Intracranial pressure (ICP) increases following traumatic brain injury (TBI) are multifactorial, and may be due to cytotoxic oedema, vasogenic oedema and/or vascular engorgement. We addressed the issue of the contribution of vascular engorgement to ICP elevation by directly measuring cerebral blood volume (CBV) with 15O positron emission tomography (15O-PET) in well-characterised patients with head injury.
We undertook 61 15O-PET studies with measurement of CBF and CBV; ten in healthy volunteers and 51 in patients with TBI (15 within 24 hours, and 36 at 2 – 6 days post injury). The median (range) admission Glasgow Coma Score (GCS) in patients was 7(3–13), but, at the stage when imaging was undertaken, all patients required sedation and/or ventilatory support for ICP elevation or reductions in GCS to < 8. Patients received protocol driven therapy aimed at maintaining ICP < 20 mmHg, cerebral perfusion pressure (CPP) > 70 mmHg, and PaCO2 at ∼ 4.5 kPa.
PET studies were undertaken using steady-state techniques. Parametric maps of CBF, CBV, CMRO2 and OEF were calculated using a blood-brain partition coefficient for H215O of 0.95 and a small to large vessel haematocrit ratio of 0.85. A standardised map with 12 regions of interest (ROI) was constructed for the supratentorial compartment, based on vascular territories and anatomical structures.
Patients had a mean (± SD) ICP of 17 ± 6 mmHg, a CPP of 74 ± 8 mmHg, and a significantly lower PaCO2 than controls (4.5 ± 0.4 vs. 5.5 ± 0.4 kPa; p < 0.001). Despite these differences in PaCO2, when compared to controls, mean CBV in patient ROIs was significantly higher, both before and after 24 hours post injury (3.3 ± 0.8 ml/100 ml vs. 3.8 ± 0.7 and 3.9 ± 0.7 ml/100 ml, respectively; p < 0.01 at both time points). Differences between controls and patient groups were preserved when values were compared after averaging across all ROIs in individual patients (p < 0.05 and < 0.001, for early and late head injury, respectively). CBV was directly, rather than inversely, related to CBF at both time points (R: 0.56 and 0.48, respectively; p < 0.0001 for both). These findings suggest dysautoregulation or metabolically coupled changes in CBF, rather than autoregulatory vasodilatation in response to classical macrovascular ischaemia. CBV was inversely related to ICP (R: 0.39; p < 0.01), suggesting that vasodilatation was not the dominant cause of intracranial hypertension across the population of patients studied.
We show significant increases in CBV following TBI; both within and after 24 hours post TBI. While previous reports of low CBV in head injury may have been confounded by the indirect estimation of CBV, 1 our data support previous suggestions 1 that CBV increases may not be the dominant cause of intracranial hypertension in all subjects. Our findings are fully consistent with microvascular ischaemia, but the concordance of high CBF and CBV values suggests that CBV increases are mainly the consequence of abnormal vascular function, rather than macrovascular ischaemia.