Conference Coverage

Be judicious with empiric antibiotics for febrile neutropenia



SAN FRANCISCO– Empiric antibiotic therapy for febrile neutropenia, a common and life-threatening complication of chemotherapy, hasn’t really changed much in 20 years, according to Alison Freifeld, MD, director of the section of oncology infectious diseases at the University of Nebraska, Omaha.

A closeup of various antibiotics spilling from a pill bottle luchschen/Thinkstock

Antibiotic resistance has become a major problem over that time. Multidrug-resistant, gram-negative blood stream infections are not uncommon, particularly with extended-spectrum, beta-lactamase–producing Escherichia coli and Klebsiella pneumoniae. Carbapenemase-producing Enterobacteriaceae are also on the rise, among others.

“Our standard empiric antibiotics” – ceftazidime, cefepime, piperacillin/tazobactam, and carbapenems – “are generally not active against these organisms, putting us in a major dilemma about what to do” with patients who have them, Dr. Freifeld said.

“Our goal at the moment is to unpack this ship, take some of these loads of antibiotics off, and figure out how we can more effectively bridge the gap between risk factors and outcomes, with fewer and more stringently applied targeted antibiotics,” she said at ID Week, an annual scientific meeting on infectious diseases.

Dr. Freifeld shared her advice at the meeting on what to do as that plays out. The main driver is to protect the remaining potency of current antibiotics without sacrificing patient care while also keeping new options in reserve for the sickest patients, so “we do not overuse these precious commodities.”

For one thing, it’s okay to shorten treatment – traditionally around 2 weeks, until the absolute neutrophil count (ANC) tops 500 cells/mcg – once the fever abates and cultures turn negative, even if the ANC remains low.

A recent trial put the approach to the test. A total of 78 patients had their antibiotics stopped after they had been free of fever for 72 hours, with normal vital signs and no other signs of infection; 79 in the control group had usual care, continuing treatment until their ANC recovered.

Early withdrawal shortened treatment by about 3 days and there were no statistically significant differences in mortality, with one death in the short-arm group and three in the long-arm group. Over half of the patients in the short-arm group were neutropenic when antibiotics were discontinued.

Serious adverse events, meanwhile, were far less common in the short-arm group (18 vs. 38). The take-home lesson is that “interventions to shorten duration of empiric antibiotics are safe and effective and important to implement now,” Dr. Freifeld said (Lancet Haematol. 2017 Dec;4(12):e573-83).

Also, “use escalation and deescalation approaches,” she said. The basic idea is to begin with monotherapy – cefepime or piperacillin/tazobactam – in uncomplicated cases, bumped up as necessary, and, in complicated cases, to start with broad, multidrug regimens, deescalated as culture reports and other information comes in (Haematologica. 2013 Dec;98(12):1826-35).

Finally, fluoroquinolone prophylaxis, “once considered the wonder of the world,” Dr. Freifeld said, needs to be limited to the highest-risk patients, particularly those with neutropenia expected to last a week or more. It does seem to lower the rates of fever and bloodstream infections, but recent investigations have shown no mortality benefit, and fluoroquinolone prophylaxis makes patients more likely to be colonized by multidrug-resistant bacteria. Many centers have opted against it, even in higher-risk patients (J Infect. 2018 Jan;76(1):20-37).

Dr. Freifeld serves on a data adjudication committee for Merck, and reported research support from the company.

Next Article: