Sinus culture poorly predicts resident microbiota.

BACKGROUND Chronic rhinosinusitis (CRS) is an inflammatory disorder of the paranasal sinuses in which bacteria are implicated. Culture-based assays are commonly used in clinical and research practice; however, culture conditions may not accurately detect the full range of microorganisms present in a sample. The objective of this study was to determine the accuracy of clinical culture of CRS specimens compared with DNA-based molecular techniques. METHODS Ethmoid samples from 54 CRS patients collected during endoscopic sinus surgery were analyzed by both clinical culture and 16S ribosomal RNA (rRNA) gene sequencing. The association between 16S relative abundance and detection by culture was determined using logistic regression. RESULTS Each subject had an average of 3 isolates identified by bacterial culture and 21.5 ± 12.5 species identified by 16S sequencing. On average, 1.6 dominant taxa (>10% abundance) per subject were identified using molecular techniques, but only 47.7% of these taxa were identified by culture. Low abundance taxa (abundance <1%) were detected in only 4.5% of cultures. The odds that any organism would be detected by culture were 2.3 times higher with each 10% increase in relative abundance (p < 0.01). Conversely, only 29.5% of isolates identified by culture represented the dominant species, whereas 40% accounted for species with 1% to 10% abundance. Interestingly, 12% of isolates detected by culture were not identified by 16S pyrosequencing. CONCLUSION Standard clinical culture is a poor representation of resident microbiota. The incorporation of modern culture-independent techniques into clinical and research practices provides additional information that may be relevant for CRS.