Massive lower gastrointestinal bleeding treated successfully with transcatheter arterial embolisation

Helena M O'Dwyer, Mark Logan, Michael J Lee

Key Words: GI Bleeding, Angiography , Embolisation,

Abstract

Lower gastrointestinal bleeding is a frequent cause of hospital admissions. Its diagnosis is difficult, often requiring endoscopy, angiography and isotope scans.

Advances in radiological equipment and techniques mean percutaneous embolotherapy, previously considered a high-risk procedure in the colon, may provide an alternative treatment option to surgery.

Introduction

Lower gastrointestinal bleeding, defined as bleeding distal to the ligament of Treitz, can be catastrophic and is considered a surgical emergency. Red blood cell scintigraphy or selective angiography may aid diagnosis, when endoscopy is inconclusive. Embolisation has rarely been performed until recently because of fears of intestinal ischaemia and infarction.

Recently reports have appeared documenting successful embolisation for colonic bleeding. The availability of microcatheters for highly selective arteriography has made this procedure safer. We present a high-risk patient with acute lower gastrointestinal bleeding, deemed a non surgical candidate, who was treated successfully with embolisation.

Case Report

A 65 year old female presented to the emergency department with severe diarrhoea, dehydration, and confusion. Laboratory tests revealed she was septic, with deteriorating renal function and a metabolic acidosis. This patient had significant co-morbidity with a diagnosis of Crohns disease for which she had a right hemicolectomy 22 years previously. Her cardiac history included a myocardial infarct 5 years previously, mixed aortic valve disease, hypertension and rheumatic fever in childhood. She also had chronic renal failure with proteinuria and a diagnosis of renal amyloid made on biopsy 13 years previously. She was treated for TB in 1958.

Initial radiological investigations, including a CT Brain and Abdomen, revealed cerebral vasculopathic changes and free fluid within the abdomen and pelvis. A lumbar puncture revealed no abnormality. In the absence of CT evidence of perforation or intra abdominal collection, mesenteric ischaemia was suspected. The patient had gram-negative septicaemia, remained acidotic with a tender abdomen and was transferred to Intensive care for optimisation, intubation and haemodialysis, prior to exploratory laparotomy. At laparotomy there were no features of active Crohns disease. The duodenum, small bowel, anastomosis, left colon and rectum were normal with no evidence of ischaemia. The intra abdominal organs, including uterus, ovaries and tubes, were unremarkable. No abnormality was seen in the lesser sac and clear fluid in the peritoneal cavity was negative on gram stain.

On Day 4 of admission, she developed severe bleeding per rectum. Gastroscopy revealed no abnormality and dark red blood with minimal clots but no active bleeding point was identified at colonoscopy. Over the next 2 days, total resuscitation included 21 units of packed RBCs, 21 units of fresh frozen plasma and 6 units of platelets.

A diagnostic angiogram was performed on day 6 of admission. A standard 5-Fr angiographic catheter (Simmons, TerumoTM, Leuven, Belgium) was used for the initial diagnostic angiogram, which identified a subtle focal, frank bleeding point in the medial aspect of the superior mesenteric artery territory at the expected location of an anastomosis secondary to prior hemicolectomy (Figure 1).

Figure 1 A focal, frank bleeding point is identified in the medial aspect of the superior mesenteric artery territory at the expected location of the previous iliocolic anastomosis (arrow).

Figure 2 After superselective cannulation of the abnormal area with a microcatheter microcoils (arrow) were used to exclude the abnormal area. Repeat angiography shows no filling of the abnormality.

This area was embolised using a 2.7F microcatheter (Tracker-18, Boston-Scientific, Cork, Ireland) placed just proximal to the site of bleeding, and five 2 x 20mm microcoils (MReyeTM embolization coils, Cook, Leuven, Belgium) were placed to exclude the abnormal area. The patient was monitored for symptoms and signs of intestinal ischaemia or infarction following the procedure (abdominal pain/tenderness, peritoneal signs, fever, nausea). She continued to remain septic with diarrhoea, dependent on haemodialysis. There was no further bleeding. All cultures remained negative. Unfortunately, the patient died 17 days post admission from multiorgan failure.

Discussion

A lower gastrointestinal bleed is defined as bleeding occurring anywhere from the ligament of Treitz to the anus (1,2). Embolisation for treatment of upper GI bleeding is a well recognised procedure, however until recently surgical resection remained the standard treatment in lower GI haemorrhage (2). The reported mortality following emergency colonic resection for bleeding ranges from 10 to 36% (3). The availability of superselective arteriography has made embolisation safer and a therapeutic choice particularily in poor surgical candidates.

The causes of lower GI bleeding are listed in Table 1.

Table 1

Some lesions may be more suitable for embolotherapy. Diverticular bleeding usually has a single feeding vessel and therefore easier to control with a superselective approach. However, arteriovenous malformations have multiple vessels requiring embolisation and therefore more prone to rebleeding (2).

Endoscopy is routinely performed in the investigation of lower GI bleeding, however in the emergency situation lesions may be missed in the unprepared colon. Even in the prepared colon, colonscopy fails to demonstrate 10 - 40% of bleeding sources (1,4).

Radionuclide scanning with technetium 99m (99mTc) sulphur colloid or (99mTc)-labelled erythrocytes is frequently used to localise mild to moderate or intermittent bleeding. Nuclear medicine examinations have the advantage of being easy to perform, non invasive and involve no patient preparation (1).

Angiography may be both diagnostic and therapeutic. It offers an important radiologic component in the investigation of these patients particularly where the bleeding site is obscure or undetermined at colonoscopy. A minimal bleeding rate of 0.5ml/min is necessary for angiographic detection, however for optimal sensitivity bleeding should be 1ml/min, equivalent to 3 units of blood per day (1,4). Nicholson et al found a high correlation between haemodynamic status and positive angiography (r = 1 for a blood pressure of less than 100mmHg), suggesting angiography in the haemodynamically stable patient may be a waste of time (3). The superior mesenteric artery (SMA) is cannulated first because most lower GI bleeds originate from the right colon, which is supplied by this artery. The bladder should be emptied by insertion of a Foley catheter prior to inferior mesenteric artery (IMA) injection to avoid missing a bleeding site behind the contrast filled bladder. If SMA and IMA injections fail to show extravastation, the coeliac artery should be cannulated as this may supply the middle colic which supplies the distal duodenum. Therapeutic angiography is most strongly indicated in frail or severely ill patients who are poor surgical candidates, but is increasingly offered to all patients with acute GI bleeds.

In the past, embolisation for lower GI bleeding was controversial due to fears of bowel ischaemia. The large bowel does not have the rich and overlapping arterial supply characteristic of the stomach and duodenum. This relative lack of a rich network of collaterals has been felt to render the gut more susceptible to ischaemic insult. There are no studies defining a safe position to embolise, but most operators aim to place the catheter tip in the vasa recta beyond the intestinal arcades.

There are two methods used to attempt to control bleeding, either by pharmacological vasoconstrictive therapy or by embolisation.

Vasopressin is infused selectively into the main trunk of the SMA or IMA, depending on the documented site of bleeding. Arrest of bleeding is reported in up to 90% of cases, with rebleeding reported in up to 20% of cases (1,5). However this treatment is not without side effects and requires the catheter to remain in situ for 24-48 hours and the patient to be monitored in ICU. The side-effects hypertension, bradycardia, coronary ischaemia, arrythmias, pulmonary oedema, oliguria, hyponatremia, peripheral acrocyanosis, and bowel cramps may be difficult to differentiate from those of impending bowel ischaemia and infarction (1,4).

Initially described in 1978, the technique of selective embolization for lower GI bleeding has progressed considerably. Newer coaxial microcatheters and fine torqueable guidewires allow super selective catheterisation of vessels. Catheter materials with less resistance reducing damage and spasm of small vessels have also evolved. Modern digital "roadmapping" assists catheter manipulation through the arterial tree. There are a wide variety of embolic agents available and improved digital angiographic equipment, contrast material and pharmacological agents increase procedure sensitivity. Both polyvinyl alcohol particles and platinum microcoils can be inserted through microcatheters. In our institution, microcoils are the embolic agent of choice as they can be clearly seen on fluoroscopy, and are easily and accurately positioned. There is also less risk of extreme distal occlusion which can occur with particles less than 100mm (5). There is a wide availability of different coil sizes and these can be well-matched to vessel calibre. Technical failure may occur because of several factors including vessel spasm, spontaneous cessation of bleeding and vascular tortuosity (6).

Embolization studies in the 1980s reported colonic infarction at a rate of between 0 and 20 percent (2). Newer catheter technology has decreased but not eliminated the likelihood of colonic ischaemia developing.

Luchtefeld et al report a success rate of 82% using embolisation for identified lower GI haemorrhage in 26 patients. Major complications were rebleeding (6%) and colonic necrosis (6%) (2). Gordon et al reported better success rates of 93% in 14 patients with no clinically apparent bowel infarctions . Interestingly the commonest site of bleeding in this series was at surgical anastomoses followed by diverticular disease (5). Peck et al achieved haemostasis in 71%, with no rebleeding within 30 days in 48%. This series also suggested that the location of bleeding may be a predictor of success, with proximal jejunal and caecal sites having less favourable outcomes (6).

Conclusion

Superselective arterial embolisation for lower GI bleeding remains a viable and possible definitive treatment option in selected patients. Procedure safety is increased by the use of microcoils and a willingness to abort if suitable catheter position cannot be attained. Although still reserved for high surgical risk patients, current literature suggests this procedure should be considered in all patients presenting with massive GI haemorrhage.

References

1. Lefkovitz Z, Cappell MS, Kaplan M, Mitty H, Gerard P. Radiology in the diagnosis and therapy of gastrointestinal bleeding. Gastroenterology Clinics of North America 2000; 29:489-512.

2. Luchtefeld MA, Senagore AJ, Szomstein M, Fedeson B, Van Erp J, Rupp S. Evaluation of Transarterial Embolisation for Lower Gastrointestinal Bleeding. Dis Colon Rectum 2000; 43:532-536.

3. Nicholson AA, Ettles DF, Hartley JE, Curzon I, Lee PWR, Duthie GS, Monson JRT. Transcatheter coil embolotherapy: a safe and effective option for major colonic haemorrhage. GUT 1998; 43:79-84.

4. Defreyne L, Vanlangenhove P, De Vos M, Pattyn P, Van Maele G, Decruyenaere J, Troisi R, Kunnen M. Embolization as a First Approach with Endoscopically Unmanageable Acute Nonvariceal Gastrointestinal Hemorrhage. Radiology 2001; 218: 739-748.

5. Gordan RL, Ahl KL, Kerlan RK Jr, Wilson MW, LaBerge JM, Sandhu JS, Ring EJ, Welton ML. Selectve Arterial Embolization for the Control of Lower Gastrointestinal Bleeding. Am J Surg 1997; 174:24-28.

6. Peck DJ, McLoughlin RF, Hughson MN, Rankin RN. Percutaneous embolotherapy of lower gastrointestinal haemorrhage. JVIR 1998; 9:747-751.