Preoperative Antibiotics and Mechanical Bowel Preparation Impact the Colonic Mucosa–Associated Microbiota but Not Anastomotic Leak Rate After Colorectal Resection

BACKGROUND: Alterations in the GI microbiome have been associated with increased anastomotic leak risk. Oral antibiotics and bowel preparations may both reduce anastomotic leak rates. OBJECTIVE: In patients undergoing colorectal surgery, we aimed to examine the impact of oral antibiotic use, bowel preparation, and other perioperative factors on the mucosa-associated microbiota and investigate the association with anastomotic leak rates. DESIGN: We opportunistically sampled Australian patients undergoing colorectal resection for any indication of anastomosis; mucosal swabs were taken from the proximal and distal extent of the resected specimen immediately after extraction. SETTINGS: A single-site, public tertiary referral hospital of 694 beds with an accredited specialist colorectal unit in Newcastle, Australia. PATIENTS: Of 192 patients, 31 patients received oral antibiotics presurgery (19 of whom also underwent mechanical preparation). In total, 31 patients received mechanical preparation and 58 received enema presurgery. Ninety-one patients did not receive antibiotics or undergo bowel preparation. MAIN OUTCOME MEASURES: The influence of patient factors on mucosa-associated microbiota composition and associations to rate of anastomotic leaks. RESULTS: The leak rate was 15% (N = 28/192). Preoperative oral antibiotic manipulation and bowel preparation minimally affected the mucosa-associated microbiota, but neither were associated with anastomotic leaks. Erythromycin (N = 14/192) had the greatest impact on the mucosa-associated microbiota. Of the different bowel preparation regimens, excluding patients given antibiotics, only Fleet enemas (N = 58/161) had a significant impact on the mucosa-associated microbiota. Bowel preparation was not associated with the occurrence of anastomotic leaks. There were only subtle differences observed in the mucosa-associated microbiota between patients who did and those who did not have anastomotic leaks. LIMITATIONS: A modest number of patients who received antibiotics were included. CONCLUSIONS: This is the largest series of routine colorectal anastomoses for which the mucosa-associated microbiota adjacent to the anastomoses was examined. We show that antibiotics and bowel preparation in the preoperative period exert only limited effects on the mucosa-associated microbiota and conclude that there are no obvious mucosa-associated microbiota characteristics that are predictive of an anastomotic leak. See Video Abstract.