Over the past two decades, dramatic changes have occurred in medicine and, in particular, in the field of hepatopancreatobiliary diseases. With respect to cancer, the role that oncogenes and tumour suppressor genes play in pathogenesis has become apparent. With respect to imaging, computerised tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET) have all evolved substantially. Similarly, interventional radiology and endoscopy have come of age. At the same time, the ability to perform safe liver surgery and transplantation has become routine. In addition, major advances have occurred in the more focal delivery of radiotherapy as well as in the modes of delivery and in the array of available chemother-apeutic agents. Despite these major advances, the overall outcome of patients with gallbladder and bile duct malignancies has not changed significantly over the same time period.
During the 1980s and 1990s, major changes occurred in world communications. The acceptance of one language in which to communicate at international meetings, the introduction of fax and, of course, the development of the internet have been great advances. Nevertheless, our ability to communicate effectively about biliary malignancies continues to be hampered by confusion in classification. For all malignancies, we have at least three staging systems, the American (AJCC), the European (UICC) and the Japanese. Moreover, for biliary malignancies, we have multiple individual systems, such as the Bismuth classification, which stratifies tumours by perihilar bile duct involvement and the new Memorial tumour classification, which includes vascular encasement. However, both of these systems ignore intrahepatic (peripheral) and distal bile duct tumours as well as lymph node and distant metastases. In addition, tumours that begin in the gallbladder and secondarily involve the bile duct are often included with cholangiocarcinomas. Agreement on and consistent application of one classification system will be necessary for us to make progress in the future.
Another major factor that inhibits progress in our understanding and treatment of biliary malignancies is their relative rarity. In most parts of the world gallbladder cancer occurs in only 2–3/100000 individuals. Interestingly, data from both the USA and the UK suggest that the incidence of cholangiocarcinoma may be increasing. However, this apparent increase may only represent an improved ability to differentiate cholangiocarcinoma from advanced gallbladder cancer. Nevertheless, the fact that both bile duct and gallbladder cancers are (1) uncommon compared with hepatocellular and pancreatic cancer and (2) very rare compared with breast, prostate, lung and colon cancer has inhibited our progress. Nevertheless, the opportunity exists to focus on areas of high incidence such as north-east Thailand, India, Chile, West Virginia and New Mexico to learn more about these rare malignancies. Similarly, further study of patients with premalignant conditions such as those with an abnormal pancreatobiliary duct junction, gallbladder polyps, choledochal cysts, intrahepatic stones, primary sclerosing cholangitis (PSC) and inflammatory bowel disease should enhance our knowledge of the pathogenesis of these malignancies.
During the past 10 years, multiple strategies have enhanced our knowledge of the pathogenesis of pancreatobiliary malignancies. These strategies have included the systematic study of (1) human tumours, (2) cell lines, (3) familial and high-risk patients and (4) animal models. For pancreatic cancer, these strategies have led to a progression hypothesis based on K-ras and Her-2-neu oncogene initiation, subsequent p53 and p21 tumour suppressor gene inactivations and, in some patients, trypsinogen, DPC4 and BRAC2 gene alterations. Human biliary tumours also have been studied, but the frequency of K-ras and Her-2-neu oncogenes activation is less than with pancreatic tumours, whereas the p53 tumour suppressor gene does play a role in approximately 50% of the patients. On the other hand, recent data from cholangiocarcinomas arising in PSC patients suggest that p16 tumour suppressor gene over-expression is much more common than K-ras or p53 abnormalities. However, the systematic study of biliary cancer cell lines and high-risk patients as well as the development of good animal models has occurred at a slower pace than with pancreatic cancer. In addition, familial biliary cancer has not been described. Nevertheless, new gene chip and microarray technologies provide the opportunity for a quantum leap in our knowledge of the genetic pathogenesis of biliary malignancy. With respect to environmental risk factors, smoking has been clearly implicated for pancreatic cancer. For biliary malignancies, radiation exposure and chemical carcinogens, such as dioxin, may be key for cancer initiation and progression.
The pathological diagnosis of biliary malignancies can be quite difficult due, in part, to the relative paucity of cancer cells. Routine cytology of brushings and small punch biopsies have a relatively low sensitivity. Adding K-ras and p53 analysis has not been particularly helpful, because both these markers are present in less than half or about half of the tumours, respectively. As p16 abnormalities are quite common, probing biliary epithelial cells for pl6 alterations maybe more fruitful. One report suggests that PET scanning can document the development of cholangiocarcinoma in patients with PSC. However, further studies of PET scanning in patients at high risk for biliary malignancies need to be performed. Similarly, more data are required to determine the value of serum tumour markers such as CEA and CA-19-9, but the development of a biliary malignancy-specific tumour marker would be ideal.
The ability of spiral CT with three-dimensional reconstruction and MR to detect liver metastases and vascular invasion has improved dramatically. CT and MR cholangiography are now available, and MRC has been shown to be equal to direct cholangiography (endoscopic retrograde [ERC] or percutaneous transhepatic [PTC]) in detecting common duct stones. However, the utility of MRC to accurately to define the intrahepatic ductal involvement of biliary malignancies has yet to be proven. Similarly, the advantage of staging laparoscopy in patients with more advanced gallbladder cancers has been documented. However, the role of laparoscopy in the evaluation of patients with early stage gallbladder cancer and in those with intrahepatic, perihilar and distal cholangiocarcinoma remains controversial.
Debate also continues regarding the value of preoperative biliary drainage in patients with biliary malignancies. Prospective randomised trials, which have been underpowered and have included more patients with distal obstruction, have failed to demonstrate a benefit for preoperative biliary drainage. In addition, large retrospective series of patients undergoing pancreatoduodenectomy have documented increased infective and/or pancreatic complications. Nevertheless, patients with proximal biliary malignancies have higher bilirubin levels and intrabiliary pressures and are more prone to cholangitis. Thus, the relative value of preoperative biliary drainage in patients with perihilar biliary malignancy with and without the atrophy/hypertrophy complex still needs to be determined.
Over the past decade, the concept of preoperative portal vein embolisation (PVE) for patients undergoing major liver resection has been introduced and adopted in many hepatobiliary units. One prospective randomised trail has not shown any advantage of preoperative PVE in patients with a normal liver undergoing a right hepatic lobectomy. On the other hand, morbidity was reduced in the subset of patients with hepatic fibrosis. However, the exact role and timing of preoperative PVE in patients with obstructive jaundice due to a perihilar cholangiocarcinoma or gallbladder cancer has yet to clarified. Should these patients undergo preoperative PVE? If so, should biliary drainage be performed first? Should the period between PVE and operation be longer or shorter with jaundice, malignancy and/or cholangitis? All these questions have yet to be adequately answered.
As with all tumours, the results of resection of gallbladder cancer and cholangiocarcinoma are best when the margins are negative microscopically, i.e. R0. This goal can be achieved with an ‘extended cholecystectomy’ in patients with stage I and II gallbladder cancer. However, in more advanced stage gallbladder malignancies, especially those presenting with obstructive jaundice, the ability to achieve an R0 resection is significantly reduced. Similarly, the likelihood of accomplishing an R0 resection is greater with intrahepatic and distal cholangiocarcinoma lesions rather than with perihilar. There has been general acceptance of the need to perform a liver resection, including the caudate lobe especially with left-sided biliary tumours, to achieve an R0 resection. The importance of minimising blood loss during resection has also become a basic principle. However, surgeons vary considerably on technical details – such as the use of the ultrasonic dissector, the harmonic scalpel, Ligaclips – as well as on temporary vascular occlusion techniques such as the Pringle manoeuvre, total vascular exclusion and ischaemic preconditioning.
Several groups have demonstrated that resection of portions of the portal vein or the head of the pancreas to achieve an R0 resection can be performed. However, an overall analysis of these extensive operations suggests that morbidity and mortality is increased and long-term survival is decreased compared with patients with smaller tumours undergoing more conventional operations. Similarly, the results of liver transplantation for biliary malignancies have been disappointing. Whether the results of transplantation will be improved in carefully selected patients treated with pre-operative chemoradiation remains to be seen.
With respect to palliative treatment, multiple options continue to be advocated by surgeons, interventional radiologists and endoscopists. In patients with locally unresectable perihilar tumours some surgeons favour a segment III bypass, whereas others prefer placement of transhepatic stents into a Roux-en-Y hepaticojejunostomy anastomosed below the tumour. In the absence of the atrophy/hypertrophy complex, most experts advocate drainage of both hepatic lobes. Many also agree that this goal is more easily achieved in proximal tumours by the percutaneous as opposed to the endoscopic route. In recent years metal stents have tended to replace plastic stents for palliation; however, with more proximal tumours the application of two or more metal stents can be technically difficult. Although several retrospective studies have been published, no prospective randomised trials comparing surgical, percutaneous or endoscopic palliation have been reported.
To date, chemotherapy, radiotherapy, chemoradiation and photodynamic therapy have not been proven to prolong survival in patients with gallbladder or bile duct malignancies. Many retrospective studies have suggested that one or more of these treatments is beneficial. However, treated patients generally have smaller, often resected tumours and good performance status, whereas untreated patients usually have larger, unresected and sometimes metastatic tumours and poor performance status. Thus, these non-randomised studies favour whatever treatment has been applied. Several innovative methods to deliver chemotherapy (arterial or chemo-embolisation) or radiotherapy (intra-operative, brachytherapy or conformal) have been introduced. However, no prospective randomised data support any of these options.
In summary, modest progress has been made in the past 20 years in our ability to diagnose, stage and safely resect biliary malignancies. However, major improvements still need to be made in classification, understanding the pathogenesis, early diagnosis, pre-operative preparation, intra-operative management, determining the best methods for palliation and understanding the role of adjuvant therapy. To accomplish these goals, worldwide cooperation will be needed to study high-risk patients and the rare populations with a very high incidence. These efforts should reveal the multiple genetic and environmental factors that cause these devastating malignancies. Armed with this information, hepatobiliary experts of the future should be able to establish preventive strategies and early diagnosis. Similarly, a multidisciplinary approach will be required to achieve the best treatment and/or palliation for individual patients. Because of their rarity, multicontinental trials will be necessary to determine the role of adjuvant therapy for biliary malignancies. Thus, dramatic progress willbe possible over the next two decades if we all work together.