ESBL-E are rising worldwide as a cause of bloodstream community acquired, health related or nosocomial acquired infections
]. In this context information about optimal therapeutic options for these infections in daily practice and their repercussion on outcome is needed. Several studies about ESBL-E bloodstream infections have analyzed the adequacy of empirical antibiotic therapy, risk factors for mortality and relationship of adequacy and outcome, with dissimilar results
]. Discordance among series is relevant and is probably related, at least in part, with the limited number of cases in most of the studies, differences in the selected population, and differences in definitions. The present study represents one of the largest series even communicated of patients with ESBL-E bacteremia. Our results corroborate the relatively high mortality of ESBL-E bloodstream infections and underscore the relevance of adequacy of antibiotic treatment on mortality in ESBL-E bacteremia. Our study also reveals a role for beta-lactam/beta-lactam inhibitor combinations in the treatment of ESBL-E infections.
A triangle of relationships among resistance, adequacy and mortality in bloodstream infections has previously emphasized as one of the potential causes for the higher mortality of infections caused by MDR agents, including ESBL-E
]. The general assumption of the causal connections of this triangle is the basis for the recommendation of wide spectrum antibiotic as empirical therapy in patients at risk of MDR infections. In ESBL-E infections carbapenems are widely considered as the drugs of choice on the basis on study cohorts, most of them retrospective, which suggests that third generations cephalosporins, irrespectively of in vitro
susceptibility, and quinolones, are less effective for the treatment of these infections than carbapenems
]. In vitro
tests that indicate an inoculum effect for beta-lactam/beta-lactam inhibitor combinations, cephalosporins, and to a lesser extent with the quinolones, also support this practice
]. As community acquired, health related and hospital acquired ESBL-E bloodstream infections are increasing, a raise in the empirical use of carbapenems in patients with suspected bloodstream infections and risk factors for ESBL-E acquisition is expected. In this hallmark with clinicians forced to use the carbapenems as the first choice for empirical treatment of severe infections, a future rise of the resistances to carbapenems has been predicted
The exact role of beta-lactam/beta-lactam inhibitor combinations in the treatment of ESBL-E infections is matter of discussion
]. Comparative clinical trials are lacking, and some recent reports suggest that beta-lactam/beta-lactam inhibitor combinations are potentially useful in the treatment of in vitro
]. A potential role of beta-lactam/beta-lactam inhibitor combinations in patients with ESBL-E bloodstream infections is suggested by our data. In our study mortality is consistently lower in patients treated empirically with carbapenems, or beta-lactam/beta-lactam inhibitor combinations when ESBL-E are susceptible although there are many confounding factors that impede us concluding consistently a role of beta-lactam/beta-lactam inhibitor combinations in empirical therapy for patients with ESBL-E bacteremia.
Another question that merit mention is the possibility of adjusting of antibiotic therapy using beta-lactam/beta-lactam inhibitor combinations or even quinolones after evaluating the susceptibility pattern of the ESBL-E in a “de-escalating” strategy. Our data suggest that beta-lactam/beta-lactam inhibitor combinations are potentially useful for treating ESBL-E susceptible to these antimicrobials, and could be an option for de-escalating.
Previous case series have detected a high variability in the origin of bacteremia, and while in the older reports the most frequent focus was respiratory tract
] in some of the more recent studies, as in the present one, the most frequent origin of ESBL-E bacteremia are urinary and biliary tract
]. This fact can be related to the higher incidence of hospital-acquired cases (i.e. ventilator associated pneumonia) at the beginning of the 2000’, and to the more recent “epidemic” increase of health-related and community acquired E-ESBL infections
]. In relation to empirical antibiotic therapy this is relevant because an effect of adequacy related with the origin of infection has been previously detected
], and suggested by our data. Its cause is unknown, although the lower mortality of ESBL-E bacteremia of urinary origin could influence this effect. Also we have detected a threefold increase in the number of cases of ESBL-E bacteremia between 2004-2005 and 2006-2008. A similar increase has been detected in previous studies.
Our study had several limitations. The study is retrospective and the comparisons between groups might be biased. The absence of a centralized laboratory to characterize ESBL-E strains and the absence of MICs are also limitation. Although the number of cases with ESBL-E is high in relation with previous studies; some subgroups have small size what may not have allowed us to find significant differences, especially regarding in hospital mortality.
In view of the increasing use of empirical carbapenems due to the risk of ESBL-E infections, information from clinical prospective studies about optimal empirical therapy for ESBL-E severe infections is needed. However the feasibility of these studies is uncertain as they imply the recruitment of a large number of patients with suspected ESBL-E infections to allow the collection of enough cases of microbiology proved infections for analysis. Meanwhile equilibrium among efforts to optimize initial therapy and efforts to promote judicious use of antimicrobial should be outweighed, and the role beta-lactam/beta-lactam inhibitor combinations in ESBL-E infections should be considered in specific hallmarks, essentially in de-escalating.