Coronary magnetic resonance angiography (coronary MRA) can detect, noninvas
ively, a high proportion of severe stenotic lesions found on coronary angio
grams. However, quantitative evaluation of coronary artery stenosis by coro
nary MRA has been performed only in a small number of patients. This study
was designed to determine whether coronary MRA can assess the degree of ste
nosis using the two-dimensional segmented turbo-FLASH method (2D method). W
e studied 108 patients with technically adequate coronary MRA images. The b
lood flow signal intensity on coronary MRA was classified as markedly decre
ased, moderately decreased, or normal. The severity of coronary artery sten
osis was determined by the caliper method, and coronary stenosis was rated
using a seven-point scale (0%, 25%, 50%, 75%, 90%, 99%, and 100%) in accord
ance with the American Heart Association classification system. Patients we
re classified into three groups: normal coronary artery (0%-25% stenosis),
moderate stenosis (50%-75% stenosis), and severe stenosis (90%-100% stenosi
s). The degree of stenosis on coronary angiography and the decrease in coro
nary MRA signal intensity were compared. The right coronary artery was eval
uated in 64 patients and the left coronary artery in 73 patients. When a ma
rked or moderate decrease in coronary MRA blood flow signal intensity was d
efined as indicating stenosis, the sensitivity and specificity of coronary
MRA for detecting angiographically severe stenosis were 85% and 80%, respec
tively. A moderate decrease in coronary MRA blood flow signal intensity det
ected angiographically moderate stenoses with a sensitivity of 38% and a sp
ecificity of 83%. Coronary MRA can detect a high proportion of severe steno
ses but only a low proportion of moderate stenoses. Technical improvements
are required before coronary MRA can be used clinically.