ABSTRACT Acute pancreatitis (AP) patients can be classified into mild or severe outcome based on the presence of organ failure days during hospitalization. However, there are no biomarkers that can predict severity outcome. This study aim is to perform an unbiased characterization of the circulating immunome of AP patients at hospital admission aiming to identify novel biomarkers which could predict disease outcome prior to the onset of organ damage and clinical deterioration. Peripheral blood mononuclear cells (PBMC) were collected from newly hospitalized AP patients. Based on their evolution, patients were classified following the Modified Atlanta classification into mild acute pancreatitis (mAP) ( n = 20) as well as moderately and/or severe acute pancreatitis (ms‐sAP) ( n = 20). PBMC were analysed by high‐dimensional spectral cytometry with unsupervised dimensionality reduction and clustering algorithms based on the subsequent outcome of the patients. Results were further validated by classical gating and functional approaches with a smaller cohort ( n = 8 mAP and n = 8 ms‐sAP). A total of 120 different immune cell clusters were identified and characterized in AP patients at hospitalization. Following computational analysis and classical gating validation approaches, circulating CD2 + CD8 dim NKT‐Like cells were found to be specifically reduced in ms‐sAP patients. Moreover, NKT‐Like cell subsets from these patients displayed an IL‐15 induced expansion of NKG2a and NKp30 referred to mAP. Patients with AP at hospitalization with subsequent ms‐sAP display a specific reduction of NKT‐Like cells coupled with an expanded function. Hence, quantifying the phenotype and/or function of circulating NKT‐Like cells during hospital admission could serve as a predictive biomarker for AP severity, aiding in early stratification and personalized treatment strategies.
Castro et al. (Tue,) studied this question.