Physicians Patient and Public Members Annual Meeting Industry DCR About Us
Home > Physicians > Education > Core Subjects > Gastrointestinal Bleeding
Print

Gastrointestinal Bleeding

Amy L. Halverson, MD
Assistant Professor of Surgery
The Feinberg School of Medicine
Northwestern University

Initial Assessment

The initial management of patients presenting with acute lower gastrointestinal hemorrhage is resuscitation. This includes establishing large boar intravenous access, blood transfusion as necessary and laboratory testing to evaluate for bleeding diatheses. All patients with massive lower gastrointestinal hemorrhage should have a nasogastric tube placed to evaluate for an upper gastrointestinal source of bleeding. A negative gastric lavage may miss bleeding from a duodenal ulcer due to pylorospasm preventing regurgitation of blood. A gastric aspirate containing bile without blood decreases but does not eliminate the possibility of a false negative lavage.

A simultaneous history and physical exam should be performed. Prior episodes of gastrointestinal bleeding, peptic ulcer disease, gastritis, enteritis, colitis, and previous upper or lower endoscopy should be reviewed. A history of hepatic disease, coronary artery disease and prior abdominopelvic radiation may be of particular significance. Patients should specifically be asked about the use of warfarin, Plavix, aspirin or nonsteroidal analgesic medications.

Differential Diagnosis

The incidence of lower gastrointestinal hemorrhage is unknown. Diverticular bleeding is the most common etiology of lower gastrointestinal hemorrhage accounting for approximately 30-50% of individuals treated for lower GI hemorrhage. In a review of etiologies of lower gastrointestinal bleeding Vernava et al. reported that inflammatory bowel disease was an etiology in 21% followed by colonic neoplasia and benign anorectal disease.(1) In several series, arteriovenous malformations have accounted for only 1 to 13 percent of diagnoses. (2) (3) (4)In considering etiology of lower gastrointestinal bleeding one must also consider upper GI sources as well such as gastric or duodenal ulcers or small bowel lesions.

Diverticulosis

Diverticulosis is present in approximately 40 percent of the individuals over age 60. The risk of diverticular associated bleeding is estimated to be between 4 and 48 percent.(5;6) Diverticular bleeding resolves spontaneously in 80 percent of episodes. The risk of a second episode of bleeding is approximately 25 percent. Although most diverticula are in the sigmoid colon, 50 percent to 95 percent of bleeding diverticula are to the right of the splenic flexure. (7) (6)Observation alone is recommended after one episode of diverticular bleeding. After a second episode however, the risk of bleeding a third time increases to 50 percent and resection of the affected segment of colon is recommended. The risk of bleeding from diverticular disease is increased three-fold in individuals taking nonsteroidal analgesic medication compared to matched controls. Wilcox et al. showed that nonsteroidal medications increase the risk of diverticular bleeding to the same degree that they increase the risk of bleeding from a duodenal ulcer.(8)

Inflammatory Bowel Disease

Bleeding from inflammatory bowel disease usually presents as bloody diarrhea. As many as 6 percent of patients with either Crohn's colitis or ulcerative colitis may have severe gastrointestinal hemorrhage. In the series by Rossini et al., 45 of 311 patients (14%) with massive lower gastrointestinal hemorrhage diagnosed by colonoscopy had ulcerative colitis. Robert et al. reported that hemorrhage due to ulcerative colitis resolves spontaneously in 50 percent of episodes. Resection is recommended after one episode because of a 35% chance of rebleeding.(9) The reported incidence of massive bleeding associated with Crohn's disease ranges from 1 percent to 13 percent. Bleeding is more common in patients with colonic involvement that among those with small bowel disease alone (1.9% vs. 0.7%, p<0.001). (10)

The recommended operation for bleeding from severe colitis is a total abdominal colectomy. Proctectomy generally should not be performed in the acute setting given the increased morbidity in these usually ill patients. An interval proctectomy with ileal pouch reconstruction may be performed in individuals with ulcerative colitis. Bleeding from the residual rectal stump usually resolves following colon resection. Only in unusual cases of severe bleeding from the rectum should proctectomy be considered. Colectomy with ileoproctostomy can be performed in patients with Crohn's colitis if the rectum is not the source of bleeding and there is not extensive proctitis. Segmental small bowel resection is appropriate for individuals with small bowel Crohn's disease when bleeding has been identified in a diseased segment of small bowel.

Radiation Proctitis

Severe bleeding from radiation proctitis is unusual and is often difficult to manage. Topical medications are generally not helpful for controlling bleeding from severe proctitis. Argon and Nd:YAG laser have been successfully used to treat hemorrhagic proctitis. Some patients will require repeated treatments. Reported minor associated complications include ileus, abdominal pain and tenesmus. (11;12)

Topical application of a 4% formalin solution is also effective for patients with hemorrhagic proctitis. Approximately 500 ml of the formalin solution is instilled in 50 ml aliquots into the rectum via a rigid proctoscope. The solution is left in the rectum for 30 seconds then aspirated, and the rectum irrigated with saline. This is repeated 10 times. Care is taken to avoid dripping formalin onto the perianal skin. Reported success rates range between 75 percent and 95 percent.(11-13)While this technique is generally safe, severe complications such as rectal necrosis and formation of a rectovaginal fistula have been reported.(14)

When endoscopic measures fail, severe hemorrhage may improve with a diverting colostomy alone. Ultimately proctectomy may be required to control hemorrhage. Surgery for radiation proctitis is associated with a high morbidity. Beart et al. reported a 44 percent morbidity rate following colostomy alone and 80 percent morbidity with more aggressive operations. The mortality rate in this series was 13 percent.(15)

Neoplasms

Bleeding associated with adenomatous polyps or adenocarcinoma of the colon and rectum usually presents as melena or bright red blood per rectum. Although lower gastrointestinal hemorrhage is unusual with colorectal neoplasia, it reportedly accounts for between 7 and 33 percent of cases of lower gastrointestinal hemorrhage.(3;16)

Arterial venous malformations

Arteriovenous malformations are degenerative lesions that occur due to chronic partial obstruction of submucosal veins on the colon wall. With time these veins become dilated and tortuous. The precapillary sphincters become incompetent and arteriovenous communications develop.(16)Arteriovenous malformations occur most frequently in the cecum. This may be explained by the increased wall tension associated with the lager diameter of the cecum according to Laplace's law.

Bleeding caused by arteriovenous malformations is often slow and intermittent. Only about 15 percent of individuals with arteriovenous malformations develop lower gastrointestinal hemorrhage.(17) Several series reported that arteriovenous malformations account for only about 2 percent of cases of massive lower gastrointestinal bleeding.(16) Significant bleeding due to these vascular ectasias spontaneously resolves in 85-90 percent of patients.(18) Up to 85 percent of these individuals with have a second episode of bleeding. Therefore, arteriovenous malformations identified in individuals following an episode of lower gastrointestinal hemorrhage should be treated. The treatment of choice is colonoscopic coagulation. Colectomy should be reserved for individuals with repeated bleeding episodes.

Anorectal Disease

Significant lower gastrointestinal bleeding due to benign anorectal disease is unusual. Yet, Vernava et al reported that 11 percent of 17,941 patients with lower gastrointestinal bleeding also had anorectal disease defined as hemorrhoids, fissure and fistula-in-ano.{Vernava AM, Longo WE, et al. 1996 27 /id} All patients presenting with lower gastrointestinal bleeding should undergo digital rectal examination, anoscopy and rigid proctoscopy. The presence of anorectal pathology does not preclude the presence of a more proximal source of bleeding and evaluation of the entire colon should be completed even with an identified anorectal lesion.

Anorectal varices may be confused with hemorrhoids and can cause significant bleeding. The incidence of anorectal varices ranges between 78 and 89 percent in individuals with portal hypertension.(19;20) Bleeding from anorectal varices can be controlled by over sewing the lesion or injection sclerotherapy. The definitive treatment for persistent bleeding from lower gastrointestinal varices is portal decompression.

Upper GI sources of bleeding

An upper gastrointestinal source of bleeding (i.e., proximal to the ligament of Treitz) is identified in 10 to 15 percent of individuals who present with bleeding per rectum. Associated symptoms such as hematemesis or epigastric pain are always present with upper gastrointestinal bleeding. If a definitive source of bleeding is not identified in the lower gastrointestinal tract, an esophagogastroduodenoscopy should be performed to evaluate the stomach and duodenum. Vernava et al. even recommends this procedure on all patients undergoing surgery for lower gastrointestinal bleeding even when a lower gastrointestinal source has been identified.(16)

Small Intestinal Hemorrhage

When colonic and upper gastrointestinal evaluations fail to identify a source of bleeding, a small intestinal source should be considered. Angiodysplasia is the most common cause of hemorrhage in the small intestine, occurring in 70 to 80 percent of all episodes.(21) Other causes of small bowel bleeding include diverticula, Meckel's diverticulum, neoplasia, enteritis and aortoenteric fistulas.

Evaluation and Treatment

Colonoscopy

Colonoscopy is the procedure of choice for evaluation of lower gastrointestinal bleeding. Successful visualization of a source of bleeding is optimized by colonic lavage. Reported rates of identifying a source of bleeding range from 74 to 82 percent. (16) Four liters of GoLytely? (Braintree Laboratories, Inc., Braintree, MA) orally or via a nasogastric tube is the lavage method of choice for most authors. Prior to colonoscopy, patients should be transfused for severe anemia and any coagulopathy should be corrected. Colonoscopy should not be attempted in an unstable patient or in a patient with severe, active colonic inflammation.

Endoscopic coagulation is the treatment of choice for bleeding arteriovenous malformation and telangiectasias. The risk of rebleeding after endoscopic coagulation of angiodylplasias ranges for 13 to 53 percent. The risk of colonic perforation following endoscopic coagulation is approximately 2 percent.(22)

Reported endoscopic treatments for diverticular bleeding include coagulation, injection of epinephrine and placement of an endoscopic hemoclip. Jensen et al. reports results of urgent colonoscopy in 48 patients with severe hematochezia and diverticulosis. In all patients colonoscopy was performed within 6-12 hours after hospitalization or the diagnosis of hematochezia. Patients with active bleeding on colonoscopy were treated with epinephrine or bipolar coagulation if a visible vessel was seen. The authors recommend tattooing the area for easy identification should rebleeding occur.(23)

Enteroscopy

A negative colonoscopy in an individual with clinical evidence on ongoing bleeding should be followed by an upper endoscopy to evaluate for bleeding from the stomach or duodenum. If standard upper endoscopy is unsuccessful in identifying a bleeding point then small-bowel entersocopy should be considered. This procedure may be done using either a special enteroscope or pediatric colonoscope. Authors using this technique have reported successful identification of bleeding sources in approximately 25 percent. (24;25)

A new diagnostic tool is the video capsule endoscopy. The M2A Given capsule is an ingestible capsule that contains a color camera, light source, battery and radio transmitter. The capsule transmits digital images to a data recorder unit worn by the patient. Recent studies using this device have shown a diagnostic yield of 55 percent to 66 percent in patients with occult gastrointestinal bleeding who had failed attempts at diagnosis with conventional endoscopy and radiographic small bowel studies. (26;27)

Nuclear Scintigraphy

The use of nuclear scintigraphy for the detection of hemorrhage was introduced in the 1970's. It reportedly can detect bleeding as slow as 0.1 ml/minute.(28) 99mTc-labled red blood cell scanning can be detected on images up to 24 hours after injection. This technique can detect intermittent as well as active bleeding with a sensitivity of 80 to 98 percent.(29) The accuracy of blood cell scanning in localizing the site of hemorrhage ranges between 37 and 98 percent.(16) The relative imprecision in locating a bleeding site is the main limiting factor of scintigraphy. Often it is used as a preliminary test to confirm the presence of bleeding prior to proceeding with angiography. Scintigraphic evaluation should not be performed in patients with rapid bleeding or hemodynamic instability.

Selective Mesenteric Angiography

Selective mesenteric angiography is not as sensitive to low rate or intermittent bleeding compared to red cell scans, but it can provide precise localization of bleeding sites and can be used for therapeutic intervention. Angiography can detect hemorrhage at a rate of 0.5-1 ml/minute. Intravenous contrast is injected via a trans femoral catheter first into the superior mesenteric artery followed by the inferior mesenteric artery. If no bleeding site is identified then the celiac axis is studied. A bleeding site is identified by extravasation of contrast material into the lumen of the bowel. The sensitivity of angiography ranges from 34 to 86 percent, with a complication rate of approximately 2 percent. False negative examinations thought to be caused by intermittent hemorrhage due to either vasospasm or intermittent clotting. Various agents such as heparin, tolazoline vasodilation, urokinase and streptokinase have been used in attempts to increase the diagnostic sensitivity of angiography. Koval et al. reported an improved efficacy from 32 percent to 65 percent using pharmacologic assisted angiography. These authors reported an increased transfusion requirement and increased complications with these more aggressive measures.(30)

After localization of a bleeding site, therapeutic options to control hemorrhage include selective infusion of vasopressin or superselective embolization. Vasopressin (Pitressin, Parke-Davis, Morris Plains, NJ) when arterially infused causes arteriolar vasoconstriction and bowel wall contraction. Vasopressin infusion is started at 0.2 units/minute and can be increased to 0.4 units/minute. After infusion for 20 to 30 minutes, a repeat angiogram is obtained to evaluate for cessation of bleeding. If there is no evidence of ongoing bleeding the vasopressin infusion is continued for 6 to 12 hours. At that point the infusion range is decreased to 0.1 units/minute for an additional 6 to 12 hours. The vasopressin infusion is then discontinued and saline is infused for several additional hours. The reported success rate of vasopressin infusion is approximately 80 percent. Rebleeding occurs in approximately 30 percent of patients.(31)

The use of vasopressin infusion is limited by its high risk of complications. Most important, vasopressin infusion can lead to myocardial ischemia and peripheral vascular ischemia. Other complications include mesenteric thrombosis, intestinal infarction hyponatremia and death. Vasopressin infusion is contraindicated in individuals with coronary artery disease or peripheral vascular disease. The risks associated with vasopressin infusion have made arterial embolization the preferred angiographic method for controlling bleeding.

Early experience with angiographic embolization was reported a relatively high incidence of recurrence and potential complications primarily related to intestinal ischemia and infarction. In recent years significant improvements in embolization technique such as smaller catheters and materials with less resistance that prevent damage to and spasms of small vessels have made super selective embolizations safer. A recent study by Luchtefeld et al. reported a 53 percent success rate of positive angiograms. Embolization was successful in stopping bleeding in 14 of 17 bleeding episodes (82%). In this series there was one patient who required surgery for intestinal necrosis. (32)

Surgery

Indications for surgery: Approximately 10 to 25 percent of patients presenting with acute lower gastrointestinal bleeding will require urgent surgery. Indications for surgery in the acute setting include patients requiring greater than 1500 ml of blood transfusion for initial resuscitation with ongoing bleeding, a requirement of six or more units of red blood cells, ongoing bleeding for 72 hours or rebleeding within one week of initial cessation of bleeding. (16)

Operative strategy: If a precise site of bleeding has been located, segmental resection is appropriate. If the patient requires surgical intervention and the bleeding site has not been localized then a subtotal colectomy is indicated. Depending on the patient's overall condition, an end ileostomy or a ileoproctoscopy with or without a diverting loop ileostomy may be performed. A blind segmental colectomy alone has been associated with a rebleeding rate as high as 75 percent and mortality rates of 20 to 50 percent.(16) Intraoperative diagnostic methods used to identify a site of bleeding include intraoperative colonoscopy, esophagogastroduodenoscopy and enteroscopy. Patients should be placed in the modified lithotomy position to facilitate intraoperative colonoscopy. With intraoperative enteroscopy the small bowel may be fed over a scope. Care must be taken to avoid tearing of the mesentery. Transillumination of the bowel wall with dimmed operating room lights may help to identify vascular malformations in the bowel wall.(4) Remzi et al. recently reported on the use of highly selective angiography with methylene blue injection to localize an occult small bowel bleeding site in a patient with Crohn's disease.(33)

Outcome: Segmental colon resection following localization of a bleeding site has been associated with rebleeding rates of 0 to 15 percent and mortality rates of 0 to 13 percent. Small studies looking at outcome following subtotal colectomy for management of lower gastrointestinal bleeding reported rebleeding rates from 0 percent in most series to 60 percent in one series. The mortality rates ranged from 0 to 40 percent. (16)

Summary

When a patient presents with acute lower gastrointestinal bleeding successful treatment is dependent on initial aggressive resuscitation, appropriate interventions to identify and possible treat the site of bleeding, and surgical treatment based on the findings of diagnostic testing. Colonoscopy following colonic lavage is the initial diagnostic procedure of choice in a stable patient. If colonoscopy in not feasible or is nondiagnostic 99Tcm red blood cell scan or selective angiography may be performed depending on whether the patient has apparent minimal to moderate bleeding versus massive hemorrhage. A segmental colon resection is appropriate when a bleeding source has been identified. If the bleeding site cannot be located then subtotal colectomy is indicated.

References

  1. Vernava AM I, Longo WE, Virgo KS JF. A nationwide study of the incidence and etiology of lower gastrointestinal bleeding. Surg Res Commun 1996; 18:113-120.
  2. Sebastian Domingo JJ, Santos CL, Castellanos FD, Perez dA, V, Menchen P, Senent C et al. [Angiodysplasia of the colon. Experience in 60 cases]. Rev Esp Enferm Dig 1990; 77(2):125-128.
  3. Rossini FP, Ferrari A. [Emergency colonoscopy in the diagnosis of enterorrhagias]. Minerva Dietol Gastroenterol 1979; 25(4):369-372.
  4. Church JM. Vascular Malformations: Afteriovenous Malformations, Hmangiomas, and Varices. Endoscopy of the Colon, Rectum, and Anus. New York: Igaku-Shoin, 1995: 379-397.
  5. Gennaro AR, Rosemond GP. Colonic diverticula and hemorrhage. Dis Colon Rectum 1973; 16(5):409-415.
  6. McGuire HH, Jr., Haynes BW, Jr. Massive hemorrhage for diverticulosis of the colon: guidelines for therapy based on bleeding patterns observed in fifty cases. Ann Surg 1972; 175(6):847-855.
  7. Casarella WJ, Galloway SJ, Taxin RN, Follett DA, Pollock EJ, Seaman WB. "Lower" gastrointestinal tract hemorrhage: new concepts based on arteriography. Am J Roentgenol Radium Ther Nucl Med 1974; 121(2):357-368.
  8. Wilcox MC, Alexander LN, Cotsonis GA, Clark SW. Nonsteroidal antiinflammatory drugs are associated with both uper and lower gastrintestinal bleeding. Dig Dis Sci 1997; 42(5):990-997.
  9. Robert JH, Sachar DB, Aufses AH, Jr., Greenstein AJ. Management of severe hemorrhage in ulcerative colitis. Am J Surg 1990; 159(6):550-555.
  10. Robert JH, Sachar DB, Greenstein AJ. Severe gastrointestinal hemorrhage in Crohn's disease. Annals of Surgery 1991; 213:207-211.
  11. Kaassis M, Oberti E, Burtin P, Boyer J. Argon plasma coagulation for the treatment of hemorrhagic radiation proctitis. Endoscopy 2000; 32(9):673-676.
  12. Venkatesh KS, Ramanujam P. Endoscopic therapy for radiation proctitis-induced hemorrhage in patients with prostatic carcinoma using argon plasma coagulator application. Surg Endosc 2002; 16(4):707-710.
  13. Saclarides TJ, King DG, Franklin JL DA. Formalin Instillation for Refractory Radiation -Induced Hemorrhagic Proctitis: Report of 16 Patients. Dis Colon Rectum 1996; 39(2):196-199.
  14. Luna-Perez P, Rodriguez-Ramirez SE. Formalin instillation for refractory radiation-induced hemorrhagic proctitis. J Surg Oncol 2002; 80(1):41-44.
  15. Jao SW, Beart RW Jr, Gunderson LL. Surgical treatment of radiation injuries of the colon and rectum. Dis Colon Rectum 1986; 151(2):272-277.
  16. Vernava AM, III, Moore BA, Longo WE, Johnson FE. Lower gastrointestinal bleeding. Dis Colon Rectum 1997; 40(7):846-858.
  17. Reinus JF, Brandt LJ. Vascular ectasias and diverticulosis. Common causes of lower intestinal bleeding. Gastroenterol Clin North Am 1994; 23(1):1-20.
  18. Boley SJ, DiBiase A, Brandt LJ, Sammartano RJ. Lower intestinal bleeding in the elderly. Am J Surg 1979; 137(1):57-64.
  19. Chawla Y, Dilawari JB. Anorectal varices - their frequency in cirrhotic and noncirrhotic portal hypertension. Gut 1991; 32:309-311.
  20. Hosking SW, Johnson AG. Bleeding anorectal varices--a misunderstood condition. Surgery 1988; 104(1):70-73.
  21. Lewis B, Waye JD. Bleeding from the small intestine. In: Sugawa C, Schuman BM, Lucas CE, editors. Gastrointestinal bleeding. New York: Igaku-Shoin, 1992: 178-188.
  22. Naveau S, Aubert A, Poynard T, Chaput JC. Long-term results of treatment of vascular malformations of the gastrointestinal tract by neodymium YAG laser photocoagulation. Dig Dis Sci 1990; 35(7):821-826.
  23. Jensen DM MGJRKT. Urgent colonoscopy for the diagnosis and treatment of severe diverticular hemorrhage. New England Journal of Medicine 2000; 342(2):78-82.
  24. Berner JS, Mauer K, Lewis BS. Push and sonde enteroscopy for the diagnosis of obscure gastrointestinal bleeding. Am J Gastroenterol 1994; 89(12):2139-2142.
  25. Chong J, Tagle M, Barkin JS, Reiner DK. Small bowel push-type fiberoptic enteroscopy for patients with occult gastrointestinal bleeding or suspected small bowel pathology. Am J Gastroenterol 1994; 89(12):2143-2146.
  26. Ell C, Remke S, May A, Helou L, Henrich R, Mayer G. The first prospective controlled trial comparing wireless capsule endoscopy with push enteroscopy in chronic gastrointestinal bleeding. Endoscopy 2002; 34(9):685-689.
  27. Fireman Z, Mahajna E, Broide E, Shapiro M, Fich L, Sternberg A et al. Diagnosing small bowel Crohn's disease with wireless capsule endoscopy. Gut 2003; 52(3):390-392.
  28. Alavi A. Radionuclide localization of gastrointestinal hemorrhage. Radiology 1982; 142(3):801-803.
  29. Kester RR WJSJ. The 99mTc-labeled RBC scan. A diagnostic method for lower gastrointestinal bleeding. Dis Colon Rectum 1984; 27:47-52.
  30. Koval G, Benner KG, Rosch J, Kozak BE. Aggressive angiographic diagnosis in acute lower gastrointestinal hemorrhage. Dig Dis Sci 1987; 32(3):248-253.
  31. Molgaard CP. Mesenteric angiography for the diagnosis and treatment of lower gastrointestinal bleeding. Seminars in Colon & Rectal Surgery 1997; 8:164-171.
  32. Luchtefeld MA, Senagore AJ, Szomstein M, Fedeson B, Van Erp J, Rupp S. Evaluation of transarterial embolization for lower gastrointestinal bleeding. Dis Colon Rectum 2000; 43(4):532-534.
  33. Remzi FH, Dietz DW, Unal E, Levitin A, Sands MJ, Fazio VW. Combined use of preoperative provocative angiography and highly selective methylene blue injection to localize an occult small-bowel bleeding site in a patient with Crohn's disease: report of a case. Dis Colon Rectum 2003; 46(2):260-263.