Jernej Avsenik1, Fajko F Bajrović, Primož Gradišek, Nataša Kejžar, Katarina Šurlan Popović. 1. From the Clinical Institute of Radiology (J.A., K.Š.P.), University Medical Centre Ljubljana; Department of Radiology (J.A., K.Š.P.), Faculty of Medicine, University of Ljubljana; Division of Neurology (F.F.B.), University Medical Centre Ljubljana; Institute of Pathophysiology (F.F.B.), Faculty of Medicine, University of Ljubljana; Clinical Department of Anaesthesiology and Intensive Therapy (P.G.), Centre for Intensive Therapy, University Medical Centre Ljubljana; Department of Anaesthesiology with Reanimatology (P.G.), Faculty of Medicine, University of Ljubljana and Institute for Biostatistics and Medical Informatics (N.K.), Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia.
Abstract
BACKGROUND: Currently established prognostic models in traumatic brain injury (TBI) include noncontrast computed tomography (CT) which is insensitive to early perfusion alterations associated with secondary brain injury. Perfusion CT (PCT) on the other hand offers insight into early perfusion abnormalities. We hypothesized that adding CT perfusion and permeability data to the established outcome predictors improves the performance of the prognostic model. METHODS: A prospective cohort study of consecutive 50 adult patients with head injury and Glasgow Coma Scale score of 12 or less was performed at a single Level 1 Trauma Centre. Perfusion CT was added to routine control CT 12 hours to 24 hours after admission. Region of interest analysis was performed in six major vascular territories on perfusion and permeability parametric maps. Glasgow Outcome Scale (GOS) was used 6 months later to categorize patients' functional outcomes to favorable (GOS score > 3) or unfavorable (GOS score ≤ 3). We defined core prognostic model, consisting of age, motor Glasgow Coma Scale score, pupillary reactivity, and CT Rotterdam Score. Next, we added perfusion and permeability data as predictors and compared updated models to the core model using cross-validated areas under the receiver operator curves (cv-AUC). RESULTS: Significant advantage over core model was shown by the model, containing both mean cerebral extravascular-extracellular volume per unit of tissue volume and cerebral blood volume of the least perfused arterial territory in addition to core predictors (cv-AUC, 0.75; 95% confidence interval, 0.51-0.84 vs. 0.6; 95% confidence interval, 0.37-0.74). CONCLUSION: The development of cerebral ischemia and traumatic cerebral edema constitutes the secondary brain injury and represents the target for therapeutic interventions. Our results suggest that adding CT perfusion and permeability data to the established outcome predictors improves the performance of the prognostic model in the setting of moderate and severe TBI. LEVEL OF EVIDENCE: Prognostic study, level III.
BACKGROUND: Currently established prognostic models in traumatic brain injury (TBI) include noncontrast computed tomography (CT) which is insensitive to early perfusion alterations associated with secondary brain injury. Perfusion CT (PCT) on the other hand offers insight into early perfusion abnormalities. We hypothesized that adding CT perfusion and permeability data to the established outcome predictors improves the performance of the prognostic model. METHODS: A prospective cohort study of consecutive 50 adult patients with head injury and Glasgow Coma Scale score of 12 or less was performed at a single Level 1 Trauma Centre. Perfusion CT was added to routine control CT 12 hours to 24 hours after admission. Region of interest analysis was performed in six major vascular territories on perfusion and permeability parametric maps. Glasgow Outcome Scale (GOS) was used 6 months later to categorize patients' functional outcomes to favorable (GOS score > 3) or unfavorable (GOS score ≤ 3). We defined core prognostic model, consisting of age, motor Glasgow Coma Scale score, pupillary reactivity, and CT Rotterdam Score. Next, we added perfusion and permeability data as predictors and compared updated models to the core model using cross-validated areas under the receiver operator curves (cv-AUC). RESULTS: Significant advantage over core model was shown by the model, containing both mean cerebral extravascular-extracellular volume per unit of tissue volume and cerebral blood volume of the least perfused arterial territory in addition to core predictors (cv-AUC, 0.75; 95% confidence interval, 0.51-0.84 vs. 0.6; 95% confidence interval, 0.37-0.74). CONCLUSION: The development of cerebral ischemia and traumatic cerebral edema constitutes the secondary brain injury and represents the target for therapeutic interventions. Our results suggest that adding CT perfusion and permeability data to the established outcome predictors improves the performance of the prognostic model in the setting of moderate and severe TBI. LEVEL OF EVIDENCE: Prognostic study, level III.