27 Non-invasive Diagnosis of Pneumocystis Pneumonia in an Oncology Patient Population

Background Pneumocystis jirovecii pneumonia (PCP) is an opportunistic infection commonly affecting immunocompromised individuals. Invasive procedures are usually required to obtain lower respiratory specimens such as bronchial wash and bronchoalveolar lavage to diagnose PCP. However, this invasive procedure is sometimes not possible in severely ill patients especially patients who are thrombocytopenic. Hence, there is a need for a less invasive sample to diagnose PCP. The majority of previous studies on PCP diagnosis have been conducted mostly among HIV-infected patient population. This study compared the use of induced sputum and bronchial wash in diagnosing PCP in oncology patients. Methods Induced sputum and bronchial wash specimens were collected from patients presenting to a cancer hospital in New York between October 2019 and November 2021. DNA extraction from induced sputum and bronchial wash was done using the BioMerieux eMag Nucleic Acid Extraction System. The Roche Light-cycler 2.0 Real-Time PCR Thermal Cycler system was used to detect P. jirovecii DNA. Results In total, 324 PCP from 289 unique patients were tested in sputum (n = 324; 289 unique patients) and bronchial wash (n = 69; 66 unique patients). Of these, 30 (30/324, 9.3%; 29/289, 10% unique patients) and 9 (9/69, 13%; 8/66, 12% unique patients) PCP positives were detected in sputum and bronchial wash respectively. Further analyses of patient-matched sputum and bronchial wash specimens showed that the induced sputum sensitivity was 72.2%, specificity was 92.9%, positive predictive value was 66.7%, negative predictive value was 94.5% and the accuracy was 89.6%, for the diagnosis of PCP, compared with bronchial wash. Conclusion Detection of P. jirovecii in lower respiratory tract samples especially using minimally invasive collected samples like induced sputum is crucial to the timely diagnosis of PCP and monitoring of patients with chronic infections. Our results suggest that induced sputum is a useful minimally invasive sample for diagnosing PCP infection in our patient population and this has potential application in routine diagnostic laboratory. Comparing PCP diagnosis in lower respiratory samples using a larger patient population (>3500 tests from >2000 unique patients tested in 6 years) to further our understanding of the utility of induced sputum in PCP diagnosis is the current focus of our laboratory.