Thank you very much for the kind introduction. And my task is going to be addressing the glucocorticoid-induced hyperglycemia, and I've chosen to discuss the pathogenesis, management, and prevention. I have no complex of interest in relationship to the contents of this presentation, and I thought when I set the presentation to address the impact, the recognition, the practical management, and prevention of glucocorticoid-induced hyperglycemia in people with diabetes and those without diabetes. So these are the outlines. We'll go through them as we go along. So to start, let us address how glucocorticoids affect blood glucose levels, and I think some of the presentation may sound simple for people, but I think some of it is meant for the completion. The literature you can find, people will be talking about steroid-induced hyperglycemia, steroid-induced diabetes, glucocorticoid-induced diabetes, glucocorticoid-induced hyperglycemia. Ideally, if you are talking about induced diabetes, it means it is hyperglycemia in the diabetes range, and hyperglycemia could be just high blood glucose less than the diabetes range, or in somebody who is known to have diabetes, and when you have actually given them steroids, you have caused them to have hyperglycemia. Some of the literature will also include the new-onset diabetes after transplantation as part of this complex, but I think for today's discussion, I'm not going to be including that. It's always nice to check what people have been doing, so this is a graph I downloaded last night from PubMed, and this just shows the increasing interest in the steroid-induced hyperglycemia, and I put in all the possible search points there, or terms there, and you can see that there has been, since 1960 until this year, there has been continuous rise in the steroid or glucocorticoid-induced hyperglycemia. The impact of glucocorticoid-induced hyperglycemia is fairly serious, and I have seen in my own clinic patients who were admitted, and they have high blood glucose throughout their admission, and only discovered by their maid doing their flash glucose monitoring, and we downloaded that. Throughout their stay in the hospital, the blood glucose was running between 200 to 300. So the scenarios that have been studied in relationship to this subject includes many scenarios, including generally hospitalized patients, patients with cancer, patients in end-of-life, naturally pregnancy, neurological conditions treated with steroids, rheumatological conditions, and a lot of focus has been on the post-transplant scenarios. The impact of glucocorticoid-induced hyperglycemia, if we take just one example, for instance, the post-transplant patients, it reduces survival, it increases the rejection, it increases the risk of infections, and increases the risk of cardiovascular events. And you can translate this into many of the other conditions. The pathogenesis, I lost, I tried to find, last night I tried to find the reference where I got this diagram from, but I think I could not find it in my files, but I think I just made some modification in it to show the impact on insulin deficiency above the blue line, and impact on insulin resistance below the blue line. For instance, at the pancreas, there will be decrease in insulin synthesis, increased beta cell apoptosis, or destruction. At the liver, there is increased gluconeogenesis. At the muscles, there is reduction in glycogen synthesis, reduction in the glucose uptake, and increased breakdown of muscles and proteolysis. At the fat tissues, there is increased lipolysis, decreased fat accumulation in peripheral stores. All of these, as you well know, that actually will cause increase in the blood glucose. There is another way of looking at this. You could actually look at the, on your left-hand side, the glucose influx, increasing the glucose influx into the blood, caused by the increasing gluconeogenesis, increased appetite, and also increased effect on glucagonepinephrine. On your right-hand side, you can see the impact on glucose removal, reduction in the glucose removal from the blood, because of the reduction in the insulin production and secretion, reduction in insulin sensitivity, and decrease in the glucose uptake by the tissues. All of these will cause hyperglycemia. I think in a conference like this, it may seem offensive to show such a slide, but I think I just want to pick two points only, that we are here talking about the glucocorticoid effect of the steroids. We're not talking about the mineralocorticoid, and that's why I put that in the red box there. The other thing, it's very important from the management and the assessment, assessment and the management point of view, to distinguish between short, intermediate, and long-acting steroids, as you see in here. This is a very nice graph shown by Epper and colleagues in a recent, very extensive review, which I recommend for people to read and use for their own teaching sessions. Here it shows the short-acting, intermediate-acting, long-acting with examples such as hydrocortisone, prednisolone, and dexamethasone, and shows the hyperglycemic effect there in hours, the onset. For instance, the short-acting preparations, the onset is very quick, at one hour. It peaks at three hours and starts to go away between five to six hours. The intermediate-acting ones start their hyperglycemic effect around four hours, peak at around eight, and start to go away by 12 to 16 hours. Dexamethasone is very variable, but starts late and continues for 24 to 36 hours. So showing this would actually show the cartoon, the glycemic profiles of the glucocorticoids given once daily and twice daily. You can see that clearly, the short effect in the hydrocortisone, the intermediate effect for the prednisolone, and the long-acting effect for the dexamethasone. It also shows a distinction between when you are using the medication for short, for once or twice daily. Okay, so we need to know who is at increased risk for glucocorticoid-induced hyperglycemia and diabetes. And this is getting old now, but this is a very nice meta-analysis, looking at the hyperglycemia induced by glucocorticoids in non-diabetic patients by Liu and colleagues. And I actually just make the summary from it, looking at the overall frequency. They've included about 13 studies. Most of them are observational studies or retrospective studies, and they found actually the average of steroid-induced hyperglycemia is around 32.3%, and diabetes around 18.6%. So we are talking about one-third of our patients who are treated with steroids from these studies, and these studies are being affected by hyperglycemia, and nearly one in five are affected by diabetes. They have identified that risk is very variable, particularly looking at causing, caused by the glucocorticoid attributes, by the patient's characteristics, by the underlying or the disease process that's being treated. And I'll elaborate this in the following slide here. So these are the comments on the glucocorticoid patients and disease factors. For instance, if we start by the glucocorticoid attributes, the type, the dosage of the glucocorticoids, with a strong and high dose, are more likely to cause hyperglycemia. The route of administration, naturally systemic, particularly IV and oral, more than inhaled. But also particularly notorious is the intra-articular glucocorticoids. The dosing regimen, whether it's continuous and the concomitant medications, particularly for those who are post-transplant patients and patients with other conditions, with diuretics, immunosuppressives, i.e. medications that have inherently the risk of increasing hyperglycemia. When you look at individual patients, older people get it more. A prior history of impaired fasting glucose, impaired glucose tolerance, gestational diabetes, high BMI, abdominal obesity, high triglycerides, a history of polycystic ovary, history of insulin resistance, hemoglobin A1c more than 6%, family history of diabetes, also specific races and ethnicities, impaired renal function, the coexistence of hypertension. And also, if on a previous exposure to steroids a patient has developed hyperglycemia, then on this time also they're very likely to have. Also some other collection of infections, post-transplant patients, chronic obstructive pulmonary disease, and smokers, have all been identified in previous studies that they increase the risk of hyperglycemia. I'll show you some of this data here. This is the frequency of hyperglycemia in hospitalized patients who are receiving high dose glucocorticoids based on the history of diabetes. And if you look here on the three sets of curves, the one that's shown overall is on my side here. And also when you look at those with history of no history of diabetes compared with those who have a history of diabetes, you can see the finding of blood glucose more than 200 at least once in the red columns or twice in the yellow columns is actually increased. So anybody who has history of diabetes is at increased risk of having hyperglycemia during treatment. Also, if you take a more elaborate classification such as no diabetes, pre-diabetes, and diabetes, you can see that the risk of steroid-induced hyperglycemia, shown here by the red or the amber color, is increasing as you move from no diabetes to pre-diabetes to diabetes. So I think knowing the underlying diabetes or glucose tolerance status is very important. Here's how to recognize that. This is the important slide that gives us the background information. For instance, if you look here at this continuous glucose monitoring of glycemic pattern in the patients who are receiving prednisolone for chronic obstructive airway disease. It was black and white, but I decided to color it for ease. If you can see at the blue triangles, this is the team patients with no diabetes, no prednisolone. These are the controls. When you look at the 40 patients with no diabetes, on prednisolone, you can see there is a rise in the blood glucose when comparing with the controls, whereas the seven people in the squares, the red squares, who has diabetes and on prednisolone, you can see that actually the blood glucoses are much higher. The other thing also should be illustrated more in the following slide, that actually there's a specific rise in the blood glucoses, which is around the meals, after the meals, at both lunch and the evening meals. This is a very, very focused study looking at what time that we need to look for hyperglycemia. Here's the glucocorticoid-induced diabetes and related risk factors in patients with neurological diseases. They included 25 patients, 13 of whom developed diabetes. And when they looked at the highest point of hyperglycemia, they found actually the two hours after lunch is the most important time. And this is very important from practical point of view. So this is the time where to look for hyperglycemia. When we come to management, I am going to actually go through the management of hyperglycemia steroid and steroid-induced hyperglycemia as suggested by the British group. Here is the identification criteria as the same as the ADA, the International Criteria for Diabetes. When we start steroids, we check for blood glucose. If it is more than 11.1 millimoles on two or more occasions, we question whether the patient is preexisting or not. If it is preexisting, this is what we call steroid-induced hyperglycemia. If no known history of diabetes, then this is the steroid-induced diabetes. The ADA criteria are applicable to this situation with two cautionary points. Number one, it is not appropriate to do an oral glucose transfer test for a patient who is sick and while in hospital, number one. And number two, the hyperglycemia may occur very fast, not to be reflected in HPA1C. Therefore, monitoring in the hospital is more important than anything else. The assessment for patients who know previous history of diabetes, we need to check hemoglobin A1C when we decide to treat with steroids so that we know where we are standing from the start. This is in general, but it is much more important for people with high risk. When we start the steroids, we need to do the continuous blood glucose, capillary blood glucose monitoring before and after meals. The recommendation from the British group saying actually before and after meals, but I think most experts favor the after meals because the phenomenon itself is a postprandial phenomenon and the fasting blood glucose is rarely moving up. If it is less than 12, then you can consider, depending on the situation, whether you want to stop or monitor for about two to three more days. If it is consistently more than 12, then actually you need to frequency of the testing and you decide on the treatment. Patients who have diabetes, to reassess the control, what's the situation, what's the medication that they are taking before, are they on oral agents, are they on insulin, and then check the blood glucose four times a day with a great focus on the postprandial. Patients with type 1 diabetes, the focus is much more and you would expect to increase the doses of the insulin. The target is around 6 to 10 millimoles per liter, although sometimes it is acceptable to go slightly above. And end of life, the emphasis is on the symptom control and also the capillary blood glucose range is quite much more relaxed. Once steroids are reduced or discontinued, it's very important that actually we should not allow patients to go to hypoglycemia, and therefore the monitoring should continue and the medication, if it was increased from before, it needs to be reduced down in tandem with the reduction in the blood glucose, the dose of the steroids. Naturally, if not sure, we encourage other specialties to involve the diabetes, but I think sometimes we have to keep an eye on them because sometimes you just give one advice and you expect them to follow it and come back to you, but I think the patients can be lost to control. The strategies will depend on the previous diabetes history, the current glycemic control status, the current illness itself, whether you are talking about an inpatient or outpatient, and also the steroid themselves, the doses, the likely duration, the strength, and also previous experience and responses before. I'm not going to show you this in detail, but I think just the question is to ask, is the patient known to have diabetes? If no, then there's one pathway. If yes, there is type 2 diabetes, there's one pathway. And if yes, there is type 1 diabetes, there's another pathway. I'll just take you through this one by one. Here is, for instance, if the patient is not known for diabetes before, then monitoring is based on the criteria, as we said, fasting and postprandial. The recommendation in the BRIDGE practice is based on sulfonylurea, and particularly they recommend the Glycolyzide because it is widely used. This is not echoed in the other international recommendations because, for instance, in North America, Glycolyzide is not available, and therefore doesn't feature very much in those. And then you adjust the medication as required, and if you're not meeting the target, to consider the insulin. For patients who are type 2 diabetes, naturally if no hypoglycemic symptoms and not on any sulfonylurea or insulin, then to start on Glycolyzide. Nowadays, most of the patients are on certain medications. We don't have patients who are actually not being treated with some medication. And naturally, the recommendation, this is for the UK guidelines, to seek specialist advice if patients are not responding. Those who are on once daily nighttime insulin, it is important, and I see that for the last three years in my hospital, where there is a lot of cancer patients, the insulin is not being moved to the morning. It should be moved to the morning because this is the right time. If you leave it in the evening, there's more risk to have hypoglycemia. If the patients are on twice daily insulin, then you need to increase the morning dose by about 10% to 20% and monitor and adjust the dose as required. Again, you escalate up the medication as required. Sometimes patients may need to go on four times daily, i.e. basal bolus regimen. For type 1 diabetes, one has to be very careful because sometimes these patients can actually go on diabetic ketoacidosis simply because of the steroids. If a patient is going on twice daily insulin, it's not really commonly used nowadays, but certainly here, but the morning dose needs to be increased. Monitoring has to be more intense and to meet the targets of 6 to 10 millimoles. If patients are on basal insulin, make sure that the basal is given in the morning and it is increased by about 10% to 20% to start with. This is the article I referred to. I actually recommend that you make a note of the first author and the year. I think it is an excellent article that I should recommend. It gives a lot of practical points. I will not go through the whole set of medications for a diabetes audience, but I think I want to just talk for all of these groups. They have advantages and disadvantages. There is evidence base for any of them, availability, cost, and suitability, matching the steroids with the medication. Last month, I gave a similar talk like this. Actually, while I was preparing, I came across this article from Russia, which is reanalyzing all the data and proposing a pathogenesis-based glucose-lowering therapy for this particular group. It is a very extensive article and I recommend that people also have a look at it. The time that the patient is going to leave hospital, then we have to de-escalate the medication, but at the same time teach the patient how to assess the blood glucose. Sometimes giving them information like this, how to reduce the dose according to the prednisolone dose, then actually it can be helpful. I think the British group actually proposed some discharge notes with a lot of information of particular value. How to prevent it? We can think about lifestyle, steroid stewardship, but also metformin has evidence. I will not discuss the point one and two because of time, but I just want to show you this slide here, showing the metformin preventing the glucocorticoid-induced hyperglycemia using metformin, 850 milligrams twice a day. It's a small study. There are 17 patients versus 12 placebo, and I'll show you the summary of the data here. If you look at the placebo on your left-hand side, you can see versus the baseline at four weeks, there is a rise in the blood glucose at the same time points, whereas in patients who are treated with metformin, you can see that actually the lines are almost, not almost, actually overlapping between the baseline and the four weeks. In summary, the use of glucocorticoids is very common in clinical practice in cancer patients, rheumatological conditions, immunological conditions, renal conditions, and also pulmonary conditions. They can impact adversely on glycemic management in those patients who are living with diabetes and those without previous diagnosed diabetes. Therefore, surveillance is required for both of these populations to try and mitigate against the harmful effects of hyperglycemia. Because of the complexities of the various glucocorticoids used, their power and their duration of action, and also matching that with the patient's own diabetes and the underlying disease, then individualized glycemic target is required. And the safety when glucocorticoids are lowered to prevent risk of hyperglycemia is of paramount importance. With that, I come to an end, but I would like to plug in the Journal of Diabetes and Endocrine Practice. This is the publication of the Gulf Association of Endocrinology and Diabetes, which is the body that has taken over the ACE Gulf chapter when the ACE has actually abolished its chapters. And I'm the editor-in-chief, and I would welcome your submissions. Thank you very much.

glucocorticoid-induced hyperglycemia

pathogenesis

management strategies

insulin resistance

risk factors

individualized care