Study objective: To assess the value of Aspergillus polymerase chain reaction (PCR) test performed on the BAL in diagnosing invasive pulmonary aspergillosis (IPA).
Design: Between January 1996 and 1997, we prospectively followed up 249 cancer patients with pulmonary infiltrates suggestive of pneumonia. Bronchoscopy with fungal stains, cultures, and PCR was performed on all patients. PCR was used for the detection of Aspergillus mitochondrial and alkaline protease gene DNA. The PCR products were visualized either directly on polyacrylamide gel or after Southern transfer and probing with specific probes for mitochondrial and alkaline protease DNA.
Results: The 249 patients consisted of 10 patients with proven IPA (tissue invasion), 22 patients with probable IPA (microbiologic culture), 18 patients with possible IPA (consistent clinical and radiologic findings), and 199 control patients with no evidence of IPA. PCR positivity was strongly associated with all forms of IPA (p < 0.002). The sensitivity, specificity, positive predictive value, and negative predictive value of PCR were 80%, 93%, 38%, and 99%, respectively, for proven IPA, and 64%, 93%, 52%, and 96%, respectively, for probable IPA. Southern blotting analysis did not improve the diagnostic yield of the PCR test.
Conclusion: PCR performed on BAL is associated with high specificity and negative predictive value for IPA. The low positive predictive value could be related to the transient colonizing presence of aspergilli in the respiratory tract. The sensitivity correlates with the certainty of the diagnosis based on tissue invasion.