Simplifying COVID Immunology, One Metaphor at a Time

; Abraham Verghese, MD; Shane Crotty, PhD


August 03, 2021

This transcript has been edited for clarity.

Eric J. Topol, MD: Hello. This is Eric Topol with my co-host, Abraham Verghese, for our Medscape podcast, Medicine and the Machine. Today we welcome Professor Shane Crotty. Shane has a remarkable background at the Massachusetts Institute of Technology (MIT), University of California San Francisco, the Emory Vaccine Center, and now here in La Jolla at the La Jolla Institute for Immunology (LJI). He's been there for about 18 years now. Shane, welcome. It's great to have you with us.

Shane Crotty, PhD: Thanks for having me.

Topol: If you are active on Twitter and if you read the major journals on COVID, Shane is there, whether it's in Science magazine, Cell, or you name it. On Twitter he's a great educator about immunology and virology.

Shane, you took on a big mission to try to educate the social media world beyond the usual academics, and you're writing papers in top-tier journals. I want to get your thoughts about the challenges COVID presents to the medical professionals who don't necessarily have expertise in immunology. Maybe you can address this mission you've been on.

Crotty: I didn't do much on Twitter for a long while. But the science on Twitter is pretty great, particularly during COVID-19, because there is a lot of good filtering, good preprints, and even some people publishing data directly on Twitter. So Twitter definitely has had scientific value.

But public education is different and definitely harder. I've always found communication to be important to the public. At MIT I majored in biology and in writing; the writing part was in the interest of communicating to nonscientists. For me during COVID, that came to a head when the Pfizer and Moderna vaccine clinical trial data became available but the vaccines didn't yet have emergency use authorization. This was right around Thanksgiving last year. These vaccines looked incredible, but a lot of what was in the news and on Twitter was about fear of RNA vaccines. They were being called genetic vaccines, and I didn't see that being corrected anywhere or better communication about them.

At that point, I'd worked about 200 days straight since March. I was taking Thanksgiving Day off and I thought, Why don't I write something for Twitter on RNA vaccine safety, and what an RNA vaccine actually is and how it works? The safety data from the clinical trials were phenomenal. So that's what I did. It was seen about a million times in the first 24 hours and 3 million times in 72 hours. That's when I referred to RNA as being like a Post-it Note. It's just a temporary message that the cell writes instructions on. There are around 5000 Post-it Notes inside every single cell, and once they're read, they are shredded and thrown away. It's not a permanent change to the cell.

That analogy was used in newspaper and television reports all over. That swung it for me to try to provide Twitter posts that were educational, even though the immunology gets really complicated really fast.

Abraham Verghese, MD: Comparing RNA to Post-it Notes is exactly the kind of colorful metaphor that not just the public but all of us need to truly understand immunology, given that we studied it so long ago. I remember the first time I heard someone describe an activated immune cell as being the difference between a Piper Cub plane and an F-16. That was the moment I understood what activation really means. You've done a nice job of explaining immunology in terms that all of us can understand.

Topol: Seeing your communication at that juncture was quite important. You then reviewed the mRNA vaccines preceding the current ones and provided some grounding. But even today, some people are still trying to call this gene therapy. There's this misinformation that has been propagated that the mRNA goes into the brain and causes headaches and brain fog. It's just endless in terms of the misinformation. So we're thankful to have you trying to set it straight.

Give us a quick review of the sequence of events when someone gets COVID — the interferon response, the innate immunity and adaptive immunity, the antibody B-cell and T-cell responses.

Crotty: Sure. I work from a simplified model. In terms of protective immunity, the pathogenesis of the virus always matters. I see it as an infection in two stages: One is the upper respiratory tract and oral infection, and the second is the lung infection. Clearly the lung infection and the pneumonia are where the hospitalizations and deaths come in.

It's also a different immunologic battle in those two places. Temporally, you are infected first in the nasal passages and then quickly the rest of the upper respiratory tract and the oral mucosa. Any virus that can cause disease in humans has to have at least one good invasion trick. Frequently, understanding that trick is important for understanding the pathogenesis and understanding protective immunity. For this virus, it's been pretty clear from early on that its biggest trick is avoiding an early type I/type III interferon response. SARS-CoV-2 is an unusually stealthy virus and that gives it a bigger head-start on your immune system than the average virus.

In general, your innate immunity is important to dampen down the levels of viral replication and keep it from spreading further by activating those antiviral pathways in neighboring cells and tissues. But also, almost all of the adaptive immune response — your antibody and T-cell response — is completely dependent on that first innate immune signal. The adaptive immune response does not start until those innate immune alarm bells go off. You can think of it as a burglar alarm or a fire alarm. Until that alarm goes off, your B-cell and T-cell fire trucks are just sitting parked there because they don't do anything until they hear that alarm. That timing is a big deal, because the virus is replicating on an exponential scale and the adaptive immune response also replicates on an exponential scale. So any of those time lags matter.

Our group, along with Alex Sette's group, was the first to show that people who weren't hospitalized but had symptomatic COVID-19 make a T-cell response — both CD4 and CD8 — and a neutralizing antibody response to the virus pretty quickly, and of a magnitude that largely fits our expectations. If you said, okay, there's a new respiratory viral infection that on average is causing something like cold or flu-like symptoms, how big a T-cell and antibody response might you expect to control that? It looked to be in that range and with the qualities we would expect for those T cells and B cells.

Vaccines vs the Delta Variant

Topol: Fast-forward and we are well into the mRNA vaccination era. Hundreds of millions of people have received them. And we're also well into the Delta variant. Can you tell us your thoughts about how Delta is different from the ancestral strains — Alpha, Beta, Gamma — with respect to the vaccine protection and breakthrough infections?

Crotty: That is definitely the topic at the top of the moment. The quickest answer is that Delta does have a partial immune evasion from antibodies. It probably doesn't have any meaningful evasion from T cells. Its antibody evasion is somewhere between the very modest amount we see with Alpha and the more severe amount we see with Beta. The most detailed analysis of RNA vaccine protection from Delta is from Public Health England, which calculated 88% efficacy still against Delta for symptomatic cases, essentially no change in protective efficacy against hospitalization that is still in that 95% range. That seems consistent with expectations from in vitro data, from what's been happening in the United States in terms of cases and protective efficacy against other variants.

Protection against the Alpha variant was probably only down about 5% for the RNA vaccines, whereas protection against Beta was down about 20% in a big study in Qatar. So the data from England largely matched expectations. There's definitely a flip side, which is the data coming out of Israel and more concern from the Israeli government about how much protection there is. We'll have to see how it plays out, but that definitely, as far as I can tell, gets pretty deep into the epidemiology of what's the proper way to calculate efficacy and how well you match the cases.

I like the editorial in The New England Journal that commented on the Public Health England paper and their choice of tests and negative controls for how they came up with the efficacy rates. It's conceivable that there's a bit of a difference there. Moreover, in England, they took this pretty bold strategy of giving the two doses 12 weeks apart instead of 3 weeks apart, whereas in Israel, they followed the clinical trial schedule with the 3-week interval. As a result, more people in Israel are further out from their last dose than in England. It will be interesting to see more head-to-head comparisons of that over time and how it affects protective efficacy.

The fact that even with substantial drops in antibody titers over 6 months and that antibodies aren't quite as good against Delta as against Alpha, or the original strain, the observation that protection against hospitalizations and, by proxy, deaths hasn't changed is consistent with the notion that there are a lot of different parts of the immune system that are contributing to protection at that hospitalization stage.

That takes us back to the model and thinking about this as a disease in two phases. You have the upper respiratory tract infection and then the lung infection, and they're different. And while this is a virus that replicates quickly in the upper respiratory tract, it's a pretty slow disease in the lungs. It's usually 7-14 days before you run into trouble with pneumonia. Being a slow disease in lung tissue means there's a lot more time for different branches of the immune system to contribute to control, whereas in the upper respiratory tract, you've only got 3-5 days before people become symptomatic and transmitting.

So although stopping that upper respiratory tract infection and preventing symptomatic cases is a pretty high bar, preventing hospitalization is a much lower bar because the kinetics are so different. There's a lot more time for CD4 T cells, CD8 T cells, memory B cells, and recall antibody responses to contribute to stopping the virus in 14 days vs stopping the virus in 3 days. Immunologically, that is why the level of protection against hospitalizations and death remains high even when protection against symptomatic cases drops for different variants.

A Change in Vaccine Delivery?

Verghese: You have looked at administering vaccines very slowly or over a long period of time for other viruses that have a slow pathogenesis. Do you see us changing the means by which we administer these vaccines, not just the intervals but by infusion instead of inoculation?

Crotty: With our HIV vaccine, we asked whether we would get a different immune response if the immune system sees the antigen over a different period of time. Working with Darrell Irvine, our MIT collaborator, we compared a single bolus injection with a conventional syringe with providing the same amount of antigen and adjuvant over a 14-day window instead. The results were dramatically different. All of the monkeys made neutralizing antibodies in one case (14-day administration) and none of them in the other case. That sort of delivery of antigen over a 14-day period is more like what would happen during an infection. It matches what your immune system has essentially evolved to do. One thing we learned by doing that was that when you deliver the same amount of protein, the same amount of antigen, over a longer window of time your immune system gets a lot better at retaining that antigen, and as a result, it gets better at sustaining a protective immune response, even though in actuality it's the same amount of material.

Those results have been repeated in multiple experimental systems, so there's a lot of interest in trying to figure out whether that can be mimicked in other contexts. The thought at the moment is that the RNA vaccines aren't doing that. The reports of the RNA vaccines have been that the RNA probably results in expression of protein for just a couple of days. If that's true, and if that could be changed to protein being expressed over a longer period of a week or two, we might be able to produce a big improvement in the quality of the immune response just by changing the kinetics.

The Importance of Memory

Verghese: Your widely cited Cell paper early on in the pandemic showed a fairly robust response to this virus across the whole immune system and that there was good memory. For the benefit of people like me for whom Immunology101 was a long time ago, it's interesting that we're focusing more on T cells than on B cells, when most of us think of this vaccine as being in the business of generating antibodies from B cells. Explain the importance of the memory component and, if you would, expand on that a bit.

Crotty: For any virus that's susceptible to neutralizing antibodies, your perfect vaccine is a vaccine that elicits high levels of neutralizing antibodies forever. If you can do that and you can get essentially sterilizing immunity, you have such high neutralizing antibodies, you stop the virus at the front door, and that's the end of it. So neutralizing antibodies are definitely important.

In the real world, that frequently doesn't happen. Either you can't generate that high level of neutralizing antibodies, or you can't maintain them long-term, or you end up with variants or mutants that can get around some of that. So that then raises the question: If the virus gets past those neutralizing antibodies, that frontline defense, do you have additional lines of defense and/or do you have redundancies and backup systems in place?

For an acute infection like SARS-CoV-2, that's where the T cells come into play. Once you are infected, the T cells become more important than the antibodies because now you have a lot of virus inside cells, generally inaccessible to the antibodies. Combined with that, it's generally very hard for acute viruses to evolve and escape T cells because there are so many different epitopes available for the T cells to recognize. And my T cells will recognize different epitopes than your T cells. So if the virus could escape some of my T-cell epitopes, they wouldn't be escaping yours, and thus there's not that much evolutionary advantage. So it makes sense for there to be a high level of attention on the antibodies, along with a recognition that they're not the only source of protection, particularly when you shift from talking about protection from infection vs protection from hospitalization-level disease.

Medically, one context to frame that in is the many people on B-cell depletion therapies for a variety of clinical reasons. Those people either can't make any neutralizing antibodies against the virus or make very low levels of neutralizing antibodies against the virus. Yet, in general, they haven't had much greater risk for poor outcomes with COVID-19 compared with healthy people. Now, there certainly is some elevated risk, so you definitely want to have those antibodies. But if it was all about those antibodies, those people would be at a much elevated risk. Yet they're not, suggesting that if they get infected, there are other components of the immune system that can serve as compensation for lack of a great neutralizing antibody response. That's the reason for attention to T cells.

In our original paper we focused on T cells in part because everyone was working on a spike vaccine. In the absence of data, that makes sense, but if it turns out that spike is a poor antigen for T-cell epitopes, you have to have a T-cell response to get an antibody response. If you don't have a CD4 T-cell response, which is my lab's expertise, you usually can't get those neutralizing antibodies. If that's a problem, you want to know that quickly. Instead, we saw 100% of people had good CD4 T-cell responses and spike was a very good T-cell target. One of the things we said in that paper was that the good T-cell response probably indicates that spike-only vaccines are a reasonable target. That was one of the reasons to call on the T cells.

T Cells and Booster Shots

Topol: T cells are rarely measured. There's no commercial assay. Your group at the LJI have been leaders in this space. One of the things that's comforting about the variants such as Delta is that the T-cell responses are strong and seem to not be affected nearly as much as the neutralizing antibodies. Of course, you would think, oh, well, good — we won't have to worry about this. But then there's still tremendous concern about widespread need for boosters. So if our on-demand system is so good on these reserves, and T cells are not so affected because of these diverse epitopes and the ability to deal with the evolution of the virus, why are we going to need boosters?

Crotty: There are multiple LJI studies on this general topic. One recent study led by Alex Sette showed that, yes, with all the variants there's a negligible effect on T-cell recognition of them. And a preprint we put out just a couple of weeks ago I think is the first report of T-cell memory 6 months out from an RNA vaccine. One of the things we haven't known about RNA vaccines is how long they generate memory for the virus. That's a big consideration when you're talking about the need for boosters. We studied the low-dose Moderna vaccine at 25 µg and, quite impressively, we found essentially no change in T-cell memory between 1 month post-vaccination and 6 months post-vaccination. So it looks like the RNA vaccines actually generate impressively stable T-cell memory, which is a positive sign.

If these backup systems, so to speak, the military reserves, are so good, why would you need boosters? I'd say that comes down to two things. First, there's no proof that T cells are providing that protection. It's much harder to show scientific proof for T cells than antibodies because you can purify them and passively transfer them into another person or a monkey and have direct experimental proof. It is 30 times more resource-intensive to study a T-cell response than an antibody response. It's harder to get samples, and biologically it's tougher to prove. So you have to hedge somewhat because there isn't enough proof on that side.

Second, it comes down to the definition of protection. One level of protection is hospitalization or death. Another level of protection is obviously infection and/or symptomatic infection. It's less likely that T cells are contributing protection against symptomatic disease and certainly against infection because they take some time to kick in. They have to wait until you're infected. Then they have to detect the infection. Then they have to get to the site of the infection and expand and whatnot. Normally, that's going to take time, which is fine for prevention of hospitalization, but this is a virus that's fast at replicating in your upper respiratory tract and transmitting in just a handful of days. So if the goal is to prevent symptomatic disease and transmission, you want to bias things more toward that antibody component.

Topol: You were one of the first groups to show the sanguine long-term protection going out to 9 months. And the other issue here is that people, when they've had confirmed COVID, they say, well, I'm good to go. I don't need a dose of a vaccine. Could the vaccine take that to yet another level, and is a prior COVID infection plus one dose of mRNA vaccine better than no infection and two doses of mRNA vaccine?

Crotty: Last question first. Is prior COVID plus one dose of mRNA vaccine better than, well, anything else? Yes. I summarized that literature in a one-page piece in Science a few weeks ago, which I called "Hybrid Immunity." Eric, you and some others have called it super-human immunity, which is bulletproof, and is fair enough. The immune responses in those individuals — the T-cell and antibody responses — are dramatically higher than what you get from either infection or immunization.

Do people with prior COVID need vaccination? I think it's a topic about which reasonable people can disagree because the epidemiologic data so far have been out to 8 or 9 months post-infection. Those people are still something like 93%-100% protected, depending on the study. Those are relatively large studies, so that's pretty good evidence.

Against variants, however, that protection drops, and how much that protection drops really has not been quantified well. The best paper on the topic is the Science paper on the Manaus, Brazil, outbreak, which suggested something like a 30% drop in protective efficacy against the P-1 variant compared with the original virus in terms of reinfection.

That's obviously a big drop if it's occurring in other contexts and with other variants, whereas the vaccine seems to be more consistent than that. In our Science paper, where we measured the immune responses, the immune memory in people who had had COVID, it was quite encouraging 8 months out.

The status of the discourse before our paper was centered on the possibility that people don't have any memory at all and maybe don't have any protective immunity. Maybe coronaviruses are just strange in this way. We didn't think that would be the case immunologically, but obviously you want data. That was the largest ever study of immune memory to a virus over 6 months, with each of these different immunologic compartments. We showed that people do have antibodies over that period — a number of groups have shown that. We showed they have memory B cells over that period. If anything, they have more at 6 months than at 1 month. We also showed that memory T cells were retained rather well over that time period, such that we projected that people who have had COVID are likely to have a reasonable amount of protective immunity for years into the future; it wouldn't necessarily prevent infection or even symptomatic disease but it would prevent serious outcomes.

A caveat that we put on that is that there were 100-fold differences between individuals in those components of immunologic memory, and we don't know which ones are most important over time. So if you show me 100 people who had COVID, some of those people will have very low immune memory but I don't know which ones. A serum antibody test can't tell me. Public health officials always want to lean on the safety side, in which case the recommendation is certainly that all those people should be vaccinated. If they want to roll the dice and not get vaccinated, most of them probably are protected. That's where the data are. That was a fairly long answer, but it's one of my favorite topics.

The Importance of Communicating

Verghese: You are remarkable in the way that you simplify and help the rest of us understand these difficult immunologic concepts. I can't let you go without asking you about your biography of David Baltimore. Most of our listeners are probably not aware that you wrote this wonderful biography, Ahead of the Curve, about a remarkable figure in science. You're living through an era where multiple scientific superstars, including my co-host, have emerged as faces of this pandemic. Will you tell that story at some point?

Crotty: Excellent question. The answer is no. Writing a book takes a lot of time and focus. Academic science jobs take 110% of your available time. I found that if you're working on cutting-edge science, it's just not compatible to do research and writing. Certainly, though, I'm super-excited to communicate with nonscientists via Twitter and other venues. And I tell people that there's just one person to follow on Twitter and it's Eric Topol. In terms of COVID, it's incredible how consistent and fast he is in finding the good stuff.

In terms of writing a biography of David Baltimore, I'm proud of the book. I wrote it for nonscientists. I was trying to use the life of David Baltimore as a prism into molecular biology, RNA viruses, HIV, cancer, and immunology — he worked in all of those areas. It was a fun book to write. I looked around and saw that few biographies of living scientists had been written for nonscientists. Biographies can be fun, but science moves fast. Reading a biography of a scientist from 100 years ago doesn't give you insights into the types of things that are going on nowadays. That was my goal.

Topol: One of the reasons you and your group at the LJI are so influential in the pandemic is that not only are you doing unique work in immunology as kind of a T-cell capital, but also because of your background in communications, you're able to come up with ways to explain all this. This is not easy at all. I don't know any discipline that's become more challenging and arcane than immunology. Everything about it is a real challenge.

I also want to mention that on your wall, there's a picture of Tony Fauci holding up one of your papers. Can you tell us that story?

Crotty: Tony Fauci held up our first paper on COVID, published in Cell, in a congressional hearing, to try to explain the importance of T cells to politicians. And you can see all our names on the paper he's holding up. People here loved it. I was super-excited about it. The president of LJI, Mitch Kronenberg, asked Fauci if he'd be willing to sign a photo of it, and he did.

Topol: As much as the pandemic has taken its toll on so many lives, it's also brought out some extraordinary science, and you've been a real leader in that. To see a scientific paper being shown at a congressional hearing — I don't remember ever seeing that. So, congratulations. We're thrilled to have had a chance to discuss this with you. It's been a great educational session. Keep up the great work. We need you, and we need to learn more about the immunologic responses to COVID. This pandemic isn't over. We have other phases to get through and there will be more challenges in the years to come.

Crotty: I want to say that one reason for the group's success during COVID is Alex Sette, Daniela Weiskopf, and a whole bunch of people at LJI who have a long history of being good collaborators and easy people to trust. We were able to gather people together quickly and get people to trust us, to provide samples and whatnot. To have Alex Sette in the building — one of the top people in the world at identifying human T-cell epitopes — allowed us to attack this at full speed and focus on trying to get to solutions as opposed to worrying about a lot of other issues. This has been a real privilege to work on.

Verghese: Keep coming up with those Post-it Note metaphors for RNA.

Topol: Burglar alarms, fire engines. That helps us understand. Thanks so much.

Eric J. Topol, MD, is one of the top 10 most cited researchers in medicine and frequently writes about technology in healthcare, including in his latest book, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again.

Abraham Verghese, MD, is a critically acclaimed best-selling author and a physician with an international reputation for his focus on healing in an era when technology often overwhelms the human side of medicine.

Shane Crotty, PhD, started his biology career as a junior in high school when he was accepted to the National Science Foundation's Young Scholars program. During the year, he spent one weekend a month working at the UCLA/USC joint marine biology lab on Catalina Island studying sharks. To learn more about his current research and his view of COVID, follow him on Twitter at @profshanecrotty

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.