QS208 - Anastomotic Leak Occurs Independently from Microbiota Shifts Associated with Surgical Bowel Preparation

Purpose/Background: Following bowel surgery, microbiota-driven complications including anastomotic leak (AL) and surgical site infections (SSI) remain major sources of morbidity and mortality. For more than 50 years surgical bowel preparation (SBP), the combination of mechanical bowel preparation (MBP) and oral antibiotics (OA), has been administered with the assumption that gut decontamination serves to reduce post-surgical complications. Using a murine model of colon surgery, we investigated the components of SBP individually and in combination to determine microbiota compositional changes associated with the development of anastomotic dehiscence. We hypothesized that mice administered SBP would have a significant reduction in AL and SSI, and that this benefit would be increased in Western diet fed mice with resultant gut dysbiosis.

Methods/Interventions: C57BL/6J mice (n = 144) were randomized to a 10-12 week feeding period of either standard diet (SD; n = 72) or Western diet (WD; n = 72) before colon anastomosis (Figure 1a). Twenty-four hours prior to surgery, mice received one of four interventions: water (control), MBP, OA (neomycin and metronidazole each at concentrations of 1mg/mL), or both MBP and OA. Fecal microbiota were characterized by 16S rRNA sequencing at 4 timepoints: prior to starting bowel preparation treatment, day of surgery (DOS), post-operative day (POD) 4, and POD7. Upon euthanasia on POD7, mice were assessed for AL and wound healing.

Results/Outcomes: AL rates were not significantly different between SD and WD mice (44% versus 30%, p = 0.08) or among bowel preparation groups (MBP 49%; OA 23%; SBP 44%; water 31%; p = 0.09) although there was a trend toward fewer ALs in mice treated with OA alone compared to the other treatment groups (p = 0.07). When stratified by diet, the effect of bowel preparation on AL showed no differences (p = 0.27 and 0.39 for SD and WD, respectively). SD mice had significantly fewer non-healed wounds compared to WD mice (p < 0.01) although no differences were observed between bowel preparation groups (p = 0.63) or bowel preparation groups stratified by diet (p = 0.84 and 0.71 for SD and WD, respectively). Among SD groups, no significant differences in genera were observed at DOS or POD4 between mice with or without AL (Kruskal-Wallis p ≥ 0.066) (Figure 1b). Bacteroides had significantly greater relative abundance in SBP-treated WD-fed mice who experienced AL on POD4 (p = 0.04) while the relative abundance of Paramuribaculum was greater in those that did not have AL (p = 0.01) (Figure 1c).

Conclusion/Discussion: Despite changes observed in the microbiome, no differences in the rates of AL or wound breakdown were identified from the administration of any component of bowel preparation. However, we did identify bacterial genera that may be causally associated with AL. Future studies are needed to potentially develop therapeutic interventions that mitigate pathogenic microbial shifts with the goal of reducing postoperative complications.