Routine screening for cystic fibrosis-related diabetes

Introduction

Cystic fibrosis-related diabetes (CFRD) is an important and well-established complication of cystic fibrosis (CF) affecting up to 35% of patients by the third decade of life.^1,2 Prevalence increases with age and is associated with pancreatic insufficiency, female gender, liver disease and the DF508 mutation.^3–7

DECLARATIONS

Competing interests

None declared

Funding

None

Ethical approval

Not applicable

Guarantor

DP

Contributorship

DP is the sole contributor

CFRD has a negative impact on lung function, weight and survival and is associated with late complications such as retinopathy, nephropathy, microalbuminuria and neuropathy.^6,8–15 Decline in respiratory function and nutritional health has been shown to develop several years prior to presentation, although most of this data has been collated from unscreened populations.^{3,11,13,16–18} The pre-diabetic state may contribute directly to clinical decline as a consequence of abnormal glucose metabolism or alternatively, the decline in nutrition and lung function may result from increased inflammation and nutritional requirements in patients with poorer health. ¹⁴

A delay in the diagnosis of CFRD is commonly due to an absence of symptoms. Early identification through screening is therefore essential as prompt intervention with insulin can impact on weight, lung function and overall prognosis.^17–20 It has also been suggested that insulin therapy may be beneficial in patients with CF and impaired glucose tolerance.^21,22

In this short review, three key areas will be addressed. These include defining CFRD in the context of type 1 and 2 diabetes, determining the most effective screening tools and addressing the impact of screening on outcome.

What is CFRD?

The high prevalence of impaired glucose tolerance and variable insulin resistance in patients with CF makes the diagnosis of CFRD difficult. CFRD is a distinct form of diabetes which shares features of both type 1 and type 2 diabetes. It is associated with a combination of insulinopenia and insulin resistance.^{12,14,23–26} To complicate matters, there is a high prevalence of abnormal glucose metabolism in patients without CFRD, with transient elevation of blood glucose occurring in associated with infection, feeding and steroid therapy. Impaired insulin secretion has also been described in non-diabetic patients with CF.^14,27

There are various situations that can lead to an accentuation of abnormal glucose metabolism and the development of diabetes in CF. These include gestational diabetes, steroid-induced diabetes, feeding-related diabetes, infection and abnormal glucose metabolism resulting in postprandial hypoglycaemia. Proactive detection of CFRD with routine screening should therefore be implemented as an essential part of clinical practice.

Screening for diabetes

The glucose tolerance test

Diabetes is defined by a level of chronic hypergly-caemia which if not treated would result in complications such as microvascular disease. Current World Health Organization (WHO) diagnostic criteria are based on the oral glucose tolerance test (OGTT) where a positive result equates to a fasting plasma glucose ≥ 7.0 mmol/L or the two-hour plasma glucose ≥ 11.1 mmol/L. In contrast an impaired glucose tolerance test (IGTT) is defined as two-hour plasma glucose level >7.8 mmol/L and < 11.1 mmol/L. Theoretically it should therefore be easy to screen for CFRD by undertaking an OGTT.

Unfortunately there are several reasons why direct extrapolation of the WHO diagnostic criteria is inappropriate in the context of CF. Firstly a diabetic OGTT (DGTT) as defined by a two-hour plasma glucose ≥ 11.1mmol/L does not equate to a diagnosis of CFRD. Some patients with CF will have normal blood glucose profiles despite a DGTT and both an IGTT and DGTT can return to normal on further testing. ²⁸ Raised fasting plasma glucose levels are often absent in patients with CFRD. In a recent multicentre audit of 392 patients undergoing OGTT, 82% of patients with CFRD were identified exclusively on the basis of the two-hour plasma glucose level, with only 12% having solely an elevated fasting glucose. ¹ All patients with a DGTT underwent routine blood glucose monitoring with 20/34 (59%) having diabetic profiles.

The prevalence of IGTT is variable and up to 56% of tests can revert to normal by the next annual GTT. ²⁸ Of 1419 OGTT carried out in the Leeds Adult CF unit, 60.2% were normal, 23.3% were impaired 11.3% were diabetic and 4.2% showed reactive hypoglycaemia at two hours. However the significance of an IGTT should not be underestimated, as it is associated with a higher risk for developing CFRD and may result in worsening lung function. ¹⁴

HbA1c

Glycosylated haemoglobin (HbA1c) has been used as a screening test for CFRD but it remains too unreliable.^28–30 HbA1c levels do not appear to differentiate between asymptomatic patients with impaired and normal OGTT. Those studies reporting a high prevalence of raised HbA1c levels in patients with newly-diagnosed CFRD were carried out in unscreened populations where the diagnosis was delayed. This is highlighted by Yung et al. who found elevated HbA1c levels in nine out of 12 patients with newly-diagnosed CFRD. ³¹ The diagnosis of CFRD was already significantly delayed as the majority of patients were symptomatic at the time of diagnosis. Three patients had glycosuria, six had symptoms of hyperglycaemia and one had recent weight loss. This contrasts with studies carried out in populations undergoing annual OGTT, where HbA1c was found to be elevated in only a minority of cases.^29,32,33 Solomon et al. prospectively followed up 94 patients with CF and found no difference in the mean HbA1c level between patients with and without CFRD. ³⁴ Other studies have confirmed that both HbA1c and random blood glucose is not a substitute for early diagnosis of CFRD with OGTT and not an appropriate screening test.^10,33

Random glucose and serial blood glucose monitoring

High random glucose can occur despite normal OGTT and raised levels are often absent in CFRD.^28,32,35 Random glucose measurements are therefore not appropriate for routine screening purposes.^22,32,36 In contrast, serial blood glucose monitoring is helpful in identifying individuals with CFRD and should be carried out in patients with DGTT. Blood glucose monitoring is also essential for individuals starting oral steroids, feeds and in those with concerning symptoms.

Continuous glucose monitoring

More recently there has been significant interest in continuous glucose monitoring system (CGMS) which can provide a detailed glucose profile for any individual. The test has proved useful as a research tool, demonstrating pathological glucose excursion in patients with both IGTT and NGTT. ³⁷ CGMS does not appear to be advantageous in detection of CFRD as compared with OGTT^37,38 but may prove an effective screening tool for the future. More work is needed to establish the repeatability of CGMS, accuracy of blood glucose measurement, simplification of the equipment and correlation with clinical outcome.

Why screen?

Early studies in the 1990s demonstrated a fall in lung function and nutritional status prior to diagnosis of CFRD.^17,18 This decline was insidious and occurred up to 4.5 years prior to diagnosis. ¹⁶ Similar findings have been reported in more recent publications which also confirm clinical improvement with early intervention. ¹⁷ While Lanng et al. demonstrated improvement in body mass index (BMI) and a reduction in the number of infections following insulin therapy, lung function did not fully recover suggesting a long-term impact on health. ²⁰ Rolon et al. reported a significant fall in lung function and nutritional status six months before diagnosis of CFRD with both parameters improving significantly with insulin therapy. ¹⁸

In contrast, a recent case control study from the Leeds unit found no significant difference in weight, lung function, body mass index (BMI) or intravenous treatment between patients with and without CFRD in the six years prior to diagnosis. ³⁹ One of the crucial features of this study was the fact that the population had undergone routine screening with OGTT as well as blood glucose monitoring for over a decade. These observations suggest that a decline in lung function and nutrition is not inevitable in the pre-diabetic state if screening and appropriate care is instituted early.

Conclusion

Despite overwhelming evidence for screening and the presence of national guidelines, clinical practice varies significantly across the UK. ⁴⁰ This variation may be a reflection of the complexity and variability of glucose metabolism in CF and the absence of a simple and specific screening tool. Currently the OGTT is the screening test of choice. However since in CF a DGTT does not always indicate CFRD in asymptomatic patients, glucose monitoring is required to confirm the diagnosis. In Leeds, we recommend pre and postprandial blood glucose monitoring in patients with a DGTT result. If blood glucose levels are normal or there is an IGTT, then we recommend a repeat OGTT after six months. Serial blood glucose measurements and OGTT are also undertaken where clinically appropriate.

It has been suggested that an annual OGTT is too time and resource consuming for it to be deployed effectively. While it should be acknowledged that not all patients will actually undergo an annual test, it is not acceptable to use this as an excuse for not doing the test especially as it has now been deployed successfully across CF units worldwide.

Appropriate resources need to be invested into screening patients with CF for CFRD so that the adverse impact of late diagnosis on health status can be negated. Waiting for patients to present with diabetic symptoms in the absence of a screening tool is no longer acceptable practice.