Incyte Corporation (INCY) Earnings Call Transcript & Summary

Hello, and welcome to the Incyte Data Highlights from SITC Teleconference and Webcast. [Operator Instructions] It's now my pleasure to turn the call over to Christine Chiou, Head of Investor Relations. Please go ahead.

Thank you, Kevin. Good evening, and welcome to Incyte's Oral PD-L1 Program Highlights Conference Call and Webcast at SITC. The slides presented today are available for download on the Investors section of our website. Joining me on the call today are Herve and Steven, who will deliver our prepared remarks; and Jeff Jackson, our Vice President of Translational Sciences, who will join us for the Q&A. Before we begin, I'd like to remind you that some of the statements made during the call today are forward-looking statements and are subject to a number of risks and uncertainties that may cause our actual results to differ materially, including those described in our 10-Q for the period ended September 30, 2021, and from time to time in our other SEC documents. We will now begin the call with Herve.

Thank you, Christine, and good evening, everyone. So today, we are presenting data and perspective on our oral PD-L1 franchise, which consists of a portfolio of 3 oral PD-L1 inhibitors in clinical development. 550, our most advanced candidate is the first oral PD-L 1 inhibitor to demonstrate clinical activity. We are conducting in parallel, development of our 2 other oral PD-L 1 inhibitor, as we work to identify the best candidates for full development. 550 and is undergoing dose schedule optimization and the Phase 2 study is also underway, while both 280 and 318 are in dose escalation. An oral PD-L 1 inhibitor would be important for safety reasons and for economic convenience and reimbursement reasons. And as you will hear, we are making good progress and 2022 will be the year when key decision will be made on which product to move forward on potential registration path. And now, I'll turn the call over to Steven.

Thank you, Herve, and good evening, everyone. We have focused on development on oral small-molecule inhibitors of the PD-1/PD-L1 access. The potential advantage of an oral inhibitor includes better tissue penetration by a small molecule compared to an antibody, which may improve potency. Here on Slide 7, using confocal microscopy, we're able to show you oral PD-L1 inhibitor induced internalization of PD-L1. 550 is able to bind to the PD-L1 receptor, leading to dimerization and internalization of the receptors from the cells. Internalization starts within one hour and increases over time, as you can see towards the bottom of the slide. On Slide 8, we are showing a set of pharmacodynamic biomarkers that demonstrate T-cell activation in patients treated with 550. On the left-hand side, you can see the increase in plasma concentrations of the interferon related cytokines, as exemplified by CXCL9 following 550 treatment. And in the middle, increases in the expression of multiple interferon regulated genes in peripheral blood in a dose-dependent manner as measured by RNA Seq. And on the right, we are showing increases in proliferation of the circulating T-cell, as measured by Ki67, which were observed post-treatment for many patients. Moving on to the data for 550 that was presented earlier today at SITC. Here on Slide 9, the study designed for the dose escalation and expansion study. The Part 1 dose escalation evaluated 100 mg daily to 800 mg twice daily in a modified 3 + 3 design. It was report in dose expansion phase. In the Part 2 dose expansion phase, we looked at both I-O experienced in I-O-naive patients. And in the Part 3 dose expansion, we focused on a cohort of I-O-naive patients, whose tumors may be more likely to respond based on genetic status with microsatellite instability high and mismatch repair deficient solid tumors. And in Part 4, looking at human papilloma virus positive tumors like squamous cell anal carcinoma or head and neck cancers, as we have seen early signals in HPV-positive tumors. Turning to Slide 10. We had enrolled in the study at the cutoff of April 9 this year, a total of 79 patients, of whom 27 were in the dose escalation phase and 52 patients were in the dose expansion phase. Specifically, Part 2A had 10 patients, Part 2B had 33 patients, and Part 3 had 9 patients. Over 60% of the 79 patients were heavily pretreated with 2 or more lines of prior therapy, and approximately 17% of patients had prior I-O treatment. The study itself is ongoing with greater than 120 patients enrolled now. Here, we see the breakdown of the number of patients treated per dose level, with the majority of patients being treated with either 200 mg twice daily or 400 mg twice daily. Moving on to safety and treatment-related adverse events. Any grade treatment-related adverse event occurred in 58% of patients with nausea and fatigue being the most common. Immune-related safety signals, such as rash and pruritus indicate to us that we are achieving activity with regard to target modulation. Most treatment-related adverse events seen in this ongoing Phase 1 study were consistent with what has been seen with PD-1/PD-L1 monoclonal antibodies, including immune-related adverse events, with the exception of an increased rate of primarily low-grade peripheral neuropathy. Here, you see on this slide, 5 patients experienced a peripheral sensory neuropathy. Moving to Slide 13 and the occurrence of immune-related adverse events. We're showing the data that 15 patients or 19% experienced an immune-related adverse event. 10 patients in total or 13% experienced an immune-related adverse event of peripheral neuropathy, this includes both motor and sensory neuropathies, of which all were less than or equal to Grade 3 and all Grade 2 and 3 cases were resolved or improved. In some cases, these patients required a dose reduction, some are dose interruption and some required steroid treatment, as you can see on the slide. Only 2 patients had to discontinue therapy. 5 of the 10 patients continued their treatment uninterrupted. Of the 79 patients enrolled, 68 were included in the efficacy-evaluable population, which included all solid tumor participants enrolled in the study, who had received at least one dose of 550, completed a baseline scan, and met at least one of the following criteria: at least one post-baseline scan or had been on the study for a minimum of 63 days of follow-up or had discontinued from treatment. As you can see, there were 11 patients who were ongoing in the study but had no post-baseline scan information entered at the time of data cutoff and they were, hence, excluded from the efficacy evaluable analysis. Among the efficacy-evaluable population, where we had assessed 100 mg daily to 800 mg twice daily, the response was seen in 8 of the 68 evaluable patients and disease control greater than 12 weeks was seen in a total of 13 patients. Looking at the I-O treatment naive population as part of the 2B expansion cohort, there were 3 responses out of 14 patients treated with 400 mg twice daily, and in Part 3 in the MSI-high deficient mismatch repair I-O-naive patients, we saw 3 responses out of 5 patients treated with 400 mg twice daily. There were 2 additional responses this cutoff, one at 800 mg twice daily in Part 1 and one at 400 mg twice daily in Part 2A. Again, a very small end, but promising activity. This is the first oral PD-L 1 inhibitor to demonstrate clinical responses. Of the 8 overall responses, the majority were at the 400 mg twice daily dose level, 7 of the 8 were in I-O treatment naive and include human papilloma virus positive cancers, MSI-high cancers and deficient mismatch repair cancers. At the time of data cutoff, the majority of patients still had ongoing responses as designated with the plus sign on the right. On Slide 16, this is a swimmer plot of time to first response and the duration of response seen from the efficacy evaluable population, which was updated in the middle of this calendar year. As you can see, 5 of 8 responses are still ongoing, and there are responding patients who have been on therapy greater than 280 days. To summarize, 550 is showing activity and its safety profile is consistent with monoclonal antibodies with the exception of the peripheral neuropathy cases. We are working to optimize the dose schedule for 550 to attain the right therapeutic ratio, including evaluating intermittent dosing with the 400 mg BID dose. In parallel, we also have our 2 other oral PD-L1 inhibitors in dose escalation phase, and I'd like to spend some time talking about those developments as well. 550 is the most advanced compound, while 280 and 318 are catching up quickly, we have enrolled more than 35 patients on 280 already and more than a dozen on 318. The compounds work by the same mechanism of action, and we are seeing preclinical data that suggests the compounds are working as they should in terms of target modulation. In terms of differences, one is more chemically distinct from the others. On Slide 19, early preclinical data in the same mouse model shows that treatment with 280 and 318 induces comparable antitumor activity as was seen with 550. Although still very early in this program, we are seeing tumor shrinkage in patients treated with 280 and with 318. And to date, we have seen no evidence of peripheral neuropathy with either compound. In conclusion, we have 3 programs ongoing, and the studies are enrolling well. Over the next few months, we expect to have data available that will allow us to select our lead programs. And based on the clinical profile, we would be in a position to determine, which indications we would move forward into full development. We look forward to continue updating you on this very exciting program. With that, I'd like to turn the call back to Herve.

Thank you, Steven. So on Slide 22, the top depicted PD-1, PD-L1 antibody used across tumor types in the US in 2021. So PD-1, PD-L1 use is increasing obviously worldwide, and it's driven by combination and monotherapy use in adjuvant and consolidation. So on this chart, monotherapy use is depicted in dark blue, using combination with an oral agent is in light blue and PD-1 or PD-L1 use in combination with an injectable is shown in light red or pink. Clearly, an oral product could offer potential benefits for patients every time A PD-L1 therapy or PD-1 therapy is used either as a monotherapy agent or when used in combination with an oral agent. Taking a quick look at lung cancer, for example. In 2021, 50% of treated non-small cell lung cancer patients received the PD-1, PD-L1 antibody and 72% of the PD-1 or PD-L1 use is as a monotherapy regimen. For example, with Durva in Stage 3 consolidation or in Stage 4 after the fourth cycle of chemo plus PD-1. Taking another example in renal cell carcinoma, around 50% of patients are treated with a PD-1 or PD-L1 therapy, 57% of those treated with the PD-1, PD-L1 are treated with a monotherapy regimen, while 41% are treated with an antibody in combination with an oral TKI. The utility of an oral PD-L1 combination, it may not offer a lot of value in combination with IV chemotherapy, but if you look across the entirety of the US PD-1 and PD-L1 market, 80% of the current market would be amenable to an oral PD-L1 agent. Turning to Slide 23. Here, we want to show where the potential benefits of an oral PD-L1 inhibitor would bring more value either to different geographic regions in combination or as the monotherapy agent. The oral route of administration could lead to ease of dosing and no office visit requirement as well as no administration costs, which is more important in ex US regions, where reimbursement system favor oral administration. There is also an opportunity with oral-oral combination, where the same economic and convenience argument applies with the addition of the opportunity to create fixed-dose combination products. On the safety side, an oral small molecule inhibitor with a short half-life provides a switch off option if needed, while antibodies remain in circulation for a long period after their injection. This allows for easier titration and for better management of immune-related aid. And lastly, there is a theoretical benefit for the potential of increased tumor penetration, and we will conduct additional studies to see if this could lead to increased efficacy. Overall, the opportunity for an oral PD-L1 is substantial, and we are very encouraged by the data coming from our oral PD-L1 franchise. And with that, I'd like to turn the call over to Q&A.

[Operator Instructions] Our first question today is coming from Salveen Richter from Goldman Sachs.

This is Matt on for Salveen. Congrats on the data. Could you just briefly discuss how we should view the 21% ORR in treatment-naive patients in comparison to, say, Keytruda or other PD-L1's? And then also discuss maybe the safety results, were you expecting to see the safety profile or were you anticipating something perhaps better than the antibodies? And then finally, you recently initiated the treatment-naive Phase 2 in specific tumor types, was that based on this data?

Yes, Matt, thank you for the questions. It's Steven. So firstly, obviously, when you report out a study, you use the formal terms like overall response rate as it should be, but I think when you do early studies, you can see many patients either fall out early or as you saw, there were 11 patients had no post baseline scan information. So I wouldn't focus per se on the percentage, I'm not saying that by any way as an excuse, it's just they can be a little tricky when you look at different numerators and denominators. What is really encouraging is that we had 8 real confirmed resist responses. If you look at among 68, you get that response rate we determined overall of around 12% or as you allude to in the more focused groups, it can go upwards of 20% plus. But again, the bottom line is with 5 of the first oral PD-L1 agent, we have an active immunotherapy and that the responses are durable. So I think that's point number one. In terms of getting more precision around the response rate, I'm going to just answer your third question first, before I come back to safety. I think you have to look in more homogeneous, more likely I-O responsive, less potentially beat up patients to get more precision on what we call benchmarking data, in particular tumor types, whether they be lung or renal or melanoma, et cetera. So as I alluded to in my presentation, we've already enrolled another 50 patients compared to when this data cut was, and we'll have a good sense when we next present the data, what the true efficacy of the compound in those settings. But again, very encouraged by these responses and the durability thereof. Just to quickly also say to you, for the Phase 2, yes, it is built on the study and to be repetitive, it's just a cleaner, more homogeneous way of looking at it as opposed to a dose range in early study. In terms of safety, you never know when you take a new compound into the clinic, what you're going to see, because they're all chemically distinct. On the encouraging side, although not good for patients, per se, is we see immune respond or immune-related adverse events. And that's what you want to see that your drug is modulating the immune system and doing so in a fashion that causes what you would see with other immune modulators, so rash, pruritus, potentially endocrinopathies that are immune related, et cetera. So again, while not great for patients, if you know that you're then doing the target modulation that you want, what was a little unusual and unexpected were these 10 cases of peripheral neuropathy, 5 of which were sensory. And it's a little higher than what's been reported for IV checkpoints, whether the PD-1 or PD-L1, which is usually only a few percentage points. So that's on the straight sort of quantitative side. But in keeping with what we keep saying the 10-hour half-life of the drug, the moment it's held for an adverse event, you see pretty rapid resolution thereof and you can see only 4 of the 10 actually required steroids on top of that to get to resolution. So it just talks to the fact that when you switch off, as Herve was saying, you get a -- the drug washes out and resolution occurs of the neuropathy. Clearly, it's not a side effect we want patients to experience, which is why we spend time and continue to spend time looking at dose and scheduling work to see if we can weave the therapeutic ratio a little bit better with 550. We also took 2 other compounds into the clinic 280 and 318 and 318's chemically distinct from the other 2. 280 is a little ahead of 318. As I said in the presentation, there are more than 30 patients treated with that and then about a dozen patients treated with the last compound. And we've seen tumor reduction with both, which again encouraging in terms of the mix of action. And to date, have not seen peripheral neuropathy. Obviously, that's caveated by the smaller numbers and the duration. But the idea is to achieve the right therapeutic agent -- ratio with one or more of these agents that gives us the efficacy, a good safety profile that we can take forward. And that's what we're committing to doing, hopefully, sometime in 2022. So I think I answered all your questions.

Next question today is coming from [ Eva Privitera ] from Cowen and Company.

My question is actually about the neuropathy signal as well, just wanted to get a few more details. Can you talk about the time of the onset and whether or not it appears to be cumulative and how reversible it is and perhaps compare it to the course and severity seen with taxane?

Sure. Eva, it's Steven again. So it's -- in terms of time of onset, it was variable. If I look across the 10 cases, I think the earliest onset was around day-38 with the -- with one of the patients, and then there was a later onset in another patient, for example, at day-108. So it's -- there's some variability in time of onset, firstly. We don't see any cumulative nature thereof. Once recognized and managed either with reduction, interruption, steroids or in 2 cases discontinuation, there is no cumulative nature and almost all had complete resolution. Just to tell you that the grading of neuropathy is such that Grade 1 is actually asymptomatic. It's only Grade 2 onwards that's symptomatic, but the Grade 2s and 3s all resolved or dramatically improved, and there's no cumulative nature. So it talks to reversibility, clearly, in fact, 5 of the 10 again, not that we want patients to have this, but 5 of the 10, were able to continue dosing interrupted -- without interruption, sorry, and then had a resolution. So that's another interesting phenomenon. It's completely different from taxane neuropathy, which involves a different process in terms of demyelination, it tends to be more severe and tends to be reversible, so much so that it doesn't allow you to use the drugs in an ongoing basis or afterwards. And then again, just to say, while we're really interested in the oral nature in the short half-life, particularly in adjuvant and maintenance settings, when you're trying to cure patients, if you -- in that setting, obviously, you really don't want to run into severe toxicities with curative patients. And so if you have a drug that you can switch off quickly, if you run into trouble, we view that as a potential huge advantage. So again, I think I got all your questions.

Our next question today is coming from Kripa Devarakonda from Truist Securities.

I had a question about, you said 318 is chemically, structurally different from 550 and 280. Can you tell us a little bit more about how 280 differs from 550, it's small patient numbers, but you don't seem to see the peripheral neuropathy in these patients. And maybe provide some time line for when we can see data from 280, 318 and also a data update from the additional patients you've enrolled for 550. And one additional question, if you now have a PD-1 antibody as well as an oral PD-L1 inhibitor, which seems to have activity. Based on all the data that you've seen so far, do you have a strategy for development or would you rather wait to see the data from all 3 compounds before going ahead with the strategy?

Yes. Kripa, it's Steven. I'll start off. I hope I got all your questions charted down here, so I'll work through it methodically. So firstly, just again, to say, for 550, there were 79 patients in this cutoff in April. There are more than, I think, about 126 are now, and the study continues to enroll, and we will update that definitely in 2022. In terms of enrollment now, there are at doses that we know are pharmacologically active. And obviously, we're targeting I-O responsive tumors. So we'll be able, as I said earlier, to give you a much better sense in a more homogeneous population at a more pharmacologically active dose of the 550 activity, but we are experimenting with scheduling changes like one week on, one week off or other types of scheduled modulations to see if we can ameliorate the neuropathy signal with 550. The other 2 compounds are enrolling incredibly well. As I said, 280 has more than 3 dozen patients on it and 318, about a dozen, and we'll continue to do that, and we commit again to show you data in 2022 for those compounds as well, because they're enrolling well. And obviously, we build on the learnings from 550. So all of them next year should have substantive updates in terms of both efficacy and safety for those compounds. If you take that all together, we'll be in a position, as Herve said, right upfront to make development decisions, whether it may end up being with more than one compound, we'll see. There's potential optionality and interest in some non-oncology indications as well for oral PD-L1's in terms of enhancing, for example, a virally directed therapy and things like hepatitis. In terms of your question on PD-1 comparativeness, I'm sort of paraphrasing what you said, I mean that's obviously of an interest, huge interest to us to patients and to you guys and I think that will come largely from the Phase 2 for 550, because will be again, at active doses in homogeneous populations. But also with 318 and 280, we can now because we know we have active compounds, people are much more likely to put I-O responsive patients on it around the world, and we'll get a much quicker sense than we got from 550 of comparative like activity compared to PD-1 in similar settings.

On the IV oral complementarity in our own portfolio, I mean, obviously, as I said, I mean, chemo combination can be done with an IV product. And except for the reversibility benefit that you get from the oral potentially, it has no other clear benefit. And sometimes some of the chemo based on cisplatin in lung cancer are not very easy to combine with oral product, because of the side effect of the chemo. So when we see it, for example, in non-small cell lung cancer is that if you do a chemo plus IV antibody, you can then do the next first 4 cycle, I mean, that's the way it's used today, you can do the next 18 months of consolidation with an oral product. So there could be a sort of IV for the chemo combination, followed by oral for consolidation. That could be something that we are looking at.

And Kripa, one other question you asked up-front, I didn't address, differences. So we haven't published the structures for 280 and 318, but we can say that 280 is most similar to 550 in terms of its structure. It also is most similar in terms of its PK behavior, both have half-lives around 10 hours. 318 is most structurally distinct compared to the other 2 and we'll have to wait until we publish it to give you more details. But the half-life thereof is about half, so about 5 hours compared to the other 2. And then translation, I can ask Jeff Jackson here to comment if he wants to. But MOA wise, they all, at least in our hands, look like they do the same things in terms of what I outlined upfront for 550, in terms of the MOA.

Yes, that's correct, Steven. They are block PD-1, PD-L1 binding and have the same effects on T-cell activation and reinvigoration.

Our next question today is coming from Leonid Timashev from RBC Capital Markets.

Hi, this is Leo on for Brian. I guess I wanted to ask a little bit more about how your oral PD-1 is going to combine with some of the assets that you have in your portfolio. I guess, are there any combinations that you're particularly excited about, such as with adenosine or some of the bispecifics that you're working on in partnership? And I guess sort of to follow up on the previous question, you guys also have retifanlimab, which originally you had as part of a maintenance strategy. I'm kind of curious how you're thinking about development of oral PD-1 in combination with retifanlimab for maintenance? And maybe does the signal of peripheral neuropathy sort of interfere with this maintenance approach?

Leonid, it's Steven Stein. Again, I'll start off and others may want to add. Again, as Herve was alluding to upfront, oral-oral combinations have been of enormous interest to us from the beginning and whether they be with assets we have in our own pipeline or other assets, for example, directed at renal cell carcinoma, there's a lot of oral VEGF TKIs that are used that would potentially combine well with an oral PD-L1. In our own pipeline, the ones we may end up doing more work with once we have the dose and schedule we want to use, you alluded to one already that may be of interest for adenosine, we have an oral small module A2A, A2B inhibitor and in antibody CD73 inhibitor, so they hit the adenosine parkway from different angles and you get profound suppression of adenosine production and hopefully, modulate the microenvironment in the right way. And there, you could again, think about using an oral PD-L1 in that setting. But really, across our portfolio, all the oral combinations may be in play. I didn't fully understand your ready maintenance approach question, but I look at it because for retifanlimab, currently we have programs in our squamous cell anal carcinoma, MSI-high endometrial and Merkel cell plus lung program. But for an oral PD-L1, Herve used an example of Durva in the PACIFIC study in, for example, Stage 3 lung cancer, where after chemoradiation, you have upwards of 18 months of IV PD-L1 inhibition used there. If you think about using an oral in that setting, it would lend itself perfectly. And then just to be repetitive, obviously, the efficacy is paramount. But on the safety side, if you ran into trouble, the ability to switch off, we think would be a powerful differentiator there. So we think these oral PD-L1's may lend themselves to all sorts of adjuvant and maintenance use. And obviously, there are places where the IV checkpoints are established, but there are still many areas in addition that haven't been addressed yet in terms of either maintenance or adjuvant settings. So across the board there.

Our next question is coming from Jay Olson from Oppenheimer.

Can you comment on whether or not you found any correlation between response to 550 and PD-L1 expression levels? And then, can you talk about what sort of oral combinations you may have looked at preclinically and if you could comment on any potential synergies you may have seen with TKIs or KRAS G12C inhibitors or any other oral combos you may have looked at preclinically?

Jay, Hi, it's Steven. I'll start off. Jeff may -- will likely add to what I say. So it's early. So we didn't only select patients by PD-L1 positive or negative in the early part of the study until we did the later cohorts I alluded to, which were enriched, either in -- for I-O sensitive tumors or MSI-high, deficient mismatch repair for HPV positive. But even in those settings, we took all comers, and we'll do more correlative work afterwards, which Jeff may want to speak to. To date, to my knowledge, just to get to the exactness of your question, we haven't seen -- amongst the 8 responders the strict correlation with PD-L1 expression level and 550 activity, but we probably need more data and more experience to get to that. In terms of preclinical synergies, I'm not sure we've published any work in that regard yet. I'll see if Jeff wants to add anything to what I said.

Yes. I would just echo, Steven, what you said. I mean, while we're assessing PD-L1 expression and other potential biomarkers, tumor mutational burden and gene signatures, the numbers of patients that we have that are in I-O sensitive indications don't really allow us to establish any kind of correlation with those biomarkers at this point.

Our next question is coming from Reni Benjamin from JMP Securities.

Steven, you had one patient who had prior I-O treatment and still got a response. I'm kind of curious, are there any learnings here, especially in regards to potential post I-O use of these oral PD-L1's? And I have a follow-up.

Ren, it's Steven. Yes, I mean, obviously, that remains in the broader context a massive unmet medical need, right. Patients who prior I-O treatment treated, who either primary refractory or end up being resistant. People have tried different combinations, et cetera, there. And to date, nobody's really had a big pop in terms of efficacy, I don't know yet what to make of the one case. It was a colon adenocarcinoma that had prior I-O treatment that ended up responding. So I think with the NF1 in terms of that response, it's probably a little too early to tell. And we will have a greater collective experience in 2022 to present for you, whether we'll address that unmet need or not remains, I think, to be seen. So -- while interest in it is in NF1 one at the moment. I'll just say that.

Got it. And as we think about future development and how registrational studies might ultimately be done. Should we be thinking about it in terms of larger IV, I-O and an oral targeted therapy versus oral plus oral targeted therapy as a superiority study, a noninferiority study or do you think there might be a faster path to market, where it's a little bit more of either biomarker-driven or comparability studies driven?

Yes. Ren, it's Steven. I'll start and Herve likely will add some. Obviously, we haven't yet made the decisions, which we said we'll make next year. In my mind, obviously, one way to go and many are doing this, including ourselves in some settings, is a smaller niche tumor setting. But I think, yes, and this is just hypothesis like at the moment, we'll have to do our development plan next year. But is maintenance adjuvant settings, you can either go to areas, where there's no established care standard yet and go against placebo in those settings, because there's no case and just like Durva did in PACIFIC in Stage 3 lung or potentially, which I think would you alluding to is go head-to-head against those in some type of non-inferiority setting or more like replication studies as well. So all of those have potentially been entertained. And I think once we get more comfort on which of the compounds is the correct one, which is the right therapeutic ratio, then we'll be able to tell you in 2022, which way we will go. But we haven't yet decided, so all of the above. My own feeling is, again, these oral maintenance setting is of huge interest.

Got it. And then I guess just, one, trying to understand the unmet need. You guys make a very good argument regarding the differentiation, right, of an oral PL-1 and the ease of dosing and the like. And so is the unmet need -- I wouldn't think it's compliance, but is it -- is it compliance? Is it mainly convenience? Are physicians kind of through your salesforce feedback or just through your key opinion leader panels, are they telling you that this is something that they're looking for? Or is it a situation more where you guys see an opportunity where you can bring something to the market and then kind of -- even though doctors aren't necessarily looking for it, once they see it, we'll start using it more.

Yes, there's a bit of everything. One thing I'll say is there's a whole globe out there, not the US alone, and there are parts of the world, where oral use of checkpoints may be particularly helpful in terms of freeing up chair time in chemotherapy suites, et cetera. But also -- and you said it correctly, currently, when all you have is to go away on a maintenance regimen, but keep coming back for injections is what you have. Once you have an alternative, which is oral only, I think it would be enticing from a convenience point of view as well. What I was trying to say, maybe clumsily in my presentation was, in the adjuvant setting, your aim is to increase care. That's what you're trying to do is to improve cure rates. When you do that, what -- efficacy is obviously paramount, and I'll say that again. But occasionally, you'll run into, unfortunately, with any therapy, but even with checkpoints, really severe toxicity that could be life-threatening and that's not a great outcome in an adjuvant curve to set in. So we think a potential differentiated there as long as you have the efficacy piece, is the ability to avert that by switching off the immune stimulation that you give in really quickly and avoiding a dramatic life-threatening toxicity. So that's the one-way I thought about it. I don't -- Herve, do you want to add anything to that?

No, I think it's relatively clear. On the clinical profile, there is a hypothesis of better efficacy that we have absolutely no data to speak about today, but that we will be checking, because there is a different penetration. It's not the same mechanism. So that's something we need to work on to identify that if it is there. And there is something that we know already, and we know it for the patients we have treated, is that on the safety side, the switch off aspect of it can become very important. Now, who are the patients who benefit most from the ability to switch off. They are early stage curable patients, that's certainly one category. And now, there is another one that is emerging is I-O-I-O combination, where we know that the on-target immune toxicities can be very limiting and sometimes very, very problematic, where being able to switch off one of the 2 components of the combination could be very useful. So that's the safety efficacy kind of a profile, if you want to think about it that way. And then you have the convenience aspect, which is what it is. But on top of convenience and the cost of infusion and the fact that it is freeing up chairs in hospital, which sometimes is very important. There are in many parts of the world, reimbursement aspect to this, where, in fact, the IV product injected in the hospital are part of the hospital budget, and prescription product like an oral product would be outside of that budget. If we are able to show that we have some sort of similar profile clinically with the switch off opportunity and that it's something that goes through a different reimbursement system, I think it could be used very widely. So that's sort of the way we are thinking about it now.

[Operator Instructions] Our next question is coming from Matt Phipps from William Blair.

This is Rob Andrew here on for Matt Phipps. A bit of a follow-up from some of the earlier questions on development here. You've mentioned you're now up to 120 patients for 850 here, giving us a bit of insight into where you're out with enrollment or 280 and 318. So maybe first, you could just kind of provide some perspective on what you think you need to get to in terms of patient enrollment for these other 2 assets to reach the decision on what goes forward? Second, is the Phase 1 plan for 280 and 318 going to look similar in structure to the Phase 1 here for 850 in terms of dose expansion and such like. And then lastly, Steven, you've just mentioned a couple of times, building on some of the learnings from this Phase 1 with 850, maybe you could just elaborate a little bit more on what you think those learnings are and whether you think they can kind of expedite the kind of early development for 280 and 318 here?

Yes Rob, it's Steven. Thank you for the question. So the reason that we're now doing more schedule modification is largely to try and improve the therapeutic ratio as regards to the peripheral neuropathy. We're extremely encouraged by the activity, the rest of the profile is very similar to IV checkpoints. If for 280 and 318, we don't run into that, for reasons related to their chemical structure or something else, then we won't have to do that scheduling work. So as soon as we're in a pharmacologically active range, and we're already seeing tumor reduction, we can go a whole lot faster and just expand out that way. So that would be the chief learning thereof. The other thing is a little bit more an operational thing is the way studies work is physicians will put their patients on, obviously with equipoise, but they -- once there's an expectation of benefit, right? So in the beginning, we didn't know these things worked at all, right? So you have heavily beaten up Phase 1 patients coming in, were very unlikely to respond. Now that we've demonstrated and published efficacy, and we're already seeing this with our program, because we share with investigators what's going on, they enroll patients. And so that part will help us operationally in a very, very strong way. In terms of numbers, it's going to be hard for me to commit now exactly, but you have 3 dozen patients on 280, a dozen on 318. If we continue enrolling this way and we get appreciable numbers of between 80 and 100 patients on both programs in an early part of next year, then we'll be sitting on a very good data set to make decisions. We may not even need that many patients. That's my sort of reaction to your question.

Our next question is coming from Vikram Purohit from Morgan Stanley.

Great. So you just spoke a bit about the volume of data you might have by 2022 from 3 different molecules, but I was wondering if you could speak in a bit more detail about how you plan to compare and contrast the data sets you might have at that point in order to prioritize the 1 or 2 molecules that you think are the best? And then secondly, a point of clarification. You mentioned that 6 of the 8 responses were in the 400 mg BID dose. And I wanted to see if you could clarify how many of the 68 efficacy-evaluable patients had received the 800 mg dosing, either BID or QD?

Yes. It's Steven. I'm going to -- I'll come back to the second part of your question in a moment. What we'll do is, it's a little tricky to answer your first question in terms of compare and contrast, because it will be driven by -- as is always the case both efficacy and safety. So if we end up with, for example, either 280 or 318, having a similar degree of efficacy, but no peripheral neuropathy, then the decision is easy, right, from that regard. Efficacy, in general, in oncology always trumps safety. And so if you end up with an even greater efficacy signal on one of the other compounds, that's one you'd likely take forward. So it's a little hard to say, on its own, the drivers because it comes in both those columns. What you want is an appreciable numbers of patients, which we're pretty confident we're getting now and an ability to have some durability of follow-up to see responses, duration of responses and the safety profile. In terms of how many patients received the 800 dose in, that was largely, I think, in the Part 1 phase of the study. And I think it was of the 79, one patient at 800 daily and another 6 patients at 800 BIG, so a total of 7 patients at the 800 dose, and then we went back down again.

Okay. Understood.

To tell you, maybe this will help, at 200 mg from a translational perspective and above, you start seeing the chemokine and T-cell changes you want. So from a PD point of view, you get in the receptor occupancy you want saturation above 200 and definitely at 400 and all the PD things you want. So you're in that pharmacologically active range already. And 800 was just going above that to see if anything else happened there.

Thank you. We've reached the end of our question-and-answer session. I'd like to turn the floor back over to Christine for any further closing comments.

Thank you all for joining us tonight for a SITC update and for all of your questions. We look forward to speaking with you at upcoming conferences. Thank you, and goodbye.

That does conclude today's teleconference and webcast. You may disconnect your line at this time, and have a wonderful day. We thank you for your participation today.