13. Evaluation of Etest for Antibiotic Susceptibility and Minimum Inhibitory Concentration (MIC) Agreement Against pseudomonas Aeruginosa (psa) from Patients with Cystic Fibrosis (CF)

Abstract

Background. CF acute pulmonary exacerbations are often caused by PSA, including multi-drug resistant strains. Optimal antibiotic therapy is required to return lung function and should be guided by in vitro susceptibility results. There are sparse data on the performance of Etest relative to reference broth microdilution (BMD) for many newer drugs against CF PSA. Herein, we describe Etest performance with 10 anti-PSA antibiotics against CF isolates. Methods. Contemporary, clinical PSA (n=105) isolated during pulmonary exacerbation from patients with CF were acquired from 3 US hospitals. MICs were assessed by BMD (reference) and Etest for aztreonam (ATM), cefepime (FEP), ceftazidime (CAZ), ceftazidime/avibactam (CZA), ceftolozane/tazobactam (C/T), ciprofloxacin (CIP), levofloxacin (LVX), meropenem (MEM), piperacillin/tazobactam (TZP), and tobramycin (TOB). Each respective MIC was completed in at least triplicate using the same inoculum between methods. Modal MICs for each method were compared by rates of essential agreement (EA), categorical agreement (CA), minor error (miE), major error (ME), and very major error (VME) rates. All miE, ME and VME were adjusted if within EA. Results. Of the 105 PSA, 46% had a mucoid phenotype. Results are summarized in the Table. Median modal Etest MICs read 0-1 dilution higher (IQR: 0-1) than BMD. CA and EA ranged from 64-93% and 63-86%, respectively. Single VMEs occurred for ATM (2.9%) and CAZ (4.2%). For CZA, 2 VMEs were observed and both were within EA. Major errors were ≤3% except for ATM (3.3%), MEM (3.3%), CZA (5.3% with adjusted ME 2.1%∗), and FEP (13%). Minor error rates were < 10% except for TZP, CIP, LEV, TOB, and FEP (13-29%), for which majority of miE were within EA (3/14, 11/16, 10/18, 13/19, 20/31, respectively). Performance was similar for non-mucoid and mucoid populations. Etest Performance Conclusion. Etest methods performed well for most antibiotics against this challenging collection of PSA from CF patients. Laboratories should be cautious of miE and ME that may occur with certain antibiotics. Furthermore, our observations suggest laboratories confirm CZA results for isolates with MICs near the breakpoint.