Good morning, everyone. Thank you so much for joining our session today. My name is Emily Perriello. For my portion of this talk, I'm going to be discussing diabetic ketoacidosis or DKA, but our session is all about putting a new twist on some of our critical care classics. So in keeping with the times, I know keto and sugar-free things are all the rage nowadays, so I figured sugar-free DKA would be a great topic to chat about this morning. Per the guidelines for this presentation, I had to include this slide, so hello, this is me. I'm an assistant professor in the Department of Pharmacy Practice at the University of St. Joseph School of Pharmacy and Physician Assistant Studies here in Hartford, Connecticut, and my practice site is in the Hartford Hospital Emergency Department. So most of my perspective is from the emergency room, and I am happy to share that perspective with you all today in the treatment of these patients. I have no financial disclosures or conflicts of interest associated with this presentation. Our objectives for today, we are going to be able to define euglycemic DKA or EDKA, as you will see it abbreviated throughout the presentation. We're going to describe some risk factors and pathophysiology associated with EDKA, summarize the diagnosis and treatment options, as well as identify some prevention strategies for patients who are at risk of developing EDKA. So let's talk DKA. I really want to go into some in-depth review of our guidelines and our treatment algorithms, and I think if we really break this down. No, I'm totally kidding. I know that I am speaking to a bunch of critical care professionals. I don't want to bore you all with this information. So we are going to jump right in to what is euglycemic diabetic ketoacidosis. So our big difference here is going to be the relative normal glycemia that you see with EDKA. Typically, normal glycemia, we mean less than 250 milligrams per deciliter. Some references go as high as 300. What you're not going to see is the typical, highly elevated, over 500 blood sugars. But everything else is pretty similar. As you can imagine, both of them have acidosis right in the name. APH of less than 7.3, a low bicarb of less than 18 milliequivalents per liter. We also expect to see ketosis with both of these conditions. And our signs and initial presenting symptoms are often very similar. You can see Kussmaul breathing. Nausea, vomiting is very common. So really, our big differentiating factor here is the normal glycemia. Now, when I say euglycemic DKA, what is the first thing that I imagine most of you are thinking about? And it is our sodium glucose cotransporter 2, or SGLT2, inhibitors. These drugs came on the market, and now our EDs and our ICUs have seen more and more of these patients popping up. However, this isn't actually a new phenomenon. Back in 1973, Monroe et al. described actually the first case series that looked at euglycemic DKA. So they were looking at regular patients who they described as having severe diabetic metabolic decompensation. But what they noticed was that a cohort of these patients, 37 in total, had a blood glucose of less than 300 milligrams per deciliter. Ultimately, they were managed very similarly to our other DKA patients with fluids, insulin, electrolytes, etc. And what they wanted to do was look at some of the causes. Why did these patients have this presentation? And they saw a lot of the common things we associate with DKA today. So decreased carbohydrates, pregnancy, infection, anything stressful that can tip the scales or mess with the balance of glucose and insulin in your body. Since then, many other reports have come out describing EDKA. So it's clearly not just our SGLT2 inhibitors that are causing this. So how does this really come about? When we look at our pathophys, generally, you are going to see an insulin deficiency along with insulin resistance. So this is similar to traditional DKA where you get less insulin. Therefore, you're going to break down your free fatty acids, we get to ketosis, etc. The same can still be true for EDKA. But something that sets EDKA apart is you might have decreased glucose availability, which is generally the opposite of a hyperglycemic DKA. This could be during a fasting state that's usually associated with some stressor, it could be infection, or even back in that very first case study, one patient had a dental abscess and he said he couldn't open his mouth because it was too painful and he wasn't able to eat. So any of these precipitating factors, in addition to all of our normal counter-regulatory hormones that we see in traditional DKA, so glucagon, our catecholamines, cortisol, growth hormone, all of that can put you at risk. The other thing that we see in this condition is increased urinary glucose excretion. And speaking of urinary glucose excretion, let's talk about where our SGLT2s fit into this picture. So this is your nephron, or this is more so a rough sketch of a nephron that I drew in PowerPoint. I did my best. So this is SGLT2. 90% of glucose reabsorption happens via SGLT2 in your proximal tubule. There's also an SGLT1. It hangs out in your distal tubule and it takes care of the remaining 10%, but we're focusing on your SGLT2 here. Our inhibitors are those things that end in flozen that I have listed on the slide. These are the ones available in the United States, but there are different ones available in other countries. And so by inhibiting SGLT2, we increase our urinary excretion of glucose. So how do these contribute to DKA? Well, when we see increased urinary glucose, we have less circulating glucose. Therefore, less insulin is produced by your body, and that in and of itself predisposes us to a ketogenic state. Also, when patients start these medications, they are more likely to reduce their insulin doses to compensate for the new urinary excretion of the glucose. We also see increased glucagon production with these medications, and we know that precipitates hepatic gluconeogenesis, glycogenolysis, and lipolysis. But the mechanism for why these agents increase glucagon really isn't clear. We also know that just based on this mechanism alone, we increase our urinary glucose excretion. So it's likely a combination of factors that contribute to the DKA itself, but the increased urinary excretion of glucose is likely what leads to the euglycemia that we see. These drugs have been around since about 2013. They gained a lot of popularity because of their cardiac and renal benefits. Then in 2015, after some of these reports of EDKA to the FDA adverse event reporting system came out, the FDA puts out this drug safety communication. And it says FDA warns that SGLT2 inhibitors for diabetes may result in a serious condition of too much acid in the blood. And after this communication came out, the labeling for all of these drugs was changed to include this warning. Then in 2016, the American Association of Clinical Endocrinology and the American College of Endocrinology followed this up with their own position statement after reviewing the literature. They stand firm that this is still very infrequent and prescribing of the SGLT2 shouldn't change because they show such great benefit. The position statement does, however, provide some guidance on how to diagnose and manage our EDKA patients associated with these medications. So diagnosis. What do we do when we have a known diabetic present to the ED or ICU and they're looking sick? DKA is always going to be on your differential. So we check a finger stick. The finger stick comes back at 200 or so. We're going to send off some formal labs to make sure that's correct. But then DKA is probably dropping pretty far to the bottom of your differential. Now with it being so common to have people on SGLT2s, this is something we need to keep in mind so we don't delay critical care. So diagnosis is key. This is a big take home point and let me tell you why. There's not a ton of literature out there on EDKA. It's lots of case series, case reports, etc. Nothing is super robust, but we do know that's oftentimes the case with our rarer disease stage such as this one. So we're left with this just kind of hodgepodge of smaller case series and case reports. What the literature is showing is a delay in the recognition of this disease state. So in this particular case series, we had 10 patients between their late 30s and early 70s, so a pretty broad range. They were all on either conagliflozin or dipagliflozin and they had a relative normal glycemia. Now I know some of these, and it depends on your hospital and your unit, a couple of these values would probably flag as high. The 403 is a little high to me, the 347 perhaps. But most of these, a 149, that's not going to be a patient that I'm super worried about DKA in. Ultimately, all of these patients were found to be acidotic. They had an elevated beta hydroxybutyrate level and were ultimately diagnosed with DKA. Some common triggers they identified were a Roux-en-Y gastric bypass or other surgery, infection, or just profound nausea, vomiting. One case did not have a trigger identified, but DKA was diagnosed in all of them. Big takeaway from this is that in greater than 50% of the cases, diagnosis and treatment was delayed due to these relatively normal glucose levels. The authors didn't specify how long the delay was, but we all know our complications of acidosis. Your body doesn't like being acidic. You get vasodilation, cardiac decompensation, respiratory issues, et cetera. This series didn't go into specifics on what the actual complications of the delayed diagnosis were. So that brings me to my next study. And as I mentioned before, most of our available literature is case reports. So there's a lot of stuff that says we diagnosed a case of EDKA. Here's what happened. Here's what might've caused it. Here's what we did about it. The commonality is that all of these are talking about a delay in diagnosis, but not what happens when you delay the diagnosis. So the few case reports that I found that do go into this are summarized on this slide. So this first case was admitted for unstable angina with a plan for cabbage. There was a peri-op exam done and it was determined she needed a tooth extraction. It was unclear why the tooth extraction was needed. I don't know if maybe it was for an intubation for the surgery, but ultimately tooth was removed and this caused decreased PO intake. Seven days later, she developed tachypnea, vomiting, hypotension, and an anion gap acidosis with a normal blood glucose of around 178. All kinds of supportive therapy were initiated at this point. She needed to be intubated. Vasopressors were started and a bicarb drip was started as well. She ended up on CRRT for the acidosis, but nothing seemed to be helping until a beta-hydroxybutyrate was checked on day two of her ICU admission. And this is what led to our diagnosis of EDKA. Insulin and dextrose were started. She subsequently underwent impella placement and VA ECMO in addition to a PCI and a stent for a subsequent MI that occurred. She also had a hemorrhagic stroke on day 10 of her ICU admit. So in this case, it's unclear whether the DKA or the cardiac ischemia occurred first. What we do know is acidosis negatively impacts the cardiovascular system causing vasodilation, which may have contributed to the ischemia. And the patient sat for at least a day in the ICU before the diagnosis was made on day two. For our second case, this person was admitted for a GJ bypass for duodenal cancer. They were on NPO for surgery. And then this person was again admitted to the ICU with an anion gap acidosis. And similar to our first case, the diagnosis of DKA wasn't recognized until a beta-hydroxybutyrate was checked on day two because blood sugars were otherwise normal. This person ultimately required a pacemaker placement after a cardiac sinus arrest. And here they believe the acidosis was probably the cause of the sinus arrest because it resolved once the acidosis resolved. So in both of these cases, the authors felt that the DKA contributed to the life-threatening complications that were seen. Now in case three, you'll see ipragloflozin is noted on the slide as the offending SGLT2. This is not available here in the United States. This case was over in Japan where that is available. This patient collapsed while sightseeing due to a V-fib arrest and his labs at baseline, so when he first came into the hospital, were acidotic with a normal sugar. So from this information, the authors felt that he didn't have an MI that caused the arrest, and then he got DKA subsequent to that. Because his labs were abnormal on admission, they felt that the DKA was what contributed to the MI. The MI caused the arrest. Now this is all based on author assumptions of what came first. Lots of chicken or the egg types of scenarios here. However, the first two cases, DKA wasn't diagnosed until ICU day two after the patient was already acidotic for a while. So we know we have increased ICU stays and potentially unnecessary ICU admissions if our diagnosis is delayed, along with prolonged acidosis, which our body doesn't like. So the big takeaway here, I will reiterate from a previous slide, diagnosis is the key. So how do we not miss this diagnosis? How do we make sure that in our practices that we are catching this as soon as possible? First off, always get a medical history. DKA is more common in our patients who have type 1 diabetes, but the SGLT2s aren't approved for type 1. They might be used off-label for this, but oftentimes type 2 diabetics are who's presenting with these, so it's important to look out for that. So what type of diabetes do they have? Obviously, what medications are they on? We all know this whole thing's been about the SGLT2s, so don't forget to look for those. And then our symptoms. Lots of these case reports report nausea and vomiting like we see with traditional DKA, and we're also going to want to make sure we're looking for our usual DKA triggers. Alcohol use, liver disease, starvation, pregnancy. Have they been in PO? Are they under stress, etc.? For our labs, we're not reinventing the wheel here. Get all the same labs that you would for normal DKA. We want a pH. We want to know that they're acidotic. You want a beta-hydroxybutyrate or any serum ketone that you measure at your institution, and you want to know their anion gap in addition to whatever metabolic panel you run to make sure we're getting all of our electrolytes. And the big thing here is not to get confused by the confounders. So don't get steered off track by a normal blood glucose, and it's important to note that patients who are on SGLT2 inhibitors may actually reabsorb urine ketones. So serum ketones are very important here because those urine ketones might not be super reliable. For our treatment, once you catch the diagnosis, treatment is easy peasy. It's pretty much the same as regular DKA. You want crystalloid fluids. Ideally, we like a nice balanced fluid like LR. We don't want to flood a person with normal saline and end up with a hyperchloremic metabolic acidosis complicating things down the line. Of course, we want to give insulin. Same doses for regular DKA. The 0.05 to 0.1 units per kilo per hour. Of course, you want to monitor potassium anytime you're starting insulin and supplement that beforehand if necessary. Dextrose is your big thing to remember here. These patients are euglycemic. You're going to need dextrose likely right away when you're starting your insulin therapy. If D5 isn't cutting it and they're getting hypoglycemic, you can bolus as necessary and increase your maintenance to D10. This is just like regular DKA. You still need the insulin to resolve it, but don't forget about the extra dextrose requirement. And the great thing is AACE and ACE recommend the same treatment doses as in the current DKA guidelines. So crystalloids, you want one to two liters in the first one to two hours. Once you get to your maintenance fluids, 150 to 250 mLs an hour sounds great along with the D5. Don't forget the dextrose. And keep an eye on all of your electrolytes. We typically think of potassium dropping, but phosphorus has a tendency to decline here as well. So I know I said that we weren't going to talk about this, but we are just going to briefly mention it so I can touch on where our consistencies are with regular DKA. So like I mentioned, fluids, great. We want to determine hydration status and get them fluid right away. You don't really have to worry as much about a corrected sodium here because they should be relatively normal glycemic. For insulin, same thing, same dose. Keep that looking great. Potassium, don't forget about your potassium and supplement as necessary. Bicarb, there's no real data to support bicarb use here. Generally, we're going to use it if your person's hemodynamically unstable and we want them to respond to some catecholamines. Or if they're very hyperkalemic, then sure. Otherwise, we don't have a ton of great data. Again, our big takeaway here, when the serum glucose reaches 200 in traditional DKA is when you would start adding dextrose back into your maintenance fluids. But here, these patients are likely going to present with the sugar already in this range. So don't forget your sugar. So then the big question is, what do I do with the SGLT2 afterwards? There's not a ton of great info here. Many references say you should discontinue the therapy when EDKA is diagnosed and initiate DKA treatment. But what they don't specify is if this is a permanent discontinuation or if a retrial later is acceptable. So the references I found that mention it long at all has a great review article on it and they recommend restarting after the resolution of DKA. They say that's fine. The European Medicines Agency says don't restart it unless another cause of DKA is identified. And our AACE and ACE position statement just doesn't address it. If you look at your SGLT2 drug labels, they don't list it as a contraindication. So this isn't super helpful. I get that. I would go on a case by case basis here. Have the risk benefit discussion with the patient. Can you identify another cause of what might have precipitated the DKA event? That is when you might want to consider either restarting or permanently discontinuing depending on the patient scenario. So this doesn't happen. It is important to remember prevention. Make sure that you are stopping your SGLT2 inhibitors 24 hours before a potential trigger. So if the person's going to have an invasive procedure, if they're going to have extreme physical activity, etc. If your patient's going to go run a marathon tomorrow, probably want to stop the SGLT2. If the patient is undergoing emergency surgery or if they are NPO for longer than you anticipated and you can't stop it 24 hours ahead of time, just discontinue it and then monitor your serum ketones. That is going to be a big point to help you keep an eye and catch this before it gets ahead of itself. And you need to start with some more advanced therapies. Patients should obviously be avoiding excess alcohol and also your very low carbohydrate or ketogenic diet. So I know keto is super hip right now, but it's not really appropriate for these patients. So to summarize, EDKA is a rare but serious adverse event that's been associated with the use of our SGLT2 inhibitors. Diagnosis can often be delayed, causing increased morbidity. Treatment is very similar to DKA, but don't forget your dextrose. Prevention through discontinuation of the drug and appropriate monitoring is imperative. And it's also great to recognize it early if the preventative measures weren't taken. So if your patient's going for a big surgery, don't forget to hold the SGLT2. And at the same time, if they're in your ICU or your ED and you know they're on this SGLT2 and they just had the big surgery or they haven't been eating or they just ran the marathon, ask if they stopped their SGLT2. Check your anion gap, your blood gas, your ketones. Recognize where others may have missed the preventative measures so you don't miss the diagnosis. And that is all I have for you. I am pre-recording this, so I believe there is a way for you all to ask me questions during the conference. But if not, my email is on the slide here for your reference, and I am happy to entertain questions that way as well. Thank you so much for your attention.

euglycemic diabetic ketoacidosis

EDKA

ketosis

acidosis

blood glucose levels

insulin deficiency