464 - Therapeutic Plasma Exchange in Septic Neonates on ECMO: Survival Trends From a Subgroup Analysis

Publication Number: 2452.464

Chelsea K. Bitler, UPMC Children's Hospital of Pittsburgh, Pittsburgh, PA, United States; John Ibrahim, University of Texas Southwestern Medical School, Dallas, TX, United States; Kara M. Marangoni, UPMC Childrens Hospital of Pittsburgh, Pittsburgh, PA, United States; Amanda Monteverde, UPMC Childrens Hospital of Pittsburgh, Pittsburgh, PA, United States; Jennifer Troutman, UPMC Childrens Hospital of Pittsburgh, Pittsburgh, PA, United States; Adriana Phillips, UPMC Childrens Hospital of Pittsburgh, Pittsburgh, PA, United States; Abeer Azzuqa, UPMC Childrens Hospital of Pittsburgh, Sewickley, PA, United States; Burhan Mahmood, UPMC Children's Hospital of Pittsburgh, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States, PA, United States

Background: Therapeutic plasma exchange (TPE) has gained increasing interest as an adjunctive therapy in neonatal extracorporeal membrane oxygenation (ECMO), particularly in neonates with sepsis and thrombocytopenia-associated multiorgan failure (TAMOF). However, its safety and impact on clinical outcomes in this population remain poorly defined.

Objective: To assess the impact of TPE on clinical outcomes among neonates with sepsis on ECMO.

Design/Methods: Single-center, retrospective cohort review of neonates undergoing Veno-Arterial ECMO for cardiorespiratory failure between January 2013 and June 2025. Patients were stratified according to the presence of an underlying diagnosis of sepsis or infection and receipt of TPE. Statistical analysis performed in Stata v19.5.

Results: Demographic characteristics between groups were similar apart from the Pittsburgh Index for Pre-ECMO Risk (PIPER) scores and prevalence of renal impairment (Table 1). Beyond infection status, the distribution of diagnoses also differed: non-septic neonates without TPE were more likely to have congenital diaphragmatic hernia (CDH), whereas septic neonates without TPE were more likely to have been diagnosed with hypoxic-ischemic encephalopathy (HIE), meconium aspiration syndrome (MAS), or pulmonary hypertension (PHTN) (Table 1). Among 63 neonates with an infectious or sepsis diagnosis, 17.5% received TPE. Septic TPE recipients were more likely to have a confirmed infectious etiology, while culture-negative sepsis predominated in septic neonates not receiving TPE (Table 1). Bacterial infections were more common than viral infections among all septic neonatal ECMO patients, particularly in those who received TPE (Table 1). TPE indications varied (Table 2). In adjusted analyses, there were no significant differences in discharge survival or neurologic injury between groups (Figure 1).

Conclusion(s): After adjusting for severity of illness and the presence of a confirmed pathogen, survival among septic neonates receiving TPE was comparable to that of both non-septic groups. The septic non-TPE group demonstrated a non-significant trend toward higher survival relative to the other three groups, a finding that warrants further evaluation. After controlling for the presence of abnormal neuroimaging findings present prior to ECMO initiation and before TPE, no significant differences in neurologic injury (hemorrhage or ischemia) were observed between groups. These findings suggest that TPE does not confer additional mortality or neurologic risk in this population and support further multi-center evaluation of its therapeutic role during neonatal ECMO.

Table 1.
Demographic characteristics and diagnoses by subgroup.

Table 2.
TPE indications by subgroup.

Figure 1. Survival and Neurological Outcomes
Figure 1A. Forest plot demonstrating the adjusted odds ratios (95% CI) for discharge survival among neonatal ECMO subgroups. Odds ratios are shown relative to the reference group of non-septic neonates who did not receive TPE adjusting for PIPER score and the presence of a confirmed infectious pathogen via culture or PCR. Figure 1B. Forest plot showing the adjusted odds ratios (95% CI) for evidence of neurologic injury on neurodiagnostic imaging in neonatal ECMO subgroups. Odds ratios are shown relative to the reference group of non-septic neonates who did not receive TPE adjusting for noted any neuroimaging abnormality prior to initiation of ECMO or prior to apheresis but after cannulation.