Patterns of infection in patients with aplastic anemia and the emergence of Aspergillus as a major cause of death.

Patterns of infection were studied in 150 patients with aplastic anemia who were admitted to the Clinical Hematology Branch, National Institutes of Health, between January 1978 and December 1989 for immunosuppressive therapy. Sixty percent of the patients were males, 71% were white, their mean age was 33.6 years (median, 27.5; range, 1-75), and 83% had severe aplastic anemia. One hundred three patients developed 1 or more febrile episodes during the study period. The risk factors for developing a febrile episode included a low Absolute Neutrophil Count (ANC) and Absolute Monocyte Count (AMC) at admission and the presence of an indwelling central venous catheter (Hickman-Broviack or Port-A-Cath). A total of 289 febrile events were studied, including unexplained fever (FUO) in 89 (31%), microbiologically documented infection (MBDI) in 137 (47%), and clinically documented infection (CDI) in 63 patients (22%). Compared to documented infections (MBDI) or CDI), FUO events were associated with a higher frequency of rigors, signs and symptoms of serum sickness, and treatment regimens known to cause fevers. None of the FUO events had a fatal outcome, even if the antibiotic therapy was discontinued before day 7. Among CDI events, bacteria were the most commonly defined etiologic agent (67%), followed by fungi (23%), viruses (7%), and parasites (3%). The patterns of bacterial infections in patients with aplastic anemia were similar to those observed in patients with cancer-related neutropenia. Twenty-one patients (15%) developed invasive fungal infections (aspergillus, 11; candida, 7; and both, 3), which were fatal in 19 (90%). Fungal infections accounted for 30% of the secondary infectious events and for 55% of fatal infectious events. The only identifiable risk factors for developing a fungal infection were the degree of neutropenia and monocytopenia at initial admission or final evaluation. Invasive pulmonary aspergillosis developed despite empirical amphotericin B therapy and was associated with a high incidence of fatal pulmonary hemorrhage (10 of 13 patients [77%]). Infection was responsible for 36 (62%) of the deaths observed during the study period and hemorrhage alone for 4 (7%). However, 20 of the patients who died of infection had concomitant hemorrhage. No significant drop in ANC, AMC, or platelet count could be demonstrated during a fatal infectious event as compared to a nonfatal infectious event. Invasive fungal infections, predominantly with aspergillus and candida, emerged in our study as the major causes of mortality in patients with aplastic anemia. Without bone marrow recovery the prognosis associated with invasive mycoses was grave.