Neutrophil CD64 for Daily Surveillance of Systemic Infection and Necrotizing Enterocolitis in Preterm Infants

Abstract

BACKGROUND Early detection and treatment of infected preterm infants could decrease morbidity and mortality. Neutrophil CD64 has been shown to be an excellent early diagnostic biomarker of late-onset sepsis (LOS) and necrotizing enterocolitis (NEC). We aimed to study whether using CD64 as a daily surveillance biomarker could predict LOS/NEC before clinical manifestation. METHODS We collected 0.1 mL whole blood from very low birth weight (VLBW) infants from day 7 postnatal age until routine daily blood tests were no longer required. Four categories of responses were defined: proven sepsis, clinical sepsis, nonsepsis/non-NEC, and asymptomatic CD64 activation. RESULTS A total of 146 infants were consecutively recruited and 155 episodes of sepsis evaluation were performed. The biomarker screening utility, sensitivity, specificity, positive predictive value, and negative predictive value for surveillance of LOS/NEC using a cutoff of 5655 antibody-PE (phycoerythrin) molecules bound/cell were 89%, 98%, 41%, and 99.8%, respectively. LOS/NEC was detected a mean of 1.5 days before clinical presentation. However, 63 episodes of CD64 activation occurred in asymptomatic infants who would not otherwise have required sepsis evaluations. CONCLUSIONS As a surveillance biomarker, neutrophil CD64 detected LOS/NEC 1.5 days before clinical presentation, but at the expense of performing 41% additional sepsis evaluations. This was mainly attributed to an unexpected group of asymptomatic infants with CD64 activation, who recovered spontaneously and did not require antimicrobial treatment. The latter group has not been previously recognized in VLBW infants and could represent subclinical infection secondary to transient bacterial translocation or mild viral infection.

Rapid Identification of Bacterial Antibiotic Resistance by qPCR in Infants with Gram-Negative Septicaemia: A Proof-of-Concept Study

“…Infants admitted to our unit between September 2011 and December 2014 who were evaluated for suspected LOS had a sample of 0.5 ml whole blood collected simultaneously at the initial sepsis evaluation. Sepsis evaluations were performed upon recognition of clinical features of neonatal sepsis in accordance with our unit protocol, as described in previous studies [6,20], before commencement of broad-spectrum empirical antibiotics. In addition, 7 infants who were previously recruited into our past study [8] according to the same criteria had residual blood samples archived.…”

Understanding and Improving Diagnostic Tests: The Clinician Perspective

“…A daily blood test was feasible and ethically approved because a very tiny amount of capillary blood (0.05 mL equivalent to 1–2 drops of blood) was used, and the timing of specimen collection could coincide exactly with other daily blood-taking procedures. Importantly, we have also proven with paired capillary/venous samples that neutrophil CD64 levels had excellent statistical “agreement” (not just a correlation) between the two types of specimens ( r = 0.99), as a quantitative cut-off value is a prerequisite for decision-making [3]. Perhaps, the most difficult and crucial issue in any surveillance study, especially for screening of acute life-threatening disease, is to act on the result in asymptomatic patients.…”

“…Nonetheless, this may not always be possible. For example, we have performed a daily surveillance study using neutrophil CD64 as the blood biomarker for screening for systemic infection and NEC in preterm infants [3]. A daily blood test was feasible and ethically approved because a very tiny amount of capillary blood (0.05 mL equivalent to 1–2 drops of blood) was used, and the timing of specimen collection could coincide exactly with other daily blood-taking procedures.…”

“…Currently, the majority of these tests are concentrated on “early” diagnosis [1, 2], but some are designed for screening or surveillance purposes, i.e. before the initial clinical presentation [3, 4], differentiating a specific subset of patients in a condition [5], and predicting the severity or prognosis of an illness at its onset [6]. All of these aspects are of great clinical and practical importance to neonatologists in decision-making.…”

“…a study with good clinical utilities, and (iii) to recognise recent major advances in this field and to explore future developments and applications. For ease of understanding and reference, “neonatal infection” [1, 3-6] and a closely related pro-inflammatory condition, “necrotising enterocolitis (NEC)” [2], are frequently quoted for illustration purposes, though other specific conditions are also used as examples in this article.…”