OK, it’s time to address the topic of challenge testing against the coronavirus epidemic. Let’s define our terms: by this, I mean “deliberately exposing volunteers to infection by the virus in order to evaluate prophylactic treatments by comparison to controls”. I will say at the outset that I am not in favor of this idea, so that might save some people the time of reading this post. I’ll also say that my objections exactly overlap with those raised here at Stat, so if you’ve read that one you will probably not come across much new here, either.
This is not an unknown technique in anti-infective therapy; challenge trials are in fact run at times when targeting less dangerous viral infections. Here is a WHO guidance document on the idea, and here is a key section to quote:
It is important to note that not all diseases for which vaccines might be developed are suitable for conducting human challenge trials. In many cases, human challenge with a virulent or even an attenuated organism would not be considered ethical or safe. For example, if an organism causes a disease with a high case fatality rate (or there is a long and uncertain latency period) and there are no existing therapies to prevent or ameliorate disease and preclude death, then it would not be appropriate to consider human challenge trials with such an organism. However, a human challenge trial might be considered when the disease an organism causes has an acute onset, can be readily and objectively detected, and existing efficacious treatments (whether curative or palliative) can be administered at an appropriate juncture in disease development to prevent significant morbidity (and eliminate mortality).
There’s the rub. We do not really have effective treatments for the coronavirus, and the more we understand about its effects in humans the less sure we are about its long-term effects. There are clearly people who have much less severe and people who have far more severe acute courses of disease, and the same goes for the long-term sequelae. And while we know a few major risk factors to sort these out (age, obesity, cardiovascular conditions), we do not understand nearly enough. Children are largely protected – but some children have major complications. Healthy 30-year-olds are at far less risk than 85-year-olds, no doubt whatsoever – but some healthy 30-year-olds have had an awful time with the virus, and we don’t know why. If we could really identify a cohort of prospective patients for whom we could say “Yeah, you folks will be OK – you’re not going to get that sick, and we have something to treat you if you do”, then that would be one thing. But we don’t.
That’s one point that has to be made up front. Now we have to ask what challenge trials would be for in our situation. The idea is that they would speed up vaccine development, giving us more data more quickly in Phase II trials (which is where we really start testing to see if any given vaccine candidate can be protective). But I don’t think we need to run challenge trials at all – not under current conditions. We have areas where the coronavirus is spreading so well through the population that I don’t see the advantage. Sure, a controlled challenge would provide you with an experimental setup where the control group almost certainly gets infected and you can compare them with the vaccine recipients. But as things stand, the main difference between doing that and testing in many of the hard-hit locations is that you would just need to round up a smaller N for the challenge trial. And yeah, that’s going to be cheaper, and perhaps take a shorter amount of time. But will it? Getting a human challenge trial off the ground against the coronavirus will not be the work of a moment – there are many places that would simply not sign off on such a design. What sort of waivers will the participants need to sign? Remember, you’ve made them sick deliberately, so you’re more on the hook than ever to provide their treatment.
I realize that there is a web site where people have been volunteering for just such trials. This is going to sound tiresome, but it’s not as simple as that. Challenge trials would be conducted just like any other clinical trial: taking place at one or more medical facilities, with interviews and full medical histories and exams for all participants. Yes, the “1 Day Sooner” folks are listing >30,000 names at the moment, but only a certain section of those people would be logistically or medically suitable/available – it doesn’t mean that there’s a line of 30,000 people ready to go. To be honest, I think that if someone actually were to announce a challenge trial that rounding up enough volunteers might not be the rate-limiting step, but that’s the step that 1 Day Sooner is addressing.
Now consider who those volunteers would be – who they would have to be. Young, healthy people – who else? We’re already stepping close to the line with the whole idea, and giving it to anyone in a more at-risk population would surely step over it. But even the Phase I safety trials of the vaccines have brought in age groups that would be unacceptable in a challenge trial for Phase II, and the Phase II trials – since they’re taking place out “in the wild” can similarly enroll people who would never be in a deliberate challenge. This means that the data you would obtain from such a challenge is skewed heavily away from some of the people that you are going to want to treat first. Both efficacy and safety can (and do) vary by age, existing medical conditions and other variables that you’re going to have to make sure to not address in the challenge trial.
The FDA guidance for coronavirus vaccine trials release the other day actually does take up this idea (page 18). They have it in the break-glass-in-case-of-emergency category, if there is just no other way to demonstrate efficacy:
If it is no longer possible to demonstrate vaccine effectiveness by way of conducting clinical disease endpoint efficacy studies, the use of a controlled human infection model to obtain evidence to support vaccine efficacy may be considered. However, many issues, including logistical, human subject protection, ethical, and scientific issues, would need to be satisfactorily addressed.
Yep. In the end, despite the heroic sound of the idea, I think that such trials are (1) still ethically questionable, (2) fraught with legal complications, (3) are for those reasons and others not likely to run meaningfully faster than regular Phase II trials, and (4) will produce data that may not be able to be extrapolated to the real-world situation. I don’t see the point.