MacroGenics, Inc. (MGNX) Earnings Call Transcript & Summary

Okay. Good afternoon, and welcome to Barclays Global Healthcare Conference. My name is Peter Lawson. I'm one of the biotech equity analyst at Barclays. I just wanted to thank everyone for taking the time out in the day today. You can e-mail me with questions, [email protected], or you can always ping me on Bloomberg. Gives me a great pleasure to introduce the management team from MacroGenics. So Scott Koenig, thank you so much.

I guess my initial question is always around the differentiation in the oncology space. So maybe you could talk through what you view as the point of differentiation for the company and kind of the key strengths.

Well, thank you very much, Peter. It's always a pleasure speaking with you, and thanks for the Barclays for hosting us today. We're in a wonderful position now in the history of the company. This is moving past our 20th year in operation. As you know, we are a fully integrated biotech company, doing everything from early discovery of targets and platforms through early preclinical, late-stage development. We have commercial manufacturing facility. And as you know, in December, we had our first approved product that we developed from inception, called MARGENZA, which is getting launched literally in the next week. The company has built from ground up its core platform technologies. Our Fc engineering technology that's incorporated in 2 of our molecules to enhance antibody-dependent cellular cytotoxicity killing of tumor cells which can nicely be combined with other molecules. We have a bispecific technology, which we call DARTs, that is being applied broadly, both to combination checkpoint molecules as well as molecules that redirect killing through a CD3-based mechanism. And then we have now 2 molecules using antibody drug conjugate technology. So in total, this is a company that has 8 clinical molecules, of which we've shown activity -- considerable activity in 7 of the 8. The 8th one is still in dose escalation. So we hope that becomes 8 out of 8. And what's very nice also is that 7 in 8 molecules, we hold the majority of the commercial rights for marketing of these molecules, plus -- which you'll hear in the future some very exciting preclinical molecules going forward.

What if we could skip to B7-H3, seems to be a massive topic and ADC as well? So it's kind of everything, a lot of questions around that. So I guess, initially, kind of why B7-H3? What drove you there? And then, I guess, secondly, kind of what derisks it for you?

Beautiful. So Peter, as you know, as we've spoken over the years, this is one of the targets that we've held in high esteem from working now over 12 years on this molecule. We had observed very early on that from various immunization strategies, we had huge numbers of antibodies, irrespective of different tumor types that we employed. As -- this molecule was originally discovered in 2001. Over the years, it's been shown in the literature to be an extremely important molecule as a predictor for clinical outcome where patients who have high levels of B7-H3 tend to do much poor than those with low levels. It has been shown to -- it's a ligand in the B7 family. So just to put in perspective, it's structurally similar to PD-L1. It's viewed to have immunomodulatory effects in inhibiting T cell function. It's also been noted that in addition to poor clinical outcome, there is increased association with metastatic disease. What's also very nice about this is that if you pick the right epitope to this target, you see very little expression on normal tissues. And in addition to dividing tumor cells in the tumor niche, you'll also find it on many of the endothelial cells aligned with the vasculature of the tumors. And so when we conducted our own studies of over 1,500 tumor specimens, we saw a high expression on most solid tumor types. And so years ago, we've now launched into the development of multiple molecules to this target. We have a molecule called enoblituzumab, uses the same exact Fc engineering that we have in margetuximab. That just started a study in combination with our checkpoint molecules in front-line head and neck. In patients getting PD-L1, they will get -- that are PD-L1 positive, they will get our anti-PD-1 license insight called retifanlimab. And those individuals with PD-L1 negative will get our tebotelimab, our PD-1 LAG-3 bispecific. So that's our first molecule target, the B7-H3. Historically, we also had a molecule that was a DART bispecific for CD3 redirected killing. We have no longer pursued that because as we got data on the ADC, we thought that at the time of the development, 2 of those that were showing activity were sufficient for us to move forward. But we did see a number of different tumor types have responses to the CD3 redirected B7-H3 molecule. So again, encouraging that targeted B7-H3 shows validation of why to pursue that target. So getting back to your question about ADCs, a number of years ago, we made the commitment to actually look at other companies' technologies using various linker/toxin technology. And we ultimately decided to pursue a linker/toxin produced by a company called Synthon, which is now called Byondis. They were developing a HER2-specific molecule. The toxin is a cleavable linker called duocarmycin. It alkylates DNA and promotes interruption of cell division and promotes apoptosis of the cells and infects mitochondrial function in those cells as well. Being a cleavable linker, as those cells die, the toxin will be released in the local environment so that even B7-H3 cells that are not expressing it and that are associated with the tumor would be killed as well. So it had all the characteristics that we wanted. So following that through, as you know, we started a dose escalation study a little over a year ago that we presented our first data in that dose escalation at ASCO last year. We dosed at that point at 2 -- up to 2 to 3 mgs per kg in an all-comer study. And what we were very pleased to see is that of 7 patients who had late-stage castration-resistant prostate cancer, 5 out of the 7 patients dosed either at 2 mgs per kg or 3 mgs per kg had a greater than 50% reduction in PSA levels. And we also saw some anti-tumor shrinkage in a number of tumors as well. As you also know is that of those 7 patients, 6 of 7 only had bony disease, only 1 was evaluable by RECIST criteria. And that particular patient who had a 50% reduction or greater than 50% reduction, also had a 29% reduction in the tumor volume by RECIST. So what we decided to do at that point was to look at the pharmacokinetics of the molecule. We found that both at 2 and 3 milligrams per kg, we saw very favorable pharmacokinetics. And so rather than finishing out the dose escalation and given that we saw manageable toxicity at both 2 and 3 level, we said we would move into expansion cohorts. And so we decided to expand into 40 patients with castration-resistant prostate cancer. And then add 2 new cohorts; 20 patients worth of patients with late-stage non-small cell lung cancer and 20 patients with triple-negative breast cancer. Because we had to put amendments into that -- for that protocol, we didn't get that go ahead till that October. So the first patient started to get enrolled in the prostate patients in November. Christmas came, we didn't enroll much in December. And then in January, things started up again, and we've got a nice enrollment rate this year. And the same thing in terms of the lung cancer cohort. And we've enrolled also several patients in the triple-negative breast. So because this is obviously a target of interest and the program of interest, what we made a strong effort to be able to provide updates on a very frequent basis. We did put a placeholder abstract into ASCO. And I should also note that after ASCO last year, we've completed our dose escalation in that we dosed up the 4 mgs per kg, so we have some patients that were treated in the second half of 2020. Several of those patients are actually still on treatment. And what we included in the abstract is data on that cohort as well. So if the abstract is accepted, we will -- and presumably, it would be a poster given the limited new data incorporated. We would include the 4 mg per kg dose and then whatever patients that we could evaluate on a reasonable basis over the first couple of months of this year. So -- and then we are -- we'll be providing updates later this year. ASCO is not the only meeting that's being held this year, so.

Okay. Good. There's a lot of questions around those, really interesting. So the placeholder abstract, you kind of mentioned you've kind of gone up to that higher dose level in the second half. And those patients are still on treatment as of now?

Some of them are, yes.

As of now kind of thing?

As of now, some of them are. Yes, I can't -- I don't know the exact numbers, but I know some of them are. Yes.

That's interesting. What kind of durability do you want to see to kind of convince KOLs. I know we've had KOL calls. Just curious about what you're hearing as well and what you want to see in that data?

My answer is more is better, right, so -- ultimately. But -- so now we're speaking about prostate because given the broad expression pattern, obviously, we have great hopes that this will be very useful in treating lots of different tumors, solid tumor types. So let's just focus on the prostate story, which is, if I look at late-stage prostate patients and the history behind that, first is the PSA effect, then is the radiographic PFS and then obviously, overall survival. If you look gleaning data from a similar line of therapy, which is [ Carbo Taxol ] studies, they were seeing approximately 30% -- I think 37% overall PSA50 reduction. I would say that -- I don't remember the exact number depending on the study, but they were looking at maintaining that for several months, somewhere around 3 months plus or minus. Clearly, that would be the minimum bar going forward. Obviously, if we hit a higher percentage, you have tolerability for shorter periods of time. But if we can see more than 30% -- 40%, 50% -- I mean, 50% would be fantastic -- of a 50% reduction. And then again, if we can see up to 6 months, that would be terrific. And then beyond that, even better. And of course, what we would like to see is associated effects on tumor reduction? And then -- and ultimately, then overall survival benefit as well, so.

Got you. Okay. I guess the other question I have around that is we've -- I guess, Daiichi has moved into that space, Y-mAbs to some degree, and we're kind of also hearing Xencor is one of the unlabeled products. So just what else do you look at that -- in that space and think this helps derisk going after B7-H3?

Well, as you know, one of the biggest questions we were asked all the years is why are you working on B7-H3 and nobody else is. And so now we've got a lot of people working on it. So in a way, people say, "Oh, maybe they were right." I know we were right because I knew we were seeing responses from both the enobli and the redirected killing. And then when we got the ADC, we know that there's a good target. So with regard to Daiichi, I think it's nicely validating what we've heard, and this is all second, third-hand anecdotal from research spokespersons at the company without formal presentation of data, is that they have seen response. They've dose escalated. They haven't hit dose-limiting toxicity. They had dosed up to what we heard is initially 8 mgs and then we heard 12 mgs. And more recently, we heard higher. And we heard that they had some objective responses in head and neck and in esophageal cancer and that they were going to do expansion cohorts in those and maybe lung cancer as well. So again, back to the point, for enobli, for instance, we had some terrific responses that we showed at SITC 2.5 years ago in late-line head and neck patients. We saw a 33% response in combination with pembro, which is much greater than the standard of pembro and NIVO in that. And as you know, in frontline, head and neck, you have about a 35% response rate that got the approval of pembro with chemo on the front line. So again, in that setting, you hear Daiichi has an effect. We had this effect with enobli. It's a nice validation going forward.

Perfect. And of the expansion trials you've opened, why triple-negative? Because I think from the ASCO data you were seeing non-small cell lung cancer kind of signal...

Yes. We saw 2 patients with non-small cell lung cancer, both had tumor reduction. We had no patients with triple-negative breast. We had huge number of breast cancer specimens that we stained for -- by IHC. And this was one that had a very nice penetration with high expression level in the majority of tumors. We're talking hundreds of specimens we looked at in triple-negative breast. On top of that, as you know, we're launching in breast cancer for HER2-positive with MARGENZA. So it's nice from a commercialization strategy to be there. And as you know, the data is not overwhelming with the checkpoints plus nab-paclitaxel in frontline -- in first-line metastatic. You're getting an effect there in a PD-L1 positive population. So there's need for additional molecules. Obviously, sacituzumab is now out there for the Trop-2 but there'll be needs for additional agent populations.

Yes. Why do you think B7-H3 works? And why do you think it kind of -- at least initially, it seems, works really well in prostate?

The prostate is serendipitous. Turned out the first patient that we had responses was at 2 mgs per kg was a prostate patient and that investigator had a lot of patients and had no alternative to give them. There's not a lot of alternatives in late-stage prostate cancer. So we kind of put more patients on and that's how we got to this result. So again, I see this as just an entry point, and it happened to be serendipitous. So another tumor came in. We'll be talking about another tumor at this point.

So you don't find that B7-H3 lights up more significantly in prostate?

No. It's actually probably sort of in the mid-range. We see probably higher expression on head and neck cancers. But across the board, as -- if you look at our data set, 75% to 100% of the solid tumor specimens, whether it's a common tumor or a rare tumor has some degree of B7-H3 expression. And then you look at the grade of expression, this one is sort of -- prostate is mid. Having said that, we have done historically in collaboration with Johns Hopkins Group, we did a study in neoadjuvant prostate cancer that's actually still ongoing following those patients. And we saw very good B7-H3 expression on newly diagnosed patients as well as late-line metastatic patients. So as opposed to many other markers, this is one that doesn't drop. If anything, it's stable and may goes up a little bit, but a high percentage, 80% to 100% of the patients have B7-H3. And that, in fact, is probably higher than PSMA, which has, obviously, been always the historical standard for marker expression.

How clean is the expression on tumors versus nontumors?

So it varies in terms of which epitope, which antibody you have. There are a lot of dirty -- so if you look at the literature, a lot of people were dissuaded, because they thought, "Oh, it's expressed in a lot of normal tissues." That's not the case. As I told you, when we started out, we had a large array of antibodies that we selected and actually ultimately picked 2 different variable domains, one for the antibody alone or with the Fc engineering, as well as the redirected DART molecules. And then another for the ADC because in the ADC, you want that to be incorporated into the cells. And those were particularly picked because they had a particularly cleanest pattern on normal tissues. So again, with dirty antibodies, you can sometimes see higher level of expression in liver. You'll see in other tissues as well. We saw a fairly clean pattern. We see very -- the highest expression on the membrane was a -- I think that we had a 1 plus in the adrenal cortex. We had a 1 plus in the pituitary. You can see it in activated dendritic cells, 1 plus there in those cases. So pretty modest expression. I mean, to compare it to, it's even a less expression that you see, for instance, on HER2 on normal tissues. So that's kind of the range. And to get -- now you can get -- you can have dirty antibodies and see more expression on other things, but that's sort of gives you a feel for what we saw.

And when you refer to dirty antibodies, that's just hitting a different epitope. That's how you kind of...

Epitope that might, of course, react with other things, just -- that's the way I refer to.

Okay. And is B7-H3 in any way kind of mutated? Is that what you think you're hitting with an epitope?

No. There's no evidence of mutations. And it's a -- seems to be a very important protein phylogenetically and ontogenetically. It's -- if you go look down all the vertebrates and actually get to invertebrates, there's huge conservation through all the vertebrates in terms of the sequence. So it's been maintained for millions of years. It's probably very important in a developing embryo because if you knock out expression in a vertebrate is to do a mouse model, those embryos get killed. So it's -- it may ward off the immune response against that. That's one view of the mechanistic value of it. And as you go into an adulthood, it seems to downregulate expression. Now you'll see transcripts in virtually all cells. You will see protein in the cytoplasm of many, many cells. But the tumor seems to up-regulate, and there was a paper a number of years ago, from the Memorial Sloan Kettering Group that showed that a MicroRNA29 was important as a regulator, and that was inhibited and allowed the up-regulation on the surface in tumors as one mechanism for regulation of control of the protein.

Thank you. Really -- we have a really interesting discussion. We've got 1 minute left. So we have to talk about the rest of the pipeline in that.

So you want to pick a favorite tackle? Or should I talk very fast about the other 7 molecule?

No, I think we've covered a great area. So I wanted to just thank you, Scott. It's been a fantastic conversation that whets everyone's kind of appetite to talk more about, deeper in the story, but we've hit that kind of 25-minute mark. But thank you so much, Scott. Much appreciated. Thanks for being here virtually at least at the Barclays Healthcare Conference.

Yes. Thanks.

And I hand it back to the operator. Thank you so much.