New effective non-surgical treatments are needed for patients whose obesity and type 2 diabetes (T2DM) do not respond to current medical therapies. We propose a randomised controlled trial of Endobarrier, an implantable intestinal device that separates ingested food from contacting the first 60cm of intestine where sited and that mimics some of the clinical effects of bariatric surgery (improved metabolic control with weight loss) with or without continued use of the GLP-1 receptor agonist (GLP-1RA) Liraglutide 1.2mg vs Liraglutide 1.8mg without the device in obese patients with T2DM who remain with suboptimal glycaemic control despite current conventional diabetes treatment, in an NHS setting.

Seventy-two patients with T2DM and obesity (HbA1c≥7.5%, BMI≥35kg/m2) despite previous GLP-1RA therapy will be studied over 24 months and randomised to receive Endobarrier with continued Liraglutide 1.2mg for 12 months; Endobarrier alone for 12 months; or Liraglutide 1.8mg without Endobarrier.

We will investigate potential mechanisms of action and their time course as part of the study by repeated measures of: 1. insulin resistance measures, gut peptides, bile acids; 2. energy intake and nutritional composition; 3. liver fat stores, 4. intestinal inflammation and permeability measures.

The data will inform clinical use of the device and development of new treatments for T2DM and obesity.

Inclusion Criteria:

• type 2 diabetes with latest HbA1c ≥7.5% (≥58mmol/mol)
• obesity with latest BMI ≥35 Kg/m2 (≥30 Kg/m2 for those of South Asian origin)
• liraglutide therapy for at least 6 months - HbA1c and weight trend data should be available
• stable weight and HbA1c in preceding 3 months (<3 Kg reduction in weight and <0.3% reduction in HbA1c)

Exclusion Criteria:

• <18 years of age
• abnormal intestinal anatomy e.g. Crohn's disease
• contraindication to oesophago-gastroduodenoscopy
• previous bariatric surgery or bowel surgery
• active infection
• anticoagulation therapy which cannot be discontinued/ coagulopathy INR >1.3
• eGFR <30
• known portal hypertension
• previous pancreatitis or amylase > 3 times the upper limit of normal
• uncontrolled cardiovascular disease
• lactating or pregnant females
• patients taking aspirin in whom it should continue (e.g. active ischaemic heart disease or cerebrovascular disease)
• excess anaesthetic risk as identified by the anaesthetist or investigator (e.g. uncontrolled obstructive sleep apnoea)