Recurring Anastomotic Leak-A Prospective Clinicopathological Investigation of a Distinct Disease Pattern

Risk factors for gastrointestinal perforation and anastomotic leak in patients submitted to cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC)

BACKGROUND

Gastrointestinal leak is one of the most feared complications after cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS-HIPEC) and harbors significant postoperative morbidity and mortality. We aim to identify risk-factors for anastomotic leak (AL) and gastrointestinal perforation (GP) to optimize postoperative outcomes of this population.

METHODS

We performed a retrospective analysis of 1043 consecutive patients submitted to CRS in a single institution. Potential risk factors for AL and GP, both related to patient overall condition, disease status and surgical technique were reviewed.

RESULTS

Anastomotic leaks were identified in 5.2% of patients, and GPs in 7.0%. The independent risk-factors for AL were age at surgery (OR1.40; CI95% 1.10-1.79); peritoneal cancer index (PCI) (OR1.04, CI95% 1.01-1.07); Cisplatin dose >240 mg during HIPEC (OR3.53; CI95% 1.47-8.56) and the presence of colorectal (CR) or colo-colic (CC) anastomosis (OR5.09; CI95% 2.71-9.53, and 4.58; CI95% 1.22-17.24 respectively). Male gender and intraoperative red blood cell transfusions were the only independent risk factors for GP identified (OR1.70; CI95% 1.04-2.78 and 1.06; CI95% 1.01-1.12, respectively). Regarding 30-day and 90-day postoperative mortality, independent risk-factors were mainly related to patient's overall condition.

CONCLUSION

Gastrointestinal leaks are a frequent source of postoperative morbidity, mainly at the expense of GP. A careful and systematic intraoperative revision of all potential gastrointestinal injuries is equally critical to perfecting anastomotic fashioning techniques to decrease gastrointestinal complication rates. We identified multiple risk-factors for AL and GP related to disease status and patient condition. Our study suggests that patient-related conditions are of paramount relevance, highlighting the importance of patient selection and preoperative patient optimization.

Inhibition of HIF-prolyl hydroxylases improves healing of intestinal anastomoses

Anastomotic leakage (AL) accounts for a major part of in-house mortality in patients undergoing colorectal surgery. Local ischemia and abdominal sepsis are common risk factors contributing to AL and are characterized by upregulation of the hypoxia-inducible factor (HIF) pathway. The HIF pathway is critically regulated by HIF-prolyl hydroxylases (PHDs). Here, we investigated the significance of PHDs and the effects of pharmacologic PHD inhibition (PHI) during anastomotic healing. Ischemic or septic colonic anastomoses were created in mice by ligation of mesenteric vessels or lipopolysaccharide-induced abdominal sepsis, respectively. Genetic PHD deficiency (Phd1-/-, Phd2+/-, and Phd3-/-) or PHI were applied to manipulate PHD activity. Pharmacologic PHI and genetic PHD2 haplodeficiency (Phd2+/-) significantly improved healing of ischemic or septic colonic anastomoses, as indicated by increased bursting pressure and reduced AL rates. Only Phd2+/- (but not PHI or Phd1-/-) protected from sepsis-related mortality. Mechanistically, PHI and Phd2+/- induced immunomodulatory (M2) polarization of macrophages, resulting in increased collagen content and attenuated inflammation-driven immune cell recruitment. We conclude that PHI improves healing of colonic anastomoses in ischemic or septic conditions by Phd2+/--mediated M2 polarization of macrophages, conferring a favorable microenvironment for anastomotic healing. Patients with critically perfused colorectal anastomosis or abdominal sepsis could benefit from pharmacologic PHI.