Clinical relevance of new diagnostic methods for bloodstream infections

The presence of bacteria and bacterial products in circulating blood has been known of for decades. Detection and identification of bacteria based on circulating nucleic acids has thus been a constant dream and remains an ongoing challenge. Two approaches are discussed here. The first is based on the detection of circulating bacterial nucleic acids by hybridisation on a microarray. This method offers broad identification capabilities, provided the probes are carefully chosen and validated, but still suffers from relatively low detection sensitivity. The second approach is represented by target-amplification methods yielding high detection sensitivities and broad-range species identification. First attempts were compromised by their exquisite sensitivity that detected nucleic acid traces present in the PCR reagents and prompted the development of numerous protocols to get rid of them. Preliminary studies using more recent home-brew or even commercial assays suggest that their exquisite detection sensitivity still leads to a very high number of 'false-positive' cases, when compared to cultures. Some of these 'false-positive' samples are likely to be true biological events of as yet unknown clinical significance. To date, and provided careful technical validation is performed, these costly assays offer very interesting potential in certain defined niches that are discussed in this article. Taken together, these novel diagnostic approaches will certainly warrant extensive clinically driven research programmes and will improve our comprehension of host-pathogen interactions. (C) 2007 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.