9. Evaluation of Synergy Testing Methods for Clinical Labs to Determine Susceptibility of Extensively Drug-resistant Gram-negatives to Ceftazidime/ Avibactam and Aztreonam Combination Therapy

Background Carbapenem-resistant Enterobacterales (CRE) and Pseudomonas aeruginosa (CR-PA) producing Metallo-β-lactamases (MBLs) cause severe nosocomial infections with no defined treatment. Combination therapy with ceftazidime/ avibactam (CZA) and aztreonam (ATM) is a potential option, but there is no approved, feasible, synergy testing method for clinical labs to guide clinical decision making. Here, we evaluate the performance of 4 synergy testing methods using gradient-strips or disks. Methods We used 10 representative Enterobacterales strains, namely, E. coli, K. pneumoniae, and E. cloacea, and 6 PA strains harboring MBL, GES or non-MBL enzymes (Fig 1). 4 strains were successfully treated with CZA-ATM in case reports, the rest were from the CDC AR Bank. Four synergy testing methods were evaluated, i) Disk stack (DS), ii) Disk elution (DE), iii) Gradient-strip Stack (SS), iv) Gradient-strip Cross (SX) (Fig 1). All methods were run side-by-side as per CLSI guidelines with broth microdilution (BMD) as the reference. Data is the mean of 3 replicates. Synergy is defined as a strain that is resistant (R) to ATM but drops to ≤ the susceptible (S) breakpoint (Table 1) in the presence of CZA (Fig 2). Categorical agreement (CA), very major error (VME), major error (ME), minor error (MI) were calculated across methods for CZA-ATM synergy relative to BMD. Summary of synergy testing methods evaluated CLSI Breakpoints used for this study Results All CRE with NDM and PA with GES were ATM-R, CZA-R and S to the CZA-ATM combination. PA with NDM or VIM remained R to CZA-ATM likely due to other mechanisms of resistance. CA was high for DE (100%), SS (81%, MI 19%), and SX (88%, MI 13%) but low for DS (25%, ME 54%, MI 31%). Representative strains are shown (Fig 2, Table 2). Removing PA, CA for DE, SS, and SX was 100% and 20% for DS. Representative results of strains with each synergy testing method. Representative data of strains displaying synergy (green) or no synergy (red) Conclusion Overall, DE was the most reliable method for CZA-ATM synergy testing, and could be a valuable tool in low-resource labs. SS and SX were reliable but prone to technical error. DS had the worst performance. Disks and gradient-strips had identical performance across brands. We propose an algorithm for ATM-R, CZA-R, and MBL-positive CRE, where CZA-ATM synergy testing may be beneficial to guide therapy. These methods are reliable qualitative indicators of the presence or absence of synergy. Synergy testing is not recommended for CR-PA due to complex resistance profiles. Disclosures Cesar A. Arias, M.D., MSc, Ph.D., FIDSA, Entasis Therapeutics (Scientific Research Study Investigator)MeMed (Scientific Research Study Investigator)Merck (Grant/Research Support) William R. Miller, MD, Entasis Therapeutics (Scientific Research Study Investigator)Merck (Grant/Research Support)Shionogi (Advisor or Review Panel member)