2184. Effect of Adjusted CLSI Breakpoints on Center-wide Fluoroquinolone Susceptibilities for Enterobacteriaceae and Pseudomonas aeruginosa: Is It Time to Restrict Empiric Fluoroquinolone Use?

Abstract

BackgroundIn January 2019, the Clinical and Laboratory Standards Institute (CLSI) lowered breakpoints of fluoroquinolones (FQ) for Enterobacteriaceae (EB)and Pseudomonas aeruginosa (PsAr). Automated commercial antimicrobial susceptibility testing (cAST) can only report levofloxacin MIC ≤1 μg/mL according to 2018 breakpoints. Updated panels will not be available until FDA approval. Laboratories and antimicrobial stewardship programs (ASP) must decide how to implement the adoption of new FQ breakpoints.MethodsAll microbiologic isolates of EB or PsAr were collected March 11, 2019–March 22, 2019. Manual E-tests and cAST by Microscan were performed for levofloxacin. Susceptible or non-susceptible (NS) isolates to levofloxacin were identified using 2018 and 2019 CLSI breakpoints [Table1]. Data were analyzed by Microsoft Excel and SPSS.Results159 isolates were analyzed. Isolates were predominantly from urine (78%) and females (68%). E.coli was the most frequently identified organism (40%) [Figure 1]. Application of new breakpoints changed 7% of EB (P = 0.083) and 16% of PsAr (P = 0.238) from susceptible to NS. [Figure 2] PsAr had higher rates of FQ NS compared with EB with new breakpoints (37% vs. 22%, P = 0.132, not shown). Non-urine and inpatient isolates had decreased susceptibility of ≥10% (P = 0.168 and 0.117). Lack of other resistance mechanism (ESBL, CRE, MDRO) was associated with a change in susceptibility (P = 0.036). Diagnosis of active cancer or immunosuppression (IS) was strongly associated with FQ NS by new breakpoints (P = 0.019, not shown). Patients with cancer or IS had the largest decrease in susceptibility (18%, P = 0.069) with application of new breakpoints. [Table 2].ConclusionUpdated CLSI FQ breakpoints resulted in trends of increased resistance. PsAr, non-urine and inpatient isolates were disproportionately affected. Implementation of manual MIC determination from select clinical isolates may be useful until approval of updated cAST. Diagnosis of cancer or IS was associated with FQ NS by new breakpoints. Restriction of FQ for empiric use may be considered, particularly in patients with cancer or IS. Larger studies are needed to determine the long-term effects of adjusted FQ breakpoints, but are anticipated to have significant impacts on ASP. Disclosures All authors: No reported disclosures.