Special Considerations of Anastomotic Leak in Patients with Rectal Cancer

Abstract

Anastomotic leak in patients with rectal cancer has the potential to cause worse oncologic outcomes in addition to major morbidity and mortality risk of this dreaded complication. Anatomic location of the rectal cancer determines the ability to perform a restorative operation and the height of the anastomosis in relation to the anal canal. Clinical staging dictates the need for neoadjuvant treatment (such as chemotherapy and radiation) which may also contribute to anastomotic leak risk. In addition to oncologic outcomes, anastomotic leak can impact bowel function, the need for permanent stoma, and long-term quality of life. This study will discuss special considerations for anastomotic leak prevention and clinical implications of this complication in patients with rectal cancer.

Anastomotic complications in patients with rectal cancer deserve special consideration due to the potential for worse oncologic outcomes and bowel function. Leak prevention starts with appropriate preoperative evaluation of the patient and tumor, and intervention to change risk factors, when able. Appropriate intraoperative decision-making and sound surgical technique influence the outcome. If a leak does occur, rapid recognition and management help alleviate some of the negative long-term consequences. This chapter will discuss considerations for anastomotic leak prevention, management, and clinical outcomes in patients with rectal cancer.

Anastomotic Leak Prevention

The best anastomotic leak management strategy in all patients is prevention. Both clinical judgment, as well as technical expertise, must be employed. For rectal cancer patients, this clinical judgment starts in the preoperative planning phase and is continued into the operating room during the surgery, as well as in postoperative care. Patients with rectal cancer should be discussed in a multidisciplinary team setting to assess the patient's need for radiation, chemotherapy, and potential need for extended resection margins. These, and several other factors have been associated with an increased anastomotic leak risk. Some factors are modifiable, and others are not.

Preoperative Factors

When assessing a rectal cancer patient preoperatively, the surgeon must consider factors that will affect the patient's risk of leak and try to optimize any modifiable risk factors ( Table 1 ). Such factors include smoking history, malnutrition, glucose control in diabetic patients, optimization of medications from comorbid diseases (e.g., steroid use), draining any sources of sepsis (e.g., perforated cancers), and awaiting reduction in inflammatory changes when possible. Discussion of the increased risk associated with these factors with the patient can be a motivating factor for improving patient buy in and will set more realistic expectations for postoperative outcomes.

Table 1. Risk factors associated with anastomotic leak.

Modifiable	Not modifiable
Smoking status	Male sex
Malnutrition	Emergency surgery
Glucose control for diabetics	Location of anastomosis
Steroid use	Radiation history
Presence of abscess/inflammation	History of vascular disease
	Comorbid disease burden

There are several factors associated with leak in rectal cancer patients that cannot be mitigated. Some of these factors include certain comorbid disease burdens, history or need for preoperative radiation, emergency surgery, male sex, and location of the anastomosis. ^{1 2 3 4} Male sex poses a higher risk for leak in low pelvic anastomosis, likely due to the greater technical difficulty with the narrow pelvis, especially when there is a bulky tumor. Similarly, the technical difficulty of performing a low or ultralow pelvic anastomosis causes a higher risk for anastomotic leak compared with colorectal or ileocolic anastomoses. Other risk factors to consider are history of pelvic radiation. While these risk factors cannot be prevented, they can be anticipated to allow for better preoperative counseling and planning. The timing of surgery after completion of radiation therapy may have an impact on leak rates. In a randomized trial, patients who had surgery within 8 weeks after completion of chemoradiotherapy were found to have a higher leak rate of 10.8% compared with 4.5% in patients who had their operation 12 weeks after completion of neoadjuvant treatment. ⁵

Preoperatively, the surgeon should evaluate the risk of leak and have a plan in mind that perhaps an anastomosis will not be done at this surgery but deferred until a more favorable medical and surgical situation. High-risk anastomoses should also be considered for fecal diversion, such as a diverting loop ileostomy. The authors' practice is to divert any colorectal anastomosis below the peritoneal reflection. These discussions with patients should be held in the preoperative visit and informed consent obtained, even if ultimately there is no diversion. While diverting stomas are felt by some to decrease the consequences of leak rather than the incidence of leak, a 2010 Cochrane review by Montedori et al demonstrated that diverting ileostomy or colostomy decreases not only the need for urgent reoperation, but also the rate of anastomotic leak. ⁶ Temporary and permanent stoma markings should be performed in the preoperative setting considering the location in both the sitting and standing positions to optimize the stoma function and care given to the individual's body habitus.

The surgeon cannot alter a patient's presentation in an emergent rather than elective setting, and patients with rectal cancer often present with obstruction and perforation. The decision to offer resection with anastomosis versus diversion with or without drainage can drastically affect a patient's oncologic and functional outcomes. Rectal cancer patients who present in the emergent setting often have advanced disease that is either locally unresectable or metastatic. ⁷ Performing an emergency resection in the setting of inflammation from a perforation is likely to result in positive margins. In addition, colorectal or coloanal anastomosis using dilated bowel with poor tissue quality due to an obstructing cancer has a higher risk of leak. In the emergency setting, patients should be staged after initial fluid resuscitation for stabilization. Diversion and drainage are surgical options that allow expedient treatment with chemotherapy and radiation prior to resection and should be considered when clinically appropriate.

There is novel research evaluating the role of the gut microbiome in both the risk of rectal cancer and risk of anastomotic leak. Zhang and colleagues found that oral antibiotic use is associated with decreased risk of rectal cancer. ⁸ Gaines and colleagues have identified bacteria that produce collagen-digesting enzymes that break down host tissue extracellular matrix proteins. ⁹ In animal models, it appears that these bacterial enzymes contribute to anastomotic leak. Further research in this area is needed to evaluate the microbiome composition of rectal cancer patients and its impact on anastomotic leak.

Intraoperative Factors

Several intraoperative factors relevant to rectal cancer patients have been associated with an increased risk of anastomotic leak. The surgeon must be aware of these risks and if present during the operation. As discussed above, the need for emergency surgery cannot be changed; however, it should be recognized that it is associated with an increased risk of leak with an odds ratio of 1.8 compared with elective resection and anastomosis. ⁴ Most intraoperative risk factors are unexpected and develop with difficulties or complications during the case. A prospective study of 616 patients undergoing colorectal resection, intraoperative hypoxia with oxygen saturation <90% for more than 5 minutes was associated with anastomotic leak. Other factors shown to be associated with an anastomotic leak included intraoperative hypotension with a systolic blood pressure of <85 mm Hg, or mean arterial pressure of <60 mm Hg, and metabolic acidosis (pH < 7.30). ^{10 11} The surgeon should communicate with the anesthesia team and assess the patient's general stability and tolerance of the operation prior to performing an anastomosis and to inform the decision to perform a proximal diversion prior to completion of the case.

Clinical and Technical Aspects of Anastomotic Construction

The quality of the anastomosis starts well before its construction as gentle handling of tissue, preservation and maximization of perfusion, and providing adequate mobilization for reach all contribute to a successful anastomosis. Several factors need to be considered and are of particular importance in patients with rectal cancer. First, if the patient is hemodynamically unstable, requiring transfusion, requiring a prolonged operation, or is otherwise unfit to undergo an anastomosis, an end stoma should be strongly considered. The surgeon should communicate with the anesthesia team prior to anastomosis creation to determine if the patient is on vasopressor support and determine if any blood products have been required during the case. In circumstances where performing an anastomosis is determined to be unwise, a nonabsorbable marking stitch can be placed on the distal end of bowel to allow for easier visualization during a potential future reversal surgery. This is more common in patients undergoing emergency surgery for perforated or obstructed cancer or patients undergoing redo surgery.

When proceeding with an anastomosis, the surgeon should assess the ends of the bowel for tissue quality and blood flow. Radiation, inflammation, and sepsis can all result in poor tissue for reconnection and these should be avoided, resecting back to healthy bowel when able. Patients who undergo neoadjuvant radiation will have radiation changes in the distal aspect of the anastomosis. The blood flow to the proximal bowel to be used in the anastomosis must be adequate. Assessment of blood flow can be performed by cutting the bowel or epiploic appendage next to the edge of the bowel to assess for venous and arterial bleeding. One should expect to find brisk, “nuisance” bleeding in the submucosal tissue. Bowel color and palpable flow in the mesenteric arteries can be used to assess perfusion. One simple test of adequate perfusion for colon anastomoses is to loosely clamp the marginal artery before dividing it, then opening the clam and checking for flow. If there is concern for adequate blood flow, evaluation by Doppler ultrasound of vessels leading to the cut edge of bowel is a simple and inexpensive ways to assess for venous and arterial blood flow. Some surgeons utilize fluorescence imaging with indocyanine green (ICG) to evaluate perfusion, although this is not the routine practice of the authors.

Another basic technical principle to anastomosis creation is the lack of tension on the anastomosis which is achieved by creating enough laxity of the bowel to be reconnected. With low colorectal or coloanal anastomoses, achieving a floppy colon conduit can be a challenge. Several maneuvers can be employed to assist in reducing tension at the site of anastomosis. For gaining length in the descending colon, it is often helpful to perform a high ligation of the mesenteric vessels, division of the left colic artery, and high ligation of the inferior mesenteric vein reduces tension at the level of the mesentery. Splenic flexure mobilization improves mobility with the potential of avoiding division of other blood vessels which could result in reduced blood flow to the distal bowel. For gaining length in the transverse or right colon, the surgeon may have to divide the middle colic artery. Full mobilization of the hepatic flexure with rotation of the right colon and even passing the colon conduit behind the ileum, also known as a retroileal anastomosis, provides a technique to gain length.

Every pelvic anastomosis should be assessed for completeness of anastomotic tissue doughnuts (when a stapled anastomosis is done) and for any defects by endoscopy with a leak test. A leak test is performed by submerging the anastomosis under saline or water, occluding the bowel proximal to the anastomosis, filling the bowel with air, and assessing for any bubbles under the liquid. The bowel must be filled enough to ensure full expansion of the anastomosis under pressure. The authors' preference is to use a flexible sigmoidoscope to insufflate for the leak test, as this also allows for visualization of the anastomosis for integrity, hemostasis, and patency. For low coloanal anastomoses, a leak test may not be feasible due to its location, but visual inspection and palpation of the anastomosis should be performed to assess for defects.

If a defect in the anastomosis is identified during a leak test, the surgeon must remedy the situation. A study by Ricciardi et al ¹² revealed that of 825 left-sided colorectal anastomoses, 65 had air leaks on intraoperative testing (7.9%). For patients who had a positive air leak test, suture repair alone was associated with a higher rate of postoperative clinical leak when compared with repair with diversion or redo of the anastomosis (12.2 vs. 0 and 0%; p  = 0.19). ¹² A repeat leak test should be performed to assess for continued defects after a repair is performed. If the anastomosis is resected and redone, the surgeon must ensure that the loss of bowel length does not lead to tension on the anastomosis. In addition, if a repair is performed or the redo anastomosis leads to a very distal anastomosis, consideration of a proximal diverting ostomy must be made as well.

Management and Outcomes of Anastomotic Leaks in Rectal Cancer Patients

The management of patients with anastomotic leak depends on the acuity of presentation, underlying clinical and anatomic factors, and the treatment modalities available and familiar to the surgeon. The first step is to resuscitate the patient and alleviate sepsis as needed. The fact that the patient has a cancer diagnosis should not delay the timeliness of intervention. Management options vary and include transanal drainage, transgluteal or transabdominal drainage via interventional radiology, placement of a transanal/endoscopic vacuum-assisted closure (VAC) device, washout and drainage, and resection or redo of the anastomosis. Detailed management of colorectal and coloanal anastomotic leaks are discussed elsewhere in this issue.

Oncologic Outcomes after Anastomotic Leak after Surgery for Rectal Cancer

Mounting evidence supports that anastomotic leak after surgery for rectal cancer has significant negative impact on oncologic outcomes. Two relatively recent meta-analyses provide similar findings. Lu et al reviewed 13,655 patients across 11 studies including 5 prospective cohorts demonstrated that an anastomotic leak was associated with an increased risk of local recurrence (odds ratio [OR] = 1.6, p  < 0.01), and higher cancer-specific mortality (OR = 1.3, p  < 0.05), but not distant recurrence. ¹³ Wang et al reported on 11,353 patients from 14 studies. Anastomotic leak was associated with higher local recurrence (hazard ratio [HR] = 1.71), decreased cancer-specific survival (HR = 1.3), and overall survival (HR = 1.67) but not distant recurrence. ¹⁴

The definitive biological reason for why anastomotic leaks are associated with, or possibly directly cause, is that the worse oncologic outcomes remain unknown. There may be a mechanical aspect where the opening in the bowel provides an opportunity for exfoliated cancer cells to adhere to the exposed luminal surface. It is believed that the inflammation and reaction to the leak provides an impetus for cancer cell tracking and adherence. The inflammatory response also causes host immunosuppression which impacts the ability to kill residual cancer cells, and thus perhaps contributes to recurrence. Lastly, the complications associated with anastomotic leak delay or obviate the use of adjuvant chemotherapy. A recent study of 532 rectal cancer patients reported a 5-year overall survival of 67.2% for patients with an anastomotic leak, compared with 86.5% for those who did not ( p  < 0.001). ¹⁵ More severe leaks were independently associated with delay or no delivery of adjuvant chemotherapy and was also an independent predictor of worse overall survival.

Bowel Function, Quality of Life, and Need for Permanent Stoma

In the early time period after diagnosis of anastomotic leak, patients often require unanticipated painful procedures (e.g., percutaneous drain placement for infection), delayed or permanent stomas, and unexpectedly prolonged periods of pain and fatigue, impacting patients' quality of life. These delays in recovery and repeat hospital visits lead to disruptions to the patient's life and the lives of their family or support system. Financial hardship occurs even with universal insurance coverage from indirect costs to care. ¹⁶

These unexpected outcomes after surgery impacts patients in the long term as well. Over time inflammation from the leak leads to pelvic fibrosis. These fibrotic changes can affect nerve and muscle tissue of the pelvis, leading to chronic pain and changes to bowel function, including fecal continence. A study from Sweden including 1,180 rectal cancer patients evaluated the effects of defecatory, sexual, and urinary dysfunction 2 years after restorative proctectomy. Of this group, 7.5% developed an anastomotic leak. On multivariate and linear regression modeling, patients with leak had an increased risk of fecal incontinence (OR = 2.27) and reduced sexual activity ( p  = 0.003). ¹⁷

Approximately 5 to 10% of patients who undergo a restorative procedure for rectal cancer will require a permanent stoma. This can be secondary to chronic infection, inflammation, formation of sinus tracts, stenosis, or poor function. A study from Korea evaluated 2,528 consecutive rectal cancer patients who underwent low anterior resection with temporary diverting loop ileostomy for rectal cancer. Anastomotic complications were an independent risk factor for permanent stoma ( p  = 0.001). ^{18 19} In another multivariate analysis of 673 patients with rectal cancer, anastomotic leakage was independently associated with the need for a permanent stoma ( p  < 0.001). ²⁰

There is a paucity of information regarding the impact of anastomotic leak on rectal cancer patient quality of life; however, several studies have evaluated the quality of life of various treatments in rectal cancer patients. A quality of life study of rectal cancer patients by Feddern and colleagues found that patients undergoing low anterior resection had a significantly worse quality of life regarding bowel function on both univariate (OR = 3.38; p  < 0.001) and multivariate (OR = 3.71; p  < 0.001) analysis compared with abdominoperineal resection. In addition, low anterior resection patients had a lower global health status (OR = 1.2; p  = 0.03) compared with abdominoperineal resection patients. ²¹ A population-level study of health-related quality of life in 6,713 rectal cancer patients in England reported higher levels of severe sexual difficulties and worse health-related quality of life in patients with a stoma. This study found that patients who had radiation therapy in addition to surgery had worse bowel control (43.6 vs. 33.0%) and higher rates of severe urinary leakage (7.2 vs. 3.5%). ²² Future studies assessing the quality of life impact of anastomotic leak on bowel function and global health status would improve our understanding of the impact of this complication. When considering a rectal cancer patient who has suffered from a leak, the surgeon should consider the short- and long-term impact on a patient's quality of life and counsel the patient on the potential trade-offs of permanent stoma versus bowel continuity if their symptoms do not improve with time.

Conclusion

Anastomotic leak after surgery for rectal cancer can have devastating consequences both acutely and in the longer term. Surgeons must carefully consider the preoperative, intraoperative, and postoperative factors that can be optimized when caring for these patients. Early detection and appropriate initial management of anastomotic leak is the key to reduce early morbidity and mortality, and limit the consequences of longer term functional and oncologic outcomes.