Project Title

Proteomic-based Etiological biomarkers of Acute pneumonia in Kenyan children (PEAK)

Project Objectives

Aims of the proposed study To develop a protein microarray to measure the expression level of host proteins in the respiratory tract of infants and children with pneumonia To retrospectively screen samples obtained from infants and children with viral (n=250) and bacterial pneumonia (n=250) in a blinded manner and to identify the top three predictors of pneumonia etiology To prospectively apply the top three predictive biomarkers in children admitted with clinical pneumonia in Kilifi County hospital over period of six months.

Host Organisation

EDCTP Project Call

Career Development Fellowship (CDF)

Study Design

clinical trial

Project Summary

Pneumonia is a major cause of paediatric mortality in low and middle income countries. Pneumonia can be caused by viral or bacterial infections and its clinical management depends on the underlying microbial aetiology. In developed countries, these infections are often diagnosed using sophisticated platforms including multiplex real-time PCR for virus diagnosis and blood cultures for the diagnosis of bacterial pneumonia. Most lower income countries lack widespread access to these diagnostic platforms. In many clinical settings, pneumonia diagnosis relies on assessment of clinical features, which overlap considerably and cannot be used to distinguish between pneumonia of viral or bacterial aetiology. This situation has led to the widespread use of presumptive antibiotic treatment, a practice that is thought to be the major driver of antimicrobial resistance (AMR). In this study, we identified host level biomarkers that could be used to characterize the underlying aetiology of pneumonia as bacterial or viral. We developed a microarray chip to quantify the expression level of 15 host proteins that were identified in a previous pilot study as being differentially expressed in viral and bacterial pneumonia. We used this chip to retrospectively analyze protein expression levels in 250 children who had previously been admitted to hospitals with virologically-confirmed viral pneumonia and 250 children who had previously been hospitalized with microbiologically confirmed bacterial pneumonia. We then used these data to identify 3 biomarker proteins (CES2, CSNK2B, CTSG ) that exhibited the best performance in distinguishing between aetiologies in this retrospective cohort. We then carried out a prospective study to measure the expression of these markers in real time in a population of children who were admitted to hospital with severe pneumonia. The results of this study showed that children with bacterial pneumonia overexpressed nasopharyngeal CTSG and CES2 relative to children with viral pneumonia. These markers could distinguish between aetiologies with a sensitivity of 70 and 59% respectively and a specificity of 72 and 63% respectively. These results indicate that these markers may provide an important avenue for accurately classifying the microbial aetiology of pneumonia and reducing the spread of AMR.

Current Organisation

Current Job Title

Biography