Empiric antibiotic and antifungal therapy for cancer patients with prolonged fever and granulocytopenia

Abstract

Since early diagnosis of even a disseminated fungal infection is difficult and treatment often ineffective in a patient with persistent granulocytopenia, we have prospectively evaluated continued antibiotic therapy and early empiric antifungal therapy in patients with prolonged fever and granulocytopenia. Between November 1975 and December 1979, all patients with fever (oral temperature >38 °C three times per 24 hours or >38.5 °C once) plus granulocytopenia (polymorphonuclear leukocytes < 500/mm 3), were evaluated and began an empiric antibiotic regimen consisting of Keflin ®, gentamicin and carbenicillin (KGC). Of the 652 episodes of fever and granulocytopenia (in 271 patients), initial evaluation failed to define an infectious etiology in 323 (49.5 percent). In 50 of the patients in whom initial evaluation did not demonstrate an infectious etiology, fever and granulocytopenia continued after seven days of therapy with KGC, without evidence for the etiology of their persistent fever. These patients were randomized to either discontinue receiving KGC (Group 1); continue receiving KGC (Group 2); continue receiving KGC with the addition of empiric amphotericin B (Group 3). The duration of granulocytopenia was comparable in the three groups (median 24 days, range 8 to 51 days). Clinically or microbiologically demonstrable infections occurred in nine of 16 patients who discontinued the KGC regimen (Group 1) (six also experienced shock, p < 0.01) compared with six of 16 patients who continued the KGC regimen (Group 2) (five in whom fungal infection developed), and in two of 18 patients who continued the KGC regimen plus amphotericin B (Group 3). The incidence of infections was less for patients receiving KGC plus amphotericin B than for patients who discontinued the KGC regimen (p = 0.013). Empiric amphotericin B therapy was also evaluated for its effectiveness in patients whose initial evaluation revealed an infectious etiology with fungal colonization throughout their alimentary tract but in whom fever and granulocytopenia remained despite at least seven days of therapy with appropriate antibiotics. In addition, the postmortem records of all patients dying between 1970 and 1979 were reviewed to ascertain the cause of death and the type of antimicrobial therapy received prior to death. Only one death due to fungal invasion occurred, when therapy with amphotericin B was instituted after one week of broad-spectrum antibiotic therapy. Collectively, these data suggest that continuing antibiotic therapy reduces early bacterial infections in patients with persistent fever and granulocytopenia and that empiric antifungal therapy also appears necessary to prevent fungal superinfections and to control clinically undetected fungal invasion.