A host-directed diagnostic methodology, which uses only 30 µL of whole blood, and enables rapid and accurate detection of sepsis in neonates and other subjects. The test removes false negative results associated with current gold standard microbiological methods.
Current treatment protocols for sepsis in vulnerable patients oblige administration of broad spectrum antibiotics at the first clinical signs. This leads to unnecessary antibiotic use, can cause liver problems, bacterial resistance, and may adversely affect the developing microbiome in neonatal patients. Furthermore, the blood culture protocol requires a large blood sample draw, takes 3-4 days to confirm a diagnosis, and is susceptible to false negative results due to the reliance on pathogen detection. An improved diagnostic tool for sepsis is therefore required.
Edinburgh researchers have developed a new test, based on monitoring up/down-regulation of a small panel of host genetic markers from three biological pathways using commercial multiplex microarray platforms. It is rapid (hours to diagnosis); sensitive (detects host RNA expression); specific for bacterial infection (reducing unnecessary antibiotic usage); only uses a small sample volume (30 µL whole blood); and is more likely to reduce false negatives (due to stability/presence of host RNA).
The marker panel was initially validated using a combination of statistical analysis and empirical measurement of 62 neonatal clinical samples, showing 98% accuracy. It was further validated on independent patient samples and shown to be consistent across several different expression platforms. The marker set has also been ‘blind validated’ for clinical utility, using samples from clinically suspected cases of sepsis that tested negative by cell culture (possible false negative samples). Further studies (222 cases, across three separate African/European studies) have confirmed it also provides robust and accurate detection of sepsis in children, infants and adults.
PCT filed in April 2015 (Reference WO/2015/155517 A1)Publications:
- Nat. Commun. 5:4649, doi:10.1038/ncomms5649
- Genomics Data 3 (2015) 41–48, doi:10.1016/j.gdata.2014.11.003
- Up to 98% accuracy with as few as 4 gene markers.
- Validated across three microarray platforms using clinical samples.
- ‘Blind validated’ on potential false negative samples.
The University of Edinburgh is seeking commercial partners to license this technology and/or collaborate on further development of the technology for commercial use.