Скачать 0.59 Mb.
DR. MARCUS: Good morning. And I would like to offer my profound respect at being able to appear before this advisory panel.
My name is Alan Marcus. I am a consultant to MiniMed. My trip to today's meeting was paid for by the company. And as a point of information, as is company policy, a policy followed even by our founder, Al Mann, I flew coach.
My family and I have no ownership interest in MiniMed.
What I hope to outline today are four distinct points; No. 1, the clinical utility of standard blood glucose monitoring and the potential benefits to be added to patient care by the addition of the information obtained by continuous glucose monitoring; the CGMS implementation process, as it will occur in a real clinical practice; the use of patient glucose profiles as part of the education and care of people with diabetes and my conclusion, which I hope you will share, that this is a device that will benefit the entire diabetes care team and its most important member, the patient with diabetes.
As you have seen this slide before, this slide depicts the variation between those people who have diabetes and their glucose fluctuations and those people who don't have diabetes and the relatively small fluctuations that occur in their daily blood glucose. The mission that all of us who were part of the diabetes care team is to approximate as closely as possible the blood glucoses of those people who don't have diabetes.
This prophetic mission statement was given to us more than seven decades ago by Dr. Joslin. Recently, as the result of the DCCT, the UKPDS and the Kumomoto(?) study, all of which have demonstrated that by achieving an average blood sugar or a hemoglobin A1c of less than 8 percent, you can forestall or prevent complications, stabilize complications and, indeed, reverse complications.
Our target has been to approximate that hemoglobin A1c, which we know will prevent danger to the patient. Unfortunately, as has recently been demonstrated in NHANES, the average hemoglobin A1c in the United States is 9.6 percent and only 10 percent of people who have diabetes are actually achieving hemoglobin A1c's at the target range that we desire.
Now, how many of self blood glucose meter measurements need to be within the target range as defined by the DCT of 70 to 120 before meals and 180 postprandial, to achieve a hemoglobin A1c target as set for us out 7 percent. Well, this study was initially performed and authored and then published by the researchers and the diabetes care team at Barbara Davis Center.
They had brought a young adolescent female, who in her and her parents' goal of achieving normal blood sugars suffered recurrent hypoglycemia, accompanied frequently by seizures. So, they set forth and analyzed a number of blood glucose meter measurements that were required to change the hemoglobin A1c of 8.5 percent demonstrated on the right pie graph, to a hemoglobin A1c of 7 percent to the left.
The lightest area demonstrates those below target, the area within target range and the area above target range. As you can see, it takes to go from hemoglobin A1c of 8.5 to hemoglobin A1c of 7 percent, that those within the target range of 70 to a 180 need to increase by about 17 percent.
Unfortunately, that is accompanied by a below target range increase of 50 percent or an increase in the chance occurrence of hypoglycemia by 50 percent, while achieving an above target range reduction of 30 percent.
Now, if not for the introduction two decades ago of self blood glucose meter measurements, none of the studies that I have spoken about earlier and, indeed, no studies looking at control of diabetes and its effect on complications could have been performed. Unfortunately, although we recommend our patients to perform four to six self blood glucose meter measurements per day as a form of gathering information that will benefit them and other members of the diabetes health care team.
The average number of blood glucose meter measurements performed by people who have diabetes in the United States, who take insulin, extrapolating from Maureen Harris's data in Diabetes in America is approximately 1.8. In other words that gives us two points during a 24 hour period to look at, to analyze, to make suggestions or revisions in our overall diabetes management.
The CGMS system, which obtains 288 points per day allows us to look at trends in profiles that occur during that day, allowing us to make better interventions and for the patient to make better adjustments and the entire member of the -- all members of the diabetes care team to improve the care of the patient with diabetes to achieve our ultimate goal.
How would the implementation process occur? Well, the candidate would need to be selected. It could be a patient who has recurrent hypoglycemia, poor control, a patient who is undergoing changes in glucose control because of status of menses. It could be a patient who is traveling unaware of the influence of food varying. Patients would be selected on the basis of need.
The CGMS would be prescribed in a prescription form by the physician. He would prescribe one sensor, two sensors, what is needed for the period of time, knowing that each sensor would monitor a three day period. So, for longer periods of time, he would prescribe two or more sensors as deemed medically necessary.
The patient would then leave the office after first entering the calibration or initial two glucose meter measurements and perform three days of normal, everyday living activity. During these normal activities, not only would meter values be entered, but other things, such as the type of food taken, activity, whether they were happy or angry, whether they slept well, whether they got ill, medications that were taken, all the variables that enter into the patient's ability to control their blood glucose.
Following this, an office visit would occur. The data would be downloaded in a glucose profile report, together with the other reports demonstrated to you by Dr. Mastrototaro would be generated. The entire diabetes care team would have the ability to interpret that report, the nutritionist, the diabetologist, the practitioner, the patient, most importantly, the nurse. the diabetic nurse educator and the nurse practitioner.
Recommended therapy changes would be based upon all available data, self blood glucose monitoring, as well as CGMS data. The information of the fluctuations caused by alterations in food and exercise could be analyzed and adjusted. Several days or weeks could pass. Assessment of the patient would then occur and if the physician felt necessary could write another prescription for CGMS monitoring to reassess the value of the interventions that were performed by all members of the diabetes care team.
The effectiveness of these interventions could then be confirmed. I am going to give you two patients and three examples of self blood glucose meter measurements that were obtained during the study and interpret them as a clinician would. As you can see on the bottom line, are the meal times. These are the self blood glucose meter measurements obtained and this patient actually performing four blood glucose meter measurements is in the top 15 percent of all patients with diabetes who require insulin or take insulin in the United States because 85 percent do not perform this number of self blood glucose meters.
So, if you looked at these four points, you would notice that they are above the target range at breakfast, lunch and dinner, of 150 and actually at bedtime at about 120. If I were to make a change as a clinician, my impulse would be, well, blood sugar is a little bit high before breakfast. I am going to increase the insulin at bedtime, maybe give a long lasting injection of NPH, possibly adjust the insulin that I am giving at breakfast.
Now let's see, with the added information obtained by CGMS, what would have been the results of my information, limited as it was, and the consequences of my actions? Well, by giving more insulin over the bedtime period, this period of hypoglycemia unaware to the patient and the diabetes health care team would have been profoundly worsened by administration of insulin.
The hyperglycemia that occurs postprandially after breakfast could have been missed, could be adjusted by dietary interventions rather than just by interventions in insulin administration. If I had targeted those areas to perform self blood glucose meter measurements, I would have been able, as this patient was because of the alarm, to pick up these areas of low blood sugar trending and high blood sugars and, therefore, to make an intervention that would have been appropriate for the patient and not based upon limited information.
Here is that same patient again. Only in this time, clearly the blood glucose at breakfast and lunch are in the area above 200 that we would all consider hyperglycemic. Once again, the blood sugar at dinner, somewhat in the lower range, what adjustment would I make? An elevated blood glucose in the morning, once again consistently, I would probably give an increase in NPH. I would try and obtain a 3:00 a.m. glucose, but failing that in my attempt to improve diabetes control and prevent complications, I would have made the same changes that you saw earlier. The result of those changes on this day would have proved just as catastrophic.
As you can see, now the overnight hypoglycemia occurs from the midnight time to the early morning awakening. For this patient, this trend and pattern for their glucose excursions prevents a significant risk; hypoglycemia, as we know, being a major cause of mortality in diabetes, as well as complications in terms of acute care costs, admissions to ERs and lost of frontal cortex tissue as demonstrated by magnetic resonance imaging.
If I had been signalled or been alerted by CGMS to instruct the patient to perform those very valuable self blood glucose meter measurements at specific times, I would have picked up areas of low blood sugar, areas of high blood sugar. I would have known when to ask the patient to check to achieve the most from a very limited resource.
Here is our second patient. On this patient you can see these blood glucoses are actually within the excellent range and probably no intervention would be necessary, just reviewing the glucose measurements obtained by self blood glucose meter measurement.
Add to this information the CGMS and you would see that even though these four dots, these four points of glucose measurement are actually okay, there is hyperglycemia occurring postprandially again on breakfast and severe near hypoglycemia occurring throughout the early morning hours.
This overnight refining of hypoglycemia is totally unavailable by the current methods that we have now unless we were to ask the patient to check their self blood glucose meter measurements continuously through this period of time. As clinicians, all of us know that we are able to achieve this for a short period of time when we care for patients, for instance, who have diabetes in pregnancy.
Unfortunately, we are all also too aware that after delivery the frequency of self blood glucose meter measurements falls off sharply. We need to add to our information and to the patients information, give them a reason why it is appropriate and important for there to be this metering and monitoring.
Once again, targeting this area shows the correlation of the self blood glucose meter measurements for both hyperglycemia and the significant hypoglycemia or near hypoglycemia. The fluctuation in glucose, as a clinician, one could imply is the body's desperate attempt to prevent hypoglycemia, something that we know the body does by the release of counter-regulatory hormones.
Through the mean absolute error, the correlation coefficient, by being able to view these trend plots in the modal day, the physician would be able to make use of the data to determine if this data provided by CGMS would be weighted heavily, not so heavily, how to incorporate it into the overall scheme and plan for diabetes management.
In conclusion, conventional blood glucose monitoring, looking at three snapshots, not looking at the movie that occurs for 24 hours is not adequate for assessing patients overall glycemic control. Optimal self blood glucose monitoring guided by CGMS, aided by CGMS may decrease the risk of glucose excursions, both hyper and hypo.
The CGMS provides information that allows the physician to make more informed decisions regarding patient treatment regimens and to educate the patient as to daily trends, which occur in glucose fluctuations. Clearly, our role as clinicians is to enable the patient with a chronic disease to care for themselves on an ongoing 24 hour a day, seven day a week basis. The CGMS provides trending information, which targets times for glycemic control and allows the patient to have the luxury of additional information, which will enable them to do so.
The data obtained from the multi-center clinical trial provides a reasonable assurance that the CGMS is safe and effective for its intended use as currently studied and functioning.
The diabetes care team includes a very important member, the person who has diabetes. This information allows that person to learn more about what is going on inside their body and the various activities that take place, insulin, exercise, changes in health status, food and nutritional intake that affect and how they affect their glucose levels, ultimately affecting their morbidity and mortality.
So, we end as we began with the indications for usage. People with diabetes mellitus, clinicians and other health care providers for them need to actively participate in the care of people with diabetes. This is an additional tool, which when added to my toolbox in caring for people with diabetes will allow me to give them additional information to seek the goal that we all want.
DR. NIPPER: Thank you, Dr. Marcus, and thank you to the MiniMed presenters.
At this time we are going to break. I am sure Dr. Habig will be happy about that. We are going to reconvene promptly at 10:40 and listen to the FDA presenters.
Thank you very much.
DR. NIPPER: Thank you for helping us maintain our schedule.
Our next presenter will be the Food and Drug Administration. Dr. Gutman has already been introduced. He is the director of the Division of Clinical Laboratory Devices, Office of Device Evaluation. We are pleased to have him present to the panel.
Agenda Item: FDA Presentation -- Overview and Introduction
DR. GUTMAN: Good morning.
Since the introduction of the first home test systems for glucose measurement in the early 1960s, this technology has been a cornerstone in modern diabetic management and has provided patients and physicians with one of the most powerful and significant tools for improving outcomes in this important disease.
The merits of tight control with the use of home measurement systems were well demonstrated in the studies reported in the Diabetes Control and Complications Trial. Results of this study have been extrapolated to management of Type 2, as well as Type 1, patients and serve as important grounding for both the diagnostic and therapeutic approaches and objectives being applied to dealing with the disease process to date.
The introduction of minimally invasive or non-invasive techniques for monitoring glucose at home or at the bedside have profound potential impact for diabetic medical care. This technology offers potential improvements in the quality of life, enhanced control to increase frequency of testing or access for testing in a broader range of patients and the potential to develop new insights in both the treatment and biology of this complicated disease process.
Recognition of the importance of these new technologies is not limited to those directly involved in the diabetic care community. It is widely known that many groups are working on various devices in this area of new technology and the agent is very committed to interacting with sponsors at any and all stages of development to help bring these important advances to market.
In addition, FDA is interested in working with sponsors on new tools to replace, to supplement or to complement existing testing methodologies in the pursuit of better ways to manage diabetes.
The submission being considered today is for us a first of a kind device under review for minimally invasive glucose measurement. The submission was jointly reviewed by members of two offices and three divisions. We had input from both the Division of Clinical Laboratory Devices, the Division of Dental Infection Control and General Hospital Devices and the Division of Biostatistics in the Office of Surveillance and Biometrics.
The intended use, as stated by the sponsor -- and I will try to quote the sponsor -- is to continuously record glucose levels in patients with diabetes mellitus. This information is intended to supplement, not replace blood glucose information obtained, using standard home glucose monitoring devices. The information collected by the MiniMed sensor will be downloaded and displayed by a computer and reviewed by health care professionals.
This information may allow identifications of glucose level excursions above or below the desired range, facilitating therapy adjustments, which may minimize these excursions.
Based on discussions with the sponsor, it is our understanding that this device is intended for prescription use only, will not allow readings to be made available directly to patients in real time, provide readings that will be available for review by physicians only after the entire recording interval, which now is suggested at 72 hours, is currently intended for occasional rather than everyday use, is to be used only as a supplement to and not a replacement for standard invasive glucose meters and, finally, is not intended to change patient management based on the numbers generated alone, but will guide future management of the patient, based on responses to trends noticed; that is, these trends or patterns may be used to suggest when to take finger stick measurements to better manage the patient.
FDA is looking for input on how to understand and evaluate the performance of this groundbreaking technology. In particular, if this panel determines is ready for market, we seek advice on how to label it. If this panel determines it is not quite ready for market, we seek assistance in determining what further steps might be required to characterize, to evaluate and to label it.
This device has been studied in a small population. The clinical trial is in just over 60 patient, but has generated a large data set, literally thousands of data points. Several different means of data analysis have been applied by the sponsor to this set. Most recently, as they discussed with you this morning, they have introduced a linear regression calibration technique and they have evaluated data using an interpretative model based on identification of five zones, including three to signal hyperglycemia, hypoglycemia and uglycemia and two to signal uncertainty in glucose measurement status.
We are grateful as an agency, the company has been quite forthcoming. They have, in fact, shared all of their data sets with us in electronic form and as they have indicated, the most recent interpretation came into the Agency this past February and we continue to work with them and to work internally on the data management. We would view this, frankly, as a work in progress in terms of our trying to understand the calibration, to understand the performance and understand the labeling.
We are seeking input on a wide range of issues. We have some specific questions I would like to pose globally and then specifically, but, first, I would like to share with you, have our statistician share with you some of our analysis of that generous data set.
John Dawson, I believe.
DR. NIPPER: Thank you, Dr. Gutman.
|Veterinary medicine advisory committee||Ranch hand advisory committee|
|Advisory committee on immunization practices||National Vaccine Advisory Committee (nvac)|
|External Advisory Committee on Cities and Communities||Wildlife Diversity Policy Advisory Committee|
|Schedule 5 Appendix c other Medical Devices||Peer reviewed by the Arizona Department of Commerce Economic Research Advisory Committee|
|Food and drug administration national institutes of health advisory Committee on: transmissible spongiform||Advisory Committee, Cuyahoga Valley School-to-Career Consortium, Broadview Heights, Ohio 1996-2002|