A simple relationship between radiological arteriovenous malformation hemodynamics and clinical presentation: a prospective, blinded analysis of 31 cases
Js. Norris et al., A simple relationship between radiological arteriovenous malformation hemodynamics and clinical presentation: a prospective, blinded analysis of 31 cases, J NEUROSURG, 90(4), 1999, pp. 673-679
Object. The authors sought to establish prospectively whether there is a si
mple relationship between radiological features of brain arteriovenous malf
ormation (AVM) hemodynamics and a patient's clinical presentation.
Methods. Thirty-one consecutive patients with AVMs underwent cerebral angio
graphy at 3.8 frames/second during each standardized injection of contrast
material. Contrast dilution curves were derived from the image sequences by
using regions of interest (ROIs) traced on arteries feeding and veins drai
ning the AVM nidus. Angiographic parameters were then analyzed in a blinded
fashion. These parameters included the times required to reach the peak co
ntrast density, the contrast decay time, and fractions thereof, in the ROI
for each vessel. The authors determined whether these parameters, the arter
iovenous transit time, and/or AVM size were related to patients' presentati
on with hemorrhage (11 patients), seizure (11 patients), or other clinical
symptoms (nine patients). Statistically significant results were found only
in analyses of arterial phase times to reach peak contrast density. Analys
es of venous parameters, AVM size, and nidus transit time showed trends but
no statistical significance. Arterial filling with contrast material was s
ignificantly slower in patients presenting with hemorrhage (mean 50%, 80%,
and 100% of time to peak +/- standard error [SE] - 1.19 +/- 0.13, 1.97 +/-
0.18, and 3.04 +/- 0.34 seconds, respectively) compared with patients prese
nting with seizures (mean 50%, 80%, and 100% of time to peak +/- SE = 0.80
+/- 0.12, 1.32 +/- 0.18, and 1.95 +/- 0.29 seconds, respectively) according
to analysis of variance (p < 0.05) and post-hoc t-tests (p < 0.05) for eac
h parameter. Patients who presented with other symptoms had intermediate ar
terial filling times.
Conclusions. These simple hemodynamic parameters, which can be obtained wit
hout added risk to the patient, may help identify a subset of individuals i
n whom AVMs pose a higher risk of future hemorrhage and who may therefore w
arrant more expeditious treatment.