Wave Life Sciences Ltd. Q1 FY2021 Earnings Call

Good morning and welcome to the Wave Life Sciences First Quarter 2021 Financial Results Conference Call. At this time, all participants are in a listen-only mode. As a reminder, this call is being recorded and webcast. I'll now turn the call over to Kate Rausch, Head of Investor Relations at Wave Life Sciences. Please go ahead.

Thank you, operator. Good morning and thank you for joining us today to discuss our recent business progress and review Wave's first quarter 2021 operating results. On the call with me today are Dr. Paul Bolno, Wave's President and Chief Executive Officer; Dr. Mike Panzara, Chief Medical Officer, Head of Therapeutics Discovery and Development; and Kyle Moran, Chief Financial Officer. This morning we issued a news release detailing our first quarter financial results and provided a slide presentation to accompany this webcast. These materials are available in the Investors section of our website www.wavelifesciences.com. Before we begin, I would like to remind you that discussions during this conference call will include forward-looking statements. These statements are subject to a number of risks and uncertainties that could cause our actual results to differ materially from those described in these forward-looking statements. The factors that could cause actual results to differ are discussed in the press release issued today and in our SEC filings, including our Annual Report on Form 10-K for the year ended December 31, 2020. We undertake no obligation to update or revise any forward-looking statements for any reason. I'd now like to turn the call over to Paul. Paul?

Thanks, Kate. Good morning, everyone, on the call. Thank you for joining us. During the call today, I will provide some opening remarks, after which Mike will give an update on our clinical trials and Kyle will briefly review our financials. It has been an incredibly productive start to the year for Wave as we advanced three cross-generation stereopure oligonucleotides into clinical development. We have formally initiated clinical trials for WVE-004, our C9orf72 candidate in ALS and FTD; and WVE-003, our SNP3 candidate in Huntington's disease. We've also received important regulatory approvals towards initiating our third PN chemistry program, targeting exon 53 in DMD, WVE-N531. These clinical trials are designed to enable rapid proof-of-concept using biomarker-driven adaptive designs and are the first investigational candidates designed with our novel PN backbone chemistry modification. Next year, we expect that data from these clinical trials will enable decision-making about next steps for these programs as well as provide insight into PN chemistry across different modalities, tissue types and targets. We've also made substantial progress with our endogenous ADAR editing capability, which we believe is the most advanced in its class. We have generated a breadth of RNA editing data, demonstrating activity across in vivo and in vitro systems, including in vivo editing in the central nervous system. Much of this data is being presented in an oral presentation tomorrow, May 14, at the ASGCT Annual Meeting. Our first ADAR editing program for alpha-1 antitrypsin disease has generated promising initial results and we are on track to share in vivo data this quarter. Our PRISM platform is unique and differentiated from others developing RNA therapeutics. At our foundation we set out to embrace rather than ignore the reality and importance of stereochemistry that exists in each and every oligonucleotide. In choosing control for the three-dimensional orientations of backbone linkages and advancing single-isomer therapeutics, we can apply the principles of rational drug design to our pipeline candidates, which is impossible with mixture-based oligonucleotides. This resolution enables us to define distinct profiles for our stereopure molecules and we now have several years of clinical data to further inform our platform. Earlier this year, we announced the discontinuation of our remaining first-generation programs following the results of the PRECISION-HD trials. While we only saw modest and inconsistent reductions of mutant huntingtin, it is important to note there were no clinically meaningful trends in disease progression or laboratory values such as elevations in CSF white blood cells, proteins, and neurofilament light chain or NFL. There were, however, suggestions of allele selectivity, underscoring the precision enabled by our platform. In our next-generation programs, we have prioritized the use of in vivo models during preclinical development to ensure we advance clinical candidates that will reach the desired site of action and engage target. In addition to the wealth of data collected over the past several years, we are also leveraging an influx of new talent in oligonucleotide therapeutics to further advance our understanding of design principles, pharmacology and toxicology. The application of PN backbone chemistry modifications in the context of controlling stereochemistry was a major advancement that emerged from our platform. Based on what we have seen preclinically, this innovation has the potential to significantly improve the profile of therapeutic oligonucleotides, independent of sequence, tissue type or modality. Separately, our ADAR editing capability further expands our toolkit beyond silencing and splicing, enabling us to select the best modality to address the root cause of genetic diseases. We anticipate sharing more on PN chemistry and ADAR editing at a Research Day later this year. Our current pipeline is comprised of programs designed with next generation PRISM including PN chemistry. I'm extremely proud of how quickly we have advanced this innovation to the clinic and we are rapidly approaching the first of many opportunities for clinical proof-of-concept of PN chemistry. I'd now like to turn the call over to Mike Panzara for an update on our neurology programs. Mike?

Thanks, Paul. The foundational work that has been done throughout the evolution of PRISM has provided us with a diverse and robust neurology-focused portfolio that is currently moving through stages of preclinical discovery and clinical development as illustrated here. As Paul just mentioned, all of our current discovery-stage and preclinical programs utilize PN chemistry, including the multiple discovery programs in collaboration with our partner Takeda. These programs are yielding exciting results, including target engagement and distribution in the central nervous system of nonhuman primates, which further validate our approach. Our Therapeutics Discovery portfolio continues to build upon this progress to maximize the potential of oligonucleotide therapeutics for the treatment of neurological disorders with high unmet need. Now, I'd like to discuss the programs currently in clinical development with three next-generation candidates. Our development organization is focused on site activation and initiating dosing simultaneously in three clinical trials across four disease areas: C9orf72-associated amyotrophic lateral sclerosis or ALS and frontotemporal dementia or FTD with WVE-004, our candidate targeting C9orf72 hexanucleotide repeat expansions; Huntington's disease with WVE-003 our SNP3-targeting candidate; and Duchenne muscular dystrophy with WVE-N531 our exon 53 skipping candidate. Each of these clinical candidates incorporates PN chemistry and the availability of relevant preclinical models has enabled a greater understanding of PK/PD relationships to guide development. Further, the learnings from our first-generation programs are being incorporated to mitigate risk and more efficiently execute our plans. Starting with C9orf72. Our clinical candidate WVE-004 is designed to target a hexanucleotide repeat expansion in the C9orf72 transcript, which is one of the most common genetic causes of ALS and FTD. These expansions drive the common pathophysiology underlying these two diverse and devastating phenotypes. 004 is the first C9orf72 candidate being advanced simultaneously in a single basket-like study for both C9 ALS and C9 FTD. C9orf72 mutations lead to multiple drivers of toxicity. The hexanucleotide repeat-containing RNA transcripts deposit in tissues and are toxic on their own, but they are also translated into long dipeptide repeat proteins or DPR proteins that trigger cellular toxicity through a variety of downstream mechanisms. 004 selectively targets the pre-mRNA of variant transcripts that contain the hexanucleotide expansion with the goal of suppressing both the RNA and DPR-associated toxicities, while allowing C9 protein expression. In the first quarter our foundational work to identify and validate the targeting site used to achieve this selective knockdown was published in Nature Communications. The preclinical data demonstrates 004's ability to rapidly and durably reduce over 90% of the DPR poly-GP in the spinal cord and reduce at least 80% of poly-GP in the cortex. This effect lasted at least six months after two intracerebroventricular injections of 004 administered seven days apart at the start of the study. C9orf72 protein was unchanged over the same period. The effect of 004 on poly-GP in the CSF is a key endpoint in our clinical study, so we are looking forward to assessing the impact of treatment in humans, given these promising preclinical results. These results along with data from nonhuman primates have also allowed us to start at a dose that our clinical trial predicted to be pharmacologically active. This week at the European Network to Cure ALS Virtual Meeting or ENCALS we introduced FOCUS-C9, an adaptive trial that is designed to enable faster optimization of dosing frequency of 004 based upon review of unblinded data throughout the study. FOCUS-C9 is a Phase Ib/IIa global, multicenter randomized double-blind, placebo-controlled trial, in which we are planning to enroll approximately 50 patients with documented C9orf72 expansions and confirmed ALS, FTD or mixed phenotypes. FOCUS-C9 includes single- and multiple-ascending dose portions of 004 administered intrathecally. At points throughout the study, based upon predefined data-driven milestones, an independent committee will review unblinded data to determine the next single-dose level to be escalated to and the optimal frequency in the next multi-dose cohort, meaning whether the dosing interval should be monthly or less frequent. Samples are collected for biomarker analysis at multiple time points within both the single- and multi-ascending dose portions to enable the assessments required to make these recommendations. Regulatory and ethics approvals have been received and clinical site activation is underway, so we anticipate dosing soon. I'll now turn over to WVE-003, our allele-selective candidate for Huntington's disease, which is designed to selectively lower mutant huntingtin while preserving wild-type huntingtin. The presentations, posters and feedback from experts at the recent virtual CHDI HD Therapeutics Conference only served to bolster our confidence that we are pursuing the right approach to Huntington's. Let's review what we know. Patients with Huntington's disease have expanded CAG repeats in their huntingtin gene that lead to production of mutant huntingtin protein. This is a monogenic autosomal dominant genetic disease that is fully penetrant and affects the entire brain. Preserving wild-type huntingtin is as important as lowering mutant huntingtin. Evidence supports that Huntington's disease is driven by both the gain of function of mutant huntingtin protein and the loss of function of wild-type protein, which is essential for homeostasis of the central nervous system. Wild-type protein is critical for the protection of neurons that are under stress and plays a key role in trafficking synaptic proteins and vesicles including the production and transport of brain-derived neurotrophic factor or BDNF in the cortex. Wild-type protein is also critical for the formation and function of cilia, which control the flow of CSF and help maintain CNS homeostasis. In healthy individuals these important functions of wild-type huntingtin balance out the collective stresses based on the central nervous system. However in the case of HD, there is the added burden of the mutant protein itself. Those living with HD have been subjected to decades of toxic stress that come with mutant huntingtin protein years prior to symptom onset. Looking at levels of wild-type in healthy individuals or models that lack the effect of mutant protein does not adequately represent the role the healthy protein plays in the context of Huntington's disease. This smaller protective reservoir of wild-type huntingtin eventually loses the battle to the expected stresses placed on the CNS along with the toxic effects of mutant huntingtin resulting in disease progression. If one thinks about this as a push-pull of positive and negative factors in this balance of wild-type and mutant huntingtin in the CNS, it stands to reason that depletion of wild-type protein along with mutant protein, as with non-selective approaches, would shift the balance towards disease progression, erasing any benefit or even potentially accelerating decline. This has been our hypothesis since we began our HD program. The data that are emerging support our position and make us resolute in our differentiated approach to treating this disease. 003 has been improved over our prior SNP-targeting candidates by applying PN backbone chemistry modifications in the context of control over stereochemistry. Further, the presence of the SNP in a relevant animal model has allowed us to do in vivo preclinical work to determine a dose predicted to be pharmacologically active right from the start of the first cohort. The slide illustrates some of these in vitro and in vivo data demonstrating that 003 is highly selective for mutant huntingtin and able to achieve potent and durable knockdown of mutant huntingtin in vivo in the BACHD mouse model. We investigated this model knowing that there were several limitations including the fact that it contains multiple properties of the mHTT gene, some of which do not have the SNP3 variant. Nonetheless, we observed potent and durable knockdown of mutant huntingtin in the striatum out to 12 weeks, with a similar effect observed in the cortex. Nonhuman primates do not have SNP3 and as such we are not able to evaluate the pharmacodynamic effects in this model. Therefore, we use the concentration data from the BACHD mouse compared with tissue concentrations in the CNS of nonhuman primates and then use PK/PD modeling to estimate tissue concentrations required to achieve striatal and cortical knockdown in humans with 003. These analyses are guiding the starting dose and dosing regimen planned for our clinical trial. While target engagement studies in the CNS of nonhuman primates were not possible for 003, they were possible for our most advanced therapeutic candidate in our CNS discovery collaboration with Takeda, WVE-005. Like 003, this candidate uses PN chemistry. But unlike 003 the human transcript targeted by 005 is homologous to the monkey sequence allowing us to assess target engagement throughout the CNS. In this study for an undisclosed target nonhuman primates received a single 12-milligram intrathecal injection of 005. One month after administration, we observed that the candidate was widely distributed throughout the CNS and led to substantial knockdown of target including in the striatum. This experiment once again highlights the potential of this next-generation chemistry. In the first quarter we received regulatory and ethics approvals to initiate SELECT-HD, a Phase Ib/IIa global multicenter randomized double-blind placebo-controlled trial of 003 in early manifest HD. We are targeting enrollment of approximately 36 patients carrying SNP3 in association with expanded CAG repeat. Patients from the PRECISION-HD studies will be able to transition to SELECT-HD after a washout period assuming they meet other inclusion and exclusion criteria. Unfortunately, based on the recently disclosed safety and efficacy data, patients who received active treatment with tominersen in the GENERATION-HD study will not be permitted to enroll in SELECT-HD. Although those who received placebo in GENERATION-HD are eligible to be screened for study entry. Like FOCUS-C9, SELECT-HD has an adaptive design to enable optimization of dosing frequency and more rapid determination of target engagement. An independent committee will evaluate unblinded data on an ongoing basis to guide dose escalation and dosing interval in each cohort. Key objectives, in addition to safety and tolerability, include plasma PK, CSF exposure of 003 and changes in key biomarkers including mutant huntingtin, wild-type huntingtin and neurofilament light chain over the course of the study. Clinical site activation is underway and we expect to dose our first patient soon. WVE-N531 is our systemically administered candidate for patients with Duchenne Muscular Dystrophy or DMD that are amenable to exon 53 skipping. This is also our first stereopure splicing candidate, designed applying PN chemistry. As we have shared previously, when applying this format to an exon 23-targeting surrogate treatment of an aggressive double knockout or DKO mouse model lacking both utrophin and dystrophin, we observed rescue of mice treated with 75 milligrams every other week as compared with PBS or first-generation chemistry dosed at 150 milligrams per kilogram weekly. Once again, application of the PN backbone modifications had a profound effect. In March 2021, we initiated clinical development of N531 with the submission of a clinical trial application. Since then we've received regulatory approval for an open-label clinical trial that is powered to evaluate whether N531 dosed every other week increases dystrophin production in up to 15 boys with DMD. The trial will also assess drug concentration in muscle and initial safety. Dosing is expected to initiate this year. I will now pass the call back over to Paul to discuss our ADAR editing capability and upcoming milestones there. Paul?

Thanks, Mike. We continue to generate compelling RNA editing results with our ADAR editing capability which we believe has many advantages and positions us at the forefront of this space. As a reminder, our approach to RNA editing employs short fully chemically modified oligonucleotides to recruit endogenous RNA editing enzymes called ADAR. Our ADAR editing compounds are optimized using our proprietary stereochemistry and PN chemistry, which enables us to avoid delivery vehicles such as AAV vectors or nanoparticles and allows us to leverage established manufacturing processes. To date, we have demonstrated editing activity across in vivo and in vitro systems including durable RNA editing of up to 50% in nonhuman primates with GalNAc-conjugated oligonucleotides. Our ADAR editing oligonucleotides are also highly specific. Our practical approach to RNA editing opens the door to a number of therapeutic applications, such as restoring or modifying protein function and up-regulation of protein expression. These applications greatly expand the landscape of disease variants that we can potentially address and we are advancing discovery work for multiple ADAR editing targets as well as evaluating new potential targets. Our first ADAR editing program uses a GalNAc-conjugated oligonucleotide to correct a single RNA-based mutation in the mRNA coded by the SERPINA1 Z allele, which triggers alpha-1 antitrypsin deficiency or AATD. ADAR editing is a simple and efficient approach to treating this disease by simultaneously reducing aggregation of the mutated misfolded alpha-1 protein and increasing circulating levels of wild-type alpha-1 antitrypsin protein thus addressing both the liver and lung manifestations of AATD, all while avoiding the risk of permanent off-target changes to DNA. Last year, we successfully demonstrated upwards of 60% editing of the SERPINA1 Z allele transcripts to wild-type in hepatocytes in vitro, which led to a threefold increase in functional wild-type AAT protein. Encouraged by these results, we moved forward to successfully develop a proprietary transgenic mouse model containing both humanized SERPINA1 and humanized ADAR that enables pharmacokinetic and pharmacodynamic assessment of human sequences in vivo. We are on track to share in vivo data from this model in the first half of 2021 and we expect to present additional data at a scientific congress later this year. These in vivo results are expected to enable lead candidate optimization as well as inform potential preclinical development studies and timelines. In summary, 2021 is a year of execution for Wave in a busy time as our next-generation pipeline advances in the clinic. As you heard today, we are advancing three unique clinical programs that will each provide insight into our novel PN chemistry and potentially rapid proof-of-concept and clinical validation of our platform with biomarker data. We're making excellent progress with our ADAR editing capability. In addition to the expected in vivo data update for AATD that I just mentioned, I look forward to speaking further about our RNA editing platform at a Research Day later this year, which we expect to share more details about on our next quarterly call. I will now turn the call over to Kyle Moran, our Chief Financial Officer to report our financial results before turning the call over to questions. Kyle?

Thanks, Paul. We ended the first quarter with $148.5 million in cash and cash equivalents and marketable securities. This balance does not include an additional $30 million in committed research support that we received in early April under our collaboration with Takeda. Our total operating expenses for the first quarter 2021 were $43.4 million as compared to $54.2 million last year. R&D expenses were $33.4 million as compared to $41.2 million in the same period in 2020. This decrease was primarily related to a decrease in external expenses related to our discontinued suvodirsen program as well as decreases in compensation-related and other external expenses, partially offset by increases related to our clinical and preclinical activities for our HD programs and C9orf72 program for ALS and FTD. G&A expenses were $10.1 million for the first quarter of 2021 as compared to $13 million last year with the decrease driven by lower compensation-related and other external expenses. Finally, we continue to expect that our existing cash and cash equivalents, together with our expected and committed cash from our existing collaboration, will enable us to fund our operating and capital expenditure requirements into the second quarter of 2023. As a reminder, this does not include potential milestone payments or other uncommitted payments under our Takeda collaboration.

Thanks, Kyle. With that, we'll open up the call for questions. Operator?

Thank you. Operator instructions were provided. Our first question will come from Salim Syed from Mizuho. You may begin.

Hi. Good morning. This is Michael Lin on for Salim. Thanks so much for taking our questions. A few if possible. First on the C9 trial design, just wondering about the protocol and how adaptive these trials would be, are they being written to be able to enroll many — the hundreds of patients to potentially be converted to registrational? I'll follow-up after that.

Thank you. I'll pass the call over to Mike.

Yes. Hi. So the way the study is designed, as you can see, it allows the study to be expanded as necessary to collect additional information. We made it flexible to enable us to really, once we get recommendations from the independent committee on next steps, to be able to adapt the study as necessary. So, I think we'll have to wait to see what the data show, but it is our intention to make it adaptable and flexible to enable us to direct it the way we need to, to understand the clinical meaningfulness in both ALS and FTD.

So, I can't expand on number of patients based on the committee's assessment then. So, if that's the specific question, yes.

Great. And one on AATD on the upcoming data. What will you be looking for specifically to move forward into the clinic or not? And how would you prioritize this if it did move forward versus the other pipeline programs?

So the prioritization for AATD and the reason we worked on generating the model was really driven on identifying the production of the protein. So again, not working on what percent editing is. That's interesting, but what drives the progress on a medicine is does it generate the protein. So, one will be protein production. Other features that we'll be evaluating obviously are not just protein production, but protein function. So we'll be able to look at a number of those assessments and that's going to guide our decision on translation.

Okay, great. Thank you. And last one on frontotemporal dementia mentioned today. After the donanemab data, how are you thinking about treatment, specifically from the ASO side? Maybe what should we think about in terms of potential targets?

So just for clarification, I want to make sure on — you might be thinking about frontotemporal dementia. I just want to make sure we're not thinking about Alzheimer's. We have FTD, so frontotemporal dementia otherwise. I want to be careful that we didn't discuss Alzheimer's disease as a target therapeutic. But happy to discuss antisense treatment and therapeutics for the treatment of CNS and neurologic diseases more broadly, if that's the question, but I just want to make sure we get your question correct.

Yes, that would be helpful.

I was going to say, yes, as you see with the FOCUS-C9 study, the emphasis on FTD, getting patients in and actually the cortical effects that we saw in the preclinical models make us very excited about being able to access the CNS to be able to have an effect on FTD. We're going to be using some of the clinical outcome measures that you'd want to measure cognitive change in our FTD study to be able to get at that question.

If we just step back and think about neurologic diseases more broadly, one of the compelling data sets that Mike shared in addition to a number of the in vivo mouse studies is we see really good intrathecal distribution across the brain. As we think about distributing to the regions of the brain that are necessary for a whole variety of neurologic diseases, we see intrathecal administration delivering this. It's not different than some of our colleagues in this space discuss around distribution. So we do believe antisense oligonucleotides can distribute. What we see with the PN is this broad distribution and the addition of highly controlled stereochemistry. The data we're generating about the value of stereochemistry is that all the assessments of target engagement we're seeing are with the actual compound, because we're not dealing with a mixture of hundreds of thousands of different molecules that can distribute differently. By having a single drug, we know that the knockdown that we're seeing is based on the design of that single oligonucleotide. We've also seen the benefit of durability. As Mike said, in the adaptive design piece, we're going to be testing and exploring that in the clinic. As we think about chronic administration of medicines across different diseases, dosing frequency is important. Being able not to sacrifice potency for dosing frequency is something that we're excited to continue to explore in the clinic. So as we think about the future of antisense oligonucleotides for the treatment of neurologic diseases we think the future is promising in a data-driven way and we'll be assessing that further across three clinical studies currently. I hope that answers your question. We're always happy to explore that more, but I think the future looks bright for treating genetic diseases with oligonucleotides.

That's super helpful. I guess what I was referencing was just the mention of Alzheimer's in the press release along with other CNS indications?

Okay. Sorry. I think we are talking — now I know where you're going which is the holistic list of targets that we've been exploring in large indications. We mentioned, for example, large neurologic indications such as Parkinson's, Alzheimer's and others. We are focused on the pipeline that we're exploring as opposed to what the potential is. Apologies for our misunderstanding of your question.

Our next question will come from the line of Joon Lee from Truist. You may begin.

Hey, thanks for taking our questions. For all of your CNS programs in addition to the new and improved backbone chemistry, have you considered a different route of delivery? Another company has recently disclosed proceeding with intracerebroventricular route using an intraventricular catheter (Ommaya reservoir), I believe for Huntington's disease. It does give you more direct access compared to intrathecal and I'm not aware of any IP that prevents you from doing that. So, I would appreciate if you could provide some perspective there? And I have a follow-up question.

Yes. There's always opportunities to think about different approaches. When we think about permanent catheter placements and other drilling procedures for delivery, our first question always is: what problem are we trying to solve? As we currently think about the data we've generated to date, intrathecal access to the central nervous system is available. We've shown that in intrathecal nonhuman primate studies. We've looked at intracerebroventricular injections in mice and we see correlation in terms of distribution in CNS tissue. So to date, given the durability we are assessing in the clinic, permanent intracerebroventricular catheters can have risks. They can get clogged and they need to be changed. So it really comes down to what problem we're trying to solve. As we think about intrathecal distribution we've demonstrated across multiple tissues utilizing the PN enhancements on our medicines themselves, we don't right now see that we need an intraventricular catheter. As we look at durability and different indications, we could want to solve other problems in the future. Could come up, but right now that doesn't seem to address the solution that we have with the current administration. Mike, do you want to add anything?

No, I'll just say the same thing. In general you want to go for the simplest form of administration that gives you the access you're looking for. Given the progress we've made with accessing all parts of the brain in these diseases through intrathecal administration, that would be the simplest approach. Especially as Paul indicates, when you have durability of effect after single doses that leads us to the approach that we don't need to do intracerebroventricular administration with a catheter and we can do what's necessary with intrathecal dosing that any neurologist can perform.

We've had the benefit of extensive time to explore multiple animal models across multiple species. ICV in mice and intrathecal in nonhuman primate data, alongside clinical experience with intrathecal distribution, all support our approach. The PN backbone improvements translate into animal PK/PD that we are now exploring in the clinic through adaptive designs that are going to give us answers.

Got it. And looking forward to your ADAR presentation tomorrow at ASGCT. Can you talk about some differences between the approach you're taking with GalNAc-conjugated guide RNA versus circular guide RNA that's being used by another company? I guess we'll find out tomorrow, but just wanted to get any input that you can share on what we should be focusing on tomorrow and some perspectives on the pros and cons of different approaches? Thank you.

I can't speak necessarily for the pros and cons of other approaches. I can speak about our approach. Importantly, GalNAc is a targeting ligand for hepatocytes. One of the advantages when we built the ADAR platform from the beginning is that where short oligonucleotides go and distribute, we can generate an editing capability there. We've done and looked at that in vivo in the CNS, with non-GalNAc-conjugated distribution and correction. For AATD specifically, it's a hepatic target in hepatocytes because GalNAc then becomes an advantage, as we can target the cell type of interest. From a platform perspective, the platform can exist without GalNAc, but we are using GalNAc where we think about the liver specifically. We don't need viral vectors or other delivery vehicles. We take learnings around stereo-controlled, PN-containing oligonucleotides and apply that to the ADAR platform and use GalNAc conjugation for hepatocyte targeting. We'll be presenting more tomorrow and we're excited to share those data.

Great. Thank you.

Our next question will come from Mani Foroohar from SVB Leerink. You may begin.

Hi guys. Thanks for taking the question. I'll ask one quick one on enrollment. When we look at rolling over placebo patients with GENERATION-HD into SELECT, how many of those patients are still eligible? I worry a little bit that, obviously, Huntington is a relentlessly progressive disease. So some proportion of those patients may now be meaningfully more severe than they were when they were first screened for GENERATION. And secondarily whenever — based on whatever portion of those patients are likely in your view to roll over, how much of a head start does that give you on enrollment in SELECT? And are there other mechanisms that you can pursue or changes in trial process to accelerate the completion of enrollment there given there were a couple of delays and hiccups along the way for GENERATION previously?

I'll pass the question to Mike.

Regarding the movement from GENERATION-HD, right now the way it stands is patients don't have been disclosed whether they've been receiving placebo or active treatment. We're hopeful that will happen and then patients can make the choice about whether they want to transition. We're not exactly sure when that will be, but we're hoping that happens relatively soon as disclosure to all the patients about what they've been receiving occurs. There is the possibility some patients will have progressed out of our inclusion criteria, which is why they will have to be rescreened for both inclusion and exclusion, including whether they have SNP3. That will be an important criterion. As we think about the overall population, we'd expect about 40% of the GENERATION population to have the appropriate SNP. Regarding other operational things to help move things along, we've learned a lot from the GENERATION and PRECISION-HD1 and PRECISION-HD2 studies. We've learned how to operationalize intrathecal administration more efficiently. We have a lot of site overlap between GENERATION-HD and PRECISION-HD, which has allowed our physicians to get experienced in the screening process. We have new laboratories to do SNP identification. There's a variety of things that will help accelerate recruitment for SELECT-HD, including the addition of regions with higher representations of SNP3. The adaptive design also allows an independent committee to evaluate unblinded data and guide next steps, which is a big change versus how PRECISION-HD was run.

All right. Thanks.

Our next question will be from the line of Paul Matteis from Stifel. You may begin.

Thanks for taking the question. This is Alex on for Paul. A couple of questions on SNP3. Appreciate the 005 target engagement data in nonhuman primate, but I was wondering if you could maybe quantify a little bit some of the biodistribution that you mentioned you're seeing in nonhuman primates with 003 that you're using for the PK/PD modeling knowing that's not really a disease model, but anything you can say there would be really helpful? And then secondarily, I'm curious if 003 targets the exon one fragment of mHTT. And generally your thoughts on the exon 1 side as a potential driver of pathology in HD? Thanks.

Mike, do you want to start?

Regarding the nonhuman primate data, even though it's not target engagement data for the SNP3 compounds, we are clearly able to achieve concentrations throughout the brain, including in deep gray structures such as the striatum, that would be predicted to engage target based on what we have from the BACHD model. There is a very large window to dose escalate to engage target, and we are quite comfortable based on the preclinical data we have that we are able to get into the regions that matter and engage targets. We are starting at a dose that we believe is pharmacologically active from the beginning and then the independent committee will tell us how close we are to that and allow adjustments as necessary. In terms of exon 1, a lot of the exon 1 data comes from postmortem material and seems most relevant in extremely long expansions. Nonetheless, when we're bringing down mutant protein, we would expect to also have an impact on exon 1.

Antisense oligonucleotides can reach intronic and exonic transcripts, so we should therefore hit exon 1. As Mike alluded to, the most important piece is that the data are still unclear beyond very long repeats. What we've learned in our prior clinical work is that wild-type sparing is an important consideration. We've enhanced PN chemistry for distribution, durability and potency, and we believe mutant protein suppression while preserving wild-type is a compelling approach. We'll generate clinical data with SNP3 using PN chemistry to assess this approach.

Great. Thanks so much.

Our next question will come from the line of Eun Yang from Jefferies. You may begin.

Thank you. I have a few questions. First on the AATD program. You mentioned it's not just editing efficiency, you've already shown 50% in nonhuman primates, but producing wild-type AAT protein and function. Do you know what level or number of wild-type AAT protein is needed in order to see some clinical benefits? What's the kind of a minimum level? Second question is on 004. You are already doing site activations. When should we expect clinical data from this trial? And lastly the additional $30 million you received from Takeda in April, how would that be amortized in the income statement? Would that be similar to what you've done in the past?

I'll start with the AATD question and then pass to Mike and Kyle for the others. When we think about alpha-1 antitrypsin, we're looking at protein levels. The SERPINA1 model allows us to study humanized sequences in vivo. Generally, there's a threshold around 11 micromolar of AAT protein that has been considered clinically relevant based on replacement therapy experience. Those infusion products often require frequent dosing because the protein levels tail off, so sustained correction that keeps patients above a protective threshold could be beneficial for both liver and lung manifestations. The key for us in the in vivo data is not just percent editing but demonstration of production of functional wild-type protein in the animal model. We're taking a deliberate approach to generate the preclinical data that will inform how this program transitions. This program also demonstrates Wave's ability to bring ADAR editing into patients across a variety of indications, including CNS indications where we may not require GalNAc. I'll pass to Mike on your question about timing for 004 data and then Kyle on the accounting.

We have three biomarker studies underway with adaptive designs that will provide a continuous flow of data. Specifically regarding 004, we're going to be dosing soon and we'll have a good sense of how the studies are progressing throughout the year. What we disclose and when will be informed by how the studies proceed, the types of data that come in and whether there's a meaningful number of patients at any given point in time. Feedback from the independent committees will come throughout the course of this year and next and will guide when and how we disclose data. We anticipate data that will enable decision-making and provide greater insights into the chemistry.

The adaptive design principle means the independent committee will evaluate data on an ongoing basis and may trigger sharing at multiple points. That makes precise timing less specific today, but we'll provide guidance as the trials progress.

Eun, thanks for your question. Your assumption is correct. We would account for the $30 million in the same way we've accounted for upfront cash payments and other committed cash payments to date under GAAP — recognizing revenue and amortization as appropriate.

Thank you.

Our next question is from the line of Luca Issi from RBC. You may begin.

Oh, great. Thanks for taking the question. This is Lisa on for Luca here. A couple of questions for us. First one, we have obviously seen the Phase 3 data from Roche and Ionis and it seems to me that there may have been a detrimental impact to patients versus placebo as placebo directionally outperformed the high dose and we have seen a dose-dependent increase in ventricular volume in the brain. Did you have the same read on this data? And if so, do you think that it's possible that a wild-type sparing approach may not have caused a detrimental impact here? Second, you've obviously discontinued SNP1 and SNP2 with the old chemistry. Just wondering what is the plan going forward? Are you planning to explore the PN chemistry for SNP1 and SNP2 or only for SNP3?

Like everybody else, we can only view what has been disclosed publicly. Objectively, what was reported was dose-dependent changes in clinical measurements including cortical and striatal measures. There's a question beyond distribution. We think about two considerations — one is wild-type sparing and the other is the non-selective approach. Studies in normal animals that have shown safety with 50% reductions don't fully capture the human disease context. The reductions being studied in the clinic occur in the setting where patients already have stressors and dysfunction from mutant protein. Removing wild-type protein in that context could be problematic. That's why we believe a wild-type sparing approach is important. Regarding SNP1 and SNP2, we can apply PN chemistry and have generated early data. The measured approach right now is to get data with PN chemistry with SNP3 and use that as guidance before investing additional resources into SNP1 and SNP2, but we are prepared to evaluate the totality of HD allele-specific approaches.

We've been concerned from the beginning that a non-selective approach could have detrimental effects. We've said that before, and it's part of why we're pursuing a wild-type-sparing strategy. There are many reasons that could explain what happened in that Phase III study, but effects on wild-type protein must be considered. We have tools to study this and we're hoping the community asks these questions because it's important.

Great. Thanks for taking my questions.

Thank you. I see no further questions in the queue. I'd like to turn the call back over to Dr. Paul Bolno for any closing remarks.

Thanks everyone for joining the call this morning to review our first quarter 2021 corporate updates. And thank you to our Wave employees for their hard work and commitment to patients. We look forward to speaking to you all again soon. Have a nice day. Thank you.

This concludes today's conference call. Thank you for participating. You may now disconnect.