This study is a randomized trial that evaluates the effect of metformin addition or not to standard care on the duration of diabetes remission after bariatric surgery.

In addition to significant weight loss, several randomized control trials (RCTs) have demonstrated that bariatric surgery can reverse or at least improve type 2 diabetes (T2D). Despite the variability in study design and patient characteristics of these RCTs, there is a consistent favorable effect of surgery compared to medical treatment for weight loss, change in HbA1c, reduction in diabetes medications, remission of metabolic syndrome and improvement in quality of life. Diabetes remission rate is estimated from 15 to 45 % according to the 4 available RCT including the most used surgery (Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG)) with at least three to five years of follow-up. These results mean that more than half of patients with type 2 diabetes are still or newly diagnosed with diabetic after surgery and that extending time of diabetes remission after bariatric surgery is of major concern.

No RCT has explored yet an intervention to extend diabetes remission. Apart from bariatric surgery, metformin is unequivocally recommended to treat both diabetes and pre-diabetes along with lifestyle interventions. Results of the Diabetes Prevention Program trial showed that metformin reduces diabetes incidence by 31% in obese patients with pre-diabetes. We hypothesized that metformin might extend the duration of diabetes remission after bariatric surgery.

The study is a randomized, controlled, open-labeled, multicenter trial.

Patients fulfilling the inclusion criteria and without any of the exclusion criteria will be randomized.

Patients will receive:

• Standardized care plus metformin treatment if randomized in the experimental group given for 3 years
• Standardized care alone if randomized in the reference group

Primary objective is to demonstrate that metformin increases the proportion of patients with T2D remission compared to standard care among ex-T2D patients operated of BS, after a 3-year period of treatment.

Secondary objectives are:

• To assess the proportion of patients with T2D partial or complete remission with metformin compared to standard care in ex-T2D patients operated of BS, after 1 and 2 years of treatment.
• To assess body weight and metabolic parameters in metformin group versus standard care.
• To assess tolerance, nutritional status and adherence to metformin in intervention group versus standard care.
• To assess micro and macroangiopathy at 3 years.
• To assess quality of life changes from baseline at 1, 2 and 3 years.
• To assess the accuracy of long term prediction score (i.e. prolonged remission assessed at the end of the study with the Ad-DiaRem score)
• To explore gut contribution to metformin metabolic effect by: (i) gut microbiota differences (diversity, composition and function) between metformin treated and non-treated individuals and (ii) measurements of metformin-induced enterohormones secretion

Patients are followed up every 6 months during 3 years in both arms. If diabetes is diagnosed during the follow-up (HbA1c > 6.5 %), the primary endpoint of the study is obtained meaning end of diabetes remission but patients will be still followed up to the end of protocol to monitor the secondary endpoints. When remission is over, the care defined by the protocol (ie metformin + standardized care or standardized care alone) should be stopped. In both groups, when remission is over, management of the disease has to be adapted according to physician's and patient's preference whatever the arm of randomization.

Inclusion Criteria:

• Adults 18-70 years old
• Having undergone gastric bypass or sleeve gastrectomy 12 to 36 +/-3 months before inclusion
• "ex-T2D" treated with at least one anti-diabetic drug before bariatric surgery or HbA1c ≥ 6.5 % before bariatric surgery
• HbA1C < 6.5 % at inclusion with no anti-hyperglycemic medications for the last three months
• Written consent

Exclusion Criteria:

• Known type 1 diabetes

• Pregnancy and breastfeeding

• Estimated glomerular filtration rate<44 ml/min (MDRD)

• Known intolerance to metformin

• Known contraindication to metformin:

  • Acute metabolic acidosis
  • Acute affection which could lead to renal deterioration (ex: dehydration, serious infection, shock, intravascular administration of iodinated contrast agent within the last 48 hours)
  • Acute or chronic disease which could lead to a tissue hypoxia (ex : severe cardiac insufficiency, severe respiratory insufficiency, myocardial infarction within the last 3 months, shock)
  • Hepatocellular insufficiency
  • Prothrombin ratio ≤ 50%
  • SGOT or SGPT levels ≥ 10 times the upper limits of the normal range
  • Alcohol use disorder

• Medications and medical conditions likely to confound the assessment of diabetes:

  • glucocorticoids treatment
  • renal graft
  • Cushing's syndrome
  • acromegaly
  • fasting plasma triglyceride > 600 mg/dl despite treatment

• Patient under legal protection

Individual participant data (IPD) that underlie results in publication could be shared. IPD detailed in the protocol could be shared for the purpose of a metaanalysis.

Two years after the last publication

Data sharing must be accepted by the sponsor and the principal investigator (PI) based on a scientific project and scientific involvement of the PI team. Collaboration will be fostered.

Teams wishing obtain IPD must meet the sponsor and PI team to present scientifics (and commercial) purpose, IPD needed, format of data transmission, and timeframe. Technical feasibility and financial support will be discussed before mandatory contractualization. Processing of shared data must comply with European General Data Protection Regulation (GDPR).