Delayed CSF lavage for arteriographic and morphological vasospasm after experimental SAH.

Abstract

Irrigation of the subarachnoid space after aneurysmal subarachnoid hemorrhage (SAH) has been reported to alleviate subsequent arterial vasospasm. The authors have investigated the effect of lavage of the cerebrospinal fluid (CSF) space in the two-hemorrhage canine model of vasospasm. Twelve dogs had basilar cistern lavage with 120 cc of artificial CSF 24 hours after each of two SAH's, and 12 control dogs had two sequential SAH's without intervening lavage of clot. The amount of clot on the ventral brain stem was evaluated at sacrifice and was graded from 0 (no clot) to 4 (maximum clot) to assess the adequacy of clot removal. Dogs that had undergone lavage had a median grade of 1 (range Grade 0 to 2); control dogs had a median grade of 2 (range Grade 1 to 3.5, p less than 0.001. Wilcoxon rank sum test), indicating significant reduction of gross clot by lavage. The neurological findings were graded from 0 to 5, based on meningismus, ataxia, paresis, and cranial nerve deficits. No significant differences in neurological grade were found on any day between the two groups. Satisfactory angiograms were obtained before and 7 days after hemorrhage and were controlled for blood pressure and blood gases; these showed significant spasm in both groups. There was a mean reduction (+/- standard deviation) of 21.6% +/- 16.2% in basilar artery diameter in control dogs, compared to a 28.8% +/- 15.1% reduction in dogs with lavage (difference not significant, t-test). There was a strong, but insignificant, trend toward reduction of endothelial desquamation in the basilar and middle cerebral arteries in dogs with lavage compared to control animals (p = 0.06). Corrugation and tearing of the elastica, thickened intima, intimal fibroplasia, vacuolization of the endothelial or smooth-muscle cells, and presence of blood cells in the adventitia occurred similarly in both groups. It appears that cisternal lavage 24 hours after hemorrhage in this model has no effect on the angiographic, neurological, or most morphological sequelae of SAH, in spite of evidence for removal of clot as seen at sacrifice. Any postulated interaction of clot and vessel resulting in chronic vasospasm must occur before this time. Evaluation of the effect of much earlier lavage (for instance, 1 hour after hemorrhage) may elucidate the point at which vasospasm is instigated after SAH, and help in determining what factors cause vasospasm.