Direct contact dissolution using MTBE is a safe, effective treatment for many patients with cholesterol gallstones when performed by clinicians experienced with this technique. This may be the treatment of choice for many patients at high risk for general anaesthesia or surgery. Cholecystostomy using local anaesthesia with subsequent stone extraction is an alternative requiring prolonged catheter drainage, and is associated with some risk and discomfort (Hawkyard et al, 1990). Most experts concur that laparoscopic cholecystectomy cannot be performed safely and comfortably without general anaesthesia under most circumstances. A second small population of patients are inordinately apprehensive about general anaesthesia or surgical removal of their gallbladder and refuse standard therapy in spite of recurrent biliary symptoms. Although direct contact dissolution is generally well tolerated, it may require several days of paramedical attention with medical supervision. This procedure will be most efficiently, effectively and comfortably performed by an experienced team and is, therefore, probably best provided by referral centres with physicians sufficiently interested in the treatment of biliary tract stone disease to develop expertise with this method.

Within the past 7 years, gallbladder lithotripsy by shockwaves has been proven to be a safe and effective non-invasive therapy for selected patients with gallstone disease. While regulatory decisions prevent shockwave therapy from being used more frequently in the USA, the number of patients treated in Europe and Asia is increasing constantly. At our institution, a relatively constant number of about 250 new patients per year have been treated since 1988 (Figure 4). About 20% of patients with gallstones are suitable for shockwave therapy according to present criteria. The rate of evacuation of all fragments is determined by the initial stone number and stone size, the success at stone fragmentation, adjuvant bile acid dissolution therapy, and gallbladder contractility. In contrast to laparoscopic cholecystectomy (Dubois et al, 1989; Perissat et al, 1989; Southern Surgeons Club, 1991), shockwave therapy does not require general anaesthesia. And in contrast to direct contact dissolution therapy of gallbladder stones using MTBE (Thistle et al, 1989), lithotripsy is non-invasive. In the majority of patients, complete fragment disappearance takes several months. Preliminary analyses of the cost-effectiveness of lithotripsy have revealed that lithotripsy, including retreatments and bile acid medication for recurrent stones, costs about as much as open cholecystectomy (Rothschild et al, 1990; Bass et al, 1991). The ideal patient for gallbladder lithotripsy has a single radiolucent stone < or = 20-25 mm in diameter in a functioning gallbladder (Figure 1). In patients with such stones, nearly all studies have confirmed a favourable outcome with rapid clearance of all fragments and a relatively low rate of stone recurrence. For carefully selected patients, extracorporeal shockwave lithotripsy is therefore an attractive non-invasive therapy.

Oral cholelitholytic bile acid therapy has become established treatment for selected patients with cholesterol gallstones. The treatment finds its clinical application both alone and in combination with ESWL. UDCA alone or, less commonly, a combination of this bile acid with CDCA is used. Optimal results can be expected only in carefully selected patients. Bile acid dissolution therapy is most successful in patients with radiolucent gallstones which are < or = 0.5 cm in diameter or are shown by OCG to be floating. Dissolution is seldom seen when the stones are > 1 cm in size. Cholelitholytic treatment in combination with ESWL yields optimal results in single radiolucent gallstones which are not greater than 2 cm. ESWL thus makes it possible to use medical treatment effectively in single 1-2 cm gallstones when bile acids alone would not be successful. Bile acid treatment is extremely safe, especially if UDCA is given without the addition of CDCA.

The most certain symptomatic manifestation of gallstones is episodic upper abdominal pain. Characteristically, this pain is severe and located in the epigastrium and/or the right upper quadrant. The onset is relatively abrupt and often awakens the patient from sleep. The pain is steady in intensity, may radiate to the upper back, be associated with nausea and lasts for hours to up to a day. Dyspeptic symptoms of indigestion, belching, bloating, abdominal discomfort, heartburn and specific food intolerance are common in persons with gallstones, but are probably unrelated to the stones themselves and frequently persist after surgery. Many, if not most, persons with gallstones have no history of pain attacks. Persons discovered to have gallstones in the absence of typical symptoms appear to have an annual incidence of biliary pain of 2-5% during the initial years of follow-up, with perhaps a declining rate thereafter. Gallstone-related complications occur at a rate of less than 1% annually. Those whose stones are symptomatic at discovery have a more severe course, with approximately 6-10% suffering recurrent symptoms each year and 2% biliary complications. The far higher rates of symptom development reported in a few studies raise the possibility that these incidence estimates may be too low. The best predictors of future biliary pain are a history of pain at the time of diagnosis, female gender and possibly obesity. The risk of acute cholecystitis appears to be greater in those with large solitary stones, that of biliary pancreatitis in those with multiple small stones, and that of gallbladder cancer in those with large stones of any number. Drugs that inhibit the synthesis of prostaglandins may now be the treatment of choice in patients with gallstones who are suffering acute pain attacks. Persistent dyspeptic symptoms occur frequently following cholecystectomy. A prolonged history of such symptoms prior to surgery and evidence of significant psychological distress appear to be the best predictors of unsatisfactory outcome.

In recent years, there has been a rapid growth in knowledge about the subcellular mechanisms involved in pancreatic ductal secretion of bicarbonate. The mechanisms governing anion transport across the luminal membrane of duct cells have been well characterized. Evidence suggests that the cystic fibrosis transmembrane conductance regulator is a cyclic adenosine monophosphate-regulated Cl- conductance in the luminal membrane that plays a pivotal role in ductal bicarbonate secretion by recirculating the Cl- imported into duct cells through Cl(-)-HCO3- exchange. The mechanisms governing ion transfer across the basolateral plasma membrane of duct cells are less well defined. There is some evidence suggesting that secretin may cause exocytotic insertion of proton pumps into the basolateral plasma membrane. Once inserted into the plasma membrane, proton pumps could engage in primary active electrogenic H+ ion transport to interstitial tissue while secondary active HCO3- secretion occurs over the luminal membrane through Cl(-)-HCO3- exchangers coupled in parallel with the cystic fibrosis transmembrane conductance regulator. These and other subcellular phenomena related to ductal secretory function are reviewed.

Endoscopic sphincterotomy and percutaneous approaches to the biliary tract have revolutionized the treatment of bile duct stones. Both the endoscopic and transhepatic approaches are less invasive than open surgery. This is an advantage for the mostly elderly and frail patients with common bile duct stones. Other patients with intrahepatic stones, e.g. young patients with oriental lithiasis, may also profit from the non-surgical approach. In this latter group it is often difficult for the surgeon to obtain access to the stone-bearing bile ducts. Due to the anatomical situation, size or impaction of stones the non-surgical approach, including mechanical disintegration, may primarily fail. Several techniques such as intracorporeal lithotripsy using electrohydraulic probes or laser light, extracorporeal shockwave lithotripsy or direct contact dissolution are now available and often allow complete clearance of the bile ducts. If a kidney lithotripter with radiographic devices is available, it should be used after an attempt at mechanical lithotripsy has failed (Figure 1). According to the literature, experience with this method is greater than with any other 'third-step approach'. The procedure is simple, relatively safe and successful in approximately 80% of patients. However, in at least one third of patients, several sessions have to be performed and further endoscopy is frequently required for extraction of fragments. Intracorporeal techniques may become the procedure of choice in the future, at least in patients with common bile duct stones. At the moment, however, the different devices are still not fully developed and too susceptible to damage. A further major drawback, especially with high-energy electrohydraulic intracorporeal lithotripsy, is the danger of bile duct injury or even perforation, so that most procedures must be performed under optical control. The use of contact dissolution cannot generally be recommended. Treatment with mono-octanoin or modified mono-octanoin solvents takes too long, is often not successful and has a high rate of side-effects. MTBE may shorten the procedure considerably, but is suitable only for cholesterol stones, and the danger of spill-over into the intestine with absorption and systemic side-effects has to be weighed against the probability of success.

Free-floating tumor debris or mucobilia as a cause of intermittent obstruction has been described infrequently. A patient with intermittent jaundice caused by tumor emboli from an intrahepatic polypoid mucinous cholangiocarcinoma is presented. Symptoms of intermittent jaundice and midepigastric pain persisted over 5 years despite an initial cholecystectomy and common bile duct exploration before definitive diagnosis and treatment of an hepatic trisegmentectomy (segments II, III, and IV). Intraductal mucin was confirmed intraoperatively and pathologically as the cause of the obstructive jaundice. The patient remains asymptomatic and without evidence of disease more than 5 years postoperatively. This report of a predominantly mucin-producing intrahepatic cholangiocarcinoma details a rare protracted clinical course of intermittent biliary obstruction from mucus emboli and highlights the possibility of long-term survival after complete resection.

Extensive epidemiological and experimental studies have suggested that some chemical agents, nutritional deficiencies, and physical factors are associated with the development of esophageal cancer (EC). Recent evidence also suggests an etiologic role of certain microorganisms in esophageal carcinogenesis either by producing carcinogens or promotors or by acting directly on the host cells. The mutagenic and carcinogenic effects of several fungi and bacteria isolated from the grains and foodstuffs in high-risk areas have been shown by in vitro and in vivo studies. Certain viruses, e.g., human papillomavirus, herpes simplex virus, cytomegalovirus, and Epstein-Barr virus, have been implicated in the pathogenesis of a variety of human cancers, and all of them are known to produce tumors in animals and cell transformation in vitro. These viruses also have been shown to infect the esophageal epithelium. Therefore, although many of the key issues of their mechanisms of action are unclear as yet, they should be considered potential etiologic agents of EC. The present review summarizes the data available on the etiology of EC, emphasizing the current evidence implicating an etiologic role of microorganisms in the pathogenesis of this malignancy.

Gastrointestinal symptoms occur in a large number of patients with food allergies. Immediate hypersensitivity mechanisms may give rise to the nausea, vomiting, abdominal pain, and diarrhea experienced by these patients. However, there are limited human data about the pathophysiological basis for these symptoms. Most of the available information comes from a variety of animal models. This article reviews the literature using models of intestinal food hypersensitivity, as well as human studies, that have contributed to our understanding of the pathophysiological mechanisms in gastrointestinal food hypersensitivity.

Ascites indicates the accumulation of fluid in the peritoneal cavity, due to a wide range of causes. These causes can be classified according to the presence of portal hypertension, severe blood dyscrasia and peritoneal disease. Cirrhosis is the most frequent cause of ascites. The occurrence of ascites in cirrhosis is due to portal hypertension, which is responsible for the increase in hydrostatic pressure at the sinusoidal level and the alterations of splanchnic and systemic haemodynamics. These latter include increased splanchnic inflow, reduced systemic resistance and increased plasma volume and cardiac output. Portal hypertension also plays a major role in determining sodium retention, which occurs in the setting of increased RAA system and SNS activity. The mechanisms by which portal hypertension leads to the activation of antinatriuretic factors and sodium retention are not completely understood; three main hypotheses have been proposed to explain this relationship, namely the underfilling, the overflow and the peripheral arterial vasodilatation theories. In patients with cirrhosis and ascites, there is an overall activation of the renal prostaglandin system, which probably acts to maintain renal haemodynamics and GFR by counteracting the vasoconstricting effects of AII and noradrenaline on renal circulation. In advanced stages, ascites may become refractory to medical treatment and renal function shows a progressive impairment and eventually acute renal failure, the so-called HRS, due to a marked vasoconstriction of the renal arteries and the opening of the intrarenal-arteriovenous (A-V) shunts. In this condition, the reduced renal synthesis of vasodilating prostaglandins is probably of pathogenic importance. Treatment of ascites is usually based on bed rest, low-sodium diet and administration of aldosterone antagonists and loop diuretics. A sequential treatment of ascites based on the progressive addition of more potent drugs is the best way to relieve ascites while avoiding potentially dangerous side-effects. Patients who fail to respond to the above manoeuvres are said to have refractory ascites. Current treatment of this latter condition is mainly based on therapeutic paracentesis and the application of the LeVeen shunt, but long-term results are unsatisfactory.

Gastric varices (GV) are a common (20%) accompaniment of portal hypertension; they are more often seen in those patients who bleed than in those who do not (27% versus 4%, p < 0.01). They can develop in both segmental and generalized portal hypertension. Depending on their location and relation with oesophageal varices, GVs can be classified as gastrooesophageal varices (GOV) and isolated gastric varices (IGV); each of these can be further subdivided as follows: GOV1 (extension of oesophageal varices along lesser curve) and GOV2 (extension of oesophageal varices towards fundus); and IGV1 (varices in the fundus) and IGV2 (isolated varices anywhere in the stomach). The common presentation of GVs is variceal bleeding and encephalopathy. In comparison with oesophageal varices, GVs bleed significantly less often (64% versus 25%, p < 0.01) but more severely (2.9 +/- 0.3 versus 4.8 +/- 0.6 transfusion units, p < 0.01). Patients with GOV2 and IGV1 bleed more often than patients with other types of GVs. Sclerotherapy for oesophageal varices can significantly influence the natural history of GVs. GOV1, or lesser curve varices, disappear in the majority of cases (59%) after obliteration of oesophageal varices. In those with persisting GOV1, the incidence of bleeding and mortality is high and these patients require gastric variceal sclerotherapy (GVS). During oesophageal variceal sclerotherapy, bleeding can occasionally be induced from GVs. After obliteration of oesophageal varices, recurrence as GVs (secondary GVs) can occur in about 9% of patients. Emergency GVS is quite effective in controlling acute bleeding from GVs, more so than balloon tamponade. Potent sclerosants like tetradecyl sulphate and alcohol and a glue, bucrylate, have been quite effective. Elective GVS can achieve obliteration of GVs in nearly 70% of patients. Rebleeding and ulceration are common complications of GVS; probably related to incomplete obliteration and mucosal injury respectively. Splenectomy is quite effective in treating GVs due to segmental protal hypertension. For GV bleeding due to generalized portal hypertension, a shunt operation is often effective. TIPS procedure appear to be a very promising therapy for GV bleeding. Liver transplantation may be a superior alternative to sclerotherapy and shunt surgery for gastric varices.

In this chapter, we have tried to indicate the role of the portosystemic shunt in the treatment of portal hypertension. The conclusions are evident: in the last 10 years it has lost its role as leader in the treatment of portal hypertension. However, some firm statements can be made. The selective shunt is an operation that provides both good variceal decompression and satisfactory maintenance of liver function. Its results in great part depend on the skill of the surgeon. Only a patient with good liver function (Child's classes A and B) is a candidate for shunt surgery, with, very occasionally, a patient with severe disease (class C). In an emergency, the operation is used only after failure of sclerotherapy, but it must be used at the right time before the patient's condition has deteriorated. In the prevention of variceal rebleeding, the selective shunt or sclerotherapy can be routine measures. The choice between the two treatments depends on the patient's willingness and the ability of the institution to perform both procedures successfully. If sclerotherapy is chosen, the institution must be able to rapidly rescue a sclerotherapy failure by shunt surgery. Liver transplantation is probably the treatment of the future, but it is at present impossible to suggest that the procedure is feasible for all patients with variceal bleeding and severe liver disease.

There is now substantial clinical evidence to suggest that portal hypertensive gastropathy is an important source of gastrointestinal bleeding in patients with portal hypertension. Although a relatively uncommon presenting feature in such patients, it appears to become progressively more frequent and important the longer such patients with bleeding oesophageal varices survive after treatment by endoscopic sclerotherapy. It is now being increasingly recognized as the most important cause of haemorrhage after oesophageal varices in such patients. The endoscopic and histological characteristics of the condition are now well established but from a clinical point of view it is important to distinguish it from a number of other disorders. The pathogenesis of portal hypertensive gastropathy is poorly understood; venous congestion secondary to portal hypertension undoubtedly plays an important role but this is not thought to account entirely for the condition since abnormalities in the arterial blood supply are also observed. Many abnormalities in gastric mucosal function have been reported but it is unclear whether these are secondary disturbances or whether they play an important primary role in the development of the condition. Animal studies to date have not been helpful due to the lack of a satisfactory experimental model. Portocaval shunt surgery cures portal hypertensive gastropathy but propranolol has been shown to be highly effective in controlling haemorrhage from this condition and should now be considered the treatment of choice. The mechanism of action is unclear, and it remains to be shown whether other beta-blockers, or indeed any other drugs, are useful in treating this disorder.