| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, The University of Chicago Hospitals, 5841 S. Maryland
Ave., MC 2026, Chicago, IL 60637.
2
Central Kentucky Radiology, 2365 Harrodsburg Rd., Ste. B125, Lexington, KY
40504.
3
Racine Radiologist Group, 3803 Spring St., Rm. 208, Racine, WI 53405.
Received January 24, 2001;
accepted after revision April 4, 2001.
Address correspondence to B. Funaki.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Twenty-seven patients with severe colonic bleeding due to diverticular disease (n = 19), angiodysplasia (n = 6), cecal ulcer (n = 1), or unknown cause (n = 1) underwent attempted microcoil embolization (n = 25). Microcatheters were used in all procedures, and embolization was performed at the level of the vasa recta or the marginal artery of Drummond. Branches of the superior mesenteric artery were embolized in 12 patients, branches of the inferior mesenteric artery were embolized in 12 patients, and branches of both the superior and inferior mesenteric arteries were embolized in one patient.
RESULTS. Technical success was achieved in 93% (25/27) of the procedures. However, immediate hemostasis occurred in 96% (26/27) of patients because in one failed procedure, an occlusive dissection of the inferior mesenteric artery arrested bleeding. Three patients rebled within 24 hr. One patient was treated with endoscopic cauterization, and two patients underwent right hemicolectomy. One patient who underwent right hemicolectomy for rebleeding had ischemic changes found on pathologic analysis of the resected specimen, and a second patient who underwent embolization of branches of the superior and inferior mesenteric arteries developed bowel infarction requiring left hemicolectomy. Prolonged clinical success occurred in 81% (22/27) of patients.
CONCLUSION. Therapeutic microcoil embolization for severe colonic hemorrhage is an effective and well-tolerated procedure.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Before all procedures, all patients received blood transfusions for hemorrhage. IV midazolam hydrochloride (Versed; Roche Pharmaceuticals, Nutley, NJ) and fentanyl citrate (Sublimaze; Abbott Laboratories, North Chicago, IL) were given for sedation and analgesia. Oxygen saturation, blood pressure, and heart rate and rhythm were monitored in all patients by noninvasive means.
The right or left common femoral artery was punctured using the Seldinger technique, and a 5-French vascular sheath was placed. Catheters were used on the basis of the attending interventional radiologist's preference; but in general, a 5-French RC-1 catheter (Boston Scientific, Natick, MA) was used for superior mesenteric artery catheterization, and a 5-French Rosch Inferior Mesenteric or Simmons 1 or 2 catheter (Boston Scientific) was used for inferior mesenteric artery catheterization. All angiograms were obtained using digital subtraction angiography. If obtained, nuclear medicine studies were used to guide catheterization (Fig. 1A,1B,1C,1D). If the bleeding source appeared to be in the distal colon, the inferior mesenteric artery was catheterized initially; and if bleeding appeared to originate in the proximal colon, the superior mesenteric artery was treated first. IV glucagon (1 mg) was administered at the discretion of the radiologist to decrease bowel peristalsis.
|
|
|
|
Once the bleeding source was localized angiographically, a 3-French microcatheter (Target 18 or 325; Boston Scientific) was advanced coaxially to the bleeding site through the indwelling 5-French catheter. Attempts were made to position the catheter as close to the bleeding site as possible. In the inferior mesenteric artery distribution, the catheter was advanced to the marginal artery or the terminal artery if possible; and in the superior mesenteric artery, the catheter was advanced to the vasa rectum (Fig. 2A,2B,2C). When spasm occurred, intraarterial papaverine (25 mg) was administered. Embolization was performed only when the catheter could be advanced to or distal to the mesenteric border of the colon (in other words, to the marginal artery or vasa rectae). Microcoils (complex helical fibered platinum coil; Boston Scientific) ranging in size from 2 to 3 mm x 20 to 30 mm were deployed. Embolization was performed until no further arterial extravasation was seen.
|
|
|
Technical success was defined as cessation of bleeding as seen on angiography. Clinical success was defined as cessation of bleeding without further surgical or endoscopic intervention. Minor ischemic complications were defined as events that required no further therapy. Major ischemic complications were defined as events that required a prolonged hospital stay or additional treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Prolonged clinical success occurred in 81% (22/27) of patients. Three patients rebled within 24 hr. The first patient was an 84-year-old woman who had undergone embolization for severe rectal bleeding. At the time of embolization, she was hypotensive, with a hematocrit of 10%. Because she had a barium impaction in the sigmoid colon and rectum from a small-bowel study performed 1 month earlier, colonoscopy was attempted initially instead of angiography but failed to identify the bleeding site because of hemorrhage. Angiography was performed after failed colonoscopy, but residual barium limited optimal delineation of both the bleeding site and the arterial anatomy (Fig. 4A,4B,4C). Using multiple obliquities, we identified three third- and fourth-order feeding arteries and embolized them with microcoils. Hemostasis occurred, although additional arteries in the region of the affected area of the bowel were obscured by residual barium. The patient's condition stabilized and improved overnight. Less severe rebleeding occurred the next day, and a second colonoscopy disclosed the site of hemorrhage, which was treated with endoscopic cauterization. The patient has had no further bleeding at the 2-month follow-up.
|
|
|
A second patient with diverticulosis rebled 2 days after initial embolization, and repeated angiography revealed perfusion through the previously placed microcoils. Additional microcoils were deployed and hemostasis was again achieved, but the patient rebled 24 hr later and then underwent a right hemicolectomy. Pathologic analysis of the resected specimen revealed mucosal ulcerations consistent with ischemic changes.
The third patient, a 56-year-old man, rebled 3 days after embolization from the cecum. A right hemicolectomy was performed and pathologic analysis revealed a cecal ulcer, cecal polyp, and appendicitis. Bleeding was thought to originate from the cecal ulcer but was probably exacerbated by coexisting acute appendicitis.
Delayed (>1 week) rebleeding occurred in two patients at 2 and 4 months after embolization. Both patients had extensive sigmoid diverticulosis and experienced a single episode of bleeding that was treated conservatively. Neither patient has had additional hemorrhage at the 9- and 16-month follow-ups.
Minor ischemic complications occurred in two patients. An asymptomatic ulcer in the right colon was discovered incidentally on colonoscopy in a 50-year-old man 4 months after successful embolization. The exact cause of the ulcer was unclear. The ulcer may have been the source of the bleeding or may have been ischemic in nature. A colonic stricture detected at colonoscopy developed in one patient 1 year after successful embolization.
Major ischemic complications occurred in two patients. In one patient, ischemic changes were noted in a segment of right colon that was resected for rebleeding. Although it is uncertain whether the ischemia would have required treatment, we chose to include it as a major complication because surgery was pursued to correct recurrent bleeding. Bowel infarction requiring hemicolectomy occurred in a 68-year-old woman with lymphoma (Fig. 5A,5B,5C,5D). At the time of angiography, the patient was considered a prohibitive surgical risk, and embolization was performed as a life-saving measure. During the procedure, the inferior mesenteric artery was catheterized, but because of vasospasm, a microcatheter could not be advanced to the bleeding site in the splenic flexure of the colon. Thus, a microcoil was deployed in the marginal artery approximately 5 cm from the source of the bleeding. The superior mesenteric artery was then catheterized, and an angiogram showed active extravasation. Additional microcoils were deployed at the site of the bleeding, and hemostasis occurred. The patient experinced continual abdominal pain after the procedure. Approximately 1 week after embolization, colonoscopy revealed diffuse ischemia of the left colon, and a left hemicolectomy was performed.
|
|
|
|
Follow-up longer than 1 month was available in 23 patients. Total follow-up was 235 months (range, 1 day-57 months; mean follow-up, 8.7 months). Colonoscopy was performed before embolization in 22 patients and after embolization in six (24%) of 25 patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patients with colonic hemorrhage can be divided into three broad groups. The first group encompasses most patients and is characterized by minor bleeding that can be treated conservatively [15]. The second group has severe life-threatening intermittent bleeding, with hemodynamic stability between episodes of hemorrhage. Because of the intermittent nature of bleeding in this group, 99mTc-RBC scans are helpful before angiography is performed. Nuclear medicine scanning is more sensitive than angiography and is advantageous for intermittent bleeding because imaging is performed continuously during a 1- to 2-hr period. Angiography is performed within 1 hr of a study with positive results. Using this algorithm, the number of negative angiograms with negative results decreases [16]. The third group consists of patients with continual active bleeding. These patients are typically hypotensive [11] and require urgent angiography.
The objective when performing colonic embolization is to achieve a compromise between selective arterial inflow reduction and maintenance of collateral arterial blood flow. In other words, arterial inflow must be decreased enough to allow hemostasis, but not to the degree to cause total devascularization. This goal underscores the importance of both an intact coagulation cascade and superselective embolization. We routinely correct coagulopathy before therapy using a 3-French microcatheter that enables superselective embolization of the marginal arteries or vasa recta. The preferred level of embolization remains controversial. In general, we deploy embolic agents as distally as possible and do not routinely attempt embolization unless a microcatheter has been advanced to or beyond the border of the colon. The risk of infarction is related to both the embolic agent and the proximity of embolization. Ischemia and infarction may result from embolization of proximal colonic branches supplying a large area of bowel or embolization of multiple distal arteries that are devoid of collateral flow [4, 13]. In most early reported series, embolization was performed in the proximal superior or inferior mesenteric arteries because microcatheters facilitating superselective embolization were not yet available. The procedural complication in our series—a dissection of the inferior mesenteric artery—appeared to function identically to a proximal embolization because bleeding immediately stopped in this patient. Proximal embolization has proven to be effective; however, when encountered, ischemic complications may be significant because arterial supply to a long bowel segment is affected [4].
Various agents have been used for colonic embolization. We favor microcoils because they are readily seen, allowing precise deployment and, in most cases, fulfilling the objective of decreasing the perfusion pressure while allowing enough collateral flow to preclude infarction. Polyvinyl alcohol and Gelfoam (Upjohn, Kalamazoo, MI) pledgets may also be used successfully, but in our experience, these agents have disadvantages. Both agents are more difficult to control than microcoils because neither can be directly seen. An additional disadvantage of polyvinyl alcohol particles is that small particles, as well as larger particles (if they fragment) may reach intramural circulation beyond the level of collateralization [17] or may reflux into nontarget arteries.
Superselective embolization may be technically demanding, especially in elderly patients who have moderate to severe atherosclerotic disease. The colonic arterial arcade is tortuous, and smaller arteries are prone to spasm. We were unable to superselectively catheterize a bleeding artery in one patient despite the use of vasodilators and were unable to advance a microcatheter to the bleeding site in a second patient. In two additional patients, embolization was initially successful, but rebleeding occurred within 24 hr. We suspect hemostasis occurred initially because of the deployed microcoils acting in combination with arterial vasospasm. Once spasm abated, rebleeding ensued. In one of these patients, hemorrhage was arrested a second time with additional embolization, but rebleeding again occurred. Both patients ultimately underwent hemicolectomy.
Two of the three patients in our series who experienced early recurrent hemorrhage after embolization bled from cecal lesions (angiodysplasia and a cecal ulcer). Peck et al. [5] suggested that bleeding from the cecum might not be as responsive to embolotherapy as bleeding from other sites in the gastrointestinal tract because of the predominance of angiodysplasia that occurs in the cecum. In their experience, rebleeding occurred in 75% (3/4) of cecal embolizations. The third patient with early rebleeding in our series had a barium impaction that precluded thorough evaluation of the feeding arteries. Embolization was pursued after failed colonoscopy. The patient had severe comorbidities and was not considered a surgical candidate. Although prolonged hemostasis did not occur after embolization, rebleeding was much less severe, enabling successful endoscopic treatment of the bleeding site.
In this series, immediate hemostasis was achieved in 97% (30/31) of patients, and even when rebleeding occurred, embolotherapy proved beneficial because it allowed patients to be stabilized for surgery or slowed bleeding enough to facilitate endoscopic therapy. Although it appears that superselective embolization decreases the likelihood of ischemia, it does not eliminate the risk altogether.
One instance of bowel infarction occurred in our series, and three additional patients had other ischemic complications. Ischemic complications with superselective embolization have also been reported in the literature [11,12,13]. Our patient who suffered infarction underwent embolization of branches of both the superior and inferior mesenteric arteries. The patient was considered a prohibitive surgical risk at the time of embolization because of blood loss and comorbidity. Superselective embolization was technically unsuccessful; thus, arterial perfusion to a relatively long segment of bowel was ultimately compromised, leading to ischemia. Because submucosal collateral blood flow may be preserved only when arteries to a short segment of bowel are embolized and ischemia may be unpredictable, we believe that attempts should be directed to occlude arteries to the shortest segment of bowel possible. In a hemodynamically stable patient, we believe that other therapies are preferable to embolization when perfusion to a long segment of bowel must be compromised. However, any treatment must be analyzed in the context of the patient's underlying clinical condition and in comparison with the success and complications associated with alternative treatment. In our patient who suffered bowel infarction, colonoscopy had already failed, and vasopressin was relatively contraindicated because of coronary artery disease.
The cause of colonic hemorrhage has important implications for both the short- and long-term success of the procedure. Because embolization treats a symptom of the underlying disorder rather than the disease itself, late recurrent bleeding may occur in some patients, particularly those with multifocal disease (e.g., extensive diverticulosis and bowel metastases). Nonetheless, like many percutaneous procedures, successful embolization does not preclude repeated attempts should bleeding recur; and, when successful, embolization often averts bowel resection and associated colostomy. In our series, diverticular disease was the most common cause of bleeding. These patients were ideally suited for embolization because focal bleeding is typically arrested with two or three microcoils. Our patients with bleeding diverticula exhibited a good initial response to embolization. Repeated hemorrhage was uncommon, occurring in only two patients at 2 and 4 months after successful embolization.
A recent study by Jensen et al. [18] described the endoscopic treatment of colonic diverticular hemorrhage using epinephrine injections or bipolar coagulation in 10 patients. Endoscopic treatment was effective, and none of these patients had recurrent bleeding or required surgery. However, several disadvantages are associated with this technique. First, high-volume bowel lavage is necessary for 6-12 hr before an urgent colonoscopy. Second, this technique remains unproven for other types of hemorrhage. In contrast, embolization can be performed without delay and has been proven to effectively stop bleeding due to any cause, not just diverticular hemorrhage. Angiography does not preclude endoscopy in the event that the bleeding site is not seen. A limitation of endovascular therapy is that the patient must be actively bleeding at the time of the procedure to allow identification and treatment. Colonoscopy may identify a bleeding source by revealing adherent clots, prominent vessels, or a mass. Angiography is well suited for patients with severe hemorrhage who may not be able to tolerate prolonged high-volume lavage, whereas colonoscopy may more accurately characterize intermittent bleeding. Therefore, these minimally invasive techniques should be regarded as complementary rather than competing therapies.
One weakness of this study is that a relatively small percentage (24%, 6/25) of our patients underwent colonoscopy after embolization. Therefore, the full extent of ischemic complications is likely underestimated. One patient developed a stricture at the site of embolization; other undetected episodes of subclinical bowel ischemia may also have occurred. Nonetheless, nearly all patients in our series were elderly and frail, with extensive comorbidity. We believe any undiagnosed complications that were clinically asymptomatic in this population are of debatable significance.
In summary, our results agree with others that superselective embolization of colonic hemorrhage using coaxial microcatheters is safe and effective. This procedure should be the initial treatment of choice in any patient with bleeding severe enough to require angiography and in whom superselective catheterization is possible. With superselective embolization, the risk of symptomatic ischemia is minimal. If superselective catheterization is unsuccessful or if microcoil embolization fails to control bleeding, less selective embolization should be used judiciously because the risk of ischemic complication increases. Nonetheless, less selective embolization may be valuable in many patients because it may allow a patient's condition to stabilize before surgery.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. J. Laing, T. Tobias, D. I. Rosenblum, W. L. Banker, L. Tseng, and S. W. Tamarkin Acute Gastrointestinal Bleeding: Emerging Role of Multidetector CT Angiography and Review of Current Imaging Techniques RadioGraphics, July 1, 2007; 27(4): 1055 - 1070. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G Annamalai and I Robertson Acute gastrointestinal haemorrhage: investigation and treatment Imaging, August 1, 2004; 16(3): 264 - 270. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
