Oncolytics Biotech Inc. (ONCY)

[onco_investor]

Feb16 conf call transcript H&N Phase III UK Approval

Thanks again to RJC for the great transcribing work

Topics addressed in the Q&A session include:

- Adaptive Design, and Related Penalties:
- How Adaptive Design got into the SPA:
- Enrollment Rate:
- Start of Enrollment:
- US/UK requirements:
- UK start-up:
- Site Distribution:
- Belgium:
- Patient Distribution:
- Total number of Sites:
- NSCLC:
- Trial Cost:
- Partnering:
- Uniform Enrollment:
- Next Presentations:
- And in summary …

Full transcript of the Q&A session from Oncolytics' Feb 16, 2010 UK Phase III H&N Approval conference call. The audio is currently available at: http://www.newswire.ca/en/webcast/viewEvent.cgi?eventID=2967140

Q&A session, following a brief Phase III Clinical Program Update:

Adaptive Design, and Related Penalties:

{6:13} Miguel Nakamovitz(spelling?) [Rodman and Renshaw]: Could you discus a little bit more the details of this Trial in terms of the adaptive process … why this design was chosen, and perhaps if you could talk a bit about if there are any statistical penalties for examining the data after a certain number of patients have been enrolled?

Brad Thompson [Oncolytics Biotech]: The adaptive clinical designs have a very specific use and utility, and that is in indications where there is a fairly firm linkage between absolute response (like partial responses, stable disease, complete responses, time to progression), and overall survival. In indications where you don't have that tight linkage, I think you would not use an adaptive design. What an adaptive design does, is allows you to use that linkage (and we're relying primarily on time to progression linkage with overall survival), to predict from early data the statistical likelihood that you're going to succeed in the study, or have already reached an end point. So what it does, is it allows us to (in this case) enroll in a Phase IIIa / Phase IIIb structure. We'll enroll the first 80 patients, which will be split evenly between the control group (which is Carbo/Tax of 40 patients), and the test group which is Reolysin + Carbo/Tax at 40 patients . And then on the day that you enroll the 80th patient, you take a look at the time to progression data that you have at that time, for progression free survival, and that would be for all of them obviously on the 80th patient –you don't have any data, on patient 75, you have very short term data. If you actually do a statistical analysis about the probability of success, based on the fact that you have better time to progression data than you do with raw survival data at that point.

That gives you 3 outcomes at that point. One is that you stop the study because you actually believe you're at the end point … like you will reach statistical significance. The second is that you stop the study because there's absolutely no evidence that there's anything going on, and the third being that there is evidence of something going on but that you haven't reached statistical significance at 80 patients. And that's what our estimation is of the most likely outcome … that third one.

The adaptive design allows you to do that kind of statistical analysis which allows you to immediately verify, and proceed into the second stage of the study. The advantage of the second stage of the study is that it gives you a fixed end point; not being a fixed survival benefit, but a fixed statistical probability, so you're aiming for a p of .05 rather than a lifespan extension of say, of 16 weeks. So the number of patients that you enroll can vary to get to that p, so it ranges from 100 to 400 patients in the second stage of the study with the most statistically likely being 200. What that allows you to do is not basically exceed your end point when you've already reached the end point, and keep enrolling patients in a clinical study where they have a 50% chance of not getting the new triplet [Reo/Carbo/Tax], and so, say at 200 patients, you have statistical significance, then you're done, and you can stop the study then, as appose to carrying on and enroll another 200 patients, half of which won't get this hopefully better standard of care.

So I think it's a much more patient friendly study. It gets you to end points (good or bad), quicker. With respect to statistical penalties, of course in the first 80 patients there is no penalty, because you're looking at it after the 80 patients, and what we're planning on doing is: we'll treat the Phase IIIa data independently, the Phase IIIb data independently, and then there's the data merge, where you merge the data from both groups. We have actually three different looks at the data from the merger. There's an infinitesimal statistical penalty in the second stage, because what you're looking at is time to progression data, not the overall survival data as the end points. Actually your statistical penalty is not your primary end point; it's on your secondary end point. So your primary end point, which is overall survival, actually doesn't have really any penalty at all associated with it. So it's quite a powerful statistical tool, but the key thing is selecting where you use it, and I think that's where people have to be very cautious about the use of adaptive design. If you have an alternate earlier read … some indicator, some measure that is very tightly linked to the real outcome, that in this case being time to progression being linked to overall survival apparently, then you can use this kind of adaptive design. If you don't have that tight linkage, say with some of the other types of cancers, like say non-small lung which has a long tail of survivors already, on not as much of a tight link as between the time to progression and death, then I think that you should probably stick with a traditional statistical package.

How Adaptive Design got into the SPA:

{11:15} Miguel Nakamovitz: That was a good answer. Is this something that you proposed to the Agency, or did they suggest it, or was it more of a collaboration to come up with this design.

Brad Thompson: Well they've been actually very interested in adaptive design for quite some time and the Agency has been … candidly … the FDA gets criticized for not wanting to get things through quicker, but they're really just interested in us sticking to "the gold standard", which is overall survival, and that's really what we're looking at, but they're also being very aggressive about looking for ways to move studies along quicker with just different statistical analysis, so they've been very receptive to adaptive statistics. They don't propose anything. I mean, they're always reactive. So we knew that they would be probably very responsive to this kind of design. They have approved a couple of other cases for infectious disease areas and that. So the scientist in me thinks of this is highly elegant, I mean I'm not a MD, I'm a scientist. I've always been interested in using adaptive design in an appropriate occasion, and this was a very good application of it. Don Berry, who is actually the statistician who is kind of the grandfather of adaptive design, is actually the statistician who worked with us on this, and we co-presented it to the Agency. It was kind of a weird kind of collaboration, but the FDA really is seriously interested in looking at ways of accelerating clinical development, and this is a specific tool that I think they're quite happy that people are starting to look at.

Enrollment Rate:

{12:37} Miguel Nakamovitz: Thanks. And then just one final question in terms of the rates of enrollment: I mean, for the first Phase with the fixed number of patients, obviously it's the second Phase that will depend on how the thing adapts, but could you just give some remarks there about how long you think the study will take to enroll?

Brad Thompson: We're hoping to enroll 10 patients a month for the first stage, so the primary enrollment should take 8 months … assuming we get the number of sites we want on, and everything like that, and it looks like we should be able to do that. So, 8 months for the 80 patients.

The second stage, we're aiming to double that enrollment on a monthly basis, and we'll be bringing on additional sites while we're doing the first stage, in anticipation of the second stage ongoing. So we'll have that 8 month period to double up on the site number that will be involved with the second stage of the study. Assuming it's the 200ish patients at 20 patients per month, you're looking at a 10 month primary enrollment for the second stage (for the most statistically likely). It could be as short as 5 months if it's 100 patients, and 20 months if it's 400 patients, but 200 in the second stage is the most likely enrollment.

Start of Enrollment:

{13:40} Miguel Nakamovitz: Got it. And just one final question … Is it your anticipation that we could see the first patient enrolled in this study by the end of the first quarter?

Brad Thompson: With a bit of finger crossing luck, and stuff like that, that is actually possible, though we've resisted giving a forecast of a start date until we're absolutely certain about when it's going to happen. Of course, this UK Approval was a key element of that. We've been going through the process of the US sites since last fall when we got the SPA, so we're very close to having sites enrollable right now, in the States. We want to just have more sites kind of in parallel going on to it, so I think that it's possible we could still do it next month, like Q1, or it might fly into early Q2. So that's why we've been kind of saying H1, 2010 … because whether it's March 31, or April 15, or something, I mean, it's likely to be not far out of that band.

US/UK requirements:

{14:50} Neil Maruoka [Canaccord-Adams]: Based on your preliminary discussions with the European regulators, have they indicated any requirements for approval of Reolysin that differs from that outlined by the FDA?

Brad Thompson: Well, it's identical. We submitted the identical protocol … literally identical, and we had one comment come back just requesting we confirm about the water we used to make the product being made to European standards. In Europe, they distill their water, and in the States they make it by reverse osmosis. Of course, all the manufacturers do both, just so they can sell it to both. That was the only comment we had.

UK start-up:

{15:22} Neil Maruoka: How long a delay do you expect in getting the UK sites on line, compared to the US?

Brad Thompson: Well, the US is a little more linear, so we needed the SPA first before anybody would touch it. The US is, in many ways, the slower jurisdiction to get anything up and running. In the UK we were actually able to parallel track, and we also had the advantage in the UK that many of the sites in the UK had already run patients, either on our Phase II head and neck, or one of the other multiple studies we'd run in the UK. So it's a lot faster process for us to get sites up and running. We're already part way through the process with many of the sites in the UK, so I would expect that a lot of those UK sites will come on near-simultaneously, or very soon thereafter with the US sites.

Site Distribution:

{16:25} Alan Ridgeway [RBC Capital Markets]: What's the breakdown on the number of sites in the US, the UK, and Belgium?

Brad Thompson: I don't have the exact numbers … it's a little bit in flux, but you're looking at around 4 to 6 sites in Belgium, as the anticipation. The Approval process in Belgium is completely different than anywhere else on the planet, so I don't even want to get into that, it's a very much backwards process. The Belgian regulatory authority is almost the last in line, as opposed to other places where they're first in line. Sort of 8 to 10 sites in each of the two other jurisdictions is the current kind of split. That could shift now a little bit. We're getting a lot of secondary interest popping up now, in both the UK and in the US. I think we'll probably put on as many sites as we can in the first part of the study to make sure that we meet or exceed enrollment.

Belgium:

{17:20} Alan Ridgeway: So, on the Belgian front then, will you be starting the Trial before essentially getting approval? Like would you start it in the UK and in the US, and just go ahead and then bring Belgium on when they're ready, or would you wait for them?

Brad Thompson: I think it's more the minimum number of sites … when we get up into the teens with the number of sites that are enrollable, then we'll go ahead and start enrolling, because the other sites will come on very quickly thereafter. Belgium's kind of odd. For all I know, we might actually enroll in Belgium first. It's a very very different process there, and of course with two languages for your filings, it makes it more interesting. The insurance is totally different. I've learned more about insurance in the last two months in Belgium than I thought I knew totally. It's just a different jurisdiction to operate in. We could start actually start enrolling in Belgium, with the way things are going, before the other two countries if we wanted to. It's a very dynamic process right now.

Patient Distribution:
{18:06} Alan Ridgeway: And as far as a breakdown of geographies for the first 80 patients … it doesn't matter at all? … you just enroll 80 where you can get them?

Brad Thompson: Yup, that's exactly correct. The only thing you have to watch out for when you're running multi-jurisdictionals is that you file in the US first. The FDA is … "rather reluctant" would be an understatement to Approve studies, and to join already existing studies, whereas the other jurisdictions aren't, so that's why we filed in the US first. Once we got the SPA in the States … that was kind of locked in stone, and we went ahead and filed in the other jurisdictions.

Total number of Sites:

{18:37} Alan Ridgeway: And now when you talk about doubling the number of sites, you're talking about potentially going up to 50?

Brad Thompson: Yup. We're trying to be very cautious on the number of patients per site enrolled. If you had 25 sites, and break it into 80 patients … If you're only counting on a site enrolling a patient every two to two and a half months, then you'd comfortably meet your enrollment for that time period. Sites always say they can enroll faster than they do, in our experience, and we're actually using the enrollment rate that we're actually getting out of our existing studies, then downgrading the time on that and using that as our calculation to come up with the number of sites

NSCLC:


{19:15} Alan Ridgeway: On the non-small cell lung cancer in the US … the Phase II, with the 3 new sites could you potentially give us an update on when you might think that that would be enrolled, and when we might see some data from that study?

Brad Thompson: I think that the bulk of the enrollment should be done next quarter, and certainly enough that we'll have enough data to be making some decisions about what to do with that moving forward. So I think we could get a preliminary read on the data as early as the end of next quarter, because the primary response rate is the primary outcome of that study … so you get a very quick read on the patients … whether they respond or not.

It's an interesting study. You're pre-screening, which is something we've never been able to do before. When we first started studies in Canada, very early on, we were specifically prohibited by Health Canada from pre-screening, and the screening technologies (for EGFR in particular), were rather not usually held in those days, and nobody knew the importance of KRas to start off, and now, we can actually pre-screen rather efficiently for the first time, there's an actual linkage with important indications that people didn't know about, and I'm just thrilled that we actually have a study where we can actually pre-screen patients, and see that pre-screened patients, Carbo/Tax which does the immune system "knockdown" thing that is helpful, that makes tumors leaky which is helpful, and they work in synergy with Reo on an individual basis in the lung, where Reo is preferentially cleared.

You might expect that I have some fairly high hopes for that particular study, but the pre-screen part is really a major step forward for everybody in our field, and specifically for us it's quite helpful.

Trial Cost:


{21:08} David Miller [Biotech Stock Research]: How far into the Pivotal Trial will your current cash resources get you?

Brad Thompson: We have enough cash to do the full 80 patients, and the full data analysis on the 80 patients. So, you're looking at about a 6 month window after the last patient is enrolled, which gets you certainly past a true response rate, true PFS, and the Kaplan-Meier should have spread far enough based on the anticipated death rate to get a full data analysis on that. We'll be actually enrolling patients in the second stage during that, and our burn covers off on that as well. That means, if you're looking at a milestone perspective, that gets us to that point.

It costs us around US$70,000 per patient for that study, and so you're looking at the primary enrollment in the first 80 patients is US$5.6 million, and you're probably enrolling another 50 – 60 – 70 patients at least into the second stage; so just to make the numbers easy, you're adding another $5.6 million of enrollment while you're waiting on analysis on the first part.

Partnering:

{22:25} David Miller: What's the current status on partnership talks?

Brad Thompson: We're talking. Getting the SPA was very helpful. I think it really helped lead to, at least the perception of decreased regulatory risk. We've been having people tell us for quite some time that they believed that the agent's clearly active, and I think that that's the case … you can make it, and it's safe, and we've got a great IP base and all that sort of stuff, but with a perception that with a new class of agent, and in this case a live agent … you know: what would the FDA do … and the SPA answers that question. The SPA had other uses than just getting the FDA to approve it, and I think that it's bearing some fruit in the partnering world.

David Miller: Do you have any sense of when one of them is going to bite … you know: in the middle of the initial phase, after the initial phase, after you've already figured it out?

Brad Thompson: I think that we have companies indicating all of those options that we're currently discussing things with. Some people would prefer that we don't have Phase III data before they partner with us because of course that changes the price tag. Some are actually more interested in the non-small cell lung than they are in the head and neck, so I think there will be some interest exhibited there later … the middle of the year hopefully we'll have interim data, again under confidentiality, that people are seeing our data in real time … so they'll know more about it than anybody on this call except for me. Or the equivalent of "my hair's wet" on a date request, which means they want pivotal Phase III data which means I don't need them anymore. Most of the people at the table are in those first two categories: they would prefer us not to have [Phase III] data, or are waiting for non-small cell lung data.

Uniform Enrollment:

{24:08} David Miller: One of the things that occurs to me with this Trial design is if you end up having a not linear Trial enrollment, or a back-end loaded Trial, that the initial look might be a little difficult. Is there any adjustment in statistical procedures, or otherwise, that you can look at that?

Brad Thompson: You can readjust the stats. I mean, you can always readjust stats, as long as the overall statistical plan is careful. It's best if we stick to the 10 per month. If it goes 2 per month, and then 60 patients in the last month, then we'll have to insert a bit of a wait there, at the end, clearly. But as long as we stay within broad parameters around that sort of 10 per month range, then we'll be ok. So that's just probability. On that poster session – I can't remember when it was … the last time we talked about the Phase II data out of the UK, we segregated the data, the investigators didn't have a clue why I wanted to do that, but I mean, I have reasons because I think stats. And there was a clear linkage, beginning to show between response rates and lifespan in those 19 patients, and that's continuing to develop and it's getting more accentuated, and there's always been a clear linkage between response rate and time to progression obviously. We're pretty confident that as long as we stick close to our enrollment that we'll have the kind of linkage that you need really to do, with adaptive design.

{25:44} David Miller: Are you going to update us on the number of patients you have enrolled as you go along?

Brad Thompson: Absolutely! I'm going to have fun doing that. I don't normally, as you know, talk about how many patients we have as we go along, but I'm planning to let people know about enrollment as we go out, as it's important for people to evaluate it, and for timelines with respect to other reasons.

Next Presentations:

{26:00} David Miller: Can you talk about when and where the next Reolysin data presentations might be?

Brad Thompson: No I can't. <round of chuckles here> We're waiting here for a couple of things for ASCO, and we're waiting to hear on a couple of things for other meetings, and some of the studies will be reporting sooner than expected, and some will be reporting later than expected. It's pretty dynamic. We have a number of sites that will be reporting this year, which I'm very much looking forward to.

And in summary …

{26:45} Brad Thompson: Well, I'd just like to thank everybody for joining us today. I think we really have a, I would say … interesting year ahead of us at Oncolytics. We have a number of our earlier stage studies, and other kind of interesting studies that will be reporting, and I can't tell you how thrilled we all are to be finally, right on the cusp of starting to enroll in our first Phase III study. So, again, I thank everybody for their attention, and interest in our company over the last few years … and stay in touch, `cause it's gonna to be fun.