Pape: Hi, this is Joy Pape with Diabetes In Control. We’re at AACE 2018 in Boston and I have the pleasure today to speak with Dr. Kathleen Wyne, just like the wine some of us drink, but most people know her as Kittie. So, Kittie, we’ll just call you that. Now you did Chair at another talk – Reproductive and Endocrinology and Daily Endocrine Practice – and you’ll be moderating today, Horse or Zebra: Unusual Lipid Disorders. But we’re not going to be talking about those today; what we’re going to be talking about is what is near and dear to your heart and what you like to talk about. So, here it goes: Why do you think adult-onset type 1 diabetes is an important topic?
Wyne: So, I think it is really, really important because people don’t realize it exists. Everybody thinks type 1 diabetes is only diagnosed in children. At least 50 percent of type 1 diabetes is diagnosed in adults, and, that diagnosis is usually missed. I cannot tell you how many times I’ve gone into a hospital room, I’ve reviewed the chart, I’ve looked at the pattern of their sugars and I see just this weird variability that I know is type 1. I say to the patient, “How long have you had type 1?” And they say, “Oh no, I don’t have type 1, I have type 2.” And I say, “Why do you say that?” They say, “Well my doctor says I am too old, so it has to be type 2.” And I tell them, “You have type 1 and I will prove it.” But that happens so often and also sometimes patients are proactive, and they come to my clinic, having requested a referral because they say, “I’m 40-years-old, I’ve never been fat, why would I have type 2 diabetes?” Now they might still have type 2 diabetes, maybe they have Asian ancestry where our standards for overweight and obese are lower. But it’s a fair question. The same thing with an obese person who has type 2 who needs very low insulin doses, maybe actually has type 1. So, it does change the way we manage them.
Pape: So, you say that you’re going to prove to them that they have type 1. How do you do that?
Wyne: That’s actually a very good question because most people think you diagnose type 1 by doing a C-peptide and if the C-peptide is undetectable, then by definition, they have type 1. That’s not true. First of all, someone who has had type 2 for many years could have an undetectable C-peptide. Second of all, any time you measure C-peptide, it must have a simultaneous glucose. You need a context. So, this person who is an adult onset type 1, may have a C-peptide that’s reported in the normal range. But if the simultaneous glucose is 410, then it’s inappropriately low which is not what you would see in a newly diagnosed type 2 where it would be inappropriately high. So, first of all, your C-peptide needs to be in a context. Second of all, that alone may not convince you it’s type 1. But what you do is you go ahead, and you draw the antibodies. We now have five antibodies that we use to characterize type 1 diabetes. Not everybody has all five antibodies. Everybody knows the GAD65 antibody, they know the ICA antibody. There’s also IA-2, ZnT8, which is the newest one and stands for Zinc Transporter 8, and there’s also the Insulin Autoantibodies. So, we measure all five of those in all of our newly diagnosed type 1s or suspected type 1s, using them to confirm the diagnosis. There will be an occasional person who is negative for all of those antibodies, so, we believe there are still more antibodies to be found.
There’s a couple papers in the literature describing a new antibody called, Tetraspanin-7, not commercially available. So far, data suggests some minor antibody, but again, for those antibody-negative people, there’s hope that there are more antibodies coming.
Now when we look at these five antibodies, as I mentioned, not everybody has all five, but I’ve very often, in my adult-onset type 1s, found that their GAD65 negative, ICA negative, so if you had only done the two classic antibodies, you would have said, “Oh you’re type 2, you’re not type 1.” But, they come up positive for IA-2 and ZnT8.
Pape: Very interesting. So, I have a question to ask and I have heard in the literature and what people tell me; When you do the C-peptide and the antibodies, should it be fasting or non-fasting?
Wyne: So, the antibodies do not need to be fasting. Ideally, to be clean, you would want to do the C-peptide in a fasting situation. Here’s the reason why I do it fasting: I’m actually doing the C-peptide, not to make the diagnosis, but in case I need it to qualify them for an insulin pump. To qualify for an insulin pump, the C-peptide must be drawn with a glucose below 225 mg/dl. So, that’s why I am drawing at fasting. I’m hoping that I’ll catch those two, and generally if insurance requires it, it just has to be drawn within six months prior to the application. What that is also telling you is I’m pushing my newly diagnosed patients to pump and CGM as early as possible. This raises the next questions of: Should we be doing that with the newly diagnosed type 1s because many of the adult-onset type 1s have a prolonged honeymoon – they may honeymoon for five years, whereas in the childhood-onset, the classic honeymoon is one to two years. So, maybe they don’t need insulin at first. Maybe they don’t need to go to a pump, but I want that option as early as possible. So, if I am going to measure the C-peptide, I’m going to do it fasting, so it serves that other purpose if I need it.
Pape: So, why is DKA no longer sufficient to make a diagnosis?
Wyne: That’s actually an interesting question too because for many of us when we train, DKA by definition was type 1 or we would say it was pathognomonic of type 1. So, I did part of my training in Texas where some years ago, and I’m not going to say how many, we started seeing certain minority groups coming in with DKA. They required very high doses of insulin – very insulin resistant. They settle out on the insulin infusion at 4 to 6 units an hour, whereas a true type 1 would be at .5 to 1. If you were to do C-peptide at that point, it would be undetectable. You discharge them on insulin, two weeks later, they’re calling in that they’re hypoglycemic. Three months later, their C-peptide is very, very high because they regained their beta cell function; they’re behaving like type 2s. What we’ve learned is that certain people with type 2, if they go untreated long enough, with their high free fatty acids and lipotoxicity and glucotoxicity, the pancreas just stops making insulin. It basically says, forget this I am not going to do this anymore. So, they go into DKA because it’s the absence of insulin that gives you the DKA, not the high sugars. So, once you resolve the glucotoxicity and the lipotoxicity, the pancreas starts making insulin again, and it could keep making it for quite a few years. The best description of this ketosis-prone type 2 diabetes, I think, comes from a cohort from sub-Saharan Africa where they described these lean people who come in, insulin resistant, come in with DKA, and they do beautiful for about five years, and somewhere at about five to seven years, the A1C starts to jump up and then it goes up high. But the key is ketosis-prone type 2 is a very different phenotype than the classic type 1. So, you don’t know what the cause of the DKA is – is it complete loss of beta cells or is it temporary loss of beta cell function? C-peptide doesn’t tell you does it?
Pape: No, and that begs another question: What about the people who are prone to DKA, people who have type 2 diabetes on SGLT2?
Wyne: Oh! So that’s our euglycemic DKA. Euglycemic DKA has been in the literature for many years, we just kind of all forgot about it, but it has always been there. The key there is that with the recognition, the SGLT2s brought it out that we need to remember these exists. We have found a couple of the predictors, but more importantly, we need to rethink and change the way we teach DKA to our medical students, our residents, and this includes internal medicine, family medicine, and emergency medicine. They need to understand that glucose is not required for DKA. If you look at the physiology, the biochemical changes, everything that causes DKA, causes hyperglycemia. But, they are not bidirectional arrows. The glucose is not causing the DKA, it’s the lack of insulin that’s causing the DKA.
So, what are the predictors that we have? One is eating a low-carb diet on one of those agents. The other one, quite simply, is not taking your insulin, which is going to cause DKA anyway, right? But that’s one of the components or just simply taking too little insulin. Something that needs to be kept in mind is that when you’re hyperglycemic, if you’re drinking enough fluids, you can maintain your sugar around 250. So, when my patients come into the ER, and they say, “I have type 1 diabetes, I’m throwing up, my stomach hurts, I think I have DKA,” and the ER says, “Well, your sugar is below 350, go home,” that’s the perfect example of why we need to retrain people to understand it’s not sugar. Just because the sugar is below 350, doesn’t mean they are not in DKA. That’s the lesson we need to learn.
Pape: Very good. And so, what is the age of the oldest person you have diagnosed with type 1 diabetes?
Wyne: So, earlier this year, I diagnosed a 76-year-old man. But in my training, I participated in the diagnosis of someone who was 92 years old with autoimmune type 1. And the reason I say participated is because I wasn’t the primary person doing the consult who asked the questions, that was actually the fellow who did the consult. But I got to help in the evaluation. So, for me myself, the first one that I thought might be, I would say is 76. But I have seen as old as 92 presenting with autoimmune type 1 diabetes. Let me give you the flip side of that, though: What’s the youngest type 2 I’ve ever seen?
Pape: I don’t know.
Wyne: So, in Texas, we had children as young as two-years-old with type 2 diabetes. So, age does not tell you the type of diabetes. Patients don’t come in with a number on their forehead. You have to use your clinical skills to decide what kind and then do your test to prove that you’re right.
Pape: When do you start screening for microvascular complications in adult onset type 1?
Wyne: So, you know the data we have with type 1 diabetes suggests that microvascular complications don’t start until after at least five years of hyperglycemia. So, the easy answer would be to say five years. But, first of all, you have to put in the context of the patient and any other diseases, comorbidities, risk factors. So, if I have a slender 26-year-old who’s never been sick a day in their life, I’m probably not going to screen for any microvascular complications. If I have a 58-year-old man, odds are he’s got a little bit of hypertension, right? So, I am going to check him out for microalbuminuria – I’m going to look at his lipids carefully. I’m probably not going to do a formal retinal exam; not for at least five years. Do I need to do neuropathy screening at diagnosis? Probably not for a type 1. So, to me the big one is the microalbumin, and that’s a function of age and other comorbidities.
Pape: And when do you start statins on an adult-onset type 1?
Wyne: (Laughs) The statins are a big question. Patients with type 1 diabetes tend to be very tech-savvy, science-savvy, medical-savvy, so they already have some thoughts when they come in. So here is the challenge with statins: Everybody says when you have diabetes you must be on a statin. Well, the majority of that data is in people with type 2 diabetes. And I would say that the standard-of-care for people with type 2 diabetes is everybody gets a statin. Doesn’t matter what their lipids are, their A1C, their age, well not totally, the teenagers, you might wait a little bit depending on their situation – their A1C, their comorbidities. Type 2 diabetes, again, it’s going to go back to age and comorbidities. If I have a 25-year-old who is slender, healthy and active, that type 1 diabetes does not automatically get them a statin, OK? That 58-year-old man probably needs to be on one anyway, irregardless of the diabetes, right? So, the issue becomes age 40. So, for the general population, age 40 is kind of the break point of when you should start thinking about statins. So, the challenge becomes, if the person is 38-years-old and newly diagnosed, should they be on a statin? And the short answer is: You have to look at their comorbidities and risk factors. So that’s the newly diagnosed person. What about the person who was diagnosed at age 20 and they’re now 40-years-old? Does that person now need to go on a statin? This is where it gets difficult. There’s data from Europe and from the Untied States and some of the data suggests that after 15 or 20 years of type 1 diabetes, maybe you should be on a statin. One thing to me that is pretty clear from the data is if you have had the diabetes for five years and your A1C is still uncontrolled, you need to be on a statin. I don’t think anyone is going to have any problems with that. If you are still running an A1C of 9 to 12 after five years, you’re probably not going to come under control and it’s probably a good idea to be on a statin. What about the person who has had it for 20 years, maybe had some rough times in the late teens, early 20s, but has been doing well, and maybe has been doing well for 10 years. Their BMI is below 21, their systolic blood pressure is 110, they are physically active, their A1C is in the low 6s, does that person really need to be on a statin? That’s the big question right now in type 1 diabetes.
The standard risk calculators don’t account for the difference between type 1 and type 2. One of the things we did in the AACE lipid guidelines, the update in 2017, is we made a point of pulling type 1 and type 2 apart, which no other guideline has done, to try and give people those tools. What we’ve found was there actually is a risk calculator that’s dedicated to type 1 diabetes; it’s been developed in Denmark, it’s available free online, it’s actually published in Circulation, I think in 2016. We use that to estimate CV risk in our patients and document the risk and track it. If you take that risk calculator and just kind of put in different possibilities, what comes through very clearly is, as soon as the evidence of kidney disease, cardiovascular disease escalates. This was also shown with the MACES study. But, microalbuminuria alone doesn’t escalate your cardiovascular disease. The moment the estimated GFR starts to drop, then your cardiovascular disease risk goes up. So macroalbuminuria with preserved GFR, I’m going to treat those people, I’m scared that their GFR is still preserved. With micro, I am going to give it some time to see if we can make it go away; if we can’t make it go away, I am going to probably put them on medication. But duration of disease alone isn’t necessarily enough to me in the context of glucose control and also asking the questions, “Do they have hypertension, do they smoke, do they have any other complications?” If they’re developed retinopathy, I’m going to give them a statin. If they’ve developed significant neuropathy, I’m going to give them a statin. I’m talking the person who has trace neuropathy, maybe has one red dot in their eye after 20 years, they may not need a statin. It’s complicated, isn’t it?
Pape: Yes, it’s complicated! (Laughs) I feel like we just got a bite of a lot more that we can learn about and we hope to learn more from you, I can’t thank you enough.
Wyne: Well, thank you!