During the past several decades, total mesorectal excision has dramatically increased the proportion of sphincter-saving procedures applied in the treatment of rectal carcinoma. However, an increased risk of anastomotic dehiscence (AD) is associated with total mesorectal excision which potentially endangers the blood supply of the rectal stump. AD requiring reoperation, the most invasive type of intervention, should be considered as an important point in clinical research. These patients are exposed to the risks of a second anesthetic and operative intervention, and they are usually managed by Hartmann procedure with takedown of anastomosis and future operation to restore intestinal continuity. Several authors have also reported data showing impaired oncologic outcome for patients with AD who underwent reoperation.^1–3

This study aims to perform a systematic review of studies that describe AD requiring reoperation and its subsequent mortality compared with the overall postoperative mortality after anterior resection for rectal carcinoma. Given the impact that total mesorectal excision and stapling devices have had on AD, only studies using these techniques after the year 2000 will be included in this review.

Methods

Results

Included studies

We identified 807 studies, of which 41 were included. The most common reason for not meeting the inclusion criteria was that the study described the rate of AD without providing specific data on the intervention of AD. Two studies with the same affiliation and duplicate time of patients' date as two other studies were also excluded.^4,5 A total of 39 studies were included in our analysis.^6–44

The included studies had a total population of 24,232 patients, with sample sizes ranging from 82 to 2729 patients. Four randomized, controlled trials were included, and the number of nonrandomized prospective and retrospective clinical trials was 12 and 23, respectively (Table 1).

Table 1 Details of the included articles

View Fullscreen

Incidence of AD

We used the number of ADs as defined in the studies. A total of 2075 dehiscences were confirmed in 39 studies. The overall pooled rate of AD was 8.6%. A large variation in AD rates was observed. The highest and lowest reported AD rates were 20.6% and 1.2%, respectively (Fig. 1).

View Figure

• Download as Powerpoint
• Download Figure

Mortality caused by AD and overall postoperative mortality

The total number of patients who underwent anterior resection in the 25 studies was 14,159. Dehiscence occurred in 1194 patients, of whom 715 required reoperation. A total of 59 patients died as a direct consequence of AD. Thus, the pooled mortality rate in AD patients requiring reoperation reached as high as 8.3% (range, 0%–54.6%). The pooled mortality caused by AD in the patients of the 25 studies was 0.4% (range, 0%–1.8%), and the overall postoperative mortality rate after anterior resection was 1.3% (range, 0%–3.0%). Thus, almost one third of the overall postoperative mortality was caused by AD. This percentage ranged from 0% to 100% (Fig. 2).

View Figure

• Download as Powerpoint
• Download Figure

Discussion

AD is a predominant cause of morbidity and mortality after anterior resection. This effect is more obvious for ADs requiring reoperation, which commonly require additional operative interventions and a prolonged hospital stay.^41,46,47 Aside from its adverse effects on late functional results, AD also impairs the long-term prognosis of patients with rectal carcinoma, particularly when reoperation is required.^1–3 The mechanisms of the negative prognostic effects of AD requiring reoperation include more pronounced immune system suppression caused by reoperation, more marked inflammation associated with AD requiring surgical revision, such as endotoxin that switches off apoptosis and thus leaves mitosis to shift the balance in favor of tumor growth, or transmission of viable exfoliated carcinoma cells during further surgery.^48,49

The incidence of AD after anterior resection is an essential measure to determine the clinical value of different operative and perioperative interventions because of its effect on the postoperative course and long-term survival of a patient. Thus, the incidence of AD after anterior resection is often selected as a primary end point in clinical trials. However, the results of different studies varied, largely because of the lack of objectives and easily applicable definitions of AD, making comparisons complicated. Therefore, clear conclusions on the type of operative and perioperative management preferred in daily practice were hindered. In this analysis, we found that the definitions of AD and postoperative mortality varied over studies. Definitions were mainly based on clinical symptoms, and several authors disagreed about whether pelvic abscess or rectovaginal fistulas should be considered as AD.^8,22,36 To standardize reporting of clinical studies, the International Study Group of Rectal Cancer proposed a generally acceptable definition in 2010,⁴⁵ which can be adopted by clinicians for their studies and help the readers in comparing results of different reports. However, in this analysis, we found that only one author reported on ADs based on the standard definition.

Previous studies showed that rectal carcinoma patients prefer to be involved in the decision-making process and to be informed on the risks of different treatment options.⁵⁰ Despite the vast body of evidence on the potential risk factors, AD remains difficult to predict in an individual patient. Because the selection process of high-risk patients remains imperfect, we emphasize that each patient should be counseled preoperatively on the risk of AD, the consequences including mortality, and the possibility of reoperation. Thus, clinicians need reliable and accurate information on the rates of AD requiring reoperation and the associated mortality after resection for rectal carcinoma.

Patients of AD who require reoperation are often quite ill. Most often, these patients have purulent/fecal drain contents in their drains and exhibit markedly increased infection parameters (leukocytosis, C-reactive protein). They usually have abdominal pain and fever, and subsequently develop signs of peritonitis (tenderness to palpation, abdominal wall rigidity, tachycardia, etc). If operative reintervention with control of the septic source is delayed or not performed, the clinical condition of these patients will deteriorate and ultimately result in sepsis with clinical signs of hypothermia, leukopenia, and organ failure. Imaging studies, such as computed tomography with transrectal instillation of contrast, reveal considerable leakage at the anastomotic site, with fluid collection in the pelvis.

In this thorough overview of studies describing AD requiring reoperation and subsequent mortality following resection of rectal carcinoma, we found that patients have a considerable pooled rate of 5.4% for AD requiring reoperation, which indicates that 63.1% of AD patients need reoperation. Moreover, 89.6% of patients with AD requiring reoperation need temporary or permanent diverting stoma construction. We also found that AD considerably contributed to mortality after rectal carcinoma surgery, accounting for almost one third of all postoperative mortality. AD is still the most frequent cause of death after anterior resection in this analysis, with a pooled postoperative mortality rate of 8.3% in patients with AD requiring reoperation. This finding is supported by and comparable with the results of a nationwide population-based study in Sweden exclusively assessing patients who died following anterior resection of the rectum.⁵¹ The study describes 140 of 6833 patients who died following rectal excision with anastomosis during the initial hospital stay or within 30 days, of whom 42% (59 patients) had AD as a cause of mortality, a result highly similar to our findings.

Although accurate prediction of risk is impossible, certain factors are known to influence AD rates. The clinical AD rate slightly increased after introduction of the total mesorectal excision technique.⁵² Based on the cause analysis of AD after anterior resection for rectal carcinoma, numerous parameters, such as distance of the anastomosis from the anal verge, tension of the anastomosis, bacterial contamination, anastomotic blood flow, surgical techniques, and defunctioning stoma, were reported to be responsible for the adverse effects on anastomotic healing.^{20–24,34–37} Early urgent therapeutic intervention is required to avoid the life-threatening consequences of clinical AD. For severe dehiscence without septic complications, a loop ileostomy or colostomy is necessary to prevent fecal contamination, especially for patients without defunctioning stoma in the initial operation. When a broad anastomotic insufficiency is accompanied by sepsis or peritonitis, a revision of the anastomosis is necessary. Only an aggressive surgical treatment can save the life of a patient under these circumstances. Hartmann operation is referred to as the “gold standard” technique when sepsis is present.^18,42,53

The limitation of this article is that the publications were all sourced from English-language papers, and non–English-language papers were excluded in the analysis. A formal meta-analysis will provide more powerful evidence. The methodology used in the present study was not as powerful as a meta-analysis. However, our systematic review provides the summarized data on the AD rate and on mortality directly from the original publications.

Thus, the present study can be used for counseling patients preoperatively. With a considerable risk of reoperation and postoperative mortality, AD remains the most serious complication following anterior resection of the rectum for rectal carcinoma. Frequent feedback of these outcomes can support clinicians when informing patients about the risks of different treatment options and can facilitate the decision-making process.