Pancreatic cancer

• Pancreatic ductal adenocarcinoma (PDAC, the common exocrine cancer of the pancreas) is usually diagnosed late, after it has invaded vessels or spread, because it produces few early symptoms and the tumour sits deep behind the stomach.
• It accounts for over 85 percent of pancreatic cancers and arises from the duct cells; roughly two thirds start in the head of the pancreas, where they often first show as obstructive jaundice.
• The classic picture is painless obstructive jaundice, or epigastric pain that bores into the back, with steady weight loss, anorexia and, increasingly, new diabetes in an older adult.
• Multiphase pancreatic-protocol CT is the best initial test; endoscopic ultrasound (EUS) with biopsy, MRI/MRCP and ERCP for biliary stenting complete the work-up and staging.
• Curative treatment is possible only in the minority whose tumour is resectable: Whipple pancreaticoduodenectomy for small head tumours, or distal pancreatectomy, usually followed by adjuvant chemotherapy. Five-year survival overall is only about 5 to 15 percent; even after complete resection 5-year survival is only about 17 to 24 percent.

By the end of this page you should be able to:

1. Describe who gets pancreatic cancer, where it arises in the pancreas, and why it is usually found late.
2. Explain how smoking, chronic pancreatitis, obesity, diabetes and hereditary syndromes raise risk, and which cystic precursors can become cancer.
3. Link each symptom and sign to its mechanism (jaundice, back pain, weight loss, Trousseau, Courvoisier).
4. Choose the right first, confirmatory and staging investigations for a patient with suspected or proven disease.
5. Classify a tumour as resectable, borderline resectable, locally advanced or metastatic, and state the management for each.
6. Outline adjuvant, neoadjuvant and metastatic chemotherapy, biliary stenting, enzyme replacement and palliative care.
7. Compare pancreatic ductal adenocarcinoma with pancreatic neuroendocrine tumours, chronic pancreatitis, cholangiocarcinoma and ampullary cancer, and list the screening criteria for high-risk individuals.

About 85 to 90 percent of cancers arising in the pancreas are ductal adenocarcinomas; most of the rest are rarer subtypes (acinar cell carcinoma, cystadenocarcinoma) or neuroendocrine tumours. ^[2] The tumour is usually a firm, pale, poorly defined mass in a gland surrounded by dense fibrous stroma; that desmoplasia is part of why it is detected late and why it resists drugs.

Exocrine pancreatic cancer is classified clinically by where it is in the body (head, body or tail) and, more importantly, by resectability: resectable, borderline resectable, locally advanced and metastatic. That system, not the simple TNM stage alone, is what drives treatment.

Pancreatic cancer is uncommon but highly lethal. Globally it is the seventh leading cause of death among cancers, with about 466,000 deaths a year, and European incidence and mortality rates are among the highest in the world. ^[2] In the United States it is the third leading cause of cancer death, and PDAC is projected to become the second leading cause of cancer-related death by 2030. ^[6]

• Age. It is rare before the age of 50 and most common from the sixth decade onward, with peak incidence in the 70s to 80s. ^[2] Hereditary forms can present in younger adults, especially when there is an affected first-degree relative diagnosed before age 50. ^[2]
• Sex. Men are affected more often than women (GLOBOCAN 2020 age-standardised incidence 5.7 versus 4.1 per 100,000). ^[2]
• Site in the pancreas. Most ductal adenocarcinomas arise in the head of the pancreas, so they often present first as obstructive jaundice; body and tail tumours present later and are usually large at diagnosis. ^[3]
• Trend. The number of cases is projected to rise, with roughly 38,000 additional cases across Europe by 2040. ^[2]
• Survival. Although the 5-year survival rate has improved from 5-6 percent in the last decade to about 12 percent, overall 5-year survival is still only about 5 to 15 percent. ^[4][6] Only about 20 percent of tumours are resectable when first found. ^[6]

Pancreatic cancer is a disease of ageing, tobacco, obesity, longstanding inflammation and, in a minority, inherited errors. About 80 percent of new diagnoses are advanced or metastatic, and most cases have no single identifiable cause beyond these broad exposures. ^[4]

• Age and male sex as above. ^[2]
• Family history. The risk is about 9 times higher with one affected first-degree relative, 32 times higher with three or more, and rises further if a relative was diagnosed before age 50. ^[2]
• Hereditary syndromes and susceptibility genes. Mutations in BRCA2 raise relative risk 3- to 4-fold; BRCA1 is also implicated in many centres. PALB2 is a core homologous recombination partner of BRCA2. Other syndromes: (i) Peutz-Jeghers syndrome (STK11/LKB1) raises relative risk up to 132-fold, the highest of any of them; (ii) hereditary pancreatitis with PRSS1 mutations carries very high risk (RR 69); (iii) familial atypical multiple mole melanoma (FAMMM) with CDKN2A (p16) mutations (RR 13-22); (iv) Lynch syndrome (mismatch repair genes) also carries increased risk and should prompt surveillance when a first-degree relative has pancreatic cancer. ^[1][2]
• Hereditary pancreatic cancer overall. Hereditary factors account for about 10 percent of cases. ^[2] Pathogenic germline variants are found in roughly 5-6 percent of unselected patients as BRCA1/2 or PALB2 mutations, rising to 5-19 percent in those with a family history of pancreatic cancer. ^[4]
• Chronic pancreatitis itself is a risk factor (see below), with relative risks as high as 69 in hereditary pancreatitis.

• Smoking. The best-established modifiable risk factor; summary relative risk about 1.66 versus non-smokers. ^[2]
• Obesity. Risk rises by 10 percent or more for each 5 kg/m2 increase in body mass index, and is higher still in established obesity; in a large pooled analysis 20-50 percent higher. ^[2]
• New-onset diabetes in an older adult, especially accompanied by weight loss, is an important presentation as well as a risk factor; long-standing diabetes is associated with a 1.5- to 2-fold increase in risk. ^[2]
• Chronic alcohol producing chronic pancreatitis, dietary patterns and poor physical activity contribute through obesity and diabetes. Heavy occupational exposure to nickel and some chemicals may matter. ^[2]

Some pancreatic cysts are neoplastic and can progress over years to invasive cancer; these are the most tractable targets for early detection.

• Intraductal papillary mucinous neoplasm (IPMN) is a mucin-producing cystic tumour of the pancreatic ducts. Main-duct and mixed IPMNs carry the highest risk of progression; side-branch IPMNs of small size can be watched. ^[1]
• Mucinous cystic neoplasm (MCN), with ovarian-type stroma and no communication with the duct, usually occurs in women and can progress to invasive cancer. ^[1]

Red-flag cyst features are jaundice with cystic lesions in the head, an enhancing solid component, or a main pancreatic duct 10 mm or more in diameter: these people should be referred for resection. ^[1]

The leading model is the PanIN progression: a normal duct cell becomes a low-grade precancer, then high-grade dysplasia, then invasive adenocarcinoma over many years, gaining KRAS, p53, CDKN2A and SMAD4 alterations. That long latency is why detection in a precursor stage is possible, at least in theory; it also explains why most patients present only when the tumour has reached vessels or metastasised.

• Obstruction. A mass in the head blocks the common bile duct and pancreatic duct. Bile backs up (jaundice, pale stools, dark urine, pruritus); pancreatic enzymes and bicarbonate cannot reach the gut (steatorrhea, weight loss, often fat-soluble vitamin deficiency).
• Pain. Tumours invade the peritoneal nerve plexus and retroperitoneal structures; hence the deep boring epigastric pain through to the back, worse lying flat, partially relieved by leaning forward.
• Wasting. Maldigestion plus inflammatory cytokines (cachexia) explain rapid weight loss and anorexia even with normal intake.
• Endocrine failure. Islet destruction and insulin resistance produce new diabetes, often with weight loss rather than gain.
• Thrombosis. PDAC releases procoagulant mucins, producing migratory superficial thrombophlebitis (Trousseau syndrome) and a heightened risk of both venous and arterial thromboembolism.

• Gross: a pale, ill-defined scirrhous mass, usually in the head, with a dilated pancreatic duct and bile duct upstream. The gallbladder is often distended (Courvoisier).
• Microscopic: malignant glandular cells in a dense fibrous stroma with perineural and lymphovascular invasion. Lymph-node metastases are common at the time of surgery.

Because the organ is deep and symptoms are non-specific, the diagnosis is usually late. Symptoms are often absent or non-specific until the tumour is advanced, and dyspepsia, depression and weight loss are easily attributed to other causes.

• Painless obstructive jaundice. Dark urine, pale stools, yellow sclerae and pruritus; pain is absent or mild in head tumours. ^[1]
• Epigastric pain that radiates to the back, often relieved by sitting forwards or leaning. ^[3]
• Weight loss and anorexia (often profound and rapid), nausea, vomiting.
• Steatorrhea and pale stools with loss of fat-soluble vitamins; the gut fails completely without biliary and pancreatic secretions.
• New diabetes or worsening glycaemic control in an older adult.
• Depression, sometimes the presenting feature months before diagnosis.
• Migratory thrombophlebitis (Trousseau syndrome): crops of painful, superficial red nodules that appear in different sites over days; a useful clinical clue.
• Late signs: ascites from peritoneal spread, hepatomegaly and metastatic nodules, gastrointestinal bleeding.

• Jaundice with scleral icterus; scratch marks and xanthomata from chronic cholestasis.
• Courvoisier sign: a smooth, non-tender, palpable gallbladder in a jaundiced patient (the duct is obstructed below the gallbladder, which dilates). It favours malignancy over gallstone disease, in which the gallbladder is usually scarred and not palpable.
• Virchow node (left supraclavicular), Sister Mary Joseph nodule (periumbilical), Blumer shelf (rectal metastasis) in advanced disease.
• Cachexia, hepatomegaly, ascites, epigastric mass.

Ductal adenocarcinoma is usually one of a small group of causes of pancreatic and peripancreatic mass. The differential turns on imaging and on where the lesion sits.

The aim is to diagnose, to stage, and to decide whether the patient is fit for resection. Most of this can be done without surgery.

• Transabdominal ultrasound is a common first imaging study in jaundice, but NICE advises that people with obstructive jaundice and suspected pancreatic cancer should be offered a pancreatic-protocol CT before the bile duct is drained. ^[1]
• Multiphase pancreatic-protocol CT of the chest, abdomen and pelvis is the best initial test for a suspected pancreatic mass: it defines size, vascular contact and metastases, and serves as the staging scan. ^[1] It should be done before biliary drainage whenever possible, because preoperative decompression is not needed if the patient proceeds to early surgery. ^[1]

• EUS with EUS-guided biopsy is the most sensitive test for small lesions and the preferred route for tissue diagnosis when CT is inconclusive. ^[2] It also images vascular contact well.
• MRI and MRCP add detail on liver metastases, pancreatic cysts and duct anatomy; MRCP is especially useful for cystic precursors and for anyone needing a non-iodinated study. ^[1][2]
• FDG-PET/CT is reserved for patients with localised disease on CT who are being considered for cancer treatment, and for clarifying indeterminate liver or peritoneal disease. ^[1]
• ERCP is mainly therapeutic: biliary stenting to relieve obstruction. Brushings can be taken for cytology when there is no tissue diagnosis. ^[1]
• CA 19-9 is not a diagnostic test: sensitivity does not exceed 75 percent and specificity 85 percent, and levels vary with cholestasis and inflammation. It is used for monitoring response, providing a baseline, and, after treatment, to track recurrence. ^[2] About one in ten people cannot synthesise CA 19-9 at all. ^[3] A level above 500 IU/mL before treatment makes an anatomically resectable tumour biologically borderline; levels that high most often reflect a neoplastic process rather than a benign cyst. ^[2][3]
• Laparoscopy with laparoscopic ultrasound finds small-volume peritoneal or liver metastases that CT misses: in one four-year series it changed management in 1 in 5 patients. ^[3]
• CT chest is included in the staging CT to look for pulmonary metastases. ^[1]

Newly diagnosed patients who have not yet had a pancreatic-protocol CT should have one covering the chest, abdomen and pelvis, followed by FDG-PET/CT if they will receive surgery, radiotherapy or systemic therapy. ^[1] Liver MRI and EUS are added when the CT leaves doubt about T or N stage. ^[1]

PDAC is staged using the TNM system (tumour extent, nodal spread and metastasis), but management follows the resectability classification, which depends on naming the exact vascular contacts. Whether or not a patient can have surgery is decided by a specialist pancreatic multidisciplinary team on multiphase CT.

Two other considerations widen the definition of "borderline":

• Biological: a pre-treatment CA 19-9 above 500 IU/mL and biopsy-proven regional lymph-node metastases make an anatomically resectable tumour "biologically" borderline. ^[3]
• Conditional: the patient's performance status and comorbidity (ECOG) determine whether they are fit for a major resection, regardless of anatomy. ^[3]

Adequate sampling for staging usually requires examination of at least 12 lymph nodes (the ISGPS recommends at least 15 for the pancreatic head). ^[3]

All patients should be discussed by a specialist pancreatic cancer multidisciplinary team that partners with local cancer units. ^[1] Care is offered and consented to jointly; decisions about whether and when to operate are shared with the patient.

NICE currently recommends neoadjuvant therapy only within clinical trials for both resectable and borderline-resectable disease. ^[1] International practice, reflected in the NCCN and other major guidelines, increasingly favours neoadjuvant treatment (typically mFOLFIRINOX or a gemcitabine-based regimen such as gemcitabine plus nab-paclitaxel) for borderline-resectable disease, because it has been shown to reduce vascular involvement and increase the rate of margin-negative resection without increasing post-operative morbidity; however, no single regimen is clearly preferred, and the benefit over immediate surgery remains a matter of debate. ^[3] Where a tumour is anatomically resectable, upfront surgery followed by adjuvant chemotherapy remains standard in many centres.

Surgery is the only curative treatment, possible in only about 20 percent of patients. ^[6]

• Whipple pancreaticoduodenectomy (pylorus-preserving or standard) is the operation for a small tumour of the head of the pancreas without vascular involvement or distant metastases. ^[1] It removes the pancreas head, duodenum, gallbladder and bile duct, with reconstruction (pancreatic, biliary and gastric enteric) to the jejunum. Vascular resection of the superior mesenteric or portal vein may be added when contact is reconstructable.
• Distal pancreatectomy is used for body and tail tumours; splenic vessels are taken together with the spleen.
• Total pancreatectomy or central pancreatectomy are reserved for special cases because they are not oncologically better and cause brittle diabetes and exocrine failure.

During attempted resection that proves the tumour is unresectable, surgical biliary bypass or prophylactic gastrojejunostomy (if the duodenum is threatened) may be added in the same operation. ^[1] Open surgery remains the standard; minimally invasive resection is controversial. ^[3]

A critical surgical teaching point: the surgeon handles the tumour as little as possible before the aorta is clamped and removes it before any other procedure, and removes the stalk's attachment with a margin of adjacent tissue.

• For patients with resectable tumours, jaundice and good fitness, resect soon rather than pre-operatively drain the bile duct. ^[1]
• If drainage is needed before resection, use endoscopically placed self-expanding metal stents; choose a fully covered metal stent when it may need to be removed later. ^[1]
• For unresectable tumours, offer endoscopically placed self-expanding metal stents rather than surgical bypass. ^[1]
• Consider surgical bypass if an unresectable tumour is found during an attempted resection. ^[1]

• Adjuvant therapy after resection: start once the patient has recovered, with the aim of completing all 6 cycles. Gemcitabine plus capecitabine is the standard adjuvant combination. ^[1] Consider gemcitabine alone if the patient cannot tolerate combination chemotherapy. ^[1] (NICE noted in 2018 that off-label use applied to this combination.) In some guidelines (ESMO) adjuvant modified FOLFIRINOX is also standard; NICE in 2018 did not recommend it. Modified FOLFIRINOX is widely used off-label and is a reasonable option where available and where the patient's performance status is good. ^[5]
• Metastatic disease. For patients with an ECOG performance status of 0 to 1 and a bilirubin below 1.5 times the upper limit of normal and no major comorbidities, ESMO recommends FOLFIRINOX or NALIRIFOX, with gemcitabine plus nab-paclitaxel also an option; gemcitabine plus nab-paclitaxel or gemcitabine alone is used for less fit patients. NICE offers FOLFIRINOX for metastatic disease with ECOG 0 to 1, and gemcitabine plus nab-paclitaxel (through its technology appraisal TA476) where other combination chemotherapy is unsuitable. ^[1][5]
• Olaparib maintenance is the targeted option for patients with a germline BRCA1/2 mutation whose metastatic disease has not progressed on first-line platinum-based chemotherapy. In the POLO trial, maintenance olaparib lengthened progression-free survival compared with placebo (7.4 versus 3.8 months), with similar quality of life. ^[4]
• Locally advanced disease. Offer systemic combination chemotherapy where the patient can tolerate it; consider gemcitabine otherwise; capecitabine can be used as radiosensitiser if chemoradiotherapy is being given. ^[1]
• Second-line. Consider an oxaliplatin-based regimen if oxaliplatin was not used first line; consider a gemcitabine-based regimen after FOLFIRINOX progression. ^[1] Liposomal irinotecan with 5-FU and leucovorin is not recommended by NICE (TA440). ^[1]
• NTRK fusion-positive tumours may be treated with larotrectinib under NICE technology appraisal TA630. ^[1]

• Pain: strong opioids by WHO ladder; EUS-guided or image-guided percutaneous coeliac plexus neurolysis for uncontrolled pain, unacceptable opioid side-effects, or escalating dose needs. ^[1]
• Nutrition: enteric-coated pancreatin for unresectable disease and around resection; early enteral (including oral and tube) feeding rather than parenteral after pancreaticoduodenectomy when the gut works. ^[1] Fish oils are not recommended for weight loss. ^[1]
• Duodenal obstruction: gastrojejunostomy is preferred over duodenal stent in a more favourable prognosis; duodenal stenting is otherwise appropriate. ^[1]

Specialist follow-up after resection looks for problems resulting from surgery, and new or unresolved symptoms should prompt specialist investigation. ^[1] CA 19-9, CT and clinical assessment are used at intervals, with careful attention to the fact that CA 19-9 does not prove cure.

• Biliary and duodenal obstruction, jaundice and cholangitis.
• Pancreatic exocrine and endocrine failure with steatorrhea and diabetes.
• Venous thromboembolism, including pulmonary embolism, is common; use prophylaxis where indicated. ^[1]
• Tumour haemorrhage, gastric outlet obstruction.
• Post-pancreatectomy complications: pancreatic fistula, delayed gastric emptying, haemorrhage, diabetes, anastomotic leak and sepsis.
• Metastatic spread to the liver, peritoneum and lungs.

Even with modern therapy, pancreatic cancer is lethal. Overall 5-year survival is approximately 5 to 15 percent. ^[6] In BRCA-associated and platinum-sensitive disease the outlook is better. ^[4] Population-based modelling from a Swiss cancer registry found that rising incidence over 40 years was partly offset by longer survival, but the disease remained almost uniformly fatal once metastasised; our own discussion of numbers therefore keeps to those published above.

Tumour stage and performance status are the dominant factors; body/tail presentation and N-stage worsen outcomes, while resection with negative margins and adjuvant therapy are the only steps that meaningfully improve survival. ^[4]

Population screening is not recommended for pancreatic cancer: incidence is too low, survival even in detected early disease is poor, tests are imperfect, and costs are high. ^[2] Screening is instead offered to high-risk individuals, by criteria that combine inherited susceptibility and family history.

Offer surveillance to people who have:

• hereditary pancreatitis with a PRSS1 mutation;
• BRCA1, BRCA2, PALB2 or CDKN2A (p16) mutations and one or more first-degree relatives with pancreatic cancer;
• Peutz-Jeghers syndrome.

Consider surveillance for:

• two or more first-degree relatives with pancreatic cancer across two or more generations;
• Lynch syndrome (MLH1, MSH2, MSH6 or PMS2 mutations) with any first-degree relative with pancreatic cancer.

Those who had pancreatic cancer themselves should be asked about affected first-degree relatives. ^[1]

• People without hereditary pancreatitis: MRI/MRCP or EUS.
• People with hereditary pancreatitis and a PRSS1 mutation: pancreatic-protocol CT (EUS is not offered in hereditary pancreatitis). ^[1]
• Higher-risk groups (BRCA2, Peutz-Jeghers, familial clustering): magnetic resonance, with EUS; many programmes start at age 40 to 50, or 10 years younger than the youngest relative's age at diagnosis. ^[2]
• The CAPS consortium recommends 12-monthly screening when no concerning lesion is present, shortening to 3 to 6 months for small lesions, and CA 19-9 at each visit. ^[2]

There is no screening test for the general population, so prevention is risk-factor reduction: stop smoking, keep a healthy weight, treat type 2 diabetes, and limit alcohol. Chronic pancreatitis should be managed to reduce inflammation and malabsorption. For high-risk families, refer to clinical genetics before considering surveillance.

• Older adults. Comorbidity and frailty dominate the decision to operate; surgery or combination chemotherapy is offered only when the patient can tolerate it. ^[1]
• Children and young adults. PDAC is unusual under 40; when it occurs, test for a hereditary syndrome and consider genetic counselling for the family. ^[4]
• Hereditary-cancer families. Germline and somatic testing is recommended by NCCN and ASCO; about one fifth of PDAC has a pathogenic germline variant. ^[4] First-degree relatives of mutation carriers should be offered genetic counselling and, if appropriate, surveillance.

• Diagnosis. A transabdominal ultrasound can suggest obstruction and a mass, but it often misses small lesions and cannot stage vascular contact accurately. Where CT is available, a pancreatic-protocol CT is preferred. ^[1] When it is not, the best fallback is clinical assessment plus the cheapest available cross-sectional imaging, with a low threshold for referral to a centre that has CT and EUS.
• Tissue diagnosis. Even in centres without EUS, ERCP brushings or a transcutaneous or surgical biopsy may be the only route to diagnosis. ^[1]
• Biliary relief. Endoscopically placed self-expanding metal stents before chemotherapy can be life-saving; where endoscopy is unavailable, surgical bypass (choledochojejunostomy or cholecystojejunostomy) may be needed. ^[1]
• Surgery. Remove patients who are resectable to a centre with a pancreatic surgeon and an intensive care unit. ^[3]
• Chemotherapy. Where combination regimens are not available or affordable, gemcitabine monotherapy still offers benefit in the metastatic setting. ^[1]
• Symptom control matters. Pancreatin replacement, coeliac plexus block or opioid analgesia, hydration and nutritional support, and access to hospice or palliative care are the practical measures that change quality of life at far lower cost than anticancer therapy. ^[1]

• Pancreatic ductal adenocarcinoma is the commonest (over 85 percent) and most lethal cancer of the pancreas, with 5-year survival of only about 5 to 15 percent. ^[2][6]
• It arises most often in the head, and therefore presents with jaundice or with deep back pain, weight loss and new diabetes. ^[2]
• Risk rises with age over 50, tobacco, obesity, chronic pancreatitis, longstanding diabetes and inherited susceptibility (BRCA2, PALB2, Peutz-Jeghers, hereditary pancreatitis, FAMMM, Lynch). ^[1][2]
• Cystic precursors (IPMN, mucinous cystic neoplasm) can progress to invasive disease; a main pancreatic duct of 10 mm or more or an enhancing solid component warrants referral for resection. ^[1]
• Multiphase pancreatic-protocol CT is the best initial test; EUS with biopsy, MRI/MRCP and ERCP complete diagnosis and staging; CA 19-9 is only a monitor, with sensitivity 75 percent or less and specificity 85 percent or less. ^[1][2]
• Resectability, not TNM stage alone, drives management: resectable, borderline, locally advanced, metastatic. ^[3]
• Curative surgery is only Whipple pancreaticoduodenectomy for small non-vascular head tumours, and distal pancreatectomy for body/tail tumours. ^[1]
• Adjuvant gemcitabine plus capecitabine is standard; metastatic disease gets FOLFIRINOX for fit patients and gemcitabine with or without nab-paclitaxel for less fit ones; germline BRCA disease may benefit from olaparib maintenance. ^[1][4][5]
• Coeliac plexus block, biliary stenting, enzyme replacement and palliative care are core parts of management at every stage. ^[1]
• Screening is for high-risk families only, not the general population. ^[1][2]