Aevi Genomic Medicine, Inc . (NASDAQ:GNMX)
Q1 2017 Earnings Conference Call
May 10, 2017 8:30 AM ET
Executives
Michael Cola President and Chief Executive Officer
Brian Piper Chief Financial Officer
Garry Neil Chief Scientific Officer
Analysts
Jason Butler JMP Securities
Brian Marks Zacks Investment Research
Operator
Good day and welcome to the Aevi Genomic Medicine First Quarter Conference Call and Business Update. Todays conference is being recorded. At this time, I would like to turn the conference over to Mr. Brian Piper, Chief Financial Officer. Please go ahead sir.
Brian Piper
Thank you, Gina. Good morning everyone and welcome to the conference call. As a reminder, a copy of todays presentation can be found on the Aevi Genomics website. Participants on todays call are Chief Executive Officer, Mike Cola; Chief Scientific Officer, Garry Neil; and myself Chief Financial Officer, Brian Piper.
Before we begin, I would like to direct your attention to Slide two and remind you that todays discussion will include statements about the companys future expectations, plans, and prospects that constitute forward-looking statements for purposes of the Safe Harbor provisions under the Private Securities Litigation Reform Act of 1995.
And with that, I would now like to turn the call over to Aevi Genomic Medicines CEO, Mike Cola. Mike?
Michael Cola
Thank you, Brian and welcome everyone to our corporate update. Im on slide three. We have a fairly tight agenda today, mostly focused on 001. We will take you through the SAGA trial summary, although we missed our primary endpoint, were very excited by the genetics subset of responders that weve been working on over the last six weeks. Garry will take you through that analysis and our steps forward. I will also readdress the market potential. We do have a reduced set of genes, but we still think its a very significant and compelling business story moving forward its a very significant portion of the ADHD market.
I will briefly touch on 002 as were just getting that trial off the ground and Brian will take you through the Q1 results of operation. As you know, our mission is focused on translating genetic discoveries into novel therapies to improve the life of children and adults with pediatric onset life altering diseases. We do that through our collaboration with Childrens Hospital of Philadelphia and that collaboration over the last 2.5 years has yielded two programs on slide five; that is 001 which well spend the bulk of the time on today and AEVI-002 or anti-LIGHT program.
Garry will take you through our plans for studying the genetic subset within ADHD that weve identified as responders. We expect that data to be out mid-2018 at the latest. It will explain the trial design and why we think we have some opportunity to do better than what we did in the SAGA trial as far as timing. 22q Deletion Syndrome, as you know, weve had struggles to recruit into that trial. These patients have been physically very fragile. Parents have not been really that excited about taking any chances with their child as far as changing therapy. We will talk about two patients in the coming months in top-line data.
Garry will also walk you through the story around contactin-4 or CNTN4, in autism spectrum disorder. We have an unusual finding, a novel finding in ADHD as far as the prevalence of that particular gene that causes disease in ADHD, we think its a very important gene in autism. And we are doing work with CHOP today to understand its phenotype and prevalence in autism. Hope to have discussions with the agency in the middle of the year.
And then as I mentioned earlier, well talk through Severe Pediatric Crohns Disease, as that program is just getting off the ground with initial data the second half of 2017. This will be open label data and will be based on some end points that are harder than what were currently working with.
And with that, Ill turn it over to Garry.
Garry Neil
Thanks, Mike. Im now on slide seven. So this is a high level schematic of our AEVI-001 ADHD development plan and youll notice that the plan is very similar to our original plan but has been updated to include a new Phase 2 trial shown here in dark blue on the right. And Ill describe this new study in more detail a little later. Our plan remains otherwise very similar to the previous plan, in that we expect to ultimately perform two pivotal Phase 3 trials in a genetically selected subset of ADHD patients, one in 6 to 12 year olds and the other in 6 to 17 year olds along with a longer term safety trial for first approval in ADHD. Phase 3 will now be delayed until we confirm the findings that our post-hoc analysis in the new Phase 2 trial.
Now also as a reminder, we began our development program last year with an ADHD Phenotype/Genotype study shown in the upper left corner in children aged 6 to 17. And this recently completed non-interventional study has several objectives mainly one; confirm the prevalence of mGluR CMV mutations and individual genes in U.S. children with ADHD; two, better characterize the mGluR CMV mutation positive phenotype; and three, to find suitable patients for our clinical interventional study, and we did all of that. We ultimately enrolled 1,876 kids from sites across the country, all of them had ADHD. We genotyped them, analyzed their phenotype drug use and so on. And this patient pool was of course used to recruit the patients for SAGA.
The Phenotype/Genotype study taught us a lot about how to recruit these types of patients, about mGluR gene distribution and the phenotype in ADHD. For example, we learned that the genes are not evenly distributed across the 273 gene network. In fact, 75% of the patients had a CMV in just 1 of 25 genes. And one gene contacted four, well talk about that later, accounted for 20% of all the mutation positive kids with ADHD. We also saw differences in the ADHD phenotype at the gene level and were still understanding that. Ill provide more details on this on subsequent slides. In these analysis, theyve all been critical for the interpretation of our responder analyses and setting the path forward.
Slide eight. As I mentioned, the Phenotype/Genotype study was the source of patients for the SAGA trial which is showed here schematically. As a brief reminder, the trial was a multi-center 24 site, six week randomized placebo controlled parallel design study to assess the efficacy and safety of AVEI-001 in adolescents with genetic mutations impacting the mGluR network in ADHD. It was a one-to-one randomization design, either drug or placebo, we enrolled 96 patients; 46 randomized to drugs, 50 to placebo.
There were two phases to the trial, a four week dose optimization phase where subjects received AVEI-001 or placebo, initiated at 100 milligrams twice daily and - weekly up to 400 twice weekly based on clinical response and tolerability. Then at the end of week four, there was a second two week maintenance phase and subjects in that phase continued with their optimized dose for an additional two weeks. The large majority of subjects reached the highest 400 milligram twice daily dose. The primary endpoint in this trial, as youll recall, was improvement in the ADHD-RS and the key secondary endpoint was reduction in the Clinical Global Impression of Improvement, CGI-I.
Slide nine. So, as you also know, and as weve previously reported, the SAGA trial did not meet its primary endpoint. There was numerical superiority favoring the drug with an absolute reduction of about 15 points in ADHD-RS on active treatment and about 12 on placebo, but that was not significant. Interestingly, when we looked at the response on the inattention subscale of the ADHD-RS, there was a nearly significant trends favoring patients on drug, but that also just fell slightly short of significant. But we also pre-specified two responder analyses in our analysis plan.
And as I mentioned last time, responder analysis provide several benefits in assessing and interpreting clinical trial data, particularly in trials that rely on rating scale endpoints and those may have a high placebo or variable response. But the main reason for doing this was because we suspected that not all of the genes in the mGluR network would predict response to AEVI-001, even though they are highly predictive of the disease. Not knowing what genes might respond, we did not pre-specify individual response apriori.
Now in these analyses, these responder analyses, a significant high proportion of AEVI-001 patients responded by both ADHD-RS and CGI-I, 70% on drug versus placebo, 42% for ADHD-RS and 57% on drug versus 33% on placebo for CGI-I both of those were highly significant. So take it together, we felt these findings strongly suggest that there must be a responding population and we were able to very quickly exclude factors such as gender and weight, and the most likely explanation that remained and the one that we had foreseen was a genetic responder sub-population. We plan to do such an analysis in any event, although we had hoped to do it in the context of a positive result overall.
Lets move on to slide 11 and talk about the additional post-hoc analyses that we did. So, again, we wanted to focus on genetic responders. We approached the analysis in the following way; we knew that we had CMV in the total of 44 genes from the 273 gene network in the SAGA study. We also knew that responding patients on drug had a very robust response with an ADHD-RS reduction of over 20 versus only about 0.5 on average for the non-responders. So we decided to focus on genes that were sufficiently prevalent for analysis, had at least one patient on drug and one on placebo, and that showed a meaningful response to AEVI-001. We included some GRM genes and contactin-4 because these were both prevalent and showed a good response to the drug but we also included some other neurodevelopmental genes.
Now doing it this way, we knew that we would miss some responding genes because the numbers were too small for analysis in this study, but we felt it would be best at this stage to focus on those genes that we thought would give the best likelihood of clinical response. Taking this approach, we identified a 9-gene subset which comprises, as Mike said, about 10% of the pediatric ADHD population or about 40% to 50% of all of the mGluR positive ADHD patients which is a very substantial and study-able population. As Ill show you in the subsequent slides, this responder sub-population had a very robust response to drug.
Lets go to slide 12. So heres the frequency of the responder genes to 001. At this point, as we file IP, we are only revealing the identity of one of these genes, contactin-4 which is shown on the left, most prevalent one. As I mentioned earlier, about three quarters of the population seems to express the CMV in one of only, in this case 24 genes, and you can see here that our analysis identified the responders 8 of the 15 most frequent genes. And then the SAGA trial list accounted for about 40% of all of the mGluR patients that were studied.
Slide 13. Once again the most prevalent CMV in both the Phenotype/Genotype and the SAGA trial was contactin-4 which accounted for about 22% of the entire mGluR CMV positive population and about 5% of trials of ADHD all by itself. Interestingly, patients with this CMV may also have a more severe phenotype.
Lets go to slide 14. Here is an analysis of ADHD-RS reduction in the SAGA trial, limiting the analysis to the nine genes of interest that I just mentioned. As you can see, this analysis is highly statistically significant and has a placebo subtracted reduction in ADHD-RS of nearly 12 in week six. And as noted on the slide, you can see that 43% of patients in this subset had a contactin-4 CMV 14% were one of the GRMs and 42% were under neurodevelopmental genes.
Lets go to slide 15 and put this in some context. I wanted to show you here the change in ADHD-RS, the primary endpoint in SAGA in three adolescent trials. Were comparing SAGA now to two additional adolescent ADHD trials. We limit the comparison to adolescent trials because adolescents typically have a higher placebo response and lower treatment effect size than pediatric patients do.
So we wanted to make this an apples-to-apples comparison. When we compare AEVI-001 in the 9-gene subset that I just showed you to these two highly effective and successful ADHD medications Vyvanse and Adderall XR, we can see that AEVI-001 had a much greater placebo subtracted reduction total ADHD-RS and in both the inattention hyperactivity subscales that either of those drugs. Of course, its not a head to head comparison, but we were struck by the fact that the magnitude of the treatment effect didnt, in our subset, exceeded that of or even high dose Vyvanse with respect to overall change with ADHD-RS and both the inattention and hyperactivity subscale. So we think we were pretty struck by that finding.
Lets go to slide 16. We also looked at the responder analysis in the 9-gene subset. As a reminder, we used standard response definitions namely, an ADHD-RS reduction of 30% or more and a CGI improvement rather, to 1 very much improved or 2, much improved on the 7 point Likert scale. And as you can see, the response rate of the 9-gene subset were 89% and 72% respectively ADHD-RS CGI-I compared to only 21% and 13% of placebo, both very significantly superior to placebo.
Lets go on to slide 17. Now I want to talk a little bit about contactin-4. Again, we were particularly intrigued by the discovery that contactin-4 was so prevalent in our ADHD population, both in the Phenotype/Genotype study and the SAGA trial and the gene is very interesting. It encodes an anchored neuronal membrane protein that functions as a cell adhesion molecule. The developing nervous system requires the formation of many complex interconnections and network and contactin-4 appears to play a key role in the formation of axon connections and synapses in developing the nervous system.
Mutations in this gene have been previously with a variety of neurological conditions, including neurodevelopmental delay, autism spectrum, bipolar disease and schizophrenia and we also had seen those associations in the network more broadly. Now the CHOP team was the first to identify an association between ADHD and CMV affecting this gene, but because we use the higher resolution technology, namely the Illumina Omni 2.5 chip in our genotyping, we were able to detect smaller CMVs in this gene and this is probably why we detected a higher frequency of the gene than the CHOP team originally did. Consequently, we estimate, as I said earlier, the frequency of the CMV to be 5% of the overall ADHD population.
Next slide, slide 18. Equally interesting, when you look at the phenotype associated with contactin-4 mutated, it gets with the mutation in contactin-4 with ADHD, the observation that these kids seem to have a higher substantially and significantly higher prevalence of emotional dysregulation symptoms including disruptive behavior, anger control, risk taking, inappropriate movements and sounds as compared to the non-mGluR CMV positive ADHD kids. And we previously reported this in a broad population but were now seeing it as contactin-4 drives the substantial part of this. So were doing more work to better understand this phenotype. We do know though that the emotional dysregulation phenotype and ADHD often predicts worst life outcomes for the patients.
Next slide, slide 19. So given the high prevalence and the unique features of this gene, we chose to analyze it independent of the other genes in the selected nine gene subset. So we looked at it alone. And we observed that the AEVI-001 response in patients with CMV associated with this gene was higher than with any other. In fact, all six patients who were randomized to receive AEVI-001 responded vigorously as compared to 3 in 12 patients randomized to placebo. And the magnitude of response was the highest seen with any gene, so even with data on only 18 patients, the reduction in the primary endpoint of ADHD-RS, SAGA was statistically significant.
And going to slide 20, again, when we do responder analyses using just patients with the mutation in this gene, you will see that we have a very statistically and clinically significant difference between placebo and drug for both ADHD-RS and CGI, 100% and 83% respectively.
Lets go to slide 21 and just to summarize, we were able to identify our responder subset of patients who bear a CMV in one of nine genes. Taken together, this subset accounted for about 40% of all mGluR CMV positive ADHD patients and about 10% ADHD patients overall. So this was a very substantial population. Response rates in patients in this subset were very high, but were noticeably higher in patients with a CMV in one particular gene, contactin-4. All patients with a CMV and contactin-4 in the SAGA trial responded to AEVI-001. The prevalence of contactin-4, CMV in our population of ADHD patients was greater than 20% of the overall mGluR positive patients and accounts for 5% of all ADHD patients.
So by itself, this is a very substantial and study-able population. Moreover, preliminary data suggests that these patients may have a more severe disease phenotype characterized by higher prevalence of emotional dysregulation. So we believe that the best path forward for the development of AEVI-001 in ADHD lies in enrichment of patients with CMV in these genes in future study. Not to say that other genes in the mGluR network could also not predict or could predict response to AEVI-001, these can also be elucidated in future studies, but we feel its necessary to first concentrate on patients with mutations that have been observed in patients who did show response. I want to emphasize that not all genes predict response to AEVI-001, we remain convinced that the 273 gene mGluR network is highly predicted for the disease ADHD.
Slide 22, we talk about the next steps for AEVI-001 in ADHD and autism. Were currently finalizing the design of the new Phase 2 trial I mentioned to confirm our post-hoc analyses in the SAGA trial. Right now we plan a multi-center randomized placebo controlled trial in patients aged 6 to 17, use of younger patients which should now be enabled by our juvi talks and pediatric pk studies should result in better effect size and lower placebo response. We plan to use the stage adaptive design with the initial focus on patients with contactin-4 CMV. The N should be around 40 to 70 patients.
Subsequent stages will enroll patients with a CMV in the remaining genes in the 9-gene subset. We expect to get this study underway in the second half of this year, 2017, with top-line data expected mid-2018. As I noted, contactin-4 looks particularly interesting in autism spectrum disease, its prevalence maybe even higher in that population than it is in ADHD. Were currently genotyping ASD patients at CHOP with the Omni 2.5 chip and studying their phenotype in more detail. Depending on the data, we plan a discussion later this year with FDA regarding a potential orphan indication for contactin-4 related autism.
So with that, Ill turn it back to Mike.
Michael Cola
Thank you, Garry. Were on slide 23. As I mentioned earlier, with the 9-gene subset, we have a reduced set of patients that were looking at approximately 10% of the market. So we still think its a very significant and compelling business case for the 9-gene subset. Looking at current sales of about $11 billion, growing slowly and about 90% of this sales are stimulant, we still think there is a very high unmet need for a non-stimulant that is effective. We have about 6 million patients in the 6 to 17 category and about 10 million adult patients.
If you take 10% of that, there are about 1.6 million patients total using current premium pricing of about $15 a day and current compliance and adherence which we think we can improve on, about five to seven scripts per year, still end up with a total addressable market of $2 billion to $3 billion which is obviously sizeable. As we understand more about other diseases with these genes, particularly ASD, we think there is a lifecycle opportunities in those diseases as well.
And with that Ill move on 002 in Severe Pediatric Onset Crohns Disease. This is the program that we licensed in the middle of last year in 2016. Its based on a discovery at Childrens Hospital of Philadelphia that in very severe early onset Crohns disease and IBD in general, you have a loss of function indicated by genetic mutation in decoy receptor 3. Decoy receptor 3 is a buffer protein that reduces inflammatory cytokines particularly light that are causal in IBD. And when you have left DcR3 and have more light, you have these severe pediatric onset symptoms. We think that comprises anywhere from 10% to 15% of those patients.
We did begin our search looking for a DcR3 analog, we couldnt find one with an appropriate profile, particularly for children and so we moved on to monoclonal antibodies that known down light. And we were lucky to find one from Kyowa Hakko Kirin which was just coming back from a collaboration with Sanofi Aventis. We were able to get that deal done in the middle of last year and weve made great progress over the last nine months to get us toward the clinical trial. Obviously, we successfully reinitiated and transferred the IND with a new protocol.
We have re-qualified the clinical trial material which was the major part of the work over the last nine months and we are cleared by the FDA to start a trial, still have some administrative work at CHOP to get this thing up and running, but were moving through that quite rapidly and we expect this study starts at the end of this month, beginning of June. And initial data in the second half of 2017 with endoscopic - in Crohns Disease Activity Index as our primary endpoints.
And with that, Im going to turn it over to Brian for the quarterly financial update.
Brian Piper
Thanks, Mike and Ill walk us through the results of operations from Q1. R&D expenses for the quarter were $7.9 million, increasing from $7 million for the same period in 2016. Not surprisingly, this is mainly due to increased spend on clinical activities related to completion of the SAGA for AEVI-001 and to a lesser extent, initial cost start-up cost for the AEVI-002 program. We do anticipate clinical spend should decrease going forward as the majority of clinical activities related to the SAGA trial have ceded.
G&A expenses for Q1 17 were $3 million decreasing from $4.2 million for the same period in 2016. This is mainly due to severance benefits recorded in 2016. Cash we recorded at the end of March 31, $29.2 million, importantly, we do feel that these cash resources will be sufficient to fund operations through Q2 2018 which would in course include delivering top-line data by mid-2018 in the Phase 2 trial to confirm the genetic responders to AEVI-001 that Mike and Garry described earlier as well as initial data by the second half of this year from the signal finding trial of AEVI-002 in Severe Pediatric Onset Crohns Disease.
And with that, I will turn it back to Mike for any final comments and Q&A.
Michael Cola
Thank you, Brian. On slide 28, you see our upcoming events on our pipeline chart again, very busy time ahead of us, getting a trial up and running in the mGluR positive genetic subset, particularly with the Crohns program starting as well, with initial data in the second half of the year. We expect success in these trials and we continue to be very confident in our ability to execute again with this plan.
And with that, Ill turn it back to the moderator. Thank you.
Question-and-Answer Session
Operator
Thank you. [Operator Instructions]. And well take our first question from Jason Butler from JMP Securities.
Jason Butler
Hi. Thanks for taking my questions and thanks for going through all the details in the prepared comments. Just first question, you mentioned that in the CNTN4 patients, there are more disruptive behaviors. Can you talk about whether the severity of ADHD symptoms as a whole is, is worse than the patient population or what their response to current ADHD treatment option is relative to the border ADHD population?
Garry Neil
Thats still something that were analyzing. I mean we see this with a much higher odds ratio of these symptoms which are prevalent to begin with, in ADHD patients. So it is a source of considerable morbidity in those patients.
Michael Cola
But we do, in the phenotypic data that Garry showed on slide 18, know that approximately two-thirds of the patients that were in the phenotypic study were on stimulants at the time of the questionnaire or had been on stimulants. So, although we are doing actual EMR work with Childrens Hospital now to understand how theyre responding to therapy, the increased odds ratio for these more severe symptoms you would expect they are not responding to stimulant therapy very well, but we have to clarify that going forward.
Jason Butler
Okay, great. And then from the next Phase 2 trial, when we see results mid-18, will that be from just the first cohort of CNTN4 patients or will that be from old patients in the trial?
Garry Neil
Yeah, our approach Jason is to win on contactin-4 first and I dont have an answer for that because were going to do it as an adaptive design. Once we go on CNTN4, well go to the other eight genes. So our hope is to do it as quickly as possible, were working through the details of that study right now, its kind of coming together real-time, but obviously our goal would be to provide much data as we possibly can, as quickly as we possibly can.
Jason Butler
Okay, great. And then just last question from me, can you just give us an idea of how many patients well see from the 002 trial later this year?
Garry Neil
So there are cohorts of four patients. We do have a fairly lengthy washout requirement than we have to do the first patient sequentially, but were hoping were going to get us as many of that first cohort done as we can this year.
Jason Butler
Okay, great. Thanks for taking the questions.
Michael Cola
If you remember Jason, its also an open label trial. So we will report that data as it comes in, we dont necessarily have to wait for the first four patients.
Jason Butler
Great. Thanks.
Operator
And we will take our next question from Brian Marks from Zacks Investment Research.
Brian Marks
Hi. Good morning guys. Among the children that fit into the CNTN4 cohort, is there much comorbidity with ADHD and autism or any other disorders such that it might present challenges in designing in ADHD study just with the CNTN4 cohort?
Garry Neil
Well, we had no problem enrolling them with our enrollment criteria in the SAGA trial. But were interested in the fact that, as Mike said, were really looking at a phenotype that appears to have more emotional dysregulation. Were also looking to find better ways to be able to evaluate the response to therapy on those specific symptoms which arent always perfectly evaluated by the existing scales. Were going to be thinking about some supplemental scales to use in those patients. But no, I dont think it will be a problem getting patients in the study, but we are very interested, as I said, were doing much more deep phenotyping looking at autism and other comorbidities associated with this gene and there may be a spectrum of neurodevelopmental disorders which would provide a lot of opportunity for us for AEVI-001.
Brian Marks
Okay. In terms of endpoints in the study, will they be similar to SAGA?
Garry Neil
ADHD trials are fairly stereotypical in that everybody uses the regulatory gold standards which are ADHD-RS and CGI-I as their primary endpoints. But we also, as I said, well also be using some additional endpoints that are more specifically targeted in some of the symptoms of interest that weve already mentioned
Brian Marks
I think you guys said, when you talked about the SAGA top-line results indicated that there could be potential to increase the dose or increase the treatment duration. Is that potentially still in the cards maybe not with this upcoming CNTN4 study but potentially with the follow-on 9-gene study?
Garry Neil
Go here to read the rest:
Aevi Genomic Medicine's (GNMX) CEO Michael Cola on Q1 2017 Results - Earnings Call Transcript - Seeking Alpha
