Medical Devices Advisory Committee

НазваниеMedical Devices Advisory Committee
Размер0.59 Mb.
1   ...   5   6   7   8   9   10   11   12   ...   16

DR. GROSS: It is consistent with what the sponsor presented.

DR. JANOSKY: Okay, but not in terms of some of the other information that we heard today.

DR. GROSS: I believe that -- well, and let me -- yes, let me clarify the distinction. Dr. Campbell and John Dawson both presented regression results, R squared, slope and intercept. That information used predicted glucose values from the sensor that had already been calibrated using a fixed intercept regression calibration, but then a subsequent regression analysis was performed on the resulting paired sensor and meter values.

Obviously, those regressions were not unrestrained because they were being used to evaluate the performance of the sensor. But the regression calibration itself and the issue that you raise in terms of the range of predictor values, that regression is done with a fixed intercept.

DR. JANOSKY: Okay. That clarifies it.

Just one other quick question.

You presented to us patient demographic information on your final study. Have you looked at any differential effects between the performance and patient characteristics?

DR. GROSS: We have not stratified the results by those characteristics.

DR. JANOSKY: Would you expect any differential effects?

MR. GREGG: Dr. Mestman, would you please comment on this?

DR. MESTMAN: No, I don't think that we will expect any -- Jorge Mestman -- I don't expect any difference between different groups, patients.

DR. NIPPER: Thank you.

Dr. Gutman, you are going to pass, right?

Dr. Everett.

DR. EVERETT: Again, my questions deal largely with the patient demographics that you used. Initially, it looks like you picked 64 patients. Is that correct?

DR. GROSS: No, there were 62 enrolled in the study.

DR. EVERETT: 62. How many of those actually had Type 1 diabetes and how many had Type 2?

DR. GROSS: I believe we had four subjects with Type 2 diabetes.

DR. EVERETT: Was there any particular reason for selecting such a low number?

MR. GREGG: Dr. Mestman.

DR. MESTMAN: Jorge Mestman.

We tried to select patients with Type 1 because we felt that they were going to give you more information with this type of system. That is the reason we selected more Type 1s than Type 2. We have some patients Type 2, insulin treated patients that were part of this study.

DR. EVERETT: Is a system being recommended to using both types of patients?

DR. MESTMAN: Jorge Mestman.

Personally, from the clinical point of view, yes, I think that the system will be very useful for both Type 1 and Type 2 patients.

DR. EVERETT: Under the section "Age," you have 22 to 74. What you didn't show was a scattergram showing exactly how many patients are 22, how many are 74 or any age in between. How do I know the ages are not skewed more toward 24 or skewed more toward 74?

DR. MESTMAN: Jorge Mestman. I don't know. We have this data available. We have the data. We can show you the data. Do you have the data with the different --

DR. GROSS: This is Todd Gross.

I don't have a distribution of age prepared. As is typically seen in our studies, both with this product and with others, we see a wide range of age. We also see an average of 44 years with a standard deviation of 11. That in and of itself doesn't guarantee that we don't have 60 patients who are 44 and one who is 22 and one who is 74. That is -- I mean, it is unlikely but we would have to provide the distribution of age.

DR. EVERETT: Are you suggesting then that the circulatory system in a 22 year old is equivalent to a 74 year old patient?

MR. GREGG: Dr. Mestman.

DR. MESTMAN: Jorge Mestman.

Regarding diabetes control, fluctuations in blood sugars, yes, I think it is similar. You know, everybody have patients, a Type 1, age 70 or thereabouts and the fluctuations in blood sugars are very, very similar.

Indeed, one of the most difficult patients with Type 1 to treat is elderly because he appears to have in my experience more fluctuation, more difficult to treat.

DR. EVERETT: And then there is race. Do you think race plays no difference in diabetes since what I notice is a majority of the patients are caucasian and then you have three African Americans, one Asian. Is there any reason why you didn't include more of the other racial groups?

DR. MESTMAN: Jorge Mestman.

Well, remember that most of the patients that we selected for the group or we invite to participate were Type 1. So, this is what you see in Type 1 diabetes. There are more caucasians. But everybody was invited to participate in the study.

DR. EVERETT: But you have to decide what groups of patients you are going to recommend to use the system. Do you have any idea of the limitation as it relates to race or age?

DR. MESTMAN: Jorge Mestman.

I don't think there will be any limitations because of age or race. I think that the system will be very useful for every person with diabetes that needs to be controlled.

DR. EVERETT: Do you have any data to suggest that you are correct?

DR. MESTMAN: The only data that I have is the data that -- all the patients that we have studied, but I don't have any particular data to support what you are asking. I personally believe from the clinical point of view that the system will be very useful to a person with diabetes regardless of ethnicity or age.

DR. EVERETT: I guess the thing that really kind of bothers me is this duration of disease. Generally when patients have diabetes, generally, the longer they have the disease, the more complications they develop, particularly as it relates to what we actually find in the bloodstream, as well as the damage done to the blood vessels.

In this particular study, it is, what, .4 years to 54 years?

DR. MESTMAN: Correct.

DR. EVERETT: That is a long time. Did you stratify the data at all as it relates to whether your system worked better based on how long you have had the disease and, perhaps, did it work so well the longer you have had the disease?

DR. GROSS: As I said before, we have not yet -- I am sorry -- Todd Gross -- we have not stratified the results. The results that we provided are aggregate across all patients and across these range of demographic characteristics, which we are proud to say are quite wide in the cases of age and duration of diabetes, as you have identified. But we have not yet stratified the performance results by those variables.

DR. EVERETT: You are recommending that it be used regardless of how long they have had the disease but you have no evidence of how well it works based on how long you have the disease. Is that correct?

DR. GROSS: The only evidence that we have is the aggregate evidence across this group of patients but not -- we do not have a -- we have no evidence that it performs differentially within these stratifications, just that the performance that we have provided is aggregate across those ranges.


A couple of other questions. When you implanted the device -- did you bring the device with you?

MR. GREGG: We do have that for you.

John, why don't you provide the device.

DR. EVERETT: My question is if you accidentally implant the device in some tissue other than subcutaneous or adipose tissue, what happens with the sensitivity of the device?

DR. GROSS: I am sorry. Could you repeat the question?

DR. EVERETT: If you accidentally implant the device in some other tissue than where it belongs, what happens with the sensitivity and specificity of this instrument?

DR. GROSS: We will have to get Dr. Mastrototaro back.

MR. GREGG: The question is in the event that it is inserted in other than interstitial fluid --


MR. GREGG: -- what is the resulting reaction in terms of sensitivity?

DR. EVERETT: Yes. For example, in a very thin person, I would assume it is easy to implant the device in muscle as adipose tissue.

DR. MASTROTOTARO: This is John Mastrototaro.

Actually each of you are free to take it out of the bag if you haven't, but please don't stick yourself. There is a needle guard on all of the devices.

We have actually studied the sensor in some of the feasibility studies. We had a couple of triathletes who participated in the study. So, they didn't have a whole heck of a lot of tissue there on top of their muscle. But the way the sensor is inserted, if you look at it, it is angled. The needle is angled and basically the way that it is put in is you pinch up an area of the skin and then you insert the needle into the area that you have picked up off of the muscle. So, we have really never had one that was inserted in the muscle.

It has always been in the subcutaneous tissue or in the adipose tissue below the skin.

MS. LAPPALAINEN: For the record, what was passed out to the panel is the glucose sensor assembly and the monitor?


DR. EVERETT: What part of this is the sensor?

DR. MASTROTOTARO: This is John Mastrototaro.

Basically, there is the needle introducer, which you can see, which I am holding here -- I have some fake skin, which I am going to just take and do a little insertion here, just to show you how it goes in and what is left -- and then I will pull it out and show you what is left behind in the tissue.

DR. NIPPER: John, I want to know if it is FDA approved fake skin.


DR. MASTROTOTARO: It is actually the subject of next week's panel meeting.


It is inserted at an angle under the skin like that and then the needle is withdrawn and it leaves behind the sensor. Now, the sensor is housed within a flexible tube and the active region of the sensor is exposed in the tube. So, this part here is what resides in the subcutaneous tissue. When you are using it, you can easily pinch the skin up and just slide that in at an angle.

DR. EVERETT: And then two other short questions.

Cholesterol levels. Diabetic patients frequently have abnormal cholesterols. That is they are usually high. Does that have any effect on the sensitivity of the electrode?

DR. MASTROTOTARO: John Mastrototaro.

Well, the sensor resides in the interstitial fluid and one of the things actually that we were a little bit potentially concerned about is profusion, blood flow, and it is one of the reasons that we have primarily used the abdomen and not looked at some of the extremities for the system. But in the patients we have studied, we did collect the cholesterol levels of each of the patients. I don't have that information on me, but we haven't seen any data that would suggest that there is a problem with that, based on where the sensor resides in the body and the fact that it is in the interstitial fluid.

DR. EVERETT: One last question.

Any chance of the electrode itself or the sensor breaking off in the skin?

DR. MASTROTOTARO: John Mastrototaro.

I am happy to say that we have never had that and because the sensor is housed within that tube material, really the only way you could do it is if you got underneath it with a blade and cut the tubing, but I don't see that as potentially happening. So, no. What happens when you pull one of these out is the tube with the sensor imbedded inside of it come out as one piece.

DR. EVERETT: Not a single time has one broken off, is that correct?


DR. EVERETT: You haven't had one break off a single time?

DR. MASTROTOTARO: That is correct.

DR. EVERETT: Okay. Thanks.

DR. NIPPER: Thank you, Dr. Everett.

Dr. Clement.

DR. CLEMENT: Thanks. Steve Clement.

I promise, I just have one calibration question. I am tired of hearing of it. But one of the things that intrigued me on going through the Volume 1 of 2, mailed October 30th, 1998, which actually had the latest manual. I am sure it is going to be updated from then. It mentioned about the calibration part. It said that calibrated the glucose sensor that simply two readings are done and those numbers are averaged and entered in.

From paging through pages and pages of all these graphs, it looked like generally the people have obviously a very big peak after postprandial and then once their insulin kicks, their levels go down. Also, from the preliminary information, it said on using the single point technique that regardless of how a calibration is done is that if there is an error where the blood sugars are actually rising, there is a greater chance that calibration would cause an error in the entire system.

Now, with the linear regression method, I am sure that attenuates that effect to some extent, but still you mentioned that the first point is used as sort of like the stake in the ground type of thing. Is that the -- are you referring to the first point, that when you actually insert it and initiate it, is that value counted more than the other points when the actual calibration calculation is done?

DR. MASTROTOTARO: This is John Mastrototaro.

When they enter a meter value that they put in as a calibration value -- and in the monitor you can put "cal" "yes or no" -- each time they do that. So, at the very beginning of the first day and then let's say they do that again at the beginning of the second day, that stake in the ground simply means that it is performing a real time sensitivity check with that particular value that is put in.

So, even with the regression approach, where all the values entered on a day are used retrospectively to generate the trend plot, that one value that is entered as a calibration value is also used in real time as a sensitivity measure.


One of the suggestions I thought from reading this is that it looks like if this is done fasting, which I assume is when it was supposed to be done, was that -- should that be included in the instructions that it should be done at fasting, assuming that would be where the patient's actual blood sugar is more of a plateau as opposed to going up and down, which may actually cause an error in it?

MR. GREGG: Go ahead, John.

DR. MASTROTOTARO: This is John Mastrototaro.

Yes. For all intents and purposes, the more stable the blood sugars are at the time that they enter that value, the better off that measure will be. So, it may be a good idea if it doesn't say it that it should be preprandial.

DR. CLEMENT: Okay. I have got past the calibration stuff.

One of the practical issues we see on patients, I will give you a scenario of a patient of mine, who is in very good control, on intensive monitoring, intensive therapy, uses an insulin pump. He keeps one meter in his pocket all the time. He keeps a meter in his desk. He keeps a meter on his bedside. They may be from different manufacturers, for example. Some of those meter results he may be entering as calibration events or maybe other events.

How would you respond to someone -- this is real world stuff that happens all the time -- how would you --

DR. NIPPER: Dr. Clement, excuse me for interrupting. You are still talking about calibration whether you know it or not.


DR. CLEMENT: Sorry, sir.

DR. NIPPER: You are allowed. Just as an analytical chemist, I had to throw that in there.
1   ...   5   6   7   8   9   10   11   12   ...   16


Medical Devices Advisory Committee iconVeterinary medicine advisory committee

Medical Devices Advisory Committee iconRanch hand advisory committee

Medical Devices Advisory Committee iconAdvisory committee on immunization practices

Medical Devices Advisory Committee iconNational Vaccine Advisory Committee (nvac)

Medical Devices Advisory Committee iconExternal Advisory Committee on Cities and Communities

Medical Devices Advisory Committee iconWildlife Diversity Policy Advisory Committee

Medical Devices Advisory Committee iconSchedule 5 Appendix c other Medical Devices

Medical Devices Advisory Committee iconPeer reviewed by the Arizona Department of Commerce Economic Research Advisory Committee

Medical Devices Advisory Committee iconFood and drug administration national institutes of health advisory Committee on: transmissible spongiform

Medical Devices Advisory Committee iconAdvisory Committee, Cuyahoga Valley School-to-Career Consortium, Broadview Heights, Ohio 1996-2002

Разместите кнопку на своём сайте:

База данных защищена авторским правом © 2014
обратиться к администрации
Главная страница