Скачать 485.78 Kb.
DR. RIVIERE: I would think there's got to be some databases out there that link through--there's different databases on controlled--again, there's got to be databases that a pregnant woman is put on this drug and taken off that drug, there's a reason relative to response. That might get at this a lot faster.
DR. UHL: There are lots of problems, though, with the longitudinal databases, and the purpose of those databases are for billing. And to be able to extract information for epidemiologic purposes, although they're used, there are limitations to their use. And you would have to know what code you would be looking for for failed efficacy, and that's not usually coded. If a woman's being treated, say, for pneumonia and has failure to improve in 72 hours on this antibiotic and is switched to another one, the code would probably still say pneumonia, and you'd have to capture the switch in therapies so that the clinical record has to be linked to pharmacy record as well. And there's a lot of problems with that.
Hence, our reason for using the PK/PD approach where we're not looking at efficacy in the pregnant woman, we're looking at similar--how can you get a similar kinetic profile at different stages of gestation because, from a PK/PD standpoint, that should translate into efficacy. As long as you know what PD parameter for the antimicrobial interest is the important one. Is it time above MIC? Is it an AUC to MIC? That type of thing. So we don't have to do an efficacy study to get to that. We can assume that if you have the appropriate dose that you'll have the same efficacy.
DR. ROSENBERG: Again, just a question. Is there anything you can do in the hospital setting that, again, because of risk/benefit and because of the controlled environment you could learn about certain things that might be relevant more generally by using a hospital setting where maybe more serious infection or pregnant women are now in a controlled situation where you're getting--you're constantly tapped into the bloodstream, draw as often as you want? Is there anything that could be done there that would help--
DR. WOOD: I think that's similar to some of--the antihypertensive studies, the women are in a very high-risk pregnancy situation.
DR. ROSENBERG: They're in the hospital setting--
DR. WOOD: Well, they're not in the hospital setting, but they're in a very--an outpatient clinic that specializes in high-risk pregnancy.
Do you want to comment?
DR. UHL: They're done like any other kinetic study is done. They're done in a clinical research unit. But if what you're saying is you think it would be easier to study, I mean, maybe--that's what I'm interpreting. It would be easier, potentially, to study pregnant women in the hospital setting.
The flip side to that would be you still have to get IRB approval, and that seems to be the biggest stumbling block. And would these women be potentially at higher risk than women with the same condition that didn't warrant that they were hospitalized? It's a huge can of worms. Could you more easily get the data that you want? Could you more easily get blood draws? Possibly. But you'd still have to go through the same administrative issues.
DR. WOOD: I think the point about the risk, if the woman is hospitalized for her health or for, you know, potential miscarriage going on, you know, the likelihood of interest of participating in a clinical trial versus a PK/PD study is probably less than one who is at a high-risk pregnancy but not to the point where she's in the hospital and worrying about a crisis.
DR. ROSENBERG: No, I didn't mean that. I meant more in a hospital because she has pneumonia; therefore, you've got to treat her with antibiotic, irrespective of her pregnancy; therefore, you're going to get the data you want, not because she's in for another reason. She's in for the reason where the high risk makes you want to have to treat and, therefore, she's going to be treated anyway because it's her risk that's now forcing the treatment. Now the question is: Can you use that population to get the data?
DR. MILLER: Just to interject here, we are doing IV gent, and I believe all of those women, even in this day and age, will be hospitalized. Maybe with HMOs I guess I should--but they are going to be ill and they are going to be in the hospital, and they're going to be receiving IV gentamicin, and the consortium of COEs are going to be addressing that by getting five or ten women from each of the different sites in order to drive our population PK approach for that drug.
DR. DAVIS: I guess you have to be careful if you so focus the group to those who happen to be available, because in the end, the question is: Is it gentamicin that you really want to know about, or does it--because it is a highly used one, or is it just because those patients who get on it are so sick and clearly use it anyway, so--but I think you have to define where you're trying to go with the data, or does it represent a group of drugs and so you're going to look at this as a class, because you're not going to be able to test every drug. So if you only test ten, do you test the ten you can get, or do those ten represent something or how you're going to interpret the results of those ten in light of all the other drugs? But if they're the top ten, you don't have to interpret in light of other drugs because they are the ones most commonly used and will stand on their own? And I'm not sure what the premise is.
DR. WOOD: I think we've taken it from a couple of--I mean, I think we've combined those. We're taking it from a couple of different strategies. The first time around we said we can find a few studies and we want to look at products that are commonly used in chronic conditions because we think those are a high-priority set of conditions. And based on the proposals we got, we ended up with the two antihypertensives that we're looking at.
With the counterterrorism thing, again, it was, you know, we have now a group and, you know, based on a priority setting of both the stockpile and what we think vulnerable populations might be exposed to or might need to be exposed to in a terrorist attack, and then, again, waited to see what was available, what's convenient within that group and what's doable--rather than what's convenient, maybe the word doable, and we sort of went that strategy. But there probably are other strategies. You know, we--and we could continue on and sort of dig down in those areas, or we could move on to other classes or priorities.
DR. LANGER: I had a question that relates more to the second than the first, but is there--and just my naivete. Has there been good research or significant research done on sort of trying to model, like say you understand the physical, chemical properties of a drug and you understand it's half-life, you know, so there's--say you know pharmacokinetics and biodistribution in a non-pregnant individual, could you--has there been any good modeling studies to say if you have that kind of information that you could then make even a reasonable prediction on a pregnant one?
DR. WOOD: I haven't heard of any. Have you?
DR. UHL: I don't believe so. To stretch that a little further, there has been a lot of discussion about, for example, drugs into breast milk based on physical/chemical properties and such. And in that case, there's enough discussion about the differences of across animal species and animal milk that we're not sure how well they correlate.
As far as pregnancy is concerned, the changes are not static. They're very dynamic changes throughout the nine months of gestation that to model that would be very problematic.
DR. LANGER: I'm sure it would be very difficult, but, I mean, the question is--but I wonder if that's--I guess what I'm trying to get at, is that a direction of research that's worth pursuing? In other words, let's say you could encourage NIH to put out a request for proposal. I'm sort of thinking in answer to the second question, you get NIH to put out a request for proposals to try to do those kinds of models. It seems to me that that's something that pharmacokineticists or chemical engineers might be interested in doing if they were encouraged to do it. I guess the question would be whether it would be of any value to you.
DR. UHL: I'm not familiar with anything in the literature about doing modeling like that on the pregnancy side. There's a lot of information in the literature about physical/chemical modeling and PK/PD as far as lactation is concerned, but this would--to do some pregnancies sounds to me like it would be creating a new field or--which is good, but with limited resources, where do you want to go with that?
DR. WOOD: I think your point about NIH--
DR. LANGER: But that's number two, not number one. I'm not suggesting that the FDA fund it, but what I'm thinking about is actually, you know,--
DR. WOOD: That might be less scary to people.
DR. LANGER: Yes, exactly. But it seems to me NIH has initiatives where they want to encourage research in the general area of women's health, seems to me--you know, they just established like a new sort of bioengineering and radiology. It just seems to me that that's the kind of research that could get done, you know, and maybe there's models that would be of some value. That's really what I wanted. So then the question is: Could you initiate discussions with NIH to try to--you know, they often put out these requests for proposals to see if people can come up with some good ideas.
DR. WOOD: Yes, they're putting something out. Has it gone out? Not out. Coming out soon. The NICHD is very interested in the concept of sort of pregnancy as sort of--pregnant women as therapeutic orphans, and they--but their interest is clearly much broader than FDA's in terms of understanding the pharmacology of pregnancy, our interest being, you know, as evidenced by the research that we fund, being fairly targeted at how we're going to get useful information that can ultimately go on a label that can help with dosing and clinical practice, their interest being much broader in terms of the whole physiology, pharmacology of pregnancy and how that interacts with medications. And they are planning to put out an RFA with a very broad mission and a very general RFA in this arena. And I think your point is well taken, that that idea of modeling may not be in there, and it would be something worth to go back and look and encourage them to do because that's an interesting question.
DR. LANGER: Yes, because I would think myself, just speaking for the bioengineering community, and I also think that there's a whole group of people in schools of pharmacy, pharmacokineticists, you know, that that would be a real good challenge, that there might be some contributions that could be made there. So if they were encouraged to do it, you know, that might give you--because it just seemed to me when you were talking earlier, you know, you realize you're talking about changes in fluid and things like that, but maybe there's ways to get some--I don't know--
DR. WOOD: Yes, and depending on where we are, we always need to go a step further.
DR. RIVIERE: There's been studies done on--toxicology studies and teratogens and placental transfer, and I don't remember when, but I've seen PD/PK models of modeling pregnancy, that that kind of work, you know, physiological-based model--
DR. WOOD: And then moving it beyond, because I think the--
DR. RIVIERE: But I don't--that's the problem. Animals and humans is a huge jump.
DR. LANGER: Right, that's one jump. But, see, the other way of thinking about it is--I'm not just thinking about one drug. In other words, I think what you'd like to think about is if you understood certain fundamental parameters, like if you understood the physical/chemical properties of it, like, say, you know, that it will lead to diffusion coefficients, partitioning. You know, so then--and, you know, the half-life then maybe there's some predictions that you might--you know, what I'm thinking about is sort of a simple model--I mean, it wouldn't be that simple, obviously, but could you plug something into the computer some day--
DR. RIVIERE: Some people have done some of that. The problem, again, is it's done in rodents, and then--
DR. LANGER: No, but I'm--yeah.
DR. RIVIERE: --the physiology is so--and you can do that. I agree. But there has been some work done on that.
DR. WOOD: But, again, that's not asking the question about how the physiology of the mother changes in terms of her PK/PD--
DR. LANGER: That's right.
DR. WOOD: It's more placental transfer, which is, again, more akin to the lactation--
DR. LANGER: It's just one question. I'm trying to think about a broader context, I guess.
DR. WOOD: That's an excellent idea.
DR. LANGER: Maybe we can see if there's more answers or suggestions to one, two, or any of the questions. Do people have any other comments on the first or second question, or the third? Do we want--
DR. WOOD: I guess I would also have a question. Does this--you know, we've had the experience both in terms of the labeling regulation and doing this research, that there is a very understandable hesitancy for people to take on this challenge. I guess in terms of how--you know, does explanation relieve that anxiety or is it--you know, or is it something we're never going to be able to overcome?
DR. FEIGAL: Could I just ask a question on the third one? One informational question. Has there been a prospective registry larger than the acyclovir registry, or is that the largest?
DR. WOOD: That's the largest.
DR. FEIGAL: That one sort of illustrates kind of the problems with the registry strategy because it took them about ten years to get 1,000 exposed women, and then only about a third of those were exposed in the first trimester. And so they ended up with really only the power to exclude a doubling of all birth defects if the hypothesis was that it would cause all those defects to evenly go up, and it only had a power to pick up a ten-fold increase in any specific common birth defect.
And so I think the setting in which exposure cohorts are useful is something to take a close look at, because when we then looked at whether or not that registry should continue to get more power and you realize that power is the function of the squaring the sample size, to double the power we would have had to have gone, you know, to another 4,000 women. Obviously it's a very small fraction of the people who were exposed.
My other question is if you don't start with an exposure cohort, you can start with an outcome cohort, and there are birth defect registries. And if you actually said for a second that, well, most of the exposures identified in the birth defect registry aren't going to be the cause of the birth defects, because we're not making much progress there. But it is an interesting source of information about what drugs those people took, because they very systematically interviewed, and there, when you're interested in the outcome, the dilemma is that they better remember the drugs they took, and so you've got recall bias and you have problems doing causality for the outcome. But if you say I'm not interested in the outcome, I'm interested in finding out what drugs pregnant women are exposed to, and you realize a third of all women--not a third, 3 percent of women have a birth defect, and you've got states where reporting is mandatory, like California, you may be able to get quite a bit of your exposure targets to select the drugs that are feasible to study as a byproduct of some other kind of research. So that's another suggestion.
|Nutraceuticals, nutrigenomics, Public Health Nutrition, Clinical and Therapeutic Nutrition, Institutional Food Administration, Food Science, Food Safety, Food Toxicology and Quality Control||Project Number: s-295 Title: Enhancing Food Safety Through Control of Food-borne Disease Agents Duration: October 1, 2000 to September 30, 2005|
|The microbiological safety and quality of food / M. Barbara Lund, Tony C. Baird-Parker, Grahame W. Gould||Ф кгма 4/3-21/02 пп кгма 4/02|
«Actions for safety and preservation of food and biological value of foodstuff. An estimation of high quality of a flour, bread,...
|Information Systems & Modeling for Counterterrorism||Food and nutrition in food allergy foreword|
|Kudler Fine Food is a western United States based company, which is situated in San Diego metropolitan area. It furnishes quality food products at the||Joint Convention on the Safety of Spent Fuel Management and on the Safety of Radioactive Waste Management|
|Visionary grassroots education initiatives||United States Ocean Observing Initiatives – a look to the Future|