Impact of Fluoroquinolone Prophylaxis on Neutropenic Fever, Infections, and Antimicrobial Resistance in Newly Diagnosed AML Patients

Introduction: Fluoroquinolone prophylaxis is recommended during induction chemotherapy for patients with acute myeloid leukemia (AML) to reduce risk of neutropenic fever and systemic bacterial infections. We evaluated the effectiveness of primary fluoroquinolone prophylaxis in an area with high fluoroquinolone resistance.

Materials and methods: We performed a retrospective chart review of newly diagnosed adult AML patients who received frontline therapy at Mount Sinai Hospital in New York, NY, between 2012 and 2019. Primary outcome was development of neutropenic fever. Secondary outcomes were development of systemic bacterial infections and infections with multidrug-resistant organisms and Clostridioides difficile. Infectious outcomes were collected through 6 months after therapy initiation. We estimated the effect of fluoroquinolone prophylaxis with a time-dependent Cox proportional hazards model.

Results: Of 121 included patients, 87 received antibiotic prophylaxis and 34 did not. There was no difference in baseline characteristics, although the prophylaxis group had longer neutropenia duration (median 30 vs. 23 days, P = .013). The prophylaxis group had a reduced risk of neutropenic fever (hazard ratio 0.59, P = .039). The prophylaxis group had fewer gram-positive (P = .043) and gram-negative (P = .049) bloodstream infections and fewer clinically documented infections during frontline therapy (P = .005) and follow-up (P = .026). There was no difference in incidence of C. difficile or infection with fluoroquinolone-resistant or multidrug-resistant organisms. There was no mortality difference between groups.

Conclusion: In an area with high fluoroquinolone resistance, primary fluoroquinolone prophylaxis in newly diagnosed AML patients reduced the risk of neutropenic fever and systemic bacterial infections without increased antimicrobial resistance. Prospective, randomized studies are needed to confirm these observations.