IDEAYA Biosciences, Inc. (IDYA) Earnings Call Transcript & Summary

Good day, and welcome to the IDEAYA Biosciences IDE397 Preliminary Clinical Data Update Conference Call. Today's conference is being recorded. At this time, I would like to turn the call over to Yujiro Hata, CEO, IDEAYA Biosciences. Please go ahead.

Good morning, everyone. I'll be making some forward-looking statements today. And today, we're absolutely thrilled to provide a program update on IDE397 as well as our Q1 2022 earnings update. Full year end 2021 has been an absolutely transformational year for the company. We've been able to advance 2 programs in the clinic as well as make significant progress across 2 additional late-stage preclinical programs and PARG and Pol Theta. As part of this key advancement across our portfolio, today, we'll be giving an additional update on the IDE397 Phase I dose escalation program. Before we begin, I wanted to thank all of the employees of IDEAYA and our partner, GSK, for its contributions to the IDE397 program, both through our IDEAYA, GSK MAT2A joint development committee as well as the broader joint steering committee. GSK has been a phenomenal partner and invaluable in advancing our thinking on this program from both a monotherapy and combination development perspective. We also thank GSK for its review and approval of the presentation materials we will share today. The focus of today's IDE397 program update will be specifically on the data required to complete the GSK option data package, which includes key preclinical pharmacodynamic data, clinical PK, clinical adverse events and clinical plasma and tumor pharmacodynamic data. At the end of this session, we will have an open Q&A session with our analysts, and we do ask for the portion of IDE397 to keep the scope of the cautions to be within the data categories and what's enclosed in this presentation. And that is the content that has been reviewed and approved by GSK under our partnership agreement. In terms of the agenda for today, I will first start with providing a high-level profile on the GSK option data package requirements and a high-level overview of the collaboration economics. Next, Dr. Matt Maurer, our Head of Clinical Oncology, will provide a walk-through on the Phase I/II clinical development plan, the protocol amendment that has been submitted to the FDA to enable both the monotherapy expansion phase as well as the clinical combination initiation for this program. Next, Matt will provide an overview of the clinical pharmacokinetic data as well as the favorable clinical adverse event profile that we've referred to date. After that, Dr. Michael White, our Chief Scientific Officer, will provide a walk-through of the IDE397 pharmacodynamic data including a very robust data set around the preclinical tumor pharmacodynamic data as well as the clinical plasma and tumor pharmacodynamic data. And then lastly, a PK/PD analysis as well as projected clinical efficacy time the preclinical projections to our clinical both pharmacodynamic preliminary data. And then at the end, I will provide a summary of the presentation today as well as next steps for this program. And then we will transition to the next phase of the presentation. I can give us a summary from our Q1 2022 earnings update, which Paul Stone, our Chief Financial Officer, will walk through. Moving on to the next slide on Slide 3. This is a summary of the GSK option data package, both in terms of timing and scope of the data supporting the GSK opt-in decision. And as noted earlier, today's clinical data update on IDE397 will be focused on these key parameters and data categories that are required under our agreement for the option data package. In terms of preclinical data, the contents include both nonclinical safety, pharmacokinetic data and pharmacodynamic data as well as in vivo efficacy. This data package is now complete and is being uploaded currently to the online data room. Next, the clinical data requirements for the [ MAT2A ] option data package with GSK, the safety and tolerability data, pharmacokinetic data and pharmacodynamic data, both including plasma as well as tumor pharmacodynamic data. Based on where we are in the dose escalation, which is we are currently enrolling cohort 6, we are targeting expansion for the monotherapy expansion phase in midyear 2022. In addition, based on where we are on the assembly of the data package to GSK, we also anticipate that we will deliver the option data package to GSK midyear 2022. If GSK exercises an option, there will be a $50 million option exercise fee due subject to HSR clearance. At GSK elects to opt in, the cost split would modify to an 80%, 20% cost share with GSK responsible for 80% of the costs moving forward and IDEAYA for 20%. Next, there will be a set of milestone payments that could be potentially due, including $465 million in development and regulatory milestones as well as $475 million sales milestones. In addition, we have retained 50-50 U.S. profit splits as well as ex U.S. high single-digit to subteen double-digit percentage royalty. We believe this partnership structure with GSK in the scenario of opt-in, it's highly value accretive for our shareholders and enables a very cost-efficient fashion to advance this program in the clinic. So with that, I will now pass it along to Dr. Matt Maurer, our Vice President, Head of Clinical Oncology to go through the next phase of the presentation. Matt, we'll pass it on to you.

Yes. Thanks, Yujiro. So on this slide, you can see the sketch of the 397 development plan showing a comprehensive approach to proof-of-concept evaluation of both monotherapy and combinations, which were supported by our preclinical evaluations. A parallel mono and combo strategy provides multiple shots on goal and optimizes our probability of uncovering transformational opportunities that can be rapidly brought to registrational follow-up studies. The tumor types of main interest include non-small cell lung cancer, in esophagogastric carcinomas for both [ monotherapy ] testing as well as in combination with taxanes. These choices were based upon our preclinical monotherapy activity as well as the high unmet medical need and large addressable MTAP-deleted populations. It should be noted that MTAP-deleted cancers have a relatively poor prognosis and do not respond well to immunotherapy given that they are generally cold tumors and thus require novel therapies. You can see that for each tumor type we are combining with the standard of care taxane, docetaxel for lung and paclitaxel for esophagogastric. We and others have validated preclinical combinatorial activity with MAT2A inhibition and taxanes and, thus, are excited about the prospect of clinical benefit. In addition, we have monitored every basket cohorts for exploration across other tumor types and an opportunistic evaluation of another combination partner in which we have demonstrated preclinical synergy. We are targeting launch of expansion in combination testing in the middle of this year. Next slide. Shown on this slide are representations of our preclinical pharmacokinetic data. Key points to highlight are the well-behaved steady-state concentrations over time curves in the left panel, and dose proportional increase in exposure shown as fold increases on the right panel. We are most excited about the ability to deliver dosing that has reached our target efficacious exposures based on preclinical data. Dose cohort 5, which is our -- in our target range has been cleared in preliminary tolerability, and we are currently enrolling into cohort 6 where the first patient has cleared the DLT period. You can see the normalized relative to fold increase in exposures across a 20-fold range. IDE397 shows flat and consistent temporal PK profiles across doses with a low Cmax and Cmin ratios. In short, the favorable PK profile supports an acceptable clinical dosing regimen. Next slide. And this slide provides a bit more detail on the safety experience thus far among the patients treated through cohort 5. The panel on the right shows the number of patients that have experienced drug-related adverse events per investigator broken out by grade. The side effect profile so far has been manageable with low-grade GI events in mild thrombocytopenia that have not affected dosing. Specifically, dose-related AEs occurring in greater than 10% patients include low-grade nausea, decreased appetite, diarrhea, dehydration and thrombocytopenia. There has been only 1 drug-related Grade 3 event of asthenia. There have been no drug-related SAEs or dose-living toxicities to date. We believe these data set the drug up for appropriate monotherapy testing as well as a therapeutic window useful for combination testing. This emerging safety profile also compares favorably to prior drugs targeting PRMT5 and MAT2A, which were limited by either hematologic or liver toxicities. So with that, I'll pass it over to Mike to walk through some preclinical data as well as our clinical PD data.

Thank you, Matt. Good morning, everyone. I'll take a few minutes to walk you through our preclinical and clinical pharmacodynamic biomarker data. The key pharmacodynamic markers for the IDE397 program are directly related to the mechanism of action underpinning the MAT2A synthetic lethal interaction with MTAP deletion. As illustrated in the schematic on the left, upon MTAP deletion, which happens in 15% of solid tumors, the cellular MTAP substrate, MTA, accumulates and partially inhibits the essential enzyme, PRMT5, in tumor cells, which is Hit 1 of this paralyzed synthetic lethal opportunity. The MAT2A inhibitor, IDE397, reduces the availability of cellular SAM, which PRMT5 needs to activate splicing factors via symmetric dimethyl arginine modification, SDMA. That's Hit 2. So this one-two punch selectively disrupts essential pre-mRNA splicing in MTAP-deleted tumors. So you can see that in action on the right, no inhibition of SDMA accumulation in MTAP wild-type cells even with high doses of IDE397, and strong suppression of SDMA accumulation in MTAP-deleted cells even at low doses. Notably, as shown on the far right, the disruption of essential splicing in MTAP-deleted cancer cells by IDE397 is equivalent to that observed with direct SAM competitive PRMT5 inhibitors. In contrast, IDE397 has very limited activity in MTAP wild-type cells consistent with the synthetic lethal mechanism. The same can't be said for the PRMT5 inhibitor, which strongly perturbs mRNA splicing in all cells. So given this mechanism of action in patients on trial, we follow drug-dependent MAT2A inhibition by measuring the concentration of SAM in plasma, that's our peripheral PD. And we follow the extent of inhibition of PRMT5 by measuring the abundance of SDMA in cancer cells from tumor biopsies. That's our tumor PD. So on the next slide, please [indiscernible]. In Slide 8, you can see, in preclinical tumor models and in patient biopsies, we measure tumor SDMA by immunohistochemistry with an anti-SDMA antibody. In our workhorse, preclinical MTAP-deleted lung cancer model on the left, we validated dose-dependent reduction of tumor SDMA by IDE397 with complete inhibition of tumor SDMA and tumor growth by 30 milligrams per kilogram. That same dose extinguish tumor SDMA across a subset of MTAP-deleted patient-derived xenograft models, as shown on the right, demonstrating that MAT2A inhibition is sufficient to block PRMT5 activity in MTAP-deleted tumors in vivo. 9 of 12 MTAP-deleted models evaluated had 80% to 100% reduction of tumor SDMA as compared to time-matched vehicle controls. We'll give you 2 [indiscernible] from one of our key indications of interest on the next couple of slides. On Slide 9, non-small cell lung cancer is a patient population of high interest for the IDE397 program. And in this MTAP-deleted lung squamous cell carcinoma PDX model, we achieved 92% tumor growth inhibition by end of study. With respect to the PD, the vehicle arm, middle panel, showed extensive SDMA staining with an H-score of 85. The tumor SDMA H-score was reduced to 0 by IDE397 treatment, as shown on the right, indicating 100% suppression of the production of SDMA in the tumor. Now on Slide 10. Here is another example in the same disease indication where we saw a tumor regression by end of study. This tumor had an SDMA H-score of 250 in control animals, quite high. The max is 300, as you can see in the middle panel. 95% of that baseline signal is gone in residual IDE397 treated tumors, as you can see on the far right. On Slide 11, as I mentioned earlier, in the clinical Phase I dose escalation study, we're following plasma SAM as a peripheral PD biomarker and tissue SDMA as the tumor PD biomarker. We've observed significant plasma SAM reduction on treatment across cohort 1 through cohort 5 to date. As an example, you can see the deep and consistent post-treatment plasma SAM reduction among the evaluable patients in our recently cleared cohort, cohort 5 on the left, with a mean reduction of 77% after 3 weeks of therapy. On the right, you can see that across all evaluable patients on trial to date, the percent plasma SAM inhibition is associated with increasing drug exposure, indicating exposure-dependent target modulation. On the next slide, we'll move to the tumor PD from the trial. So here on Slide 12, we were very pleased to see exposure associated tumor SDMA reduction in target tumor types. Our first predose and on-treatment paired biopsies came from a pancreas cancer patient from cohort 3, shown on the far left, and displayed little detectable modulation of tumor SDMA on treatment. The cohort 4 pancreas cancer patient, middle panel, showed both nuclear and cytoplasmic SDMA staining in the predose tumor biopsy. Those dark brown ovals are SDMA-positive nuclei, which are surrounded by lighter brown cytoplasmic SDMA staining in the pretreatment biopsy shown in the top panel. This patient's AUC plasma drug exposure is about 2.5-fold higher than the pancreas cancer patient in cohort 3. Notably, the cytoplasmic staining is absent on treatment with also a detectable, although modest reduction in nuclear SDMA staining, middle bottom panel, indicating an emerging pharmacological response to IDE397 in the tumor. Finally, on the right panel, at the higher IDE397 exposures, you can see almost complete loss of SDMA staining in the post-treatment biopsy from this lung cancer patient in cohort 5. The plasma drug exposure in this patient is almost sixfold higher than what we saw in cohort 3 patients. We'll go in a deeper dive on the next slide. So Slide 13 first. It's worth noting that the steady state exposures or patients in cohort 5 are getting within the predicted efficacious concentration range based on preclinical in vivo efficacy studies using a representative MTAP-deleted lung cancer model. This patient was on the high side of exposures observed in cohort 5, and the plasma SAM reduction is also on the high end at 79%. The central lab pathologists tumor SDMA score for the pre-dose biopsy was 280. As shown in the bar graph on the right, the vast majority of tumor cells in the predose biopsy had the maximum SDMA-staining intensity receiving scores of 3 plus. In contrast, the H-score for the on-treatment biopsy was only 13 and, in the bar graph shown on the right, the vast majority of tumor cells in this sample were completely negative for SDMA receiving scores of 0. I'll sum up on the next slide. So Slide 14, preclinically. We see strong tumor growth control as well as tumor regressions in multiple MTAP-deleted CDX and PDX tumor models using once-daily oral IDE397 doses from 10 milligrams per kilogram and above. Significant reduction of tumor SDMA is detectable at 3 milligrams per kilogram and is maximally suppressed by 30 milligrams per kilogram. We've plotted those exposure response details for the MTAP lung cancer model H838 on the right. Clinically, IDE397 achieved plasma SAM reduction of 60% or more across cohort 1 through 5 in the Phase I. We've plotted the steady-state exposures for the patients discussed in this presentation on the right, benchmarked against the preclinical observations. Clinical tumor SDMA is showing an exposure-dependent relationship with 95% reduction of tumor SDMA in cohort 5. So based on these observations, cohort 5 is in the target exposure range for efficacy, and we are currently exploring a higher dose in cohort 6. So with that, I'll hand the presentation back to Yujiro.

Great, Mike. Thank you so much for that walk-through, and thank you, Matt, for the walk-through of the clinical data as well. In terms of key next milestones for IDE397. First is to deliver the GSK option package midyear 2022. And then second, to initiate both the monotherapy expansion and combination studies in midyear 2022. And then as noted earlier, the protocol amendment to the FDA has been submitted to enable this next phase of the clinical program. In terms of a summary of the presentation today, we believe we've demonstrated IDE397 shows robust suppression of tumor pharmacodynamics across multiple MTAP-deleted PDX models, including 100% reduction of tumor SDMA. We have completed the preclinical package for the option data package and the data has been uploaded to the option data package data room. Next, IDE397 has demonstrated a favorable human PK profile that we believe supports once-daily dosing as noted on clinicaltrials.gov. IDE397 has demonstrated robust clinical plasma pharmacodynamics across all cohorts and exposure-dependent tumor pharmacodynamics. Today, we've also observed a very favorable human AE profile, including no serious adverse events, and the maximum tolerated dose has not yet been reached. Importantly, the summary of this data, we believe we are seeing an emerging preliminary risk benefit profile versus historical. On the benefit side, as noted earlier, we have seen clinical activity, including robust plasma and tumor pharmacodynamics and, as reported earlier, tumor shrinkage in multiple patients. From a risk profile side, we believe we have observed a favorable AE profile, including, again, no SAEs and MTD that has not yet been achieved. To contextualize this risk/benefit profile, we wanted to highlight 2 additional molecules in this pathway, including GSK's PRMT5 inhibitor, who had an expansion dose selected at 400 mg once a day. And as noted here at that expansion dose, you saw approximately 26% SAEs and 53% Grade 3/4 AEs. For AG-270, of the 200 mg QD expansion dose, they observed just under 20% Grade 3 or higher AEs and at the next dose level of 200 mg BID, a 67% Grade 3 or higher AE. At the outset of this program, IDEAYA's belief has been generating a molecule that has sufficient exposure, that can deliver robust pharmacodynamic modulation with an AE profile that is superior from previous compounds would provide a unique profile to interrogate this opportunity within MTAP deletion. We believe this interim clinical data positions the molecule well to enter into the next phase of development. And we believe this favorable risk/benefit profile will position IDE397 to potentially evaluate unique combinations as well as to explore potentially earlier lines of treatment as monotherapy agent. With that, this is the conclusion of the prepared remarks, so the IDE397 Phase I clinical program update. And with that, I will pass it along to Paul Stone, our Chief Financial Officer, to provide an update across the portfolio as well as our financials as it relates to our Q1 2022 earnings release update.

Good morning, everyone. I'm Paul Stone, Chief Financial Officer of IDEAYA. I'm going to report on our financial results for full year 2021 as well as a corporate update. In 2021, for the full year, we had operating expenses of $78 million. This lumped us with a cash balance of $368 million, which provides us with the runway for planned operations into 2025. From a program perspective, our darovasertib Phase II clinical and our preclinical programs are each maturing. Starting with darovasertib, we are currently enrolling patients in a Phase II clinical evaluation of darovasertib in combination with crizotinib in collaboration with Pfizer under clinical collaboration and supply arrangement. We are evaluating that combination in both metastatic uveal melanoma as well as GNAQ/11 mutated skin melanoma. We also opened recently a monotherapy evaluation of darovasertib in primary uveal melanoma. And we are considering and evaluating expansion opportunities preclinically, both with KRAS inhibitors and KRAS-driven tumors, and in combination with crizotinib and potentially other cMET inhibitors in cMET-driven tumors. In terms of key updates and catalysts, we announced yesterday that we've expanded our relationship with Pfizer on our clinical collaboration and supply arrangement to support a potential registrational trial in metastatic uveal melanoma as well as to support a Phase I clinical evaluation of the combination in cMET-driven tumors subject to preclinical validation. Finally, for this program, we are targeting FDA guidance as well as a clinical data update in midyear 2022. Our next [ development ] advance program is IDE161, which is a wholly owned program targeting PARP for patients that have HRD or BRCA mutations. In this program, we've observed in vivo efficacy with enhanced tumor growth inhibition and/or tumor regressions in multiple niraparib-resistant models, which is a PARP inhibitor. We are targeting IND in Q4 of this year. In our Pol Theta program, also partnered with GSK, we've demonstrated in vivo efficacy in combination with niraparib, which is GSK's PARP inhibitor, and we are in IND-enabled studies first half of this year. Importantly, for this program with GSK, we have the potential for up to $20 million in aggregate additional milestones from preclinical through early Phase I. Finally, on Werner Helicase, we are targeting a development candidate with GSK in collaboration with them in 2023. And similarly for this program, we have the potential for up to $20 million in aggregate milestone payments from preclinical to early Phase I. We're excited to see these programs mature.

Great. And Paul, anything additional for the financial results or is that the conclusion?

I think we covered the financial results. As I said, $78 million of operating expenses for 2021. Cash into 2025 with $368 million on the balance sheet for this year.

Okay. Great. Thank you, Paul, for that summary. And this will be the end of our prepared remarks. And with that, the operator will open it up for the analyst Q&A session.

[Operator Instructions] We will take our first question from Anupam Rama with JPMorgan.

Just a logistical question for me. Can you remind us, one, sort of the data upload is complete and, two, GSK by midyear, how long GSK has to make a formal opt-in decision? And will you be disclosing when the package is delivered or just ultimately the GSK decision?

Yes. Yes, Anupam, so the timeframe for GSK's review period has not been disclosed. But I would say that we've been in very close communication with GSK. Towards the end of last year, based on that communication, we've begun building the data room early. And a lot of the data room has already been occupied, including preclinical data as well as several key pieces of the clinical data. So our view is that, one, the dose is selected for the expansion phase of this trial. It will be an incremental review, which we would hope would enable an efficient -- a time-efficient decision by GSK on the opt-in. In terms of would we provide guidance on when the option package is delivered relative to the decision by GSK? We have not made a decision on that, Anupam, at this point. But I think our hope is that those 2 events may be in fairly close proximity. And based on that, I think it may not be a bad baseline assumption that we would make an announcement as part of an actual GSK decision on the option package.

And our next question comes from Maury Raycroft with Jefferies.

I was just wondering if you can talk more about -- what exactly is going to be submitted to GSK. It sounds like it's subject to initiation of expansion cohorts or MTD, and you're starting dosing with dose level 6. And so I just wanted to see if you can talk more about what else could still impact or be included in the GSK data package. And will you include any data from cohort 6 in the package?

Yes. So I think in terms of data categories that will be provided in the option package that's noted in the schedule in our agreement was as outlined in one of the earlier slides of the presentation. In terms of cohort, Maury, here, I would say it depends on how cohorts goes. So we are currently in that cohort, as I believe, as noted in the print materials. We have cleared the first patient on the DLT window. My understanding is that second patient will likely clear sometime next week if they get through next week. And so it depends on -- if 6 is a possible dose for expansion. If it's up to 5 and we end up selecting 5 in expansion dose, I would say the data that what was presented today will be largely the [indiscernible]. If it's a scenario that we decide to escalate into cohort 7, and we clear that cohort, obviously that cohort will be included as well. I would say at this point, the baseline working assumption that we are internally utilizing for planning for this next phase of development is that the potential to expand this current cohort we're rolling, which is cohort 6.

Got it. That's very helpful. And also, I just wanted to check to see if you're seeing anything additional on impact on bilirubin. And you've mentioned the UGT1A1 biomarker in the past. I'm just wondering if you can remind me if you've assessed that in the clinic and if you're seeing any differences versus AG-270 on those measures?

Sure. Matt, do you want to answer that in terms of liver safety profile to date versus AG-270 and our clinical experience there?

Yes. So we don't have the liability on the liver enzyme. So -- and we have not seen any significant liver toxicity to date. So we're encouraged by that.

And we will take our next question from Charles Zhu with Guggenheim Securities.

My first one, I want to ask on the tumor SDMA down modulation signals that you've disclosed so far. I guess in addition to potentially looking at it from the perspective of percent tumor SDMA reduction, how should we also think about these tumor SDMA levels not only as a percent reduction, but also on an absolute level? For example, the cohort 4 pancreatic cancer patients, you have showed some level of knockdown, but post treatment, it looks like the absolute amount could still be perhaps comparable to what some other tumors maybe on the pretreatment perspective. So I guess could you provide some thoughts on that front as well, not only on the percent knockdown, but also on absolute amounts post treatment?

Mike, do you want to take that?

Yes. Absolutely. Thank you for that question. With respect to the absolute SDMA signal, when we look at that signal, we are really looking at the proteins that have been modified by that symmetric dimethyl arginine. And our therapeutic mechanism of action requires that dimethylation to occur in order to support the pre-mRNA splicing that will allow the tumor to thrive. So in terms of absolute concentrations of SDMA in the tumor, preclinically, we find that the better you suppress those absolute values of SDMA, the better response you have with respect to tumor shrinkage. So in terms of what the absolute values are in the clinical arena and whether or not there's going to be indication-specific thresholds for efficacy remains to be seen. But from the preclinical studies, we do know that the more we can suppress tumor SDMA, the better efficacy we see in those preclinical models, both the CDX and the PDX.

Got it. Understood. And I guess as well, I'm also curious then what are the potential implications on a pretreatment H-score impact? And would we -- could we still potentially expect to see efficacy in a low baseline H-score tumors?

You want me to take that, Yujiro?

Sure. Yes.

From what we have seen on the preclinical side, when we have a response to the MAT2A inhibitor, we have a suppression of the tumor SDMA. We have found tumors that have H-scores of 80 to 100 that grow well and become suppressed, and we have tumor control with the MAT2A inhibitor treatment, I showed you an example of that, with one of those lung cancer, squamous cell lung cancer samples. We also have tumors that have a H-scores close to 300, which is the maximum H-score that you can have, and we find that we can suppress that all the way down to single or low double digits and that corresponds to efficacy. So the absolute concentration of SDMA in any given sample, we think, is going to be variable. It's going to be based on a snapshot in time with respect to the turnover rate of the proteins that are methylated as well as the activity of the enzymes involved is going to be related to the speed of the growth of the tumor. But we do find that we see efficacy irrespective of the starting point of the SDMA signal as long as we are able to suppress that with sufficient exposure to IDE397.

Got it. And maybe just one last one, if I could squeeze it in. I may have missed it, but what's the potential significance of observing what appears to be 100% knockdown of cytoplasmic SDMA while also knocking down 12% of the nuclear SDMA in that cohort 4 patient?

Mike, you want to take that again?

Sure. I'll take that one, too. What we're seeing in that patient is signs of activity of IDE397 in the tumor that is indicated by a reduction in the signal of the cytosol and a little bit in the nucleus. We don't understand the mechanistic relationship there with respect to whether a tumor has a cytosolic signal or not. So what we can say at this point is in that cohort 4 patient, we are seeing emerging signs of productive pharmacology vis-a-vis the drug is getting to the site of action for the intended therapeutic mechanism, is engaging the target and is having a consequence on the pharmacology. That's the extent of the conclusion that we wish to draw.

our next question comes from Joel Beatty with RW Baird.