Refining outcome prediction after traumatic brain injury with machine learning algorithms.

Outcome after traumatic brain injury (TBI) is typically assessed using the Glasgow outcome scale extended (GOSE) with levels from 1 (death) to 8 (upper good recovery). Outcome prediction has classically been dichotomized into either dead/alive or favorable/unfavorable outcome. Binary outcome prediction models limit the possibility of detecting subtle yet significant improvements. We set out to explore different machine learning methods with the purpose of mapping their predictions to the full 8 grade scale GOSE following TBI. The models were set up using the variables: age, GCS-motor score, pupillary reaction, and Marshall CT score. For model setup and internal validation, a total of 866 patients could be included. For external validation, a cohort of 369 patients were included from Leuven, Belgium, and a cohort of 573 patients from the US multi-center ProTECT III study. Our findings indicate that proportional odds logistic regression (POLR), random forest regression, and a neural network model achieved accuracy values of 0.3-0.35 when applied to internal data, compared to the random baseline which is 0.125 for eight categories. The models demonstrated satisfactory performance during external validation in the data from Leuven, however, their performance were not satisfactory when applied to the ProTECT III dataset.

Ibuprofen attenuates the inflammatory response and allows formation of migratory neuroblasts from grafted stem cells after traumatic brain injury.

PURPOSE: There is hope for neural stem and progenitor cells (NSPC) to enhance regeneration when transplanted to the injured brain after traumatic brain injury (TBI). So far, the therapeutic effects of NSPC transplantation have been hampered mainly by the notable death of the transplanted cells. Neuroinflammation may lead to additional cell death after TBI and we hypothesized that survival of grafted NSPC could be enhanced by anti-inflammatory treatment. METHODS: Mice that were subjected to controlled cortical impact TBI and grafted with NSPC, were treated with the non-steroidal anti-inflammatory drug ibuprofen. RESULTS: Ibuprofen was found to down-regulate the TBI-induced inflammatory response. In addition, migrating neuroblasts from transplanted cells were observed near the contusion and in the ipsilateral hippocampus in ibuprofen-treated animals only, suggesting that the anti-inflammatory treatment had beneficial effects on graft survival and/or differentiation. However, Morris Water Maze performance or TBI-induced tissue loss was not influenced by ibuprofen treatment. CONCLUSIONS: Our data suggests that anti-inflammatory strategies may be a complement to enhance the outcome for the cell transplants following TBI.

Functional outcome is impaired following traumatic brain injury in aging Nogo-A/B-deficient mice.

Increasing age is associated with a poor prognosis following traumatic brain injury (TBI). CNS axons may recover poorly following TBI due to expression of myelin-derived inhibitors to axonal outgrowth such as Nogo-A. To study the role of Nogo-A/B in the pathophysiological response of the elderly to TBI, 1-year-old mice deficient in Nogo-A/B (Nogo-A/B homozygous(-/-) mice), Nogo-A/B heterozygous(-/+) mice, and age-matched wild-type (WT) littermate controls were subjected to a controlled cortical impact (CCI) TBI. Sham-injured WT mice (7 months old) and 12 month old naïve Nogo-A/B(-/-) and Nogo-A/B(-/+) served as controls. Neurological motor function was evaluated up to 3 weeks, and cognitive function, hemispheric tissue loss, myelin staining and hippocampal beta-amyloid (A beta) immunohistochemistry were evaluated at 4 weeks post-injury. In WT littermates, TBI significantly impaired learning ability at 4 weeks and neurological motor function up to 2 weeks post-injury and caused a significant loss of hemispheric tissue. Following TBI, Nogo-A/B(-/-) mice showed significantly less recovery from neurological motor and cognitive deficits compared to brain-injured WT mice. Naïve Nogo-A/B(-/-) and Nogo-A/B(-/+) mice quickly learned the MWM task in contrast to brain-injured Nogo-A/B(-/-) mice who failed to learn the MWM task at 4 weeks post-injury. Hemispheric tissue loss and cortical lesion volume were similar among the brain-injured genotypes. Neither TBI nor the absence of NogoA/B caused an increased A beta expression. Myelin staining showed a reduced area and density in the corpus callosum in brain-injured Nogo-A/B(-/-) animals compared to their littermate controls. These novel and unexpected behavioral results demonstrate that the absence of Nogo-A/B may negatively influence outcome, possibly related to hypomyelination, following TBI in mice and suggest a complex role for this myelin-associated axonal growth inhibitor following TBI.