Clarification of VMAC Questions - Moderator, Dr. Michael Doyle, Acting Chair
DR. DOYLE: All right. Thank you, Aleta. First of all, I want to thank all of our public presenters for their comments. They are very much appreciated and a very important part of this process.
I think we should begin by asking just the general question: Do any of you on the Advisory Committee have any remaining questions, from a general perspective, of the presenters? And if so, we will pursue those.
After that, we will then bore down and address the two questions, of which one is a two-part question, and again do that by general questions first and then I would like to have each presenter give a perspective for each question that is asked so we will have that for the record.
So let us open it up for just general questions regarding what we have heard today or any comments regarding what we have heard in general today. Yes, Dr. Poppenga?
DR. POPPENGA: I just have one question of the Pfizer representatives. Is -- are there any ongoing studies to further characterize sensitive individuals or populations with regard to this particular product --
DR. SUTHERLAND: Nothing --
DR. POPPENGA: -- the better to find the mechanisms of reaction?
DR. SUTHERLAND: Oh. I will answer it in two ways. One, nothing ongoing specifically right now other than the continued monitoring of adverse events and digging into each event as it occurs and trying to understand the drug in that manner, like we do all of our products.
But I think one misnomer that seems to have been conveyed is that there were no studies, safety studies or toxicologic studies, done post-approval, and there were multiple studies to try to understand the allergic potential of the drug, the impact of residual solvents, receptor binding studies, pharmacokinetic studies, repeat dosing studies to demonstrate that there wasn’t accumulation and that there wasn’t an increase in adverse events with repeated dosing.
So there is a database out there that has been openly shared with the Center, discussed with the Center, on file. All of that information is available and so there have been a lot of studies done to help us better understand the adverse effect profile relative to the allergic potential.
DR. DOYLE: Okay. Any other questions? Dr. Senior?
DR. HARTOGENSIS: I would like to add to that. Can I add to that?
DR. DOYLE: Oh, please, yes.
DR. HARTOGENSIS: I just wanted to clarify the Freedom of Information summary that refers to no new data, that is actually the Freedom of Information that refers back to the label changes, just specifically for those label changes. There was not data for that -- to support the label changes. And what Steve is talking about were other studies that were evaluated. So just want to clear that up.
DR. DOYLE: Okay. Dr. Senior?
DR. SENIOR: I noticed that the reporting from Banfield, the database that they keep, was an order of magnitude higher than the reporting that -- of adverse effects that was reported to the Center for Veterinary Medicine. And I just wondered if -- did I hear that right? Is that difference in fact correct? Are those numbers correct?
DR. DOYLE: Anyone can answer that question?
DR. HODGE: I will speak to that question. I think I talked about that when I talked about the incidence rates.
Obviously, the Banfield data really talks about the animals that were actually treated, so there will be differences in that, the reaction rates. You know, if I compare what I have, if I look at my vaccine reaction rate divided by the doses distributed, obviously those aren’t always equivalent to the doses administered. When you look at the Banfield database, that is actually the number of patients, and so the reaction rate is going to be higher.
It is also what is in the database, whether it is reported after 3 days or 30 days, and we are relying even under an enhanced PV program, pharmacovigilance program, that the people actually have to pick up the phone and call.
So that is not a surprise to me. If I look at any of the literature, you will find that discrepancy.
DR. SENIOR: So we could expect that the actual adverse events that took place would be 10 times what is reported under the --
DR. HODGE: Not necessarily. And those may be -- and maybe Dr. Glickman can talk a little bit about the differences as well. But I would not necessarily be expecting that the real rate may be any number higher because these are the reported ones that people feel are associated with a particular product, not just events that are occurring of any type in the patient population that has been treated.
DR. DOYLE: Thank you, Dr. Hodge. Do you care to comment, Dr. Glickman?
DR. GLICKMAN: Yes. To partially explain the differences, 1, I think it reflects the difference between passive surveillance and active surveillance. So, for example, when we were looking at -- systematically at the Banfield records, if an animal vomited 3 days, 28 days, following the administration of a heartworm preventive, we counted that. It is very unlikely an owner would have reported that to anybody so that through the FDA database or the Pfizer database, that would never appear.
But we counted every event that the doctors or technicians recorded in the medical record and then compared rates.
DR. DOYLE: Any other --- one more?
DR. SENIOR: This -- it is alluded to several times and both in oral presentations and also in some of these written materials that we have been given that there is missing data. And I want to clarify that.
I -- in reading through *Laurie Renta’s presentation, there is -- she refers to slides or information that Dr. Hampshire showed her superiors in Wyeth which can be found on an url that is given. And I checked that url and it doesn’t lead anywhere. It is -- it doesn’t come up.
So I just wanted to clarify: Is there any data that we have not been given? Is there anything out there that we have not been given relative to the experience with ProHeart 6 prior to its withdrawal in 2004?
DR. McCHESNEY: I think all the FDA website has gone -- undergone changes so that a url -- it is regrettable that a url doesn’t work, but it is not surprising.
That data is all available. It was -- it has been on the website. So I don’t think -- I don’t believe there is anything we are having -- if it is, we can surely make that available. That should be able to come up.
DR. MURTAUGH: Well, to follow up with the same question, there were several references in the docket materials we have to increased mast cell tumors, for example, that -- and other suggestions of increasing rates of adverse events referred to in this presentation of Dr. Hampshire's. And so I had a similar question regarding the -- I will call it -- well, the actual status of information that was alluded to primarily in the report from Senator Grassley’s investigation.
DR. DOYLE: Thank you, Dr. Murtaugh. Any comments?
DR. HODGE: I will comment to the best of my ability. But regarding the mast cell tumor, this was an epidemiological study, again using the Banfield data, with Dr. Glickman, and they did find a slight increase in mast cell tumors. Whether or not that was significant or not, it was not great and it certainly did not trigger for us a cause for alarm.
One thing that I wanted to point out that hasn’t really come up is that Moxidectin is a macrocytic lactone, as are all the heartworm preventative products, and a lot of what we are talking about today are really class effects of all of those products, of which Pfizer has another one as well. But just to put that into perspective, that all of what we are talking about today really are class effects of that class of products.
DR. DOYLE: Thank you, Dr. Hodge. Any other questions? Dr. Murtaugh, is that sufficient? Dr. Senior?
DR. SENIOR: There was a mention made of some reporting required by Fort Dodge Animal Health relative to manufacturing. Was that manufacturing reporting related in any way to production of this product? It was one of the comments from the floor.
DR. McCHESNEY: Yes, I think the answer is to the product prior to withdrawal from the market. Since then, in the past 18 months since this RiskMAP has been in, Fort Dodge -- when the product still belonged to Fort Dodge -- the facility was re-inspected because we wanted to be able to, one, satisfy ourselves that they were reporting everything they were getting and looking at the facility. And there were -- John, there were a few problems found, but nothing that was deemed to be -- it was discussed with them and they provided the information.
DR. BAKER: Right. There was a -- they lacked the -- I think the lot information on the ADE reports for the actual product that was given. So Fort Dodge went back and recovered as much as they could on the ADE reports so we could see what lots were given to what animals on the reports that were received. That was one of the things we had asked for in the RiskMAP, is they keep a close eye on the lots to see if there was any manufacturing relationships we could identify through the lot information. So they did correct that.
DR. McCHESNEY: And the other point -- I don’t want to speak for Pfizer, but the other point is that the product that is sold internationally was, I think, and still continues to be made, at the original facility, and that was the facility FDA inspected sometime last year.
DR. DOYLE: All right. Thank you, Dr. McChesney and Dr. Baker. Another comment or question by Dr. Senior?
DR. SENIOR: Yes, if you don’t mind -- last one.
The -- and this goes to the Dr. Glickman statement and it refers to the possible distinguishing features that you might apply to a RiskMAP. And he led us to believe that Banfield hospitals used 3 different heartworm prevention medications, and correct me if I didn’t hear this properly, but there would be the ProHeart 6 and then alternative 1 and alternative 2.
Also, he mentioned that it could well be that the clinicians in the hospitals were selecting when they saw a high-risk patient or some distinguishing feature where they would select one preventive over another.
I am just wondering: What criteria was the computer that guides the treatment programs -- what criteria is in those computers to guide the clinician regarding prescribing that would have -- that would select for one heartworm preventive versus another? And so, you know, has Banfield identified some safety selection process within their system?
DR. DOYLE: Anyone care to comment? Dr. Glickman?
DR. GLICKMAN: I can address it at least for the time when we did the study. There were really only two products that Banfield was using at the time. One was Ivermectin, an oral heartworm, and the other was injectable Moxidectin, ProHeart 6. There was a second oral but it was very infrequently used.
Both were in the formulary and were intended for use in healthy dogs. And the decision on which to make was up to the owner, depending on whether they preferred oral or injectable, and they were told that the -- what would be required with each. And it was really the owner that made the decisions, nothing in the computer making the decision, nor was there any selection factor based on health.
There was a selection factor, though, based on health for all heartworm preventives. Someone remarked, and I forget who, that -- I think the statistician FDA -- isn’t it interesting that the dogs that did not get a heartworm preventive had a much higher rate of, for example, liver events and other events? That is to be expected, because it was the dogs that were in the wellness program who were deemed to be healthy who were given a heartworm preventive of any kind. If they were sick, the veterinarians generally deferred heartworm preventive in order to treat whatever the problem was in the dog. So that is not surprising.
DR. DOYLE: I thank you, Dr. Glickman. Any other questions? Yes, Dr. Apley?
DR. APLEY: Thank you. Back to the previous data before the original withdrawal. Dr. Baker in his slide, where you compare June of ’01 to September of ’04 to the recent RiskMAP data, those data would be the ones we are talking about that would have been presented at the first VMAC?
DR. BAKER: Yes, sir.
DR. APLEY: We talked a little bit about the -- this potential for carcinogenesis. I don’t see any of those data represented on that chart. Was it is not on the radar? Did you not have any?
R. BAKER: These were the elements that were discussed at the 2005 VMAC, these clinical signs.
DR. APLEY: Those exact signs?
DR. BAKER: Those exact signs, yes.
DR. APLEY: So I am obviously still struggling to get a foothold of expected -- and obviously versus observed, and struggling to put this in relation to other drugs.
I am looking for -- and realize that the numerators and denominators can be apples and oranges, but, Dr. Hodge, you mentioned that this is a class characteristic. Could you compare? Can you give me an idea of -- to Pfizer’s class representative that is in this group, therapeutic group?
DR. HODGE: You are kind of putting me behind -- with a rock and a hard place.
DR. APLEY: Yes, I know. I put you in a spot! Don’t worry. I am fixing to do it to others, too, so don’t worry.
DR. HODGE: I will tell you this, that because of the RiskMAP, the classification and the description of a safety profile is more difficult, and with the ProHeart 6 we have at this point a really clear safety profile for this product.
Do we see this type of reaction with our other products and the industry people have to go? No -- just kidding! Basically, a lot of the same effects.
We get -- of course, our product is a topical product, so when we rank our adverse events, we are going to get of course more application site disorders with a topical product. That is not a big secret.
We get digestive upsets, as we do with Moxidectin, and you can get also allergic phenomenon, neurological events. And so, really, it is sifting those out and creating safety profiles.
Now, what is interesting as well is this is also simplified because we are really only looking at the dog, and for our other products we have the cat and the dog, and so there is a distinct safety profile for the cat and 1 for the dog, and within those, it is fairly over a period of time and really looking at -- ProHeart now is not a new product, so I would venture to say that the safety profile we see today, particularly after the 8 months on the market, is really -- and with the international data behind that, is really the clear, specific safety profile for this product, and other things that need to be monitored -- as now Revolution has been on the market for 10 years. And we also have identified out of that 10-year body of data the safety profile of that product. And then, of course, being in practice and using some of the other products, all of these events are typical of what you see with this class of products.
DR. APLEY: Thank you. It is probably between the rock and the hard place, but for the Office of Surveillance.
Are we -- maybe I got the answer but I am still not able to internalize it -- what is the trip -- and I think, Dr. Baker, you mentioned that you had ways of ranking kind of cumulative -- the adverse effects that are reported --
DR. BAKER: --- the assessment?
DR. APLEY: Yes. What is the trip wire again? How much different from the other class representatives for this use for heartworm prevention, how much difference does there need to be before you go, oops, it is time for us now to consider our first RiskMAP? Are we talking 3 in 10,000, 2 in 10,000? I am hung up on trying to figure that out.
DR. BAKER: That is a very good question. I don’t think I can give you a definitive answer there for what is the actual threshold.
I think when you start seeing safety -- you know, these safety signals, and you have a very robust pharmacovigilance program that detects them, you know, and that is what -- you know, we are evolving to a signal detection data mining-type software system where we can pick up these events quickly that occur disproportionately more than is expected. And that is what data mining is really all about, is when you look for a drug event pair that is carrying disproportionately more than other drug event pairs for other drugs and you can pick it up as a signal. And we are not there yet, but that is where we want to get, is using data mining to help identify these signals, because they just -- they identify those, that disproportional occurrence of some drug event pair. And, you know, we are not there yet, but that is going to help us get to where you are talking about, so --
DR. APLEY: And maybe you can’t answer this because of confidentiality of other products, but, for example, when this 1 you said, “it is time to act on this,” were we 10 times the other products, were we 100 times the other products, 1,000 times the other products?
DR. McCHESNEY: I don’t necessarily think it is “is it 10, 100, 1,000, 100,000, 1,000,000 times?” I think it is the seriousness of the reactions.
So if you look at other products, you look at some other drugs, you know, you might have a lot of reactions, but if death is high, that would surely be a concern. And I think when we went back and we have seen a decrease in the death rates, I think that is what triggers it, is we look at sort of on the whole. It is not so many, it is not the -- it is not “is there a certain number that triggers it?” It is the seriousness of the responses that we are seeing as a whole and kind of where they fall. And if this -- you know, back in 2000, 2004, death was a significant number of the events and things that may have led to death.
So I think what we are seeing now in this once -- since we moved to the RiskMAP -- so in the interim, we sort of had to satisfy ourselves that something was occurring, or had occurred, or was being out there, that gave us some sense that if we put it on the RiskMAP and we followed it closely, we would be able to act quickly in these areas where they were serious.
I think what we have seen is we have seen an increase in anaphylaxis/anaphylactoid reactions go up if you look at these numbers, and admittedly the denominators are little tough on these, but we have seen death go down by fourfold. And, you know, whether that is better awareness by the veterinarian to react to when they see an anaphylaxis --
So I think that is where we are trying to go with this, and that is -- those are the kind of candidates for RiskMAPs, in my view, where you have something you can really get your hands around and focus on and make a limited change and see how that affects it and then react quickly to it. I just don’t think there is an actual number you can put your finger on.
DR. HARTOGENSIS: And I just wanted to add to that. And in addition, as, Doctor, you mentioned, as you said, it was certainly the severity of the adverse reactions that we were most concerned about, particularly death and anaphylaxis. However, we are seeing, you know, a large amount of lack of effectiveness reports for some of the other products in comparison to this one.
So that is why this RiskMAP has fit in nicely to balance that risk/benefit ratio, because there is the benefit of, you know, of potential less owner compliance issues and effectiveness issues with this product and while managing the risks, which can be -- you know, obviously were severe in some cases, death by anaphylaxis as Dr. Baker’s presentation pointed out. Just in 2004, there were 20 cases of dogs that died from anaphylaxis. And we didn’t see that this time, which, you know, helps point out that we may have mitigated that risk by educating the veterinarian and the owners of what to look out for and be aware of that potential.
DR. APLEY: Did I misunderstand that 20 deaths, anaphylaxis, in 2004, that was for other drugs?
DR. HARTOGENSIS: No, that was --
DR. APLEY: That was for this drug?
DR. HARTOGENSIS: That was for this drug.
DR. APLEY: Okay, I had misunderstood --
DR. HARTOGENSIS: Correct.
DR. APLEY: -- that that is for this drug alone --
DR. HARTOGENSIS: That is for this drug.
DR. APLEY: -- in 2004?
DR. HARTOGENSIS: Correct.
DR. APLEY: And in the RiskMAP data now, you have identified none --
DR. HARTOGENSIS: Correct.
DR. APLEY: -- that have been reported, death due to anaphylaxis?
DR. HARTOGENSIS: Correct.
DR. APLEY: Thank you.
DR. HARTOGENSIS: And you could attribute that to the awareness.
DR. APLEY: I had misunderstood that. Thank you.
DR. BAKER: But two of them suffered anaphylactic shock in the RiskMAP data. Two of them actually went into, you know, vascular collapse.
DR. DOYLE: Okay. Thank you, Dr. Apley. Any other questions in general?
DR. DOYLE: Are there any other clarifications we need regarding the VMAC presentations? Dr. Baker, do you have anything you want to say?
DR. BAKER: I think one thought that Dr. Hodge mentioned, they use the term “safety profile,” so they characterize safety profiles. Well, you know, the term we presented earlier was “case series,” which is, you know, it is a summary of all the clinical manifestations that identify risk factors. And, you know, hopefully, you know, you generate that information, then you can educate practitioners about -- regarding that information so that when they are in their clinic, you know, and they have -- you know, they know what is going on in their local community in terms of heartworm disease and what is happening with their patients, well, they can weigh that.
That is what we are trying to do, is allow veterinarians to have that ability to look at their patient population, look at what they are dealing with, you know, from the infectious disease part with, you know, heartworm disease, and then look at the safety profile which we term the “case series” and see how those two things fit. And that is -- you know, in their professional judgment, you know, talking with the pet owner, explaining these things, that they could make the best choice for that individual animal. So that is where we hope this information gets to.
DR. HODGE: Yes. And I just might clarify a little on just pharmacovigilance in general.
It would be nice if there was a point at which something tripped an action or a number or a statistic that could be applied to this data. All of the signal detection software relies on reporting ratios and so forth and they do generate signals.
But the standard, the gold standard, for signal detection in this field is really individual case review. And then out of that you develop a case series, and then out of that you strengthen or develop your signal. And once that is done, then that can lead to further research, risk minimization plans et cetera, things that you can do to prevent or mitigate certain reactions.
But anything above that is higher level, but drilling down to read each individual case, it doesn’t get any better than that.
So, you know -- and I have always been in your camp as well, wanting numbers, statistics, and I have tried. But the truth of the matter is there is nothing that can get you away from reading each individual case, making a decision about it, and then, based on that, taking action.
DR. BAKER: Can I say one -- maybe one follow-up with that?
You know, if you -- I hate to go back to that guidance document I talked about, that CDER/CBER guidance document on pharmaco-epidemiologic assessment, but one case can be, you know, a signal, one individual case. And that is why you have to look -- you know, especially for a rare event, that is why you need to -- you know, we use all this signal -- which we are trying to get, the signal detection software, you know, data mining software, but that only allows you to help identify these signals. But then you have to, like I said, drill down into those individual cases and make your assessment of those individually.
But one -- you know, one case can be a signal, one. So that is why you can’t go back to using any magic number, because one individual case -- and it is in the guidance doc. That document really goes into depth about talking about these issues that I talked about.
DR. DOYLE: Okay, thank you, Dr. Hodge and Dr. Baker. Dr. Apley, did you have another comment? Sorry to take the mike away from him.
DR. APLEY: Two quick questions again for our speakers.
I have heard during the day and read various opinions on whether reformulation had an effect or with taking out the residual solvents. It is tempting to draw a temporal association in there, understanding shelf life, time to be used et cetera, with the actual deaths, a decrease under the RiskMAP data as in the original data before 2004. Am I stepping out and making an unwise temporal association in thinking that the reformulation actually has changed the characteristics of this drug?
DR. HODGE: I will go for it. I think you are absolutely correct.
I think that -- basically, I think we were dealing with a multi-factorial problem and that the formulation was probably one of the major ones. Other ones, I believe, are -- and it is Fort Dodge after this event actually revamped their pharmacovigilance system. They got PV Works. They reworked their entire pharmacovigilance group and got new members on board from the human side.
So this experience for them created, basically, the -- a much, much better pharmacovigilance system because they didn’t have the data prior to the voluntary recall to defend the product in what they were seeing.
But the formulation, the pharmacovigilance processes, and then the RiskMAP, then the removal of confounding factors has led to what we have today.
DR. DOYLE: Thank you, Dr. Hodge. Any other questions?
DR. DOYLE: Any other points of clarification from the panel?