Comparison of different microbiological procedures for the diagnosis of Pneumocystis jirovecii pneumonia on bronchoalveolar-lavage fluid.

ABSTRACT

BACKGROUND: The current diagnostic gold standard for Pneumocystis jirovecii is represented by microscopic visualization of the fungus from clinical respiratory samples, as bronchoalveolar-lavage fluid, defining "proven" P. jirovecii pneumonia, whereas qPCR allows defining "probable" diagnosis, as it is unable to discriminate infection from colonization. However, molecular methods, such as end-point PCR and qPCR, are faster, easier to perform and interpret, thus allowing the laboratory to give back the clinician useful microbiological data in a shorter time. The present study aims at comparing microscopy with molecular assays and beta-D-glucan diagnostic performance on bronchoalveolar-lavage fluids from patients with suspected Pneumocystis jirovecii pneumonia. Bronchoalveolar-lavage fluid from eighteen high-risk and four negative control subjects underwent Grocott-Gomori's methenamine silver-staining, end-point PCR, RT-PCR, and beta-D-glucan assay. RESULTS: All the microscopically positive bronchoalveolar-lavage samples (50%) also resulted positive by end-point and real time PCR and all, but two, resulted positive also by beta-D-glucan quantification. End-point PCR and RT-PCR detected 10 (55%) and 11 (61%) out of the 18 samples, respectively, thus showing an enhanced sensitivity in comparison to microscopy. All RT-PCR with a Ct < 27 were confirmed microscopically, whereas samples with a Ct ≥ 27 were not. CONCLUSIONS: Our work highlights the need of reshaping and redefining the role of molecular diagnostics in a peculiar clinical setting, like P. jirovecii infection, which is a rare but also severe and rapidly progressive clinical condition affecting immunocompromised hosts that would largely benefit from a faster diagnosis. Strictly selected patients, according to the inclusion criteria, resulting negative by molecular methods could be ruled out for P. jirovecii pneumonia.