Should conventional angiography be the gold standard for carotid stenosis?

To compare conventional angiography (CA) and rotational angiography (RA) to assess the degree of angiographically-measured stenosis versus cross-sectional area (CSA) stenosis in an in vitro carotid model.

Various grades of stenosis were created by adhering different amounts of silicone rubber sealant onto the inner wall of clear, radiolucent tubes. Following 2- and 3-projection CA and 20-projection RA, the tubes were transected at the actual maximum stenosis. The cross-sectional areas were digitally photographed, and CSA stenosis was calculated using ImageJ planimeter software. The differences among CA, RA, and CSA stenosis measurements were compared statistically.

There was no significant difference between RA and CSA stenosis measurements (p=0.46). Conventional angiography with 2 or 3 projections between 0 degrees and 90 degrees underestimated the severity of disease in 19 (63%) of 30 samples. The maximum stenosis percentage was significantly lower in CA versus RA (p<0.0001 in 2-projection, p<0.0003 in 3-projection) and in CA versus CSA stenosis (p<0.0004 in 2-projection, p<0.001 in 3-projection). The maximum stenosis percentages measured by RA were less than CSA stenosis in 5 (71.4%) of 7 tubes (p=NS) containing 50% to 69% stenoses. Eight tubes had mountain-shaped lesions, which was significantly overestimated by RA (11.5%+/-9.7%, p<0.012).

CA with 2 or 3 projections significantly underestimates the maximum stenosis in an in vitro model. RA may overestimate disease in patients with mountain-shaped plaques and may underestimate disease if the stenosis is <70%. Our data suggest that CA should not be the gold standard for the qualification of carotid endarterectomy in asymptomatic patients, nor for vascular laboratory quality assurance analysis.