Hello. Thank you to Dr Gavin for his kind introduction. Uh And thank you for the previous speakers for setting the stage for what I'm going to be talking about. I'm Vivian fonseca from new Orleans Louisiana. And I'm going to build on the data you've heard from the previous speakers about using uh continuous glucose monitoring in type two diabetes focusing on type two. And I'd like to thank Dr Miller for telling us a little bit and showing us the data about using it in type two wit and without incident. So and I particularly want to thank Dr Miller for sharing the cases with with me. Uh She's provided the slides of the pictures, the A. GPS etcetera. But the case discussion is mine and that's my interpretation. I'm sure many of you might have a different interpretation and that could make for some good discussion but that's what clinical medicine is about. And you're my disclosures and I'd like to start by talking a little bit about some real world data. Uh and I'll come back to some real world data also at the end, when we talk of people who are not on insulin, but let me start with people are insulin Using a freestyle libre, real flash glucose monitoring. And these were abstract presented at the recent 80TD meeting. So it's very new data. Uh The first of them is a very large study done in the UK using UK liberal view uh in people over the age of 18 with type two diabetes who had at least five days of sensor readings uh last year. And they what it showed well what they were looking at was time and range. Uh and The overall time and range was a little bit about 51%. And they wanted to see what the impact of using flash glucose monitoring was. And what they found was that the target Of greater than 70% uh was achieved in more people. It went up from 51%. The mean went up to about almost 58%. Uh and and and the impact of the pandemic. I mean this is very focused on what happened in the pandemic and this has been reported in other countries. The pandemic affected people with diabetes a lot more than people without diabetes. But many people became very conscious of this fact that became emerged very early when when the covid started and people started attempting to control diabetes using more C. G. M. And flash glucose monitoring. And what you can see here is that in all age groups between january 2020 and june 2020 time. And range improves significantly and the proportion of people meeting Uh more than 70% time and range with less than 4% below range also improved significantly. Not quite where we want it to be, but still a significant improvement, including elderly people, many of whom have type two diabetes requiring insulin. A lot of challenges in managing their diabetes, et cetera. And we're very vulnerable to the effects of the pandemic. So uh hopefully better control of diabetes will keep away such severe infections. The second study is a prospective observation study in Italy, a smaller number of patients 322 we type two diabetes from hospital sites. Uh They had uh poor glycemic control with a one CS between eight and 12 despite taking basal bolus insulin regimen for at least one year and after 3 to 6 months. Uh About 230 of them had continued using freestyle library. And uh they also had a you know, they've broken down into intervention group and control group. And what you see is that those who had the intervention had a significant improvement in A one C. Compared to those using self monitoring of blood glucose And years of study from Canada looking at doing chart reviews again from population based. Uh but it's a much smaller number of 91 patients using basil insulin uh from six diabetes centers. And uh they found that the average A one C was quite high at 8.9 prior to using uh freestyle, liberating this relatively elderly population who had moderate obesity and had been using insulin for at least two about 3 to 4 years and the uh years. The results, the A one C. Was significantly reduced after the intervention by 10.8%. Uh And subgroup analysis by baseline A one C showing that those uh even below 9% had a significant reduction of 90.5 and those uh as very often as seen in uh clinical trials and in data such as this those with very high A one CS appear to do much better because they have more to fall. But overall within 3 to 6 months you see a significantly reduced A one C. So let me now with that background move on to the case studies and I'll come back to look at some real world data and people not taking insulin. Okay, So the first patient is a 66 year old male with type two diabetes who has a moderate renal impairment and is taking self injurious. Only Metformin had been stopped as the G. F. R. dropped below 45 And the a. one c. was discrepancy with the glucose tests. Uh The A one C appeared to be low but when the patient itself monitoring the glucose level was low. And uh the same was being seen as you can see in the G. P. Which I'll discuss in more detail involved. Now this is not an uncommon scenario. People with CKD continue to advance. Uh The cellphone doria has to be used because people cannot take Metformin. They don't want to take insulin just because they can't take Metformin. They prefer to have another oral agent S. E. L. T. To innovators by that stage. Are very effective in slowing the progression of CKD but are less effective in achieving glycemic control. Self aneurysm metabolized in the kidney. The pharmacogenetics change. So they get hypoglycemia at different times that are unexpected and get high hyperglycemia at other times. Another important thing is that A one C becomes unreliable due to anemia and Hamal Asus and they are developing other complications right now. But it is getting worse, neuropathy is getting worse supply. Civic control doesn't matter to prevent blindness and amputations, et cetera. So yes, the ADP report in more detail, the glucose management indicator, uhh is 7.3, which is higher than the reported A one C in this patient, which ranged uh from 5 to 6% which is probably falsely low because the average glucose was around 167. And if you look at the diamond range, it's somewhat suboptimal being 63. Some laws, not much in no very, very low readings, but the fair number of high readings, including some that are very high. And if you look at the profile, you do see some lows, particularly at night, which is a problem in people with renal disease and highs in the afternoon and evening, uh particularly postprandial lee after the lunch and dinner on some days. Not on also, some days it's reasonably good. So we need to then move on to look at the daily glucose profiles to understand this a little bit better. So this is the way I approach it. I look at the A. G. P. I find it very useful, but when things are not quite fitting, I do look at the daily glucose profile and you still see this pattern of a noontime rise in blood glucose, which is fairly consistent and some drops at night on some days. And the pattern insights give you some very clear indicators. Here's the problem time, the afternoons for hypoglycemia and the low glucose can be seen at various times, but mainly mainly at night. And when you look at the low glucose events, there were four such events Compareding up to 105 minutes, which in somebody who's got renal disease, probably got heart disease might make them susceptible to ischemia or arrhythmias. And the average glucose is still quite hard. What we noticed from this case is that there is considerable variability, both intra day and into the intraday is very obvious. But from day to day there is some variation. There's clearly unrecognized hypoglycemia, particularly at night, very poor postprandial control. And it clearly points out that using a self injurious alone is problematic. It doesn't take care of the postprandial glucose adequately and sets up a public senior several hours later. So we need to consider alternatives such as perhaps a long acting GLP one receptor agonist. The patient in tolerated to take care of that postprandial glucose excursion. We could reduce the Sultan urea dose particularly at night to avoid that policy and maybe also consider a bedtime snack. And while you're discussing diet, maybe talk about shifting some of that carbohydrate that's at like taken at lunch and dinner two other times during the day, maybe at breakfast and at bedtime. So hopefully by understanding what's actually going on through the day with that patient, you can take better decisions rather than relying on the false A. One C. And the misleading occasional self blood glucose monitoring in such a patient. So let me move on to another fairly common scenario. Less common perhaps than the renal impairment. Uh ketosis prone patients with type two diabetes, 49 years old On Basal insulin takes an occasional small bowlers of uh rapid acting insulin when the glucose is very high as an a one c. 8%. Yet the blood glucose is not always very high. In fact, a lot of the time when uh she tests is within range. And if you look at the A. G. P, the average glucose is not bad at 1 60 within a one C. Of 7.2. Which suggests again some discrepancy with what's happening with the A. One C. And when you look at the Uh diamond range, it's almost that 68% within the target range. But some eyes and no lows in this particular patient. So it's mainly a problem of hyperglycemia, which we need to look at it a little bit more details. So, a question that I ask in a patient like that who's had episodes of ketosis, is this patient really got type 1? 5% of type two patients turned out to be get anybody positive and really have type one, which means that they probably need a little bit more of basil insulin rather than just pushing I'm sorry, More bolus insulin rather than just pushing the basal insulin. And uh the G. M. E. Does not reflect the glucose speaks and for that matter neither does the A. One C do that very well, glucose tends to be high because they are insulin deficient. Unfortunately, this patient is not getting hypoglycemia, but that's because she's not taking enough insulin at the right time and she might be prone to hypoglycemia if she did it the other way. So if you look at the G. P. U. C. Here that there is some very variability, a lot of the time the patient is well within range. But uh there is some days when the glucose is high, fortunately no hypoglycemia. So you want to know what are those days and what's happening? And you notice that when you look at the daily glucose profile that it's the weekends, the blood look hype after breakfast on saturday and sunday, a little bit of variability during the week. But again, uh on saturday and sunday uh patient tends to be high in this case uh friday high during the night. We need to find out what happened on thursday night. So having a little bit more information on the patient's carbohydrate intake at particular times. Perhaps asking the patient to keep a diary might help because that will help you to taylor uh the appropriate use of a rapid acting insulin bowlers on a scheduled basis, rather just using it randomly when the patient happens to see a high blood glucose which may lead to overcorrection. He has the glucose pattern insight and you you notice that it points to some degree of variability uh in this, but let's no problem of hypoglycemia. And you see that again on the snapshot a lot of the time this patient is doing reasonably well. But I get back to the fundamental question that you see you, this patient has got insulin deficiency uh possibly related to maybe a low C peptide or get antibody and not having the glucose secretary capacity to avoid postprandial discussions when a high carbohydrate load is taken, leading to some erratic pay peaks on on weekends. So this is a good opportunity to discuss diet, particularly high carbohydrate on those days, keeping a diet record and review along with the C. G. M. The next time she comes you could split the basal insulin to cover twice twice a day. Consider basal bolus therapy. But we have alternatives to take care of some of these peaks. A long acting GLP one receptor agonists or an S. E. L. T. Two in a bit of might help. However you need to bear in mind that patients like this at risk of ketoacidosis. So uh some mitigation of that risk might be needed such as keystone monitoring etcetera, which the patient may not be willing to do on a regular basis but should be aware of and uh maybe make sure the patient has ketone testing equipment. So let's move on to the 3rd case. This is a 59 year old male with type two diabetes for 10 years. Now on basal bolus therapy may be that the duration of diabetes was inaccurately estimated and it had diabetes longer and advanced to this late stage of basal bolus therapy. A one C. is not bad at 7 3. But uh he's getting some hypoglycemia at night that he actually recognizes and might actually uh not recognize it at other times. The average glucose is not bad. And the A one C glucose management indicator is close to the A one c maybe a bit lower, indicating some possibility of low blood sugars which you see there. But there's a lot of variability and that is really the problem here. A lot of the time the patient is well within target. But there's Hiestand lows that you want to to address. This is uh a little bit more detail. And uh In particular, I want to point out a few very high blood sugars are greater than 250. So maybe you want to try and identify when those are occurring. So here, you can see that in the overall profile, the GDP have dropped during the night, particularly some hypoglycemia around three am on some days, uh sort of consistent rise uh in the morning after breakfast and another rise after after lunch and again another rise after dinner. And very often the patients when they were doing uh self blood glucose monitoring. We're not testing at these times. We missed them and and we missed the hypoglycemia at night. If they were asymptomatic. If you look at the daily glucose profile, you find that it is somewhat erratic, unpredictable. Uh some days high higher on monday and Wednesday, but not so high on thursday, with some drop in blood glucose as some other times dropping at night. Other times tending to drop towards the afternoon. You can see that uh year on this Tuesday and this friday, so a considerable fluctuation and variation. And we frequently see this in people on basal bolus therapy. In general, basal bolus therapy has not been a tremendous success in Type two diabetes. They have more problems with hypoglycemia and hyperglycemia. In these patients perhaps uh more of a problem than in type one where uh people learn how to how to manage their diabetes better. If you look at the glucose pattern insights, you see that the problem times tend to be the mornings, uh there's some lows and and a lot of variability, uh large standard deviation, but that variability is occurring at many times during the day except at night, where it tends to be more on the low side. So you've got a lot of work to do with this patient address the hypoglycemia, particularly during the night and in the afternoon, uh and try to eliminate some of the peaks to reduce the hypoglycemia. Uh sorry, hyperglycemia. So uhh in summary, this patient has a lot of variability intraday and and today. Unrecognized hypoglycemia, night and afternoons, poor postprandial glucose control. So we need to identify the reasons for the variability. Diet incorrect carb counting and overcorrection with insulin. Okay, may be delayed meals and some stress, which is something that we find very hard to quantify as compared to carbohydrates. It's very hard to know what people stress levels are. A diary would be helpful. The patient is high at bedtime but might drop during the night. Again, too much basil insulin. Again, you could consider adding long acting GLP one If tolerated an SLT two in a bit of to reduce those excursions and a bedtime snack, making sure that it's not refined carbohydrates. Maybe using some protein or corn starch to eliminate the nocturnal hypoglycemia. So what are these cases? Teachers about Type two diabetes. First of all? Uh there's a considerable heterogeneity in this disease. It's so variable. We learned so much when we start testing people. The glycemic patterns vary a lot, giving us insight into the disease process for the individual patient. Is it the lack of insulin that's in response to a meal? That's the problem. Is the meal a problem? Is it the real disease? Even though the patients not taking insulin as you saw in the first patient died and stress have a considerable impact on glucose patterns and help the patients understand this. When they do flash flash glucose monitoring, they themselves are then willing to change their diet uh quite considerably in my experience. So see GM may help devise an effective and personalized treatment strategy in these patients. Let me turn to now fairly recent publication Of real world data and people with type two diabetes, we're not taking insulin. Uh And this comes from three countries in europe where you see a reduction in A one C and it's in million multiple leader. Yeah, So this is roughly a reduction of about 1.3% in in Percentage stones of the a. one c. And when you look at the breakdown of changing a one c. Uh you know those who start off with a very high one C tend to have a greater benefit, as I pointed out earlier. Otherwise it doesn't matter about their age gender whether they're taking insulin for a long time or shorter time, obese or not obese, they all benefit from uh a flash glucose monitoring. Can we put this into context? The real world data which is the randomized clinical trials And here's the analysis that attempts to do that. Uh Here's what happens in a randomized clinical trial, use intervention uh with a control group. Everybody's being followed very consistently. You're doing uh monitoring in a very consistent kind of way. And your uh this particular study looked at people uh uh We're using flash glucose monitoring and compared them with self monitoring of blood glucose. And the good thing about this study was that at baseline, the groups were well matched, including a one C being about 7.8 uh in both groups now bear in mind these are people not taking it taking insulin. And you may wonder if do they really need to have uhh flash glucose market? Will it really help them? And you see here the the devices were provided for us uh 12 weeks and then uh there was a significant difference in in a one C dropped. Both groups dropped as the study effect greater drop in those using flash glucose monitoring. That then persisted. They made persistent changes uh perhaps in their diet and and other things exited lifestyle that will allow a longer term improvement in glycemic control. And when you focus in on what areas will help, you can see that diamond range increased significantly in those doing flash blood glucose monitoring. And when you look at various subgroups, uh looking at hyperglycemia, it was less a uh huh uh greater than 1 80. Less. Greater than 2 40. Less greater than 300. Uh in the group doing flash glucose monitoring. So you're eliminating those large peaks which you probably never saw in people doing self blood glucose monitoring. And another strength of this study is they looked at patients treatment satisfaction. Uh was it disturbing to them to be testing uh and knowing their blood glucose so many times. And the total score clearly showed an improvement in their satisfaction. They like the convenience and flexibility of not having to stick their finger uh and using the device and uh a higher proportion of them said they would recommend it to other people. And yes, the real world data again, reduction in A one C. After startling flash glucose monitoring showing a big drop in A one C. Particularly in those who have very high A one C. Uh And it didn't matter where they were an insulin and not on insulin. They both had a benefit a little bit more of a drop in those not taking insulin uh and overall a good outcome. So in summary flash glucose monitoring provides us with a lot of useful information in clinical practice that lead to actionable items. We can, as you can see we can very easily make changes in diet behavioral change. Uh pharmacological therapies can be changed or added that will allow much better glucose control, reduce the risk of hypoglycemia. And it together produce the risk of long term complications of diabetes. So I want to thank you all for your attention and I hope you enjoyed this program, a combination of good, strong real clinical trial evidence, along with some real world data and a case discussion that I hope will help you in your clinical practice.
Related Presenters
