Subscribe to eTOC


SAN FRANCISCO — Use of diffusion-weighted imaging (DWI) and perfusion MRI (PI) immediately after an ischemic stroke show that lesions in areas of the brain once thought to cause aphasia may not necessarily be so directly involved.

“We used to figure out what part of the brain was affected by studying aphasic stroke patients with chronic lesions,” said Argye Hillis, MD, Associate Professor of Neurology at Johns Hopkins University in Baltimore, MD.

Previous studies suggested an association between lesions in the insula and apraxia of speech. In a widely cited report, Nina F. Dronkers, MD, Professor of Neurology and Linguistics at University of California-Davis, compared two groups of stroke patients – those with speech deficits and those without. The left precentral gyrus of the insula was a common area of infarction in MRI and CT studies of 25 patients with “apraxia of speech,” in which patients have problems coordinating the movements involved but recognize the sounds of speech, including their own errors. The disorder is attributed to impaired planning of articulatory movements (Nature 1996;384(6605):159–161).

Dr. Hillis used DWI and PI within 24 hours of stroke to identify which cerebral areas are essential for specific language functions and which contribute to speech but are not essential. She discussed her findings here at last year's Annual Meeting of the American Neurological Association, where she was awarded the Derek Denny-Brown Neurological Scholar Award for her work.


Dr. Hillis presented imaging studies in two series of patients with nonlacunar left middle cerebral artery stroke. Forty had insular damage and 40 did not.

Speech language pathologists and neurologists evaluated the patients for apraxia of speech during this acute phase. The investigators were blinded to imaging results. Dr. Hillis reported that, unlike the earlier studies of the effects of chronic stroke, there was no association between speech apraxia and damage to the anterior insula. Instead, apraxia of speech was associated with hypoperfusion or infarct of the Broca area in patients with or without insular damage.

Dr. Hillis said the findings suggested that restoring blood flow to the Broca area should reverse the symptoms. This observation prompted a new study of 24 patients with damage to the left caudate with or without damage to surrounding subcortical tissue. Patients again had DWI, PI, and language assessment within 24 hours of the stroke.

Among patients with caudate infarcts, those with Broca aphasia had hypoperfusion in the post-inferior frontal cortex and precentral gyrus with normal blood flow in the superior temporal gyrus. In those with Wernicke aphasia, hypoperfusion was evident in the posterior superior temporal gyrus and there was normal blood flow in the Broca area.

In addition, within seven days of caudate stroke, the researchers induced an increase in blood pressure with phenylephrine or intravenous fluids alone to increase blood flow in the areas of hypoperfusion seen on PI. They increased blood pressure very slowly, in 10-percent increments, to a mean arterial pressure of 130 mmHg.

“We saw improvement almost immediately – in less than 30 minutes,” Dr. Hillis reported. This approach to the treatment of post-stroke aphasia is preliminary and needs much more study, she cautioned.

Dr. Hillis said the imaging studies – done before there is any reorganization or recovery – enabled investigators to identify cortical regions essential for specific components of speech and language. “The findings allowed us to re-evaluate the role of the insula and subcortical infarcts in aphasia. It also enabled us to identify neural correlates of selective speech deficits due to small areas of ischemia,” she said.


“These studies on post-stroke aphasia by Dr. Hillis are original, creative, rich, and exciting,” said Martin L. Albert, MD, PhD, Professor of Neurology and Director of the Harold Goodglass Aphasia Research Center at Boston University School of Medicine.

“Dr. Hillis and her colleagues demonstrated improvement in language function within 30 minutes of treatment by carefully increasing blood flow in areas of cerebral hypoperfusion (by carefully increasing blood pressure).”

“Although these studies are preliminary, they represent an important step toward improving language function in patients with aphasia, a condition affecting more than one million people in this country – many of whom are unable to be gainfully employed.”

Traditional studies of aphasia were based on static information about the brain, usually in the chronic stage of the illness, Dr. Albert continued. “The use of diffusion-weighted imaging and perfusion MRI immediately after an ischemic stroke contributes to the newly emerging picture of brain-language relationship by filling in the gaps between what is known about language in the healthy brain and what is known about the brain in the chronic stages of aphasia,” he said. “The report adds an important dimension to theories of the cognitive neuroscience of language.”

Commenting on the study by e-mail, Dr. Dronkers noted that few studies have looked at speech and language abilities during the first few days after a stroke. “The work by Dr. Hillis allows us to assess the effects of acute stroke and compare them to the long-term effects we see in chronic patients,” she said.

“In terms of the insula and aphasia, we must keep two things in mind,” she continued. “First, the insula is actually a large area of cortex. The area we found related to speech functions in 1996 is only a very tiny part within the anterior insula and has still not been evaluated in acute patients. A direct comparison between and chronic patients with lesions in this discrete area would, indeed, be an important comparison and would tell us more about the reorganization of speech functions after damage to this circumscribed area.”

She added that it is also important to remember that previous studies did not relate the insula to the language disorders of aphasia, but rather to apraxia of speech. “The latter is a motor speech deficit in which patients have trouble coordinating articulatory movements, not in formulating thought into language,” she said.

“Dr. Hillis is virtually the only person working at this level of sophistication of language testing and imaging,” Howard Kirshner, MD, Professor of Neurology at Vanderbilt University Medical Center in Nashville, TN, told Neurology Today. “She is in the unique position of being a trained speech language pathologist and a neurologist.”

“Diffusion-weighted imaging is being used widely today,” Dr. Kirshner said. “It more or less shows damage. Bright areas probably already have an infarct or damage that is imminent. But perfusion imaging is not widely done. Using a contrast agent with PI, usually gadolinium, you can see where the blood is going … A diffusion-perfusion mismatch, where the perfusion deficit is large and the diffusion deficit is small, means that the tissue may be salvageable,” Dr. Kirshner explained. “Dr. Hillis is showing more evidence of that … She is showing that behavior also follows those deficits.”

“These data contradict earlier evidence from studies in the chronic phase of stroke associating the insula with apraxia of speech,” he continued. “In addition, the results cast doubt on the findings of aphasia with chronic lesions restricted to subcortical structures such as the caudate nucleus, anterior limb of the internal capsule, and putamen.”

“We need the right tools to evaluate [cortical damage in aphasia],” Dr. Hillis told meeting attendees. “The exciting thing is, we can now do things to improve cognition, even days after the stroke when the cortex is hypoperfused but not yet infarcted.”


  • ✓ In a new study, Dr. Argye Hillis, of Johns Hopkins University, used DWI and PI within 24 hours of stroke to identify which cerebral areas are essential for specific language functions and which contribute to speech but are not essential.
  • ✓ Experts in aphasia say the findings should enable further research to assess the effects of acute stroke and compare them to the long-term effects we see in chronic patients.


• Dronkers NF. A new brain region for coordinating speech articulation. Nature 1996;384:159–161.