Bariatric-metabolic surgery versus conventional medical treatment in obese patients with type 2 diabetes: 5 year follow-up of an open-label, single-centre, randomised controlled trial

Background: Randomised controlled trials have shown that bariatric surgery is more effective than conventional treatment for the short-term control of type-2 diabetes. However, published studies are characterised by a relatively short follow-up. We aimed to assess 5 year outcomes from our randomised trial designed to compare surgery with conventional medical treatment for the treatment of type 2 diabetes in obese patients.

Methods: We did our open-label, randomised controlled trial at one diabetes centre in Italy. Patients aged 30-60 years with a body-mass index of 35 kg/m(2) or more and a history of type 2 diabetes lasting at least 5 years were randomly assigned (1:1:1), via a computer-generated randomisation procedure, to receive either medical treatment or surgery by Roux-en-Y gastric bypass or biliopancreatic diversion. Participants were aware of treatment allocation before the operation and study investigators were aware from the point of randomisation. The primary endpoint was the rate of diabetes remission at 2 years, defined as a glycated haemaglobin A1c (HbA1c) concentration of 6·5% or less (≤47·5 mmol/mol) and a fasting glucose concentration of 5·6 mmol/L or less without active pharmacological treatment for 1 year. Here we analyse glycaemic and metabolic control, cardiovascular risk, medication use, quality of life, and long-term complications 5 years after randomisation. Analysis was by intention to treat for the primary endpoint and by per protocol for the 5 year follow-up. This study is registered with ClinicalTrials.gov, number NCT00888836.

Findings: Between April 27, 2009, and Oct 31, 2009, we randomly assigned 60 patients to receive either medical treatment (n=20) or surgery by gastric bypass (n=20) or biliopancreatic diversion (n=20); 53 (88%) patients completed 5 years' follow-up. Overall, 19 (50%) of the 38 surgical patients (seven [37%] of 19 in the gastric bypass group and 12 [63%] of 19 in the bilipancreatic diversion group) maintained diabetes remission at 5 years, compared with none of the 15 medically treated patients (p=0·0007). We recorded relapse of hyperglycaemia in eight (53%) of the 15 patients who achieved 2 year remission in the gastric bypass group and seven (37%) of the 19 patients who achieved 2 year remission in the biliopancreatic diversion group. Eight (42%) patients who underwent gastric bypass and 13 (68%) patients who underwent biliopancreatic diversion had an HbA1c concentration of 6·5% or less (≤47·5 mmol/mol) with or without medication, compared with four (27%) medically treated patients (p=0·0457). Surgical patients lost more weight than medically treated patients, but weight changes did not predict diabetes remission or relapse after surgery. Both surgical procedures were associated with significantly lower plasma lipids, cardiovascular risk, and medication use. Five major complications of diabetes (including one fatal myocardial infarction) arose in four (27%) patients in the medical group compared with only one complication in the gastric bypass group and no complications in the biliopancreatic diversion group. No late complications or deaths occurred in the surgery groups. Nutritional side-effects were noted mainly after biliopancreatic diversion.

Interpretation: Surgery is more effective than medical treatment for the long-term control of obese patients with type 2 diabetes and should be considered in the treatment algorithm of this disease. However, continued monitoring of glycaemic control is warranted because of potential relapse of hyperglycaemia.

Funding: Catholic University of Rome.