Is Chronic Pelvic Sepsis Complicating Low Anterior Resection of Rectal Cancer Preventable?

Abstract

The combination of anatomical inaccessibility, less-than-optimal blood supply, tightly closed anal sphincters below a low anastomosis, and an infected haematoma is likely to be contributory to anastomotic leakage following low anterior resection of the rectum for rectal cancer. Although under-reported, chronic pelvic sepsis complicating low anterior resection of the rectum is still a major problem associated with impaired quality of life. It should be avoided as much as possible, in addition to the fact that it is more difficult to manage surgically than acute sepsis. Primary preventive measures are well established. Secondary prevention of chronic pelvic sepsis is achieved by early diagnosis and active management of the anastomotic leak. However, optimal postoperative management cannot fully eliminate chronic sinuses or delayed reactivation leaks. With chronic leakage, major restorative redo-anastomosis or ablative abdominal perineal resection is required and 20% of patients will require a permanent stoma.

1. Introduction

Anastomotic leakage (AL) is still a major problem in colorectal surgery. The incidence ranges between 3 and 25%, indicating a multifactorial aetiology (tissue perfusion, radiotherapy, patient frailty, technical factors, total mesorectal excision), heterogeneity and inconsistency of definitions [1,2]. Advances in rectal cancer surgery, total mesorectal excision (TME) and the need for only a 1 cm margin below the tumour [3] has led to low colorectal anastomosis (CRA) and manual coloanal anastomosis (CAA), with or without intersphincteric resection, being commonly performed [4]. AL has a significant impact on both short-term postoperative recovery and long-term functional and oncological outcomes [2,3,4]. The effect of an anastomotic leak depends on whether it is a high anastomotic leak (intraperitoneal), which usually presents as peritonitis, or a low anastomotic leak (extraperitoneal), the presence or absence of stoma, and the time of detection (early or late). Other clinical features that may influence the formation of fistulae or prognosis include, but are not limited to, simple or complex sinus tracts, high-output or low-output cavities, fistulous or non-fistulous leaks and irradiated or non-irradiated tissue. Low colorectal anastomosis and coloanal anastomosis performed after proctectomy are at a higher risk of postoperative anastomotic leakage as compared to intraperitoneal anastomosis [1,2,3,4]. Primary preventive measures for AL include tension-free anastomosis, ensuring perfusion, technical precision, appropriate diversion and patient optimisation. Secondary prevention of chronic pelvic sepsis includes early detection, early endoscopy, early diversion, endoscopic vacuum-assisted surgical correction (EVASC) and active drainage of the acute anastomotic leakage. A diverting stoma has been shown not only to mitigate the consequences of a leak but also to decrease the leakage rate after rectal surgery [5,6,7]. As a result, the majority of surgeons routinely use a diverting stoma and pelvic drainage for mid- and low-rectal resection to reduce the risk and sequelae of leakage [6]. The risks to health and subsequent function from an unprotected anastomotic leak occurring in 10–20% of low anastomosis (below the pelvic floor) are such that the stoma is advocated in all cases. Diverting ileostomy is omitted in carefully selected patients only [5,6,7,8]. A low anastomotic leak is more difficult to diagnose because of the sealing-off by the pelvic inlet and contained leakage. Although it may have mild clinical consequences, particularly in the presence of a defunctioning ileostomy, it can lead to chronic pelvic sepsis [9]. Chronic pelvic sepsis presents sub-clinically from a silent or occult leak, as a missed small sinus when the stoma is reversed, or a reactivation leak after a year [10,11]. Early postoperative AL is observed in 13 to 28% of patients after sphincter-saving proctectomy. This is a risk factor for a permanent stoma highlighted in the study analysing long-term outcomes of the patients included in the Dutch TME trial [12,13]. A Dutch snapshot study demonstrated an acute anastomotic leak rate of 13% but an 8% leak rate after 1 year (i.e., a chronic leak). Pelvic sepsis 30 days after total TME was 17.5% and half of these patients (9.5%) did not heal and developed chronic presacral sinus [14]. The GRECCAR 5 trial [15] showed a 20% overall anastomotic leak rate after 1 year. There was also 8% delayed leakage on long-term follow-up. These suggest that there is no actual effect on anastomotic leak rate in the presence or absence of a stoma. This is corroborated by the fact that anastomotic leak may be radiological (occult), subclinical or clinical [16]. The stoma mostly mitigates the sequelae of an anastomotic leak and in fact delays the diagnosis [7,17,18]. In addition, the residual pelvic inflammation or scarring from a subclinical anastomotic leak may render the ultimate closure of any diverting stoma not technically possible. The optimal timing of ileostomy reversal is generally 3–6 months after initial surgery and should not be delayed even if the patient is receiving adjuvant chemotherapy [7,17]. Further research on quality-of-life outcomes and long-term benefits is necessary to help define which patients are suitable for early stoma reversal. Prior to closure, a water-soluble contrast should document healing of the anastomosis, and a flexible endoscopic examination will ensure that a stricture or stenosis has not formed during the period of diversion [1,2]. A chronic leak may be defined as an issue with the anastomosis after 1 year or fibrosis of the anastomosis after 3 weeks [11,19,20]. Chronic pelvic sepsis is usually defined as any pelvic abscess or fistula tract persisting for 6 months after low CRA or CAA [21]. The incidence of chronic pelvic sepsis is 10% and manifests as a sinus, fibrosis, fistula or a major low anterior resection syndrome (LARS) [11,19]. Low anterior resection syndrome (LARS) is a condition of anorectal dysfunction that occurs frequently following anal sphincter-preserving surgery for rectal cancer and can reduce patients’ quality of life. It includes symptoms such as faecal incontinence, rectal tenesmus (incomplete defaecation or the sensation of incomplete defaecation), obstructed defaecation, changes in stool frequency or consistency, unpredictable bowel function and painful defaecation (dyschezia) [22]. The sepsis spreads locally through the route of least resistance and the symptoms may vary from pain in the hips, back and pelvis to a life-threatening necrotizing soft tissue infection (fasciitis). Central sepsis may occur, especially if there was superimposed radiotherapy, and may disseminate into the sciatic notch, causing sciatica and chronic bone destruction. Chronic pelvic sepsis may also emanate from an abscess at the rectal stump of a low Hartmann’s procedure, which may extend laterally to the hips or centrally fistulate into the vagina [14,20]. Chronic pelvic sepsis and the commonly associated rectovaginal fistula (RVF) have a high risk of leakage after standard redo-CRA or CAA because of the proximity of the anastomosis to the repaired RVF and the surrounding sepsis. The transanal colonic pull-through with delayed coloanal anastomosis (Turnbull–Cutait) procedure would avoid these impediments. It would provide a 79% success rate of restoration of continuity without recurrent chronic sepsis or RVF, acceptable function with low LARS score and the avoidance of a permanent stoma [22]. The majority of patients having a DCAA had a diverting stoma either performed during the DCAA or during a previous procedure. This conflicts with the concept of DCAA, which was initially described as a way to avoid a diverting stoma after rectal surgery for Hirshsprung’s disease in children [23] and later adapted by Cutait et al. for the management of rectal cancer in adults [24]. However, in the presence of chronic pelvic sepsis, and given the risk of septic recurrence in this situation, the protection of a diverting stoma is advocated [25]. Thus, DCAA could be proposed as an alternative to standard redo-CRA or CAA in failed pelvic anastomoses, especially in cases of impaired pelvic conditions such as chronic pelvic sepsis or RVF [21]. For a fibrotic anastomotic stricture without pelvic sepsis, a standard redo-colorectal or coloanal anastomosis has good outcomes [26].

2. Prevention

The combination of anatomical inaccessibility, less-than-optimal blood supply, tightly closed anal sphincters below a low anastomosis and an infected haematoma is likely to be contributory to AL following low anterior resection of the rectum for rectal cancer. Primary and secondary preventive measures for anastomotic leakage have distinct objectives and limitations. The discussion focuses on secondary prevention, i.e., early postoperative detection and management. The judicious use of abdominal drains is important. It may be counterproductive to attempt to drain an anastomosis or the general peritoneal cavity. The adhesions that occur in the healing process of the anastomosis or general peritoneal cavity will attract the peritoneal drain (foreign body), which may physically damage the anastomosis or small bowel [27]. Redivac suction drains below the peritoneal reflection against haematoma formation are deliberately not placed in the vicinity of anastomoses and are removed after 48 h [28,29,30]. A Cochrane review from 2004 showed a lack of benefit of abdominal drains but demonstrated no definite harm [27]. A series of almost 1000 patients failed to demonstrate associations between drains and infectious complications, although there was a marginal association of the use of drains with leaks (14% vs. 8%) [29]. Partly because of the perineal microbial flora, a perineal wound is more likely to be infected via a perineal drain than via a transabdominal suction wound drain following abdominoperineal resection for malignancy [31]. A transanal drainage tube may decrease endoluminal pressure but does not protect against anastomotic leakage in patients without a diverting stoma [32,33].

Chronic pelvic sepsis can be prevented by the early detection and early active management of the acute episode of the anastomotic leak. This entails an early computed tomography (CT) scan with rectal contrast to confirm the leak and the acute phase protein (C-Reactive Protein, CRP) trend [34]. However, despite a CRP cut-off level of 124 mg/dL on day 3/4, there are still missed leaks due to heterogeneity of the underlying pathology (inflammatory bowel disease, cancer, infection, emergency or elective surgery), the site (colon, rectum), type of surgery (minimally invasive or open), neoadjuvant chemoradiotherapy, stoma or no stoma, etc. [1,2]. If there is no leak, there will be a decreasing CRP trend by day 2/3. If there is no stoma, there will be a higher value of CRP on day 2/3. There may, however, be a leak with a defunctioning stoma in situ even with a decreasing CRP demonstrated by rigid video sigmoidoscopy [18,35,36]. A phase II study demonstrating the importance of routine rigid video sigmoidoscopy in the early postoperative period showed a missed leak rate of 10% at 2 weeks post anterior resection. Thus, routine endoscopic imaging is required on postoperative day 4/5 in the ward and at 2/3 weeks in the outpatient clinic as it has been demonstrated that subclinical leaks occur with ileostomies at this time [37]. Early management is also best at this time [18,20]. If there is a leak, an immediate diverting ileostomy is performed laparoscopically if not present. An exploratory laparoscopy is required if there is abdominal contamination and to exclude an internal hernia predisposing the anastomotic leak. Passive management entails the insertion of a transgluteal or transanal (Malecot) drain into the defect. The latter is usually associated with late diagnosis, a large defect and greater symptoms. The problems with passive drainage of an acute leak are firstly, 30% will require redo-drainage because of the difficulty in knowing when to take out the drain, secondly, 10% will require redo-acute surgery, and thirdly, only 15% may have the anastomosis preserved [19,34,38]. The preservation of the anastomosis during leakage management may lead to a very low rate of permanent stoma (<5%) [39,40]. However, despite this anastomosis conservation, postoperative leakage can lead to fibrotic stricture, chronic leakage or RVF [21]. These anastomosis-conserving measures may, in addition, postpone or preclude stoma closure. Active management is a shorter pathway and likely more cost-effective, although formal economic analyses are limited. Following an immediate laparoscopic ileostomy, the sepsis is kept under control with an open-pored polyurethrane endosponge connected to a controlled negative pressure system inserted into the defect. This would rapidly decompress the septic cavity, stimulate granulation and facilitate mucosal approximation within 2–3 weeks following confirmation of healing by a CT scan with rectal contrast [18]. There will be less rectal fibrosis and a better quality of life for the patient. However, with a large defect or an associated fistula, it is wise to additionally insert an 8f drain to evacuate the sepsis rapidly prior to suturing the granulation tissue transanally 2 weeks following an endoscopy [21,40]. If there is a residual small sinus, a redo-endoscopic vacuum therapy is required, and if this fails, a redo-surgery is required [18,20]. For a high anastomotic leak, an endoscopic vacuum-assisted surgical correction (EVASC) stent is utilised to quickly collapse the septic cavity. A pilot endoscopic vacuum therapy (EVT) study without a covering stoma showed some promise [41]. Following granulation, the mucosa is approximated using a high transanal platform. Although active management requires active monitoring in a centralised high-volume unit with the means of rapid intervention when required, the approach avoids the short- and long-term risks of an ileostomy [18]. These include infection, dehydration and electrolyte imbalance predisposing renal failure, obstruction, malnutrition, stoma problems, closure complications and a probable risk factor for LARS [42,43]. It can complicate adjuvant chemotherapy by increasing the risk of severe high ileostomy output, dehydration and deterioration of renal function compromising adherence [42]. Early stoma restitution (<14 days) may be a safe and feasible alternative in highly selective cases after good perioperative counselling and shared decision-making [44]. The innovation of an implantable biodegradable wireless anastomotic sensor for an early anastomotic leak will be invaluable and complimentary [45].

Chronic pelvic sepsis can be substantially reduced if anastomotic integrity is maintained at one year. Although there is still limited adoption and modest evidence, routine video endoscopic imaging in the early postoperative period (day 4) and at 2–3 weeks in the outpatient clinic is of value. Endosponge-assisted early closure of an acute low anastomotic leak by 2–3 weeks has the best outcome in preventing chronic pelvis sepsis. Nonetheless, optimal postoperative management cannot fully eliminate chronic sinuses or delayed reactivation leaks. With chronic leakage, major restorative redo-anastomosis or ablative abdominoperineal resection is required, although 20% will require a permanent stoma.