*Purpose: Respiratory infections are common after lung transplant and contribute to mortality and allograft failure. 16S ribosomal RNA (rRNA) gene sequencing could facilitate diagnosis of bacterial infections, increasing detection of bacteria not identified via routine cultures. We hypothesized that 16S rRNA sequencing would reveal bacteria that cultures did not.

*Methods: 99 bronchoalveolar lavage (BAL) samples were collected from 69 adult lung transplant recipients (LTRs) from January 2017 – June 2018 for a single center retrospective study using 16S rRNA gene sequencing. Sequences were classified taxonomically, and relative abundances were calculated; relative abundance of at least 10% was considered prevalent. Conventional bacterial culture results were obtained from medical records. Samples were categorized as “infected” if 2 or more criteria were met: fever, increased sputum, change in sputum appearance, increased shortness of breath, new radiographic infiltrate, or physician concern for infection. Routine, asymptomatic samples were considered “surveillance.” Cultures and sequencing results were compared to determine if bacteria that grew in culture were also identified by sequencing, or if additional prevalent bacteria were revealed by sequencing. Calculations were performed using chi-square and fisher exact tests.

*Results: Out of 99 BAL samples, 19 were considered infected. In infected samples, there was 100% concordance between cultured organisms and sequencing results. In contrast, only 67% of surveillance samples had the same bacteria identified in culture and sequencing. While 53% (10/19) of infected samples grew only normal respiratory flora (NRF) in culture, 90% (17/19) of them had at least one prevalent bacteria identified by sequencing that was not considered NRF. Sequencing also revealed prevalent organisms in 100% (2/2) of infected samples that had no growth in culture. For surveillance samples, sequencing identified at least one prominent bacteria in 85% (45/53) of samples that only grew NRF and 71% (10/14) of samples with negative cultures. These rates were similar for infected and surveillance samples (p=0.67 for NRF, p=1.0 for negative cultures). For infected samples, the most commonly identified bacteria by sequencing were anaerobes and Haemophilus. Surveillance samples often included anaerobes, Staphylococcus aureus, Haemophilus, and Streptococcus pneumoniae.

*Conclusions: In both infected and surveillance samples from LTR, 16S sequencing revealed abundant bacteria that were not detected in standard cultures. These data suggest that 16S sequencing of BAL may reveal additional bacterial information that may guide antibiotic treatment decisions.

« Back to 2020 American Transplant Congress