DCCT and EDIC: The Diabetes Control and Complications Trial and Follow-up Study

Dr. Doron Schneider and Dr. Jack Leahy discuss the Diabetes Control and Complications Trial and the follow up study, EDIC. These landmark, studies performed 30 years ago, have stood the test of time and still influences the standard of care today.


Dr. Doron Schneider (DS): Hi, I’m Dr. Doron Schneider. I’m a general internist at Abington Health right outside of Philadelphia, and today we’re going to talk about a landmark trial called the DCCT trial, and here to discuss this landmark trial with me is Dr. Jack Leahy.

Dr. Jack Leahy (JL): Hi Doron. I’m Jack Leahy, and just to introduce myself I’m the head of endocrine here at the University of Vermont in Burlington, Vermont.
DS: Fabulous. Well, thank you for spending a few minutes with us here today, Dr. Leahy. We wanted to really reflect on the history of the DCCT given that we’re coming up on really 30 years of achievement that reflects both the DCCT and the continuation study called the EDIC study.  So, just to give everyone a foundation about what the trial is and the EDIC study is, can you tell us a little bit about really what was the construct of the DCCT? Can you give us a little bit about who were the patients, what were the questions they were trying to answer back 30 years ago when the trial had just launched?
JL: Yeah, I actually love that you said it’s a landmark trial. I think from my perspective and probably the diabetes perspective in general it’s really a stunning clinical trial that has clearly stood the test of time, and many people still view [it] as one of the great clinical trials in our field. So this is a trial that was designed to deal with a really controversial question many years ago—a question that actually younger people have really sort of forgotten, which was: Is there actually a proven benefit of intensive blood glucose control in patients who have diabetes, and if there is what actually is that benefit? And the question back then was more focused on microvascular complications as opposed to macrovascular complications because really the classic diabetes complications back in the ‘80s and ‘90s related to issues in terms of the eyes and kidneys and legs and that kind of thing. This was a trial of taking patients with type 1 diabetes broken into two subgroups. So [in] about half of the patients—and this is a US trial and on average, I think, there were 1440 patients or so that were enrolled—about half of them were absolutely free of any microvascular complications and in particular [from any] retinal microvascular complications. And the other half had had their illness for a little more on average than five years, and they had some very, very modest retinal complications—nothing terribly overt. So the design of the trial was to intensify blood glucose control in a subgroup and then look at the outcomes, primarily in terms of retinal health, but also many of the other complications in terms of peripheral vascular issues, neuropathy, and also in terms of kidneys, and then compare that safety in terms of rates of hypoglycemia and really anything else that might have been identified.
Now, what’s really crucial in sort of thinking about this trial is it was published in the New England Journal in 1993, and it was performed over an average of six to seven years. So you go back six or seven years, and it really was performed from the end of the ‘80s through the beginning of the ‘90s and thus was designed in protocol even before that. So this is a time preceding so much of the standard technology in therapeutic things that we use today. So, intensive therapy back then was really defined as, “Well, you can’t use NPH insulin” [neutral protamine Hagedorn insulin, an intermediate-acting form of insulin]. It was kind of known that NPH was a difficult insulin. So intensive therapy in the mid-1980s was either putting people on an insulin pump and that was about a little more than half the patients or giving them a basal bolus insulin program. But [at] that time the current analog insulins we have didn’t exist, so [the] basal bolus insulin program was mostly using Ultralente insulin [a long-acting form of insulin] and then regular insulin at meals (multi-shot insulin program with some blood glucose testing), as opposed to conventional control, which back then was really NPH and regular [insulin] with virtually no blood glucose testing. And then, again, they [the patients] were followed for an average of about seven years to see what happened.
The general sort of results were that in the intensive group they were able actually to get an average hemoglobin A1C over all of those years of about 7, which is actually stunning if you kind of think about it, as opposed to the conventional group, which were on average of about 9, so they had a 2 percentage point difference. And they just saw huge reductions in microvascular complications—stunning reductions in both retinal problems and in early kidney problems and in peripheral neuropathy of more than 50 plus percent for all of those over the 6 to 7 years of the trial. Th[is was] the beginning of the whole concept that has now driven diabetes care since that time,  [in] which intensive blood glucose control has a major if not dominant impact on risk of microvascular complications. We still sort of argue over the issue related to macrovascular, although I’m guessing we’ll talk about later. The EDIC trials actually shed some light on that.
DS:  Well, what a wonderful intro to the trial—its design, the primary endpoints and then some of the secondary looks at harm. If you can just tell us a little bit about the patients who were enrolled in the sense of their duration of diabetes. You referenced that they were relatively free of some complications, but can you tell us just so that we understand about how generalizable this trial is to the patients we see in the office. First, of all you did mention they were type 1, but how far into their disease were they? Were they young, were they middle aged, where they older? Can you just give us a frame of reference at a time of entry to the trial?
JL: Well, you know, generally fairly young into early adult years. I think on average children were not included, but adolescents were into early adult years. Again, half the group were free of any kind of known complications that had a duration of less than five years of the illness. The other half had very, very modest complications and, I think, were allowed to have a duration anywhere from 5 to 15 years is my memory. So, the kind of patients I think we would think about [are] pretty early in the course of type 1 diabetes and fairly young and, actually, that will be important as we start to think about the EDIC trial—but pretty early in the disease. This is in no way a trial of taking people with long-standing type 1 diabetes; established complications; high risk of cardiovascular disease, and then study[ing] the impact of intensive blood glucose control. That’s not this trial. This is when someone’s pretty early in the course of the disease and still very healthy related to this disease. [It’s looking at] how intensive therapy might have a benefit at that time.
DS: Excellent. That does help frame out the trial design. So, you clearly reference the outstanding benefit that we saw in microvascular disease in the order of 50, 60, almost 70 percent reduction. In that benefit we did see that it did come at some degree of cost as it relates particularly to hypoglycemia and weight. If we can spend just a few moments talking about each one of those. Let’s start with hypoglycemia. It was a pre-specified endpoint that was collected prospectively. What did we learn about really driving down that blood sugar to a mean of 7? What happened to hypoglycemia?
JL: So the answer to that is they had a much higher frequency of hypoglycemia, and a particular serious hypoglycemia, but, I think, the proper answer to that must frame the timeframe of when all this occurred. So, as I already said the only insulins which were available are really the insulins which precede the analogs which we have today—so just by definition not as safe as what we think about with our current analogs. So not only is it a fold difference in terms of rate of hypoglycemia that I’ll mention in a second, but also the absolute numbers actually have importance as we think about how things have really evolved over the last 25 years since this trial.
So what was seen back then [was] a three-fold higher rate of serious hypoglycemia and admissions to the hospital for hypoglycemia versus the control group. And I think when I try and teach this trial to students and to the residents, I try and point out to them this is not junky hypoglycemia. You know, the definition of serious hypoglycemia is, traditionally then—and has continued since then—as requiring help from someone else. But if you really talk to patients these are events that they don’t forget or that their spouses or other people who’ve been involved don’t forget. These are seizures. These are unconsciousness. These are car accidents. I mean these are major events. And back then, a three-fold higher rate, I mean, that is very, very serious in terms of that observation….It was clear intensive therapy back then came with a significant price in terms of hypoglycemia.
The other issue that I think is interesting is that when you look at the absolute rate as reported back in the trial, it’s usually reported as episodes per hundred patient years, and I can’t remember the exact number, but it was somewhere in the range of about 50 or 60 episodes per 100 patient years. So what that meant is with state-of-the-art intensive therapy in a, you know, very carefully monitored clinical trial back at that time—back in the 80s and early 90s—that people who were in the intensive group would have on average an episode of major hypoglycemia a little bit less than every two years, which in today’s world would be shockingly unacceptable because as we now move forward and just sort of move in today’s world and think about what clinical trials might tell us in type 1 diabetes with intensive therapy with all of the monitoring devices we have and with all of the better insulins, you know, we typically think about a rate of 2, 3—somewhere in that range—certainly 1/10 or less than what we saw back then. So the rate of hypoglycemia was very significant back then and clearly balanced all the positives.
DS: And before you finish this segment on hypoglycemia is there any data about any long-term sequelae from hypoglycemia in the sense of impact on cognitive status, dementia, or any other long lasting effect as to patients who did have hypoglycemic events as you described?
JL: Well, that’s a hard question and I’m not sure how to exactly answer that. I’m not sure whether there are any reported data from this trial on that specific issue. If there are I can’t answer that. And I think in some respects we have more of a general concern. I think all of us in this business of seeing people who have had repeated bouts of hypoglycemia and have had probably some medical and/or neurological sequelae related to that. On the other hand, I think that the prevailing opinion would be just the opposite: that people tolerate bouts of serious hypoglycemia pretty well. We do know that there’s a relationship maybe with dementia in the diabetes world, probably more [with] type 2 diabetes. I don’t think that’s related to hypoglycemia, but on the other hand I guess one could argue a little bit we don’t exactly know, and there was great fear years ago that if you look at children with type 1 diabetes and do IQ tests they seem to be a little bit less good than control children, but on the other hand, if you now do studies against other children with chronic illnesses they kind of match up. So I guess my feeling in the answer to that is it’s never desirable to have any kind of significant hypoglycemia problems in patients, but it is not clear to me there’s any lasting sequelae in those patients.
DS: Right, and that’s my understanding as a primary care doc as well regarding both in general and as it relates to this trial that they did not find [lasting sequelae] specifically in the hypoglycemia group.
JL: Yeah, if I might, let me just actually jump in with one more thing if I might because there is another part to the trial related to hypoglycemia I probably should identify. So if you actually go back and read that trial because of this increasing rate of hypoglycemia in the intensive therapy and again a three-fold increase in the intensively treated patients. If you actually look at the New England Journal paper there’s an interesting figure, which is showing the benefits in terms of microvascular reduction as graphed against the risks of hypoglycemia when your x axis is the hemoglobin A1C. So the conclusion of that study was the optimal hemoglobin A1C in a patient with type 1 diabetes was 7 because then you have near total reduction in risk of retinal disease as opposed to less of a risk of serious hypoglycemia than if you further lower the A1C. So, back then we were pretty comfortable: shoot for 7, don’t go much lower, you know, do the best you can. I think in today’s world probably the discussion is not quite so absolute because our current insulins are a bit safer and actually the EDIC trial again has sort of raised some issues: maybe a bit lower is better.
DS: Right. So let’s move now on to weight: what happened over the course of DCCT? And then we will talk about the extension trial in a moment. That, I believe, was one of the other major costs, if you will—correct?
JL: You are exactly correct. Everything I’ve said so far at least in terms of the risk of hypoglycemia was predicted. I mean, no one’s going to be surprised that when you lower A1C dramatically in these patients that, sure, hopefully, you would see benefits, which they did, but it’s going to come with a cost, and the cost was hypoglycemia. What was not predicted and actually was a big shocker is that there also came with a substantial risk of weight gain. And if you go to the original trial—the New England Journal paper and just kind of quickly read—what you will see is that on average about a 5-pound weight gain (so, remember this is about a 6- to 7-year trial) in the conventionally treated patients as opposed to about a 10-pound weight gain or thereabout in the intensively treated patients.
So, when you first read that, I think you could sort of say, “Oh, well, you know it’s there, but it’s not stunningly huge.” So even though it was there, it was reported as an outcome. I don’t think people initially sort of appreciated that this is kind of much of an issue. And then starting two years later: lots and lots of subgroup analyses, different kinds of papers, take this huge database and start to analyze it, and there was a paper published a couple of years later which I think really put this into a capsule that we had not anticipated. And that was they started to look at the weight gain in the patient population in quartiles. And so, there was a quartile of people who didn’t gain any weight, so, you know, they were fine with the intensive therapy. And then the opposite quartile the average weight gain was very, very substantial. I can’t remember the exact number, but probably about 20 pounds on average. And actually when you went to that quartile, not only were they gaining weight; they were also having manifestations of many of the metabolic issues we sort of link to weight gain, such as blood pressure increases, such as a more metabolic lipid profile. And for the first time I think people’s eyes opened up to realize well, wow, there maybe is some overlap between the metabolic background that we then had typically linked really only to type 2 diabetes. Maybe in some patients with type 1 diabetes there’s actually a similar kind of predisposition, which might be brought out through intensive insulin therapy.
Now we again move into today’s world, and that’s a huge topic of conversation now because our teenagers and young people with type 1 diabetes are getting bigger and bigger and are having more of the metabolic sequelae that one sort of thinks about with obesity and insulin resistance. It’s not unique to type 2. We’re seeing it more and more in type 1 diabetes. So that’s the other issue with weight. It’s really emerged years after the trial to recognize this is a major risk of people who are on intensive insulin therapy if they have the genetic predisposition to be at risk for that.
DS: Now, what I’d like to turn to is the next phase after the DCCT trial formally ended. There was a wonderful natural observation study called the EDIC study where there were additional insights about such things as weight that you just described as well as the primary endpoints that were described earlier in our podcast. So can you please tell us a little bit about what happened in the EDIC study, which is a continuation observational trial? How many years did it go for, and what happened to the two arms of the DCCT?
JL:  Yeah, so this actually just I think a fabulous story. I am of the generation, I think you are too quite frankly, where when the DCCT was published I mean it was incredible news in 1993. It changed our world. It changed how we thought about things. But somehow we thought it was a complete story and that the story was if you intensively treat people with type 1 diabetes in a way that was doable, and increasingly doable as treatments got better, that you [would] have a real[ly] important protective effect against microvascular complications. But in the original trial there was no benefit for macrovascular complications and, in part, that was because these were pretty young, for the most part healthy, people who, you know, were not terribly at risk for cardiovascular events and so, you know, you’re not going to see much of a protective effect if not much was happening in the control group.
So that’s kind of what we thought was the finished story and so what the EDIC trial did was to take people who were now finishing the DCCT trial and continue to follow them, but no longer in any kind of controlled environment. So essentially the trial stops, and all of the intervention stops so that the people who were in the intensive group, you know, within about a year A1C started to loosen up a little bit and sort of drifted up to the high 7s because that’s what happens. It’s not easy to keep an A1C of 7 percent with type 1 diabetes. It’s really hard work and when you’re no longer in the clinical trial then it just loosened up a little bit. So it rose a bit. Then the people who were in the control group who previously had an A1C of 9 percent they started to hear all of the wonderful things attached to the better control and so they just naturally improved because, I think, the medical profession had adopted that, you know,  this is important. So their A1Cs came down to the high 7s. And so somewhat serendipitously these two groups within a year or so essentially equalized in terms of hemoglobin A1C and actually equalized in terms of sort of treatment approaches. So they became one and the same now. There was no obvious difference between the two.
And then they were just followed…And they were followed for close to another ten years. Data w[ere] periodically collected to understand the different health issues of these individuals and see how they’re doing. So that’s the design with some pretty interesting surprises.
DS:  Yeah, so a very crisp review of that observational follow-up study which lasted almost 20 years, and you eluded to the interesting findings. So, let’s start off first with the primary endpoints regarding microvascular complications. What did we see as it relates to them?
JL: So, I guess maybe to try and understand why these are interesting, you might take one step back and say, “Well, what might you have predicted?” And I think probably people would say, “Well, the prediction would be then when you have this difference in A1C, this intensive therapy difference, you’ll see a protective effect against something, i.e., microvascular complications as well as identified. But now, if the difference in A1C goes away then in theory the protective effect will go away. It may take a period of time, but the concept would be that it will go away. I mean you need ongoing intensive therapy to maintain the benefits. And that is not at all what was seen. The initial protection against microvascular complications, and the ones that have been mostly focused on are retinal, but there’s been also neuropathy and in some of these, they continue. Something about that early period of intensive therapy gives an ongoing effect that does not require maintenance of that intensive difference in A1C, which is kind of stunning if you think about it.
DS: Right, and I believe that that concept has been one that has been named metabolic memory. I believe that it has really fundamentally altered the way that we think about the early and aggressive treatment of patients who are just diagnosed really for all the above reasons that you’ve just articulated. This is a lasting impact over decades to really get that initial early control. The beta cells have memory.
So, let’s then focus on the big other shock that you’ve alluded to a couple of times regarding macrovascular complications. Now, you did clearly state that this is a very young population. They didn’t have a lot of co-morbidities. They didn’t have a lot of cardiac risk factors. So we didn’t see a lot of endpoints as it relates to cardiac outcomes in the DCCT. What did we see as it relates to cardiovascular outcomes in EDIC?
JL:  So, I love how you framed the question because the shock— this is just more than shock. I mean this is just really an altering event in our world, which is this same lingering metabolic memory, or whatever you want to call it, was true for cardiovascular disease, and so all of a sudden because the trial now drifted out to another ten-plus years from the original seven years of the DCCT, you’re getting older people who are having higher cardiovascular event rates and the same protection—sizable protection; we’re not talking something little, dinky, I mean, we’re talking at the sort of 30 to 40 percent protective range—was still there years after the trial was over. And again I emphasize that not only had A1C come back to be equal during the ten years they weren’t in the trial, but blood pressures were the same, lipids were basically the same, I mean, none of the cardiovascular risk factors that we think about easily explains any of this. There was something magical—magical!—about that first six to seven years of intensive blood glucose control in these patients, which continued to be present 20 years later, and I don’t think we have a time where it’s ended yet. I mean, maybe it’s going to go forever. So [this was] a stunning observation that now covered not just microvascular complications, but macrovascular complications and totally moved the conversation of intensive blood glucose control and protection…[to cover] not just…microvascular issues, but long-term macrovascular issues.
DS:  So it really crossed the gamut of microvascular and macrovascular improvements in endpoints and thus was a trial that really shook the landscape and will never be repeated. I’m quite confident of that. As we finish up here I can’t help but just have you comment, if you would, around the type 2 diabetic patient. Do any of the things that we just described really bear out in type 2 diabetes? This is a majority of new onset diabetes is type 2 and this was a landmark trial. Can we glean anything as it relates to an even larger population—as it relates to micro- or macrovascular—from DCCT or EDIC?
JL: So the wonderful way to answer that is “more or less”: more or less everything I just said can be applied to type 2 diabetes as long as you’re willing to be a little creative in terms of lining up the similarities. So, as soon as the DCCT came out, and again that was in a New England Journal paper in 1993, I think the results were so widely and stunningly accepted that intensive blood glucose control, as possible that many years ago, was viewed as now the standard of care for type 1 diabetes. Fine. But immediately, the question became, “Does this have any relevance to type 2 diabetes?,” because, of course, the vast, vast, vast majority of patients around the world have type 2 diabetes, and also, you know, we didn’t use insulin quite the way for type 2 diabetes then that we do now so we didn’t quite think about intensive therapy the same way for type 2 diabetes we did for type 1. And then, of course, in older folks there was always fear about additional risks of hypoglycemia in terms of what that might mean for cardiovascular risks on and on.
So, it turns out we asked does it work in type 2 and then five years later in 1998 was the publication of the famous UKPDS, the United Kingdom Prospective Diabetes Study that was not actually perfectly analogous to the DCCT—it was a little bit of a different trial—but it was looking at first drug therapy in type 2 diabetes in people who were very early in the course of the disease, looking at an attained hemoglobin A1C with intensive therapy of 7 versus about 8 in people with conventional therapy. So not quite the same delta that we had in type 2, but with very similar results, which showed a relationship between intensive blood glucose control and prevention against microvascular complications in type 2 diabetes. No clear prevention of macrovascular, although that’s a very, very complicated subject. How one group took metformin and maybe metformin showed some cardiovascular protection—that’s open to a bit of debate—and there was a reduction in myocardial infarction in this trial in absolute terms, but not statistical terms.
So again, we were sort of left with well, intensive blood glucose control seems to be good also in type 2 for microvascular. And then the UKPDS did exactly what the DCCT did, which is to basically wait about another eight to nine years. Again, the trial was finished. People were, you know, let out of it. Their A1Cs normalized in the intensive group versus the control groups. Everything else normalized including blood pressure, and then they were actually studied eight or nine years later with the same finding, which was continued, now proven, protection against cardiovascular disease in the people who had had the A1Cs that were intensively controlled versus those who hadn’t. Same molecular mimicry.
And actually one of the things that gets lost in this conversation is if you read the original UKPDS what you took away from that is that blood pressure was a better control target than A1C because blood pressure during the trial did actually seem to confer some cardiovascular benefits while blood glucose control didn’t really. But at the end of the continuation phase, that eight years later, when A1C had normalized, blood pressure had normalized, the cardiovascular protection lingered despite A1Cs coming together, but the blood pressure related one didn’t. So if there is a benefit of intensive blood pressure control in diabetes you need to keep it controlled. There isn’t this sort of long-term memory.
And then as a final closing of the loop there’s actually another trial we talk a lot about—something called the Steno-2 trial, which is a small trial of taking people in Scandinavia and trying to intensively control not just their blood glucose, but blood pressure and lipids, and at the end of that trial there was pretty good cardiovascular protection, but then everything was stopped. They were studied an average of seven or eight years later, and they still had this lingering cardiovascular benefit. So this whole idea of mimicry, or at least this sort of molecular protection, has been shown not just in type 1 diabetes, but type 2 diabetes. It seems to be a biological fact.
DS: Well, Jack, I really want to thank you for reviewing the DCCT trial, its primary endpoints, its secondary endpoints, the EDIC study and really giving us the context of those studies as it relates to the type 2 diabetic patient and that review. And these are landmark trials and have absolutely set the standard of care as it relates to the goals of treatment, as it relates to how [to] monitor patients over time, and to help us all be grounded in fact—and not fiction, but fact—as to what expectations we can get from that type of control. This trial, the DCCT, in the observation period set the standard as you just articulated, emulated by other trials and clearly in its thirtieth anniversary is still holding the test of time.
But what I’d like to do is wrap it up right now and refer our listeners to betacellsindiabetes.org for additional information. And at this point, again, one final thank you to Dr. Leahy for an eloquent review of the trials. We look forward to seeing you next time. Thank you, Dr. Leahy.
JL: Thank you very much.