Gossamer Bio, Inc. (GOSS) Earnings Call Transcript & Summary

Thank you for standing by. My name is Dina, and I will be your conference operator today. At this time, I would like to welcome everyone to the Gossamer Bio Inc. PROSERA Phase 3 Topline Results Call. [Operator Instructions] It is now my pleasure to turn today's program over to Bryan Giraudo. Please go ahead.

Good morning, everyone, and thank you for joining us today. Earlier this morning, we issued a press release announcing top line results from PROSERA, our Phase III study of seralutinib pulmonary arterial hypertension. The press release is available on the Investors section of our website. Joining me today are Gossamer executives, Faheem Hasnain, Dr. Richard Aranda, Bob Smith, Dr. Rob Roscigno and Caryn Peterson. Before we begin, I'd like to remind you that we'll be making forward-looking statements on today's call. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those expressed or implied. For a discussion of these risks, please refer to our SEC filings and today's press release. We undertake no obligation to update or revise these forward-looking statements. With that, I'll turn the call over to Faheem.

Thanks, Bryan. This morning, we announced top line results from PROSERA, our Phase III registrational study of seralutinib in patients with pulmonary arterial hypertension or PAH. At week 24, seralutinib demonstrated a numerical improvement in 6-minute walk distance of approximately 13.3 meters versus placebo with a p-value of 0.032. While we saw an improvement relative to placebo, the p-value did not meet the prespecified statistical alpha threshold of 0.025, and therefore, the study is not statistically significant. And all other p-values we'll be discussing today are nominal. The overall treatment effect and statistical parameters were materially diluted by an outsized placebo response and meaningful regional heterogeneity, which compressed the pool placebo-adjusted difference. Now consistent with the Phase II TORREY study in the prespecified intermediate and high-risk subgroup, which was an N of 234 patients as defined by REVEAL Lite 2 Risk Score, seralutinib demonstrated a clinically meaningful placebo-adjusted improvement in 6-minute walk distance of 20 meters at week 24. The p-value was equal to 0.0207. Across key secondary endpoints, including time to clinical worsening in NT-proBNP results consistently favored seralutinib. And in that intermediate and high-risk subgroup, 3 of 4 key secondary endpoints also had a p-value below 0.0125, underscoring seralutinib's activity in patients with higher baseline risk. And to be very clear, this is not a novel or post hoc subgroup. This is a well-established clinically relevant patient population. In sotatercept's label expansion study, HYPERION, newly diagnosed patients were required to be intermediate or high risk at baseline as defined by REVEAL Lite 2 or its European cousin to compare a score. And this increased separation between drug and active arms on 6-minute walk distance is a finding that we see across all populations of increased risk in disease, including functional Class III amongst others. And although we won't get into the slides today, we also had compelling results in the subgroup of patients in the study with PAH associated with connective tissue disease. This patient group accounts for roughly 30% of the U.S. patient population and their disease is typically more difficult to treat than other etiologies. That was seen recently in the STELLAR study of sotatercept where PAH associated with connective tissue disease patients were the worst performing subgroup, demonstrating an 8-meter improvement on drug. In contrast, the subgroup with CTD was actually the top performer in PROSERA with a 37-meter placebo-adjusted improvement. Now this makes sense biologically given the anti-inflammatory and antifibrotic components of seralutinib's mechanism of action and the fibrotic nature of connective tissue disease. So certainly more to come here as we dig into the data. Furthermore, in the patients that were on background sotatercept, we also saw a dramatic placebo-adjusted 6-minute walk result, consistent with the theoretical biological synergy between the 2 mechanisms of action, but we must acknowledge that was a very limited single-digit patient set. Also, safety was generally consistent with prior experience. Now given this outcome, we are now focused on 3 priorities. Firstly, fully understanding the PROSERA data set. secondly, engaging with the FDA to discuss the results and their implications and certainly carefully evaluating our strategic options and resource allocation as a company. To ensure we are prioritizing our resources on the Group 1 PAH opportunity and fully understanding the PROSERA data set, we've made the decision to pause the SERANATA Phase III enrollment. Importantly, this decision does not reflect our belief in seralutinib's potential in fibrotic disease. In fact, just the opposite, which PROSERA, of course, reinforced. Given the strength of the signal we observed in patients with connective tissue disease, this especially makes sense. This population has significant overlap with the Group 3 pulmonary hypertension population. Now before I hand it over to Richard to walk through the data in more detail, I want to say a sincere thank you to the patients, caregivers, investigators and study teams who made this program possible. Their contribution to the advancement of PAH treatment is absolutely incredibly meaningful, and it's deeply appreciated. With that, I'll turn it over to Richard.

Thanks, Faheem. PROSERA was a randomized, double-blind, placebo-controlled global Phase III study in patients with WHO Functional Class II and III who were already on background therapy. The study was designed to enroll 175 patients per arm, though we ultimately enrolled 390 patients randomized 1:1 to seralutinib or placebo and treated for up to 48 weeks. The primary endpoint was change from baseline in 6-minute walk distance at week 24 in the intent-to-treat population. Key secondary endpoints included time to clinical worsening, clinical improvement, change in NT-proBNP and reduction in REVEAL Lite 2 Risk Score. The study enrolled a heavily treated prevalent PAH population, including 55% on triple or quadruple therapy and 61% on background prostacyclin. Turning to patient disposition, we randomized and dosed 390 patients in total across the study with 193 assigned to placebo and 197 to seralutinib. There was a slight differential in patients that completed week 24 between the placebo and seralutinib arms with more patients withdrawing prior to week 24 on the seralutinib arm driven by protocol-specified criteria for liver enzyme elevation, which tended to occur early on in the treatment course, consistent with past studies. After week 24, more subjects from the placebo arm dropped out as compared to the seralutinib arm with the leading cause being disease progression. Notably, over the course of the study, no subjects from the seralutinib arm withdrew early due to disease progression as compared to 12 subjects in the placebo arm. Overall, retention across both treatment groups was strong through week 24 and end of study. In the appendix of this deck, we have a set of baseline demographics and disease characteristics. Generally, we enrolled the patient population we sought to enroll and the treatment arms were well balanced. As a reminder, our primary endpoint is the change from baseline in 6-minute walk distance at week 24 in the intent-to-treat population. At week 24, we saw improvement of 28.2 meters in the seralutinib arm versus 13.5 meters in the placebo arm. The placebo-adjusted improvement using the Hodges-Lehmann approach was 13.3 meters with p-value equal to 0.0320. The prespecified 2-sided alpha was 0.025, and as a result, the primary endpoint is considered not statistically significant. The magnitude of the placebo response was unexpected and greater than what has been seen in past PAH studies, which we'll cover in more detail in a couple of slides. In looking at 6-minute walk distance at weeks 12 and 16, the Hodges-Lehmann treatment effect is 9.6 meters at week 12 and 11.9 meters at week 16, respectively. Notably, at week 16, the p-value achieved the prespecified 0.025 threshold. As many of you know, many clinical trials reference a p-value less than 0.05 as a common benchmark. For PROSERA, we set a more stringent threshold of p-value less than 0.025. That reflects the higher level of statistical certainty typically expected in a single registrational study where regulators generally look for strong evidence. Even though we didn't cross the prespecified statistical bar, we did see a measurable separation versus placebo that's consistent with drug activity. A p-value of 0.032 is still below the more commonly cited 0.05 benchmark, which signals a degree of statistical confidence. This slide is a visual representation of the 6-minute walk distance results that were generated in the study through week 48. The lines represent the observed mean changes at week 12, week 24 and week 48. The triangles represent the imputed median values used in the primary analysis at week 24. As you can see from the observed means, those who remained on study through week 48 continue to improve, and there appears to be continued separation of the curves. We need to acknowledge the limited sample size out to week 48, given the rollover into the open-label extension by study design and any early withdrawals. These data out to week 48 are consistent with the longer-term TORREY Phase II data with seralutinib, and they may be in part reflective of seralutinib's mechanism of action as a potential reverse remodeling agent that can lead to increased improvements over time. Here, we put the PROSERA placebo performance into context. As you can see, in PROSERA, the placebo arm showed a larger improvement that is often seen in many other Phase III PH trials where placebo frequently remains near baseline or declined slightly over time. The unusually strong placebo improvement in PROSERA reduced the placebo-adjusted difference and is an important factor in interpreting why a numerically positive effect did not clear the prespecified statistical bar. To further understand our placebo response, we evaluated the placebo response by prespecified geographic region groupings and noted it differed across the regions. In North America, the placebo performance was more aligned with typical modern PH trials and the overall treatment effect was most pronounced with a 25.9-meter placebo-adjusted improvement in 6-minute walk distance. In other regions, particularly Latin America, outsized placebo improvements materially compressed the pool treatment difference. Turning to a summary of our key secondary endpoints. The results consistently favored seralutinib across all key secondary endpoints, including time to clinical worsening, NT-proBNP, reduction in risk score and clinical improvement. NT-proBNP at week 24 showed an estimated location shift of negative 124 nanograms per liter versus placebo with a nominal p-value of 0.002 and separation favoring seralutinib was evident as early as week 4. Time to clinical worsening, clinical improvement and REVEAL Lite 2 Risk Score measures also favored seralutinib. The consistency across functional, biomarker and clinical endpoints strengthens the overall efficacy profile. Now let's focus on the prespecified intermediate and high-risk subgroup as defined by REVEAL Lite 2 Risk Score greater or equal to 6, which was comprised of 234 patients in PROSERA. Here, we see a more pronounced and clinically meaningful effect at week 24. In this subgroup, seralutinib demonstrated a 20-meter placebo-adjusted improvement in 6-minute walk distance at week 24 with a p-value of 0.0207. Importantly, these patients represent a readily identifiable population with considerable unmet need and mortality risk, making the signal particularly compelling from a clinical perspective. In contrast, in the low-risk patients, while patients on drug showed a change from baseline in 6-minute walk distance of 29 meters, patients on placebo showed a change from baseline of 20 meters at week 24, materially compressing the ability of the drug arm to show a noteworthy placebo-adjusted improvement. As before, this slide represents the data in a more visual manner. The line represents the observed mean changes at week 12 and 24 and the triangles represent the imputed median value used in the primary analysis at week 24. Focusing on the intermediate and high-risk group, one can see that seralutinib treatment results in continual improvement over time. The placebo improves modestly at week 12 and then it plateaus. This allows seralutinib to demonstrate a clear difference from placebo. In contrast, in the low-risk subgroup, seralutinib treatment also results in continual improvement over time. However, there is a high placebo response at weeks 12 and 24, which obscures any treatment effect. The data in these 2 subgroups based on risk score shows that the drug effect itself is actually quite consistent across risk groups, but the placebo behavior is very different. So this slide helps reinforce an important point. The PROSERA outcome is less about a lack of drug activity and more about where and how placebo performance influence detectability, particularly in a heavily treated lower-risk population. To help put the PROSERA 6-minute walk distance results into broader clinical context, we thought it would be useful to step back and compare what we observed with seralutinib against other approved add-on therapies in PAH. This slide summarizes reported 6-minute walk distance changes from select pivotal studies, recognizing upfront that cross-trial comparisons have important limitations, including differences in patient populations, background therapies, study designs, endpoints and statistical approaches. With that in mind, there are a few points worth highlighting. First, when we look at the active arm 6-minute walk distance improvements, the magnitude we observed with seralutinib is consistent with and comparable to approved PAH therapies. This is especially relevant given the degree of background therapy in PROSERA, where the majority of patients were already on triple therapy and many were receiving prostacyclins and a few on sotatercept. In that context, we believe the elevated placebo response represents a meaningful headwind when interpreting placebo-adjusted outcomes. Importantly, when we focus on patients with higher baseline disease severity where placebo effects are more muted, the treatment effect with seralutinib becomes clearer and the observed delta in 6-minute walk distance aligns well with therapies that are now standard of care. Concordant with the 6-minute walk distance data, we were pleased to see that in the prespecified intermediate to high-risk subgroup, we see clear evidence for clinically meaningful and statistically significant treatment effects with seralutinib across key secondary endpoints compared to placebo. Starting with NT-proBNP, the magnitude of reduction increases substantially in this population with a decrease of approximately 266 nanograms per liter, which is more than double what we observed in the overall study population. Given the strong relationship between NT-proBNP, right ventricular stress and long-term outcomes in PAH, we view this as particularly meaningful signal and clinically important. We also see this enhanced effect play out in clinical improvement where patients treated with seralutinib were over 3x more likely to improve compared to placebo. Similarly, the likelihood of achieving a greater or equal 1 point reduction in REVEAL Lite 2 Risk Score was approximately twofold higher with seralutinib versus placebo in this subgroup, supporting the potential of seralutinib to address morbidity and mortality risk based on REVEAL Lite 2 Risk Score improvement. Finally, for time to clinical worsening, the hazard ratio again numerically favored seralutinib. This convergence across multiple measures is a key reason we believe the PROSERA results are clinically meaningful and it directly informs our confidence in seralutinib. Next, reviewing our safety. Overall, seralutinib was generally well tolerated in PROSERA. This table is an overall summary of treatment-emergent adverse events, or TEAEs. TEAEs reported in 86.5% of patients receiving seralutinib and 80.5% of patients receiving placebo. Severe AEs were similar between treatment groups. TEAEs leading to discontinuation in AESIs were greater in the seralutinib arm compared to placebo. The majority of these events were related to hepatic adverse events and transaminase elevations, which will be discussed further. Serious adverse events occurred in 16.0% of seralutinib treated patients and 18.9% of placebo treated patients and deaths were similar between the groups. This slide summarizes the incidence of serious adverse events by preferred term that have been reported in greater than 2 subjects receiving seralutinib. Review of the terms did not reveal any unexpected events or new safety signal. The most frequently reported TEAEs occurring in greater or equal to 5% of patients on seralutinib are summarized here. The most frequent adverse event for seralutinib was cough at 37% and 13.7% with placebo. Higher rates of ALT and AST increases were also observed, 14.5% and 14.0% on seralutinib versus 0.5% each on placebo. Other events such as headache, nausea and diarrhea were generally similar between arms. Taken together, these data support that seralutinib is generally well tolerated with a safety profile that is manageable and consistent with prior experience. Based on laboratory analysis, liver transaminase elevations of greater or equal to 3x the upper limit of normal were observed in 13% of patients receiving seralutinib and 1% of patients receiving placebo. These elevations generally occur during the first 3 months of treatment and resolved with drug interruption or discontinuation. In the context of existing treatments, we find that the overall safety profile of seralutinib is favorable as an add-on therapy in the PH patient population. In considering transaminase elevations greater or equal to 3x the upper limit of normal, bosentan, for example, an endothelin receptor antagonist carries baseline and monthly liver tests monitoring under a REMS program, because at the 250 milligram b.i.d. bosentan dose, 14% of patients had ALT or AST elevations greater than 3x the upper limit of normal and 7% had elevations greater than 8x the upper limit of normal. Clinicians are well accustomed with routine monitoring when using therapies with potential hepatic effects given that ERAs are part of frontline combination therapy in PAH. Overall, seralutinib was generally well tolerated and the safety profile is broadly consistent with prior experience. Now I'll hand it back to Faheem to discuss next steps.

Yes. Thanks, Richard. All right. Let me summarize the key takeaways from the PROSERA study. Those study just missed its primary endpoint, we firmly believe the results show clear evidence of efficacy, especially in a patient population that has already been heavily treated. Importantly, PROSERA confirmed what we saw in the TORREY Phase II trial. Patients with more severe baseline disease experienced a greater distinction between seralutinib and placebo. This enhanced separation highlights the potential benefit seralutinib can offer for those with advanced forms of disease. Now when it comes to safety and tolerability, seralutinib appears to be acceptable as an add-on therapy in pulmonary arterial hypertension. The main safety findings such as cough and liver enzyme elevations are well understood by PAH clinicians given the established profiles of existing therapies. Taken together, the results from both PROSERA and TORREY support a positive risk-benefit profile for seralutinib. This is especially important as it represents a new mechanism of action for a progressive disease where there is still significant unmet need. Now from a commercial perspective, we believe seralutinib continues to represent a meaningful opportunity in PAH, particularly in patients with higher baseline risk who face the greatest morbidity and mortality. In PROSERA, we saw a clinically relevant effect on 6-minute walk distance and consistent benefit across multiple endpoints in this identifiable population, achieved without vasodilation or hemoglobin-mediated mechanisms. And importantly, these results also highlight potential differentiation in patient groups such as those with connective tissue disease who have historically derived less benefit from existing therapies. So as we tried to make clear this morning, PROSERA did not deliver the results we had hoped for, but it does provide substantial evidence of a clinically meaningful effect in high-risk patients living with PAH. Our immediate next steps are to complete our deeper analysis of the PROSERA data across endpoints and subgroups and of course, to engage with regulators to discuss the results and understand their perspectives. We also await the results from the CT FRI study, the lung imaging study, which will assess blood volume distribution in the lungs. We believe this will be an important piece of context as we interpret the overall data set. Now I want to close by once again thanking the patients, investigators and study teams who contributed to the development of seralutinib as well as our employees, partners and shareholders for their continued engagement. We thank you again for your time today. Operator, you may open the line to questions.

[Operator Instructions] Our first question comes from the line of Joe Schwartz with Leerink Partners.

So given you enrolled PROSERA to include REVEAL Lite scores of 5 or greater in order to avoid enrolling mild patients who you expected to respond less to seralutinib and you narrowly missed the more rigorous p-value of 0.025 there in the overall population and now we're doing the subgroup analysis in patients with REVEAL Lite of 6 or greater. I'm wondering a few things. First, if you can help us understand the practical differences between these 2 different patient populations. Second, what does each patient population represent relative to the total PH population in treatment today? And how did you decide to choose the cutoff of 5 when designing PROSERA while including a cutoff of 6 for the prespecified analysis?

So Joe, let me start. In the plan that was agreed to with the FDA, we had used REVEAL Lite 5 as an entry criteria. In our statistical analysis plan, we were able to look at the median of REVEAL Lite as a subgroup of prespecified subgroup analysis. That median was 6. So 5 and greater was an enrollment criteria, 6 was what the data actually -- or the patient population and the data generated. So that's the reason why 6 being the median of all patients was where the line was cut. So...

And remember, the patients in this study that were 6 or greater represented the higher-risk patients -- represented 2/3 of the patient population in PROSERA.

And your second part of your question, Joe?

No, I think that...

He wanted to know the difference between 5 and 6....

This is Rob. 5 still includes the upper end of low risk. So some of those patients filtered into the study by being a REVEAL Lite 2 of 5.

And then another one on efficacy. I'm just wondering how many patients in PROSERA were on sotatercept? And how did they perform?

So 6 patients, we had 5 on drug, 1 on placebo. Those that were on drug had a mean improvement of over 70 meters.

Okay. And then on safety...

It's very consistent with what we saw in our preclinical work, where we think the -- this combination represents not just an additive effect, but actually a synergistic effect, not only did we see that preclinically, but there were a lot of anecdotal evidence of that, and now we're starting to see it in the trial setting.

And to your question about patient population and commercial as far as the more intermediate and high-risk patients as part of the pie, Bob?

Yes, Joe. Yes, the commercial opportunity is significant in this high-risk patient population. We're seeing a large number of the patients that are treated with triple or four therapies are typically functional Class III patients, which fit into this analysis.

And I just want to make one thing really clear because Joe, it's a great question, and it really stimulates an important point. That is that we clearly have now proven that in the higher-risk patient population, the sicker patients, this drug shows [indiscernible] in that 24-week window. But the data also suggests that over time, even the lower-risk patients, the less sick patients improve over time. And as we take a look at our 48-week data, we see the trend that we saw in TORREY, where over time, patients got better, not just the sicker patients, but also the patients who were less sick in TORREY improved over time. So it's not like this drug would only be used in the most sick patients, physicians would want to use this drug even in the lesser sick patients to prevent longer-term progression because of the underlying effect we're having in both lung and heart.

Our next question comes from the line of Yasmeen Rahimi with Piper Sandler.

Two questions for you. I guess the first question is, you had -- when you originally designed PROSERA, it was enriched for this high-risk population where we found really the robust results. So at the time when you engaged with the agency, did you ask, yes, you want us to include a broader population, but prespecified. Was there a communication around saying, yes, even if you hit stat sick on your prespecified high-risk population, you can file, but we would like you for a consistency basis on an overall population. Would love to kind of understand at the time when you designed it, what the FDA's view were on the high risk population, which obviously is the highest unmet need. And Faheem, thank you for the comments that ultimately the drug could be used across. So that's one. Sorry, for the long commentary. Two -- second question is 48-week data as well as the CT substudy is going to be really helpful. Maybe help us conceptualize how much visibility do you have currently on a blinded basis on the -- or any color on when that data could come and whether you need that data as a gatekeeper to engage with the agency in terms of next steps. So if you could provide some color on when we could get that and which you want to wait for that before you discuss next steps. I appreciate color on these both topics.

Thank you, Yasmeen. This is Caryn Peterson. In terms of the enrichment strategy and the discussions with FDA prior to initiating the Phase III trial, they were in full agreement with our target population and our eligibility criteria. So high-risk population, it's well identified, and they were in agreement with our selection criteria.

And Yas, in regards to the FRI data, we have not seen those results. As you know, they had to be adjudicated by central read. We anticipate that we'll be able to have those results early in the second quarter.

Yes. And Yas, back to the FDA and our approach there. We believe that -- well, first off, I think we would all recognize that this is a disease of incredible progression. Mechanistically, the need for a new mechanism, it attacks this disease in a different way, that affects the underlying mechanics of this disease is important. The evidence that we've got in place, we think, combined with the TORREY study together will represent a compelling package for the FDA. We have 2 studies now that are in concordance with one another, both meeting the statistical threshold for 2 studies that the FDA would -- we think the FDA should consider as clinically meaningful and relevant for this patient population.

The next question comes from the line of Paul Choi with Goldman Sachs.

My first question is in terms of the precedents in the sort of broader cardiovascular space and PAH in particular, can you direct us to any regulatory path or precedents that where you've seen subsets of patients approved, including in your case, such as a higher-risk population relative to a study that may have studied evaluated a broader population? And my second question on the financial side, with your decision to pause enrollment in SERANATA, can you give us an updated view on what your cash runway might look like? And then any updated thoughts on how to treat the convertible due in June of next year?

Yes, Paul. So again, we will end the first quarter with roughly $105 million of cash on the balance sheet. In regards to runway, we are obviously going through the exercise right now to evaluate what that will allow us to accomplish as well as time lines for potential interactions with regulators. So more to come on that. But again, about $105 million on the balance sheet at the end of March. Caryn?

Yes. With regard to the patient population in the high-risk group here, there's probably not precedent, but this represents a high unmet medical need. And I think that in itself is a very important subset that the FDA looks at in the overall population. So we believe that there is a place for this drug. I think the FDA will probably see the data and support that high unmet medical need.

And Paul, there is, of course, plenty of precedent on FDA approving drugs that miss their primary endpoint, but it's typically in diseases where the unmet need is high, and it's first-in-class drug.

Our next question comes from the line of Andreas Argyrides with Oppenheimer.

I was hoping for more clear-cut results here, but clearly an active drug. Just if you can give us a little bit on the placebo response here, just how the geographical breakdown compared to expectations at the time of enrollment? And then I have one follow-up.

Yes. So I think, Andreas, you recall that we made a significant investment in Latin America following the very, very significant results that we saw in the STELLAR study for sotatercept, where that was a geography that patients benefited the most. So certainly, for us to see an almost parity between the placebo rate and the treatment rate and how the statistical plan using Hodges-Lehmann works where that ended up reducing the treatment effect by 8 meters. So adding 8 meters on the placebo side was extremely, extremely disturbing to our team because, again, we expected to have an effect that has been seen in most PAH studies where Latin America is the best performing geography. We're still early in the investigation of what happened in Latin America, and we certainly will have more to come as we continue that work. And even in other geographies, we saw a higher-than-normal placebo rate. Importantly, in the places where PAH treatment has been quite frankly, the most mature in North America and Western Europe, Australia, we are seeing what would be historically comparable placebo rates. So certainly, there is something that happened in Latin America. We have to understand it, and we will obviously engage not only with the investigators there, their sites as well with PPD who you recall was also CRO that did the STELLAR study. So more to come, but that was probably the most surprising and disappointing finding. Faheem?

Yes, Paul. What's really fascinating about what happened in Latin America, we saw a substantial number of super responders on placebo with over 100-meter walk improvements, which is really kind of quite fascinating. But what I think is really interesting, which really shows the impact of this drug is that over time, we start to see a separation. Even though we had that substantial placebo effect, as we look at the Latin America data out to week 48, we actually see improvement on the drug arm, just to give you kind of a sense, when we do an apples-to-apples comparison, the placebo effect starts to catch up on these patients at week 48. So they have a 40-meter improvement on placebo at week 48, that drops to 15 meters. But the drug effect goes from a 50-meter improvement up to a 66-meter improvement. So you start to see the separation of placebo and drug over time. And we think that might be related to and consistent with what we see in the less sick patients that as we're affecting physiological -- having those physiological effects in the lung and the heart, the sicker patients will take a little longer for that response to occur. And we saw that in TORREY, and we seem to be seeing it here again.

So what we need to do, Andreas, is really unpackage where, in fact, those patients that were enrolled in Latin America, a REVEAL Lite 5 or greater, it's obviously disappointing. It's obviously extremely frustrating, and it is incumbent upon the Gossamer team with our friends at PPD and Chiesi to understand what happened because this result and all of our KOL thought partners upon seeing this were stunned by what happened in Latin America.

Thanks for your question...

I was going to ask a follow-up on the -- on any feedback on KOLs. But before I do, can you just also and you mentioned this in the press release, but talk a little bit about how kind of this geographical discrepancies impacted SERANATA. So I mean, I know it's a pause. You're partnering with Chiesi. There's a communication there. Is there an opportunity to continue in ILD, but focusing on certain geographies? Just maybe some kind of thoughts just because we know that there's a big opportunity in ILD.

Yes. Look, as I said in my remarks, we are absolutely convinced that seralutinib may play and should play an important role treatment in patients with PH associated with interstitial lung disease. not only because of the fact that we know that we are having effect in PH, but also our antifibrotic effect, the anti-inflammatory effect seems especially well suited for that patient population. So our decision to pause is not related to our conviction in the drug in that patient population. Just the contrary, we continue to be quite optimistic about that. But it's really about resourcing and making sure that our first priority is getting PROSERA over the finish line and getting it approved by the FDA. And that is -- at this point in time, this is where all -- I mean, obviously, we're going to go through all of the financials and making sure that we are prioritizing everything for that effect first. At the point in time where we think we've got financial resources to be able to initiate SERANATA again, that will be something that we will be very excited about.

And Andreas, I think to underscore that point, if you look at the results we had in the CTD population, where, again, fibrosis and inflammation are significant, that benefit in CTD is a read-through to what could happen in PH-ILD. Quite frankly, the results that we showed in that connective tissue disease population are unprecedented when it comes to efficacy, just look at what happened with the STELLAR study, which only had an 8-meter improvement to our 37 meters. So ultimately, the pause for SERANATA, as Faheem said, is to understand not only the financial implications and what we need to do to get the PAH indication over the finish line. But importantly, to your point, when you have issues like we had with study conduct in places like Latin America, we need to understand that and revisit those assumptions when it comes to starting another significant Phase III.

Our next question comes from the line of Olivia Brayer with Cantor Fitzgerald.

On week 48 data, can you give us any actual details around what that number is, both in terms of treatment effect and p-value? It looks like it's in the high 20s, maybe even close to 30 based on the slides in terms of treatment effect, but would be curious if you guys can put some numbers around that. And then on time to clinical worsening, how important is that going to be as part of your conversation with the FDA? I assume they'll have some questions just around that p-value. Obviously, it is a secondary, not a primary. But I know that they look at TTCW pretty closely in this patient population. But it would be great to hear your thoughts on just the importance of that endpoint considering that it wasn't stat sig in either the overall or the intermediate to high-risk subgroup.

Rich, why don't you start with CTD and then I'll get some of the numbers for Olivia. But need on time to clinical worsening question, how important that is...

Time to clinical...

Yes, so I think that the -- overall, the time to clinical worsening in CTD, we didn't really look at that yet. We just looked...

No, I'm sorry, the question is the implications of -- how important is time to clinical worsening to -- overall.

Okay. Apologies. It's -- it was one of our key secondary endpoints. It's a different way beyond 6-minute walk to look at the effect of the drug. It's around hospitalizations, escalation of therapy. So it will -- it is an important endpoint, particularly for the European regulators. It's something that will be considered by the FDA, but we do have, I believe, other than time to clinical worsening, other secondary -- our primary and other secondary endpoints that would be adequate to get approval.

Yes. And I would just add on, too, on the time of clinical worsening, Olivia, is in the study, there was only 39 events. It was a low number of events at week 24. We probably need to see an event rate up in the 50s to be able to show any sort of statistical significance. But in that higher-risk patient population, those with REVEAL 6 and above, we saw a very strong trend in time to clinical worsening. But again, the event numbers were just too small to show statistical difference.

Yes. And I think that is also compounded by, again, the high placebo rate. That is also probably an implication of some of these patients not having clinical events. So again, it comes back to placebo. In regards to your question about patients at week 48, so you can see on the slides that we have the numbers, 57 on seralutinib, 65 on drug. It was roughly a -- on the overall patient population, a 30-meter improvement the statistical value -- a different statistical test is using ANCOVA without imputations at -- run at 0.02. So statistically significant and more importantly, very clinically meaningful.

Our next question comes from the line of Patrick Trucchio with H.C. Wainwright.

Just a clarification question from us and then a couple of follow-up questions. First, I was wondering what the prespecified placebo assumption was in your power model and how far the observed 13.5 meter placebo deviate from that assumption? Secondly, in the REVEAL greater than 6 subgroup, I was wondering if the treatment effect was consistent across regions and background therapy strata. And then just lastly, I was wondering, as you've had discussions with the FDA, did they indicate an openness to a risk-based label? And what could the scenarios around a path forward to approval look like -- could one of those look like an approval with a confirmatory trial? Or how should we think about the pathway forward and in which population?

Thanks, Patrick. So the -- in regards to the placebo assumptions we had, we based our assumptions based on the STELLAR study. So we were expecting a placebo rate of plus/minus 2 meters. Your second question?

It was about the risk score geographic distribution.

Yes. So we're still unpackaging that, but certainly, we saw higher risk in Western Europe and North America. That being said, I think we also saw, frankly, better study conduct in those geographies, which is why we see the placebo rate that was near the bounds of our assumptions. In regards to regulatory, I'll let Caryn speak to that.

Path -- the approval path.

Yes. We're going to have obviously a conversation with FDA. We believe that we have between TORREY and PROSERA identified a high-risk population, certainly with an unmet medical need. And regards to how that looks as an indication statement, hard to know at this point, but I do believe that we have a compelling package to put in front of the FDA and determine a population where this drug is needed.

And I'll say, Patrick, as it relates to the need in the marketplace, we did have a confidential discussion with our steering committee, represented 15 of the top KOLs across the globe that reviewed this data with us. And I can tell you that there was an unequivocal support for the profile that we have here with seralutinib and all of the members saying that we absolutely need to go for approval to get this drug to their patients. So we -- part of our -- a big part of our confidence lies in the fact that we believe we have tremendous support in the clinical community for the profile of this drug as it stands today. And it speaks to the unmet need. It speaks to the fact that sotatercept is not a cure. We only have vasodilation, and we have a lot of drugs to use on these patients, but it still isn't having the desired effect, which is trying to delay progression, slow progression and improve patients' quality of life. This drug has the potential to make a big difference here.

Your next question is from the line of Ellie Merle with Barclays.

How do you interpret the imbalance in liver enzyme elevations? And does this suggest that there is systemic exposure of seralutinib? And if you could clarify what grade the liver enzyme elevations were? And then turning to the imaging substudy as we sort of relate those near term, how should we interpret what good data would be from this and the implications for the conversations with the FDA?

Yes, I can address the liver enzymes. It's -- first of all, the liver enzymes is consistent with small molecule TKI that's metabolized by the liver. Our mechanism is because of its profile thought to be immune mediated. So the percentage you see that we saw is very consistent with other TKIs and even other drugs such as bosentan. We also know that once you remove drug or interrupt drugs, the enzymes resolve completely and fairly rapidly.

In regards to the FRI data, we think it will be very, very important. Our expectation would be to replicate Ellie, what we saw in the TORREY study. Obviously, we've disclosed that while TORREY was limited to about 18 patients, a little over 120 patients of images, which we think will be very important because it will be a second check, if you will, on an objective view of what's going on with the patients. If you look at both the overall patient population as well as the -- those that had a REVEAL Lite of 6 or greater, the one objective endpoint that we were highly statistically significant on was the change in NT-proBNP, which is the biomarker for right heart health and function. We do believe that, that FRI data could be a very important assist in helping to explain not only the placebo effect, but also bring a level of objectivity for our discussions with regulators about what happened in the PROSERA study. So we, one, are very confident that, that data should be good like we saw in TORREY, but also is an important pillar to our regulatory strategy.

Our next question comes from the line of Laura Chico with Wedbush Securities.

So one more on the placebo response. Slide 9, you had that great picture of 6-minute walk distance placebo results from other studies at week 24. Do you have any sense as to how the week 48 placebo responses for other programs might fare? And I guess I'm just trying to understand if the week 48 placebo response you're seeing is also elevated or if that is more in line with what we would be expecting from other trials? And then just a housekeeping question. The PVR data was not collected in PROSERA, correct?

So PVR was not collected. In regards to other studies, 48 weeks, not a lot of folks did what we did, where you kept placebo-controlled data to week 48, it would be really an apples and oranges comparison because you'd be comparing it to open-label extension data, right? So for example, in the sotatercept data sets, most of their 6-minute walk data is open label. We were one of the first sponsors to go out to week 48 on a placebo-controlled basis. But what we can say is across all geographies over time, placebo starts to behave normally, specifically, as Faheem said, in Latin America, where when you look at those patients that had a week 24 walk and stayed in the study to week 48, which is roughly about 29 or 30 patients, you see placebo at week 48 starting to behave like you would have expected. So I also believe that, that week 48 data is another important pillar for our discussions with regulators because it starts to meter out the placebo effect and also continues to show continued improvement for patients on drug. So ultimately, we do think that, that week 48 endpoint is really, really important for our ability to say this drug has an important place in the marketplace for patients because of that long-term efficacy. And again, that placebo effect starting to normalize longer term.

Yes. And just on that placebo effect, I just want to bring everybody's attention back to specifically the North American results where we saw minus 3.9 meter, which was kind of more in line -- minus 3.9 meter placebo effect, which was more in line with our assumptions. And it's really quite interesting in North America. We obviously had a pretty substantial improvement on the seralutinib arm of almost 26 meters, almost 26 meters in North America, and that's on an n of 75 patients. So more normalized placebo effect, you really then get to see the true drug effect that is going on here.

Yes. And I think to add to that, quite frankly, the drug did what we asked it to do.

Absolutely...

The drug performed. We wanted to go north of 20 meters. And in the geographies where the study was run as well as it could be, we were well north of 25. It is placebo and is Latin America that has made this challenging.

And just to be clear, in every region, we were north of 20 meters on the drug effect.

Our final question comes from the line of Vamil Divan with Guggenheim.

So I got a couple of questions following up on some of the prior commentary. So one on the safety side, following on Ellie's question earlier on the liver safety. I noticed also on one of the slides that for the treatment-emergent adverse events leading to discontinuation of the drug or the investigational product and also for a special interest, the rates were quite a bit higher, about 3x higher for both of those in the seralutinib arm as compared to placebo. So just curious if you can give any more details there in terms of what is driving the discontinuations, especially. Second was around -- you mentioned your partner, Chiesi. I'm just curious if you've had any conversations around how they're thinking about the program now on a go-forward basis? And if they had any hand in also pausing the ILD trial. And then my last question was just again, going back to the question that was asked earlier around the debt coming due next year. I know, Bryan, you mentioned the cash on hand right now. Just curious if you can share any thoughts on how you're thinking about at this point, kind of navigating things through the debt payment in June next year.

Yes. So I'll take the Chiesi, the debt piece, and Rich could talk about the AEs. So clearly, we have a debt obligation due in May of 2027. We'll engage with both shareholders and bondholders to see what options we have. I would say that Chiesi remains very supportive of the seralutinib program. And certainly, as we are unpackaging this data, they have affirmed their commitment both commercially and financially. So that will also obviously play into the financing piece of that. And yes, they agreed with Gossamer about pausing SERANATA because, again, we need to interpret the PROSERA results and to see if we need to change some of the assumptions around the SERANATA study. But Rich, do you want to answer the question about the TEs?

Yes. So I think the question was around we had specifically for the hepatic, there's AESIs and then there's AEs and then there's the transaminase and the numbers are different. That's because they're looking at different components of liver safety. So the real value that we should be looking at is the 3x laboratory value because that's what's typically most relevant. The adverse events, that captures everything that exists, for example, that is related to the liver. I think the question was also around the discontinuation rate due to hepatic events, which was 7%. So we had protocol-specified discontinuation or withdrawal rates. And I'm going to say upfront that, that we were a bit more conservative in our trial because we wanted to understand the liver safety. So typically, you don't need to withdraw the trial unless you're above 8x. What we did is under certain circumstances because of monitoring or lapse of monitoring at sites, we recommended that the patient would withdraw after their enzymes come down because we weren't confident that they were going to be followed appropriately.

Okay. So maybe just a quick follow-up. So about 7% of discontinuations were due to liver. Can you comment on because it's a total 15% in the seralutinib discontinued due to treatment-emergent AE. What can you comment on what the other 8% were due to or a sense of what they're from?

No. So one is the AE table is 15%. So that's reported by the sites, right? And then the other number is the discontinuation rate, which is 7%, right? So the discontinuation rate is really reflective more of the 12% transaminase elevation that is based on laboratory parameters.

There are no further questions in queue. I will now hand the call back to Faheem Hasnain for closing remarks.

All right. Thank you very much, and thank you all for participating, and thanks for all the thoughtful questions. Look, I'll just close by saying we are absolutely confident that we have a drug that is making a difference and has potential to make a huge difference for patients living with PAH. Certainly, we've got very strong evidence in the fact in the 24-week period that this drug makes a difference in the high-risk patients. But I also fundamentally believe -- we believe, and most importantly, I think the clinical community believes that we are seeing with this drug a reverse remodeling effect, having an effect both in the lungs and the heart, a drug that has both antifibrotic qualities as well as anti-inflammatory qualities and a drug that will be used certainly in the higher-risk patients, but also a drug that will be used in patients to prevent longer-term progression. And this disease area, these patients are so desperately in need of another solution because, unfortunately, the solutions we have really aren't sufficient. And even after seralutinib, we will continue to need to innovate as an industry. I want to thank you all. Thank you for the time you spent with us and look forward to follow-up questions and dialogue with you as we go forward.

Thank you.

Thank you again for joining us today. This does conclude today's presentation. You may now disconnect.