Skip to Main Content

The pandemic has shined a spotlight on the biopharmaceutical industry, proving its value through the speed of vaccine development to a society that ranked it even lower in its esteem than the federal government. But when it comes to drugs for Covid-19, how has the industry performed? Josh Schimmer has been analyzing the industry for decades now with the firm Evercore ISI, and he was early to flag the risks of this new virus in the beginning of 2020. He wrote in a research note “very mixed feelings about how the biotech industry delivered during the pandemic.”

We were intrigued by that, so we invited Schimmer to “The Readout LOUD” to have a conversation. This interview has been edited and condensed for clarity.

Even before we saw how the last year actually played out, you predicted that clinical trial data for Covid therapeutics would be a mess. Why did you expect that at the outset?


If you look at the history of development of antivirals and if you consider the number of antivirals we actually have on the market, which is very few, and appreciate just how little we really understand about that intersection of virology and immunology and how difficult those trials can can be to run in the first place. That is evident just in the wide spectrum of how Covid presents from asymptomatic to catastrophic.

Trying to navigate through all that with fairly murky science, I kind of had a sense that this was going to be a lot trickier on the therapeutic side. Much more straightforward on the vaccine side, though.


Unfortunately, you were right for the most part. You say report after report of nearly useless trial data has been trumpeted as success. We’re not asking you to name names unless you really want to, but can you give us some of the broad examples?

Most [Covid-19] patients recover very successfully, and often very rapidly. So when you have a small dataset, it’s really hard to tease out a therapeutic effect. As companies are generating these datasets, they bring their own biases. But the reality is, when you step back and look at all this variability, teasing out signal from noise in a small trial is is incredibly hard.

And just going back to another example, think of Alzheimer’s. How could you ever possibly get a clear signal in an Alzheimer’s trial outside of some truly miraculous recoveries which never really happen? The contrast here is that miraculous recoveries of Covid happen, and they happen often. And so it’s really, really challenging to discern a treatment effect. You can understand then why companies may be looking at their data saying, “Wow, isn’t this exciting?” But for those of us who are a little bit more experienced, looking at data, trying to figure out, OK, is this a real signal or could this be noise? It can be very frustrating. It almost feels like you’re on opposite sides of some battlefield with the data in the middle trying to convince each other and everyone around you what you’re actually seeing.

I think the truth is that as far as the data is unfolded, there have been way more failures than successes. How many drugs do we have approved? You know, not many. It might have been a little hard to predict a priori what was going to work, because we don’t have as good an understanding of biology as we need. And so we did need to throw a lot of spaghetti against the Covid window and just kind of see what was going to stick. Part of the challenge is that the biotech industry is so vast now, we’ve had so many companies throwing their spaghetti against that window, and it’s almost getting in their way of finding those pieces that are really going to stick.

To that point, acting FDA Commissioner Janet Woodcock has said similar things about the spaghetti situation, basically that just 6% of all the clinical trials done for Covid therapeutics would yield data that the FDA would actually consider actionable. Whose responsibility is it to organize the system better?

That’s a great question. If we want to centralize it, it needs to be centralized. It needs to be government-mandated, government-run. And that also means run by people. We could assemble that, but it would have to essentially be command and control, and that’s not exactly congruent with with the capital system that that we have — which actually has been the most prolific in terms of in terms of drug development. So now the question is, how do you know if that were really the right way to succeed? You’d be taking fewer shots on goal. But each shot on goal would give you better information. And maybe at the end of the day, there needs to be some compromise between between those.

There’s been so much hope for an antiviral against SARS-CoV-2, basically a pill that you could take to cure the disease in the early days or even prevent it if given post-exposure. Why have these antivirals taken so much longer to develop than, say, antibody drugs or vaccines?

My gestalt is that the window of intervention for a direct antiviral is extremely narrow. Right? A vaccine is very long. Give it once, and whatever. Five months later, three months later, you’re protected. You can’t do that with a direct antiviral. We can’t be all walking around taking those every single day. So we need to figure out exactly when to start them and exactly when they’re useful. And in many settings, as the infection is being established, own bodies are mounting an immune response. Our own bodies are the antiviral. I think the challenge is just trying to find that sweet spot of intervention pre-hospital, but also being cognizant of the fact that most people recover on their own. Most people don’t need the antiviral. And so how do you hone in on that on that clinical trial design that gives you a clear and compelling answer. And I think we’re kind of seeing that throughout all the clinical trial results.

So the antibody drugs were also expected to play a pretty big role in the pandemic, but the treatments themselves haven’t really taken hold across the country, it seems. Why do you think or are you surprised that they haven’t? And what do you think the barriers still are?

I’m guessing it’s kind of related, right? Who do you treat? We don’t have supply to treat everyone. So you’re going to want to focus on those highest-risk individuals at the right time, in the right setting, and that logistically itself is is challenging.

You called the industry “a wonderful mess of innovation that we’ve seen through the pandemic.” As we come through the pandemic, and maybe try to prepare better for the next one, do you see lessons that could make us take this jumble of innovation and better organize it so we can actually get answers and maybe better drugs faster?

The reality is that all of this is murky. And so there’s always going to be that element of uncertainty. For me, some of the noise was more in the form of the way that the data was presented. It felt a little aggressive or ambitious in terms of the conclusions that were drawn from small data sets or even some larger data sets relative to what the data was really telling us. And so if I could wave my magic wand, I would suggest that maybe companies tone down that messaging. I know it’s going to be really hard for them, though, because they’re probably looking at their data. They’re probably very excited. Maybe they’ve got pressures on the company to to deliver value to shareholders. And this is one way that they see to do it. That’s where that frustration kicks in. They’ve got their objectives and motivations. We’ve got slightly different. We’re really all trying to get ahead and predict the future and get to the truth a little faster than others. And managing that is is just an aspect of the industry in general.