Possible Nosocomial Transmission of Pneumocystis jirovecii

To the Editor: Diversity of genotypes among Pneumocystis jirovecii (human-specifi c Pneumo-cystis species) isolates mainly involves internal transcribed spacer (ITS) loci (1). Type Eg, one of the most frequently detected ITS genotypes, has been found worldwide (2). The locus of dihydropteroate synthase (DHPS) is also of interest because DHPS is the target of sulfonamides, the main drugs used to treat Pneumocystis pneumonia (PCP). Studies of the DHPS locus have found mutations at positions 165 and 171, which confer potentially lower sensitivity to sulfonamides to mutant P. jirovecii organisms (3). Airborne transmission of Pneumocystis ssp. has been demonstrated among animals and probably occurs among humans (4). Reports of clusters of PCP cases in hospitals (4,5) provide a rationale for considering the possibility of nosocomial P. jirovecii infections. Moreover, we recently quantifi ed P. jirovecii in the air surrounding patients with PCP (6). Our fi ndings suggested that the fungus is exhaled from infected patients and then spreads into their surrounding air. Because matches of P. jirovecii genotypes between pulmonary and air samples would strengthen these fi ndings, we conducted DHPS and ITS typing of P. jirovecii isolates from PCP patients and from the air in their close environment. We assayed P. jirovecii DNA that we had previously detected in pulmonary samples (bronchoalveolar lavage and induced sputum) from 15 PCP patients and in 15 air samples collected 1 meter from each patient's head (6). ITS genotyping was based on sequence analysis of ITS 1 and 2 regions after amplifi cation with a nested PCR, cloning, and sequencing, as described (7). ITS alleles were identifi ed by using the typing system by Lee et al. (2). DHPS genotyping was based on a PCR restriction fragment-length polymorphism assay that enables detection of mutations at positions 165 and 171, as described (8). Among the 15 pulmonary samples, ITS genotyping was successful for all 15; among these, 8 ITS genotypes were identifi ed (Table). Type Eg was most frequently identifi ed. Mixed infections, which correspond to detection of >1 genotype in a given sample, were detected in 5 samples. DHPS genotyping was successful for all 15 pulmonary samples. A wild genotype was identifi ed in 9 samples, a 165 mutant genotype in 1 sample, and a 171 mutant genotype in 2 samples. Mixed infections were identifi ed in the 3 remaining samples. Among the 15 room air samples, ITS genotyping was successful for 7; among these, …