Diagnostic Markers of Neonatal Sepsi

The use of diagnostic markers in neonatal sepsis has remained an important area of research over the past decades. Many infection markers are components of the inflammatory cascade and reflect the host's immunological status and response to infection. Cytokines, acute phase reactants and leucocyte surface antigens are the most frequently investigated candidate markers of neonatal sepsis. The mainstay of treatment of neonatal sepsis remains intravenous antibiotics and supportive therapy. One of the main difficulties in the management of neonatal sepsis is that the clinical features of sepsis can be subtle and non-specific early in the course of the infection. This can lead to delayed initiation of antibiotic therapy. In addition, there are important limitations of the current gold standard for the diagnosis of neonatal sepsis, i.e., blood culture. Difficulties include the long turnaround time required for conventional culture techniques. This means that broad spectrum antibiotics cannot be rationalised or escalated appropriately at a very early stage of the infection. This thesis is intended to cover the various strategies of diagnostics that I have investigated in the context of neonatal sepsis. Strategies include: (i) as a screening test as part of a surveillance programme, (ii) in combination with other clinical tests to improve diagnostic utilities for specific conditions, (iii) in combination with other novel diagnostic tests to facilitate clinical decision-making and (iv) as a test to obtain pathogen specific information such as identity and antibiotic resistance patterns. The main hypotheses of the studies highlighted in this thesis are:;1. Highly sensitive biomarkers in the early phase of neonatal sepsis, such as neutrophil CD64, can effectively screen for late-onset sepsis (LOS)/necrotising enterocolitis (NEC) in asymptomatic infants.;2. Combination of non-specific inflammatory markers with routine clinical investigations can help target specific inflammatory conditions, such as intra-abdominal inflammation/sepsis.;3. "Hypothesis-free" marker discovery approaches can enable neonatologists to discover novel and useful neonatal sepsis markers.;4. Combination of sepsis markers with complementary properties can further increase diagnostic utilities.;5. Application of molecular techniques to the investigation of neonatal sepsis can yield actionable and clinically relevant information early in the course of neonatal sepsis, such as Gram-identity and antibiotic resistance profile of the invading pathogen.;In chapter 2 the possibility of screening for LOS/NEC as routine surveillance of high-risk infants in order to detect infection before clinical presentation (hypothesis 1) is explored. The marker used in the study was neutrophil CD64 measured in preterm infants at daily intervals. The main findings of the study were that while LOS/NEC could be detected 1.5 days before clinical presentation, it would be at the "cost" of 41% increase in sepsis evaluations and antibiotics use than current levels. Chapter 3 illustrates one of the strategies that could be employed to facilitate diagnosis of intra-abdominal inflammation/sepsis by combining an early, sensitive biomarker with a routine clinical investigation (hypothesis 2). In this study, neutrophil CD64 was combined with C-reactive protein and abdominal radiographs and it was found that the combination with the optimal diagnostic utilities was neutrophil CD64 and abdominal radiographs. This combination could achieve high sensitivity and negative predictive value (both 0.99) while achieving a reasonable specificity (0.8). This suggests that frontline clinicians could confidently rule out intra-abdominal inflammation/sepsis shortly after clinical presentation with this diagnostic combination. The use of marker panels (in this thesis, taken to mean combinations of two or more markers) to improve diagnostic utilities during the early phase of LOS/NEC is illustrated in chapter 4. By using a hypothesis-free proteomic approach, proteins associated with LOS/NEC were discovered (hypothesis 3). Serum amyloid A and apolipoprotein C2 were two of the most useful markers found, and were subsequently combined in a diagnostic score (ApoSAA score) by using logistic regression modelling. The score was found to perform well with area under receiver operating characteristics curve of 0.93 in the prospective cohort validation phase (hypothesis 4). In chapters 5 and 6, I describe how the use of quantitative polymerase chain reaction can enable us to obtain actionable information about the invading pathogen, such as Gram-identity and antibiotic resistance profile shortly after clinical presentation, so that prompt escalation or rationalisation of antibiotic therapy is possible (hypothesis 5). In chapter 5, we were able to demonstrate that appropriately designed gene probes could directly detect bacterial genetic materials in neonatal blood, thereby helping the clinician diagnose Gram-negative and Gram-positive septicaemia with high specificity. In chapter 6 the primer/probe sets could be designed to accurately predict antibiotic resistance patterns in common Gram-negative pathogens encountered in the neonatal setting. (Abstract shortened by ProQuest.).