Validation of Molecular and Protein Biomarkers in Sepsis

Summary

Participation Requirements

Description

Sepsis, defined as life threatening organ failure resulting from a dysregulated host response to infection, remains a leading cause of death in critically ill patients and has been included as a health priority in a 2017 WHO resolution. Diagnosis of this disorder is challenging because the clinical ...

Sepsis, defined as life threatening organ failure resulting from a dysregulated host response to infection, remains a leading cause of death in critically ill patients and has been included as a health priority in a 2017 WHO resolution. Diagnosis of this disorder is challenging because the clinical signs and symptoms of systemic inflammation in sepsis overlap with those of non-infectious critical conditions i.e. severe inflammatory response syndrome (SIRS) e.g. cardiac arrest and burns. Early and accurate diagnosis of sepsis is critical for improving patient outcomes and reducing antibiotic usage. Delays in antibiotic administration are associated with worse outcomes; however paradoxically, indiscriminate prescription of antibiotics to patients without bacterial infections increases both rates of morbidity and antimicrobial resistance. The rate of inappropriate antibiotic prescriptions in the hospital setting is estimated at 30 to 50% and would be decreased by access to improved diagnostic tests. There is currently no gold standard laboratory test that can broadly determine the presence and type of infection. Although new polymerase chain reaction (PCR)-based molecular diagnostics can profile pathogens directly from blood culture, they suffer from sensitivity issues due to dependence on sufficient numbers of pathogens in the blood sample. They are also limited to detection of a discrete range of pathogens. As a result, there is a growing focus on molecular diagnostics that profile the host immune response. Current sepsis groupings are based on clinical criteria such as the presence of shock, infection source, or organ failure, but such groupings may not represent the underlying biology driving the host response. They have also failed to adequately match patients for novel interventions. If the heterogeneity of sepsis truly reflects heterogeneity in the host response, characterisation of these underlying host response types will be fundamental to enabling precision sepsis therapeutics. In a previous multi-centre, clinical-temporal study in three cohorts of patients admitted to the intensive clinical care unit (ICU); (i) out of hospital cardiac arrest (n=36 - SIRS group)) (ii) pulmonary sepsis (n=84) (iii) abdominal sepsis (n=64) and 30 healthy controls, validated potential host immune biomarkers. Using 202 samples from these cohorts, the investigators derived a set of gene biomarkers which can identify patients with severe inflammation and discriminate sepsis from non-infected inflammation across a broad range of clinical conditions. Other biomarkers have been identified for use for other purposes e.g. prognosis/severity. Our patent arising from this work has been filed and entered PCT stage. From these patented markers a parsimonious set of 17 genes has been further delineated, which are under further evaluation. A sub-panel of two gene entities has been identified that can accurately detect severe inflammation using receiver operating characteristic/area under the curve (ROC) analysis with a value of approximately 0.98. A panel of three/four gene entities has been identified for discrimination of SIRS from all sepsis types (ROC 0.89-0.92), all depending on sensitivity or specificity range settings. Better diagnostics for sepsis-driven inflammation are needed in both inpatient and outpatient settings. In low-acuity outpatient settings, contributing circa 80% of total UK antibiotic use, a simple diagnostic to discriminate a septic inflammatory process from an innocuous, self-limiting condition, would assist in appropriate antimicrobial use, appropriate triage, avoiding further investigations, and appropriate escalation / admissions. In higher-acuity settings, causes of non-infectious inflammation are important to exclude; a decision model for antibiotic prescription should include a non-infected, non-healthy cause. A reliable diagnostic, such as ours, needs to distinguish all three presentations: non-infected inflammation, sepsis, and relative health. It will represent a major step-change in provision of diagnostic/stratification capability, vastly improve decision and patient management pathways and potentially reducing antibiotic overuse in the acute medical and critical-care environment. These biomarkers once validated in an independent cohort via qPCR for mRNA and their commensurate proteins, together with an accompanied easy-to-use, clinically oriented scoring system will represent a complete data package which can be rolled out internally, subject to the appropriate further accreditation and/or leveraged for development of point-of-care devices by commercial partners. This latter option could prove useful for dissemination of the test to other patent-appropriate global territories. There is an urgent need to validate these findings in several ways; Using an independent patient cohort, where laboratory scientists are blinded to the clinical phenotypes of the recruited patients and clinicians are not aware of the gene expression data. Using a bioinformatics approach to validate the results in already published datasets This project will use different approaches to validate these novel SIRS or sepsis-associated biomarkers identified by Artificial Neural Network (ANN) and parametric data mining of previously published datasets and further validated previously from a previous well characterised clinical cohort; Public Health England laboratories will assess biomarker mRNA expression using a qPCR approach, with RNA purified from patient and control whole blood Public Health England and Cardiff laboratories will assess protein biomarkers using ELISA assays Proteomic analysis of blood by external collaborators Further data analysis will be conducted using ROC curve analysis and arithmetic algorithms and/or other statistical/bioinformatics methods. Assessment of specificity and sensitivity and positive and negative predictive values using well established methods will also be conducted to evaluate the performance of the biomarker panels in discriminating patient control and disease groups.

Tracking Information