The oxygen reactivity index indicates disturbed local perfusion regulation after aneurysmal subarachnoid hemorrhage: an observational cohort study

Background Cerebral autoregulation (CA) can be impaired in patients with delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH). The Pressure Reactivity Index (PRx, correlation of blood pressure and intracranial pressure) and Oxygen Reactivity Index (ORx, correlation of cerebral perfusion pressure and brain tissue oxygenation, PbtO2) are both believed to estimate CA. We hypothesized that CA could be poorer in hypoperfused territories during DCI and that ORx and PRx may not be equally effective in detecting such local variances. Methods ORx and PRx were compared daily in 76 patients with aSAH with or without DCI until the time of DCI diagnosis. The ICP/PbtO2-probes of DCI patients were retrospectively stratified by being in or outside areas of hypoperfusion via CT perfusion image, resulting in three groups: DCI + /probe + (DCI patients, probe located inside the hypoperfused area), DCI + /probe− (probe outside the hypoperfused area), DCI− (no DCI). Results PRx and ORx were not correlated (r = − 0.01, p = 0.56). Mean ORx but not PRx was highest when the probe was located in a hypoperfused area (ORx DCI + /probe + 0.28 ± 0.13 vs. DCI + /probe− 0.18 ± 0.15, p < 0.05; PRx DCI + /probe + 0.12 ± 0.17 vs. DCI + /probe− 0.06 ± 0.20, p = 0.35). PRx detected poorer autoregulation during the early phase with relatively higher ICP (days 1–3 after hemorrhage) but did not differentiate the three groups on the following days when ICP was lower on average. ORx was higher in the DCI + /probe + group than in the other two groups from day 3 onward. ORx and PRx did not differ between patients with DCI, whose probe was located elsewhere, and patients without DCI (ORx DCI + /probe− 0.18 ± 0.15 vs. DCI− 0.20 ± 0.14; p = 0.50; PRx DCI + /probe− 0.06 ± 0.20 vs. DCI− 0.08 ± 0.17, p = 0.35). Conclusions PRx and ORx are not interchangeable measures of autoregulation, as they likely measure different homeostatic mechanisms. PRx represents the classical cerebrovascular reactivity and might be better suited to detect disturbed autoregulation during phases with moderately elevated ICP. Autoregulation may be poorer in territories affected by DCI. These local perfusion disturbances leading up to DCI may be more readily detected by ORx than PRx. Further research should investigate their robustness to detect DCI and to serve as a basis for autoregulation-targeted treatment after aSAH.