Prognostic markers of pediatric meningococcal sepsis

Abstract

Having available tools to determine the prognosis of pediatric meningococcal sepsis at admission to the Intensive Care Unit or during the course of the disease constitutes a clinical necessity. Recently, new readily measurable circulating biomarkers have been described as an additional tool for severity classification and prediction of mortality in meningococcal disease. These biomarkers have been associated with increased risk of mortality scores and a number of organ failures in heterogeneous samples of critically ill children. In future, genetic markers may be used for identification of high-risk patients by creating prediction rules for clinical course and sequelae, and potentially provide more insight in the complex immune response in meningococcal sepsis. We briefly summarize the data pointing at the emerging genome-wide expression profiling studies and review the prognostic value of the main markers investigated in pediatric meningococcal sepsis putting them in the current frame of sepsis in general.

Keywords::

• chemokine
• children
• endothelial marker
• gene-expression
• genome
• interleukin
• meningococcal
• prognostic marker
• sepsis

Meningococcal infection remains a significant health problem, with an extensive mortality and morbidity among children worldwide, even in the era of new-sophisticated vaccines and advanced critical care management. Encouragingly, its mortality has fallen in recent years. The case fatality rate is reported to be 5–10% as a result of several factors such as: the centralization of care of seriously ill children in pediatric intensive care units (PICUs), the development of therapeutic protocols and the dissemination of guidelines regarding early recognition and management [1]. Of those who survive, 10–20% develop permanent sequelae, including skin scars, limb amputation, seizures and brain damage [2,3].

Sepsis is an elusive syndrome rather than a discreet pathologic entity. Its clinical presentation is the culmination of a complex network of interactions between the infecting microorganism and the host immune, inflammatory and coagulation responses. Although great progresses have been made in understanding the pathophysiology of meningococcal septic shock, the knowledge about the mechanisms that trigger the onset of a fulminant course is still very limited. There are compelling evidences of genetic influence in determining the varied outcome of meningococcal sepsis among individuals. Several polymorphisms of the implicated genes are undoubtedly influencing the invasive phenotypes of the disease. Markers of inflammation, immune response and endothelial dysfunction or coagulation imbalance are currently used, in clinical or investigation setting, in order to accurately define the course of sepsis. However, the overwhelming interest on sepsis biomarkers is partly derived from the technical revolution in genetics technology during the last few years. Thus, genome-wide studies of representative meningococcal isolates using high-throughput sequencing are beginning to provide details on the relationship of invasive phenotype and genotype in this fascinating organism and how this relationship has evolved [4]. In addition, the microarray technique, applied in septic patients, allows genome-wide expression profile studies and serves as a valuable tool in further elucidating the underlying pathophysiology of sepsis as well as in discovering newer more accurate biomarkers [5].

Numerous genetic association studies have shown that single-nucleotide polymorphisms in innate immunity genes were associated with susceptibility to and outcome of meningococcal disease in children [6]. Genetic variation in genes encoding for pathogen-recognizing receptors , such as Toll-like receptors can modify inflammatory responses in several cell types, such as dendritic cells and macrophages, all expressing pathogen-recognizing receptors triggering an intracellular signaling cascade resulting in the transcription of pro-inflammatory cytokines and chemokines [7]. It has also been shown that multiple chaperones or co-chaperones, including heat-shock protein (Hsp) 72, tend to form a complex with HS factor 1 monomers. Once inside the nucleus, HS factor 1 binds to a heat-shock element in the promoter of Hsp genes, leading to the synthesis of more Hsps. Newly generated Hsps migrate to the cytosol and interfere with stress-induced apoptotic programs. The most accepted mechanisms for the release of Hsp72 are via extracellular vesicles derived from the plasma membrane (ectosomes), via exosomes or the lysosome–endosome ATP-binding cassette system. Outside cells, Hsp72 induces the activation of macrophages, monocytes, dendritic cells and natural killer cells. Accordingly, extracellular Hsps act as a ‘danger signal,’ further activating immune competent cells through lipoprotein TLR4/CD14-dependent signaling [8]. Unfortunately, immunity-related genes are subject to epigenetic regulation, potentially rendering the host immune-deficient for a long period after the initial sepsis challenge [9]. In addition, extensive recombination, especially changes in virulence genes and capsule genes, could lead to important functional consequences in strains that appear identical by standard classification methods, such as multilocus sequence typing [10]. It has been also found that recombination events may alter penicillin susceptibility and/or intragenic vaccine targets, raising concerns that recombination events could impair treatment efficacy and lead to vaccine escape of Neisseria meningitides.

The clinical significance of a prognostic biomarker should ideally be directly proportional to disease severity and the therapeutic options for that disease, allowing more interventional therapies in those patients with worse prediction. Unfortunately, in the current clinical setting, there is not such a widely-used prognostic marker for pediatric sepsis. Recently, Wong et al have proposed IL-8 as a robust stratification biomarker predicting survival in pediatric septic shock within 24 h of admission in PICU [11]. Traditional markers, however, are more widely used to predict pediatric sepsis outcome. Lower C-reactive protein levels were recorded in non-survivors than in survivors, reflecting the fulminant course of the disease in those with low C-reactive protein admission levels [12]. Procalcitonin admission levels were successfully linked with severity of disease in children with meningococcal sepsis, while the persistence of high procalcitonin levels despite therapy was associated with increased prediction of mortality risk scores [13]. Differences in plasma plasminogen activator inhibitor type 1 levels, reliably predicting mortality in patients with similar levels of TNF-α production, were assumed to be due to polymorphisms in the plasminogen activator inhibitor type 1 gene (SERPINE1) [14]. In contrast, the substantial inter-individual variability of plasma thrombomodulin antigen levels diminishes its prognostic value [15]. Two polymorphisms (–1645C>T and –1641A>G) in the promoter region of the protein C gene (PROC) showed a significant age-dependent association with susceptibility and severity but not with outcome [16]. Nonetheless, the increased incidence of myocardial ischemia associated with severe coagulopathy in acute meningococcemia in pediatric patients suggested early replacement therapy with protein C, AIII, factor VIII or fibrinogen [17]. Thus, a substantial amount of work, including validation, remains to be done in order to timely organize unbiased biomarkers datasets for sepsis using effective patient-stratification strategies. Marshall has proposed that a key challenge in the field is to reduce and manage sepsis heterogeneity by more accurately stratifying patients for the purpose of robust clinical research [18].

Abnormal flow patterns have also been seen in patients presenting with severe sepsis and septic shock due to microcirculatory flow pattern dysfunction. There were negative correlations between the ICAM-1, VCAM-1, E-selectin levels and the microcirculatory values at the time of admission to PICU [19]. E-selectin, ICAM-1 and VCAM-1 were increased in both early and late pediatric sepsis [20], whereas meningococcal septic shock was associated with a significant and persistent increase in circulating soluble co-inhibitory immune receptor carcinoembryonic antigen-related cell-adhesion molecule 1 concentration from 24–48 h up to day 7–8 following PICU admission [21]. Closely related to microcirculatory dysfunction, admission lactate levels correlated with septic shock and prolonged hospital stay or fatal outcome [22]. Surprisingly, though, hypoinsulinemia rather than sepsis-related insulin resistance has been demonstrated in meningococcal septic shock [23]. In addition, in children with sepsis the early metabolic pattern was characterized by a high (nCD64, glucose, triglycerides)–low (cholesterol, HDL, LDL) biomarker combination in contrast to the moderate pattern of traumatic brain injury, in which only glucose increases, combined with moderate cholesterol–lipoprotein decrease [24].

A logic explanation of these inflammatory-metabolic derangements could be a dynamic and ever-shifting balance between pro-inflammatory cytokines and anti-inflammatory components of the human immune system that can be tilted by time-related dysfunction of one or more cytokines. The regulation of inflammation by cytokines and cytokine inhibitors might further be complicated by the fact that the immune system has redundant pathways with multiple elements having similar physiologic effects. Thus, although during acute illness HSP70/inflammatory mediators were significantly increased and glutamine significantly depleted, glutamine and HSP70 were not correlated [25]. We have recently shown that high doses of glutamine in septic human peripheral blood mononuclear cells could not induce any of the type-1 T helper (Th-1), Th-2 and Th-17 cytokines, although it suppressed Hsp72 [26]. In contrast, lipoprotein induced IL-6 and Th-2 (IL-10) but not Th-1 (IFN-g), Th-17 (IL-17), Hsp72 or Hsp90-α. Importantly, although extremely high levels of IL-1Ra and IL-10 have been detected in patients with meningococcal septic shock only the homozygous IL-10 genotype-1082 A/A has been associated with disease severity alone or combined with FcγRIIa genotypes [27]. Similarly, although the –174 C/G polymorphism in the promoter region of IL-6 gene has been associated (-174 G allele) with increased IL-6 release and the homozygous G/G phenotype with the risk of death among infected patients, it was not shown to increase the risk of invasive meningococcal infection.

Efforts to derive and validate a multi-biomarker sepsis outcome risk model should probably be based on large cohorts of children with septic shock, thereby including the meningococcal sepsis sub-cohort. Population genomics, the application of ‘next-generation’ sequencing methods to large numbers of meningococcal isolates, is currently generating novel insights into the biology of the meningococcus with implications for improved understanding of the biology of this disease and the development of novel methods for its control [4]. In this setting, the newly tried PERSEVERE, a risk model for estimating mortality probability in pediatric septic shock, using five biomarkers measured within 24 h of clinical presentation, could potentially serve as an adjunct to physiological assessments for monitoring how risk for poor outcomes changes during early interventions in meningococcal septic shock [5].

Today, a magnitude of biomarkers has been tested in sepsis patients and in meningococcal sepsis, as well. However, not a single biomarker approach has yet met the desired clinical benefits. Therefore, multiple stratification biomarkers based on genome-wide expression profiling are under active investigation and present exciting future possibilities [28]. To sum up, in the current frame of sepsis, it is unlikely that any single biomarker would serve as a certain predictor of outcome to the whole children population. The complexity of interactions during the host immune response and the host genetic variability make this effort rather impossible. A multi-biomarker aspect may perform better in predicting outcome and the adoption of pediatric patients’ subclasses, according to their genome expression profile, is possibly a reasonable approach in order to identify high- and low-risk groups. Unfortunately, a marked variability in differential gene expression has recently been observed between time points and between patients revealing dynamic expression changes during the evolution of sepsis [29]. Thus, while there was evidence of time-dependent changes in expected gene networks including those involving immune responses and inflammatory pathways, temporal variation was also evident in specific ‘biomarkers’ that have been proposed for diagnostic and risk stratification functions, seriously limiting a static or single time point biomarker estimation. It seems, therefore, that serial estimations or more comprehensive network approaches may be required to optimize risk stratification in complex, time-critical conditions such as evolving meningococcal sepsis [29].

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.