Skip to main content

Personalised Risk assessment in febrile illness to Optimise Real-life Management across the European Union

Periodic Reporting for period 2 - PERFORM (Personalised Risk assessment in febrile illness to Optimise Real-life Management across the European Union)

Reporting period: 2017-07-01 to 2018-12-31

Fever is the commonest reason that children are brought to medical care. Although the vast majority of episodes of fever in children are caused by viral infections which resolve without treatment, hidden in the vast number of children with fever attending hospitals and medical services are a small proportion whose fever is caused by potentially life threatening bacterial infection. A major problem in the provision of paediatric services worldwide is detecting the small number of children with bacterial infections among the more common viral illnesses. Nearly a quarter of all children in hospital emergency departments have a febrile illness and less than 5-10% of these turn out to be bacterial in origin. Current diagnosis largely relies on culture of bacteria from blood, urine or spinal fluid, but the results of bacterial cultures may take 48 hours or more to become available, so do not influence the decision on whether to administer antibiotics. Blood tests such as white cell count and protein markers such as c-reactive protein have poor sensitivity and specificity. In the absence of reliable tests a large proportion of febrile children are treated with antibiotics when in fact they suffer from viral infections, while others with serious bacterial infections are sent away only to return critically ill.
PERFORM's aim is to apply new molecular and protein methods to identify biomarkers in the blood of children with fever to distinguish bacterial from viral infection. Rather than attempting to identify the causative bacteria, PERFORM aims to identify the pattern of genes and proteins activated by the host infection to identify a” signature” distinguishing bacterial from viral infection. Pilot data showed that as few as two genes expressed in the peripheral blood during infection might accurately distinguish bacterial from viral infection. By studying a cohort of 2000+ children recruited in a previous grant, applying sophisticated RNA expression and proteomic methods as well as bioinformatic analysis to identify a small number of genes or proteins that distinguish bacterial from viral infection. We then plan to validate these biomarkers of bacterial infection in a newly recruited cohort of over 5000 children from countries in Europe and West Africa. In the final phase of the project we aim to translate the most promising protein and RNA expression markers into simple rapid, affordable tests that could be further validated prior to introduction into clinical use in Europe.
By month 36, PERFORM established patient recruitment in 9 European countries, the Gambia and Nepal. The clinical protocol for collection and handling patient data, samples and subsequent analysis received ethical approval in all partner countries and all regulatory approvals to start patient recruitment were concluded. An electronic database for recording patient information and sample tracking was set-up. So far, over 4800 patients had been recruited with samples for protein and RNA extraction available.
Through recruitment of a new patient cohort, analysis of the previously recruited EUCLIDS cohort started, through the biobank of RNA and plasma samples previously collected. Initial study of the genes activated during the course of infection confirmed our previous findings that the pattern of RNA expression in blood can distinguish bacterial from viral infection. Proteomic analysis by Mass spectrometry identified a number of proteins that distinguish bacterial from viral infection in the “discovery cohort” and the results are currently being validated both in a second validation cohort by mass spectrometry, and by immunoassay on the PERFORM prospective cohort.
An important aim of PERFORM is to develop a rapid test for the diagnostic RNA signature. Our bioinformatics team has identified from the RNA expression data small signatures containing candidate transcripts. Although the final RNA signature will still be refined as more RNA expression data becomes available, work has begun on the development of a rapid test using the RTPCR based methods through industrial partner, Biomerieux. Initial results show that RNA diagnostic signature can be detected rapidly and with high accuracy. This confirmation that RNA signatures detected by RNA sequencing and micro array can also be detected using technology appropriate for POC devices provides strong support for a rapid test.
In addition to the studies to identify RNA and protein signatures distinguishing bacterial from viral infection, PERFORM is conducting an observational study to understand how fever in children is managed in different European settings. This study collected data from over 40000 emergency department attendances in 9 European countries and ongoing analysis is providing information on clinical management in different settings and also on resource use and anti-microbial prescribing.
The diagnostic WP is applying state-of-the-art molecular diagnostics to detect bacteria, viruses and other pathogens in samples of blood, and throat swabs from all children recruited. This data supplements the routine diagnostics performed in each participating hospital’s microbiology laboratories, leading to more accurate assignment of patients into specific diagnostic groups. Micropathology has developed novel molecular assays for a wide range of viruses and bacteria. Over 1000 patients have complete screening of blood, nasopharyngeal secretions for this extensive panel of pathogens with initial analysis providing a new picture of the causes of fever in Europe.
A major outcome has been to establish a bioinformatic pipeline combining clinical, laboratory, RNA and protein information for subsequent analysis including a large biobank of over 5000 samples. A joint database, online patient recruitment database and secure system for storage of all protein and molecular data has been set-up. PERFORM has met its patient recruitment numbers and collected data from partner European countries and West Africa, made progress in its RNA and protein analysis, and the bioinformatic analysis of all RNA expression, molecular microbiology and protein data.
PERFORM has set in progress the largest RNA expression study ever undertaken to identify the pattern of host genes expressed in the blood of children with fever to accurately distinguish bacterial from viral infection as well as distinguishing infections from inflammatory causes of fever. Data so far confirmed that bacterial infection can be distinguished from viral infection with a small number of RNA transcripts. Concurrent proteomic analysis has identified novel protein biomarkers which may improve distinction of bacterial from viral infections. Our observational study on management of febrile children across European countries provides a unique picture of how children with fever are managed across Europe and West Africa and informs ongoing studies on how to improve management. The major expected impact of this action will be identification and development of improved methods to diagnose bacterial infection and distinguish bacterial infection from other causes of fever in children. The work is likely to result in improved model for care of febrile children across Europe and ultimately in reduction of the use of unnecessary antibiotics in those children suffering from viral infection, therefore contributing to the global effort to reduce anti-microbial resistance.
The PERFORM Clinical Network