The postoperative angiograms in 66 patients who underwent craniotomy f
or clipping of 78 cerebral aneurysms were reviewed. Indications for ur
gent postoperative angiography included neurological deficit or repeat
subarachnoid hemorrhage. Routine postoperative angiograms were carrie
d out in the remaining patients. Postoperative angiograms were reviewe
d to determine the incidence of unexpected findings such as unclipped
aneurysms, residual aneurysms, and unforeseen major vessel occlusions.
Logistic regression analysis was used to test if the following were f
actors that predicted an unexpected finding on postoperative angiograp
hy: aneurysm site or size; the intraoperative impression that residual
aneurysm was left or a major vessel was occluded; intraoperative aneu
rysm rupture; opening or needle aspiration of the aneurysm after clipp
ing; or development of a new neurological deficit after surgery. Kappa
values were calculated to assess the agreement between some of these
clinical factors and unexpected angiographic findings. Unexpected resi
dual aneurysms were seen in three (4%) of the 78 occlusions. In additi
on, three aneurysms were completely unclipped (4%); these three patien
ts were returned to the operating room and had their aneurysms success
fully obliterated. There were nine unexpected major vessel occlusions
(12%); six of these resulted in disabling stroke and two patients died
. Of six major arteries considered to be occluded intraoperatively and
shown to be occluded by postoperative angiography, two were associate
d with cerebral infarction. Logistic regression analysis showed that a
new postoperative neurological deficit predicted an unforeseen vessel
occlusion on postoperative angiography. Factors could not be identifi
ed that predicted unexpected residual aneurysm or unclipped aneurysm.
The inability to predict accurately the presence of residual or unclip
ped aneurysm suggests that all patients should undergo postoperative a
ngiography. Since a new postoperative neurological deficit is one fact
or predicting unexpected arterial occlusion, intraoperative angiograph
y may be necessary to help reduce the incidence of stroke after aneury
sm surgery. With study of more patients or of factors not examined in
this series, it may be possible to select cases more accurately for in
traoperative or postoperative angiography.