John Roberts, MD, Editor

In the past year, major advances have been reported in almost every area of gastroenterology, including esophageal disease, gastric disease, lower gastrointestinal tract disease, and biliary tract disease.

Long-term therapy with proton-pump inhibitors has now been shown to control symptoms and maintain healing in patients with erosive esophagitis. Treatment regimens for Helicobacter pylori infection should achieve an eradication rate of at least 90%, and the search for simpler, effective regimens continues. It is well known that the risk for colorectal cancer is higher among patients whose first-degree relatives have had this type of cancer. A key study has shown that relatives of patients with adenomatous polyps are also at increased risk for developing colorectal cancer. Chronic unexplained diarrhea is common among patients with AIDS, and the best approach to diagnosis has been elusive. An important study has shown that clinical variables can help predict which patients with AIDS and diarrhea will benefit from endoscopic studies. Endoscopic biliary sphincterotomy is the procedure used most frequently to remove common bile duct stones, and more than 150 000 sphincterotomies are done each year in the United States. A seminal paper has demonstrated that complications of the procedure are related to several factors, including the technical skill of the endoscopist.

Esophageal Disease

In 1996, important findings in the area of esophageal disease included the further elucidation of the benefits and risks of proton-pump inhibitors; the discovery that some patients with achalasia benefit from injection of botulinum toxin; and the finding that the antiosteoporosis agent alendronate causes esophagitis, albeit rarely.

Lansoprazole Maintained Healing

Robinson M, Lanza F, Avner D, Haber M. Effective maintenance treatment of reflux esophagitis with low-dose lansoprazole. A randomized, double-blind, placebo-controlled trial. Ann Intern Med. 1996; 124:859-67.

Erosive esophagitis is a chronic disorder, and up to 80% of patients have relapse within 6 months after completing short-term therapy. Relapses can be treated as they occur. Maintenance therapy with various acid suppressants has been attempted, but the results have been disappointing. Robinson and colleagues sought to determine the effectiveness of the proton-pump inhibitor lansoprazole in preventing recurrent esophagitis.

In a clinical trial, 170 patients with endoscopically verified erosive esophagitis that healed after 8 weeks of acid suppression therapy were randomly assigned to receive either lansoprazole (15 or 30 mg/d) or placebo for as long as 12 months. Symptoms were evaluated and endoscopy was done 1, 2, 3, 6, 9, and 12 months after treatment. Endoscopy was also done if symptoms recurred.

One month after treatment cessation, 98% of the patients in the high-dose lansoprazole group and 93% of those in the low-dose group remained healed compared with 45% of patients in the placebo group. Symptom scores correlated well with endoscopic findings. At 12 months, 90% of the patients who received high-dose lansoprazole remained healed compared with 79% of patients who received low-dose lansoprazole and 24% of patients who received placebo. No serious adverse effects occurred.

Lansoprazole at a dosage of either 15 mg/d or 30 mg/d controls symptoms and maintains healing of erosive esophagitis. Whether such long-term treatment will result in fewer complications, such as Barrett esophagus, remains unknown.

Helicobacter pylori plus Omeprazole Created Risk for Atrophic Gastritis

Kuipers EJ, Lundell L, Klinkenberg-Knol EC, Havu N, Festen HP, Liedman B, et al. Atrophic gastritis and Helicobacter pylori infection in patients with reflux esophagitis treated with omeprazole or fundoplication. N Engl J Med. 1996; 334:1018-22.

Helicobacter pylori infection recasts the structure of the gastric mucosa, and the evidence is sufficient to declare that H. pylori has some role in the development of gastric cancer [1]. In addition, evidence suggests that long-term use (as long as 5 years) of proton-pump inhibitors results in atrophic gastritis, argyrophil hyperplasia, and elevated gastrin levels. These effects may present another risk for gastric cancer. Previous studies have suggested that acid suppression in patients infected with H. pylori accelerates the development of atrophic gastritis and argyrophil hyperplasia.

Kuipers and colleagues assessed the effect of omeprazole on the stomachs of patients with gastroesophageal reflux disease who were or were not infected with H. pylori. In a nonrandomized study, patients with endoscopically verified reflux disease underwent fundoplication (n = 72) or received omeprazole (n = 105), 20 to 40 mg/d. Helicobacter pylori status was determined, and patients were followed for an average of 5 years. No patient received treatment for H. pylori infection.

Among patients undergoing fundoplication, 31 (43%) were positive for H. pylori. One of these patients had atrophic gastritis at study onset, but no other patient developed this condition. Among patients receiving omeprazole, 59 (56%) were positive for H. pylori; 18 of the 59 developed atrophic gastritis. Two of the 46 patients not infected with H. pylori developed atrophic gastritis, but this outcome was not considered to represent a difference from the fundoplication group.

Long-term omeprazole treatment is associated with development of atrophic gastritis in patients positive for H. pylori. If clinicians are going to use long-term proton-pump inhibitor treatment for maintenance therapy in patients with gastroesophageal reflux disease, it would be prudent to determine H. pylori status. If the patients are infected, I would first prescribe treatment to eradicate H. pylori.

Botulinum Toxin Was Effective in Some Patients with Achalasia

Pasricha PJ, Rai R, Ravich WJ, Hendrix TR, Kalloo AN. Botulinum toxin for achalasia: long-term outcome and predictors of response. Gastroenterology. 1996; 110:1410-5.

Botulinum toxin, a potent inhibitor of acetylcholine release from nerve endings, can reduce smooth-muscle tone in the gastrointestinal tract. Injection of botulinum toxin into the lower esophageal sphincter provides effective short-term treatment of achalasia.

Pasricha and associates assessed long-term outcomes in patients with achalasia who were treated with botulinum toxin. Thirty-one patients with achalasia were followed for a median of 30 months. Twenty-eight patients responded initially, and 20 of these remained improved after 3 months. Almost all of these 20 responders had had relapse by 28 months, but 15 responded to reinjection. Older patients and those with vigorous achalasia were more likely to respond to treatment (in contrast to classic achalasia, vigorous achalasia is characterized by simultaneous contractions that are large in amplitude and repetitive, as in diffuse esophageal spasm). Duration of illness, previous esophageal dilatations, and baseline radiologic appearance did not predict treatment success.

Botulinum toxin therefore seems to be an effective treatment for achalasia in about two thirds of patients, with an average duration of response of about 16 months. Patient age and the type of achalasia influenced the likelihood of response. Studies that directly compare botulinum toxin with pneumatic dilatation and surgery are now needed to assess the long-term efficacy and cost-effectiveness of this treatment. In an elderly patient with comorbid disease in whom surgery presents substantial risk, however, treatment with botulinum toxin is reasonable.

Alendronate Caused Esophagitis, but Rarely

de Groen PC, Lubbe DF, Hirsch LJ, Daifotis A, Stephenson W, Freedholm D, et al. Esophagitis associated with the use of alendronate. N Engl J Med. 1996; 335:1016-21.

The bisphosphonate alendronate increases bone mineral density and prevents vertebral fractures in women with postmenopausal osteoporosis [2]. However, it has also been associated with the development of esophagitis and gastritis.

In this surveillance study of 475 000 patients worldwide, 199 patients had adverse effects related to the esophagus (number needed to treat [NNT] to cause an adverse effect, 2387). Fifty-one of the cases were classified as severe (NNT, 9313). Endoscopic findings included erosions, ulcerations, exudative inflammation, and thickening of the esophagus.

The chemical esophagitis associated with alendronate seems fairly uncommon. However, it is prudent to tell patients the following: Do not chew or suck the tablet, swallow the tablet with 6 to 8 ounces of water, do not lie down for at least 30 minutes after taking the medication, do not take alendronate at bedtime, and stop taking the drug if problems with swallowing or esophageal pain occur.

Gastric Disease

Developments in the field of gastric disease in 1996 included the following: The treatment regimens for H. pylori-related ulcers continued to become simpler, the clinical effects of over-the-counter antacid preparations were carefully tested, the risk for rebleeding after acute upper gastrointestinal hemorrhage was stratified to better predict which patients could be discharged early from the hospital, and it was found that neither the use of low-dose aspirin nor the use of warfarin confounded fecal occult blood testing.

Short-Term Treatment Cured Duodenal Ulcers

Labenz J, Idstrom JP, Tillenburg B, Peitz U, Adamek RJ, Borsch G. One-week low-dose triple therapy for Helicobacter pylori is sufficient for relief from symptoms and healing of duodenal ulcers. Aliment Pharmacol Ther. 1997; 11:89-93.

At least 75% of duodenal ulcers and gastric ulcers are caused by H. pylori. The recurrence rate of duodenal ulcer after eradication of H. pylori is only about 2% per year; this rate contrasts sharply with the 25% rate seen 10 years ago with histamine-2 (H₂) blockers. Some popular treatment regimens are shown in Table 1. Because most regimens are complicated and last at least 10 days, compliance remains a problem.

Labenz and colleagues therefore tested whether 1-week, low-dose triple therapy for H. pylori would heal duodenal ulcers and relieve symptoms. Fifty-nine outpatients who had duodenal ulcers and had positive results on rapid urease testing for H. pylori were enrolled in a clinical trial. All patients received omeprazole (20 mg twice daily), clarithromycin (250 mg twice daily), and metronidazole (400 mg twice daily) for 1 week. Patients were then randomly assigned to receive either omeprazole (20 mg twice daily) or placebo for another 3 weeks.

The overall H. pylori cure rate was 96% (95% CI, 87% to 100%); no difference was seen between the treatment groups. After 2 weeks, duodenal ulcer healing was confirmed by endoscopy in 91% (CI, 80% to 100%) of patients receiving omeprazole and in 76% (CI, 60% to 91%) of patients receiving placebo (P = 0.14). All ulcers had healed by 4 weeks. Relief of dyspeptic symptoms and adverse events did not differ between the treatment groups.

One week of triple therapy with omeprazole, clarithromycin, and metronidazole at low and simple doses is highly effective in curing H. pylori-related duodenal ulcer. The continuation of acid suppression therapy beyond the first week may not be necessary. An earlier, single-arm study [3] showed similar cure rates with a 4-day course of quadruple therapy, although the issue of resistance to metronidazole was unanswered. Therefore, I recommend that these patients be treated with anti-H. pylori therapy for at least 1 week.

Histamine-2 Blockers: Slower Onset, Longer Duration

Feldman M. Comparison of the effects of over-the-counter famotidine and calcium carbonate antacid on postprandial gastric acid. A randomized controlled trial. JAMA. 1996; 275:1428-31.

Americans are being inundated with advertisements for over-the-counter H₂ blockers. Pharmaceutical companies are spending $500 million a year in advertising for these products, but are they actually superior to other antacid products?

Feldman compared the onset of action, potency, and duration of effect of famotidine with those of calcium carbonate. Eighteen healthy volunteers received three regimens (placebo, calcium carbonate [4000 mg], or famotidine [10 mg]) in random order, with approximately 1 week between studies. All agents were given 1 hour after the participants consumed a test meal. Two identical meals were consumed 2.5 and 6 hours after the initial medication was given. Gastric acid secretion and acid neutralization were measured by titration of intragastric acid.

Calcium carbonate had a rapid onset of action (30 minutes), but the effect lasted only 60 minutes. Famotidine had a delayed onset of action (90 minutes) but a more prolonged duration of effect (at least 9 hours). Peak effect of famotidine occurred at 3.5 hours. Peak potencies of famotidine and calcium carbonate were similar.

The author suggests that the most effective treatment would be a combination of an antacid with an H₂ blocker. Cost-effectiveness, however, remains an issue: Histamine-2 blockers, especially those sold over the counter, are very expensive, although the price is decreasing. More studies on these types of treatments for common dyspeptic symptoms are now being conducted.

Low Risks for Repeat Episodes of Gastrointestinal Bleeding Were Described

Rockall TA, Logan RF, Devlin HB, Northfield TC. Selection of patients for early discharge or outpatient care after acute upper gastrointestinal haemorrhage. National Audit of Acute Upper Gastrointestinal Haemorrhage. Lancet. 1996; 347:1138-40.

Acute upper gastrointestinal hemorrhage is a common medical emergency that develops with various degrees of severity. Patients who arrive at the emergency department vomiting blood are generally hospitalized until the risk for further bleeding is defined.

This study was undertaken to identify patients at low risk for further bleeding in order to avoid unnecessary hospitalizations. In a British national audit, Rockall and colleagues used a previously validated scoring system to analyze 2531 patients with acute upper gastrointestinal hemorrhage who were admitted after presentation to the emergency department and who underwent diagnostic esophagogastroduodenoscopy. Point scores were assigned according to age, presence of shock, comorbid conditions, diagnosis, and esophagogastroduodenoscopic criteria. Specific scores are listed in Table 2, and results of the study are summarized in Table 3.

A low-risk score identifies many patients with acute upper gastrointestinal hemorrhage who have low risk for rebleeding and death. Early endoscopy-possibly done in or from the emergency department-and rapid discharge of these patients may provide opportunities for considerable cost savings.

Aspirin Did Not Confound Occult Blood Tests

Greenberg PD, Cello JP, Rockey DC. Asymptomatic chronic gastrointestinal blood loss in patients taking aspirin or warfarin for cardiovascular disease. Am J Med. 1996; 100:598-604.

Upper gastrointestinal bleeding occurs more commonly in patients who take aspirin. No clinically useful dose of aspirin seems free from the risk for bleeding. As little as 37 mg of aspirin, for example, can cause gastrointestinal bleeding. Greenberg and colleagues sought to determine whether low-dose aspirin or warfarin causes a substantial degree of occult gastrointestinal blood loss and, if so, whether this loss affects occult stool blood testing.

The researchers enrolled 100 patients older than 40 years of age into a prospective crossover study. All patients underwent baseline quantitative stool blood testing and were then assigned to one of three groups. Patients who were not using aspirin were asked to take regular aspirin, 325 mg/d, for 2 months (group 1). Patients initially taking aspirin, 325 mg/d, were switched to 8 mg/d for 2 months and vice versa (group 2). Patients in group 3 took warfarin only for 4 to 6 weeks. Loss of blood in stools was measured at the end of the study.

Results are summarized in Table 4. The rate of hemoccult-positive results on stool tests did not increase over normal with either aspirin or warfarin therapy. Standard guaiac-based stool tests reliably detect blood at a level of about 3 mg of hemoglobin/g of stool [4].

Aspirin caused a small but clinically insignificant increase in occult fecal blood. Warfarin did not seem to increase blood loss, and the small amount of blood that was lost is unlikely to interfere with fecal occult blood tests. Positive results on fecal occult blood tests in patients taking either low-dose aspirin or warfarin should not be assumed to be false positive, and these patients should be managed in the same manner as those not taking these medications. Other studies have shown that aspirin or warfarin tends to unmask lesions present in the lower gastrointestinal tract.

Famotidine Prevented NSAID-Related Ulcers

Taha AS, Hudson N, Hawkey CJ, Swannell AJ, Trye PN, Cottrell J, et al. Famotidine for the prevention of gastric and duodenal ulcers caused by nonsteroidal antiinflammatory drugs. N Engl J Med. 1996; 334:1435-9.

In the general population, therapy with a nonsteroidal anti-inflammatory drug (NSAID) is associated with about a 1% incidence of gastrointestinal bleeding. In patients who are older than 60 years of age or have previously had a gastrointestinal bleeding episode or ulcer, the incidence increases to 3% to 4%. Among patients who have all three of these risk factors and have cardiovascular disease, the incidence increases to 9%. Misoprostol has been shown to provide effective prevention, but this drug has numerous adverse effects [5]. Ranitidine has been shown to effectively prevent duodenal ulcers but not gastric ulcers. [6]

Taha and colleagues sought to determine whether famotidine prevents peptic ulcers (both gastric and duodenal) in patients receiving long-term NSAID therapy for arthritis. In a clinical trial, 285 patients with either rheumatoid arthritis (82%) or osteoarthritis (18%) who were taking NSAIDs were randomly assigned to one of three groups: famotidine, 20 mg twice daily; famotidine, 40 mg twice daily; or placebo. Patients underwent interviews and endoscopy at baseline and after 4, 12, and 24 weeks of therapy. The primary end point was the cumulative incidence of gastric or duodenal ulceration at 24 weeks. Results are shown in Table 5.

High doses of famotidine were effective in preventing both gastric and duodenal ulcers in patients with arthritis who were receiving long-term NSAID treatment. A trial of famotidine would seem to be helpful for patients who must take NSAIDs, especially if they have had dyspepsia. Whether long-term treatment with other H₂ blockers in high doses or with proton-pump inhibitors will be equally effective is still unknown.

Scintigraphy Was Sensitive for Gastrinomas

Gibril F, Reynolds JC, Doppman JL, Chen CC, Venzon DJ, Termanini B, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas. A prospective study. Ann Intern Med. 1996; 125:26-34.

Many tumors-pancreatic, endocrine, and carcinoid tumors; lymphomas; and meningiomas-have a high density of somatostatin receptors and thus can be imaged by using somatostatin receptor scintigraphy. Gastrinomas, which cause the Zollinger-Ellison syndrome, also have high concentrations of somatostatin receptors.

This group of researchers from the National Institutes of Health compared somatostatin receptor scintigraphy with other imaging methods for the localization of gastrinomas in patients with the Zollinger-Ellison syndrome.

Eighty consecutive patients with the Zollinger-Ellison syndrome underwent ultrasonography, computed tomography, magnetic resonance imaging, angiography, bone scanning, and somatostatin receptor scintigraphy.

Gastrinomas were identified by ultrasonography in 19% of patients, by computed tomography in 38%, by magnetic resonance imaging in 45%, by angiography in 40%, and by somatostatin receptor scintigraphy in 70%.

Somatostatin receptor scintigraphy is the most sensitive method for imaging either primary or metastatic lesions in patients with the Zollinger-Ellison syndrome. It should probably be the first imaging procedure used in patients suspected of having the syndrome. In a follow-up study [7], the same researchers found that somatostatin receptor scintigraphy altered management in 57 of 122 patients with the Zollinger-Ellison syndrome. It altered the preoperative plan in about half of the cases, confirmed a cure after surgery in 4 of 22 patients, and indicated that 24 of 50 patients were not cured after surgery. Somatostatin receptor scintigraphy also clarified suspicious lesions detected by computed tomography in 6 of 12 patients and identified metastatic liver disease in 15 of 25 patients who had otherwise normal results on imaging studies.

Lower Gastrointestinal Tract Disease

In 1996, researchers found that adenomatous polyps predispose close relatives to a higher risk for colorectal cancer. Other notable findings included the discovery of two independent risk factors that predicted a poor prognosis in ulcerative colitis and the finding that the cause of HIV-related diarrhea was more often diagnosed among relatively sicker patients by use of two specific clinical variables.

Polyps Were Associated with Higher Family Risk for Cancer

Winawer SJ, Zauber AG, Gerdes H, O'Brien MJ, Gottlieb LS, Sternberg SS, et al. Risk of colorectal cancer in the families of patients with adenomatous polyps. National Polyp Study Workgroup. N Engl J Med. 1996; 334:82-7.

It is well known that the risk for colorectal cancer is higher among patients whose first-degree relatives have had colorectal cancer. Adenomatous polyps in patients themselves are another risk factor. Winawer and colleagues sought to determine the magnitude of the risk for colorectal cancer among family members of patients who had colorectal adenomatous polyps removed and to identify characteristics associated with that risk.

A random sample of 1199 National Polyp Study participants who had been found to have adenomatous polyps were interviewed for information on colorectal cancer in their parents and siblings. About 14% of these persons were excluded for various reasons. Analysis of the 1031 patient interviews showed that the age-adjusted relative risk for colorectal cancer in first-degree relatives was 1.78 (CI, 1.18 to 2.67). The relative risk in siblings of patients whose polyps were found before 60 years of age was 2.59 (CI, 1.46 to 4.58), and risk tended to increase inversely with the age of the index patient in whom the polyp was found.

First-degree relatives of patients with adenomatous polyps are at increased risk for colorectal cancer, especially if the polyp is found before 60 years of age.

Stool Frequency Predicted Outcomes of Ulcerative Colitis

Travis SP, Farrant JM, Ricketts C, Nolan DJ, Mortensen NM, Kettlewell MG, et al. Predicting outcome in severe ulcerative colitis. Gut. 1996; 38:905-10.

Table 6 lists the drugs commonly used against ulcerative colitis and Crohn disease. Sulfasalazine remains an important standby, but the adverse effects, including headache, nausea, azoospermia, hemolytic anemia, and pancreatitis, can be severe. The newer 5-aminosalicylic acid analogues (mesalamine products) can be given by enema, by suppository, and orally. The dose of these compounds needs to be higher in patients with Crohn disease and patients with very active ulcerative colitis. Corticosteroids are often used to treat these diseases, but they are often used incorrectly (doses that are too high, length of treatment that is too short, and doses that are tapered too rapidly). The usual dose in initiating therapy is 1 mg/kg of body weight.

Metronidazole has been found to be effective maintenance therapy for Crohn disease, particularly if coupled with a 5-aminosalicylic acid analogue. Immunosuppressive drugs, such as azathioprine and 6-mercaptopurine, are also effective in Crohn disease. The potential risk for cancer in these patients, who are receiving long-term immunosuppressive therapy, does not seem great. Cyclosporine has been disappointing in the treatment of inflammatory bowel disease, especially as maintenance therapy.

Despite therapeutic advances, it remains difficult to predict at an early stage which patients with severe ulcerative colitis will respond poorly to intensive medical therapy and will require colectomy. Travis and colleagues assessed whether early patterns or change in inflammatory markers or other variables could predict the need for surgery.

For 51 consecutive episodes of severe colitis in 49 patients, 100 mg of hydrocortisone was administered intravenously every 6 hours and 100 mg was administered rectally twice daily. Treatment was continued for 5 to 7 days until it was clear that the patient had responded or that colectomy was needed.

Complete response occurred in 21 patients. Stool frequency and C-reactive protein levels correlated strongly with the need for colectomy. Colectomy was required in 85% of patients who had more than 8 stools on day 3 or who had between 3 and 8 stools and a C-reactive protein level greater than 45 mg/L. After 7 days of treatment, patients with more than 3 stools a day or visible blood had a 60% chance of continuous symptoms and a 40% chance of colectomy in the following months.

Approach to HIV-Related Diarrhea Was Described

Wilcox CM, Schwartz DA, Cotsonis G, Thompson SE 3d. Chronic unexplained diarrhea in human immunodeficiency virus infection: determination of the best diagnostic approach. Gastroenterology. 1996; 110:30-7.

Chronic, unexplained diarrhea is a common complication of HIV infection, but the best diagnostic approach is unknown. Wilcox and colleagues evaluated the diagnostic yield of upper and lower gastrointestinal endoscopy in HIV-positive patients with chronic diarrhea. They sought to determine the clinical features of patients in whom endoscopy may be diagnostic and to evaluate long-term outcomes and response to currently available therapy.

Forty-eight consecutive HIV-infected patients who had chronic diarrhea underwent upper endoscopy to the jejunum and colonoscopy with biopsy. Probable causes of diarrhea were found in 21 patients (44% [CI, 30% to 58%]). Colonoscopy identified a cause in 13 patients, most commonly cytomegalovirus colitis. Upper gastrointestinal endoscopy with biopsy identified microsporidiosis in 7 patients and cryptosporidiosis in 2. Weight loss and duration of diarrhea were independent indicators that a diagnosis would be made. Patients with CD4 lymphocyte counts less than 100 cells/mm³ were more likely to receive a diagnosis. Median survival was 8.3 months (CI, 3.9 to 9.5 months) in patients with a diagnosis and 14.2 months (CI, 9.6 to 43.1 months) in those without a diagnosis. Of the 25 patients without a diagnosis who could be followed, diarrhea improved with treatment of symptoms in 9 (38%).

Clinical variables helped predict which HIV-positive patients with diarrhea would benefit from endoscopy. These patients tended to be have more severe HIV disease. Patients without weight loss and without markedly decreased CD4 lymphocyte counts are unlikely to benefit from endoscopy but are likely to benefit from treatment of the symptoms of diarrhea.

Biliary Tract Disease

Endoscopic biliary sphincterotomy has become very common, and the risk factors for complications were described in 1996. In addition, a European group reported the clinical factors associated with poor outcomes in primary sclerosing cholangitis.

Risk for Complications of Sphincterotomy Was 9.8%

Freeman ML, Nelson DB, Sherman S, Haber GB, Herman ME, Dorsher PJ, et al. Complications of endoscopic sphincterotomy. N Engl J Med. 1996; 335:909-18.

Endoscopic biliary sphincterotomy has become the procedure most frequently used to remove common bile duct stones. It is also used to treat various causes of strictures in the region of the sphincter of Oddi. Approximately 150 000 such sphincterotomies are performed each year in the United States. Freeman and colleagues sought to identify risk factors for complications of biliary sphincterotomy and the outcomes of these complications.

In a 3-year, 17-center surveillance study, 2347 patients undergoing endoscopic biliary sphincterotomy were followed for complications within 30 days after the procedure. Complications occurred in 9.8% of patients; the most frequent complications were pancreatitis (5.4%) and hemorrhage (2.0%). Death related to the procedure occurred in 0.4% of patients. Multivariate analysis revealed five independent risk factors for complications, as shown in Table 7. Endoscopists who performed more than one sphincterotomy per week had lower complication rates (8.4% compared with 11.1% of those who performed less than one sphincterotomy per week).

Complications of endoscopic biliary sphincterotomy seem to be related both to the clinical indication for the procedure (as suggested by the need for precut sphincterotomy, combined procedures, and suspected sphincter dysfunction) and the technical skill of the endoscopist.

Prognosis of Primary Sclerosing Cholangitis Was Described

Broome U, Olsson L, Loof L, Bodemar G, Hultcrantz R, Danielsson A, et al. Natural history and prognostic factors in 305 Swedish patients with primary sclerosing cholangitis. Gut. 1996; 38:610-5.

Primary sclerosing cholangitis is a chronic, destructive biliary disease with a highly variable natural history. Broome and colleagues described the natural history and outcome of 305 patients with primary sclerosing cholangitis and evaluated the prognostic significance of clinical, biochemical, and histologic findings at the time of diagnosis.

At the time of diagnosis, 134 (44%) of the patients were asymptomatic. In these patients, the diagnostic workup had typically been initiated because an elevated alkaline phosphatase level had been found on routine laboratory studies. However, 29 of these 134 patients (22%) became symptomatic during the 5-year follow-up. For all patients, the median duration of survival from time of diagnosis to death or liver transplantation was 12 years. Cholangiocarcinoma developed in 8% of patients. Patient age, serum bilirubin level, and histologic stage at time of diagnosis were independent prognostic factors. A serum bilirubin level greater than 1.2 mg/dL suggested that the disease was in an accelerated phase and that the patient would require transplantation within 6 years.

Dr. Roberts (Series Editor): Madrona Medical Group, 4465 Cordata Parkway, Bellingham, WA 98226.