Abstract
ABSTRACTBackgroundPeople with Multiple Long-Term Conditions (MLTC) experience higher rates of organ failure and death following cardiac surgery. The aim of this study was to identify disease processes in MLTC associated with increased susceptibility to organ injury post-surgery.MethodsUnsupervised machine learning methods were applied to pre-surgery biomarkers of haematological, cardiac, liver, and renal disease from four intensively phenotyped cardiac surgery cohorts. The resulting MLTC clusters were then matched to a fifth cohort where single nuclei RNA sequencing (snRNAseq) was performed on atrial biopsies collected at surgery.ResultsK-means clustering identified two MLTC clusters. Cluster 1 had lower rates of chronic kidney disease and anaemia and increased immune system activation pre-surgery. Cluster 2 had more severe cardiorenal disease, anaemia, and elevated biomarkers of immunological ageing pre-surgery. Cluster 2 had significantly higher rates of organ injury relative to Cluster 1. The results were consistent across internal and external validation analyses.Analysis of snRNAseq data in biopsies from Cluster 1 demonstrated enrichment for immune response genes in cardiomyocytes, naive T/B lymphocytes and progenitor cells, and activation of non-tissue resident macrophages relative to Cluster 2. Cluster 2 showed enrichment for senescent/ effector memory T cells, dysregulated activation of tissue-resident macrophages, and cardiomyocyte dedifferentiation relative to Cluster 1.In UK Biobank, genetic modification of genes differentially expressed between the two MLTC phenotypes altered 90-day mortality post-surgery.ConclusionsImmune system homeostasis determines susceptibility to organ injury and death in people with MLTC undergoing cardiac surgery and represents a previously unrecognised target for organ protection interventions.
Publisher
Cold Spring Harbor Laboratory