Obstruction of the colon as a separate clinical entity is singled out for discussion for two reasons. Firstly, it brings its own serious complications (the most dramatic being perforation), which are not directly related to the underlying cause. Secondly, it introduces special elements that influence the timing and limit the selection of procedures in managing the actual cause of obstruction.

Obstruction of the intestine at any level leads to progressive distension of the proximal bowel with fluid and gas. In the small intestine, distension may result in early vomiting if the site is proximal or in the sequestration of large amounts of fluid if it is more distal. Neither distension nor sequestration is an important elements in the presentation and course of obstruction of the colon.

The obstructed colon has less peristaltic activity, and dilatation tends to cause fewer and less severe cramps than small bowel obstruction. In the majority of patients the ileocaecal valve prevents decompression of the colon into the small intestine. This situation results in closed loop obstruction, which is especially hazardous because it can result in massive dilatation of a segment of the colon without the signs and symptoms usually associated with intestinal obstruction. Secondary dilatation of the small intestine with its consequences may follow, but it is slow to develop and is usually unimportant.

Colon obstruction presents as an emergency, especially because of the associated complication of perforation. Even when intraluminal pressure is constant, the tension on a segment of colon wall increases with an increase in diameter; the relatively larger diameter and thinner wall of the caecum and ascending colon make these the most frequent sites of perforation. In the final stages before perforation there is an embarrassment of circulation, which can also contribute to rupture of the wall. The site of perforation may be in a more distal segment of the colon, usually just proximal to an obstructing lesion such as a carcinoma.

During barium studies under the fluoroscope the normal ascending colon dilates readily up to only about 9 cm in diameter. In chronic obstruction, much larger degrees of distension may nevertheless be quite harmless. Still, when plain radiographs show distension of the ascending colon beyond 10 cm, the possibility of rupture must be considered as imminent. If the capacity of the right colon is exceeded, the serosa between the taeniae splits, perforations, usually punctuate, appear in the mucosa, and patches of haemorrhage and even infarction are also commonly observed. Perforations of this type initially lead to progressive contamination of the peritoneal cavity, but may not actually decompress the obstructed colon.

The clinical course of colon obstruction is often quite insidious. Up to 80 per cent of cases are caused by a carcinoma. Many of these patients will have experienced prior episodes of partial obstruction and may not be aware that complete obstruction has developed until 2 or 3 days have passed. Although lower abdominal cramps are usually reported, these may be mild. Cessation of the passage of stool or gas is the most frequent complaint, and an observant patient will also report abdominal distension. Physical examination shows a quiet abdomen with obvious tympany and distension, sometimes most pronounced over the ascending colon. There may be tenderness over compromised bowel, the caecum or the area proximal to a tumour. Rectal examination may disclose a low rectal tumour as the cause of obstruction. More commonly, the rectum feels capacious and empty. In the presence of a tumour or compromised bowel, there may be traces of blood. The physical examination should include a careful search for abdominal wall hernias, especially in any old surgical incisions.

Plain radiographs of the abdomen are usually helpful in colon obstruction. They may confirm or support the diagnosis and point to the site and aetiology; they can also suggest the presence of perforation by showing free air and can demonstrate a dangerously distended right colon. If an expeditious barium study is obtained, it will help identify the actual site and nature of obstruction (Fig. 1) 1101. The barium enema examination should be done without preparation and care must be taken not to push barium above the obstruction.

In practice, it is difficult to assess the likelihood of colon perforation. When obstruction is partial, sequential evaluation of symptoms and the presence and degree of tenderness over the right colon can suggest impending perforation. If obstruction is complete, and especially if decompression into the small intestine is absent, urgent surgical exploration with decompression is the first consideration.

In the more economically developed areas of the world, carcinoma is by far the most frequent cause of colon obstruction. Carcinomas of the left colon are the most likely to lead to obstruction, although those in other areas also cause obstruction, the least frequent being those of the rectum. Most patients in whom the obstruction involves the distal colon have a presentation that includes an initial period of symptoms of partial obstruction. Final obstruction reflects obturation of the narrow residual channel by faecal material. Obstruction in the right or transverse colon is more likely to occur without preceding symptoms, perhaps because of the liquid nature of the stool.

The prognosis of treated obstructing carcinomas is not very different from that of non-obstructing lesions of similar location and extent. Prognosis may be estimated by use of the Dukes' or other classification. The early operative risk is much higher in these patients especially with respect to perforation.

Complete acute colon obstruction is rarely caused by diverticular disease, although it sometimes occurs due to obturation of a strictured oedematous segment. Incomplete obstruction is more usual, and there may be associated partial or complete obstruction of an adherent loop of small intestine.

Differentiation between a carcinoma and diverticular disease of the sigmoid may be very difficult and can lead to an incomplete or delayed resection in a patient incorrectly diagnosed as having diverticulitis.

Volvulus of either the ileum, caecum and right colon or of the sigmoid colon almost always leads to acute obstruction. The twist of 180° or more around the mesenteric axis imparts a similar twist in the longitudinal axis; this is the immediate cause of obstruction. Because vascular compromise may result, there is a danger of perforation in neglected cases. Expert interpretation of the plain radiograph will usually establish the diagnosis. Contrast enemas may be needed when there is considerable dilatation of the proximal colon or small intestine (Fig. 2) 1102. Volvulus of the right colon generally reflects a failure of fixation of the colon. Sigmoid volvulus is probably an acquired abnormality but is usually observed in patients with an elongated colon. Spontaneous reduction of all forms of volvulus will often occur, relieving obstruction; recurrence is the rule.

Only a few cases of colon obstruction are due to causes other than cancer diverticulitis, and volvulus. The uncommon causes include incarceration in hernias, strictures such as those that follow vascular compromise, metastatic malignant tumours, especially those of pelvic origin, radiation strictures, and various inflammatory conditions. Adhesions, although a common cause of small bowel obstruction, are a rare cause of colon obstruction.

Management should be directed at relief of obstruction before perforation occurs. Secondary objectives include treatment of the underlying cause and restoration of the continuity of the colon with the smallest number of operations in the shortest time feasible. Choices between the various surgical alternatives reflect the urgency of the situation, the condition of the patient, the state of the bowel to be used for anastomosis, and the potential for cure of a malignancy. Impending or actual rupture of the right colon is another major modifying factor. As with anastomosis in elective resections, tension, impaired circulation, and local contamination may all dictate a staged procedure. An anastomosis between distended ileum and normal colon is usually safe: one between distended and normal colon much less so.

In the treatment of obstructing tumours in the caecum, ascending or right transverse colon, a right colectomy extended across to the transverse colon as needed is the usual procedure. Even if the colon is distended, an adequate dissection for cure is feasible, and an anastomosis of the ileum to normal and usually empty collapsed distal colon is safe. If the caecum is perforated or when other factors make the risk exceptionally high, a resection with ileostomy carries less immediate risk, although a second stage is required for reanastomosis.

If the obstructing lesion is in the region of the splenic flexure and down to the mid-sigmoid colon, a right colectomy with extension to include the lesion (a subtotal colectomy) with ileocolonic anastomosis may still be selected. In the young and healthy patient seen by an experienced surgeon subtotal colectomy is an acceptably low risk and troublesome diarrhoea afterward is the exception. A segmental resection with direct anastomosis using distended and unprepped proximal bowel carries a high risk. The safe procedure is to perform a resection with an end colostomy and with reanastomosis as a second stage. A proximal transverse colostomy or a caecectomy and ileostomy for caecal perforation with staged later section may be dictated by the general or local operating conditions. Recent reports suggest that antegrade intraoperative irrigation may be used in selected cases to establish safe conditions for a direct anastomosis.

When a tumour is present in the distal sigmoid colon or upper rectum, a single stage operation is seldom feasible. If a satisfactory dissection can be carried out under the operating conditions, a resection with an end proximal colostomy should be done. Otherwise, a proximal defunctioning colostomy alone is done. If the defunctioning colostomy can be carried out in the sigmoid colon rather than in the transverse colon, a single subsequent operation may be possible: many surgeons consider it unsafe to have two separate colon suture lines because of an increased risk of postoperative leakage. A prior colostomy remote from the operative anastomoses makes a third stage necessary.

When the obstructing tumour is at or below the peritoneal reflection, fortunately an unusual site for obstruction, a two- or three-stage operation is generally the best choice because a satisfactory dissection for the resection of a malignant tumour is often neither feasible nor safe under these conditions. If hepatic or other unresectable metastases are present, however, a Hartmann procedure (resection of the tumour with closure of the rectum and an end sigmoid colostomy) is a good single stage choice. Alternatively, a proximal transverse colostomy must be performed as a first stage, with resection of the tumour deferred until proper conditions can be secured.

In the presence of colonic obstruction there is always a risk when the caecum is not directly examined: severe damage to the wall caused by distension can result in perforation occurring even after distal decompression. Most operations should therefore include visualization of the caecum. Some minor transmural injuries can be managed by imbrication or by the placement of a caecostomy tube. Right colectomy is necessary with more advanced injuries or perforation. The site of colon obstruction and its management and the general condition of the patient determine whether an ileostomy or an ileocolonic anastomosis is indicated.

Complete obstruction from diverticulitis is, in fact, rare. In most cases, obstruction is incomplete, and an element of inflammation and oedema will respond to antibiotics and to ‘putting the colon at rest’. When non-surgical treatments fail, a primary resection can almost always be performed safely. An extensive dissection of mesenteric nodes is not necessary.

Conditions for a primary unprotected colonic or colorectal anastomosis often cannot be met. A safe colon anastomosis is one that can be left in a clean, normal field: there must be no gross faecal contamination, nor any exudate and purulent material from an abscess. Most cases of obstructive diverticulitis are best managed by primary resection and end colostomy. Continuity is restored at a second operation several weeks later. When there is a large abscess or the local inflammatory reaction precludes a safe dissection, a defunctioning transverse colostomy with drainage of the abscess as a first stage may be the only safe alternative. Another alternative in selected patients is percutaneous drainage of the abscess, with later resection if obstruction resolves.

Obstruction due to volvulus of the ileocaecal area requires prompt diagnosis and management in order to avoid the otherwise common complication of perforation. The diagnosis can be made on plain radiograph, but a contrast study is usually obtained. Once the diagnosis is established, an urgent laparotomy is needed. Although colonoscopic decompression or reduction by a column of barium may be feasible, the possibility of mural compromise and the likelihood of recurrence make resection with primary anastomosis the favoured management in almost all cases. In the presence of perforation with extensive peritoneal contamination or when other factors dictate a more expeditious procedure, resection with ileostomy and later reanastomosis may be the best choice.

When the caecum is intact and undamaged, caecopexy offers an alternative to resection. A long segment of the ascending colon should be secured to the left abdominal wall by raising a peritoneal flap and forming a retroperitoneal track for the colon, fixing the cut edge of the peritoneum to the anterior aspect of the colon with multiple sutures.

The management of sigmoid volvulus is in some ways quite different from that of a caecal volvulus. Perforation is rare and decompression through a rigid sigmoidoscope, perhaps by inserting a well lubricated tube through the lumen, is easily accomplished in most patients. The tube can be left in place. In chronically debilitated patients, simple decompression by this technique, when needed, may be the best choice.

If decompression is successful and definitive treatment is selected, the procedure would be an elective sigmoid resection after routine bowel preparation. An unprotected primary resection and anastomosis can usually be undertaken safely. Fixation of the sigmoid colon without resection is usually not successful because of marked redundancy and the adjacent location of the colon segments in the proximal and distal ends of the volvulus.

When decompression fails or when there is evidence of infarction or perforation, management is difficult. Reduction of the loop at operation can result in perforation and massive faecal contamination of the peritoneal cavity. If a tube is placed in the rectum before surgery, it can sometimes be manipulated into the loop by the surgeon for decompression. Primary anastomosis is seldom feasible after emergency resection. Most surgeons perform an end sigmoid colostomy with staged restoration of colon continuity. An alternative in the absence of perforation or of bowel infarction would be intraoperative decompression with the rectal tube, followed by an elective resection after proper preparation of the colon and patient.

Volvulus of the transverse colon is rare, and principles of management are the same as for caecal volvulus.

Numerous other anatomical and pathological causes of acute colon obstruction are encountered (Table 1) 346. Although certain principles for their management can be re-emphasized, the surgeon is left to analyse and resolve each as an individual problem. Whatever the primary site of obstruction, the possibility of caecal perforation should be considered; failure to visualize the caecum always carry a risk. Healing of any anastomosis is uncertain when it cannot be placed in a normal, uncontaminated area of the peritoneal cavity. Even a diverting proximal colostomy will not ensure healing of a questionable anastomosis. The additive risks of staged operations favour a one-stage procedure, but only under near optimal conditions of anastomosis. Finally, bypassing an obstructed segment of colon does not always preclude the risk of subsequent perforation, especially when an ileocolonic bypass is used.

Albers JH, Smith LL, Carter R. Perforation of the cecum. Ann Surg 1956; 143: 251–5.

Ballantyne GH, Brandner MD, Beart RW Jr, Ilstrup DM. Volvulus of the colon. Incidence and mortality. Ann Surg 1985; 202: 83–92.

Howard RS, Catto J. Cecal volvulus. A case for nonresectional therapy. Arch Surg 1980; 115: 273–7.

Koruth NM, Krukowski ZH, Younger GG, Hendry WS, Logir JR, Jones PF. Intra-operative colonic irrigation in the management of left-sided large bowel emergencies. Br J Surg 1985; 72: 708–11.

Morgan JP, Jenkins N, Lewis P, Aubrey D. Management of obstructing carcinoma of the left colon by extended right hemicolectomy. Am J Surg 1985; 149: 327–9.

Saini S, Mueller PR, Wittenberg J, Butch RS, Rodkey GV, Welch CE. Percutaneous drainage of diverticular abscess. An adjunct to surgical therapy. Arch Surg 1986; 121: 475–8.

Savage PT. Immediate resection with an end-to-end anastomosis for carcinoma of the large bowel presenting with acute obstruction. Proc R Soc Med 1967; 60: 207.

Wolmark N, Wieand HS, Rockette HE, Fisher B, Glass A, Lawrence W. The prognostic significance of tumor location and bowel obstruction in Dukes B and C colorectal cancer. Findings from the NSABP clinical trials. Ann Surg 1983; 198: 743–52.

Wright HK. The functional consequences of colectomy. Am J Surg 1975; 130: 532–4.