STAMPEDE

Published

Schauer PR, et al. "Bariatric Surgery versus Intensive Medical Therapy in Obese Patients with Diabetes". The New England Journal of Medicine. 2012. 366(17):1567-76.
PubMed • Full text • PDF

Clinical Question

In obese patients with uncontrolled T2DM, does bariatric surgery plus intensive medical therapy improve HbA1c better than intensive medical therapy alone?

Bottom Line

In moderately obese patients with uncontrolled T2DM, bariatric surgery plus intensive medical therapy resulted in a lower HbA1c than medical therapy alone at one and three years. Despite this study's drawbacks including short-term follow-up and the use of HbA1c as the primary endpoint, this and other studies suggest that bariatric surgery should be considered as part of comprehensive diabetes management.

Major Points

The use of bariatric surgery to treat and potentially reverse diabetes mellitus has been an attractive proposal for some time. Bariatric surgery results in reversal of diabetes in a majority of selected obese patients with T2DM (>70%) and is associated with minimal perioperative mortality. These observations prompted a number of randomized studies of weight-loss surgery in obese diabetic patients.

Published in 2012, the Surgical Treatment and Medications Potentially Eradicate Diabetes Efficiently (STAMPEDE) trial randomized 150 moderately obese patients (BMI 27-43 kg/m²) and poorly controlled T2DM to one of three groups: 1) intensive medical therapy, 2) intensive medical therapy plus Roux-en-Y gastric bypass, or 3) intensive medical therapy plus sleeve gastrectomy. At one year of follow-up, patients who received bariatric surgery were more likely to lose weight, have lower average HbA1c levels, use fewer diabetic medications, and use less insulin, compared to patients that received intensive medical therapy alone. Specifically, approximately 40% of bariatric surgery patients achieved a HbA1c ≤ 6% compared to just 12% of those treated with modern intensive medical therapy. Among patients who underwent bariatric surgery, those who received Roux-en-Y gastric bypass used fewer diabetic medications compared to patients who received a sleeve gastrectomy. Adverse events were similar across groups, although surgery was associated with a higher risk of anemia and an increased rate of hospitalization. In a subsequent report, the authors present 3-year data that demonstrates a sustained improvement in HbA1c among patients undergoing bariatric surgery.

STAMPEDE was a reasonably large study but methodologically limited by its single-center design and use of a single surgeon to perform all operations. As such, generalizability is limited. Further, the selection of HbA1c as a primary endpoint is suboptimal in a disease where the most relevant outcomes are overall survival and microvascular and macrovascular outcomes. Nevertheless, in a disease where these outcomes occur over many years and perhaps decades, a surrogate marker is necessary, and HbA1c is a validated surrogate.

Despite these qualifications, STAMPEDE demonstrated a significant and durable reduction in HbA1c among obese diabetic patients who underwent bariatric surgery. In addition to the improvements in diabetes parameters, bariatric surgery was associated with significant weight loss as well as resolution or improvement in hypertension and dyslipidemia. Although surgical weight loss cannot be recommended as a general strategy among obese diabetic patients, the benefits appear to outweigh the risks in selected patients. As a result, the American Diabetes Association and other guideline-publishing institutions have recommended that providers consider bariatric surgery as part of comprehensive diabetes management.

5 year follow-up from STAMPEDE published in 2017 described ongoing superiority of bariatric surgery versus medical therapy alone for HbA1c and BMI reduction as well as superiority for reduction medications, improvement in cholesterol, and QOL.

Guidelines

American Diabetes Association (2015) adapted

• Bariatric surgery may be considered for adults with BMI >35 kg/m² and type 2 diabetes, especially if diabetes or associated comorbidities are difficult to control with lifestyle and pharmacological therapy. (Level B)
• Patients with type 2 diabetes who have undergone bariatric surgery need lifelong lifestyle support and medical monitoring. (Level B)
• Although small trials have shown glycemic benefit of bariatric surgery in patients with type 2 diabetes and BMI 30–35 kg/m², there is currently insufficient evidence to generally recommend surgery in patients with BMI <35 kg/m². (Level E [expert opinion])

Design

• Single-center, non-blinded, randomized, controlled trial
• N=150
  • Intensive medical therapy (n=41)
  • Intensive medical therapy and Roux-en-Y gastric bypass (n=50)
  • Intensive medical therapy and sleeve gastrectomy (n=49)
• Setting: Single center at Cleveland Clinic
• Enrollment: 2007-2011
• Follow-up: 1 year
• Primary outcome: HbA1c <6.0% ± diabetes medications

Population

Inclusion Criteria

• T2DM with HbA1c ≥ 7.0%
• Age 20-60 years
• BMI: 27-43 kg/m²

Exclusion Criteria

• Previous bariatric or other complex abdominal surgery
• Poorly controlled psychiatric or medical comorbidities

Baseline Characteristics

• Mean age: 49±8 years
• Female: 67%
• White: 73%
• Mean A1C: 9.2±1.5%
• Duration of diabetes: 8.5 years
• Use of insulin at baseline: 44%
• Current smoker: 32.4%
• Mean BMI: 36±3.5 (36% BMI<35)
• Metabolic syndrome: 92%
  • Mean waist circumference: 115.0 cm
  • Mean waist-to-hip ratio: 0.96
  • Dyslipidemia: 84%
  • Hypertension: 63.3%

Interventions

• Randomized to one of three groups:
  • Intensive medical therapy (IMT)
  • IMT plus Roux-en-Y gastric bypass
  • IMT plus sleeve gastrectomy
• Patients were counseled by a diabetes educator and encouraged to participate in a Weight Watchers program.
• Patients were treated with antihypertensive and lipid-lowering medications according to ADA guidelines (SBP <130, DBP <80, LDL < 100)
• Patients randomized to undergo bariatric surgery were evaluated by psychological, nutritional, and surgical services
  • After gastric bypass, supplements included a multivitamin, iron, vitamin B12, and calcium citrate with vitamin D.
  • After sleeve gastrectomy, supplements included a multivitamin and vitamin B12.

Outcomes

Comparisons are IMT vs. IMT + Roux-en-Y gastric bypass vs. IMT + sleeve gastrectomy.

Primary Outcome

Target Hb_A1C ≤ 6.0% at 12 months

IMT vs. IMT plus Roux-en-Y gastric bypass: 12% vs. 42% (P=0.002)

IMT vs. IMT plus sleeve gastrectomy: 12% vs. 37% (P=0.008)

No significant difference between surgical groups.

Target Hb_A1c of ≤6.0% at 3 years

IMT vs. IMT plus Roux-en-Y gastric bypass: 5% vs. 38% (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 5% vs. 24% (P=0.01)

Glycemic relapse (defined as patients having a A1C ≤6.0% at 12 months, but not at 3 years) was 80% (4/5) in the medical therapy group, 24% (5/21) in the gastric bypass group, and 50% (9/18) in the sleeve group

Secondary Outcomes at 12 months

Mean Hb_A1c

IMT vs. IMT plus Roux-en-Y gastric bypass: 7.5% vs. 6.4% (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 7.5% vs. 6.6% (P<0.001)

Median fasting plasma glucose

IMT vs. IMT plus Roux-en-Y gastric bypass: 120 vs. 99 (P=0.001)

IMT vs. IMT plus sleeve gastrectomy: 120 vs. 97 (P=0.02)

Mean number of diabetes medications

IMT vs. IMT plus Roux-en-Y gastric bypass: 3.0 vs. 0.3 (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 3.0 vs. 0.9 (P<0.001)

Insulin use

IMT vs. IMT plus Roux-en-Y gastric bypass: 38% vs. 4% (P<0.01)

IMT vs. IMT plus sleeve gastrectomy: 38% vs. 8% (P<0.01)

% change in HDL

IMT vs. IMT plus Roux-en-Y gastric bypass: 11.3±25.7 vs. 28.5±22.7 (P=0.001)

IMT vs. IMT plus sleeve gastrectomy: 11.3±25.7 vs. 28.4±21.9 (P=0.001)

% change in HS-CRP

IMT vs. IMT plus Roux-en-Y gastric bypass: -33.2 vs. -84 (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: -33.2 vs. -80 (P<0.001)

Body weight (kg)

IMT vs. IMT plus Roux-en-Y gastric bypass: 99.0±16.4 vs. 77.3±13.0 (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 99.0±16.4 vs. 75.5±12.9 (P<0.001)

Median change in triglycerides

IMT vs. IMT plus Roux-en-Y gastric bypass: -14 vs. -44 (P=0.002)

IMT vs. IMT plus sleeve gastrectomy: -14 vs. -42 (P=0.08)

Secondary Outcomes at 3 Years

Median fasting plasma glucose

IMT vs. IMT plus Roux-en-Y gastric bypass: 132 vs. 100 (P=0.001)

IMT vs. IMT plus sleeve gastrectomy: 132 vs. 106 (P=0.007)

% change in LDL

IMT vs. IMT plus Roux-en-Y gastric bypass: 2.5±29.9 vs. 16.9±54.5 (P=0.14)

IMT vs. IMT plus sleeve gastrectomy: 2.5±29.9 vs. 14.5±52.2 (P=0.20)

% change in HDL

IMT vs. IMT plus Roux-en-Y gastric bypass: 4.6±20.7 vs. 34.7±27.3 (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 4.6±20.7 vs. 35.0±31.0 (P<0.001)

Body weight (kg)

IMT vs. IMT plus Roux-en-Y gastric bypass: 100.2±16.6 vs. 80.6±15.5 (P<0.001)

IMT vs. IMT plus sleeve gastrectomy: 100.2±16.6 vs. 79.3±15.1 (P<0.001)

Change in SBP

IMT vs. IMT plus Roux-en-Y gastric bypass: 0.63±22.63 vs. 1.29±20.38 (P=0.88)

IMT vs. IMT plus sleeve gastrectomy: 0.63±22.63 vs. -4.43±20.69 (P=0.27)

Change in DBP

IMT vs. IMT plus Roux-en-Y gastric bypass: -6.48±12.33 vs. -4.25±10.57 (P=0.36)

IMT vs. IMT plus sleeve gastrectomy: -6.48±12.33 vs. -6.25±13.30 (P=0.94)

Mean number of diabetes medications

Roux-en-Y gastric bypass vs. sleeve gastrectomy: 0.48±0.80 vs. 1.02±1.01 (P=0.02)

Number of patients achieving levels of Hb_A1C ≤7.0% without diabetes medications

Roux-en-Y gastric bypass vs. sleeve gastrectomy: 58% vs. 33% (P=0.01)

Subgroup Analysis

Reductions of Hb_A1C, medication use, and BMI in the surgical groups were similar between those with BMI<35 and BMI>35.

Adverse Events at 3 Years

Comparisons are IMT vs. IMT vs. IMT plus Roux-en-Y gastric bypass vs. IMT plus sleeve gastrectomy.

Excessive weight gain: 16% vs. 0% vs. 0%

Hypoglycema: 91% vs. 64% vs. 82% (severe hypoglycemia requiring intervention: 0% vs. 2% vs. 0%)

Bowel obstruction: 2% vs. 2% vs. 2%

Stricture: 0% vs. 2% vs. 2%

Ulcer: 2% vs. 8% vs. 0%

GI leak: 0% vs. 4% vs. 0%

Intraabdominal bleeding: 0% vs. 2% vs. 2%

Dumping syndrome: 0% vs. 8% vs. 2%

Gallstone diseases: 0% vs. 2% vs. 2%

Stroke: 0% vs. 0% vs. 2%

Retinopathy: 0% vs. 2% vs. 4%

Nephropathy: 9% vs. 14% vs. 10%

Foot ulcer: 0% vs. 4% vs. 2%

Anemia: 14% vs. 16% vs. 31%

IV treatment for dehydration: 7% vs. 14% vs. 8%

Wound infection: 0% vs. 2% vs. 0%

Hernia: 2% vs. 6% vs. 2%

Pneumonia: 0% vs. 4% vs. 2%

Renal calculus: 14% vs. 10% vs. 8%

Cancer: 5% vs. 4% vs. 4%

No deaths, excessive weight loss or hypoalbuminemia, or life-threatening complications were reported in any of the treatment groups.

Criticisms

• Single-center study with only a single surgeon performing all of the bariatric operations; generalizability is therefore limited
• Long-term efficacy and safety cannot be determined in this duration of time
• Not adequately powered to detect differences in macrovascular outcomes
• Not adequately powered to detect small differences between the two bariatric surgeries
• HbA1c is an inadequate primary endpoint; long-term survival is the desired endpoint

Funding

Ethicon Endo-Surgery with support from LifeScan, the Cleveland Clinic, and the National Institutes of Health