Objective: Impairment of cerebral autoregulation (CA) appears to be an important cause for secondary ischemia after subarachnoid hemorrhage (SAH). It has been shown that graded CA impairment is predictive of outcome. Little is known about whether such impairment is present, what causes CA impairment, whether it precedes vasospasm, and whether it is predictive of outcome in patients with severe aneurysmal SAH.
Design: Prospective, controlled study.
Setting: Neurosurgical intensive care unit.
Patients: Twelve patients after aneurysmal subarachnoid hemorrhage, 40 controls.
Interventions: Recording of cerebral blood flow velocities and continuous measurement of arterial blood pressure at a controlled ventilatory frequency of six per minute to standardize the influence of intrathoracic pressure changes on blood pressure.
Measurements and main results: We calculated the phase shift angles (deltaphidegrees) between slow (0.1 Hz) arterial blood pressure and cerebral blood flow velocity waves measured by transcranial Doppler ultrasound in the middle cerebral artery during a) posthemorrhage days (PHD) 1-6 (early or prevasospasm phase), and b) during PHD 7-13 (late or vasospasm phase) using a 6/min ventilation protocol, and in 40 controls spontaneously ventilating at the same rate. deltaphi <30 degrees indicated lost CA. Mean flow velocities >100 cm/sec were considered vasospasm. We combined early and late measurements to assess the CA relationship with low cerebral perfusion pressure (CPP) and/or vasospasm. We assessed the Glasgow Outcome Scale (GOS) score at discharge (1 = worst, 5 = best). The admission Hunt and Hess score was 3.6 +/- 0.7. GOS scores were n = 3 (GOS 1), n = 2 (GOS 2), n = 5 (GOS 3), n = 1 (GOS 4), and n = 1 (GOS 5). In the early phase, deltaphi was 40.4 +/- 19.8 degrees (left), and 40.4 +/- 19.2 degrees (right). CPP was 69.4 +/- 10.9, intracranial pressure (ICP) was 6.7 +/- 2.8 mm Hg. In the late phase, deltaphi worsened in six patients and none improved: 32.1 +/- 21 degrees (left), and 26.9 +/- 17.2 degrees (right); CPP was 68.1 +/- 12.1, ICP was 7.5 +/- 3.7 mm Hg. CA was significantly impaired in both phases when compared with normal subjects (deltaphi: 65.7 +/- 24.5 degrees; p < .01 for early, p < .001 for late phase). In the early phase, seven of eight patients in whom autoregulation was intact had a GOS >2 at discharge and disturbed CA on at least one side was predictive of either vegetative condition at discharge or death (p < .01). In the late phase, deltaphi was no longer predictive of outcome. Spasm was present in 8 of 17 vessels (47%) in which CA was lost; no spasm was found in 25 of 28 vessels (89%) in which CA was intact (p < .01). A low CPP was present in 6 of 17 vessels (35%) in which CA was lost; a normal CPP was found in 21 of 27 vessels (78%) in which CA was intact (p > .05, NS). However, 14 of 17 vessels (82%) with lost CA showed spasm and/or low CPP while only 8 of 27 cases (30%) with intact CA had either spasm or low CPP (p < .001).
Conclusions: CA can be assessed in a graded fashion in SAH patients. CA impairment precedes vasospasm; ongoing vasospasm worsens CA. CA assessment early after subarachnoid hemorrhage, within PHD 1-6, is predictive of outcome whereas late assessment is not. CA impairment is associated with cerebral vasospasm and low CPP.