Activity Transcript

Bruce W. Bode, MD, FACE: Hello. I'm Dr Bruce Bode, president and CEO at Atlanta Diabetes Associates, and I'm also associate professor of medicine at Emory University. Welcome to this program titled, “Expanding Diabetes Technology Access and Adoption to Reduce Health Inequities and Improve Patient Outcomes.” Joining me today are Dr Jennifer Sherr, who is a professor of pediatrics in the Department of Pediatrics at Yale University School of Medicine. Dr Sherr spends the majority of her time conducting clinical research on new technologies and therapeutics to improve the lives of people living with diabetes. Jen, thank you for joining us.

Jennifer L. Sherr, MD, PhD: Hi, Dr Bode. Great to be here.

Dr Bode: And Dr Jones, who is an associate professor in the division of endocrinology at Cincinnati Children's Hospital Medical Center. Dr Jones focuses on type 1 diabetes and how social determinants of health impact chronic diseases, specifically diabetes. Welcome.

The goal for the next 25 to 30 minutes is talking about insulin delivery systems, specifically automated insulin delivery systems, also known as AID. And then, we're also going to talk about the social determinants of health and the importance of individualizing diabetes technology. Overall, the goal from the American Diabetes [Association] is to have an HbA1c under 7%. Ideally, before eating, your glucose should be 80 to 130 mg/dL, and after eating it should be less than 180 mg/dL. Most people with type 1 are using CGM. And when you look at CGM, the goal would have a 70% time in range between 70 and 180 mg/dL, and to have less than 4% below 70 mg/dL and less than 1% below 54 mg/dL. And also, you want to minimize hyperglycemia with less than 25% greater than 180 mg/dL, and less than 5% greater than 250 mg/dL.

And so, when using these automated insulin delivery systems, you have an insulin delivery system, you also have a CGM, and in this system, there is an algorithm. And so, every 5 minutes a sensor glucose is going into the algorithm and making insulin to get a goal, ideally between 110 and 150 mg/dL. There's other terminology that people use as being artificial pancreas or hybrid closed loop and it's known as hybrid because you still need to cover your meals either by putting in carbs or putting in insulin.

Before we move on, please take a moment to answer the following question that will soon appear on the screen.

Jen is now going to talk about FDA approved AID systems.

Dr Sherr: Thank you, Bruce. To review currently commercially available systems, we're going to go in the order by which they achieved regulatory approval. So, the first AID was the 670G, which was approved in September of 2016. This system uses a PID algorithm with insulin feedback. It has adaptive limits which are impacted by the total daily dose, and an estimate of fasting glucose and plasma insulin concentration at the time of fasting. The system is a treat-to-target system with a target glucose of 120 mg/dL. An algorithm-driven basal insulin delivery is provided every 5 minutes. Adjustable parameters are the insulin to carbohydrate ratio and the active insulin time but preset basal rates do not impact automation. It uses the Guardian 3 sensor, which requires a minimum of 2 calibrations per day. Safety constraints with the system lead to system mandated exits. And in 2020, approval was lowered to age 2 with release of the 770G, which uses Bluetooth instead of radio frequency signal transmission.

The next system that was approved is the t:slim X2 Control-IQ. This system was approved in December 2019. The system uses an MPC algorithm and has a treat-to-range approach. The usual range that the algorithm targets is 112.5 to 160 mg/dL, and the system relies on preset basal rates to adjust algorithmic insulin delivery every 5 minutes. Additionally, the system provides an automated correction bolus if glucose is predicted to be greater than 180 mg/dL. And this has provided once an hour if there's been no bolus in the past hour. Adjustable parameters are basal rates, insulin to carb ratios, and correction factors. Active insulin time is hard set at 5 hours. It uses the G6, so no calibrations are required, and there is algorithm-based learning based on the total daily dose. In 2022, a mobile phone bolus feature was released.

Finally, in January 2022, the Omnipod 5 was approved. This is the first tubeless AID. With this system, there's an MPC algorithm that's located on the pod, which contains insulin and is attached directly to the person's body like a patch. The system is a treat-to-target with targets ranging from 110 to 150 in 10 mg/dL increments. Targets can be varied throughout the day. And the system automates basal insulin delivery every 5 minutes, and also has a smart bolus calculator that incorporates both the CGM value and trend into the algorithmic determination of bolus insulin delivery. Adjustable parameters are the insulin to carb ratio, system targets, active insulin time and correction factors. Again, with this system, preset basals do not impact automation. It uses the G6, so no calibrations are required. There's algorithm learning based on the total daily dose.

Dr Bode: Thank you, Jen.

Nana, what do the guidelines say about the use of these technologies?

Dr Jones: The American Diabetes Association, which really guides our standards of care, have really expanded upon who is eligible for current pump therapy. So, if you need to normalize blood glucoses, and so your HbA1c’s are above target, including above 6.5% or above 7%, you have wide glycemic excursions. So, your coefficient of variation or the wideness of your blood sugar trends also is indicative of if you need an automated insulin delivery system. If you have hypoglycemia unawareness, thus you can't tell when your blood sugars are low, or even when you have hyperglycemic awareness, sometimes that can even be an indication. And of course if you want flexibility in your insulin regimen.

In the 2023 update, it specifically addresses knowing and growing literature on reductions in HbA1c and improvements in time in range with its use. So really, AID should be offered to all patients with type 1 diabetes, especially to bring the HbA1c into a target and to reduce those glycemic excursions. There's no one size fits all model for diabetes technology.

Dr Bode: As you've mentioned, Nana, advances in technology could be of benefit to a lot of individuals and are recommended in guidelines, but we know that underrepresented minorities, patients on public insurance and those with less economic stability are less likely to reach diabetes goals; why is this, Nana?

Dr Jones: Yeah. So, you're right, Bruce. And it's odd, because pumps are wonderful, continuous glucose monitors are amazing, sensor augmented pumps are great. And then the automated insulin delivery systems that Jen talked about are transformational. Plus, with these advancements, there really is no ideal candidate. Everybody is a good candidate. So why is it that underrepresented minorities, those on public insurance and those with other socioeconomic instabilities are not able to get this technology? And that's really because of access. Despite the fact that diabetes technology is ideal for everyone, it really isn't. You must have access to this technology. Just because you're a good candidate doesn't mean that every candidate wins. There really are health disparities in the access to diabetes technology, and we really want to understand why there are such gaps, right? Differences in this access are primarily based on insurance coverage. If you're on public insurance, public insurance tends to lag behind diabetes technology and FDA approvals of those devices. Implicit bias. So, these are the subconscious, sometimes maybe even conscious stereotypes that providers may have upon patients, especially if they deem them not have high health literacy, not have good, safe environment, or maybe based on racial implicit biases that we may not be conscious of that really don't align with access to such life-changing devices. And then of course, socioeconomic disparities. These automated insulin delivery systems, CGMs, even smart insulin pens, are not inexpensive. And so, if you have a high copay, if you have a high deductible or if you're on public insurance and simply cannot pay for those technologies, you don't have access to them.

If we take a look at data, especially from the Type 1 Diabetes Exchange, which really took a hard look back in the 2020s about diabetes technology use amongst racial ethnic groups, you can see that 47% of non-Hispanic whites were using continuous glucose monitoring back in 2020 compared to 30% in Hispanics and 19% in non-Hispanic blacks. And then, if we took a look at JCEM, 62% of non-Hispanic white patients were using CGM so, an even greater number, but still only 37% of Hispanics and 13% of non-Hispanic blacks were using continuous glucose monitoring.

And then, if we flip over to pump use, the Type 1 Diabetes Exchange recognized that 45% of non-Hispanic whites were using pumps compared to 30% and 28% of Hispanic and non-Hispanic blacks, respectively. And the JCEM December article, in that same year, they noted that 54% of non-Hispanic whites were using pumps compared to only 22% in the Hispanic population, and a dire 7% in the black population. So again, time and time again, we see that certain populations have less access to diabetes technology. And really that's just a fact. It's not an opinion, it's not a comment, it truly is a fact based on numbers that we have seen time and time again in our diabetes journals.

Dr Sherr: I'll just add that in terms of choosing the best technology, I think the best way we do that is to provide individuals with diabetes with all the information about all the devices that are available. So constantly providing education. When individuals come to see me in clinic, they know I'm going to bring up technology. They know I'm going to tell them all that's out there. And I'm going to talk about specific features of each individual system to find out if one might work for them. And maybe they're not interested in trying it at that day, but maybe I'm planting a seed. If somebody wants a system, I'm going to go ahead and fight, and that means that I'm going to go ahead and put in the order, but then I'm going to work on prior authorizations, do peer-to-peers, and also really cite the literature - provide the ADA standards of care, the AACE guidelines, the ISPAD guidelines; we have so much literature that really supports that these technologies should be used. So, providing an individualized option to best personalize the care of the person with diabetes.

Dr Jones: Absolutely, Jen. But I do want to acknowledge that there may be some providers that would be uncomfortable offering a diabetes technology that is not covered by the patient's insurance. So, you know there's a device that a patient wants but can't get because of the insurance, we do have to take on that burden for the patient and really fight for the patient to get that diabetes technology. And we can move the needle. And so, it is okay to offer a diabetes technology to a patient you know whose insurance doesn't cover it so that the patient understands that this technology exists and understands that, eventually, they might be able to get access to it, and then really have their rights laid out before them.

Dr Bode: Nana, you have a case of this from a person, 19 years old, diagnosed at age 12. You want to talk about that?

Dr Jones: Yes. Absolutely. So, let's take a raw example of how health disparities affect diabetes patients. So, this is a case of a 19-year-old. He was diagnosed with diabetes at the age of 12. Non-Hispanic black male on public insurance with a low median household income. They lived in a high poverty neighborhood, single parent household. He was in school in an underperforming school district and had very low literacy and numeracy. By the time that I took over his care, he was known to have a mood disorder and had experienced both physical and an emotional trauma. He had been on multiple daily injections without diabetes technology and had sub-optimal glycemic control with an average... And yes, I said an average HbA1c of 15.3%. There were multiple challenges to the care of this patient, with multiple episodes of hospital utilization, including emergency room department visits, hospital medicine admissions as well as admissions for diabetes. During an inpatient stay, which was quite novel... [because] we had never offered diabetes technology to start at an inpatient admission, but we took the risk, and did it because we understood that it was best for the patient. So, during an inpatient admission, we started diabetes technology, an automated insulin delivery system. Our educators really went to bat for this patient and started it at a starting HbA1c of 16.4%. Within 3 to 4 months, that HbA1c was down to 7.4% with that addition of that diabetes technology. But then guess what? Those other social determinants and other social influences of health were not completely remediated, and so, the pharmacy failed to deliver his diabetes technology. And so, there was a gap in coverage of his diabetes technology that we were unaware of, again, not having the means to contact us in between clinic visits. And then by the time he returned to a clinic visit, HbA1c was back up to 13.2% because he had to transition back to multiple daily injections. And then even at that visit, he could not pick up sensors. The pharmacy that was around the corner from his home where he had access to get to, sensors were out of stock and therefore, again, could not start the technology. An HbA1c at this point currently remained between 11.2% to 12.5%. This is a poignant example of how those social influences have impacted the care and diabetes delivery of these automated insulin delivery systems. Really sad and something that we really should be continuing to work on to fight against.

Dr Bode: Nana, I totally agree. You just have to keep fighting and you are. You've been doing this your whole life, fighting for this.

Dr Jones: Absolutely. Absolutely.

Dr Bode: Jen, let's continue our discussion with a few more cases.

Dr Sherr: Thanks, Bruce. The first case I want to share is one of a 17-year-old girl that was diagnosed with type 1 in 2006. In recent years with adolescence, she's had suboptimal glycemia with reduced engagement in her diabetes care regimen. At the time that I'm seeing her, she's on conventional pump therapy and real-time CGM, using the suspend before low system, and this is back in January 2020. And at that time, her HbA1c is 7.2%. The challenges are, dad's really concerned and excited for her to transition to a hybrid closed loop as she was in her junior year of high school. He keeps thinking college is right around the corner. She acknowledges the fact that she hesitates to bolus for carbs that she's eating because of a fear of hypoglycemia. So, she transitioned to the t:slim X2 Control-IQ in the spring of 2020. What we're seeing here is data on suspend before low in clinical care prior to that transition. What you'll notice is that her average glucose is 208 mg/dL, her time in target range is 39.4%. And you'll just see that across the day, she's really running on that upper level of that target range. That median is just sitting right above that 180 mg/dL. Taking a look, as a bar graph, we see here that she's in range about 40% of the time. And when we look at her insulin delivery metrics, we find that her total daily dose is 52.6 units. 46% of her insulin is coming from basal insulin and she's entering about 145 g of carbs. What happens with the transition to Control-IQ? Here we're seeing data from 2021. Please note it does span the holidays, so I'm going to give her some credit for that. And what you'll see here is that her average glucose is 164 mg/dL and she's spending 66% of her time in target range. The overnight glycemia is excellent with really a reduction in those ribbons that we see representing her standard deviation for her glucose across the time of day. And you'll see that her total daily dose really hasn't changed much and she's entering about 133 g of carbs. When we take a look at her data, what we find is that there's these periods where automated corrections, indicated in the red blocks, are occurring. And so, we notice she's still struggling to engage with carb bolusing, but despite that fact, with use of this automated insulin delivery system, her time in target range has increased 16%, and that means 3 more hours a day in range. And what does this mean in terms of her glucose management indicator? It's 7.2%. It's the same place she started. And so, in truth, this speaks to the fact that time in target range tells us so much more about the lived experience for individuals with diabetes. How does she feel Control-IQ has improved her care? Overnight control has vastly improved. Despite occasionally missing carb entry, she continues to see improvement in glycemia. And the family felt much more confident as she prepared to leave for college in the fall of 2021. Now, I did see her in 2022, but then she didn't see me from January '22 until just earlier this month. And what was notable is that she's still achieving those same metrics, 165 mg/dL for her average glucose and 65% time in target range. Not surprisingly, she's continuing to struggle with bolusing for meals. And so, I indicated we want to see her more frequently. And the benefit of that is maybe I could have discussed options to upgrade to the mobile phone bolus feature that's now available. She stated, as a college student, she's on her phone a lot more. And so, this feature could come in very handy to help improve her engagement with her diabetes management.

Recently, data have been published that supports the benefits of t:slim X2 Control-IQ during real-world use. Using a cohort of more than 19,000 individuals with type 1 diabetes, data was collected on time in target range. These individuals were using Basal-IQ, a predictive low glucose suspense system, and had 1 month of data prior to their initiation of Control-IQ. And there was 90 days of data with AID use. In these graphs, we see time in target range on the Y axis and time on the X axis. On the left-hand side, we see plots on age of users. As a pediatric endocrinologist, what stands out to me is that the kids had the lowest time in range, but across all users, regardless of age, there was a rapid rise in time in range with system use. On the right-hand side, groups are divided by the glucose management indicator, which provides an estimated HbA1c based on sensor glucose values. Not surprisingly, those with the highest GMI saw the largest rise in their time in range, but again here, we see that all had improvements in time in range. The findings of pivotal trials that supported regulatory approval of these devices are echoed in these real-world data sets.

Dr Bode: Thank you for sharing your cases, Jen and Nana.

Before we move on, please take a moment to answer the following question that will soon appear on-screen.

Jen, top line updates on recently presented data from 2022 on AID systems in patients with type 2 diabetes.

Dr Sherr: So up until now, the data presented has primarily focused on use of AID in those with type 1 diabetes, but this past year, headlines at many conferences showed the resounding benefit of use of AID in people with type 2 diabetes. Clinical trials of the use of Omnipod 5 in people with type 2 diabetes were presented at ADA [and ATTD]. Improvements were evident for those on multiple daily injections at baseline. But more striking, was the increase in time in range by 7.6 hours a day for those on basal only insulin at baseline.

In January 2023, data was published on use of the CamAPS HX2 full closed loop system where in a cohort of 28 participants with type 2 diabetes, time in range was 32% with usual care, and increased to 66% with this fully automated system. And what it means by a fully automated system is users didn't need to announce any meals or enter any carbs. They just were able to set it and let it work. So truly astounding findings.

Finally, real-world data of people type 2 using the t:slim X2 Control-IQ and 780G have shown improvements in glycemic metrics with AID use.

Dr Bode: Nana, for individuals with type 2 diabetes, what should we consider when they're on insulin therapy?

Dr Jones: As providers, we are proponents of equitable access to diabetes technology, and that includes patients with insulin-dependent type 2 diabetes. Remember, everyone is a candidate. There is no longer you can get a pump if 1, 2, 3, 4, 5, if you're checking blood sugars 1, 2, 3, 4, 5. The question is now, what are the barriers to implementing diabetes technology for my patient? Not, can my patient have diabetes technology? They absolutely can. Now of course, we do have to live in the real world, and we really do understand that there are social influences of health that prevent the healthcare system from being able to deliver this diabetes technology to the patient. And we're not going to be able to fix all those social influences by just having the passion to do so. There are real life barriers to diabetes technology. In reality, consider other options such as smart insulin pens, standalone continuous glucose monitors, and even flash glucose monitoring systems.

Dr Bode: Jen, what AID systems are in development and undergoing FDA review currently?

Dr Sherr: Thanks so much. So now we're going to transition to devices that are not yet commercially available. First, we'll review the 780G, which has a PID algorithm with insulin feedback with fuzzy logic algorithm incorporated. Benefits of this system over the 670G include varied targets of 100, 110 or 120 mg/dL. An automated basal is modulated to achieve these targets. However, if automated basal is maxed out and the individual is greater than 120 mg/dL, there's the ability to have automated correction boluses administered every 5 minutes. The system was initially approved in many other countries with a Guardian 3 sensor, which again required 2 calibrations. But now, the system has been available in other countries with a Guardian 4, which requires no calibrations. The system display has been revamped from the 670G with more icon-based menus and fewer button pushes. And pivotal trials have shown improvement in time in range with 22% of total bolus dosing coming from those automated corrections. While the 780G is not approved in the US, it is in the other countries.

So, leveraging real-world uploads from individuals in 22 countries, the impact of the 780G was explored. This is data from CareLink from August 2020 to July 2021 with a total of 12,870 users. System use led to achievement of 75% time in target range. When broken down by age less than 15 and over 15, time in range was roughly 73 and 76% respectively. Those with a system target of 100 and an active insulin time of 2 hours, the settings that are considered to be optimal had even higher time in range achieved, 78 and 81% respectively. Automated correction accounted for 25% of total daily bolus insulin in this cohort. Further, this publication explored a different cohort and looked at changes pre- and post-initiation of 780G. System use led to 11% increase, meaning 2.6 more hours a day. Finally, a third cohort was assessed to see if benefits of system use are sustained. As you can see, over 6 months' time, glycemic improvements are persistent with device use.

Dr Bode: What are the other future technologies coming to the market? And what should we know about them?

Dr Sherr: Thanks, Bruce. First, we have the G7, which achieved regulatory approval in December 2022. This sensor is smaller. It has a disposable integrated transmitter with each sensor and a warmup period of 30 minutes, which is really helpful because we need that sensor data to automate insulin delivery. Announcements have been made regarding integration of this device into both the t:slim X2 Control-IQ and the Omnipod 5 system. Building on the t:slim X2, we've heard of new devices to be released, including the Mobi, which will be an on-body reusable patch pump with a disposable reservoir. The device is waterproof and will have a button to allow direct bolusing from the patch pump itself. It's 50% smaller than the t:slim X2. It will have smartphone control and inductive charging like how smartphones charge. The t:slim X3 has also been announced. This device will have improved battery life, a USB-C port, wireless software updates, modernized user interfaced, enhanced technology, and it will hold 300 units. I think that's going to be really important for individuals with type 2 diabetes who may require more insulin. Another AID that has received great attention is the iLet Bionic Pancreas. This device differentiates itself from others in a few ways. It uses an MPC algorithm, but to start on this therapy, the system's initiated with just the user's weight. Meal announcements are qualitative and described as meal type, breakfast, lunch or dinner and meal size, usual for me, less or more. The target glucose is 120 mg/dL and can be shifted up or down. And all correction doses are completely automated when the user goes ahead and enters a meal bolus. Data on the pivotal trial of this device was published in the New England Journal of Medicine in September 2022. Finally, I wanted to mention Tidepool Loop. Tidepool is an organization that was founded to allow those with diabetes to have greater access to their device data by creating a singular platform to upload all devices and has been working to bring Loop to market. Loop is an open-source app that runs on a smartphone. The app was developed by the do-it-yourself community. It uses a model prediction that anticipates future glucose based on the effects of delivered insulin, user entered carbohydrates and 2 forms of short-term adaption, glucose momentum and reverse correction. Temporary basal rates are used to drive glucose towards a user-specified glucose target. This interoperable algorithm was just cleared by the FDA on January 24^th. This is a huge step forward in allowing for greater choices in care.

Dr Jones: All right. Jen, you have spoken so eloquently on the roadmap to these AID systems and development, and those undergoing FDA review. With my eye on equity, I worry that the rapid pace of diabetes tech development widens the gaps but doesn't close them. How are we, as providers, to ensure that these newer systems don't leave our vulnerable patients behind?

Dr Sherr: I think we need to highlight the potential long-term health economic outcomes that can be achieved if we achieve glycemic targets and minimize both acute complications, including diabetic ketoacidosis and severe hypoglycemia, as well as long-term complications of diabetes. So, the time is now to invest in people with diabetes to keep them safe and healthy, and to help people with diabetes understand the technologies that are out there and overcome any barriers they may perceive in terms of what the technology may be like and help educate them on what it will help them achieve.

Dr Jones: I really appreciate that insight, Jen. Patients should know that diabetes technology offers the opportunity to improve self-care through empowerment. 99% of the time that you spend with diabetes is not in my four-walled clinic room, it's actually in the real world. And these systems create freedom to move through their lived experiences more seamlessly. Providers should know that not all technology is created equal and not all technology is delivered equally. There exist systemic structures in place within our healthcare system and beyond, which limit our patient's capacity for good health. The most effective modality to address these gaps is one, identify them. So, you have to know, right? Have a systematic screening for social determinants of health in your clinic and in your workspace. And two, you have to mediate them. Ensure measurable interventions to address these identified barriers. And then, providers should really undergo implicit bias training and have an appreciable understanding of how these social influences impact the health and wellbeing of their patients, where they live, work and play.

Dr Bode: Jen, do you have any practical tips to optimize these systems?

Dr Sherr: Thanks so much, Bruce. I think this is critical. The biggest tip I can provide is really to set up reasonable expectations of what automated insulin delivery can and cannot do for people with diabetes and their families. We should be considering making insulin to carbohydrate ratios more aggressive. Typically, we have needed to account for the fact that basal insulin will always be provided, but with these systems we can go ahead and give a more aggressive bolus knowing that automated basal can be held back. We need to reinforce the need to bolus prior to meals and talk about strategies if people forget to bolus for a meal. Sensor wear is essential to automation. Help people with diabetes determine how to keep the sensor on as much as possible. Some of that may be related to insurance issues and making sure they have continued access to the system. Other issues could be related to tapes and adhesives and making sure that their skin integrity is okay and that the devices are staying on for their intended duration. For most of these systems, a shorter active insulin time of 2 to 3 hours is likely appropriate. And then, it's really critical to understand things that systems may struggle with. Exercise is still very much difficult to manage and it's going to take a lot of time to figure out how it will work best for an individual. We know many of these systems have the option of an exercise mode or an activity mode or a temp target, and ideally these should be set at least an hour prior to activity. And then, it's really critical to guide people with diabetes not to overtreat lows when on these systems, because if they overtreat lows, there's going to be a rise in glucose. And with these automated insulin delivery systems, that is going to lead to a rise in insulin delivery, which is going to perpetuate this rollercoaster they're experiencing. So, highlighting that they likely need only eight to 10 g of carbs to treat a low.

Dr Bode: Nana, what are your clinical pearls and takeaways?

Dr Jones: I think those clinical pearls that Jen alluded to are very, very important with these automated insulin delivery systems. But again, with an eye on equity, I want to make sure everybody understands that it's unacceptable to have inequities in the access to diabetes technology. The healthcare system needs to prioritize improving technology access from the healthcare system to the public insurance payers. And then, we also need to make systematic changes to our healthcare infrastructure that's needed to improve the diabetes technology access just like in my case, when my diabetes educators walked over from the clinic space into the inpatient space and did pump start education. We really need to think about novel and innovative ways to be able to deliver diabetes technology to our patients.

Dr Bode: Thank you for this great discussion. I'll just provide some closing comments in summary for our learners. Technology is rapidly developing in the field of diabetes. Glucose sensing is standard of care and it's going to get better over times. Hybrid closed loop systems are improving, and most people will achieve a time in range greater than 70%. The goal is to reduce the burden, both for the patient as well as the HCP, is crucial. The progression from hybrid closed loop to fully closed loop is undergoing. And one of the most important things is that you need to recognize racial, ethnic and socioeconomic disparities in use in excess with these technologies. FDA has now made that all of these trials have to have at least 25% not being Caucasian.

And, thank you for participating in this activity. Please continue on to answer the questions that follow and complete the evaluation. Thank you everyone.

This transcript has not been copyedited.