New Evidence for Redrawing the Language Map of the Brain
ARTICLE IN BRIEF
A research team has updated the traditional map of language comprehension in the brain based on new research in patients with primary progressive aphasia, showing that word comprehension is actually located in the left anterior temporal lobe and sentence comprehension is distributed widely throughout the language network.
For more than a century, neurologists have focused primarily on two areas of the left hemisphere of the brain to understand language processing: Broca's area, in the left posterior frontal lobe, for language production, and Wernicke's area, in the posterior left temporal lobe, for language reception and comprehension.
But a new study of individuals with primary progressive aphasia (PPA) — a condition that most often occurs when either Alzheimer's disease or frontotemporal lobar degeneration affects the brain's language centers — suggests that this language map needs to be redrawn.
Using quantitative magnetic resonance imaging (MRI) and language testing, world-renowned PPA expert Marek-Marsel Mesulam, MD, FAAN, director of Northwestern University's Cognitive Neurology and Alzheimer's Disease Center, and his team found that single-word processing is actually located entirely outside of Wernicke's area, in the left anterior temporal lobe. And sentence comprehension resides diffusely throughout the language network, they wrote in a paper published online first on June 25 in Brain.
In the past, the understanding of the brain's language centers was based largely on studies of aphasia in stroke patients. And the conventional wisdom has been that significant damage to Wernicke's area results in the loss of both single-word comprehension and sentence comprehension.
Aphasia studies found that people whose strokes had affected Wernicke's region couldn't explain what common words like “mirror” or “supermarket” meant. Even if they understood what certain words meant, they were usually unable to understand those words in a sentence. For example, a patient with a stroke showing changes in Wernicke's region might be able to correctly identify a key and a lock, but they would not understand what to do when told, “Put the key in the lock,” because they would be unable to understand the construction of the sentence.
This construction didn't make sense to Dr. Mesulam. “When we looked at our PPA patients who had atrophy, cortical thinning, or neuronal loss in Wernicke's area, word comprehension was completely spared,” he told Neurology Today. “Some of the patients had abnormalities in sentence comprehension, but others did not, and it was in no way as severe as what was described in the stroke literature.”
Dr. Mesulam and neuroscientist Emily Rogalski, PhD, an associate professor in the Cognitive Neurology and Alzheimer's Disease Center, launched a systematic study measuring cortical volume and thickness throughout the brain in 72 PPA patients. They found that patients who had lost cortical thickness in Wernicke's area had been spared single-word comprehension and variable sentence comprehension impairments.
“We then did a regression analysis correlating cortical thinning with impairment of comprehension, and we found that the correlations with sentence comprehension were quite widespread,” Dr. Mesulam said. “They included parts of Wernicke's area, but also Broca's area. But the correlations with word comprehension impairments were very focal, centered in the anterior part of the left temporal lobe.”
So why do stroke patients with damage to Wernicke's area lose both single-word comprehension and sentence comprehension, if that is not the primary center for these functions? “Our conclusion was that stroke destroys not only the cortex in this area, but also the underlying fiber pathways that interconnect language centers that work with one another,” Dr. Mesulam explained. “A stroke in Wernicke's is a double whammy: Cortical neurons are lost, and there is also interference with communications throughout the language area, resulting in severe language comprehension impairment.”
The study does not suggest any specific changes in the diagnosis or assessment of either stroke or PPA, Dr. Mesulam said, but it provides additional insights into the organization of function in the brain.
“For many of us who go into neurology, what lies in our heart of hearts is to figure out how in the world the brain works,” he said. “Our study sheds a slightly different light on the organization of language comprehension in the brain, and it also goes to show that there are no absolute answers about brain organization, especially in the higher cortical functions. Each approach — dealing with models such as stroke, epilepsy, or neurodegenerative disease — gives a slightly different insight on organization of function in the brain.”
These conclusions are very important, said Christopher M. Filley, MD, FAAN, a professor of neurology and director of the Behavioral Neurology Section at the University of Colorado School of Medicine. “Presenting a study of 72 patients with PPA, so well characterized with cortical measures and language assessments, is remarkable in itself,” he said. “And these findings clearly indicate that the conventional view of Wernicke's aphasia is not entirely accurate, at least in terms of the cortical representation of language. Wernicke's aphasia in the traditional sense, related to a stroke, typically involves damage to the subcortical white matter, as this paper points out. But if one looks at degenerative cortical disease, as occurs with PPA, then the pathologic anatomy is different. Presumably the white matter is relatively spared.”
Dr. Mesulam and his team were careful to articulate some of the limitations of the study, said Anne Foundas, MD, a professor and chair of neurology and cognitive neuroscience at the University of Missouri Kansas City School of Medicine.
“These are two different brain injury mechanisms. In a population that has a slower neurodegenerative process, like PPA patients, you would expect manifestations to be different from acute processes like left hemispheric stroke patients,” she said. “In PPA, there is also ongoing degeneration, whereas with stroke there may be neuroplasticity and remodeling of networks that could impact the structure-function relationships.”
Other areas may also contribute to the behaviors and deficits Dr. Mesulam's team examined, she added. “For example, there is a lot of converging evidence that object recognition is mediated by a more bilateral network that involves the left ventral temporal lobes, while category-specific naming appears to be more of a left anterior temporal lobe location. In addition, there is significant evidence that the inferior frontal cortex — the more classic Broca's area — has a large impact on grammatical structure and language comprehension at the sentence level.”
Bridging the more classic understanding of language zones with newer imaging technology and well-controlled studies assessing specific behavioral dissociations will help clinicians develop cognitive rehabilitation strategies that target specific populations, Dr. Foundas said.
“The more we understand about specific cognitive constructs, the more we can apply that knowledge to a variety of clinical populations in brain injury, and ultimately use that to develop targeted treatments,” she said. “Understanding that there are larger networks in language helps us to understand that brain injury-induced language deficits are more [part] of a network model. This gives us a better idea of how we can help our patients function better, whether their injury is a neurodegenerative or an acute process, and give them a better quality of life.”
EXPERTS: ON A NEW LANGUAGE MAP OF THE BRAIN