Article Text

Download PDFPDF

Colonic embolisation: useful but caution required
  1. Department of Imaging,
  2. Hammersmith Hospitals NHS Trust,
  3. Hammersmith Hospital,
  4. Du Cane Road,
  5. London W12 0HS, UK

Statistics from

See article on page79

The value of diagnostic visceral angiography in the investigation and management of acute and chronic gastrointestinal bleeding is well established, its use in this role having been first described over 35 years ago.1 It is possible to identify the source of haemorrhage in 87% of patients actively bleeding at the time of the study, and to identify a lesion which could be responsible for the problem in 74% of those not bleeding at the time of investigation.2 These figures are dependent on a number of factors, including the use of meticulous angiographic technique, the quality of the imaging equipment and the experience of the operator in both the acquisition and interpretation of data.3

In a patient thought to be actively bleeding at the time of the angiogram, the vessel most likely to be responsible is examined first—for example, the inferior mesenteric artery (IMA) in rectal bleeding, the coeliac axis in haematemesis. If a source is not found then the other major visceral vessels are studied in turn. If the superior mesenteric artery (SMA) and IMA studies appear normal in a case of bleeding thought clinically to be of lower gastrointestinal origin, the coeliac axis should also be studied, not only because there may be aberrant vascular anatomy (the coeliac axis and its tributaries can give rise to middle colic and jejunal arteries), but also because bleeding from the upper gastrointestinal tract may present in an unusual way and should be excluded.1 ,4 Portal hypertension, for instance, may be best revealed on a coeliac study. In the patient not actively bleeding it is vital to examine all three major vessels, the coeliac axis, SMA and IMA, with selective and superselective studies being performed as appropriate.

Therapeutic vascular embolisation in the management of gastrointestinal bleeding was first described by Bookstein et al in 1974.5 They used autologous blood clot in nine cases, controlling the bleeding in the three patients with haemorrhage from the lower gastrointestinal tract. Embolisation procedures require the operator to achieve superselective catheterisation of the vessel or vessels involved in the pathological process to maximise the efficacy of the manoeuvre and minimise the risk of complications, such as ischaemic necrosis or the inadvertent embolisation of neighbouring structures. All potential routes of collateral blood supply need to be examined and embolised where necessary. Advances in catheter and guidewire technology (e.g. effective co-axial catheter systems and hydrophilic guidewires) have made it increasingly possible to manipulate catheters into small and tortuous vessels and this has greatly increased the safety and efficacy of the method. Embolic agents have also been refined and minute embolisation coils are available which can be delivered through very fine coaxial catheter systems.6 These advances have revived interest in the use of embolisation in the lower gastrointestinal tract,7 ,8 formerly an area in which the technique was regarded as hazardous owing to the relative paucity of the collateral blood supply in some areas.

Rosenkrantz et al in 19829described three cases of colonic mucosal necrosis in a total of 23 patients embolised and this incidence of serious complications has deterred many radiologists from using the technique in the lower bowel. Their paper and subsequent publications deserve greater scrutiny. It is important to look at the figures in the context of the underlying clinical condition and in comparison with the success and complications associated with alternative treatment—for example, vasopressin infusion, or emergency partial or total colectomy.

In their review of the literature, Guy and colleagues7found reported morbidity and mortality rates for emergency hemicolectomy of 50% and 30%, respectively. Vasopressin infusion achieved haemostasis in up to 90% of cases, but with a rebleeding rate of 20% and an associated complication rate as high as 43%. Complications included myocardial infarction, cardiac arrhythmia, cardiorespiratory arrest, and visceral infarction. In patients treated by therapeutic embolisation, colonic infarction and ischaemia was reported in up to 20% of cases. In their own series, nine patients underwent successful embolisation. There were no cases of procedurally related colonic infarction and only two cases of endoscopically diagnosed asymptomatic mucosal ischaemia. They attributed their success to advances in catheter and guidewire technology enabling highly selective catheterisation thus minimising the area of tissue embolised.

In a recently published series, Gordon and colleagues8reviewed 17 patients considered for embolisation. In three cases they did not achieve a highly selective catheter position and therefore did not proceed. In 13 (93%) of the remaining 14 cases haemostasis was achieved. There were no episodes of recurrent bleeding and no evidence of bowel ischaemia or stricture formation either endoscopically or clinically on follow up (mean 10 months). In all cases microcoils were delivered through fine coaxial catheters. The authors emphasised the need for highly selective catherisation, but advised against the use of fine particulate emboli which may occlude mucosal capillaries.

The series reported in this issue (see page 79) by Nicholsonet al adds further weight to the growing evidence that with meticulous technique and modern equipment, embolisation is an effective treatment in acute lower gastrointestinal haemorrhage with a comparatively low major complication rate.

It is important to emphasise a number of issues which affect the safety and success of embolisation procedures. Clinical liaison is paramount. Careful patient selection is vital for embolisation, the technique being reserved for those subjects in whom conservative management has failed, and emergency surgery is contemplated. A high order of technical expertise is essential and an appropriate range of catheters, guidewires and embolic agents should be readily available. Adequate follow up is essential not only to detect complications such as stricture development, but also to determine the underlying pathology responsible for the haemorrhage if this is not already known. Embolisation is not the treatment of choice for an operable colonic carcinoma!

Finally, embolisation may in many cases represent definitive treatment, but in others may usefully convert an emergency colectomy carrying a high risk in a desperately sick patient to an elective, more limited, resection in a stable patient. The inclusion of vascular intervention from 1 April 1998 in NCEPOD will bring all vascular interventional procedures under close scrutiny, a development welcomed by active interventionists who wish to maintain and raise the standards of their discipline.


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles