The provides robust evidence that transfusion is safe in hospitalized patients with COVID-19 -- except it still hasn't been published in a peer-reviewed journal. Michael Joyner, MD, principal investigator for the Expanded Access Program at Mayo Clinic and lead author on the manuscript, speaks with 鶹ý's Editor-in-Chief Marty Makary, MD, MPH, of Johns Hopkins in Baltimore, about the findings and the surrounding context.
Following is a transcript of their remarks; note that errors are possible:
Makary: Hi, I am Marty Makary, Editor-in-Chief of 鶹ý. I'm privileged to be with Dr. Mike Joyner today. He's the Vice Chair of Anesthesiology at the Mayo Clinic and he's Head of the National Convalescent Plasma Program. Mike, great to see you.
Joyner: Marty, it's really a pleasure to visit with you.
Makary: Mike, tell me about what's going on with convalescent plasma. There is a lot of debate right now, a lot of controversy, but it didn't really start that way. It started off as a scientific journey. Tell me where things are right now.
Joyner: Right. What happened, Marty, is your colleague at Hopkins, Arturo Casadevall, in late February published an op-ed about the use of convalescent blood products in the 1930s to stem an outbreak of measles at a boys' school. It's a long story, but Arturo's a good friend of mine. I had read this op-ed and I emailed him and I said, "Arturo, do you realize the blood banking infrastructure in the United States is big enough and robust enough you could even do this at scale?"
Makary: This was back in February or March?
Joyner: Yeah. This is... late February he wrote it, so he and I start this dialogue. Your other colleague, Shmuel Shoham, and others at Hopkins were thinking about a prophylactic protocol because historically the use of passive immunity products has been very helpful in prevention of disease. I'm old enough to have gotten a gamma globulin shot for hepatitis when I was at high school in the 1970s.
Anyways, it goes on. We start generating protocols, Hopkins has a protocol -- we have some dialogue with the FDA -- and all of a sudden people start doing this at Euston Methodist and also at Mount Sinai in New York. There becomes this rash of compassionate use EINDs being submitted to the FDA.
On March 30th, Dr. Peter Marks pinged me and said, "Are you interested in taking a look at this Expanded Access Program?" The Expanded Access Program is designed to give people promising therapies and to understand more about them in places where clinical trials and that sort of experimental medicine just isn't done: small towns, community hospitals, and so forth.
We took a look at it. I contacted my IRB chair because we needed IRB approval. He said, "We can do this. We can be the..." Mayo meaning, "we," can be the home IRB. Before you know it, in about a week we had this whole electronic enrollment and case report system set up, IRB approval, and things took off.
Then, Marty, people have to recognize one of the original motivations for this was to stop the FDA or take the burden off them of doing 20, 30, 40, 50, 100 compassionate use INDs per day and try to rationalize this and get some safety data. Our goals when we started were to provide access to promising data and get some safety information.
Our initial IRB cap was 5,000 patients and we just assumed that as this got going trials would be occurring concurrently or be planned and started and so forth. Things just exploded and they exploded for a couple of reasons. #1 is the blood bankers really got mobilized. #2 is everybody realized this was sort of the first best biological shot on goal. The remdesivir data wasn't out and so forth, Marty.
Then what happened is two groups... the Orthodox Jewish community in New York City, led by a man named Chaim Lebovits, realized that their early exposure to the pandemic left a lot of recovered, reasonably healthy people who could donate and they donated at scale. Then, Diana Berrent from Survivor Corps mobilized other survivors through Facebook, social media and so forth. All of a sudden, by the end of April and May, there was actually enough supply, and things just exploded from there.
Makary: Wow, so people were actually invited through social media after they posted that they had COVID? They...
Joyner: Right, and there was a huge network there in the... what Diana did through Facebook and other mechanisms was do that. Then she and Chaim almost began to match potential survivors who met the criteria for donation with, literally, blood bank seats and time slots in the Northeast because that's where things were really hot and heavy.
Chaim, who is in the wholesale shoe business, has superior logistics skills and he was able to help the blood bankers really think about things in a different way. A remarkable individual.
Then things take off, and we then published a paper in the JCI in the middle of May, Marty, showing that the rates of transfusion-related acute lung injury and transfusion-associated cardiac overload were not scary. We started then, in the late May and early June, having enough data to begin to look for signals of efficacy, time to transfusion, and volume transfused. Then, as June progressed, what do we know about a dose-response relationship between antibody levels in the plasma and outcomes?
Makary: Why is that important? Why is the dose-dependent relationship with survival so important?
Joyner: I think the idea is that what we did is really this demonstration project which became a huge registry. Now, the question then becomes, Marty, what elements of the registry have some randomization, pseudo-randomization, some sort of way to think about things that could have some parallels, at least, with a trial?
Once you started seeing that people that were otherwise similar, some had got plasma with low levels of antibodies, some medium and some high, you could start asking yourself, "Is there a dose-response relationship?" If you think it's the antibodies in the convalescent plasma that are helpful, then you would think that you would have some sort of inverse relationship between antibody levels and survival. That's what we found. I mean, it's tricky. The Israelis have reported some similar things and there has been other smaller trials showing the same thing.
Makary: Convalescent plasma has been around for, what, about a century now?
Joyner: You bet.
Makary: What has the experience been with other conditions, like Ebola?
Joyner: Right. The first thing everybody needs to recognize is that it always happens in a pandemic, there's always a rush, and there's always limited data. It started in 1918 with the flu pandemic. There was some evidence of efficacy and there's a beautiful meta-analysis, pooled data analysis, from 2006 on this.
There is some evidence it's worked in SARS, some it's worked in MERS, and some mixed evidence in Ebola. But more recently, at the end of last year, there was a nice trial in the New England Journal with monoclonal antibodies, again, suggesting that they work, so, again, passive immunity worked.
Probably the best evidence comes from Argentine hemorrhagic fever, which was more endemic and not a pandemic disease, and the fatality rate was about 15% to 20% in the 1960s and early 70s in Argentina. With convalescent plasma, they knocked that down to 1% or 2%, so that's really the very best evidence in a more controlled situation.
Makary: How did you develop an interest in doing this on a broad scale? You're a clinician and you're busy helping run a department at Mayo Clinic. I'm sure you've got a lot on your plate.
Joyner: I stumbled into it because I knew Arturo. For some reason, Peter Marks from FDA, who's been a real pleasure to work with and just a superior individual. We're so lucky he's involved in this and is also going to be involved in the vaccine decision-making. Peter pinged me and really it's just dominoes lined up.
Scott Wright, our IRB chair, was catalytic because he had some ideas about how to run clinical trials more efficiently. We had some people that were expert at making online questionnaires that could turn into case report forms. We have a statistical colleague who I'd worked with in the past, our Rickey Carter, and then the donors appeared, and then the blood bankers were willing.
There are five or six or eight kind of self-assembled dominoes that got aligned and each one fell in sequence, and the whole thing could have been much smaller if those dominoes hadn't self-aligned and fallen in the way that they did. We really stumbled into this.
Makary: You're basically showing up at work, doing your job, and within two months your name is on, what, 100,000 consent forms for convalescent plasma?
Joyner: It took like five or six months, yeah, and I'm not a clinical trialist. We do a lot of studies and my lab does studies in humans, but it's sort of small-end physiology study. Prior to this, what we've been working on was we'd found these people with rare forms of hemoglobin that's left-shifted. We all remember from medical school when you're hypoxic, a right shift in the oxygen hemoglobin dissociation curve is a good thing and it protects you from hypoxia. That always bothered me because animals that are adapted to high altitude are left-shifted, like the bar-headed goose that flies over Mount Everest. We found these patients with hemoglobin Malmo and we'd just finished a series of studies showing that these people were in fact incredibly hypoxia-resistant.
I think that's the other thing is that I had this skillset and had done human studies, and was more than willing to kind of challenging conventional wisdom or jumping with both feet. Really, one of the people that connected me and Arturo was the author David Epstein, who wrote the book most recently, Range. David is a friend of both Arturo and me. He was the one who tweeted Arturo's op-ed and I picked it up off of David's Twitter feed. That's how the whole thing started, so I think I was really lucky to not be an expert.
Makary: It's really amazing.
Joyner: ... know what was impossible, so we just started doing things.
Makary: The collegiality in medicine is really pretty impressive. I think a lot of times we talk to friends, friends from residency, med school, talk to colleagues in other departments, and really try to cross-pollinate, so it's really interesting to see that.
Joyner: Well, that's what happened. It's interesting, too, thinking about the history of cardiopulmonary bypass because the original heart-lung machine was designed at Hopkins and there were some kinks that were worked out at Mayo, so I thought that was an interesting parallel from the 1940s and 50s.
But other people showed up. Nigel Paneth from Michigan State, Jeff Henderson, Brenda Grossman from Washington, [INAUDIBLE]. I mentioned Shmuel, and people, really, from all over the country. We have a call every Friday, or every other Friday now, where it's about 250 or 300 people from all over, and people have shared data, shared experiences. It's really been interesting to see people put their normal rivalries and, "Who's going to be first author on the paper, get the grant?" Put that aside for a little while, while people attack this problem.
Makary: Are there any respected people in the field who absolutely do not believe convalescent plasma works at all, that it has no activity against the virus?
Joyner: I think in terms of no activity against virus, that's pretty clear-cut.
Makary: It's clear-cut that it has some activity?
Joyner: Yeah. Even in studies, Marty, that don't really show any kind of mortality or outcomes benefit, every study that's been done where they've measured viral load it drops, viral load goes down. In some of the less positive studies, even small studies, viral load goes down, patients become afebrile sooner, and the oxygen requirements tend to go down, so there are these surrogate markers.
You're a transplant surgeon and I think about how people get excited when a drug shrinks a tumor. It may not do anything to overall survival. But if it shrinks a tumor, people get pretty excited about it. I think the least positive studies in this area, at least do the equivalent of shrink the tumors, make people physiologically better, at least temporarily.
Makary: If you had COVID-19 and were in the hospital, would you get convalescent plasma?
Joyner: I think what you would want to... with the following caveats. I'd want it early, I'd want to make sure I had enough, and now that we were able to test it, I'd want to test it. Then, again, thinking more broadly about passive immunity, it's also only in a preprint press release form. But Lilly had some very interesting data about an outpatient study. People who were minimally symptomatic, and with their monoclonal antibody preparation, they were able to reduce the hospital admissions from around 6% to less than 2%.
I also think if you look at the authors on some of these papers from the 20s and 30s, it's names like Flexner, Cecil, as in Cecil's Textbook of Medicine. I think if these guys came back to life and you said, "Dr. Cecil, guess what we just found?" He'd look at me, Arturo, and others and say, "Well geez, boys, I'm sure glad you were able to recapitulate what we learned in the 20s and 30s before there were antibiotics and vaccines and things like that."
Makary: Should convalescent plasma be a standard of care for people hospitalized and very sick with their condition?
Joyner: I think you don't necessarily want to go all the way there. I think what you want to do is not call things a standard of care because I'm not sure there is a standard of care. The remdesivir data, while promising, isn't a home run. We have some other therapies that are coming in, so I think people have to kind of keep some sort of balance, and continue to try to do the trials, and do the trials in an intelligent way. I think you could also ask one unit versus two units and that sort of thing.
I think the other thing to remember, and this is why I'm so pleased at some of the refined products, our hyperimmune globulins are coming online to be studied. Because, remember, this is an uncharacterized product, given at sort of a guesstimate dose. It wasn't like there was extensive pharmacokinetics studies beforehand in sort of an all-comers population.
I think what we've been able to do with the Expanded Access Program is start to say, "Look, the odds are it's going to probably work best in this sort of patient." Again, you as a surgeon will understand this, that you can throw the kitchen sink at people when you're attempting to do salvage therapy in the ICU and not a whole lot of things work. Not a whole lot. That isn't to say occasionally you don't get a remarkable response, but early, I think, is almost certainly better and there are probably some limitations to salvage therapy.
Makary: How many patients have received convalescent plasma in the United States to date?
Joyner: We are pretty confident that over 80,000 have been transfused through the Expanded Access Program and that number... because we're still really having to curate the data. Then there's been other INDs, other programs, to give it, so I think probably close to 100,000 would be my guess.
Makary: Why did an initial group of patients not get randomized when there might have been a limited supply of convalescent plasma?
Joyner: Right. I think that's a great question. As I look back, and hindsight's 20/20, and as we think about what might happen next time, people were sort of flying by and this is sort of the best biological shot on goal.
But certainly by June, we knew that time to treatment probably mattered, that rescue or salvage therapy was probably not the best use case. We knew that based on historical data and what was emerging from the EAP. What I think, probably in retrospect, should have happened about June 15th is that the academic centers should have been trialized.
In other words, we had 2,800 sites out of 6,200 sites in the United States. The vast majority of those places have never done a clinical trial, have little interest in doing a clinical trial, and don't have the infrastructure, Marty, to do one.
I think what should have happened sometime in the middle of June is the CTSAs, for example, Clinical and Translational Science Award institutions -- and there's about 60 and they are the named places -- they should have huddled and figured out a way to do the clinical trials at those places.
Makary: You set up this...
Joyner: Go ahead.
Makary: You set up this study really quickly and I'm really impressed how fast it took off at a time when we were really in a health emergency. Would a randomized, controlled trial have been longer to set up because of all the added steps that are required?
Joyner: Yes, correct. You have to have enough material. It's got to be pre-positioned. Even things like testing people in a rapid turnaround testing to make sure patients are qualified early on, site preparation, so on and so forth. I think that that could have been done in early June with the big academic centers and you could have had almost a two-track.
1) is let's do the randomized, controlled trial in the places that can do it and where we can build the infrastructure or at least have some of the infrastructure to do it. Then 2) continue to let the people in Duluth, Minnesota, Lawton, Oklahoma, Yuma, Arizona, and those sorts of places use the EAP. I think that would have been a more ideal solution.
Makary: Here we are here with an experience now of about 80,000 Americans who have received convalescent plasma, with a strong safety track record and a dose-dependent survival benefit when compared to patients at a similar state in their illness. At this point, would it be ethical to randomize patients to not receive it at this point?
Joyner: Yeah. I think that given the nature of the data, given the nature of what needs to be done with RCTs, and the way we think about proof now, you can make a strong case for randomization.
On the other hand -- and what got me started thinking about time to transfusion as a potential modulator of outcomes -- is think about what happened in Iraq and Afghanistan in the late 2000s when Secretary of Defense Robert Gates, a PhD, not a scientist, learned about the golden hour in trauma. He found out that only about 20% of seriously injured soldiers and marines were getting transported to definitive care in less than an hour.
He made it a priority and over the course of about a year 70% were subsequently transferred in less than an hour. You can see the mortality drop in terms of severely injured battlefield patients as the time to transport got faster. The question is would you ever want to do a randomized trial on something like that?
Makary: No.
Joyner: Exactly. Would you ever want to do a randomized trial on like time to balloon in acute ischemic conditions in the heart? No, so I think your assertions... I'm an anesthesiologist, so our definition of what requires a randomized, controlled trial may be a little different than people who practice a slower motion of medicine.
Makary: Now, your study. I read it. It's posted on MedRxiv and it's a 35,000-patient experience. It includes patients from Mayo and many of these sites, including my hospital, Johns Hopkins. Now, I don't speak for Johns Hopkins, but I will say the results are pretty impressive. You submitted that article when for publication?
Joyner: We submitted it for publication about three weeks ago and we've resubmitted it last week. We got nice initial reviews at a major journal and we've made the editorial changes, which were significant. As you know, the major journals are more than pleased to give you significant and detailed input about how they would like the paper changed and some additional analysis. Our initial findings hold, and we're continuing to work through the additional data we've collected.
Makary: I think I saw August 10th was the date when you posted the...
Joyner: Correct.
Makary: Now, we're at September 21st. Why are the journals taking so long? Why can't they do a 24-hour review when we have 1,000 Americans dying per day?
Joyner: I would leave that to the journals and I would simply say that the journal we've worked with, several of the senior people involved in the paper have worked closely with the editor-in-chief and I think that they're moving very quickly. They had some concerns, which I think were legitimate, to think about the analysis and we were pleased to repeat the analysis or alter the analysis to address those concerns.
I also think that's what the preprint world has really begun to do for us, really help us in that regard, because if you present the primary data the way we did, people then are able to take a look at it and think about it on their own.
Makary: Soon after the data was reviewed by the FDA, there was an emergency use authorization granted. All of a sudden, maybe because we're in an election season, maybe because there was a tweet shortly after your data was submitted to the FDA, all of a sudden it became political. What happened there?
Joyner: It's hard to know for sure because you don't know what was going on inside the Department of Health and Human Services and their interactions with the White House. I don't have a complete picture of that, Marty.
I think the main thing to emphasize to people is what are the criteria for an emergency use authorization? Unmet medical need, safety, and at least some biological plausibility that there is potential efficacy, so clearly convalescent plasma meets those criteria.
We had actually been discussing the data as it emerged, starting in June, with the FDA about when they might want to move in that direction. There was actually a CMS code for convalescent plasma that developed in very late July or early August, and how things got hung up and ended the dialogue between FDA, BARDA, and the NIH, and then how the political process got involved is just unfortunate.
Makary: Was the emergency use authorization for convalescent plasma too soon, just at the right time, or too late?
Joyner: I think that you could make a strong case that it could have been done earlier. But again, the regulators have very challenging jobs. They're getting pulled in a lot of different directions. They have to try to evaluate what's going on here.
One of the things -- it's a little bit like what you mentioned about the journals -- is we live in a world where things are done in series, and bureaucracies operate in series, journals operate in series, our standard medical practice, in many ways, operates in series. When all of a sudden you start asking people to do things in parallel, it's for everybody involved very challenging and I'm actually impressed to the extent that that's actually occurred.
The real question, I think, when we go back to normal, is why do things under normal circumstances have to be so slow when you could do many things in parallel that we now do in series?
Makary: Well, congratulations, Mike, on what was a very impressive study. This version I read on MDRX was, I thought, a very sound study. When I read that, I thought, "This is entirely logical that all the boxes are checked for the FDA to issue an emergency use authorization." When they did so, I was so bothered at how so many folks criticized the FDA.
It seemed as if they did absolutely the right thing at the right time. Maybe they should have done it even a little earlier. But then to see people question the credibility, not just of the FDA, but of the entire vaccine approval process because of that sort of critique.
Joyner: Correct.
Makary: Now we've got people questioning whether or not data safety monitoring boards at the vaccine studies are credible. We've got all kinds of suspicion, conspiracy theories. I found that very disturbing. What were your thoughts on it?
Joyner: I think my main thought is I have had the privilege of working with senior physician scientists embedded in these organizations, career individuals, a number who have had outstanding academic careers and done other things as well. I have been super impressed by these individuals and I have a lot of faith in them.
You're always worried about political pressure for whatever reasons. Think back to Gerald Ford and the flu vaccine, again, which I remember I was in high school and so forth. You're always concerned about things like that, but I have great faith in Dr. Marks, his team, in our colleagues, in the FDA.
The interactions with NIH have been maybe a... they've questioned the data a little bit more, but, again, I have great faith in their search for the truth as well. I think at the level, for lack of a better word, of the technocrats, I've been very impressed and it's been an absolute pleasure to work with these individuals.
Makary: Is our research infrastructure in this country set up for a health emergency? We've got doctors who have said, "I'd love to study this." But it seems like they're up against a lot. You've seemed to navigate the system and move a study along very quickly, but it seemed like we've got a lot of basic studies done very late in this pandemic like rates of asymptomatic spread, viral shedding rates, basic things, the nuts and bolts of this virus. That research was very late.
Joyner: Marty, I think a couple things. One is it's always good to score a 15 on a 10-point extrovert scale when there is a controversy, so you're able to leverage maybe that skillsets. I've run large things before, so that was helpful. I have been Vice Dean for Clinical Research, helped reform our IRB, and been Co-PR of our CTSA, so I've run big things before. I think that was quite helpful in my own personal experience.
I think the other thing you have to ask yourself is just the focus of the country in general. People say, "Well, what do you think about the national response, or the scientific response, or whatever?" We've had the world's greatest slow-motion public health crisis, the opioid epidemic, for the last 20 years and we've really struggled to address that. We've struggled to address obesity. We've struggled to address suicide and health disparities of many types, including infant mortality.
If you look at our inability to address these sorts of slow-motion, indolent things with some pretty obvious solutions and then all of a sudden you have to, for lack of a better word, play up-tempo, or run the two-minute drill with a pandemic, is it any surprise that people have struggled?
Now, could you have done it better? For sure. Should there have been some planning? For sure. Since 2000, we've had anthrax, Zika, Ebola several times, SARS, MERS, and now this, and people were worried about H1N1, so at least seven things. The question is fool me once, shame on you. Fool me twice, shame on me.
Are we going to set up some [warm?] infrastructure, whether it's knowing where the ventilators are, to the source of epidemiology you mentioned, to convalescent plasma, to refined products, to monoclonals, to vaccines the next time? I hope we do learn our lessons.
Makary: There's a piece in 鶹ý this week that we put out talking about some of the research that's come out on coronavirus and how tragic, how late it's been. If you think about dexamethasone, we learned about that in June when we could have learned about that in April. The recent study on outdoor dining being associated with an increased risk of acquiring the infection, why are we learning about that in September? That was a survey study of 300 individuals. Are there some basic things we can do to tighten up our research systems to make them more adaptable, more resilient?
Joyner: Yeah, we need to a) have some infrastructure just be [warm?] and b) I think we've got to really get the CDC back where it was 30, 40 years ago. I think we have to, at the level of NIH, think about where our priorities are in terms of public health. This is a little bit of an aside. But 60% to 70% of cancers are possibly preventable and only 6% of the National Cancer Institute budget is spent on prevention.
I think we do have to recalibrate a little bit. We have to look at the funding of our state public health departments and I also think... I'm lucky enough to have grown up post-Sputnik, so everybody got critical thinking and quantitative education in high school and in college. I think we've really got to come back to that so people can actually evaluate the data that's coming out, and put some sort of faith in scientific thinking, and help people like Arturo and others who've made a lot of media appearances just buck up their credibility and get those messages across.
Makary: That's great. Mike, any other thoughts on where we stand right now in our research systems in the United States or on convalescent plasma?
Joyner: I think that the work on convalescent plasma continues and we're pleased to have contributed. I look forward to just more work on passive immunity in general, both for convalescent plasma and starting to think about how we could utilize that in the future for niche diseases, rare infectious diseases, maybe to help us stop some of the antibiotic resistance, and so forth. I think we have many opportunities there.
I think hopefully, when this is all done, people can do a kind of an after-action look back and start asking ourselves what sort of balance we want in our research portfolio between pure basic research, which is super important, sort of applied intermediate research, and then really very practical things that can help us right away. Figure out how we'd balance that portfolio and then what infrastructure we need to maintain or develop so that when this happens again we don't have to start doing 0 to 60. We can at least have a kind of an idler and have the pilot light on.
Makary: Do you think that we're going to see a very major breakthrough with some therapeutic before we see a vaccine?
Joyner: I think the vaccines are pretty close. I think the issue with the vaccine, in talking with individuals who are not scientists, but work in the drug distribution and logistics thing, is going to be a) how good is the vaccine? Hopefully, we'll get something that's better than the flu vaccine -- but I doubt it will be the measles vaccine -- and then I think getting it manufactured and distributed. Two shots of a product that has to be stored and transported at -80 is very challenging logistically.
Makary: Yeah, we could be looking at a freezer crisis.
Joyner: One of the individuals I have been in contact with who's really expert at this sort of logistics, he's pointed that out to me, is do we have enough freezers and things like that. Enough needles and so forth. Time will tell, but I'm hopeful that we'll get some good news here late fall, early winter, and then we'll be able to get people vaccinated. It will have a... more than a marginally effective vaccine.
Makary: Speaking of -80 degrees, what was the name of that bird that flies at Mount Everest again that you mentioned?
Joyner: Bar-headed goose.
Makary: Bar-headed goose.
Joyner: Look up the bar-headed goose and you'll see one of the great physiological experiments of Nature.
Makary: Have you ever seen one or do they fly around Minnesota?
Joyner: No, I've seen... No, we don't have them in Minnesota. I've seen them filmed. But llamas and alpacas are left-shifted. There are a number of other animals that live at a high altitude and have adapted to life at a high altitude that are left-shifted. It's really a fascinating problem and hopefully, when this all settles down, we can go back and find more patients with hemoglobin Malmo and continue to study them.
Makary: I'm glad we've got experts on oxygen exchange like you out there, Mike, and great to see you, and thanks so much for joining.
Joyner: It's been a great pleasure to visit with you, Marty. I really appreciate it.
Makary: Thanks so much.