21 April 2011. Traumatic brain injury (TBI) has become a hot topic. Between the high number of military personnel exposed to blasts from homemade bombs and a new awareness of the seriousness of concussive sports injuries, the disorder has grabbed the attention of many researchers. One especially troubling question is whether brain injury seeds the development of dementia. The data to date are suggestive, but far from conclusive.

An important factor to keep in mind is the nature of the brain trauma. “We should not assume that the processes occurring after a single severe injury are identical to those occurring after repetitive concussive injuries,” said David Brody of Washington University in St. Louis, Missouri. Brody is currently comparing mouse models of each type of head trauma, and said that preliminary data suggest there are substantial differences between the two, although it is too early to give details. One feature shared in common is axonal injury—these long, thin nerve fiber tracts appear to be supremely vulnerable to shearing, stretching, and twisting damage (see Smith et al., 2003). Yet even for axons, Brody said, the exact nature of the injury may differ in severe versus mild TBI.

Aβ deposits (red) and APP (green) at a site of axonal injury in mouse brain. Image credit: David Brody and Hien Tran

Severe TBI has been better studied, with a more extensive literature. “Most of the evidence for increased risk of dementia of the Alzheimer’s type is after single severe injuries,” Brody points out (see, e.g., Mortimer et al., 1991; Jellinger et al., 2001; Jellinger et al., 2001; and Fleminger et al., 2003). In one study of almost 2,000 World War II veterans, those with moderate to severe head injuries had twice the risk of developing dementia, and for those with an ApoE4 allele, the risk seemed to increase further (see Plassman et al., 2000).

Numerous papers agree that ApoE4 increases TBI-related dementia risk (see, e.g., Isoniemi et al., 2006; Mauri et al., 2006; and Luukinen et al., 2008), with one influential study finding that people with head injuries had a 10-fold increase in AD risk if they also carried an ApoE4 allele (see Mayeux et al., 1995). The finding is not universal, with a few small or short-term studies showing no association (see, e.g., Chamelian et al., 2004 and Han et al., 2007). Overall, however, the data imply a role for AD-related processes after head injury.

Does the increased risk for AD mean that severe brain injury triggers Alzheimer’s pathology? Not necessarily, Brody said. Such injuries might merely lower a person’s overall brain health and cognitive reserve, leading to an earlier onset of AD symptoms later in life. Or some third factor could be at work. For example, a recent review posited that sleep apnea, which is common among TBI patients, may be a source of cognitive impairment and dementia risk after brain injury (see O’Hara et al., 2009). This idea would be encouraging, since sleep apnea is treatable.

Several studies do point to a possible link between severe brain injury and Alzheimer’s disease processes, however. In animal models, amyloid precursor protein (APP) accumulates in injured axons, where it may be snipped by β-secretase, presenilin-1, or caspase-3 to make Aβ (see Stone et al., 2002 and Chen et al., 2004). Inhibiting caspase-3 lowers Aβ levels after brain injury and improves recovery in mice (see Abrahamson et al., 2006). Likewise, administration of anti-APP antibody to rats with rattled brains reduces neuron and glia death and improves cognitive function (see Itoh et al., 2009 and Itoh et al., 2009). In a mouse study, inhibition of APP-cleaving β- and γ-secretases had similar beneficial effects (see ARF related news story on Loane et al., 2009). These findings suggest a role for amyloid pathology in the deficits caused by head injury, at least in animal models.

In humans, researchers have seen contradictory changes in the levels of Aβ40 and Aβ42 in cerebrospinal fluid after severe brain injury, with some studies showing an increase and some a decrease (see Emmerling et al., 2000; Franz et al., 2003; and Olsson et al., 2004). The literature on amyloid changes in the brain is also confusing. In about one-third of brain samples surgically removed within hours of severe TBI, researchers have found diffuse Aβ plaques (see Ikonomovic et al., 2004). TBI brains containing plaques also show higher levels of soluble Aβ42 (see DeKosky et al., 2007). Brody and collaborators measured levels of extracellular Aβ in brain interstitial fluid taken from drainage catheters of trauma patients in intensive care, however, and found peptide levels were lower after the brain injury, when the injured brain’s activity was low, and gradually rose as the brain recovered and resumed activity (see ARF related news story on Brody et al., 2008; Schwetye et al., 2010; and reviewed in Magnoni and Brody, 2010).

“It is possible that in humans with TBI, extracellular amyloid-β levels are reduced in the soluble form, yet intracellular amyloid-β levels are increased in an insoluble form,” Brody said, emphasizing that the brain contains different pools of Aβ. “That would fit with all the available animal and human literature.” He notes there is no data as yet on the oligomerization status of the soluble Aβ, a key factor for determining its toxicity.

Despite the research to date, it is not clear how injury-associated amyloid release relates to AD. Do diffuse post-injury deposits go on to seed Alzheimer’s-like plaques? At least one study that examined TBI patients surviving up to three years after injury suggests not, finding that, despite the presence of large amounts of APP and Aβ, there were “virtually no” Aβ plaques in these brains (see Chen et al., 2008). This would suggest that post-injury deposits may be a fleeting phenomenon that the brain subsequently clears. Likewise, at autopsy, only about one-third of people who have had severe brain trauma show amyloid deposits, Brody said.

Mark Burns, at Georgetown University in Washington, DC, wrote to ARF, “This leads to a very confusing picture of why the brain becomes susceptible to AD after TBI. I suspect the answer may be that the brain has suffered a severe stress that leaves it primed and vulnerable to other stressors as time progresses.” This might lead to an earlier onset of AD, Burns suggested.

However, Sam Gandy at Mount Sinai Medical Center in New York City told ARF that he is convinced by the data that Aβ42 accumulation triggered by brain injury does play a role in susceptibility to AD. Cognitive reserve may be lost as well, Gandy wrote, both through amyloid-related and non-amyloid pathways.

Brody points to the need for more standardized, validated animal models of brain injury to sort all this out. Such models are in development (see, e.g., Brody et al., 2007), but much more work will be needed to answer these questions. To find out what role tau plays in chronically concussed brains, see Part 3.—Madolyn Bowman Rogers.

This is Part 2 of a six-part series. See also Part 1, Part 3, Part 4, Part 5, Part 6. See PDF of entire series.