Objective: Hemoglobin contributes to vasospasm after subarachnoid hemorrhage. One mechanism may involve binding of nitric oxide, destruction of nitric oxide, or both. Support for this mechanism would be evidence that nitric oxide donors prevent vasospasm. This study attempted to provide such evidence.
Methods: A randomized, blinded study was conducted in which 13 monkeys underwent cerebral angiography and creation of a right subarachnoid hemorrhage. Subcutaneous osmotic pumps were implanted to deliver sodium nitroprusside (n = 7) or vehicle (n = 6) via catheters into the right basal cisterns. Seven days later, angiography was repeated, and the animals were humanely killed. Levels of cyclic nucleotides, hemoglobins, and thiocyanate were measured.
Results: Significant vasospasm of the right middle cerebral artery was present in animals treated with sodium nitroprusside (35 +/- 22% reduction in diameter, P < 0.05, paired t test) and placebo (28 +/- 20% reduction, P < 0.05, not significantly different from nitroprusside group by unpaired t test). Adequate delivery of sodium nitroprusside was supported by the finding of a significant increase in cyclic guanosine monophosphate levels in the cerebral arteries of treated animals compared with placebo (P < 0.05, unpaired t test). Thiocyanate was not present in significantly increased amounts in animals treated with nitroprusside, although this group did display elevated concentrations of nitrosyl hemoglobin (measured by electron paramagnetic resonance spectroscopy) and cyanomethemoglobin (measured by spectrophotometry) in the cerebrospinal fluid on Day 7.
Conclusion: The lack of effect of sodium nitroprusside was not the result of inadequate drug delivery because cyclic guanosine monophosphate levels were significantly increased in vasospastic arteries. Vasospasm may not have been prevented because of a toxic effect of sodium nitroprusside metabolites, involvement of smooth muscle relaxation or contraction processes downstream of cyclic guanosine monophosphate, or both.