<p>Coronary plaque progression often deviates from the diagnostic range of clinical examinations, resulting in the need for emergent follow-ups and increased mortality among outlier patients. This study first analyzes clinical data of fractional flow reserve, optical coherence tomography (OCT), and computed tomography from 180 patients over a 9&#xa0;years period, with an average follow-up of 2 years. When the plaque progression and healthy control groups were compared in a 1:1 match (n = 10 each), analysis of single artery images failed to detect plaque progression. Therefore, 100–200 image frames were used to reconstruct patient-specific coronary anatomy using computational fluid dynamics, enabling prognostic assessment of plaque progression. The results suggest that steep plaque slopes promote plaque progression by increasing shear stress and hemodynamic disturbance accompanied by greater macrophage recruitment than gentle slopes, which was validated by a microfluidic model. Moreover, increased activation and fusogenic potential of macrophages in steep plagues justified the development of fusogenic macrophage (FM)-vesicles to detect plaque progression. The fusion potential enables FM-vesicles to self-target fusogenic macrophages more efficiently than in vitro compared to macrophages, showing superiority over liposomes and macrophage-vesicles. The membrane fusion mechanism facilitates lysosomal escape, allowing prolonged cytosolic retention without degradation. Rabbit carotid ligation was used to produce steep and gentle plaques by controlling the incision direction. When gold nanoparticles were loaded into FM-vesicles and injected into these arteries ex vivo, OCT accurately imaged the plaque slope and detected changes in signal intensity. This study presents the translational potential of FM-vesicles for clinical application by reducing diagnostic outliers in the detection of plaque progression.</p> Graphical abstract <p></p>