Optimal cerebrospinal fluid magnesium ion concentration for vasodilatory effect and duration after intracisternal injection of magnesium sulfate solution in a canine subarachnoid hemorrhage model.

OBJECT: The optimal CSF Mg(++) concentration for vasodilation of spastic cerebral arteries after subarachnoid hemorrhage (SAH) and its duration are unknown. The temporal profile of the vasodilatory effect and optimal CSF Mg(++) concentration after the intracisternal injection of MgSO(4) solution were investigated in an SAH model in canines.

METHODS

Cerebral vasospasm was induced by experimental SAH using a 2-hemorrhage model in 26 female beagles. On Day 7, 0.5 ml/kg of 15, 10, 5, or 0 mmol/L MgSO(4) in Ringer solution was injected into the cerebellomedullary cistern. Angiography was performed on Day 1 (before SAH) and before and 1, 3, and 6 hours after the intracisternal injection on Day 7 to measure arterial diameters of the basilar artery (BA), superior cerebellar artery (SCA), and vertebral artery (VA). Cerebrospinal fluid Mg(++) was also measured at the same time.

RESULTS

Arterial diameters of the BA, SCA, and VA were significantly decreased by vasospasm on Day 7. Arterial diameter ratios (ratio of arterial diameter after MgSO(4) injection to diameter before injection on Day 7) of the BA and SCA at 1 and 3 hours after and the VA at 1 hour after intracisternal injection of the MgSO(4) solution were positively correlated with the CSF Mg(++) concentration. All arterial diameter ratios, except 1 point of the SCA, exceeded 1 if the CSF Mg(++) concentration was > 3 mEq/L at 1 hour after injection. Animals with CSF Mg(++) concentrations > 3 mEq/L at 1 hour after injection (11 dogs) showed significantly increased arterial diameters of the BA at 1 and 3 hours after and of the SCA and VA at 1, 3, and 6 hours after injection, as compared with the diameters before injection. The CSF Mg(++) concentration significantly increased at 1 hour (3.73 ± 0.69 mEq/L, p < 0.01) and 3 hours (2.05 ± 0.35 mEq/L, p < 0.01) after the intracisternal injection as compared with the baseline value (1.41 ± 0.20 mEq/L).

CONCLUSIONS

The reversible effect of an intracisternal injection of MgSO(4) solution on the spastic artery requires CSF Mg(++) concentrations > 3 mEq/L. The vasodilatory effect continues for 3-6 hours after injection. These results suggest that the continuous infusion or intermittent intracisternal injection of MgSO(4) is needed to maintain the optimal CSF Mg(++) concentration and constantly ameliorate cerebral vasospasm.