Identification of completed coronary stent healing by optical coherence tomography - validation study with histopathology in porcine model of coronary restenosis.

Abstract

BACKGROUND: The potential of optical coherence tomography (OCT) is limited by incomplete validation with histopathology. The study aimed to assess whether OCT can identify a completely healed coronary stent. MATERIAL AND METHODS: The study included 40 swine and total of 106 stents (82 DES, 24 BMS). Follow-up OCT and histopathology examination was done after 28 days (n = 53) and 90-days (n = 53). 273 frames were matched between histopathology and OCT. Histopathologic criteria for completed healing: high endothelialization score (2-3), low inflammation score (0-1), low fibrin score (0-1), high neointimal smooth muscle score (2-3) and lack of binary restenosis. RESULTS: Predictors of a well healed stent were presence of homogenous neointima (OR: 2.53) and the total number of struts per section (OR: 1.11). The presence of microvessels (OR: 0.28) and increasing neointimal area (OR: 0.65) predicted incomplete healing. Cutoff values were identified: inflammation score was the highest in segments with neointima thickness over 0.35 mm. Persistent fibrin deposits were found at 28-day follow-up in DES with less than 13 embedded struts per section and neointima area less than 2.234 mm². The number of embedded struts per section showed a positive correlation with the healing score, while both protruding covered and protruding uncovered struts showed a negative correlation. CONCLUSIONS: OCT demonstrated moderate ability to predict completed stent healing.