PROJECT SUMMARY In the United States over 500,000 cardiac surgeries are performed annually. Postoperative acute kidney injury (AKI) occurs in ~20% of cardiac surgical patients. Several cohort and registry studies have reported that AKI defined by rise in serum creatinine after cardiac surgery associates with development of chronic kidney disease and cardiovascular events. However, it is unclear which patients will develop long-term adverse kidney events, adverse cardiovascular events, or both. In recent years blood and urine AKI biomarkers have been identified that reflect different aspects of AKI biology and that detect serum creatinine defined AKI earlier than rise in serum creatinine or detect subclinical AKI that is not revealed by traditional serum creatinine assessment. Clinical outcomes and biomarker research has focused mainly on detecting cardiac surgery- associated AKI itself, but not on predicting which patients are at greatest risk for long-term adverse kidney and cardiovascular outcomes after AKI. The central hypothesis of this proposal is that perioperative blood and urine AKI biomarkers significantly associate with increased long-term (2 to 5 years) postoperative major adverse kidney events (MAKE) and major adverse cardiovascular events (MACE). MAKE is defined as the composite of dialysis, death, renal hospitalization, or ≥ 30 day postoperative eGFR decline >25% from preoperative baseline. MACE is defined as the composite of death or hospitalizations for heart failure, myocardial infarction, coronary revascularization, arrhythmia, or stroke. We propose a prospective observational cohort study of 610 patients undergoing cardiac surgery at UT Southwestern Medical Center who will be followed for a minimum of 2 years and up to 5 years following cardiac surgery. Based on preliminary data we will assess plasma NT-pro- B-type natriuretic peptide, plasma intact fibroblast growth factor 23, serum cystatin C, urine TIMP-2*IGFBP7, and urine Kidney Injury Molecule-1 AKI biomarkers preoperatively and at 5 postoperative time points. This proposal will address three specific aims: 1) To determine the association between in-hospital AKI biomarkers and occurrence of MAKE during long-term follow-up; 2) To determine the association between in-hospital AKI biomarkers and occurrence of MACE during long-term follow-up; and 3) To develop clinical prediction models for long-term MAKE and MACE after cardiac surgery. In addition to traditional regression modeling, we will use machine learning that leverages detailed perioperative data including time-varying intraoperative and intensive care unit clinical data and blood and urine AKI biomarker data to create high performing prediction models. Our proposal is significant because knowing what blood and urine AKI biomarkers and clinical parameters accurately predict long-term major adverse kidney and cardiovascular outcomes after cardiac surgery provides the foundation for clinical trials that will identify eff...