Reporter's Notebook: RSNA 2017
Radiology Today
Vol. 19 No. 1 P. 6

Editor's note: This article is based on materials distributed for press conferences at RSNA 2017 in Chicago.

MRI Uncovers Brain Abnormalities in People With Depression
and Anxiety

Researchers using MRI have discovered a common pattern of structural abnormalities in the brains of people with depression and social anxiety, according to a study presented at the RSNA annual meeting.

Major depressive disorder (MDD), often simply referred to as depression, is a serious condition. The condition can have a debilitating effect on an individual's life. People with MDD often lose interest in activities they once enjoyed and sometimes find it difficult just to get out of bed. At times, they may feel suicidal. About 16 million Americans have MDD.

Social anxiety disorder (SAD) is an intense fear of being watched and judged by others. Symptoms can be extreme enough to interfere with daily activities. People with this disorder have difficulty developing and maintaining social and romantic relationships. About 15 million American adults have social anxiety disorder.

Both conditions share some clinical symptoms, suggesting the two disorders may have similar brain mechanisms. However, few studies have directly compared the brain structural effects of the two disorders, according to the author of the new study, Youjin Zhao, MD, PhD, from Sichuan University in Chengdu, China.

Zhao and coauthor Su Lui, MD, used MRI to assess alterations in the brain's gray matter among MDD and SAD patients. They focused on the thickness of the cortex, which is the outer layer of the cerebrum, or principal part of the brain.

The researchers acquired and analyzed high-resolution images from 37 MDD patients, 24 SAD patients, and 41 healthy control individuals. Both MDD and SAD patients, relative to healthy controls, showed gray matter abnormalities in the brain's salience and dorsal attention networks. The salience network is a collection of brain regions that determines which stimuli are deserving of our attention, while the dorsal attention network plays an important role in focus and attentiveness.

"Our findings provide preliminary evidence of common and specific gray matter changes in MDD and SAD patients," Zhao said. "Future studies with larger sample sizes combined with machine learning analysis may further aid the diagnostic and prognostic value of structural MRI."

The differences between the MDD and SAD patients and the healthy controls related to either thickening or thinning of the cortex. For instance, both MDD and SAD patients, relative to healthy controls, showed cortical thickening in the insular cortex, a brain region vital to perception and self-awareness.

It is still unclear exactly what the relationship is between the clinical manifestations of MDD and SAD and cortical thickening in brain regions like the anterior cingulate cortex, a part of the brain associated with emotion, Zhao said.

"First, it is possible that a greater cortical thickness may reflect a compensatory mechanism that is related to inflammation or other aspects of the pathophysiology," she said. "Second, greater anterior cingulate cortical thickness could be the result of both the continuous coping efforts and emotion regulation attempts of MDD and SAD patients."

As for cortical thinning, Zhao said that other research provides convincing evidence to support the theory that reduced cortical layer thickness in some brain regions may result in the decreased thickness of the frontal lobe, a large part of the brain that is involved in a variety of functions, including emotion.

The researchers also found disorder-specific involvement of the brain's "fear circuitry" in patients with SAD and involvement of the visual recognition network in patients with MDD. Alterations in the brain within the region of the visual recognition network might be related to impaired selective attention and working memory in MDD, according to Zhao.

"The visual recognition network is involved in emotional facial processing, which is crucial for social functioning," she said. "Depression has been associated with structural alterations in these regions."

No Evidence That Gadolinium Causes Neurologic Harm
There is no evidence that accumulation in the brain of the element gadolinium speeds cognitive decline, according to a new study presented at the annual meeting of RSNA.

Introduced in 1988 as a means of enhancing MRI images, gadolinium-based contrast media are commonly used today. "It's estimated that approximately 400 million doses of gadolinium have been administered since 1988," said the study's lead author, Robert J. McDonald, MD, PhD, neuroradiologist at the Mayo Clinic in Rochester, Minnesota. "Gadolinium contrast material is used in 40% to 50% of MRI scans performed today."

Scientists previously believed gadolinium contrast material could not cross the blood-brain barrier, the semipermeable membrane that selectively filters materials in the bloodstream from entering extracellular fluid in the brain and central nervous system. However, recent studies, including one from McDonald and colleagues, found that traces of gadolinium could be retained in the brain for years after MRI.

On September 8, 2017, the FDA recommended adding a warning to labels about gadolinium retention in various organs, including the brain, for gadolinium-based contrast agents used during MRI. The FDA highlighted several specific patient populations at greater risk, including children and pregnant women. Still, little is known about the health effects, if any, of gadolinium that is retained in the brain.

For this study, McDonald and colleagues set out to identify the neurotoxic potential of intracranial gadolinium deposition following IV administration of gadolinium-based contrast agents during MRI.

The researchers used the Mayo Clinic Study of Aging (MCSA), the world's largest prospective population-based cohort on aging, to study the effects of gadolinium exposure on neurologic and neurocognitive function.

All MCSA participants underwent extensive neurologic evaluation and neuropsychological testing at baseline and 15-month follow-up intervals. Neurologic and neurocognitive scores were compared using standard methods between MCSA patients with no history of prior gadolinium exposure and those who underwent prior MRI with gadolinium-based contrast agents. Progression from normal cognitive status to mild cognitive impairment and dementia was assessed using multistate Markov model analysis.

The study included 4,261 cognitively normal men and women, between the ages of 50 and 90 with a mean age of 72. Mean length of study participation was 3.7 years. Of the 4,261 participants, 1,092 (25.6%) had received one or more doses of gadolinium-based contrast agents, with at least one participant receiving as many as 28 prior doses. Median time since first gadolinium exposure was 5.6 years.

After adjusting for age, sex, education level, baseline neurocognitive performance, and other factors, gadolinium exposure was not a significant predictor of cognitive decline, dementia, diminished neuropsychological performance, or diminished motor performance. No dose-related effects were observed among these metrics. Gadolinium exposure was not an independent risk factor in the rate of cognitive decline from normal cognitive status to dementia in this study group.

"Right now, there is concern over the safety of gadolinium-based contrast agents, particularly relating to gadolinium retention in the brain and other tissues," McDonald said. "This study provides useful data that at the reasonable doses 95% of the population is likely to receive in their lifetime, there is no evidence at this point that gadolinium retention in the brain is associated with adverse clinical outcomes."

Coauthors are Jennifer S. McDonald, PhD; Terry Therneau, PhD; Laurence J. Eckel, MD; David F. Kallmes, MD; Rickey Carter, PhD; Clifford R. Jack Jr, MD; and Ronald C. Petersen, MD, PhD.

This paper is the winner of the Kuo York Chynn Neuroradiology Research Award.

New Studies Show Brain Impact of Youth Football
School-age football players with a history of concussion and high-impact exposure undergo brain changes after one season of play, according to two new studies conducted at University of Texas (UT) Southwestern Medical Center in Dallas and Wake Forest University in Winston-Salem, North Carolina.

Presented at the RSNA annual meeting, both studies analyzed the default mode network (DMN), a network of brain regions that is active during wakeful rest. Changes in the DMN are observed in patients with mental disorders. Decreased connectivity within the network is also associated with traumatic brain injury.

"The DMN exists in the deep gray matter areas of the brain," according to Elizabeth M. Davenport, PhD, a postdoctoral researcher in the Advanced NeuroScience Imaging Research (ANSIR) lab at UT Southwestern's O'Donnell Brain Institute. "It includes structures that activate when we are awake and engage in introspection or processing emotions, which are activities that are important for brain health."

In the first study, researchers studied youth football players without history of concussion to identify the effect of repeated subconcussive impacts on the DMN.

"Over a season of football, players are exposed to numerous head impacts. The vast majority of these do not result in concussion," said Gowtham Krishnan Murugesan, a PhD student in biomedical engineering and a member of the ANSIR lab. "This work adds to a growing body of literature indicating that subconcussive head impacts can have an effect on the brain. This is a highly understudied area at the youth and high school level."

For the study, 26 youth football players (ages 9 to 13) were outfitted with the Head Impact Telemetry System (HITS) for an entire football season. HITS helmets are lined with accelerometers or sensors that measure the magnitude, location, and direction of impacts to the head. Impact data from the helmets were used to calculate a risk of concussion exposure for each player.

Players were equally divided into high and low concussion exposure groups. Players with a history of concussion were excluded. A third group of 13 noncontact sport controls was established. Pre- and postseason resting functional MRI (fMRI) scans were performed on all players and controls, and connectivity within the DMN subcomponents was analyzed.

The researchers used machine learning, a type of artificial intelligence that allows computers to perform analyses based on existing relationships of data, to analyze the fMRI data.

"Machine learning has a lot to add to our research because it gives us a fresh perspective and an ability to analyze the complex relationships within the data," Murugesan said. "Our results suggest an increasing functional change in the brain with increasing head impact exposure."

Five machine learning classification algorithms were used to predict whether players were in the high-exposure, low-exposure, or noncontact groups based on the fMRI results. The algorithm discriminated between high-impact exposure and noncontact with 82% accuracy, and low-impact exposure and noncontact with 70% accuracy. The results suggest an increasing functional change with increasing head-impact exposure.

"The brains of these youth and adolescent athletes are undergoing rapid maturation in this age range. This study demonstrates that playing a season of contact sports at the youth level can produce neuroimaging brain changes, particularly for the DMN," Murugesan said.

In the second study, 20 high school football players (median age 16.9) wore helmets outfitted with HITS for a season. Of the 20 players, five had experienced at least one concussion, and 15 had no history of concussion.

Before and following the season, the players underwent an eight-minute magnetoencephalography (MEG) scan, which records and analyzes the magnetic fields produced by brain activity. Researchers then analyzed the MEG power associated with the eight brain regions of the DMN.

Postseason, the five players with a history of concussion had significantly lower connectivity between DMN regions. Players with no history of concussion had, on average, an increase in DMN connectivity.

The results demonstrate that concussions from previous years can influence the changes occurring in the brain during the current season, suggesting that there are longitudinal effects of concussion that affect brain function.

"The brain's default mode network changes differently as a result of previous concussion," Davenport said. "Previous concussion seems to prime the brain for additional changes. Concussion history may be affecting the brain's ability to compensate for subconcussive impacts."

Both researchers said larger data sets, longitudinal studies that follow young football players, and research that combines both MEG and fMRI are needed to better understand the complex factors involved in concussions.

Murugesan's coauthors are Afarin Famili; Elizabeth M. Davenport, PhD; Ben Wagner, BM; Christopher T. Whitlow, MD, PhD; Joel Stitzel; Jillian Urban; Joseph A. Maldjian, MD; and Albert Montillo, PhD.

Davenport's coauthors are Jillian Urban; Ben Wagner, BM; Mark A. Espeland, PhD; Alexander K. Powers, MD; Christopher T. Whitlow, MD, PhD; Joel Stitzel; and Joseph A. Maldjian, MD. This work is part of several National Institutes of Health-funded studies of youth and high school football, with additional funding from the Childress Institute for Pediatric Trauma. Data collection was performed by researchers at Wake Forest University, and data analysis was done by researchers at UT Southwestern.