Limitations of OCT in identifying and quantifying lipid components: an in vivo comparison study with IVUS-NIRS.

AIMS

We aimed to assess the agreement between IVUS-NIRS and OCT to assess lipid plaques in patients with acute coronary syndromes or stable angina. In addition, the impact of both macrophages and calcifications was investigated.

METHODS AND RESULTS

Forty-three patients undergoing both IVUS-NIRS and OCT assessment of the culprit and/or non-culprit coronary lesions were enrolled. Cross-sections from lipid plaques, calcified plaques and normal-appearing vessel tracts were identified and matched with the two imaging techniques. Lipid arc was measured by both IVUS-NIRS and OCT. Macrophage presence and calcifications were also investigated with OCT. OCT detected a lipid plaque in 90 cross-sections (48.9%), with a sensitivity of 85.5% and a specificity of 69.7% as compared with IVUS-NIRS. The percentage of OCT false positive was 20.1% and of false negative was 4.9% for lipid plaque detection. The Pearson correlation coefficient for lipid arc was 0.675, p=0.0001. Macrophages were detected in 73% of OCT false positive cross-sections. Conversely, calcifications were present in 66.7% of OCT false negative cross-sections. The variability of lipid arc was independently associated with macrophages (beta=0.295, p=0.013).

CONCLUSIONS

Agreement between IVUS-NIRS and OCT for lipid plaque detection is suboptimal. The presence of macrophages and superficial calcifications on OCT negatively affects lipid detection.