Optical coherence tomography and artificial intelligence for calcium quantification in coronary disease of diabetic patients. (3rd October 2022)

• Record Type: Journal Article
• Authors: Di Muro, F M; Demola, P; Nardi, G; Ciardetti, N; Meucci, F; Stolcova, M; Ristalli, F; Di Mario, C; Mattesini, A
• Is Part Of: European heart journal. Volume 43(2022)Supplement 2
• Publication Date: 2022-10-03
• Subjects: Cardiology -- Periodicals; Heart -- Diseases -- Periodicals; 616.12005
• DOI: 10.1093/eurheartj/ehac544.1204
• Languages: English
• ISSNs: 0195-668X

Abstract

Abstract: Background: Diabetes mellitus (DM) is associated with increased cardiovascular morbidity and mortality. Coronary artery disease in diabetic patients is characterized by a greater burden of lipidic plaques and calcifications. Little is known on the quantitative and qualitative characteristics of calcific plaques in diabetics vs non diabetics. The recent application of Artificial Intelligence (AI) to optical coherence tomography (OCT) enables unique evaluation of coronary calcification. Purpose: To compare qualitative and quantitative characteristics of coronary calcified plaques in diabetic and non-diabetic patients using AI-OCT. Methods and material: We recruited 78 patients admitted for chronic coronary syndrome (CCS) or acute coronary syndrome (ACS) undergone intracoronary imaging with OCT between January 2019 to October 2021. Differences in plaques characteristics assessed by Artificial Intelligence applied at OCT runs were compared in DM and non-DM population using generalized estimating equations. To estimate the burden of calcification we classified the calcific lesions according to the Fujino score, an OCT based calcium scoring system. Results: A total of 78 patients were included (54 non-DM lesions, 29 DM lesions). The culprit lesion was examined by OCT in all patients without any peri- or postprocedural complications. The population was homogeneous for cardiovascular risk factors even if we observed a higher prevalence of peripheral arterial disease (PAD) in the DM cohort (22.2% vs 2% p value 0.003). There were no statistical differences in previous PCI or CABG but we observed more multivessel PCI in the history of DM patients if compared with non-diabetic ones (33.3% vs 11.8% p value 0.021). The clinical presentation in DM groups was more often unstable angina (22.2% vs 0% p value <0.001) while STEMI, NSTEMI or CCS had the same prevalence in the two cohorts. At baseline angiography, patients with diabetes had more often multivessel disease (29.6% vs 17.6% p=0.014) with all the vessels equally involved. There were no qualitative differences in plaque morphology but using the Fujino score to estimate the calcium burden in the two population we found hardest calcific plaques expressed by higher Fujino score more frequently in DM patients compared to non-DM ones (50% vs 26.9%, p=0.04 of Fujino score 4). Conclusion: DM has an impact on atherosclerotic process and plaque remodeling. Applying AI methods at OCT plaque analysis, we can extract important and standardized information on calcium burden in diabetic. This might help the interventional cardiologist in image interpretation, therapeutic strategy decision, improving workflow and clinical outcomes. Funding Acknowledgement: Type of funding sources: Public Institution(s). Main funding source(s): University of Florence