Background: Sevoflurane produces direct vasodilation of coronary arter
ies in vitro and decreases coronary vascular resistance in vivo pharma
cologic properties that may contribute to the development of ''coronar
y steal.'' This investigation examined the effects of sevoflurane on t
he distribution of regional myocardial perfusion in chronically instru
mented dogs with steal-prone coronary artery anatomy. Methods: Dogs we
re chronically instrumented for measurement of aortic and left ventric
ular pressure, diastolic coronary blood flow velocity and subendocardi
al segment length. After recovery from surgery, dogs underwent repetit
ive, brief, left anterior descending coronary artery (LAD) occlusions
via an implanted hydraulic vascular occluder to enhance collateral dev
elopment. A progressive left circumflex coronary artery (LCCA) stenosi
s was also obtained using an ameroid constrictor. After development of
LCCA stenosis, the LAD was totally occluded to produce a model of mul
tivessel coronary artery disease. Systemic hemodynamics, regional cont
ractile function and myocardial perfusion measured with radioactive mi
crospheres were assessed in the conscious state and during sevoflurane
anesthesia at 1.0 and 1.5 MAC with and without restoration of arteria
l blood pressure and heart rate to conscious levels. Results: Total LA
D occlusion with simultaneous LCCA stenosis increased heart rate, mean
arterial pressure, left ventricular systolic and end-diastolic pressu
res, end-diastolic segment length, and rate-pressure product in consci
ous dogs. Subsequent administration of sevoflurane caused dose-related
decreases in arterial pressure, left ventricular systolic pressure, d
ouble product, and peak rate of increase of left ventricular pressure
at 50 mmHg. Perfusion of normal myocardium was unchanged during sevofl
urane anesthesia. In contrast, sevoflurane caused dose-dependent decre
ases in blood flow to myocardium supplied by the stenotic LCCA, which
returned to control levels after restoration of heart rate and arteria
l pressure. No reduction in collaterally derived blood flow to the occ
luded region was produced by 1.0 or 1.5 MAC sevoflurane. No redistribu
tion of blood flow away from the occluded LAD region to normal or sten
otic myocardium occurred during sevoflurane anesthesia. In fact, incre
ases in the ratio of blood flow between occluded and normal zones or o
ccluded and stenotic zones were observed in the subepicardium during 1
.5 MAC sevoflurane with maintenance of the heart rate and arterial pre
ssure at conscious levels. Conclusions: The results demonstrate that s
evoflurane does not reduce or abnormally redistribute myocardial blood
flow derived from coronary collateral vessels in a chronically instru
mented canine model of multivessel coronary artery obstruction.