Cefepime Does Not Increase Mortality in Patients With Febrile Neutropenia

Alice Goodman

November 04, 2009

November 4, 2009 (Philadelphia, Pennsylvania) — Despite concerns raised after a meta-analysis suggested that increased mortality is associated with cefepime use in patients with febrile neutropenia and other infections, a new, more extensive meta-analysis should lay that fear to rest, researchers reported here at the Infectious Diseases Society of America (IDSA) 47th Annual Meeting.

"We found no significant increase in death among cefepime-treated patients with febrile neutropenia, [compared with those treated with] comparator drugs, either in trial data or in patient-level data," stated Peter Kim, MD, from the US Food and Drug Administration (FDA). "Our study found no plausible reason for mortality related to the drug's safety or lack of efficacy."

Cefepime is a fourth-generation cephalosporin approved by the FDA in 1996. It is widely used to treat pneumonia, urinary tract infection, skin and skin-structure infections, complicated gastrointestinal infections, and, empirically, febrile neutropenia, Dr. Kim explained.

In 2007, a meta-analysis of 38 trials found that 30-day mortality was higher in patients treated with cefepime than in those treated with other beta-lactam antibiotics (Lancet Infect Dis. 2007;7:338-348), so the FDA decided to undertake a more thorough meta-analysis to determine if febrile neutropenia patients treated with cefepime had higher mortality.

The FDA meta-analysis was based on 88 trials from inside and outside the United States, and included the 38 trials from the 2007 analysis. This FDA meta-analysis involved more than 9000 patients treated with cefepime, and more than 8000 patients treated with comparator drugs. In addition, the FDA meta-analysis included patient-level data from 35 trials, 7 of them on febrile neutropenia; the 2007 meta-analysis had no patient-level data.

Of the 88 trials examined in the FDA meta-analysis, 24 were in febrile neutropenia and 22 were in pneumonia. Ceftazidime was the comparator agent in 47 trials.

Thirty-day all-cause mortality was 6.2% for cefepime and 6% for the comparator agents. In the 24 febrile neutropenia trials, 30-day all-cause mortality was 6.5% and 5.6%, respectively.

Skin and skin-structure infections were the only clinical conditions associated with a significant increase in mortality related to cefepime use, but Dr. Kim cautioned that this was based on only 2 clinical trials with a small number of subjects.

Looking at patient-level data from 35 trials (5000+ treated with cefepime and 4000 with comparator drugs), 30-day all-cause mortality with cefepime was 5.63% and with comparator drugs was 5.68%. In the 7 febrile neutropenia trials with patient-level data, mortality rates were 7.9% and 6.5%, respectively.

Dr. Kim said that the data suggested that mortality is lower with cefepime in the United States than in other countries, and that the patients who died in febrile neutropenia trials appeared to die of underlying causes or cancer.

"This FDA meta-analysis [of cefepime] is very important. This drug is safe and there was no difference in 30-day all-cause mortality using a drug that also treats Gram-negative organisms," said Paul G. Auwaerter, MD, IDSA program chair and clinical director of the Division of Infectious Diseases at Johns Hopkins University School of Medicine in Baltimore, Maryland.

A cautionary note came from Richard Whitley, MD, from the University of Alabama at Birmingham, who said: "Cefepime is a good antibiotic, but we should restrict its use to those who really need it. This drug was used in high-risk patients already at risk for death, and that may be why a signal was raised in the earlier meta-analysis. A meta-analysis is not a randomized controlled trial. Nothing should be put in the drinking water."

Dr. Kim and Dr. Auwaerter have disclosed no relevant financial relationships. Dr. Whitley reports financial ties with Gilead, 3-V, Biosciences, and Chimerics.

Infectious Diseases Society of America (IDSA) 47th Annual Meeting: Abstracts 550 and 694. Presented October 30, 2009.