Skip to Content

Two Diabetes Drug Mysteries

Here are some data to file under “Drugs do things that we don’t expect”. The SGLT-2 inhibitors are a class of diabetes medications that work by inhibiting the sodium/glucose transporter 2 protein in the kidneys. That keeps glucose from being reabsorbed there; instead, more of it is removed in the urine, and that lowers circulating glucose levels. One side effect, as you might imagine, is an increased risk of urinary tract infections, but overall, the class seems to have a lot of beneficial effects.

Too many beneficial effects, actually. One of the major drugs in this category,  Jardiance (empagliflozin) from Boehringer and Lilly, has recently been the subject of a big outcomes trial by the two companies. And the results were good – the drug reduced cardiovascular mortality, all-causes mortality, and hospitalizations from heart failure. Good news! But when the team dug further into the data, things got weird. You’d think that these benefits would be due to reductions in glycosylated hemoglobin (HbA1c), lower LDL cholesterol, lower blood pressure, etc. But when they corrected for all these factors, the effects persisted.

“It’s quite clear that the results that we see” from the drugmakers’ Empa-Reg Outcome study—including the 38% reduction in the risk of cardiovascular death—“really is not explained through these classical risk factors we have all been aware of for some decades now,” Thomas Seck, Boehringer’s VP of clinical development and medical affairs for its primary care unit, said in an interview.

So what the heck is it explained by? At this point, no one knows. This is reminiscent of the situation with statins, whose good outcomes are not completely explained by their reduction of LDL levels. This should serve as a reminder that (1) there are a lot of biochemical mechanisms that we don’t know about yet and (2) the ones that we know about aren’t necessarily as important as we’ve made them out to be.

Meanwhile, at the same ADA meeting where these results where released, J&J presented data on their own SGLT-2 inhibitor, Invokana (canaglifozin). And with this one, too, patients were notably less likely to suffer cardiovascular events, which is good news. But there was also an unexpected increased in the risk of amputation (which is already a risk in advanced Type II diabetes patients). This is not something that’s turned up with the other SGLT-2 compounds so far, and is also a mystery.

We do not know what a new drug is going to do, not really, until it’s gone into a large patient population. And that means, most of the time, until it’s made it to the market. Clinical trials are absolutely necessary to clear out the biggest, most noteworthy problems, and will show you the biggest, most noteworthy benefits that can be shown in the time it takes to run the trial. But the longer, more subtle things (or the ones that happen in very low incidence) will only appear once the drug is out there in the real world, being taken by a large number of people under all kinds of conditions.

26 comments on “Two Diabetes Drug Mysteries”

  1. Eric says:

    The increase in amputation risk with canagliflozin was a secondary endpoint added to the CANVAS trial mid-study at the request of the FDA. This wasn’t specifically examined in the earlier EMPA-REG trial because the FDA (and Lilly-Boehringer) weren’t aware of the possible increased risk. It’s too early to know if it’s a class effect, but either way it still supports Derek’s point – we don’t always know what new drugs are going to do.

  2. Anon says:

    Unintended consequences that are good! Just curious if these class of drug can also lower triglycerides?

    1. Anon says:

      SGLT2is tend to mildly increase triglycerides and LDL-C.

      1. Chemcat says:

        Derek’s mention of the increased risk of amputation for canagliflozin users made me think of a different undesirable side effect of SGLT2 inhibitors that received coverage recently. A retrospective assessment found an approximately doubled risk for diabetic ketoacidosis in patients 180 days after initiating SGLT2 inhibitor treatment, as compared to patients starting DPP4 inhibitor treatment. The findings, published in a June 8 correspondence in the New England Journal of Medicine, support a 2015 FDA drug safety communication about this potential side effect for SGLT2 patients.

        Decreasing blood glucose with SGLT2 inhibitors could cause: 1.) a decrease in insulin production, which causes lipid breakdown and ketone body synthesis; and 2.) an increase in glucagon levels, which drives lipolysis and ketone body synthesis. However, if lipolysis is happening as a result of lower insulin levels and higher glucagon levels, why would triglyceride levels be elevated (instead of lowered) in patients using SGLT2 inhibitors?

        Good topic today. We might not be able to predict how a drug is going to act, but boy, is it interesting to attempt an explanation once we see it at work. 🙂

        1. Placebo says:

          Thank you for providing the reference

  3. Mark Thorson says:

    Reducing serum glucose would reduce oxidative stress in the vascular endothelium, which in turn would reduce the activity of uncoupled eNOS relative to coupled eNOS, which would increase generation of nitric oxide at the expense of — get ready for it — peroxynitrite, The One True Cause Of All Disease! Arf, arf!

    1. Zach says:

      Okay Lan- Heyyy, wait a minute.

    2. Lane Simonian says:

      I decided to take a couple of personal snark days off, but this is the right idea, Mark.

      In the brain, high levels of glucose and sodium lead to high levels of myo-inositol and high levels of myo-inositol lead to the formation of amyloid and more importantly to the production of peroxynitrite and the two are intertwined. In fact, high levels of myo-inositol and low levels of the antioxidant glutathione are in tandem the best predictors for the likely onset of Alzheimer’s disease

      Peroxynitrite is of course not the cause of all diseases, but through cysteine oxidation, tryosine nitration, DNA damage, inflammation, and damage to the immune system, peroxynitrite plays a role in some poorly treated or presently incurable diseases.

      Since it is the weekend I will post an abstract rather than a link to the very long article, but for those interested in part or all of the review it is well worth the time.

      “The discovery that mammalian cells have the ability to synthesize the free radical nitric oxide (NO) has stimulated an extraordinary impetus for scientific research in all the fields of biology and medicine. Since its early description as an endothelial-derived relaxing factor, NO has emerged as a fundamental signaling device regulating virtually every critical cellular function, as well as a potent mediator of cellular damage in a wide range of conditions. Recent evidence indicates that most of the cytotoxicity attributed to NO is rather due to peroxynitrite, produced from the diffusion-controlled reaction between NO and another free radical, the superoxide anion. Peroxynitrite interacts with lipids, DNA, and proteins via direct oxidative reactions or via indirect, radical-mediated mechanisms. These reactions trigger cellular responses ranging from subtle modulations of cell signaling to overwhelming oxidative injury, committing cells to necrosis or apoptosis. In vivo, peroxynitrite generation represents a crucial pathogenic mechanism in conditions such as stroke, myocardial infarction, chronic heart failure, diabetes, circulatory shock, chronic inflammatory diseases, cancer, and neurodegenerative disorders. Hence, novel pharmacological strategies aimed at removing peroxynitrite might represent powerful therapeutic tools in the future. Evidence supporting these novel roles of NO and peroxynitrite is presented in detail in this review.”

      The small molecule, phytochemical revolution in medicine is not over after all.

  4. Anonymous says:

    It is possible that the sugar-based structures of these molecules have other biological actions, such as preventing bacteria from binding to epithelial cells in the GI tract, changing the GI microbiome, and leading to changes in entero-endocrine metabolic signaling through altered bacterial metabolism. They may also have an impact on glycosylation at cell surfaces, which could impact the microvasculature and lead to changes in peripheral circulation. Some SGLT inhibitors have been shown to have positive effects on liver fat deposition and NASH, which could account for additional cardiovascular benefit.

  5. dearieme says:

    It’s odd. Every now and again evidence turns up implying quite clearly that the Lipid Hypothesis explanation of CVD is wrong-headed. And yet everyone marvels at what they choose to call a paradox instead of shouting “It’s false – we’ve been entirely wrong for decades now”.

    Contrast Alzheimer’s, where people are prepared to consider the possibility that the amyloid hypothesis is wrong.

    Why the difference? Is it really just about money, careers, and so on?

    1. Some idiot says:

      Very good point… My feeling is the difference is like this:

      In the case of the lipid hypothesis, the hypothesis lead to drugs which help treat the condition. Later on, we found out that the underlying hypothesis was wrong… Hence the paradox…

      With Alzheimer’s, no hypothesis has yet lead to a useful treatment, so the game is different…!

      So in one case, the hypothesis was wrong, but inadvertently lead to useful treatments, whereas in the other case, no treatments have been forthcoming (which tends to suggest that the main hypotheses over the last decade or two need a decent shake, to put it mildly).

      1. SteveM says:

        Right. Root around this Low Carb/High Fat diet site:

        I recently went on a LCHF diet just to lose some weight. (It works.) The presentations at the site done by the professionals related to the profound errors in the existing diabetes/lipid treatment paradigms are very compelling. The Dr. Eric Westman (Duke) videos present an honest clinician’s take on an LCHF approach to obesity management. (Subscription required, but it’s free for a month.)

        And the anecdotal “success stories” from the T2 diabetics who were completely free of the disease symptoms after only weeks on the diet are amazing.

        Data accumulation is gaining to suggest the biggest opportunity for diabetes management is life-style, not drugs. Too bad there’s essentially no money in it.

        BTW, I read recently that 3rd party researchers obtained the raw data from the anti-depressant clinical trials that the Pharma companies held back (i.e., did not publish the report at clinicaltrials dot gov). That comprehensive meta-analysis without the Pharma gaming shows that they don’t work either.

        1. The Other White Meats says:

          You’re referring to the book “The Emperor’s New Drug” by Kirsch. A good read for anyone suffering depression, the family of, or professionals.

          1. tangent says:

            Sounds interesting, where did they publish their analysis?

          2. SteveM says:

            TOWM, thanks much for the reference.

            Tangent, a search on the author Kirsch turns up this informal review of his analysis by him:

            ncbi dot nlm dot nih dot gov/pmc/articles/PMC4172306/

            The institutional and economic inertia that maintain the apparently discredited paradigms for diabetes and depression treatment are ginormous.

          3. tangent says:

            Thanks, Steve.

            You know, a central piece of what he’s saying there is: we asked FDA for the unpublished trials, and combining those with the published trials resulted in no effect or tiny effect. I wonder what a person on the FDA end would say about it. They saw all of this data and approved the drugs; what is the difference of methodology?

        2. Dimapurgoa says:

          Some references pls, for emerging data that suggest diet/lifestyle is the best intervention in T2D.

  6. hse says:

    Empagliflozin, canagliflozin, and dapagliflozin all have remarkably similar structures. Maybe the unknown mechanism driving amputations is a class effect, but is essentially seen only in cana patients due to the 10-20-fold higher doses for that drug (100/300 mg vs 10/25 mg for empa and 5/10 mg for dapa.) If so, one might expect to see a greater incidence of amputations in patients on 300 vs 100 mg cana. Has this been seen or is it too early to say?

    1. milkshaken says:

      canagliflozin has more electron rich ring system than the other two drugs, so there could be hydroxylated metabolites specific to that drug – for example, maybe one of these oxidized thiophenes affects platelet aggregation in a bad way that leads to increased vascular problems underlying diabetic complications resulting in amputations.

      There is a chance it is not class related. And even if it was – do you remember statins having rhabdomyolysis mechanism based side-effect? The incidence was markedly different across the drug class, and it was cerivastatin – high potency lipophilic one from Bayer – that killed lots of patients and had to withdrawn, the other statins seem more than order of magnitude safer if you stay away from grapefruit juice.

      1. Barry says:

        The key to the statin therapeutic indices was tissue distribution. The better tolerated ones have greater exposure in the liver where they’re useful (and correspondingly less exposure in other tissues where toxicity shows up). Looking at these structures, I would expect all of them to be rapidly cleared from the plasma compartment and to have very little exposure in most tissues (although there’s always that hepatic first pass). These highly hydrophilic cmpd’s are likely to flush rapidly out of circulation through the glomeruli.
        The puzzle to me is that they can be efficacious given once a day.

        1. Eric says:

          Canagliflozin is predominantly metabolized in the liver and to a lesser degree in the kidney. It has a 10-13 hour half-life in plasma which makes it perfectly suitable for once per day dosing. You can’t always estimate PK by looking at a structure.

  7. Emjeff says:

    Something that might not be immediately obvious to the med chem readers of this blog, but , in my opinion,the canagliflozin results on amputations considerably weaken the use of glycosylated hemoglobin as a surrogate end-point for diabetes. Decreasing the percent of glycosylated hemoglobin is supposed to predict the lowering of the devastating effects of diabetes, such as amputations, blindness, etc. The fact that amputations increased in this study in the face of presumably lowered Hgba1c means that, once again, we are carrying increased risk into Phase 3 during the development of drugs to treat diabetes. If I were in this development space, I’d be very concerned.

  8. Barry says:

    I’m baffled, looking at the structures that these can be efficacious at once a day. The Na+/glucose transporter is expressed on the basolateral surface of the nephron, accessible only once the drug has passed to the urine side. Are these rapidly cleared from the circulation, but then stay bound for hours in the porter?

  9. SteveDoc22 says:

    Is it conceivable (and I am not positing this, just asking) that the statistical analysis being used to “control” for multiple risk factors is inherently flawed?

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.