In Which Knives Are Effectively Used to Treat Diabetes: STAMPEDE at 5-Year Follow Up

Hi all! It’s Bill. Today’s piece comes at you courtesy of Dr. Priya Joshi, one of my fellow members of the Primary Care track here at Penn. Priya has kind of a Doogie Howser vibe going, given that she is like 11 years younger than I am and is way too smart. She’s psyched about writing more stuff, and we’re psyched to have her on board. So, here we go:

STAMPEDE On: Five-year Outcomes for Bariatric Surgery vs. Intensive Medical Therapy for Diabetes

Let’s all agree that weight loss is hard, diabetes makes it harder, and a lack of evidence comparing weight loss therapies makes it one of the hardest topics to broach in the office. But a new five year outcome update by Philip Schauer and the STAMPEDE investigators (“Great band name!” – Bill) tests the durability of the findings from the landmark trial of intensive medical therapy versus bariatric surgery for type two diabetes. So fear not, STAMPEDE On (ba dum tsh), and think again about surgical options for anyone with BMI 27 – 43 who can afford surgery*.

*Predictable comment about insurance and health care policy in the current political climate.

Too long Didn’t Read (TL;DR)

At five year follow up of this single-center randomized control trial, in the 134 remaining  patients out of 150 who originally underwent intensive medical therapy, sleeve gastrectomy, or roux-en-y gastric bypass, surgical options continued to be clearly superior in achieving and maintaining glycemic control, with additional benefits in weight loss, blood pressure control, and, for what it’s worth, lipids. (Read: you can downgrade from high intensity to moderate intensity atorvastatin). The adverse effects of surgery are very real, but don’t effect all patients, and are certainly not out of proportion to its benefits.

Too long Did* Read (TL;D*R)

Overall, the average patient who comes to your office has tried and failed to lose weight and get their diabetes under better control. At some point after referring them to a nutritionist, talking about weight loss programs, and thinking about medications, you look at their BMI a second time and wish there was a magic pill. We can talk about magic pills at a separate time, but for now let’s cut the small talk, peal apart some fascial layers, and talk surgery.

The STAMPEDE trial split 150 patients in a 1:1:1 parallel study looking at intensive medical therapy, sleeve gastrectomy, and Roux-en-Y Bypass surgery done at your friendly neighborhood Cleveland Clinic. To control for the heterogeneity of surgical skill, every procedure for the entire study was performed by the same, single surgeon, who is probably one of the busiest men alive. For the most part patients who joined the study stayed in their assigned therapy category, with one exception (Sigh, we’re all looking for The One) who bailed on the medical and swapped for the surgical. These motivated individuals sported an average weight of 100 kg, age of 50 years, and A1c just under 10%, with an average BMI in the mid-30s. After they got randomized to each group, they were followed for five years.

The data has already been reported at 1- and 3-year follow up, and did not look good for medical management: at 1 year, the primary end point of a hemoglobin A1c of 6.0% or less happened in 12% (5 of 41 patients) in the medical-therapy group versus 42% (21 of 50 patients) in the gastric-bypass group (P=0.002) and 37% (18 of 49 patients) in the sleeve-gastrectomy group (P=0.008). “But an A1c of 6.0% is an unreasonable endpoint with no data to support its association with better long-term patient outcomes!” I hear you cry. Astute! However, given that mean A1c in the surgery groups was 6.5 +/- 1% vs 7.5 +/- 1.8% in the medical therapy group, let’s put away the pitchforks and admit that surgery looked better.

At the end of five years, these benefits are largely preserved. As Darth Vader once said, “The force is strong with this [surgeon]” because people in either surgical group had significantly lower A1cs, need for glycemic control medications, and waist lines. (“I think she means lower waist circumferences, but I didn’t read the appendix so maybe she means they wore low rise jeans, I dunno I’m leaving it as-is.” -Bill) Before you raise your skepticals (skeptical spectacles) to Darth Vader, remember that Yoda once also said “Judge me by my size, do you?” and sadly got randomized to the intensive medical therapy group with only a 5kg mean weight loss. I know you think there can’t be another Star Wars reference, and you’re wrong because that one patient I mentioned before who jumped from medical therapy to surgical intervention was in fact Yoda who also said “[screw this], Do. Or do not. There is no try.” and then lost 20kg, 2 (points?) of A1c, and stopped needing medications to treat his diabetes.

I feel like that third Star Wars reference may have made things a little hard to follow. So try this figure, that has almost nothing to do with Star Wars.


Schauer et al., N Engl J Med 2017; 376:641-651

Let’s pause for a sec graph D, above. Even patients in the surgery group whose BMI at a given point in the study was > 35 (i.e. who in general did not have sustained weight loss), A1c levels were more likely to be controlled in line with those surgical patients with a BMI < 35 than with patients in medical therapy. This may indicate that surgery has a separate benefit on glycemic control above and beyond weight loss.

Hm. Something Smells. It’s Not Fishy. It Just Smells.  

Now I know what you’re saying – many wake up and think about it every morning: Dumping Syndrome. Specifically, what was the risk of this as a complication of the surgical arm. And let me tell you, it’s not as shitty as I thought it would be. And that pun? Totally intentional. I own that pun.

Dumping Syndrome: 8% (1/12 patients) vs I always thought it was 50%.

But how intense was their medical therapy? Were they actually going to get anyone to lose weight?

Medical Therapy:

  • Baseline diabetes education class
  • Metformin > Actos Exenatide Sulfonylurea or Repaglinide Insulin +/- pioglitazone for goal A1c less than 6%

There’s that 6% again. Not exactly a standard-of-care A1c goal, given that most of us are happy under 7%. Perhaps that led more patients to be on insulin, decreasing the weight loss in the medical therapy group, except that we can see fewer patients were on insulin at the end of the intervention than the start of it. So perhaps I’m making excuses for how bad the surgeons are kicking our asses on this one. Besides, let’s all take a deep breath and remember that those who volunteer for a study and have follow up every three months for a year then six months thereafter will have at baseline a higher success rate with weight loss, so it’s really hard to argue that this medical therapy group might have underperformed what we see in practice. Meanwhile,those who got surgery lost 20kg (about 50lbs) with A1cs in the low 7s and fewer medications, whereas those who got medications and counseling lost 5kg (about 10lbs). The difference is significant both statistically and clinically.

This Seems Like a Good Place to Stop

Overall, the evidence is still a part of a small sample size that predominantly looks at motivated individuals over a period of five years. The risk-benefit discussion remains a careful one that should include that 1/12 patients may live a life of pooping one’s pants post-eating. Shit gets real. But for those who can afford it, there is at least some evidence that suggests surgery may be of more use than another discussion about carb counting. And if they’re motivated and able to undergo the procedure like the 150 patients in STAMPEDE, they may be very glad they did.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s