[Investigation of Pneumocystis jirovecii in clinical specimens by different methods]

Pneumocystis jirovecii causes pneumonia in premature, newborn or malnourished children, as well as in immunocompromised subjects such as chemotherapy receiving, transplant and AIDS patients. Since the mortality and morbidity rates of Pneumocystis pneumonia (PCP) in these patients were high, rapid and accurate diagnosis is important. The aim of this study was to evaluate the diagnostic value of Giemsa staining (GS), direct fluorescent antibody (DFA) assay, (1→3)-β-D-Glucan (BDG) test and real-time polymerase chain reaction (PCR) for the detection of P.jirovecii in clinical specimens. A total of 100 PCP-suspected patients with underlying diseases who were followed-up in outpatient and inpatient clinics of our hospital between December 2008-July 2010 were included in the study. All the patients (66 male, 34 female; mean age: 42.04 years) were under long-term immunosuppressive drug therapy due to their hematological malignancies, kidney transplantation, neutropenia or chronic diseases. Respiratory samples [86 bronchoalveolar lavage (BAL), 8 endotracheal aspirate, 1 nasotracheal aspirate, 3 pleural, 2 lung biopsy samples] obtained from the patients have been studied with GS (Merck, Germany), DFA (Pneumo Cel, Cellabs, Australia) and PCR (primers targeting MSG gene, LightCycler, Roche, USA), while serum samples (n= 100) with BDG (Fungitell, ACC Inc, USA) and PCR methods. In BAL samples two were found positive by GS, DFA and PCR, and six were positive only by PCR, yielding a total positivity in 8 (8%) samples. All of the sera were negative with PCR, however 29 of them were positive (> 80 pg/ml), five were equivocal (61-79 pg/ml) and 66 were negative (< 60 pg/ml) with BDG test. Eight patients with positive results in BAL-PCR were also positive with BDG test. Although the agreement between GS and DFA was high (κ= 1), it was observed as low between PCR and DFA (κ= 0.38), DFA and BDG (κ= 0.07), BAL-PCR and BDG (κ= 0.28). DFA taken as the gold standard, the sensitivity and specificity values of GS, PCR and BDG methods were calculated as 100% and 100%; 100% and 93%; 100% and 67%, respectively. In the ROC analysis performed for BDG test, with DFA and BAL-PCR taken as the gold standards, the sensitivity, specificity and cut-off values of BDG were estimated as 100%, 93.9% and 494 pg/ml, and 100%, 72.8% and 62 pg/ml, respectively. Our data indicated that, overall specificity was high (100%) when using GS and DFA tests together, while the sensitivity has been elevated to 93% with the additional use of PCR and BDG tests, in the diagnosis of PCP-suspected patients. In conclusion, combination of all these tests should be performed for the laboratory diagnosis of P.jirovecii.