The severity of head injury in traumatic brain injury (TBI) is significantly linked with the risk of developing posttraumatic epilepsy (PTE) and seizures, and PTE itself further worsens outcomes at 12 months, findings from an analysis of a large, prospective database suggest.
“We found that patients essentially have a 10-times greater risk of developing posttraumatic epilepsy and seizures at 12 months (post-injury) if the presenting Glasgow Coma Scale (GCS) is less than 8,” said first author John F. Burke, MD, PhD, University of California, San Francisco, in presenting the findings as part of the virtual edition of the American Association of Neurological Surgeons Annual Meeting.
While posttraumatic epilepsy represents an estimated 20% of all cases of symptomatic epilepsy, many questions remain on those most at risk and on the long-term effects of posttraumatic epilepsy on TBI outcomes.
To probe those issues, Burke and colleagues turned to the multicenter TRACK-TBI database, which has prospective, longitudinal data on more than 2700 patients with traumatic brain injuries and is considered the largest source of prospective data on posttraumatic epilepsy.
Using the criteria of no previous epilepsy and having 12 months of follow-up, the team identified 1493 patients with TBI. In addition, investigators identified 182 orthopedic controls (included and prospectively followed because they have injuries but not specifically head trauma) and 210 controls who are friends of the patients and who do not have injuries but allow researchers to control for socio-economic and environmental factors.
Of the 1493 patients with TBI, 41 (2.7%) were determined to have posttraumatic epilepsy, assessed according to a NINDS epilepsy screening questionnaire, which is designed to identify patients with PTE symptoms. There were no reports of epilepsy symptoms using the screening tool among the controls.
Burke noted that the 2.7% was in agreement with historical reports.
In comparing patients with TBI who did and did not have posttraumatic epilepsy, no differences were observed in the groups in terms of gender, although there was a trend toward younger age among those with PTE (mean age 35.4 years with posttraumatic injury vs 41.5 without; P = .05).
A major risk factor for the development of posttraumatic epilepsy was presenting GCS scores.
Among those with scores of less than 8, indicative of severe injury, the rate of posttraumatic epilepsy was 6% at 6 months and 12.5% at 12 months.
In contrast, those with TBI presenting with GCS scores between 13 and 15, indicative of minor injury, had an incidence of posttraumatic epilepsy of 0.9% at 6 months and 1.4% at 12 months.
Imaging findings in the two groups showed that hemorrhage detected on CT imaging was associated with a significantly higher risk for posttraumatic epilepsy (P < .001).
“The main take-way is that any hemorrhage in the brain is a major risk factor for developing seizures,” Burke said. “Whether it is subdural, epidural blood, subarachnoid or contusion, any blood confers a very [high] risk for developing seizures.”
Posttraumatic epilepsy was linked to poorer longer-term outcomes even for patients with lesser injury: Among those with TBI and GCS of 13 to 15, the mean Glasgow Outcome Scale Extended (GOSE) score at 12 months among those without posttraumatic epilepsy was 7, indicative of a good recovery with minor defects, whereas the mean GOSE score for those with PTE was 4.6, indicative of moderate to severe disability (P < .001).
“It was surprising to us that PTE-positive patients had a very significant decrease in GOSE compared to PTE-negative patients,” Burke said. “There was a nearly 2-point drop in the GOSE and that was extremely significant.”
A multivariate analysis showed there was still a significant independent risk for a poor GOSE score with posttraumatic epilepsy after controlling for GCS score, head CT findings, and age (P < .001).
The authors also looked at mood outcomes using the Brief Symptom Inventory-18 (BSI-18), which showed significant worse effect in those with posttraumatic epilepsy after multivariate adjustment (P = .01). Additionally, a highly significant worse effect in cognitive outcomes on the Rivermead cognitive metric was observed with posttraumatic epilepsy (P = .001).
“On all metrics tested, posttraumatic epilepsy worsened outcomes,” Burke said.
He noted that the study has some key limitations, including the 12-month follow-up. (A previous study showed a linear increase in posttraumatic follow-up up to 30 years.)
“The fact that we found 41 patients at 12 months indicates there are probably more that are out there who are going to develop seizures, but because we don’t have the follow-up we can’t look at that,” he said.
Although the screening questionnaires are effective, “the issue is these people are not being seen by an epileptologist or having scalp EEG done, and we need a more accurate way to do this.”
A new study, TRACK-TBI EPI, will address those limitations and a host of other issues, with a 5-year follow-up, he noted.
Commenting on the study as a discussant, neurosurgeon Uzma Samadani, MD, PhD, of the Minneapolis VA Medical Center and CentraCare in Minneapolis, suggested that the future work should focus on issues including the wide-ranging mechanisms that could explain the seizure activity.
“For example, it’s known that posttraumatic epilepsy or seizures can be triggered by abnormal conductivity due to multiple different mechanisms associated with brain injury, such as endocrine dysfunction, cortical spreading depression, and many others,” said Samadani, who has been a researcher on the TRACK-TBI study.
Factors ranging from genetic differences to comorbid conditions such as alcoholism can play a role in brain injury susceptibility, Samadani added.
Furthermore, outcome measures currently available simply may not capture the unknown nuances of brain injury, she said.
“We have to ask, are these an all-or-none phenomena, or is aberrant electrical activity after brain injury a continuum of dysfunction?” Samadani speculated.
“I would caution that we are likely underestimating the non-easily measurable consequences of brain injury,” she said. “And the better we can quantitate susceptibility, classify the nature of injury and target acute management, the less posttraumatic epilepsy/aberrant electrical activity our patients will have.”
Burke and Samadani have disclosed no relevant financial relationships.
Virtual edition of the American Association of Neurological Surgeons Annual Meeting: Abstract 526.