Well, thank you for attending our conference on Beyond CGM and I'm going to be talking about dual sensing technology as the cornerstone of adult DKA prevention. I am Doctor Carol Wisham, clinical professor of medicine at the University of Washington in Spokane, Washington, and these are my disclosures. So Rich already talked about the burden of diabetic ketoacidosis in patients with both type 1 and type 2 diabetes, and just want to highlight that in this older study, 1 in 5 individuals with a diabetic ketoacidosis admission were patients with type 2 diabetes, and furthermore, there's a higher Mortality in patients with type 2 diabetes compared to type 1 diabetes. There is also evidence to suggest that a single episode of DKA and particularly recurrent episodes are associated with increased risk for MC, for advanced kidney disease, for cognitive impairment, and obviously for recurrent diabetic ketoacidosis. We do actually have newer data which demonstrates an increased risk in DKA both in type 1 and type 2. It originally appeared as if DKA risk was declining, but as of the most recent data, there is a further tick upwards in type 1 and. Type 2 and in this particular data set from England, 1 in 4 admissions for DKA was attributed to patients with type 2 diabetes. So obviously this remains an ongoing concern and it may be an increasing concern in patients with type 2 diabetes. So there is a review of who is at risk for diabetic ketoacidosis that which are reviewed. Just want to highlight the high risk clinical groups, including adolescents and young adults, which David will be discussing later. Patients with pregnancy, but also older patients, frail individuals with long standing type 2 diabetes are at increased risk. We know that certainly medical treatments such as insulin pump therapy. SGLT-2 inhibitors, and mental health disorders and treatment for mental health disorders, as well as chronic kidney disease can also lead to increased risk. And finally, there are social and behavioral drivers. Rural populations are associated with increased risk, perhaps related to inadequate resources for healthcare delivery. There's also issues related to social determinants of health that seem to be. Related to increased risk, as well as importantly, low carbohydrate and ketogenic diets. Rich already discussed, and I want to reiterate that amongst recurrent diabetic ketoacidosis, those account for 20% or more of admissions for diabetic ketoacidosis. So obviously, the risk for DKA is multifactorial, and we need to address all these issues as we try to work to prevent DKA. This is a more recent study which I think is interesting. It looks at DKA in a real world practice in Germany. So this study looked at 179 patients with 203 DKA episodes. The patient breakdown, you can see the majority of them were pre-existing patients with type 1 diabetes. Those are the patients we really should be able to most easily prevent DKA and. But also there were pre-existing type 2, about 14%, and there were some new onset type 2s as well. There were patients with LA as well as a small number of patients with pancreogenic diabetes. The risk patterns in type 1 diabetes is that 71% of the recurrent DKA occurred in patients with type 1 diabetes. 51% of them occurred despite the fact that they were doing continuous glucose monitoring. 24% of them occurred in patients on insulin pump or AID systems. 21% of them, only 21% actually had ketone testing available, but only 6% of them were using them regularly. The main triggers for DKA was poor adherence in 56%, as well as pre-existing infections in 32%. There were some patients who had euglycemic DKA from SGLT2 inhibitors. None of them reported receiving any education about ketone monitoring. There was a relatively high mortality occurring in about 2.3% of individuals, and this was highest in the patients who had type 2 diabetes and multiple other comorbidities. Finally, the key gaps that were identified were related to poor patient education. Inadequate testing and then lack of education on what to do when patients were sick or sick day rules as we call it. So really important that we prioritize the education to increase ketone monitoring, provide the education on what to do when patients are sick. This is what's going to be needed for us to prevent DKA. And getting back to some of the information that Rich already reviewed, just want to highlight that again in this particular study, in patients with type 1 diabetes in the type 1 diabetes exchange registry, which I would say represents a pretty highly educated group, they only 40% of patients would check their ketones when they're nauseated or vomiting, and fewer than 30% check their blood sugars when their blood And fewer than 30% checked their ketones when their blood sugars were over 300 mg per deciliter. And if you look at this breakdown across ages, whether it's due to symptoms or related to high blood sugar, they were high in all the groups, but they tended to be the highest in the oldest individual, those that are greater than 50 years of age. So this identifies again another gap in our education of our patients who are older and seemingly maybe we don't think about their risk of decay. So, really important to focus on that group as well. In another particular study looking at the fact that even though patients can be aware of their risk, they still don't necessarily act on it. So in this particular study, again, this was from an endocrine clinic. These were really well managed patients with diabetes. 62% of them had history of DKA and 93% of them were on CGM. They suggested that 95% of patients were aware of decay as an issue, but only 3% tested regularly. 30% of them misattributed their symptoms to other conditions. 39% said that they were regularly reminded to actually monitor ketones, but that didn't necessarily translate to action. The main barriers discovered in this study were poor symptom recognition, lack of equipment, and what was considered time constraints, and this tended to be the highest amongst the patients on insulin pumps and AID systems and those that were in the best control as measured by CGM. So, in summary, in order to prevent decay, we're going to need more than just continuous ketone sensing. Patients need to have specific awareness of their personal risk and recognize their early symptoms. They need to have the testing equipment available and test when symptoms or risk occur. And finally, and very importantly, they need to know how to respond to the test results and when to contact the healthcare team. The healthcare professionals also need to have specific actions. They need to identify their patients at risk. DKA and including those patients with type 2 diabetes, they need to educate their patients on clear guidance on ketone testing, recommend the appropriate ketone testing supplies, and then very importantly, this needs to be something that is reinforced on a regular basis. So at every visit or perhaps at least once a year, patients need to be reminded of the importance of ketone testing and knowing when to contact you when abnormal. So the most common precipitating factors for decay are something that both the people with diabetes, especially those that are on intensive insulin therapies, as well as their healthcare professionals need to be aware of. Those, those risks include missed or underdosed insulin, which of course could be related to the presentation of new onset diabetes, those with mental health disorders, and then insulin pump users where they might have site failures. They need to understand that in acute illness or injury, infections, trauma, surgery, major events like myocardial infarction or pulmonary embolism, those, those are also high risk for developing ketoacidosis. Medications, corticosteroids, high dose thiazides, sympathomimetic agents, atypical antipsychotics, checkpoint inhibitors, as well as SGLT-2 inhibitors, and then finally, those folks who are using cocaine or cannabis are also at increased risk for DKA. Now Rich reviewed a study to show that CGM lowers DKA events in type 1 diabetes. This is a more recent study that just corroborates what he reviewed, and that showed that there was in this study after CGM a 60% reduction in the DKA events. Now, why might this happen? Well, CGM does not detect ketones, so it's not that they understand that they have high ketones and that they're acting on them. Most likely it's related to them acting on a correction dose of insulin, which is only going to work if the glucose is truly high enough to trigger that correction dose. But it is true that the CGM does improve the visibility of the glucose and therefore allows people to know when their glucose levels are high and need to be adjusted. Now this is a study that we are presenting at the ADA this year and looking at CGM use in Medicare patients with type 2 diabetes who are on prandial insulin and showing how it is associated with a reduction in DKA events. So again, this is a fee for service Medicare population, and we looked at all of these patients over a 4-year period of time. Uh, we excluded patients with type 1 diabetes, end stage renal disease or patients less than 65. And what we found was that 1.6% of these individuals with type 2 diabetes had one or more DKA events versus when they started CGM, that number dropped to 0.7%. So this again is almost a 60% reduction in the risk of DKA associated with CGM, much like the patients with type 1 diabetes. And again, much like the patients with type 1 diabetes, these patients were probably triggered to take additional insulin when their glucose levels are high, which further lowered ketones that might be present. But as Rich already described, ketone levels can be elevated before the glucose is at range, as demonstrated in this particular patient. So it is important to know that CGM is not going to catch all of the patients with ketoacidosis or ketosis because some of those may have elevated levels before the glucose levels actually trigger a correction event. And in this again recent study looking at the correlation between blood glucose and beta hydroxybutyrate rise or risk for DKA, they took patients with type 1 diabetes and did insulin pump suspensions and then measured their beta hydroxybutyrate, and what they discovered was that 89% had what would be considered an elevated beta hydroxybutyrate of 0.6 millimole per liter after five hours of insulin pump suspension. 44% reached 1 millimole at 7 hours, and the beta hydroxybutyrate reached that elevated level of 1, while glucose levels were still less than or equal to 250 mg per deciliter. So it was important to know again in this particular study that you could have glucose levels that might not trigger a correction value and still have elevated ketones. Now, the 3 presenters at this symposium all participated in this international consensus paper where we came up with a recommendation for application of this new technology of continuous ketone monitoring. The first molecule that is formed in the development of ketosis is acetoacetic acid. And that is then metabolized either through reduction to the primary ketone body, which is beta hydroxybutyrate, or with spontaneous decarboxylation can form acetone. And as you can see, the amount of beta hydroxybutyrate compared to acetone is markedly higher, and this is important to understand that the measurement of acetone is not necessarily going to accurately reflect what's going on in terms of the total amount of ketosis. So again, Rich, described the current methods of monitoring ketones, what their pros were, what their limitations were, and I just want to highlight that by using urine ketone testing, you could delay the recognition of early ketosis, as well as not being able to follow the resolution of ketosis because it clears more slowly. So beta hydroxybutyrate with current measurement of using a finger stick, has the advantage of being able to identify the early ketosis as well as being able to follow resolution, but it involves a finger stick and involves carrying a meter and carrying strips, and most of our patients now, especially with type 1 diabetes, are on CGM. They do not necessarily have or want to be carrying any additional equipment around. I thought this was an interesting figure demonstrating the different levels of beta hydroxybutyrate that might develop under certain physiologic states, and as you can see between the Fed state, early and later fasting and ongoing fasting, you can see. Increases of levels that would be consistent with the diagnosis of DKA that is exceeding 3 millimoles per liter with additional stress on top of that, you can get further rise in the beta hydroxybutyrate and then finally extremely high levels in cases of diabetic ketoacidosis. So, getting back to the consensus panel's description, the first thing they really wanted to describe is what kind of thresholds should we be looking at to describe the person's risk for diabetic ketoacidosis. So these are the thresholds they came up with. So, anything that was under 6 millimolar would be considered normal. Between 6 and 1.5 would be considered elevated ketones. Between 1.5 and 3 would be considered high ketones. And greater than equal to 3 would be considered urgent high ketones. They they reviewed the symptoms that patients might look at, look for in terms of the different levels of ketosis and then the actions that should happen and a lot of that revolved around the use of um liquids, carbohydrates, and insulin to block the formation of ketones and then recommended a course immediate attention. Whether it's in the ER or whether it is uh reaching out to your healthcare professional, they recommended, you know, at the very high levels that they, they have some kind of input from their healthcare professional. They furthermore went into more detail for what patients might be able to do for home monitoring, and I'm not going to go through all of these, but they, you can see as the ketone levels increase and the glucose levels increase, there's increased fluids recommended. They recommended that you take, that they take the usual. Insulin to cover their carbohydrates, but as the glucose and ketones increase, they recommend relatively more supplemental insulin as described by percent of their total daily dose, so 5 to 20% depending upon the level of ketones and glucose. I think this is a really helpful table that one could provide their patients in order for them to be able to manage their ketosis at home, and even mild ketoacidosis can be managed at home as long as the patient has the information on how to do so. They can keep food and fluids down, and they know when to go to the emergency room. We also suggested there be specific graphics that be included in whatever reader or whatever is showing the ketone levels, and that is increasing, stable, or falling, much like we have on CGM. The level that we recommended would be significant would be greater than or less than 0.4 millimole per liter per hour. Now, putting all this together, there is a vision for an integrated continuous glucose and ketone monitoring system involving one ketone glucose monitor. A mobile app or some reader that would be able to read both the glucose and the ketones, and in this case, the glucose is on the top and the ketones is on the bottom. That would ideally be associated with cloud software which would allow a healthcare professional to evaluate what's going on with the patient real time if they are contacted and there is an issue or in follow up when they come into the office to review what happened when the patient had ketosis and how they responded to it so further education could have. So further education could be done. Importantly, they also include a caregiver app, which is going to be very important for those that are at extremes of age so that a caregiver could monitor and make sure that the patient is remaining safe. Now what are the benefits of continuous ketone monitoring? Obviously there are several, as we've discussed, we can identify early rise in ketones, early warning of insulin delivery failure, very important for patients on insulin pumps, as we know this happens very frequently in clinical practice, can prevent acute hospitalizations. The high-risk patients may benefit the most and we want to really start with, I'm probably getting this in the hands of all patients with type 1 diabetes, but let's not forget the patients with type 2 diabetes, long duration, those that have multiple daily insulin doses and now we have many patients with type 2 diabetes on insulin pumps. This has the advantage for the patient as they don't have as much equipment to have to carry around because it's going to be with them at all times. And then most of the insulin pump manufacturers have already stated that they want to utilize the information from a continuous ketone monitor into their algorithms to help give patients further information on how to address the ketone measurements. Benefits aren't without challenges, and so there's several challenges and some unknowns about what this is going to mean for us. So obviously patients have to be educated so that they know what the numbers mean. You don't want a phone call every time somebody hits 0.7 millimole per liter. The healthcare professionals, they need to understand it so that when their patients do contact them, they know how to respond. We need to understand the difference between physiologic and pathologic. We need to make sure that whatever we're using represents accurate measurements, and luckily, the data suggests that's the case. What do we do with patients with type 2 diabetes versus those with type 1? And then of course, how do we manage your glycemic DKA and furthermore, we need to be concerned about reimbursement. So again, we need to address all of these challenges to fully appreciate the benefits of continuous ketone monitoring. So to finish up, I'd like to talk about a case, this is a case that was published not too long ago. Um, so this was a continuous ketone monitoring. So this was a patient who was in a study for continuous ketone monitoring, and she was type 1, she was on SGLT2 inhibitors and a low carbohydrate diet, and she participated in regular physical activity. And so what you can see on day 113 of the study, if you look at the bottom where her glucose levels are, you can see there's only a very small amount of time where her glucose levels are above range. The first one, which occurred somewhere around 11 o'clock, um, she was, again, remember she's on a low ketone diet, but she exercised, her glucose actually went up and her ketone levels went up, and her ketone levels approached that 3 millimolar that as Rich described, defines the potential for a diabetic ketoacidosis. Spontaneously, the levels came down, but they fluctuated in the elevated range all day. She went to bed. The levels went up, and she awakened the next morning with a level that was around 2.2 millimolar. Again, the levels of the ketones continued to fluctuate through the day despite glucose levels that were within target range all through the day. Now, what triggered her to take carbohydrate and insulin at 10 o'clock is not entirely clear, but she did respond. She said she took some carbohydrates, she took some insulin. And you can see how quickly the ketone levels came down, despite the fact that glucose levels actually went up from the ingested carbohydrates. So once again, this particular study showed that there were 14 hours of ketosis with a peak level of 2.9 millimole and showed how well and quickly you could measure resolution with the continuous ketone monitoring. So again, patients with type 1 diabetes by themselves are at high risk, but each of these other issues, the automated insulin delivery, the SGLT-2 inhibitor, low carbohydrate, and increased physical activity all contributed to the fact that this woman was demonstrating elevated ketone levels throughout the day. So, in summary, it is important to understand that DKA is preventable and although we have shown a significant decrease in risk for DKA in patients with both type 1 and type 2 diabetes, the risk does remain. We will hopefully shortly have the ability to measure ketones real time, and this can obviously give us additional insight and give the patient additional insight to reduce their risk for DKA. So dual ketone. So dual glucose ketone sensing may reduce DKA risk by alerting the patient with diabetes of early detection of rising ketones, enabling them to give timely supplemental insulin and fluid intake. It allows them to identify site issues earlier, but they need to understand, again, clear guidance on when to contact the healthcare professionals. So education of this patient is essential. They need to understand how important it is to test ketones, how to interpret the results, and when and how to act. So early detection and informed action can prevent DKA.
Related Presenters
