Ml. Schroder et al., THRESHOLDS FOR CEREBRAL-ISCHEMIA AFTER SEVERE HEAD-INJURY - RELATIONSHIP WITH LATE CT FINDINGS AND OUTCOME, Journal of neurotrauma, 13(1), 1996, pp. 17-23
Cerebral ischemic insults occur in at least 30% of severely head injur
ed patients at a very early stage following trauma and are associated
with early death. To date, the threshold for ischemia of 18 mL/100 g/m
in used in human head injury studies has been adopted from animal stud
ies (by temporary occlusion of the middle cerebral artery). Since the
traumatized brain becomes more susceptible to irreversible damage if a
ccompanied by ischemia one may question whether the threshold for isch
emic vulnerability is higher than 18 mL/100 g/min. Cerebral ischemia c
an cause atrophy. Therefore, the authors obtained computerized tomogra
phy (CT) scans in 33 comatose head-injured patients (Glasgow Coma Scor
e of 8 or less) at least 3 months following injury and compared ventri
cle sizes (as a reflection of atrophy) with cerebral blood flow (CBF)
obtained within 4 h (average 2.3 +/- 0.8 h) after injury. Ventricular
measurements were performed in three fashions: the third ventricular s
ize (cm), the bicaudate cerebral ventricular index (BCVI), and the hem
ispheric ventricular index (HCVI). No significant correlation was foun
d between early CBF and any of the ventricle sizes. Applying a multipl
e correlation analysis with four independent parameters [CBF, CBF/time
postinjury, CBF/(time postinjury)(2), age], only age emerged as a sig
nificant indicator for predicting ventricle size (p < 0.001). We also
compared CBF data, obtained within 4 h after trauma, from survivors at
3 months after injury (mean CBF of 32 mL/100 g/min) with CBF data fro
m nonsurvivors (CBF 20 mL/100 g/min). The difference in CBF between su
rvivors and nonsurvivors was significant at p < 0.001 (Wilcoxon rank-s
um test). The proportion of patients with CBF less than or equal to 20
mL/100 g/min was 56% in the nonsurvivors and only 5% in survivors. Th
e difference in the proportions was significant at p < 0.001 (chi-squa
re test). We conclude that a measure of atrophy does not correlate wit
h ultra-early CBF. However, based on the clear distinction between sur
vivors and nonsurvivors, we suggest the threshold for ischemia after h
ead injury be redefined as a CBF of 20 mL/100 g/min.