A modest proposal for a clinical trial program for Long Covid

Time to stop beating around the bush

I had a convo over DM at #Bluesky about how I'd design and what to test in #longcovid clinical trials. While many have moved on after the vaccines and the emergence of the omicron soup of variants, for many of those with #longcovid life is still hell. In fact, given a sufficient number of reinfections many vaccinated will eventually develop #longcovid. We are still beating around the bush about Covid19 in general and #longcovid in particular, so whatever is worth, here are my thoughts about population selection, design, intervention, outcomes for long-covid trials and the rationale to do things this way. As always discource about my non-nuanced thoughts is welcome.

Population : Invite people who self identify as having #longcovid . No other restriction other than the obvious (exclude populations who cannot consent or those who are vulnerable to coercion.

Design for #longcovid #RCT: Stratify individuals via evidence of

• persistence (which many have via antigen or sequencing testing in plasma, stool or other accessible biological matrix)

• T cell Exhaustion markers

• Evidence of vascular involvement(VEGF/caveolin/VVO relevant biomarkers) and

Randomize by strata to

1. 2 antivirals with different MoA : protease inhibitors + remdisivir with and without maraviroc

2. metformin

3. For those we can type the antigen (MassSpec) or RNA to one of the variants, add the monoclonal that was effective against that variant which is the putative cause of their #longcovid
   

Outcomes:

1. Symptom burden, functioning and time to resolution of what individuals claim their LC was

2. Antigen , RNA levels, inflammation and vascular markers

3. Correlation between 2 and 1 (joint biomarker and time to event analysis) & get an idea of sensitivity/specifity of said markers for #longcovid and as biomarkers of response
   

Exploratory outcomes: It is important to use this data to probe the possibility of a surrogate for #longcovid , so I propose we

1. Generate a surrogate similar to the HIV TLOVR or snapshot from the virus and inflammatory biomarkers derived which will be used in subsequent trials to derive somewhat objective markers of response

2. This surrogate will hopefuly synthesize virological and biomarker data to a reasonable accurate and precise predictor with good discriminatory activity of those who respond versus those who do not and perhaps identify components of therapy that work better given the “levels” of the surrogate

Rationale: The approach tests multiple interventions and hypotheses about the biology of #longcovid (e.g. persistence). If any of the interventions work, the design will tells us if we can a priori identify those who may respond & establish surrogates for simplification trials: therapies with shorter durations, fewer components and identify the non-responders who should be offered participation in trials of investigational antivirals, RNA interference therapies or other monoclonals/biologics to be developed.

BlueSky Thread

Mastodon link

Twitter link (protected)