Epidemiology and Risk Factors

Although the 10^th most common cause of cancer in the United States, pancreatic cancer continues to far outpace its incidence rate with respect to mortality. In 2011, an estimated 37,660 patients died of pancreatic cancer, making it the fourth most common cause of cancer death in men and women. The incidence and mortality rates have unfortunately remained almost unchanged for the past several decades. The majority of these tumors (85%) are adenocarcinomas arising from the ductal epithelium. In this presentation we will only focus on pancreatic adenocarcinoma.

The following factors have been associated with increased risk of pancreatic adenocarcinoma:

• Tobacco use
• Obesity
• Chronic pancreatitis

Males and females have essentially equal incidence. African Americans have a higher incidence than white Americans. Approximately 90% of pancreatic adenocarcinoma is sporadic, while 10% is familial. Genes most commonly associated with familial cases of pancreatic adenocarcinoma are BRCA-1, BRCA-2, and CDKN2A/p16.

Molecular Pathogenesis

Multiple genetic mutations have been found in patients with pancreatic adenocarcinoma. KRAS is one of the most commonly mutated genes in pancreatic adenocarcinoma, occurring sporadically in more than 90% of tumors.(1) This is an activating mutation and appears to happen early in the pathogenesis of pancreatic cancer as it is also found in non-invasive precursors such as intraductal papillary mucinous neoplasms.

Loss of function of tumor suppressor genes is also frequently found. The most common lost gene is the p16/CDKN2A gene on chromosome 9p, which is somatically inactivated in almost all pancreatic cancers.(2) The loss of CDKN2A can be somatic or as part of the familial atypical mole melanoma syndrome (FAMM) in which patients are at increased risk of both pancreatic cancer and melanoma. Other less common mutations are TP53, SMAD4, and BRCA, especially BRCA-2.

Clinical Manifestations and Diagnosis
Pancreatic cancer is typically diagnosed late in the course of disease due to its mostly asymptomatic course of development. Eventual symptoms include jaundice (from pancreatic head obstruction of the biliary ducts), abdominal pain, weight loss, and sometimes development of diabetes because of the involvement of the normal pancreas by non-insulin secreting malignancy. Therefore, development of new onset adult diabetes associated with weight loss should instigate further evaluation for possible underlying pancreatic malignancy.

Initial workup usually includes CAT scan (CT) or right upper quadrant (RUQ) ultrasound if the patient is presenting with new onset jaundice. Diagnosis requires tissue biopsy typically obtained by endoscopic retrograde cholangiopancreatography (ERCP) or by an endoscopic ultrasound (EUS)-guided biopsy. ERCP can be both diagnostic and therapeutic when employed to place stents for relieving pressure in the biliary duct. EUS, as explained below, can also be therapeutic if used for celiac plexus neurolysis and pain control.

There has been great interest in the use of serum tumor markers for the diagnosis of pancreatic cancer. CA 19-9, a sialylated Lewis (a) blood group antigen, is commonly expressed and shed in hepatobiliary or hepatic cancers, and is the most tested marker in the setting of pancreatic cancer. CA 19-9 is not, however, tumor specific or 100% sensitive.(3) It can be elevated in benign conditions such as pancreatitis or biliary obstruction from other causes such as cholelithiasis. It can also be falsely negative in individuals who are negative for Lewis antigens. Therefore, CA 19-9 is not used to make the diagnosis of pancreatic cancer. CEA, carcinoembryonic antigen, is also sometimes elevated in patients with pancreatic cancer but is more commonly utilized for colorectal cancer, as it is not as sensitive or specific for pancreatic cancer.

Staging
The American Joint Committee on Cancer (AJCC) has published TNM staging criteria for pancreatic cancer, similar to staging systems for other solid tumors (http://www.cancer.gov/cancertopics/pdq/treatment/pancreatic/HealthProfessional/page3). However, in practice, a more simplified staging system, based on the resectability of the tumor, is used. Following imaging by triple phase abdominal CT or MRI and typically chest CT, the disease is classified as: 1) resectable, 2) borderline resectable (tumors involving nearby structures and not clearly resectable or unresectable), 3) locally advanced (involving nearby structure to the point of being clearly unresectable despite absence of metastatic disease), and 4) metastatic.

Treatment - Surgical Resection
Surgical resection is currently the only possibly curative technique for pancreatic cancer. However, at the time of diagnosis, nearly 80% of patients cannot proceed with surgery due to either vascular involvement or metastatic spread. Institutions differ in their use of surgery for pancreatic cancer with various degrees of local involvement. Important structures near the pancreas that are closely analyzed for involvement on imaging are the celiac trunk, superior mesenteric artery (SMA), superior mesenteric vein (SMV), hepatic artery and the portal vein. Patients with clearly resectable tumor have a clear fat plane around the celiac axis, hepatic artery and SMA and no evidence of SMV or portal vein abutment by the tumor. While tumor circumventing more than 180 degrees of the SMV may be resected and reconstructed, such involvement of the arteries (SMA, celiac axis, etc.) is not amendable to reconstruction and is deemed unresectable.

The most common procedure performed is the Whipple, which involves resection of the pancreatic head in addition to the first and second portions of the duodenum, the common bile duct and the gallbladder. The distal portion of the stomach is also removed in the “classic” Whipple. The following YouTube video provides an explanation of the procedure: (http://www.youtube.com/watch?v=x4NJTNLq3xw). A different surgery is needed for pancreatic cancers in the pancreatic body or tail, which involves distal pancreatectomy, often concurrent with splenectomy, for complete dissection of lymph nodes present along the course of the splenic arteries.

Role for Adjuvant Therapy
Even in early stage pancreatic cancer that has undergone optimal resection, median survival ranges from 15 to 19 months with a 5-year survival of only 20%.(4) Therefore, adjuvant chemotherapy, radiation therapy or concurrent chemoradiation have been used in attempts to cure microscopic disease post resection. Unfortunately, the results for surgery alone have not been impressive. One of the positive trials was the large phase III CONKO-001 trial in which 368 patients were randomly assigned to adjuvant gemcitabine for six months versus observation following macroscopically complete resection. Eighty percent of patients had R0 resection (no microscopic disease at the margins of resected malignancy). Even so, underscoring the difficulty of achieving a cure in pancreatic cancer patients, 74% of the patients in the gemcitabine group and 92% of the control group developed recurrent disease during median follow-up of 53 months. Median disease-free survival was significantly improved in the patients receiving gemcitabine (13.4 months vs. 6.9 months, p<0.01). No overall survival difference was shown.(5)

Given similar modest improvements seen in other trials, the NCCN (National Comprehensive Cancer Network) currently endorses the use of adjuvant chemotherapy with either gemcitabine or 5-FU/leucovorin, although gemcitabine is preferred because of its favorable toxicity profile. Concurrent chemotherapy and radiation are also accepted and commonly practiced in the United States.(6) At present, adjuvant studies comparing chemotherapy only versus chemotherapy combined with radiation are ongoing.

Role for Neoadjuvant Therapy
There has also been interest in neoadjuvant therapy for disease deemed resectable. Theoretically, neoadjuvant therapy for resectable disease should downsize the tumor and thereby increase the possibility of margin-free resection. Furthermore, neoadjuvant therapy in seemingly resectable disease may allow time for visualization of progressive/metastatic disease on subsequent imaging during therapy that was presumably there at the time of diagnosis but was too small to visualize on imaging. By this rational, we would spare some patients from a surgery that wouldn’t have benefited them.

Most studies of neoadjuvant therapy have been retrospective, while the few prospective studies show improved completion rate of chemotherapy and decreased rates of surgery due to presence of more advanced disease on subsequent imaging. No survival benefit, however, has been demonstrated in neoadjuvant therapy versus adjuvant therapy. Currently, researchers are recruiting for a phase III study that compares adjuvant therapy to a combination of neoadjuvant and adjuvant therapy in resectable disease. This trial should help answer the question of the role of neoadjuvant therapy in resectable pancreatic cancer (http://clinicaltrials.gov/ct2/show/NCT01314027?term=NCT01314027&rank=1).

Neoadjuvant therapy in the setting of borderline resectable disease makes the most sense, as it would potentially render a borderline case resectable. No prospective phase III trials have addressed this scenario, however. The current approach varies by institution but it appears that most physicians prefer neoadjuvant therapy as opposed to immediate surgery. Fortunately, there are several ongoing trials assessing R0 resection rate following neoadjuvant chemotherapy in patients with borderline resectable disease to establish the possible benefits of this approach (for example, Clinicaltrials.gov NCT01359007, using FOLFIRINOX chemotherapy).

Systemic Treatment for Locally Advanced and Metastatic Disease
Systemic chemotherapy is used in the management of non-resectable patients. It is mostly limited to patients with good performance status (ECOG 0-1). Gemcitabine was established as the first line option both for survival benefit, as well as for relief of symptoms, with the publication of a randomized trial in 1997 comparing it to 5-FU, the gold standard at that time.(7) The combination of gemcitabine with various cytotoxic and molecularly targeted agents has failed to build on benefit from gemcitabine alone. The only combination currently approved by the FDA is gemcitabine with erlotinib, which showed a very modest improvement in survival from 5.91 months to 6.24 months in a phase III study comparing it to gemcitabine alone.(8)

Until recently, gemcitabine and the less commonly used combination with erlotinib were the only regimens for unresectable pancreatic cancer. In May of 2011, a French group published the results of the randomized phase III PRODIGE trial that showed dramatic improvement in both median progression-free survival (6.4 months vs. 3.3 months) and median overall survival (11.1 months vs. 6.8 months) in favor of the group receiving FOLFIRNOX vs. the group receiving gemcitabine.(9) FOLFIRINOX is a biweekly regimen consisting of 5-FU/leucovorin, irinotecan and oxaliplatin. It is more toxic than gemcitabine, however, with high rates of neutropenia and diarrhea among other side effects. Therefore, only patients with high performance status should be offered this regimen.

Chemoradiation for Locally Advanced Disease
It is unclear whether chemoradiation offers benefits compared with chemotherapy alone in patients with non-metastatic locally advanced disease. It may be helpful for patients with poorly controlled pain or local obstructive symptoms but it is not clear whether it prolongs overall survival. Both 5-FU and gemcitabine have been used as radiosensitizers and are endorsed by the NCCN guidelines.

Palliation of Locally Advanced and Metastatic Disease
Somewhat unique to pancreatic cancer is the requirement for substantial palliative procedures in the course of illness. Symptomatic biliary obstruction, for example, occurs in more than 65% of patients with pancreatic cancer.(10) Cancer-related pain is also a major symptom that sometimes cannot be managed on oral pain medications alone. Therefore, an understanding of currently available options for palliation of these symptoms is crucial to management of patients with pancreatic cancer.

Biliary or Gastric Obstruction
The majority of unresected patients will eventually develop biliary obstruction due to growth of the tumor in the pancreatic head. In most cases, a biliary stent – either plastic or metal – can be placed to alleviate the obstruction. Metal stents are wider in diameter so remain patent longer. However, they become embedded in the biliary duct and if not removed during eventual resection, they may not be easily replaced upon re-occlusion. Plastic stents, on the other hand, are typically occluded within three months of insertion and are usually replaced at regular intervals as a result. One randomized trial of 100 patients at a single center comparing plastic stents and self-expanding metal stents showed the median patency time of 1.8 months vs. 3.6 months in favor of metal stents.(11) Metal stents are also typically more expensive than plastic stents.

Sometimes a biliary stent cannot be placed as a result of ERCP failure due to the mass obstructing the gastric outlet. In this situation, percutaneous biliary drainage is often needed. An open biliary-enteric bypass can also be performed, usually if patients are undergoing surgery for resection but found to have unresectable disease during the operation.

Symptomatic gastric outlet obstruction also occurs in 10-25% of patients. Similar to biliary stents, an enteral stent can relieve the obstruction. Another alternative for a fit patient with a life expectancy of 3-6 months is duodenal bypass surgery with or without a jejunostomy tube.(12)

Cancer-Associated Pain
Most patients with pancreatic cancer experience cancer-related pain, often secondary to the cancer’s infiltration of retroperitoneal nerves of the upper abdomen. When usual narcotic pain medications are not enough or titration is hampered by opioid-related adverse events, celiac plexus neurolysis should be considered. The celiac plexus transmits pain sensation originating from the pancreas and most of the abdominal viscera except for the left colon, rectum and pelvic organs. It was originally described in 1919 by Kappis et al. in Germany. It is now done commonly for pancreatic cancer or pancreatitis either percutaneously, surgically or under endoscopic ultrasound (EUS) guidance. The procedure involves injection of a combination of a local anesthetic, such as bupivacaine, and a neurolytic, usually an alcohol. Several randomized trials have shown improved pain relief in celiac plexus neurolysis, In a recent study of 96 patients, half were randomized to EUS-guided celiac plexus neurolysis at the time of diagnostic EUS if unresectable adenocarcinoma was confirmed.(13) These patients reported better pain control at three months suggesting that neurolysis at the time of diagnosis may be beneficial and eliminate need for a repeat EUS. The article by Arcidiacono et al. provides a description of the EUS guided method with associated imaging.(14)

Pancreatic Insufficiency
Pancreatic insufficiency occurs in up to 94% of patients undergoing pancreatic surgery.(15) It also occurs in unresected patients as a result of tumor-induced damage to the pancreatic parenchyma or blockage of the pancreatic duct. Symptoms of pancreatic insufficiency result in steatorrhea, abdominal cramps, weight loss and malnutrition. Therapy is based on symptoms, rather than diagnostic testing, given the high prevalence of this problem in patients with pancreatic cancer. Enteric-coated preparations of pancreatic enzymes are taken orally (25,000 to 75,000 units of lipase for a main meal and 10,000 to 25,000 units for a snack) with half of the dose taken at the start of the meal and half taken in the middle of the meal.(15)

Thromboembolic Disease
Venous thromboembolic (VTE) disease risk is elevated in all patients with cancer and pancreatic cancer is a notorious culprit. The NCCN panel recommends low molecular weight heparin (LMWH) instead of warfarin (Coumadin) for patients with pancreas cancer who develop VTE. This recommendation is supported by two landmark studies, one in a group of patients with different types of cancer and another in patients with pancreatic cancer. The first study showed a two-fold decrease in incidence of recurrent VTE at six months in patients with advanced cancer and a prior VTE who were treated with LMWH compared to an oral anticoagulant (Coumadin).(16) In another study, prophylactic LMWH was used in patients with advanced pancreatic cancer and the regimen significantly decreased the risk of developing symptomatic VTE.(17) This study is currently published in abstract form only and prophylactic treatment is not yet endorsed by the NCCN.

Other Symptoms
Depression and malnutrition are common in patients with pancreatic cancer. Consulting palliative care is recommended to help manage these symptoms, as well as to arrange for hospice care in later stages of the disease.

Summary

Pancreatic cancer continues to be one of the most lethal cancers in the twenty-first century despite recent advances. Surgery remains the only chance at cure but most completely resected patients still die of their disease. Adjuvant therapies are usually used to improve the results of surgery alone. Systemic chemotherapy with gemcitabine and more recently, FOLFIRINOX, are recommended when resection is not an option. Until more effective therapies are found, managing symptoms such as pain and biliary obstruction are key elements in the management of patients with pancreatic cancer.

Suggested Reading
NCCN Guidelines on Treatment of Pancreatic Adenocarcinoma (http://www.nccn.org/professionals/physician_gls/f_guidelines.asp)