P557 Increased risk of anastomotic leak in Crohn’s disease patients unable to complete preoperative mechanical bowel preparation

I. Iesalnieks1, F. Marek2, Z. Kala2, L. Kunovsky3

1Department of Surgery, München Klinik Bogenhausen, Munich, Germany, 2Faculty of Medicine, University Hospital Brno, Masaryk University, Brno, Czech Republic, 3Department of Gastroenterology and Internal Medicine, Faculty of Medicine, University Hospital Brno, Masaryk University, Brno, Czech Republic

Background

To assess the risk of postoperative anastomotic leak in Crohn’s disease patients unable to complete the preoperative mechanical bowel preparation (MBP): a prospective observational study from two referral centres in Germany and the Czech Republic.

Methods

Preoperative MBP was used routinely in all Crohn’s disease patients undergoing elective ileocolic or colorectal resections completed by the formation of an anastomosis since 6/2016. The MBP consisted of 2 L Polyethyleneglycol (PEG) solution combined with two doses of oral antibiotics Metronidazole and Paromomycin. The MBP was defined as incomplete when patients were not able to drink the whole amount of PEG solution due to side effects or complications. The primary endpoint was occurrence of anastomotic leak. The secondary endpoint was the incidence of postoperative intraabdominal septic complications (IASC) which were defined as an anastomotic leak, intraabdominal abscess or fistula and peritonitis.

Results

Between 6/2016 and 11/2019, 96 Crohn’s disease patients underwent elective ileocolic or colorectal resections after receiving preoperative MBP and oral antibiotics. Twenty-four (25%) developed complications of MBP, mostly vomiting; 17 patients (18%) were not able to complete MBP. The presence of extraintestinal disease manifestations (Hazard Ratio 4.8, p = 0.029), preoperative weight loss (HR 5.7, p = 0.019) and female sex (HR 13.3, p = 0.005) were associated with an increased probability not to be able to complete MBP. Postoperative anastomotic leak occurred in 2 patients (2%). The risk of anastomotic leak was significantly higher in patients unable to complete MBP (12%) as compared with patients with complete MBP (0%, p = 0.03). Postoperative IASC occurred in 7 patients (7%). Patient unable to complete preoperative MBP were at higher risk to develop IASC; however, the difference was not statistically significant (18% vs. 5%, p = 0.10)

Conclusion

The anastomotic leak rate is very low when preoperative mechanical bowel preparation and oral antibiotics have been used. However, patients not able to complete MBP might be at an increased risk.