Legionella antibiotic susceptibility testing: is it time for international standardization and evidence-based guidance?

Legionella pneumophila, a Gram-negative bacillus, is the causative agent of Legionnaire's disease, a form of severe community-acquired pneumonia. Infection can have high morbidity, with a high proportion of patients requiring ICU admission, and up to 10% mortality, which is exacerbated by the lack of efficacy of typical empirical antibiotic therapy against Legionella spp. The fastidious nature of the entire Legionellaceae family historically required inclusion of activated charcoal in the solid medium to remove growth inhibitors, which inherently interferes with accurate antimicrobial susceptibility determination, an acknowledged methodological shortfall, now rectified by a new solid medium that gives results comparable to those of microbroth dilution. Here, as an international Legionella community (with authors representing various international reference laboratories, countries and clinical stakeholders for diagnosis and treatment of legionellosis), we set out recommendations for the standardization of antimicrobial susceptibility testing methods, guidelines and reference strains to facilitate an improved era of antibiotic resistance determination.

A multicenter study of viable PCR using propidium monoazide to detect Legionella in water samples.

Legionella quantification in environmental samples is overestimated by qPCR. Combination with a viable dye, such as Propidium monoazide (PMA), could make qPCR (named then vPCR) very reliable. In this multicentre study 717 artificial water samples, spiked with fixed concentrations of Legionella and interfering bacterial flora, were analysed by qPCR, vPCR and culture and data were compared by statistical analysis. A heat-treatment at 55 °C for 10 minutes was also performed to obtain viable and not-viable bacteria. When data of vPCR were compared with those of culture and qPCR, statistical analysis showed significant differences (P < 0.001). However, although the heat-treatment caused an abatement of CFU/mL ≤1 to 1 log10 unit, the comparison between untreated and heat-treated samples analysed by vPCR highlighted non-significant differences (P > 0.05). Overall this study provided a good experimental reproducibility of vPCR but also highlighted limits of PMA in the discriminating capability of dead and live bacteria, making vPCR not completely reliable.

Legionella pneumonia and serum procalcitonin.

Procalcitonin is a specific marker of severe bacterial infections with systemic inflammation. Quantitative evaluation of serum procalcitonin was performed in 140 patients with Legionella pneumonia. Positive values (>0.5 ng/mL) were found in 57.1% of the subjects. The positive sample rate was higher in early sera (73.9% in the first week) and decreased progressively. Mean values and the positive sample rate (69.8%) were higher in sera collected in the first 2 weeks of disease. Procalcitonin was positive in 86.7% of legionellosis cases confirmed by culture and mean positive values were significantly higher (P < 0.001) than those of culture-negative subjects. We conclude that procalcitonin can represent a useful prognostic marker for severe suspected Legionella pneumonia in the first 2 weeks of disease.