Pancreatic, Hepatic, and Biliary Carcinoma

April 1, 2005

Pancreatic, hepatic, and biliary carcinomas in adults represent three of the most challenging malignancies facing the oncologist. Although groups at high risk for these malignancies are recognized, screening and early-detection strategies have not been successful.

Pancreatic CancerHepatocellular CarcinomaGallbladder CarcinomaReferences

Pancreatic, hepatic, and biliary carcinomas in adults represent three of the most challenging malignancies facing the oncologist. Although groups at high risk for these malignancies are recognized, screening and early-detection strategies have not been successful. For each neoplasm, surgery represents the only practical curative treatment option. Radiation and chemotherapy have been helpful only in selected clinical circumstances. Hopefully, our evolving understanding of the molecular biology and cellular biochemistry of these neoplasms will provide new approaches for early detection and therapy. This section will review the current approaches to management of these three gastrointestinal malignancies.

Pancreatic Cancer

Pancreatic cancer is the second most common gastrointestinal cancer and the fourth leading cause of cancer death in the United States. The incidence of pancreatic cancer is exceeded only by that of lung, colorectal, skin, prostate, and breast cancers. It is estimated that 24,000 new cases of pancreatic cancer will be diagnosed in the United States during 1995 [1]. The median survival of patients with this disease is 3 to 4 months, and the 5-year survival rate is only 3% [2]. One reason for the dismal prognosis is that the initial symptoms of this disease are nonspecific and, thus, the disease is usually advanced with obvious or occult metastases established before a diagnosis is ever made. Currently, most clinical efforts are diagnostic and palliative. The development of successful treatment strategies will be based on the evolving understanding of the molecular events involved in cellular transformation, tumor progression, and regulation of neoplastic growth.

Epidemiology and Etiology

The incidence of cancer of the pancreas increases with age. Risk increases after age 30 years, with most cases ocurring between the ages of 65 and 79. However, the disease has also been reported in younger individuals, including children [1].

The ratio of males to females affected differs according to age, varying from 2:1 for patients younger than 40 years of age to 1:1 for patients older than 80 years. The slight male predominance prevails in both whites and nonwhites. Pancreatic cancer is more common in Hispanic and African-American than in white populations. The incidence in blacks is 14.9 per 100,000, as compared with 8.7 per 100,000 in whites [2].

A variation in incidence and mortality among different religious groups has also been observed. There is an increased incidence of pancreatic cancer among Jews in New York City and in Israel and a lower mortality rate is lower among Mormons [3].

The incidence in countries of origin and in first- and second-generation immigrants to the United States has been examined. The rate in the first-generation immigrants rapidly increases to the rate of US whites.

The cause of pancreatic carcinoma remains uncertain, but several factors have been implicated. Cigarette smoking has been associated with an increased risk of pancreatic carcinoma [4,5]. A study from Veterans Administration hospitals showed almost twice the rate of pancreatic carcinoma for heavy cigarette smokers (ie, at least two packs daily) as for nonsmokers. The risk increases as the level of cigarrette smoking increases, and the excess risk levels off 10 to 15 years after cessation of smoking [6].

Dietary factors also appear to play a role in the development of this disease. In animal studies, dietary fat enhances the experimental production of azaserine-induced pancreatic tumors [7]. Although coffee consumption was believed to be a strong etiologic factor [8], more recent studies have failed to support this observation [9]. Alcohol has not been conclusively associated with pancreatic cancer [5].

A third factor possibly involved in the development of pancreatic carcinoma is prior gastrectomy [10] for benign conditions, which may increase the risk of developing pancreatic cancer by two- to fivefold. It has been postulated that the postgastrectomy achlorhydric environment favors the colonization of bacteria that reduce nitrate-containing compounds to N-nitroso compounds. The latter are believed to be carcinogenic [11]. Another explanation for the correlation between gastrectomy and pancreatic cancer is the increased plasma cholecystokinin (CCK) levels associated with prior gastrectomy. CCK stimulates the growth of normal pancreatic cells and, in animal models, is believed to mediate the growth stimulus in carcinogenesis [12]. Inhibition of pancreatic cancer by a CCK antagonist has also been observed in experimental studies [13].

Besides smoking, diet, and prior gastrectomy, certain disease states, including chronic pancreatitis and diabetes mellitus, also have been associated with pancreatic carcinoma. Calcifications associated with chronic pancreatitis have been found in 3% of patients with documented pancreatic carcinoma [14]. About 15% of pancreatic carcinoma patients are diabetics, and, in more than half, the onset of clinical diabetes precedes the diagnosis of pancreatic carcinoma by less than 3 months [15].

Finally, occupational exposure to solvents, petroleum compounds, beta-naphthylamine, and benzidine has been found to be associated with pancreatic cancer [16,17]. The nitrosamines are recognized as potent pancreatic carcinogens in hamsters [11]. Azaserine has produced pancreatic tumors in rats [18]. Indeed, exposure to these industrial chemicals for 10 years or more may increase the risk of pancreatic carcinoma fivefold [16].


Pancreatic cancer arises from both the exocrine and endocrine parenchyma of the pancreas [19]. Approximately 95% of pancreatic cancers occur within the exocrine portion of the pancreas and may arise from ductal epithelium, acinar cells, connective tissue, or lymphatic deposits. The most common pancreatic cancer is a ductal adenocarcinoma, which accounts for about 80% of all pancreatic cancers [20]. Most carcinomas arise in the proximal portion of the pancreas, which includes the head, neck, and uncinate process. Only 20% of pancreatic carcinomas arise in the body, and only 5% to 10% in the tail [21].

The majority of patients with pancreatic carcinoma present with advanced disease. At presentation, 85% of patients have clinically obvious metastases or micrometastases. The usual breakdown of patients after staging is: disease apparently confined to the pancreas, 20%; locally advanced disease, 40%; and visceral metastases, 40%. The most common sites of metastases are the liver, peritoneum, lymph nodes, and lung.

Clinical Features

The initial symptoms, which include anorexia, weight loss, abdominal pain, and jaundice, are generally vague and nonspecific. As a result, two thirds of patients experience symptoms for at least 2 months before the diagnosis is made. Weight loss, often gradual and progressive, is one of the earliest and most frequently unappreciated symptoms. A typical patient with pancreatic carcinoma has lost more than 10% of his or her body weight by the time of diagnosis. Abdominal pain, the most common symptom, is present in 70% to 80% of patients. The pain is due to local tumor infiltration into the retroperitoneum and splanchnic nerve plexus. Severe pain is often considered a sign that the tumor is not resectable.

Jaundice secondary to biliary obstruction can present as either an early or a late symptom, depending upon the tumor location. Associated symptoms of dark urine and pale stools occur. Gastric outlet obstruction and duodenal obstruction can occur in as many as 25% of pancreatic head cancers and are usually secondary to local tumor invasion and motility problems from infiltration of the splanchnic nerves.

Other occasional findings include a palpable gallbladder at presentation (Courvoisier's sign)[22], splenomegaly, depression, and a higher frequency of venous thrombosis and migratory thrombophlebitis (Trousseau's sign)[14].


Computed tomography (CT), ultrasonography, endoscopic retrograde cholangiopancreatography, and fine-needle aspiration biopsy have all been used successfully to diagnose pancreatic cancer. However, CT of the abdomen is the most useful procedure for diagnosis and staging. The advantages of CT scanning are that it is not operator dependent, it is not limited by stomach or bowel gas, and it will demonstrate liver metastases as small as 2 cm, involvement of peripancreatic lymph nodes, perivascular invasion, and lymphadenopathy [23]. It can also reveal dilation of the pancreatic duct and the site of obstruction in 88% to 97% of instances. However, CT has certain limitations, as not all pancreatic carcinomas are observed as masses and not all masses are pancreatic carcinomas. Nevertheless, a review of 100 cases of pancreatic carcinoma showed a false-negative rate for CT of only 5% [24].

Unlike CT, ultrasonography does not involve ionizing radiation. For the patient presenting with jaundice, ultrasonography is more accurate than CT in distinguishing obstructive from nonobstructive jaundice [25]. However, the efficacy of ultrasonography decreases with obesity and excessive bowel gas [26,27].

Although endoscopic retrograde cholangiopancreatography and fine-needle aspiration biopsy are invasive procedures, they do offer certain advantages. Endoscopic retrograde cholangiopancreatography (ERCP) has high sensitivity in diagnosing pancreatic carcinoma (94%) [28]. It can localize the tumor and detect the site of ductal obstruction. Furthermore, endoscopic retrograde cholangiopancreatography permits the aspiration of pancreatic secretions for cytologic examination. Access to bile also allows measurement of other cancer-derived factors such as the K-ras oncogene. As reported in one series [29], the false-negative rate for ERCP is only 3%. Although the radiographic diagnosis of pancreatic cancer can generally be made using a combination of CT, ultrasonography, and endoscopic retrograde cholangiopancreatography, fine-needle aspiration biopsy, which has a sensitivity of 86% [30], can provide a histologic diagnosis.

Serologic Markers

The search for a specific tumor marker for pancreatic cancer that might aid in screening and early diagnosis has not yet been successful, although several serologic markers are helpful in managing the disease. Carcinoembryonic antigen (CEA), a high-molecular-weight glycoprotein normally found in fetal tissue, has been studied most extensively. It is elevated (greater than 2.5 mg/mL) in only 40% to 50% of patients with pancreatic cancer. It can be elevated in pancreatitis as well as in many other benign intestinal disorders and in biliary, gastric, hepatocellular, and colorectal carcinomas.

Other serologic markers associated with pancreatic carcinoma include CA 19-9 and the ratio of testosterone to dihydrotestosterone. CA 19-9 is a mucinous glycoprotein with a half-life of less than 1 day that is associated with a variety of malignancies (pancreatic, hepatobiliary, gastric, and colorectal). The sensitivity of CA 19-9 for pancreatic cancer has been found to range from 67.6% to 92% (37 U/mL is the upper limit of normal) [31]. In general, the level of CA 19-9 increases with more advanced disease stage; the CA 19-9 level in patients with stage I or II disease is usually within the normal range. Hence, the sensitivity of this marker in diagnosing early, resectable pancreatic tumors is lower [31,32]. When 37 U/mL is used as the cutoff point, the CA 19-9 level is also elevated in 4.3% to 28% of patients with chronic pancreatitis and in 21% to 35% of those with extrapancreatic gastrointestinal cancers [31,32]. Raising the cutoff level improves the specificity of CA 19-9 for pancreatic carcinoma.

Another serologic marker that can be analyzed is the ratio of testosterone to dihydrotestosterone, which has been shown to be less than 5 in 67% of men with pancreatic cancer [33]. This is related to the increased level of 5-alpha-reductase associated with pancreatic cancer [34]. Compared with CA 19-9, the ratio of testosterone to dihydrotestosterone is less sensitive but more specific for pancreatic cancer [33].

Two other antibodies, DU-PAN-2 and SPAN-1, have been screened in patients with pancreatic carcinoma. Both were raised against human pancreatic cancer cell lines. The specificity of these newer markers in patients with pancreatic cancer is comparable to that of CA 19-9 [35,36]. However, the antibodies are reactive against several other types of cancer, principally gastrointestinal neoplasms, and against many other benign conditions.

Hence, no single marker alone is sufficiently sensitive or specific to be used as a screening test, but combining these markers can help to confirm the diagnosis of pancreatic carcinoma.


Staging is useful in choosing treatment, assessing prognosis, and comparing the results of different treatment programs. The most commonly used staging system for pancreatic cancer is the tumor-node-metastasis (TNM) classification devised in 1981 by the American Joint Committee for Cancer Staging and End Results Reporting (Table 1).

TXPrimary tumor cannot be assessed
T0No evidence of primary tumor
T1Tumor limited to the pancreas
T1aTumor measuring 2 cm or less in greatest dimension
T1bTumor > 2 cm in greatest dimension
T2Tumor extends directly to the duodenum, bile duct, or peripancreatic tissues
T3Tumor extends directly to the stomach, spleen, colon, or adjacent large vessels
NXRegional lymph nodes cannot be assessed
N0No regional lymph-node metastasis
N1Regional lymph-node metastasis
MXPresence of distant metastasis cannot be assessed
M0No distant metastasis
M1Distant metastasis
Stage groupingTNM
Stage IT1 T2N0 N0M0 M0
Stage IIT3N0M0
Stage IIIAny TN1M0
Stage IVAny TAny NM1

In the TNM system, the primary tumor status is defined by extension through the pancreatic capsule; nodal status is defined by the presence of regional pancreatic lymph-node involvement; and metastatic disease status is defined by the presence of distal lymph node, peritoneal, or visceral disease. Only stage I disease, which is localized within the pancreatic capsule, is amenable to resection.


The approach to therapy differs in patients with pancreatic carcinoma depending on the stage of their disease at presentation.

Resectable Disease: As mentioned, only patients with T1-2N0M0 disease are considered to have resectable disease, and surgery is the only potentially curative modality. In general, curative surgery is feasible only in patients with cancer of the pancreatic head, since patients with cancer of the pancreatic tail or body invariably present with advanced disease. Even among those patients whose disease is resectable, 90% die of tumor recurrence within 1 to 2 years [19,37]. The median survival after curative resection is 12 to 18 months.

The site of recurrence is often local. In one study, local recurrence was the sole site of failure in 27% of patients who underwent resection and was a component of failure in 70% [38]. Whipple resection or pancreatoduodenectomy is considered the standard surgical procedure for cancer of the head of the pancreas [39].

Adjuvant radiotherapy: Most patients who undergo curative resection subsequently die of recurrent disease, which indicates the need for effective postoperative adjuvant therapy. In 1974, the Gastrointestinal Tumor Study Group (GITSG)[40] evaluated the role of adjuvant radiation and chemotherapy following curative resection. A total of 43 patients were randomized to either a control arm for observation or a treatment arm in which 40 Gy of radiation was given in a split course, with intravenous fluorouracil administered for the first 3 days of each radiation course and on a weekly basis beginning 1 month after the completion of radiotherapy and continuing for 2 years thereafter. Survival was longer in the treatment group than in the observation group (21 vs 10.9 months). Because of the small size of the study, 30 additional patients [41] with better performance status were subsequently registered, and the results again showed improved survival for patients in the treatment arm. Although the statistical power in these two trials is weak and subject to criticism, this treatment approach has become common practice.

Adjuvant chemotherapy: Whereas combined postoperative radiation and chemotherapy for pancreatic carcinoma has been systematically evaluated, adjuvant chemotherapy alone has not. However, before effective adjuvant therapy strategies can be developed, it is necessary to find an effective regimen for the treatment of advanced pancreatic cancer. No such regimen is known at the present time.

Unresectable Disease: Patients with unresectable disease are treated palliatively. Biliary obstruction in the presence of unresectable disease is relieved by endoscopic placement of a stent. Percutaneous biliary drainage is best avoided in a patient in whom no further therapy is planned, because it results in recurrent infections. Duodenal obstruction, which is uncommon, can be relieved with bypass surgery or with recently developed laparascopic techniques.

Chemotherapy: As mentioned earlier, the results of chemotherapy for pancreatic carcinoma have been disappointing. This disease is sensitive to only a few agents. Table 2 shows those drugs that have produced measurable responses in at least 10% of patients in phase II studies.

DrugNumber of patientsResponse (%)

Fluorouracil offers the best response rate: 28% in a collective series, [42] with a range of 0% to 67%. The wide range is probably due to patient selection rather than to differences in scheduling or response criteria. However, the numbers of patients in these mostly uncontrolled trials are small.

Another agent with low-level activity against pancreatic cancer is mitomycin (Mutamycin), for which response rates range from 10% to 38%, with an average response rate of 21% [43]. Responses with streptozocin (Zanosar) and doxorubicin (Adriamycin, Rubex) have been reported to be 11% [44] and 13% [43], respectively. However, the responses seen with all of these single agents are rarely complete, and the response durations are usually short. Complete responses have been reported with ifosfamide (Ifex), although the rates differed widely in the two studies reported so far: A trial involving 25 patients reported a response rate of 60% (with 4% complete responses) to ifosfamide [45], whereas another trial involving 29 patients reported a response rate of 22% (with 3.5% complete responses)[46].

Since fluorouracil and mitomycin are the most effective single agents for pancreatic carcinoma, most combination chemotherapy regimens have included these two agents. In one of the earliest studies, FAM (fluorouracil, doxorubicin, and mitomycin) produced a response rate of 37% [47]. Thirty-nine patients with advanced pancreatic cancer were treated with fluorouracil, 600 mg/m² intravenously (IV) on days 1, 8, 29, and 36; doxorubicin, 30 mg/m² IV on days 1 and 29; and mitomycin, 10 mg/m² IV on day 1. The cycle was repeated every 8 weeks. No complete responses were observed, and the median survival of the responders was 12 months. However, a subsequent study by the North Central Cancer Treatment Group comparing fluorouracil alone vs fluorouracil plus doxorubicin vs FAM failed to detect a survival advantage for combination chemotherapy over single-agent fluorouracil [48].

Another chemotherapeutic combination, SMF (fluorouracil, mitomycin, and streptozocin), was originally reported by Wiggans et al [49] to yield a response rate of 43%. In this study, 23 patients with advanced pancreatic cancer were treated with streptozocin, 1 g/m² IV on days 1, 8, 29, and 36; mitomycin, 10 mg/m² IV on day 1; and fluorouracil, 600 mg/m² IV on days 1, 8, 29, and 36. The cycle was repeated every 8 weeks. A subsequent study by the GITSG [50] comparing the efficacy of FAM with that of SMF failed, however, to reproduce those promising initial results. The responses for the two regimens were 13% and 15%, respectively.

Other attempts to improve the results of chemotherapy have centered on modifying the dosing schedule. In general, the drug whose schedule is modified has been fluorouracil, which may be given either in a bolus or as a continuous infusion. Based on the rationale that fluorouracil is a cell cycle-specific drug with a short half-life, it should be more effective if given as an infusion. Nevertheless, response rates in trials where this drug was administered by continuous infusion varied from 0% to 39%. The trials were limited by their small sample sizes [51].

Modulations of fluorouracil by leucovorin (folinic acid) [52,53] and N-phosphonacetyl-L-aspartate (PALA)[54] in the treatment of pancreatic cancer have also been tried. Leucovorin increases the intracellular concentration of 5, 10-methyltetrahydrofolic acid, which results in prolonged inhibition of thymidylate synthetase by the fluorouracil metabolite FdUMP. However, results with high-dose leucovorin and fluorouracil in patients with advanced pancreatic carcinoma have been poor [52,53].

PALA inhibits the enzyme aspartate transcarbamylase, resulting in a reduction in the pool of uridine triphosphate and cytosine triphosphate, an effect that could enhance the therapeutic index of fluorouracil. However, a study by the Southwest Oncology Group (SWOG) reported a response rate of only 5% when PALA was given with high-dose fluorouracil as a short-term infusion [55].

Only a few trials have utilized intra-arterial chemotherapy. In those that did, results were disappointing and did not show a survival benefit [56,57].

Finally, new compounds continue to be elevated. Gemcitabine (2,2´-difluorodeoxycytidine) is a novel pyrimidine antimetabolite that demonstrated preclinical antitumor activity in murine solid tumor models and responses in patients with pancreatic cancer during phase I evaluation. A phase II trial by Casper et al was performed in chemo-naive patients with adenocarcinoma of the pancreas [57a]. Gemcitabine, 800 mg/m², was administered intravenously weekly for 3 consecutive weeks every 28 days. A response rate of 11% was observed (all partial responses). Despite these marginal response rates, interesting improvements in pain and performance status were frequently observed. Based on these observations, a randomized trial of gemcitabine vs fluorouracil was conducted, which demonstrated significant improvements in symptom control and survival with gemcitabine [57b]. Thus, gemcitabine should be considered for the front-line management of symptomatic patients with advanced pancreatic adenocarcinoma.

Hormonal therapy: Hormonal therapy also has potential as palliative treatment for pancreatic carcinoma. Results with experimental models have suggested that pancreatic cancer is a hormone-dependent tumor [58,59]. Theve et al [60] treated 14 patients with advanced pancreatic carcinoma with tamoxifen and reported that median survival was prolonged to 8.5 months (3 patients survived for 22 months). In another study of tamoxifen, however, Miller and Benz [61] did not achieve such promising results.

Reports that somatostatin (octreotide) inhibits pancreatic carcinoma growth in animals [62] suggest that it may have potential use in treating pancreatic cancer. CCK and secretin stimulate the growth of the exocrine pancreas in animals. Although the role of these hormones in human tumorigenesis is uncertain, CCK can stimulate the growth of a human pancreatic adenocarcinoma cell line [63]. Among the actions of somatostatin and its analogs is the inhibition of the secretion or action of gastrointestinal hormones, including CCK. In two human pancreatic adenocarcinoma cell lines, growth inhibition was seen with somatostatin. Direct inhibitors of CCK have also been evaluated recently, but no effect on disease progression or quality of life has been documented [64].

Radiotherapy: Palliative radiotherapy has been used to control pain in patients with pancreatic carcinoma. Radiotherapy by itself does not improve survival but, when combined with fluorouracil, does [65]. In a GITSG study comparing 6,000 cGy alone with fluorouracil plus 4,000 or 6,000 cGy, in patients with locally unresectable pancreatic cancer, median survival was longer in both chemoradiation treatment arms than in the radiotherapy alone arm (10 vs 5.5 months).

Other investigational radiotherapy techniques have been used in attempts to improve the management of pancreatic cancer, with mixed results. One of these, intraoperative radiotherapy, has been reported to enable a higher dose of radiation to be delivered to the disease site without injuring neighboring tissues [66]. In a study at the Mayo Clinic [67], 159 patients were randomly assigned to either intraoperative radiotherapy plus external-beam radiation, or with external-beam radiation alone. Although local control was superior in the intraoperative radiotherapy group, this improved local control did not translate into improved survival. Finally, other attempts to improve survival in pancreatic cancer, such as ultrasonically guided percutaneous implantation of iodine-125 seeds, have not yielded significant results.

Biologic Therapy: There has been considerable interest in the use of monoclonal antibodies to treat patients with pancreatic cancer. Monoclonal antibodies are believed to kill the tumor target either indirectly, by interaction with cells that affect antibody-dependent cell-mediated cytotoxicity (ADCC) and natural killer cell activity, or directly by complement-dependent cytolysis [68,69]. Two monoclonal antibodies, MoAb 494/32 and MoAb CO 17-1A, show strong reactivity against pancreatic cancer.

MoAb 494/32 is a murine antibody of the immunoglobulin (Ig) G1 isotype that binds strongly to human pancreatic cancer cells and mediates ADCC. A partial response was seen in a patient with pancreatic cancer during phase I-II evaluation of this antibody. In a prospective, randomized trial in patients who had undergone pancreatic resection, median survival of the 29 patients who received postoperative MoAb 494/32 was 14.2 months, not significantly different from the 12.9-month median survival of the 32 control patients [70].

Another important monoclonal antibody is MoAb 17-1A, which is a murine IgG2a that targets a 37-kd glycoprotein on the surface of a variety of human gastrointestinal adenocarcinomas [71]. It does not mediate complement-dependent lysis of tumor cells, but it does participate in ADCC. In preliminary clinical studies of MoAb 17-1A, some responses were seen in patients with pancreatic carcinoma [69]. Another opportunity for therapeutic intervention would be using antibodies against the epidermal growth factor receptor. It has been found that pancreatic cancer cells overexpress the epidermal growth factor receptor [72]. This receptor is a transmembrane glycoprotein expressed by a variety of normal and malignant cells.

Future Targets: As molecular biologists and biochemists develop better understanding of pancreatic cancer, new targets for therapy will emerge. One such target is the K-ras oncogene. It has been found that at least 80% of pancreatic adenocarcinomas contain a mutated K-ras gene [73]. This oncogene plays an important role in the development and progression of pancreatic cancer and is a target for new therapeutic approaches.

Pain Control: Pain is a symptom experienced by virtually all patients with pancreatic carcinoma at some time in the course of their disease. It is often what motivates the patient to seek medical attention. Pain may arise when the tumor infiltrates into the retroperitoneum or the splanchnic nerve plexus. It is typically burning and severe, and if there is nerve compression, it may be accompanied by dysesthesia and hyperesthesia in the area of innervation. Anticonvulsants and tricyclic antidepressants are useful in treating this type of pain [74,75]. In addition, steroids can reduce the edema associated with pancreatic cancer and lessen the compression and pain.

Pain may also be due to gastric or small bowel obstruction or bile duct dilatation, which leads to intense muscular spasm. This, in turn, leads to ischemia with cellular breakdown and liberation of pain-producing substances. This type of pain is diffuse and poorly localized and is referred to the dermatomes that share the same innervation as the affected viscus.

Treatment for such pain involves the generous use of narcotics, which can be given in two ways. The oral route is preferred, as it is painless and convenient. The use of long-acting narcotics, such as controlled-release morphine, usually achieves better pain control. In general, the required dose is titrated with a short-acting morphine preparation given every 4 hours. If oral administration fails to control the pain, parenteral administration can be used. Under these circumstances, the drug is preferably given via a continuous infusion pump, since the pump can be programmed to administer the drug at a background infusion rate and allow for a boost dose to be given on demand by the patient.

In patients with intractable pain, more radical procedures, such as a celiac axis block, can be carried out under fluoroscopic guidance. The celiac plexus is a dense network of nerves between the two large ganglia formed by the splanchnic nerves. The splanchnic nerves and sympathetic trunks are the sole mediators of pain arising from the pancreas and the extrahepatic biliary ducts. Therefore, by blocking the celiac axis, pain can be blocked. Pain relief with this procedure is greater than 90% in most reported series [76-78].

Radiotherapy also has been used to control pain [65]. Usually administered with a radiosensitizer, the radiation acts by shrinking the tumor and relieving compression of nerves. The disadvantage of radiotherapy treatment is delayed relief. Nevertheless, this method is quite effective when it is well coordinated.

Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is one of the most common malignancies; it causes an estimated 1,250,000 deaths every year worldwide. The distribution of HCC shows striking geographic variations. Countries and populations are grouped according to incidence rates, ie, high (20 or more per 100,000 males per year) in China, Southeast Asia, Western and Southern Africa, and in Chinese populations in Singapore, Taiwan, and Hong Kong; intermediate (6 to 19 per 100,000 males per year) in Japan, Bulgaria, Poland, France, Hungary, Yugoslavia, Czechoslovakia, Belgium, and Austria and among New Zealand Maoris and Hawaiians and Chinese living in the United States; and low (fewer than 5 cases per 100,000 males per year) in the United Kingdom, the United States, Canada, Australia, Israel, Scandinavia, Latin America, India, and New Zealand [79].

Worldwide, HCC occurs in approximatley three times as many males as females. The incidence increases with age independent of risk until it levels off in the elderly. In high-incidence regions, there is a marked shift toward increased incidence in the younger age group.

Risk Factors

HCC is almost always associated with chronic underlying liver disease, principally hepatitis B and hepatitis C. There is a positive correlation between the geographic pattern of hepatitis B surface antigen (HBsAG) and the incidence of HCC. Molecular biologic studies have shown integrated hepatitis B virus DNA in the livers of patients with chronic hepatitis and hepatocellular carcinoma [80].

Aflatoxin, a mycotoxin resulting from Aspergillus fungi, appears to be an important cocarcinogen in rural Africans. Aflatoxin-contaminated food has been reported to produce liver tumors in trout, rats, and rhesus monkeys fed aflatoxin B1 over a long period of time [81].

In the West, HCC generally occurs in the setting of cirrhosis. Eighty to 90% of cases are related to ethanol-induced cirrhosis. Other risk factors include hemochromatosis and parasitic infections of the liver.


HCC occurs in two gross patterns, a diffuse form and a nodular form. The diffuse form may evolve from a unicentric lesion [82].

Of all the histologic patterns of HCC, only the fibrolamellar variant is of importance in treatment and clinical behavior. This variant is more frequent in females, occurs at an earlier age, is not associated with cirrhosis, and tends to be solitary and, thus, more amenable to surgical resection. These fibrolamellar carcinomas, characterized histologically by eosinophilic polygonal-shaped cells separated by lamellar fibrosis, have better prognosis than do the other histologic types of HCC [82].

Invasion of the portal vein has been found in 14% of resected HCC specimens when the lesion was smaller than 2.0 cm and in 71% of specimens when the lesion was larger than 5.1 cm. These thrombi may involve the hepatic vein, vena cava, and portal vein [83].

Clinical Features

The most common symptoms in patients with HCC are abdominal pain (91%), ascites (43%), weight loss (35%), weakness (31%), fullness and anorexia (27%), vomiting (8%), and jaundice (7%). About one third of patients are asymptomatic. Metastatic disease can present as malignant ascites, skeletal pain, dyspnea with pulmonary involvement, and neurologic abnormalities due to brain metastases [84].

Ascites may occur as a consequence of underlying chronic liver disease or may be due to a rapidly expanding tumor. Other less common symptoms include fever of unknown origin, intra-abdominal hemorrhage, upper gastrointestinal bleeding, bone pain, coma secondary to hepatic failure, hypercalcemia, and respiratory symptoms. Jaundice occurs infrequently and usually is secondary to the underlying liver disease.

Among the physical signs present in patients with HCC, hepatomegaly is the most frequent, occurring in 50% to 90% of patients. Signs of cirrhosis such as spider angiomas and gynecomastia are common. Abdominal bruits can arise from the HCC, presumably secondary to the increased vascularity. Ascites can occur as part of the underlying liver disease, caused by a hemoperitoneum, or as malignant ascites. Splenomegaly may occur as a result of associated portal hypertension from the underlying liver disease. Weight loss and fever are also common, particularly with rapidly growing or large tumors.The Budd-Chiari syndrome, Virchows or Trosier's nodes, and cutaneous metastases have also been reported.

A number of paraneoplastic syndromes have been reported in association with HCC. The most important ones include hypoglycemia (also caused by end-stage liver failure), erythrocytosis, hypercalcemia, hypercholesterolemia, dysfibrinogenemia, carcinoid syndrome, increased thyroxine-binding globulin, sexual changes, and porphyria cutanea tarda.

The changes in laboratory values with HCC are very similar to those with cirrhosis; they include elevated levels of alkaline phosphatase, transaminases, and bilirubin. These changes are usually present in about 50% of patients and predict a short survival [85].

Alpha fetoprotein (AFP) and alpha-1-globulin are the tumor markers usually associated with HCC [86]. In humans, they are normally detected in the embryo beginning 6 weeks after fertilization and disappear a few weeks after birth. Although AFP levels are also elevated in germ-cell tumors and during pregnancy, the immunoassay will still detect between 70% and 90% of all HCCs. Reports have associated normal levels of AFP with improved survival. Postresection AFP levels are used in monitoring clinical progress.


Hepatocellular carcinomas are best visualized on CT scans. CT can demonstrate primary lesions larger than 1 cm and may also identify compression or invasion of the portal or hepatic veins and extrahepatic metastatic disease, particularly to the hepatoduodenal ligament and lungs [87].

Ultrasonography is inexpensive and noninvasive, but the hyperechoic appearance of HCC may lead to confusion with benign as well as metastatic tumors. HCC may show as a focal form that appears on ultrasound as a rounded or lobular mass lesion, often multiple, with high- and low-level echoes. Ultrasound may show both hepatic veins and portal veins and occasionally demonstrates intravascular tumor. HCC may also present as a diffuse form with changes that may be subtle and indistinguishable from the diffuse changes seen with cirrhosis and chronic active hepatitis.

Magnetic resonance imaging has not as yet been demonstrated to be more useful than CT scanning in the liver. The most sensitive tests are CT portography and lipoidol CT scanning. These are considered mandatory if resection is contemplated since they pick up small lesions in the apparently normal residual lobe and are reported to be capable of diagnosing 50% more lesions than can be diagnosed with CT scanning alone.

Laparoscopy is more invasive, but it provides direct visualization of the liver, peritoneal cavity, and viscera and allows for a percutaneous needle biopsy under direct vision. It may result in more accurate staging and the avoidance of unnecessary laparotomies.

Percutaneous biopsy techniques include fine-needle aspiration cytology and core-needle biopsy. Needle biopsy may be performed directed by ultrasound or CT, blindly, or, as already mentioned, under direct vision using the laparoscope.


The staging system for hepatic tumors was developed by the Union Internationale Contre le Cancer (UICC), and it essentially follows the TNM system (Table 3).

TXPrimary tumor cannot be assessed
T0No evidence of primary tumor
T1Solitary tumor 2 cm or less in greatest dimension without vascular invasion
T2Solitary tumor 2 cm or less in greatest dimension with vascular invasion; or Multiple tumors limited to one lobe, none > 2 cm in greatest dimension, without vascular invasion; or A solitary tumor > 2 cm in greatest dimension without vascular invasion
T3Solitary tumor > 2 cm in greatest dimension with vascular invasion; or Multiple tumors limited to one lobe, none > 2 cm in greatest dimension, with vascular invasion; or Multiple tumors limited to one lobe, any > 2 cm in greatest dimension, with or without vascular invasion
T4Multiple tumors in more than one lobe, or tumor(s) involving a major branch of the portal or hepatic vein(s)
NXRegional lymph nodes cannot be assessed
N0No regional lymph-node metastasis
N1Regional lymph-node metastasis
MXPresence of distant metastasis cannot be assessed
M0No distant metastasis
M1Distant metastasis
Stage groupingTNM
Stage IT1N0M0
Stage IIT2N0M0
Stage IIIT1 T2 T3 T3N1 N1 N0 N1M0 M0 M0 M0
Stage IVAT4Any NM0
Stage IVBAny TAny NM1

Stage I solitary tumors smaller than 2 cm in diameter without vascular invasion have the best prognosis. Multiple tumors, vascular invasion, either microscopic or macroscopic, and lymph-node spread are considered adverse prognostic factors.


Surgery: Surgery is the only curative modality for HCC, but its use depends mainly on tumor size and location and the condition of the uninvolved liver. Small, encapsulated HCCs may be cured with resection 50% of the time, but these tumors are rare. Patients with clinical stage I or II tumors should be considered for resection.

As mentioned, another limiting factor for many patients with HCC is the hepatic reserve available. Diffuse HCC and advanced cirrhosis are generally considered unresectable. The goal in patients with cirrhosis is to preserve as much functioning liver tissue as possible to avoid postoperative liver failure [88]. Subsegmental or segmental resections are preferable in patients with small lesions and cirrhosis. Intraoperative hepatic ultrasound can help detect satellite or metastatic tumors and allow exact localization and ligation of segmental portal and hepatic veins during minimal resections.

Operative mortality increases with the extent of resection and with a decrease in the degree of function of the remaining liver tissue and has been reported to range from 5% to 33% [89]. The type and extent of tumor and the coexistence of cirrhosis or hepatitis will determine the long-term survival. One-year survival rates have been reported to be about 80%, with 5-year survival rates ranging from 30% to 46%.

There is no role for adjuvant chemotherapy or radiotherapy following curative resection.

Cryosurgery: Cryosurgery involves the in situ destruction of tumor by the application of subzero temperatures. With the use of liquid nitrogen, temperatures less than –20ºC can be achieved. This procedure will lead to the death of both tumor and nontumor tissue either immediately or during the thawing period. Multiple freeze-thaw cycles are used to increase the percentage of tumor tissues destroyed.

Theoretically, cryosurgery should be an ideal modality for the treatment of patients with cirrhosis who have inadequate hepatic reserve and multifocal tumor. Instead, the complications associated with this procedure and the very low survival reported in the literature have shown that cryosurgery offers no clear improvement over standard surgery and that its use should be limited to the research setting [90].

Chemotherapy: Most patients with HCC are not candidates for curative surgery, so chemotherapy is their only option. Unfortunately, currently there is no very effective chemotherapy for HCC [91]. Doxorubicin, still considered the single-agent standard for HCC, has only a 20% response rate. No other single agent has yet demonstrated a higher rate. Combination chemotherapy has been found to increase toxicity with no improvement in survival. Interferon alfa-2a (Roferon-A) also has been used in HCC with no impact on patient survival [92,93].

Another approach to the treatment of HCC is hepatic intra-arterial infusion, in which high concentrations of chemotherapy are delivered directly to the liver. Hepatic intra-arterial infusion of agents such as floxuridine (FUDR), mitomycin, and interferon have produced reported response rates as high as 50% [94,95]. Unfortunately, these responses are not usually durable. Another drawback of this method is that these treatments require either long-term percutaneous catheterization of the hepatic artery or laparotomy for the placement of an infusion device, which tends to be costly and prone to complications. Hepatic intra-arterial infusion is currently not recommended as standard practice.

Chemoembolization, the administration of chemotherapy concomitant with percutaneous hepatic artery embolization, is yet another approach to treating HCC. The major blood supply to HCC is derived from the hepatic artery, while normal liver parenchyma is supplied primarily by the portal vein. Thus, occlusion of the hepatic artery can accomplish relatively selective tumor ischemia.

Gelatin sponge is the most widely used vaso-occlusive material, although, other substances such as starch, polyvinyl alcohol, collagen, chemotherapy microspheres, and autologous blood clot have also been tried. The different sizes of these particles predict for a different level of arterial blockage, and their differing degrees of degradability predict for differences in the duration of blockade. The optimal material has yet to be determined.

Chemoembolization for HCC produces partial responses in approximately 50% of patients. Nonetheless, no definitive study has demonstrated a superiority of chemoembolization over hepatic intra-arterial infusion, other palliative therapies, or the state-of-the-art supportive care for HCC.

Radiotherapy: External-beam radiotherapy has a role in the palliative management of HCC, but this role has been limited by the total organ dose tolerance of radiation therapy. Radiation hepatitis with subsequent fibrosis is one of the possible complications seen when doses of 3 to 30 Gy are delivered. Conformal radiotherapy ports can be used to minimize the beam scatter and allow the delivery of therapeutic doses of radiation to solitary lesions.

A new technique being used is radiation-labeled antibodies to ferritin. Ferritin is not only found in normal tissues but also synthesized and secreted by HCC. Use of the (iodine-131 polyclonal antiferritin antibody [96] has produced a 48% remission rate in phase I and II studies. It has also downstaged a small percentage of inoperable cancers to resectable ones. When a radiolabeled antibody-chemotherapy regimen was compared with chemotherapy alone, no survival advantage was found with the combination therapy [97]. Attempts were made to substitute the iodine isotope for a beta-emitting yttrium-90 antiferritin isotope, but it was found that the antiferritin moeity was no longer selective for HCC. Current investigation is focused on the use of a human monoclonal antiferritin antibody.

Liver Transplantation: Liver transplantation has been used as an alternative to resection in patients with advanced bilobar HCC or HCC in the presence of advanced cirrhosis [98]. The results of this strategy, however, have been discouraging, with 3-year survival rates amounting to less than 50% and long-term survival rates of approximately 20% [99].

Gallbladder Carcinoma

Carcinoma of the gallbladder was first described by Maximilian de Stoll in 1777 [100]. It is now the fourth most common tumor of the gastrointestinal tract. Together with liver cancer, carcinoma of the gallbladder will account for an estimated 18,500 new cancers and 14,200 estimated US deaths in 1995 [1]. Gallbladder carcinoma is an aggressive disease, and the prognosis for patients is dismal, with an overall 5-year survival rate of less than 5% and a median survival of 6 months [101,102].

Epidemiology and Etiology

Carcinoma of the gallbladder is a disease of older people. Incidence peaks in the seventh decade, with three times as many women as men affected. There is no difference in the age of onset between males and females [103]. The incidence is higher in white women than in black women and 10 times higher in Mexicans, American Indians, and Alaskan natives, suggesting a genetic basis of susceptibility [4,104].

The female predominance in this disease has suggested an association between benign and malignant gallbladder disease. Silk et al [102] reviewed their 22-year experience with gallbladder carcinoma at Roswell Park Cancer Institute and found that cholelithiasis was present in 48 (68%) of 69 patients. A calcified (porcelain) gallbladder was found in seven patients (10%). Another study by Paraskevopoulos et al [105] found gallstones in 18 (86%) of 21 patients with histologically proven carcinoma. Other authors have emphasized the occurrence of carcinoma in porcelain gallbladders [106], but because some of the patients with porcelain gallbladders also have gallstones, it is difficult to attribute an exact role to each of these factors. Cholecystitis has also been considered one of the contributing factors in this disease. Silk et al found acute and/or chronic cholecystitis in 54 (76%) of the 69 patients in their series.

Other possible contributing factors for this disease include chemical carcinogens, especially nitrosamines and methylcholanthrene, and inflammatory processes like ulcerative colitis in which there is a strong association with carcinoma of the gallbladder.


Overall, 80% to 85% of gallbladder carcinomas are adenocarcinomas, and most are of the scirrhous type. The anaplastic, squamous-cell, and adenoacanthoma types are rare. In the series from Roswell Park [102], 67 (94%) of the patients had adenocarcinoma, and 4 (6%) had squamous-cell carcinoma. In a series by Yamaguchi et al [107], 29 patients had adenocarcinoma and 2 patients had adenosquamous carcinoma. Carcinoma was limited to the fundus in 21 patients, to the body in 1 patient, to the neck in 2 patients, and involved the entire organ in 6 patients.

Natural History

Gallbladder carcinoma is usually detected at an advanced stage, when the prognosis is poor. Many gallbladder carcinomas are discovered by pathologists as an incidental finding after the removal of the gallbladder for chronic cholecystitis. In some patients, the cancer is diagnosed at autopsy.

Gallbladder cancer can be found as a polypoid projection into the lumen of the gallbladder or as a diffuse thickening of the wall of the organ, with or without extension into the liver and other adjacent organs. The liver and the regional lymph nodes are the most common sites of involvement, followed by peritoneal carcinomatosis.

Cancer invasion of the liver is usually by direct extension of the disease, by spread via the lymph channels or Luschka's ducts, or by a combination of these processes [108]. Liver metastases far from the gallbladder bed are seen in half of patients.

Invasion of the local lymph nodes by the tumor tends to occur early. The lymphatic drainage is to the lymph nodes along the cystic and common bile ducts and then through the pancreaticoduodenal nodes to the paraaortic nodes. The cystic node is invaded in 42% to 79% of cases. Almost half of patients have an extension of the tumor to the common bile duct. However, invasion to the colon, pancreas, and stomach is seen in only 10% to 20% of patients [109].

Clinical Features

Patients with gallbladder carcinoma may be asymptomatic or may present with abdominal pain, jaundice, weight loss, anorexia, or nausea and vomiting. Others may present with a right upper quadrant mass or complications such as gastrointestinal hemorrhage. Some of these symptoms and signs make distinguishing between carcinoma and benign disease a difficult task.

Serum chemistries are not always very helpful because serum bilirubin levels are often normal. An elevated alkaline phosphatase level often accompanies a normal serum bilirubin level in this disease [103].


Ultrasonography of the gallbladder can be helpful in the diagnosis of gallbladder cancer. It may show a localized thickening or a mass. Unfortunately, the appearance of gallbladder carcinoma or ultrasound is difficult or impossible to differentiate from the wide spectrum of appearances associated with gallstones [105]. In the series from Paraskevopoulos [105], 21 of 3,197 patients who had routine cholecystectomies were found to have carcinoma of the gallbladder at operation. The correct preoperative diagnosis was made in only two patients despite the fact that all but one had ultrasound examination, which revealed gallstones in 18.

CT scanning can also demonstrate gallbladder abnormalities. It can aid by identifying irregular thickening of the gallbladder wall, masses in the gallbladder region, and intraluminal masses [110]. Endoscopic retrograde cholangiopancreatography may demonstrate filling defects in the gallbladder or evidence of infiltration of the common bile duct.

Angiography can demonstrate irregularities and tumor vessels arising from the cystic artery. However, it is not considered a screening procedure since by the time the diagnosis is made, the tumor is incurable [110].

Another method used to diagnose gallbladder carcinoma, is an upper gastrointestinal contrast study, which may show displacement of stomach and duodenum when extensive disease is present; however, findings are not specific.


There are two commonly used staging systems for gallbladder carcinoma: the TNM staging system of the International Union Against Cancer and the Nevin staging system (Table 4)[101].

TisCarcinoma in situ
T1Tumor invades the mucosal or muscular layer
T1aLimited to mucosa
T1bInvades muscular layer
T2Invades perimuscular connective tissue
T3Invades serosa and/or one organ, liver < 2 cm
T4Invades 2 or more organs, or liver > 2 cm
N1AHepatoduodenal ligament nodes: cystic duct, pericholedocal, or hilar
N1BOther regional nodes: peripancreatic (head only), periportal celiac, or superior mesenteric
Stage 0TisN0M0
Stage IT1N0M0
Stage IIT2N0M0
Stage IIIT1 T2 T3N1 N1 Any NM0 M0 M0
Stage IVT4 Any TAny N Any NM0 M1

In the classification system of Nevin and Moran, stage I tumors are localized to the mucosa; stage II tumors penetrate the muscularis; stage III tumors involve all three layers; stage IV disease involves metastases in the cystic duct lymph nodes; and stage V disease involves invasion of metastases to liver or adjacent or distant organs.

The management of gallbladder carcinoma largely depends upon the stage of disease at presentation. In the series of Silk et al, [102] stage of disease was the only factor that consistently influenced survival.


Surgery: Surgical resection is the primary treatment modality for carcinoma of the gallbladder. Cholecystectomy may be all that is required for stage I and II disease. It is generally effective in those cases in which the carcinoma is an incidental finding on histopathologic examination of a gallbladder removed for symptomatic benign disease. Shieh [111] reported a 66.6% 5-year survival rate for patients whose tumors were confined in the wall of the gallbladder and were found incidentally by the examining pathologist.

Some authors believe that patients with gallbladder carcinoma invading beyond the mucosal layer may benefit from more radical procedures such as extended cholecystectomy with regional lymph-node dissection and resection of the gallbladder bed. Once the lesion invades the gallbladder serosa it is considered incurable. After cholecystectomy or radical surgery the 5-year survival rate is 10% to 30%, with locoregional recurrences seen in about 80% of the cases [109]. Pancreaticoduodenectomy has also been recommended when lymphadenopathy posterior to the pancreas or duodenal invasion is present.

Radiotherapy: The role of radiation as either primary or adjuvant therapy for this disease has been difficult to evaluate because of small series and the difficulty in assessing tumor response. Nevertheless available data thus far have suggested an important role for radiation as palliative and adjuvant therapy for gallbladder cancer.

Vaittinen [103] reported on 24 patients treated with surgery alone and 7 patients treated with surgery and postoperative external irradiation and found the average survivals of 29 and 63 months, respectively. Hanna and Rider [112] reported on 51 patients with cancer of the gallbladder, of whom 35 received radiation therapy. Irradiation increased the total survivals of patients treated curatively and palliatively. Bosset [109] reported five of seven gallbladder cancer patients free of disease at 5, 9, 11, 31, and 58 months after complete resection and postoperative radiotherapy to 5,400 cGy in 30 fractions. Todoroki [113] reported on 17 patients with stage IV gallbladder carcinomas who received 2,000 to 3,000 cGy of intraoperative radiotherapy; 10 of the patients also received 3,600 cGy of external-beam irradiation. The 3-year survival rate for the radiation-treated group was 10%, whereas none of nine stage IV patients treated with resection alone were alive at 3 years.

These results suggest an increased benefit in terms of survival and palliation in patients receiving postoperative radiation as either palliative or adjuvant therapy.

Chemotherapy: Chemotherapy provides little benefit in advanced gallbladder cancer or as an adjuvant treatment after surgical resection. Fluorouracil, considered the most active single agent in the treatment of advanced gallbladder disease, produces only transient responses in about 20% of patients. Other agents that have been used include mitomycin and doxorubicin. For patients with localized disease, postoperative radiotherapy combined with fluorouracil may offer some benefit. Smoron [114] observed one patient with a survival of 6 years after external postoperative irradiation and chemotheraphy with fluorouracil.

Newer drugs and alternate routes of delivery are essential if there is to be any success with this line of treatment.



1. American Cancer Society: Cancer Statistics-1995. J Am Cancer Society 45:8–30, 1995.

2. Gloeckler LA, Hankey BF, Miller BA, et al: Cancer Statistics Review, 1973–1988 (NIH Publ 91-2789). Bethesda, National Institutes of Health, 1988.

3. Enstrom JE: Cancer and total mortality among active Mormons. Cancer 42:1943–1951, 1978.

4. Krain L: The rising incidence of carcinoma of the pancreas: An epidemiologic appraisal. Am J Gastroenterol 54:500, 1970.

5. Wynder E et al: An epidemiologic evaluation of the causes of cancer of the pancreas. Cancer Res 35:2228, 1975.

6. Kahn H: The Dorn study of smoking and mortality among US veterans: Report on eight and one-half years of observation. NCI Monogr 19:1, 1966.

7. Roebuck BD, Yager JD Jr, Longnecker DS, et al: Promotion by unsaturated fat of asazerine-induced pancreatic carcinogenesis in the rat. Cancer Res 41: 3961–3966, 1981.

8. MacMahon B, Yen S, Trichopoulos D, et al: Coffee and cancer of the pancreas. N Engl J Med 304:630–633, 1981.

9. Goldskin HR: No association between coffee and cancer of the pancreas (letter). N Engl J Med 306:997, 1982.

10. Offerhaus GJA, Tersmette AC, Tersmette KWF, et al: Gastric, pancreatic and colorectal carcinogenesis following remote peptic ulcer surgery. Mod Pathol 1: 352–356, 1988.

11. Pour P, Althoff J, Kruger F, et al: The effect of N-nitrosobis-(2-oxopropyl)-almine after oral administration to hamsters. Cancer Lett 2:323, 1977.

12. Pour P, Lawson T, Helgeson S, et al: Effect of cholecystokinin on pancreatic carcinogenesis in the hamster model. Carcinogenesis 9:5197–5201, 1988.

13. Nio Y, Tsubono M, Morimoto H, et al: Loxiglumide (CR1505), a cholecystokinin antagonist, specifically inhibits the growth of human pancreatic cancer cell lines xenografted into nude mice. Cancer 72:3599–3606, 1993.

14. Robin A, Scott J, Rosenfeld D: The ocurrence of carcinoma of the pancreas in chronic pancreatitis. Radiology 94:289, 1970.

15. Karmody A, Kyle J: The association between carcinoma of the pancreas and diabetes mellitus. Br J Surg 56:362, 1969.

16. Brooks J: Cancer of the pancreas, in Brooks JR (ed): Surgery of the Pancreas, p 263. Philadelphia, WB Saunders, 1983.

17. Mancuso T, El-Attar A: Cohort study of workers exposed to betanaphthylamine and benzidine. J Occup Med 9:277–285, 1967.

18. Longnecker D, Curphey T: Adenocarcinoma of the pancreas in asazerine-treated rats. Cancer Res 35:2249, 1975.

19. Cubilla AL, Fitzgerald PJ: Surgical pathology of tumors of the exocrine pancreas in Moosa AR (ed): Tumors of the Pancreas, pp 159–193. Baltimore, Williams & Wilkins, 1980.

20. Legg MA: Pathology of the pancreas, in Brooks JR (ed): Surgery of the Pancreas pp 41–77. Philadelphia, WB Saunders, 1983.

21. Howard JM, Jordan GL: Cancer of the pancreas. Curr Probl Cancer 2–1, 1977.

22. Moertel CG: Exocrine pancreas, in Holland JF (ed): Cancer Medicine, pp 1792–1804. Philadelphia, Lea & Febiger, 1982.

23. Hessel SJ, Siegelman SS, McNeil BJ, et al: Prospective evaluation of computed tomography and ultrasound of the pancreas. Radiology 143:129–133, 1982.

24. Ward EM, Stephens DH, Sheedy PR: Computed tomographic characteristics of pancreatic carcinoma: An analysis of 100 cases. Radiographics 3:547, 1983.

25. Taylor KJW, Rosenfield AT: Grey scale ultrasonography in the diagnosis of jaundice. Arch Surg 112: 820–825, 1977.

26. Lees WR: Pancreatic ultrasonography. Clin Gastroenterol 13:763–789, 1984.

27. Rosch T, Lightdale CJ, Botet JF, et al: Localization of pancreatic endocrine tumors by endoscopic ultrasonography. N Engl J Med 326:1721–1726, 1992.

28. Freeny PC, Ball TJ: Endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC) in the evaluation of suspected pancreatic carcinoma: Diagnostic limitations and contemporary roles. Cancer 47:1666–1678, 1981.

29. Freeny PC, Lawson TL: Radiology of the Pancreas, p 457. New York, Springer-Verlag, 1982.

30. Mitty H, Efremidis S, Yeh HC: Impact of fine-needle biopsy on the management of patients with carcinoma of the pancreas. Am J Roentgenol 137:119–121, 1981.

31. Pasquali C, Sperti C, Alfano D, et al: Evaluation of carbohydrate antigens 19-9 and CA-125 in patients with pancreatic cancer. Pancreas 2:34–37, 1987.

32. Malesci A, Tommasini MA, Bonato C, et al: Determination of CA 19-9 antigen in serum and pancreatic juice for differential diagnosis of pancreatic adenocarcinoma from chronic pancreatitis. Gastroenterol 92:60–67, 1987.

33. Robles-Diaz G, Diaz-Sanchez V, Fernandez-del Castillo C, et al: Serum testosterone:Dihydrotestosterone ratio and CA 19-9 in the diagnosis of pancreatic cancer. Am J Gastroenterol 86:591–594, 1991.

34. Igbal MJ, Greenway B, Wilkinson ML, et al: Sex steroid enzymes aromatase and 5 alpha reductase in the pancreas: A comparison of normal adult, fetal, and malignant tissue. Clin Sci (Lolch) 65:71–75, 1983.

35. Kiriyama S, Hayakewa T, Kondo T, et al: Usefulness of a new tumor marker, SPAN-1, for the diagnosis of pancreatic cancer. Cancer 65:1557–1561, 1990.

36. Mahvi DM, Meyers WC, Bast RC, et al: Therapeutic efficacy as defined by a serodiagnostic test utilizing a monoclonal antibody in carcinoma of the pancreas. Ann Surg 202:440, 1985.

37. Romond EH, Mendelsohn LA, MacDonald JS: Adjuvant therapy of gastrointestinal cancer. Cancer Treat Res 33:273–295, 1987.

38. Griffin JF, Smalley SR, Jewell W, et al: Patterns of failure after curative resection of pancreatic carcinoma. Cancer 66:56–61, 1990.

39. Whipple AO, Parsons WB, Mullins CR: Treatment of carcinoma of the ampulla of Vater. Ann Surg 102:763, 1935.

40. Gastrointestinal Tumor Study Group: Pancreatic cancer: Adjuvant combined radiation and chemotherapy following curative resection. Arch Surg 120:899–903, 1985.

41. Gastrointestinal Tumor Study Group: Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Cancer 59:2006–2010, 1987.

42. Carter SK, Connis RL: Adenocarcinoma of the pancreas: Current therapeutic approaches, prognostic variables and criteria of response, in Staquet MJ (ed): Cancer Therapy: Prognostic Factors and Criteria of Response, pp 235–237. New York, Raven Press, 1975.

43. Crooke ST, Bradner WT: Mitomycin C: A review. Cancer Treat Rev 3:121–139, 1976.

44. Smith FP, Schein PS: Chemotherapy of pancreatic cancer. Semin Oncol 6:368–377, 1979.

45. Mawla NGE: Ifosfamide in advanced pancreatic cancer. Cancer Chemother Pharmacol 18(suppl 2):55–56, 1986.

46. Loehrer PJ, Williams SD, Einhorn L, et al: Ifosfamide: An active drug in the treatment of adenocarcinoma of the pancreas. J Clin Oncol 3:367–372, 1985.

47. Frederick PS et al: 5-Fluorouracil, adriamycin and mitomycin C (FAM) chemotherapy for advanced adenocarcinoma of the pancreas. Cancer 46:2014–2018, 1980.

48. Cullinan SA et al: A comparison of three chemotherapeutic regimens in the treatment of advanced pancreatic and gastric carcinoma. JAMA 253:2061–2067, 1985.

49. Wiggans G, Woolley PV, Macdonald JS, et al: Phase II trial of streptozotocin, mitomycin and 5-fluorouracil (SMF) in the treatment of advanced pancreatic cancer. Cancer 41:387–391, 1978.

50. Gastrointestinal Tumor Study Group: Phase II studies of drug combinations in advanced pancreatic carcinoma: Fluorouracil plus doxorubicin plus mitomycin C and two regimens of streptozotocin plus mitomycin C plus fluorouracil. J Clin Oncol 4:1794–1798, 1986.

51. Hansen RM: Gastric and pancreatic carcinomas, in Lokich JJ (ed): Cancer Chemotheraphy by Infusion, 2nd ed, pp 340–357. Chicago, Precept Press, 1990.

52. Crown J, Casper ES, Botet J, et al: Lack of efficacy of high-dose leucovorin and fluorouracil in patients with advanced pancreatic adenocarcinoma. J Clin Oncol 9: 1682–1686, 1991.

53. Decaprio JA, Mayer RJ, Gonin R, et al: Fluorouracil and high-dose leucovorin in previously untreated patients with advanced adenocarcinoma of the pancreas: Results of a phase II trial. J Clin Oncol 9:2128–2133, 1991.

54. Ardalan B, Glazer RI, Kenle TW, et al: Synergistic effect of 5-fluorouracil and N-(phosphonacetyl)-1-aspartate on cell growth and ribonucleic acid synthesis in a human mammary carcinoma. Biochem Pharmacol 30:2045–2049, 1981.

55. Morrell LM, Bach A, Richman SP, et al: A phase II multiinstitutional trial of low dose N-(phosphonacetyl)-1-aspartate and high dose 5-fluorouracil as a short term infusion in the treatment of adenocarcinoma of the pancreas: A Southwest Oncology Group Study. Cancer 67:363–366, 1991.

56. Bengmark S, Andren-Sandberg A: Infusion chemotheraphy in inoperable pancreatic carcinoma: Recent results. Cancer Res 86:13–14, 1983.

57. Theodors A, Bukowski R, Hewlett J, et al: Intermittent regional infusion of chemotherapy for pancreatic adenocarcinoma. Am J Clin Oncol 5:555–558, 1982.

57a. Casper ES, Green MR, Kelsen DP, et al: Phase II trial of gemcitabine (2,2´-difluorodeoxycytidine) in patients with adenocarcinoma of the pancreas. Invest New Drugs 12(1):29–34, 1994.

57b. Moore M, Andersen J, Burris H, et al: A randomized trial of gemcitabine vs 5-FU as first-line therapy in advanced pancreatic cancer (abstract). Proc Am Soc Clin Oncol 14:199, 1995.

58. Benz C, Hollander C, Miller B: Endocrine-responsive pancreatic carcinoma: Steroid binding and cytotoxicity studies in human tumor cell lines. Cancer Res 46:2276–2281, 1986.

59. Benz C, Wiznitzer I: Steroid binding and cytotoxicity in cultured human pancreatic carcinomas (abstract). J Steroid Biochem Mol Biol 19(suppl):125S, 1983.

60. Theve NO, Pousette A, Carlstrom K: Adenocarcinoma of the pancreas-a hormone sensitive tumor? A preliminary report on Nolvadex treatment. Clin Oncol (R Coll Radiol) 9:193, 1983.

61. Miller B, Benz C: Endocrine treatment of pancreatic carcinoma (abstract). Proc Am Soc Clin Oncol 4:90, 1985.

62. Upp JR, Olson D, Poston GJ, et al: Inhibition of growth of two human pancreatic adenocarcinomas in vivo by somatostatin analog SMS 201–995. Am J Surg 142:308–311, 1988.

63. Smith JP, Solomon TE, Baghari S, Kramer S: Cholecystokinin stimulates growth of human pancreatic adenocarcinoma SW-1990. Dig Dis Sci 35:1377–1384, 1990.

64. Abbruzzese JL, Gholson CF, Daugherty K, et al: A pilot trial of the cholecystokinin receptor antagonist ML-329 in patients with advanced pancreatic cancer. Pancreas 7:165–171, 1992.

65. Moertel CG et al: Therapy of locally unresectable pancreatic cancer: A randomised comparison of high-dose (6000 R) radiation alone, moderate dose radiation (4000 R plus 5-fluorouracil) and high-dose radiation plus 5-fluorouracil: The Gastrointestinal Tumor Study Group. Cancer 48:1705–1710, 1981.

66. Shibamoto Y, Manabe T, Baba N, et al: High dose external beam and intraoperative radiotherapy in the treatment of resectable and unresectable pancreatic cancer. Int J Radiat Oncol Biol Phys 19:605–611, 1990.

67. Roldan GE, et al: External beam versus intraoperative and external beam irradiation for locally advanced pancreatic cancer. Cancer 61:1110–1116, 1988.

68. Bosslet K, Kern HF, Kanzy A, et al: A monoclonal antibody with binding and inhibitory activity towards human pancreatic carcinoma cells. Cancer Immunol Immunother 23:185–191, 1986.

69. Herlyn DM, Koprowski H: IgG 2a monoclonal antibodies inhibit human tumor growth through interaction with effector cells. Proc Natl Acad Sci USA 79:4761–4795, 1982.

70. Buchler M, Friess H, Schultheiss HK, et al: A randomized controlled trial of adjuvant immunotheraphy (murine monoclonal antibody 494/32) in resectable pancreatic cancer. Cancer 68:1507–1512, 1991.

71. Gottlinger H, Funke I, Johnson J P, et al: The epithelial cell surface antigen 17-1A, a target for antibody mediated tumor therapy: Its biochemical nature, tissue distribution and recognition by different monoclonal antibodies. Int J Cancer 38:47–53, 1986.

72. Korz M, Meltzer P, Trent J: Enhanced expression of epidermal growth factor receptor correlates with alterations of chromosome 7 in human pancreatic cancer. Proc Natl Acad Sci USA 83:5141–5144, 1986.

73. Almoquera C, Shibata D, Forrester K, et al: Most human carcinomas of the exocrine pancreas contain mutant c-k-ras genes. Cell 53:49, 1988.

74. Budd K: Psychotropic drugs in the treatment of chronic pain. Anaesthesia 33:531, 1978.

75. Serdlow M: The treatment of shooting pain. Postgrad Med J 56:159–161, 1980.

76. Ischia S, Luzzani A, Ischia A, et al: A new approach to the neurolytic block of the coeliac plexus: The transaortic technique. Pain 16:333–341, 1983.

77. Leung JWC, Bowen-Wright M, Aveling W, et al: Coeliac plexus block for pain in pancreatic cancer and chronic pancreatitis. Br J Surg 70:730–732, 1983.

78. Thompson GE, Moore DC, Bridenbaugh LD, et al: Abdominal pain and alcohol coeliac plexus block. Anaesthesia Analgesia 56:1–5, 1977.

79. International Union Against Cancer: Workshop on Biology of Human Cancer. Rep. 17: Hepatocellular carcinoma. Geneva, 1982.

80. Beasley RP et al: Prevention of perinatally transmitted hepatitis B virus infections with hepatitis B immune globulin and hepatitis B vaccine. Lancet 2:1099–1122, 1983.

81. Adamson RC, Corree P, et al: Carcinogenicity of aflatoxin B1 in rhesus monkeys: Two additional cases of primary liver cancer. J Natl Cancer Inst 57:67–78, 1976.

82. Beazley R, Cohn I Jr: Tumors of the pancreas, gallbladder, and extrahepatic ducts. ACS Textbook of Clin Onc 16:219–231, 1994.

83. Okuda K, Ryu M, Takayoshi T: Surgical management of hepatoma: The Japanese experience, in Wanebo JH (ed): Hepatic and Biliary Surgery, pp 219–238. New York, Marcel Dekker, 1987.

84. Okuda K, Obata H, et al: Prognosis of primary hepatocellular carcinoma. Hepatology 4:3–6S, 1984.

85. Chin H, Cheng E, Gellar N: Hepatocellular carcinoma: Statistical analysis of 78 consecutive patients (abstract). Proc Am Soc Clin Oncol 3:6, 1984.

86. Buamah PK, Cornell C, James OFW, et al: Serial Serum AFP heterogeneity changes in patients with hepatocellular carcinoma during chemotherapy. Cancer Chemother Pharmacol 17:182–184, 1986.

87. Hosoki T, Chatani M, Mori S: Dynamic computerized tomography of hepatocellular carcinoma. Am J Radiol 139:1099–1106, 1982.

88. Tsuzuki T, Ogata Y, et al: Hepatic resection in 125 patients. Arch Surg 119:1025–1032, 1984.

89. Adson MA: Primary hepatocellular cancers: Western experience in Blumgart LH (eds): Surgery of the Liver and Biliary Tract, p 1155. New York, Churchill-Livingstone.

90. Zhou XD, Tang ZY, Yu YQ, et al: Clinical evaluation of cryosurgery in the treatment of primary liver cancer. Cancer 61:1889–1892, 1988.

91. Lewis BJ, Friedman MA: Current status of chemotheraphy for hepatoma, in Ogawa M (ed): Chemotherapy of Hepatic Tumors, pp 63–74. Princeton, NJ, Excerpta Medica, 1984.

92. Gastrointestinal Tumor Study Group: A prospective trial of recombinant human interferon alpha 2B in previously untreated patients with hepatocellular carcinoma. Cancer 66:135–139, 1990.

93. Kardinal CG, Moertel CG, et al: Combined doxorubicin and alpha-interferon therapy of advanced hepatocellular carcinoma. Cancer 71:2187–2190, 1993.

94. Atiq OT, Kemeny N, Niedzwiecki D, et al: Treatment of unresectable primary liver with intrahepatic fluorodeoxyuridine and mitomycin C through an implanted pump. Cancer 69:920–924, 1992.

95. Yodono H, Sasaki T, et al: Arterial infusion chemotherapy for advanced hepatocellular carcinoma with EPF and EAP therapies. Cancer Chemother Pharmacol 31:S89–92, 1992.

96. Order SE, Stillwagon GB, Klein JL, et al: Iodine 131 antiferritin, a new treatment modality in hepatoma: An RTOG study. J Clin Oncol 3:1573–1582, 1985.

97. Order SE, Pajak T, et al: A randomized prospective trial comparing full dose chemotheraphy to 131 I antiferritin: An RTOG study. Int J Radiat Oncol Biol Phys 20:953–963, 1991.

98. Iwatsuki S, Starzl TE, Sheahan DG, et al: Hepatic resection versus transplantation for hepatocellular carcinoma. Ann Surg 214:221–229, 1991.

99. Penn I: Hepatic transplantation for primary and metastatic cancers of the liver. Surgery 110:726–735, 1991.

100. de Stoll M: Rationis Mendendi in Nosocomio practico vendobonensi. Part I Lugduni Batavorium, Haak et Socios et A et J. Honkoop, 1788.

101. Nevin JE, Moran TJ: Carcinoma of the gallbladder: Staging, treatment and prognosis. Cancer 37:141–148, 1976.

102. Silk YN et al: Carcinoma of the gallbladder: The Roswell Park experience. Ann Surg 210:751–757, 1989.

103. Vaittinen E: Carcinoma of the gallbladder: A study of 390 cases diagnosed in Finland 1953–1967. Ann Chir Gynaecol 168(suppl):1–18, 1970.

104. Krain LS: Gallbladder and extrahepatic bile duct carcinoma: Analysis of 1808 cases. Geriatrics 27: 1111–1117, 1972.

105. Paraskevopoulos JA, et al: Primary carcinoma of the gallbladder: A 10-year experience. Ann R Coll Surg Engl 74:222–224, 1992.

106. Polk HC: Carcinoma in the calcified gallbladder. Gastroenterology 50:582, 1966.

107. Yamaguchi K, Tsuneyoshi M: Subclinical gallbladder carcinoma. Am J Surg 163: 382–386, 1992.

108. Fahim RB, McDonald JR, Richard JC, et al: Carcinoma of the gallbladder: A study of its modes of spread. Ann Surg 156:114–124, 1962.

109. Bosset JF et al: Primary carcinoma of the gallbladder: Adjuvant postoperative external irradiation. Cancer 64:1843–1847, 1989.

110. Itai Y, Araki K, et al: Computed tomography of gallbladder carcinoma. Radiology 137:713–718, 1980.

111. Shieh CJ, Dunn E, Standard JE: Primary cancer of the gallbladder: A review of a 16-year experience at the Waterbury Hospital Health Center. Cancer 47:996–1004, 1981.

112. Hanna SS, Rider WD: Carcinoma of the gallbladder or extrahepatic bile ducts: The role of radiotherapy. Can Med Assoc J 118:59–61, 1978.

113. Todoroki T et al: Intraoperative radiotherapy for advanced carcinoma of the biliary system. Cancer 46: 2179–2184, 1980.

114. Smoron GL: Radiation therapy of carcinoma of the gallbladder and biliary tract. Cancer 40:1422–1424, 1977.