Neurobehavioral phenotypes may identify individuals with TBI at risk for poor outcomes

Qualitatively distinct profiles of symptoms and cognitive functioning in patients with traumatic brain injury may identify those at risk for poor outcomes, according to results of a prospective cohort study published in JAMA Network Open.

“It has been very difficult to develop treatments that help patients recover from traumatic brain injuries, and we believe one reason is that there is a lot of variability in the effects of brain injuries across patients,” Lindsay D. Nelson, PhD, associate professor of neurosurgery at Medical College of Wisconsin, told Healio Neurology. “Prior treatment studies have used a one-size-fits-all approach, treating diverse patients the same way. In this study, we were looking for subgroups of TBI patients that have distinct profiles of clinical signs and symptoms, which we believe will help us develop more personalized treatments for TBI.”

Specifically, Nelson and colleagues sought to assess whether distinct neurobehavioral phenotypes were identifiable 2 weeks after TBI, as well as to determine the extent to which early neurobehavioral phenotypes affect 6-month outcomes. They analyzed data from 1,757 patients who presented to 18 U.S. level I trauma centers within 24 hours of TBI between 2014 and 2019, as part of the Transforming Research and Clinical Knowledge in TBI study. The patients had mean age of 39.9 years, and 67.4% were men. The researchers derived latent profiles using common dimensions of neurobehavioral functioning at 2 weeks after injury, which they evaluated via the NIH TBI Common Data Elements. Satisfaction With Life Scale (SWLS), Quality of Life after Brain Injury-Overall Scale (QOLIBRI-OS), Glasgow Outcome Scale-Extended (GOSE) and Rivermead Post-Concussion Symptoms Questionnaire (RPQ) served as 6-month outcomes.

Results of the latent profile analysis showed the following distinct neurobehavioral phenotypes at 2 weeks after injury:

• emotionally resilient (n = 419);
• cognitively impaired (n = 368);
• cognitively resilient (n = 620); and
• neuropsychiatrically distressed (n = 350).

Combining the latent profile group with models that included demographic characteristics, medical history, Glasgow Coma Scale score and other injury characteristics significantly improved the estimation of the association with 6-month outcomes, with R² increases of between 0.09 and 0.19 on the GOSE, 0.12 and 0.22 on the SWLS, 0.14 and 0.32 on the QOLIBRI-OS and 0.13 and 0.34 on the RPQ.

“Incorporating information about the clinical subgroups identified in this study significantly improved our ability to predict patients’ later outcomes,” Nelson said. “This information could help clinicians provide more accurate prognoses to patients.”