Cerebral Autoregulation in Severe Traumatic Brain Injury: Source of Index,Timing,and Relationship to Spreading Depolarization

Abstract

Impaired cerebral autoregulation is a contributor to secondary injury in traumatic brain injury (TBI) and is associated with worse outcomes. The mechanism of injury progression from impaired autoregulation is suggested to be related to triggering of spreading depolarization (SDs) events—massive waves of cellular dysfunction implicated in secondary injury. The most-studied continuous measurement of autoregulation is the pressure reactivity index (PRx), a rolling correlation between waveform level mean arterial pressure (MAP) and intracranial pressure. However, alternative indices using correlation of MAP with brain tissue oxygen tension (PbtO₂), probe-derived cerebral blood flow (CBF), and regional brain oxygen saturation (rSO₂) can also be performed and may have advantages. Our objective was to evaluate the predictive value of different multimodal autoregulation indices for 30-day mortality in TBI and to assess the relationship between impaired autoregulation and SD. Autoregulation indices including PRx, oxygen reactivity index (ORx; from PbtO₂), CBF reactivity (CBFRx; from thermal diffusion), and oxygen saturation reactivity index (OSRx; from rSO₂) were obtained in a cohort of 218 patients with TBI requiring invasive monitoring. Associations between autoregulation indices and 30-day mortality were evaluated using multivariable logistic regression models, adjusting for age and admission Glasgow Coma Scale (GCS). A subset of postsurgical patients with SD monitoring was evaluated to assess for a potential relationship between SD and autoregulation. Within the first 24 h, univariable analysis demonstrated significant associations between 30-day mortality and ORx (p = 0.0078) or OSRx (p = 0.0014), but not other indices. In multivariable models adjusting for age and admission GCS, both ORx (p = 0.044) and OSRx (p = 0.027) remained significant predictors of 30-day mortality. In the overall monitoring time, only ORx (p = 0.008) and PRx from the parenchymal monitor (p = 0.027) were significantly associated with mortality. In patients with SD monitoring, we noted an inflection in the predicted probability of SD below MAP 85 mmHg, which may be supportive of the lower limit of autoregulation being a threshold for increased SD probability. We conclude that early (first 24 h) impairment in ORx and OSRx is associated with worse outcomes in TBI, while ORx and PRx are more strongly associated throughout the window of invasive monitoring. This effect may be mediated by increased SD with impaired autoregulation. OSRx can be derived from noninvasive near-infrared spectroscopy monitoring. This presents an opportunity to assess and optimize impaired autoregulation earlier in a patient’s course without invasive monitoring. This could enhance outcome prediction and guide therapeutic strategies, particularly early after TBI.