Carcinoma of the pancreas is due to ductal adenocarcinoma in ~90% of cases. The condition is clinically silent until the tumour has reached an advanced stage, when it presents with weight loss, pain, and/or jaundice. Late presentation and early spread largely explain its poor prognosis, with ~5% survival at 5 years. The condition is about twice as common in those with obesity or type 2 diabetes as in those without. Conversely, diabetes is present in up to 50% of those with carcinoma of the pancreas, and the diagnosis of diabetes often precedes the diagnosis of cancer. Tumour-associated humoral factors conferring insulin resistance may be responsible. Progression to pancreatic cancer has been shown to involve a cascade of gene mutations, of which KRAS is the most important. These occur in parallel with the development of progressive premalignant histological changes known as pancreatic intraepithelial neoplasia (PanIN) lesions. Early PanIN lesions are relatively common in the general population, and reflect the prevalence of other risk factors for pancreatic cancer. Since humoral markers are unavailable, pancreatic biopsy is impractical, and imaging techniques are expensive and carry the risk of false positives, possibilities for screening are currently limited.
Pancreatic carcinoma was described in association with diabetes at the end of the 19th century, and it was assumed until recently that diabetes was simply due to destruction of the gland. Early descriptions already made it clear, however, that the tail of the pancreas (well supplied with islets) often escapes the tumour (typically located in the head of the pancreas), and that apparently healthy islets can survive amidst a mass of tumour tissue.
The two conditions are typically diagnosed in close temporal association with one another. Thus, diabetes may be diagnosed up to 3 years before diagnosis of the cancer, at or around the time of cancer diagnosis, or shortly thereafter. Pancreatic cancer is also associated with the introduction of insulin to treat type 2 diabetes. A systematic review and meta-analysis representing data for 1,332,120 people and 41,947 cancers found that pancreatic cancer risk was increased among new users of insulin (RR: 3.18, 95%CI: 3.27-3.71), but not among longer term users. Pancreatic carcinoma can and does develop in established diabetes, but the risk does not appear to differ from that of the non-diabetic population.
The histological and epidemiological evidence therefore suggest that carcinoma of the pancreas causes or worsens diabetes (reverse causation), and that diabetes itself may not predispose to this form of cancer. The working hypothesis is that humoral factors released by the tumour promote the development or worsening of diabetes.
An association between pancreatic cancer and use of the GLP-1 based therapies has been reported from two regulatory databases but remains controversial (see GLP-1 based therapies and cancer).
Pancreatic cancer is twice as common in obesity or type 2 diabetes as in the general population. The association with diabetes is complex, since chronic pancreatitis (often clinically silent) predisposes to both conditions, and because the tumour can be a cause of the diabetes rather than a consequence.
The incidence of pancreatic ductal adenocarcinoma is around 8-12/100,000/yr, with higher rates in some populations, including African-Americans and Maori. Peak age of onset is 65-75 years of age, and men are more commonly affected than women. A small fall in incidence has been described in western countries, but the incidence is increasing in some parts of Asia; cigarette smoking is one possible influence.
Huxley et al performed a meta-analysis looking specifically at type 2 diabetes and pancreatic cancer. The authors included 36 studies (17 case-control and 19 cohort studies) from 1966 to 2005, with information on over 9,000 people with pancreatic cancer. Their results support a strong association between type 2 diabetes and pancreatic cancer (summary OR: 1.82; 95% CI: 1.66-1.89).
Pancreatic cancer is a silent condition that typically presents late. It is the fourth most lethal cancer worldwide, and caused an estimated 217,000 deaths in 2000. The 5 year survival rate is ~6%, so the mortality of the condition closely reflects its incidence.
Ris factors can be inherited or acquired. Around 10% of pancreatic cancer is familial, and the risk of pancreatic cancer is increased 6-fold in those with 2 affected relatives. The BRCA2 gene conveys a 3.5-fold risk of pancreatic cancer, which is also a feature of some rare genetic syndromes, including Peutz-Jegher syndrome. It is more common in some ethnic groups.
75% of cancers develop in the head of the pancreas, and in most cases few symptoms appear until the tumour is well advanced – typically 3 cm in diameter at diagnosis. Presentation is with weight loss, painless jaundice, or with upper abdominal pain radiating to the back. Obstruction of the bile duct produces marked dilatation of the bile duct, and the gall bladder may become so distended as to be palpable (Courvoisier’s sign). Blockage of the pancreatic duct results in loss of exocrine pancreatic function and may cause acute pancreatitis. The pain may become so intractable as to require pancreatectomy for its relief. Other features include thromboembolic disease (Trousseau’s sign) and depression.
Surgery is feasible for only 5-10% of patients, and late diagnosis is the major reason for the dire prognosis of this form of cancer.
Glucose intolerance is present in 80% of cases, and diabetes in ~50%; detection of the diabetes typically precedes diagnosis of the tumour by up to 5 years1. Circulating humoral factors are likely to be responsible, but have yet to be characterised.
Unexplained weight loss in a person with diabetes should arouse clinical suspicion; by clinical tradition the three causes to consider are poor glucose control, thyrotoxicosis and carcinoma of the pancreas. Unfortunately, most people with diabetes and pancreatic cancer are overweight at diagnosis, and recent weight loss is characteristic of new-onset type 2 diabetes.
Pancreatic cancer arises as the consequence of the progressive accumulation of somatic genetic mutations affecting pancreatic parenchymal cells. Similar processes underly all the major cancers. As with a number of other cancers, chronic inflammation may drive this process by accelerating cell turnover and processes affecting DNA injury and repair. The cumulative nature of the process is reflected in the histological changes associated with the early development of pancreatic cancer, which closely reflect the accumulation of gene mutations. These changes have been formally outlined and defined in the Pancreatic Intraepithelial Neoplasia (PanIN) system.
This staging system describes the progression from the normal duct epithelium to squamous metaplasia, in which the cells which normally line the ducts are replaced by mature squamous or transitional epithelium.
PanIN 1-A: these are characterized by flat epithelial lesions composed of tall columnar cells with basally located nuclei and abundant supranuclear mucin. This is considered a transitional form, and its neoplastic nature has not been established.
PanIN 1-B: These epithelial lesions have a papillary, micropapillary or basally pseudostratified architecture, but are otherwise identical to PanIN-1A.
PanIN 2: These mucinous epithelial lesions may be flat or papillary, and cytologically, by definition, these lesions must have some nuclear abnormalities which fall short of those seen in PanIN-3. Mitoses are rare.
PanIN 3: These lesions are usually papillary or micropapillary, but may rarely be flat. Budding off of small clusters of epithelial cells into the lumen and luminal necroses should suggest this diagnosis. These lesions are characterized by a loss of nuclear polarity, dystrophic goblet cells, mitoses which may occasionally be abnormal, nuclear irregularities and prominent nucleoli.
PanIN lesions are common in the ageing population, and are present in 16% of healthy individuals, as against 60% with chronic pancreatitis and 82% of those with adenocarcinoma.
Kras is mutated in ~85% of people with pancreatic cancer. Kras is a moleclar "on-off" switch that becomes set in the "on" position once mutated. This in turn activates NK-κB, a process that is thought to produce a pro-tumour microenvironment by promoting generation of new blood vessels, inflammation, and tissue repair as seen in inherited pancreatitis, a condition associated with the development of cancer.
Characteristic mutation cascades (L) and two specific examples (R)Characteristic mutation cascades (L) and two specific examples (R)The cascade of gene defects predisposing to pancreatic cancer is now well characterised, and can be reproduced by insertion of the Kras gene in knock-out mouse models.
Diagnosis and Treatment
Diagnosis is by abdominal CT followed by ERCP. Radiological staging procedures are then performed. Fewer than 20% of tumours are resectable. The most common operation for a lesion in the head of the pancreas is a Whipple's procedure, which involves removal of the whole pancreas and much of the duodenum. Gastric or biliary bypass are used to relieve jaundice, and coeliac axis block or pancreatectomy are used to treat intractable pain. Gemcitabine is the most widely used chemotherapeutic agent.
As yet no humoral marker is available for screening. Repeated CT scans may be offered to those at very high risk, for example those with hereditary pancreatitis, but its value is limited by a relatively high rate of false positives.
^ Chari ST et al. New onset diabetes: a potential clue to the early diagnosis of pancreatic cancer. Lancet Oncology 2009;10(1):88-95
^ Colmers IN et al. Insulin use and cancer risk in patients with type 2 diabetes: a systematic review and meta-analysis of observational studies. Diabetes Metab 2012;38:485-506
^ Hariharan D et al. Analysis of mortality rates for pancreatic cancer around the world. HPB 2008;58-62 //www.ncbi.nlm.nih.gov/pmc/articles/PMC2504856/pdf/MHPB10-058.pdf
^ Huxley R et al. Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. Br J Cancer 2005;92:2076–2083.
^ Hruban RH et al. Update on familial pancreatic cancer. Adv Surg 2010;44:293-311
^ Wang F et al. The relationship between diabetes and pancreatic cancer. Molecular Cancer 2003;2:4
^ Jones S et al. Core signalling pathways in human pancreatic cancer. Science 2008;321(5897):1801-6