Seeking Alpha
We cover over 5K calls/quarter
Profile| Send Message| ()  

Executives

Kevin Gorman - President & CEO

Jane Sorenson - IR

Chris O'Brien - CMO

Analysts

Thomas Wei - Jefferies

Robyn Karnauskas - Deutsche Bank

Phil Nadeau - Cowen & Company

Jon LeCroy - MKM

Joshua Schimmer - Leerink Swann

Neurocrine Biosciences Inc. (NBIX) Phase II Results of VMAT2 Inhibitor NBI-98854 for Treatment of Tardive Dyskinesia Call March 27, 2012 8:00 AM ET

Operator

Good day everyone and welcome to this Neurocrine announces Phase II Results of the VMAT2 Inhibitor call. At this time all sites are on a listen-only mode (Operator Instructions). As a reminder today's call may be recorded and it's now my pleasure to introduce our first speaker, President and CEO, Mr. Kevin Gorman. Please go ahead sir.

Kevin Gorman

Thank you very much and thank everyone for joining us this morning. I'm joined here with Chris O'Brien, our Chief Medical Officer and Tim Coughlin, our CFO. Before we get started, I would like Jane Sorenson to read our Safe Harbor statement please.

Jane Sorenson

Good morning. I want to remind you of Neurocrine’s Safe Harbor precautions. Certain statements made in the course of this conference call that state the company's or management's intentions, hopes, believes, expectations or predictions of the future are forward-looking statements which are subject to risks and uncertainties. Information concerning factors that could cause the actual results to differ materially from those contained in or implied by the forward-looking statements is contained in the company's SEC filings, including but not limited to the company's annual report on Form 10-K and quarterly reports on Form 10-Q.

Copies of these filings may be obtained by visiting the Investor Relations page on the company's website at www.neurocrine.com. Any forward-looking statements are made only as of today's date and we undertake no obligation to update these forward-looking statements to reflect subsequent events or circumstances. Kevin?

Kevin Gorman

Thank you, Jane. So you have had a chance to see the press release from last night on our VMAT2 trial. The dataset was not as clean as we would have expected and we are compelled to perform an ad hoc analysis, but it's in the ad hoc analysis that the drug displayed substantial efficacy. We are going to spend the majority of this call with your questions. We'll start out with Chris giving you a brief overview. One last thing before Chris goes through the trial with you is that you may have also seen this morning that we have another press release. We are very pleased to see that Abbott has received just minor comments from the FDA on their SPA for elagolix. They have already responded to the agency and expect to receive full approval on the SPA shortly, all aspects of the starting of this trial, Abbott is going full speed ahead and we expect this trial as Abbott has directed us to start next quarter. So without any more delay, Chris as I said is going to give a brief overview of our VMAT2 trial and then we would like to take your questions. Chris?

Chris O'Brien

Thanks very much, Kevin and good morning. Thanks to those for joining us. If you are having difficulty of tracking through the webex or you need to refer to the slides that I am going to be using, they are also available on our website posted this morning. So we will walk through those. So let's start by moving to slide number three which is entitled The Tardive Dyskinesia Study, but I wanted to remind people of the intent, the purpose of the design of this early Phase II study. If you recall, when we started, the study kicked off last fall, the initiative was to see if we could get proof of concept very early in this drug discovery program.

At the time, when we had two weeks of tox data and [APIM] solution. So we designed a small cross-over trial that would allow us to detect a 3 point difference in patient score when they were either in the placebo period of the cross-over trial or the active period of the cross-over trial. So in this case, individual patients are used as their own control or so called within subject comparison and with that in mind, what we know about the AIMS, we knew that the sample size estimate would be approximately 8 subjects in each arm of the study for a total of approximately 32 people for randomization.

As you will hear, we go through the data today there were 37 subjects that actually qualify for randomization and of that, the intention to treat or ITT population was a total of 32 subjects. ITT in this case defined as subjects who had gotten drug and who had data from both variants, period one and period two. Subjects were randomized either a 12.5 mg or 50 mg and they were also randomized into the sequence of exposure. So they could have gotten placebo first and then active or vice versa.

So four groups of eight turns out that we ended up having an intent to treat population of 32 and we will walk through the numbers with them as it stands. The schematic on slide four since we showed you, use that again. There are two treatment periods and the important thing is that the primary endpoint of the study was a comparison of the scores at the end of one treatment period to the scores at the end of another treatment period. So active versus placebo. Again powered to detect a three-point difference and as you will hear patients came into the study with baseline scores in the 14 to 15 range on the AIMS with a standard deviation of just under four points.

And this is fairly typical to what we had expected based on our earlier open-label trial with that kind of baseline score and standard deviation. So if we turn to the slide entitled primary efficacy, slide number five, you can see that as we analyze this statistical output and put it through the normal kind of QA/QC review, we noted some anomalies with data. What I mean by this is when you line up the data by the 8 sites that were participants in the study, we see that one of the sites had patterns of patient response that were quite different than the other sites.

So our first reaction as part of the QA process is to say did something happen with randomization that would have disturbed the way the drug was supplied in a blinded crossover fashion. We checked carefully the randomization and in fact we looked at the pharmacokinetic data and you see in fact that randomisation was preserved and that exposure to study drug was exactly what we would have predicted based on our earlier Phase 1 studies. So that was okay. So then we looked at, if there is something about the AIMS scoring that is anomalous and as you may be aware in most Tardive Dyskinesia studies conducted over the last 30-40 years, the AIMS score is conducted during the structured neurologic exam by a clinical rater who is trained in doing the AIMS and that's exactly what was done here. In this case, we also happen to have videotaped the patients, subjects during the conduct of the AIMS exam. So we went back to multiple sites and looked at the scores on the AIMS and looked at the videos and noticed that the one site that had the anomalous pattern of AIMS scores that, there was a disconnect between the recorded score and the available information from the video. At the other sites the video in the AIMS tracked very nicely, exactly what you would expect in a clinical trial environment.

So we have this challenge for us in what to do and here for an early Phase II study, designed to tell us do we have evidence that there is a mechanistic impact of VMAT2 inhibition on the movement disorder’s interest, we had this inconsistent and inaccurately applied AIMS at one site. We simply conducted a post-hoc analysis with that site removed from the analysis.

So we used the pre-specified statistical analysis plan, applied it to the data from seven sites instead of eight and saw this striking evidence of efficacy at the 50 milligram dose and exactly as one would predict with VMAT2 inhibition, based on what's known about this mechanism and hyperkinetic movement disorders.

You will see on the slide at the bottom that the difference for this group of patients on 50 milligrams between their placebo period and their active period was 4.2 points. On the next slide, slide 6, this is displayed in a tabular fashion on the left is the LS means, the difference between the LS means and the P value for all sites that are on the right is the post-hoc analysis with the excluded site removed.

You can see the striking impact of this change and this is very much in line with what we had anticipated when we did our sample size estimates and calculations in the beginning. You may also recall from one of our earlier conference call that in this early Phase II trial we chose two doses that were rather far apart, based on our open label and Phase I experience.

The 50 milligrams dose was chosen because we had seen a nice effect in the open label trial with 50 milligrams and that was well into the exposure range that we thought would give us the beneficial effects we needed, at the same time being recently well tolerated.

The 12.5 milligrams was chosen because we thought this was right on the threshold of what might be the minimally effective dose and in a sense you can use this as a comparison opportunity in a crossover trial and indeed that’s what was borne out. The 12.5 mg dose is very similar in terms of what we are seeing to the placebo dose. When you look at individual subjects you will see that there are some subjects that appeared to have responded nicely to a 12.5 milligrams dose individually but as a group the mean change is comparable to placebo.

So that helps us tremendously as we go into the next trial. We know how to take advantage of this information and use it for dose selection going forward.

So let’s go to the next slide and put this information in context. So we had just looked the difference between the scores when the patient was in the placebo periods versus when they were in the active period.

But you remember that all of these patients who obviously came into this study and we had some baseline scores to look at and I think for a clinicians and for patients and for those of you who are familiar with the Tardive Dyskinesia literature, its often helpful to look at what does that change from baseline represent and what kind of effect was there from placebo or simply having participated in this trial.

And again, I would remind you that the baseline scores were in the 14 to 15 range and so what you see is that when the group is on placebo or low-dose active medication, the change is a reduction of four or five points. So about a third improvement and very consistent with what’s in the literature for placebo effect, about a 30% improvement in AIMS.

On the other hand, the 50 milligram group was associated with a nine to 10 point improvement in AIMS. Here shown as a mean of minus 9.2, change from baseline. And obviously this is highly, statistically significant and certainly in keeping with the comments we were hearing from investigators, patients and care-givers who participated in this study, and told us that they were remarkably improved during two weeks of the four-week trial and we heard this repeatedly from multiple sites.

So this very consistent with the robust treatment effect, both clinically, meaningful and statistically, significantly better.

Now, you may recall that in addition to the AIMS, we had a number of other scales that we put in this early Phase II trial., We wanted to look at a clinician global impression of change around Tardive Dyskinesia, this CGI-TD, the patient global impression of change, as well as the other safety scales, which we will get to in a minute.

Here what you can see our two tables on this slide number 8. At the top, its the table with all the sites and at the bottom is the table with the excluded sites removed and you can see that the post-hoc analysis shows what appears to be a fairly consistent dose response pattern, namely a 40% or so improvement in the placebo group with improvement being defined on the 7-point Likert scale is very much improved or much improved and this goes up to 65% or so with the 12.5 milligram group and 80% with the 50 milligram group.

So this will be consistent with what I see treating patients in the clinic with the VMAT2 inhibition. About two-thirds of patients have a robust response and are satisfied with the response to therapy and this is borne out here in the post-hoc analysis.

Let’s go on the next slide about safety and tolerability. Again, I mentioned 37 subjects were randomized, the ITT population was 32, that is, we had at least some contributing data from both periods and everybody had at least one active dose of drug and 31 of those subjects actually completed both periods. So you see the safety data here is for all subjects who had participated in the trial.

The overall reporting of adverse events reflected that the drug was reasonably well tolerated. In fact 32% is actually quite a bit lower than I would have anticipated in the trial with patients who have schizophrenia, who are taking multiple medications for their psychiatric problems. They were all on a variety of anti-psychotic drugs including typical such as haloperidol, but most commonly atypical, the most common one was Risperdal, but also Seroquel and Abilify and several others.

Now it’s interesting, I mentioned the surprise at the rather low rate of AE reporting; these study patients were coming into the site every single day for the duration of the study. So for one month they came in everyday because they had to have their study medication provided. It was drug Active Pharmaceutical Ingredient, API powdered mixed in with Diet 7UP. This study was started before the solid dose formulation was available.

And so, what you see is patients coming in everyday very intense study, multiple hours in the clinic each visit and normally in that setting you see a high frequency of AE reporting because there are many opportunities to talk to the investigator and study coordinator. And so, somewhat of a surprise that the number is as you see here, you can see that the placebo period was associated with AE frequency of 17% reasonably comparable in the 12.5 milligram and although the same number of subjects reported AE’s in the 12.5 milligram and the 50 milligram period. One of the subjects reported more AE’s so the overall frequency of various AE’s is highest in the 50 milligram group at 32%.

We did have one of the 50 milligram subjects discontinue study medication and participation in the trial because of an adverse event of motor restlessness, so called akathisia. This occurred early in period one. He was randomized to 50 milligrams, discontinued on day seven. That was judged as probably or possibly a drug related adverse event. There were no treatment emergent or treatment related serious adverse events. The one SAE we did have occurred after the treatment periods in the patients with the history of emphysema and she had an exacerbation of her COPD.

I mentioned that there were a variety of safety scales; the Barnes Akathisia Scale, the Simpson Angus Scale for Parkinsonism, the Calgary Depression Scale for Schizophrenia patients; the Columbia Suicidality Scale, all of these four scales had low baseline stores at entry and remained either stable or improved during the course of the study. So during two weeks of active dosing in this trial there was no signal of any emergent depression, suicidality, Parkinsonism etcetera.

The Brief Psychiatric Rating Scale or BPRS was where you would expect for study of stable patients with schizophrenia basically in the low 30s and it more or less remained unchanged over the course of the study. The mean went from 32 to 28 which is a slight improvement, but I would interpret that as being no change in the context of this trial, so very nice safety and tolerability.

Next slide simply summarizes the pharmacokinetics. The drug performed as we had expected based on our earlier Phase I data. There were no obvious drug-drug interaction problems and no significant or clear differences in exposure based on gender, smoking, BMI, etcetera. We have some [sub-DNA] typing data which we are looking at and this will help us make plans for our second Phase II study, our next Phase II study as we go forward.

So let’s go to the conclusion slide before we get to time for some question and answers. We are very pleased. We would consider this a successful study in that. Our primary objective for starting this trial was, do we have proved that this drug reduces Tardive Dyskinesia and is safe and well tolerated?

And the answer is, yes. We have seen a significant reduction in the range that we would expect for an effective dose of the VMAT2 Inhibitor in patients with moderately severe Tardive Dyskinesia. We obviously were troubled by the inconsistent and inaccurate application of the AIMS by one of the sites and had the post-hoc analysis which informs us so we can move on to the next Phase II trial.

The magnitude of effect, the responder rate, the consistency of the patient responses and the safety profile are all exactly what we are looking for in terms of moving forward and it is our intent to conduct the next Phase IIb trial as planned starting in Q3. This will be a 12-week trial. It will be a large study. We will include a certainly the 50 milligram dose, most likely the 25 milligram dose, but we are still in the midst of PK/PD analysis and we will use that inform the final bill selection by the time we get to study start later in the summer.

So at that point, one of the thing I guess, I wanted to mention that early Phase II trials serve a very important purpose. They tell you what you need to pay attention to and what were the kind of things and issues you might run into in subsequent trials. If this trial had been a larger study, we probably would have never have detected the inconsistencies at this one site; it just would have been lost in the mix of the study and would have been a clear positive result with 50 or 60 subjects and we would have carried on regardless. But since it was a small study and having one site have this inconsistency, we were able to see this and we can use this information to benefit future trials.

So will continue to do what has always been done when using the AIMS and what was done in our trial; all the sites went through a training and certification process for the investigators, all the sites were monitored closely, but what is usually not done is real-time monitor, the video comparing it to the AIMS score. Normally, the investigator does the score and that’s the end of the story. In our case, we fortunately have these videos and I think we’ll build this in future trials where we can actually monitor as the study is going on and make sure things are tracking together obviously in a blinded fashion.

So at that point I would like to summarize.

Kevin Gorman

Yeah at this point I think we are ready for your questions and we are very happy to take them now.

Question-and-Answer Session

Operator

(Operator Instructions). And we will take our next question from the site of Ian Somaiya with Piper Jaffray. Please go ahead, your line is open.

Unidentified Analyst

This is actually [Doug Cameron] for Ian. I was hoping that you could perhaps provide more details on how the video reviews were conducted and what you saw on them and how that compared to the AIMS scoring?

Chris O'Brien

Sure we will be happy to do that. So as I mentioned the clinician who is doing the AIMS scoring does that as part of a structured neurologic examination and they have the seven body regions and they rate each body region for movement intensity on a 0-4 score. And while that is going on, a videotape of the patient is running. So there is about a 10 minute videotape that is part of the source document. So what we did, when we noticed this data anomaly, we went back to multiple sites and reviewed the videos because we have videos for every AIMS exam for every subject and compared the scores to what is available on the video.

And normally the clinical exam and the video don't match exactly. The video of a patient tends not to have quite the dynamic range that the clinician gets during the exam. The clinicians are better at detecting the subtleties and movement and you can hear for example the respiratory dyskinesia that may not be evident on the video. But what we did was we went out to the sites and reviewed the videos of the patients at the various time points and just to make sure that they were tracking with what the scores show. And in this case, one of the sites, they didn’t track together. For example you might see a video where a patient has minimal dyskinetic movement and yet has a moderately high score. In the next video they look about the same and the score has come down. Whereas at the other site, the videos and the scores tracked fairly nicely together, exactly what you would expect in a clinical study.

Operator

We'll take the next question from the site of Thomas Wei with Jefferies. Please go ahead, your line is open.

Thomas Wei - Jefferies

Thanks, I had a couple of questions, still just trying to reconcile a couple of the data points here. I think you have done a very nice job of explaining what had happened, but I am still a little bit confused about given the problems and AIMS scoring at that one site, when you look at the data, it seems to be that the 50 mg dose was the data that was very markedly affected, but the data for the placebo arm doesn't actually change from the ITT to the post hoc analysis and that would seem a little bit unusual given the cross-over design and the fact that the same patients would have gotten both placebo and 50 mg, yet the erratic scoring only really seem to affect their scores when they were actually given drug. Can you help us understand that a little bit better?

Chris O'Brien

Sure, Thomas, so what happened with those scores is that there was an inconsistency in how the AIMS was applied. So the site when they were in the active period, that was when the biggest discrepancy was in place and so you have to know that there were five subjects on 50 mg at the site by randomisation and they happen to have the biggest discrepancy during the 50 mg period. That discrepancy is not as evident in the placebo period and that again reflects I think the inconsistency of how the exam was done.

Let me give you an example. That is a, it would be subject who happened to have a high score at well on active drug and a high score while on placebo. When you look at the video, if you were to use that as an approximation and score in the video, you end up seeing a very different pattern of response, you might see a high score on placebo and a low score on active and that difference obviously doesn't affect the placebo mean, but it affects the active mean.

Thomas Wei - Jefferies

That's helpful and then the second thing in the data that I am still a little bit confused about, have you had a chance to look a little bit more closely at the PGIC, so investigator you know it seems as that the investigator had a problem with scoring AIMS, but then the PGIC is a patient reported number and that also seems to have been affected at that one center or at least the data too seemed to change for 50 milligram for those patients who are excluded. Have you had a chance to look at that yet?

Chris O'Brien

Yeah, two points on that [Thomas], so on the one hand if you look at the PGIC data the patients clearly are saying we are feeling better. I mean patients generally were happy being in this trial and overall they were talking that at least half the patients said much improved or very much improved. What you don't see was this cut of the 87-point Likert scales are about the minimally improved and no change than minimally worse, where do those fit in.

The fact is most of the patients clustered in the scores of no change, minimally improved or much improved and so all it takes is one subject removed and there you see the big difference in these scales. I don't think that we can rely on the data from this site given the nature of the AIMS problem and I haven't spent too much time coming to micro-analyze the PGIC data. I see it as not terribly helpful in the context of this crossover trial. It’s just and particular since we know that one of the sites had this problem with the data.

Thomas Wei - Jefferies

That's fair and then just lastly you had mentioned maybe a stricter, real-time monitoring of videos and any AIMS scoring, while the study is underway but I guess that's still you do you have an errant investigator in the scoring that still make rates and data is using your larger studies, so you would end up having to [correct] some data or accept poor data and so it’s caught. Is there anything that you can do more in advance of that happening to ensure that your investigators are scoring the way the other investigators are? Is there some way to have the data ultimately centrally reviewed, any thoughts there?

Kevin Gorman

Yes. Thank you. You bring up three very good points. So the first is what do we do in advance as far as training? The standard for clinical trials using AIMS has always been to have training and certification of the raters before starting the trial and that was done here. The investigators, all of these investigators have been involved in clinical trials with AIMS before. They all went through a process of scoring a standard video of patient with Tardive Dyskinesia and had to pass the certification test of what their scores were getting in. So that’s been the standard for years. That’s what was done here.

Could that be done with some enhancements? Absolutely. We’re beginning that process of improving that for the next trial, but I think the key was the second point that you made was. Normally in clinical trials, you have variability on AIMS and if your trial is big enough, even you have, like the phrase used was, an errant investigator. If you have that, that’s okay because the trial was big enough that any deviation by one side or one investigator, it is not enough to throw off the efficacy [safe zone].

In fact in this trial, as I said, if we had a 50 or 60 subject trial, we would have probably not even seen this. It would have been so strikingly positive that this would have been part of the background noise. But having said that, obviously we do want to try to assure the quality of the data coming in on the primary endpoint and one of the ways we plan to do that is have our trained monitors, while the study is going on, looking at the video and comparing it to the actual AIMS score that the clinician puts down and seeing if they’re tracking together.

Obviously, there are differences between video and real clinical examination and to the extent that people have tried to control for that, one of the things that has been done, which we are considering is that you have an independent, rating physician at the site that is separate from the treating physician. So they just come in and do the AIMS assessment, then they leave, whereas the other investigators talking to the patient about the feeling and what the side effects are etcetera. So that’s one avenue.

The third possibility is one you mentioned and that is having a blinded, central rater score video tapes and this has been done in movement disorder trials over the years. I think there is a rule for that. The challenge as I mentioned is, sometimes the videos don’t capture what's actually going on with the patient at the time and historically, when people have done video rating, they find that the dynamic range of the instrument is compressed and may not be as useful a tool for the clinical trials.

So there have been attempts to look at other objective measures of Tardive Dyskinesia, for example, lingual force recording devices or computer based video algorithms for accessing facial muscle movement; none of these things have panned out very well and so currently the FDA is expecting the AIMS will be the goal standard and the primary end point going forward with ratings done by the clinician.

Operator

Our next question comes from the side of Robyn Karnauskas with Deutsche Bank. Please go ahead, your line is open

Robyn Karnauskas - Deutsche Bank

I guess my first question is, you noted that you are using a different formulation actually the capsule formulation versus the powder in Phase IIb; and can you think about any differences or how that might affect on the outcome and maybe PK when you go forward in the next Phase IIb study?

Chris O'Brien

Sure. Thanks Robyn. So the capsule formulation that we have as a solid dose for the next study has been through already Phase I relative bio-availability study, so we have compared capsule to the powder and solution formulation. So we have a good understanding of the PK profile of the capsule and it’s exactly what we would want it is unlikely to produce any differences in exposure compared to API and solution. That’s from a strictly PK point of view.

But I do think as you bring up an interesting point, when you have patients that come in every day to see the powder end solution and that daily contact with the coordinator, the pharmacist, the investigator that tends to engender a feeling of trust and comfort on the patient’s level when they come back to clinic day after day after day for a month and for movement disorders, in particular hyperkinetic movement disorders that are susceptible to the effects of anxiety and stress; establishing trust and becoming comfortable at the site tend to be associated with the decrease in the intensity of the involuntary movements. And that maybe that this daily dosing on the daily visit account for some degree of the placebo effect that we saw in this trial. So it’s our expectation in the Phase IIb that with a capsule and less frequent clinic visits that maybe less of a factor.

Robyn Karnauskas - Deutsche Bank

And to follow-up on that, how did the placebo respond and differ from previous studies you've seen in this indication?

Chris O'Brien

That's a good question. There are two things to keep in mind. If you look at most of the product Tardive Dyskinesia trials, most of them are multi-center parallel group trials with placebo and active drugs. In general, the placebo effect is in the 20% to 40% range so let’s just say 30% which turns out to be similar to what we saw.

Now in contrast, there are a few well controlled crossover trials that were done and if you look at those carefully you will see that the ones that have a very low placebo response rate are single site, single rated trial, so you have very little variability there and the placebo effects are quite small. Obviously, in our study we were not a single site, single rates trial and our placebo effect was much more consistent with what's seen in the literature with multi-site trial.

Robyn Karnauskas - Deutsche Bank

And when patients were split back off those drug; did their symptoms come back a 100% and if so how long did it take?

Chris O'Brien

So the last day of study drug dosing was day 28, with patients happened to be on active drug during that period and we saw them again on day 35. Likewise for patients who randomized the study drug in the first period, their last day of dosing would have been day 14 and we saw them again on day 22 and day 29 and day 35.

So we had a lot of opportunity to see patients after they stop drug and if once you remove the one site that had the inconsistent use of AIMS, we see a return Dyskinesia over about a two week period in most of the patients.

There is some variability, not everyone returned to 100% and that’s something we need to understand is that a pharmacodynamic effect, is that a placebo effect basically by being comfortable and in a trial do you get a, you know you get some measures and long-term improvement as long as you are still seeing investigator on a regular basis; I am not sure yet, but obviously we will use that information when we do our power calculation and sample size estimation for Phase IIb.

Robyn Karnauskas - Deutsche Bank

And my last question sort of is taken a little bit from (inaudible) question. So if you look at the PGIC versus the AIMS on a patient-by-patient basis, you know you noted that some people did not feel better and maybe I think you said 25% which is sort of in line with what you have seen before, people didn't really respond that well from the patient reported outcomes, but when you look at these two different scores, do they correlate on a patient-by-patient basis meaning same patients that improved on the AIMS are the patients that really improved on the PGIC?

Chris O'Brien

So the short answer is the CGI, the Clinician Global Impression tracks quite well with the AIMS. THE PGIC, the patients generally all felt better and so for the patients that improved a lot on their AIMS, yes their PGIC improved a lot. Unfortunately or not because they felt better, the PGIC showed patients felt better even if their AIMS score had gone back towards baseline and so they simply were comfortable being in the study and generally felt better at the end of the trial. So it's not a great correlation on the PGIC.

Operator

Our next question comes from the site of Phil Nadeau with Cowen & Company. Please go ahead, your line is open.

Phil Nadeau - Cowen & Company

First on the post hoc analysis, I don't think you have told us exactly how many patients were excluded in that analysis. What was the number and what was the split between 12.5 and 50 mg patients?

Chris O'Brien

Sure thanks Phil. So the ITT population for the 8 sites was 32 subjects and with a one site excluded it was 25 subjects. So seven subjects were excluded and five of them were on 50 mg and two were on 12.5 mg.

Phil Nadeau - Cowen & Company

And in the 25 subjects that were remaining after the post hoc or after the exclusion, was the effect size consistent in those 25 subjects, so I guess one risk of removing seven subjects is, obviously in the end get smaller and it may be possible to drive a good result using a smaller subset of patients. So would the effect of, again was it consistent in all 25 and would the affect still have been there, had you excluded kind of the seven best responding subjects as well?

Chris O'Brien

Yes that's an interesting question. You know we are down to relatively small numbers in the 50 mg group, you are talking about ten subjects. And the consistency of response in the 50 mg subjects was very consistent. There wasn't a lot of variability there. So you know for a first small subject size, it was pretty tight.

Phil Nadeau - Cowen & Company

And what happened to so this 32 patients in the ITT, 37 I guess were randomized, so five didn’t get all the analysis, what happened to those five?

Chris O'Brien

Sure. Yeah so one subject as I mentioned discontinued because of a side effect in period one, of the (inaudible). One subject was discontinued because of narcotic drug seeking behaviors, that was in violation of the protocol. And three subjects withdrew consent meaning that they didn't want to keep coming back to the clinic everyday. They didn't have, it wasn't because of a side effect or because of something, one of the other categories for discontinuation. It was simply withdrew consent.

Phil Nadeau - Cowen & Company

And then on the side-effect side, it does seem like there was higher rate of adverse events for drug and placebo, although it’s unclear to me exactly what those events were, what were the treatment emerging?

Chris O'Brien

The main one that’s different from placebo was headache and that was tend to be mild and obviously not associated with discontinuation but truly it was being driven, the actual number of subjects reporting adverse events in the 12.5 milligram period and the 50 milligram period was the same. It’s just that one subject reported multiple different side-effects and hence the percentage goes up.

Phil Nadeau - Cowen & Company

Okay. And then one last question kind of a follow up to one that Thomas asked and that’s on the clinical group impression. That also should difference between the post-hoc and the ITT analysis, is the clinical group impression, is that part of the AIMS score or is that a entirely separate measure and if it is a separate measure, did you have any way to verify what this one physician was doing on that measure?

Chris O'Brien

So, the CGI is a separate scale. That’s a 7-point Likert scale, four being no change, seven being very much worse, one being very much better and it’s an impression by the clinician of how the subject is doing compared to how they were before starting treatment.

The, it is the instructions for that CGI is it’s to be based on TD, not their schizophrenia status or their general health or anything like that. So it should track fairly closely with the AIMS as general because it is measuring TD.

Phil Nadeau - Cowen & Company

Did you expect a correlation between CGI and AIMS at the site or was that also somewhat off?

Chris O'Brien

It’s a little off.

Operator

Next, we will go to the site of Jon LeCroy with MKM. Please go ahead. Your line is open.

Jon LeCroy - MKM

Thanks for taking my question. Most of my questions have been answered. I just wanted to make sure all the eight sites were in the US. Is that correct?

Kevin Gorman

Yes John they are.

Operator

And our last question comes from the site of Joshua Schimmer with Leerink. Please go ahead your line is open.

Joshua Schimmer - Leerink Swann

Good morning thanks for taking my questions. Just wondering when that final analysis of that one site is expected to complete and do you think you will have additional updates?

Kevin Gorman

The review is an ongoing process, Josh. I am not sure what kind of updates you were thinking about. I mean for our purposes the key for the study was, did we have sufficient evidence that we have a drug that works, so we can move on to the larger Phase IIb study and while we will continue to analyze this data and try to learn from it, to minimize problems going forward. It is not I am expecting some kind of a wholesale announcement of new discoveries or anything. So may be you can tell me what you are thinking about?

Joshua Schimmer - Leerink Swann

I am just kind of wondering if that, it sounds like may be you don’t expect to find a root cause as to what went wrong in the center but if you did whether you might announce that?

Kevin Gorman

I guess we may never really understand what happened. Obviously if we have some believing insight we’d probably share that but I am not sure that any major announcement would come out of this.

Chris O'Brien

And also I would like to add Josh that while we continue to investigate this, clearly the primary reason why we are investigating it is, so it helps inform for the next study to learn as much as we can about this. We do respect and will continue to respect the confidentiality that is involved here with our investigators.

Joshua Schimmer - Leerink Swann

Were there any patients in whom or who responded better to placebo versus drug either at that one site or at other sites?

Chris O'Brien

Well, at that one site, yeah and as I said the scores were very inconsistent, so if you could call an AIMS score that was lower on placebo and higher on Active or response, yeah, but when you look at the video and you don’t see the dyskinetic movement when they supposedly had high score it’s hard to interpret that as a response.

Joshua Schimmer - Leerink Swann

And was that something unique to that site with any other patients who…?

Chris O'Brien

Yeah, the other as I mentioned when we look at the videos from the other sites and compare them to the AIMS, well, the scores are not identical. They at least track together; when the patient is worse on the AIMS they are worse on the video and vice-versa.

Joshua Schimmer - Leerink Swann

How is the dose of the antipsychotic or concomitant antipsychotic control during the trial?

Chris O'Brien

The requirement for this protocol was that patients were on stable doses of their concomitant medications for at least 30 days before randomization and not allowed to change during the trial. We had some rules that if a patient had a drug added that might have been a confounding anti-dyskinesia medication that would have been deemed an important protocol deviation and they would have been excluded before that database locked. Likewise if their antipsychotic dose changed, they would have been excluded from the ITT population. So nobody’s antipsychotic dose changed during this trial.

Operator

And we have no further questions at this time.

Kevin Gorman

So, I would like to thank everyone for participating this morning and all the good questions we had. We are here and we are available to take your questions throughout the rest of the day and the week. We are going to be in the Needham Conference in New York next week and we’ll be presenting this data. Following that, I think as our next meeting is in the beginning of May at the Deutsche Bank Conference in Boston.

So we have a number of opportunities where we can continue to get together and continue to take your questions on this and be able to roll-out our thoughts and our plans forward.

I would just like to close by acknowledging that while it’s distasteful for every to everyone including us that have to perform a post-hoc analysis on a trial. This is a small exploratory Phase II trial. We feel it was entirely justified in this case and I hope from our explanation you are gaining that understanding also.

We strongly believe that the drug has clearly demonstrated its proof-of-concept and has reaffirmed our confidence in this mechanism and to the drug and we are going to be moving forward as we previously described in the large Phase II program. It’s informed us on dose selection and it’s given us a number of very valuable information points.

So with that, I will close and again I thank you for your attention and your thoughtful questions. Take care.

Operator

This does conclude your teleconference. Thank you for your participation. You may now disconnect.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!

Source: Neurocrine Biosciences' CEO Reviews Phase II Results of VMAT2 Inhibitor NBI-98854 for Treatment of Tardive Dyskinesia (Transcript)
This Transcript
All Transcripts