Friday, March 11, 2016

Diabetes Research: The Twisting Path

Wouldn't it be wonderful for those of us who are pancreatically challenged if someone discovered a drug that would make our beta cells (the cells in the pancreas that produce insulin) multiply without multiplying too much (ie, cancer)?

Well, in spring 2013, a group from Harvard reported just that in mice. They called the hormone that caused beta cell replication in mouse liver and fat betatrophin (the protein was already known by other names, such as ANGPTL8 and lipasin). Probably in part because the research came from a respected research lab, it caused a lot of excitement and hope, including articles in the popular press suggesting that type 1 diabetes would soon be cured (we've all heard that one before), and other labs followed up on the Harvard research.

The saga of betatrophin is interesting in that it illustrates the pitfalls of scientific research, especially in today's world, where the research can be so complex.

In fall 2013,  another lab showed that mice in which the beta trophin gene was knocked out didn't have any changes in glucose metabolism.

And in fall 2014, a third lab reported that betatrophin did not cause beta cell proliferation.

The Harvard group then retracted their claim that betatrophin caused significant beta cell expansion, saying that further work had shown the same thing as the third lab. They said that when they used more mice, they found that some mice responded and others didn't. When they used only 7 mice they happened to have mostly responders; when they used 52 mice they found lower beta cell replication rates.

Later, an analysis of the saga asked whether a mouse system was in fact the best system in which to study human beta cell replication. And indeed another study had shown that although mouse beta cells responded dramatically to betatrophin, human beta cells were completely unresponsive.

An editorial in the journal Diabetes called this "the elephant in the room." Many studies are done in mice, and humans don't always respond like mice. They include a long list of compounds that induce robust beta cell replication in rodents but not in humans. Agreeing that despite these problems we still need to study rodents because we can more quickly get information that would be impossible in human studies, they cautioned about  not forgetting the elephant in the room.

I suspect most readers won't want to slog through all these papers. So what does this long saga offer us nonrodent patients?

I think it's a cautionary tale. We need to learn to take the results of research with a grain of salt. No one study, even a study by respected researchers in the top of their field, is definitive. It must be replicated in other labs.

This is especially true today, when research often involves dozens of different researchers, sometimes working in different labs, performing very complex techniques including manipulating genes. A mistake in just one of the myriad techniques involved could throw all the results off.

We need to be especially cautious about the popular press summaries of complex research. The popular press called the initial Harvard study a "breakthrough," and apparently physicians were overwhelmed with patients, or the parents of patients, wanting to try betatrophin.

We also need to be cautious about mouse studies. Mice aren't humans, and although they sometimes do react just like we do, other times they respond quite differently. The mouse studies just give researchers ideas that they can then try on human volunteers. Only when the human studies are done, with good safety and therapeutic results, can we begin to hope that someone will develop the compound under study.

This saga also shows how medical research isn't always a straight line from an idea to a treatment. There may be many dead ends, side trips, restarts, and disagreements. Even if several labs get the same results, there can be disagreements about the interpretion of the results.

So it's good to read reports of new research, but one must be careful about attributing too much credance to any one study. And if study A appears to show one thing and study B appears to show the opposite, we shouldn't throw up our hands and reject all scientific research. The path to the truth is twisting; we have to accept that.




  1. Most of the elderly and long term diabetics I've known have opined that a cure for diabetes has been five years away for well over 20 years now. Most of these "cures" turn out to be things that work in mice but are not replicated in humans. In fact virtually *everything* appears to cure diabetic mice.

    Perhaps a valid field of study would be what makes mice so damn special?

  2. Doug Melton has indicated that he's not interested in curing mice.