HIV and ID Observations

In the podcast I did with Helen Branswell — Infectious Diseases and global health reporter for STAT — she mentioned that the flow of scientific information for the COVID-19 pandemic made the commonly cited “drinking from a fire hose” analogy somehow inadequate.

Since she’s from Canada, I offered Niagara Falls as an alternative, to which she cleverly responded.

Yeah. It’s like standing out at the falls, trying to fill a glass of water. It’s overwhelming and it’s impossible to keep up.

She then mentioned the ascendancy of the preprint servers, which of course is part of this deluge. Never in the history of medical research has so much non–peer-reviewed data made its way to the public’s insatiably hungry (or should I say thirsty, to continue the analogy) view.

And it’s not just preprints. Let’s start with remdesivir, the repurposed antiviral making its way quickly through clinical trials — though apparently not quickly enough.

Here’s what we know as of today, April 27, 2020:

• There is one (count ’em) peer-reviewed published paper on use of remdesivir in COVID-19. It appeared right here in the august pages of the parent journal of this site, the venerable New England Journal of Medicine. It’s an uncontrolled expanded access compassionate use study showing that it might work, and likely isn’t too harmful. That’s all we can say from this controversial publication — controversial because under normal circumstances, the New England Journal of Medicine does not publish studies like this. But they would argue (and I agree), these are far from normal circumstances.
• Optimistic comments from a site investigator conducting a remdesivir study became public. During what sounds like a faculty meeting or medical grand rounds, the researcher said people in the study responded promptly to remdesivir treatment. Under ordinary times, this information would not have left the walls of the hospital, or even been mentioned at all — but these are not ordinary times! We conduct all meetings online, making them easy to record (and to leak, um, share). In the annals of “How is current life so bizarrely different from pre-COVID-19 life?,” let’s remember this one.
• Discouraging results of a partially completed study appeared on the World Health Organization’s site. More accurately — briefly appeared; it has now been taken down, but can still be read because someone took a screen shot, and shared it. The target sample size for this randomized trial was intended to be 453, but only 237 enrolled — whether the study was stopped due to a futility analysis or because the incidence of COVID-19 in China fell is not clear (the company statement implies the latter). Regardless, based on this now infamous screen shot, the drug did not significantly improve clinical outcomes. And yes, we’re now gleaning information from a blurry screen shot. Gosh, how we’ve changed.

Published, peer-reviewed data on remdesivir is expected soon — looking forward to that.

Speaking of looking for data, how about this one about tocilizumab, the humanized monoclonal antibody against the interleukin-6 receptor:

#Covid19 | #Tocilizumab improves significantly clinical outcomes of patients with moderate or severe COVID-19 pneumonia. First results of the #CORIMUNO-TOCI open-label randomized controlled trial https://t.co/L403gJ7rfF pic.twitter.com/TLiL8PAmCU

— AP-HP (@APHP) April 27, 2020

That’s all we get — just the title slide of what looks to be a PowerPoint presentation. Someone else offered additional data, perhaps an investigator?

Good news. 129 moderate or severe patients w/ #COVID19. 65 in #tocilizumab arm, 64 in control arm. Endpoint: death or need for mechanical ventilation by D14. Tocilizumab significantly superior to control arm. Publication will follow soon.

Tantalizing! But is there more? Peer-reviewed journal next, or preprint server?

If that’s not enough for a Monday a.m., we also have this announcement about sarilumab, another monoclonal antibody against the interleukin-6 receptor. Here’s an excerpt from what is a quite complex (and confusing, at least to me) summary:

Analysis of clinical outcomes in the Phase 2 trial was exploratory and pre-specified to focus on the “severe” and “critical” groups. In the preliminary Phase 2 analysis, Kevzara [sarilumab] had no notable benefit on clinical outcomes when combining the “severe” and “critical” groups, versus placebo. However, there were negative trends for most outcomes in the “severe” group, while there were positive trends for all outcomes in the “critical” group (see table below).

Got that? Negative and positive trends.

So why is this happening? To overstate the obvious, we are all so desperate for good news on therapy for this terrible infection that we hang on every released word. And the rapidity of communication these days allows such preliminary messages to appear instantly!

But there’s definitely something missing, too. Actually, a lot missing.

Let’s take the tocilizumab study: No details on inclusion or exclusion criteria. No sample size calculation or statistical methods section. Nothing on screen failures. Nothing on dosing, duration of follow-up, concomitant therapies. No “Table 1” of baseline characteristics. No information on clinical outcomes. No safety data.

And of course, no peer review. Oh well.

With sarilumab, we have more information, but it’s so convoluted I don’t think anyone can make an assessment of whether this treatment is helpful, or harmful, or in what populations.

And for remdesivir? One grows tired of getting “critical” medical information from images that probably should have been communicated by Snapchat, if that.

Though come to think of it — what a great analogy. Because nothing ever dies on the internet.

Hey, listen to Helen Branswell tell her story!

Transcript here. Subscribe here.