Morning and welcome to Nuvalent Conference Call. At this time, all participants are in a listen-only mode. There will be a question-and-answer session after the prepared remarks. Please be advised that this call is being recorded at the company's request. I would now like to turn the call over to Alex Balcom, CFO of Nuvalent. You may begin.
Thank you all for joining us today. Earlier this morning, we issued a press release announcing pivotal data from our ongoing ARROS-1 trial of Zidesamtinib for patients with ROS1-positive non-small cell lung cancer. The press release and the slides that we will be using during today's call are available in the investor section of our website at nuvalent.com. On the call with me today is our CEO, Jim Porter, our Chief Medical Officer, Dr. Christopher Turner, and Chief Development Officer, Darlene Noci, who will join for the Q&A session. During the call, we will make forward-looking statements related to our current expectations and plans, which are subject to risks and uncertainties. We will also make certain forward-looking statements about the potential attributes and benefits of our product candidate and the format and timing of our development activities and clinical trials.
Actual results may differ materially due to various important factors, including those described in the risk factors section of our public filings, including our Form 10-Q filed in May. These statements represent our views as of this call only and should not be relied upon as representing our views as of any date in the future. We undertake no obligation to publicly update any forward-looking statements. Let me now turn the call over to Jim.
Thanks, Alex, and thank you all for joining us today. It's my honor to be speaking with you today as we report our first pivotal data set as a company. Nuvalent was founded with a bold mission to discover, develop, and deliver a portfolio of precisely targeted therapies for patients with cancer. The foundation of the company is anchored on a deep expertise in chemistry and structure-based drug design. Through innovative chemistry, we are able to take on tough chemistry problems to unlock new solutions for patients with cancer. Our approach focuses on clinically proven kinase targets, which we believe can accelerate both the discovery and development phases of making new medicines. Today, there are over 90 FDA-approved kinase inhibitors, and over 70 of them are for patients with cancer.
In order to understand where we can make a difference in this space, we established relationships with leading physician scientists from the outset of building the company. These are the experts who developed earlier generation kinase inhibitors, and we learned from their perspectives on the limitations of existing kinase inhibitor therapies and the medical needs of the patients they treat. These early conversations shaped the trajectory of the company and have served as a driving force behind the rationale for each of our programs as we work toward our ultimate goal of maximizing the potential for patient impact. Turning to slide four, our programs are grounded in a central premise. Imagine a future where patients and their providers do not have to choose. In our early conversations with leading physician scientists, it became clear that limitations of currently available targeted therapies lead to difficult trade-offs in treatment decisions.
We learned that these limitations include the emergence of on-target resistance mutations, inadequate CNS penetrance to control brain metastases, and off-target adverse events that can be treatment-limiting. With our expertise in chemistry and structure-based drug design, we knew that each of these limitations had the potential to be addressed through intentional, elective design, and innovative chemistry. With this, the founding hypothesis of Nuvalent was clear. By overcoming limitations of currently available therapies, we believe we could go beyond the trade-offs and deliver a best-in-class molecule that had the potential to drive deeper, more durable responses for patients with cancer, move up the treatment paradigm, and maximize the potential for patient impact. In these early conversations, physician scientists also helped us identify ROS1 as not only a space with a clear medical need, but one that exemplified the opportunity to disrupt a treatment paradigm with a potential best-in-class profile.
The current landscape is summarized on slide five. Today, there are four FDA-approved therapies for patients with ROS1-positive non-small cell lung cancer. However, each requires physicians to make trade-offs in their treatment decisions. Crizotinib was the first TKI approved for ROS1-positive non-small cell lung cancer and remains the frontline standard of care. Its key limitation is that it is not highly brain penetrating, and up to 55% of patients will experience progression with CNS disease. Secondary, treatment-emergent ROS1 resistance mutations have also been identified as a key driver of disease progression after crizotinib, with approximately 40% of patients progressing with the G2032R ROS1 mutation. The other FDA-approved therapies, entrectinib, repotrectinib, and taletrectinib, were developed as brain penetrant inhibitors to address crizotinib's key limitations.
However, all are dual TRK/ ROS1 inhibitors, and inhibition of TRK in the brain has been associated with a broad spectrum of CNS toxicities that can be treatment-limiting. To date, no therapy has managed to supplant crizotinib as a frontline standard of care. Our solution to these limitations is zidesamtinib. As outlined on slide six, zidesamtinib is the first and only ROS1 selective compound designed to address the combined medical needs of hitting the ROS1 oncogenic driver, addressing or preventing ROS1 resistance mutations, having excellent brain penetrance, and avoiding TRK-related neurotoxicities. To evaluate each area of its target product profile, as well as its ability to move up the treatment paradigm, we are investigating zidesamtinib in our ARROS1 clinical trial, which was designed as a global open-label single-arm phase I-II study.
As detailed on slide seven, the phase I dose escalation portion of our trial enrolled heavily TKI pretreated patients, many of whom had exhausted all available therapies. It was designed to determine the recommended phase II dose. Preliminary data from the phase I portion of the trial was first presented at the EORTC-NCI-AACR Symposium in 2022, with updated data presented at the ESMO meeting in 2024. These data demonstrated preliminary clinical proof of concept supporting each component of our target product profile in a heavily pretreated patient population. Importantly, these preliminary data also supported the potential for zidesamtinib to improve outcomes when used in earlier lines of therapy and the inclusion of cohorts in the phase II portion of the trial for patients who are less heavily pretreated, as well as those who were TKI- naive.
Enrollment in the global open-label multicohort phase II portion of ARROS1 began in September 2023, with a recommended phase II dose of 100 milligrams once daily. The trial was primarily designed to support registration in both the TKI pretreated and TKI- naive advanced ROS1-positive non-small cell lung cancer populations. An exploratory cohort for other ROS1-positive solid tumors was also included. Following discussion at a recent pre-NDA meeting with the FDA, we have aligned on an initial registration path for TKI pretreated patients with advanced ROS1-positive non-small cell lung cancer. The patient populations for the pivotal data set are detailed on slide eight. The pivotal safety population includes 432 patients with ROS1-positive non-small cell lung cancer treated at the recommended phase II dose by our data cutoff date of March 21st, 2025. The primary efficacy analysis population includes 117 patients with measurable disease by blinded independent central review, or BICR.
These patients must have received zidesamtinib at the recommended phase II dose by May 31st, 2024, to allow for at least six months of duration of response or DOR follow-up for nearly all responders by the data cutoff date. We have been encouraged by the support from investigators and patients for our ARROS1 trial, as evidenced by the robust enrollment beyond the patients included in our primary efficacy analysis population. This enthusiasm has been incredibly humbling and reinforces to us that a clear need remains for patients with ROS1-positive non-small cell lung cancer. With that, let's dive into our top-line data results. ARROS1 is a multi-regional trial with phase I and phase II patients enrolled across North America, Europe, Asia, and Australia. Key characteristics of the primary analysis population are described on slide nine. Overall, patients had a median of two prior lines of therapy, and 53% had received prior chemotherapy.
50% of patients had received two or more prior ROS1 TKIs, of which 93% had received prior lorlatinib, repotrectinib, or taletrectinib. Notably, the heavy exposure to prior chemotherapy and ROS1 TKIs beyond crizotinib and entrectinib in the global patient population continues to differentiate ARROS1 from other ROS1 studies in TKI pretreated patients. In addition to evaluation of activity across patients who were TKI pretreated, we evaluated activity in subsets with key drivers of disease progression. 36% of patients had a secondary ROS1 resistance mutation, and 49% had active CNS disease, including cases of disease progression following treatment with the brain-penetrant TKIs entrectinib, lorlatinib, repotrectinib, and/or taletrectinib. The top-line efficacy results observed in the overall TKI pretreated population are detailed on slide 10. Overall response rate, or ORR, and duration of response were evaluated by blinded independent central review.
In 117 patients, the ORR was 44%, and the estimated probability of continued response, or DOR rate, was 78% at 12 months and 62% at 18 months by Kaplan-Meier estimates. The median DOR is continuing to mature, and we are encouraged by the potential for durable responses even in this more heavily pretreated population. This overall TKI pretreated population contains a few notable subsets that differentiate it from the population studied by the available ROS1 TKIs. Fifty-eight patients had received two or more prior ROS1 TKIs. In this subset, the ORR was 38%. The overall patient population also included 17 patients treated with prior repotrectinib, where the ORR was 47%, and seven patients previously treated with taletrectinib, where the ORR was 43%. There are no approved therapies for these populations, and to our knowledge, no comparable data sets are available.
It is our continued belief that any activity in these heavily pretreated patients who have likely exhausted all available therapies indicates the potential for differentiation and improved activity in earlier lines of therapy. This belief is supported by the durability responses observed with zidesamtinib in the second-line setting, as outlined on slide 11. Fifty-five patients received zidesamtinib following treatment with the most commonly used front-line ROS1 TKIs, crizotinib and entrectinib, with or without prior chemotherapy. This represents a realistic second-line population. In this subset, an ORR of 51% was observed. Notably, the DOR rates at six, 12, and 18 months were all 93%. We are highly encouraged by this potential durability in the second line, particularly when considering that crizotinib's median duration of response is approximately 18 months in the front-line setting.
We note that this population is differentiated from those studied by the available ROS1 TKIs based on the nearly even split between patients receiving crizotinib versus entrectinib and the history of prior chemotherapy in nearly half of the patients. The subset of 28 patients who received prior crizotinib only, with or without chemotherapy, is most similar to the TKI pretreated populations reported by repotrectinib and taletrectinib. Here, an ORR of 68% was observed with no progression events among responders. Overall, these findings support zidesamtinib's potential to drive durable responses in the second-line setting, where there is currently no standard of care in the continued investigation of zidesamtinib in TKI- naive patients. Our belief in the potential for the improved activity of zidesamtinib in earlier lines of therapy is further supported by its demonstrated activity against key drivers of disease progression.
Slide 12 summarizes the activity of zidesamtinib against the most common ROS1 resistance mutation, G2032R. This G2032R mutation is present in approximately 40% of patients progressing on crizotinib and has also been reported following treatment with other ROS1 TKIs. In 26 evaluable patients with a ROS1 G2032R mutation, the ORR was 54%, and the DOR rate at 12 months was 60%. This included responses in patients who had received two or more prior ROS1 TKIs, including lorlatinib or repotrectinib, which was designed with the goal of addressing G2032R mutations. To better characterize the specific activity of zidesamtinib against the G2032R mutation, we looked at the subset of patients who had received crizotinib or entrectinib as their only ROS1 TKI and were therefore naive to a TKI designed with activity against the G2032R mutation.
In this subset of six patients, the ORR was 83%, and the DOR rate at 12 months was 80%. Overall, these data support the potential for zidesamtinib to provide an option for patients who may have progressed in earlier lines of therapy with a ROS1 mutation. In addition, we believe that using a TKI with broad coverage of ROS1 resistance mutations in earlier lines of therapy has the potential to extend the durability response by preventing the emergence of on-target resistance. Slide 13 summarizes the activity of zidesamtinib against CNS metastases, another key driver of disease progression. In 56 intracranial response evaluable patients that had measurable CNS lesions by blinded independent central review at baseline, the intracranial ORR was 48%, including 11 intracranial complete responses, or CRs, for a CR rate of 20%. The intracranial DOR rate at 12 months was 71%.
CNS responses were also observed in patients who had received one or more prior brain-penetrant TKIs, including entrectinib, lorlatinib, repotrectinib, or taletrectinib, where any additional activity is encouraging. To better characterize the specific activity of zidesamtinib against CNS metastases, we looked at a subset of patients who had received crizotinib as their only prior ROS1 TKI and were therefore naive to a highly brain-penetrant TKI. In this subset of 13 patients, the intracranial ORR was 85%, including seven intracranial CRs, for a CR rate of 54%. There was only one progression event among CNS responders for a DOR rate at 12 months of 91%. Overall, these data support the potential for zidesamtinib to address active CNS disease.
In addition, we believe that using a highly brain-penetrant and CNS-active TKI in earlier lines of therapy has the potential to extend the durability of response by preventing disease progression due to CNS metastases. As the potential for treatment durability improves, so does the importance of long-term tolerability. As shown on slide 14, zidesamtinib was generally well tolerated, and the safety profile remains consistent with prior reports. Treatment-emergent adverse events occurring in 15% or more of patients were reported by preferred and group terms. Among the 432 pooled patients with ROS1-positive non-small cell lung cancer treated at the recommended phase II dose, the most frequent treatment-emergent adverse events were peripheral edema in 36% of patients, constipation in 17% of patients, blood CPK increase and fatigue, both occurring in 16% of patients, and dyspnea in 15% of patients.
Treatment emergent adverse events requiring dose modification were infrequent, with dose reductions occurring in 10% of patients and only 2% of patients discontinuing treatment. We believe these data reflect the highly differentiated nature of zidesamtinib's safety profile from other available ROS1 TKIs and are consistent with its highly ROS1-selective TRK-sparing design. Importantly, these data reinforce our belief in the potential to keep patients on therapeutic doses of zidesamtinib, further providing the potential to drive deep, durable responses. Turning to slide 15, we're incredibly encouraged by these pivotal data for TKI pretreated patients with ROS1-positive non-small cell lung cancer. We believe these data confirm the clinical proof of concept previously reported in our preliminary phase I data sets, including the demonstration of meaningful clinical responses for TKI pretreated patients, clinical activity in patients with key drivers of disease progression, and a generally well-tolerated safety profile consistent with zidesamtinib's ROS1-selective TRK-sparing design.
Importantly, we believe these pivotal data now demonstrate that addressing each of these medical needs has the potential to extend durability of response, particularly when used in earlier lines of therapy, and to avoid the need for treating physicians and their patients to choose between efficacy and tolerability. Our ultimate goal is to bring zidesamtinib to all ROS1-positive patients. Today, we are excited to also share early data from our ongoing TKI-naive cohort that further supports the potential for zidesamtinib to improve outcomes in earlier lines of therapy. These preliminary findings are highlighted on slide 16. As of August 31st, 2024, 35 patients had enrolled in the ongoing TKI-naive cohort of the phase II portion of the ARROS1 trial. For these patients, treatment with zidesamtinib delivered an ORR of 89%, including three CRs for a CR rate of 9%.
Durability of response ranged from 1.9 months-13.9 months, with a 96% DOR rate at 6 months and 12 months. Additionally, an intracranial ORR of 83% was observed in six patients with measurable intracranial lesions, including four CRs for a CR rate of 67%. Intracranial DOR ranged from 4.6 months-11.1 months, with no CNS progression among intracranial responders. Taken together, we believe these data continue to support the potential for zidesamtinib to be a generally well-tolerated therapy that can drive deep, durable responses for patients across the treatment paradigm. We are incredibly pleased with these pivotal data for TKI pretreated patients and preliminary data for TKI- naive patients and have outlined our anticipated upcoming milestones for zidesamtinib on slide 17.
With pivotal data in hand, we expect to initiate the rolling submission of our planned NDA in July 2025, seeking approval of zidesamtinib for TKI pretreated patients with advanced ROS1-positive non-small cell lung cancer. We aim to complete the rolling NDA submission in the third quarter of 2025. Driven by our goal to deliver therapies to patients as quickly as possible, we are actively pursuing opportunities to accelerate development and streamline regulatory review. To that end, we're pleased to share that the FDA has accepted our request to participate in the Real-Time Oncology Review, or RTOR, pilot program. The RTOR program is an initiative of the FDA's Oncology Center of Excellence, which aims to provide a more efficient review process to ensure that safe and effective treatments are available to patients as early as possible while improving review quality and engaging in early, iterative communication with the applicant.
In parallel, we continue to engage with the FDA on the potential for a line-agnostic indication as data from our TKI- naive cohort mature. Enrollment in this cohort is ongoing, with a total of 104 patients enrolled as of June 16, 2025. As shown on slide 18, a total of 539 patients have been enrolled across the phase I and phase II portions of our ARROS1 study as of June 16, 2025. The clear enthusiasm and interest that we have experienced throughout the ARROS1 study starts to paint the picture of a compelling commercial opportunity as we build towards launch readiness.
We brought on our head of commercial early last year and are well underway with our commercial preparedness efforts to support a potential first approval and commercial launch for zidesamtinib in 2026, including building our commercial and field teams, identifying key prescribers and sites of care, and initiating discussions with healthcare decision-makers. We'll look to provide additional details on our commercial plans in the future. Turning to slide 19, while today's call has focused on pivotal data for zidesamtinib, we continue to advance a broader portfolio of potentially best-in-class products with complementary initial indications in non-small cell lung cancer. Later this year, we expect to report pivotal data from the TKI pretreated cohorts of our ALKOVE-1 phase I-II study of Neladalkib for advanced ALK-positive non-small cell lung cancer.
The potential for multiple near-term launches into lung cancer indications presents a unique opportunity to efficiently build commercial capabilities that leverage the clear synergies between our ROS1 and ALK programs. Beyond a potential first approval for TKI pretreated patients, our goal has always been to deliver new treatment options to all patients with ALK-positive non-small cell lung cancer. Our ALKAZAR study is designed to support the potential global registration of Neladalkib for the treatment of patients with TKI- naive ALK-positive non-small cell lung cancer. This phase III randomized controlled trial will enroll approximately 450 patients who will be randomized one-to-one to receive Neladalkib or alectinib, a current standard of care in the frontline setting. Given ALKAZARs global registration design, we've been working to ensure that the protocol and study materials are harmonized across participating countries prior to dosing the first patient.
Today, we're pleased to share that we have advanced clinical startup activities for the ALKAZAR trial. We continue to activate sites and expect to begin enrolling patients early in the second half of 2025. Beyond our parallel lead programs for ROS1 and ALK, we are excited to advance our third program for HER2-altered non-small cell lung cancer, where we are greatly encouraged by the potential for TKIs to transform the treatment paradigm. Our HEROEX-1 phase I trial for NVL-330 is progressing well, and we look forward to providing updates on that trial in the future. In closing, this is a remarkable time in the evolution of our company. Bringing a drug from ideation to pivotal data in just a few short years is a rare opportunity that would not be possible without the tireless dedication of our team to making an impact on the lives of patients.
To the patients, caregivers, and investigators who are participating in our clinical trials, we thank you for all of your contributions. We share in your hope for additional years with loved ones. We look forward to sharing more in the months and quarters ahead as we advance towards our goal of becoming a fully integrated commercial-stage biopharmaceutical company with the ability to discover, develop, and deliver our pipeline of precisely targeted therapies to patients with cancer. Thank you, and we would be happy to take your questions.
Thank you. Ladies and gentlemen, we will now begin the question and answer session. To ask a question, you may press the star followed by the number one on your telephone keypad. If you're using a speakerphone, please pick up your handset before pressing any keys. To withdraw your question, please press the star followed by the number two.
With that, our first question comes from the line of Anupam Rama with JPMorgan. Please go ahead.
Hi guys. Congrats on the data. This is Priyanka on for Anupam. Just a couple of questions. Can you put into context the durability you are seeing in the refractory setting relative to standard of care options? Are there certain common characteristics in the patients with intracranial complete responses that you would have us f ocus on? Thanks.
Sure. First of all, thanks, Priyanka. Yeah, let's take the first one first. The question was around the durability that we're seeing in the context of the available therapies. We're quite encouraged by what we've seen with respect to durability. Remember, the idea here was what the physicians had guided us to do is to try to design a compound that can address those key drivers of treatment progression on standard of care crizotinib.
That is, crizotinib has limited brain penetrance, and there's the emergence of ROS1 resistance mutations. So we optimize zidesamtinib to have excellent brain penetrance and to address a broad spectrum of ROS1 mutations. In addition, what they told us very clearly is that the other drugs that have been developed in this space had limitations in that they often hit off-targets that are dose-limiting, right? And namely being TRK off-target. TRK inhibition in the brain leads to a broad spectrum of neurotoxicities.
He said if we had a single compound that can do each one of those things, hit the original fusion, hit the ROS1 resistance mutations, have activity in the brain, but importantly, be a well-tolerated drug so patients can continuously receive therapy, have limited dose reductions and discontinuations, that would give us the best opportunity to drive deep and durable responses no matter the line of therapy, right? We'll start in the broadest patient population. In that broadest patient population, these are patients that are heavily pretreated. Half the patients received at least two prior ROS1 TKIs. 93% of those patients actually had received other brain-penetrant drugs like repotrectinib, lorlatinib, and taletrectinib. In that population, we're seeing 44% response rate, and the DOR at the 18-month landmark is 62%. Quite encouraging.
I should point out that no other drug has reported data in this patient population. This is a unique patient population compared to any other study. In that dataset, we included both patients that had received prior repotrectinib, where nearly half of the 17 patients had responded, 47% response rate, and taletrectinib. Three of the seven patients there had also responded. Again, a unique dataset, and we're seeing durable responses. If you look at slightly less heavily pretreated patients, patients that received one prior TKI, those are patients that received either half of them with crizotinib, the other half entrectinib, the response rate there was 51%. The DOR at the 18-month landmark really stands out. It's 93% at that 18-month landmark.
Now, looking at the subsets there, the other comparative datasets that have been for drugs that have been approved mostly just enrolled crizotinib-only patients, right? Those studies were largely just patients that received one prior line of therapy, and it was mainly consisting of crizotinib. In our most comparative dataset, that would be the patients that received crizotinib-only. There were 28 of those patients. The response there was 68%, and there were actually no progression events among responders. Taken into context, in those previously treated datasets, we're quite encouraged by what we're seeing. We have also shared today some limited data in the TKI naive setting. High response rates there. The durability is still early, but it's trending in the right direction. At that DOR landmark, we're at 96% at 12 months.
Certainly encouraged by what we're seeing there, but we're going to let that data continue to mature. To your second question was around the CNS activity. The CNS activity was a key component of our target product profile. As a company with a deep expertise in chemistry, we really were inspired by the drug lorlatinib in the ALK setting. We knew that the chemist at Pfizer designed an incredibly brain-penetrant drug. The preclinical data and the clinical data from that compound suggests excellent brain penetrance and deep and durable responses in the intracranial compartment in the ALK non-small cell lung cancer patient population. That was our guiding principle. Could we make a drug as brain-penetrant as lorlatinib? By all of our preclinical characterization, we do indeed have that, right?
What we are quite encouraged to see is that that is playing out with our clinical data. We designed a drug to be highly brain-penetrant, and what we are seeing thus far is quite encouraging. We start, again, in the broader, heavily pretreated patient population. We had 56 patients with measurable CNS lesions at baseline, and 48% of those patients had an intracranial response. 20% of those patients had an intracranial complete response. Keep in mind, in that broader patient population, many of those patients have already received brain-penetrant drugs. We saw intracranial activity in patients previously treated with brain-penetrant drugs like entrectinib, lorlatinib, repotrectinib, and taletrectinib. There is no comparable dataset out there that we are aware of in the ROS1-positive non-small cell lung cancer space.
If you go to less heavily pretreated patients that have received only prior crizotinib, we saw an 85% intracranial response rate, and 54% of those patients had intracranial complete responses. If you go to the naive setting, a smaller sample set here, only six patients that had measurable lesions at baseline. Five of those patients, or 83%, had an intracranial response, and four of six, or 67%, had an intracranial complete response. Across every variable, we believe we are seeing encouraging signs of intracranial activity, and that was what we had hoped to do by the design of a highly brain-penetrant drug like zidesamtinib.
Thank you so much for answering our question.
Thanks, Priyanka.
Your next question comes from the line of Andrew Berens with Leerink Partners. Please go ahead.
Thanks. Congrats on the impressive data.
I think you just touched on a fair amount of this already, but maybe more directly, I'd like to see how you think the data stack up against some of the competition you mentioned. Obviously, there's lots of ways to slice and dice the data in these trials. They seem particularly sensitive to prior exposure with TKIs, especially those active at some of the known resistance mechanisms like G2032R. You could help us put the data in perspective versus some of the other agents like taletrectinib. One other thing stood out. The data you showed suggests patients treated with entrectinib are harder to treat with zidesamtinib downstream. This appears to be maintained even if there's a G2032R mutation, which entrectinib is not active against. Just wondering if you have any ideas.
It may be too early to know, but what could be driving this increased difficulty treating these patients? Thanks.
Sure thing. Thanks, Andy. Thanks for the questions. Remember, we're trying to design a drug for all patients, right? We heard loud and clear from physicians that no drug had ever been purposely designed for ROS1 non-small cell lung cancer patients. The other approved therapies are largely repurposed drugs that hit multiple kinases that also happen to have activity against ROS1. From the outset, we knew exactly what the challenges were for ROS1 lung cancer patients, and we designed a solution for those patients. Now, if we want to be a drug for all patients, we certainly want to be inclusive in our clinical study.
We can't be a drug for everyone and then say, "We're only going to take the healthiest of these patients." We want to make sure that any patient that has ROS1 non-small cell lung cancer is included in the ARROS1 study. That being said, we end up enrolling some difficult-to-treat patients. Half the patients have received at least two prior ROS1 TKIs, and nearly half those patients have received chemotherapy as well. It is a unique dataset compared to what other drugs have reported. It is difficult to do some apples-to-apples comparison. Let's start with some of those, how to look at the data in context of the available therapies. I think first and foremost, what we hear from physicians is the safety profile, right? Other drugs have been developed that address one or more of those issues that I pointed out, either ROS1 mutations or CNS disease.
Now, what all those drugs have in common is that they hit multiple kinase targets, namely the ROS1 TRK dual inhibitors. By inhibition of TRK in the brain, it leads to a broad spectrum of neurotoxicities. That has really been the challenge for the development and the uptake of the other drugs in this space, is the safety profile, right? The physicians have said, "We largely stick with crizotinib as a standard of care because it is difficult to manage these patients and keep them on therapy." Some of the more recently approved drugs hit even more off-targets that lead to additional safety issues. Our challenge here was, could we make a compound that was very selective for ROS1 and ROS1 mutations? By doing so, we would have a drug that patients could stay on therapy.
We believe that's the best chance to drive deep and durable responses in these patients. That's what we're seeing with zidesamtinib. The safety profile is unique here. We have low discontinuation rates due to adverse events, 2%. Low dose reductions due to treatment-emergent adverse events at 2%. We're quite encouraged with this safety profile because it is indicative of a ROS1 selective TRK-sparing design, which is what we set out to do. That is probably the most prominent thing to focus on when understanding the other compounds in this treatment landscape as compared to zidesamtinib. Now, with respect to activity, we are seeing all the hallmarks that we are looking for. We're seeing activity in the brain. I just went through that in the last question of some of the activity we're seeing in the brain, even beyond the other brain-penetrant drugs.
What really stands out there is the high CR rates we're seeing in the brain, which I think is unique compared to other ROS1 drugs. With respect to ROS1 mutations, for patients that just received crizotinib or entrectinib, we're seeing an 83% response rate for patients that have the ROS1 G2032R mutation. That's certainly what we hope to see, is that our drug was designed to be active in that patient population. We're seeing, indeed, activity in that population. As I mentioned before, most of the other comparative datasets are largely enrolling crizotinib-only patients that have only received one prior TKI. Some of those datasets included chemotherapy. Some of those datasets did not. Our one prior patient population, we're quite encouraged what we're seeing.
We're seeing 51% response rate, and we're seeing at the 12-month and 18-month DOR landmark, we're seeing 93% patients still in response at those landmarks. If we go more specifically to that crizotinib-only population, we're seeing a 68% response rate with no progression events among responders. It is difficult to do cross-trial comparisons, but looking at that data, we're quite encouraged of how zidesamtinib stacks up with any option available for ROS1 patients.
Okay. Any idea what's going on in the post-entrectinib setting? Why those patients, even though the drug's not active at G2032R, you're seeing more difficulty treating them, even if you haven't identified G2032R mutation?
Yeah. My apologies, Andy. I forgot the second part. Yeah, on the post-entrectinib, we're seeing activity beyond all the other therapies. We have activity beyond crizotinib, entrectinib, repotrectinib, lorlatinib, taletrectinib.
Regardless of what the prior treatment history is, zidesamtinib has shown activity, activity against mutations, activity in the brain. What I think you're pointing out is that in a small sample set of patients that received prior entrectinib, we had a 33% response rate. It's a little bit lower than what we see with crizotinib-only patients at 68%. Keep in mind that patients that progress on crizotinib can progress for two common ways. One is ROS1 resistance mutations. Another is CNS disease. Our drug was designed to be highly brain-penetrant. It might make sense that you are going to have a slightly higher activity in that patient population. For patients with prior entrectinib, there could be a variety of reasons why those patients might progress.
If they have ROS1-driven disease, meaning they have a ROS1 mutation, there's a good chance that our drug would be active in that patient population. If they have other heterogeneous mechanisms, what I mean by that is bypass mechanisms signaling outside of ROS1, it's going to be harder for a drug like zidesamtinib as a monotherapy to be active in that patient population. With that small sample set, it's difficult to predict how many patients have purely ROS1-driven disease versus other heterogeneous mechanisms. We'll continue to look into that, but I think across the board, we're encouraged by the activity we're seeing beyond all the other therapies.
Okay. Thanks. Congrats on the data again.
Thanks, Andy.
Your next question comes from the line of Divya Rao with TD Cowen. Please go ahead.
Hi, team. Congrats on all the data. This is Divya on for Mark.
We have two questions. One is, we found it really noteworthy that the efficacy in brain mets appears really strong across all subgroups. Any thoughts on why that might be the case, and is that a trend seen with other TKIs? Our second question is, based on FDA summary documents for repotrectinib, they had about 18 months of follow-up prior to filing for first-line approval, which is comparable to the 21 months that taletrectinib had. Would it be fair to assume a similar timeline for zidesamtinib, which would suggest a filing in the first line sometime in the second half of 2026?
Thanks, Divya. Thanks for the questions. We take the brain mets first. As mentioned before, the goal here was to design a drug that's highly brain-penetrant. We were really inspired by the data that lorlatinib has generated in the ALK non-small cell lung cancer space.
If you look through that literature, you'll see that lorlatinib can drive high CR rates in patients with CNS disease, right? That is different than what you see for other ALK TKIs, right? We wanted to do something similar with zidesamtinib. Have excellent brain penetrance to drive deep and durable regressions in patients that have CNS disease, or conversely, to prevent the disease from spreading to the brain, right, if you have excellent brain penetration. We're seeing that with zidesamtinib. We're seeing very high response rates in the brain, even beyond the other brain-penetrant TKIs. We're seeing high CR rates in the brain, particularly in those less heavily pretreated patients, patients that just received crizotinib or in the TKI- naive setting. It is working as we had hoped, as we're seeing that intracranial activity. Other TKIs have reported intracranial activity.
I do not think there's a lot of data available on those high rates of complete responses from some of those other therapies. Now, to your second question, the amount of follow-up, this really speaks to the execution of this program. We've really been humbled by the enthusiasm for the ARROS1 study. We just started the clinical development a couple of years ago on this program. This program's been moving incredibly fast. It is really due to the outstanding execution of the Nuvalent team, the collaboration from the investigators globally on this study, as well as the great support we enjoy from the patient advocacy groups, which help point patients to the ARROS1 study as well as our other studies. We had started the phase II in September of 2023. In September 2023, opening up sites globally, we had very, very expedited enrollment.
In fact, enrollment went so fast that we significantly over-enrolled the study. Now, we knew in the precedent space that other drugs in the ROS1 non-small cell lung cancer space followed TKI- naive patients for at least 12 months post-response, right? We knew that was a guiding principle that we should keep in mind for our TKI- naive dataset. That being said, there is no clear standard of care for previously treated ROS1 patients. We designed our drug to be active there. We had breakthrough designation in that previously treated setting, patients that received at least two prior ROS1 TKIs. We said there's an opportunity here. We can cut the data earlier for previously treated patients. We aligned with the FDA on what that appropriate amount of follow-up was.
As you saw, we had selected 117 patients in that analysis patient population that were enrolled by May of last year. We followed those patients for six months post-response, right? That gave us an earlier snapshot of previously treated patients. We recognize we can let the TKI- naive data continue to mature. The idea here was we have the opportunity to potentially seek approval for previously treated indication. Then subsequently, we'll work with regulators around that TKI- naive data as all patients have had at least 12 months of follow-up post-response. That is how we viewed it. Do not want to guide specifically to when that data cut for TKI- naive patients will come, but we are going to work with regulators to get there as fast as possible.
That is helpful. Thank you so much.
Your next question comes from the line of Peter Lawson with Barclays.
Please go ahead. Great.
Thanks so much. Thanks for taking my question. I wonder if you could just elaborate upon the patients that responded after prior TKIs. Is that linked to CNS activity or overcome resistance mechanisms?
Yeah. Thanks, Peter. I think it's really due to the three main components of our target product profile. It's having activity against ROS1 mutations. It's having excellent brain penetrance to treat patients that progress with CNS disease or prevent disease from spreading to the brain. Just as importantly, it is the safety profile. That is what has held back some of these other drugs. It's difficult to keep patients on therapy continuously if they're constantly experiencing these off-target dose-limiting toxicities. They're either getting dose interruptions or they're getting dose reductions, or in some cases, they're actually discontinuing the therapy.
In those instances, if the patient's not getting continuous therapy, there is a chance that their disease could progress. If you can then treat them with a drug that is well tolerated, that patients can stay on continuous therapy, you have a better opportunity to drive deep and durable responses. I think all three of those things were really key, we believed, to having a drug that can work in any line of therapy, especially in these previously treated patients.
Gotcha. Thank you. On the first-line data, when's the next update we get around that? When do you expect to kind of give an update around FDA discussions around path to approval for first-line?
Let me turn that over to Alex.
Yeah. Thanks for the question, Peter. We plan to present the pivotal data from the ARROS1 trial at a future medical meeting.
We have not provided the specifics yet, but we will look to provide updates in the future.
Would you give additional guidance around your FDA discussions around moving that into frontline?
Yeah. As Jim mentioned, the strategy there, and we will look to provide updates there as well.
Perfect. Okay. Thank you so much. Congrats.
Thanks, Peter.
Thank you.
Question comes from the line of Richard Law with Goldman Sachs. Please go ahead.
Hey, guys. Good morning and congrats from me as well. Is there any comments on the preliminary survival trends that you can share, and will any survival data be included in the submission? Also, another question is, we saw that taletrectinib was just listed as a preferred agent in both 1L and subsequent therapy in NCCN and guideline. Is your expectation that zidesamtinib be included in the guideline by approval?
Is there any possibility that it could be recommended in the first line even before your approval there as a lower category? Thank you.
Thanks, Rich. I think the first question was, do we have progression-free survival data available, and will that be included in any submissions? First of all, any data we generate on the program, of course, will be provided to the FDA and global regulators as part of any submission. Let me just tell you how we think about a single-arm phase two, which we think is consistent with how regulators often view these types of studies. What regulators often look at is in a single-arm study, can you shrink tumors? That is response rate, and for how long? That is duration of response. Progression-free survival is often an endpoint reserved for comparing two different drugs in a randomized fashion, right?
That's why we're not necessarily leaning into the progression-free survival in this updated dataset because often regulators will not view that in their judgment of the, sorry, in the construction of the prescribing information. For most single-arm study approvals that we're aware of in the oncology space, ORR and DOR, as well as safety, seem to be the main data included in most of those approved therapies. To your second point about NCCN, we're certainly going to work with all potential collaborators out there to get this drug listed in the guidelines. It's too early for us to comment specifically on the strategy or timelines there, but we certainly want to get this drug available for all patients as fast as possible, including in the frontline. We will work with all potential collaborators out there to make sure that folks are aware of the attributes of zidesamtinib.
Great. Thank you.
Your next question comes from the line of David Nierengarten with Wedbush. Please go ahead.
Hey, thanks for taking the question. Most of mine have been asked, but I did have one last one on the nature of the patients on the entrectinib versus other treatments and the difference in overall response rate. Was there any difference in chemotherapy use or duration of therapy in the prior setting? Just wondering if patients who are a little bit sicker, if they took longer on entrectinib or had a chemotherapy treatment at a higher rate. Thanks.
Thanks, David. Yeah. Very insightful question. As you point out, in that patient population that received one prior, nearly half the patients have received prior chemotherapy. As I mentioned before, there could be a number of reasons why patients might be progressing on entrectinib.
It could be on-target resistance, which zidesamtinib has been designed to address, or it could be other bypass mechanisms. It could be heterogeneous disease, which a monotherapy is not going to be a good option for those patients regardless of which ROS1 TKI you're using. Difficult for us to dig into that at this point in this top-line update. We'll continue to do so, but nonetheless, we're quite encouraged by we're seeing activity beyond all the other therapies. We would think that points to if the patient does have ROS1-driven disease, and our drug was designed to broadly cover ROS1 resistance mutations, both in the periphery as well as in the CNS, then there's a good chance those patients could respond to zidesamtinib. I think across the board, we've been encouraged by that.
Your next questio n comes from the line of Charles Zhu with LifeSci Capital. Please go ahead.
Hey, good morning, guys. Congratulations on this update and reinforcing your best-in-class profile. I had a couple of questions from me. Regarding the NDA filing package, would you have the opportunity to provide an updated cut, potentially with more patients as well relative to the March 21st cutoff date you are reporting here, or is the present data package that you presented today what you intend on filing with the FDA? Secondly, on commercial and field force readiness, could you also provide a little bit more color around that build-out, sizing, geography, and to what extent you might have to further upsize that as you expand into second-line plus ALK, TKI- naive ROS1, and potentially front-line ALK way down in the future? Thank you.
Thanks, Charles. Yeah. Two questions there.
One is our NDA strategy. Let's take that one first, and I'm going to turn that over to Darlene.
Around TKI- naive. Yeah, sure. So the TKI- naive cohort, as Jim mentioned, was incorporated into the ARROS1 pivotal study design, and these patients started enrolling in the phase two portion of the study. We are looking forward to our first initial NDA submission for the company, and that's to support the pre-treated indication. We do plan to submit all of the data from the pivotal study at that time. We are really encouraged by the preliminary TKI- naive data that we've shared, and we continue to have open dialogue with FDA about the potential for a line-agnostic indication as we believe there remains an unmet medical need for these patients.
We value the relationship that we have with the FDA, and we will continue to work closely as these data mature to enable a line-agnostic pathway.
The second question, Charles, was around our commercial build strategy. Let's have Alex take that.
Thanks for the question, Charles. The goal here is to build a fully integrated company. We brought on our Head of Commercial early last year and have continued to build out the team. Our commercial readiness efforts are well underway to support the first approval and launch for zidesamtinib in 2026, including building out commercial and field teams, identifying key prescribers and sites of care, and initiating discussions with healthcare decision-makers. Specific to the sizing of the sales force, we haven't provided specifics yet. However, I would note there's many synergies to our three development programs in non-small cell lung cancer.
We are also able to leverage learnings from the precedents set by other drug launches in this space. We will look to provide updates on this in the future.
Thanks for taking the questions, and congrats again.
Thanks, Charles.
Your next question comes from the line of Roger Sung with Jefferies. Please go ahead.
Great. My congrats to the dataset. Thanks for taking the question. A couple of quick ones. First one is maybe I missed or maybe because the n is small. Can you comment on the activity beyond the G2032R mutation if you have any dataset?
Another question is just to follow on the previous one, for your NDA submission for the second-plus line ROS1, how the FDA will view the data from the additional patient beyond the primary analysis package, other 117 patients in ongoing pivotal study, understanding you'll over-enroll those patients in this population. Thank you.
Sure. Thanks, Roger. Let's talk first about the G2032R. Remember, that's a key component of our target product profile, to make sure that we have broad coverage of ROS1 mutations. We did share activity in that subset. Across the entire study population, regardless of how many prior ROS1 TKIs the patients received, plus or minus chemotherapy, there were 26 patients that had the ROS1 G2032R resistance mutation. 54% of those patients responded. At that DOR landmark of 12 months, 60% of the patients were still in response.
Now, we saw responses following at least two prior ROS1 TKIs, plus or minus chemotherapy. That also included patients that received prior repotrectinib. Keep in mind, repotrectinib was also designed to address that G2032R mutation. We're seeing activity beyond repotrectinib in patients that have that mutation. Now, in a subset of patients that are less heavily pretreated, meaning they've only received one prior ROS1 TKI, and that could have been crizotinib or entrectinib, there's a smaller sample set, only six patients. In those six patients, five of them responded, an 83% response rate and an 80% DOR at that 12-month landmark. In that actual sample set, only one patient had a progression event among the responders. For us, it was pretty convincing that our drug is active here against that ROS1 resistance mutation.
The second question you had asked was around our NDA submission strategy, and it was particularly around beyond the primary analysis. Let me start, and then I'll turn it over to Darlene. As you might have seen, we had vastly over-enrolled the study. As of May of last year, we thought we had what was the primary analysis patient population, 117 previously treated patients. We went and then followed those patients for at least six months beyond response. That was the pivotal data we're reporting today and will be used to guide our submission with the FDA. What happens to all the additional patients beyond that additional data cut? We wanted to make sure we continued to provide access to zidesamtinib. Clearly, we enjoyed robust enrollment in the study. There's a lot of enthusiasm and momentum on the study.
We wanted to maintain continued access to zidesamtinib. We kept the cohorts open. We kept enrolling. That gives us an opportunity to generate additional data, additional patient and physician experience with zidesamtinib. Those are all good things, right? Regarding our submission, maybe Darlene can talk a little bit more about the mechanics of that.
Yeah, sure. As Jim mentioned, we are presenting the top-line efficacy and safety data, but all data collected as of the NDA data cutoff in March of 2025 will be included in the NDA. FDA will have the opportunity to review all of the data that we've collected as of that point.
Thank you.
Thanks, Roger.
Your next question comes from the line of David Dai with UBS begin?
Great. Thanks for squeezing me in. I also want to add my congratulations on the dataset here.
Two questions for me. First, on the TKI naive front-line data, it seems to be pretty impressive. Let me just help us understand how do you think that 89% OR and then about 96% 12-month DOR could translate to progression-free survival in the long run? Just on the safety side, just to clarify, have you seen any neurotoxicity such as dizziness or dysgeusia that other competitors have shown?
Sure. Thanks, David. We'll start with the TKI- naive front-line data, then we'll go to the safety question. On the TKI- naive data, 35 patients that were enrolled as of August of last year. The reason August is we wanted to make sure we had sufficient follow-up on those patients to really understand what we're learning from a durability perspective. Let's start with the response rate. The response rate was 89%.
I should point out that any ROS1 TKI will have high response rates in the front-line setting. There's nothing unique about our 89% response rate in front-line patients. All drugs have high response rates in the front-line setting. Some may differ in how many patients receive prior chemotherapy, and that can knock the response rate down a little bit. In general, you should expect high response rates because if the patient has ROS1-driven disease, the drug is going to be active. What's important is how durable those responses are. There are certain things that impact that durability. Do you have good coverage of ROS1 resistance mutations? You could prevent the disease progressing through that on-target acquired resistance. Do you have high brain penetrance to prevent disease from spreading to the brain?
If the patients are presenting with CNS disease, making sure you can have activity with those patients that have disease in their brain. Maybe most importantly is the safety profile, right? It's very important if you're going to put a patient on drug for hopefully years to come that the drug is well tolerated, right? If a physician has to manage many different safety challenges in these patients, it becomes a real problem for the physician and more importantly for the patient. We heard loud and clear that was a design goal for our program. Make sure the drug is very selective for ROS1. It avoids off-targets like TRK. It avoids any other off-targets because those off-targets can drive adverse events. They can drive dose-limiting toxicities. If the patient has an interruption or reduction or discontinuation, their disease can progress.
What's really important here is that safety profile because that will impact the durability. Now, we have an early snapshot of the durability. We provided 6-month and 12-month landmarks. It's difficult to make a lot out of that at this point, but it's at 96% at 12 months. That's certainly intriguing, but we'll continue to let that data mature. Other thing that we learned from the TKI- naive data is we do see those high CNS response rates. That's what we had hoped to see. We see five of the six patients respond, and four of those patients actually had an intracranial complete response. We're on the right track there with respect to brain penetrance. Now, your second question was around the safety, and I'm going to turn it over to Chris. Yeah.
The zidesamtinib safety profile is differentiated and has substantially reduced toxicities, as Jim mentioned, including a reduced number of dose reductions, treatment discontinuations from what's currently available. What we show in the safety table is treatment emergent, which is regardless of the causality, of 15% or higher, grouping similar terms together, and none of these terms are TRK-related. Overall, the data continue to show that zidesamtinib is well tolerated and consistent with the ROS1-selective TRK-sparing design.
Great. Thank you so much. Congrats again.
Thanks, David.
Your next question comes from the line of Michael Schmidt with Guggenheim Securities. Please go ahead.
Hey, guys. This is Rosy on from Michael. Congrats on the progress. Just compared to the phase one data, are there any notable differences in the baseline patient characteristics versus the earlier part of the ROS1 trial?
Do you anticipate the pretreatment profiles of patients to perhaps evolve in the real world, just given the recent FDA approval in terms of which ROS1 TKIs are being used? Thank you.
Thanks, Rosy. Yeah, the baseline populations are distinct and unique compared to any other ROS1 study that we're aware of. It's actually only really consistent with our phase one, where we also enrolled a very difficult-to-treat patient population. And our drug was active. It was active in all the variables there, whether it be mutations, CNS disease, and it had that differentiated ROS1 selective TRK-sparing safety profile. We wanted to make sure we made a drug for any ROS1 lung cancer patient. We were widely inclusive in our protocol to make sure that if you had a ROS1 non-small cell lung cancer, you were eligible for the ARROS1 study.
That being said, we ended up enrolling many patients that had taken multiple ROS1 TKIs, many patients that have taken chemotherapy. I think the number was 50% of the patients have received at least two prior ROS1 TKIs, and nearly half the patients have received prior chemotherapy. Of those patients that received two prior ROS1 TKIs, 93% of them had received the brain-penetrant drugs like lorlatinib, repotrectinib, taletrectinib. That dataset does not exist for any other drug that we're aware of in the ROS1 space. What's so special in this update is that we're seeing activity in that difficult-to-treat patient population. Again, that was by design. We hope to design a compound that could be active for any ROS1 lung cancer patient, regardless of the line of treatment. How will that continue to evolve in the real-world setting?
I imagine you might see more use of some of the newer generation ROS1 TKIs as physicians try to understand, can they navigate those safety challenges for those drugs? We already generated a dataset that shows our drug is active beyond any of the other therapies: crizotinib, entrectinib, lorlatinib, repotrectinib, taletrectinib, active in patients that have CNS disease, active in patients with ROS1 mutations. We think we are on the right track with respect to understanding the population here.
Great. Thanks. Congrats on the progress.
Thanks, Rosy.
Your next question comes from the line of Sean Lee with H.C. Wainw right. Please go ahead.
Hey, good morning. This is Sean Lee for RK, and congrats on the results. I just have two quick questions.
First is, can we expect any additional future updates from this current cohort that you're presenting today, as well as what additional steps do you need to take to complete the NDA submission, especially regards on the CMC front? Thanks.
Thanks, Sean. Yeah, I'll take the first one. So any future updates? We're really excited about this update today. It is the pivotal data. The next milestone for us is the NDA submission. As you heard today, we guided to NDA submission, rolling submission starting in July. That leads right into your second question. Let me turn that over to Darlene. What are the next steps regarding our NDA submission strategy?
Yeah, sure. We're really excited about our first initial NDA submission for the company for the pretreated indication. As Jim said, we're initiating the rolling submission starting next month, and we target completion in Q3.
With regard to the TKI- naive potential, as we said earlier, we remain encouraged by the preliminary data that we've shared, and we're continuing to have open dialogue with the FDA about that potential as the data mature. Again, we believe there remains an unmet medical need for these patients. We are going to continue to get zidesamtinib to as many patients as quickly as possible.
Great. Thanks again for taking my questions.
Thanks, Sean.
We have no further questions at this time. I would like to turn it back to Jim Porter for closing remarks.
Great. Thank you all for joining us today. We are proud of the progress we have made across our pipeline and appreciate your continued support and engagement. We look forward to speaking with you in the weeks and months ahead. Thanks.
Thank you, presenters.
Ladies and gentlemen, this concludes today's conference call. Thank you all for joining. You may now disconnect.