 Hi, I'm Cindy Patel. I'm a medical oncologist here at the University of California, San Diego, and I'm joined here by my colleagues, Dr. Karen Yoon and Dr. Karen McAllen. And today we're going to be discussing work that Dr. Yoon actually led as a resident and now a fellow on the rapid onset type 1 diabetes that can occur with immune checkpoint blockade, in particular with anti-PD1 therapy. This was work she published in OncoTarget on May 1st, volume 11 at page 2740. And so with that, Dr. Yoon, do you want to talk a bit about what led you to want to do this project and how this particularly affects patients? Yeah, thanks Dr. Patel. You know, as immunotherapies being used more for different cancer treatments, we decided to look at type 1 diabetes because this is a rare immune related adverse event, and we wanted to kind of characterize the incidence and the kinetics of this disease. There's been case reports about this immune related adverse event with about 0.2 percent of patients getting this wear side effect on immunotherapy, and we wanted to look more into that, trying to identify certain risk factors, identify any kind of lab or imaging abnormalities that could pretend risk for patients who would end up developing this side effect. And so what we did with our research was we looked in our electronic medical record and identified 1,327 patients who had received immunotherapy between the year of 2013 to 2018. And from those patients, we further narrowed from those patients, we then identified patients with a blood glucose of greater than 200 and a bicarb of less than 18 to see from there of whether or not we can extract from those patients who had developed type 1 diabetes related to immunotherapy. And so how we kind of then proceeded was to then look into the charts of these patients, and we identified five patients who were found to have type 1 diabetes related to immunotherapy, and that was an incidence of 0.38 percent of patients that we looked at. And so in terms of the characteristics of the patients who had type 1 diabetes, four had new onset diabetes, and one patient had a history of type 2 diabetes and then determined to have type 1 diabetes related to immunotherapy. And then in terms of the things that we were looking for, we were trying to see whether or not there were any trends in hyperglycemia. We looked at amylase and lipase levels to evaluate for any preceding pancreatitis prior to the diagnosis of type 1 diabetes. And we also looked at CT imaging prior around the time that they were diagnosed with diabetes. And so from our results, we found that of the five patients, four of them actually had elevated A1C levels at the time of diagnosis at a mean of 9.1 percent. And for most of our patients, the A1C remained persistently elevated over time. Looking at the glycemia levels prior to the onset of diabetes, we didn't see any persistent trends or elevated blood glucose levels prior to the onset of diabetes. And there was no significant trend in amylase or lipase levels to suggest any pancreatitis. And looking at the imaging prior to the onset of diabetes, there was no evidence of pancreatitis prior to the diagnosis of diabetes. And so we also evaluated autoantibodies and found that two of the patients had GAD positive antibodies. And so these are autoantibodies that can be seen in autoimmune diabetes. And interesting enough, it was these patients who at the onset of immunotherapy initiation, those who are GAD antibody positive were found to have diabetes onset earlier than those who are GAD antibody negative at 20 days and at 106 days versus those who are GAD antibody negative who develop diabetes more than a year later. Great. Dr. McAllen, so you've treated a lot of patients with diabetes. Can you describe in your practice, in terms of the types of fulminant diabetes cases Dr. Yoon described, how they differ? How are they similar? And really what can, you know, practicing oncologists, hospitalists, ER docs, what are the take home messages on how we can try to keep these patients safe with a rare but devastating side effect? So I think the first thing is that the presentation tends to be extremely fulminant, as Dr. Yoon said. And if you look at patients who haven't received any kind of immunotherapy, but just have typical type one diabetes onset, they're also often extremely fulminant. So most of these patients with their new diagnosis of type one diabetes, be it from immunotherapy or not, get admitted to hospital and treated for their diabetic ketoacidosis and they get treated with fluids and insulin and then they get discharged. And they're usually seen in the office by me or one of my colleagues a week or so later. And as the weeks go by, the patients take quite two quite different paths. So a typical patient with type one diabetes has what we call a honeymoon period, where they get substantially better and their insulin doses are much reduced. But you guys as patients, the oncology patients, they start off bad and they don't get any better. And it tends to be a very stubborn form of type one diabetes. I think since you wrote your paper, I've seen many more patients and all of them I've seen one person who seems to have had a sort of a honeymoon. But almost everybody remains with this brittle, very difficult to manage diabetes. I haven't seen anybody get off insulin. And typical type one patient who hasn't received immunotherapy, there may be, they may get off insulin for six months or more. The other thing that I've noticed is that actually, contrary to the literature, the literature would suggest that the immunotherapy patients do have typical insulin requirements like a standard type one, which is, you know, a half to three quarters a unit per kg. But I have a couple of those patients who are requiring substantially more. So, and then the other major difference in our research, GAD antibody was not the norm. But if you have a patient who develops type one diabetes without immunotherapy at diagnosis, GAD antibodies or other autoimmune markers, that would be the norm. So this is a little bit, it's an atypical diabetes. And it's probably the worst side effect, the worst endocrine side effect you can get from immunotherapy. I see a lot more thyroiditis and a lot more pituitary problems. And those are so much more easy to manage. But the type one diabetes is it's a difficult disorder. Absolutely. And so from the medical oncology side of things, this is actually one of the more difficult diagnoses to make. Normally, we'll get a spot glucose, for example, when a patient gets their immunotherapy treatment, but we rarely catch these patients at the time at which they're getting their spot glucose at the time of their infusion. They typically present in ER, vague symptoms and their blood sugars are in the several hundreds, they're acidotic. And so they really present them as you guys have so eloquently described with this really fulminant full on DKA. And our experience has been that exactly Dr. Mark Howe and that these are very difficult to treat patients. Now, one of the questions is these patients are on cancer immunotherapy for a life threatening cancer. And many of these patients, and there is this unusual association that those patients that get some of these rare autoimmune conditions immune related adverse events or IRAs, as we describe, are ones that actually just derive durable benefit from their therapy. And part of the reason we're able to manage their diabetes for so long and track they one sees is a disease that in other terms may have just had a survival measured in months. Now they're actually living years and in the sense that diabetes is a chronic illness, their cancer to some degree has become a chronic illness as well. They're now complicated by a relatively brittle diabetes, high insulin requirements. And in fact, you know, one of the challenges we face is we're going to pause their immunotherapy, right? And this obviously is not an immune really adverse event that responds to steroids. In fact, the hyperglycemia is actually more pronounced, right? So this is actually a paradoxical immune really adverse event for most of us where our gut reaction is trying to tamp in the immune response. And so for both of you, maybe a divine commenting, you know, if you were the ER doc or you're the person on consults, how do you think about steroids? And for me for you, Dr. McAllen and Dr. Ian, how do you think about re-challenge of immunotherapy for these patients? So I'll start with the glucocorticoids. I think there has not been any evidence in the literature that use of glucocorticoids does anything other than wreak havoc with the blood sugars. So I give steroids a big thumbs down. And as to your second question, I know Dr. Yoon is going to answer it, but endocrinologists come up against oncologists trying to stop their wonderful therapies because of endocrine side effects. Don't ever do that because we endocrinologists can treat her treat through. So you guys should always do what you need to do, but I'll let her answer. Yeah, it's from this from our studies. So for patients who had resumed immunotherapy after the diagnosis of diabetes, actually all of them had progression of disease. And so I don't think the diabetes will preclude them from being back on immunotherapy. I know the NCCN guidelines recommends that holding immunotherapy until blood glucose are in the normal range, but I don't think that the diagnosis of type 1 diabetes related to immunotherapy should necessarily preclude someone from continuing their treatment. And in a sense, Dr. McAllen, the damage to the beta is already, right? They already sailed immunologically. And so sadly, I guess, the ability to resume. Now, one of the points I do want to make about immune-related adverse events is it's roughly the rule of thirds, meaning if you do re-challenge a patient with immunotherapy, 30 patients will actually not have an additional IRA if you look across data sets. A third of patients will have the same IRA, though at this point the damage to the beta islets, beta islet cells has already happened. And so it's unlikely to make the diabetes more brittle than it already is for most patients. But then a third can actually get a different immune-related adverse event, right? So they can get pneumonitis, myocarditis, thyroiditis. And so I think the discussion to have with your patient really is for this particular immune-related adverse event, once it happens on continuation, is really driven by the risk of a new immune-related adverse event that could happen because, sadly, the ship has already sailed in terms of the neuropathy here. And so with that, maybe I can go around and ask each of you your final thoughts, maybe a final pearl as you take care of these patients and what we can leave our readers with in terms of diagnostic and management strategies. And congratulate Dr. Yoon on her work as a resident now, a fellow here. Yeah. Yeah. So I think, you know, in terms of the research, I think some of the teaching points that we got from this was in terms of trying to identify patients who are going to end up developing type 1 diabetes-related to immunotherapy, it's important to always assess for symptoms of hyperglycemia and DKA, like polyuria, polydipsia. It's really going to be hard to, because it happens so quickly and so rapidly, it's going to be hard to follow A1C levels or glucose levels to really identify those who are going to end up developing type 1 diabetes. Dr. McAllen? Yeah. I mean, I think patient education is really the only way to go, but I do think that a lot of these, many of your patients are older and have so many comorbidities that are trying to decide why they don't feel well in the morning is really, really difficult. And because the incidence of this is not high, it's less than 1%, it would seem it's really not feasible to have everybody checking their blood sugars. So I think you're relying on patient education and that's going to be difficult, but there is a certain mortality from acute diabetic ketoacidosis, not specifically related to immunotherapy, but, and there have been studies that show that if you educate people about symptoms of DKA, you can reduce the mortality. I think there was a study from Italy, but I think this is a different population because most of the populations that we're talking about are healthy young people and this is not that population. So it is tricky, but I agree with Dr. Yoon. Great. Well, thanks to my colleagues, Dr. McAllen and Dr. Yoon, and appreciate your time and this great work. Thank you, Dr. Yoon. Thank you. Great. Thanks. Thanks.