PROJECT SUMMARY
Melioidosis is a tropical infection caused by inoculation, inhalation, or ingestion of the Gram-negative soil
saprophyte and Tier 1 select agent Burkholderia pseudomallei (Bps). The overall melioidosis mortality rate
exceeds 40% in endemic areas of southeast Asia such as northeast (NE) Thailand (despite appropriate
treatment), and modeling indicates that 165,000 cases of human melioidosis occur annually worldwide. For
decades, diagnosis of melioidosis has required culture of Bps from a clinical specimen. This may take several
days, delaying appropriate treatment, and in many resource-limited settings the necessary microbiology
facilities for bacterial culture and identification are not available. Few suitable non-culture-based diagnostics
exist. Performing a case-control analysis nested within a prospective, single-center cohort study of patients
hospitalized with infection (Ubon-Sepsis study), we have developed a five-metabolite signature in plasma that,
in preliminary studies, has extremely high accuracy differentiating melioidosis from other causes of infection
(misclassification error rate=2.0%, AUC=0.99). These results suggest that measuring a limited number of
circulating metabolites during initial presentation has significant diagnostic potential for melioidosis. We
hypothesize that this host metabolomic signature is a novel and accurate diagnostic tool in identifying patients
with melioidosis. To test this potentially high impact hypothesis, we will leverage our singular expertise in
melioidosis diagnostics and human immunology, and an unparalleled multi-center melioidosis research
infrastructure in NE Thailand in the following specific aims: 1) Replicate the diagnostic accuracy of the five-
metabolite melioidosis signature in plasma for the detection of melioidosis in additional cases in the Ubon-
Sepsis study. 2) Validate the diagnostic accuracy of the five-metabolite melioidosis signature in plasma for the
detection of melioidosis in an independent prospective multi-center study. 3) Use an unbiased approach to
identify additional signatures of melioidosis in plasma. If our hypothesis is proven, the use of blood metabolites
as a melioidosis diagnostic could greatly enhance our present approaches to identifying melioidosis.
Subsequent development of this tool for clinical use could have a profoundly beneficial impact on the burden of
this often-lethal infectious disease.