How Has Data Misuse Hurt COVID-19 Efforts?

John Whyte, MD, MPH; Eric J. Topol, MD


September 15, 2020

Editor's note: Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.

  • Because only a tiny fraction of a person's antibodies have the ability to neutralize the coronavirus, convalescent plasma is not the best treatment for patients with COVID-19. It's also an expensive and arduous process.

  • Recommendations for treatment, such as the use of hydroxychloroquine or convalescent plasma, need to come with randomized trials and sufficient data.

  • In a small remdesivir study of about 1000 patients, researchers changed the primary outcome from survival to recovery time. Without prospective, randomized data, the indications for remdesivir were broadened under emergency use approval.

  • Interferon is showing the most promise for preventive treatment. Inhaled interferon would be given to patients soon after a diagnosis, before severe illness takes hold.

  • AstraZeneca's phase 3 COVID-19 vaccine trial was put on hold temporarily after a participant showed signs of transverse myelitis.

This transcript has been edited for clarity.

John Whyte, MD, MPH: You're watching Coronavirus in Context. I'm Dr John Whyte, chief medical officer at WebMD. Today, I'm joined by my friend and colleague, Dr Eric Topol, the founder and director of the Scripps Research Translational Institute and an avid tweeter. Eric, thanks for joining me.

Eric J. Topol, MD: It's always good to be with you, John. Thanks.

Whyte: I call you an avid tweeter because your Twitter feed is terrific. You succinctly give us the latest data and the latest evidence about where we are with COVID. But we were talking right before we started about how often, sadly, there are a lot of negative responses to what you're presenting as facts and data.

Topol: Thanks, John. I've been on Twitter for over a decade. I've actually had a chance — almost 11 years now — to kind of see the trends, at least from my perspective. Until the pandemic, there was a little bit of antiscience that I would see in response to posts.

What happened during these months in 2020 has been nothing short of outrageous, horrific. Basically every day, I get attacked for facts, truth, evidence. Many days I've thought, "I just better stop." I use Twitter as a way to archive what I read. Years ago, it became my main archival means of tracking the things that I want to look back at. So, I might just go back and do that — I don't know. It's been challenging.

Whyte: I wanted to get your thoughts, because you're deep in the data on all of this, on where your best guess is on treatment. Even on your criticisms on convalescent plasma, we've seen the same issues in terms of remdesivir and others — since when do we change outcomes, Eric, at the end of a study design? That we're saying we're looking at death, and no, we're not looking at death anymore. Now we're looking at time to leave the hospital. How is that good science?

Topol: It's not good science. When we make antibodies from COVID-19, we make lots of antibodies, and not just IgM but of course IgG. It turns out that when you look in-depth at these antibodies — any clones any of us would make — most of them have no neutralizing capacity. It's only a tiny fraction in most people that do the job, that take down the virus. The problem is that you give this convalescent plasma to a person who's sick, and most of the antibodies they're going to get are not going to do anything.

Now, unfortunately, it's such a great altruistic thing: I've been sick. I've had an infection. I want to help other people. And it's been, of course, nurtured. That part's really good. The problem is there's this illusion that the plasma infusion is going to have all this potent neutralizing antibody. That's not the case. That's why we have programs of neutralizing antibodies that are very potent, that are proven, pure, basically ideal. This is a very suboptimal...

Whyte: Ideal, but also, if we're injecting monoclonal antibodies, they also can have untoward effects as well.

Topol: Sure. So can plasma. You know, some of the problems that we can't screen out that could be virus pathogens, for example, we may not know that for years, just like what happened with hepatitis C. And who knows? So the point is that it's not safe.

Monoclonal antibodies engineered, that are potent, they may not be [without side effects]. But at least it's pure. That is, it's pure IgG neutralizing antibodies, not diluted with a very small fraction. Biologically, it doesn't have legs that this going to be saving lives in a big proportion. Maybe it'll have a small modest effect. But, that's about as good as we can ask for.

It's actually a big deal to get this convalescent plasma fractionated to patients. It's an expensive and arduous task. You saw happened with hydroxychloroquine. That was another US Food and Drug Administration (FDA) emergency approval that had to be revoked because of lack of safety and lack of efficacy months later, pressured by the President — outrageous aggressive promotion of hydroxychloroquine.

Whyte: But to be fair, even in good times — Dr [Margaret] Hamburg, whom you know very well, would say that at the FDA, you're criticized either way. You either approve a drug too fast or you approve a drug too slowly, right?

Topol: Yeah.

Whyte: People could say, "It's an emergency, right? We're in a pandemic that occurs once a century, Eric. We have to try things."

Topol: No, I'm with you, John. And I don't want to hold back anything.

Whyte: No! I'm glad — I don't think you are holding back. You shouldn't.

Topol: I just want to make sure that we have a signal, because hydroxychloroquine we knew could have some cardiac arrhythmia side effects. We knew there were more cardiac arrests in certain studies — like a doubling. And it wasn't like a slam dunk there.

Whyte: What regulatory flexibility should they have used? And we should want to use regulatory flexibility during these times. So what was the right mechanism?

Topol: The right mechanism would be — in each of these cases, we have insufficient data — from randomized trials. There certainly weren't any randomized trials of hydroxychloroquine at the time. Now, remdesivir was a different story because as you pointed out, John, aptly, they changed the goalposts from survival in a relatively small trial of 1000 patients approximately to recovery time. That was significant. The P value for survival was .06. So, there was this trend.

Unfortunately, it was not a large enough trial to determine survival. And what was happening — after the convalescent plasma, to add insult to injury — here again, the commissioner of FDA issues "we're broadening the indications for remdesivir under emergency approval" based on no prospective, meaningful, randomized data, just a very small study.

Whyte: I didn't think you would be a supporter of data that are trending because you have to have your data points, right? If we always said, "Something [has a P value of] .06," that might not even get published.

Topol: Well, I don't think there's any magical P value.

Whyte: No.

Topol: .05. The fact that the mortality was reduced to that extent and the fact that it helped people recover, you could say, "Okay, well, remdesivir helps in severe COVID-19." And you want more replication. Fine. That, I can see. But the point about them giving it a broadened indication, essentially by the emergency use authorization (EUA), to say that anyone who goes in the hospital should get remdesivir... This is expensive. It's in short supply. The data are weak, flimsy. I mean, this is just one more thing in this [trend of] "I don't have the evidence. I don't have the data. But we're going to give approvals for things."

Now, your point I can't emphasize enough, which is in a pandemic you don't want to withhold anything that could help people. But at the same time, you don't want to promote things that lack evidence. There is a fine, delicate balance. Dexamethasone, for example, went through the right trial: over 6000 people, and it improved survival substantially. Those are the kind of data we should have. That's the data we need to be certain, to be confident of anything.

Whyte: Well, let's give some confidence to people now that we've criticized all potential options, Eric. Where do you see therapeutics going? On your radar screen what do you think has the best chance in terms of ...

Topol: I do think there are some good things on the horizon. The first that we touched on were the neutralizing antibodies, because they are so potent. They are good because as soon as you have the diagnosis, you can basically squash the infection. So you wouldn't progress, and they should be helpful even in someone who's in the early phases, not severe.

Whyte: They won't develop immunity, though, correct?

Topol: No, but they should be good for at least a couple of months, which is important in a person who otherwise could be very high risk for morbidity and mortality. But the problem with those is they're relatively expensive. They're not so easy to make at scale. The dream that I talked about with Tony Fauci — I know you did talk with him as well — was that you could actually give it to yourself subcutaneously, that we would have it available to all high-risk people if need be, that kind of thing. But we're not going to have that. That would be the fantasy of a very effective treatment.

Now, the other [goal], which is preventive. What I emphasize here is the treatments we have so far — dexamethasone and the more modestly effective remdesivir — provide late benefits when someone's really sick. We want treatments that prevent this toll of the disease of COVID-19.

The one that has the best data right now is interferon. Interferon is, as you know, the natural, innate immunity. It's the first line of defense, which consists of cytokines and interferon. It turns out that this nasty virus is very good at deactivating your interferon response. Some people don't have a very good interferon response at all, particularly with age and male sex and all sorts of risk factors. Some people genetically, even young and healthy people, don't have the ideal interferon response. So, the idea here is to give people an inhaled interferon early. There are many studies now, both observational and small randomized studies, that support that. I think that will emerge — I hope — as useful.

There are many other repurposed drugs out there. But those two — monoclonal antibodies and interferon — basically supplement, if you will. Those look the most promising right now. I wish they were going to be definitive evidence, like, yesterday. We're still a little bit a ways from that.

Whyte: Let's talk about the vaccine and where you think we are. I want to first address the elephant in the room of a clinical hold on AstraZeneca's vaccine development program. That's a good thing, isn't it? Isn't that showing the system is working?

Topol: Yeah.

Whyte: This is a serious adverse event: transverse myelitis. Not completely uncommon in a vaccine trial. Trial was put on hold. Success, in a way, that the process is working?

Topol: Well, we don't know too many details. We know that in July, there was a temporary hold on the trial. A participant had multiple sclerosis. We don't know too much about that; it was deemed not to be related to the vaccine. Now we have a cousin of multiple sclerosis (MS) with this transverse myelitis, which could be from a virus. It could be from an immune disorder. We don't know. It's a woman who's apparently getting much better. But we don't know exactly what was going on in terms of temporal sequence.

Until you know, you definitely want to put the trial on hold. And you're right. This is the way it's supposed to work. It does work. I mean, trials all the time have temporary holds, large trials. You basically wait. You don't enroll anyone else until you really study what happened because you don't want to have something like that happen again.

So yes, we don't know enough right now. But I do think that there are a lot of things about the vaccine trials that it's critical we do know. We don't have the data analysis plans from these companies.

And these trial lists, these are done by academics. But they are working with companies. We don't have their stopping rules that are known as part of that data analysis plan. It's not been transparent except for what's put on

Whyte: It's common that there isn't transparency in development programs. But given a pandemic, we can't be having information parsed out by press releases and abstracts. Everyone has to know the data, both good and bad. As you know, when things aren't approved, typically we don't get the information. But in this situation, wouldn't you agree we have to really argue for complete transparency?

Topol: Right. This is a very delicate matter in many dimensions. For one, as we discussed, we have this FDA foundation trust issue. We have a president who is using this vaccine as a way to promote his election campaign. So we have that in the background.

We also have the issue of making the call prematurely on a trial whereby if it's based on a limited number of infections — these are moderate or severe infections — you're extrapolating that to tens of millions of healthy people that you're going to give a vaccine to. Once you stop the trial, it's like you can't go back. And then, people say, "Well, that's it." You find out potentially bad things later, you lose the trust of the entire public, which is already compromised and fragile. So, everything's on the line. All the high-quality, high-velocity science that...

Whyte: Which we should celebrate, right? Look at the innovation that we've had to get to this point of phase 3. What's your prediction though, Eric? When do you think, based on your knowledge of the process, as well as what's going on, that we would have a drug for — and I'm going to parse words for the audience — true approval, not an emergency use authorization? So, full approval based on a phase 3 trial, probably with phase 4 postmarket surveillance. When's that time frame?

Topol: I think we'll see it if we let things proceed naturally in the first part of the new year — any time in that first quarter, perhaps, ideally with enough confidence that the efficacy is solid and that we haven't run into any significant, adverse, serious, proportionate adverse reactions. It's impossible to have zero adverse events. The hope is that it's exceptionally low.

Whyte: No serious adverse events.

Topol: Yes. I think once you treat 100 million people around the world, or billions of people, you're going to see antibody-enhanced immune reactions or immunopathologic reactions. We're likely to see that. Hopefully it's at 0.000 whatever percent incidence. I want to see this thing where you don't interrupt a natural experiment that you design, because if you stop early, the chances that you have stopped it at a point of exaggerated benefit...

Whyte: What are you optimistic about?

Topol: I'm actually optimistic about everything. You wouldn't know it.

Whyte: It doesn't sound like it. I'm teasing you.

Topol: No, I'm optimistic we'll have a vaccine. It'll be safe, with those caveats I mentioned, and effective, if we are patient. I'm optimistic we will have other repurposed drugs, such as new monoclonal antibodies. I'm optimistic that, with the right leadership, we will model and get more people in the country to use the measures that we need to do, such as wearing masks and trying to keep distance, avoiding crowds, and all that sort of stuff.

If we get things on track, we can prevail over this. I'm confident of that. We've been very impatient, very divided in terms of this — as you stated at the start — the animus and antiscience. We have got to get much better education and buy-in of the public that such things as masks and vaccines are good. Right now, we have people who are not believing that.

Whyte: We don't want to be antiscience; that's what we started [our conversation] with. But we also don't want to be sloppy science too, as we've talked about in terms of data analysis.

Dr Topol, I want to thank you for sharing your insights. I want to thank you for helping educate us here today, as well as every day that you do on Twitter. We're going to check in with you, if we can, in a couple months to see where we are in the latest data around vaccines.

Topol: Thank you, John. I'm always happy to talk to you, and I look forward to more interactions.

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.