Understanding long-term outcomes in traumatic injury

Traumatic brain injury (TBI) is a complex and impactful medical condition, capable of altering the life of a patient long after the initial trauma. This thesis strives to increase our understanding of long-term outcomes following TBI, from mild to severe cases. The studies aimed to explore predictive factors that influence patient recovery and to evaluate potential treatments targeting the chronic neuroinflammation that follows such injuries.

The research project started with an experimental study (Study I) using a rodent model to assess the chronic inflammatory response after a penetrating brain injury. We employed a treatment known as Resolvin D1, a lipid mediator derived from omega-3 fatty acids, hypothesized to aid in resolving inflammation and promoting tissue repair. Although results did not show significant changes in inflammation levels or tissue loss between treated and control groups, the study provided important insights into the timeline and complexities of chronic inflammation in TBI.

In the clinical studies (Studies II-IV), we examined data from trauma registries to evaluate patient outcomes following TBI. One key focus was the relationship between pre-injury health, as measured by the American Society of Anesthesiologists (ASA) score, and patient outcomes. The ASA-score is based on the burden of comorbidities and rates a patient's overall health status.

The second study (first clinical study) explored health-related quality of life (HRQoL) two years post-TBI, an often overlooked aspect of recovery. Using tools like the RAND-36 and EQ-5D questionnaires, we assessed how 170 trauma patients perceived their physical, emotional, and social health over the long term. Interestingly, we found that TBI patients sometimes reported better HRQoL outcomes than those who experienced non-TBI (NTBI) trauma, particularly in physical functioning and daily role limitations. This could be attributed to factors such as cognitive biases, reduced expectations due to aging, or an impaired ability to fully recognize deficits after brain injury. A high ASA-score (indicating worse health before the injury) was strongly associated with a reduced HRQoL in both the TBI and NTBI cohort. Increased injury severity, measured with the head value of the abbreviated injury score (AIS), showed a trend to association with lower HRQoL, but this was not statistically significant. We found no difference in symptoms of depression, assessed with the self-assessed Montgomary-Åsberg depression score (MADRS-S), between any of the groups.

In the third study, we included 823 trauma patients and examined 90-day mortality after mild TBI with intracranial findings (complicated mTBI) compared to NTBI. We found ASA-score to be strongly associated with higher mortality rates, independent from age and injury severity, an association we did not see in the NTBI cohort. This difference could be either due to a treatment bias and under-triage of mTBI patients, a selection bias excluding frail NTBI patients, or that the ASA-score captures an important vulnerability in mTBI patients. AIS (head) was not independently associated with mortality. We also tested the predictive value of the Trauma and Injury Severity Score (TRISS). The accuracy of TRISS was very low, which suggests that this score is not useful in milder trauma.

In the fourth and final study we tested the added value of ASA-score to the well- established International Mission for Prognosis and Analysis of Clinical Trials in TBI (IMPACT) model, on a cohort of 720 patients with moderate to severe TBI (msTBI). Here we confirmed a strong and independent association between ASA-score and both 90-day mortality and 1-year functional outcome (measured with Glasgow Outcome Score). The inclusion of ASA-scoring yielded a significantly increased estimated explained variance of the already comprehensive IMPACT model. Patients with an ASA-score of 3 or above had a pronounced increase in mortality compared to healthier individuals, which held true even in younger patients. In this cohort, TRISS proved to be strongly associated with GOS, and even more to mortality, highlighting the importance of the overall burden of injury.

While these studies provided valuable insights, limitations such as the lack of baseline HRQoL data and the challenges in differentiating chronic inflammation types in experimental models point to avenues for future research. The findings stress the necessity of early and tailored interventions to improve long-term outcomes and reinforce the importance of considering a patient's overall health in management and outcome prediction.

By investigating both the biological underpinnings and clinical outcomes of TBI, this work aims to contribute to more refined and effective approaches to treatment and rehabilitation. Understanding these long-term trajectories can enhance care protocols, allowing for a more personalised approach that accounts for individual health profiles, ultimately improving patient outcomes and quality of life following TBI.

List of scientific papers

I. Kiwanuka O, Cheung J, Hånell A, Lindblad C. Immunological response and tissue loss in a rodent model of chronic traumatic brain injury treated with resolvin. [Manuscript]

II. Kiwanuka O, Lassarén P, Thelin EP, Hånell A, Sandblom G, Fagerdahl A, Boström L. Long-term health-related quality of life after trauma with and without traumatic brain injury: a prospective cohort study. Sci Rep. 2023. 13(1): p. 2986. https://doi.org/10.1038/s41598-023-30082-4

III. Kiwanuka O, Lassarén P, Hånell A, Boström L, Thelin EP. ASA-score is associated with 90-day mortality after complicated mild traumatic brain injury - a retrospective cohort study. Acta Neurochir (Wien). 2024. 166(1): p. 363. https://doi.org/10.1007/s00701-024-06247-z

IV. Kiwanuka O, Lassarén P, Fletcher-Sandersjöö A, Tatter C, Tjerkaski J, Nelson DW, Thelin EP. ASA-score is an independent predictor of 1-year outcome after moderate-to-severe Traumatic Brain Injury. Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine. (2025) 33:25. https://doi.org/10.1186/s13049-025-01338-x