Diagnostic yield of bronchoscopy with bronchoalveolar lavage in febrile patients with hematologic malignancies and pulmonary infiltrates.

2008 
Infectious complications are a major cause of morbidity and mortality in immunosuppressed patients. Febrile patients with hematologic malignancies and pulmonary infiltrates have high mortality rates, especially if mechanical ventilation is required. The diagnostic value of fiberoptic bronchoscopy (FOB) with bronchoalveolar lavage (BAL) in these patients is controversial. We retrospectively analyzed the microbiological results of BAL samples obtained during 249 FOB examinations from 199 febrile patients with hematologic malignancies and pulmonary infiltrates (underlying diseases: acute leukemia 103 patients, lymphoma 84 patients, other malignancies 12 patients). Two hundred forty-six examinations could be evaluated. Seventy-three out of 246 BAL samples were sterile; 55 samples showed microbiological findings classified as contamination or colonization. One hundred eighteen samples showed positive microbiological results of bacteria and/or fungi classified as causative pathogens. Thereof, in 70 samples, only bacterial pathogens were detectable (Gram-positive, 35; Gram-negative, 30; mixed Gram-positive and Gram-negative, 5). Thirteen samples showed both fungi and bacterial pathogens. In 33 samples, only fungi were detectable, thereof, in 15 samples Aspergillus species, in 16 samples Candida species, and in 2 both. In two samples, a viral pathogen could be detected. Three nonlethal complications (bleeding, arrhythmia) occurred that required early termination of FOB. In 94 (38.2%) patient episodes, antibiotic treatment was modified as a result of microbiological findings in BAL samples. Our results show that FOB with BAL is a valuable diagnostic tool with low complication rates in high-risk febrile patients with hematologic malignancies and pulmonary infiltrates, contributing crucial results for the individual case, and also improving epidemiologic knowledge.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    35
    References
    65
    Citations
    NaN
    KQI
    []