- TABLE OF CONTENTS
- Pancreatic Cancer
- Incidence and Epidemiology
- Etiology and Risk Factors
- Signs and Symptoms
- Screening and Diagnosis
- Staging and Prognosis
- Surgical treatment of resectable disease
- Surgical palliation
- Neoadjuvant and adjuvant therapies
- Treatment of unresectable lesions
- Treatment of metastatic adenocarcinoma
- Pancreatic Cystic Neoplasms
- Pancreatic Endocrine Tumors
- Carcinoid Tumors of the GI Tract
- Adrenocortical Carcinoma
- Suggested Reading
Surgical treatment of resectable disease
The rate of resection for curative intent ranges from 10% to more than 75%, with the higher percentage resulting from both a more aggressive approach and better preoperative staging for resectability. Also, there is growing evidence that patients with potentially resectable pancreatic cancer have a shorter hospital stay, reduced surgical mortality, and an overall better outcome if the surgery is performed at "high-volume" medical centers staffed by surgeons who treat approximately 16 operable cases per year.
Extended resections may include portal or superior mesenteric vessels, the colon, the adrenal glands, or the stomach. If resection of adjacent organs or tissues results in the conversion of a positive to a negative resection margin, it is of great potential benefit to the patient. With regard to the extent of lymph node dissection, several recent, prospective, randomized studies have shown an increase in postoperative morbidity, but with no improvement in overall survival in patients undergoing an extended lymph node dissection.
Currently, pancreatic lesions may be classified as resectable, borderline resectable, and unresectable. The definiton of borderline resectable tumors includes impingement or abutment on the SMV/portal vein, short segment venous occlusion with a suitable proximal and distal vein for reconstruction, minimal involvement of the gastroduodenal artery and hepatic artery, and < 180° involvement of the SMA. Criteria for unresectability include detection of distant metastases and circumferential involvement of the SMA, hepatic artery, or celiac artery.
Determination of resectability. The initial approach to surgery for pancreatic carcinoma includes a determination of resectability. This determination should be first made preoperatively with high-quality CT or MRI and, perhaps, EUS, although EUS is not as reliable as the other modalities for assessing vascular involvement. Operative determination of resectability includes careful examination of the liver, porta hepatis, and portal and superior mesenteric vessels. The head of the pancreas and uncinate process are mobilized by an extensive Kocher maneuver to evaluate the head of the pancreas. The SMA is palpated, and its relationship to the tumor is assessed. The hepatic artery and celiac trunk are examined to make certain there is no vascular encasement.
• Intraoperative biopsy—Most patients with resectable pancreatic tumors can successfully undergo pancreaticoduodenectomy without an intraoperative biopsy. A time-consuming frozen section interpretation may not be informative, and histologic confirmation may be impossible with small lesions associated with peritumoral pancreatitis. Most large series of pancreaticoduodenectomy for carcinoma include resections of benign pathology, based on clinical judgment. A negative fine-needle cytology should not deter an experienced surgeon from proceeding with resection. In patients without a histologic diagnosis, however, a biopsy is warranted in patients considered unresectable who are undergoing an attempt at resection.
• Whipple vs pylorus-preserving procedure—If the tumor is deemed to be resectable, a standard pancreaticoduodenectomy (Whipple procedure) or pylorus-preserving Whipple (PPW) procedure is performed. The PPW option theoretically eliminates the nutritional problems caused by a reduced gastric reservoir and gastric dumping, but has not been shown to alter long-term nutritional status. If there is any doubt about cancer proximity or blood supply to the pylorus, an antrectomy should be performed. If the tumor approaches the pylorus or involves the subpyloric nodes, classic antrectomy is recommended. Recent prospective randomized studies have shown therthat there is no significant difference in clinical or oncologic outcomes between pylorus-preserving and standard Whipple resection.
• Reconstruction technique—The most common reconstruction technique after a Whipple resection utilizes the proximal jejunum in an ante or retrocolic position to complete the pancreaticojejunostomy, which is followed by a hepaticojejunostomy and gastrojejunostomy. Pancreaticogastrostomy is also an effective and safe means of creating the anastomosis.
• Postoperative complications—Analysis of national databases indicate that the average post-operative mortality rate following pancreaticoduodenectomy is 6% to 8%, however studies from high-volume centers report a mortality rate of < 2%.
The leading causes of postoperative mortality include postoperative sepsis, hemorrhage, and cardiovascular events. Most of the septic complications arise from pancreaticojejunostomy leaks.
In many series, early delayed gastric emptying is the leading cause of morbidity for pylorus-preserving procedures. The number two cause of morbidity, seen in 5% to 15% of all patients, is a leak or fistula from the pancreatic anastomosis. Today, with appropriate drainage and nutritional support, more than 95% of pancreatic fistulas will heal using conservative measures.
An analysis of 200 patients who underwent resection of pancreatic adenocarcinoma in the era prior to adjuvant therapy found that the most important factors influencing long-term survival were the diameter of the primary tumor, status of the resected lymph nodes, and status of the resected margins. Patients with tumors < 3 cm in diameter had significantly longer median survival and 5-year survival rates (21 months and 28%, respectively) than those with tumors < 5% for those with positive nodes. Patients with no lymph node involvement had a 5-year survival rate of 36%, as compared with < 5% for those with positive nodes. Patients who underwent resections with negative margins had a 5-year survival rate of 26%, vs 8% for those with positive margins. The type of resection (pylorus-preserving vs standard Whipple procedure) did not influence survival.
• Body and tail tumors—Tumors in the body and tail of the pancreas are typically larger than tumors in the head of the pancreas and are often metastatic on presentation. For patients who are surgical candidates, resection employs a distal pancreatectomy with concomitant splenectomy. Because of the large size of these tumors, removal may require resection of adjacent organs. Of note, the rate of pancreatic leaks following distal pancreatectomy is approximately 30% to 40%, although, with appropriate treatment, they resolve using conservative measures.
Surgical palliation is also considered in patients undergoing exploration with curative intent. Jaundice, gastric obstruction, and pain may be alleviated by surgical palliation.
Biliary tract obstruction. Either a choledochojejunostomy or cholecystojejunostomy can be used to bypass the biliary obstruction. Recurrent jaundice and cholangitis are less likely to develop when the common duct is used for decompression. Nonoperative means of biliary decompression can often be accomplished with endoscopically placed expandable metallic stents. The use of metal stents should be reserved for patients who have been fully evaluated and deemed unresectable by a multidisciplinary team.
Duodenal obstruction. Although duodenal obstruction is rare as a presenting symptom, duodenal involvement may occur eventually in 25% of patients. Some investigators believe that prophylactic bypasses are safe and should be performed in all patients. One phase III trial supports prophylactic bypass, but the subject remains controversial.
Pain relief. Severe back pain may be an incapacitating symptom. Pain relief may be achieved by alcohol(Drug information on alcohol) injection of the celiac plexus, which may be performed intraoperatively, percutaneously, or endoscopically. An intraoperative injection of 25 mL of ethanol (95%) on both sides of the celiac axis will assuage tumor pain. (For further discussion of these techniques, see the "Pain Management" chapter.)
• Preoperative chemoradiation therapy—Several single-institution studies have evaluated the role of preoperative irradiation in conjunction with fluorouracil(Drug information on fluorouracil) (5-FU)- and gemcitabine(Drug information on gemcitabine)-based chemotherapy. In these studies, 60% to 80% of the lesions were completely resected 1 to 1.5 months after the completion of chemoradiotherapy. Median survival has ranged from 16 to 36 months, but no phase III trials have been conducted to evaluate preoperative therapy vs postoperative sequencing.
The appropriate radiation dose and volume are dependent on the planned concurrent chemotherapy. For patients undergoing concurrent 5-FU–based neoadjuvant chemoradiation, doses of 4,500 to 5,000 cGy are appropriate with inclusion of both the primary tumor and elective nodal coverage. Although interest in concurrent gemcitabine-based approaches was initially tapered by high rates of GI toxicity, treatment with full-dose gemcitabine (eg, 1000 mg/m2) appears to be well tolerated when radiation treatment volumes are limited to the primary tumor and omit elective nodal coverage. There are research initiatives to further address the role of neoadjuvant chemotherapy. For example, the recent data demonstrating the high response rate from the FOLFIRINOX regimen (5-FU, leucovorin, irinotecan(Drug information on irinotecan), and oxaliplatin(Drug information on oxaliplatin)) in metastatic pancreatic cancer are of interest to determine if this combination will enhance resectability when used in the neoadjuvant setting, and are the subject of a new NCI GI Intergroup study of FOLFIRINOX followed by chemoradiation and surgery.
A randomized phase III trial including 342 patients (Conroy T et al: N Engl J Med 2011) compared the combination chemotherapy regimen of 5-FU, irinotecan, leucovorin, and oxaliplatin (FOLFIRINOX) to standard gemcitabine, showing significant improvement in response rates (31.8% vs 11.3%, P < .001), progression-free survival (6.4 months vs 3.3 months, P < .001) and survival (11.1 months vs 6.8 months, P < .001). This regimen is best reserved for patients with good performance, given the potential toxicity .
• Postoperative chemoradiation therapy—The role of radiotherapy in the adjuvant setting has been called into question by several European trials. Criticism of these studies includes lack of quality control and suboptimal chemotherapy and radiotherapy schedules in the arms that included radiotherapy. In addition, for most GI sites, the preferred sequencing of modalities has been to give several months of full-dose multidrug chemotherapy before proceeding to radiotherapy with concurrent chemotherapy, a sequencing that was not used in the European trials.
A recently activated clinical trial being conducted by the European Organisation for Research and Treatment of Cancer (EORTC), Radiation Therapy Oncology Group (RTOG), and the Southwestern Oncology Group (SWOG) addresses all of these criticisms. Postoperative care for patients with cancer of the pancreatic head will be double-randomized, with all patients receiving six cycles of gemcitabine. The first randomization will be to receive or not receive concurrent erlotinib with gemcitabine. The second randomization will be whether to administer either no further therapy or radiotherapy (50.4 Gy/28 fractions) with concurrent infusional fluorouracil or equivalent capecitabine(Drug information on capecitabine) (Xeloda). All radiotherapy treatment plans will be centrally reviewed in advance of treatment. Thus, this trial is evaluating the contribution of postoperative radiotherapy using standard-of-care techniques with proactive quality control.
The American College of Surgeons Oncology Group (ACOSOG) Z05031 trial was a phase II trial that tested a regimen of radiation therapy, cisplatin(Drug information on cisplatin), interferon-alfa, and 5-FU in patients with resected pancreatic cancer. The trial demonstrated a median survival of 27.1 months, which was the longest survival reported with use of adjuvant therapy in a cooperative group trial. However, use of this regimen was associated with significant toxicity, and only 56% of patients completed the entire treatment course.
In addition, the GI Intergroupcompleted a randomized phase II trial to explore new combinations incorporating the monoclonal antibodies bevacizumab(Drug information on bevacizumab) (Avastin) and cetuximab(Drug information on cetuximab) (Erbitux), each given with gemcitabine; irradiation was given with oral capecitabine. Theresults suggest no added benefit with the addition of either bevacizumab or cetuximab.
Postoperative chemotherapy. A German randomized phase III trial including 368 patients with resected pancreatic cancer compared postoperative gemcitabine given for 6 months vs observation. Disease-free survival was significantly greater for patients receiving postoperative gemcitabine (13.4 months vs 6.9 months; P < .001), including for patients with either R0 or R1 resection. Overall survival, however, was not significantly different between the gemcitabine and control groups (22.1 months vs 20.2 months; P = .06). In an updated analysis, the benefits of gemcitabine as compared with observation remained. Despite an improvement in median survival of only 2 months noted among treated patients (22.8 months vs 20.2 months; P = .005), a 5-year survival of 21% in the treatment arm vs 9% in the observation arm was reported.
The European Study Group for Pancreatic Cancer (ESPAC)-3 was a multicenter randomized phase III trial which compared adjuvant fluorouracil vs gemcitabine following resection of pancreatic cancer. Following R0/R1 resection, patients were randomized to 5-FU (425 mg/m2) bolus on days 1-5 every 28 days vs gemcitabine (1,000 mg/m2) given intravenously on days 1, 8, and 15 every 4 weeks. Both regimens were continued for 6 months. A total of 1,088 randomized patients were enrolled after a minimum follow-up of 2 years, median survival for the 5-FU/leucovorin group was 23 months, compared with 23.6 months for the gemcitabine group. There was no statistical difference between the two study arms.
Another report from the ESPAC-3 periampullary adenocarcinoma trial including 428 patients determined that adjuvant chemotherapy (5FU or gemcitabine) vs observation did not show a survival difference in the primary analysis; however, multivariable analysis adjusting for prognostic variables showed a survival benefit favoring adjuvant chemotherapy.
Locally advanced but potentially resectable lesions. These lesions comprise 10% to 15% of cases presenting to physicians. Data from phase II preoperative chemoradiotherapy trials indicate that trimodal therapy is crucial for margin-free resection and long-term survival. A meta-analysis from the ESPAC-1 trial showed that chemotherapy alone is ineffective for patients who have had resection for microscopic disease at a margin (R1), thus adding further support for both chemotherapy and radiation therapy for these borderline resectable patients. As mentioned, the NCI GI Intergroup is conducting a study of FOLFIRINOX followed by chemoradiation followed by surgery for borderline resectable patients.
Irradiation. Radiation therapy can prolong and/or improve quality of life in some patients with unresectable adenocarcinoma of the pancreas. It is better combined with chemotherapy. Long-term survival is, unfortunately, highly unusual.
Chemoradiation. The addition of chemotherapy to radiation therapy has been shown to improve the survival of patients with unresectable pancreatic adenocarcinoma. In a Gastrointestinal Tumor Study Group (GITSG) trial of unresectable disease conducted in the 1970s, moderate-dose radiation (4,000 cGy) with 5-FU chemotherapy significantly improved survival, as compared with higher doses of radiation (6,000 cGy) and no chemotherapy (median survival, 9.6 vs 5.2 months). Given the study era, radiotherapy was administered using two dimensional split-course techniques. GITSG has also compared chemotherapy plus irradiation with chemotherapy alone and demonstrated a significant improvement with combined-modality therapy (median survival, 42 vs 32 weeks).
A systematic review of 11 trials including 794 patients with locally advanced pancreatic cancer using radiation/combined-modality therapy showed a survival benefit for chemoradiation over irradiation alone, but there was no significant advantage for chemoradiation followed by chemotherapy compared with chemotherapy alone.
The Eastern Cooperative Oncology Group (ECOG) conducted a randomized, prospective trial in patients with locally advanced, unresectable, nonmetastatic pancreatic adenocarcinoma. Patients were randomized to receive either gemcitabine monotherapy or radiation therapy, given concurrently with and followed by gemcitabine. The overall survival was 11.1 months in the combination-therapy group and 9.2 months in the chemotherapy-only group (P = .034). Despite the statistical significance of these findings, this trial was plagued by poor accrual and was terminated early.
Based on these data, many practitioners would favor several months of gemcitabine-based chemotherapy before proceeding to chemoradiotherapy. This allows early delivery of optimal chemotherapy and spares the patient destined to rapid dissemination of disease from the morbidity associated with use of concurrent chemoradiotherapy. If radiotherapy is given concurrently with 5-FU, elective nodal doses of 4,500 cGy and primary tumor doses of 5,400 cGy based on bowel tolerance are appropriate. In all cases, delineation of the target volume should incorporate all available diagnostic imaging. Image-guided radiation therapy (IGRT) is appropriate to ensure patient setup and daily target visualization.
• Approaches under investigation—At present, clinical investigators are considering a variety of chemoradiation therapy approaches including the addition of new targeted therapies in early phase trials. The benefit of irradiation for patients with locally advanced disease, however, remains of interest for clinical trials because of toxicity concerns and the relatively brief survival rates.
Objective response by size criteria is difficult to achieve with radiotherapy. This may be due to the substantial sclerotic component associated with pancreatic adenocarcinomas rather than true radiation resistance. Regardless, until it becomes possible to achieve a high incidence of objective clinical responses, it will be difficult to persuade surgeons to attempt resections after chemoradiotherapy for unequivocal clinical T4 lesions. Recent results reported from the University of Michigan and Rush University Medical Center are of particular interest in this regard. The Michigan group has a long interest in concurrent radiotherapy and gemcitabine because of the drug's radiosensitizing properties. Previous work from the group suggested that normal tissue tolerance limits the ability to give full doses of both radiotherapy and gemcitabine concurrently. However, in their most recent work, these investigators have taken advantage of technical developments in radiation oncology by using IMRT, breath-holding techniques to limit respiratory excursion, and online daily setup verification to treat macroscopic disease with tight planning target volumes. Radiation doses were successfully escalated to 5,500 cGy in 25 fractions over 5 weeks with gemcitabine (1,000 mg/m2 over 100 minutes) given on weeks 1, 2, 4 and 5. Objective responses were seen in 52% of 27 cases, with a mean overall survival of 23 months, including two cases converted to resectable status with minimal residual disease at surgery.
The dose of gemcitabine that can be given concurrently with irradiation depends on the volume and dose of radiation. If full doses of gemcitabine (1,000 mg/m2/wk) are given concurrently with irradiation, the volume of radiation must be markedly reduced to avoid unacceptable GI toxicity. In the Michigan series, the treatment volume included the primary tumor only plus a 1-cm planning target volume (PTV) margin, with omission of nodal coverage.
Pancreatic adenocarcinoma is still one of the most frustrating, resistant solid neoplasms to treat, and therapy for metastatic disease remains palliative. Few agents have demonstrated activity in > 10% of patients diagnosed with this disease.
Chemotherapy. Because metastatic pancreatic carcinoma is incurable, the anticipated risks of chemotherapy, which are often substantial, must be balanced against the gains that may be achieved; unfortunately, these are few. Patients who are debilitated due to their underlying or comorbid disease should not be offered chemotherapy, because their likelihood of deriving any benefit is exceedingly slim. However, patients who desire therapy and who, while symptomatic, still have good performance status may be offered "standard" chemotherapy (Table 2), or, if possible, they should be encouraged to participate in a clinical trial.
• 5-FU—Historically, single-agent 5-FU has been associated with a response rate of up to 25% in pancreatic cancer. The use of 5-FU, doxorubicin(Drug information on doxorubicin), and mitomycin(Drug information on mitomycin) (FAM) and 5-FU plus doxorubicin offer no advantage over 5-FU alone. 5-FU plus leucovorin appears to be ineffective.
• Gemcitabine—Gemcitabine is indicated for the treatment of locally advanced or metastatic pancreatic adenocarcinoma. Gemcitabine was compared with 5-FU in a group of 126 previously untreated patients and showed a small, but statistically significant, improvement in response rate. Median survival in the gemcitabine group was 5.7 months, with 18% of patients alive at 12 months, as compared with 4.4 months in the group receiving 5-FU, with 2% of patients alive at 12 months. Perhaps more importantly, clinical benefit response (a composite measurement of pain, performance status, and weight) occurred in 23.8% of the gemcitabine-treated group vs 4.8% of the 5-FU–treated group. Due to its palliative potential, gemcitabine has become the standard of care for patients with unresectable, metastatic pancreatic adenocarcinoma.
• Combination therapy—There have been a number of attempts to improve the therapeutic outcome for patients with metastatic pancreatic cancer by comparing promising combinations of agents in randomized clinical trials. Unfortunately, the results have been disappointing. The ECOG compared gemcitabine with or without 5-FU, demonstrating a median survival of 5.4 months for gemcitabine vs 6.7 months for the combination; however, this difference was not statistically significant. Another trial explored the addition of irinotecan to gemcitabine. There was no survival benefit when this regimen was compared with gemcitabine alone, although the combination did increase the tumor response rate (16.1% vs 4.4%; P < .001).
A meta-analysis of 15 randomized trials showed a significant survival benefit for patients with advanced pancreatic cancer and a good performance status who received gemcitabine either with a platinum analog (hazard ratio [HR] = 0.85; P = .01) or a fluoropyrimidine (HR = 0.9; P = .03).
A European randomized trial of 319 patients with advanced pancreatic cancer compared capecitabine plus gemcitabine vs gemcitabine alone and showed no difference in overall survival (8.4 vs 7.2 months; P = .234). The combination showed improved median overall survival in patients with good performance status (10.1 vs 7.4 months; P = .014).
A phase III study of 565 patients compared gemcitabine vs the combination of gemcitabine plus the multitargeted antifolate pemetrexed(Drug information on pemetrexed) (Alimta), and demonstrated a significant response benefit with the combination (14.8% vs 7.1%; P = .004). However, overall and progression-free survival rates were comparable. There was increased hematologic toxicity with the combination.
Three Intergroup metastatic pancreatic cancer trials have been reported. ECOG completed a trial of gemcitabine vs fixed-rate infusion gemcitabine vs fixed-rate gemcitabine plus oxaliplatin, accruing 832 patients. At a median follow-up of 12.2 months, neither of the two investigational regimens was significantly better than standard gemcitabine (both groups had approximately 1 month longer median survival). Cancer and Leukemia Group B (CALGB) compared bevacizumab plus gemcitabine vs gemcitabine alone, and SWOG evaluated gemcitabine with or without cetuximab. These studies reported no difference in outcome with the addition of an antibody.
In a phase III trial, 607 patients with metastatic pancreatic cancer were given gemcitabine and erlotinib (Tarceva) with either bevacizumab or placebo. The results, published in 2009 by Van Cutsem et al, showed that patients receiving bevacizumab had significantly longer progression-free survival, but their improvement in overall survival was not significant. In the Charité Onkologie Clinical Studies in GI Cancer (CONKO) 003 trial, presented at the 2008 ASCO meeting, 168 patients who progressed after gemcitabine therapy received 5-FU and leucovorin with or without oxaliplatin (Eloxatin). The study demonstrated a significant advantage in progression-free survival favoring patients receiving oxaliplatin.
The NCIC has presented a randomized phase III study comparing gemcitabine with or without erlotinib in 530 patients with metastatic pancreatic cancer. The combination produced improvement in both overall (6.24 vs 5.91 months; HR = 0.82; P = .038) and progression-free survival (HR = 0.77; P = .007). As with other epidermal growth factor receptor-targeted agents, rash was associated with response. Diarrhea increased with the combination.
Patients with good performance status who progress after gemcitabine-based therapy may respond to second-line therapies including FOLFOX, FOLFIRI, and capecitabine, as noted in small studies.
• Agents with marginal activity—Agents with marginal activity include mitomycin, doxorubicin, ifosfamide(Drug information on ifosfamide), streptozocin (Zanosar), and docetaxel(Drug information on docetaxel) (Taxotere). To date, monoclonal antibody therapy and hormonal manipulation have been ineffective.
• Novel approaches—A progressively better understanding of the molecular biology of pancreatic cancer has revealed numerous new therapeutic targets. Some agents currently being studied include vaccines, mTOR inhibitors, hedgehog pathway inhibitors, and inhibitors of multiple tyrosine kinases as examples. In addition, Abraxane (nanoparticle paclitaxel(Drug information on paclitaxel)) and EndoTAG-1, a cationic liposomal paclitaxel agent, have shown early promise when combined with gemcitabine, and are currently being tested in a randomized phase III trial.