The serum proteome of VA-ECMO patients changes over time and allows differentiation of survivors and non-survivors: an observational study.

BACKGROUND: Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is applied in patients with refractory hemodynamic failure. Exposure of blood components to high shear stress and the large extracorporeal surfaces in the ECMO circuit trigger a complex inflammatory response syndrome and coagulopathy which are believed to worsen the already poor prognosis of these patients. Mass spectrometry-based proteomics allow a detailed characterization of the serum proteome as it provides the identity and concentration of large numbers of individual proteins at the same time. In this study, we aimed to characterize the serum proteome of patients receiving VA-ECMO. METHODS: Serum samples were collected on day 1 and day 3 after initiation of VA-ECMO. Samples underwent immunoaffinity based depletion for the 14 most abundant serum proteins, in-solution digestion and PreOmics clean-up. A spectral library was built with multiple measurements of a master-mix sample using variable mass windows. Individual samples were measured in data independent acquisition (DIA) mode. Raw files were analyzed by DIA-neural network. Unique proteins were log transformed and quantile normalized. Differential expression analysis was conducted with the LIMMA-R package. ROAST was applied to generate gene ontology enrichment analyses. RESULTS: Fourteen VA-ECMO patients and six healthy controls were recruited. Seven patients survived. Three hundred and fifty-one unique proteins were identified. One hundred and thirty-seven proteins were differentially expressed between VA-ECMO patients and controls. One hundred and forty-five proteins were differentially expressed on day 3 compared to day 1. Many of the differentially expressed proteins were involved in coagulation and the inflammatory response. The serum proteomes of survivors and non-survivors on day 3 differed from each other according to partial least-squares discriminant analysis (PLS-DA) and 48 proteins were differentially expressed. Many of these proteins have also been ascribed to processes in coagulation and inflammation (e.g., Factor IX, Protein-C, Kallikrein, SERPINA10, SEMA4B, Complement C3, Complement Factor D and MASP-1). CONCLUSION: The serum proteome of VA-ECMO patients displays major changes compared to controls and changes from day 1 until day 3. Many changes in the serum proteome are related to inflammation and coagulation. Survivors and non-survivors can be differentiated according to their serum proteomes using PLS-DA analysis on day 3. Our results build the basis for future studies using mass-spectrometry based serum proteomics as a tool to identify novel prognostic biomarkers. TRIAL REGISTRATION: DRKS00011106.

Acute, Chronic, and Treated Aortic Diseases Present Distinguishable Serum Proteome Fingerprints with Protein Profiles That Correlate with Disease Severity.

Aortic diseases are a rare but potentially life-threatening condition. We present a serum proteomic study for a spectrum of aortic diseases including thoracic aortic aneurysms (n = 11), chronic dissections (n = 9), acute aortic dissections (n = 11), and surgically treated dissections (n = 19) as well as healthy controls (n = 10) and patients of coronary heart disease (n = 10) to represent non-aortic cardiovascular disease. In total, we identified and quantified 425 proteins across all 70 samples. The different aortic diseases represented distinguishable proteome profiles. We identified protein clusters that positively or negatively correlate with disease severity, including increase of cytosolic tissue leakage proteins and decrease of components of the coagulation and complement system. Further, we identified a serum proteome fingerprint of acute aortic dissections, consisting, among others, of enriched inflammatory markers such as C-reactive protein and members of the S100 protein family. The study underlines the applicability of serum proteomics for the investigation of aortic diseases and highlights the possibility to establish disease-specific prognostic markers.