Ischaemic colitis is an inflammatory condition produced by interruption of the blood supply to the colon insufficient to cause full thickness tissue death. It most commonly affects those in the sixth to the eighth decades of life and is thus being seen with increasing frequency in our progressively elderly population.

Ischaemic colitis may be caused by occlusion of a major artery, small vessel disease, venous obstruction, ‘low flow’ states, or intestinal obstruction (Fig. 1) 1064. In each case, the mucosa and sub-mucosa are predominantly affected, the extent of injury being determined by the severity and longevity of the insult. Ischaemia reduces the integrity of the mucosa and allows invasion by pathogenic organisms such as clostridia, which are normal constituents of colonic flora. These processes produce inflammation and mucosal ulceration which may resolve completely. Alternatively the insult can result in permanent injury with healing by fibrosis and subsequent stricture formation. Rarely, necrotizing colitis develops, which can spread to affect areas of the colon which are not ischaemic.

Although any part of the colon can be affected, the splenic flexure is particularly susceptible to ischaemic injury because it is the site of the watershed between the superior mesenteric artery, supplying the transverse colon, and the inferior mesenteric artery which supplies the descending colon. These vessels are linked by a marginal artery, but this is frequently absent or poorly developed at the splenic flexure. Occlusion of either major artery or their feeding branches (middle colic artery from the superior mesenteric artery and left colic artery from the inferior mesenteric artery) can therefore result in ischaemia. This point is of particular relevance during aortic and colorectal surgery if the inferior mesenteric artery is ligated. During aortic surgery it is important to confirm pulsatile flow in the superior mesenteric and marginal arteries prior to ligating a patent inferior mesenteric artery. If this is absent or if doubt exists then the inferior mesenteric artery should be reimplanted into the graft.

A typical patient is 50 years of age or more and complains of left-sided abdominal pain which is acute in onset and started in the left iliac fossa. Loose stools, which characteristically contain dark blood as well as clots, may be passed. There may be a history of previous similar episodes, or of peripheral or cardiovascular disease, or collagen vascular disease, especially if the symptoms are atypical.

As ischaemic colitis is predominantly a disorder of colonic mucosa and submucosa it is not usually associated with a major systemic upset, but a low grade pyrexia and tachycardia should be expected. On abdominal examination, the affected colon is tender and may be palpable. Dark blood will be present per rectum. Signs of peripheral vascular disease or other associated conditions should be sought.

It is important to first establish the diagnosis and then determine the presence of any treatable aetiological factors.

A plain abdominal radiograph and a contrast enema are the most useful investigations in the initial stages of this disorder. ‘Thumb-printing’ is diagnostic and is most often seen at the splenic flexure (Fig. 2) 1065. It is present at an early stage (from 3 days) and is the result of submucosal oedema and haemorrhage which produce swellings that project into the bowel lumen. These are clearly seen in contrast studies.

Later, mucosal ulceration and irregularity may develop and these can resemble the appearances of ulcerative colitis or Crohn's disease. However, ulcerative colitis invariably affects the rectum and there is loss of the normal colonic haustral pattern while in Crohn's disease deep ulcers resemble ‘rose thorns’ and areas of affected colon are separated by normal bowel. These features are not seen in ischaemic colitis.

Although many of the features of ischaemic colitis are reversible, stricture formation, if it occurs, is not and causes further diagnostic problems. Ischaemic strictures are often long, uniform and have smooth, gradual beginnings and ends, an appearance called ‘funnelling’. However, these findings do not exclude carcinoma; this diagnosis should be considered, particularly if only a short segment of colon is affected. The role of angiography is not established. Although it can be valuable in isolated cases where significant, symptomatic occlusive lesions are revealed, there is generally no correlation between the appearance of vessels at angiography and the integrity of the colonic blood supply.

Ischaemic lesions are usually beyond the reach of the rigid sigmoidoscope, but colonoscopy can be used to visualize and biopsy affected colon. In the early stages of ischaemia, the mucosa will be heaped up, oedematous, and bluish purple (the ‘thumb-prints’ seen radiologically). It will bleed on contact with the endoscope or other instruments. Later, ulceration as well as strictures may be seen.

It should be noted that some of these diagnoses, for example carcinoma, are also possible aetiological factors for ischaemic colitis.

This will be determined by the mode of presentation, which in turn reflects the underlying stage of the ischaemic process. Conservative management is the mainstay of treatment for those seen with acute symptoms. The patient is rested in bed and given intravenous fluids. Broad-spectrum antibiotics are often administered, although there is no conclusive evidence to suggest that they influence outcome. There is no place for anticoagulation or steroid administration unless this is indicated by an underlying disorder such as vasculitis.

It is very rare for ischaemic colitis to progress to frank colonic gangrene, but all patients should be monitored frequently to assess progress. If the injury is transient then resolution occurs after a few days to a week. More severe insults lead to stricture formation. These require investigation and treatment if they produce symptoms or if there is diagnostic doubt. Excision followed by end-to-end anastomosis is safe, although it is essential to ensure the viability and vascularity of the resection margins. If malignancy is excluded, then the resection can be limited to the affected segment but a more radical excision should be performed if there is continuing diagnostic uncertainty.

Marston A. Vascular disease of the colon. In: Marston A, ed. Vascular Disease of the Gut. London: Edward Arnold, 1986: 158–70.