Congenital Heart disease (CHD) is a leading cause of childhood death. Substantial morbidity and mortality relates to the postoperative course. For example, only 70% of neonates survive to hospital discharge after their first complex surgery for single ventricle heart disease. Adverse systemic inflammatory responses are highly exaggerated in some children postoperatively. This inflammation is pathological, results in leaky blood vessels and fluid overload, toxin release as well as cell damage contributing to lung, heart and kidney injury. Reasons why some children develop this amplified systemic inflammatory response after heart surgery while others do not are poorly understood. Mechanisms for how cardiopulmonary bypass and surgery drive this inflammation are also inadequately characterized. Currently, there are no existing methods to predict patients at high-risk for acute adverse postoperative complications, let alone adjust our management to mitigate these effects. Instead, our postoperative care approach is a one-size fits all, reactive process 'after' patients become inflamed or adverse events occur.

Proteins in a patient's blood participate in and reflect acute inflammatory responses. In other pediatric conditions, protein biomarkers have been shown to both predict and monitor inflammation and adverse outcomes, and importantly predict responsiveness to anti-inflammatory drug therapies. This is the premise of precision medicine. Personalizing treatment to each individual patient.

New technologies now allow the levels of tens of thousands of proteins to be measured from a few drops of blood. In this proposal the investigators will identify predictors of adverse events after heart surgery by quantifying protein levels and their changes after surgery. It is now possible to detect those proteins with the greatest variability in the postoperative course over time, and between patients, as well as those that are associated with adverse outcomes. The most informative proteins will yield insights into the causes of the inflammatory response. The investigators anticipate identifying protein plasma biomarkers in pathways associated with inflammation, metabolism, blood vessel function and the immune system as these may be key mechanisms involved. Advanced understanding of these mechanisms is critical to deriving targeted therapies to prevent or mitigate inflammatory responses.

The investigators will also collect patient clinical data, such as age, cardiac anatomy, and duration of surgery. By combining this clinical information with blood protein profiles, the investigators will be able to develop a model predicting patients at highest risk for adverse postoperative events using machine learning approaches. The overarching goal of this research integrating clinical and bench research is ultimately to translate precision medicine approaches to the Cardiac ICU. Guiding personalized care of high-risk patients by enabling clinicians to anticipate outcomes and tailor decision-making at the bedside will undoubtably improve outcomes in CHD.

Acute inflammation following cardiopulmonary bypass (CPB) contributes to morbidity and mortality; however, mechanisms eliciting these responses and biomarkers predicting at-risk patients remain unidentified. Proteins are disease-associated circulating factors in patients' blood that both participate in and reflect inflammation. Existing data suggests proteomic biomarkers hold promise for predicting and monitoring inflammation and adverse outcomes; however, further research is needed to validate biomarkers in pediatric cardiac disease, elucidate underlying mechanisms of CPB-induced inflammation, and translate them to clinical practice.

Specific Aims: Our overarching goal is to integrate quantitative biomarkers with clinical medicine to enhance care of patients with congenital heart disease (CHD) by personalizing their treatment. The investigators propose a first step towards a precision medicine paradigm for children with CHD, many of whom experience inflammation post-CPB. The investigators hypothesize that proteomic profiling will detect biomarkers that identify patients at highest risk for exaggerated systemic inflammatory responses to CPB. The investigators will conduct a pilot feasibility study integrating unbiased proteome-wide discovery approaches, to offer a deep and unbiased view of the proteomic landscape underlying pathological processes post-CPB (Fig. 1).

AIM 1: To prospectively enroll patients with CHD undergoing CPB in a study of quantitative proteomic profiles.

Hypothesis: Conducting an observational study utilizing unbiased proteome-wide discovery approaches in children in the cardiac intensive care unit (CICU) undergoing CPB is feasible. The investigators will identify 80 children in 3 diagnosis cohorts. Those undergoing stage 1 palliation (S1P) for hypoplastic left heart syndrome (HLHS), or biventricular (BiV) repair both have elevated risk of exaggerated inflammatory responses with postoperative courses characterized by fluid overload, prolonged mechanical ventilation, fevers, and capillary leak, and those with D-transposition of the great arteries (DTGA) post arterial switch operation (ASO) have lower risk. The investigators will collect demographic and clinical data and obtain blood samples at four time points: pre-CPB, immediately post-CPB, 12 hrs post-CPB, and 24 hrs post-CPB. Plasma will be analyzed on the SomaScan v5.0 platform, which measures 10,778 proteins from a 55 µL plasma sample.

AIM 2: To enhance understanding of inflammatory responses in children undergoing CPB by characterizing proteomic changes. Hypothesis: Temporal intra-patient proteomic responses to CPB and their relationships with outcomes will be characterized by the SomaScan platform. The investigators will describe dynamic intra-patient temporal variability in protein levels from pre- to post-CPB. Next, the investigators will identify proteins that demonstrate the greatest variability between patients overall, across timepoints, and between patients, stratified by outcome. To discriminate candidate biomarkers of postoperative inflammation and adverse outcomes, the investigators will study three clinical endpoints: time to successful extubation, extent of fluid overload in the 72 hours post-CPB, and a composite binary adverse outcome. The biomarker search will be unbiased and broad, with control for false discovery. Predictive proteins will be validated by targeted multiplex immunoassay and pathways described using systems biology approaches.

AIM 3: To develop a machine learning classifier predicting patient outcomes based on proteomic profiles and clinical variables. Hypothesis: Adverse outcomes following CPB will be predicted by a combination of clinical and proteomic biomarkers. The investigators will use supervised machine learning classifier approaches to discriminate and define the combination of proteomic and peri-operative clinical features (e.g. age, cardiac morphology/surgery, CPB, and circulatory arrest times) that accurately differentiate individuals at high risk of acute post-operative adverse events.

Inclusion Criteria:

• consent from parents
• cardiac surgical criteria and age criteria; elective BiV repair in patients aged >1 and <5 years or standard risk S1P/ASO

Exclusion Criteria:

• preoperative ventilation or vasoactive support or ECMO

Prospective patients undergoing S1P for HLHS, BiV repair, or ASO will be identified. Inclusion: elective BiV repair in patients aged >1 and <5 years or standard risk S1P/ASO. Exclusion: preoperative ventilation or vasoactive support. With primary treating team approval, the investigators will approach parents of eligible patients to discuss study participation with written informed consent obtained if they agree to enroll.

The investigators may share deidentified high dimensional proteomics data with the broader scientific and clinical community.