Risk factors for mortality in patients with Stenotrophomonas maltophilia bacteremia and clinical impact of quinolone-resistant strains

Background: Stenotrophomonas maltophilia is an important nosocomial pathogen. This pathogen has intrinsic or acquired resistance to a number of antibiotics classes. Furthermore, Stenotrophomonas infections have been associated with high mortality, mainly in immunocompromised patients. Accordingly, we conducted a retrospective cohort study on the clinical data, microbiological characteristics, and outcomes of patients with S. maltophilia (SM) bacteremia. Methods: A retrospective cohort study was conducted at two tertiary care referral hospitals in Seoul, South Korea. Data were collected between January 2006 and December 2015 from electric medical records. Our analysis aimed to identify the risk factors associated with crude mortality, as well as the predictive factors of quinolone-resistant strains in SM bacteremia patients. Results: A total of 126 bacteremia patients were enrolled in the study. The mortality rate was 65.1%. On multivariable analysis, hypoalbuminemia (odds ratio [OR], 5.090; 95% confidence interval [CI], 1.321-19.621; P = 0.018), hematologic malignancy (OR, 35.567; 95% CI, 2.517-502.515; P = 0.008) and quinolone-resistant strains (OR, 7.785; 95% CI, 1.278-47.432; P = 0.026) were independent risk factors for mortality. Alternatively, usage of an empirical regimen with quinolone (OR, 0.172; 95% CI, 0.034-0.875; P = 0.034) was an independent protective factor for mortality. The multivariable analysis of predictive factors revealed that high Charlson comorbidity index (OR, 1.190; 95% CI, 1.040-1.361; P = 0.011) and indwelling of a central venous catheter (CVC) (OR, 3.303; 95% CI, 1.194-9.139; P = 0.021) were independent predisposing factors associated with quinolone-resistant strains in SM bacteremia patients. Conclusions: Our findings suggest that a high Charlson comorbidity score and indwelling of a CVC were significantly independent predictors of quinolone-resistant strains in SM bacteremia patients. Therefore, we need to carefully consider the antibiotic use in SM bacteremia patients with these predictive factors.