-S.

Naming and word comprehension selectively correlated with MTG-ATL connectivity, and grammar and repetition correlated with IFG-MTG connectivity.

"Through our methodology we can demonstrate in which areas of the brain aphasic symptoms are more related to functional derangement as opposed to structural damage," Dr. Bonakdarpour said. "We think those areas are more likely to respond to therapeutic interventions."

He added that the findings "can potentially be used for two purposes: 1) Planning of interventions such as transcranial magnetic stimulation (TMS), and 2) Measurement of response to therapeutic interventions from speech therapy, to drugs, or neuromodulation (TMS)."

"We need to continue our research on better understanding of abnormal physiology in neurodegenerative diseases that cause memory loss or PPA," Dr. Bonakdarpour said. "Results from such investigations can help us with better management of these devastating diseases."

Dr. Hugo Botha of the Mayo Clinic in Rochester, Minnesota, who has also studied disrupted functional connectivity in PPA, told Reuters Health by email, "This paper is the latest in a decade or more of research supporting the idea that degenerative diseases target large-scale brain networks or systems. In the case of PPA, there have been important papers showing that the language network is affected, along with some other networks involved in speech and language."

"This paper adds further evidence by showing how specific aspects of language impairment map onto disruption of connections between language areas," he said. "It is interesting, because the authors found correlations between specific domains and connectivity across PPA subtypes. I think it is important given the heterogeneity of PPA and the fact that a large number of patients cannot be accurately classified into one of the canonical subtypes."

"The three canonical PPA subtypes represent pure or idealized manifestations of dysfunction of specific parts of the language network, but in many cases the disease process affects multiple subregions of the language network as well as non-language networks," Dr. Botha said. "It is helpful for clinicians and researchers to realize that some language deficits may help localize the problem, and hence help formulate a differential diagnosis, regardless of whether or not a patient neatly fits into one of the three canonical subtypes."


Cortex 2019.