590. Activities of Cefepime-Taniborbactam and Ceftazidime-Avibactam against Cefepime-Resistant Respiratory Gram-Negative Pathogens in a Hollow Fiber Infection Model

Abstract

Background. Cefepime-taniborbactam (FTB) combines cefepime (FEP), a fourth generation cephalosporin with taniborbactam, a novel inhibitor of metallo (MBL)- and serine β-lactamases (SBL). FTB 2.5g IV q8h was safe and effective in adults with complicated urinary tract infection in a Phase 3 trial (NCT03840148). FTB is also under development for hospital-acquired and ventilator-associated bacterial pneumonia (HABP/VABP). Methods. An in vitro hollowfiber infection model (HFM) was used to assess resistance emergence in MBL- and/or SBL-producing Klebsiella pneumoniae (KP, n=5) and Pseudomonas aeruginosa (PA, n =3) treated with humanized exposures of FTB or ceftazidime-avibactam (CZA). Dense ( ≥ 7 log10 CFU/mL) log phase cultures were inoculated into HFM cartridges and treated with human equivalent doses of FEP (2g q8h), FTB (2.5g q8h), or CZA (2.5g q8h) for 4 days. KP strains collectively harbored NDM (n =2), VIM (n =1), CTX-M (n =4), SHV-ESBL (n =1), CMY (n =1), KPC (n =1), and OXA-48 (n =1). PA strains produced either VIM (n =1), CTX-M (n =1), or KPC (n =1). Pharmacokinetic profiles of FTB and CZA in HFMs, based on free drug exposures in plasma of healthy volunteers, were confirmed by LC-MS/MS. For CZA HFMs with MBL-producing strains, EDTA was added to sequester zinc to restore CZA susceptibility (MIC ≤ 8 mg/L). Viable bacteria were quantitated by serial dilution plating; subpopulations with elevated MICs ( ≥ 4x) were monitored on FTB- or CZA-supplemented agar. Results. FEP, FTB, and CZA MICs ranged 16 to > 128 mg/L, 0.5 -8 mg/L, and 2 to > 128 mg/L, respectively. FEP was inactive (n =7) or bacteriostatic (n =1, FEP MIC =16 mg/L). FTB was bactericidal ( ≥ 3 log kill) against all 8 strains; subpopulations with elevated FTB MICs were not detected. Against MBL producers, CZA was inactive without EDTA and was bacteriostatic (n =2) or bactericidal (n =2) only when EDTA was added to disable MBLs. Against non-MBL producers, all of which were CZA-susceptible, CZA was either bactericidal (n =1) or bacteriostatic (n =2) or allowed growth due to emergence of resistance (n =1). Conclusion. Humanized exposures of FTB in a HFM were bactericidal against high inocula of MBL- and/or SBL-producing, multidrug-resistant respiratory pathogens and prevented emergence of resistance for 4 days. The results support development of FTB for HABP/VABP.