Bristol-Myers Squibb (BMY) Presents at Goldman Sachs 38th Annual Global Healthcare Brokers Conference (Transcript)

About: Bristol-Myers Squibb Company (BMY)
by: SA Transcripts

Bristol-Myers Squibb Company (NYSE:BMY) Goldman Sachs 38th Annual Global Healthcare Conference Call June 14, 2017 1:00 PM ET


Thomas Lynch - Chief Scientific Officer


Jamilu Rubin - Goldman Sachs

Jamilu Rubin

All right. Thank you, everybody for joining us and what I think will be a really exciting session with Dr. Tom Lynch, the Chief Scientific Officer of Bristol-Myers Squibb and especially coming off what was I think an exciting ASCO 2017 just last weekend. A lot of questions around Bristol-Myers' comprehensive program.

So, great opportunity, very timely to have you here Tom. So, thank you very much for making the trip. I know you wanted to open it up with some opening comments first.

Thomas Lynch

Jamie, thank you very much for the chance to speak about Bristol-Myers Squibb. I could not be more excited to be the Chief Scientific Officer of BMS. I think this is an extraordinary time in medicine. It's a transformative time in our understanding of diseases and our ability to bring treatments to patients.

So, when I think about research and development at Bristol-Myers, I think of several important priorities. Many of these were shown at our ASCO meeting this year, but first, we're going to work hard to develop Opdivo and Yervoy as an important combination therapy for patients with cancer.

Second, what you saw at ASCO this year is the emergence of a pipeline, a pipeline that's incredibly rich. A pipeline that has multiple new agents in it that are able to impact cancer. For example, LAG-3; for example, IDO; for example, anti-KIR; for example, GITR; example of four new targets that we think are important in drugs that can make a difference in how we approach cancer.

Third, we are committed to looking at combinations and developing combinations that can impact cancer. Fourth, it's important remembering that we're not just a cancer company, but building and developing outside of cancer is important. Fifth, I think it's importance of translational medicine and remembering that we've got to continue to develop biomarkers to better understand, which patients are going to benefit most from our medicines as we go forward.

So, I also want to make a couple of comments as well about lung cancer since I know that's the top of mind of many people in this room. I'm a lung cancer person and I basically have devoted my career to treating patients with non-small cell lung cancer doing trials in non-small cell lung cancer and particularly thinking about translational medicine in non-small cell lung cancer.

I believe strongly that Bristol-Myers Squibb has a dynamic program in non-small cell lung cancer. We have both the 227 trial and the 568 trial give us a great number of options of which to approach this disease. We have a broad and diverse program of monotherapy of I-O therapy, of I-O chemotherapy in patients with non-small cell lung cancer, and so I think that that program is well positioned at this point.

I want to specifically just address the question of Study 12, which I get asked a lot about in term of the importance of Study 12 and what my thinking is about Study 12 that was reported at ASCO this year, and I look at Study 12 as a lung cancer expert and I see a study that showed that both monotherapy and combination immunotherapy is able to deliver two-year survivals, which compare very favorably with historical chemo controls, and in my opinion, that study supports the design of Study 227, which is essentially the trial we were comparing in expressors and non-expressors comparing our drugs to chemotherapy.

I think it's very difficult to draw many more conclusions about Study 12, and particularly when asked about the differences in the high expressor groups that were looked at and the people with more than 50% PD-L1 expression and again 12 patients, 13 patients in each arm, I think it's very, very difficult to make comparisons between those two sections of the trial, given the incredibly small numbers.

So, my take on 12, it supports the design of 227 and 227 gives Bristol-Myers many opportunities to see our program in lung cancer evolve.

Jamilu Rubin

Okay. That was a great, great opening. I may just take the opportunity of you being here and addressing sort of specifically some of the issues, the concerns that are out there. You did have a very comprehensive presentation at ASCO, but as you know investors had a hard time getting past CheckMate-012.

And I think you have to understand in the context of the failed CheckMate-026 study and your decision of not to go ahead with the accelerating filing strategy, investors left the conference less confident about CheckMate-022 -- CheckMate-027, sorry, and your position in mind.

So that's just understanding context. So, let me give you another chance to explain what happened, clearly very small numbers, but the question is why did the OS survival curve fall off so much in the over 50% expressors from one to two years, I think with 60% OS and that compared to 75% OS in the mono group.

So, when I look at the data, why would I needed the combination when I can get a 75% response from the mono group?

Thomas Lynch

So, Jamie, I'll say listen, when you got 12 and 13 patients, 12 and 13 patients, all it takes is one or two patients to have a survival event to switch those curves around, one or two patients can change by 20% to 25% depending on where they are in that curve, can change the way those curves look.

And so, I just think you can't positive or negative drive any specifics in terms of comparing 12 and 13 patients, it's just not statistically sound, and so I think that is the problem. Remember, survival incorporates much more than PFS.

Survival looks at deaths from many other causes as well, and we don't know what all the contributing factors were at that point. So again, when the numbers are that small, I just don't think you can make that conclusion.

Again, getting back to the concept of what we're doing in lung cancer, we're looking at nivolumab as a single agent, Opdivo and Yervoy together as a combination, nivolumab combined with chemotherapy, both in expressors and non-expressors in all comers.

So, I really believe Jamie that what we have is a program in lung cancer and I just don't think you can read that much in Study 12.

Jamilu Rubin

If you had shown a much higher survival in the combo versus the mono, would you have said to us don't pay attention, these are small numbers?

Thomas Lynch

I really hope I would because if I didn't, I wouldn't have any credibility.

Jamilu Rubin

So, the results are interesting because in all three patient groups, we saw strong PFS in all three sub patient groups, but it didn't translate into higher overall survival and our understanding, at least my understanding of PFS is that it's a pretty good predictor of overall survival. Is that not correct?

Thomas Lynch

So, Jamie, I think PFS is an important endpoint in cancer in general. I think as we look at the difference between PFS and OS, when I think back when I was running the Fellowship Program at Mass General and I would teach my fellows about the difference between PFS and OS, sure we would love to see every trial be OS based.

But we know that the market and the world, the way we treat patients has changed and that PFS really becomes the endpoint that's much more related to the activity of the actual drug at hand.

When possible, we love the design trials for OS, but there are so many treatments that people get after they have whatever the indicated treatment is that I believe PFS will increasingly become important.

Look at breast cancer, in breast cancer you almost never see a drug now approved on overall survival. It's almost always on PFS because of that.

Jamilu Rubin

So, we also saw that six week dosing of Yervoy, this is back to 12 performed worse than that 12-week dosing of Yervoy, yet CheckMate-227 uses six-week dosing. So, can you explain why that happens?

Thomas Lynch

So, I'll say a couple things. I think we believe as we develop the Opdivo-Yervoy combination, and I'll remind you and Jamie you know this extremely well is that Opdivo and Yervoy, the only I-O, I-O combination to show a survival benefit in patients with cancer.

So, in melanoma, we've shown that the combination of Opdivo and Yervoy together, it gives you a survival benefit. So, our feeling in putting together the combination was to put together a combination that will give CTLA-4 enough of our presence and enough drug treatment to be able to contribute to the activity in that setting.

I think it's again very difficult to compare the 6 to 12 they were accrued at different time points. There were some imbalances in terms of the types of patients that enter at different times of the study.

So, I believe that the Q6 regimen is a great regimen to go forward with. I don't believe from 12 you can necessarily say that 12 weeks might be better.

Jamilu Rubin

Okay. The comparison for CheckMate-227 will be chemo not monotherapy, but even chemo and both I think Keynote 24 and Keynote 21G performed better than what we have historically seen with chemo.

And I think that Bristol-Myers had cited as the historical OS rate of something like 15% to 20%, but it seems that we're seeing survival which looks much better than that. And if memory serves me right, I think overall survival for chemo and one of the Keynote studies was around 55% and after 18 months was 43%.

So that’s a higher bar. So, I know that you said CheckMate-227 is compared against chemo, but how do you think about chemo as a comparator and just given how high the bar has become?

Thomas Lynch

So, Jamie again a great question and one that I'm really happy to get because as a lung cancer doctor struggled for years to get two-year survivals of more than 10%. So, seeing these kinds of numbers that we're seeing in these trials now, where we're getting two-year survivals of lung cancer of 30% and 40% I think is terrific.

But yet again you have to remember when you're looking at two-year survival in lung cancer with a group starting in chemotherapy, many of those patients then crossed over to get immunotherapies.

So, the impact of second line I-O treatment, I think is very profound and so -- and remember in different trials, some trials mandate crossover okay, so a trial like 26 mandated crossover. Other trials crossovers’ active selection are the discretion of the investigators.

So, for example in 227, we don't mandate crossover in the United States if you're on the chemotherapy arm, we would imagine that most American oncologists will choose a PD-1 drug in second line.

In Europe, in many places I think they will as well. I think few will choose to use the combination since it's not done as often in second line, but I do think when you look at overall survival, the impact of crossover becomes very challenging when you look at that and so I think that that's something which is again a good thing for patients overall. We want to see all these arms do as well as they can.

Jamilu Rubin

But even PFS that we saw in Keynote 24 for chemo was about I think 8.9 months, so it was pretty high.

Thomas Lynch

That's a great point. When you get PFS' that are high, to me what does that show? It shows the impact of patient selection. You selected a group of patients who are more likely to benefit from chemotherapy or from I-O therapy, good performance status as a way of selecting a better group of patients.

Jamilu Rubin

So, have you learned those lessons from 26 where there was poor patient selection? Can you -- Keynote 21G was an all comers patient population. So, they weren’t higher expressors who were selected. The PFS from chemo was 8.9 months.

Thomas Lynch

Keynote 21G was in patients who had non-squamous lung cancers. So, it's a different type of lung cancer, and I would argue that in 26 because that's the study where there is probably the most discussion, biggest difference is we did not -- we selected a group of patients, we analyzed by 5% instead of 50% okay.

And so, I think these are your biggest difference there, and I said this at the ASCO Investors Meeting afterwards. I said it in an interview for Forbes Magazine, and I think that when you see differences like this, it's almost always in patient selection. And the one thing we know about the groups is 50% selection is different than 5% selection.

Jamilu Rubin

It seems that chemo combo is working. We've seen that certainly with 21G. There will be confirmatory trials that we put out later this year. Bristol’s set was that IO-IO would deliver deeper and more sustained responses. Yet again, we seem to be seeing deep and sustainable responses with chemo combos with Keynote 21G showing a trend towards overall survival.

So, has your thinking started to change around this and I know you've added a chemo arm to 227, but what are your observations about the data that we've seen in chemo combos.

Thomas Lynch

So, it's a great question and I think couple things. One is I still believe that I-O, I-O combinations have the potential to deliver excellent duration of response. I am hopeful and I don't think it's proven yet that chemo combos will necessarily deliver the same duration responsible.

We won't know that till we see the results of our large study of -- Merck's large study of, Roche's large study. The truth will come out in terms of what the duration of response is, but I think the thing that's made I-O, I-O so appealing in other diseases such as melanoma, which is where we have the most data is the duration, the durability of the responses.

Now in lung cancer, there is good reason to believe the chemotherapy plus an I-O agent may have the unique profile and I'm very excited about this. It's something which we have. We have chemo combinations with Opdivo.

We also have in trial 568, a combination with Opdivo, Yervoy and chemotherapy together, a quad combination. So, we believe fully that chemotherapy plus I-O makes good sense as does I-O, I-O and that's why we do the trials.

Jamilu Rubin

So just last on CheckMate-012 and then we're going to move on, but what is to develop that trial that gives you so much comfort or conversely what sort of questions does it raise in your own mind? And is it even fair to compare these two groups, the combination versus the monotherapy?

My understanding was that these patients were enrolled at different times, but what is it about that trial that makes you feel so good about 227 or conversely what raises the question.

Thomas Lynch

So, Jamie the one thing I would say about 12 is remember what Study 12 is. Study 12 is a Phase 1/2 trial that was used to generate data to justify and to inform a large Phase 3 program in lung cancer.

That was why Study 12 was done. Study 12 is not done to compare the different groups that were entered into Study 12 okay. I think that's really important to keep in mind why was Study 12 done. It was not done to make interest from comparisons.

Only the randomized Phase 3 trials are going to give us the definitive answers to know how our drug works and that's true with every company. You need to do large randomized Phase 3 trials to be able to get the kind of answers in a common disease like lung cancer.

And so, what I find in Study 12, I find that Study 12 supports the design of the Study 227 and I think that's very clear in my mind.

Jamilu Rubin

You weren’t at the company when -- you were on the Board, but you were not Head of Research when decisions were made about the design of CheckMate-026. So, you have some healthy distance from that. So, I just want to ask you, I think people are still kind of confused as to why that trial failed so miserably and you talked about the 5% versus the over 50%.

But even when you look at subs group analysis, because patients in the over 50% did not show a benefit over chemo. So, what do you think went wrong with that trial and most importantly, are you confident that the lessons learnt from that trial's failure has been incorporated into 227 as the risk as you can possibly make it?

Thomas Lynch

So, Jamie it's a great question. I get asked about 26 a lot and my thinking about 26 is the following, this is our strategy of looking at a expressor population, particularly a 5% expressor population resulted in a group of patients who didn't benefit from monotherapy in that setting and looking at the 50% expressor group did benefit from monotherapy in that setting.

And so, I think that's the single biggest difference in that trial and remember 227 was designed after -- well before the result of 26 were known and so I think that in my mind those are the two big differences between 26 and keynote 24 which Merck ran.

Now we looked at a number and as a lung cancer doctor, when I think about what other kinds of differences we have looked up and down, we have looked at histology, we've looked at potential common use of radiation, we've looked at other drugs that were used at the same time, we look at dose intensity schedules, we looked at all these other areas and we really haven't been able to find anything that could clearly explain this except for the 5% and the 50%.

The one area that I was talking to a key opinion leader at ASCO about could there be differences in the timing of the biopsies and some biopsies are biopsies we could use by the older archived tissues and then the Merck trial could use, could that introduce a difference? Possibly could, but at the end of the day, you can't prove that. I can't prove that. No one can prove that.

At the end of the day it's 5% and 50%. In my mind, that explains the difference.

Jamilu Rubin

For CheckMate-227, are you relying, are you relying on your internal patient selection? Are you doing that yourself inhouse or are you relying on a CRO and I think and correct me if I am wrong that with CheckMate-026, the patient selection was done by an outside vendor?

Thomas Lynch

What do you mean by patient selection?

Jamilu Rubin

Choosing patients for the trial, enrolling patients for the trial.

Thomas Lynch

So, enrolling patients, the choice of patients for a trail comes from the investigators. So, the investigators are the people who choose the patients for the trial.

Jamilu Rubin

So, the CRO has nothing to do with it.

Thomas Lynch

The CRO helps process, helps us with the data management and the study design, but the doctors are the ones who choose which patients go on a given trial and many sites have several trials open.

Jamilu Rubin

You've made it very clear that you're very committed to CTLA-4 and I just want you to, there were a number of abstracts and posters and presentations at ASCO in melanoma brain meds and high colorectal cancer patients and others that did show a convincing higher response with the combination versus monotherapy.

But at the same time, I think there are as many questions around CTLA-4. 067 did show an overall survival benefit versus Yervoy and not versus Nivo. So just if you could kind of why are you so committed to CTLA-4 and it was interesting that you introduced at the Analyst meeting, two new CTLA-4 assets that might be less toxic than Yervoy.

So, if you could kind of talk about your confidence and mostly why it's still so interesting to you?

Thomas Lynch

So, Jamie that's a great question and I'll start off by saying couple things. Fundamentally I am medical oncologist, I am doctor who takes care of patients. There is nothing that drives good decision-making more than survival and for seeing a combination of drugs or treatment approach that improves survival and we have evidence in melanoma using two I-O drugs together, Opdivo and Yervoy together improves survival and improves survival compared to the standard of care when the trial was designed, which was Yervoy alone.

So, the combination improves survival but I'll also point in 67 that we get a trend towards survival benefit compared to Opdivo alone. The study was not powered or designed to look at that and we also saw benefit in the patients who had BRAF mutation, which we think is important as well.

So, I believe that that survival signal is important and plus Jamie as you just said, in a number of other tumor types, we're seeing higher response rates. Now it's our job as a company to do Phase 3 trials and demonstrate that those response rate benefits translate into PFS and survival benefits in other areas.

So, we're actively working in doing that and you're going to be seeing data coming out over the next 12 to 18 months in small cell lung cancer, head and neck cancers, in hepatoma, in non-small cell lung cancer looking at the combination.

And we saw data as well at ASCO looking at the combination of mesothelioma that shows benefit. So I think that's why asking the question why does Tom Lynch like CTLA-4? It's because of survival.

Point number two, we always should be working on improving the therapeutic index of a drug. I don't care what the drug is. I don't care if the drug is Aspirin. We should be working on improving the therapeutic index.

Really important to think about this. So, with CTLA-4 we have two follow-on compounds. We had the new formulation, which is a non-fucosylated CTLA-4, which we believe has increased potency.

Now we hope that increased potency may translate to a better therapeutic index and obviously trials are going to be -- need to be done to show that. We also have the CytomX compound, which is a CTLA-4 compound that's a pro-body that gets cleaved by the protease and the tumor and therefore may have more antitumor effect and less off-target side effects.

And again, because we believe so deeply in the target we're developing these two additional compounds. So, I think it shows very much that Bristol-Myers is committed to looking at CTLA-4 and as I said, in terms of research priorities, one of the most important research priorities for Bristol-Myers Squibb is to develop Opdivo and Yervoy in many solid tumors across the Board over the next several years.

Jamilu Rubin

So, the next related trial is MYSTIC that's coming out in a couple weeks and it would be probably the first large randomized trial to confirm or challenge the I-O thesis. Is that how you are thinking about it? You're still sitting in my shoes and that trial is negative. What does that tell me about 227?

Thomas Lynch

So obviously sitting in your shoes, your guys jobs are so much harder than anybody else's. I do feel for you. What I'll say about MYSTIC is the following, is that yes, it's looking at the hypothesis of CTLA-4 and PD-1, but it's a fundamentally different trial than what our trial is.

It's using two drugs which are different than our two drugs. So, they're not the same drugs. Second, our trials are much larger trial than MYSTIC is. So, the sizes are different. Third, we spend a lot of time at Bristol-Myers Squibb looking at schedule and dose and sequence and how you do -- how you put these two drugs together.

And so, we really have to see what data from MYSTIC looks like. What our data from 227 looks like and how this evolves. So again, I think it's very dangerous to make these cost trial comparisons and again I know you have to, but I know it's very hard and very difficult to do that.

Jamilu Rubin

So, if it's positive, I should not assume that's a good time for 227?

Thomas Lynch

Well I would think, if it's positive, that would be something that…

Jamilu Rubin

So, if it's negative, it's not -- I shouldn’t lead into it. If it's positive, it's a good sign.

Thomas Lynch

The best way to answer a question like that is to really think about what's right for patients? If it's positive, that's a great thing for patients with lung cancer.

Jamilu Rubin

I think we have a question.

Thomas Lynch

They're coming with a mic for you.

Question-and-Answer Session

Q - Unidentified Analyst

So, one of the questions that has not been addressed, is this whole business about testing for PD-1 expression. So, the immunohistochemistry test from my understanding has a lot of variability to CLIA test from site to site while within the site, it may be okay, but between sites it's all over the place.

So as an anecdote of patient trying to qualify for a PD-1 drug failed to meet the marker in one CLIA test versus going to another where they met it. And recently you're also talking about looking at tumor burden potentially going forward.

So, A, what can you say about the immunohistochemistry as a reliable test for all these trials? And what else can you look at in the future as you do this new trial?

Thomas Lynch

So, thank you for that question. So, I'll break it down to two parts. The first part talks about immunohistochemistry chemistry. So, I believe immunohistochemistry test while they're fairly easy to do, PD-1 is not the only example of immunohistochemistry test that has variability.

Just think back EGFR expression in colorectal cancer and lung cancer and other IHC test, they can be challenging and so there are, there is a fair amount of subjectivity that comes with some immunohistochemistry tests.

Doesn't mean they're not useful because they're useful in many cases, but there is a certain amount of difficulty in interpreting this. The second point is what about other emerging biomarkers and one of the biomarkers we were very excited about it and you mentioned it at AACR we reported data looking at total mutation of tumor mutational burden can be.

So looking at the number of mutations that existed with cancer and we talk this, as Jamie mentioned, we talk about this at our ASCO Investor Meeting where we talked about how increased tumor mutational burden may predict for the presence of neoantigens, neoantigens that could drive an immune response and so as you're pointing out, one of the advantages of a test like TMB, is it's more objective in terms of in fact you get a numeric score of the number of missense mutations present in a given sample.

So, it may be more objective than in immunohistochemistry test. So again, very excited about that target that's not only did we see it in data that was presented at AACR, but a number of investigators including groups at Memorial Sloan Kettering have remarked on the importance of tumor mutational burden.

So, one that we'll be excited to see how this develops over the next several years.

Unidentified Analyst


Thomas Lynch

Yes, so our data was retrospective correct.

Unidentified Analyst

[indiscernible] you're basically going blind going forward.

Thomas Lynch

So, I would argue that when you develop any, so this is one of the key things as I said earlier, the importance of investing in translational meds because we've got to let the science guide us.

We've got to let the science take us where it's going to get us. In the very beginning, absolutely it requires a tremendous amount of work and using different platforms, using different tumor types to determine what are the right cutoffs for tumor mutational burden.

The other issue is what about the advent of liquid biopsies? Can you get tumor mutational burden from a liquid biopsy as opposed to a tumor sample? All really good question and again some of the reasons that we're doing things such as working with GRAIL and working with Foundation Medicine and working with the QIAGEN to be able to develop these tests, to be able to guide our drug development because I think it's incredibly important.

Jamilu Rubin

I think there is a different question.

Unidentified Analyst

I had a question and also had a clarification, but the question first is given the totality of data you have in house and everything else you've seen from the peer group, do you feel more comfortable about the chemo arm in 227 today or do you feel more comfortable about the CTLA-4 arm for high expressers in 227?

Thomas Lynch

So that's a great question and similar to actually the question that Jamie asked, just a few seconds ago, I think both these arms are worth exploring. We need to get answers in terms of how to best treat patients with cancer.

We are going to have a large trial of the combination I-O, I-O combination in high expressers and greater than 1% expressers and we're also going to have a data with chemotherapy in all comers that will come out of our program, our family of trials.

It's not, so Bristol-Myers Squibb, we don't have all our eggs in one basket. We've got monotherapy. We have I-O, I-O. We have I-O chemo and as which turns out to be better, even I suspect, I suspect there is going to be a situation where our doctors are going to sit back.

They're going to look at Mrs. Jones or Mr. Jones and they're going to say these patients perhaps it's I-O alone. For this patient, maybe it's I-O, I-O and for this patient maybe it's I-O and chemotherapy because what we're learning is lung cancer is an incredibly hard disease to treat.

It's not just one disease but many diseases and this was pointed out by the gentlemen earlier, the growth of understanding of the biology throughout driving additional biomarker sets I think we'll continue to make this an incredibly interesting area as it unfolds.

Jamilu Rubin

Tom, I think that investors underestimate the opportunity for Opdivo as monotherapy and frontline lung. Is that still, we've kind of written it off in our models. But do you have the opportunity to show a statistically significant benefit with monotherapy in 227 and the highest expressors?

Thomas Lynch

Well Jamie, that's a great question. As you point out, we do have a monotherapy arm in 227 that we will get. I have to see where the events take us, but we do have a monotherapy arm compared to chemotherapy in 227.

Jamilu Rubin

Okay. Good. Another really important trial that report out later this year is frontline renal cell carcinoma, Yervoy and Nivo; tell us are there any lessons to be learned about renal, is renal cell carcinoma as difficult a tumor to treat in frontline as lung cancer is?

What are the comparisons, what are the differences between the two? How do you feel about that combo strategy?

Thomas Lynch

So, great question Jamie about -- thinking about kidney cancer. So, kidney cancer when I go back to my training I'll never forget seeing my first couple of patients with kidney cancer back in 1992 and I turned to my mentor and said, now what we do here?

He said well, there is a drug called Velban, single agent Velban. So how well does it work? He said the response is about 3% okay. So, things have changed dramatically in kidney cancer. Who could imagine that we would have five or six tyrosine kinase inhibitor, the [indiscernible] tyrosine kinase inhibitors that work in kidney cancer.

Who would have imagined that immuno oncology would have the kind of impact that it's had now, now we did have some signals from I-O2, there were some I-O2 signals in kidney cancer in the past, but now we have this remarkable opportunity with I-O drugs.

And so Nivolumab we've seen to be an effective drug and we again, I am eagerly waiting for the events to occur plus be able to see whether or not there is a benefit to adding CTLA-4 to Opdivo in that setting and again we are hoping for the end of the year, but again it's up to the events.

Jamilu Rubin

You guys seem super excited about the LAG-3 opportunity at ASCO but some investors walked away from your approach or kind of underwhelmed by the 13% overall response rate across the population and especially in comparison to a competitor asset, albeit at smaller numbers by Syndex, which I think showed that 30% who are in the same PD-1 refractory population

Tell us why you're so excited about this? I think the excitement is really due to the LAG-3 expressor population, which we don't really know much about, but talk about the LAG-3 program and why it's generated so much excitement at Bristol-Myers?

Thomas Lynch

So, LAG-3 is important for a couple of reasons. I am glad you brought this up. So, when I think back to thinking about I-O as a concept and you think about PD-1, so you have -- you give an anti-PD-1 agent that drives an immune response.

T cells are designed to become exhausted and so T cell exhaustion is a major feature of resistance to I-O agents. And so, a compound like anti-LAG-3 directly impacts the targeted T cell exhaustion and it has the potential to reverse resistance. So mechanistically I find this incredibly appealing to look at in this setting.

At ASCO, we presented data to show the response rate with our anti-LAG-3 agent, yes it was 13% overall, but when you look at it by expressors of LAG-3 20% versus 7%. So, the biomarker turned out to be very important.

I think other drugs like the Syndex compound, which looks specifically at an epigenetic modifier, yes very exciting as well albeit 13 patients versus a much larger group of patients that we presented in our trial.

You also have to remember that how we define resistance may be different between our sample and other samples and in this second line setting, we just now are learning what the definitions and what the types of patients we're going to see in this area will be.

Jamilu Rubin

And this in combination with Nivolumab or…

Thomas Lynch

Yes, I should say it's essential clarifying, our LAG-3 data is just like the Syndex data. It was in combination with ours is with Nivolumab, there is with Pembro.

Jamilu Rubin

Okay. And what about in the naïve patient population. This was refractory…

Thomas Lynch

So, what we're doing with LAG-3 now, we're expanding it into melanoma in the I-O naïve population. We're looking at a number of solid tumors and we're looking at LAG-3 plus Nivolumab both in the I-O naïve as well as in the prior I-O and we're developing the biomarker to try to get a better biomarker to be able to select our patients better.

Jamilu Rubin

So just one last question. You rattled off a bunch of next assets, the next new waive of assets, LAG-3, GITR, IDO, anti-KIR, Lirilumab, which of those four or five are you most excited about?

Thomas Lynch

So, it's like asking, which of your children do you love the most okay. I think these are all outstanding assets. They're all very early. I like the fact that LAG-3 shows some benefit from prior I-O, but we have a couple of -- at least one GITR response in prior I-O as well.

So, let's see where the data takes us. Let's follow the biology that's always going to be the best way to get there.

Jamilu Rubin

When are we going to find out, is the IDO for this along since I think you announced you're going to start a registrational trial, is that for this along in terms of clinical trial development?

Thomas Lynch

So, the IDO as you know we have two shots with IDO. We have a great relationship with Incyte and we're developing a data with Incyte, IDO plus Nivolumab and we've shown our commitment for the compound to the target by purchasing complexes, what is now the Bristol IDO.

We need a name for our drug as well. So, I hope the people will do naming come up with a good name for it. So yes, I think IDO is pretty far along. It's in registrational plans and we look forward to seeing those data.

Jamilu Rubin

Okay. We have run out of time. Thank you so much.

Thomas Lynch

Thank you.