A major cause of morbidity and mortality in neutropenic patients is serious infection. As the neutrophil count decreases below [10.sup.6] per L, the risk of infection increases, with the greatest risk occurring in patients with neutrophil counts below [10.sup.5] per L. Profound neutropenia usually occurs in patients with hematologic disorders but is increasingly being seen in patients with other neoplastic disorders who are receiving myelosuppressive therapy. Currently, febrile neutropenic patients are treated promptly with empiric antibiotic therapy. Rolston and colleagues compared the effectiveness of four antibiotic regimens in the treatment of febrile patients with neutropenia.
The study included 567 adult cancer patients who (1) were febrile and neutropenic (with absolute neutrophil counts less than [10.sup.6] per L), (2) had received myelosuppressive therapy and were expected to become neutropenic or (3) had infection but could not mount an adequate febrile response. None of the patients had received previous antibiotic therapy for the current febrile episode. Pregnant patients, patients less than 16 years of age and patients with allergies to beta-lactam antibiotics were excluded from the study.
Patients were randomly assigned to receive imipenem, ceftazidime, ceftazidime plus amikacin, or imipenem plus amikacin. Patients who showed a response to therapy were treated for a minimum of seven days or for four days after resolution of all signs and symptoms. Patients with a proven Bacteroides infection who were not receiving imipenem were given additional treatment with clindamycin or metronidazole. Patients with a known gram-positive infection who showed no improvement after 48 hours of the assigned regimen were given vancomycin. Response to therapy was evaluated by an investigator who was unaware of which antibiotics patients were receiving. Response was defined as the resolution of all clinical and laboratory evidence of infection without a change in the antibiotic regimen.
Overall, the most effective regimen was imipenem and amikacin, with a response rate of 76 percent. However, this response rate was not significantly better than the rate with imipenem (72 percent) or ceftazidime and amikacin (71 percent). The least effective regimen was ceftazidime alone, which had a response rate of 59 percent.
Single-agent gram-positive infections occurred in 101 cases, with staphylococci or streptococci accounting for 91 percent of these infections. Single-agent gram-negative bacilli accounted for 658 infections. The majority of patients with uncomplicated gram-negative infections responded to therapy, while the majority of patients with complicated gram-negative infections did not respond to therapy. The most favorable prognostic factor for recovery from infection was improvement in the neutrophil count. Unfavorable prognostic factors included gram-positive infection, acute leukemia, pulmonary or enteric infection, and single-agent therapy with ceftazidime.
The authors conclude that imipenem is more effective than ceftazidime in the initial empiric treatment of febrile neutropenic patients. Single-agent therapy with imipenem was as effective as combination therapy with amikacin and ceftazidime or amikacin and imipenem. The reduced efficacy of ceftazidime in this study may be partially explained by the increased frequency of gram-positive infections. Initial therapy with vancomycin does not appear to be beneficial since the majority of patients who failed to respond to the initial regimen responded to vancomycin given later in the course of the infection. (Archives of Internal Medicine, February 1992, vol. 152, p. 283.)
COPYRIGHT 1992 American Academy of Family Physicians
COPYRIGHT 2004 Gale Group