This study aims to test the assumptions that: (1) coronary arterial fl
ow is attenuated in an early activated region by ventricular pacing; (
2) asynchronous mechanical activation caused by ventricular pacing und
er controlled perfusion pressure and intact coronary tone is associate
d with reduced coronary flow compared to atrial pacing; and (3) abolis
hment of vascular tone under controlled perfusion pressure diminishes
the expected difference in blood flow between atrial and ventricular p
acing. Blood flow velocity (BFV) in the left anterior descending (LAD)
and the left circumflex arteries (CFX) and a wall thickening index we
re measured in 14 open-chest dogs under normal conditions, and constan
t perfusion pressure. Four pacing sites were used: right atrium (RA(p)
), mid-right ventricle (RV(p)), mid-left ventricle (LV(p)), and left v
entricular apex (Apex(p)). Pacing modes were either sequential ventric
uloatrial (VA) (protocol A, n = 7), or sequential atrioventricular (AV
) (protocol B, n = 7), with a shorter AV difference (30 msec) than nor
mal. Results: BFV was decreased in the LAD during RV(p) and Apex(p) pa
cing by 9.7%-12.9% versus RA(p) and by 11.6%-14.6% versus LV(p) (P < 0
.05). No BFV variations were observed in the CFX. Flow velocity conduc
tance (FVC = mean blood flow velocity divided by the mean aortic press
ure) was higher by 16%-28% in the CFX for the three ventricular pacing
sites versus the atrial pacing, and higher by 14.1% +/- 6.1% only in
LV(p) versus RA(p) pacing in the LAD (P < 0.05). Wall thickening index
reduced during ventricular pacing in all three ventricular sites by 5
0%-64% (P < 0.05) compared to atrial pacing. Under constant perfusion
pressure, LAD blood flow decreased with ventricular pacing as compared
to right atrial pacing; this was particularly pronounced during the d
iastolic phase (16.6%-45.5%, P < 0.02). Normalized oscillatory flow am
plitude (OFA(n)) was reduced in RV(p) pacing compared to RA(p) and LV(
p) pacing (16.2 +/- 3.5 and 21.7% +/- 4.1%, respectively, P < 0.03). T
he variations in blood flow and OFA(n) disappeared with adenosine-medi
ated maximum vasodilatation. Summary: (1) Mean and phasic flows are re
duced in the early activated LAD region by ventricular pacing (RV(p),
Apex(p)). (2) Under controlled perfusion pressure and intact vascular
tone, ventricular pacing compromises blood flow compared with atrial p
acing. (3) This effect disappears when vascular tone is eliminated by
intracoronary injection of adenosine, suggesting that the coronary aut
oregulation is responsible for some of the effects.