As much as 80% of people with pancreatic cancer have diabetes or glucose intolerance at diagnosis (DeMeo, 2001). Yet, whether diabetes is causally associated with pancreatic cancer or simply a consequence of it remains unclear.
The likely cause of diabetes in the context of exocrine pancreatic cancer is believed to be multifactorial, including reduced release of insulin in response to stimuli (more from islet dysfunction than islet destruction by the tumour), and peripheral resistance, which improves following resection of the tumour (Wang et al, 2003). Potentially, there may also be diabetogenic peptides found in people with pancreatic cancer (Wang et al, 2003).
Diabetes and pancreatic cancer
In a retrospective cohort study, Gupta et al (2006) observed that 149 of 36631 individuals newly diagnosed with diabetes received a diagnosis of pancreatic cancer over the subsequent 6-year period. The risk was 2.2-fold higher in those newly diagnosed with diabetes than those without diabetes (95% confidence interval 1.84–2.56) and was highest during the first 2 years after diabetes diagnosis. The authors estimated one additional pancreatic cancer diagnosis for every 332 new diagnoses of diabetes over 6 years.
Questions have been raised regarding the possibility of distinguishing this sub-group at increased risk of pancreatic cancer from other people with type 2 diabetes (Chari et al, 2008). If this were possible, people newly diagnosed with diabetes could be screened to detect pancreatic cancer at a stage where it may be operable (Chari et al, 2008). However, there are insufficient and conflicting data supporting improved outcomes following such early detection of pancreatic cancer at the time of diabetes diagnosis (Damiano et al, 2004; Chari et al, 2005; Pelaez-Luna et al, 2007).
Routine imaging (a mixture of ultrasound, computed tomography and magnetic resonance imaging) of all people with new onset acute diabetes (n=115) over the age of 50 years, identified pancreatic cancer in 5.2% of patients in a study by Damiano et al (2004). Whether targeted screening based on the presence of other risk factors for pancreatic cancer – including presence of gastrointestinal symptoms, weight loss, or early need for insulin – would have increased the rate of cancer identified, has not been tested.
Guidance published in Gut (Pancreatic section of the British Society of Gastroenterology et al, 2005) has grade B recommendations that the “diagnosis of pancreatic cancer should be considered in patients with adult onset diabetes who have no predisposing features or family history of diabetes”. They go on to recommend that unexplained diabetes in people over 50 years of age, with no family history, obesity or steroid usage should be referred to a local cancer unit. However, this recommendation is not currently included in NICE (2005) referral guidance for suspected cancer. Where there is suspicion, clinicians, in addition to examining the abdomen, should have a low threshold to seek specialist help and to consider an ultrasound examination.
There is no clear guidance on how best to manage diabetes in the context of pancreatic cancer. The primary concern should be resolution of hyperglycaemic symptoms, safe blood glucose levels and the avoidance of hypoglycaemia. Long-term complications are unlikely to be a major concern in this group. Fluctuations in glucose levels with the potential need for chemotherapy, decreasing appetite and change in weight and dietary input may warrant follow-up.
With recent evidence suggesting a mixed pattern of both islet dysfunction and peripheral resistance (Wang et al, 2003), there may be benefit in using oral agents before initiating insulin. However, with the nature of this illness, many people are likely to require insulin, and progression may be faster than in type 2 diabetes (Noy and Bilezikian, 1994).
The case study in Box 1 highlights the fact that, although diabetes is very common, occasionally it may be secondary to another cause, such as pancreatic malignancy.
Despite the fact that there is no evidence to support early diagnosis of cancer improving outcome, with most cases of diabetes now being diagnosed and managed in primary care, practitioners should be aware that there may be secondary causes for diabetes. It also highlights the need for further research in how best to manage this subgroup, and the potential differences compared with other people with type 2 diabetes.