Objectives:
This work aims to describe the value of a nasopharyngeal aspirate (NPA), as a non-invasive sample, for diagnosing lung primary infection by the fungus Pneumocystis in immunocompetent infants. Most infants develop a lung primary infection before 6 months of age that goes undetected. There is no culture method for microbiologic diagnosis of this infection. Diagnosis relies on Pneumocystis-DNA-amplification of a NPA sample. However, the diagnostic yield of a NPA varies with age reaching a peak of 45% at 2 - 5 months of age in healthy live infants while examination of autopsy lung tissue samples from infants dying in the community give lung-yields over 90% at comparable ages. This yield discrepancy warrants clarification. Studies in immunosuppressed rats document numerical correlation of Pneumocystis organisms obtained from NPA and lung samples but, whether NPA predicts lung infection in immunocompetent infants has not been studied because lung biopsies cannot be obtained from healthy infants. Therefore, we studied paired NPA and lung samples from dead autopsied infants. 
Methods:
The protocol was approved by the University of Chile School of Medicine Ethics Committee. Paired nasopharyngeal aspirate samples and lung biopsies were prospectively obtained from presumably immunocompetent infants dying in the community and undergoing a medico-legal autopsy at the Medico-Legal Service (Coroner´s office) of Santiago between 2012 and 2017. Pneumocystis diagnosis was sought in NPA pooled samples from both nares and in lung biopsy samples of 3% of the infant lung weight (average weight of the sample 0.84, median 0.75 and range 0.32 - 1.5 grams) simultaneously obtained from 27 infants of a mean age 3.55 months (median 3.05, range 1.0 to 8.9 months) dying in the community using P. jirovecii-DNA amplification of the mtLSUrRNA by nested-PCR. mtLSUrRNA was reamplified using Platinum Taq DNA polymerase enzyme and the amplification product purified for sequencing using Wizard kit SV gel and PCR clean-up system (Promega).
Results:
Pneumocystis jirovecii DNA amplification results from upper and lower samples were concordant in 22 (81.5%) infants and were Pneumocystis-positive in 16 (59.3%) and negative in 6 (22.2%). Five infants (18.5%) had discordant NPA/Lung results (4/0 and 0/1 positive NPA). Sensitivity, Specificity, NPV, and PPV for NPA were 80.0%, 85.7%, 60.0% and 94.1%, respectively. The mitochondrial gene was successfully sequenced in paired samples from 3 of the 16 Pneumocystis-positive concordant infants and confirmed same 80, 81, 85, 88, and 248 allelic positions in Pneumocystis paired isolates.
Conclusion:
Results validate double-NPA sample for diagnosis of Pneumocystis primary lung infection with 80% sensitivity and a positive predictive value of 94%. They provide a first reference to compare reports of Pneumocystis incidence in NPA or lung samples from infants. Identical allelic positions in a small proportion of paired samples whose isolates were successfully sequenced suggest Pneumocystis isolates in lung and NPA samples were not different. Use of NPA sampling at other ages, especially in adults that have a very mild lung burden of Pneumocystis requires further validation. Less-invasive sampling is desirable.

1 / Predictive value of a nasopharyngeal aspirate sample for diagnosis of pulmonary Pneumocystis in infants: A prospective autopsy study.