P197 Similarity and difference in in-stent neoatherosclerosis assessment between near-infrared spectroscopy and optical coherence tomography

Abstract Backgrounds Near-infrared spectroscopy (NIRS) has been implicated to assess the composition of the atherosclerotic plaques in native coronary arteries. However, little has been reported about the assessment of neoatherosclerosis within stents by NIRS. Optical coherence tomography (OCT) is one of the current clinical standard modalities to identify neoatherosclerosis. We sought to compare NIRS and OCT assessments in the detection and evaluation of in-stent neoatherosclerosis. Methods We retrospectively analyzed 54 in-stent lesions in 52 patients who underwent both NIRS and OCT examinations before percutaneous coronary intervention (PCI). All stented segment in both PCI target and non-target lesions were investigated. NIRS analysis included lipid core burden index (LCBI)4mm. NIRS-derived lipid-rich plaque (NIRS-lipid) is defined as LCBI4mm more than 400 according to the previous studies. OCT-derived lipid-laden neointima (OCT-lipid) was defined as having a signal-poor region with diffuse border precluding visualization of stent struts of more than 90 degrees and 1mm in longitudinal length. OCT analysis included the presence or absence of lipid-laden neointima, calcification, ruptured plaque, thrombus, heterogeneous neointima and measurement of lipid length and lipid arc. Lipid volume index (LVI) was defined by OCT as the averaged lipid arc multiplied by lipid length. Max LCBI4mm were compared in lesions with co-registered OCT findings and correlations between max LCBI4mm and LVI were determined. Additionally, OCT findings predicting max LCBI4mm more than 400 were determined. Results Max LCBI4mm was 202 (7.5-460). OCT-lipid was observed in 31 (57.4%) lesions. LVI was 1283 (475-4725). Lesions with OCT-lipid has significantly higher LCBI than those without OCT-lipid (343 [106-593] vs 42 [0-260], p = 0.003). On the other hand, max LCBI were not significantly different according to the presence or absence of calcification, ruptured plaque, thrombus and heterogeneous neointima (60 [39-537] vs 227 [6-451], p = 0.913; 441 [43-547] vs 202 [6-435], p = 0.328; 245 [10-548] vs 251 [42-446], p = 0.990; 202 [40-440] vs 235 [3-471], p = 0.970, respectively). Linear regression analysis showed significant correlation between LVI and max LCBI4mm (P < 0.001). Receiver operating characteristic (ROC) analysis showed the best cutoff max LCBI4mm value for predicting OCT-lipid was 148 (Area under the curve [AUC] = 0.735). ROC analysis showed the best cutoff lipid length and maximum lipid arch value for predicting LCBI4mm more than 400 was 1.9 mm and 128 degrees, (AUC = 0.741, 0.732, respectively) respectively. Conclusion There is good agreement between NIRS and OCT for detection of neoatherosclerosis in stented lesions. In the in-stent lesions, the cut-off value of max LCBI for predicting OCT-lipid was smaller than the previously reported value in de novo lesions. Application of different cut-off value may be considered in in-stent lesions.