Performance evaluation and further development of the PCR and microarray-based Prove-it™ Sepsis assay

Abstract

Introduction: Superior clinical management of septic patients may result from the more rapid speciation of organisms in positive blood cultures. The Prove-it TM Sepsis assay is a rapid, broad-range PCR and microarray-based assay designed to identify the majority of bacterial pathogens responsible for sepsis from positive blood cultures. The pathogen panel covers 50 Gram-negative and Gram-positive bacterial species; 24 at the species level and 26 at the taxon level. It also reports methicillin resistance by detecting the mecA gene in addition to differentiating between Staphylococcus aureus, Staphylococcus epidermidis and coagulase negative staphylococci (CNS). DNA is extracted from a positive blood culture and the topoisomerase and mecA gene regions are amplified by PCR. These PCR products are subsequently overlaid on the Prove-it TM TubeArray, where hybridization is detected in a single reaction. Detection is by solidstate hardware, and interpretation and reporting is via Prove-it TM Advisor software. The assay time is 3 hours including the DNA extraction. Objectives: To conduct a performance evaluation study for the Prove-it TM Sepsis assay according to the EN 13612-standard: "Performance evaluation of in vitro diagnostic medical devices" and to compare obtained results to those of current sepsis diagnostics based on conventional culturing. We evaluated the sensitivity, specificity and time to result of Prove-it TM Sepsis in two major teaching hospitals in Helsinki and London: Helsinki University Central Hospital Laboratory (HUSLAB), and University College London Hospital (UCLH), respectively. Methods: 3318 blood samples collected from patients with suspected sepsis were analysed. Blood culture bottles of BacT/ALERT 3D (bio-Merieux) in HUSLAB and BACTEC 924 (Becton Dickinson) in UCLH were incubated for a total of 6 days or until flagged as positive. In both laboratories, DNA was extracted from 100 mul of blood culture using the automated DNA extraction instrument easyMAG (bioMeri-eux) prior to the Prove-it TM Sepsis assay. Conventional blood cultures were conducted in parallel and results were only revealed for the comparison at statistical analysis stage. Discordant results between the conventional methods and Prove-it TM Sepsis were further studied by DNA sequencing and case-by-case review of original microbiology laboratory data. Results: Of the analyzed 3298 blood cultures, 2107 yielded a pathogen by conventional techniques. Of these, 302 samples contained microbes not covered by Prove-it TM Sepsis, and an additional 137 cultures contained more than one organism. Sensitivity and specificity for Prove-it TM Sepsis were 94,7% and 98,7%, respectively. Of particular significance was the assay's faultless ability to differentiate MRSA from MSSA and from coagulase negative staphylococci. Furthermore, it provided a result on average one day earlier than the reference methods. Conclusions: Prove-it TM Sepsis was considered to be a fast, robust, and high performance diagnostic platform which is easily implemented into everyday laboratory workflow. Both study sites identified patient cases where timely information provided by Prove-it TM Sepsis would have significantly improved patient management. Examples here include more rational management and antibiotic choice subsequent to earlier differentiation of 'Gram-positive cocci in clumps' into MRSA, MSSA, or coagulase negative staphylococci, and earlier speciation of Gram-negative organisms. After the performance evaluation study, the pathogen panel of Prove-itTM Sepsis is further configured for detection of Candida spp. and new bacterial targets, i.e., Neisseria non-meningitidis, Kingella kingae, and Propionibacterium acnes. The assay identifies now 60 out of the 302 samples that were not covered during the performance evaluation, increasing the pathogen coverage from 86% to 89%. Prove-it TM Sepsis has also been evaluated with highly promising results for the diagnostics of bone and joint infections using bone biopsy and articular fluid as sample types. These evaluation studies have demonstrated excellent performance for this DNA-based diagnostic platform for various diagnostic applications. The earlier speciation provided by Prove-it TM Sepsis could contribute to faster, more evidence-based patient management and positive outcomes.