Dr. Juneja did a spectacular job on the pathogenesis and progression of diabetes. That was really a great talk, I thought.

But I wanted to underline a couple of points that will tie into some of the other things I’m going to talk about. And that is, if there is a natural history of progression in type 2 diabetes, and this progression has to do with failure of the beta cells, as I think it does, then, you know, are we just building sand castles near the shore that are always going to get knocked down? Are we doomed to failure in treating diabetes? Or is there some possibility that there may be things in the future that we could do this. And I’ll talk about drugs that are available now. I’ll make some mention of work that’s ongoing that I think makes this a hopeful area.

When you talk about beta cell dysfunction in diabetes, there’s this huge body of literature, mostly cross-sectional studies. We got 10 patients from the clinic, and we got 10 people from the cafeteria that weren’t diabetic that looked kind of like the diabetics. We did X study on them, and we measured beta cell function. And these were studies largely done in the seventies and eighties to really try and define the lesion that we see in type 2 diabetes.

What do we see? Well, the beta cell in a diabetic person is insensitive to glucose. Normal people have a very, very responsive beta cell. And they may vary even amongst normal people. But small changes in circulating glucose activate the beta cell to synthesize insulin, to secrete insulin, and to do this in a progressive manner. As the glucose goes up, the beta cell fires progressively more. In diabetes, that’s lost. You don’t see the first-phase insulin secretion.

Again, first-phase insulin secretion is bandied about a lot. It’s an artificial term. We see it in the clinical research center only when you give a bolus of glucose. And then you can see a nice, rapid spike of insulin in 10 minutes. It doesn’t happen in everyday life. When you eat your lunch, you’re not going to have a rapid pulse of insulin come out.

But it’s a very good marker of the beta cell being sensitive to changes of glucose. And it’s a very good marker of beta cell health. We know that because people who are developing diabetes or people that have type 2 diabetes lose this. We also know, if you do a ramp of glucose, progressively increase the blood sugar over time, people with diabetes have a blunted response. Their beta cell doesn’t respond to insulin. This is sort of the hallmark functional defect. It doesn’t require more or less beta cells. These beta cells just don’t perform normally.

Diabetic patients have a decreased maximal capacity to secrete insulin. These tests are a little crazy. We’ve done some ourselves. You raise the blood sugar to four or five hundred. And then you give arginine on top of it. You throw everything but the kitchen sink into the bloodstream and see how much insulin can come out.

This was noted a long time ago to uncover a real deficit in diabetic insulin secretion. And this may be related not only to dysfunctional beta cells, but also to decreased beta cell mass. People with diabetes have impaired pulsatile secretion, oscillatory insulin secretion. And although this was for a long time thought to be a parlor trick, a laboratory demonstration—oh, isn’t it neat that beta cell function is pulsatile? — in fact, there’s good data now that suggests that up-and-down secretion of hormones probably is tied in with the sensitivity of body tissues to those hormones, and that losing normal pulsatility is a bad thing in diabetes.

And then finally, I’ll show you some data that normal augmentation of insulin secretion after a meal, what we call the incretin effect, is reduced in diabetes. This is just one demonstration. So, this is people in the CRC getting a bolus of glucose right here, about 25 grams, essentially the contents of a glass of orange juice, but pushed immediately as a bolus of insulin. And in healthy people, you see this rapid spike of insulin secretion that peaks within 10 minutes. And this is called first-phase insulin secretion. Again, an artificial construct, but a marker of beta cell health. These are lean people; here’s their first phase. These are heavy people; here’s their first phase. First phase in nondiabetic people tends to be proportional to insulin sensitivity. If you’re heavy and insulin resistant, you compensate by secreting more insulin.

These are diabetic patients down here. They have no first-phase insulin secretion. Even the heavy ones who have some later insulin secretion, as the glucose is higher, some second phase, this early spike in insulin secretion is gone. So, a demonstration of a functional defect in the diabetic beta cell.

This is the incretin effect. It’s a long-known physiologic phenomenon. It states the following: If you take healthy people and you give them an oral glucose tolerance test, and then you bring them back the next week and give them IV glucose to precisely match the plasma glucose during the OGT, you get a big difference in insulin secretion. Oral glucose or oral meals, taking carbohydrate through the gut, causes about a 50, or sometimes 70 percent increase in insulin secretion at a given insulin level. This is due to secretion of gut hormones and activation of neural stimulus to the beta cell.

But it’s very important, I think, and why healthy people can control their blood sugar so tightly around a meal. I mean, face it, unless you’ve had an oral glucose tolerance test, you may not have seen a blood sugar much over 130 in your life. Because even an hour, 2 hours after Thanksgiving dinner, most people that don’t have diabetes are controlling their blood sugar in a very tight range. Glucose homeostasis is quite robust. We forget about that, because we look at glucose homeostasis in diabetes so much.

So, here’s the incretin effect in diabetes. These patients were hyperglycemic at baseline. And of course, their blood sugars went off the curve after they took the OGT. Nonetheless, you could match that with IV glucose. But here you don’t see so much augmentation. In fact, on average, it’s about a 7 percent change in insulin secretion in this diabetic population. Another demonstration of a functional beta cell defect in people with diabetes.

I think up until the nineties, until the UKPDS, there was this notion that people with type 2 diabetes had islets; they had beta cells; they secreted insulin. But they had all these functional deficits so that the beta cell output was never quite sufficient. And I think it took longitudinal studies to actually get a handle on the fact that, yeah, they do have functional deficits, but it’s not static. It gets worse. And that’s something that I emphasize to patients when I tell them right now.

And again, to get back to the preventive and staving-off point. I try to emphasize to people that unless they really engage with their disease, their diabetes isn’t going to be like it is today. If they follow the normal course, it will get worse over time. And this is one of the best studies, I think, to show this. This is from Weyer and Pratley in Arizona studying the Pima Indians, who a lot of you have heard of.

The Pima Indians are a fairly genetically discrete Indian tribe that have the highest rates of ob
esity and diabetes in the world. They probably also have the highest participation rates in clinical research in diabetes in the world. An enormous amount of what we know about diabetes comes from the Pimas.

So, this is a longitudinal study, where they took a group of middle-aged Pima Indians who had normal glucose tolerance. And it seems like darned near everybody who lives in that area has a glucose clamp and a beta cell function test, kind of annually, maybe on their birthday. I don’t know. They seem to be more regular than people donating blood. What you can then see here is, they then followed these patients over 5 years.

And they found that although everybody had normal glucose tolerance to start with, most of them were obese. The mean BMI was over 30. And most of them were insulin resistant. This is data from a clamp. I can tell you this is moderately insulin resistant. If they then follow them over time and saw who got impaired glucose tolerance and who got diabetes and restratified them, they found that the diabetic patients were more insulin resistant.

More importantly, what they see is this drop-off in the beta cell response over time. That is, if you’re going to get diabetes, the thing that’s most dramatic is your ability to continue to secrete appropriate amounts of insulin. And again, when you look at these data down here, these are people who over time maintained normal glucose tolerance. In general, the people got heavier, and they got more insulin resistant. But if you’re not prone to diabetes, you compensate. You, actually, your beta cell secretes more insulin, and you keep your blood sugar normal, if you can compensate. If your beta cell function declines over time, you fall off the curve and become diabetic.

And again, you’ve seen these data before. I think they’re terribly important, because I think this is where the game shifts in terms of our understanding of diabetes. When you follow the patients in the UKPDS and you see, well, these guys drew the short straw, and they just got diet and exercise. And that was 1980s prescriptions of diet and exercise. And of course we would expect that their blood sugars would go higher over time.

But here’s a patient on sulfonylurea and insulin, and by god, the same thing happens to them after an initial response, the A1c goes up. We see this in clinic every day, right? So, Mr. Jones gets referred from primary care because he’s got new-onset diabetes, education, et cetera. We intervene. We put him on a drug. See him back in a year. Is he doing okay? And then pretty soon, we have to put him on another drug. He does okay for while, and then another drug, and then eventually he’s on three or four shots of insulin a day. And that’s the progressive course of type 2 diabetes.

Again, what’s important is that for the most part, people who are progressing over time are insulin resistant to begin with. It’s not that they’re going from BMIs of 30 to 50 to cause this progression. For the most part, they’re overweight to begin with, and they’re insulin resistant to begin with. And that doesn’t change a heck of a lot. What changes is their beta cell function, their ability to secrete adequate insulin for their degree of insulin resistance. And this doesn’t seem to be affected by any of our sort of conventional drugs, sulfonylurea, metformin, or insulin. If you have people treated conventionally for diabetes, and again, by conventionally, I mean the usual drugs and the usual A1cs, kind of 7.0 to 9.0, there seems to be progressive beta cell failure.

So, besides the longitudinal data, it turns out that there’s a lot of pathologic data that supports the beta cell and a decrease in the number of beta cells in the cause of diabetes. So, this is the famous Butler study using a big autopsy sample from Mayo Clinic and looking at people that were obese but nondiabetic, obese with impaired fasting glucose, and obese with type 2 diabetes. Simply sectioning the pancreas, staining the islets for insulin, and then measuring the area: how big were the islets, how many beta cells were there.

Well, if this is the standard, by the time you have a person with impaired fasting glucose, they’re down to about 50 percent of the normal complement of beta cells. By the time they have type 2 diabetes, they’ve fallen off even further. And you can see in lean people, again, the number of beta cells in a nondiabetic is less than an obese person. Again, the current thinking is, if you’re obese and insulin resistant, your beta cells hypertrophy. That for three quarters of the population, or 80 percent of the population, that’s the normal response. This keeps your blood sugar in the normal range, even with obesity and insulin resistant.

Normal people have smaller beta cells. They don’t need as many. They’re more sensitive. And if they have diabetes, they have less beta cells.

You know, the interesting thing is, this paper made a big splash. You can go back and see data like this published in the fifties. So, this observation, oftentimes the stuff that the pathologists do doesn’t get out of the basement and see the light of day. But this is a consistent finding that’s been noted in recent studies. But you can find it going back decades in the literature.

So, it does seem to be that type 2 diabetes is similar to type 1 diabetes, and there have been a number of papers talking about unifying hypotheses in that absolute numbers of beta cells are diminished in this disease.

This is just to make a point about not only are the islets smaller in diabetes, here’s a nice, normal islet, you can see it sort of surrounded by a capsule here, the nuclei of these densely packed mostly beta cells, 65, 70 percent of the cells in an islet are beta cells. This is a type 2 islet. Here again, this capsule. And you can see nuclei in here, but you see a lot of this schmutz. It looks like somebody has taken an eraser and kind of erased through this. This is islet amyloid. And again, if you go back in the literature, you’ll find pathologists talking about this back to the nineteenth century.

In type 2 diabetes, there is this weird amyloidosis that happens to the islet, so that not only are the beta cells dysfunctional, and not only are the islets smaller, but a lot of the islet mass is taken up by this amyloid stuff. And in cultured studies, the amyloid actually promotes beta cell death.

So, this is one leading hypothesis as to why the beta cells drop out. Interestingly, it’s been hard to show that there is a nice progression of amyloid in nondiabetics, and then slightly glucose intolerant, and very glucose intolerant. It would be nice to have that kind of dataset. We don’t get to look at a lot of islets, unfortunately. It’s hard to talk people into a biopsy.

So, what are the proposed mechanisms of beta cell failure? And Rattan touched on this a little bit in his talk. And I’ve just mentioned amyloidosis as one. I think most of us think that amyloidosis is what happens after you get diabetes, and the best data suggest that up until the point that you’re pretty diabetic, you don’t seem to have a lot of amyloid in your islet.

Likewise, I think of glucotoxicity as, again, a late-stage environmental toxin to the beta cell. Most of us aren’t that hyperglycemic during normal life, and people, even people with impaired glucose tolerance with an oral glucose load, it’s hard to show huge differences in their blood sugar eating normal meals. I think glucotoxicity certainly happens when the blood sugar starts to get up near 200. I think where that cutoff is, is hard to say. But to me, this seems like a late-stage thing.

Lipotoxicity is an interesting and sort of many-splendored mechanism. You could put lipids, fatty acids in with cultured beta cells. And if you give them a lot of palmitate, saturated fat, you can show that they start to be apoptotic, that they don’t function as well, that they don’t grow as well. You can put them in other fatty acids, like oleic acid, a monounsaturated fat, and it seems to be not so bad to them. So, there’s this notion that some fatty acids are toxic, some are not.

Fatty acids tend to be a little higher, we know a lot less about what the normal fluctuations are. But I think people have extended these cultured data to say, well, hell, lipotoxicity and having high amounts of lipid could be a problem. I think a lot of this feeds into this notion of what’s called reactive oxygen damage. And it’s a broad hypothesis that Brownlee and others have put forth to talk about tissue damage in response to excess. That’s the way I interpret it.

If you flood the system, the body, with calories, fuels, something’s got to happen to them. You know, to a certain extent, we’re all able to dispose of glucose, store lipids as fat, put them into storage forms where they’re safe and nontoxic. Fats should be in adipose cells. Glucose should be in muscle and liver. And the other cells should not be big storage depots.

I think what happens in a lot of people is, when you get an overabundance of the fuel, all the cells are seeing more fuel. There’s only so much you can pound through and oxidize fully. In high levels of respiration, putting things through the oxidative phosphorylation, you generate reactive oxygen species, toxic free radicals, and things that cause cell damage. And this is a big area of research in all aspects of chronic disease. But there’s a notion that in the beta cell, having high levels of glucose or lipids, other fuels, can lead to damage through reactive oxygen.

Finally, I list autoimmunity here. It turns out that people with type 2 diabetes have twice as high prevalence of anti-islet antibodies as nondiabetic people. Now, it’s a lot less than people with type 1, but there may be some level of autoimmunity causing things.

You’ll notice, one thing I didn’t put up here is beta cell exhaustion, because that’s just a term and a concept I don’t like. Endocrine cells don’t get exhausted. Most endocrine cells put under a stress, released from negative feedback, challenged by high levels of whatever they’re regulating hypertrophy. The normal beta cell response to insulin resistance is to hypertrophy. Normal response to a pituitary cell that makes ACTH, when you take away cortisol, it hypertrophies. Those are what normal beta cells do.

I think there may be a phenomenon that looks like exhaustion, but I think it has as its underpinning a defect already. So, you could say that somebody who’s prone to beta cell exhaustion has a defect on one of these key determinants of beta cell mass. They either have beta cells that don’t replicate normally, divide, double, hypertrophy. They have some problem with neogenesis, the development of new endocrine cells in the pancreas from stem cells in the pancreatic ducts. Or they have increased rates of apoptosis, cell death.

It’s these things together that determine anybody’s beta cell mass. And they play out very slowly. In the immediate prenatal time period, there’s a huge burst of replication and neogenesis, so that everybody has more beta cells just around the time of birth than they’ll ever have in the rest of their life. In the early neonatal period, then, there’s high rates of apoptosis, as you sort of scale back the cells to what you need.

Throughout life, then, these things sort of go back and forth. And most people stay in balance. They make enough. As cells get old and senesce and die through apoptosis, they’re replaced by new cells. When I was in medical school, they told us you got all the beta cells you’re going to get at birth. Hang on to them, because when the die off, you’re going to have diabetes. Now we know that’s not true. They do replicate. They turn over slowly in humans, over months to years. But it’s this balance that keeps the beta cell mass normal.

So, rather than think of it as exhaustion, my sense is that the people who are prone to diabetes, people who come from families where the genetics suggest there’s a lot of diabetes, people who have gestational diabetes, the ones that get impaired glucose tolerance when they get middle-aged spread, have some defect in one of these processes. And it’s these things that give the appearance, the inability to compensate, the appearance of beta cell exhaustion.

So that again, when you look at this natural history here, and you see this early part, everybody always thinks, ah, yes, look at this. These beta cells really grow, they hypertrophy. This is an inadequate response, given that the blood sugar is going up, right? A nondiabetic person wouldn’t tolerate the blood sugar crossing this line. They would get whatever beta cell mass is needed to keep this normal. And the fact that there’s some hypertrophy is sort of the last gasp. It’s the hard few steps of the marathon runner who’s not going to make the finish line.

And then, my sense is that there are features here, whether they’re genetic or due to reactive oxygen species or whatever, inability to compensate, that allow you to be diabetic. And then once you get diabetic, things like glucotoxicity, amyloidosis ensue, cause this progression that we dread so much.

So, this is a real problem. This is the crux of the matter right here is that we start firing away with drugs up here and for the most part, it becomes inexorable, and the waves gradually wash down our sand castle and the A1c gradually goes up. So that when you talk about beta cell dysfunction in type 2 diabetes, there’s clearly a functional problem. There are measurements of beta cell performance that don’t work. There’s also clearly a decrease in beta cell mass, whether you measure it by maximal insulin secretion or at autopsy by number of islets.

And I think type 2 diabetes is a disease where both are involved. I think mostly people that are going to get diabetes probably have some level of dysfunctional beta cells most of the time, or at least a predilection for that. They have problems with the compensatory mechanisms of islet cell growth that usually keep us from being diabetic. And it’s this composite that gives diabetes over time. And whether these things are mechanistically linked, I think it’s a big question. This is an ongoing debate right now.

So, do we have any hope? Can we make beta cells grow? Are there tonics? Well, the developmental biologists, the cell biologists are all over this, and they’re doing lots of very interesting studies, understanding the signaling pathways that are involved here, and have identified in preclinical studies a huge array of factors that will make cultured beta cells or islets grow, or at least start down the replication pathway.

The key ones for our discussion this morning are the two incretins, glucagon-like peptide-1, or GLP-1, and GIP, glucose-dependent insulinotropic peptide. Those are the basis of therapeutics; TZDs, as well. I’ll show you some interesting ones as well. Growth hormone, that’s not surprising that growth hormone would stimulate islets to grow. It stimulates a lot of things to grow. PTHrP, the cause of hypercalcemia in malignancy, turns out to be a pretty good beta cell stimulus. Prolactin is thought to play some role in the normal hypertrophy of islets in women during pregnancy.

I’m trying to think of gastrins up here, the hormone that we know of as stimulating gastric acid production is a good beta cell tonic. And the islet-transplant people are really into that when they take out the islets before transplantation, they oftentimes try to juice them with gastrin. But the point is that there’s a lot of known molecules that signal in the beta cell that seem to engage the pathways that cause growth or that prevent apoptosis.

Let’s talk a little bit about GLP receptor agonists. Dr. Marrero sort of provided an introduction to this by talking about exenatide and liraglutide, which are drugs that are available now; liraglutide not in the U.S., but I suspect will be here soon. These are both injectable molecules. They’re peptide hormones that bind to the GLP-1 receptor. GLP-1 is a native, naturally occurring hormone in the gut.

When we eat, GLP-1 is secreted. It does a bunch of things to help control blood sugar. The incretin effect, that I described earlier, is mediated about 50 percent by GLP-1. So, these drugs work by binding the GLP-1 receptor, which is a specific peptide hormone receptor that only binds to GLP-1 in these agonists. It’s expressed in the beta cell for sure. And when GLP-1 binds to this receptor on the beta cell, it drives signaling increases in cyclic AMP, et cetera, that result in insulin secretion.

The GLP-1 receptor is also expressed in other tissues, the heart, the brain, the gut. And there’s very interesting biology there that we won’t touch on today. Again, you’ve seen this picture. Just to make the point that GLP receptor agonists can cause an acute effect on the beta cell. Here they restitute first-phase insulin secretion in diabetic patients who don’t have it. You can do all these nice tricks in healthy humans, too, giving exenatide, giving GLP-1. Anything that activates the GLP-1 receptor will cause an increase in insulin secretion in the short term.

So, as acute secretagogues, these analogues are very good. And for the most part, that has been the message as to how they do their business.

Exenatide also decreases glucagon; this is a glucagon response to a meal in a diabetic subject, it’s high. Exenatide has an acute effect to decrease glucagon. GLP-1 will do the same. Liraglutide will do the same. Again, these sort of instantaneous, give a dose, see what happens, part and parcel for how we think these drugs work.

And of course, exenatide reduces A1c. We all know that. That’s why a lot of us in the room use it. Actually, for all the GLP-1 research that I’ve done and all the exenatide I’ve given to people in the research center or to animals, I don’t use it, because my hospital doesn’t have it on formulary. But my colleagues use it a lot, and so, I’ve been interested in the progression of the story.

Regardless, whether you combined exenatide with metformin, sulfonylurea, or the both, you get a nice reduction in A1c. And again, the message has been, this works because it’s an acute, insulin-secretagogue and it suppresses glucagon, and that makes the blood sugar better. This just to show where these compounds are going. This is a once-a-week, long-acting exenatide, and you can give it at a half dose or a full dose. You get this big—look at this in this—1.7 drop in A1c.

This is a product of the Amylin company that makes Byetta. We’re all sort of waiting for them to release the pivotal trials on this drug. But there’s a lot of promise. And I can tell you that in development, these once-a-week, once-every-2-weeks, once-a-month, annually with your checkup, injections of GLP agonists, are a big area in drug development right now. And one of the reasons is because they’re injectable, it would be nice to minimize the injection. But again, effective compounds for lowering blood sugar in the short term.

Again, this is a 15-week trial. Liraglutide, which is a modified GLP-1 that binds to albumin and is a once-a-day as opposed to a twice-a-day, like Byetta, does the same thing: lowers A1c.

So, a question is, are they just insulin secretagogues, sulfonylureas, with a twist, that they don’t cause hypoglycemia? If you look at beta cells, and you go through the research in the cell literature, and if you’re a clinical investigator like me and not into the intricacies of where all these arrows go, I like to think of GLP-1 signaling as two parts. There’s an acute part that’s cyclic AMP and things downstream of cyclic AMP, have a role on membrane depolarization in the beta cell and exocytosis. And that’s why we see things like a restitution of first-phase insulin secretion, et cetera, that there’s acute effects that augment glucose-stimulated insulin secretion.

But if you look at a lot of the data that’s emerging in beta cells and other cultured cells, you’ll see that there’s chronic effects. The GLP-1 drives gene transcription. Things that take hours and days to manifest. The GLP-1 blocks apoptosis. Cells that are trying to decide, you know, I’ve been here a long time, this has been hard work, I think I’m just going to go into apoptosis and get outta here.

If you give them GLP-1, sometimes you can talk them out of jumping. All right, come in off the ledge. We won’t. And then GLP-1 will also stimulate beta cell growth and make beta cells divide in culture and replicate. So, these chronic effects are very, very appealing. And it’s very nice to think, wow, wouldn’t it be great if this guy that I’m giving exenatide to because he wanted to lose weight and it’s helping his diabetes, wouldn’t it be great if I was regenerating beta cells?

And you know, we’re at that point right now in the story. Wouldn’t it be great? I don’t know; it would be great. But I don’t know if it’s happening.

But again, a huge amount of literature, of studies in preclinical models has shown that you can increase key gene transcription in beta cells in rats and cultured cells. Insulin, GLUT-2, glucokinase. This would be on everybody’s top list of things you’d really want to drive in a beta cell. That you can stimulate growth by promoting proliferation in blocking apoptosis, that’s really a great thing. And that this thing actually is not only something you take out of the culture dish and apply in rodents, and that’s really cool, because rodents are at least a step closer to humans than cultured cells.

So, this is a nice study that just shows the broad use of GLP receptor agonists on parameters of beta cell mass. These are mice. They’re db/db mice. They have a leptin receptor mutation. And they get diabetes. They get smaller islets as part of this. If you treat them with the placebo or the saline control, this is the islet size. If you treat them with exenatide, Byetta, you double the islet mass. And this is just a week’s treatment.

And again, rodents are different, obviously, than humans. The more I work with rodents, the more parallels I see. They have higher metabolic rates, higher cell turnover. Beta cells in a rodent turn over every 50 or 60 days. And so, these kinds of changes are always more dramatic and they look great. But you can see here that the proliferative response is higher. Apoptosis, the number of cells that are dying off, is about cut in half. And so that in a very short period of time, the beta cell mass, the thing you look at under the microscope when you slice the pancreas and look at the islets, has gone up substantially.

This kind of thing has been replicated over and over again in all sorts of models, whether they’re genetic models with diabetes or given toxins to rats to kill their beta cells or cutting out part of their pancreas. Things that tickle the GLP-1 receptor seem to compensate for this very nicely. And then the picture that’s worth a thousand words. These are the control animals given nothing, and these are the humongoid islets of the animals treated with exenatide for a while. Again, this great sales pitch for Amylin. Wouldn’t you love this to be your islet? And there’s the doubling in size.

We don’t know about human islets in humans. We know about human islets taken out of humans, say, and cultured as part of either research for islet transplantation or before islet transplantation. And this is what happens to a normal human islet in culture. There’re the nice round cells in the bottom of the dish, and it starts to look a little frazzled by day three, and it’s started to come apart, become completely unwound by day five. If you happen to have some GLP-1 in the dish, it holds up much better over time. Again, there seem to be things that promote the beta cell health by running through the GLP-1 receptor.

So, again, just to reiterate. We know that people with diabetes fail, that their A1c goes up over time regardless of what we treat them with. Do we know this for Byetta yet? No, we don’t. We have data like this, where this would be the clinical trial against the placebo control, and we’d see about a one point drop in A1c.

And then this is the open label that goes out to 104 weeks. And that looks smashing. That looks great. That looks nice and stable.

But I would make two points. One is, these are the winners, right? These are the people that are getting a great response to exenatide. Lots of them have lost lots of weight, and that’s why they’re sticking with it. And you’ve got to watch open label trials. Dr. Marrero made that point, as well. I actually am very dubious of looking at long-term data like this that’s not placebo controlled or controlled by an active comparator, because I think you can really be misled by only looking at the winners.

The other thing is, relatively short duration. This is 104 weeks; not 104 years. Look when things are starting to go bad here. You know, it’s in that sort of two- to four-year period where you could really see this. So, the duration of these trials is simply not long enough to make a conclusion about the effect of a GLP receptor agonist on the progression of diabetes and, presumably, the health of the beta cell.

Let’s talk about DPP-IV inhibitors, because they’re another compound that are related. I see them as totally different drugs than exenatide, and not an option.

They’re, in my view, kind of like metformin and alpha-glucosidase inhibitors, that sort of difference. But they are effective drugs.

They do raise GLP-1 levels. They raise GIP levels, the other incretin. Both of those in cultured systems stimulate beta cell growth, so these might be nice drugs.

Sitagliptin is what you know as Januvia. Vildagliptin is a similar compound that’s now available in Europe. Last I heard, there were 15 or 17 of these under development. You can’t really be a drug company if you don’t have a DPP-IV under development. You know, it’s one of those things. It’s kind of like teenagers, you gotta have a funny hat that you wear sideways. But there’s a lot of these coming, and it will be interesting to see what happens.

Again, DPP-IV is an enzyme that exists throughout the body inside of capillaries and floating in the blood. And DPP-IV takes GLP-1 under GIP, the active incretins, knocks off a couple of amino acids, and makes them inactive by and large. For GLP-1, this is really fast. So, intact GLP-1 has a very short half-life in the circulation, only a minute.

Now, exenatide and liraglutide aren’t susceptible to DPP-IV. And that’s why they’re useful drugs. They don’t get degraded in a minute. They have half-lives that are hours. But the native hormones are metabolized quickly, if you block this enzyme, as you can with a DPP-IV inhibitor, they’re all tremendously effective at this. They’ll knock down DPP-IV by 98 percent for 12 hours. They’re very good at that.

Then the active hormones exist in the circulation in higher concentrations. They’re protected. These sort of have increased the dose. And this shows just that. If you take people and you measure GLP-1, active GLP-1 after an OGT, or with a DPP-IV inhibitor, you can see that you double or triple the circulating levels of GLP-1. And presumably, that’s a good thing and accounts for these kinds of effects, the same relative magnitude, in my view, of A1c difference that you get with other drugs, you know, about 0.6 or 0.8 in a well-controlled population and up to 1.4 or 5 in a poorly controlled population. But effective glucose lowering by these drugs. And again, nobody’s proven what it is about DPP-IV inhibitors that does this. Most of us think that GLP-1 plays a large role in this because we can measure it.

But it turns out that DPP-IV metabolizes about 25 peptides. And so, those are things, a lot of them we don’t measure very well. And it may be that any one of these things, because it could be that there’s a multiple effect of DPP-IV inhibitors.

Do these things do anything to beta cell health? Besides being acute insulin secretagogues and inhibitors of glucagon, do the DPP-IV inhibitors do anything to beta cells?

Well, these are animals that have a knock out of the DPP-IV. That’s the ultimate DPP-IV inhibition. And unfortunately, we can’t do that in a clinic. But in a mouse, you can. You can knock out the DPP-IV gene. This animal doesn’t metabolize GLP-1 normally. They have very good glucose tolerance, which is kind of astonishing, because in an animal, it’s hard to make you better than normal. But these animals are better than normal in terms of glucose tolerance.

Also, if you give them streptozotocin, which is a classic beta cell toxin, that’s how we cause experimental diabetes, what you can see is, wild-type animals given a dose of strep, their blood sugars go from fastings in the low 100s to mid-200s in a few weeks. If they don’t have any DPP-IV, and presumably, if all the products the DPP-IV metabolizes are in higher amounts, they actually do better. They get hyperglycemic and then they regain here. Their insulin secretion is retained.

And, although it doesn’t show up well here, their beta cells look healthier. You can see here, and even on this poor projection, you can see that the beta cells are really dampened in the streptozotocin group. And here you can see a lot of red, which are glucagon cells, and not much green, which is insulin. But if you don’t have a DPP-IV, actually, your beta cells hang in there pretty good.

And this kind of thing can be shown, then, with a DPP-IV inhibitor. These are normal animals. After you give them strep, you can see you blasted out a lot of the beta cells. You just got a few hangers on here. If you treated them with DPP-IV before they got strep, you get more hangers on. Even if you give them the DPP-IV after, it looks better than the strep alone. And then here is just the effect in small islets where the devastation is a little more thorough. Early treatment with a DPP-IV inhibitor seems to be positive.

What I can tell you about the DPP-IV inhibitors in clinical trials is the data is not nearly as strong as it is for exenatide and GLP-1 agonists. DPP-IV inhibitors don’t raise the GLP-1 levels nearly as high as you can get with exenatide. And so, the data on these beta cell effects, growth, death, beta cell mass, are a little more muted. But they seem to be present. Again, the question is, could we see this play out over time? Do we have information that suggests that chronic treatment with a DPP-IV inhibitor alters the natural history of type 2diabetes? And by that, could we infer that some of this effect is due to promotion of beta cell health? But we don’t have that data.

Again, this beautiful, flat curve here from Bo Ahrén stretching over 52 weeks that they’ve now extended to 100-some weeks is great. But the clinical trial ended here. And these are all the winners. These are all the open-label subjects. And the people here on metformin who were failing, I don’t know what they gave these guys to hang in there and keep taking only metformin while their A1cs were going up. Maybe free parking and lunch. I don’t know. This is not an accurate comparison, because these are people who chose to stay in the study.

So, what about TZDs? And again, Dr. Marrero talked a lot about TZDs. And this is a very interesting area, partly because a lot of this research is maturing for TZDs, which have been around a lot longer than GLP agonists and DPP-IV inhibitors. And we have a little more data on which to base things. And then also because there’s been this push-back in the last year with some caveats on TZD use.

Nonetheless, studies in cultured cells support a protective effect of TZDs on beta-cell function. Unlike exenatide or GLP-1, you don’t see a robust growth effect, but you do see protection when you throw on toxins to cultured cells. They don’t seem to die; they seem to be heartier, more robust to external stressors.

In rat models treated with TZD, they also seemed to have preserved islet cell mass. And I’ll show you some of this data. It has some implications as to how beta cells go bad. And then there’s a little bit of data in humans of short duration, the TRIPOD and PIPOD being the most exemplary. But I’ll show you another twist on that in a second.

So, the notion of lipotoxicity, that high levels of circulating lipids, particularly lipids getting into cells that aren’t adipose cells, where they shouldn’t be, is derived from a huge body of work in what are called Zucker rats. Zucker rats have a defective leptin receptor mutation. So, their leptin receptor is abnormal, doesn’t receive a signal from leptin. These animals, then, are very fat, and over time develop beta cell failure.

So, here’s the Zucker at 6 weeks when the islet still looks pretty good. But by 10 weeks, and certainly 16 weeks, the islet is just a mess. It sort of atrophied. If you stained this for lipid, there would be a tremendous amount of triglyceride. These animals get what’s called “extra adipose triglyceride deposition” in their liver, their heart, their beta cells. They’re filled with fat. And this seems to cause them to apoptose at a faster rate and for the islet to sort of be damaged and die away.

Well, this is the same animal now treated with a TZD, and this lipotoxicity seems to be greatly mitigated in this term. And in fact, if we think about what TZDs really do, I mean, we think about them as insulin sensitizers, but what PPAR-gamma really does is directs adipose tissue differentiation. PPARs put fat in the places it should be, in adipose cells, essentially redirects triglyceride storage from tissues, like heart, liver, and beta cells, where it shouldn’t be, to subcutaneous fat where it should be. So that at least in this model, TZDs seem to be very effective in one genetic form of lipotoxicity, which is interesting.

The other thing is that pioglitazone, another TZD, can be used to mitigate the effects of interleukin-1, which is again, a cytokine that’s used to damage beta cells in culture. And you can see here is the control with glucose at 100, and then glucose at 100, pioglitazone does nothing to the control tissue. If you look at glucose plus IL-1beta, you can see that the number of cells that are dying that show markers of apoptosis goes through the roof.

But if you then put the IL-1beta and pioglitazone in at the same time, the pioglitazone really mitigates the apoptotic effects of IL-1beta; again, glucotoxicity, high amounts of blood sugar, or culture sugar will cause more cells to go into apoptosis, and the pioglitazone seems to mitigate that.

So, here we’ve got three of the classic mechanisms of beta cell damage–lipotoxicity, glucose toxicity, and inflammation–and the cells are all protected by a TZD.

The interesting thing about culture information; although, again, stuff in a dish is way different from patients in a clinic, but it does tell you that the compound seems to interact with the key cell. Dr. Marrero talked about that. We’re not entirely sure if the benefits we see in patients are due to improved insulin sensitivity, decreased demand on the beta cell, and not asking a diabetic beta cell to do something it can’t do, which is compensate. Or whether these drugs have direct effects. But these kinds of studies suggest that they have direct effects on beta cells.

So, here’s a nice human study that gets at this question. This is Ovalle and Bell at Birmingham, and they took a small group, about 18 patients, with poorly controlled type 2 diabetes. This is their A1c around nine at baseline. They randomized them to a shot a day of long-acting insulin or to rosiglitazone. After a short period, three months, of treatment, both groups had a significant drop in their A1c; although, not to target. But they did beta cell function tests before and after the treatment. So, they gave them a dose of IV glucose here and here, and measured how much insulin came out.

And this just shows the difference. And what it shows is that that people who got insulin, their beta cell function was crummy at the beginning and it was equally crummy at the end. The guys who got rosiglitazone just for three months all of a sudden started to secrete insulin, despite the glucose level being about the same. Again, these kind of data suggesting that TZDs may have some beneficial effect on the beta cell over time.

You know, the longitudinal studies that would show stability, durability, of the TZD effect are just starting to be released. This was shown earlier. I think one of the most important papers to come out last year, this is the ADOPT trial that took drug-naïve type 2 diabetic patients who needed to be started on therapy, that would be most new diabetic patients, and randomized them to get a sulfonylurea, metformin, or rosiglitazone. And then followed them over time, you see here, up to five years. Really nice longitudinal study.

And what they used as several parameters to define beta cell failure, either a worsening of A1c or a fasting glucose that was over 175. And you can just see the number of patients failing here. A couple of very interesting points. One, a lot of people have been knocking sulfonylureas for a long time as effective in the short term but potentially harmful in the long term. For glyburide, at least, this seems to be borne out in this trial.

Metformin and rosiglitazone do much better. Again, interestingly, these aren’t the nice, flat lines that I showed you in the open-label trials with DPP-IVs and exenatide. Some people are failing, but they’re failing at a slower rate.

The point is that these kinds of trials, what I call “durability trials,” how effective is the drug before you have to add another drug, is the best way we have to look at natural history of diabetes and make inferences about beta cell function. And these are the kind of trials that you’d need to see with a GLP-1 agonist or with a DPP-IV inhibitor before you could start to make comments that all that stuff in the culture dish and in the rodents plays out in humans. I think those studies are coming; I think we shouldn’t hold our breath. It takes a long time, and these are very hard and expensive to do.

I mean, ultimately, the best thing would be able to image the islets, right? Wouldn’t it be great if you could send your patient down, CT of the abdomen, look for bad stuff, oh, yeah, and quantitate the islet mass? That would be terrific.

That’s an active area of research. NIH has put a lot of money into it. A lot of very smart people are looking at that. Imaging gets better all the time. I feel like most of endocrinology now is chasing down extraneous nodules that were found on somebody’s high-sensitivity imaging.

The imaging of the islet is still about two or three orders of magnitude away from being sensitive enough to make useful measurements in humans. You can do it in rodents that you’ve engineered to have their islets glow, and you put the islets in the liver. You can get them on a CT scan. But the days when we can measure islets in humans with the modalities we have now are a ways off. It’s not out the question; it’s like everything. It’s a bioengineering thing that’s important, and somebody will figure out how to do it. But it’s going to be years before that happens.

Until then, what can we say about drugs that might alter the natural history of diabetes and affect beta cell health? There are drugs out there and compounds in development that seem to intersect with the growth and death pathways in beta cells. And some of these are on the market. In my view, they are still potential therapies. And if somebody is trying to get you to put patients more quickly on DPP-IVs and exenatide because it’s going to prevent the progression of the diabetes, I think they’re overstepping what the data says.

We don’t have any way now to say that those drugs will be any better over time than the drugs we’ve talked about, the conventional drugs.

No question that if you had drugs that prevented beta cell drop out, beta cell health, that you’d want to use them way early. You’d want to use them to prevent diabetes. You might want to put them in the water, because if you could think about the cost savings to not have to use four test strips a day, five medications, go to the ER once a week or once a year for a hypo, et cetera, it would be a huge cost savings.

None of our drugs do this right now. None of the conventional therapies. Whereas, it’s unknown to say what you can expect from a DPP-IV or from exenatide, it’s pretty known what you can expect from the other drugs.

The question in my mind, if you can intervene early and hold the blood sugar down, keep the A1c at seven or 6.5 right away, will that have an effect on beta cell function over time? Again, you can make extrapolations from the UKPDS. But I think those kinds of studies would be very useful. Again, they would support really putting a lot more energy into early treatment than we currently do.

Again, I think this is the Holy Grail in my view, interfering with the natural history of diabetes, doing something to make the beta cell normal. Make the beta cell compensate for things that seem to be inevitable in our society right now, which is caloric excess and lack of physical activity. Until that happens, I think we’re still struggling to fight a mighty battle, because it’s a very hard disease to approach with current therapies.