Evaluation of Aspergillus PCR Protocols for Testing Serum Specimens

A panel of human serum samples spiked with various amounts of Aspergillus fumigatus genomic DNA was distributed to 23 centers within the European Aspergillus PCR Initiative to determine analytical performance of PCR. Information regarding specific methodological components and PCR performance was requested. The information provided was made anonymous, and meta-regression analysis was performed to determine any procedural factors that significantly altered PCR performance. Ninety-seven percent of protocols were able to detect a threshold of 10 genomes/ml on at least one occasion, with 83% of protocols reproducibly detecting this concentration. Sensitivity and specificity were 86.1% and 93.6%, respectively. Positive associations between sensitivity and the use of larger sample volumes, an internal control PCR, and PCR targeting the internal transcribed spacer (ITS) region were shown. Negative associations between sensitivity and the use of larger elution volumes (≥100 μl) and PCR targeting the mitochondrial genes were demonstrated. Most Aspergillus PCR protocols used to test serum generate satisfactory analytical performance. Testing serum requires less standardization, and the specific recommendations shown in this article will only improve performance.