Expert Blog

Can Current Drug Therapy Prevent Beta Cell Failure and Bring Treatment Durability in T2DM?

John L. Leahy, MD

A hot topic in the type 2 diabetes world is whether we have in hand the tools to stop the decline in beta cell function that typifies this disease and, consequently, a therapy or therapies that successfully control blood glucose for many years – so-called treatment durability. Actually this is three topics. What are the specific mechanisms for the beta cell failure? Do any of our existing therapies, or those on the drawing board, reverse these mechanisms to slow or stop the beta cell failure? Do any of our current or potential therapies have treatment durability regardless of how they do it?

I spoke on this subject at the annual meeting of The Endocrine Society this past June. A couple of weeks later there was a multi-speaker symposium at the annual meeting of The American Diabetes Association–Chris Rhodes discussed basic beta cell biology, Stefano Del Prato argued that it was defective beta-function that caused the beta cell failure in type 2 diabetes, while Michael Nauck (standing in for Juris Meier) argued for a lowered mass of functional beta cells and David D’Alessio reviewed whether any of the existing therapies are beta cell protective. To summarize, a large number of cellular mechanisms for the beta cell failure in type 2 diabetes have been proposed, and we have therapies that at least theoretically could be protective. But are they? I’ll structure my remarks around that issue.

Improved glucose control

Many primary providers would probably consider reversal of glucose toxicity as a strategy to protect beta cells, in part because this is a well-known concept, and most providers have patients who responded incredibly well, almost magically, to early therapy. This became a hotly discussed topic in 2008 based on a study from China that intensively treated new onset patients with type 2 diabetes (A1c 9.5-9.8%) with oral agents or insulin to normalize blood glucose values, with therapy stopped two weeks later.[1] Amazingly, a year later half of the insulin-treated patients still required no pharmacologic therapy (i.e., they had fasting blood glucose less than 126 mg/dL and 2-hour postprandial glucose less than 180 mg/dL). The one-year remission rate was lower in the oral agent-treated group—27%. Whether these findings mean that insulin has superior protective potential or were a quirk of the protocol is unknown. More important to this discussion, however, was the fact that these results were paralleled by sustained improvements in beta cell function in the remission groups–prompting an ongoing discussion in the specialty diabetes world about whether early intensive therapy, especially with insulin, can alter the natural course of the beta cell dysfunction in type 2 diabetes. Additional insight will hopefully come when the ORIGIN trial finishes later this year. This trial used basal insulin versus standard oral agent therapy over eight years in more than 10,000 subjects with prediabetes or early-onset type 2 diabetes, with the primary outcome cardiovascular events but also including measures of beta cell function.[2] Whatever these findings might be, however, it’s important to remember that we’re only talking about treating early stage disease: everyone’s clinical experience with later stage type 2 diabetes is disappointing in terms of aggressive blood glucose control, stabilizing beta cell function, and reducing the need for therapy.

My Take: On balance intensive blood glucose control has little to no beta cell protective effect in the average patient with type 2 diabetes of more than a few years duration. It does have a role, but for prevention of microvascular and possibly macrovascular complications.


I’ve written on this subject in the past.[3] Several TZD treatment studies have shown incredible success (best we have) at preventing conversion of prediabetes to overt type 2 diabetes through stabilizing beta cell function. The latest is the ACT NOW trial with a 72% success rate.[4] It's worth restating that no other therapy, lifestyle or pharmaceutical, comes close to that result. Also the ADOPT trial, performed in person with recent onset type 2 diabetes, reported better durability of rosiglitazone versus metformin or glyburide as first-drug therapy[5] because of stabilized beta cell function and improvements in insulin sensitivity.[6] The mechanism is thought to be “resting” of the beta cell – the lowered work load from the improved insulin sensitivity stops the beta cell excess drive or metabolic stress, thereby protecting the beta cell mass and/or function.

My Take: TZDs might be the best we have. Also, a lot could be learned about beta cell protection from finding out exactly how TZDS do what they do. Unfortunately, it's a moot point. These drugs are disappearing from the marketplace – the latest scare is bladder cancer with pioglitazone – and with them interest in better understanding PPARg physiology and pharmacology.

Incretin therapy

Many had hoped that incretin therapy – GLP-1 receptor agonists or DPP-4 inhibitors – would bring treatment durability. This is because of the current buzz over the importance of the incretin system in post-meal blood glucose control generally, how it’s broken in type 2 diabetes, and the ability of these agents to raise beta cell mass and function in preclinical rodent studies. Patricia Brubaker discussed this in a prior blog on this website. So what about human beings? The first answer is that it’s too early to say anything definitive. Second, we’re learning that it’s really hard to design a clinical trial to know if beta cell function is in fact stabilized or improved. It must be done after the drug is stopped to eliminate a “treatment effect” – but how long is too long to guard against a benefit being lost through recurrence of the disease? What is a meaningful effect? Test against an active comparator or no treatment? And how do we know whether a benefit is due to the drug itself and not more generally from improved blood glucose control during treatment? Such questions go on and on. A group from Amsterdam has taken on the challenge, but so far the findings are not particularly encouraging. They showed one year of vildaglipitin (a DPP-4 inhibitor available outside the U.S.) followed by a 12-week washout did not induce an improvement in beta cell function.[7] Also, exenatide for one year caused a sustained improvement in beta cell function when on therapy, but not after four weeks off the drug.[8] In contrast, an abstract at the 2010 American Diabetes Association annual meeting reported that three years of exenatide therapy was associated with improved beta cell function after a four-week washout, but the effect was so modest as to be clinically meaningless from my perspective.

My Take: We don’t have evidence for incretin therapy inducing a sustained rise in beta cell mass and/or function after stopping the drugs, at least after four weeks. Given the hope and excitement that came with these drugs based on the rodent findings, that’s a big disappointment for many. However, what has been lost in this discussion is that the studies DO show a continued improvement in beta cell function when on drug. So the million dollar question (many billions is probably more correct) is how long does that treatment benefit last, and will it bring a durability that exceeds what we currently have and is clinically meaningful? That’s the information we are really waiting for.

Interleukin-1 receptor antagonists

An example of a therapy possibly on the drawing board is interleukin-1 receptor antagonists. Of course, these drugs already exist and are used in various arthritis conditions. Trying them for type 2 diabetes results from basic science, which uncovered that islet inflammation may be a pathogenic mechanism for the beta cell pathology in type 2 diabetes.[9] So the next step was trying interleukin-1 receptor antagonists in type 2 diabetes. The first high profile study was a 13-week pilot trial with anakinra in 70 persons with type 2 diabetes that showed a 0.4% lowering of A1c along with a modest improvement in stimulated c-peptide secretion.[10] However the same patients studied 39 weeks after anakinra withdrawal showed a loss of the improved c-peptide secretion in the overall group, although the proinsulin to insulin ratio was still modestly improved.[11]

My Take: Sexy and timely research topic. However the clinical results so far (admittedly very early) seem a bit underwhelming.


I finished my talk at the Endocrine Society with the following summary slide:

- Many existing diabetes therapies improve beta cell function:

  • Direct
  • Indirect effects -- Beta cell rest or reversal of glucose toxicity. Thus, existing therapies may be useful for prevention of diabetes

- Durability or beta cell protection REQUIRES proof of a reduction in the progressive decline in beta cell function in type 2 diabetes by proving all of the following:

  • Long-term drug response and attenuation of the progressive beta cell failure -- proof requires at least a five-year study
  • Meaningful sustained improvement in beta cell function after discontinuing the drug
  • Plausible biological mechanism for the beta cell protection
  • No current therapy meets these criteria