(766) Presepsin May Differentiate PGE-1-Induced Systemic Inflammation From Sepsis: A Prospective Study

Introduction: Prostaglandin E1 (PGE1) infusion maintains ductal patency in duct-dependent congenital heart disease(CHD) while awaiting surgery. Delay in surgery necessitates prolong PGE1 infusion, and predisposes to healthcare-associated infections. Considering difficulty in differentiating PGE1-induced systemic inflammation from neonatal sepsis, we planned to study presepsin and inflammatory markers in patients on PGE1 infusion for >5 days with & without sepsis.

Methods: Baseline demographic, clinical and laboratory characteristics were recorded at enrolment, then every 5 days and when sepsis was suspected, until end points (surgery or death). Sepsis defined by clinical suspicion with positive body fluid culture, and/or pneumonia on X-ray, or improvement in inflammatory markers with antibiotics. Presepsin and inflammatory markers were compared at observation points when patients did and did not have sepsis. Data were expressed as median (IQR).

Results: Thirty patients [male, 18; age, 11 (7, 17.5) days] with CHD were enrolled. PGE1 infusion was started at 5 (3, 12) days of life at 20 (12.5, 50) ng/kg/min. PGE1 was given for 21 (11, 27) days at cumulative dose of 388 (216, 691) mcg/kg. Out of 89 observation points, 78 were non-sepsis, while 11 observation points were during sepsis episodes. Cumulative dose and duration of PGE1 infusion were similar in two observation groups [648 (374, 979) vs 324 (216, 590); p=0.2 and 22 (10, 27) vs 20 (12, 27); p=0.44 respectively], so was TLC and ANC. Platelet count was lower during sepsis episodes [190 (73, 373) vs 360 (260, 480); p=0.001], while CRP (mg/dl) [38 (24, 100) vs 4 (2.5, 6.4); p< 0.001], procalcitonin (ng/ml) [2.03 (0.49, 8.12) vs 0.19 (0.10, 0.46); p< 0.001] and presepsin (ng/ml) [2.99 (2.79, 3.64) vs 1.91 (1.72, 2.103); p< 0.001] were higher. CRP and procalcitonin cut offs to identify sepsis were 19.55 mg/dl (AUC, 0.95; sensitivity, 91.7%; specificity, 92.2%) and 0.29 ng/ml (AUC, 0.95; sensitivity, 91.7%; specificity, 75.3%) respectively. Presepsin of 2.17 ng/ml had high predictability for sepsis (AUC, 0.99; sensitivity, 100%; specificity, 94.7%).

Conclusions: Prespsin and procalcitonin seem to help identify sepsis in patients on PGE1 for >5 days. CRP >20mg/dL also suggests sepsis. Findings need further confirmation.