Colonic angiodysplasia.

Colonic angiodysplasia (AD) is an important vascular lesion responsible for approximately 6.0% of cases of lower gastrointestinal hemorrhage. Lesions are usually located in the right colon and, although the pathophysiology is unknown, most are probably acquired as the result of a degenerative process associated with aging. Diagnosis is usually made during colonoscopy, but angiography can be efficacious when hemorrhage is severe. Most patients with bleeding AD are treated by endoscopy. Various methods have been used (monopolar electrocoagulation, injection therapy, contact probes, and lasers) with acceptable safety and success. perforation of the right colon is the main concern, especially with monopolar electrocoagulation and the YAG laser. In clinical practice, contact probes are used most often, although this approach has not been well studied as treatment for colonic AD.

The American College of Gastroenterology Bleeding Registry: preliminary findings.

OBJECTIVES: The American College of Gastroenterology (ACG) Institute for Clinical Research and Education conducted a survey study to assess demographics, management strategies, and outcome for patients with gastrointestinal bleeding. This pilot project was intended to determine the feasibility of surveying the ACG membership about common clinical issues. METHODS: Color-coded survey forms were sent to all ACG members and Fellows, with instructions to supply information about demographics, presenting symptoms, management, and outcome for bleeding patients and procedure-matched controls. Forms returned between June 1 and August 31, 1995, were tabulated and analyzed for differences between the bleeding group and procedure-matched controls. RESULTS: A total of 1235 forms were returned by respondents, 60% of whom were in private practice. Patient demographics indicated that bleeding patients were significantly older, more likely to be male, and more likely to use alcohol, tobacco, and prescription or over-the-counter aspirin or nonsteroidal anti-inflammatory drugs and anticoagulants than were controls. Upper GI bleeding accounted for 76% of bleeding events, with duodenal and gastric ulcers being the source in more than 50% of the upper GI bleeders. Diverticula was the most common bleeding source identified in lower GI bleeders. In the bleeding group, 78.8% were anemic, with 60.9% having hemoglobin of <10 g/dl; 31% presented with orthostatic changes in blood pressure or shock. Most bleeding subjects, regardless of source, were hospitalized, 58.2% received blood transfusions, and 45.5% received endoscopic therapy. Rebleeding (11.2%), need for surgery (7.1%), and fatalities (2.1%) were uncommon. Over-the-counter aspirin and nonsteroidal anti-inflammatory drugs were used significantly more often in the bleeding population (47.6%) than in controls (19.4%). CONCLUSIONS: The success of the GI Bleeding Registry supports the feasibility of surveying ACG members about common clinical problems. Data suggest that ACG members manage sick patients with severe gastrointestinal bleeding who require hospitalization, transfusions, and endoscopic treatment. These preliminary results will serve as an impetus to conduct further survey studies of gastrointestinal bleeding and other common digestive disease conditions.

Diverticular bleeding and the pigmented protuberance (sentinel clot): clinical implications, histopathological correlation, and results of endoscopic intervention.

UNLABELLED: Over a 3-yr period, we performed colonoscopy on five patients (mean age 71 yr) in whom a specific diverticulum that contained a pigmented protuberance (PiP) was unequivocally identified as the cause for hemorrhage. Four of these individuals had endoscopic bipolar cauterization of the PiP, and two patients had surgery. AIM: To (1) determine the clinical significance of an intradiverticular PiP, (2) correlate endoscopic features of a PiP with histopathological findings and, (3) assess results of endoscopic treatment for affected patients with lower GI bleeding. METHODS: Medical charts were reviewed to garner data and assess outcome for patients. The bleeding diverticulum in one surgical case was detected in the resected specimen, and histological examination was performed. RESULTS: The mean number of transfusions, bleeding days, and lowest hemoglobin concentration before definitive treatment was 4.8, 3.4, and 8.4 gm%, respectively. All patients rebled before treatment. In all instances, the PiP projected through the neck of the diverticulum, which was actively bleeding in two patients. Endoscopic bipolar cautery directed at the PiP achieved permanent hemostasis in three of four subjects (75%) (morbidity 0%). Endoscopic therapy failed in one subject, and a hemicolectomy was performed. Histological evaluation of the resected specimen showed erosion of a medium sized artery into the diverticulum. The PiP represented a sentinel clot (not a visible vessel) adherent to a breach in the vessel wall. A patient who had surgery instead of endoscopic therapy had a prolonged, complicated postoperative course. CONCLUSIONS: (1) The presence of an intradiverticular PiP may identify a subset of patients at risk for severe recurrent diverticular bleeding. (2) Histopathological analysis showed the PiP to be a sentinel clot rather than a visible vessel. (3) In patients with severe recurrent diverticular bleeding, endoscopic treatment of the vessel beneath this lesion may be a viable alternative to surgery.

Diverticular bleeding: are nonsteroidal anti-inflammatory drugs risk factors for hemorrhage and can colonoscopy predict outcome for patients?

OBJECTIVE: This retrospective study was performed to determine if certain endoscopic features of a bleeding diverticulum predict outcome for patients and to assess the role of nonsteroidal anti-inflammatory drugs (NSAID) as a risk factor for hemorrhage. METHODS: Over a 28-month period, colonoscopy was performed on 13 patients (mean, age 74 yr) in whom a specific diverticulum was unequivocally identified as a cause for bleeding. Endoscopic features of the affected diverticulum were recorded and correlated with outcome for patients. Drug histories were reviewed to document use of NSAID before bleeding. RESULTS: Three patients had a visible vessel located inside a diverticulum, and one subject had an adherent clot with active bleeding. These colonoscopic findings were classified as stigmata of significant hemorrhage (SSH). In the remaining nine patients the diverticula were ulcerated. This endoscopic finding was classified as stigmata of insignificant hemorrhage (SIH). Compared with patients with SIH, individuals with SSH experienced a greater number of bleeding episodes (3.5 vs 1.3, p = 0.006), had a lower initial hemoglobin concentration (8.2 vs 12.5 gm%, p = 0.009), and required more transfusions (3.3 vs 0, p = 0.04) and invasive treatments (75% with SSH were managed by endoscopy or surgery vs 0% for those with SIH, p = 0.01). Ninety-two percent of the patients were taking NSAID (100% with SSH and 89% with SIH). Seventy-five percent of subjects with SSH compared with 0% of patients with SIH had a combined exposure to NSAID and ASA (p = 0.01). CONCLUSIONS: Presence of a visible vessel or an adherent clot with active bleeding is a reliable marker for significant hemorrhage. Ulcerated diverticula are the cause of trivial bleeding, and presence of this endoscopic finding accurately predicts a benign clinical course. NSAID may be an important risk factor for diverticular bleeding. It is possible that combined exposure to NSAID and ASA results in more severe bleeding compared with use of NSAID alone.

Prevalence and natural history of colonic angiodysplasia among healthy asymptomatic people.

OBJECTIVE: The prevalence of colonic angiodysplasia (AD) among healthy asymptomatic people is unknown, and the natural history of these lesions has not been clearly defined. The purpose of our study was to determine prevalence and to review and assess the natural history of AD. METHODS: Each of the authors had previously published his own prospective study that involved screening colonoscopy for the detection of neoplasia in asymptomatic adult men and women who had never bled. All pathological lesions diagnosed by endoscopy were recorded, but only data pertaining to polyps were published. We pooled and analyzed the endoscopic findings (raw data) from those studies and assessed the natural history of the lesions by reviewing the medical charts of affected subjects to determine whether bleeding had occurred. Nine hundred sixty-four patients were evaluated (mean age, 62 yr). All were > or = 50 yr old (range, 50-79 yr), not anemic, Hemoccult negative, asymptomatic, and had full colonoscopy to the cecum. RESULTS: Eight subjects had AD (prevalence, 0.83%). Lesions in affected individuals were usually small (mean size, 4.0 mm) and most often were located in the right colon (62%). No subject bled, and all maintained a stable hemoglobin (mean, 14.6 g%) with a mean follow-up of 3 yr. CONCLUSIONS: We conclude that 1) colonic AD is uncommon among healthy asymptomatic people (prevalence, 0.83%), 2) lesions are usually small (< 10 mm) and are located proximal to the hepatic flexure, 3) the natural history for AD in these people is benign, and the risk of bleeding over a 3-yr period is low (0% in 3 yr), and 4) because of this low risk, endoscopic treatment for incidental (nonbleeding) AD is unnecessary.

A prospective assessment of results for needle-knife papillotomy and standard endoscopic sphincterotomy.

Results from studies evaluating needle-knife papillotomy are conflicting. The aim of this prospective trial was to assess the safety and efficacy of needle-knife papillotomy and compare the results with those of standard endoscopic sphincterotomy. During a period of 6 years, ERCP was performed 1680 times. Four hundred fifty-six patients had biliary obstruction and were candidates for endoscopic sphincterotomy. Cholangiography was successful in 414 subjects (91%). These individuals underwent endoscopic sphincterotomy, which was successful in 395; overall success rate for endoscopic sphincterotomy was 87% (395 of 456). Forty complications occurred (9.7%), but only 21 (5.1%) were serious. No procedure-related mortality occurred. Fifty-two of 456 patients (11%) were managed by needle-knife papillotomy. In 42 of these subjects, cholangiograms had failed or were incomplete, and endoscopic sphincterotomy was not attempted; in 10 individuals endoscopic sphincterotomy had failed. Needle-knife papillotomy was successful in 47 of 52 patients (90%). Three complications occurred (5.7%), only 2 of which (3.8%) were serious. Morbidity rates for needle-knife papillotomy and endoscopic sphincterotomy were not significantly different. No procedure-related mortality occurred. Needle-knife papillotomy increased the overall success rate for endoscopic management with sphincterotomy from 87% (success rate for endoscopic sphincterotomy alone) to 97% (success rate for endoscopic sphincterotomy and needle-knife papillotomy).(ABSTRACT TRUNCATED AT 250 WORDS)

A diagnostic approach to pancreatic cancer.

The diagnostic and staging evaluation for patients suspected of having pancreatic cancer has not been standardized. This review describes a practical algorithm for the workup of affected individuals. Computed tomography as an initial test can usually provide the most information when clinical features or biochemical or serological (CA19-9) test results suggest the presence of a neoplasm. Endoscopic retrograde cholangiopancreatography is appropriate when scans are equivocal or clearly abnormal especially in jaundiced patients. A cytological diagnosis is possible and biliary decompression can be accomplished with this procedure. Laparoscopy followed by angiography is useful for staging surgical candidates.

Angiodysplasia of the gastrointestinal tract.

Angiodysplasia is an important vascular lesion of the gut and a source of significant morbidity from bleeding. This lesion is probably responsible for approximately 6.0% of cases of lower gastrointestinal (GI) bleeding and 1.2-8.0% of cases of hemorrhage from the upper GI tract. Small bowel angiodysplasia accounts for 30-40% of cases of GI bleeding of obscure origin and represents the single most common cause for hemorrhage in this subset of patients. Lesions in the large bowel occur most often in the right colon. Their cause is unknown but most are probably acquired and the result of a degenerative process associated with aging. The incidence of colonic angiodysplasia among strictly asymptomatic individuals has never been determined and the natural history for these lesions is incompletely understood. Angiodysplasia in the upper GI tract occurs most often in the stomach and duodenum. When affected patients have been evaluated by colonoscopy concomitant lesions have been diagnosed in one-third of instances. Angiodysplasia has been purported to occur with higher frequency in patients with renal failure, von Willebrand's disease, aortic stenosis, cirrhosis, and pulmonary disease. Not all of these associations have been subjected to critical analysis, but available evidence does not support a strong relationship in most instances. Patients with bleeding angiodysplasia are occasionally treated with hormones or, more often, by endoscopic methods. Uncontrolled case studies have reported reduction or cessation of bleeding in subjects managed with conjugated estrogens. However, prospective randomized controlled trials assessing the efficacy of hormonal therapy are limited, and results from two trials conflict. Safety profiles for the endoscopic methods are acceptable, and reported efficacies are high, although not all methods have been extensively evaluated specifically for the treatment of angiodysplasia. Perforation of the right colon is a potential problem, especially for monopolar electrocoagulation and lasers.

Endoscopic therapy for patients with a post-operative biliary leak.

Twenty-three patients with a post-operative biliary leak were treated by various endoscopic methods and results were analyzed. Leaks occurred at the cystic duct in 13 patients, at the common duct in 6 patients, and at an anomalous branch of the right hepatic duct in 4 patients. Treatments included sphincterotomy alone (4 patients), stent alone (6 patients), sphincterotomy and stent (12 patients), and sphincterotomy and nasobiliary drainage catheter (1 patient). Five patients also had supplemental percutaneous catheter drainage of a biloma. All treatments were completed successfully in the absence of major morbidity, and permanent closure of the leak occurred in 100% of cases. Endoscopic therapy for patients with a post-operative biliary leak is safe and effective and should be recommended before surgical re-exploration.

The sentinel hyperplastic polyp: a marker for synchronous neoplasia in the proximal colon.

We prospectively screened 129 asymptomatic subjects (mean age 64 yr) with flexible sigmoidoscopy. Colonoscopy was performed at a later date, regardless of the sigmoidoscopic result. Our intent was 1) to establish the prevalence of proximal neoplasms in patients with and without hyperplastic polyps within reach of the 60-cm sigmoidoscope and 2) to determine whether a distal (sentinel) hyperplastic polyp predicts the presence of synchronous neoplastic polyps higher up in the colon. Our results show that 15% of asymptomatic adult subjects without polyps on sigmoidoscopy have adenomas in proximal colonic segments that can be diagnosed only by colonoscopy. By comparison, proximal neoplasms were detected in 32% (p less than 0.05) and 37% (p less than 0.05) of patients when hyperplastic or adenomatous polyps, respectively, were present on the sigmoidoscopic examination. This finding suggests that a distal (sentinel) hyperplastic polyp by itself may be a marker for neoplastic polyps in proximal colonic segments. Also, the "index" adenoma and "sentinel" hyperplastic polyp may be equivalent for predicting the presence of proximal neoplasms. The observed detection rates for these polyps were both significantly higher than expected when compared to patients who did not have polyps in the distal colon or rectum. If these results can be confirmed by a larger prospective trial, then full colonoscopy for detection of proximal neoplasms may be indicated when either an index adenoma or sentinel hyperplastic polyp is detected by sigmoidoscopy.

Prevalence and malignant potential of colorectal polyps in asymptomatic, average-risk men.

Screening colonoscopy was performed on 119 asymptomatic elderly men with no other risk factors for colonic neoplasms. Ninety adenomas were detected in 49 (41%) subjects. Mean adenoma size was 6.5 mm, with 83 (92%) less than or equal to 10 mm. Forty-nine (54%) adenomas were located proximal to the splenic flexure, and 17 (19%) were classified as tubulovillous or villous. Moderate- to high-grade dysplasia was found in 29 (32%) adenomas and was associated with larger size (p less than 0.0001) and villous architecture (p = 0.0038). Two subjects harbored adenomas with invasive cancer. Seventy-one hyperplastic polyps were found in 40 (34%) subjects. The mean size of hyperplastic polyps was 3.4 mm, and 64 (90%) were located in the left colon. We conclude that, in this population, adenomas are common lesions that are frequently small. However, substantial proportions of these adenomas may be at risk for malignant degeneration due to the presence of villous architecture and higher grades of dysplasia. Because adenomas and hyperplastic polyps are endoscopically indistinguishable, all polyps found at endoscopy should be removed or ablated.

Flexible sigmoidoscopy may be ineffective for secondary prevention of colorectal cancer in asymptomatic, average-risk men.

Asymptomatic men (N = 114) 50 years of age or older had screening for colorectal neoplasia with flexible sigmoidoscopy followed by colonoscopy regardless of the sigmoidoscopic result. Our study objective was to determine the prevalence of patients having isolated adenomatous polyps in a proximal colonic segment in the absence of a distal index neoplasm within reach of the sigmoidoscope. Through the combined use of sigmoidoscopy and colonoscopy, adenomatous polyps were detected in 47 of 114 individuals (41%). A total of 88 adenomas was found. Seventeen patients had isolated neoplasms in proximal colonic segments in the absence of distal adenomas. These patients represented 15% of screened subjects (17 of 114) and 20% of individuals who lacked adenomas on sigmoidoscopy (17 of 84). The majority of proximal neoplasms were small (less than 1.0 cm), tubular adenomas. Flexible sigmoidoscopy may be ineffective for screening asymptomatic men for neoplasia. However, it remains to be determined if a 20% miss rate (for those with a normal sigmoidoscopic examination) is significant and whether small proximal adenomas are worth finding.

A randomized comparison of multipolar electrocoagulation and injection sclerosis for the treatment of bleeding peptic ulcer.

Several options are available to the gastroenterologist for the endoscopic control of peptic ulcer hemorrhage. Sixty men (mean age, 62 years) were stratified into those with actively bleeding ulcers or ulcers with stigmata of recent hemorrhage, and then randomized to treatment with injection sclerosis or the multipolar probe until bleeding ceased. There were no significant differences between the randomized groups in terms of age, ulcer size, or transfusion requirements. We achieved hemostasis in 95% of all patients. Re-bleeding rates were 25% and 23% in patients treated with the multipolar probe and injection sclerosis, respectively. There was no significant difference in mortality (14% multipolar probe, 7% injection sclerosis). Patients over the age of 70, those whose onset of bleeding was while an inpatient, and those with an ulcer larger than 2 cm were more likely to have a poor outcome, regardless of the type of endoscopic therapy. Injection sclerosis and multipolar electrocoagulation are equally effective in controlling bleeding from peptic ulcer.

A prospective, controlled analysis of endoscopic cytotechniques for diagnosis of malignant biliary strictures.

We prospectively collected brushings and bile for cytology in 30 consecutive patients with bile duct strictures (17 malignant, 13 benign) who were assessed by endoscopic retrograde cholangiography. When appropriate, the cellular debris on stents that were removed from individuals who were managed with these devices was evaluated for malignant cells as well. Our aim was to assess the value of these endoscopic cytotechniques for making a diagnosis of obstructing cancer of the biliary tract. A cumulative total of 78 specimens were obtained. Overall, sensitivity was highest for stent (36%) and brush (33%) cytology, compared with results obtained from bile (6%). If the results for all methods are combined, 47% of patients with cancer (eight of 17) could be diagnosed by one or more cytological technique. There were no false-positive results (specificity, 100%). Our results show that brush and stent cytology are nearly equivalent for detecting cancer, but because a diagnosis is delayed until the endoprosthesis is removed (mean 3.4 months), the brush technique is preferred. Results for bile cytology are marginal. Specificity for these cytotechniques is high; therefore, a positive result by any method is sufficient evidence for cancer, and other invasive diagnostic procedures are unnecessary.

Push-enteroscopy for diagnosis of patients with gastrointestinal bleeding of obscure origin.

Push-enteroscopy using a disinfected colonoscope was performed on 39 patients with gastrointestinal bleeding of obscure origin. Our results show that: (1) A high percentage of patients (38%) have pathological lesions responsible for bleeding located in the distal duodenum and proximal jejunum, which are readily detected by push-enteroscopy. (2) Duodeno-jejunal arteriovenous malformations (AVMs) are the most common cause for bleeding, and these lesions can be conveniently cauterized through the endoscope. (3) An efficient sequence of steps for diagnosis of patients with this problem includes push-enteroscopy when the initial EGD and colonoscopy are normal followed by small bowel radiography. Mesenteric angiography and intraoperative enteroscopy can be reserved for patients with severe bleeding when push-enteroscopy and small bowel radiography are negative. We conclude that push-enteroscopy has an important role to play in the early assessment of patients with gastrointestinal bleeding of obscure origin.

Endoscopic retrograde wire-guided brush cytology for diagnosis of patients with malignant obstruction of the bile duct.

A brush was adapted for use over a guide wire to facilitate reliable acquisition of cytological specimens from tight and potentially malignant strictures encountered during ERCP. Cells collected in this manner were assessed for abnormalities indicative of cancer. We have called this technique endoscopic retrograde wire-guided brush cytology (ERWBC). Thirty-nine strictures (24 malignant, 15 benign) in 34 patients were brushed. There were no complications, and all specimens were adequate for evaluation. Sixty percent of patients with cancer were diagnosed by ERWBC. Sensitivity was highest for cholangiocarcinoma (100%), intermediate for pancreatic cancer (60%), and lowest for patients with biliary obstruction due to metastatic disease (22%). There were no false-positive results (specificity, 100%). The positive and negative predictive values were 100% and 58%, respectively, and accuracy for the test was 72%. Collection of cytological specimens using a brush with a wire guide is effective especially for diagnosis of cholangiocarcinoma. A positive result is sufficient evidence for malignancy, and other invasive diagnostic tests are unnecessary. We recommend ERWBC for brushing all strictures of unknown cause during ERCP in an effort to make a diagnosis of cancer.