Determination of Persistent Inflammation, Immunosuppression, and Catabolism Syndrome in the Pediatric Cardiac ICU

Approximately every fifteen minutes, a child is born in the United States with a congenital heart defect (CHD). The severity of the defect can vary widely, but children will often require invasive surgery for correction of their heart defects. While survival after complex CHD surgery has greatly improved over time, there remains a significant portion of children who do not survive. Reasons for this are multifactorial, but one notable cause is due to the development of severe, life-threatening infections, known as sepsis. Within the postoperative population, there is a subset of patients who have immunological dysfunction that predisposes them to infections. These patients often have persistent inflammation, immunosuppression, and catabolism syndrome (PICS) and are characterized by having a low absolute lymphocyte count (ALC), low albumin and an elevated c-reactive protein as well as a prolonged intensive care unit (ICU) stay. Utilizing an institutional database, we have previously investigated the impact of PICS on all culture-positive sepsis mortalities within a high-volume, single- center children’s hospital over the past 20 years. While a PICS phenotype occurred in nearly one-half of all culture-positive sepsis-related deaths institution-wide, it occurred in nearly three-quarters of the sepsis mortalities in the cardiovascular ICU (CVICU). Based on these findings, we will perform a deeper analysis in two institutional databases to assess the timing and risk of developing PICS with sepsis in the CHD postoperative period. These databases are based on a searchable electronic medical record (EMR) within a single institution stretching back over 25 years. Data comes in the form of structured data points, such as billing codes and encounter dates, semi-structured data such as laboratory results, or unstructured data such as progress reports. Extracted data will be analyzed with the intent of understanding baseline risk and time-to-event analysis of key characteristics and outcomes of the PICS phenotype. Aim 1 will determine the relationship of clinical and genetic variables in CHD patients with PICS postoperatively. Specifically, the association of genetic polymorphisms to postoperative ALC, as well as assessing the general prevalence of PICS within the CHD surgical population and identifying clinical and demographic risk factors associated with phenotype development will be determined. In Aim 2, known risk factors associated with postoperative PICS and sepsis will then be coupled with machine learning (ML) to identify new, previously unidentified risks. Aggregate risk factors will be used to generate prediction models of PICS after CHD surgery that can be embedded within an EMR. The overall goal of these pursuits is to create a better understanding of what features potentially lead to PICS and postoperative sepsis in CHD patients and create a practical clinical prediction tool that offers guidance to researchers and clinicians alike. With a clearer understanding of who might develop PICS in the postoperative period, targeted research questions and proactive treatment modalities can be employed to temper the effect of immune dysfunction in this critically ill and vulnerable population. Project Number: 1R21HL175625-01A1 | Fiscal Year: 2025 | NIH Institute/Center: National Heart Lung and Blood Institute (NHLBI) | Principal Investigator: Ryan Stark (+1 co-PI) | Institution: VANDERBILT UNIVERSITY MEDICAL CENTER, NASHVILLE, TN | Award Amount: $271,938 | Activity Code: R21 | Study Section: Surgery, Anesthesiology and Trauma Study Section[SAT] View on NIH RePORTER: https://reporter.nih.gov/project-details/1R21HL17562501A1