When Do I Defunction?

John Hobbis

Powerpoint File

INTRODUCTION

Anastomotic breakdown following colorectal resection may have such disastrous consequences that we must do everything we can to prevent it. Every time we resect a carcinoma of the left colon or rectum we have to ask ourselves, “Is it safe to do an anastomosis?”. If yes, then “Should the anastomosis be protected with a proximal stoma?”
Anastomoses for these cancers are constructed at one of three levels. A “high” rectal anastomosis, used for left sided colonic cancers, is taken to the top of the rectum.

The anastomosis is intraperitoneal and the rectum is supplied by the middle and inferior rectal arteries. A “low” rectal anastomosis is used for cancers involving the upper or mid rectum. The resection includes a total mesorectal excision and is taken to the distal rectum, usually about 5-7cms from the Dentate Line. The middle rectal arteries are divided and the rectal stump is supplied by the inferior rectal arteries only. An “ultra low” rectal anastomosis, constructed at or just above the ano-rectal junction, may be used for cancers of the lower rectum. The lower the anastomosis, the higher the risk of leakage.

INCIDENCE OF LEAK

The potential for leakage of colonic and colorectal anastomoses was demonstrated by a study by Fielding, published in 1980. The outcome of 1466 patients from multiple centres throughout the United Kingdom, who underwent a resection for colonic or rectal cancer, was prospectively audited. There was an overall leak rate of 13% (10.8% for anastomoses within the peritoneal cavity and 18.7% for those below the peritoneal reflection) (1). This series were used as evidence for the need for subspecialisation within general surgery.
More modern series have reported lower leak rates. Two important multicentre audits, conducted in the United Kingdom in the 1990s, were the Trent / Wales and the Wessex audits. In the Trent / Wales audit the overall leak rate was 4.9%, (7.4% for anterior resections and 3.7% for other colonic resections). In the Wessex audit the overall leak rate was 3.4% (6.9% for anterior resections and 2.6% for others). In both series the mortality following a leak was about 20%. As a result of these audits, the “Guidelines for the Management of Colorectal Cancer (2001)”, issued by the Association of Coloproctology of Great Britain and Ireland, gave the following recommendation: “Surgeons should carefully audit their leak rate for colorectal surgery and should expect to achieve an overall leak rate of below 8% for anterior resections and below 4% for other types of resection” (2).

CAUSE OF LEAKS

An anastomosis may leak because of a technical fault in its construction or because of a failure to heal. Technically sound anastomoses can be constructed either by hand suture or staples but it is harder to construct a sound hand sutured anastomosis deep in the pelvis. An air leak demonstrated by inflating the anastomosis after construction indicates an increased likelihood of a clinical leak and should be corrected with sutures (3).
A healthy left colonic pedicle that reaches the rectum without tension is essential for a successful colorectal anastomosis. In many patients the left colon will only reach the lower rectum if the inferior mesenteric artery is ligated above the origin of the left colic artery. This means that the pedicle receives its blood supply from the middle colic artery via the marginal artery. In some patients the marginal artery is deficient around the Splenic Flexure (4). Great care must be taken not to damage the marginal artery or any vessels in the mesentery when mobilising the left colon and splenic flexure. Selecting the optimum site on the left colonic pedicle to use for the anastomosis is crucial. The more distal the less tension but the more proximal the better vascularity.
The most important factor in the subsequent healing of the anastomosis is oxygenation (5). Any morbidity that reduces the perfusion of the anastomosis or causes hypoxaemia will increase the risk of leakage. Ischaemic heart disease, acute and chronic respiratory diseases, diabetes, old age, coexisting sepsis, perioperative blood transfusion, previous irradiation and technical problems with the anastomosis have all been associated with leakage (6,7).

WHAT DOES A PROXIMAL STOMA ACHIEVE?

A proximal stoma probably does not prevent leakage and certainly will not prevent the breakdown of a poorly constructed or poorly perfused anastomosis. It does reduce the clinical effects of a leak by reducing contamination and thus a higher proportion of leaks remain at the subclinical level. It also means that a leak is more likely to be managed satisfactorily without further surgery. However, severe life threatening sepsis due to anastomotic breakdown can occur, even in the presence of a proximal stoma.
The complications associated with the construction of a proximal stoma and its closure must be taken into account when deciding whether to defunction an anastomosis. Wexner reported closure site leakage in 2 patients out of 67 (3%) undergoing loop ileostomy closure and an overall complication rate of 10.8% (8). In my own series, 3 patients out of 80 (4.5%), developed significant intraperitoneal sepsis after closure of their loop ileostomy (9).

WHAT DO I DO?

In the following circumstances I regard it as unsafe to construct an anastomosis and perform Hartmann’s operation or abdominoperineal excision, leaving the patient with a permanent stoma.

Dubious blood supply to the left colonic pedicle
Comorbidity that gives a significant risk of hypoxia or hypo-perfusion in the post-operative period.
Pre-existing pelvic sepsis
Residual pelvic tumour following resection.
In more favourable circumstances I construct an anastomosis. The decision whether or not to add a loop ileostomy depends partly on the level of the anastomosis.

High rectal anastomosis
If the anastomosis is technically sound and air tight and the patient is otherwise healthy, I would not perform a proximal stoma. If the patient has moderate comorbidity (including heavy smoking) that I think may delay healing, then I perform a loop ileostomy.

Low rectal anastomosis
The relatively high risk of leakage and its disastrous consequences means that most of my patients having a low rectal anastomosis are given a loop ileostomy. For me to make an exception to this rule the patient must have no comorbidity (ASA 1), be a non smoker, the left colonic pedicle must be long enough and healthy enough to flop into the pelvis so that there is no dead space behind the neorectum and the patient must not have had pre-operative radiotherapy (T0,T1,T2 tumours only in our practice).

Ultra-low rectal anastomosis
All my patients who have an ultra-low anastomosis are given a proximal stoma and there are no exceptions. The dilemma in these patients is likely to be whether they should have an anastomosis at all. Just because it is technically possible to achieve satisfactory tumour clearance and do an anastomosis, doesn’t mean that it is necessarily the best thing to do. An anastomosis, even a defunctioned one, creates the possibility of septic complications. It requires a second operation to close the stoma, which again involves more risk. The functional result of a coloanal anastomosis may be poor. Many people cope better with a well-constructed colostomy than they would with the problems of urgency and incontinence that can occur with an ultra-low anastomosis. I am likely to advise an elderly, unfit or immobile patient to undergo an abdominoperineal excision for their low rectal cancer, rather than risk the potential complications and poor function of an ultra-low anterior resection.

POSTOPERATIVE CARE

Preventing anastomotic leakage involves high quality technical surgery and high quality decision making. At the end of the operation there still remains a very important aspect of leakage prevention; postoperative care. It is vital to prevent hypoxia and mesenteric hypoperfusion in the first 48 hours after surgery. This requires close teamworking with the anaesthetist and careful monitoring in a unit staffed with adequate numbers of adequately trained nurses.

References

1. Fielding LP, Stuart Brown S, Blesovski L, Kearney G, (1980). Anastomotic integrity after operations for large bowel cancer – a multi-centre study. BMJ 281:411-414.
2. Guidelines for the management of colorectal cancer (2001). www.acpgbi.org.uk
3. Beard JD, Nicholson ML, Sayers RD, et al, (1990).Intra-operative air testing of colorectal anastomoses: a prospective, randomised trial BRJ Surg; 77:1095-1097.
4. Griffiths JD, (1956). Surgical anatomy of the blood supply of the distal colon. Ann R Coll Surg Engl, 4; pages 241-256
5. Shandall A, Lawndes R, Young HL, (1985). Colon anastomotic healing and oxygen tension. BRJ Surg; 72: 606-609.
6. Irvin TT, Goligher JC, (1973). Aetiology of disruption of intestinal anastomoses. BRJ Surg; 60:461-464.
7. Alves A, Hannis Y, et al, (2002). Factors associated with clinically significant anastomotic leakage after large bowel resection: multivariate analysis of 707 patients. World J Surg; 26: 499-502
8. Wexner SD, Taranow TA, Johansen OB, et al, (1993). Loop ileostomy is a safe option for fecal diversion. Dis colon rectum; 36, 4: 349-354.
9. Botros MM, Hobbiss JH, (2005) Loop ileostomies. Do they prevent or promote sepsis? Colorectal Disease; 5(suppl 2), 31.