Can we really use ß-lactam/ß-lactam inhibitor combinations for the treatment of infections caused by extended-spectrum ß-lactamase-producing bacteria?

In this issue a provocative study authored by Jesús Rodrı́guez Baño and colleagues asserts that the b-lactam/b-lactamase inhibitor (BLBLI) combinations piperacillin/tazobactam and amoxicillin/clavulanate are suitable for the treatment of certain patients with bloodstream infections due to Escherichia coli possessing extended-spectrum b-lactamases (ESBLs). The authors examined patients with infections originating mostly from the urinary and biliary tracts and included both communityand hospital-acquired infections. In the case of patients treated with piperacillin/tazobactam, we read that deaths attributed to infection did not occur when the minimum inhibitory concentration (MIC) of ESBL-producing E. coli against that drug was #2 mg/L. Weren’t infections due to ESBL-producing bacteria supposed to be treated with carbapenems? As clinicians, we readily appreciate that this study has significant implications for our daily practice. These findings challenge us to revisit a fundamental therapeutic approach against antibioticresistant bacteria and reward us with a needed alternative to carbapenems for the treatment of ESBL-producing E. coli bloodstream infections. Beyond that, this study may herald a renaissance of BLBLI therapy in the era of ‘‘bad bugs, no drugs.’’ For years, the paradigm of how to treat Gram-negative bacteria–producing ESBL was shaped by the link between clinical outcomes and antibiotic MICs. On the one hand, increased mortality in the setting of elevated cephalosporin MICs, due to the hydrolytic effects of ESBLs, disqualifies the use of extended-spectrum cephalosporins (eg, ceftazidime, cefotaxime, and ceftriaxone) [1]. On the other hand, favorable outcomes result from treatment with carbapenems, which retain low MICs and are stable in the presence of ESBLs [2]. Alternatives are acceptable only if their MICs are well below the resistance breakpoint, a rare occurrence for fluoroquinolones. Piperacillin/tazobactam and amoxicillin/ clavulanate are ‘‘stable’’ against ESBLs, because these enzymes are readily inactivated or inhibited by tazobactam and clavulanate. In the clinical microbiology laboratory, this results in much lower MICs of BLBLI combinations than b-lactams alone. This property is further manifest in the simple double disc diffusion test where, for instance, ceftazidime’s zone of inhibition is increased by clavulanate. Why, then, have we avoided using BLBLI combinations to treat infections caused by ESBL-producing organisms even when the MICs indicate susceptibility? First, there are concerns about the increase in piperacillin/tazobactam MICs when bacterial inocula reach 10 colonyforming units/mL. Second, the presence of other mechanisms of b-lactam resistance in a given strain, such as AmpC enzymes (eg, CMY-2, FOX-5, ACT-1), hyperproduction and specific mutations of non-ESBL enzymes (eg, SHV or TEM), or additional ESBLs, may provide a ‘‘complex background’’ that reduces the activity of BLBLIs. Third, pharmacokinetic/pharmacodynamic studies indicate that conventional doses of BLBLIs do not achieve ‘‘targets’’ associated with satisfactory outcomes. Finally, the published record on the use of BLBLIs for the treatment of infections caused by ESBL-producing organisms is limited and of heterogeneous quality [3]. Rodrı́guez Baño et al, coping with the onslaught of ESBL-producing E. coli in their Spanish communities, provide us with their clinical experience framed in Received 13 September 2011; accepted 23 September 2011; electronically published 4 November 2011. Correspondence: Robert A. Bonomo, MD, Infectious Diseases Section, Louis Stokes Cleveland Department of Veterans Affairs Medical Center, 10701 East Blvd, Cleveland, OH 44106 ([email protected]). Clinical Infectious Diseases 2012;54(2):175–7 Published by Oxford University Press on behalf of the Infectious Diseases Society of America 2011. DOI: 10.1093/cid/cir793