The Prevalence of Legionella Species as a Co-pathogen in HIV-Associated Community-Acquired Pneumonia

BACKGROUND: Due to the lack of clinical suspicion, poor diagnostic performance, increased patient immunosuppression, and the increased chance of co-infection, HIV-associated Legionella community-acquired pneumonia (CAP) is currently under-reported. Thus, this study aimed to determine the frequency of Legionella in CAP-infected HIV patients. METHODS: Following initial diagnosis, DNA extracted from bronchoalveolar lavage (BAL) from CAP-infected HIV patients hospitalized at Hospital San Vicente Fundación in Medellin, Colombia were assayed for the presence of Legionella species (PAN Legionella, L. Anisa, L. bozemanii, L. micdadei, L. pneumophila and L. pneumophila serogroup 1) using singleplex real-time PCR (qPCR). Results were validated with agarose gel electrophoresis and reconfirmed using pre-amplification qPCR. RESULTS: Of the 59 HIV-infected individuals in the study, majority were non-smokers (64.4%), male (77.9%), and highly immunosuppressed (CD4 cell count <200 cells/μL). Initial CAP diagnoses were M. tuberculosis (37.3%), P. jiroveci (32.2%) and others (30.5%). Initial screening of pooled BAL samples indicated that majority of positive PAN Legionella were associated with M. tuberculosis and P. jiroveci. Of the 14 individual M. tuberculosis-infected patient BAL assayed, 10 were positive for PAN Legionella. Likewise, 6/9 P. jiroveci-infected BAL were also positive. Of all of the detected Legionellaceae infections, 31.3% were L. Anisa, 25.0% L. bozemanii, 18.8% L. pneumophila, and 12.5% L. micdadei, and 37.5% uncharacterized. Interestingly, none of the L. pneumophila infections were due to serogroup 1. Of note, all L. bozemanii and L. micdadei infections were associated with P. jiroveci, while all L. pneumophila infections were associated with M. tuberculosis. Legionella-infected patients had more complications and higher mortality rates compared with un-infected patients. CONCLUSION: Results indicate that Legionella are prevalent in the BAL of HIV co-infected patients. Clinicians should be aware of the possibility of the presence of Legionella—and not just L. pneumophila—in HIV-associated CAP. The role Legionella plays in clinical presentation, disease severity and inflammation remains to be determined. If further investigation supports these findings, this could change the way that CAP is managed in HIV-infected individuals. DISCLOSURES: All authors: No reported disclosures.