Fifty-eight percent of adults with type 2 diabetes who were randomized to metabolic-bariatric surgery achieved remission of their diabetes over a median of 11 years of follow-up, compared with 3% of those assigned to intensive medical and lifestyle management, according to an analysis from the ARMMS-T2D study published Oct. 20 in Diabetes Care. Patients with a shorter history of diabetes, no insulin use and greater first-year weight loss were the most likely to reach remission, the authors report.
ARMMS-T2D, the Alliance of Randomized Trials of Medicine versus Metabolic Surgery in Type 2 Diabetes, is described by the University of Oklahoma as the largest and longest-running trial of metabolic surgery for diabetes. The peer-reviewed paper analyzed 262 adults, 166 originally randomized to surgery and 96 to medical-lifestyle therapy. At baseline, 68% were women, the mean age was 49.9 years, mean body mass index was 36.4 kg/m2 and mean HbA1c was 8.5%. Remission was defined as an HbA1c below 6.5% without any glucose-lowering medication, and participants were followed for a median of 11 years and up to 14.
Beyond the headline remission rates, the paper’s purpose was to identify who benefits. In univariate time-to-event models, shorter diabetes duration, no insulin use, fewer classes of glucose-lowering medication, higher body weight, higher fasting C-peptide and insulin levels, and weight loss at one year each predicted remission (all p<0.05). In a multivariate LASSO model, three factors remained: diabetes duration at baseline, insulin use and weight change at one year (all p<0.01). Among surgery patients whose diabetes returned after remission, relapse was predicted by a longer diabetes history and a greater number of medication classes before treatment.
The study was funded by the National Institute of Diabetes and Digestive and Kidney Diseases under grant U01 DK114156. Donald Simonson of Brigham and Women’s Hospital and Harvard Medical School is lead author; John Kirwan, executive director of the OU Health Harold Hamm Diabetes Center, is the principal investigator and has led the study since it began. Participating centers include the University of Washington, Pennington Biomedical Research Center, Cleveland Clinic, Duke University, Kaiser Permanente Washington, the University of Kansas Medical Center, Weill Cornell Medicine, the University of Pittsburgh and the Joslin Diabetes Center.
“Now we have a durable answer,” Kirwan said in the university release, adding that the findings mean “we can offer surgery to the right people earlier” rather than as a last resort. The authors write that baseline predictors of remission and relapse across the three surgical procedures and medical therapy should help guide personalized treatment selection.
Why it matters
Long-term randomized data comparing surgery with medical care for diabetes are scarce, and most remission studies are observational or single-center. A 58% versus 3% gap over more than a decade is a large effect, and the predictors are clinically usable: duration of diabetes and insulin use are known at the first visit, and weight response is known within a year. The authors’ framing, and Kirwan’s, is that surgery should be considered earlier in the course of diabetes rather than after other treatments have failed.
For health systems, the predictors offer a way to prioritize patients: those with recently diagnosed diabetes who are not yet on insulin stand to gain most, while those with long-standing, heavily treated disease are less likely to achieve or keep remission. The abstract does not address cost, surgical complication rates or how the three procedures compare with one another.
The caveats
The analysis is a secondary, predictor-focused report from a cohort originally randomized in several separate trials, not a new randomized comparison. The 58% figure counts anyone who achieved remission at some point during follow-up, and the paper’s relapse analysis shows that some of those patients later lost it; relapse rates and effect sizes for the predictors, with confidence intervals, were not available in the abstract. The surgical arm was larger than the medical arm (166 versus 96). Participants had a mean BMI of 36.4, so the findings apply to people with obesity rather than to leaner patients with diabetes. The university release does not discuss limitations.
What to watch
Follow-up in ARMMS-T2D continues to 14 years for some participants. The authors say their predictors should help guide treatment selection, which would require prospective validation in other cohorts before they are used as a formal decision tool.
This article is for general information and does not replace advice from a qualified clinician. Corrections: email the editor using the address on the Contact page.
