 |
       |
|
|
ANGIOGRAPHY IN THE DIAGNOSIS AND TREATMENT OF LOWER GASTROINTESTINAL BLEEDING
Sundeep J Punamiya*, Vimal R Someshwar**
Hon. Asso. Consultant and Head, Department of Vascular/Interventional Radiology, Bombay Hospital, Mumbai. Consultant, Interventional Radiologist, Bhatia General Hospital, Lilavati Hospital, Mumbai.
INTRODUCTIONLower gastrointestinal tract bleeding generally originates from lesions in the small intestine or large bowel. Up to 10% of patients, however, have the bleeding source proximal to the ligament of Trietz-in the duodenum, stomach or even in the oesophagus.[1] Lower GI bleeding may be slow and chronic, presenting with intermittent melanotic stools and anaemia. Acute GI bleeding, on the other hand, often presents as single or recurrent episodes of massive bleeding per rectum. Occasionally, the bleeding is explosive enough to cause significant hypovolaemia and even exsanguination. Five diseases account for the vast majority of cases : arteriovenous malformations, diverticulosis, tumours, ulceration and haemorrhoids. There are numerous less common causes which include conditions such as polyps, mesenteric ischaemia, vasculitis, small and large bowel varices, Meckel’s diverticulum, aorto-enteric fistula, visceral arterial aneurysm, endometriosis, radiation-induced injury and intussusception.
Since the first intra-operative application of mesenteric angiography by Margulis more than 30 years ago, angiography has become an integral tool in the management of patients with gastrointestinal bleeding. It is used for localising the site of bleeding and then for controlling the bleeding when more conservative measures are unsuccessful.
WHEN TO CONSIDER ANGIOGRAPHY
Angiography is more likely to demonstrate the site of haemorrhage if there is ongoing and active bleeding, as determined by a continuous need for blood and fluid transfusion to maintain haemodynamic stability (> 4 units of blood per day).[2] The accuracy of angiography depends on various factors:
1. Type of study : Selective angiography of the appropriate vessel is required to identify the bleeding site. This demands not only technical expertise of the radiologist but also a familiarity with anatomical variants, pathology, and anomalous sources of blood supply.2. Type of bleeding : Only arterial and capillary bleeding can be detected by selective angiography. Venous bleeding is extremely difficult to demonstrate on angiography.
3. Rate of bleeding : It has been documented experimentally that a minimal bleeding rate of 0.5 ml/min is required for angiographic demonstration of bleeding.[3] Hence if the bleeding is not brisk enough, the chances of picking a bleeder are poor.
4. Timing of study : The timing of the angiography is vital. Majority of times, there is a minute-to-minute variation in the rate of bleeding and it is very difficult to do a bedside assessment of the activity of bleeding. If the bleeding is brisk enough during the angiogram, the site of haemorrhage can be easily identified. If the bleeding ceases or reduces even a few seconds before the procedure, the angiogram can be negative.
5. Use of technical modifications : If the bleeding is intermittent and angiogram is negative, more aggressive pharmacological methods can be used to prolong or induce bleeding during the procedure and thus demonstrate the bleeding point. Various investigators have used techniques such as systemic heparinisation, selective intra-arterial vasodilatation, and/or thrombolytic agents (streptokinase, urokinase) in an attempt to improve diagnostic yield.[4],[5]
ACUTE LOWER GI BLEEDING
In patients with acute lower GI bleed, angiography should be the procedure of choice and needs to be performed on an emergency basis when the patient is actively bleeding. Extravasation of contrast medium into the bowel lumen is the main finding to look for and is seen only when the vessel is actively bleeding (Fig. 1). If the contrast extravasation is not seen, one has to rely on indirect evidence such as abnormal blush or abnormal vasculature, which would suggest a probable bleeding source (Fig. 2).
Fig. 1. 85 year old female with colonic diverticuli. Active bleeding demonstrated by extravasation of contrast into the ascending colon.
Fig. 2. 8 year old boy presented with severe rectal bleeding. Inferior mesenteric angiogram revealing a carpet of dilated vessels, consistent with rectosigmoid haemangiomatosis.
Fig. 3a. 54 year old female with intermittent melaena. (a) Superior mesenteric angiogram (DSA) revealing a vascular tuft in the caecal region supplied by a hypertrophied ileocaecal artery.
Fig. 3b. Early filling of the ileocaecal vein during the late arterial phase. Features are characteristic of angiodysplasia.
Fig. 4. 53 year old male with anaemia and persistent occult blood in stool. Superior mesenteric angiogram demonstrating irregular tumour vessels and abnormal blush in the ileocaecal region. On subsequent surgical exploration, a large leiomyosarcoma of the ileum was resected.
Fig. 5. 32 year old patient of pancreatitis presented with intermittent bouts of severe bleeding per rectum. Superior mesenteric angiogram (DSA) shows an aneurysm of the middle colic artery (arrow), which used to bleed into the adjacent small bowel.
Fig. 6. 64 year old female with two episodes of melaena. Superior mesenteric angiogram reveals dense inflammatory blush of the proximal jejunum.
Fig. 7a. 36 year old male had presented in the emergency ward with massive bleeding per rectum. (a) Superior mesenteric angiogram shows active bleeding in the caecum.
Fig. 7b. After initial intra-arterial infusion of vasopressin into the SMA, significant vasospasm was achieved and the bleeding stopped.
Fig. 7c. On discontinuing the infusion, however, the bleeding recurred and the patient was sent for surgery.
fore considering an angiogram, patients should have digital rectal and proctosigmoidoscopic examinations to exclude an ano-rectal source of bleeding such as haemorrhoids. A dedicated colonoscopy is not preferred as the large volume of blood within the colonic lumen can obscure the vision. Also, retrograde reflux of blood in the colon may mimic a more proximal source of bleeding. An upper GI endoscopy may be necessary to rule out a source in the upper GI tract; approximately 10% of patients presenting as massive lower GI bleeding have a source in the upper tract.
Overall, the efficacy of angiography in visualising contrast extravasation is in the range of 67-92%.[6],[7] In haemodynamically stable patients, however, the bleed is more intermittent, with a minute-to-minute variation. It is difficult in such patients to determine whether the bleeding is active, as rectal bleeding may persist a few hours after the bleeding may have stopped. Angiography performed in haemodynamically stable patients has often been unrevealing, with an efficacy of 30-40%.[7],[8] More recent aggressive pharmacological techniques using systemic heparinisation, selective intra-arterial vasodilatation, and/or thrombolytic agents (streptokinase, urokinase) have improved diagnostic yield for extravasation. If the angiogram is negative in spite of these measures, the arterial access can be retained and patient shifted to the ward. This would permit a prompt repeat angiogram, if bleeding recurs.
CHRONIC LOWER GI BLEEDING
If the bleeding is intermittent and chronic, angiography is undertaken after endoscopy and/or barium fail to reveal the causes of bleeding. In these cases, the chance of picking up a bleeding source on angiography is about 45%.[9],[10] If the angiogram is negative and the patient bleeds again, a repeat angiogram would locate the bleeding source in an additional 25% of patients.[10],[11]
Unlike acute massive bleeding, contrast extravasation is not a usual feature in chronic bleeders. The angiogram would reveal a vascular abnormality that proposes the site of bleeding. Angiodysplasia is seen as a vascular tuft with an early draining, dense vein (Fig. 3). A tumour would show the characteristic features of tumour neovascularity and blush (Fig. 4). A visceral aneurysm is identified by a saccular out-pouching of the artery (Fig. 5). Inflammatory lesions are seen as abnormal mucosal hyperaemic blush, with early venous drainage (Fig. 6).
The angiographer can also help the surgeon by facilitating exact localisation of a bleeding site at surgery by selective intra-operative methylene blue injection.12 This is important mainly in lesions of the small intestine where exact intra-operative localisation may be difficult. When a lesion is detected on the angiogram, a catheter is subselectively advanced deep in the artery supplying the lesion and the patient shifted to the operation theater. At surgery, methylene blue is injected through the catheter and the segmental discolouration of the small intestine delineates the area of the lesion, which is subsequently resected.
Fig. 8a. 23 year old male recovering from typhoid had a sudden, severe bout of per rectal bleeding. (a) An emergency superior mesenteric angiogram was considered, which showed contrast leaking into the caecum, probably from an ulcer.
Fig. 8b. Particles of gelfoam were injected through the catheter positioned in the ileocaecal artery and the post-embolisation angiogram shows occlusion of the bleeding artery. The bleeding ceased immediately and later on colonoscopy showed a healing ulcer in the caecum.
Fig.9a. 54 year old female, bleeding intermittently from a caecal angiodysplasia. (a) Superior mesenteric angiogram revealing the characteristic vascular tuft and an early draining vein in the caecal region.
Fig.9b. A microcatheter is advanced into the artery that supplies the lesion and superselective angiogram confirms its presence. Tiny particles of polyvinyl alcohol were injected through this catheter.
Fig.9c. The lesion is no longer opacified on the superior mesenteric angiogram after embolisation and the bleeding never recurred.
Fig.10a. 22 year old case of leukaemia, was undergoing chemotherapy and had profuse lower GI bleeding. (a) Superior mesenteric angiogram shows massive contrast extravasation into the jejunal lumen.
Fig.10b. Microcatheter was advanced into the vasa recta that was spurting blood. A small embolisation coil was deployed into this vessel.
ANGIOGRAPHIC THERAPY
Emergency subtotal colectomy used to be the traditional treatment of acute lower GI bleeding and has been reported to be associated with a high mortality (28-47%) and significant morbidity.[13] With angiographic localisation, it was possible to limit the resection, thereby reducing the mortality to about 14%.[6],[14]
Recent progress in transcatheter therapy has further changed the perspective. Since the advent of vasopressin in the 1960s and embolisation therapy in the 1970s, radiological control of haemorrhage has either obviated emergency surgery altogether or it has replaced the more hazardous emergency colectomy to an elective limited bowel resection, considerably lowering the morbidity and mortality.
Fig.10c. Subsequent angiogram shows effective embolisation, with no more bleeding into the small intestine.
Vasopressin acts by constricting the smooth muscle of the blood vessels, as well as the smooth muscle of the bowel wall. Selective intra-arterial infusion of vasopressin into the mesenteric artery had been used widely in lower GI bleed, especially in colonic diverticulosis. With this form of treatment, the bleeding stops in 65-85% cases.[15-17] It is relatively simple and only requires positioning the catheter in the superior or inferior mesenteric artery, but requires prolonged infusion. Due to the prolonged infusion and its systemic effects, various complications have been encountered. The major problems associated with vasopressin therapy have been hypertension, bradycardia, myocardial infarction, peripheral vascular ischaemia, catheter-related thrombosis and hyponatraemia with water retention. In addition, more than 30% of these cases had a recurrence after stopping the infusion[7] (Fig. 7).
Embolisation to control acute lower GI bleeding was first described by Rosch in the early 1970s to overcome the problem of recurrent bleeding after vasopressin therapy. It involves introduction of thrombogenic material into a catheter introduced selectively into the bleeding artery and occluding it. The use of embolisation was initially reserved for cases in which vasopressin therapy failed. As experience with this technique became more widespread, primary embolisation with a variety of agents at various sites have been performed (Fig. 8).[13],[18-20] Unlike vasopressin therapy, the effect of embolisation is more immediate and does not require an indwelling catheter, but is technically much more demanding. By this technique, the catheter has to be placed in an ideal position and embolic material deposited with precision to prevent any bowel ischaemia or infarction. The estimated rate of colonic infarction after embolisation has been reported at approximately 15% in the earlier series, where embolisation of proximal vasculature was carried out.[13],[18-20] Lately, catheters of small calibre are being produced, using which it is now possible to reach up to the vasa recta of the bowel (Figs. 9, 10).[21-22] Thus the embolic material is delivered at an extremely selective site and is less likely to cause extensive mural damage compared to proximal delivery. In the few reported series and in the authors’ experience, not a single case of symptomatic bowel ischaemia has been encountered.
REFERENCES
1. Rosch J. Lower gastrointestinal bleeding. In : Dondelinger RF, Rossi P, Kirdziel JC, Wallace S, eds. Interventional Radiology. Stuttgart : George Thieme Verlag, 1990; 349-67.
2. Kadir S, Emst CB. Current concepts in angiographic management of gastrointestinal bleeding. Curr Prob Surg 1983; 20 : 281.
3. Nusbaum M, Baum S, Kuroda K, et al. Pharmacologic control of portal hypertension. Surgery 1967; 62 : 299.
4. Rosch J, Keller FS, Wawrukiewicz AS, et al. Pharmacoangiography in the diagnosis of recurrent massive lower gastrointestinal bleeding. Radiology 1982; 145 : 615.
5. Koval G, Benner KG, Rosch J, et al. Aggressive angiographic diagnosis in acute lower gastrointestinal haemorrhage. Dig Dis Sci 1987; 32 : 248.
6. Browder W, Cerise EJ, Litwin MS. Impact of emergency angiography in massive lower gastrointestinal bleeding. Ann Surg 1980; 204 : 530.
7. Athanasoulis CA. Lower gastrointestinal bleeding. In : Athanasoulis CA, Pfister RC, Greene RE, Robertson GH eds. Interventional radiology. Philadelphia : Saunders, 1982; 115-48.
8. Rahn NH, Tishler JM, Han SY, Russinovich NA. Diagnostic and interventional angiography in acute gastrointestinal haemorrhage. Radiology 1982; 143 : 361.
9. Sheedy PF, Fulton RE, Atwell DT. Angiographic evaluation of patients with chronic gastrointestinal bleeding. Am J Roentgenol 1975; 123 : 338.
10. Rollins ES, Picus D, Hicke ME, et al. Angiographic is useful in detecting the source of chronic gastrointestinal bleeding of obscure origin. Am J Roentgenol 1991; 156 : 385.
11. Lau WY, Yuen WK, Chu KW, et al. Obscure bleeding in the gastrointestinal tract originating in the small intestine. Surg Gynecol Obstet 1992; 174 : 119.
12. Athanasoulis CA, Moncure AC, Greenfield AJ, et al. Intraoperative localisation of small bowel bleeding sites with combined use of angiographic methods and methylene blue injection. Surgery 1980; 87 : 77.
13. Uflacker R. Transcatheter embolisation for treatment of acute lower gastrointestinal bleeding. Acta Radiol 1987; 28 : 425.
14. Uden P, Jiborn H, Jonsson K. Influence of selective mesenteric arteriography on the outcome of emergency surgery for massive lower gastrointestinal haemorrhage : A 15-year experience. Dis Colon Rectum 1986; 29 : 561.
15. Baum S, Nusbaum M. The control of gastrointestinal haemorrhage by selective mesenteric arterial infusion of vasopressin. Radiology 1971; 98 : 497.
16. Clarke RA, Colley DP, Eggers FM. Acute arterial gastrointestinal haemorrhage : Efficacy of transcatheter control. Am J Roentgenol 1981; 136 : 1185.
17. Gomes AS, Lois JF, McCoy RD. Angiographic treatment of gastrointestinal haemorrhage : comparison of vasopressin infusion and embolisation. Am J Roentgenol 1986; 146 : 1031.
18. Walker WJ, Goldin AR, Shaff MI, et al. Per catheter control of haemorrhage from the superior and inferior mesenteric arteries. Clin Radiol 1980; 31 : 71.
19. Rosenkrantz H, Bookstein JJ, Rosen RJ, et al. Postembolic colonic infarction. Radiology 1982; 142 : 47.
20. Lawler G, Bircher M, Spencer J, et al. Embolisation in colonic bleeding. Br J Radiol 1985; 58 : 83.
21. Guy GE, Shetty PC, Sharma RP, et al. Acute lower gastrointestinal haemorrhage : treatment by superselective embolisation with polyvinyl alcohol particles. Am J Roentgenol 1992; 159 : 521.
22. Peck DJ, McLoughlin RF, Hughson MN, Rankin RN. Percutaneous embolotherapy of lower gastrointestinal haemorrhage. J Vasc Intervent Radiol 1998; 9 : 747.
 |
