Overview

Anastomotic leak remains one of the most serious complications after restorative rectal surgery. It is widely accepted that proximal diversion with an ileostomy or colostomy reduces the clinical significance of postoperative anastomotic leaks.

Although there currently are no absolute indications for proximal diversion, the most commonly accepted risk factors for postoperative leak are:[1][2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12],[13],[14],[15]

• Comorbid medical conditions of the patients (e.g., cardiac disease, diabetes mellitus/elevated HgbA1c, anemia, malnutrition)
• Prior pelvic radiotherapy
• Level of the anastomosis (< 5 cm from the anal verge)
• Technically challenging pelvic dissection in the setting of:
  • Male sex (narrow pelvic working space
  • Obesity
  • T4 and/or large bulky tumor
  • Prior pelvic surgery
• Tobacco/nicotine use
• Immunosuppression and/or immunosuppressive medical therapy (e.g., steroids, biologic medical therapy for inflammatory disease, organ transplantation, poorly controlled human immunodeficiency virus [HIV])
• Emergent operation
• Technical mishaps with the anastomosis (e.g., incomplete stapler donut/ring, positive air leak test, need for reinforcing sutures, multiple stapler firings during a minimally invasive proctectomy; performing the anastomosis again is recommended when the procedure is anatomically and technically feasible)

A randomized trial of patients with rectal cancer and who underwent anastomoses < 8 cm from the anal verge demonstrated a leak rate of 5.8% in the stoma group; the leak rate was 16.3% in the group without stoma (p = 0.04). In this study, male gender and lack of stoma usage were associated with anastomotic leakage when they employed multivariate analysis. Of note, use of radiochemotherapy, anastomotic height, and body mass index were not found to be associated with an increased risk of developing anastomotic leak. However, the definition of anastomotic leak and the investigation for anastomotic leak were not well defined or interrogated compared with other recent publications.[3]

Many surgeons still use drains for low pelvic anastomoses. A recent systematic review on the use of prophylactic pelvic drainage after low anterior resection suggests no obvious benefit in regard to reduction in leak rate or overall complications. One study noted a reduction in the mortality rate of the drained group.[16] However, another propensity score-matched retrospective analysis of 6,157 patients concluded that even though drains increased the length of stay by 0.86 days, drains did not impact on any other secondary outcomes, such as surgical site infections, anastomotic leakage, overall morbidy, mortality, and reoperation rates. The authors of this analysis concluded that drains should not to be used routinely.[17]

Randomized controlled studies also did not find a significant difference regarding mortality, leak, and other postoperative complications in patients who were drained versus those who were not drained.[18][19][20][21] These findings led to the conclusion in the American Society of Colon and Rectal Surgeons 2023 Clinical Practice Guidelines for Enhanced Recovery after Colon and Rectal Surgery that the decision to place a drain is category 1B evidence and drains should not be routinely used.[7]

Anastomotic leak remains one of the most challenging and poorly defined complications after restorative rectal surgery. It is widely accepted that proximal diversion with ileostomy or colostomy reduces the clinical impact and subsequent consequences of postoperative low pelvic anastomotic leak. These consequences include increased rate of postoperative mortality, permanent stoma formation, and poor bowel function.[4] In addition, anastomotic leak after a restorative low anterior resection has adverse effects on rectal cancer oncologic outcomes and survival rates.[5][6]

Investigations into better understanding the etiology and prevention of low pelvic anastomotic leaks has been of increasing interest over the last several decades. In addition, alternative tools for fecal diversion are currently under investigation. These alternative tools may help us identify the development of low pelvic anastomotic leaks with internal stool diversion systems and protection minimizing the clinical consequences of the anastomotic leak. This may result in further evolution of rectal cancer surgery with selective utilization of proximal diverting stoma for only patients who develop a leak, rather than proximal diverting stoma for all high-risk patients. However, in the meantime, proximal diverting stoma remains the most widely utilized tool for patients with high-risk low pelvic anastomosis as a means to minimize the acute complications and oncologic perils associated with anastomosis leak.

The PILLAR III clinical trial aimed to evaluate the impact of indocyanine green fluoroscopy in the assessment of anastomotic perfusion and the impact on leak rate. Unfortunately, the study was closed early due to insufficient patient accrual. However, the study’s results still provide value in further evaluating and defining anastomotic leak rate. The multicenter study included 25 centers and recruited 347 patients who underwent restorative proctectomy with low pelvic anastomosis (< 10 cm from the anal verge) and then randomized patients to perfusion versus no perfusion assessment. The evaluation and definition of anastomotic leak were more critically defined and evaluated in this study compared with prior studies. Anastomosis leak was evaluated by flexible endoscopy in the early postoperative period. Any defect, regardless of clinical symptoms and/or radiographic findings, was defined as an anastomotic leak. The mean level of anastomosis was low in both groups (5.2 +/–3 cm). Anastomotic leak was reported as 9% and 9.6% in the perfusion and standard groups, respectively (p = 0.37).[8][9] While there was no statistical difference between the two groups, the PILLAR III clinical trial furthered progress in recognizing, defining, and identifying the incidence of anastomosis leak in high-risk patients undergoing restorative rectal surgery.

A similar anastomotic leak rate of 8.6% was observed in another study conducted at two centers in Egypt. This study, which collected the data prospectively, aimed to identify perioperative risk factors for anastomotic leakage and included both small intestinal and colorectal anastomosis, which may account for a slightly lower leak rate compared with the PILLAR III trial. This study identified low serum albumin levels and emergency operations as two independent risk factors for anastomotic leak.[10]

A randomized clinical trial, published in 2016, investigated patients with rectal cancer undergoing anastomoses < 8 cm from the anal verge over a 10-year period (2004–2014) with and without proximal diverting stoma. The leak rate was 5.8% in the stoma group and 16.3% in the no stoma group (p = 0.04). However, the definition of anastomotic leak and the investigation for anastomotic leak were not well defined or interrogated compared with other recent publications. In this study, male gender and lack of stoma usage for anastomosis lower than 6 cm were associated with statistically significantly higher rates of anastomotic leakage when they employed multivariate analysis.[3]

Because the rates of anastomotic leak are noted to be 5.8%–16.3% with or without diversion and given its consequences, efforts continue to evaluate techniques for prevention of anastomotic leaks. A recent randomized trial of 850 patients (the EssentiAL trial) noted a significant difference of anastomotic leak rate in those who utilized green fluorescence imaging versus those who did not. The patients all had rectal cancer tumors of < 12 cm from the anal verge and underwent minimally invasive technique for resection. The surgeons decided whether to perform diversion. All patients who underwent diversion had a contrast enema X-ray within 30 days. Those who were not diverted and had a clinical suspicion of a leak, a contrast enema or computed tomography (CT) was performed to confirm or disprove a leak. Leaks were also defined as grade A (a clinical asymptomatic leak solely confirmed by diagnostic studies), grade B (symptomatic leak that required nonsurgical intervention), or grade C (symptomatic leak that required surgical intervention). Those evaluated with indocyanine green had a significantly lower leak rate of 7.5% compared with the non-indocyanine group leak rate of 11.8%.[22]

Although the PILLAR III trial was not successful because of patient accrual, it laid the groundwork for future studies. A Japanese study showed that some interventional evaluation can help to minimize leaks. This same paper notes there are two other studies—the IntAct (multicentre EUropean) and AVOID (Netherlands)—that are underway but have not yet met accrual to report their results.[23][24]

Although there are no absolute indications for proximal fecal diversion for restorative low pelvic anastomosis after low anterior resection or proctectomy, many experienced, high-volume rectal cancer surgeons recommend proximal stool diversion consideration when two or more risk factors are present. Ultimately, the decision, recommendation, and preoperative discussion with the patient for proximal diversion should be evaluated and planned on a case-by-case basis. The most commonly accepted risk factors for postoperative low pelvic anastomotic leak were listed at the beginning of this chapter.

A recent meta-analysis evaluating the role of diverting ileostomy in rectal cancer reconstructive surgery further highlights the ongoing areas of gaps in defining and identifying low pelvic anastomotic leaks. A higher number anastomotic leaks were detected in patients without proximal diverting ileostomy in this multivariant analysis. This finding supports the conclusion that the majority of anastomotic leaks are currently identified only in patients with symptoms of anastomotic leaks, as we know that a proximal diverting ostomy does not prevent an anastomotic leak. Rather, the proximal diverting stoma facilitates minimizing the systemic complications and consequences of anastomotic leak, including emergent reoperation. This multivariant analysis highlights the deficits in our current understanding, identification, and definition of what constitutes a low pelvic anastomotic leak.[12]

Intraoperative evaluation of the low pelvic anastomosis with flexible endoscopy is widely recommended to both perform an air-leak test as well as direct inspection of the anastomosis and proximal mucosa. Another modality that may be used in early detection is a water-soluble rectal contrast fluoroscopy or cross-sectional imaging. Early identification of anastomotic leak may lead to earlier management and salvage of the low pelvic anastomosis. Furthering our understanding, definition, identification, and prevention of a low pelvic anastomotic leak are currently goals of many ongoing surgical research studies.

Last, while a diverting proximal stoma may minimize the severity of the acute and chronic clinical consequences of low pelvic anastomotic leak, a diverting loop ileostomy is not without negative impacts on clinical outcomes of its own. Diverting loop ileostomy is associated with a higher rate of hospital re-admission and acute kidney injury because of high output ileostomy and the resulting severe dehydration. In addition, postoperative complications can occur at the time of the second operation for ileostomy reversal, including wound infection, urinary tract infection, ileus, pneumonia, anastomosis leak, and hernia development. In order to avoid creating a diverting loop ileostomy in all high-risk patients for low pelvic anastomotic leaks, current clinical trials are investigating the use of internal stool diversion systems that help identify which patients will develop a leak and would then benefit from diverting loop ileostomy and which patients will not develop a leak.[12][13][14][15]