Élie Azoulay1, Anne Bergeron2, Sylvie Chevret3, Nicolas Bele3, Benoît Schlemmer3, Jean Menotti3. 1. Intensive Care Unit, Paris 7 Diderot University, Saint-Louis Teaching Hospital, Paris, France. Electronic address: elie.azoulay@sls.ap-hop-paris.fr. 2. Pulmonology Department, Paris 7 Diderot University, Saint-Louis Teaching Hospital, Paris, France. 3. Intensive Care Unit, Paris 7 Diderot University, Saint-Louis Teaching Hospital, Paris, France.
Abstract
RATIONALE: Pneumocystis jiroveci polymerase chain reaction (PCR) has higher sensitivity than conventional stains but cannot distinguish colonization from infection. METHODS: We compared P jiroveci PCR and conventional stains in HIV-uninfected immunocompromised patients. RESULTS: Among the 448 patients, 296 (66%) patients had hematologic malignancies; 72 (16.1%), bone marrow transplants; 44 (9.8%), solid tumors; 21 (4.7%), renal transplants; and 15 (3.4%) were taking immunosuppressants for systemic diseases. Diagnostic strategy consisted of BAL in 351 patients and induced sputum (IS) in 97 patients. Conventional pneumocystic pneumonia (PCP) stain was positive in 39 (8.7%) patients, including 34 with positive PCR. In addition, PCR was positive in 32 patients, including 21 with complete follow-up, of whom 14 were diagnosed with probable or definitive PCP (a 36% increase). PCR was 87.2% sensitive and 92.2% specific; positive and negative predictive values were 51.5% and 98.7%, respectively. Sensitivity and negative predictive value were 100% on IS. CONCLUSIONS: In HIV-uninfected immunocompromised patients with acute pulmonary infiltrates, P jiroveci PCR correlates with clinical evidence of PCP. A negative PCR allows withdrawing anti-PCP therapy.
RATIONALE: Pneumocystis jiroveci polymerase chain reaction (PCR) has higher sensitivity than conventional stains but cannot distinguish colonization from infection. METHODS: We compared P jiroveci PCR and conventional stains in HIV-uninfected immunocompromised patients. RESULTS: Among the 448 patients, 296 (66%) patients had hematologic malignancies; 72 (16.1%), bone marrow transplants; 44 (9.8%), solid tumors; 21 (4.7%), renal transplants; and 15 (3.4%) were taking immunosuppressants for systemic diseases. Diagnostic strategy consisted of BAL in 351 patients and induced sputum (IS) in 97 patients. Conventional pneumocystic pneumonia (PCP) stain was positive in 39 (8.7%) patients, including 34 with positive PCR. In addition, PCR was positive in 32 patients, including 21 with complete follow-up, of whom 14 were diagnosed with probable or definitive PCP (a 36% increase). PCR was 87.2% sensitive and 92.2% specific; positive and negative predictive values were 51.5% and 98.7%, respectively. Sensitivity and negative predictive value were 100% on IS. CONCLUSIONS: In HIV-uninfected immunocompromised patients with acute pulmonary infiltrates, P jiroveci PCR correlates with clinical evidence of PCP. A negative PCR allows withdrawing anti-PCP therapy.
Authors: Stefanie Desmet; Eric Van Wijngaerden; Johan Maertens; Jan Verhaegen; Eric Verbeken; Paul De Munter; Wouter Meersseman; Britt Van Meensel; Johan Van Eldere; Katrien Lagrou Journal: J Clin Microbiol Date: 2009-10-21 Impact factor: 5.948
Authors: Kelvin K W To; Sally C Y Wong; Ting Xu; Rosana W S Poon; Ka-Yi Mok; Jasper F W Chan; Vincent C C Cheng; Kwok-Hung Chan; Ivan F N Hung; Kwok-Yung Yuen Journal: J Clin Microbiol Date: 2013-02-13 Impact factor: 5.948