Pulmonary infiltrations in febrile patients with neutropenia. Risk factors and outcome under empirical antimicrobial therapy in a randomized multicenter study.

Abstract

Different empirical approaches to antimicrobial treatment of lung infiltrates in patients with neutropenia were studied within a prospective, randomized multicenter trial. Patients with neutropenia with hematologic malignancies and fever of 38.5 degrees C or higher associated with newly diagnosed lung infiltrates were randomized for an initial therapy with acylaminopenicillin plus aminoglycoside (Group A), third-generation cephalosporin plus aminoglycoside (Group B), or the double beta-lactam combination (Group C), each in combination with rifampin. Nonresponders were given empirical amphotericin B plus 5-fluorocytosine beginning on day 4, day 5, or day 6 under study. Of 295 patients entered, 91.2% were evaluable. Complete response was obtained in 61.3% with no significant difference between treatment groups. The addition of rifampin did not improve treatment results. Only 27.1% of patients achieved a complete response by antibiotic therapy without additional antifungal therapy. Fungi dominated in cases of microbiologically documented infections and were associated with a poorer outcome compared with bacterial pneumonias. The trend of leukocyte counts under study had a highly significant effect on the outcome of infection. Lung infiltrates in febrile patients with neutropenia represent a high risk of treatment failure. Persistent neutropenia has a significantly adverse effect on the outcome of infection. Incorporation of systemic antifungal agents into first-line therapy, particularly in selected high-risk subgroups, might improve future treatment results. The quality of diagnostic techniques to establish the etiology of pulmonary infiltrates needs to be improved.