Authors :
Presenting Author: Joshua Laing, BBiomedSci(hons) MBBS PhD – Monash University
Zhibin Chen, PhD – Monash Univeristy
Belinda Gabbe, PhD – Monash University
Piero Perucca, MD PhD – Univeristy of Melbourne
Patrick Kwan, MD PhD – Monash University
Terence J O'Brien, MBBS MD – School of Translational Medicine, Monash University, The Alfred Centre
Rationale:
Post-traumatic epilepsy (PTE) is a serious long-term complication of traumatic brain injury (TBI) that is associated with worse psychosocial, medical and mortality outcomes than patients with TBI who do not develop epilepsy. Identification of risk factors for PTE has potential to delineate a subgroup of individuals with TBI who may benefit from antiepileptogenic therapies, but currently there are no accurate methods to predict PTE in individual patients.
Methods:
21,210 individuals with moderate to severe TBI as per the abbreviated injury scale (AIS) were identified via a population-based Victorian State Trauma Registry from Jan 2005 to June 2023, with 2-year follow up data. Patients with PTE were identified from a subgroup of 3,782 participants with moderate to severe TBI who completed a structured interview for the development of new onset seizures. Patients with a diagnosis of epilepsy prior to TBI were excluded. Patient and injury characteristics were collected, as well as 2-year outcome measures including quality of life (EQ-5D-5L). LASSO regression was applied to predict important clinical risk factors for PTE. Post-discharge deaths were identified using the Victorian Death Index.
Results:
531 (14%) of the 3,782 individuals with moderate to severe TBI had developed PTE. Regression analysis identified that PTE was associated with younger age, early posttraumatic seizures, low height falls, medical comorbidities (CCI2+), alcohol misuse, subarachnoid haemorrhage, intracranial haemorrhage, higher injury severity scores as measured by AIS head and initial GCS, and need for neurosurgical intervention. Using the identified risk factors above during training, the prediction model had an area under the receiver operating curve of 0.75 (95% CI: 0.69-0.80) for the test set, with a sensitivity and specificity of the model of 59% and 80% respectively. The Glasgow Outcome Scale showed lower odds of a better recovery in those with PTE. The likelihood of reporting poor activity, pain and discomfort, and anxiety and depression were higher in those with PTE. When corrected for confounding risk factors, PTE had a 3% lower overall utility score for health-related quality of life. From the development of PTE, those with PTE had a 2.9 times higher risk of mortality than those without PTE (median follow up 2.86 years).
Conclusions:
PTE following moderate to severe TBI is prevalent and associated with reliably identifiable important clinical risk factors. Future interventional trials of antiepileptogenic treatments may benefit from these clinical selection criteria. Once established, PTE is independently associated with worsened morbidity and overall mortality, compared to those without PTE.
Funding: No specific support was received in support of this abstract.