Moderna, Inc. (MRNA) Earnings Call Transcript & Summary

Well, good morning, everyone, and thank you for joining us as we continue an exciting day here at the 2020 Jefferies Global Healthcare Conference. I don't know if there's anyone or any company more in the spotlight these days with the exciting developments going on at Moderna. So I'm very happy to have with us here the CEO of Moderna, Stéphane Bancel. Stéphane, thanks for joining us at such a crucial time, both in Moderna's history, both with the exciting developments in -- with your pipeline, but certainly, of course, with all the things going on in the world.

I guess I just wanted to start off with the first question or first topic, I'd say. You obviously have quickly evolved the platform and have built a huge pipeline. Maybe just spend a moment talking first since you had a Science R&D Day yesterday, before we jump into the nitty gritty of some of your programs, of course, COVID-19, et cetera. Just talk with us about the platform. And maybe just a couple of the highlights yesterday about the Science Day. That will maybe lead us into why that platform is exciting, particularly for your vaccines.

Well, Michael, good morning, and thank you for having us. Very happy to be here. So yes, Moderna platform basically relies on messenger RNA, which is a very exciting platform because if you think about it, unlike small molecule that the pharma industry has used for 100-plus years, our recombinant mRNA is an information molecule. So it carries the information into the human body to basically then hijack the ribosomes of the human cells to make proteins on demand. What we are sure already in human is we can make human protein in humans. And the best data I will point people to is the publication of a VEGF program by AstraZeneca, which is in nature. You can find the paper on our website. This program is now in Phase II. We can make antibodies, and that's very exciting program we shared last September at our annual R&D Day where we should -- we can basically make a full IgG, a full antibody in humans by coding 2 mRNA, Ab [ and lifegen ], put in the same particle. And in the cell, we make both protein that's self-assemble, and we show we made binding, neutralizing antibodies. And then, of course, vaccines. Now with COVID-19, this is our 10th vaccine in clinical studies where we're basically coding the messenger RNA, the protein of a virus, i.e., an antigen.

So yes, I mean, your technology platform essentially uses -- as I know you like to say, the software of life, but essentially the key technology to make antibodies, to make proteins, to make all of the key proteins that you could direct as a therapeutic. Maybe I would just like to delve into why that's important, of course, because the platform has rapidly, as I think you could explain, gone from bench to human with COVID-19. So maybe just explain in the last couple of months how you've done that and explain briefly what that first data showed and why you're excited about that first data, and then we'll go into some of the key questions about that data for COVID-19.

Sure. So in the first line, I mean, the sequence was available from the Chinese on January 11, 42 days after we shipped the clinical-grade products or GMP products to the NIH. And on March 16, so 2 months after the sequence was online, we started dosing humans. So that, of course, was a record time. I think Fauci has said that the fastest time before from a sequence of a new virus emerging to injecting human was 20 months. So it was 2 months versus 20 months. We started the Phase II last Friday. I expect the Phase II to recruit very quickly because we have all the safety data from the Phase I. We are dosing 50 microgram and 100 microgram, so basically select the dose for a Phase III, which we still anticipate to start in July, which if you think about it will be just again another totally incredible performance by the Moderna team, going in less than 6 months from sequence to starting the Phase III. So the key thing about the data that we released so far, so on May 14, Thursday, May 14, we basically got from the NIH with running the study, a data cut of whatever data they had at that day. We had a conference call with them. And then in the evening, they sent us a PDF with all the data, participant per participant. And what we observed first is that the vaccine at the low and medium dose, 25 and 100 microgram, was very well tolerated. At the 250 microgram, where we had less people because that cohort was started later. So we needed the data cut, we just had less people at that time. Not at the prime, but at the boost. You had a couple of people with fevers, the low-grade fevers, a bit of redness at the site of injection and not at the prime. And that's important to understand because we saw those only at the boost. And also what we saw is the binding antibodies and neutralizing antibodies. And so given we have 10x between the low dose and the high dose, it is not surprising to see at the boost given we -- you can expect we will release the data soon and maybe published, a very high level of antibody that when you go back to hit the immune system in such a short time after that you have a bit of side effect, which the immunologist will tell you is almost a good sign that you're making a lot of antibodies. And we said that this top dose, we will not be using in a Phase II. We will not be using in a Phase III. And the reason we had to pick that dose is we did 2 days of discovery between January 11 and January 13. So we walked into the clinic without the usual several quarters of years of preclinical work in many species, and so we're kind of guessing the dose. And we use the rest of the totality of the data on the rest of the platform, the 9 vaccines before the COVID vaccine to try to find a range. And we said, let's not be too cute. Let's make sure we go high enough in dose because as we're chasing the virus, as you can imagine, if the 25- and 100-microgram dose will not have given enough antibodies, you would have been silly to go back in the clinic again with a higher dose. So let's just be safe with a very high dose and then we'll zoom into the dose of Phase III as we learn more in the clinic. And so...

Can you -- yes, I was just saying, can you talk about and the data that you presented, both low, medium and high dose, commented about binding antibodies and neutralizing antibodies. I think the qualitative comments you made that binding antibodies exceeded convalescent sera, that neutralizing antibodies were at or above convalescent sera, that qualitatively, everyone thinks that's very exciting. But that, okay, well, how many patients were in convalescent sera? Convalescent sera can range over periods of time. And how many patients it was? And why not disclose it in the table? Maybe can you just help people understand to that criticism, even though the qualitative data was very good, why not disclose the data? And give us a range of what the data was?

Sure. Of course. So we did not disclose the data because the NIH is running the study, and they want to be able to publish the data. And as you know, top-tier journal will not publish data that has already been presented. The reason for which we decided with the Board of Directors and outside council to release the data as a top line level is because we believe the information was too material. There were too many people, including at the NIH, who were aware of data and wanted to ensure according to Reg FD, we had all our investors, we're going to get the data at the same time outside market hours and not having the risk of a leak because, again, when we saw the data on Thursday, we literally called the Board. And we had a Board meeting on the Friday to review the data. We have several clinicians on the Board who have the ability like the management team to appreciate the quality of the data. Basically, what we saw is for all the subjects we had at the free dose for that data cut of that Thursday night, everybody made binding antibodies that were at or above convalescent sera. But what was really, again, early days, but a very good sign for us that got us very excited, is at the low dose, we had neutralizing data for our first 4 subjects. When you run vaccine studies, because safety is very important, because you dose healthy people, what you usually do in the industry, this is not a model, I think everybody does that. You do a few sentinel subjects. In our case, it was 4, and then you pause. You observe safety before you enroll the rest of the cohorts. And because binding antibody assays are very quick, it's a few hours. It's like a regular assay that you can do at your local kind of lab quest or corp type of assay. But the neutralizing antibody, because they are live assays, they take many more days to run. And so this is why you could have the full cohort with binding assays complete, but only the sentinel, because of that pause we had as we were running -- or sorry, as the NIH was running the Phase I. So what was exciting, we said, is we said we're at or above convalescent sera. So of course, it was not one point. What the NIH did is they take a broad sample of convalescent sera, and what you have basically is band which is a lower concentration of titer that you had for the convalescent sera and the highest in that group. So you have basically people all over the place in that band. And what we said is we're at the band or above the band. That's what we think about is, we don't add a number or above a number, that doesn't make sense scientifically. There's a band of convalescent sera where we're at or above. And we also said it was dose-dependent. Meaning when you see the data, what you will expect is that at the lower dose, you will be in the band or maybe above the band. And then at the mid-dose, one of my program will be above a 25 microgram because we said it was dose-dependent and the data were very clear. It's dose-dependent manner.

And so to that side, I'd say, I don't say noise, but to the various folks who write about only 8 patients, was it possible to get more of those samples to a BSL3 lab? People said, well, if Moderna wanted to do it, they could get -- they could be first in line to those labs. Does that matter? Or what do you say to that?

Yes. And so it was, again, back to the sentinel and the way the study was designed, which is at the data cut. Those data that we received from the NIH because, again, we don't run the study, and we did not analyze the assay, the blood sample was done at the NIH lab. On that, of course, they have the data cut we had included only those subjects. And the reason we decided to do at least a top line release is, again, as I said, we were worried that with so many people aware of the data that it was not possible for the company to fulfill our obligation to ensure that we have no material information leaving the company. And so we had the Board on Friday night. The Board looked at the data and said, look, we cannot keep this under wrap because we cannot control it. If it was just our data, with just a few employees of a company knowing like we do for all of our programs, we have kept the data until we have a full data set.

This was NIH data. A lot of people at the NIH know the data, and you can't control it as opposed to people internally at your own company.

Correct. And plus, as you think about it, there's so much going on and so much attention because the entire country wants a vaccine working. I have no idea if tomorrow is going to be tweeted by somebody. It's going to be mentioned in the conference, on TV, and that's where we get our legal teams said you cannot take.

Absolutely. Now next steps with this data that everyone has read over. You said a publication, and I'm sure people are chomping at the bit to get even more data or just that initial data that was in the press release. You look at the remdesivir stuff, that was published very quickly. How fast are you guys working to get that out? And how much -- would that include more than just the 8 patients? And maybe talk to what we should expect next year.

Yes. So the goal of the NIH is to publish the entire data set from that first cohort. And they have said that -- I think Dr. Tony Fauci has said publicly that it's a matter of weeks, not months for them to publish that. As you say, for a lot of work that has happened around COVID-19 because we want to make sure we spread the data to the entire scientific community so everybody can learn at the same time quickly because, as we said, one vaccine is not going to be enough, one treatment is not going to be enough to help everybody, that a lot of journalists have put online right away before peer review the data. And be very clear in a big kind of red warning, this has not been peer review. And then to have it peer reviewed in the next kind of week or so because, again, there are people ready to go and review that data. And again, being the first U.S.-based vaccine data set that should be out there, we should expect people to want to review the data very quickly to make sure the community can use that data.

Okay. So it may be out in a nonpeer-reviewed format first is what you...

And again, I want to be careful, it's a possibility. Just I'm careful because this is not our data. At the end of the day, Dr. Tony Fauci is going to decide what happens. And that's why I want to be cautious in my words because I don't control what will most likely happen.

Makes sense. Makes sense. Okay. Now as we get that, are they expected to have more updates from those data sets, longer-term follow-up, durability or it's just up to Tony?

Yes. So whatever data we will have at the time they do a cut to put in the paper, all the data available is going to be in that paper. That's where...

So they're having another cut of that coming up.

Correct. And that cut would include all those 3 cohorts, 25, 100, 250, of the 18 to 55 year old. I'll remind you, just for completeness, that through the study as it was ongoing, the initial study with the NIH, we added 6 cohorts. We added 3 cohorts for the 55 to 70 years old, same dose, and 3 cohorts, 71 and above. So that data set would not be published yet because it started later, so just a question of time again. But the first data set we want to get out to the community is a full data set of neutralizing antibody, binding antibody safety of first precohort, healthy adults.

Yes. Absolutely. There's going to be a lot of interest in them, certainly at the above 70 patient as well, right? So now you started the Phase II or about -- I guess, yes, you did enroll the first patient, I think, was the announcement. This is I think 600 patients, 2 doses of placebo, younger patients, older patients and I'd say a much more robust data set in more patients than the NIH study. How fast can you get that data out? And importantly, when that data comes, what is the discussion with regulatory bodies later this fall or winter around that? Because everyone takes all of this upon and says, if this data is very good and has high levels of neutralizing antibodies at or above convalescent sera, what do you do as the NIH or as a government agency with that?

Yes. Those discussions will, of course, happen with, I would say, the FDA mostly. Of course, NIAID is very involved, obviously. But at the end of the day, it's an FDA decision. And if there is, let's say, an emergency use application or accelerated approval, again, that's an FDA decision. Then it will be most probably a CDC distribution of a physical product, like you've seen with the Gilead antiviral, to ensure the good allocation of the product around the country to people who need it the most. And so that will happen, of course, in the fall as we get more and more data, not only from the Phase II, but also safety data from the Phase III. As we've said, we are on track to start the Phase III in July. And the Phase III is going to be really large. So even in the early fall, you're going to have a lot of safety data, but also neutralizing data on a much bigger end if we're able to take a cut at data of a Phase III. And so that data set is important. As you know, it's typically in vaccine in Phase I to do the study in tens of people. In our case, it was 45 people. Because, again, you want to make sure the product is safe that you make neutralizing antibody so if we want a bigger exposure, as you say, 600 people into the Phase II. The piece that I think is sometimes overlooked about our vaccine technology and not always appreciated by investors. It's hard because we do not use a, let's say, like a viral system to bring in the genetic material, in our case, messenger RNA, that the variability between subjects in our technology is very, very tight. If you go back and you look at the CMV data, or the Zika data, which we presented recently at our Vaccine Day in April or any other vaccine that we published. What you will see is that every dose -- first, you have dose dependent if it's to a naive virus. I mean people have not been exposed with the virus before. And you will see a very tight band between people with the lowest level of neutralizing antibody and the highest. As you know, we are over technology because of viral particles. It really depends on people own immune system, whether they respond or they don't respond. It's kind of a hit and a miss. We do not...

This kind of relates to a question when we get to the winter, assuming you get this data, to the question of how do you -- how does the agency -- or how does the public, I mean, we want to ask a broader question because there's a lot of people listening. The public, how do you get them comfortable with: a, how many patients you have treated by then? I guess the answer is thousands because of the Phase III. But that's part one. And two, how do you get people to feel comfortable without infection data?

Yes. And I think it's about combining what we announced on May 18, which is neutralizing antibody and their titers and the animal challenge data. What we announced on May 18 is that we had full protection for the mice model. We have a nonhuman primate model running. And again, by the fall, by late summer, the data should be out published. And if we show in nonhuman primate full protection to a viral challenge where, basically for people on the call today who are not used to those models, where you give a vaccination, a prime and a boost like the human schedule, and then you wait usually 20, 30-day-ish to make sure the immune system has time to map the immune response and we find the antibody. And then you hit with a very high dose of virus. And you look at viral loads in the nose, in the throat, in the lungs, and you look at disease. And if we are able to show in nonhuman primate, like we have shared -- that we have shown with the NIH in the mice in them having full protection, then the 2 set of data, as Dr. Fauci has said publicly, provides you a basis that you have a belief that the vaccine should be working.

Right. So safety from Phase I, Phase II and fill in the blank here, thousands of people by the winter from Phase III will have been dosed.

Yes.

Mouse data, but most importantly, nonhuman primate challenge data, which will be published in the fall. And if, therefore, you are preventing infection in the challenge model, you triangulate that. There's 3 different points to feel comfortable without infection data in humans that, that will be a regulatory discussion and that the public will -- whatever number that is, there's a survey, there's 50% of people might feel comfortable to get it. So I'm sure you [indiscernible] sign of these, most of those people will feel comfortable to get that.

Yes. And if you think about it, let's say, if there is an emergency use or accelerated approval, we will be supply-constrained. And we'll talk about that. We have the partnership with Lonza. We have our own plant in Massachusetts. We're making as much product as we can. But the piece that is very important to understand is we will not have 1 billion dose ready in the warehouse by Christmas. And so whatever we have every month, as soon as we can make the product available when we get the okay from the FDA, you're going to want first to really work with the CDC very closely. So okay, who needs the vaccine first? You have health care worker, essential workers, the elderly, maybe above 75 years old, people with high comorbidity. So this is where the allocation of product has to be done by the government, not the company.

How many -- I don't want to say doses because it's up to 2 doses, but how many patients' doses equivalent could you have by the winter?

So we have not shared precisely the numbers, but what we've said is that in the Q3 time frame, we're starting to make hundreds of thousands of doses per month. In the Q4, millions of doses, in the low tens of millions of doses per month, and we'll keep ramping up. So the way you want to think about it is we said that we could go up to 1 billion doses per year output. But we'll build up capacity, so we need to buy machine. We need to hire people to scale up the manufacturing process, which we are doing all those things as we speak, if we think below that.

Thousands now and up to tens of millions.

Yes, exactly. And then you get to hundreds of million or 80 million and 90 million up to 1 billion.

Okay. A question for you, too, in the last couple of minutes. Number one, the world would hear about other vaccines coming out, mRNA-based, that's Pfizer, BioNTech as well as viral vector-based from J&J and Astra. How should the community put those into context? And how do you comment to those coming? And what does it all mean for you?

Yes. I would say, first, as I've said before many times publicly, I'm very sincerely hoping more than one vaccine makes it to the finish line because no manufacturer of the ones you named or others has the capacity to supply the planet. Nobody has. So to really put the end to this public health crisis that we're facing as a world, as a planet, we need many people to get to the finish line. The piece that I would want to look at first is efficacy. So substitutes for efficacy before you have dose Phase III readout is neutralizing antibody and titers. You cannot compare the assay of different companies precisely. But you can think about it as order of magnitude. If one is around 10 or 20, and the one is to over 200, and so the second vaccine is going to be better. You're going to make more antibodies, meaning you have less chance of having infection driving into disease and also have a duration of protection. Because if you start with a very low antibody base after vaccination and then the antibodies are going to wane over time, you might have less duration of protection. So I think you don't want to compare this company and that company saying one is 20, the other one is 22. Oh, 22 is better, no. If it's close because of the difference of your test or the assay, it's immaterial. But when you have a lot of difference, that's the type of thing that I'm going to be looking at, just for me as a CEO of a company, to understand the competitive landscape and how good our product is or not.

Right. So yes, magnitude, significant magnitude, then we could start talking. There's noise within the assays if they're close, and then durability will be the others. Now that will take time, too. But by the winter, you will have durability information on Phase I, too.

Correct, because the Phase I and Phase II have 12-month follow-up study. So you take blood from people regularly. So you're able to plot individual by individual. And then again, with [ aroba ] which we have seen on CMV. We have 12 months of data on CMV. You'll see first, it lasts very long. And 2b [ aroba ] stay very, very tight. So we don't have half the people or 20% of the people that are not protected anymore.

Yes. Fantastic. So point to the CMV data for durability. Okay. Stéphane, thank you so much for the work you're doing under a lot of scrutiny from everyone, but it's all for an important cause, as you can appreciate. So we appreciate everything you're doing and being visible and staying in front of people. Thank you for your time today.

Thank you so much for having us.

Thank you.

I say thank you. Bye-bye.