Dr. Jo Anne Welsch, currently working on vaccine development for low and middle income countries at a global health organization, discusses how vaccines are developed and tested, and the implications of the process for the coronavirus.
In this episode
- Colleen and Jo Anne discuss the process of creating vaccines and how it has changed
- We look at what this means for the current pandemic
- Jo Anne leaves us with some optimism
Timing and cues
- Opener (0:00-0:42)
- Intro (0:42-3:30)
- Interview part 1(2:26-15:14)
- Break (12:03-13:35)
- Interview part 2 (13:35-27:32)
- Outro (27:32-28:38)
Editing and music: Brian Middleton
Research and writing: Jiayu Liang and Pamela Worth
Executive producer: Rich Hayes
Host: Colleen MacDonald
Colleen: Jo Anne, welcome to the podcast.
Jo Anne: Thank you, Colleen. It's a pleasure to be here.
Colleen: So we’re in the midst of the coronavirus pandemic, and it’s all -consuming. We're all eager for a COVID-19 vaccine, and we want it tomorrow. But experts are saying it could be 12 to 18 months before we have one. And that's where you come in. You’ve been working on vaccine development for many many years. So, could you start by telling us first just what a vaccine is and then walk us through the process for developing one.
Jo Anne: Well, the basic premise of a vaccine is that you are going to make a really safe and effective vaccine that when you administer it to an individual, a healthy individual, that you are gonna get your body, your immune system to make antibodies to usually a protein, or sometimes a polysaccharide, depends on if it's a bacterial or a viral vaccine. But a vaccine, the way it works is that you basically are making your body make an antibody, so that when you are presented with a virus or an infectious disease, bacteria, virus, whatever, that your body already has antibodies, and so it can fight. And that's the bottom line, you're just trying to get those antibodies already in your body so that when you inhale a bacteria or a virus, your body already knows what to do with it.
You know, our immune systems are pretty sophisticated. And if you already have antibodies, high-quality antibodies, that will recognize the pathogen when it comes into your body, then it will be able to eliminate it from your system.
Colleen: So then, how do you go about developing a vaccine?
Jo Anne: It's a very, usually, slow and painstaking process. I think that the first thing you do is learn the most you can about the pathogen. And then, you isolate that target, and then you create it in the lab, a formulation. Like, let's say it's a polysaccharide that covers the surface of the bacteria and you know that that's a major virulence factor, a major way it gets into your body. So, you're gonna take the polysaccharide and you're gonna make a, vaccine, and then you're gonna test it in the lab.
And in order to test it in the lab, you usually have to immunize animals to do some preclinical research and find out if those animals would make an antibody to your vaccine that would then be able to recognize the bacteria or the virus. And that's the early-stage development. It can take a short time or it can take a really long time. And I think that in the case of, I would say, the fact that genome sequencing has gotten so fast and powerful, that it's really decreased the amount of time that's taken to identify the correct target on the pathogen.
Colleen: So, when you test this on an animal, how long does it take to figure out if they've developed the antibody?
Jo Anne: Well, it depends on the vaccine, but usually, you give like two doses a couple weeks apart, and then you take a small amount of their serum and test that serum against live bacteria or live virus. And, you know, you usually have different types of laboratory assays. I know we're hearing a lot about test kits these days. But it's a similar idea, that you have specific assays that have been developed that you know will give you the information you need.
Colleen: I'm sorry, can you clarify what an assay is?
Jo Anne: So an assay, I'll give an example again from Neisseria meningitis. So, in that case, there's a very well-defined test, and that test means that you mix live bacteria with human complement, and then you add your serum from your animal. And then, you mix it around, and then you basically see if the antibody that you've elicited in the mouse, usually, actually kills the bacteria. And it's a very clear readout, and it's functional. So, an assay really involves mixing different components together in order to get an answer. Does that make sense to you?
Jo Anne: Because there's other kinds of, like, more...like, there's a very standard assay that you use also. You might even do this first before you go to a "functional assay," using live bacteria or a live virus. You might just have a plate that has like 96 little wells in it. And you might just look and see if the serum from the animal has just IgG antibodies, like, has an antibody. And that's, like, a simple readout in which you identify, "Oh, antibodies there," but you don't know if it works or not.
Colleen: So it sounds like with vaccine development, you know or you can see very clearly, or right off the bat, if it's working or not.
Jo Anne: Yes. And you know, in the earlier days, I think... And, again, like, a lot of them were easy vaccines that were developed. I shouldn't say any vaccine's easy, but I think with some of these more recent vaccines, they've been a little more difficult. And, for example, there's a meningococcal B vaccine that was developed. And, in that case, they actually had to sequence the entire genome way before we had high throughput sequencing and they use software to identify potential outer membrane proteins.
And then, by brute force, they tested 600 of them, and only 3 turned out to actually have any functional antibody response. So, in that case, like, you're doing a lot of work. You're making 600 different proteins and you're seeing which ones actually will elicit a functional antibody in, usually, a mouse model, to begin with.
Colleen: So, are there different approaches to creating a vaccine?
Jo Anne: Yes. And I would say that right now we are in a really...like, it has exponentially grown, all these amazing, great ideas. And I think that a good example of this is what's going on now with this coronavirus. They learned a lot from the SARS outbreak, and they learned a lot from MERS, you know, the Middle Eastern respiratory syncytial virus. So, they had a good sense of how these coronaviruses operate and how they infect your lungs. And so by being able to quickly sequence the entire genome of this virus, they can really then hone in and say, "Okay, this is a critical gene. It makes this protein, and that protein is critically responsible for infecting human lung cells," for example.
And so, you know, I think that this new approach about going straight to the sequence and using very sophisticated software, really, that allows you to determine, "Okay, this is expressed outside and this is the critical virulence factor. And that's what they've done with this coronavirus. And that is how they went from the sequence being published by China in early January and WHO spreading it all over and letting everyone have access to it, then everybody just got busy. And so, that’s how it's pretty astonishing that we've gone from sequence to actually having some of these different types of vaccines almost ready for clinical trial.
Actually, two of them are in clinical trials. And they have different platforms. So, a platform is like they, a company, usually, they develop a kind of technology to be able to move things along. And, again, just because sequencing is so easy and so fast now, a lot of them have said, "Okay, I'm gonna use a D-activated, so a non-virulent, like, adenovirus, and I'm gonna take this critical gene from this virus, and I'm gonna splice it in there." And that's how they've made multiple vaccines.
And I'm thinking specifically of a place in Tianjin, in China, CanSino, they actually developed using the exact sequence for the Ebola outbreak to develop an Ebola vaccine that worked. And, you know, it got licensed in China. So, it worked well, and that was basically their platform, if you will. Like, they take the critical gene from the coronavirus and they plug it into this vector...we call it a vector, but, really, it's just, like, another virus that cannot cause disease, that's been deactivated.
Colleen: Can you explain what deactivating the virus is?
Jo Anne: I think historically, the approach really was to take...and this hasn't really been used for pathogenic bacteria as much as viruses, where you isolate and grow the virus, and then you deactivate it, so you kill it. And you kill it in a variety of ways. It's a chemical way to deactivate it. So, it's like a dead virus that you can then use as your vaccine that will still make an antibody.
Once you've shown that it works in animals and go through all the clinical trials, you can actually say, "Okay, we've just killed this live virus, and we can still get a good antibody response that would recognize live virus." And I think there's been advances from that. And I'm thinking about, some of our older vaccines. Like, you know, measles, there's been different approaches. And for other viral vaccines, there's been ways to deactivate it, not chemically, but by knowing the biology of the virus and then using DNA sequencing methodologies cut out what you know is the viral part.
So, it's called attenuated. They use that terminology. And I know there's a flu vaccine that we use right now, and it is licensed for kids, and it is what they call a live attenuated. And it tends to work better, so they don't actually chemically deactivate or kill the virus. They just remove all of the genes in it that have anything to do with infection.
Colleen: So, you said, historically, that's how scientists have approached vaccines. Are there new methods that a being used now?
Jo Anne: I think that now, we really go straight to the genome, we get the sequence, we figure out what's important. And you have to actually go through and really know the biology of the pathogen and understand what the critical proteins are that that virus makes so that you can target that. And in the case of this coronavirus, we know that it makes these spike proteins.
And the spike protein is the target of multiple researchers now to use that spike protein, it sticks out from the surface of the coronavirus. And they're using that as their vaccine. We call it a vaccine target because we're gonna use that spike protein to get your body to make antibodies to that spike protein so that when you are exposed to a coronavirus with that specific spike protein, that you've already seen it, you've got antibodies to it, and you will be immune to it.
Colleen: So, can you explain a little bit just what the spike protein is? I'm imagining the visual that we see of the virus that's sort of a ball with pieces coming off of it.
Jo Anne: It's a ball. It's a corona, it's a sun corona. Yeah. And the pieces that come off of it, it's a stalk, and then the spike protein at the end of the stalk. And it's a critical factor for being able to infect human respiratory cells because it actually targets this certain receptor that are present on human cells. And so, that spike protein, like, knows right where to go, and it latches on there, and then it sets off that cascade of events of being able to replicate in your body. And so, the spike protein is a critical factor for this coronavirus to infect humans. So, it's a great target for a vaccine.
Colleen: It sounds so doable.
Jo Anne: It is very doable. Well, that's the research part of it. And then, you know, this is why vaccine development is so tricky, Colleen. I think it's that it's great to show that it works in an animal model and in your functional assays. But then you can't go from research to the clinic, you've gotta go through this very regulated manufacturing process. So, that usually takes a little while, and it just depends on how equipped your facility is. When you make a vaccine, it's a really regulated process.
Any therapeutic, so any drug, any vaccine that goes into a human, has to actually be manufactured under what we call good manufacturing practice. And it's a very regulated, very strict infection control so that you'd know what you make is what you've made and you know that it's clean and you know exactly what you have. And so, when you approach, like, the FDA or a regulator you have to have this huge mass of data just on how you've manufactured the vaccine and all the preclinical data.
So once you've done that, once you know that you have a drug product that is suitable for use in humans, then you have to start your first clinical trial, which is usually less than 100 people, and it's usually healthy adults, 18 to 45 just to start with. And you've gotta show, number one, and this is your primary endpoint, is safety. And you do what we call dose-ranging. So, you're looking at what different doses are gonna make the best antibody response.
So, usually, you'd try three different doses and you have one group that gets what we call a placebo control, which is usually saline, but it could be,I guess you could even give a licensed flu vaccine because that wouldn't work against the coronavirus. And so, you usually divide your subjects into four different groups.
And you start with your placebo control and your lowest dose. And then, you stop after a week, and you meet with your data monitoring committee, and you review the data, and they say, "Okay, safe to proceed." And then, you start in with your second dose. And then, stop, review the data, get a safe to proceed, and then you do your highest dose. And, usually it could be a single dose, and then you would examine the antibody response in those individuals about three weeks after they get a dose. But if you have to do two doses, then you would give the first dose, and then you would give the second dose three weeks later, usually. And then, wait a couple of weeks before you actually look at the antibody.
So, I would say even the shortest... I'm just thinking about the trial they just started up in Seattle. So, that one, I'm pretty sure they're gonna use two doses. And there is only 45 healthy subjects, 18 to 45. And it is designed, as I said, to give the low dose to you know, stop, examine the data, ensure that it's safe, and then proceed. So, if they're gonna do two doses, three weeks apart, then that's six weeks. And then, you usually wait another two weeks after the second dose, two or three weeks. So from start to finish, you're looking at a minimum of nine weeks before you can get a sample from your subjects and test them.
Colleen: So at that point you then move on to a second clinical trial?
Jo Anne: Yep. So, let's assume that it not only was safe but it was immunogenic, i.e. it elicited functional antibodies that you know neutralize the virus or kill the bacteria. And so, you usually have to have a functional assay for that. And, at that point, you've already, you know, written up your Phase 2 protocol, you've already gotten the FDA to look at it, and give their feedback, and approve it, and you're ready to go. And you've identified your clinical trial site so that... usually, if that's successful, then you wanna just go.
And, usually you know if it's gonna be successful because you usually test the sample from the subjects in the trial after the first dose just to get a first read if there's any hint of a response. And then, your Phase 2 trial is much larger, much larger. So, you know, it's usually less than 100 for Phase 1, and then your Phase 2 usually has like 1,000 in it because you need more safety data.
Colleen: Okay. So then, you do the next test that's with a much larger number of people and all of that data is all being collected just on a much larger scale?
Jo Anne: On a much larger scale. And, ideally, you would have a comparator vaccine, if you will. So that you can show non-inferiority statistically, that you're as good or better than an already licensed vaccine. But in the case of something like this coronavirus it's a new vaccine so there won't be a comparator.
Colleen: So, going back to the sort of trials. Sо, there's your first one, your second one. What happens next?
Jo Anne: You write up your report, you go talk to the FDA, and then you move on and progress to Phase 3. And Phase 3 is even bigger. It's even bigger. It's 3,000 to 5,000. It's got dual goals in the sense that safety is no longer what we call your primary endpoint. Your primary endpoint is actually how good an antibody response do you get? And is it in greater than 85% or 90% of your subjects. And so, it changes to how effective is this vaccine.
And, in many cases we can't really show how effective it is in the public because we can't expose people to meningitis, it's too deadly. Just like we can't expose people to coronavirus and see if it works or not. So, you have to actually have a very good functional test readout to show that, "Yes, we are getting highly functional antibodies that are capable of neutralizing this virus or capable of killing that bacteria." And then, usually, you have to show advocacy out in the real world too. But you'd be licensed before that. That would be part of your post-licensure obligations, to collect data on enough people that could potentially be exposed.
Colleen: So tell me the next part of the process. After you've done all of your clinical trials, what comes next?
Jo Anne: Then, after you do your Phase 3, if the data is compelling, you go back to FDA and you submit your entire package of data, which is called a Investigational New Drug Application, IND, and get approval and licensure. And there'll be some back and forth because you have to both agree on what the right dose is and all of those things. And you have to have done your clinical trial in the age range that you think you need to give this vaccine to.
Colleen: Given the coronavirus outbreak, are procedures being handled differently?
Jo Anne: I think, from a human subject research safety perspective, absolutely not. Totally the same regulations apply. What's different is that it's urgent and it's all hands on deck.
And so, the FDA is not gonna put you in the queue for review. They are gonna front and center say, "Okay, we're gonna review your package." I think I mentioned this earlier, it's called the IND filing, and it's an Investigational New Drug. And even though it's not a drug, it's a vaccine, it falls under that category. And so, they're gonna review your IND package right away. And so, to me, I think that that's what's different. But my experience with FDA and everybody who's been on all of those committees that had been reviewing our data packages, they are meticulous, super smart, and don't let people cut corners.
Colleen: So, is there anything that gives you concern about the way that things are being communicated to the public?
Jo Anne: In this case, I think the public is not really being very informed in a direct way. I think when the scientists get on the podium and communicate, it's clear, but I don't think we can presume that there is gonna be a vaccine around the corner. It's much more likely that we will have a therapeutic, like a drug, that we can use to treat patients much faster. But even then, a drug therapeutic, it will be a few months away because we need to make sure that it is actually working.
Colleen: What are the biggest challenges when developing a vaccine?
Jo Anne: I think the biggest challenges is making sure that...you just have no idea if it's gonna work or not. I actually think the failure rate is similar to developing cancer drugs or something. I think you get to one step and 30% of them are okay after the Phase 1. Then, only like 20% get it past Phase 2. And then, you know, another less than 20% of those get it past Phase 3. So, you know, we have no guarantee and there's a lot of failures.
I do think that we're smarter now about developing drugs and vaccines. But I think the failure rate is not insignificant and it's very expensive to do.
Colleen: Jo Anne, thank you so much for joining me. This has been really informative. I don’t know, it just helps people really understand what is involved.
Jo Anne: It's really been a pleasure to talk about one of my favorite subjects in the whole world. So, thank you very much. I really appreciate the opportunity to share my information.