In contrast to previous studies [32,33,35], in our work, thenthereby imipenem was not associated with VAP due to PRPA. But the main resistance mechanism to carbapenem is loss of the porin oprD, which leads to a selective resistance to these antibiotics.Our study confirms results of the case-control study conducted by Harris et al. in that previous exposure to piperacillin: piperacillin-tazobactam was statistically associated with the isolation of piperacillin-tazobactam-resistant P. aeruginosa (OR 8.63; 95% CI 6.11 to 12.20; P < 0.0001) [33]. In an observational study comparing the relative risks of emergence of resistant P. aeruginosa associated with four individual antipseudomonal agents, Carmeli et al. demonstrated that there was an significant association between piperacillin treatment and the emergence of piperacillin resistance (HR = 5.

2; P = 0.01) [32].In our study, fluoroquinolones treatment in the week prior to VAP was an independent factor associated with hospital death. Few data could explain this result. The expression of mexAB-oprM, responsible for acquired resistance to fluoroquinolones is regulated by the quorum-sensing, and, therefore, is known to be growth-phase dependent [36,37]. We suggest that the overexpression of the porin mexAB-oprM could be synchronized with a hyper-production of bacterial virulence factors, leading to a very virulent bacterial inoculum.Results on the role of fluoroquinolones in the emergence of piperacillin-resistant P. aeruginosa were of borderline statistical significance. Trouillet et al.

suggested that receiving any fluoroquinolone may be a risk factor for acquiring piperacillin-resistant P. aeruginosa [10]. In a cohort study, Carmeli et al. also found previous ciprofloxacin treatment was a risk factor of the emergence of antibiotic-resistant P. aeruginosa (hazard ratio 9.2; P = 0.04) [32]. But in contrast to these previous studies, we did not found any role of carpapenems in the emergence of PRPA.Duration of exposure to these antibiotics should also be taken into consideration. In a case-control study conducted by Paramythiotou et al., among 34 patients with multi-drug resistant P. aeruginosa, a previous treatment with ciprofloxacin or imipenem was a significant risk factor for the acquisition of multi-drug resistance only when the duration of the treatment was longer than the median duration of treatment with these antimicrobials observed in that study [38].

Our study has several limitations. First, we used data prospectively collected in a large database that was not specifically designed for the present subject. However, all data were collected prospectively, with special attention to nosocomial infections and treatment adequacy. VAP was consistently documented by quantitative cultures of distal pulmonary specimens, and all patients were observed Batimastat until ICU and hospital discharge.