Section 1

Sarah, 57 years old, was diagnosed with type 2 diabetes two years ago, with a BMI of 27.6 kg/m². Hyperglycaemia was reasonably controlled on metformin M/R 1 g twice daily, such that her HbA_1c four months previously was 61 mmol/mol and she had chosen to focus on lifestyle measures to improve glycaemic control.

Sarah had, however, become unwell of late. She saw her GP with symptoms of marked tiredness, thirst, polyuria (especially at night), genital thrush and weight loss of half a stone over the last month or so. Fingerprick testing showed a capillary blood glucose of 17.8 mmol/L, with ketones of 0.2 mmol/L. An HbA_1c came back later at 108 mmol/mol.

What further information might be useful here? What do you think are the implications of Sarah’s sudden deterioration in glycaemic control?

Section 2

You need to check whether there has been any change in Sarah’s diet and lifestyle, and if she is continuing to take her metformin. The dramatic loss of glycaemic control does raise the question as to whether Sarah really does have type 2 diabetes. This, together with the presence of osmotic symptoms, suggests that her endogenous insulin supply may be failing, although there is no evidence (yet) of ketosis.

Ask about Sarah’s family history of diabetes; this may point towards a specific type of diabetes. It would also be worth checking Sarah’s past medical history for any clues.

—

Sarah was having no problems tolerating her metformin and could not identify any radical change in lifestyle. With a loss of just over 3 kg in weight, her BMI had fallen to 26.1 kg/m². There was no family history of diabetes or personal/family history of autoimmune disease. Around six years ago, Sarah had suffered from acute pancreatitis from excessive alcohol consumption in connection with anxiety and depression. Sarah’s current alcohol consumption was no more than 8 units/week, and she was a non-smoker.

With this information, how would you weigh up the various diagnostic possibilities for different types of diabetes?

Section 3

Type 2 diabetes is looking less likely now because of the rapid progression of hyperglycaemia and because Sarah has no family history of type 2 diabetes.

The history would be unusual for type 1 diabetes. The age of presentation is late and diabetes was diagnosed 2 years ago. Sarah has not presented with diabetic ketoacidosis (DKA) and there is no personal or family history of autoimmune disease. LADA (latent autoimmune diabetes in adults), perhaps best considered as a slowly developing form of type 1 diabetes, is certainly a possibility, although again there is no autoimmune history to support this diagnosis. LADA may present at this age and can, at least initially, be managed with oral medication. The lack of family history and later age of presentation really rule out MODY (maturity-onset diabetes of the young) as a possibility.

The history of pancreatitis raises the possibility of pancreatogenic (type 3c) diabetes (that is, diabetes secondary to pancreatic disease).

In addition to failing insulin (and other islet cell hormone) secretion, there is also damage to exocrine function in pancreatogenic diabetes. What symptoms might Sarah have that would indicate impaired digestive enzyme production?

Section 4

It would be useful to ask Sarah about gastrointestinal symptoms that would suggest maldigestion and malabsorption. Has she experienced upper abdominal pain, loose offensive stool (steatorrhoea) or bloating (1)? Weight loss is also a feature of chronic pancreatitis but, in this case, is probably attributable to Sarah’s uncontrolled diabetes. Pancreatic enzyme insufficiency (PEI) commonly pre-dates onset of diabetes in pancreatogenic diabetes (2).

What would you do next for treatment and management?

Section 5

It would be worth trying an additional oral agent, but referral to the diabetes specialist team for further investigation and management is now warranted.

—

Gliclazide 80 mg twice daily was added to Sarah’s regimen, and she was referred urgently to secondary care for consideration of an insulin start.

Gliclazide is a good choice for add-on therapy in this situation because potentially its effects in reducing hyperglycaemia and improving symptoms will be rapid. If the gliclazide is ineffective, then this would strengthen the idea of poor insulin reserve.

—

Sarah’s hyperglycaemia showed only modest improvement with gliclazide. In addition to her weight loss, Sarah admitted to persistent loose, floaty stools.

What treatments do you think Sarah now needs?

Section 6

In the hospital diabetes outpatient clinic, Sarah was commenced on basal–bolus insulin therapy, with discontinuation of the gliclazide.

Islet cell antibodies (GAD, IA2 and ZnT8 autoantibodies) were negative. This is consistent with a lack of autoimmune process, favouring pancreatogenic diabetes (or type 2 diabetes) as a diagnosis, rather than type 1 diabetes or LADA. C-peptide levels were low, indicating deficient endogenous insulin reserve, consistent with pancreatogenic diabetes (or type 1 diabetes or LADA) rather than type 2 diabetes.

A CT scan of the abdomen showed changes diagnostic of previous pancreatitis. Faecal elastase levels were reduced, indicating PEI.

Thus, a firm diagnosis of pancreatogenic diabetes (type 3c diabetes) was reached (3).

To deal with the symptoms of malabsorption due to PEI, Sarah was given oral pancreatic enzyme replacement therapy (PERT) to be taken with meals (4).

Malabsorption of fat-soluble vitamins is a problem with PEI. Which fat-soluble vitamin should Sarah be tested for, to avoid potentially serious consequences?

Section 7

Sarah should have levels of 25-hydroxyvitamin D checked. Vitamin D deficiency would place her at risk of osteoporosis and subsequent fracture (4). Vitamin D supplements are commonly required in individuals with type 3c diabetes.

Can you think of any other causes of pancreatic damage that might lead to pancreatogenic diabetes?

Section 8

Think of conditions or situations where there is global damage to the pancreas. In these situations both endocrine (hormone-secreting) and exocrine (digestive enzyme production) functions of the pancreas can be affected (5).

• Pancreatitis (acute or chronic)
• Pancreatectomy
• Trauma
• Pancreatic carcinoma
• Cystic fibrosis
• Haemochromatosis

 
Chronic pancreatitis accounts for around 75% of all cases of type 3c diabetes (6). Alcohol is a frequent cause of pancreatitis. Other causes include smoking, toxic chemicals, autoimmune disease, genetic causes and pancreatic duct obstruction (2).

The lifetime risk of developing diabetes in people with chronic pancreatitis has been estimated to be as high as 80%. NICE advises offering HbA_1c monitoring to people with chronic pancreatitis at least every 6 months (7).

Try and summarise the distinguishing clinical features of type 3c diabetes from type 1 and type 2 diabetes in terms of age, presenting features, BMI, islet cell autoimmunity and C-peptide (insulin) levels.

Section 9: Recognition and management of type 3c diabetes

The distinguishing clinical features of type 3c diabetes compared to type 1 and type 2 diabetes are summarised below (1,2).

CF=cystic fibrosis; DKA=diabetic ketoacidosis

Sarah’s case illustrates the need to undertake a careful review of the patient’s history in determining an underlying cause of diabetes, and to be open to a change in diagnosis and management. Type 3c diabetes is frequently misclassified (often as type 2 diabetes) and so prevalence is underestimated in practice (8).

It is important to recognise type 3c diabetes, as this will direct appropriate medical treatment. Insulin is ultimately likely to be required, as the underlying abnormality is compromised insulin production (rather than insulin resistance). There is a risk of decompensation to severe hyperglycaemia and DKA. In the longer term, microvascular and macrovascular complications can develop, as with other forms of diabetes.

Section 10: Clinching the diagnosis

There is no definitive diagnostic test for type 3c diabetes, but proposed major diagnostic criteria (in addition to usual requirements for diagnosing diabetes) are listed below (3,5).

• Pancreatic exocrine insufficiency (faecal elastase-1 testing)
• Pathological pancreatic imaging (endoscopic ultrasound/CT or MRI scan)
• Absence of antibodies associated with type 1 diabetes

 
In addition to reduced insulin secretion in type 3c diabetes, there is also reduced glucagon secretion. What might be your concern here?

Section 11

Endocrine dysfunction in type 3c diabetes involves all the pancreatic islet cells. In addition to loss of insulin secretion from beta-cells (causing hyperglycaemia), there is loss of glucagon from alpha-cells (blunting the response to hypoglycaemia). It is clinically very important to be aware of this reduced counter-regulatory response to hypoglycaemia, as this may lead to “brittle” diabetes (9). Glycaemic variability appears to be higher in type 3c diabetes than type 2 diabetes (10).

To counter brittle diabetes, it may be necessary to accept higher glucose and HbA_1c targets to minimise the danger of hypoglycaemia. Capillary blood glucose monitoring is advisable if medicines that can induce hypoglycaemia are being taken (insulin, sulfonylureas, meglitinides). The extension of continuous glucose monitoring (CGM), either intermittently scanned (flash) or real-time, to type 3c diabetes could be discussed with the specialist diabetes team.

Aside from managing hyperglycaemia, what other clinical concerns should you consider in type 3c diabetes?

Section 12

Type 3c diabetes is associated with an increased risk of pancreatic carcinoma (1). Refer immediately if you have concern over the possibility of pancreatic carcinoma (weight loss, epigastric pain radiating to back, jaundice, anorexia, unexpected deterioration in glycaemic control). The NICE guideline on suspected cancer recommends consideration of an urgent CT scan (to be performed within 2 weeks), or an urgent ultrasound scan if CT is not available, in people aged ≥60 years with weight loss and diarrhoea or new-onset diabetes in order to look for pancreatic carcinoma (11).

Recognition of type 3c diabetes is important to alert the clinician to the possibility of dealing with exocrine dysfunction. Referral to a gastroenterologist may be appropriate. If PEI is identified, then pancreatic enzyme replacement therapy (PERT; e.g. Creon, Nutrizym, Pancrease, Pancrex) can be taken with meals, and the response judged by relief of steatorrhoea and weight gain. Be aware that as PERT can improve digestion of carbohydrates and increase glucose levels, it may unmask diabetes in an individual with a previously normal HbA_1c (1,2,4,9). Vitamin D supplements should be prescribed to correct proven deficiency, and investigations for osteoporosis considered (1,9).

Remember to address cardiovascular risk factors (as with other types of diabetes). Smoking cessation, control of hypertension and hyperlipidaemia are of great importance.

Section 13: Management of hyperglycaemia in type 3c diabetes

The evidence base to guide management of type 3c diabetes is weak. There are no specific guidelines (other than for cystic fibrosis [CF]; see below). Treatment goals are derived from randomised controlled trials from type 1 and type 2 diabetes, and expert opinion. NICE refers to its guidelines on type 1 and type 2 diabetes, according to whether insulin is used or not, in order to guide management of type 3c diabetes, and emphasises the need for further research to inform management (7).

Lifestyle measures appropriate for type 3c diabetes include avoidance of alcohol, a high-soluble-fibre diet and regular exercise (12).

For mild hyperglycaemia, oral hypoglycaemic agents may be tried first. Here metformin remains the first-line treatment, with the benefits of a low risk of hypoglycaemia and a possible protective effect against pancreatic carcinoma (13).

The insulin secretagogues (sulfonylureas, glinides) are likely to become less effective as beta-cell function declines in type 3c diabetes, and carry the risk of hypoglycaemia. Pioglitazone is an option if there is intolerance or contraindication to metformin, but should be avoided in heart failure or if there are concerns over risk of fracture and possible bladder cancer. DPP-4 inhibitors and GLP-1 receptor agonists are generally avoided in type 3c diabetes because of concerns over pancreatitis and pancreatic carcinoma. GLP-1 RA properties of reducing appetite and weight loss are not desirable in type 3c diabetes. There is little experience using SGLT2 inhibitors in type 3c diabetes and, whilst they are potentially useful, they do carry an increased risk of DKA (1,2,4,9).

Insulin is required in most cases of type 3c diabetes to deal with insulin deficiency. It may be necessary from the outset if HbA_1c is markedly elevated. Insulin is the treatment of choice in CF-related type 3c diabetes (14).

Whilst the optimal insulin strategy for type 3c diabetes remains to be determined (7), it has been suggested this would be a basal–bolus regimen (or insulin pump) with carbohydrate counting, as employed in type 1 diabetes (1,2,9).

Individuals with type 3c diabetes can be signposted to useful information from Diabetes UK and TREND Diabetes (15,16).