Once-promising diabetes ‘breakthrough’ comes unglued with a major retraction

One of the highest-profile researchers in diabetes has retracted a paper once heralded as a breakthrough, following multiple failed attempts to reproduce its headline-grabbing results.

The retraction ends three years of scrutiny into whether a discovery by Harvard University stem cell scientist Douglas Melton was indeed a major advance in the field of diabetes, with the paper’s authors now conclusively backing away from their earlier findings.

Back in 2013, Melton’s lab reported a promising discovery: A hormone found in the liver seemed to spur the production of insulin-producing cells in mice, lighting the way for a new approach to treating diabetes. The paper, published in the journal Cell, drew attention around the world, as it suggested a means of boosting insulin by using the body’s own machinery and held out the potential to free millions of diabetes patients from regular injections.

But the claim soon lost its luster.

In 2014, an independent group of researchers found that the hormone, which Melton’s lab dubbed betatrophin, had no effect on insulin production. Melton’s team then published a follow-up that cast similar doubts on the initial findings. Earlier this year, Melton joined researchers from Baylor College of Medicine and the Harvard-affiliated Joslin Diabetes Center in publishing a strong refutation of the idea that betatrophin can mobilize a spike in insulin production.

In a retraction notice posted this week, Melton and his Harvard co-authors concede that their initial conclusion “is wrong and cannot be supported,” effectively burying the once-promising idea.

Melton, who was on vacation and unavailable for comment Tuesday, told the Retraction Watch blog he chose to pull the paper in hopes of avoiding confusion.

It’s a disappointment, like any retraction, but the gradual teardown of the betatrophin hypothesis illustrates “how scientists can work together when they disagree, and come together to move the field forward,” Melton said.

“The history of science shows it is not a linear path,” he added.

Melton is a pioneer in stem cell research, through which biological blank slates can be transformed into any type of bodily cell. He embraced the idea in the early 1990s after his son was diagnosed with Type 1 diabetes, setting out to turn moldable stem cells into beta cells, the body’s insulin factories, in hopes of crafting a cure for the disease.

That work set him down the path to betatrophin. In Melton’s initial paper, injecting mice with the new-found hormone led to a 17-fold increase in pancreatic beta cells. If the effect could be replicated and transferred to humans, Melton thought, diabetics might one day be able to trade their daily injections of insulin with a dose of betatrophin that could be administered as infrequently as once a year.

But replication never came.

That first test involved just seven mice, and when Melton’s team ran five more experiments on 52 mice, the magnitude of beta cell production fell considerably. Separately, a group of researchers led by Regeneron Pharmaceuticals’ Viktoria Gusarova tried both knocking out and dialing up the production of betatrophin, only to find that neither had a marked effect on beta cells.

To give the hypothesis a final shake, Melton recruited researchers from Baylor and Joslin to run a blinded study. In the end, all three groups came to the same conclusion: Betatrophin offered little hope as a treatment for diabetes.

Correction: A previous version of this story incorrectly described the groups that published a study in 2016 refuting the idea that betatrophin can spur the creation of insulin-producing beta cells. The researchers who worked with Melton on that study were from Baylor College of Medicine and Joslin Diabetes Center.