Aspergillus fumigatus detection by PCR in broncho-alveolar fluid

Background: The usefullness of a nested PCR for detection of Aspergillus fumigatus DNA
was evaluated in bronchoalveolar lavage (BAL) fluid during a period of two years (1996-
1998). The aim of the study was to assess the role of PCR in diagnosing invasive pulmonary
aspergillosis (IPA).
Methods: A nested PCR-based amplification of fragments of genes-encoding alkaline
proteases from A. fumigatus was used to test 167 BAL samples. All samples were checked for
the absence of amplification inhibitors. Medical, radiological, microbiological records and
autopsy findings were reviewed for assessing invasive aspergillosis. All successive patients
investigated by BAL were included in the study. They were distributed in three groups: A,
proven or probable aspergillosis (n=11); B, colonization (n=2); C, no evidence of IPA
(n=154). PCR results were compared to culture détection as gold standard and to clinical data.
Results: BAL fluids from 10 patients of group A were PCR positive. One case was falsely
negative. Among group B, one case was PCR positive, and the second one PCR negative but
had negative BAL cultures (only culture positive sputum). No false positive was detected
among group C. Comparing to culture, sensitivity was 91%, specificity, 100%, positive
predictive value, 100% and négative predictive value, 99%.
Conclusion: 1. Aspergillus fumigatus PCR in BAL fluid was an accurate test to diagnose
culture negative patients with IPA and to confirm culture positive samples; however it doesn't
make difference between infection and colonization. 2. It is an appropriate test to exclude
Aspergillus infection in patients at risk of invasive illness.