Evaluation of a real-time PCR assay for the diagnosis of Pneumocystis pneumonia.

Abstract

The aim of this study was to evaluate of the quantification of Pneumocystis jiroveci using a real-time PCR assay. We tried to verify whether quantification was really effective in differentiating between carriage and Pneumocystis pneumonia (PCP) using real-time PCR with or without sample species normalization for classifying each sample species (sputum, bronchoalveolar lavage: bronchoalveolar lavage (BAL), and total samples). Twenty-two positive samples previously examined by conventional qualitative PCR were subjected to real-time PCR. Of these 22 lower respiratory tract specimens, 10 were BAL samples and 12 were (induced) sputum samples. According to our clinical diagnostic criteria, 17 were PCP and 5 were non-PCP. In the 12 sputum samples the concentrations of Pneumocystis-specific DNA detected in the non-PCP patients did not differ significantly from those in the PCP patients. The data were normalized using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the housekeeping gene to exclude differences due to the number of human cells in collected samples. After normalization, the Pneumocystis-specific DNA/GAPDH-DNA ratio in the non-PCP patients was higher than that in the PCP patients. In the BAL samples (10 samples), the mean concentration of Pneumocystis-specific DNA detected in the PCP patients was 9.6 times higher than that in the non-PCP patients (P=0.058), and after normalization, the Pneumocystis-specific DNA/GAPDH-DNA ratio in the PCP patients did not differ significantly (P=0.19) from that in the non-PCP patients. Although the present study indicated that normalization using GAPDH might be not helpful but BAL specimens are recommended over sputum specimens for the diagnosis of Pneumocystis Pneumonia by quantification with real-time PCR.