Key points are not available for this paper at this time.
BACKGROUND AND AIMS: Cardiac sarcoidosis (CS) is poorly understood, challenging to diagnose, and portends a poor prognosis. A lack of animal models necessitates the use of residual human samples to study sarcoidosis, which in turn necessitates the use of analytical tools compatible with archival, fixed tissue. This study employed high-plex spatial protein analysis to characterize the immunologic, fibrotic, and metabolic landscape of CS. METHODS: Cardiac tissues from 64 patients-39 with CS and 25 without CS (Controls)-were analysed using the GeoMx Digital Spatial Profiler. Regions of interest (ROIs) with and without granulomatous inflammation were selected to examine protein abundance. ROIs were further divided into areas of illumination to separately analyse distinct protein profiles in myocytes, stroma, and vessels. A 79-protein panel was used to characterize immune and parenchymal cell populations. Differential protein abundance was evaluated using mixed-effects modeling. A predictive model was developed using a subset of seven proteins to distinguish inflammation-free CS ROIs from control ROIs. RESULTS: Spatial profiling demonstrated marked heterogeneity within and between sarcoidosis granulomas. Inflammation-free ROIs from CS hearts exhibited a 'distance gradient', with protein abundance varying by proximity to granulomatous inflammation. The seven-protein predictive model achieved 95.9% accuracy and an area under the receiver operating characteristic curve of 0.993 in differentiating inflammation-free CS ROIs from Controls. CONCLUSIONS: Spatial proteomics revealed novel features of the CS microenvironment and identified a distinct immune signature in tissue lacking overt inflammation. These findings may improve diagnostic accuracy in clinical biopsies from patients with suspected CS.
Peyster et al. (Wed,) studied this question.