A comparative analysis between survivors and nonsurvivors patients treated with tigecycline.

he US Food and Drug Administration (FDA) have issued a arning describing increased mortality risk associated with he use of tigecycline when compared to other drugs in the reatment of serious infections. Such increased risk was deterined using a pooled analysis of randomized clinical trials hich involved predominantly patients treated for ventilatorssociated pneumonia (VAP).1 In the present observational and retrospective study, we etermined the mortality predictor factors between survivors nd nonsurvivors patients treated with tigecycline. Patients ere enrolled in the study at 23 Latin American centers etween February 2006 and June 2009. Data collection from linical records was done using an electronic form (LatinUser® ebsite; http://www.clinicalrec.com.ar). Two tailed hypothesis esting for difference in proportions and multivariable logisic regression to identify predictors of mortality were used for nalysis. Of the 301 eligible patients, 208 patients were male and he mean age was 49.7 (range 18–86). The overall in-hospital ortality was 36.5% (110/301 nonsurvivors vs 191/301 surivors). Univariate analysis showed that age, hospitalization n intensive care unit, APACHE II score, VAP, microbiological ocumentation, previous use of carbapenems and, immunoompromise were significantly associated with mortality. The ultivariate analysis showed that only age [odds ratio (OR): .026; [95% CI 1–1.05], p = 0.0202]; VAP (OR: 2.757; [95% CI .13–6.69], p = 0.0255) and APACHE II score (OR: 1.228; [95% CI .10–1.23], p = 0.0000) were identified as independent risk facors for mortality (Table 1). In concordance with the FDA warning, our study shows that n-hospital mortality with tigecycline was significantly higher n critically ill patients with VAP. Undoubtedly, from a strictly cientific point of view, there is no evidence to support tigecyline use in these patients. In that sense, Freire et al. reported ignificantly lower cure rates in clinically evaluable patients ith VAP treated with tigecycline (47.9%) when compared to mipenem (70.1%) (−22.2% [95% CI lower limit, −39.5%]). Based on these results, tigecycline has not been approved by the FDA for the treatment of VAP.2 Considering the attainable extracellular epithelial lining fluid concentrations of tigecycline (0.37 g/mL), the current dosage of 50 mg of tigecycline twice daily is probably underdosed for the treatment of VAP caused by typical, extracellular-acting bacteria such as Acinetobacter spp (MIC90 ∼ 1 g/mL).3 Regardless of these issues, the tigecycline pharmacological and microbiological profiles encourage physicians to use the drug for VAP due to multidrug-resistant (MDR)-pathogens. In that respect, we have published that more than 50% of the tigecycline prescriptions in Argentina were for VAP, mainly due to MDR-Acinetobacter spp.4 This practice is justified by the high regional resistance rates of MDR-pathogens with limited therapeutic options (e.g. carbapenem-resistant Acinetobacter spp.; KPC-producing Enterobacteriaceae). We know that clinical trials are indispensable tools to generate new knowledge and to test therapeutic options for the care mainly of critically ill patients, however, the epidemiological situation usually force physicians to incorporate the off-label use of some antibiotics as part of their daily practice (e.g. polymyxins, fosfomycin, sulbactam and other antibiotics, alone or in combinations). In the specific case of tigecycline it is mandatory to conduct clinical trials to generate comparative evidence about tigecycline efficacy and safety in VAP using higher doses (i.e. 200 mg initial and 100-q12), as well as to determine the role of tigecycline combinations with other antibiotics (carbapenems, polymyxins, and fosfomycin among others). Observational studies although helpful, are generally not adequate to address these issues, and controlled clinical trials will be necessary to provide sufficient and reliable evidence to support approval for new tigecycline indications and dosing schedules. Nevertheless, this process usually takes several