A diverse array of new experimental approaches are now available for the detection and identification of bloodstream infections, such as matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) 
, or pyrosequencing 
. The most studied approach is multiplex real-time PCR 
. Despite the fact that this has been reported as a promising technology, validations are still required 
. Among those systems already commercially available, Septifast® has been the most studied 
, however results have been inconsistent (for review see 
In our study, we used another available technique for multiplex PCR that detects a predefined panel of the most important sepsis pathogens by electrophoretic separation of target-specific amplicons (VYOO®, SIRS lab, Jena, Germany) associated with a specific enrichment of bacterial DNA. Only two meeting abstracts 
have been published, both in 2009, on bacterial and fungal detection with this particular approach. Here we report the first study using this approach in sepsis and non-infectious SIRS patient. In the cohort of patients with an infection, 70% of patients with positive blood cultures also gave a positive result using the VYOO® technique. Interestingly, we also found a good concordance between the PCR detection and the bacteriological results observed at the site of infection in nine cases without bacteremia. Furthermore, we found two PCR results to be compatible with pathogens usually seen at the site of infection responsible for the sepsis in patients without a documented bacterial infection. Overall these results seem to be comparable to those published using Septifast® 
. Furthermore, we did not find any data in favor of bacterial translocation in non-infectious SIRS, as may have been expected, particularly in the post cardiac surgery subgroup. We were also unable to detect Candida
as described with this technique 
, and this is in accordance with a previous study using Septifast® 
, even in cancer patients who are heavily colonized by this pathogen 
. The limited capacity to detect fungal DNA may be due either to some technical problems linked to the nature of the infectious agents, the low presence of fungi within the blood stream, or the low levels of fungal circulating DNA when fungi are present within the tissues. Of note, we had very few fungal infections in our cohort.
We also measured a number of clinical and biological parameters as well as various markers and cytokines. Interestingly, the severity or organ dysfunction scores were similar in both groups and, after multivariate analysis only CRP was able to discriminate infectious from non-infectious SIRS. We have previously shown that PCT is not associated with infection after correction for shock severity (fluid balance or levels of catecholamines’ administration). Similarly to what we have already observed with endotoxins, we studied here the circulating peptidoglycan and again did not find any difference between infectious and non-infectious SIRS.
In conclusion, the VYOO® approach for detecting microbial DNA is technically demanding, requiring around 8 hours of work. Despite encouraging results, its accuracy needs to be improved as we found a significant number of false positive results. Furthermore, some patients may be infected by several pathogens and it would be dangerous not to consider those that are not detected by the VYOO® test. A limitation of our study was a rather low number of patients. Because of our results, in accordance with our statistician, we decided not to perform the scheduled validation cohort because there was a limited chance to prove something else. A larger study would be required after improvement of the test. We are at the beginning of new era of molecular diagnostics in septic patients; the results are encouraging but their current accuracy needs to be improved before they can be considered a reliable tool for clinicians and patient care.