How might a diagnostic microRNA signature be used to speed up the diagnosis of sepsis?

Abstract

Sepsis is a heterogeneous illness characterised by inflammation secondary to suspected or proven infection. A clinical and research challenge in this area is the ability to diagnose true sepsis, defined as inflammation secondary to infection. Infection is often indirectly confirmed using surrogates, whilst awaiting microbiological confirmation. microRNAs are novel molecules with a potential to be point of care rapid diagnostic test for true sepsis.

Keywords::

• diagnosis
• humans
• infection
• micrornas
• sepsis

Sepsis is a clinical syndrome defined by the simultaneous presence of infection and systemic inflammatory response [1]. The manifestations of systemic inflammatory response are protean, and the infection component is defined as a pathological process induced by invasion by a pathogenic or potentially pathogenic microorganism [1]. This illness model ignores the complex intermediate biology that is non-homeostatic [2] and phenotypically defines a final common pathway whereby organ failure and death ensues [3]. This issue coupled with the abject failure of drug trials in sepsis has resulted in calls for fundamentally readdressing the sepsis illness construct both clinically and in basic research [4–6].

Sepsis could be better conceptualized as an illness continuum, which begins when infection reaches an inflection point [7], triggering a systemic immune response that has both pro- and anti-inflammatory components, which could lead to immunological exhaustion, organ failure and death [8]. Within this paradigm, identifying and treating infection early could be achieved either using detection of pathogens or using surrogates that differentiate sterile inflammation from infected inflammation. The literature is replete with biomarkers of sepsis [9]; however, most fall short by not having either a mechanistic correlation to illness causality or specificity and often both. Thus, an unmet clinical need is the ability to differentiate between sterile and infected inflammation at the bedside. This article focuses on the potential utility of miRNA molecules or panels to differentiate between sterile inflammation and infection inflammation (true infection or true sepsis) in acute hospital admissions requiring antibiotic therapy and/or management in a critical care setting.

miRNAs are a group of single-stranded 18- to 25-nucleotide long RNA sequences that regulate post-transcriptional gene expression. miRNAs are most often located in intergenic regions but can also be found in the intronic regions of protein coding genes which the miRNAs may regulate. For example, miR-4772 family genes are located in the Chromosome q12.1 intron 5 of IL-18 receptor-associated protein [10,11]. The chromosomal location of the miRNA thus lends itself for the two key regulatory mechanisms of action, namely context-dependent transcriptional activation and (potentially reversible) default repression [10,12]. Thus, the chromosomal location and mechanisms of post-transcriptional regulation by miRNA provide a biological system where the role of miRNA in a particular pathway can be studied. In the context of infection-induced critical illness in humans, miRNA-based molecular signatures have been proposed as surrogate markers that could differentiate between sterile inflammation and sepsis [11], be diagnostic for sepsis [13], influence drug response in sepsis [14], be responsible for altered protein synthesis in muscle during critical illness [15] and potentially have prognostic utility [16,17].

Detection of infection and or danger signals by the pattern-recognition receptors (PRRs) such as toll-like receptors results in activation of clearly defined biological pathways early on the course of illness which miRNAs can influence (e.g., toll-like receptor 4 signaling, endotoxin-mediated effects and MyD88 pathways) [18–20]. These pathways are relevant to sepsis and PRR pathway-specific miRNA signatures and have a potential role as diagnostic biomarkers to differential sterile versus infected inflammation. In a heterogeneous cohort of 166 patients with sepsis, a reduction in miR-499-5p was associated with prediction of sepsis-related severity of illness [13]. In this study, at a cut-off point of 0.751, the area under the receiver operated curve for miR-499-5p was 0.686 (0.592–0.779) with a sensitivity and specificity of 68 and 65% respectively [13]. Another study using a cohort of 50 patients with culture-positive sepsis following major abdominal surgery or trauma with postcardiac surgical patients as systemic inflammatory response syndrome controls, a reduction in miR-146a and miR-223 was observed and proposed as a potential biomarker in sepsis [21]. The area under the curves (95% CI) for miR-146a and miR-223 were 0.804 (0.679, 0.928) and 0.858 (0.748, 0.968), respectively. Our group has recently shown an increase in the human-specific miRNA miR-4772-5p-iso, which has the potential to discriminate sterile and infected inflammation [11]. This was the first study to identify a novel miRNA that is increased in patients with sepsis, and when combined with downregulation of miR-150, the area under the curve was 0.89 with a specificity of 92.9% and a sensitivity of 81.8%. This miRNA signature was then replicated in an independent patient group, and these two combined cohorts gave an overall diagnostic accuracy for sepsis of 86%.

These three studies focusing on the discriminatory ability of miRNA for infected inflammation also highlight the challenges of biomarker research in sepsis. A biomarker that increases in an illness is intuitively easier to translate into clinical practice. Furthermore, the biomarker that is identified should be validated in different cohorts and in their ability to discriminate different primary sites of infection. The trajectory of illness is important as the relatively specific early pathways often coalesce in the later stages of sepsis, reducing the specificity of miRNA signature. Therefore, we propose that future studies should focus on a specific illness that causes sepsis, such as acute exacerbation of chronic lung disease where there is an inherent need to differentiate infective from non-infective exacerbations. A multipanel miRNA diagnostic test with specific miRNAs aimed at sensitivity combined with those that add specificity to diagnosing the presence of infection would be a welcome addition to clinical care.

The concept of true sepsis will evolve to match both the basic science and pre-clinical models of sepsis. It will be necessary to define the inflammatory and infectious pathways that miRNA regulate, the half-life of specific miRNAs and the utility of repeated measurements in improving diagnostic accuracy and clinical utility. The most likely clinical impact will be from their ability to differentiate inflammation secondary to infection – that is, true sepsis – thus becoming triage markers in the pre-intensive care unit setting for enhanced early management. Furthermore, in order to improve the stratification of novel therapeutics for sepsis, miRNA signatures may become part of a ‘biological passport’ used for enrolling patients with sepsis into clinical trials.

Financial & competing interests disclosure

GM Lord has received consulting fees and experimental costs from Cepheid, with whom he has collaborated on a study of sepsis diagnosis. The authors have no other 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 apart from those disclosed. 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.