Addition of exenatide to obese, insulin-treated patients can improve glycaemia and weight. Adverse events were statistically but probably not clinically significantly higher, but combination treatment was less well tolerated. Overall, exenatide was less effective in lowering HbA1c among insulin-treated patients, although significant number of insulin-treated patients still achieved significant HbA1c, weight and insulin reductions. Further research into identifying obese, insulin-treated patients who will tolerate and benefit from exenatide treatment is urgently needed.

In December 2008, to accelerate understanding of a new agent, the Association of British Clinical Diabetologists (ABCD) launched a nationwide audit on the use of exenatide in clinical practice. A password‐protected online questionnaire for collection of anonymised patient data was established and diabetes specialists in the UK were given persistent encouragement to submit data on their exenatide‐treated patients. Baseline and latest HbA1c, weight, body mass index (BMI), waist circumference, blood pressure and lipids were compared and adverse events related to exenatide were quantified. A total of 315 contributors from 126 centres submitted data on 6717 patients (54.9% male) – mean baseline age was 54.9 years, HbA1c 9.47% (80mmol/mol), weight 113.8kg, BMI 39.8kg/m2. Of these, 4551 and 4385 had dated baseline and latest HbA1c and weight respectively. Mean (±SE) HbA1c fell by 0.73±0.03% (p<0.001) and weight by 5.9±0.1kg (p<0.001) at a median (range) of 26.1(6.6–164.1) and 26.0(6.6–159.0) weeks respectively. The following parameters also showed significant falls (p<0.001): BMI 2.2±0.1kg/m2, waist circumference 5.1±0.3cm, systolic blood pressure 3.6±0.6mmHg, total cholesterol 0.16±0.03mmol/L and HDL cholesterol 0.03±0.01mmol/L. Triglycerides decreased by 0.14±0.06mmol/L (p=0.009). The change in diastolic blood pressure was not statistically significant. In all, 23.7% of patients reported gastrointestinal side effects with 7.2% having to stop exenatide permanently. Hypoglycaemia rates were 3.3% before and 5.6% after exenatide use (p<0.001). After scrutiny, one case of pancreatitis and four cases of renal failure occurring in patients on exenatide had no obvious alternate cause. All other reported side effects had <1% incidence. The rate of exenatide discontinuation was 19.9% throughout the span of the audit, most commonly due to gastrointestinal side effects (36.1%) and lack of glycaemic or weight benefit (33.8%). This large scale audit confirmed the effectiveness of exenatide in clinical use and highlighted rare associated adverse events. Importantly, we have successfully demonstrated a novel approach by a national specialist society to independently monitor the efficacy and safety of a new treatment. Copyright © 2010 John Wiley & Sons.

Introduction: There is concern that glucagon-like peptide-1 (GLP1) receptor agonists may be associated with acute pancreatitis. The data from the ABCD nationwide liraglutide audit (November 2009-June 2013; 6010 patients) provide an opportunity to assess the extent of the problem in routine clinical practice in the UK. Methods: At every patient visit, audit-contributors were invited to submit, via an electronic form, clinical data collected as part of routine clinical practice, including data on possible side effects of treatment. Cases of 'possible pancreatitis' were identified and we contacted the centres concerned to obtain full details. Results: To date, the audit has monitored 3720 years of exposure to liraglutide. There were four cases of possible pancreatitis documented from the 6010 patients on liraglutide: three patients had likely causes of pancreatitis identified and one patient had no aetiological cause. This sole case represents an incidence of 0.027/100 patient-years of exposure to liraglutide. Conclusion: In cases of acute pancreatitis of a patient on liraglutide, if another cause can be found (usually gall stones associated with obesity), the drug is not be necessarily culpable. People with Type 2 diabetes are at greater risk of acute pancreatitis (hazard ratio between 1.5 and 2.8). Thus, the possibility of liraglutide-associated pancreatitis in 'real-world' clinical practice (0.027/100 patient years) represents a very small risk.