ABCD Exenatide Audit
This, our first and highly successful large scale nationwide audit, is now completed, though it did teach us a tremendous amount as can viewed here.
About the ABCD nationwide exenatide audit
It is not easy to appreciate, looking back, the enthusiasm which greeted the arrival, in 2007, of exenatide as an agent which could be prescribed in the UK. For the first time we had an agent which both improved HbA1c and reduced weight. This was highlighted in the head to head study of exenatide against insulin glargine in which both agents reduced HbA1c by the same amount, though glargine caused weight increase while exenatide led to weight loss. By 2008 some clinical experience with exenatide had been gained by most diabetologists since its launch in the UK in 2007. ABCD realised that it was in a unique position to accelerate understanding of this new agent by pooling all the individual experiences of its many members through a national audit. ABCD had already undertaken several, relatively small, nationwide audits, and now realised that the experience gained could be scaled up, especially given the widespread enthusiasm amongst its members for the new agent. ABCD considered (2008) what might we learn from such a national audit and set up a questionnaire on a password protected subweb of its Sharepoint website for collection of anonymised patient data.
The nationwide audit was launched on 12th December 2008 and a call was sent out for data to be submitted by a deadline of 18th February 2009, with a regular countdown email sent out to all ABCD members. (to see the countdown emails in succession, once you have opened the link, just keep clicking on the slides – apply the same principle to the other links to slides in this paragraph). It is testament to the enthusiasm of clinicians for exenatide that over those two months data was submitted on 5313 patients with usable data on 3913. The first data analysis was presented by Chris Walton at a DUK satellite symposium in Glasgow on 10th March 2009 with further analysis presented by Bob Ryder at the Spring ABCD meeting in Bristol on 8th May 2009. The data continued to accumulate until the last deadline for data submission on 20th July 2009. By that time, only 7 months after launch, the audit had collected anonymised data on 6717 patients (2007-2009) submitted by 315 contributors from 126 centres. The data was presented at the Spring ABCD meeting in Newcastle on 7th May 2010. As well as nine published papers, there have also been 16 published abstracts, 12 oral presentations and 13 poster presentations.
What we learned from the exenatide and liraglutide audits
After the audit had been active for 5 years, at the ABCD autumn meeting in 2014, Bob Ryder summarised the main findings from the ABCD nationwide exenatide and liraglutide audits and the slides from that presentation are available for viewing. In summary, the findings over the 10 years since the audit was launched have been:
• The patients treated with exenatide or liraglutide in real clinical practice are much heavier and with much poorer glycaemic control than in clinical trials of these agents; nevertheless, the agents have proven to be very effective.1
• Exenatide and liraglutide were used outside NICE guidelines in substantial numbers of patients and were found to be effective in outside NICE guidelines. In particular, the agents were used with insulin (40% in the nationwide liraglutide audit) with good effect in many patient.2
• Off licence use of exenatide with insulin was found to be safe and effective in real clinical practice, a reduction in insulin dose frequently occurred, weight fell and 1 in 6 patients were able to come off insulin.3,4
• An important safety issue was uncovered in that some clinicians attempted to stop insulin when starting exenatide in order to stay within license. This led to harm to the patient in some instances - for example there are 11 reported cases of ketosis or diabetic ketoacidosis - 7 of these occurred to patients who stopped insulin at the time of exenatide initiation. Analysis of audit data allowed us to recommend that when starting exenatide in an insulin-treated patient, not to stop the insulin but rather to tail the insulin off during treatment if response to treatment allowed.5,6
• When concerns were raised in 2013 in a TV documentary and in the British Medical Journal that incretin therapies might cause pancreatic damage, we were able to publish data suggesting that in the ABCD audits there is no evidence of such a side effect and that most cases of acute pancreatitis in the ABCD exenatide and liraglutide audits had other causes for acute pancreatitis, in particular gall bladder disease.7,8,9,10,11
• Many patients with a professional driver’s licence who would lose their jobs if they went onto insulin, were able to avoid insulin, and maintain similar glycaemic outcomes and keep their jobs by using exenatide or liraglutide.12,13
• Liraglutide was found to be safe and effective among patients with moderate renal impairment, even though that was an exclusion for use at the time.14,15
• Liraglutide use was found to be associated with a reduction in alanine aminotransferase levels (an index of fat in the liver) when they are elevated.16
• Though initially liraglutide was not licensed for use with insulin and in due course it was licensed only for use with insulin determir, it was found in the audit to be effective with all the common insulin regimes, whether basal, basal-bolus or biphasic.17
• Exenatide and liraglutide may be less effective at improving glycaemic control amongst non-insulin treated South Asians.18
• Long duration of diabetes and insulin use both predict reduced response to liraglutide, with insulin use being the strongest predictor.19
• Improvements in HbA1c and weight were seen when switching from exenatide and DPP4 inhibitors to liraglutide.20
• Older age and non-metformin use were associated with more significant gastro-intestinal side effects leading to discontinuation of liraglutide treatment.21,22
• In occasional exenatide treated patients, the nausea, vomiting or diarrhoea was so severe that they developed transient acute kidney injury. No other new safety issues were uncovered.23,24
• Starting liraglutide was found to reduce 10-year cardiovascular risk as assessed by the UKPDS risk-engine.25, 26
Link to the Nationwide Liraglutide Audit - Click here
Further information
Further enquiries may be made to the ABCD nationwide audits database administrator of the project.
Papers, abstracts, presentations, posters and webcasts emanating from the audit
Click here to see the output from the audit as it evolves. The names of all the contributors will be found in the presentation slides and in the published papers.
