Activity Transcript

David C. Klonoff, MD, FACP, FRCP (Edin), Fellow AIMBE: Hi, I'm David Klonoff. I'm going to be moderator today of our session on Ketone Monitoring in Patients With Diabetes, 5 Things You Need to Know. Our faculty member is Desmond Schatz, professor at University of Florida. 

Desmond Schatz, MD: Well, it's good to be here, David. Really always nice to do programs, certainly with you, and really excited about today's topic, which is ketone monitoring in patients with diabetes.

Dr Klonoff: Ketoacidosis is a severe, potentially fatal complication of diabetes. It's defined as a pH of less than 7.3 with elevated ketones in the blood, generally 3.0 mmol/L or more. Ketones become present in these high concentrations, either in the absence of insulin or the absence of insulin action. When that happens physiologically, there's also an associated increased in glucagon. The glucagon leads to breakdown of fat in the periphery. The fat creates free fatty acids, which are converted in the liver to ketones, and if the ketone levels climb, it overwhelms the ability of the body to buffer the pH. You get a falling pH, which is dangerous to every organ in the body and can lead to collapse of every organ. 

I'm going to start with the first question for Dr Schatz. Should pediatric and adult patients with type 1 and type 2 diabetes monitor their ketones regularly?

Dr Schatz: Well, thank you very much, David. If we just look at data in terms of US adults who come to the emergency room with ketosis as well as to admissions. What we can see is that there has been an increase, certainly over the last decade or so. In 2009, the number of patients per 1000 that were seen in the ER with ketoacidosis was about 8. And in 2015, that had doubled. And we look at inpatient rates. Similarly, we see that in 2009, about 15 to 16 per 1000, and again in 2015, that had gone up to really 25. So clearly, ER visits as well as inpatients had increased. Not only have there been those increases, but there's also been an increase in mortality because of increases in hospitalization rates among persons with diabetes in percentage terms, over 50% from 2009 to 2014.

But ketoacidosis is not unique to just type 1 diabetes; we see ketoacidosis in patients with type 2 diabetes as well. And this is not only at the onset, but also in patients with established diabetes. And an older study had shown that in patients with type 1 diabetes, yes, they have perhaps more severe acidosis, and the resolution of the acidosis may be a little quicker. But in type 2 patients, over 70% of the patients were overweight or frankly obese, and type 2 diabetes was far more common in the Latino American and African American patients. And infections precipitated their admissions in close to 50% of cases. So again, take away that type 1 and type 2 are prone to develop ketoacidosis. But perhaps a much greater problem is the fact that we need to prevent these ER visits and we need to prevent these hospitalizations.

I want to reference a paper by a colleague, Dr Anastasia Albanese-O’Neill, who, in 2017, published a paper in diabetes care, which in essence showed that so few people were actually testing for ketones. When she looked at patients whose blood glucose were over 300 mg/dL. Yes, younger patients were more likely to check for ketones, but that was only about a third -- this is type 1 patients over the age of 25 -- just 7% of patients checked for ketones. Yes, there were little better rates when patients had nausea and were vomiting and were sick. And yes, the vast majority of patients under the age of 6 checked. But over the age of 25, just 26%, that's a quarter of patients actually checked for ketones. So clearly, there is a need, in answer to your question, David, to check for ketones.

Dr Klonoff: Well, ketoacidosis is a big problem. It's not being monitored for as much as it should be. 

Desmond, what do you tell clinicians to be aware of when they talk to their patients about ketones and ketoacidosis?

Dr Schatz: It's very important that people are aware of what kinds of ketones there are. And really there are three major ketones: acetoacetate, which comprises about 20% of all the ketones; acetone, about 2% of ketones; and beta-hydroxybutyrate, about 78% of ketones. But realizing that during ketoacidosis and insulin deprivation and lipolysis, that what leads to this increased production of the ketones, and especially beta-hydroxybutyrate, and the ratio between acetoacetate and beta-hydroxybutyrate really increases from 1:1 to 1:20. So it's really important that you monitor, and by far the best test is blood ketone monitoring, although it is used in the minority of cases. And using beta-hydroxybutyrate gives you a very good indication of both impending as well as actual ketoacidosis, because we know that if you were to check for ketones and measuring beta-hydroxybutyrate on a capillary blood sample below 0.6, that's really a normal range. Once we get elevations in the ketone levels in the blood, certainly the risk of ketoacidosis goes up. And we do know that there's a higher risk above 1.5. And in fact, if the values exceed 3.0 mmol/L, there's a very high risk for diabetic ketoacidosis. And in fact, medical attention must be sought immediately. But again, education is key and testing is the key as well.

Dr Klonoff: Desmond, when you teach your patients to test ketones, do you have any particular strategy for what they should do on sick days? 

Dr Schatz: Sure. Before I get there, I do want to sort of state that there are clearly risk factors for the development of ketoacidosis. And it's really important that the provider and the team taking care of patients are aware of these risk factors. And these are, most importantly, the suboptimal adherence to treatment and poor glycemic control. Unfortunately, but not surprisingly, lower socioeconomic status, psychiatric disorders, and younger patients are all risk factors, as is recurrent DKA. When we actually get there in terms of sick day guidance, it's very important that once recognized, that patients will check for ketones certainly every 3 to 4 hours. What happens is that acidosis can lead to alterations in consciousness, but dehydration is really a hallmark of ketoacidosis. So, it's very important because of the dehydration that all patients drink plenty of liquids. They should have been educated about the need for additional short-acting insulin -- that's not just to cover the high blood glucose, but additional insulin for the presence of ketones. And always, should the ketones not go away within 3 to 4 hours, to be in contact with a provider and, in fact, to seek urgent medical care when needed.

Dr Klonoff: Desmond, are there any particular points that you emphasize when you talk to patients about measuring ketones, either symptoms they should watch for or any other strategies for measuring ketones?

Dr Schatz: Yes. So, I think it's very, very important that we are aware of the symptoms of diabetic ketoacidosis. And again, early symptoms include the polys: polyuria, polydipsia, and, much later, polyphagia. But again, urinating a lot more than usual is often just the feature, and sometimes people don't realize it and that leads to dehydration. And it's really important to understand that, as you had said, Dr Klonoff, that either a relative or absolute insulin deficiency can lead to diabetic ketoacidosis. And other symptoms, such as dry skin, dry mucus membranes, nausea, vomiting, and again, you can get the so-called Kussmaul breathing, which is faster, deeper breathing, and the hallmark is this fruity-smelling breath. But overall, patients feel tired, they don't feel themselves, they can be constipated and just don't feel good. So, really nonspecific features, but really being aware of the symptoms and the signs are really key to the understanding and fact to the prevention of both ER visits as well as hospitalization.

Dr Klonoff: Desmond, you've spoken about current technology and the benefits of measuring blood ketones, especially during ketoacidosis where the acetoacetate levels might not give you a good picture. 

I'd like to comment on the question that's been raised, what is continuous ketone monitoring? What's its status? Imagine if a person could wear a sensor that would measure ketones continuously. This would be really useful information to provide to any patient who uses insulin. And it would be especially useful for someone who's had recurrent DKA for whatever reason. And this could be that they have type 1 diabetes or there's a rare type of diabetes that affects people who seem to have something similar to type 2 called ketosis-prone diabetes, also known as Flatbush diabetes. And now we're seeing SGLT2 inhibitors used widely for type 2, sometimes off label for type 1, and sometimes for other conditions. And these drugs have caused an increased incidence of ketoacidosis. And then also you have people who don't have diabetes, but they go on a ketosis-prone diet, that may be to lose weight or they feel it has other health benefits. There are some neurologic diseases that are treated with a ketosis diet, and every one of these cases could benefit from continuous ketone monitoring. 

There was a landmark article that was published last year in the Journal of Diabetes Science and Technology -- the first author was Shridhara Alva -- which expressed the feasibility of continuous ketone monitoring in subcutaneous tissue using a ketone sensor. In other words, a continuous ketone monitor. And it used the same general type of technology, which is called wired enzyme technology. It used an enzyme, not glucose oxidase. Because we're measuring ketones, it used beta-hydroxybutyrate dehydrogenase, and they put 3 sensors onto 12 different volunteers. So, there were 36 sensors that were used for 2 weeks. They asked these volunteers to be on a low-carb diet, so that was a ketosis-prone diet, and they measured a reference ketone measurement. And they compared this reference method with the results they were getting from their continuous ketone monitor. 

They had a linear response from 0 to 8 mM, and by the time you get to 9 mM, almost nobody with ketoacidosis is that high. The sensor was stable; it only had a 2.1% loss of signal over 14 days. They did have to use a single retrospective calibration, however, and ideally, you have to get a product on the market without this type of retrospective calibration. You can do prospective but not retrospective. 

They defined 2 levels of accuracy when they plotted all their data points for a reference ketone level of less than 1.5 mM, then 82.4% of the data pairs were within 0.225 mM, and 91.4% were within 0.3 mM. So given those seem like reasonable targets, although there's no official consensus, but seems reasonable, and they did very well with those targets. When the reference ketone was higher, like 1.5 mM or greater, they still had good accuracy in the +/- 20% range. 76% of data pairs with were within that range and within 30%, 89.7% of data pairs were within that range. It was a promising study.

This kind of product needs more work. It also would need to be approved by the FDA. And we need more data, especially in the high ketone range. I don't think there's enough data from this study for approval, but it's a really good start. And we also need more information about what the best reference ketone method is and how accurate are we in defining the level of performance. So then, we can see whether this product or any other product meets that level of performance. For a product to be commercially viable, I believe it needs prospective calibration. You can't calibrate after the fact or you don't know where you were. And finally, we have to test this in more types of patients, especially people who actually have ketoacidosis as opposed to volunteers.

I would like to wrap up my presentation about continuous ketone monitoring by talking about how, in the future, continuous ketone monitoring will be a digital health tool, because it's one of these sensors that will be bringing in physiological data that can be assembled on a dashboard along with other types of physiological data. It could be pressure, temperature, respiration, environmental and behavior. I also expect that we'll see the continuous ketone monitor and continuous glucose monitor combined onto a single device.

Now that we've covered ketones in various ways, I'd like to ask Dr Schatz for some closing comments.

Dr Schatz: Well again, thank you, David. Really, there's hope for the future. We have understood the pathophysiology of diabetic ketoacidosis, which is associated with insulin deficiency and excessive stress hormones, and these combined lead to a severe, life-threatening metabolic condition. And, it's important to know that appropriate fluids, insulin, and electrolytes given under guidelines will help resolve the issue. But most important is really the prevention. And part of that prevention is the measurement of ketones, which should preferably be done in the blood, although most people still use the urine. It is incredibly exciting that continuous ketone monitoring will lead to the prompt recognition of even early acidosis management and again, the prevention of DKA with its associated morbidity and mortality. 

Dr Klonoff: It's a pleasure discussing ketones. Thank you, Dr Schatz, and thank you to the audience. Please proceed to answer the post-activity assessment questions and take a moment to complete the program evaluation. Thank you.

This transcript has been copyedited.