Histological correlates of carotid plaque surface morphology on lumen contrast imaging.

BACKGROUND

Carotid angiographic plaque surface morphology is a powerful risk factor for stroke and systemic vascular risk. However, the underlying pathology is unclear, and a better understanding is required both to evaluate other forms of carotid imaging and to develop new treatments. Previous studies comparing angiographic plaque surface morphology with pathology have been small and unblinded, and the vast majority assessed only the crude macroscopic appearance of the plaque. We performed the first large study comparing angiographic surface morphology with detailed histology.

METHODS AND RESULTS

Carotid plaque surface morphology was classified as ulcerated, irregular, or smooth on 128 conventional selective carotid artery angiograms from consecutive patients undergoing endarterectomy for severe symptomatic stenosis. Blinded angiographic assessments were compared with 10 histological features recorded on detailed microscopy of the plaque using reproducible semiquantitative scales. Angiographic ulceration was associated with plaque rupture (P=0.001), intraplaque hemorrhage (P=0.001), large lipid core (P=0.005), less fibrous tissue (P=0.003), and increased instability overall (P=0.001). For example, angiographically ulcerated plaques were much more likely than smooth plaques to be ruptured (OR=15.4, 95% CI=2.7 to 87.3, P<0.001), show a large lipid core (OR=26.7, 95% CI=2.6 to 270, P<0.001) or a large hemorrhage (OR=17.0, 95% CI=2.0 to 147, P=0.02). The equivalent odds ratios for angiographically irregular versus smooth plaque were 6.3 (1.3 to 31, P=0.02), 6.7 (1.5 to 30, P=0.008), and 9.2 (1.1 to 77, P=0.02), respectively.

CONCLUSIONS

In contrast to previous studies based on macroscopic assessment, we found very strong associations between detailed histology and carotid angiographic plaque surface morphology. Plaque surface morphology on carotid angiography is a highly sensitive marker of plaque instability. Studies of the predictive value of MR- and CT-based lumen contrast plaque surface imaging are required.