2016 Section 5 Green Book

Hauser et al.

FIGURE 3. Similarity of bacterial communities between time points and sampling sites. The Morisita-Horn similarity index was used to quantify differences in microbiota in order to compare the 6-week ethmoid sample to the anterior nares, ethmoid, and nasopharynx at the time of surgery, the ethmoid at 2 weeks postoperatively, and between the ethmoid at 2 weeks postoperatively and the ethmoid at the time of surgery. The 6-week ethmoid sample was most similar to the anterior nares and ethmoid samples at the time of surgery, suggesting these may be potential sources for repopulation of bacteria postoperatively. *A value of 1 indicates identical bacterial community, and 0 is complete dissimilarity. AN = anterior nasal cavity, AN = anterior nares; NP = nasopharynx.

exhibit altered immune responses and are more suscepti- ble to other bacterial infections. 36,38 A study investigating the effects of antibiotics and steroids on the sinonasal mi- crobiome in acute exacerbations of CRS changes in the sinonasal microbiome similarly found a decrease in diver- sity at the completion of courses of antibiotics and steroids. Though this suggests that antibiotics may contribute to a decrease in diversity, resolution of the acute exacerbation may also have contributed to the changes seen, as there was no baseline steady-state sample collected. 33 There is evidence in humans to support that a more diverse micro- biome is associated with improved health outcomes and less disease burden. 35,39,40 We noted a trend toward a less complex bacterial community after surgery and antibiotics, which became more diverse again at the 6-week time point ( p = 0.09). More significant temporal differences in mi- crobial diversity may have been more evident in a larger cohort. These results suggest that treatment interventions (surgery and antibiotics) can increase bacterial diversity; whether these changes will contribute to improved clin- ical outcomes is currently under investigation. Our data illustrate an initial perturbation in the microbiome due to therapeutic intervention, from which many subjects may recover their preoperative bacterial community and some may continue to shift toward a new and different commu- nity makeup, a model that has been previously proposed for microbiome effects on disease states. 37 In humans, re- silience is variable by subsite and subject, but the nasal cavity appears to be 1 of the most stable, 41 as supported by our findings.

Possible sources for the bacteria that recolonize the si- nuses after sinus surgery include reservoirs within paranasal biofilms, colonized intramucosal sites refractory to antibi- otic therapy, within the nasopharynx, or even de novo bac- teria from extranasal sources. 8–11 We found that 6-week postoperative microbiota most closely represented the an- terior nares and ethmoid samples from the time of surgery, and were least similar to the nasopharynx, suggesting that the former sites may be critical reservoirs for sinus micro- biota. New research suggests that in multiple mucosal surfaces individuals do in fact have signature bacterial species and even strains. 42 We have previously showed that the ante- rior nares microbiome is stable over time in the absence of mitigating factors such as antibiotics 43 and also that a more diverse sinus microbial ecology is associated with healthy sinuses. 44 In this study, our preoperative and postoperative ethmoid samples were largely similar (mean M-H 0.59), indicating that this “fingerprint” may persist in the sinuses, but can be subject to milder changes. Although it is possi- ble that we may have seen more resiliency in our samples, or more recovery of baseline microbiota, if we had sam- pled the ethmoids at a later time point, it was thought that 6 weeks was enough time to reach a new postoperative steady state because microbiota in the gut have been noted to return to baseline within 2 to 4 weeks after cessation of antibiotics. 7,9,11,29,36,45 It is also possible that microbial disruptions caused by surgical and medical interventions leave a patient susceptible to recolonization by opportunis- tic pathogens. Further study will hopefully elucidate the

International Forum of Allergy & Rhinology, Vol. 6, No. 1, January 2016

124