Editors: Skeel, Roland T.
Title: Handbook of Cancer Chemotherapy, 7th Edition
Copyright 2007 Lippincott Williams & Wilkins
> Table of Contents > Section III - Chemotherapy of Human Cancer > Chapter 9 - Carcinomas of the Pancreas, Liver, Gallbladder, and Bile Ducts
Carcinomas of the Pancreas, Liver, Gallbladder, and Bile Ducts
Rekha T. Chaudhary
Carcinomas of the pancreas, liver, and biliary passages account for approximately 2% of all cases of cancer and for 5% of all cancer-related deaths in the United States. Virtually all patients with these cancers die from this disease.
I. Adenocarcinoma of the pancreas
A. Epidemiology and etiology
There were approximately 32,000 cases of pancreatic cancer in 2005 and almost all patients died from the disease. Pancreatic cancer is currently the fifth leading cause of cancer-related deaths in the United States in women and the fourth leading cause of death in men. Risk factors for pancreatic cancer include age, sex, and race. Blacks, men, and patients older than the age of 50 have a higher risk of developing pancreatic cancer. It is rare before 30 years of age, and the incidence rises throughout life, with peak occurrence during the seventh decade. Smokers have 1.6 to 3.9 times the risk of developing pancreatic cancer as compared with nonsmokers. Chronic pancreatitis is commonly associated with carcinoma of the pancreas in pathologic specimens. It is not certain whether this association is causal. Patients with familial pancreatitis appear to have a greater risk. Hereditary pancreatic cancer has been observed in rare families with an autosomal site-specific pattern, in families with BRCA2 mutations, and in families with hereditary nonpolyposis colorectal cancer. Likewise, families with p16 germline mutations may be at a higher risk of developing pancreatic cancer. More than 80% of resected pancreatic cancers harbor either activating point mutations in K-ras or inactivating mutations of the tumor suppressor genes p16, p53, and DPC4.
B. Presenting signs and symptoms
Pain is the most common presenting symptom. It occurs in three fourths of patients with carcinoma of the head of the pancreas and in virtually all patients with carcinoma of the body or tail. Usually, the pain is a dull ache in the epigastrium that radiates to the right upper quadrant when the tumor is in the head of the pancreas or to the left upper quadrant when the tumor is in the body or tail or it may be located in the lumbar region of the back. As many as one fifth of patients also present with nonspecific symptoms including weight loss, anorexia, nausea, vomiting, and constipation. The nonspecific, vague nature of these complaints may delay diagnosis for several months. Seventy percent of patients with carcinoma of the head of the pancreas have jaundice, whereas fewer than 15% of patients with carcinoma of the pancreatic body have jaundice. Physical
C. Diagnostic evaluation
Abdominal ultrasonography and computed tomography (CT) demonstrate masses in the pancreas or dilation of the pancreatic duct or the common bile duct. Sensitivity and specificity of CT are approximately 90%, whereas sensitivity and specificity of ultrasonography are somewhat less. Both tests detect relatively large mass lesions of the pancreas and usually miss 1- to 2-cm carcinomas. Endoscopic retrograde cholangiopancreatography (ERCP) demonstrates subtle ductal abnormalities; sensitivity and specificity are in excess of 90%, with biopsies detecting tumors smaller than 1 to 2 cm in diameter. Endoscopic ultrasound (EUS) may be useful for staging (i.e., nodal status), determination of major vessel invasion and at times for fine needle aspiration (FNA) for pathologic determination of tumor. To determine vascular invasion there are three options: helical CT, magnetic resonance arteriography or EUS. Percutaneous FNA of suspicious abnormalities identified on CT scan can confirm the diagnosis of pancreatic cancer, with 80% to 90% sensitivity and 100% specificity. A common histologic hallmark of pancreatic adenocarcinoma is an associated desmoplastic reaction that, in a given tumor mass, can vastly overestimate the malignant cell mass. Furthermore, pancreatic cancer may be associated with varying degrees of acute or chronic pancreatitis or cyst formation, which may make it difficult to make a diagnosis with needle aspiration and may lead to false-negative results.
D. Laboratory tests
CA 19-9 is a cell surface glycoprotein associated with pancreatic cancer. Rising serum levels may be a useful early indicator of recurrent or progressive disease once a diagnosis has been established but because of low specificity it is not used as screening method. However, there is data to support the need to obtain a CA19-9 level in all patients in whom pancreatic cancer is suspected.
E. Staging and preoperative evaluation
Staging. The primary tumor, regional lymph nodes, and potential sites of metastatic disease must be carefully assessed (Table 9.1). The staging system has been modified to better account for resectability of disease. Resectable disease is loosely defined as disease confined to the pancreas without involvement of the celiac axis or major vessels. A surgeon experienced in pancreatic surgery should evaluate each case individually when determining resectability as there are numerous clinical caveats.
Preoperative evaluation. Preoperative evaluation should be performed stepwise from least invasive to most invasive as indicated by the clinical situation. Preoperative
Table 9.1. TNM staging for pancreatic cancer
F. Primary therapy
Surgery. Three fourths of patients with pancreatic cancer are operative candidates, but only 15% to 20% have resectable tumors. Patients without evident metastatic cancer or major blood vessel involvement, whose performance status permits operative intervention, are candidates for curative surgery. Eight percent to 11% of the patients resected for cure survive 5 years. Surgical bypass procedures may also palliate obstructive jaundice and gastric outlet obstruction. ERCP stent placement may palliate obstructive jaundice.
Radiation therapy. External-beam radiation therapy can palliate unresectable carcinomas. It may also be used as a surgical adjuvant in combination with chemotherapy. Great care and expertise must be exercised to plan the radiation fields. These fields must encompass known disease without excessive involvement of adjacent normal tissue. Surgical clips placed at laparotomy or laparoscopy can guide treatment. Intraoperative external-beam radiotherapy has been successful in placing a high dose on the local tumor
Resected carcinomas. Local and distant failures are common problems after resection of pancreatic cancer. A prospective randomized study by the Gastrointestinal Tumor Study Group (GITSG) compared observation to postoperative radiochemotherapy. The study showed an overall survival benefit (median survival: 20 vs. 11 months, p = 0.03) and a 2-year survival benefit (42% vs. 15%) for the radiochemotherapy arm, but this study is criticized for small patient numbers, low radiation doses, long accrual time, and early termination. The European Organization for Research and Treatment of Cancer did a similar trial that showed a trend toward being positive in the treatment group (median survival: 17.1 vs. 12.6 months p = 0.099). This study is also criticized for low radiation doses and underpowering of the study. A complicated trial in a 2 x 2 design was completed by the European Group for Pancreatic Cancer (ESPAC-1) that is difficult to interpret and has some trial design concerns including selection bias and treatment variability. An intriguing outcome of the analysis is that the chemotherapy group seemed to have a survival benefit; however, the chemoradiotherapy group seemed to do worse than the controls. The toxicity from the combination of chemotherapy and radiation may blur the survival advantage of the chemoradiotherapy arm. Finally, a recent Radiation Therapy Oncology Group (RTOG) trial 97-04 has the benefit of modern radiation doses and the addition of gemcitabine to the chemotherapy regimen. There is no standard of care for adjuvant treatment of pancreas cancer as there is no trial yet to confirm the results of the GITSG trial. All the trials agree, however, that in the acceptable candidate, chemotherapy is useful; however, the addition of radiotherapy is still controversial.
Alternative adjuvant chemotherapy regimens (with or without radiotherapy) include the following:
Fluorouracil 225 mg/m2 by continuous IV infusion throughout radiation therapy followed by four to six courses of bolus fluorouracil weekly, or gemcitabine (1,000 mg/m2 on days 1, 8, 15 with 1-week break) or
Fluorouracil 425 mg/m2 by IV push 1 hour after leucovorin, 20 mg/m2 by IV push daily for 4 days during the first week of radiation therapy and for 3 days during the fifth week of radiation therapy followed by four to six courses of bolus fluorouracil weekly or gemcitabine (1,000 mg/m2 on days 1, 8, 15 with 1-week break) or
Gemcitabine alone (1,000 mg/m2 on days 1, 8, 15 with 1-week break) or
Capecitabine 1,500 mg/m2 daily in divided doses with radiation therapy followed by four to six courses of bolus fluorouracil weekly or gemcitabine (1,000 mg/m2 on days 1, 8, 15 with 1-week break).
Localized unresectable carcinoma. A series of randomized trials conducted by the GITSG demonstrated superior survival of patients with localized but unresectable pancreatic cancer when treated with combined-modality therapy as compared with patients treated with radiation therapy or chemotherapy alone. These clinical trials also used split-course radiation therapy. Current clinical trials do not support a specific combined-modality treatment program; however, doses of chemotherapy as discussed in Section I.F.3.b are used. A dose of 60 Gy of radiation should be delivered in a single course of external-beam radiation to gross tumor and 40 to 50 Gy to microscopic cancer. Paclitaxel is also being used at 50 mg/m2 in a 3-hour infusion along with radiotherapy of approximately 50 Gy with similar benefits and tolerability as historical controls. Neoadjuvant chemotherapy and chemoradiotherapy are being studied to improve the number of patients eligible for resection but only approximately 10% patients were deemed resectable after various chemotherapy and chemoradiotherapy regimens.
G. Chemotherapy of metastatic disease
Patients with pancreatic cancer are often poor candidates for chemotherapy because of severe weight loss, poor performance status, severe pain, lack of measurable or evaluable disease, and presence of jaundice or hepatic involvement, which may interfere with clearance of therapeutic agents. However, two recent randomized clinical trials demonstrated survival and quality-of-life benefits from chemotherapy in selected patients with advanced pancreatic cancer.
Single agents. A number of single agents have demonstrated activity although no agent has demonstrated consistent complete and partial response rates of 20% or greater using CT to measure response. Gemcitabine has been accepted as first-line therapy formetastatic pancreatic cancer in patients with adequate performance status on the basis of a phase III trial that compared bolus fluorouracil and gemcitabine, with the primary endpoint being clinical benefit score. Clinical benefit was defined as sustained (>4 weeks) improvement of one of the following parameters without worsening of any of the others: performance status, composite pain measurement (average pain intensity and narcotic analgesic use), and weight. The improvement in clinical benefit score in the gemcitabine and fluorouracil arms were 25% and 5%, respectively. In addition, there was a significant improvement in survival and in response rates, 16% versus 0%, respectively.
Combination chemotherapy. Combination chemotherapy has beeninvestigated. Themost commonly usedregimens are fluorouracil + doxorubicin + mitomycin (FAM) and streptozocin + mitomycin + fluorouracil (SMF), with response rates reported between 13% and 43%. Other popular combinations using newer agents that have been
Novel agents. Targeted agents such as bevacizumab, cetuximab, and erlotinib in combination with chemotherapy and alone are being investigated for all stages of pancreatic cancer. Erlotinib in particular has shown some promise in a randomized phase III trial of erlotinib versus chemotherapy alone. Initial reports show no increases in response rates but a modest yet statistically significant increase in 1-year survival (24% vs. 17%). This led to the approval of this medication in combination with gemcitabine for the treatment of metastatic pancreatic cancer.
Single-agent therapy with gemcitabine 1,000 mg/m2 IV of 4-week cycles of three weekly doses followed by 1-week rest is recommended for patients withmetastatic pancreatic cancer and with an Eastern Cooperative Oncology Group performance status of 0 to 2, who are not eligible for clinical trials.
Patients with a very good performance status can use a doublet such as GemOx, with gemcitabine 1,000 mg/m2 on days 1 and 8 plus oxaliplatin 130 mg/m2 over 2 hours on day 8 every 3 weeks or
Capecitabine and gemcitabine (CapGem), with capecitabine at 1,500 mg/m2 daily in twice-daily divided doses on days 1 to 14 and gemcitabine 1,000 mg/m2 on days 1 and 8 every 3 weeks or
Capecitabine and oxaliplatin (CapOx), with capecitabine at 1,500 mg/m2 daily in twice-daily divided doses on days 1 to 14 plus oxaliplatin 130 mg/m2 on day 1 every 3 weeks.
Erlotinib 100 mg PO daily plus gemcitabine 1,000 mg/m2 weekly for 3 weeks with a 1-week break.
II. Malignant islet cell carcinomas
A. Epidemiology and natural history
Islet cell neoplasms occur in approximately 1 in 100,000 people per year. These tumors cover a spectrum of neoplasms, many, but not all, of which originate from the pancreatic islets of Langerhans. Eighty percent of these tumors secrete one or more hormones excessively: most commonly insulin or gastrin, less commonly glucagon, serotonin, or adrenocorticotropic hormone, and rarely vasoactive intestinal peptides (VIPs), growth hormone releasing hormone, or somatostatin. Twenty percent are
Sixty percent of gastrinomas are malignant. Histologic appearance and tumor size do not predict malignancy; only the presence of metastatic disease confirms malignancy. Ninety percent of malignant gastrinomas have livermetastases. Other sites of spread include abdominal nodes, peritoneum, bone, and lung. Median survival from time of diagnosis of metastatic disease is approximately 2.5 years. Only 10% of insulinomas are malignant. They are usually larger than 2.5 cm, whereas benign insulinomas are generally smaller than 2.5 cm. Most glucagonomas and VIPomas are malignant (60% 80%).
B. Treatment of advanced disease
Endocrine syndromes. The first goal of treatment must be to control endocrine syndromes.
Gastric acid suppression. The H+, K+-adenosine triphosphatase inhibitors omeprazole and lansoprazole successfully control gastric acid secretion in patients with gastrinoma. Optimal doses must be individualized and periodically re-evaluated. Gastric acid secretion in the hour preceding the next dose of omeprazole or lansoprazole should be less than 10 mEq in patients who have had no previous gastric surgery and less than 5 mEq in those who have had an acid-reducing procedure. The starting dose is 60 mg/day with both agents. Doses greater than 80 mg/day should be divided. Similar, newer agents in the same class of proton pump inhibitors are equally efficacious.
Insulin suppression. Diazoxide, an insulin release inhibitor, when given at 3 to 8 mg/kg/day PO divided in three doses (e.g., 50 to 150 mg PO t.i.d.), is the therapy of choice for hypoglycemia associated with insulinoma when dietary measures fail. A diuretic should be given with diazoxide to prevent water retention.
Octreotide acetate (Sandostatin). Octreotide acetate is a somatostatin analog that inhibits gut hormone secretion. It is generally useful for carcinoid and
Chemotherapy for advanced islet cell tumors. Streptozocin is the most active single agent, with a 50% response rate. Doxorubicin is also an active agent. The combination of streptozocin and doxorubicin was demonstrated to have a superior response rate (69%), time to tumor progression (20 months), and survival time (2.2 years) than the combination of streptozocin and fluorouracil or singleagent chlorozotocin in a North Central Cancer Treatment Group study. Current recommendations for treatment are as follows:
Streptozocin 500 mg/m2 IV on days 1 to 5 and doxorubicin 50 mg/m2 IV, on days 1 and 22. Repeat every 6 weeks. Renal impairment occurs in approximately 30% of patients receiving a streptozocin-based regimen; approximately one third of those with renal insufficiency have creatinine levels higher than 2 mg/dL. Nausea and vomiting occur in approximately 60% of patients. Leukopenia occurs in approximately 75%, but only 10% have a white blood cell count of less than 1,000/ L. Stomatitis is uncommon. Liver function test abnormalities may also occur. Deaths caused by treatment are rare.
Interferon- may diminish excess hormone secretion and induce shrinkage of tumors; some trials have reported 50% response rates.
Newer approaches include bevacizumab, sunitinib, and metaiodobenzylguanidine (MIBG) radiotherapy. MIBG ([131I]metaiodobenzylguanidine) is a radiolabelled substance similar to norepinephrine that in retrospective studies has been found to produce symptomatic and objective responses in patients with metastatic carcinoid.
III. Ampullary carcinomas
In up to 80% of people, the common bile duct and main pancreatic duct empty into a common channel, the ampulla of Vater. Periampullary carcinomas can be classified according to their site of origin. Type 1 tumors originate in the ampulla of Vater or the duodenal portion of the common bile duct. Type 2 carcinomas are duodenal tumors involving the ampulla of Vater. Type 3 tumors are mixed
IV. Carcinoma of the bile ducts (cholangiocarcinoma)
A. Epidemiology and natural history
The incidence of primary biliary tree carcinoma is approximately 2 per 100,000 population. Men are affected more commonly than women. Tumors occurmost often in late-middle-aged and elderly patients. They are associated with cholelithiasis, ulcerative colitis, obesity, liver flukes, exposure to thorium oxide (Thorotrast), primary sclerosing cholangitis, and congenital anomalies of the pancreaticobiliary tree. Patients present with obstructive jaundice, except for the occasional patient with a carcinoma identified at laparotomy for cholelithiasis. Approximately half the number of bile duct tumors are located proximally. Ten percent have multicentric involvement of the bile ducts. Local invasion is common. Liver involvement occurs in nearly half of these patients. Surgical cure is uncommon. Bypass procedures or intubation of the biliary tree may offer palliation to patients whose tumors cannot be resected. Radiation therapy may relieve proximal obstruction without intubation or a bypass procedure. Combined-modality therapy with radiation and fluorouracil should be considered in patients with an unresectable but localized cancer.
B. Chemotherapy for advanced disease
Few reports are available for this unusual tumor with low response rates and median survival of approaching 6 months. Fluorouracil has response rates of approximately 10%. Newer agents and combinations have been tried including gemcitabine alone, gemcitabine and fluorouracil, gemcitabine and oxaliplatin, capecitabine and oxaliplatin, and capecitabine and gemcitabine. Outside the context of a clinical trial in patients with an acceptable performance status, one of the following regimens is recommended:
Fluorouracil 500 mg/m2 IV push on days 1 to 5 every 4 weeks or 500 mg/m2 IV weekly
Fluorouracil 400 mg/m2 IV on days 1 to 5 and streptozocin 500 mg/m2 IV on days 1 to 5
Patients with a very good performance status can use a doublet such as GemOx with gemcitabine 1,000 mg/m2 on day 1 and 8 plus oxaliplatin 130 mg/m2 over 2 hours on day 8 every 3 weeks or
Gemcitabine 1,000 mg/m2, fluorouracil 600 mg/m2, leucovorin 20 mg/m2 IV days 1, 8, and 15 every 4 weeks
Capecitabine 1,500 mg/m2 daily in twice-daily divided doses on days 1 to 14 and gemcitabine 1,000 mg/m2 on days 1 and 8 every 3 weeks or
Capecitabine 1,500 mg/m2 daily in twice-daily divided doses on days 1 to 14 plus oxaliplatin 130 mg/m2 on day 1 every 3 weeks.
V. Carcinoma of the gallbladder
A. Epidemiology and natural history
Carcinomas of the gallbladder are seen predominantly in late-middle-aged and elderly women, with the highest incidence in native Americans and the populations of Central and Eastern Europe and Israel. The areas of high frequency also report a high incidence of cholelithiasis. Patients with porcelain or calcified gallbladders identified on radiographs have a 12% to 62% risk of cancer. Carcinoma of the gallbladder most commonly presents with pain, nausea and vomiting, and weight loss. Jaundice occurs in only one third of patients. Anorexia, abdominal distention, pruritus, and melena occur in some patients. One percent of patients undergoing cholecystectomy are found to have carcinoma of the gallbladder. Overall survival is poor; less than 5% of patients who undergo resection survive 5 years. When the tumor is histologically confined to the mucosa or submucosa, survival rates of 64% at 5 years and 44% at 10 years have been reported. Gallbladder carcinomas may invade locally into the bile ducts, liver, pancreas, stomach, or duodenum. They may also spread to regional lymph nodes and distantly to liver. There is no definitive data for adjuvant therapy but usually disease that is locally advanced is treated with a combination of radiation and chemotherapy (usually fluorouracil-based).
B. Chemotherapy for advanced disease
Although there is no data that indicates that chemotherapy treatment in advanced disease is better than the best supportive care alone, chemotherapy is often used for symptom palliation. Combinations are similar to those mentioned in the preceding text for cholangiocarcinomas. Gemcitabine alone or in combination with capecitabine or leucovorin are the commonly used regimens.
VI. Primary carcinoma of the liver
Primary carcinoma of the liver is rare in the United States. There are fewer than 10,000 new patients annually, accounting for less than 2% of all malignancies. However, it is the leading cause of cancer death in parts of Africa and Asia. Ninety percent of primary cancers of the liver are hepatocellular carcinomas or hepatoma; the remaining cancers include cholangiocarcinomas (~7%), hepatoblastomas, angiosarcomas, and other sarcomas. Histologic subsets of hepatocellular carcinoma have been recognized. Fibrolamellar carcinomas occur in young patients and are more likely to be resectable and cured. Hepatocellular carcinomas are more common in men than in women. The peak occurrence is during the sixth decade, with the highest incidence during the ninth decade. There appear to be three major factors associated with hepatocellular carcinoma: viral hepatitis B and C, alcohol abuse, and aflatoxin exposure. Seventy-five percent of patients with hepatocellular carcinoma have concomitant cirrhosis, and 4% to 20% of patients with cirrhosis have hepatocellular carcinoma at autopsy, depending on the population studied. Among the patients with hepatocellular carcinoma, 15% to 80% have hepatitis B surface antigenemia. In China, the incidence of
In Africa, the increased risk appears to be related to exposure to aflatoxin, which is produced by the fungi Aspergillus flavus and Aspergillus parasiticus during improper food storage. Three percent to 27% of patients with long-standing hemochromatosis develop hepatocellular carcinoma. Anabolic steroids have also been associated with hepatocellular carcinoma. Tumors induced by anabolic steroids may retain hormone dependence and regress after withdrawal of the steroid.
Patients with primary carcinoma of the liver commonly complain of right upper quadrant pain, abdominal distention, or weight loss. The pain is usually dull or aching but may be acute and radiate to the right shoulder. Fatigue, loss of appetite, and unexplained fever may occur. Patients with underlying cirrhosis may present with hepatic decompensation: new ascites, variceal bleeding, jaundice, or encephalopathy. Rarely, patients present with paraneoplastic syndromes: erythrocytosis is themost common; hypercalcemia, hyperthyroidism, and carcinoid syndrome have also been described. Physical findings include nodular hepatomegaly with an arterial bruit and hepatic rub. Extrahepatic spread occurs in approximately 50% of patients during the course of the illness. Twenty percent of patients have lung metastases.
C. Diagnostic evaluation and screening
-Fetoprotein levels are elevated in 70% of patients and are associated with a poor prognosis. Ultrasonography and CT have a high sensitivity when lesions are larger than 2 cm; however, small lesions are frequently missed. Magnetic resonance imaging is generally equivalent to CT but at a greater cost. FNA with cytology or biopsy usually confirms the diagnosis. Screening is controversial because although some studies have shown a survival benefit with screening, there are concerns about lead-time bias. Serial -fetoprotein measurements and liver ultrasonography every 4 to 6 months should be considered in high-risk patients with hepatitis B antigenemia and with hepatitis C and cirrhosis (Child-Pugh class A and/or patients who are suitable for hepatectomy).
Staging procedures should include a chest radiograph, CT of the abdomen, complete blood cell count, blood chemistry profile, and -fetoprotein measurement (Table 9.2). If these do not disclose unresectable cancer or sites of metastatic cancer, then CT of the chest and arteriogram (upper abdominal and hepatic) should be performed to guide surgical intervention and further screening for extrahepatic involvement.
E. Primary therapy
At presentation, 25% of patients with hepatocellular carcinoma have potentially resectable lesions. Liver transplantation may permit resection of small tumors in patients with advanced cirrhosis, and survival is similar to or better than that seen after resection without transplantation Patients with hepatocellular carcinoma who meet the Milan criteria for transplantation are those patients with one nodule less than 5 cm or two to three nodules less than 3 cm and have low recurrence rates after transplantation with 5-year
Table 9.2. TNM staging for hepatocellular cancer
F. Therapy of advanced hepatocellular carcinoma
Single agents. Numerous single agents have been tested in primary hepatocellular carcinoma: alkylating agents, antimetabolites, plant alkaloids, and cisplatin have been ineffective. Doxorubicin 60 mg/m2 IV every 21 days is recommended. This regimen results in a partial response rate of approximately 2%. There is really no effective chemotherapy in this situation and patients should be considered for clinical trials.
Burris HA, Moore MJ, Andersen J, et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997;15:2403 2413.
Collier J, Sherman M. Screening for hepatocellular carcinoma. Hepatology 1998;27:273 278.
Gastrointestinal Tumor Study Group. Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Cancer 1987;59:2006 2010.
Glimelius B, Hoffman K, Sjoden PO, et al. Chemotherapy improves survival and quality of life in advanced pancreatic and biliary cancer. Ann Oncol 1996;7:593 600.
Jensen RT, Fraker DL. Zollinger Ellison syndrome: advances in treatment of gastric hypersecretion and the gastrinoma. JAMA 1994;271:1429 1435.
Kalser MH, Ellenberg SS. Pancreatic cancer: adjuvant combined radiation and chemotherapy following curative resection. Arch Surg 1985;120:899 903.
Katz MH, Savides TJ, Moossa AR, et al. An evidence-based approach to the diagnosis and staging of pancreatic cancer. Pancreatology 2005;5:576 590.
Klinkenbijl JH, Jeekel J, Shamoud T, et al. Adjuvant radiotherapy and 5-fluorouracil after curative resection of cancer of the pancreas and periampullary region: phase III trial of the EORTC gastrointestinal tract cancer cooperative group. Ann Surg 1999;230(6): 776 782.
Louvet C, Labianca R, Hammel P, et al. Gemcitabine in combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic cancer: results of a GERCOR and GISCAD phase III trial. J Clin Oncol 2005;23(15):3509 3516.
Moertel CG, Hahn RG, O'Connell MS. Therapy of locally unresectable pancreatic carcinoma: a randomized comparison of high dose (6000 rads) radiation alone, moderate dose radiation (4000 rads 5-fluorouracil), and high dose radiation 5-fluorouracil. Gastrointestinal Tumor Study Group. Cancer 1981;48:1705 1710.
Moertel CG, Lefkopoulo M, Lipsitz S, et al. Streptozocin doxorubicin, streptozocin fluorouracil, or chlorozotocin in the treatment of advanced islet-cell carcinoma. N Engl J Med 1992;326:519 523.
Neoptolemos JP, Dunn JA, Stocken DD, et al. Adjuvant chemoradiotherapy and chemotherapy in resectable pancreatic cancer: a randomised controlled trial. Lancet 2001;358(9293):1576 1585.
Rao S, Cunningham D, Hawkins RE, et al. Phase III study of fluorouracil, etoposide and leucovorin (FELV) compared to epirubicin, cisplatin and fluorouracil (ECF) in previously untreated patients with advanced biliary cancer. Br J Cancer 2005;92(9):1650 1654.
Rich T, Harris J, Abrams R, et al. Phase II study of external irradiation and weekly paclitaxel for nonmetastatic, unresectable pancreatic cancer: RTOG-98-12. Am J Clin Oncol 2004;27(1):51 56.
Safford SD, Coleman RE, Gockerman JP, et al. Iodine-131 metaiodo benzylguanidine treatment for metastatic carcinoid. Results in 98 patients. Cancer 2004;101(9):1987 1993.
Tsukuma H, Hiyama T, Tanaka S, et al. Risk factors for hepatocellular carcinoma among patients with chronic liver disease. N Engl J Med 1993;328:1797 1801.