Good day, ladies and gentlemen, and welcome to the Syndax Corporate Update Conference Call. At this time, all participants are in listen only mode. Later, we will conduct a question and answer session, and instructions will follow at that time. If anyone should require assistance during the conference at any time, please press star then zero on your touchtone telephone. As a reminder, this conference call is being recorded. I will now turn the conference over to Ms. Melissa Forst with Argot Partners. Ma'am, you may begin.
Thank you, operator. Welcome, and thank you to those of you joining us on the line and the webcast this morning for an update on results from Syndax's clinical trial. I'm Melissa Forst with Argot Partners, and with me this morning to discuss the data are Dr. Briggs Morrison, Chief Executive Officer, Dr. Peter Ordentlich, Chief Scientific Officer, and Michael Meyers, Chief Medical Officer. Also joining us on the call today is Michael Metzger, President and Chief Operating Officer of Syndax, and Richard Shea, Chief Financial Officer. This call is being accompanied by a slide deck that has been posted to the company's website. I would ask you to please turn to our forward-looking statements on slide two.
Before we begin, I would like to remind you that any statements made during this call that are not historical are considered to be forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. Actual results may differ materially from those indicated by these statements as a result of various important factors, including those discussed in the Risk Factors section in the company's most recent quarterly report on Form 10-Q, as well as other reports filed with the SEC. Any forward-looking statements represent our views as of today, May seventeenth, 2018, only. A replay of the call will be available on the company's website at syndax.com following the call. With that, please turn to slide three, and I'm pleased to turn the call over to Dr. Briggs Morrison, Chief Executive Officer.
Thank you very much, Melissa, and thank you to everyone joining us on today's call and webcast. I know it was sort of last minute. Glad you were able to make it. The purpose of today's call is to provide an update on the results of our ENCORE 601, our phase I-II clinical trial examining the activity of our Class 1 selective HDAC inhibitor, entinostat, combined with Merck's KEYTRUDA in melanoma, non-small cell lung cancer, and colorectal cancer. Last week on May eighth, we held our quarterly call to review our financial and operating results. During that call, I indicated that we would be presenting clinical data at ASCO and mentioned that the ASCO abstracts would be released on May sixteenth. The ASCO embargo lifted last night, and I hope many of you had a chance to review our abstract.
The data cutoff for those abstracts was in January of this year, and the abstracts were submitted in February. What we'd like to do today is to walk you through our most current data as of the updated data cutoff of April 24th of this year. During this call, we will not review the other aspects of our portfolio. Please refer to our call from last week for that information. A transcript of that call, along with a slide presentation, is posted on our website at www.syndax.com. Let me start on slide three, which summarizes our exciting ENCORE clinical trial program, in which we are testing entinostat in combination with PD-1 pathway antagonists, either PD-1 antibodies or PD-L1 antibodies.
We are now exploring entinostat in combination with a PD-1 antagonist in 6 different tumor types: non-small cell lung cancer, melanoma, microsatellite stable colorectal cancer, triple-negative breast cancer, ovarian cancer, and hormone receptor-positive breast cancer. The broad clinical trial program is supported by an extensive correlative science program that is designed to identify biomarkers that could predict which patients will experience a clinical benefit from our combination therapy. I will share initial results from that correlative science program on this call today. We're extremely encouraged by the results we're seeing across this clinical trial program. As we complete our initial clinical trials in each tumor type, we will prioritize which indications to advance to registration trials based upon unmet medical need, the competitive landscape, and our ability to generate an attractive return on our investment.
I'll provide details on our approach to non-small cell lung cancer, melanoma, and colorectal indications during the course of this call. I'd again like to emphasize how important we consider our correlative science program that's designed to identify biomarkers that predict which patients will experience a clinical benefit from our combination therapy. A number of companies have recently reported negative phase III clinical trials testing novel IO therapies in unselected populations of patients with phase III trials started based upon limited phase I or II trial results. We strongly believe that our ability to generate an attractive return on our investment requires us to thoroughly understand our phase II data and to make every effort to identify the patients most likely to derive clinical benefit prior to launching a registration trial. Let me now turn to slide four, which provides an overview of the ENCORE 601 trial.
After a Phase I-B safety component, we have been enrolling patients in four phase II cohorts. Patients with non-small cell lung cancer who have not yet received a PD-1 antagonist. Patients with non-small cell lung cancer whose disease has unequivocally progressed on a PD-1 antagonist and have also received chemotherapy. Patients with melanoma whose disease has unequivocally progressed on a PD-1 antagonist. Patients with microsatellite stable colorectal cancer whose disease has progressed on at least one chemotherapeutic regimen but have not yet received a PD-1 antagonist. All patients are treated with pembrolizumab, 200 milligrams IV every three weeks, and oral entinostat, five milligrams weekly.
I will not review the first cohort of patients with non-small cell lung cancer who have not yet received a PD-1 antagonist. No data from this cohort was presented at ASCO. The most recent data from that cohort was presented at SITC in November of 2017. Let's now review the data for the cohort of patients with non-small cell lung cancer whose disease has unequivocally progressed on a PD-1 antagonist. On slide six, I review the demographics of the patients with non-small cell lung cancer whose disease has unequivocally progressed on a PD-1 antagonist. In the abstract published last night, we described 57 patients. Those same 57 patients are summarized here. On our quarterly call last week, I indicated that we have now enrolled a total of 76 patients. The ASCO data represents the first 57. Slide seven shows a waterfall plot for this cohort.
In the abstract, we noted five of 57 patients had a confirmed response. Our updated data indicates that six out of 57 had a confirmed response, which represents an 11% overall response rate and a 95% confidence interval from four to 21. There's one additional patient who had an unconfirmed partial response. The median duration of response is now 4.6 months, with the longest ongoing over 14 months. Slide eight shows a swimmer lane plot where you can see seven patients still on trial as of the cutoff date. Slide seven summarizes the safety profile observed in these 57 patients. The toxicity profile is considered manageable in this population. Let me now introduce on slide 10 a biomarker that appears to predict clinical benefit in this population.
In our analyses to date, which includes measuring baseline values of 13 distinct immune cell populations, the % of classical monocytes in peripheral blood mononuclear cells is the best independent predictor of clinical benefit in this non-small cell lung cancer population. This biomarker is a simple blood test. The cells are identified by flow cytometry as being CD14 positive, CD16 negative with high expression of HLA-DR. A recent paper in "Nature Medicine" by Carsten Krieg and colleagues used an unbiased approach to identify the level of these cells as a predictor of response to immunotherapy. Slide 10 shows a plot where the baseline % of classical monocytes in peripheral blood mononuclear cells is indicated by a dot for each patient. Samples were obtained at baseline for 51 of the 57 patients.
Shown on the left panel are the values for 15 normal healthy volunteers, the six patients who had a response, and the 45 patients who did not have a response. On the right is an analysis where we defined clinical benefit as complete response, partial response, or total time on therapy of at least 24 weeks. This is a relatively standard definition of a broader clinical benefit that additionally includes patients who've had stable disease for approximately six months. In both analyses, you can see that the baseline monocyte values for the patients with clinical benefit are greater than those without benefit. This assay, of course, represents a continuous variable. We've done analyses to identify an optimal cutoff in this population. Slide 11 shows a PFS analysis using this initial cutoff.
You can see that the patients with the higher baseline percent of classical monocytes in peripheral blood mononuclear cells show both a higher response rate, 28.6% versus 5.4%, and a longer PFS, 5.4 months versus 2.5 months, than patients below the cutoff. What's tremendously exciting in this data is both the high overall response rate and particularly the PFS, which we believe is considerably longer than that observed with standard care agents are used in this population. I should note that in terms of a regulatory path forward, we believe PFS is an accepted endpoint should we conduct a randomized trial against the standard of care of chemotherapy. The 5.4 months that we observed is almost twice the best estimate we have found in the current literature for PFS when standard of care chemotherapy is used. The reference from Costantini et al.
described 115 patients who had progressed on first-line chemotherapy and then received second-line anti-PD-1 therapy. That reference is shown at the bottom of the slide. You'll note using this particular cutoff that about one-third, 14 of the 51 patients, would be considered biomarker positive and predicted to derive clinical benefit from the combination of entinostat plus pembro. Let me now discuss the potential clinical opportunity and summarize our next steps with regards to this indication or population. Slide 12 shows that the population of patients we are studying in this trial is clearly an area of growing unmet need. On this slide, we show the various approaches to newly diagnosed lung cancer, non-small cell lung cancer. Based on KEYNOTE-189, an increasing percent of patients will get pembrolizumab in combination with chemotherapy.
However, it should be noted that even in the KEYNOTE-189 data, about 80% of patients had progressive disease within 18 months of starting therapy. Some patients will receive a platinum-based regimen as their first therapy and then receive monotherapy with a PD-1 antagonist. Finally, patients with high PD-L1 levels may receive pembro monotherapy followed by a platinum-based regimen upon regression. No matter what the sequence, there are a growing number of patients who will need therapy after their disease has progressed on both a PD-1 antagonist plus chemotherapy. That's exactly the population of patients we've studied in this cohort. Our estimates are that about 84,000 patients a year are candidates for second or third-line therapy. If we use the monocyte cutoff I discussed earlier and assume about a third of patients are eligible for the entinostat pembro combination therapy, that represents about 30,000 patients a year.
Slide 13 shows the next steps we are undertaking to follow up on these exciting results. Of course, first, we need to continue our follow-up of this ongoing trial, monitoring overall response rate, PFS, and eventually, overall survival. We are already working to validate and industrialize the monocyte assay so that we can use it as a stratification marker in our next trials. We will continue to explore additional biomarkers that could improve upon the predictive value of classical monocyte. Perhaps most importantly, we plan to conduct a potential registration trial to confirm this result in additional patients. The design of such a trial is outlined on slide 13. The goal would be to validate the assay, more precisely define an appropriate cutoff value, and importantly, demonstrate that the entinostat pembro combination had superior efficacy as measured by either PFS or OS compared to standard of care agent.
The design of such a trial or trials is ongoing now. We'll say more about the final design and timeline soon. Our current best estimate is that this trial could potentially start by year-end, with top-line data potentially available in the first half of 2020. I want to emphasize that we would anticipate that such a trial would be conducted with registration intent and that such a trial could result in either an all-comers label or a biomarker-enriched label. We are also considering accelerated approval approaches based upon this data. In summary, I think it's clear that we're extremely excited about the data we are seeing in this PD-1 refractory population of patients. Let me now turn to our ENCORE 601 study in refractory melanoma. On our quarterly call last week, we indicated that we've now enrolled a total of 55 patients in this cohort.
The ASCO abstract published last night describes the first 34 patients. That is what is summarized in slide 15, the demographics of those 34 patients. Slide 16 shows a waterfall plot from this cohort. In the abstract, we noted six of 34 patients had a confirmed response, which represents an 18% overall response rate with a 95% confidence interval from 6.8 to 34.5. Those patients are shown in blue. I will note that there are three additional patients who had unconfirmed partial responses. This means that at one evaluation, their tumor had met the criteria for a partial response, but that response was not confirmed on a subsequent evaluation. These short-lived responses are not counted in our overall response rate. The median duration of response is now nine months, with the longest ongoing over 20 months. The median PFS in this cohort was 12.9 weeks.
I will note that the overall response rate in the two-thirds of patients whose disease had progressed on both a PD-1 antagonist and a CTLA-4 antagonist was also 18%, similar to the overall population. Slide 17 shows a summary line plot where you can see four patients still on trial as of the cutoff. Slide 18 summarizes the safety profile observed in these 34 patients. Let me now put these data into context. The initial data we presented last year at ASCO summarized the data from the first 13 patients treated with four response rates and a response rate of 31%. The data was relatively immature. It was not possible to accurately estimate the median duration of response. Our updated data now demonstrates an impressive median duration of response of nine months.
Again, I will note that the 65% of patients who have received both a PD-1 antagonist and a CTLA-4 antagonist, which represents a population of patients with a clear need for novel therapies. The overall response rate in that specific group of patients in our data is also 18%. Based upon continued discussions with melanoma physicians, we consistently hear that a response rate of around 20% would be considered highly clinically relevant, especially if the median duration of response exceeds six months. Nonetheless, let me again emphasize what I said earlier about our commitment to our correlative science program. We strongly believe that our ability to generate an attractive return on our investment requires us to thoroughly understand our phase II data and to make every effort to identify the patients most likely to derive clinical benefit prior to launching a registration trial.
We therefore will not be launching a registration program in melanoma at this time. We have ongoing biomarker work and have decided to follow the full 55-patient cohort to further mature while we continue our biomarker analysis. We would feel much more comfortable entering into a registration trial in melanoma once we have correlative data like I just showed you for non-small cell lung cancer. Nonetheless, I do want to outline our registration approach. Slide 19 shows the current approach to treating metastatic melanoma. We've designed a single-arm accelerated approval trial in patients who have progressed on both PD-1 and CTLA-4. The primary outcome in that trial would be overall response rate, with duration of response as a key secondary endpoint. We've also designed a randomized trial in patients who progressed on PD-1 monotherapy but have not yet received a CTLA-4 antagonist.
This trial will randomize patients to the entinostat pembro combination versus ipilimumab with a family of endpoints including overall response rate and overall survival. In summary, we remain excited about our melanoma program and look forward to updating you as data matures. Let me now turn to our results in colorectal cancer. On our call last week, I summarized the change in patient numbers for this cohort. What our ASCO abstract summarizes are the first 16 patients enrolled in this cohort, and that is what I'm showing on slide 21. Slide 22 shows a waterfall plot for this cohort. As in our abstract, we had two patients with pseudo progression, and one is a confirmed partial response. This represents a 6% overall response rate with wide confidence intervals ranging from zero to 32. The one PR is ongoing over 30 weeks.
The median PFS in this cohort was 12.3 weeks or roughly three months. Slide 23 shows a swimmer lane plot where you can see the one patient still on trial as of the cutoff date. Slide 24, again, summarizes the safety profile observed in these 16 patients. Let me now finally put this data also into context. In microsatellite stable colorectal cancer, O'Neil and colleagues recently published the results of KEYNOTE-028, in which 23 patients with PD-L1 positive microsatellite stable colorectal cancer tumors were treated with KEYTRUDA monotherapy. No responses were observed in the microsatellite stable population, and a median PFS of 1.8 months with an upper bound of 1.9 months was reported. Our initial data with the one response and a median PFS that exceeds the upper bound of the 95% confidence interval reported by O'Neil and colleagues is therefore, in our opinion, quite encouraging.
As we noted on our call last week, we will now enroll a total of 37 patients, an additional 21 on top of this first 16. If we see at least three responses in the 37 total patients, we will proceed to stage II and enroll an additional 47 patients for a total of 84. We will commence enrolling the modified stage I cohort later this quarter and anticipate making a go, no-go decision to advance stage II in the first half of 2019. I should also again emphasize that we are exploring many biomarkers in this cohort as well, with the goal of finding a way to enrich for patients who derive clinical benefit. Let me now summarize where we are with the overall ENCORE program. Slide 26 summarizes that we are moving forward rapidly in non-small cell lung cancer based upon the very exciting data I presented today.
We are allowing time for our melanoma data and biomarker analyses to mature before we commit to a registration program in melanoma. We believe we will be in a position to make that commitment before the end of this year. We are expanding enrollment in colorectal cancer. Our two other trials are randomized phase II trials in triple-negative breast cancer and ovarian cancer. We await readouts of both of those trials in the first half of next year. With that, I will open it up for questions.
Thank you. Ladies and gentlemen, if you have a question at this time, please press star followed by the 1 key on your touchtone telephone. If your question has been answered or you wish to remove yourself from the queue, press pound. Once again, if you have a question, press star and 1. Our first question comes from the line of Chris Shibutani from Cowen. Your line is now open.
Thank you for the question. Good morning. Could you educate us a little bit more in terms of this classical monocyte biomarker? In particular, I guess this is a measurement in peripheral blood. My rudimentary understanding of the sort of monocyte levels can be impacted by various situations clinically, including cardiovascular disease, infections, and whatnot. Also, monocytes are fairly heterogeneous. Can you describe kind of what the mechanistic underpinnings might be and how you're thinking about any other potential confounding factors when you have this peripheral blood measurement? Also why you think maybe this might apply for this population of patients from a lung cancer standpoint, and whether there's evidence of similar correlations or associations in other tumor types. Thanks.
Sure. Maybe we'll start with Peter or Dan. If, Peter, you can just sort of walk through some of Chris's questions.
Sure. Hi, Chris. As you know, basically, based on entinostat's mechanism of action, we are very interested on impact of myeloid cells. The work that we described here, along with the other cell types looked at, was done in Dmitry Gabrilovich's lab with sort of a question of how various myeloid population and lymphoid population of cells could relate both at baseline to the outcome of the patients as well as pre and post changes relative to mechanism of action. We had been interested in looking at MDSCs and other immune cells. Of all the ones that we looked at based on the data we have here, these classical monocytes appear to be the best predictor. What does that mean?
We looked at both the response rate and the PFS, then we sort of asked the question that you had, which is how could this relate mechanistically to what entinostat may or may not be doing relative to the function of these classical monocytes. Our current hypothesis is that based on the paper published by Krieg earlier this year and other data, is that these patients have all progressed on a prior PD-1 as well as actually prior chemotherapy. The benefit that they're getting here suggests that they're poised to respond based on these higher level of circulating monocytes, yet there's something else that's preventing their response from happening just to the checkpoint alone. We feel that entinostat's relieving whatever those immune suppressive mechanisms may be. We're still working through the science. We have tissue biopsies we're currently investigating for gene expression.
We have pre and post blood samples we're also continuing to look at. Obviously our goal is to tie some mechanism of entinostat back to these circulating levels. At least currently, the idea is these classical monocytes, which as you described, are sort of these patrolling parts of the innate immune system, just appear to be at higher levels in the responders. We just think that it just suggests almost a primed system that's ready to respond but needs something else, and that's really the relief of suppression that entinostat provides.
As far as tumor type, what do we see in the data?
yeah. Certainly we looked in our melanoma cohort. We see a trend that's similar relative to the objective response rate, so that higher baseline levels of classical monocytes associate with a response. However, we're still following the sort of PFS and overall clinical benefit to see if that also matches what we see with the lung cancer. So far all we can say is the trend relative to response seems similar. The lung is certainly clearer. We'll have to just see with a larger data set whether the PFS ends up separating or not for the two groups. In colon, the data is just too small so far with the one responder to know whether or not we see an association of high levels of monocytes.
To be clear, you're looking at this as a biomarker to be predictive of response. Is there any evidence to believe that this is also a biomarker that can be indicative of durability of response or any other dimension?
I think we're envisioning this as a biomarker of overall clinical benefit. Our goal is to identify a patient population that we may be able to then take forward that would be obviously competitive to the standard of care. Just responses alone that are short-lived wouldn't be that interesting to us. It's really the combination here of both the PFS and the overall response rate that we see in this patient population with the high monocytes. Obviously, we'll, as Rick mentioned, be looking at other parts of the sort of overall package of data to see if we can either refine the cut point through further studies or identify mechanistic reasons within the tumor tissue that could relate back to some association with these high monocytes.
Obviously, we're pretty encouraged to have found at least one of the cell types that has been previously identified as being somewhat relevant for these immune responses to be also seeming to associate with clinical benefit in our studies.
Great. Thank you.
Our next question comes from the line of David Lebovitz from Morgan Stanley. The line is now open.
Thank you. Thank you very much for taking my question. On the non-small cell lung cancer data set, was there a specific pre-specified analysis regarding this biomarker, or did this come from a post hoc evaluation just sweeping through all biomarkers, as many as you could, to determine some potential correlations with efficacy?
David, let me take that question. We are doing exploratory analyses on a number of biomarkers. As I noted in my comments, there are 13 different ones that, different immune cell subsets. There are analyses of the tumor itself, PD-L1 expression, tumor mutational burden, NanoString assays of RNA expression. It is a sort of broad and exploratory analysis. I guess for that reason, there's not really a statistical analysis of that because there's way too many tests for us to be able to correct for the multiplicity.
What we've done is to essentially go through these and essentially, are there any that look like they are able to give us the kind of correlations that we've seen? This one, by far and away, is the best predictor of the assays we've run so far. Obviously, the next step is to try to confirm this in an independent cohort, and that's the next experiment that we outlined during my prepared remarks.
I guess in a typical refractory non-small cell lung cancer population, would this be characteristic of that with respect to how many come in with high monocytes and how many come in with low?
Yeah, that's a question that we don't know the answer to. I don't think anybody We have not identified large data sets where people have characterized these specific classical monocytes. The reference I gave in the prepared remarks was to a trial in which they specifically followed patients who had first got a platinum-based chemo, then got a PD-1, sort of the way PD-1s were used when they were first approved, and then looked at the outcome of patients who received subsequent standard of care therapy. In that study, of course, they didn't look at monocytes anywhere along the way. There's work we have to do to see if we can identify independent cohorts of patients where we can look at this or generate additional data ourselves in future trials.
Okay. Just one more thing to reconfirm on melanoma. You had indicated you're going to finish the trial and evaluate that population for biomarkers, which could potentially, depending on what you find, lead to a more pivotal type trial either late this year or into next year to start?
Yeah. I think that's a fair summary.
Okay. Thanks for taking my questions.
Our next question comes from the line of Joel Beatty from Citi. Your line is now open.
Hi. Thanks. Good morning. Are you aware of any data from other trials that are not studying entinostat, but looking at this non-small cell lung cancer population to help with understanding if this monocyte biomarker is predictive of clinical outcomes in general with other drugs as well, or if this is a biomarker that's unique to entinostat?
Joel, there have been papers and literature in the past looking at monocytes as a prognostic tool in non-small cell lung cancer. In general, if you use a more general definition of monocytes, the higher the monocytes, the worse the prognosis. This is actually inverse to what people have described previously, which we think is quite interesting. Again, it is early days, and I don't know that we have not yet identified another large non-small cell lung cancer database where these specific monocytes, and again, I described for you their immunophenotype, where these specific monocytes were looked at as a correlative with the outcomes.
Thank you. That's interesting. Also, for the registrational trial design, are you able to give any sense of what would be important to show there, in terms of any type of endpoint that you might be needing to assess?
Only in general terms that we could look at overall response rate, PFS or OS. The data we have so far makes us think it's possible that we could win on a PFS trial against standard of care chemotherapy, we have more work to do to finalize that trial design.
Got it. Thank you.
Our next question comes from the line of Robert Hazlett from BTIG. Your line is now open.
Yes. Thank you for taking the question. You just touched on it, Briggs, I think a little bit. You talked about assay validation for the registrational trial. What are the other potential gating items for kicking off the non-small cell lung cancer registrational trial?
Maybe I'll let Peter talk about the biomarker, and then let Michael Meyers talk about sort of regulatory interactions and trial design, finalization stuff. Peter, you want to comment on the work on the biomarker?
Yeah, sure. Relative to gating for moving ahead, we'll clearly see the data coming in from the rest of the patients as part of the full 76 cohort. That'll give us a little bit more data, although the 51 patients' worth of data is already fairly robust. I think our goal is to work on sort of identifying the appropriate vendor and partner for us to help in terms of industrializing the assay and just refining the cut points and things like that. Relatively speaking, these are fairly well-characterized cell types. From a biomarker perspective, just collecting a little bit more data would be gating. That's probably about it.
Michael, you want to talk about sort of regulatory and clinical?
Sure. As Briggs suggested, there are a couple of different trial designs that could give us approvals both in an all-comers population or a biomarker selected population. We'll be taking those different trial designs to the FDA and to EMA, and basically get guidance from them as to what they would require in terms of an approval both in all comers, but more specifically as well in a biomarker selected population. There are a couple of ways that that can be approached. There's been an FDA guidance on that, which we've read and studied, and we'll be designing trials that speak to that guidance.
Okay. Thank you.
Our next question comes from the line of Charles Butler from Guggenheim. Your line is now open.
Yes. Thank you. Three questions, if I may. Number one, if we go back to the very first question from Chris on looking at other tumor types for elevated monocytes. The question is: If you were not able to find elevated monocytes in, for example, melanoma, what might that suggest about the biomarker use or utility to move forward in a phase III study? That's question one. The second question is, and it would be interesting, Briggs, if you could make some correlation to these data, but I believe Bristol had demonstrated nivolumab plus an anti-LAG-3 in an anti-LAG-3 positive non-small cell population also had utility in a PD-1 negative background. I'm curious if you recall that, if you would refresh us on what that meant.
It would be somewhat of a similar, I guess, maybe somewhat of a similar outcome or not, depending on what that response rate might be. The third is, if we look at CRC, I believe it was last ASCO when Roche demonstrated some interesting data with a bispecific CD3 CEA antibody and had some good responses or reasonably good responses, much better than had been seen, of course, in the past. The question is, how might one think about that in the context of the CRC data for which you have demonstrated today? Thank you very much.
Okay. No, thanks for your questions. First, the question of, if the biomarker holds up nicely in lung cancer but doesn't hold up in others, what does that mean from a development point of view and from a scientific point of view? I think if it holds up in lung cancer and we're able to take it through for approval and use it clinically, I think that's a good outcome for patients, and it would be a good outcome for us, and we would pursue that. If it doesn't work, remember, we set up this broad proof of concept trial program in multiple different tumor types based upon the sort of early data that the immunology in these different tumors may be somewhat different. It is certainly scientifically plausible that there's a marker in one tumor type that doesn't predict across all tumor types.
Obviously, if it did, that would be more advantageous, but there's no reason that it necessarily have to. I think if it doesn't, it turns out not to be as useful in melanoma as it appears to be in lung cancer, we would continue work looking for maybe a different marker that would work in melanoma. Your question about Nivo plus LAG-3. My only recollection of Nivo plus LAG-3 was in melanoma, not in non-small cell lung cancer, but I'm not an encyclopedia of all the data that BMS has ever published. They do a lot of studies. Their Nivo plus LAG-3 in melanoma in the LAG-3 positive population, I think had about an 18% response rate, which is in the same range of what we're seeing in an unselected population. Finally, your question about the CD3 bispecific.
Our understanding is that we have not seen that move into another trial. There was some initial data, again, with a low double-digit response rate, but we haven't seen that move forward into subsequent trials. Again, I think it's early for our colorectal data. We're just expanding the population. Again, I think you may remember when we first talked about the colon cancer population, there are relatively well-described immunophenotypes of different colorectal cancer subtypes. It wouldn't be, again, surprising if this combination worked in one of those subtypes but not another. It just takes a larger sample size before you can start to dissect that all out.
Thanks a lot, Briggs. Appreciate it.
Our next question comes from Madhu Kumar from B. Riley FBR. Your line is now open.
Hey, guys. Thanks for taking my questions. Following from Joel's question, kind of more pointedly, have monocyte levels been specifically examined as a marker of response and/or progression for the kinds of chemotherapies that you could potentially use in this upcoming study?
Not for this monocyte in data sets that we've identified. I think what you and Joel are both getting at, just to put it very clearly, is this a predictive biomarker or a prognostic biomarker? In other words, is it prognostic no matter what therapy you use, or is it specifically predictive for the entinostat pembro combination? That's a question that I don't think we can answer today.
Okay. Also based on the immunophenotype you guys have described, I mean, it's curious, right? Because CD14 positive HLA-DR low negative cells are in many cases a classical marker for myeloid-derived suppressor cells or MDSCs. Just remind me, how do MDSC counts perform in your biomarker analysis, and how do you think about the possibility that it's not necessarily monocyte counts per se, but really the ratio of monocytes to MDSCs that could be affecting kind of on a mechanistic basis entinostat's activity?
Peter, you want to take that one?
Yeah, sure. Far we have not seen, at least for this set of data, baseline values of MDSCs that are associated with sort of clinical benefit or non-response. We looked at 3 different types of MDSCs with Dmitry's lab. At least circulating levels don't seem to, in our studies, have an association. In terms of ratio, similarly, at least the early work we've been doing doesn't really suggest that that pops out as something that may be as relevant here. I think the more interesting question scientifically is what's happening in the tumor and is there anything that sort of predisposes the responding patients to have some type of immunosuppressive phenotype within the tumor that may relate to higher levels of certain MDSCs.
That's really what we're focusing on now through our gene expression and other analyses, because that would be obviously really exciting and interesting to tie back the tumor phenotype to the circulating immune profile, and that could help explain some of these data.
Okay, great. Thank you.
Yeah. Madhu, one other comment I would just make, we're happy to send you the reference. As we noted in the prepared remarks, there's a paper in "Nature Medicine" where they essentially took an unbiased approach Looking at a whole slew of different markers that might correlate with, we're able to predict response both by response rate and PFS. If I recall correctly, that paper was in melanoma, they identified these exact same cells. They have some discussion in the paper about why they think this is a marker. As you've noticed in our press release, Dmitry has, who's sort of an expert in these myeloid lineages, sort of feels like this may be an indicator of, as Peter said, the immune response is on. Your foot's on the gas, but there's brakes that are turning it off.
When you take those brakes away, you can allow the immune response to kick in. I think there's clearly interest in this particular marker. It's been described previously in the IO field, and we find it very interesting that it seems to be predicting clinical benefit in our lung cancer population.
Okay, great. Thanks.
Our next question comes from the line of Hartosh Singh from Oppenheimer. Your line is now open.
Great. Hey, thanks for my question. I just had one question on non-small cell lung cancer and one on melanoma. On non-small cell lung cancer, I believe it's the KEYNOTE-189 data, Briggs, that really has changed what'll be probably frontline therapy, PD-1 plus platinum-based chemo, pemetrexed, as recently presented. For your non-small cell lung cancer cohort, can you just specify what % of patients actually got PD-1 and chemo together? How much was platinum-based? Also, there's a little confusion because your abstract says PD-1 experience, but your press release says PD-1 experience for chemo. If you could just break down those percentages. I've just got a follow-up question on melanoma. Thank you.
Yeah. I'll let Michael Meyers answer that question.
All patients in that cohort had received chemotherapy and PD-1. They, however, did not receive them in combination, and that was due to the timing of the KEYNOTE-189 trial relative to ours. There's actually no thought, however, I don't think that combination is necessarily going to yield different results in this population than sequential use of chemo plus PD-1 in either direction.
Great. Thanks, Michael. Just on melanoma, it's not going to a phase III right now. Is that more a function of just the rapidly changing kind of landscape and kind of taking a step back and figuring out what's the best ROI in terms of going forward? Or is that more so seeing some sort of signal in potentially a sub-cohort in melanoma, and waiting for that data before moving forward? Or is it a combination of both? Thank you again for the questions.
Yeah. Hartosh, I think it's more the latter. Although just as you guys were probably up last night reading ASCO abstracts, so were we. I do think that there is the bar for melanoma treatments and PD-1 refractory melanomas may evolve as other agents present their data. We'll do our own homework on sort of what's coming. I think before the abstracts were presented, though, we had already made a decision based on your second comment that we'd like to sort of let the data mature, do more of this biomarker analysis work. I think the questions others have asked of if this biomarker works in melanoma the same as it works in lung cancer, and the cutoff is roughly in the same range, then it would make sense for us to run both those programs in the same way.
That's why we want to sort of wait and see a little bit more data.
Great. Thank you, Briggs.
Our next question comes from the line of Christopher Marai from Nomura. Your line is now open.
Hi, good morning. Thanks for taking the questions. I was wondering if you guys could help tie this a little bit to entinostat MOA. Previously, in studies, you've seen a significant increase in the HLA-DR expression on CD14 positive monocytes, presumably caused by entinostat. I'm wondering if that's something you're seeing here. Are these going sort of from negative to positive with respect to the HLA-DR expression? Is that durable? Any indication that that reverses in patients that respond to the entinostat combo and then go on to progress? I don't know how much data you have there. Then perhaps, tie that into some of the mechanistic understanding that you've tried to put together with respect to the IO mechanism and some of the data you've seen in breast cancer. Thank you.
Peter?
Yeah. Thanks, Chris, for that question. It was sort of one of my first thoughts as well. The data you're referring to comes from our sort of breast cancer phase II data, where we did show in a subset of patients their circulating cells, there was an increase in HLA-DR in the entinostat combination treatment versus placebo. The answer, we don't have that data yet. We are definitely looking at that. All I can say is baseline sort of HLA-DR high expression appears to be important as part of these phenotypes of these monocytes. Clearly it's an important question, and we are working on that. Relative to mechanism, we can't rule in or out the direct effect of entinostat on the monocyte population versus this idea, which is sort of attractive, that these are all patients that had prior treatment with PD-1.
The bulk of the patients did not respond to their prior PD-1. Most had stable disease or progressive disease. Despite that exposure, they still have these high monocytes, suggesting that their immune system is somehow engaged and yet not sufficient to get to a full antitumor response. We think that addition of entinostat somehow is providing that extra boost that may be needed, and it's up to us to kind of prove that. Either we can show it through looking at the baseline tumor phenotypes, suggesting that there's something else relative to the immunosuppressive environment that may be more in line with entinostat's mechanism or as we look at these pre and post or on-treatment biopsies, both from the melanoma and lung as well as the blood, sort of in a longitudinal way, it could help us understand that way relative to the mechanism.
Obviously, our goal is to tie this to some entinostat-related mechanism, and so we're working on that. Your question's sort of right on as to what our interest is as well.
Okay. Just in terms of any indication of what happens to these monocytes following patients responding, then progressing on the combo.
Yeah. We're just in the process of looking at that data. Don't know the answer to that yet.
Okay, great. You alluded to this, maybe walk us through potential steps. Thinking about the future, it sounds like there would be an opportunity to potentially desensitize patients to PD-1, since you had non-responders in the trial now responding. Walk us through maybe some thoughts to bring this combination into earlier lines of therapy. Thank you.
Yeah. Maybe I'll let Dr. Meyers give his initial thoughts on that one.
Yeah. Certainly we've talked about doing a randomized trial in melanoma in front line. That would be a combination of PD-1 or PD-1 CTLA-4 plus entinostat. Obviously, we couldn't do a single-arm trial in that indication because of the activity of the checkpoint inhibitors on their own, and it would be very difficult because of patient selection criteria to understand what a signal was. That's certainly been something that we are considering actively doing. In terms of lung, we have the preliminary results. I think there, clearly the landscape has changed, and we would look at combinations of either expanding our cohort 1, which we've said we would not do at this time, but is always under discussion or even further adding entinostat to a combination of chemotherapy plus pembrolizumab. I think both of them are very valid approaches.
I think we also know that there is a population in non-small cell lung cancer of PD-L1 low, where we have activity and where the perception is that the checkpoint inhibitors alone are probably not as efficacious as they are in high PD-L1. That would be a fertile ground as well.
Just since you touched on it, in terms of the combo of chemo, any indication from this data that patients who had prior chemo had a sort of different outcome from the combination? That's my last question. Thank you, and congrats on the data.
Well, as I said, patients on the trial received prior chemo. Most of them received chemo first line and then received the PD-1, PD-L1 inhibitor second line.
With response to their
There's no indication
their chemo is there.
Yeah, there's no indication that the response rates were different.
Okay. Thank you.
This concludes our Q&A session, I would now like to turn the call back to Dr. Briggs Morrison, Chief Executive Officer, for any further remarks.
Great. Thanks very much, everybody, for joining. The purpose here was just to sort of bring you up to date on what we'll be presenting at ASCO. Thank you so much for all of your questions, and we look forward to seeing you in Chicago, where we can continue the discussion. Thanks very much.
Ladies and gentlemen, thank you for your participation in today's conference. This does conclude the program. You may disconnect. Everyone have a great day.