13 responses to “Incretin Safety: What is the Evidence?”

  1. Just to check: no effect on hard outcomes, question of increase adverse effects, and a trial that shows worse than placebo?

  2. I agree with the above comment. The goal of diabetes treatment is to reduce meaningful adverse outcomes (ie. premature CAD), not simply to make HgA1c lab test results look better. These drugs are expensive, have side effects, and it seems that the very best that we can say about them is that at least they’re NOT WORSE than placebo when it comes to important cardiac outcomes?

  3. I agree with the previous two comments. We’re supposed to be happy that these drugs are considered merely safe from a cardiovascular standpoint, rather than actually being advantageous? Sure, I suppose there might be long term benefit in terms of microvascular endpoints but given how new and expensive these drugs are, I think I would consider them a last resort.

    Additionally, there was a recent autopsy study that purported to find potentially pre-malignant changes in pancreatic tissue in those who took incretin agents. (http://diabetes.diabetesjournals.org/content/early/2013/03/17/db12-1686.full.pdf+html) I think a longer, larger trial is needed to establish whether there is, in fact, a link between incretin agents and pancreatic neoplasia, and until that happens, these medications will not be in my first-line armamentarium.

  4. a recent therapeutic initiative article questioned the use and validity of using HgA1c as an end marker for management of type 2 diabetes…manage the patient not the blood work. It appears that these studies do just that. Does the evidence support lower of HbA1c actually lower morbidity and prolong life? Given the haze around the increased hospitalization issue I would think these drugs warrant further study before widespread use.

  5. Thank you for your comments. A number of important issues have been raised, which I will attempt to address.
    1. A1C and the risk of micro vs. macrovascular disease
    To date, incretin agents have demonstrated a similar degree of glucose lowering as other available oral diabetes medications (A1C lowering – 0.5-1% for DPP-4 inhibitors and 1-1.5% for GLP-1 receptor agonists). Glucose lowering (A1C) has been clearly linked to the risk of diabetic microvascular complications (primarily retinopathy and nephropathy). We also know that reducing A1C reduces the risk of microvascular complications (DCCT, UKPDS, ADVANCE), which is the primary justification for tighter glucose control in diabetes patients.

    The link between A1C and macrovascular disease (primarily heart attack and stroke) has not been well established. So far, no randomized controlled trial (RCT) has demonstrated a significant reduction in cardiovascular (CV) events by glucose lowering alone, no matter what agents have been used. However, long-term follow-up observational data from RCTs such as UKPDS (type 2 diabetes) and DCCT-EDIC (type 1 diabetes) suggest that improved glucose control reduces macrovascular events over 10-20 years (so-called “legacy effect” or “metabolic memory”). However, this data is not conclusive.

    2. Demonstrating benefit vs. safety of new diabetes agents
    First it should be noted that this article was intended to address safety concerns surrounding incretins. The question of efficacy, whether measured by glucose lowering or other endpoints, was beyond the scope of this brief article and was not specifically addressed.

    It should be noted that the studies of incretin safety referenced in the article (SAVOR-TIMI53 and EXAMINE) were designed as CV safety trials and were never intended or designed to demonstrate efficacy (i.e. reduction of either micro or macrovascular complications). These studies demonstrated short-term CV safety (except for the unexpected finding of increased admissions for heart failure in the SAVOR-TIMI53 trial), but the absence of short-term CV benefit is not unexpected and is consistent with other trials, such as ACCORD, ADVANCE and VADT. Furthermore, in the SAVOR-TIMI53 trial the difference in A1C between saxagliptin and placebo arms at the end of the trial was 0.2% [A1C 7.7% (saxagliptin) and 7.9% (placebo)]. This is not a statement of difference in efficacy since the placebo group required more antihyperglycemic agents including higher doses of sulfonylurea and insulin. However, it would be extremely unlikely that CV benefit would be seen for such a small A1C difference over such a short a period of time.

    Consequently, based on current evidence, it appears that short-term (3-5 year) cardiovascular benefit cannot be expected by lowering A1C alone. Whether or not long-term CV benefit can be achieved from glucose lowering remains unclear. Unfortunately, to confirm this would require a longer period of follow-up than most RCTs can realistically undertake, so we are not likely to see such a study anytime soon.

    3. Evidence of risk of pancreatitis and pancreatic cancer
    So far, the clinical evidence does not support an association between incretin use and pancreatitis or pancreatic cancer. Since the publication of the SAVOR-TIMI53 and EXAMINE trials, there have been other systematic analyses that show similar results (i.e. no association). In fact, the FDA and the European Medicines Agency (EMA) reviewed the existing data and their findings are published (N Engl J Med 2014; 370:794-797) and summarized as follows: “the FDA and EMA have explored multiple streams of data pertaining to a pancreatic safety signal associated with incretin-based drugs. Both agencies agree that assertions concerning a causal association between incretin-based drugs and pancreatitis or pancreatic cancer, as expressed recently in the scientific literature and in the media, are inconsistent with the current data.”

    The article referenced in the comments section above, raising the issue of potential expansion of exocrine and endocrine tissues of patients taking incretin agents has a number of marked methodological flaws, which I will not cover here, but have been well documented (Bonner-Weir, et al. Diabetes Obes Metab. 2014 Jul;16(7):661-6). Furthermore, several studies have been published since, which have not demonstrated a statistical association between incretins and pancreatic disease.

    1. Faillie JL, et al. Incretin based drugs and risk of acute pancreatitis in patients with type 2 diabetes: cohort study. BMJ. 2014 Apr 24;348:g2780. doi: 10.1136/bmj.g2780.
    2. Giorda CB, et al. Incretin-based therapies and acute pancreatitis risk: a systematic review and meta-analysis of observational studies. Endocrine. 2014 Aug 22. [Epub ahead of print].
    3. Meier JJ(1), Nauck MA. Risk of pancreatitis in patients treated with incretin-based therapies. Diabetologia. 2014 Jul;57(7):1320-4. doi: 10.1007/s00125-014-3231-y. Epub 2014 Apr 11.
    4. Thomsen RW, et al. Incretin-Based Therapy and Risk of Acute Pancreatitis: A Nationwide
    Population-Based Case-Control Study. Diabetes Care. 2015 Jan 29. pii: dc132983. [Epub ahead of print]
    In summary, the published article is intended to give an overview of the current understanding of incretin safety in the clinical setting. It is clear that glucose lowering reduces the risk of microvascular complications of diabetes (no matter what agents are used), but not necessarily macrovascular complications in the short term. As far as incretin safety is concerned, the weight of evidence so far suggests that this class is generally safe, but we await further results from additional CV safety trials, which will no doubt shed more light on this important topic. As with any other class of agents, it is up to each individual practitioner to make his/her own judgement regarding relative risks versus benefits of any medication he/she prescribes.

  6. Thank you Dr. Paty, for your detailed response to some of our concerns. I would like to respond to several of your points to illustrate why some of us are still hesitant about using incretin agents.

    A1C and the risk of micro vs. macrovascular disease: You point out that microvascular outcomes can be reduced by improving hyperglycemia. For many of our patients, who are being diagnosed with type 2 diabetes later in life, the lifetime risk of end-stage renal disease and significant retinopathy is relatively low, and the absolute benefit of intense A1C reduction is potentially limited. For example, this chart from Vijan et al. (1) suggests that in older patients with moderately-controlled T2DM, there appears to be little benefit to tight glycemic control with regard to significant microvascular outcomes. (I don’t know of any similar data for diabetic neuropathy, however.)

    Demonstrating benefit vs. safety of new diabetes agents
    You point out that the cardiac safety trials involving DPP4 antagonists were not designed to prove efficacy in terms of reducing CV risk. However, I would note that the designers of the SAVOR-TIMI 53 trial specifically state that the trial was powered to find a 17% risk reduction in CV events (2), which it failed to do. Perhaps this was overly optimistic wishful thinking on the part of the trial sponsors? I’m sure they were certainly hoping for a macrovascular benefit!

    Evidence of risk of pancreatitis and pancreatic cancer
    Thank you for pointing out the reanalysis of the histopathology data. It does put the original paper in a different light

    I appreciate the opportunity to have this discussion. Looking forward to hearing more about other new agents, e.g. SGLT2 inhibitors.

    Vijan S, Hofer TP, Hayward RA. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med. 1997 Nov 1;127(9):788-95.

    Scirica BM et al.The design and rationale of the saxagliptin assessment of vascular outcomes recorded in patients with diabetes mellitus-thrombolysis in myocardial infarction (SAVOR-TIMI) 53 study. Am Heart J. 2011 Nov;162(5):818-825.e6.

  7. Dr. Paty

    Could you explain how “glucose lowering (A1C) is clearly linked to the risk of diabetic microvascular complications (primarily retinopathy and nephropathy).” I thought this was only based on a small body of evidence and important clinical retinopathy and clinical nephropathy risk reduction was not shown. What specific outcomes were improved?

    Would lowering A1C by any means lead to a microvascular complication risk reduction. Did rosiglitazone reduce the risk of microvascular complications while it reduced A1C (and increased the risk of MI)?

    Looking forward to your response.


  8. While I support UBC’s attempt to provide “real time” CPD (this changed my practice etc) to busy doctors, perhaps this would be better cast as an attempt to stimulate debate.

    I found the responses more edifying than the original article. They convinced me not to use these agents until more data is available.

    Perhaps a balanced critique of each topic should be provided at the outset to ensure we are getting both sides of the story given the industry connections of many of the authors.

  9. This article and the discussion about it certainly did not “change my practice” which is to not use medications for diabetes that are expensive and have not been demonstrated to reduce risk.

  10. Great to have these online discussions! Two issues:

    1. Data from high dose (3 mg/d) liraglutide, recently approved for weight loss, suggest there is indeed an risk increased risk of pancreatitis (7 vs 1) and thyroid cancer (4 vs 1), though the numbers were too small to be anything more than a bit concerning. Of course, with malignancies, clinical trials are generally of too-short duration to be useful in demonstrating safety, and long-term randomized trials are impossibly impractical, so we will probably never know for sure. But pancreatitis is now back on my list of warnings to patients, even if the mechanism is in fact the rapid weight loss itself..

    2. I continue to be distressed by med students and residents explaining that “DPP-4 breaks down incretins” and ending their explanation there. Am I the only person concerned that this enzyme has literally five DOZEN substrates in humans, including, for example, growth hormone? DDP-4 is also known as CD26, a lymphocyte activator, possibly relevant in a population at risk of foot infection. It is also present, though with unknown function, on numerous cancer cells….will it promote these cancers or suppress them, in a patient population that is at the age where cancer is common? Combine all that with the fact that the longest randomized trial to date of a DPP-4 inhibitor was 54 weeks long, and I’ll be the first person to say “I told you so” when we start getting signals of “unexpected” adverse events.

  11. Very interesting debate!
    Considering that incretins are very expensive there are only maybe 10-15% of DM patients who can afford them.
    The benefits probably outweight risks, but we have to consider that every case is different.
    Also I believe we should educate our patients about the risks and benefits of incretins.

  12. I see the current commentary has turned more toward whether or not glycemic control has benefit. The original intent of the article was to address questions around incretin safety, and we’ve moved beyond, into a much larger (although related and important) discussion, and this is beyond the scope of this article.

    However you accept or reject them, A1C targets remain central to the diabetes guidelines around the world. If the comments we see here are any indication, the specific topic of A1C targets warrants a major debate on its own.

    The short article format on TCMP is meant as a springboard for discussion and further learning. From an editorial point of view, we felt that Dr. Paty provided a good summary of the available data (specifically addressing incretin safety, rather than A1C lowering benefits, or pharmacoeconomics) within the limits of our article format. From this comment thread, it is apparent that, for complex topics, we might need to look at more in-depth articles such as a debate format where we have 2 authors presenting the pros and cons to a specific topic.

    We’ve taken the feedback and comments to heart and will explore potential for expanded offerings in the future. If this is something you’d like to see (and even better, participate as an author!) please drop us a note at feedback@thischangedmypractice.com

    I’d like to thank the readers as well as Dr. Paty for taking their time out to comment. I hope this has been a useful educational experience.

  13. I have read Dr. Paty’s article when it came out in February, and received and reviewed the comments with some dismay. As a diabetes educator, I found Dr. Paty’s article an informative review of the research on incretins that is presently available, so I was quite surprised at the comments that this article generated. Although the responses have been referred to as debate, many of the comments felt more like attacks! Nothing in the way of other research has been given to support any of the responders’ comments, or to refute any of Dr. Paty’s practice suggestions. Many of the responses seemed to be personal opinion, rather than research based information. Although the lack of research based commentary was worrisome, what troubled me more was the personal nature of the responses. I would imagine that other potential authors will see these comments and think twice about making a submission to this very valuable resource. What a shame!

Leave a Reply