Molecular diagnostics of sepsis--where are we today?

Abstract

Rapid diagnosis of sepsis is of outstanding significance as each hour of delay of appropriate antimicrobial therapy increases mortality by 5-10%. As a result, antibiotics are started without a definitive microbial result based on clinical signs in concert with "biomarkers" with high sensitivity but a lack of specificity. Diagnostic uncertainty is compensated for by liberal use of broad spectrum antibiotics with inherent resistance as an increasing public-health problem. Blood culture reflects the current gold-standard but is positive only in approximately 20% of cases and even if positive, results are obtained too late to influence decision making. Culture-independent microbial nucleic acid amplification techniques may allow ways out of this dilemma. In addition to diagnosis of infection, "biomarkers" reflecting the host response can provide valuable information regarding prognosis, course, and response to treatment. Among available single protein markers, procalcitonin (PCT) covers these features best and a PCT-based therapeutic strategy carries potential to reduce antibiotic courses even in life-threatening infections. Recent data from transcriptomic and/or proteomic profiling would, however, indicate that marker panels derived from transcriptomic or proteomic profiling are superior to single proteins to differentiate non-infectious from sepsis-associated systemic inflammation. Multiplexed assay systems, e.g. after platform transfer from whole-genomic chips to multiplexed quantitative PCR are currently being developed with potential to improve sensitivity and specificity. Clinical utility of both, molecular tests to identify the pathogen and the ensuing host response, has still to be evaluated in prospective trials.