Phenotyping atherosclerotic plaque vulnerability has traditionally relied on histopathology to capture structural tissue features, but histology may not fully account for clinical presentation. The authors tested whether deep proteomic profiling of carotid plaques and matched plasma could reveal molecular signatures of plaque rupture and symptomatic disease that extend beyond conventional histopathology and suggest circulating biomarkers.
The study analyzed paired carotid plaque tissue and preoperative plasma from 88 patients undergoing carotid endarterectomy; 51 were classified as symptomatic and 37 as asymptomatic. Proteomic measurements were performed on both plaque and plasma samples using the Olink Explore 3072 platform. The investigators related plaque protein abundance to symptomatic presentation and to quantitative histopathological features. They compared classification performance of models based on histopathology versus models based on proteomics. The study further integrated plaque proteomics with plaque single-cell RNA-sequencing (scRNA-seq) to derive proteomic signatures of cellular abundance and explored associations between these signatures and plaque phenotypes. Finally, the authors assessed plaque–plasma concordance across the set of shared proteins.
Across 2,837 proteins quantified in plaques, 19 proteins were differentially expressed in plaques from symptomatic patients compared with asymptomatic patients. These differentially expressed proteins highlighted pathways related to neutrophil degranulation and the innate immune system. The protein with the strongest association to symptomatic presentation was FGFBP1 (log2 fold change = 1.14; P = 1.82 × 10^-6). The reported differential expression and pathway enrichment indicate an inflammatory signature distinguishing symptomatic plaques in this cohort.
Proteins associated with a composite vulnerability index derived from histopathology were enriched for inflammatory signaling pathways, including TNFα signaling through NFκB, complement activation, and IL6–JAK–STAT3 signaling. Individual protein associations with specific histopathologic features were observed: CXCL8 was linked to macrophage burden and lipid core size, while EPHB4 and PKN3 were associated with neovascularization. These mappings suggest that proteomic readouts can align with—and extend—the biological interpretation of histologic features by highlighting underlying inflammatory and vascular processes.
The authors developed models to discriminate symptomatic from asymptomatic plaques using either histopathology-derived metrics or proteomic data. The proteomics-based model outperformed the histopathology-based model, with an area under the receiver operating characteristic curve (AUC) of 0.83 versus 0.66 for the histopathology model; the difference was statistically significant (P = 0.026). This result suggests that proteomic profiling may provide better discrimination of symptomatic disease status than conventional histopathologic assessment in this dataset.
To interpret proteomic signals in the context of cellular composition, the investigators integrated plaque proteomics with plaque scRNA-seq data and developed proteomic cell-class signatures. These signatures correlated with histopathology readouts, including macrophage burden, smooth muscle cell content, and neovascularization, indicating that proteomic patterns can reflect cellular abundance and activity within plaques when linked to single-cell transcriptomic maps.
Overall correspondence between plaque and plasma protein levels was limited: the median Spearman correlation across 2,837 shared proteins was ρ = 0.11. Despite this limited concordance, selected proteins exhibited concordant associations in plasma and plaque. Notably, FGFBP1 demonstrated concordant association in plasma, marking it as a potential circulating biomarker for future risk stratification pending further validation. The authors emphasize that while most of the plaque proteome is distinct from the plasma proteome, specific proteins may nonetheless be detectable and informative in circulation.
Deep proteomic profiling of human carotid plaques identified molecular signatures of symptomatic atherosclerosis that extend beyond conventional histopathology. The findings implicate neutrophil activation and broader inflammatory signaling pathways as key determinants of plaque vulnerability. Proteomics provided superior discrimination of symptomatic versus asymptomatic plaques compared with histopathology in this cohort, and integration with single-cell transcriptomics generated cell-class signatures that aligned with histopathologic features. Although plaque and plasma proteomes were largely distinct, selected proteins including FGFBP1 may warrant further study as circulating biomarkers for clinical risk stratification.
This work is reported as a preprint and has not been certified by peer review. The study was approved by the Ethics Committee of the LMU Faculty of Medicine (approval numbers 121-09, 22-0135, 23-0772) and all participants provided written informed consent. Authors declared competing interests: M.K.G. reports consulting fees unrelated to this work; other authors reported no disclosures. The manuscript states that all data produced are contained in the present work and provides links to supplementary material and external data/code resources where available.
Limitations to note from the source: the report is a preprint and therefore preliminary; the abstract and manuscript provide the data described but further peer review and external validation are necessary before clinical application. No additional methodological or cohort details beyond those summarized here were reported in the source abstract.